From scf at FreeBSD.org Sat Mar 1 00:02:13 2008 From: scf at FreeBSD.org (Sean Farley) Date: Sat Mar 1 00:02:15 2008 Subject: cvs commit: src/tools/regression/environ envctl.c Message-ID: <200803010002.m2102D7n018356@repoman.freebsd.org> scf 2008-03-01 00:02:12 UTC FreeBSD src repository Modified files: tools/regression/environ envctl.c Log: Remove a dereference. It was unintended and a no-op. Use the correct value of errno. Although the errno value passed into printf() follows the *env() call, it is not guaranteed to be the errno from that call. When I wrote the regression tester, the environment I used did pass the errno from the call. Consolidate the print for the return code and errno into a function in the process of fixing this. Approved by: wes (mentor) Revision Changes Path 1.3 +30 -20 src/tools/regression/environ/envctl.c From jfv at FreeBSD.org Sat Mar 1 00:03:53 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Sat Mar 1 00:03:55 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200803010003.m2103rNp019580@repoman.freebsd.org> jfv 2008-03-01 00:03:53 UTC FreeBSD src repository Modified files: sys/conf NOTES Log: Add entries for em, igb, and ixgbe adapters. Revision Changes Path 1.1470 +4 -0 src/sys/conf/NOTES From jfvogel at gmail.com Sat Mar 1 00:18:53 2008 From: jfvogel at gmail.com (Jack Vogel) Date: Sat Mar 1 00:18:57 2008 Subject: cvs commit: src/sys/conf files src/sys/dev/em LICENSE e1000_80003es2lan.c e1000_80003es2lan.h e1000_82540.c e1000_82541.c e1000_82541.h e1000_82542.c e1000_82543.c e1000_82543.h e1000_82571.c e1000_82571.h e1000_82575.c ... In-Reply-To: <47C7428C.9070609@gmail.com> References: <200802292150.m1TLoC72010436@repoman.freebsd.org> <47C7428C.9070609@gmail.com> Message-ID: <2a41acea0802291552j60a2d5f3w4fc66f7a44fb6c3@mail.gmail.com> Ahhh, good point, I will do so. Thanks :) On Thu, Feb 28, 2008 at 3:23 PM, Niclas Zeising wrote: > > Jack F Vogel wrote: > > jfv 2008-02-29 21:50:11 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/conf files > > sys/dev/em LICENSE e1000_80003es2lan.c > > e1000_80003es2lan.h e1000_82540.c > > e1000_82541.c e1000_82541.h e1000_82542.c > > e1000_82543.c e1000_82543.h e1000_82571.c > > e1000_82571.h e1000_api.c e1000_api.h > > e1000_defines.h e1000_hw.h > > e1000_ich8lan.c e1000_ich8lan.h > > e1000_mac.c e1000_mac.h e1000_manage.c > > e1000_manage.h e1000_nvm.c e1000_nvm.h > > e1000_osdep.h e1000_phy.c e1000_phy.h > > e1000_regs.h if_em.c if_em.h > > sys/modules/em Makefile > > Added files: > > sys/dev/igb e1000_82575.c e1000_82575.h e1000_api.c > > e1000_api.h e1000_defines.h e1000_hw.h > > e1000_mac.c e1000_mac.h e1000_manage.c > > e1000_manage.h e1000_nvm.c e1000_nvm.h > > e1000_osdep.h e1000_phy.c e1000_phy.h > > e1000_regs.h if_igb.c if_igb.h > > sys/modules/igb Makefile > > Removed files: > > sys/dev/em e1000_82575.c e1000_82575.h > > Log: > > This change introduces a split to the Intel E1000 driver, now rather than > > just em, there is an igb driver (this follows behavior with our Linux drivers). > > All adapters up to the 82575 are supported in em, and new client/desktop support > > will continue to be in that adapter. > > > > The igb driver is for new server NICs like the 82575 and its followons. > > Advanced features for virtualization and performance will be in this driver. > > > > Also, both drivers now have shared code that is up to the latest we have > > released. Some stylistic changes as well. > > > > Enjoy :) > > > > [Snip commit details] > > Shouldn't 'device igb' go into NOTES as well? > > Thanks for your excellent support! > Regards! > Niclas > From jfv at FreeBSD.org Sat Mar 1 01:09:36 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Sat Mar 1 01:09:42 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200803010109.m2119ZtW032832@repoman.freebsd.org> jfv 2008-03-01 01:09:35 UTC FreeBSD src repository Modified files: sys/conf NOTES Log: Temporarily comment out new entries due to build problems, to be resolved next week. Revision Changes Path 1.1471 +3 -3 src/sys/conf/NOTES From jfv at FreeBSD.org Sat Mar 1 03:25:35 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Sat Mar 1 03:25:39 2008 Subject: cvs commit: src/sys/dev/igb if_igb.c Message-ID: <200803010325.m213PYme052040@repoman.freebsd.org> jfv 2008-03-01 03:25:34 UTC FreeBSD src repository Modified files: sys/dev/igb if_igb.c Log: Missed some code that is ifdef STRICT_ALIGN :( Revision Changes Path 1.2 +1 -1 src/sys/dev/igb/if_igb.c From jfv at FreeBSD.org Sat Mar 1 03:58:50 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Sat Mar 1 03:59:10 2008 Subject: cvs commit: src/sys/conf files Message-ID: <200803010358.m213wnOh053269@repoman.freebsd.org> jfv 2008-03-01 03:58:49 UTC FreeBSD src repository Modified files: sys/conf files Log: Somehow missed the stanza for 575 in the em driver Revision Changes Path 1.1273 +0 -2 src/sys/conf/files From bmah at FreeBSD.org Sat Mar 1 04:10:21 2008 From: bmah at FreeBSD.org (Bruce A. Mah) Date: Sat Mar 1 04:10:23 2008 Subject: cvs commit: src/release/doc/en_US.ISO8859-1/errata article.sgml Message-ID: <200803010410.m214AK3o054922@repoman.freebsd.org> bmah 2008-03-01 04:10:20 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) release/doc/en_US.ISO8859-1/errata article.sgml Log: Break up a long sentence to make more grammatical sense and be easier to read. Problem pointed out by: ceri Revision Changes Path 1.73.4.2 +1 -1 src/release/doc/en_US.ISO8859-1/errata/article.sgml From jfv at FreeBSD.org Sat Mar 1 04:12:55 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Sat Mar 1 04:12:58 2008 Subject: cvs commit: src/sys/dev/igb if_igb.c Message-ID: <200803010412.m214Csp8054996@repoman.freebsd.org> jfv 2008-03-01 04:12:54 UTC FreeBSD src repository Modified files: sys/dev/igb if_igb.c Log: Missing braces in link routine. Revision Changes Path 1.3 +2 -1 src/sys/dev/igb/if_igb.c From bmah at FreeBSD.org Sat Mar 1 04:33:15 2008 From: bmah at FreeBSD.org (Bruce A. Mah) Date: Sat Mar 1 04:33:17 2008 Subject: cvs commit: src/release/doc/en_US.ISO8859-1/errata article.sgml Message-ID: <200803010433.m214XF4l055813@repoman.freebsd.org> bmah 2008-03-01 04:33:15 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) release/doc/en_US.ISO8859-1/errata article.sgml Log: New errata items: Wrong KDE version number, Linux mmap(2) issues, multithreaded programs and fork(2) issue. Revision Changes Path 1.73.2.54 +12 -1 src/release/doc/en_US.ISO8859-1/errata/article.sgml From jfv at FreeBSD.org Sat Mar 1 04:36:24 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Sat Mar 1 04:36:26 2008 Subject: cvs commit: src/sys/dev/igb if_igb.c Message-ID: <200803010436.m214aOmw055959@repoman.freebsd.org> jfv 2008-03-01 04:36:24 UTC FreeBSD src repository Modified files: sys/dev/igb if_igb.c Log: An error in the poll routine turned up during LINT build Revision Changes Path 1.4 +1 -1 src/sys/dev/igb/if_igb.c From jfv at FreeBSD.org Sat Mar 1 08:23:36 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Sat Mar 1 08:23:38 2008 Subject: cvs commit: src/sys/dev/igb if_igb.c Message-ID: <200803010823.m218NZ67081040@repoman.freebsd.org> jfv 2008-03-01 08:23:35 UTC FreeBSD src repository Modified files: sys/dev/igb if_igb.c Log: Change data formating in debug code. Revision Changes Path 1.5 +12 -12 src/sys/dev/igb/if_igb.c From rwatson at FreeBSD.org Sat Mar 1 11:04:05 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:04:07 2008 Subject: cvs commit: src/sys/security/audit audit_private.h audit_syscalls.c audit_trigger.c audit_worker.c Message-ID: <200803011104.m21B448G004995@repoman.freebsd.org> rwatson 2008-03-01 11:04:04 UTC FreeBSD src repository Modified files: sys/security/audit audit_private.h audit_syscalls.c audit_trigger.c audit_worker.c Log: Rename globally exposed symbol send_trigger() to audit_send_trigger(). MFC after: 1 month Revision Changes Path 1.18 +1 -1 src/sys/security/audit/audit_private.h 1.24 +1 -1 src/sys/security/audit/audit_syscalls.c 1.6 +1 -1 src/sys/security/audit/audit_trigger.c 1.21 +4 -3 src/sys/security/audit/audit_worker.c From rwatson at FreeBSD.org Sat Mar 1 11:33:22 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:33:25 2008 Subject: cvs commit: src/sys/nfsclient nfs_vnops.c Message-ID: <200803011133.m21BXMmm006027@repoman.freebsd.org> rwatson 2008-03-01 11:33:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/nfsclient nfs_vnops.c Log: Merge nfs_vnops.c:1.277 from HEAD to RELENG_6: Remove hacks from the NFSv2/3 client intended to handle a lack of a server-side RPC retranmission cache for non-idempotent operations: these hacks substituted 0 (success) for the expected EEXIST in the event that a target name already existed for LINK, SYMLINK, and MKDIR operations, under the assumption that EEXIST represented a second application of the original RPC rather than a true failure. Background: certain NFS operations (in this case, LINK, SYMLINK, and MKDIR) are not idempotent, as they leave behind persisting state on the server that prevents them from being replayed without an error;if an UDP RPC reply is lost leading to a retransmission by theclient, the second reply will return EEXIST rather than success, asthe new object has already been created. The NFS client previouslysilently mapped the EEXIST return into success to paper over thisproblem. However, in all modern NFS server implementations, a reply cache is kept in order to retransmit the original reply to a retransmitted request, rather than performing the operation a second time, allowing this hack to be avoided. This allows link()-based filelocking over NFS to operate correctly, as an application requestingthe creation of a new link for a file to tell if it succeededatomically or not. Other NFS clients, including Solaris and Linux, generally follow this behavior for the same reasons. Most clients also now default to TCP, which also helps avoid the issue of retransmitted but non-idempotent requests in most cases. Reported by: Adam McDougall , Timo Sirainen Reviewed by: mohans Revision Changes Path 1.258.2.16 +4 -25 src/sys/nfsclient/nfs_vnops.c From brueffer at FreeBSD.org Sat Mar 1 11:35:32 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sat Mar 1 11:35:34 2008 Subject: cvs commit: src/share/man/man4 ucycom.4 Message-ID: <200803011135.m21BZUZi006159@repoman.freebsd.org> brueffer 2008-03-01 11:35:30 UTC FreeBSD src repository Modified files: share/man/man4 ucycom.4 Log: Move the HARDWARE section below the DESCRIPTION section, like it is everywhere else. MFC after: 3 days Revision Changes Path 1.5 +10 -10 src/share/man/man4/ucycom.4 From rwatson at FreeBSD.org Sat Mar 1 11:38:27 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:38:29 2008 Subject: cvs commit: src/sys/net bpf.c bpf.h Message-ID: <200803011138.m21BcQb9006229@repoman.freebsd.org> rwatson 2008-03-01 11:38:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/net bpf.c bpf.h Log: Merge bpf.c:184, bpf.h:1.49 from HEAD to RELENG_6: Remove trailing whitespace from lines in BPF. Minor tweaks to account for minor changes between RELENG_6 and HEAD. Revision Changes Path 1.153.2.13 +3 -3 src/sys/net/bpf.c 1.39.2.4 +2 -2 src/sys/net/bpf.h From rwatson at FreeBSD.org Sat Mar 1 11:40:49 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:40:52 2008 Subject: cvs commit: src/sys/security/audit audit.c audit_bsm_klib.c audit_private.h Message-ID: <200803011140.m21Benxp006381@repoman.freebsd.org> rwatson 2008-03-01 11:40:49 UTC FreeBSD src repository Modified files: sys/security/audit audit.c audit_bsm_klib.c audit_private.h Log: Add audit_prefixes to two more globally visible functions in the Audit implementation. MFC after: 1 month Revision Changes Path 1.39 +2 -2 src/sys/security/audit/audit.c 1.11 +2 -2 src/sys/security/audit/audit_bsm_klib.c 1.19 +3 -3 src/sys/security/audit/audit_private.h From rwatson at FreeBSD.org Sat Mar 1 11:43:02 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:43:04 2008 Subject: cvs commit: src/sys/kern uipc_sem.c Message-ID: <200803011143.m21Bh1d5006484@repoman.freebsd.org> rwatson 2008-03-01 11:43:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/kern uipc_sem.c Log: Merge uipc_sem.c:1.30 from HEAD to RELENG_6: Free MAC label on a POSIX semaphore when the semaphore is freed. Submitted by: jhb Revision Changes Path 1.20.2.2 +3 -0 src/sys/kern/uipc_sem.c From rwatson at FreeBSD.org Sat Mar 1 11:44:30 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:44:34 2008 Subject: cvs commit: src/sys/net netisr.h Message-ID: <200803011144.m21BiUbc006558@repoman.freebsd.org> rwatson 2008-03-01 11:44:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/net netisr.h Log: Merge netisr.h:1.34 from HEAD to RELENG_7: Update netisr comment for the SMPng world order: netisr is no longer implemented using the ISR facility, and cannot be triggered by calling splnet()/splx(). Revision Changes Path 1.33.10.1 +8 -13 src/sys/net/netisr.h From rwatson at FreeBSD.org Sat Mar 1 11:45:15 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:45:16 2008 Subject: cvs commit: src/sys/net netisr.h Message-ID: <200803011145.m21BjEOM006655@repoman.freebsd.org> rwatson 2008-03-01 11:45:14 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/net netisr.h Log: Merge netisr.h:1.34 from HEAD to RELENG_6: Update netisr comment for the SMPng world order: netisr is no longer implemented using the ISR facility, and cannot be triggered by calling splnet()/splx(). Revision Changes Path 1.33.2.1 +8 -13 src/sys/net/netisr.h From rwatson at FreeBSD.org Sat Mar 1 11:50:02 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:50:03 2008 Subject: cvs commit: src/sys/netinet tcp_usrreq.c Message-ID: <200803011150.m21Bo0JE006789@repoman.freebsd.org> rwatson 2008-03-01 11:50:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet tcp_usrreq.c Log: Merge tcp_usrreq.c:1.166 from HEAD to RELENG_7: In tcp_ctloutput(), don't hold the inpcb lock over sooptcopyin(), rather, drop the lock and then re-acquire it, revalidating TCP connection state assumptions when we do so. This avoids a potential lock order reversal (and potential deadlock, although none have been reported) due to the inpcb lock being held over a page fault. PR: 102752 Reviewed by: bz Reported by: Václav Haisman Revision Changes Path 1.163.2.3 +55 -25 src/sys/netinet/tcp_usrreq.c From rwatson at FreeBSD.org Sat Mar 1 11:50:46 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:50:48 2008 Subject: cvs commit: src/sys/kern uipc_usrreq.c Message-ID: <200803011150.m21BojCp006924@repoman.freebsd.org> rwatson 2008-03-01 11:50:45 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern uipc_usrreq.c Log: Merge uipc_usrreq.c:1.211 from HEAD to RELENG_7: Correct typos in comments. Revision Changes Path 1.206.2.2 +2 -2 src/sys/kern/uipc_usrreq.c From rwatson at FreeBSD.org Sat Mar 1 11:51:13 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:51:15 2008 Subject: cvs commit: src/sys/kern tty_pts.c uipc_usrreq.c Message-ID: <200803011151.m21BpCri006962@repoman.freebsd.org> rwatson 2008-03-01 11:51:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern tty_pts.c uipc_usrreq.c Log: Merge uipc_usrreq.c:1.212 from HEAD to RELENG_7: Remove "lock pushdown" todo item in comment -- I did that for 7.0. Revision Changes Path 1.16.2.1 +4 -0 src/sys/kern/tty_pts.c 1.206.2.3 +0 -1 src/sys/kern/uipc_usrreq.c From rwatson at FreeBSD.org Sat Mar 1 11:55:12 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 11:55:17 2008 Subject: cvs commit: src/sys/security/mac_mls mac_mls.c Message-ID: <200803011155.m21BtCR7007146@repoman.freebsd.org> rwatson 2008-03-01 11:55:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/security/mac_mls mac_mls.c Log: Merge mac_mls.c:1.99 from HEAD to RELENG_7: Properly return the error from mls_subject_privileged() in the ifnet relabel check for MLS rather than returning 0 directly. This problem didn't result in a vulnerability currently as the central implementation of ifnet relabeling also checks for UNIX privilege, and we currently don't guarantee containment for the root user in mac_mls, but we should be using the MLS definition of privilege as well as the UNIX definition in anticipation of supporting root containment at some point. Submitted by: Zhouyi Zhou Sponsored by: Google SoC 2007 Revision Changes Path 1.88.2.2 +1 -3 src/sys/security/mac_mls/mac_mls.c From brueffer at FreeBSD.org Sat Mar 1 13:05:03 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sat Mar 1 13:05:09 2008 Subject: cvs commit: src/share/man/man4 Makefile uchcom.4 Message-ID: <200803011305.m21D53wr021291@repoman.freebsd.org> brueffer 2008-03-01 13:05:03 UTC FreeBSD src repository Modified files: share/man/man4 Makefile Added files: share/man/man4 uchcom.4 Log: Manpage for the uchcom(4) driver, adapted from the one in NetBSD. Revision Changes Path 1.410 +1 -0 src/share/man/man4/Makefile 1.1 +94 -0 src/share/man/man4/uchcom.4 (new) From brueffer at FreeBSD.org Sat Mar 1 13:28:28 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sat Mar 1 13:28:30 2008 Subject: cvs commit: src/share/man/man4 ucom.4 Message-ID: <200803011328.m21DSRsJ022216@repoman.freebsd.org> brueffer 2008-03-01 13:28:27 UTC FreeBSD src repository Modified files: share/man/man4 ucom.4 Log: Xref uchcom(4). Revision Changes Path 1.9 +2 -1 src/share/man/man4/ucom.4 From rwatson at FreeBSD.org Sat Mar 1 14:52:07 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 14:52:13 2008 Subject: cvs commit: src/sys/security/mac_mls mac_mls.c Message-ID: <200803011452.m21Eq7K7038008@repoman.freebsd.org> rwatson 2008-03-01 14:52:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/security/mac_mls mac_mls.c Log: Merge mac_mls.c:1.99 from HEAD to RELENG_6: Properly return the error from mls_subject_privileged() in the ifnet relabel check for MLS rather than returning 0 directly. This problem didn't result in a vulnerability currently as the central implementation of ifnet relabeling also checks for UNIX privilege, and we currently don't guarantee containment for the root user in mac_mls, but we should be using the MLS definition of privilege as well as the UNIX definition in anticipation of supporting root containment at some point. Submitted by: Zhouyi Zhou Sponsored by: Google SoC 2007 Revision Changes Path 1.72.2.6 +1 -3 src/sys/security/mac_mls/mac_mls.c From rwatson at FreeBSD.org Sat Mar 1 15:40:53 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 15:40:57 2008 Subject: cvs commit: src/sys/kern uipc_sockbuf.c uipc_socket.c src/sys/sys socketvar.h Message-ID: <200803011540.m21FerSv041814@repoman.freebsd.org> rwatson 2008-03-01 15:40:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern uipc_sockbuf.c uipc_socket.c sys/sys socketvar.h Log: Merge uipc_sockbuf.c:1.176, uipc_socket.c:1.305, socketvar.h:1.162 from HEAD to RELENG_7: Further clean up sorflush: - Expose sbrelease_internal(), a variant of sbrelease() with no expectations about the validity of locks in the socket buffer. - Use sbrelease_internel() in sorflush(), and as a result avoid initializing and destroying a socket buffer lock for the temporary stack copy of the actual buffer, asb. - Add a comment indicating why we do what we do, and remove an XXX since things have gotten less ugly in sorflush() lately. This makes socket close cleaner, and possibly also marginally faster. Revision Changes Path 1.171.2.3 +1 -2 src/sys/kern/uipc_sockbuf.c 1.302.2.2 +12 -12 src/sys/kern/uipc_socket.c 1.158.2.2 +1 -0 src/sys/sys/socketvar.h From rwatson at FreeBSD.org Sat Mar 1 15:42:01 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 15:42:03 2008 Subject: cvs commit: src/sys/arm/xscale/ixp425 ixp425.c Message-ID: <200803011542.m21Fg1wv041883@repoman.freebsd.org> rwatson 2008-03-01 15:42:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/arm/xscale/ixp425 ixp425.c Log: Merge ixp425.c:1.9 from HEAD to RELENG_7: On the ixp425, when we fail to initialize the memory rman instance, the panic message should read "memory", not "IRQ". Revision Changes Path 1.8.2.1 +1 -1 src/sys/arm/xscale/ixp425/ixp425.c From rwatson at FreeBSD.org Sat Mar 1 15:42:24 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 15:42:27 2008 Subject: cvs commit: src/sys/arm/include _bus.h Message-ID: <200803011542.m21FgNvD041925@repoman.freebsd.org> rwatson 2008-03-01 15:42:23 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/arm/include _bus.h Log: Merge _bus.h:1.2 from HEAD to RELENG_7: Remove errant % in license comment. Revision Changes Path 1.1.10.1 +1 -1 src/sys/arm/include/_bus.h From rwatson at FreeBSD.org Sat Mar 1 15:43:01 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 1 15:43:04 2008 Subject: cvs commit: src/sys/fs/coda TODO Message-ID: <200803011543.m21Fh0be041979@repoman.freebsd.org> rwatson 2008-03-01 15:43:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda TODO Log: Merge TODO:1.6 from HEAD to RELENG_7: Add "Make MPSAFE" to the Coda todo list. Revision Changes Path 1.3.2.1 +1 -0 src/sys/fs/coda/TODO From philip at FreeBSD.org Sat Mar 1 16:17:30 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sat Mar 1 16:17:38 2008 Subject: cvs commit: src/lib/libc/sys mq_close.2 mq_getattr.2 mq_notify.2 mq_open.2 mq_receive.2 mq_send.2 mq_setattr.2 Message-ID: <200803011617.m21GHU1s044320@repoman.freebsd.org> philip 2008-03-01 16:17:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/sys mq_close.2 mq_getattr.2 mq_notify.2 mq_open.2 mq_receive.2 mq_send.2 mq_setattr.2 Log: MFC: Note, as required by our agreement with IEEE/The Open Group, that the message queue manual pages excerpt the POSIX standard. Revision Changes Path 1.4.2.1 +20 -0 src/lib/libc/sys/mq_close.2 1.4.2.1 +20 -0 src/lib/libc/sys/mq_getattr.2 1.8.2.1 +20 -0 src/lib/libc/sys/mq_notify.2 1.8.2.1 +20 -0 src/lib/libc/sys/mq_open.2 1.5.2.1 +20 -0 src/lib/libc/sys/mq_receive.2 1.4.2.1 +20 -0 src/lib/libc/sys/mq_send.2 1.5.2.1 +20 -0 src/lib/libc/sys/mq_setattr.2 From brian at FreeBSD.org Sat Mar 1 16:56:10 2008 From: brian at FreeBSD.org (Brian Somers) Date: Sat Mar 1 16:56:14 2008 Subject: cvs commit: CVSROOT access Message-ID: <200803011656.m21GuAtY054700@repoman.freebsd.org> brian 2008-03-01 16:56:10 UTC FreeBSD src repository Modified files: . access Log: Change my email address to something that works... pending Awfulhak.org being updated. Approved by: core (wilko) Revision Changes Path 1.861 +1 -1 CVSROOT/access From mav at FreeBSD.org Sat Mar 1 17:14:03 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sat Mar 1 17:14:07 2008 Subject: cvs commit: src/share/man/man4 ng_nat.4 src/sys/netgraph ng_nat.c ng_nat.h Message-ID: <200803011714.m21HE2O7056509@repoman.freebsd.org> mav 2008-03-01 17:14:02 UTC FreeBSD src repository Modified files: share/man/man4 ng_nat.4 sys/netgraph ng_nat.c ng_nat.h Log: Add support for the libalias redirect functionality. Submitted by: Vadim Goncharov Revision Changes Path 1.4 +188 -2 src/share/man/man4/ng_nat.4 1.11 +431 -1 src/sys/netgraph/ng_nat.c 1.3 +127 -0 src/sys/netgraph/ng_nat.h From piso at FreeBSD.org Sat Mar 1 17:50:11 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Sat Mar 1 17:50:15 2008 Subject: cvs commit: src/share/man/man4 ng_nat.4 src/sys/netgraph ng_nat.c ng_nat.h In-Reply-To: <200803011714.m21HE2O7056509@repoman.freebsd.org> References: <200803011714.m21HE2O7056509@repoman.freebsd.org> Message-ID: <20080301173324.GA16036@tin.it> On Sat, Mar 01, 2008 at 05:14:02PM +0000, Alexander Motin wrote: > mav 2008-03-01 17:14:02 UTC > > FreeBSD src repository > > Modified files: > share/man/man4 ng_nat.4 > sys/netgraph ng_nat.c ng_nat.h > Log: > Add support for the libalias redirect functionality. > > Submitted by: Vadim Goncharov Thanks! :) bye, P. From obrien at FreeBSD.org Sat Mar 1 18:35:28 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sat Mar 1 18:35:31 2008 Subject: cvs commit: src/sys/ufs/ufs ufs_quota.c Message-ID: <200803011835.m21IZSrx061079@repoman.freebsd.org> obrien 2008-03-01 18:35:27 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/ufs/ufs ufs_quota.c Log: Sync with HEAD - remove unneeded cases. Revision Changes Path 1.74.2.10 +5 -5 src/sys/ufs/ufs/ufs_quota.c From marck at rinet.ru Sat Mar 1 19:25:10 2008 From: marck at rinet.ru (Dmitry Morozovsky) Date: Sat Mar 1 19:25:13 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <200802290912.m1T9CsDJ074384@repoman.freebsd.org> References: <200802290912.m1T9CsDJ074384@repoman.freebsd.org> Message-ID: <20080301222414.P49885@woozle.rinet.ru> On Fri, 29 Feb 2008, Wilko Bulte wrote: WB> Modified files: WB> . access WB> Log: WB> Changes to email forwards can be made without core's approval. Please WB> use "Approved by: core (implicit)" to that type of commits. Don't you think similar comments should be added to other repositories' CVSROOT/access? Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck@FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru *** ------------------------------------------------------------------------ From attilio at FreeBSD.org Sat Mar 1 19:47:53 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Sat Mar 1 19:47:55 2008 Subject: cvs commit: src/sys/dev/usb ehci.c src/sys/fs/nwfs nwfs_subr.c src/sys/kern kern_lock.c vfs_bio.c vfs_subr.c src/sys/netncp ncp_conn.h src/sys/netsmb smb_conn.h src/sys/nfs4client nfs4_idmap.c src/sys/sys buf.h bufobj.h lockmgr.h mount.h vnode.h Message-ID: <200803011947.m21JlpmG067075@repoman.freebsd.org> attilio 2008-03-01 19:47:50 UTC FreeBSD src repository Modified files: sys/dev/usb ehci.c sys/fs/nwfs nwfs_subr.c sys/kern kern_lock.c vfs_bio.c vfs_subr.c sys/netncp ncp_conn.h sys/netsmb smb_conn.h sys/nfs4client nfs4_idmap.c sys/sys buf.h bufobj.h lockmgr.h mount.h vnode.h Log: - Handle buffer lock waiters count directly in the buffer cache instead than rely on the lockmgr support [1]: * bump the waiters only if the interlock is held * let brelvp() return the waiters count * rely on brelvp() instead than BUF_LOCKWAITERS() in order to check for the waiters number - Remove a namespace pollution introduced recently with lockmgr.h including lock.h by including lock.h directly in the consumers and making it mandatory for using lockmgr. - Modify flags accepted by lockinit(): * introduce LK_NOPROFILE which disables lock profiling for the specified lockmgr * introduce LK_QUIET which disables ktr tracing for the specified lockmgr [2] * disallow LK_SLEEPFAIL and LK_NOWAIT to be passed there so that it can only be used on a per-instance basis - Remove BUF_LOCKWAITERS() and lockwaiters() as they are no longer used This patch breaks KPI so __FreBSD_version will be bumped and manpages updated by further commits. Additively, 'struct buf' changes results in a disturbed ABI also. [2] Really, currently there is no ktr tracing in the lockmgr, but it will be added soon. [1] Submitted by: kib Tested by: pho, Andrea Barberio Revision Changes Path 1.61 +1 -0 src/sys/dev/usb/ehci.c 1.18 +1 -1 src/sys/fs/nwfs/nwfs_subr.c 1.127 +8 -18 src/sys/kern/kern_lock.c 1.537 +8 -6 src/sys/kern/vfs_bio.c 1.720 +5 -1 src/sys/kern/vfs_subr.c 1.13 +1 -0 src/sys/netncp/ncp_conn.h 1.12 +1 -0 src/sys/netsmb/smb_conn.h 1.7 +1 -0 src/sys/nfs4client/nfs4_idmap.c 1.203 +33 -16 src/sys/sys/buf.h 1.19 +1 -0 src/sys/sys/bufobj.h 1.65 +26 -19 src/sys/sys/lockmgr.h 1.229 +1 -1 src/sys/sys/mount.h 1.331 +1 -8 src/sys/sys/vnode.h From attilio at FreeBSD.org Sat Mar 1 19:53:27 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Sat Mar 1 19:53:30 2008 Subject: cvs commit: src/sys/sys param.h Message-ID: <200803011953.m21JrR84067537@repoman.freebsd.org> attilio 2008-03-01 19:53:26 UTC FreeBSD src repository Modified files: sys/sys param.h Log: Bump __FreeBSD_version in order to reflect: - lockwaiters() axing out - BUF_LOCKWAITERS() axing out - brelvp() prototype changing - lockinit() accepted arguments() range changing Revision Changes Path 1.340 +1 -1 src/sys/sys/param.h From attilio at FreeBSD.org Sat Mar 1 20:05:22 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Sat Mar 1 20:05:24 2008 Subject: cvs commit: src/share/man/man9 lock.9 Message-ID: <200803012005.m21K5Liw069756@repoman.freebsd.org> attilio 2008-03-01 20:05:20 UTC FreeBSD src repository Modified files: share/man/man9 lock.9 Log: Update lockmgr manpage with last lockmgr modifies: - Remove LK_SLEEPFAIL and LK_NOWAIT for lockinit() and add LK_QUIET and LK_NOPROFILE - Include sys/lock.h as mandatory for the lockmgr support Revision Changes Path 1.27 +9 -6 src/share/man/man9/lock.9 From attilio at FreeBSD.org Sat Mar 1 20:36:51 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Sat Mar 1 20:36:55 2008 Subject: cvs commit: src/share/man/man9 lock.9 Message-ID: <200803012036.m21Kap7S071490@repoman.freebsd.org> attilio 2008-03-01 20:36:51 UTC FreeBSD src repository Modified files: share/man/man9 lock.9 Log: - Add the informations for lockmgr_args(9) - Sort rightly lockstatus(9) in the manpage Revision Changes Path 1.28 +22 -3 src/share/man/man9/lock.9 From attilio at FreeBSD.org Sat Mar 1 20:40:02 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Sat Mar 1 20:40:04 2008 Subject: cvs commit: src/share/man/man9 Makefile Message-ID: <200803012039.m21Kducf071560@repoman.freebsd.org> attilio 2008-03-01 20:39:56 UTC FreeBSD src repository Modified files: share/man/man9 Makefile Log: Connect lockmgr_args(9) to the build. Revision Changes Path 1.331 +1 -0 src/share/man/man9/Makefile From attilio at FreeBSD.org Sat Mar 1 21:24:11 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Sat Mar 1 21:24:13 2008 Subject: cvs commit: src/share/man/man9 lock.9 Message-ID: <200803012124.m21LOAmj083212@repoman.freebsd.org> attilio 2008-03-01 21:24:08 UTC FreeBSD src repository Modified files: share/man/man9 lock.9 Log: Fix lockstatus sorting on the top list. Revision Changes Path 1.29 +1 -1 src/share/man/man9/lock.9 From gibbs at FreeBSD.org Sat Mar 1 21:58:35 2008 From: gibbs at FreeBSD.org (Justin T. Gibbs) Date: Sat Mar 1 21:58:36 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803012158.m21LwYIt084543@repoman.freebsd.org> gibbs 2008-03-01 21:58:34 UTC FreeBSD src repository Modified files: sys/i386/cpufreq est.c Log: In est_acpi_info(), initialize count before passing its pointer to CPUFREQ_DRV_SETTINGS(). The value of count on input is used to prefent overflow of the settings buffer passed into CPUFREQ_DRV_SETTINGS(). This corrects the "est: CPU supports Enhanced Speedstep, but is not recognized." error on my system. MFC after: 1 week Revision Changes Path 1.13 +1 -0 src/sys/i386/cpufreq/est.c From attilio at FreeBSD.org Sat Mar 1 22:14:46 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Sat Mar 1 22:14:49 2008 Subject: cvs commit: src/sys/sys lockmgr.h Message-ID: <200803012214.m21MEjiU086331@repoman.freebsd.org> attilio 2008-03-01 22:14:45 UTC FreeBSD src repository Modified files: sys/sys lockmgr.h Log: Split the kernel / userland interface with propert _KERNEL stub. This should have been always there, but an userland brekage for the recent lockmgr modifies showed it. Revision Changes Path 1.66 +5 -0 src/sys/sys/lockmgr.h From bz at FreeBSD.org Sat Mar 1 22:27:16 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 1 22:27:21 2008 Subject: cvs commit: src/sys/netinet tcp_input.c Message-ID: <200803012227.m21MRFO4087084@repoman.freebsd.org> bz 2008-03-01 22:27:15 UTC FreeBSD src repository Modified files: sys/netinet tcp_input.c Log: Fix indentation (whitespace changes only). MFC after: 6 days Revision Changes Path 1.372 +4 -4 src/sys/netinet/tcp_input.c From marcel at FreeBSD.org Sat Mar 1 22:54:43 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sat Mar 1 22:54:45 2008 Subject: cvs commit: src UPDATING src/sys/vm vm_map.h Message-ID: <200803012254.m21Msgus088507@repoman.freebsd.org> marcel 2008-03-01 22:54:42 UTC FreeBSD src repository Modified files: . UPDATING sys/vm vm_map.h Log: Make the vm_pmap field of struct vmspace the last field in the structure. This allows per-CPU variations of struct pmap on a single architecture without affecting the machine-independent fields. As such, the PMAP variations don't affect the ABI. They become part of it. Revision Changes Path 1.522 +8 -0 src/UPDATING 1.122 +6 -1 src/sys/vm/vm_map.h From wb at freebie.xs4all.nl Sat Mar 1 23:17:14 2008 From: wb at freebie.xs4all.nl (Wilko Bulte) Date: Sat Mar 1 23:17:19 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <20080301222414.P49885@woozle.rinet.ru> References: <200802290912.m1T9CsDJ074384@repoman.freebsd.org> <20080301222414.P49885@woozle.rinet.ru> Message-ID: <20080301231710.GA59383@freebie.xs4all.nl> Quoting Dmitry Morozovsky, who wrote on Sat, Mar 01, 2008 at 10:25:08PM +0300 .. > On Fri, 29 Feb 2008, Wilko Bulte wrote: > > WB> Modified files: > WB> . access > WB> Log: > WB> Changes to email forwards can be made without core's approval. Please > WB> use "Approved by: core (implicit)" to that type of commits. > > Don't you think similar comments should be added to other repositories' > CVSROOT/access? I'm all for that, but that is the responsibility of portmgr, doceng etc. -- Wilko Bulte wilko@FreeBSD.org From marcel at FreeBSD.org Sun Mar 2 00:52:50 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Mar 2 00:52:52 2008 Subject: cvs commit: src/sys/conf files options src/sys/geom/part g_part_vtoc8.c src/sys/sys vtoc.h Message-ID: <200803020052.m220qoYi003758@repoman.freebsd.org> marcel 2008-03-02 00:52:49 UTC FreeBSD src repository Modified files: sys/conf files options Added files: sys/geom/part g_part_vtoc8.c sys/sys vtoc.h Log: Add support for VTOC8 labels (aka sun disk labels). When a label does not have VTOC information about the partitions, it will be created. This is because the VTOC information is used for the partition type and FreeBSD's sunlabel(8) does not create nor use VTOC information. For this purpose, new tags have been added to support FreeBSD's partition types. Revision Changes Path 1.1274 +1 -0 src/sys/conf/files 1.619 +1 -0 src/sys/conf/options 1.1 +480 -0 src/sys/geom/part/g_part_vtoc8.c (new) 1.1 +106 -0 src/sys/sys/vtoc.h (new) From obrien at FreeBSD.org Sun Mar 2 01:35:38 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 2 01:35:40 2008 Subject: cvs commit: src/sys/ufs/ufs ufs_vnops.c Message-ID: <200803020135.m221Zbfo006287@repoman.freebsd.org> obrien 2008-03-02 01:35:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/ufs/ufs ufs_vnops.c Log: MFC: rev 1.278: correct typo in comment. Also sync comment layout changes and spelling fixes. Revision Changes Path 1.271.2.11 +7 -7 src/sys/ufs/ufs/ufs_vnops.c From das at FreeBSD.org Sun Mar 2 01:47:59 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Mar 2 01:48:02 2008 Subject: cvs commit: src/lib/msun/src e_sqrt.c e_sqrtl.c math.h Message-ID: <200803020147.m221lwG0006659@repoman.freebsd.org> das 2008-03-02 01:47:58 UTC FreeBSD src repository Modified files: lib/msun/src e_sqrt.c math.h Added files: lib/msun/src e_sqrtl.c Log: MI implementation of sqrtl(). This is very slow and should be overridden when hardware sqrt is available. Revision Changes Path 1.11 +8 -3 src/lib/msun/src/e_sqrt.c 1.1 +159 -0 src/lib/msun/src/e_sqrtl.c (new) 1.70 +1 -1 src/lib/msun/src/math.h From das at FreeBSD.org Sun Mar 2 01:48:09 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Mar 2 01:48:11 2008 Subject: cvs commit: src/lib/msun/amd64 e_sqrtl.S src/lib/msun/i387 e_sqrtl.S Message-ID: <200803020148.m221m8wv006713@repoman.freebsd.org> das 2008-03-02 01:48:08 UTC FreeBSD src repository Added files: lib/msun/amd64 e_sqrtl.S lib/msun/i387 e_sqrtl.S Log: MD implementations of sqrtl(). Revision Changes Path 1.1 +33 -0 src/lib/msun/amd64/e_sqrtl.S (new) 1.1 +43 -0 src/lib/msun/i387/e_sqrtl.S (new) From das at FreeBSD.org Sun Mar 2 01:48:23 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Mar 2 01:48:26 2008 Subject: cvs commit: src/lib/msun Makefile Symbol.map src/lib/msun/amd64 Makefile.inc src/lib/msun/i387 Makefile.inc src/lib/msun/man sqrt.3 Message-ID: <200803020148.m221mHaI006806@repoman.freebsd.org> das 2008-03-02 01:48:17 UTC FreeBSD src repository Modified files: lib/msun Makefile Symbol.map lib/msun/amd64 Makefile.inc lib/msun/i387 Makefile.inc lib/msun/man sqrt.3 Log: Hook up sqrtl() to the build. Revision Changes Path 1.90 +2 -2 src/lib/msun/Makefile 1.15 +1 -0 src/lib/msun/Symbol.map 1.9 +1 -1 src/lib/msun/amd64/Makefile.inc 1.12 +1 -1 src/lib/msun/i387/Makefile.inc 1.15 +19 -9 src/lib/msun/man/sqrt.3 From das at FreeBSD.ORG Sun Mar 2 01:55:24 2008 From: das at FreeBSD.ORG (David Schultz) Date: Sun Mar 2 01:55:29 2008 Subject: cvs commit: src/lib/msun/src e_sqrt.c e_sqrtl.c math.h In-Reply-To: <200803020147.m221lwG0006659@repoman.freebsd.org> References: <200803020147.m221lwG0006659@repoman.freebsd.org> Message-ID: <20080302015549.GA50747@zim.MIT.EDU> On Sun, Mar 02, 2008, David Schultz wrote: > das 2008-03-02 01:47:58 UTC > > FreeBSD src repository > > Modified files: > lib/msun/src e_sqrt.c math.h > Added files: > lib/msun/src e_sqrtl.c > Log: > MI implementation of sqrtl(). This is very slow and should > be overridden when hardware sqrt is available. Submitted by: Steve Kargl (earlier version) From obrien at FreeBSD.org Sun Mar 2 03:11:49 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 2 03:11:51 2008 Subject: cvs commit: src/contrib/cvs/src log.c Message-ID: <200803020311.m223BnEn015735@repoman.freebsd.org> obrien 2008-03-02 03:11:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) contrib/cvs/src log.c Log: MFC: rev 1.3: '-n' option which is the opposite of '-N'. Revision Changes Path 1.2.8.1 +10 -2 src/contrib/cvs/src/log.c From marcel at FreeBSD.org Sun Mar 2 05:54:34 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Mar 2 05:54:36 2008 Subject: cvs commit: src/sys/modules/geom Makefile src/sys/modules/geom/geom_part Makefile Message-ID: <200803020554.m225sYci032513@repoman.freebsd.org> marcel 2008-03-02 05:54:34 UTC FreeBSD src repository Modified files: sys/modules/geom Makefile Added files: sys/modules/geom/geom_part Makefile Log: Add geom_part module. Revision Changes Path 1.23 +1 -0 src/sys/modules/geom/Makefile 1.1 +18 -0 src/sys/modules/geom/geom_part/Makefile (new) From marcel at FreeBSD.org Sun Mar 2 05:59:25 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Mar 2 05:59:28 2008 Subject: cvs commit: src/sys/modules/geom/geom_apple Makefile src/sys/modules/geom/geom_gpt Makefile Message-ID: <200803020559.m225xPWx032743@repoman.freebsd.org> marcel 2008-03-02 05:59:25 UTC FreeBSD src repository Removed files: sys/modules/geom/geom_apple Makefile sys/modules/geom/geom_gpt Makefile Log: Remove makefiles for geom_apple and geom_gpt. These are non-existent for while. Revision Changes Path 1.2 +0 -8 src/sys/modules/geom/geom_apple/Makefile (dead) 1.2 +0 -8 src/sys/modules/geom/geom_gpt/Makefile (dead) From marcel at FreeBSD.org Sun Mar 2 06:24:30 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Mar 2 06:24:31 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200803020624.m226OTgn035162@repoman.freebsd.org> marcel 2008-03-02 06:24:29 UTC FreeBSD src repository Modified files: sys/conf NOTES Log: Add the SMI VTOC8 disk label option. Revision Changes Path 1.1472 +1 -0 src/sys/conf/NOTES From jkoshy at FreeBSD.org Sun Mar 2 06:33:11 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Sun Mar 2 06:33:15 2008 Subject: cvs commit: src/lib/libelf gelf_rel.c gelf_rela.c Message-ID: <200803020633.m226XBZZ035507@repoman.freebsd.org> jkoshy 2008-03-02 06:33:11 UTC FreeBSD src repository Modified files: lib/libelf gelf_rel.c gelf_rela.c Log: Translate the r_info field of ELF relocation records when converting between 64 and 32 bit variants. Submitted by: kaiw Revision Changes Path 1.2 +11 -2 src/lib/libelf/gelf_rel.c 1.2 +12 -2 src/lib/libelf/gelf_rela.c From kaiw at FreeBSD.org Sun Mar 2 07:01:02 2008 From: kaiw at FreeBSD.org (Kai Wang) Date: Sun Mar 2 07:01:03 2008 Subject: cvs commit: src/usr.bin/ar write.c Message-ID: <200803020701.m227115K036541@repoman.freebsd.org> kaiw 2008-03-02 07:01:01 UTC FreeBSD src repository Modified files: usr.bin/ar write.c Log: - Do not malloc buffer for 0-size member when reading from archive. - Fix a malloc buffer overrun: Use a while loop to check whether the string buffer is big enough after resizing, since doubling once might not be enough when a very long member name or symbol name is provided. - Fix typo. Reported by: Michael Plass Tested by: Michael Plass Reviewed by: jkoshy Approved by: jkoshy Revision Changes Path 1.4 +17 -11 src/usr.bin/ar/write.c From jeff at FreeBSD.org Sun Mar 2 07:19:36 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Sun Mar 2 07:19:38 2008 Subject: cvs commit: src/sys/sys sched.h src/sys/kern sched_4bsd.c sched_ule.c Message-ID: <200803020719.m227JZhe038264@repoman.freebsd.org> jeff 2008-03-02 07:19:35 UTC FreeBSD src repository Modified files: sys/sys sched.h sys/kern sched_4bsd.c sched_ule.c Log: - Add a new sched_affinity() api to be used in the upcoming cpuset implementation. - Add empty implementations of sched_affinity() to 4BSD and ULE. Sponsored by: Nokia Revision Changes Path 1.113 +5 -0 src/sys/kern/sched_4bsd.c 1.224 +5 -0 src/sys/kern/sched_ule.c 1.34 +2 -1 src/sys/sys/sched.h From jeff at FreeBSD.org Sun Mar 2 07:39:23 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Sun Mar 2 07:39:27 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map Message-ID: <200803020739.m227dNLe039427@repoman.freebsd.org> jeff 2008-03-02 07:39:22 UTC FreeBSD src repository Modified files: sys/conf files sys/kern init_main.c kern_thread.c syscalls.master sys/sys _types.h types.h proc.h lib/libc/sys Symbol.map Added files: sys/kern kern_cpuset.c sys/sys cpuset.h Log: Add cpuset, an api for thread to cpu binding and cpu resource grouping and assignment. - Add a reference to a struct cpuset in each thread that is inherited from the thread that created it. - Release the reference when the thread is destroyed. - Add prototypes for syscalls and macros for manipulating cpusets in sys/cpuset.h - Add syscalls to create, get, and set new numbered cpusets: cpuset(), cpuset_{get,set}id() - Add syscalls for getting and setting affinity masks for cpusets or individual threads: cpuid_{get,set}affinity() - Add types for the 'level' and 'which' parameters for the cpuset. This will permit expansion of the api to cover cpu masks for other objects identifiable with an id_t integer. For example, IRQs and Jails may be coming soon. - The root set 0 contains all valid cpus. All thread initially belong to cpuset 1. This permits migrating all threads off of certain cpus to reserve them for special applications. Sponsored by: Nokia Discussed with: arch, rwatson, brooks, davidxu, deischen Reviewed by: antoine Revision Changes Path 1.10 +15 -0 src/lib/libc/sys/Symbol.map 1.1275 +1 -0 src/sys/conf/files 1.291 +2 -0 src/sys/kern/init_main.c 1.1 +907 -0 src/sys/kern/kern_cpuset.c (new) 1.266 +5 -1 src/sys/kern/kern_thread.c 1.238 +13 -0 src/sys/kern/syscalls.master 1.24 +3 -0 src/sys/sys/_types.h 1.1 +146 -0 src/sys/sys/cpuset.h (new) 1.504 +2 -1 src/sys/sys/proc.h 1.98 +4 -0 src/sys/sys/types.h From jeff at FreeBSD.org Sun Mar 2 07:41:11 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Sun Mar 2 07:41:17 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h Message-ID: <200803020741.m227fAoJ039644@repoman.freebsd.org> jeff 2008-03-02 07:41:10 UTC FreeBSD src repository Modified files: sys/kern init_sysent.c syscalls.c systrace_args.c sys/sys syscall.h syscall.mk sysproto.h Log: - Regen for cpuset Sponsored by: Nokia Revision Changes Path 1.234 +6 -1 src/sys/kern/init_sysent.c 1.218 +6 -1 src/sys/kern/syscalls.c 1.18 +48 -0 src/sys/kern/systrace_args.c 1.215 +7 -2 src/sys/sys/syscall.h 1.170 +7 -2 src/sys/sys/syscall.mk 1.219 +39 -1 src/sys/sys/sysproto.h From jeff at FreeBSD.org Sun Mar 2 07:51:30 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Sun Mar 2 07:51:32 2008 Subject: cvs commit: src/usr.bin/cpuset Makefile cpuset.c Message-ID: <200803020751.m227pTiI040061@repoman.freebsd.org> jeff 2008-03-02 07:51:29 UTC FreeBSD src repository Added files: usr.bin/cpuset Makefile cpuset.c Log: Add a simple utility for manipulating cpusets. Man page will be available soon. - Lists of cpus may be specified with -l with ranges specified as low-high and commas between individual cpus and ranges. ie -l 0-2,4,6-8. - cpuset can modified -p pids, -t tids, or -s cpusetids. - cpuset can -g get the current mask for any of the above. Sponsored by: Nokia Revision Changes Path 1.1 +7 -0 src/usr.bin/cpuset/Makefile (new) 1.1 +328 -0 src/usr.bin/cpuset/cpuset.c (new) From jeff at FreeBSD.org Sun Mar 2 07:52:27 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Sun Mar 2 07:52:29 2008 Subject: cvs commit: src/usr.bin Makefile Message-ID: <200803020752.m227qQPv040101@repoman.freebsd.org> jeff 2008-03-02 07:52:26 UTC FreeBSD src repository Modified files: usr.bin Makefile Log: - Add cpuset to the build. Sponsored by: Nokia Revision Changes Path 1.311 +1 -0 src/usr.bin/Makefile From jeff at FreeBSD.org Sun Mar 2 07:58:42 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Sun Mar 2 07:58:45 2008 Subject: cvs commit: src/sys/amd64/amd64 identcpu.c mp_machdep.c src/sys/amd64/include smp.h src/sys/i386/i386 identcpu.c mp_machdep.c src/sys/i386/include smp.h src/sys/ia64/ia64 mp_machdep.c src/sys/kern sched_ule.c subr_smp.c ... Message-ID: <200803020758.m227wgH9040483@repoman.freebsd.org> jeff 2008-03-02 07:58:42 UTC FreeBSD src repository Modified files: sys/amd64/amd64 identcpu.c mp_machdep.c sys/amd64/include smp.h sys/i386/i386 identcpu.c mp_machdep.c sys/i386/include smp.h sys/ia64/ia64 mp_machdep.c sys/kern sched_ule.c subr_smp.c sys/powerpc/powerpc mp_machdep.c sys/sparc64/sparc64 mp_machdep.c sys/sun4v/sun4v mp_machdep.c sys/sys smp.h Log: - Remove the old smp cpu topology specification with a new, more flexible tree structure that encodes the level of cache sharing and other properties. - Provide several convenience functions for creating one and two level cpu trees as well as a default flat topology. The system now always has some topology. - On i386 and amd64 create a seperate level in the hierarchy for HTT and multi-core cpus. This will allow the scheduler to intelligently load balance non-uniform cores. Presently we don't detect what level of the cache hierarchy is shared at each level in the topology. - Add a mechanism for testing common topologies that have more information than the MD code is able to provide via the kern.smp.topology tunable. This should be considered a debugging tool only and not a stable api. Sponsored by: Nokia Revision Changes Path 1.158 +7 -1 src/sys/amd64/amd64/identcpu.c 1.288 +30 -41 src/sys/amd64/amd64/mp_machdep.c 1.92 +4 -1 src/sys/amd64/include/smp.h 1.184 +6 -1 src/sys/i386/i386/identcpu.c 1.283 +30 -41 src/sys/i386/i386/mp_machdep.c 1.91 +4 -1 src/sys/i386/include/smp.h 1.68 +7 -0 src/sys/ia64/ia64/mp_machdep.c 1.225 +1 -64 src/sys/kern/sched_ule.c 1.204 +188 -14 src/sys/kern/subr_smp.c 1.14 +7 -0 src/sys/powerpc/powerpc/mp_machdep.c 1.37 +7 -0 src/sys/sparc64/sparc64/mp_machdep.c 1.9 +7 -0 src/sys/sun4v/sun4v/mp_machdep.c 1.87 +32 -9 src/sys/sys/smp.h From jroberson at chesapeake.net Sun Mar 2 08:03:33 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Sun Mar 2 08:03:38 2008 Subject: cvs commit: src/sys/amd64/amd64 identcpu.c mp_machdep.c src/sys/amd64/include smp.h src/sys/i386/i386 identcpu.c mp_machdep.c src/sys/i386/include smp.h src/sys/ia64/ia64 mp_machdep.c src/sys/kern sched_ule.c subr_smp.c ... In-Reply-To: <200803020758.m227wgH9040483@repoman.freebsd.org> References: <200803020758.m227wgH9040483@repoman.freebsd.org> Message-ID: <20080301220410.P920@desktop> On Sun, 2 Mar 2008, Jeff Roberson wrote: > jeff 2008-03-02 07:58:42 UTC > > FreeBSD src repository > > Modified files: > sys/amd64/amd64 identcpu.c mp_machdep.c > sys/amd64/include smp.h > sys/i386/i386 identcpu.c mp_machdep.c > sys/i386/include smp.h > sys/ia64/ia64 mp_machdep.c > sys/kern sched_ule.c subr_smp.c > sys/powerpc/powerpc mp_machdep.c > sys/sparc64/sparc64 mp_machdep.c > sys/sun4v/sun4v mp_machdep.c > sys/sys smp.h > Log: > - Remove the old smp cpu topology specification with a new, more flexible > tree structure that encodes the level of cache sharing and other Should read 'replace'. > properties. > - Provide several convenience functions for creating one and two level > cpu trees as well as a default flat topology. The system now always > has some topology. > - On i386 and amd64 create a seperate level in the hierarchy for HTT > and multi-core cpus. This will allow the scheduler to intelligently > load balance non-uniform cores. Presently we don't detect what level > of the cache hierarchy is shared at each level in the topology. s/seperate/separate/ Hopefully I coded better than I spelled. Thanks, Jeff > - Add a mechanism for testing common topologies that have more information > than the MD code is able to provide via the kern.smp.topology tunable. > This should be considered a debugging tool only and not a stable api. > > Sponsored by: Nokia > > Revision Changes Path > 1.158 +7 -1 src/sys/amd64/amd64/identcpu.c > 1.288 +30 -41 src/sys/amd64/amd64/mp_machdep.c > 1.92 +4 -1 src/sys/amd64/include/smp.h > 1.184 +6 -1 src/sys/i386/i386/identcpu.c > 1.283 +30 -41 src/sys/i386/i386/mp_machdep.c > 1.91 +4 -1 src/sys/i386/include/smp.h > 1.68 +7 -0 src/sys/ia64/ia64/mp_machdep.c > 1.225 +1 -64 src/sys/kern/sched_ule.c > 1.204 +188 -14 src/sys/kern/subr_smp.c > 1.14 +7 -0 src/sys/powerpc/powerpc/mp_machdep.c > 1.37 +7 -0 src/sys/sparc64/sparc64/mp_machdep.c > 1.9 +7 -0 src/sys/sun4v/sun4v/mp_machdep.c > 1.87 +32 -9 src/sys/sys/smp.h > From jeff at FreeBSD.org Sun Mar 2 08:21:00 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Sun Mar 2 08:21:04 2008 Subject: cvs commit: src/sys/kern sched_ule.c Message-ID: <200803020821.m228L0Yw042389@repoman.freebsd.org> jeff 2008-03-02 08:20:59 UTC FreeBSD src repository Modified files: sys/kern sched_ule.c Log: Add support for the new cpu topology api: - When searching for affinity search backwards in the tree from the last cpu we ran on while the thread still has affinity for the group. This can take advantage of knowledge of shared L2 or L3 caches among a group of cores. - When searching for the least loaded cpu find the least loaded cpu via the least loaded path through the tree. This load balances system bus links, individual cache levels, and hyper-threaded/SMT cores. - Make the periodic balancer recursively balance the highest and lowest loaded cpu across each link. Add support for cpusets: - Convert the cpuset to a simple native cpumask_t while the kernel still only supports cpumask. - Pass the derived cpumask down through the cpu_search functions to restrict the result cpus. - Make the various steal functions resilient to failure since all threads can not run on all cpus any longer. General improvements: - Precisely track the lowest priority thread on every runq with tdq_setlowpri(). Before it was more advisory but this ended up having pathological behaviors. - Remove many #ifdef SMP conditions to simplify the code. - Get rid of the old cumbersome tdq_group. This is more naturally expressed via the cpu_group tree. Sponsored by: Nokia Testing by: kris Revision Changes Path 1.226 +443 -501 src/sys/kern/sched_ule.c From jroberson at chesapeake.net Sun Mar 2 08:27:44 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Sun Mar 2 08:27:52 2008 Subject: cvs commit: src/sys/kern sched_ule.c In-Reply-To: <200803020821.m228L0Yw042389@repoman.freebsd.org> References: <200803020821.m228L0Yw042389@repoman.freebsd.org> Message-ID: <20080301222513.Y920@desktop> On Sun, 2 Mar 2008, Jeff Roberson wrote: > jeff 2008-03-02 08:20:59 UTC > > FreeBSD src repository > > Modified files: > sys/kern sched_ule.c > Log: > Add support for the new cpu topology api: > - When searching for affinity search backwards in the tree from the last > cpu we ran on while the thread still has affinity for the group. This > can take advantage of knowledge of shared L2 or L3 caches among a > group of cores. > - When searching for the least loaded cpu find the least loaded cpu via > the least loaded path through the tree. This load balances system bus > links, individual cache levels, and hyper-threaded/SMT cores. > - Make the periodic balancer recursively balance the highest and lowest > loaded cpu across each link. > > Add support for cpusets: > - Convert the cpuset to a simple native cpumask_t while the kernel still > only supports cpumask. > - Pass the derived cpumask down through the cpu_search functions to > restrict the result cpus. > - Make the various steal functions resilient to failure since all threads > can not run on all cpus any longer. > > General improvements: > - Precisely track the lowest priority thread on every runq with > tdq_setlowpri(). Before it was more advisory but this ended up having > pathological behaviors. > - Remove many #ifdef SMP conditions to simplify the code. > - Get rid of the old cumbersome tdq_group. This is more naturally > expressed via the cpu_group tree. > With these changes ULE is the only scheduler that supports the new cpuset api. It succeeds on 4BSD but the scheduler doesn't obey the masks. I don't presently have a plan to implement it on 4BSD as it will be potentially very inefficient to search the runq for a compatible thread on every context switch. I won't object if someone else wants to implement this, otherwise I'll make the syscalls return ENOSYS if 4BSD is compiled in. The improved cpu topology load balancing is a mixed bag. On some workloads we see considerable improvements. Right now mysql suffers when it has large numbers of threads but other things seem much improved. I will be continuing to tune this however and in most cases it's a win already. Kris has done some excellent benchmarking as usual. Here you can see the improvement in postgres depending on various scheduler debug settings: http://people.freebsd.org/~kris/scaling/pgsql-16cpu.png The horrible green line is 7.0 for reference. The blue line is the same 16core machine with half of the cores disabled. Thanks, Jeff > Sponsored by: Nokia > Testing by: kris > > Revision Changes Path > 1.226 +443 -501 src/sys/kern/sched_ule.c > From bz at FreeBSD.org Sun Mar 2 08:40:53 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Mar 2 08:40:54 2008 Subject: cvs commit: src/sys/netinet tcp_input.c Message-ID: <200803020840.m228elUH043992@repoman.freebsd.org> bz 2008-03-02 08:40:47 UTC FreeBSD src repository Modified files: sys/netinet tcp_input.c Log: Some "cleanup" of tcp_mss(): - Move the assigment of the socket down before we first need it. No need to do it at the beginning and then drop out the function by one of the returns before using it 100 lines further down. - Use t_maxopd which was assigned the "tcp_mssdflt" for the corrrect AF already instead of another #ifdef ? : #endif block doing the same. - Remove an unneeded (duplicate) assignment of mss to t_maxseg just before we possibly change mss and re-do the assignment without using t_maxseg in between. Reviewed by: silby No objections: net@ (silence) MFC after: 5 days Revision Changes Path 1.373 +4 -8 src/sys/netinet/tcp_input.c From blackend at FreeBSD.org Sun Mar 2 08:54:33 2008 From: blackend at FreeBSD.org (Marc Fonvieille) Date: Sun Mar 2 08:54:35 2008 Subject: cvs commit: CVSROOT access.doc_src Message-ID: <200803020854.m228sWKJ053463@repoman.freebsd.org> blackend 2008-03-02 08:54:32 UTC FreeBSD src repository (doc committer) Modified files: . access.doc_src Log: Changes to email forwards can be made without doceng's approval. Please use "Approved by: core (implicit), doceng (implicit)" to that type of commits. Approved by: core (implicit), doceng (implicit) Revision Changes Path 1.46 +3 -0 CVSROOT/access.doc_src From blackend at FreeBSD.org Sun Mar 2 08:57:45 2008 From: blackend at FreeBSD.org (Marc Fonvieille) Date: Sun Mar 2 08:57:47 2008 Subject: cvs commit: CVSROOT access.doc_src Message-ID: <200803020857.m228viuj053764@repoman.freebsd.org> blackend 2008-03-02 08:57:44 UTC FreeBSD src repository (doc committer) Modified files: . access.doc_src Log: Add missing "core (implicit)," to the text. Pointy Hat: blackend (implicit) Approved by: core (implicit), doceng (implicit) Revision Changes Path 1.47 +2 -1 CVSROOT/access.doc_src From marck at rinet.ru Sun Mar 2 09:01:11 2008 From: marck at rinet.ru (Dmitry Morozovsky) Date: Sun Mar 2 09:01:17 2008 Subject: cvs commit: CVSROOT access.doc_src In-Reply-To: <200803020857.m228viuj053764@repoman.freebsd.org> References: <200803020857.m228viuj053764@repoman.freebsd.org> Message-ID: <20080302120034.L49885@woozle.rinet.ru> On Sun, 2 Mar 2008, Marc Fonvieille wrote: MF> blackend 2008-03-02 08:57:44 UTC MF> MF> FreeBSD src repository (doc committer) MF> MF> Modified files: MF> . access.doc_src MF> Log: MF> Add missing "core (implicit)," to the text. MF> MF> Pointy Hat: blackend (implicit) MF> Approved by: core (implicit), doceng (implicit) Implicit approval of impicit passing the implicit Pointy Hat [tm]? ;-P Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: marck@FreeBSD.org ] ------------------------------------------------------------------------ *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@rinet.ru *** ------------------------------------------------------------------------ From blackend at FreeBSD.org Sun Mar 2 09:11:01 2008 From: blackend at FreeBSD.org (Marc Fonvieille) Date: Sun Mar 2 09:11:05 2008 Subject: cvs commit: CVSROOT access.doc_src In-Reply-To: <200803020854.m228sWKJ053463@repoman.freebsd.org> References: <200803020854.m228sWKJ053463@repoman.freebsd.org> Message-ID: <20080302085614.GB1048@gothic.blackend.org> On Sun, Mar 02, 2008 at 08:54:32AM +0000, Marc Fonvieille wrote: > blackend 2008-03-02 08:54:32 UTC > > FreeBSD src repository (doc committer) > > Modified files: > . access.doc_src > Log: > Changes to email forwards can be made without doceng's approval. > Please use "Approved by: core (implicit), doceng (implicit)" to that > type of commits. > > Approved by: core (implicit), doceng (implicit) > > Revision Changes Path > 1.46 +3 -0 CVSROOT-src/access.doc_src > > http://cvsweb.FreeBSD.org/CVSROOT-src/access.doc_src.diff?r1=1.45&r2=1.46 > | --- CVSROOT-src/access.doc_src 2007/03/01 11:52:22 1.45 > | +++ CVSROOT-src/access.doc_src 2008/03/02 08:54:32 1.46 > | @@ -1,5 +1,5 @@ > | # > | -# $FreeBSD: /usr/local/www/cvsroot/FreeBSD/CVSROOT-src/access.doc_src,v 1.45 2007/03/01 11:52:22 blackend Exp $ > | +# $FreeBSD: /usr/local/www/cvsroot/FreeBSD/CVSROOT-src/access.doc_src,v 1.46 2008/03/02 08:54:32 blackend Exp $ > | # > | # This is for folks who are doc committers, but who work on docs in > | # the src/ tree as well. To be clear, this file is in addition to > | @@ -12,6 +12,9 @@ > | # > | # KEEP ALPHABETICALLY SORTED!!! > | # > | +# Changes to email forwards can be made without doceng's approval. > | +# Please use "Approved by: doceng (implicit)" to that type of commits. Sorry, ^^ I had to add "core (implicit)," :( > | +# > | blackend > | chern > | jcamou -- Marc From blackend at freebsd.org Sun Mar 2 09:11:49 2008 From: blackend at freebsd.org (Marc Fonvieille) Date: Sun Mar 2 09:11:55 2008 Subject: cvs commit: CVSROOT access.doc_src In-Reply-To: <20080302120034.L49885@woozle.rinet.ru> References: <200803020857.m228viuj053764@repoman.freebsd.org> <20080302120034.L49885@woozle.rinet.ru> Message-ID: <20080302091147.GC1048@gothic.blackend.org> On Sun, Mar 02, 2008 at 12:01:09PM +0300, Dmitry Morozovsky wrote: > On Sun, 2 Mar 2008, Marc Fonvieille wrote: > > MF> blackend 2008-03-02 08:57:44 UTC > MF> > MF> FreeBSD src repository (doc committer) > MF> > MF> Modified files: > MF> . access.doc_src > MF> Log: > MF> Add missing "core (implicit)," to the text. > MF> > MF> Pointy Hat: blackend (implicit) > MF> Approved by: core (implicit), doceng (implicit) > > Implicit approval of impicit passing the implicit Pointy Hat [tm]? ;-P > Hehe, this one was funny :) the curse of the sunday morning commit... -- Marc From cperciva at freebsd.org Sun Mar 2 10:12:17 2008 From: cperciva at freebsd.org (Colin Percival) Date: Sun Mar 2 10:12:21 2008 Subject: cvs commit: CVSROOT access.doc_src In-Reply-To: <20080302120034.L49885@woozle.rinet.ru> References: <200803020857.m228viuj053764@repoman.freebsd.org> <20080302120034.L49885@woozle.rinet.ru> Message-ID: <47CA6F62.7080901@freebsd.org> Dmitry Morozovsky wrote: > On Sun, 2 Mar 2008, Marc Fonvieille wrote: > MF> Add missing "core (implicit)," to the text. > MF> > MF> Pointy Hat: blackend (implicit) > MF> Approved by: core (implicit), doceng (implicit) > > Implicit approval of impicit passing the implicit Pointy Hat [tm]? ;-P Are you trying to elicit an admission that Marc is complicit in an illicit attempt to solicit explicit "implicit"s? Colin "perfectly licit" Percival From peterjeremy at optushome.com.au Sun Mar 2 11:00:11 2008 From: peterjeremy at optushome.com.au (Peter Jeremy) Date: Sun Mar 2 11:00:19 2008 Subject: cvs commit: src/sys/kern sched_ule.c In-Reply-To: <20080301222513.Y920@desktop> References: <200803020821.m228L0Yw042389@repoman.freebsd.org> <20080301222513.Y920@desktop> Message-ID: <20080302105958.GF67687@server.vk2pj.dyndns.org> On Sat, Mar 01, 2008 at 10:29:50PM -1000, Jeff Roberson wrote: >With these changes ULE is the only scheduler that supports the new cpuset >api. Excellent work. I didn't expect it to be implemented so quickly. > It succeeds on 4BSD but the scheduler doesn't obey the masks. I don't >presently have a plan to implement it on 4BSD as it will be potentially >very inefficient to search the runq for a compatible thread on every >context switch. I won't object if someone else wants to implement this, >otherwise I'll make the syscalls return ENOSYS if 4BSD is compiled in. I would prefer to see the project devote available resources to improving ULE - with a view to deprecating 4BSD ASAP - rather than retrofitting new features into 4BSD. IMHO, it's not clear whether requests via the cpuset API should be mandatory or advisory - I believe valid cases can be made for either approach. In the latter case, it would be more reasonable for the cpuset implementation on 4BSD to just be a no-op, rather than failing. >Kris has done some excellent benchmarking as usual. Here you can see the >improvement in postgres depending on various scheduler debug settings: > >http://people.freebsd.org/~kris/scaling/pgsql-16cpu.png The improvement is quite substantial. Congratulations Jeff. -- Peter Jeremy Please excuse any delays as the result of my ISP's inability to implement an MTA that is either RFC2821-compliant or matches their claimed behaviour. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080302/b664c4b1/attachment.pgp From peterjeremy at optushome.com.au Sun Mar 2 11:02:43 2008 From: peterjeremy at optushome.com.au (Peter Jeremy) Date: Sun Mar 2 11:02:51 2008 Subject: cvs commit: CVSROOT access.doc_src In-Reply-To: <47CA6F62.7080901@freebsd.org> References: <200803020857.m228viuj053764@repoman.freebsd.org> <20080302120034.L49885@woozle.rinet.ru> <47CA6F62.7080901@freebsd.org> Message-ID: <20080302110235.GG67687@server.vk2pj.dyndns.org> On Sun, Mar 02, 2008 at 01:12:02AM -0800, Colin Percival wrote: >Are you trying to elicit an admission that Marc is complicit in an illicit >attempt to solicit explicit "implicit"s? > >Colin "perfectly licit" Percival Another one for the fortune file I think... -- Peter Jeremy Please excuse any delays as the result of my ISP's inability to implement an MTA that is either RFC2821-compliant or matches their claimed behaviour. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080302/da1dc089/attachment.pgp From kostikbel at gmail.com Sun Mar 2 11:07:33 2008 From: kostikbel at gmail.com (Kostik Belousov) Date: Sun Mar 2 11:07:40 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <200803020739.m227dNLe039427@repoman.freebsd.org> References: <200803020739.m227dNLe039427@repoman.freebsd.org> Message-ID: <20080302110726.GI57756@deviant.kiev.zoral.com.ua> On Sun, Mar 02, 2008 at 07:39:22AM +0000, Jeff Roberson wrote: > jeff 2008-03-02 07:39:22 UTC > > FreeBSD src repository > > Modified files: > sys/conf files > sys/kern init_main.c kern_thread.c syscalls.master > sys/sys _types.h types.h proc.h > lib/libc/sys Symbol.map > Added files: > sys/kern kern_cpuset.c > sys/sys cpuset.h > Log: > Add cpuset, an api for thread to cpu binding and cpu resource grouping > and assignment. > - Add a reference to a struct cpuset in each thread that is inherited from > the thread that created it. > - Release the reference when the thread is destroyed. > - Add prototypes for syscalls and macros for manipulating cpusets in > sys/cpuset.h > - Add syscalls to create, get, and set new numbered cpusets: > cpuset(), cpuset_{get,set}id() > - Add syscalls for getting and setting affinity masks for cpusets or > individual threads: cpuid_{get,set}affinity() > - Add types for the 'level' and 'which' parameters for the cpuset. This > will permit expansion of the api to cover cpu masks for other objects > identifiable with an id_t integer. For example, IRQs and Jails may be > coming soon. > - The root set 0 contains all valid cpus. All thread initially belong to > cpuset 1. This permits migrating all threads off of certain cpus to > reserve them for special applications. > > Sponsored by: Nokia > Discussed with: arch, rwatson, brooks, davidxu, deischen > Reviewed by: antoine > > Revision Changes Path > 1.10 +15 -0 src/lib/libc/sys/Symbol.map > 1.1275 +1 -0 src/sys/conf/files > 1.291 +2 -0 src/sys/kern/init_main.c > 1.1 +907 -0 src/sys/kern/kern_cpuset.c (new) > 1.266 +5 -1 src/sys/kern/kern_thread.c > 1.238 +13 -0 src/sys/kern/syscalls.master > 1.24 +3 -0 src/sys/sys/_types.h > 1.1 +146 -0 src/sys/sys/cpuset.h (new) > 1.504 +2 -1 src/sys/sys/proc.h > 1.98 +4 -0 src/sys/sys/types.h Is the ommission of the 32-bits compat syscalls intended ? -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080302/f62327fa/attachment.pgp From ru at FreeBSD.org Sun Mar 2 11:10:47 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Sun Mar 2 11:10:50 2008 Subject: cvs commit: src Makefile.inc1 src/gnu/usr.bin/binutils/ar Makefile src/gnu/usr.bin/binutils/ranlib Makefile Message-ID: <200803021110.m22BAlaI065688@repoman.freebsd.org> ru 2008-03-02 11:10:47 UTC FreeBSD src repository Modified files: . Makefile.inc1 gnu/usr.bin/binutils/ar Makefile gnu/usr.bin/binutils/ranlib Makefile Log: Set AR=gnu-ar and RANLIB=gnu-ranlib on systems where we are forced to use GNU tools. Remove the _WITH_GNUAR knob. Prodded by: obrien Revision Changes Path 1.601 +3 -4 src/Makefile.inc1 1.18 +0 -2 src/gnu/usr.bin/binutils/ar/Makefile 1.19 +0 -2 src/gnu/usr.bin/binutils/ranlib/Makefile From philip at FreeBSD.org Sun Mar 2 11:33:55 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sun Mar 2 11:33:57 2008 Subject: cvs commit: src/sys/dev/md md.c Message-ID: <200803021133.m22BXsJJ066474@repoman.freebsd.org> philip 2008-03-02 11:33:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/md md.c Log: MFC r1.174: Zero sc->vnode if mdsetcred() fails. This fixes the panic which happens when mdcreate_vnode() calls vn_close() and mddestroy() calls it again further down the error handling path. Revision Changes Path 1.169.2.1 +1 -0 src/sys/dev/md/md.c From akiyama at FreeBSD.org Sun Mar 2 13:02:56 2008 From: akiyama at FreeBSD.org (Shunsuke Akiyama) Date: Sun Mar 2 13:02:58 2008 Subject: cvs commit: src/sys/dev/usb ucom.c Message-ID: <200803021302.m22D2um9084261@repoman.freebsd.org> akiyama 2008-03-02 13:02:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/usb ucom.c Log: MFC: rev 1.65. o Add module event handler. o Update copyright year. Revision Changes Path 1.64.2.1 +19 -2 src/sys/dev/usb/ucom.c From akiyama at FreeBSD.org Sun Mar 2 13:07:03 2008 From: akiyama at FreeBSD.org (Shunsuke Akiyama) Date: Sun Mar 2 13:07:06 2008 Subject: cvs commit: src/sys/dev/usb ucom.c Message-ID: <200803021307.m22D73Yt085651@repoman.freebsd.org> akiyama 2008-03-02 13:07:03 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/usb ucom.c Log: MFC: rev 1.65. o Add module event handler. o Update copyright year. Revision Changes Path 1.57.2.2 +19 -2 src/sys/dev/usb/ucom.c From kib at FreeBSD.org Sun Mar 2 14:00:51 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sun Mar 2 14:00:53 2008 Subject: cvs commit: src/sys/compat/linux linux_futex.c Message-ID: <200803021400.m22E0oiS088150@repoman.freebsd.org> kib 2008-03-02 14:00:50 UTC FreeBSD src repository Modified files: sys/compat/linux linux_futex.c Log: Return ENOSYS instead of 0 for the unknown futex operations. Submitted by: rdivacky Reported and tested by: Gary Stanley Revision Changes Path 1.11 +1 -1 src/sys/compat/linux/linux_futex.c From rdivacky at FreeBSD.org Sun Mar 2 14:02:19 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Sun Mar 2 14:02:22 2008 Subject: cvs commit: src/sys/compat/linux linux_futex.c In-Reply-To: <200803021400.m22E0oiS088150@repoman.freebsd.org> References: <200803021400.m22E0oiS088150@repoman.freebsd.org> Message-ID: <20080302140213.GA39790@freebsd.org> On Sun, Mar 02, 2008 at 02:00:50PM +0000, Konstantin Belousov wrote: > kib 2008-03-02 14:00:50 UTC > > FreeBSD src repository > > Modified files: > sys/compat/linux linux_futex.c > Log: > Return ENOSYS instead of 0 for the unknown futex operations. > > Submitted by: rdivacky > Reported and tested by: Gary Stanley thnx for the commit.. will you MFC this? From rwatson at FreeBSD.org Sun Mar 2 14:54:49 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 2 14:54:51 2008 Subject: cvs commit: src/sys/dev/re if_re.c Message-ID: <200803021454.m22EsmHe091411@repoman.freebsd.org> rwatson 2008-03-02 14:54:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/re if_re.c Log: Conditionally acquire Giant based on debug.mpsafenet around entry points from if_re taskqueue and other potentially Giant-free spots. If we don't do this, Giant may not be held entering KAME IPSEC, etc. This problem appeared in FreeBSD 6.2 as a result of a move to fast interrupts, and does not exist in 7.x due to not having debug.mpsafenet. PR: 118719 Reported by: Dan Lukes Reviwed by: yongari Revision Changes Path 1.46.2.39 +6 -0 src/sys/dev/re/if_re.c From wearabnet at yahoo.ca Sun Mar 2 16:31:36 2008 From: wearabnet at yahoo.ca (Abdullah Ibn Hamad Al-Marri) Date: Sun Mar 2 16:31:40 2008 Subject: cvs commit: src/sys/kern sched_ule.c Message-ID: <688753.81849.qm@web33702.mail.mud.yahoo.com> ----- Original Message ---- > From: Jeff Roberson > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; cvs-all@FreeBSD.org > Sent: Sunday, March 2, 2008 11:21:00 AM > Subject: cvs commit: src/sys/kern sched_ule.c > > jeff 2008-03-02 08:20:59 UTC > > FreeBSD src repository > > Modified files: > sys/kern sched_ule.c > Log: > Add support for the new cpu topology api: > - When searching for affinity search backwards in the tree from the last > cpu we ran on while the thread still has affinity for the group. This > can take advantage of knowledge of shared L2 or L3 caches among a > group of cores. > - When searching for the least loaded cpu find the least loaded cpu via > the least loaded path through the tree. This load balances system bus > links, individual cache levels, and hyper-threaded/SMT cores. > - Make the periodic balancer recursively balance the highest and lowest > loaded cpu across each link. > > Add support for cpusets: > - Convert the cpuset to a simple native cpumask_t while the kernel still > only supports cpumask. > - Pass the derived cpumask down through the cpu_search functions to > restrict the result cpus. > - Make the various steal functions resilient to failure since all threads > can not run on all cpus any longer. > > General improvements: > - Precisely track the lowest priority thread on every runq with > tdq_setlowpri(). Before it was more advisory but this ended up having > pathological behaviors. > - Remove many #ifdef SMP conditions to simplify the code. > - Get rid of the old cumbersome tdq_group. This is more naturally > expressed via the cpu_group tree. > > Sponsored by: Nokia > Testing by: kris > > Revision Changes Path > 1.226 +443 -501 src/sys/kern/sched_ule.c > _______________________________________________ Is it possible to have patch for RELENG_7? Regards, -Abdullah Ibn Hamad Al-Marri Arab Portal http://www.WeArab.Net/ ____________________________________________________________________________________ Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ From raj at FreeBSD.org Sun Mar 2 17:06:03 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Sun Mar 2 17:06:05 2008 Subject: cvs commit: src/sys/powerpc/aim machdep.c swtch.S trap.c trap_subr.S vm_machdep.c src/sys/powerpc/include frame.h hid.h kdb.h pcb.h pcpu.h src/sys/powerpc/powerpc db_trace.c genassym.c Message-ID: <200803021705.m22H5vbf008691@repoman.freebsd.org> raj 2008-03-02 17:05:57 UTC FreeBSD src repository Modified files: sys/powerpc/aim machdep.c swtch.S trap.c trap_subr.S vm_machdep.c sys/powerpc/include frame.h hid.h kdb.h pcb.h pcpu.h sys/powerpc/powerpc db_trace.c genassym.c Log: Unify and generalize PowerPC headers, adjust AIM code accordingly. Rework of this area is a pre-requirement for importing e500 support (and other PowerPC core variations in the future). Mainly the following headers are refactored so that we can cover for low-level differences between various machines within PowerPC architecture: Areas which use the above are adjusted and cleaned up. Credits for this rework go to marcel@ Approved by: cognet (mentor) MFp4: e500 Revision Changes Path 1.109 +18 -3 src/sys/powerpc/aim/machdep.c 1.24 +2 -2 src/sys/powerpc/aim/swtch.S 1.71 +7 -7 src/sys/powerpc/aim/trap.c 1.18 +12 -12 src/sys/powerpc/aim/trap_subr.S 1.119 +2 -2 src/sys/powerpc/aim/vm_machdep.c 1.13 +15 -5 src/sys/powerpc/include/frame.h 1.5 +39 -34 src/sys/powerpc/include/hid.h 1.5 +4 -11 src/sys/powerpc/include/kdb.h 1.10 +11 -3 src/sys/powerpc/include/pcb.h 1.25 +55 -4 src/sys/powerpc/include/pcpu.h 1.15 +20 -10 src/sys/powerpc/powerpc/db_trace.c 1.61 +55 -12 src/sys/powerpc/powerpc/genassym.c From bz at FreeBSD.org Sun Mar 2 17:12:29 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Mar 2 17:12:32 2008 Subject: cvs commit: src/sys/netipsec key.c key_debug.c Message-ID: <200803021712.m22HCT7d009087@repoman.freebsd.org> bz 2008-03-02 17:12:29 UTC FreeBSD src repository Modified files: sys/netipsec key.c key_debug.c Log: Fix bugs when allocating and passing information of current lifetime and soft lifetime [1] introduced in rev. 1.21 of key.c. Along with that, fix a related problem in key_debug printing the correct data. While there replace a printf by panic in a sanity check. PR: 120751 Submitted by: Kazuaki ODA (kazuaki aliceblue.jp) [1] MFC after: 5 days Revision Changes Path 1.29 +7 -2 src/sys/netipsec/key.c 1.6 +22 -4 src/sys/netipsec/key_debug.c From rwatson at FreeBSD.org Sun Mar 2 18:56:14 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 2 18:56:16 2008 Subject: cvs commit: src/sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c Message-ID: <200803021856.m22IuEAx019455@repoman.freebsd.org> rwatson 2008-03-02 18:56:14 UTC FreeBSD src repository Modified files: sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c Log: Replace lockmgr lock protecting smbfs node hash table with sx lock. MFC after: 1 month Revision Changes Path 1.13 +3 -1 src/sys/fs/smbfs/smbfs.h 1.38 +3 -5 src/sys/fs/smbfs/smbfs_node.c 1.44 +4 -3 src/sys/fs/smbfs/smbfs_vfsops.c From rwatson at FreeBSD.org Sun Mar 2 19:02:31 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 2 19:02:33 2008 Subject: cvs commit: src/sys/fs/nwfs nwfs_node.c Message-ID: <200803021902.m22J2U6O019672@repoman.freebsd.org> rwatson 2008-03-02 19:02:30 UTC FreeBSD src repository Modified files: sys/fs/nwfs nwfs_node.c Log: Replace lockmgr lock protecting nwfs vnode hash table with an sx lock. MFC after: 1 month Revision Changes Path 1.43 +15 -12 src/sys/fs/nwfs/nwfs_node.c From attilio at freebsd.org Sun Mar 2 19:28:17 2008 From: attilio at freebsd.org (Attilio Rao) Date: Sun Mar 2 19:28:20 2008 Subject: cvs commit: src/sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c In-Reply-To: <200803021856.m22IuEAx019455@repoman.freebsd.org> References: <200803021856.m22IuEAx019455@repoman.freebsd.org> Message-ID: <3bbf2fe10803021128s586ef668n38a426005a07177f@mail.gmail.com> 2008/3/2, Robert Watson : > rwatson 2008-03-02 18:56:14 UTC > > FreeBSD src repository > > Modified files: > sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c > Log: > Replace lockmgr lock protecting smbfs node hash table with sx lock. Thanks a lot for this and other similar changes! Attilio -- Peace can only be achieved by understanding - A. Einstein From brueffer at FreeBSD.org Sun Mar 2 20:07:49 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sun Mar 2 20:07:51 2008 Subject: cvs commit: src/share/man/man9 lock.9 Message-ID: <200803022007.m22K7mOS025355@repoman.freebsd.org> brueffer 2008-03-02 20:07:48 UTC FreeBSD src repository Modified files: share/man/man9 lock.9 Log: New sentence -> new line. Revision Changes Path 1.30 +2 -1 src/share/man/man9/lock.9 From das at FreeBSD.org Sun Mar 2 20:16:56 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Mar 2 20:17:00 2008 Subject: cvs commit: src/lib/msun/ld128 s_nanl.c Message-ID: <200803022016.m22KGuZA025693@repoman.freebsd.org> das 2008-03-02 20:16:56 UTC FreeBSD src repository Modified files: lib/msun/ld128 s_nanl.c Log: 1 << 47 needs to be written 1ULL << 47. Revision Changes Path 1.3 +1 -1 src/lib/msun/ld128/s_nanl.c From yar at comp.chem.msu.su Sun Mar 2 20:33:00 2008 From: yar at comp.chem.msu.su (Yar Tikhiy) Date: Sun Mar 2 20:33:04 2008 Subject: cvs commit: src/sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c In-Reply-To: <200803021856.m22IuEAx019455@repoman.freebsd.org> References: <200803021856.m22IuEAx019455@repoman.freebsd.org> Message-ID: On Sun, Mar 2, 2008 at 9:56 PM, Robert Watson wrote: > rwatson 2008-03-02 18:56:14 UTC > > FreeBSD src repository > > Modified files: > sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c > Log: > Replace lockmgr lock protecting smbfs node hash table with sx lock. This commit message is a textbook example of how not to write commit messages: It accurately re-tells what was obvious from the diff anyway but provides no pointers to the background for the change. Sorry for my bitching, but I just couldn't resist... :-) -- Yar From rwatson at FreeBSD.org Sun Mar 2 20:39:19 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 2 20:39:37 2008 Subject: cvs commit: src/sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c In-Reply-To: References: <200803021856.m22IuEAx019455@repoman.freebsd.org> Message-ID: <20080302203651.T31090@fledge.watson.org> On Sun, 2 Mar 2008, Yar Tikhiy wrote: > On Sun, Mar 2, 2008 at 9:56 PM, Robert Watson wrote: >> rwatson 2008-03-02 18:56:14 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/fs/smbfs smbfs.h smbfs_node.c smbfs_vfsops.c >> Log: >> Replace lockmgr lock protecting smbfs node hash table with sx lock. > > This commit message is a textbook example of how not to write commit > messages: It accurately re-tells what was obvious from the diff anyway but > provides no pointers to the background for the change. Sorry for my > bitching, but I just couldn't resist... :-) This change, the the similar change in nwfs, are part of a longer running effort to move towards using our SMPng locks rather than the less well-optimized and generally less pretty pre-SMPng primitives. I wouldn't expect any noticeable improvement in performance, and ideally no change in behavior, but it does reduce the number of unnecessary lockmgr consumers. In this context, my definition of "necessary" is where a consumer relies on specific semantics of lockmgr to function, such as LK_DRAIN, LK_UPGRADE, etc, which don't have comparable behavior with the newer primitives. Robert N M Watson Computer Laboratory University of Cambridge From das at FreeBSD.org Sun Mar 2 20:49:25 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Mar 2 20:49:28 2008 Subject: cvs commit: src/tools/regression/lib/msun test-trig.c Message-ID: <200803022049.m22KnPeq036091@repoman.freebsd.org> das 2008-03-02 20:49:25 UTC FreeBSD src repository Modified files: tools/regression/lib/msun test-trig.c Log: Don't run tests that assume <= 64-bit precision on machines with quad precision. Revision Changes Path 1.2 +20 -8 src/tools/regression/lib/msun/test-trig.c From rwatson at FreeBSD.org Sun Mar 2 21:34:17 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 2 21:34:27 2008 Subject: cvs commit: src/sys/security/audit audit_trigger.c Message-ID: <200803022134.m22LYHwo038821@repoman.freebsd.org> rwatson 2008-03-02 21:34:17 UTC FreeBSD src repository Modified files: sys/security/audit audit_trigger.c Log: Rather than copying out the full audit trigger record, which includes a queue entry field, just copy out the unsigned int that is the trigger message. In practice, auditd always requested sizeof(unsigned int), so the extra bytes were ignored, but copying them out was not the intent. MFC after: 1 month Revision Changes Path 1.7 +1 -1 src/sys/security/audit/audit_trigger.c From marcel at FreeBSD.org Sun Mar 2 21:34:58 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Mar 2 21:35:01 2008 Subject: cvs commit: src/sys/kern sched_4bsd.c Message-ID: <200803022134.m22LYvVS038901@repoman.freebsd.org> marcel 2008-03-02 21:34:57 UTC FreeBSD src repository Modified files: sys/kern sched_4bsd.c Log: Unbreak after cpuset: initialize td_cpuset in sched_fork_thread(). Revision Changes Path 1.114 +2 -0 src/sys/kern/sched_4bsd.c From fanf at FreeBSD.org Sun Mar 2 22:27:50 2008 From: fanf at FreeBSD.org (Tony Finch) Date: Sun Mar 2 22:27:52 2008 Subject: cvs commit: src/usr.bin/unifdef unifdef.c Message-ID: <200803022227.m22MRopE041905@repoman.freebsd.org> fanf 2008-03-02 22:27:49 UTC FreeBSD src repository Modified files: usr.bin/unifdef unifdef.c Log: Typo in comment spotted by Hasso Tepper. Obtained from: DragonFlyBSD Revision Changes Path 1.23 +2 -2 src/usr.bin/unifdef/unifdef.c From rwatson at FreeBSD.org Sun Mar 2 22:52:15 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 2 22:52:17 2008 Subject: cvs commit: src/sys/ufs/ufs ufs_extattr.c Message-ID: <200803022252.m22MqEKs042786@repoman.freebsd.org> rwatson 2008-03-02 22:52:14 UTC FreeBSD src repository Modified files: sys/ufs/ufs ufs_extattr.c Log: Don't auto-start or allow extattrctl for UFS2 file systems, as UFS2 has native extended attributes. This didn't interfere with the operation of UFS2 extended attributes, but the code shouldn't be running for UFS2. MFC after: 2 weeks Revision Changes Path 1.90 +18 -0 src/sys/ufs/ufs/ufs_extattr.c From mav at FreeBSD.org Sun Mar 2 23:26:36 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 2 23:26:38 2008 Subject: cvs commit: src/sys/netgraph ng_pppoe.c Message-ID: <200803022326.m22NQZMJ046972@repoman.freebsd.org> mav 2008-03-02 23:26:35 UTC FreeBSD src repository Modified files: sys/netgraph ng_pppoe.c Log: Make session ID generator to use session ID hash. Make session ID generator thread-safe. Revision Changes Path 1.93 +37 -40 src/sys/netgraph/ng_pppoe.c From davidxu at FreeBSD.org Mon Mar 3 03:24:22 2008 From: davidxu at FreeBSD.org (David Xu) Date: Mon Mar 3 03:24:25 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <200803020741.m227fAoJ039644@repoman.freebsd.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> Message-ID: <47CB6FB0.9040602@freebsd.org> Jeff Roberson wrote: > jeff 2008-03-02 07:41:10 UTC > > FreeBSD src repository > > Modified files: > sys/kern init_sysent.c syscalls.c systrace_args.c > sys/sys syscall.h syscall.mk sysproto.h > Log: > - Regen for cpuset > > Sponsored by: Nokia > > Revision Changes Path > 1.234 +6 -1 src/sys/kern/init_sysent.c > 1.218 +6 -1 src/sys/kern/syscalls.c > 1.18 +48 -0 src/sys/kern/systrace_args.c > 1.215 +7 -2 src/sys/sys/syscall.h > 1.170 +7 -2 src/sys/sys/syscall.mk > 1.219 +39 -1 src/sys/sys/sysproto.h > One question is how I can determine the size of cpuset the kernel is using ? Regards, David Xu From yongari at FreeBSD.org Mon Mar 3 03:33:58 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Mon Mar 3 03:34:03 2008 Subject: cvs commit: src/sys/dev/re if_re.c src/sys/pci if_rlreg.h Message-ID: <200803030333.m233XwcJ069091@repoman.freebsd.org> yongari 2008-03-03 03:33:58 UTC FreeBSD src repository Modified files: sys/dev/re if_re.c sys/pci if_rlreg.h Log: Implement WOL. Tested by: Fabian Keil ( freebsd-listen AT fabienkeli DOT de ) Revision Changes Path 1.104 +108 -0 src/sys/dev/re/if_re.c 1.71 +39 -1 src/sys/pci/if_rlreg.h From kmacy at FreeBSD.org Mon Mar 3 03:34:53 2008 From: kmacy at FreeBSD.org (Kip Macy) Date: Mon Mar 3 03:34:55 2008 Subject: cvs commit: src/sys/modules/cxgb/cxgb_t3fw Makefile Message-ID: <200803030334.m233Yqqd069129@repoman.freebsd.org> kmacy 2008-03-03 03:34:52 UTC FreeBSD src repository Modified files: sys/modules/cxgb/cxgb_t3fw Makefile Log: Add KMOD target. Discovered by: Takahashi Yoshihiro Revision Changes Path 1.2 +1 -0 src/sys/modules/cxgb/cxgb_t3fw/Makefile From yongari at FreeBSD.org Mon Mar 3 03:41:07 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Mon Mar 3 03:41:09 2008 Subject: cvs commit: src/sys/dev/re if_re.c src/sys/pci if_rlreg.h Message-ID: <200803030341.m233f6HH069428@repoman.freebsd.org> yongari 2008-03-03 03:41:06 UTC FreeBSD src repository Modified files: sys/dev/re if_re.c sys/pci if_rlreg.h Log: Don't allow jumbo frame on 8139C+ controller. While I'm here add a check for minimal MTU length. Revision Changes Path 1.105 +11 -3 src/sys/dev/re/if_re.c 1.72 +2 -0 src/sys/pci/if_rlreg.h From yongari at FreeBSD.org Mon Mar 3 04:15:09 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Mon Mar 3 04:15:10 2008 Subject: cvs commit: src/sys/dev/re if_re.c src/sys/pci if_rl.c Message-ID: <200803030415.m234F8Ac072319@repoman.freebsd.org> yongari 2008-03-03 04:15:08 UTC FreeBSD src repository Modified files: sys/dev/re if_re.c sys/pci if_rl.c Log: Don't map memory/IO resource in device probe and just use PCI vendor/revision/sub device id of the hardware to probe it. This is the same way as NetBSD does and it enhances readabilty a lot. Revision Changes Path 1.106 +32 -41 src/sys/dev/re/if_re.c 1.173 +18 -41 src/sys/pci/if_rl.c From jkoshy at FreeBSD.org Mon Mar 3 04:29:27 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Mon Mar 3 04:29:32 2008 Subject: cvs commit: src/lib/libelf elf_update.c Message-ID: <200803030429.m234TPnc072927@repoman.freebsd.org> jkoshy 2008-03-03 04:29:25 UTC FreeBSD src repository Modified files: lib/libelf elf_update.c Log: - Fix an off-by-one bug in _libelf_insert_section(). [1] - Update (c) years. Submitted by: kaiw [1] Revision Changes Path 1.4 +11 -5 src/lib/libelf/elf_update.c From jroberson at chesapeake.net Mon Mar 3 04:34:39 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 04:34:43 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <47CB6FB0.9040602@freebsd.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> Message-ID: <20080302183513.P920@desktop> On Mon, 3 Mar 2008, David Xu wrote: > Jeff Roberson wrote: >> jeff 2008-03-02 07:41:10 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/kern init_sysent.c syscalls.c systrace_args.c >> sys/sys syscall.h syscall.mk sysproto.h Log: >> - Regen for cpuset >> Sponsored by: Nokia >> Revision Changes Path >> 1.234 +6 -1 src/sys/kern/init_sysent.c >> 1.218 +6 -1 src/sys/kern/syscalls.c >> 1.18 +48 -0 src/sys/kern/systrace_args.c >> 1.215 +7 -2 src/sys/sys/syscall.h >> 1.170 +7 -2 src/sys/sys/syscall.mk >> 1.219 +39 -1 src/sys/sys/sysproto.h >> > > One question is how I can determine the size of cpuset the kernel is > using ? I wrote it to tolerate user masks that were much larger than the kernel mask. I set the default CPU_SETSIZE in userspace to 128 and in kernel it's MAXCPU. So in practice an application shouldn't have to redefine CPU_SETSIZE. If your set is too small the kernel will return ERANGE however. Unfortunately, if your set is larger than the kernel's CPU_MAXSIZE it'll also return ERANGE. Maybe I should use different errnos for those cases. > > Regards, > David Xu > From julian at elischer.org Mon Mar 3 04:42:37 2008 From: julian at elischer.org (Julian Elischer) Date: Mon Mar 3 04:42:45 2008 Subject: cvs commit: src/sys/kern sched_ule.c In-Reply-To: <20080302105958.GF67687@server.vk2pj.dyndns.org> References: <200803020821.m228L0Yw042389@repoman.freebsd.org> <20080301222513.Y920@desktop> <20080302105958.GF67687@server.vk2pj.dyndns.org> Message-ID: <47CB81BB.1080009@elischer.org> Peter Jeremy wrote: > >> Kris has done some excellent benchmarking as usual. Here you can see the >> improvement in postgres depending on various scheduler debug settings: >> >> http://people.freebsd.org/~kris/scaling/pgsql-16cpu.png > > The improvement is quite substantial. Congratulations Jeff. the drop off after 14 is intersting.. > From davidxu at FreeBSD.org Mon Mar 3 04:45:17 2008 From: davidxu at FreeBSD.org (David Xu) Date: Mon Mar 3 04:45:20 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080302183513.P920@desktop> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> Message-ID: <47CB82A6.4040903@freebsd.org> Jeff Roberson wrote: >> >> One question is how I can determine the size of cpuset the kernel is >> using ? > > I wrote it to tolerate user masks that were much larger than the kernel > mask. I set the default CPU_SETSIZE in userspace to 128 and in kernel > it's MAXCPU. So in practice an application shouldn't have to redefine > CPU_SETSIZE. If your set is too small the kernel will return ERANGE > however. Unfortunately, if your set is larger than the kernel's > CPU_MAXSIZE it'll also return ERANGE. Maybe I should use different > errnos for those cases. > maybe the prototype of cpuset_getaffinity() can be changed to: int cpuset_getaffinity(cpulevel_t, cpuwhich_t, id_t, int *cpusetsize, cpuset_t *); if *cpusetsize is zero, kernel just writes out the actual size the kernel is using ? From kevlo at FreeBSD.org Mon Mar 3 06:39:41 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Mon Mar 3 06:39:44 2008 Subject: cvs commit: src/sys/arm/xscale/ixp425 ixp425_iic.c Message-ID: <200803030639.m236ddlP091591@repoman.freebsd.org> kevlo 2008-03-03 06:39:36 UTC FreeBSD src repository Modified files: sys/arm/xscale/ixp425 ixp425_iic.c Log: Convert to be a 2-clause bsd-only license. Pointed out by: rwatson Revision Changes Path 1.2 +0 -7 src/sys/arm/xscale/ixp425/ixp425_iic.c From davidxu at FreeBSD.org Mon Mar 3 08:28:14 2008 From: davidxu at FreeBSD.org (David Xu) Date: Mon Mar 3 08:28:23 2008 Subject: cvs commit: src/include pthread_np.h Message-ID: <200803030828.m238SDOf098463@repoman.freebsd.org> davidxu 2008-03-03 08:28:13 UTC FreeBSD src repository Modified files: include pthread_np.h Log: Add cpu affinity APIs to get and set thread's cpu affinity mask, the functions are intented to be are compatible with glibc. Revision Changes Path 1.21 +7 -0 src/include/pthread_np.h From ru at FreeBSD.org Mon Mar 3 08:32:58 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Mon Mar 3 08:33:00 2008 Subject: cvs commit: src/usr.bin/find find.1 main.c Message-ID: <200803030832.m238Wwe2098614@repoman.freebsd.org> ru 2008-03-03 08:32:58 UTC FreeBSD src repository Modified files: usr.bin/find find.1 main.c Log: Fix SYNOPSIS and usage(). Revision Changes Path 1.86 +7 -7 src/usr.bin/find/find.1 1.16 +3 -2 src/usr.bin/find/main.c From brueffer at FreeBSD.org Mon Mar 3 08:55:51 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Mon Mar 3 08:55:57 2008 Subject: cvs commit: src/share/man/man4 fwohci.4 Message-ID: <200803030855.m238to8f008959@repoman.freebsd.org> brueffer 2008-03-03 08:55:50 UTC FreeBSD src repository Modified files: share/man/man4 fwohci.4 Log: The Sony i.LINK (CXD1947) is _not_ supported. PR: 116680 Submitted by: Jens Köhler MFC after: 3 days Revision Changes Path 1.17 +1 -3 src/share/man/man4/fwohci.4 From davidxu at FreeBSD.org Mon Mar 3 08:59:02 2008 From: davidxu at FreeBSD.org (David Xu) Date: Mon Mar 3 08:59:05 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <200803020739.m227dNLe039427@repoman.freebsd.org> References: <200803020739.m227dNLe039427@repoman.freebsd.org> Message-ID: <47CBBE21.4060104@freebsd.org> Jeff Roberson wrote: > jeff 2008-03-02 07:39:22 UTC > > FreeBSD src repository > > Modified files: > sys/conf files > sys/kern init_main.c kern_thread.c syscalls.master > sys/sys _types.h types.h proc.h > lib/libc/sys Symbol.map > Added files: > sys/kern kern_cpuset.c > sys/sys cpuset.h > Log: > Add cpuset, an api for thread to cpu binding and cpu resource grouping > and assignment. > - Add a reference to a struct cpuset in each thread that is inherited from > the thread that created it. > - Release the reference when the thread is destroyed. > - Add prototypes for syscalls and macros for manipulating cpusets in > sys/cpuset.h > - Add syscalls to create, get, and set new numbered cpusets: > cpuset(), cpuset_{get,set}id() > - Add syscalls for getting and setting affinity masks for cpusets or > individual threads: cpuid_{get,set}affinity() > - Add types for the 'level' and 'which' parameters for the cpuset. This > will permit expansion of the api to cover cpu masks for other objects > identifiable with an id_t integer. For example, IRQs and Jails may be > coming soon. > - The root set 0 contains all valid cpus. All thread initially belong to > cpuset 1. This permits migrating all threads off of certain cpus to > reserve them for special applications. > > Sponsored by: Nokia > Discussed with: arch, rwatson, brooks, davidxu, deischen > Reviewed by: antoine > The cpuset_setaffinity with command CPU_WHICH_TID may hurt another process if a weird process is out of the control. The current code intents to lookup globally a thread has exact thread id, because thread may be created and exited quickly, and thread ID is reused quickly too, it is possible the weird process gives an outdated thread ID to the API, and an irrelvant thread within another process belongs to same user gets bind to cpus, such problem is very difficult to be diagnosed when it happens, and it is an administrator is nightmare. I think there should be a CPU_WHICH_TID_LOCAL command to limit the searching scope in current process, and in most case, the CPU_WHICH_TID_LOCAL will be used instead. Regards, David Xu From davidxu at FreeBSD.org Mon Mar 3 09:05:12 2008 From: davidxu at FreeBSD.org (David Xu) Date: Mon Mar 3 09:05:24 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <47CBBE21.4060104@freebsd.org> References: <200803020739.m227dNLe039427@repoman.freebsd.org> <47CBBE21.4060104@freebsd.org> Message-ID: <47CBBF91.3020404@freebsd.org> David Xu wrote: > The cpuset_setaffinity with command CPU_WHICH_TID may hurt another > process if a weird process is out of the control. > The current code intents to lookup globally a thread has exact thread > id, because thread may be created and exited quickly, and thread ID is > reused quickly too, it is possible the weird process gives an outdated > thread ID to the API, and an irrelvant thread within another process > belongs to same user gets bind to cpus, such problem is very difficult > to be diagnosed when it happens, and it is an administrator is > nightmare. I think there should be a CPU_WHICH_TID_LOCAL command > to limit the searching scope in current process, and in most case, > the CPU_WHICH_TID_LOCAL will be used instead. > > Regards, > David Xu > > or we always pass pid to cpu_setaffinity. From sobomax at FreeBSD.org Mon Mar 3 09:15:15 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Mon Mar 3 09:15:19 2008 Subject: cvs commit: src/sys/amd64/amd64 identcpu.c mp_machdep.c src/sys/amd64/include smp.h src/sys/i386/i386 identcpu.c mp_machdep.c src/sys/i386/include smp.h src/sys/ia64/ia64 mp_machdep.c src/sys/kern sched_ule.c subr_smp.c ... In-Reply-To: <200803020758.m227wgH9040483@repoman.freebsd.org> References: <200803020758.m227wgH9040483@repoman.freebsd.org> Message-ID: <47CBC127.8010403@FreeBSD.org> Jeff Roberson wrote: > jeff 2008-03-02 07:58:42 UTC > > FreeBSD src repository > > Modified files: > sys/amd64/amd64 identcpu.c mp_machdep.c > sys/amd64/include smp.h > sys/i386/i386 identcpu.c mp_machdep.c > sys/i386/include smp.h > sys/ia64/ia64 mp_machdep.c > sys/kern sched_ule.c subr_smp.c > sys/powerpc/powerpc mp_machdep.c > sys/sparc64/sparc64 mp_machdep.c > sys/sun4v/sun4v mp_machdep.c > sys/sys smp.h > Log: > - Remove the old smp cpu topology specification with a new, more flexible > tree structure that encodes the level of cache sharing and other > properties. > - Provide several convenience functions for creating one and two level > cpu trees as well as a default flat topology. The system now always > has some topology. > - On i386 and amd64 create a seperate level in the hierarchy for HTT > and multi-core cpus. This will allow the scheduler to intelligently > load balance non-uniform cores. Presently we don't detect what level > of the cache hierarchy is shared at each level in the topology. > - Add a mechanism for testing common topologies that have more information > than the MD code is able to provide via the kern.smp.topology tunable. > This should be considered a debugging tool only and not a stable api. Cool! I am just curious if the new topology code is flexible enough to support cores that come and go on the fly? This could be useful in several scenarios, such as for example when running under multi-core aware hypervisor (e.g. Niagara), in the cases when pro-active power manager shuts down some cores to conserve power, in server applications when one of CPUs either has fried or has been unplugged, etc. -Maxim From davidxu at FreeBSD.org Mon Mar 3 09:16:29 2008 From: davidxu at FreeBSD.org (David Xu) Date: Mon Mar 3 09:16:32 2008 Subject: cvs commit: src/lib/libthr pthread.map src/lib/libthr/thread Makefile.inc thr_affinity.c Message-ID: <200803030916.m239GT4q011381@repoman.freebsd.org> davidxu 2008-03-03 09:16:29 UTC FreeBSD src repository Modified files: lib/libthr pthread.map lib/libthr/thread Makefile.inc Added files: lib/libthr/thread thr_affinity.c Log: Implement functions pthread_getaffinity_np and pthread_setaffinity_np to get and set thread's cpu affinity mask. Revision Changes Path 1.27 +3 -1 src/lib/libthr/pthread.map 1.17 +1 -0 src/lib/libthr/thread/Makefile.inc 1.1 +74 -0 src/lib/libthr/thread/thr_affinity.c (new) From rwatson at FreeBSD.org Mon Mar 3 09:25:46 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 3 09:25:50 2008 Subject: cvs commit: src/sys/amd64/amd64 identcpu.c mp_machdep.c src/sys/amd64/include smp.h src/sys/i386/i386 identcpu.c mp_machdep.c src/sys/i386/include smp.h src/sys/ia64/ia64 mp_machdep.c src/sys/kern sched_ule.c subr_smp.c ... In-Reply-To: <47CBC127.8010403@FreeBSD.org> References: <200803020758.m227wgH9040483@repoman.freebsd.org> <47CBC127.8010403@FreeBSD.org> Message-ID: <20080303092335.E66705@fledge.watson.org> On Mon, 3 Mar 2008, Maxim Sobolev wrote: > Cool! I am just curious if the new topology code is flexible enough to > support cores that come and go on the fly? This could be useful in several > scenarios, such as for example when running under multi-core aware > hypervisor (e.g. Niagara), in the cases when pro-active power manager shuts > down some cores to conserve power, in server applications when one of CPUs > either has fried or has been unplugged, etc. We have quite a bit of kernel code that expects CPUs never come and go; I've been hoping to interest people in having a devsummit session on CPU hotplugging for a couple of years now. :-) I don't see physical hotplugging as all that motivating currently, by hypervisor reallocation of CPUs *is* immediately motivating. You'll find assumptions of fixed CPUs all over our kernel -- schedulers, memory allocators, timer management, etc. A mature model for starting and stopping CPUs and allowing kernel subsystems to register with event handlers in order to start/stop their own services is necessary. Similarly, providing a way for user applications that care about CPU placement to hook into the event stream and respond appropriately is important. Robert N M Watson Computer Laboratory University of Cambridge From aryeh.friedman at gmail.com Mon Mar 3 09:28:20 2008 From: aryeh.friedman at gmail.com (Aryeh M. Friedman) Date: Mon Mar 3 09:28:24 2008 Subject: cvs commit: src/sys/amd64/amd64 identcpu.c mp_machdep.c src/sys/amd64/include smp.h src/sys/i386/i386 identcpu.c mp_machdep.c src/sys/i386/include smp.h src/sys/ia64/ia64 mp_machdep.c src/sys/kern sched_ule.c subr_smp.c ... In-Reply-To: <20080303092335.E66705@fledge.watson.org> References: <200803020758.m227wgH9040483@repoman.freebsd.org> <47CBC127.8010403@FreeBSD.org> <20080303092335.E66705@fledge.watson.org> Message-ID: <47CBC4AD.8090100@gmail.com> Robert Watson wrote: > > On Mon, 3 Mar 2008, Maxim Sobolev wrote: > >> Cool! I am just curious if the new topology code is flexible enough >> to support cores that come and go on the fly? This could be useful in >> several scenarios, such as for example when running under multi-core >> aware hypervisor (e.g. Niagara), in the cases when pro-active power >> manager shuts down some cores to conserve power, in server >> applications when one of CPUs either has fried or has been unplugged, >> etc. > > We have quite a bit of kernel code that expects CPUs never come and > go; I've been hoping to interest people in having a devsummit session > on CPU hotplugging for a couple of years now. :-) Grid computing is an other immediate application (depending on how tightly coupled the nodes are of course) From aryeh.friedman at gmail.com Mon Mar 3 09:34:37 2008 From: aryeh.friedman at gmail.com (Aryeh M. Friedman) Date: Mon Mar 3 09:34:42 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <47CBBE21.4060104@freebsd.org> References: <200803020739.m227dNLe039427@repoman.freebsd.org> <47CBBE21.4060104@freebsd.org> Message-ID: <47CBBF1E.8060701@gmail.com> David Xu wrote: > Jeff Roberson wrote: >> jeff 2008-03-02 07:39:22 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/conf files sys/kern init_main.c >> kern_thread.c syscalls.master sys/sys _types.h >> types.h proc.h lib/libc/sys Symbol.map Added files: >> sys/kern kern_cpuset.c sys/sys >> cpuset.h Log: >> Add cpuset, an api for thread to cpu binding and cpu resource grouping >> and assignment. >> - Add a reference to a struct cpuset in each thread that is >> inherited from >> the thread that created it. >> - Release the reference when the thread is destroyed. >> - Add prototypes for syscalls and macros for manipulating cpusets in >> sys/cpuset.h >> - Add syscalls to create, get, and set new numbered cpusets: >> cpuset(), cpuset_{get,set}id() >> - Add syscalls for getting and setting affinity masks for cpusets or >> individual threads: cpuid_{get,set}affinity() >> - Add types for the 'level' and 'which' parameters for the >> cpuset. This >> will permit expansion of the api to cover cpu masks for other >> objects >> identifiable with an id_t integer. For example, IRQs and Jails >> may be >> coming soon. >> - The root set 0 contains all valid cpus. All thread initially >> belong to >> cpuset 1. This permits migrating all threads off of certain >> cpus to >> reserve them for special applications. >> Sponsored by: Nokia >> Discussed with: arch, rwatson, brooks, davidxu, deischen >> Reviewed by: antoine >> > > The cpuset_setaffinity with command CPU_WHICH_TID may hurt another > process if a weird process is out of the control. > The current code intents to lookup globally a thread has exact thread > id, because thread may be created and exited quickly, and thread ID is > reused quickly too, it is possible the weird process gives an outdated > thread ID to the API, and an irrelvant thread within another process > belongs to same user gets bind to cpus, such problem is very difficult > to be diagnosed when it happens, and it is an administrator is > nightmare. I think there should be a CPU_WHICH_TID_LOCAL command > to limit the searching scope in current process, and in most case, > the CPU_WHICH_TID_LOCAL will be used instead. Would lack of this cause binaries compiled against a non-cpuset schedular to panic a cpuset one? Specifically when I was attempting to verify the performence gain I commented out all *_affinity items in sched_ule.c and kern_cpuset.c and it booted fine but panicked if any forks from the shell occured (i.e. tcsh came up fine in single user mode but *ANY* external command sent it into a panic) From dds at FreeBSD.org Mon Mar 3 10:06:53 2008 From: dds at FreeBSD.org (Diomidis Spinellis) Date: Mon Mar 3 10:06:55 2008 Subject: cvs commit: src/games/fortune/datfiles fortunes Message-ID: <200803031006.m23A6qp1014404@repoman.freebsd.org> dds 2008-03-03 10:06:52 UTC FreeBSD src repository Modified files: games/fortune/datfiles fortunes Log: Fix typo. Revision Changes Path 1.258 +1 -1 src/games/fortune/datfiles/fortunes From piso at FreeBSD.org Mon Mar 3 10:12:47 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Mon Mar 3 10:12:52 2008 Subject: cvs commit: src/sys/netinet ip_fw_pfil.c Message-ID: <200803031012.m23ACldZ014651@repoman.freebsd.org> piso 2008-03-03 10:12:46 UTC FreeBSD src repository Modified files: sys/netinet ip_fw_pfil.c Log: Raise a bit ipfw kld priority. Discussed on: net-, ipfw-. Revision Changes Path 1.27 +1 -1 src/sys/netinet/ip_fw_pfil.c From brian at FreeBSD.org Mon Mar 3 11:39:12 2008 From: brian at FreeBSD.org (Brian Somers) Date: Mon Mar 3 11:39:14 2008 Subject: cvs commit: CVSROOT access Message-ID: <200803031139.m23BdBGS018592@repoman.freebsd.org> brian 2008-03-03 11:39:11 UTC FreeBSD src repository Modified files: . access Log: Restore my real domain name. Approved by: core (wilko) Revision Changes Path 1.862 +1 -1 CVSROOT/access From ache at FreeBSD.org Mon Mar 3 11:51:32 2008 From: ache at FreeBSD.org (Andrey A. Chernov) Date: Mon Mar 3 11:51:34 2008 Subject: cvs commit: src/lib/libc/locale setrunelocale.c Message-ID: <200803031151.m23BpV2C019045@repoman.freebsd.org> ache 2008-03-03 11:51:31 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/locale setrunelocale.c Log: MFC: 1.51 - better recover from unknown encoding found (not by core dump) Revision Changes Path 1.46.2.2 +27 -3 src/lib/libc/locale/setrunelocale.c From jroberson at chesapeake.net Mon Mar 3 12:12:21 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 12:12:25 2008 Subject: cvs commit: src/sys/kern sched_4bsd.c In-Reply-To: <200803022134.m22LYvVS038901@repoman.freebsd.org> References: <200803022134.m22LYvVS038901@repoman.freebsd.org> Message-ID: <20080303021350.E920@desktop> On Sun, 2 Mar 2008, Marcel Moolenaar wrote: > marcel 2008-03-02 21:34:57 UTC > > FreeBSD src repository > > Modified files: > sys/kern sched_4bsd.c > Log: > Unbreak after cpuset: initialize td_cpuset in sched_fork_thread(). Thanks very much marcel. I just realized this error and was going to fix it when I saw that you had already. Much appreciated. Jeff > > Revision Changes Path > 1.114 +2 -0 src/sys/kern/sched_4bsd.c > From jroberson at chesapeake.net Mon Mar 3 12:16:04 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 12:16:13 2008 Subject: cvs commit: src/sys/kern sched_ule.c In-Reply-To: <47CB81BB.1080009@elischer.org> References: <200803020821.m228L0Yw042389@repoman.freebsd.org> <20080301222513.Y920@desktop> <20080302105958.GF67687@server.vk2pj.dyndns.org> <47CB81BB.1080009@elischer.org> Message-ID: <20080303021658.L920@desktop> On Sun, 2 Mar 2008, Julian Elischer wrote: > Peter Jeremy wrote: > >> >>> Kris has done some excellent benchmarking as usual. Here you can see the >>> improvement in postgres depending on various scheduler debug settings: >>> >>> http://people.freebsd.org/~kris/scaling/pgsql-16cpu.png >> >> The improvement is quite substantial. Congratulations Jeff. > > the drop off after 14 is intersting.. It's contention on various kernel locks, most likely the proc lock. Scheduling algorithms which limit concurrency at contention can improve throughput in these cases but it is to the detriment of workloads which do not suffer the same contention. The ultimate solution will be to improve proc/sleepq/signal locking for 8.0. > >> > From jroberson at chesapeake.net Mon Mar 3 12:18:48 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 12:18:55 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <20080302110726.GI57756@deviant.kiev.zoral.com.ua> References: <200803020739.m227dNLe039427@repoman.freebsd.org> <20080302110726.GI57756@deviant.kiev.zoral.com.ua> Message-ID: <20080303022041.V920@desktop> On Sun, 2 Mar 2008, Kostik Belousov wrote: > On Sun, Mar 02, 2008 at 07:39:22AM +0000, Jeff Roberson wrote: >> jeff 2008-03-02 07:39:22 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/conf files >> sys/kern init_main.c kern_thread.c syscalls.master >> sys/sys _types.h types.h proc.h >> lib/libc/sys Symbol.map >> Added files: >> sys/kern kern_cpuset.c >> sys/sys cpuset.h >> Log: >> Add cpuset, an api for thread to cpu binding and cpu resource grouping >> and assignment. >> - Add a reference to a struct cpuset in each thread that is inherited from >> the thread that created it. >> - Release the reference when the thread is destroyed. >> - Add prototypes for syscalls and macros for manipulating cpusets in >> sys/cpuset.h >> - Add syscalls to create, get, and set new numbered cpusets: >> cpuset(), cpuset_{get,set}id() >> - Add syscalls for getting and setting affinity masks for cpusets or >> individual threads: cpuid_{get,set}affinity() >> - Add types for the 'level' and 'which' parameters for the cpuset. This >> will permit expansion of the api to cover cpu masks for other objects >> identifiable with an id_t integer. For example, IRQs and Jails may be >> coming soon. >> - The root set 0 contains all valid cpus. All thread initially belong to >> cpuset 1. This permits migrating all threads off of certain cpus to >> reserve them for special applications. >> >> Sponsored by: Nokia >> Discussed with: arch, rwatson, brooks, davidxu, deischen >> Reviewed by: antoine >> >> Revision Changes Path >> 1.10 +15 -0 src/lib/libc/sys/Symbol.map >> 1.1275 +1 -0 src/sys/conf/files >> 1.291 +2 -0 src/sys/kern/init_main.c >> 1.1 +907 -0 src/sys/kern/kern_cpuset.c (new) >> 1.266 +5 -1 src/sys/kern/kern_thread.c >> 1.238 +13 -0 src/sys/kern/syscalls.master >> 1.24 +3 -0 src/sys/sys/_types.h >> 1.1 +146 -0 src/sys/sys/cpuset.h (new) >> 1.504 +2 -1 src/sys/sys/proc.h >> 1.98 +4 -0 src/sys/sys/types.h > > Is the ommission of the 32-bits compat syscalls intended ? > No, but thanks for noticing. I need to discuss something with peter first. From jroberson at chesapeake.net Mon Mar 3 12:22:41 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 12:22:50 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <47CB82A6.4040903@freebsd.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CB82A6.4040903@freebsd.org> Message-ID: <20080303022403.Y920@desktop> On Mon, 3 Mar 2008, David Xu wrote: > Jeff Roberson wrote: >>> >>> One question is how I can determine the size of cpuset the kernel is >>> using ? >> >> I wrote it to tolerate user masks that were much larger than the kernel >> mask. I set the default CPU_SETSIZE in userspace to 128 and in kernel it's >> MAXCPU. So in practice an application shouldn't have to redefine >> CPU_SETSIZE. If your set is too small the kernel will return ERANGE >> however. Unfortunately, if your set is larger than the kernel's >> CPU_MAXSIZE it'll also return ERANGE. Maybe I should use different errnos >> for those cases. >> > > maybe the prototype of cpuset_getaffinity() can be changed to: > > int cpuset_getaffinity(cpulevel_t, cpuwhich_t, id_t, int *cpusetsize, > cpuset_t *); > > if *cpusetsize is zero, kernel just writes out the actual size the > kernel is using ? > > > I don't really see this as a problem. I can define userspace to use 1024 cpus by default and we would probably never need to concern ourselves with it. After all 1024 cpus should be enough forever.. ;) From jroberson at chesapeake.net Mon Mar 3 12:26:38 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 12:26:47 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <47CBBE21.4060104@freebsd.org> References: <200803020739.m227dNLe039427@repoman.freebsd.org> <47CBBE21.4060104@freebsd.org> Message-ID: <20080303022505.P920@desktop> On Mon, 3 Mar 2008, David Xu wrote: > Jeff Roberson wrote: >> jeff 2008-03-02 07:39:22 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/conf files sys/kern init_main.c >> kern_thread.c syscalls.master sys/sys _types.h types.h >> proc.h lib/libc/sys Symbol.map Added files: >> sys/kern kern_cpuset.c sys/sys cpuset.h >> Log: >> Add cpuset, an api for thread to cpu binding and cpu resource grouping >> and assignment. >> - Add a reference to a struct cpuset in each thread that is inherited >> from >> the thread that created it. >> - Release the reference when the thread is destroyed. >> - Add prototypes for syscalls and macros for manipulating cpusets in >> sys/cpuset.h >> - Add syscalls to create, get, and set new numbered cpusets: >> cpuset(), cpuset_{get,set}id() >> - Add syscalls for getting and setting affinity masks for cpusets or >> individual threads: cpuid_{get,set}affinity() >> - Add types for the 'level' and 'which' parameters for the cpuset. This >> will permit expansion of the api to cover cpu masks for other objects >> identifiable with an id_t integer. For example, IRQs and Jails may be >> coming soon. >> - The root set 0 contains all valid cpus. All thread initially belong >> to >> cpuset 1. This permits migrating all threads off of certain cpus to >> reserve them for special applications. >> Sponsored by: Nokia >> Discussed with: arch, rwatson, brooks, davidxu, deischen >> Reviewed by: antoine >> > > The cpuset_setaffinity with command CPU_WHICH_TID may hurt another > process if a weird process is out of the control. > The current code intents to lookup globally a thread has exact thread > id, because thread may be created and exited quickly, and thread ID is reused > quickly too, it is possible the weird process gives an outdated thread ID to > the API, and an irrelvant thread within another process > belongs to same user gets bind to cpus, such problem is very difficult > to be diagnosed when it happens, and it is an administrator is > nightmare. I think there should be a CPU_WHICH_TID_LOCAL command > to limit the searching scope in current process, and in most case, > the CPU_WHICH_TID_LOCAL will be used instead. Does the tid allocator quickly recycle numbers? This would be different from the pid allocator if it does. Basically every syscall that takes a numeric id would be subject to this problem. I expect application programers will more often specify binding from within the running thread using -1. However, there is a seperate problem that is endemic with our current threading support. thread_find() requires the proc lock and returns a pointer to the found thread. However, the proc lock is not sufficient to ensure that the thread is not exiting and recycled after thread_find() returns. I believe virtually every user of this interface may be subject to races. And while were on the subject. Why is it lwpid_t anyway? We don't have lwps, we have threads. Thanks, Jeff > > Regards, > David Xu > From jroberson at chesapeake.net Mon Mar 3 12:29:40 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 12:29:49 2008 Subject: cvs commit: src/sys/amd64/amd64 identcpu.c mp_machdep.c src/sys/amd64/include smp.h src/sys/i386/i386 identcpu.c mp_machdep.c src/sys/i386/include smp.h src/sys/ia64/ia64 mp_machdep.c src/sys/kern sched_ule.c subr_smp.c ... In-Reply-To: <20080303092335.E66705@fledge.watson.org> References: <200803020758.m227wgH9040483@repoman.freebsd.org> <47CBC127.8010403@FreeBSD.org> <20080303092335.E66705@fledge.watson.org> Message-ID: <20080303022904.E920@desktop> On Mon, 3 Mar 2008, Robert Watson wrote: > > On Mon, 3 Mar 2008, Maxim Sobolev wrote: > >> Cool! I am just curious if the new topology code is flexible enough to >> support cores that come and go on the fly? This could be useful in several >> scenarios, such as for example when running under multi-core aware >> hypervisor (e.g. Niagara), in the cases when pro-active power manager shuts >> down some cores to conserve power, in server applications when one of CPUs >> either has fried or has been unplugged, etc. > > We have quite a bit of kernel code that expects CPUs never come and go; I've > been hoping to interest people in having a devsummit session on CPU > hotplugging for a couple of years now. :-) I don't see physical hotplugging > as all that motivating currently, by hypervisor reallocation of CPUs *is* > immediately motivating. You'll find assumptions of fixed CPUs all over our > kernel -- schedulers, memory allocators, timer management, etc. A mature > model for starting and stopping CPUs and allowing kernel subsystems to > register with event handlers in order to start/stop their own services is > necessary. Similarly, providing a way for user applications that care about > CPU placement to hook into the event stream and respond appropriately is > important. Just to add a little bit more about cpu migration. Presently cpuset enforces a hard limit on the cpus that a thread can run on. Also, attempts to restrict the cpu set that would leave a thread without a processor to run on fail. In the future I intend to add a force mode that would return the thread to the default set. These features can be used to migrate all of the threads in the system off of certain processors. However, there is still a lot of state, as robert has mentioned, that would have to be reclaimed. I think long term what we need is a per-cpu init chain to take care of these subsystems. I wasn't aware that we ran well enough within any virtualization environment that this might be considered a useful feature. Can anyone comment on that? Thanks, Jeff > > Robert N M Watson > Computer Laboratory > University of Cambridge > From ache at FreeBSD.org Mon Mar 3 12:51:02 2008 From: ache at FreeBSD.org (Andrey A. Chernov) Date: Mon Mar 3 12:51:07 2008 Subject: cvs commit: src/lib/libc/locale setrunelocale.c Message-ID: <200803031251.m23Cp2xr032215@repoman.freebsd.org> ache 2008-03-03 12:51:02 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libc/locale setrunelocale.c Log: MFC: 1.51 - better recover from unknown encoding, not by core dump Revision Changes Path 1.45.2.3 +27 -7 src/lib/libc/locale/setrunelocale.c From jadawin at FreeBSD.org Mon Mar 3 13:12:28 2008 From: jadawin at FreeBSD.org (Philippe Audeoud) Date: Mon Mar 3 13:12:31 2008 Subject: cvs commit: src/share/misc committers-ports.dot Message-ID: <200803031312.m23DCRmh034561@repoman.freebsd.org> jadawin 2008-03-03 13:12:27 UTC FreeBSD src repository (ports committer) Modified files: share/misc committers-ports.dot Log: - Add myself to ports committers and to tabthorpe and thierry's mentee list Approved by: thierry (co-mentor) Revision Changes Path 1.67 +5 -0 src/share/misc/committers-ports.dot From jadawin at FreeBSD.org Mon Mar 3 13:14:40 2008 From: jadawin at FreeBSD.org (Philippe Audeoud) Date: Mon Mar 3 13:14:45 2008 Subject: cvs commit: src/usr.bin/calendar/calendars calendar.freebsd Message-ID: <200803031314.m23DEdEH034668@repoman.freebsd.org> jadawin 2008-03-03 13:14:39 UTC FreeBSD src repository (ports committer) Modified files: usr.bin/calendar/calendars calendar.freebsd Log: - Add my birthday. Approved by : thierry (co-mentor) Revision Changes Path 1.227 +1 -0 src/usr.bin/calendar/calendars/calendar.freebsd From raj at FreeBSD.org Mon Mar 3 13:20:54 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Mon Mar 3 13:20:58 2008 Subject: cvs commit: src/sys/powerpc/include pmap.h psl.h pte.h sf_buf.h spr.h tlb.h trap.h trap_aim.h trap_booke.h vmparam.h Message-ID: <200803031320.m23DKrBr035044@repoman.freebsd.org> raj 2008-03-03 13:20:53 UTC FreeBSD src repository Modified files: sys/powerpc/include pmap.h psl.h pte.h sf_buf.h spr.h trap.h vmparam.h Added files: sys/powerpc/include tlb.h trap_aim.h trap_booke.h Log: Rework and extend PowerPC headers definitons towards Book-E/e500 CPUs support. Approved by: cognet (mentor) Obtained from: Juniper, Semihalf MFp4: e500 Revision Changes Path 1.21 +70 -5 src/sys/powerpc/include/pmap.h 1.5 +32 -0 src/sys/powerpc/include/psl.h 1.6 +101 -1 src/sys/powerpc/include/pte.h 1.3 +30 -0 src/sys/powerpc/include/sf_buf.h 1.8 +148 -19 src/sys/powerpc/include/spr.h 1.1 +144 -0 src/sys/powerpc/include/tlb.h (new) 1.6 +6 -107 src/sys/powerpc/include/trap.h 1.1 +106 -0 src/sys/powerpc/include/trap_aim.h (new) 1.1 +55 -0 src/sys/powerpc/include/trap_booke.h (new) 1.12 +38 -7 src/sys/powerpc/include/vmparam.h From raj at FreeBSD.org Mon Mar 3 17:17:01 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Mon Mar 3 17:17:03 2008 Subject: cvs commit: src/sys/conf options.powerpc src/sys/dev/uart uart.h uart_bus_ocp.c uart_cpu_powerpc.c src/sys/kern subr_witness.c src/sys/powerpc/booke clock.c copyinout.c interrupt.c locore.S machdep.c pmap.c support.S swtch.S trap.c trap_subr.S ... Message-ID: <200803031717.m23HH18K068262@repoman.freebsd.org> raj 2008-03-03 17:17:01 UTC FreeBSD src repository Modified files: sys/conf options.powerpc sys/dev/uart uart.h uart_cpu_powerpc.c sys/kern subr_witness.c sys/powerpc/include md_var.h mmuvar.h sys/powerpc/powerpc autoconf.c Added files: sys/dev/uart uart_bus_ocp.c sys/powerpc/booke clock.c copyinout.c interrupt.c locore.S machdep.c pmap.c support.S swtch.S trap.c trap_subr.S uio_machdep.c vm_machdep.c sys/powerpc/include ocpbus.h sys/powerpc/mpc85xx nexus.c ocpbus.c ocpbus.h opic.c pci_ocp.c Log: Initial support for Freescale PowerQUICC III MPC85xx system-on-chip family. The PQ3 is a high performance integrated communications processing system based on the e500 core, which is an embedded RISC processor that implements the 32-bit Book E definition of the PowerPC architecture. For details refer to: http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MPC8555E This port was tested and successfully run on the following members of the PQ3 family: MPC8533, MPC8541, MPC8548, MPC8555. The following major integrated peripherals are supported: * On-chip peripherals bus * OpenPIC interrupt controller * UART * Ethernet (TSEC) * Host/PCI bridge * QUICC engine (SCC functionality) This commit brings the main functionality and will be followed by individual drivers that are logically separate from this base. Approved by: cognet (mentor) Obtained from: Juniper, Semihalf MFp4: e500 Revision Changes Path 1.12 +2 -0 src/sys/conf/options.powerpc 1.7 +1 -0 src/sys/dev/uart/uart.h 1.1 +88 -0 src/sys/dev/uart/uart_bus_ocp.c (new) 1.5 +26 -0 src/sys/dev/uart/uart_cpu_powerpc.c 1.240 +3 -0 src/sys/kern/subr_witness.c 1.1 +276 -0 src/sys/powerpc/booke/clock.c (new) 1.1 +304 -0 src/sys/powerpc/booke/copyinout.c (new) 1.1 +139 -0 src/sys/powerpc/booke/interrupt.c (new) 1.1 +497 -0 src/sys/powerpc/booke/locore.S (new) 1.1 +963 -0 src/sys/powerpc/booke/machdep.c (new) 1.1 +3107 -0 src/sys/powerpc/booke/pmap.c (new) 1.1 +106 -0 src/sys/powerpc/booke/support.S (new) 1.1 +145 -0 src/sys/powerpc/booke/swtch.S (new) 1.1 +679 -0 src/sys/powerpc/booke/trap.c (new) 1.1 +842 -0 src/sys/powerpc/booke/trap_subr.S (new) 1.1 +135 -0 src/sys/powerpc/booke/uio_machdep.c (new) 1.1 +517 -0 src/sys/powerpc/booke/vm_machdep.c (new) 1.31 +1 -0 src/sys/powerpc/include/md_var.h 1.2 +1 -0 src/sys/powerpc/include/mmuvar.h 1.1 +45 -0 src/sys/powerpc/include/ocpbus.h (new) 1.1 +144 -0 src/sys/powerpc/mpc85xx/nexus.c (new) 1.1 +585 -0 src/sys/powerpc/mpc85xx/ocpbus.c (new) 1.1 +86 -0 src/sys/powerpc/mpc85xx/ocpbus.h (new) 1.1 +99 -0 src/sys/powerpc/mpc85xx/opic.c (new) 1.1 +781 -0 src/sys/powerpc/mpc85xx/pci_ocp.c (new) 1.17 +4 -3 src/sys/powerpc/powerpc/autoconf.c From bms at FreeBSD.org Mon Mar 3 17:27:43 2008 From: bms at FreeBSD.org (Bruce M. Simpson) Date: Mon Mar 3 17:27:50 2008 Subject: cvs commit: src/sys/conf options.powerpc src/sys/dev/uart uart.h uart_bus_ocp.c uart_cpu_powerpc.c src/sys/kern subr_witness.c src/sys/powerpc/booke clock.c copyinout.c interrupt.c locore.S machdep.c pmap.c support.S swtch.S trap.c trap_subr.S ... In-Reply-To: <200803031717.m23HH18K068262@repoman.freebsd.org> References: <200803031717.m23HH18K068262@repoman.freebsd.org> Message-ID: <47CC350C.8040402@FreeBSD.org> Rafal Jaworowski wrote: > Log: > Initial support for Freescale PowerQUICC III MPC85xx system-on-chip family. > Wahaeyyyy! Crack out the beers!!! From raj at FreeBSD.org Mon Mar 3 18:20:17 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Mon Mar 3 18:20:23 2008 Subject: cvs commit: src/sys/dev/ic quicc.h src/sys/dev/quicc quicc_bfe.h quicc_bfe_ocp.c quicc_bus.h quicc_core.c src/sys/dev/scc scc_bfe.h scc_bfe_quicc.c scc_bus.h scc_dev_quicc.c src/sys/dev/uart uart_bus_scc.c uart_dev_quicc.c Message-ID: <200803031820.m23IKH9q071787@repoman.freebsd.org> raj 2008-03-03 18:20:17 UTC FreeBSD src repository Modified files: sys/dev/scc scc_bfe.h scc_bus.h sys/dev/uart uart_bus_scc.c Added files: sys/dev/ic quicc.h sys/dev/quicc quicc_bfe.h quicc_bfe_ocp.c quicc_bus.h quicc_core.c sys/dev/scc scc_bfe_quicc.c scc_dev_quicc.c sys/dev/uart uart_dev_quicc.c Log: Support for Freescale QUad Integrated Communications Controller. The QUICC engine is found on various Freescale parts including MPC85xx, and provides multiple generic time-division serial channel resources, which are in turn muxed/demuxed by the Serial Communications Controller (SCC). Along with core QUICC/SCC functionality a uart(4)-compliant device driver is provided which allows for serial ports over QUICC/SCC. Approved by: cognet (mentor) Obtained from: Juniper MFp4: e500 Revision Changes Path 1.1 +111 -0 src/sys/dev/ic/quicc.h (new) 1.1 +73 -0 src/sys/dev/quicc/quicc_bfe.h (new) 1.1 +94 -0 src/sys/dev/quicc/quicc_bfe_ocp.c (new) 1.1 +39 -0 src/sys/dev/quicc/quicc_bus.h (new) 1.1 +401 -0 src/sys/dev/quicc/quicc_core.c (new) 1.7 +1 -0 src/sys/dev/scc/scc_bfe.h 1.1 +95 -0 src/sys/dev/scc/scc_bfe_quicc.c (new) 1.2 +1 -0 src/sys/dev/scc/scc_bus.h 1.1 +151 -0 src/sys/dev/scc/scc_dev_quicc.c (new) 1.2 +3 -0 src/sys/dev/uart/uart_bus_scc.c 1.1 +487 -0 src/sys/dev/uart/uart_dev_quicc.c (new) From peterjeremy at optushome.com.au Mon Mar 3 18:25:09 2008 From: peterjeremy at optushome.com.au (Peter Jeremy) Date: Mon Mar 3 18:25:19 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080303022403.Y920@desktop> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CB82A6.4040903@freebsd.org> <20080303022403.Y920@desktop> Message-ID: <20080303182453.GE90593@server.vk2pj.dyndns.org> On Mon, Mar 03, 2008 at 02:24:54AM -1000, Jeff Roberson wrote: >it. After all 1024 cpus should be enough forever.. ;) I notice the smiley but... Based on Sun's claims, they should have 1024 thread boxes in 4-6 years. They have 64-thread boxes now (8 core x 8 thread) [*] and have said they'll double threads/chip every 12-18 months. [*] I thought there was a dual-chip T5xxx which would have 128 threads) but it doesn't show up. -- Peter Jeremy Please excuse any delays as the result of my ISP's inability to implement an MTA that is either RFC2821-compliant or matches their claimed behaviour. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080303/d19617c8/attachment.pgp From rwatson at FreeBSD.org Mon Mar 3 18:33:29 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 3 18:34:00 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080303182453.GE90593@server.vk2pj.dyndns.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CB82A6.4040903@freebsd.org> <20080303022403.Y920@desktop> <20080303182453.GE90593@server.vk2pj.dyndns.org> Message-ID: <20080303183225.W82170@fledge.watson.org> On Tue, 4 Mar 2008, Peter Jeremy wrote: > On Mon, Mar 03, 2008 at 02:24:54AM -1000, Jeff Roberson wrote: >> it. After all 1024 cpus should be enough forever.. ;) > > I notice the smiley but... Based on Sun's claims, they should have 1024 > thread boxes in 4-6 years. They have 64-thread boxes now (8 core x 8 > thread) [*] and have said they'll double threads/chip every 12-18 months. > > [*] I thought there was a dual-chip T5xxx which would have 128 threads) > but it doesn't show up. It's also instructive to look at scalability issues in file descriptor related interfaces in userspace, which have a similar bitmask structure for things like select(), etc with fixed-size arrays. Robert N M Watson Computer Laboratory University of Cambridge From raj at FreeBSD.org Mon Mar 3 18:44:33 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Mon Mar 3 18:44:35 2008 Subject: cvs commit: src/sys/dev/mii ciphy.c miidevs Message-ID: <200803031844.m23IiXcg072569@repoman.freebsd.org> raj 2008-03-03 18:44:33 UTC FreeBSD src repository Modified files: sys/dev/mii ciphy.c miidevs Log: Recognize the quad-port Cicada (Vitesse) CS8204 10/100/1000TX PHY. This PHY is found on many embedded development boards: among others MPC8555CDS evaluation systems use it. Approved by: cognet (mentor) MFp4: e500 Revision Changes Path 1.10 +3 -1 src/sys/dev/mii/ciphy.c 1.49 +1 -0 src/sys/dev/mii/miidevs From raj at FreeBSD.org Mon Mar 3 19:03:45 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Mon Mar 3 19:03:48 2008 Subject: cvs commit: src/sys/dev/tsec if_tsec.c if_tsec.h if_tsecreg.h Message-ID: <200803031903.m23J3jSe080506@repoman.freebsd.org> raj 2008-03-03 19:03:44 UTC FreeBSD src repository Added files: sys/dev/tsec if_tsec.c if_tsec.h if_tsecreg.h Log: Support for Freescale integrated Three-Speed Ethernet Controller (TSEC). TSEC is the MAC engine offering 10, 100 or 1000 Mbps speed and is found on different Freescale parts (MPC83xx, MPC85xx). Depending on the silicon version there are up to four TSEC units integrated on the chip. This driver also works with the enhanced version of the controller (eTSEC), which is backwards compatible, but doesn't take advantage of its additional features (various off-loading mechanisms) at the moment. Approved by: cognet (mentor) Obtained from: Semihalf MFp4: e500 Revision Changes Path 1.1 +1648 -0 src/sys/dev/tsec/if_tsec.c (new) 1.1 +284 -0 src/sys/dev/tsec/if_tsec.h (new) 1.1 +355 -0 src/sys/dev/tsec/if_tsecreg.h (new) From xcllnt at mac.com Mon Mar 3 19:35:25 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Mon Mar 3 19:35:32 2008 Subject: cvs commit: src/sys/conf options.powerpc src/sys/dev/uart uart.h uart_bus_ocp.c uart_cpu_powerpc.c src/sys/kern subr_witness.c src/sys/powerpc/booke clock.c copyinout.c interrupt.c locore.S machdep.c pmap.c support.S swtch.S trap.c trap_subr.S ... In-Reply-To: <200803031717.m23HH18K068262@repoman.freebsd.org> References: <200803031717.m23HH18K068262@repoman.freebsd.org> Message-ID: <10411E4A-BD4B-4244-A1C2-A42CCAC2D437@mac.com> On Mar 3, 2008, at 9:17 AM, Rafal Jaworowski wrote: > raj 2008-03-03 17:17:01 UTC > > FreeBSD src repository > > Modified files: > sys/conf options.powerpc > sys/dev/uart uart.h uart_cpu_powerpc.c > sys/kern subr_witness.c > sys/powerpc/include md_var.h mmuvar.h > sys/powerpc/powerpc autoconf.c > Added files: > sys/dev/uart uart_bus_ocp.c > sys/powerpc/booke clock.c copyinout.c interrupt.c locore.S > machdep.c pmap.c support.S swtch.S trap.c > trap_subr.S uio_machdep.c vm_machdep.c > sys/powerpc/include ocpbus.h > sys/powerpc/mpc85xx nexus.c ocpbus.c ocpbus.h opic.c > pci_ocp.c > Log: > Initial support for Freescale PowerQUICC III MPC85xx system-on-chip > family. Yay! -- Marcel Moolenaar xcllnt@mac.com From mav at FreeBSD.org Mon Mar 3 19:36:04 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Mon Mar 3 19:36:06 2008 Subject: cvs commit: src/sys/netgraph ng_pppoe.c Message-ID: <200803031936.m23Ja4s9084317@repoman.freebsd.org> mav 2008-03-03 19:36:03 UTC FreeBSD src repository Modified files: sys/netgraph ng_pppoe.c Log: Use more compact LIST instead of TAILQ for session hash. Add all listening hooks into LIST to simplify searches. Use ng_findhook() instead of own equal implementation. Revision Changes Path 1.94 +32 -62 src/sys/netgraph/ng_pppoe.c From ceri at submonkey.net Mon Mar 3 19:57:35 2008 From: ceri at submonkey.net (Ceri Davies) Date: Mon Mar 3 19:57:42 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080303182453.GE90593@server.vk2pj.dyndns.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CB82A6.4040903@freebsd.org> <20080303022403.Y920@desktop> <20080303182453.GE90593@server.vk2pj.dyndns.org> Message-ID: <20080303195724.GC31250@submonkey.net> On Tue, Mar 04, 2008 at 05:24:53AM +1100, Peter Jeremy wrote: > On Mon, Mar 03, 2008 at 02:24:54AM -1000, Jeff Roberson wrote: > >it. After all 1024 cpus should be enough forever.. ;) > > I notice the smiley but... Based on Sun's claims, they should have 1024 > thread boxes in 4-6 years. They have 64-thread boxes now (8 core x 8 > thread) [*] and have said they'll double threads/chip every 12-18 months. > > [*] I thought there was a dual-chip T5xxx which would have 128 threads) > but it doesn't show up. It isn't released yet, but you're referring to the "Victoria Falls" version of the T2 processor which is due in the second fiscal half of 2008 (T5xxx refers to servers featuring the T2 CPU). Initial reports were that systems built on Victoria Falls would support 256 threads, but that no longer appears to be the case. Ceri -- That must be wonderful! I don't understand it at all. -- Moliere -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080303/288b3a25/attachment.pgp From raj at FreeBSD.org Mon Mar 3 20:40:20 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Mon Mar 3 20:40:23 2008 Subject: cvs commit: src/sys/conf files files.powerpc src/sys/powerpc/conf MPC85XX NOTES src/share/mk bsd.cpu.mk Message-ID: <200803032040.m23KeKDt088977@repoman.freebsd.org> raj 2008-03-03 20:40:20 UTC FreeBSD src repository Modified files: sys/conf files files.powerpc sys/powerpc/conf NOTES share/mk bsd.cpu.mk Added files: sys/powerpc/conf MPC85XX Log: Connect MPC85XX to the PowerPC build. The kernel config file is KERNCONF=MPC85XX, so the usual procedure applies: 1. make buildworld TARGET_ARCH=powerpc 2. make buildkernel TARGET_ARCH=powerpc TARGET_CPUTYPE=e500 KERNCONF=MPC85XX This default config uses kernel-level FPU emulation. For the soft-float world approach: 1. make buildworld TARGET_ARCH=powerpc TARGET_CPUTYPE=e500 2. disable FPU_EMU option in sys/powerpc/conf/MPC85XX 3. make buildkernel TARGET_ARCH=powerpc TARGET_CPUTYPE=e500 KERNCONF=MPC85XX Approved by: cognet (mentor) MFp4: e500 Revision Changes Path 1.65 +7 -0 src/share/mk/bsd.cpu.mk 1.1276 +4 -0 src/sys/conf/files 1.74 +20 -2 src/sys/conf/files.powerpc 1.1 +72 -0 src/sys/powerpc/conf/MPC85XX (new) 1.14 +2 -0 src/sys/powerpc/conf/NOTES From raj at FreeBSD.org Mon Mar 3 22:00:47 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Mon Mar 3 22:00:52 2008 Subject: cvs commit: src/sys/powerpc/powerpc gdb_machdep.c Message-ID: <200803032200.m23M0kMA002621@repoman.freebsd.org> raj 2008-03-03 22:00:46 UTC FreeBSD src repository Modified files: sys/powerpc/powerpc gdb_machdep.c Log: Import the omitted gdb_machdep.c for PowerPC kernel. Approved by: cognet (mentor) MFp4: e500 Revision Changes Path 1.2 +57 -2 src/sys/powerpc/powerpc/gdb_machdep.c From jroberson at chesapeake.net Mon Mar 3 22:07:55 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 3 22:08:04 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080303182453.GE90593@server.vk2pj.dyndns.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CB82A6.4040903@freebsd.org> <20080303022403.Y920@desktop> <20080303182453.GE90593@server.vk2pj.dyndns.org> Message-ID: <20080303120659.F920@desktop> On Tue, 4 Mar 2008, Peter Jeremy wrote: > On Mon, Mar 03, 2008 at 02:24:54AM -1000, Jeff Roberson wrote: >> it. After all 1024 cpus should be enough forever.. ;) > > I notice the smiley but... Based on Sun's claims, they should have 1024 > thread boxes in 4-6 years. They have 64-thread boxes now (8 core x 8 > thread) [*] and have said they'll double threads/chip every 12-18 months. > > [*] I thought there was a dual-chip T5xxx which would have 128 threads) > but it doesn't show up. We have much bigger problems than the CPU_SETSIZE in cpuset. Currently the kernel assumes it can fit all cpus in one register. So 64 is the max on any platform. We need to modify the kernel to support this large number of cpus. After that we'll still run poorly on most workloads with this many cpus due to internal lock contention. There is a lot of work to do to support platforms of this size. > > -- > Peter Jeremy > Please excuse any delays as the result of my ISP's inability to implement > an MTA that is either RFC2821-compliant or matches their claimed behaviour. > From julian at elischer.org Mon Mar 3 22:19:00 2008 From: julian at elischer.org (Julian Elischer) Date: Mon Mar 3 22:19:06 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080303120659.F920@desktop> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CB82A6.4040903@freebsd.org> <20080303022403.Y920@desktop> <20080303182453.GE90593@server.vk2pj.dyndns.org> <20080303120659.F920@desktop> Message-ID: <47CC7954.4030708@elischer.org> Jeff Roberson wrote: > > On Tue, 4 Mar 2008, Peter Jeremy wrote: > >> On Mon, Mar 03, 2008 at 02:24:54AM -1000, Jeff Roberson wrote: >>> it. After all 1024 cpus should be enough forever.. ;) >> >> I notice the smiley but... Based on Sun's claims, they should >> have 1024 thread boxes in 4-6 years. They have 64-thread boxes >> now (8 core x 8 thread) [*] and have said they'll double >> threads/chip every 12-18 months. >> >> [*] I thought there was a dual-chip T5xxx which would have 128 >> threads) but it doesn't show up. > > We have much bigger problems than the CPU_SETSIZE in cpuset. > Currently the kernel assumes it can fit all cpus in one register. > So 64 is the max on any platform. We need to modify the kernel to > support this large number of cpus. After that we'll still run > poorly on most workloads with this many cpus due to internal lock > contention. There is a lot of work to do to support platforms of > this size. the whole UMA concept breaks down. I don't know how we are going to cope when we need to start breaking workloads upint osmall subunits and farm them out. I think it's about 5 years out at most.. Cisco just announced their latest chip has 60 cores and intel is playing with one with 80.. These cores are a little like the CELL cores.. they not really designed to be used in a UMA (symetrical) OS.. I heard rumours that microsoft is diverting resources to sketch out an OS that uses these. ironically they are in the best situation because they can dedicate cores to do some of those "render-green-box-on-frog-if-full-moon" type calls they have. > > >> >> -- Peter Jeremy Please excuse any delays as the result of my >> ISP's inability to implement an MTA that is either >> RFC2821-compliant or matches their claimed behaviour. >> From piso at FreeBSD.org Mon Mar 3 22:32:02 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Mon Mar 3 22:32:06 2008 Subject: cvs commit: src/sys/netinet ip_fw_nat.c Message-ID: <200803032232.m23MW197005629@repoman.freebsd.org> piso 2008-03-03 22:32:01 UTC FreeBSD src repository Modified files: sys/netinet ip_fw_nat.c Log: When unloading kld, don't forget to flush the nat pointers. Revision Changes Path 1.2 +7 -0 src/sys/netinet/ip_fw_nat.c From marcel at FreeBSD.org Tue Mar 4 01:40:27 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Tue Mar 4 01:40:30 2008 Subject: cvs commit: src/sys/powerpc/mpc85xx opic.c Message-ID: <200803040140.m241eQqF027895@repoman.freebsd.org> marcel 2008-03-04 01:40:26 UTC FreeBSD src repository Modified files: sys/powerpc/mpc85xx opic.c Log: Add the pic_ipi method. While here, eliminate the unused openpic_ocpbus_softc struct. Revision Changes Path 1.2 +4 -9 src/sys/powerpc/mpc85xx/opic.c From davidxu at FreeBSD.org Tue Mar 4 01:57:21 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 4 01:57:24 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <20080303022505.P920@desktop> References: <200803020739.m227dNLe039427@repoman.freebsd.org> <47CBBE21.4060104@freebsd.org> <20080303022505.P920@desktop> Message-ID: <47CCACCB.7010900@freebsd.org> Jeff Roberson wrote: > Does the tid allocator quickly recycle numbers? Yes, I saw this behavior. > This would be different > from the pid allocator if it does. Basically every syscall that takes a > numeric id would be subject to this problem. > > I expect application programers will more often specify binding from > within the running thread using -1. > Don't know, but I feel it is more often the calling will be made for threads within same process, so searching it globally is not efficient in this case, and garbage id may hit an irrelevant thread in another process. > However, there is a seperate problem that is endemic with our current > threading support. thread_find() requires the proc lock and returns a > pointer to the found thread. However, the proc lock is not sufficient > to ensure that the thread is not exiting and recycled after > thread_find() returns. I believe virtually every user of this interface > may be subject to races. > threads are linked into or removed from process with proc lock held, why will a linked-in thread be recycled ? > And while were on the subject. Why is it lwpid_t anyway? We don't have > lwps, we have threads. > lwpid_t is used for thread id, I thought marcel introduced this type, by the way, gdb also understands lwpid_t, I have not found any problem with it, it is just an integer. From davidxu at FreeBSD.org Tue Mar 4 02:07:59 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 4 02:08:01 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080302183513.P920@desktop> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> Message-ID: <47CCAF49.20903@freebsd.org> Jeff Roberson wrote: >> One question is how I can determine the size of cpuset the kernel is >> using ? > > I wrote it to tolerate user masks that were much larger than the kernel > mask. I set the default CPU_SETSIZE in userspace to 128 and in kernel > it's MAXCPU. So in practice an application shouldn't have to redefine > CPU_SETSIZE. If your set is too small the kernel will return ERANGE > however. Unfortunately, if your set is larger than the kernel's > CPU_MAXSIZE it'll also return ERANGE. Maybe I should use different > errnos for those cases. > From my point, userland has to write some urgly code to guess what kernel code wants, it is rather frustrate. From marcel at FreeBSD.org Tue Mar 4 02:20:31 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Tue Mar 4 02:20:34 2008 Subject: cvs commit: src/sys/powerpc/conf NOTES Message-ID: <200803040220.m242KUrR033012@repoman.freebsd.org> marcel 2008-03-04 02:20:28 UTC FreeBSD src repository Modified files: sys/powerpc/conf NOTES Log: Comment-out cpu E500. We can't yet build it with AIM at the same time. Revision Changes Path 1.15 +1 -1 src/sys/powerpc/conf/NOTES From jroberson at chesapeake.net Tue Mar 4 02:39:35 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Tue Mar 4 02:39:43 2008 Subject: cvs commit: src/sys/conf files src/sys/kern init_main.c kern_cpuset.c kern_thread.c syscalls.master src/sys/sys _types.h cpuset.h proc.h types.h src/lib/libc/sys Symbol.map In-Reply-To: <47CCACCB.7010900@freebsd.org> References: <200803020739.m227dNLe039427@repoman.freebsd.org> <47CBBE21.4060104@freebsd.org> <20080303022505.P920@desktop> <47CCACCB.7010900@freebsd.org> Message-ID: <20080303163838.H920@desktop> On Tue, 4 Mar 2008, David Xu wrote: > Jeff Roberson wrote: > >> Does the tid allocator quickly recycle numbers? > > Yes, I saw this behavior. I see, it uses unr. We should introduce some flags to this to prevent quick recycling anyway. > > >> This would be different from the pid allocator if it does. Basically >> every syscall that takes a numeric id would be subject to this problem. >> >> I expect application programers will more often specify binding from within >> the running thread using -1. >> > > Don't know, but I feel it is more often the calling will be made > for threads within same process, so searching it globally is not > efficient in this case, and garbage id may hit an irrelevant thread in > another process. This is an application race. I think if we slow down the id recycling we should not bother with it. > >> However, there is a seperate problem that is endemic with our current >> threading support. thread_find() requires the proc lock and returns a >> pointer to the found thread. However, the proc lock is not sufficient to >> ensure that the thread is not exiting and recycled after thread_find() >> returns. I believe virtually every user of this interface may be subject >> to races. >> > > threads are linked into or removed from process with proc lock held, > why will a linked-in thread be recycled ? You are right. I thought it was only the PROC_SLOCK. But as long as the PROC_LOCK is held that's ok. > >> And while were on the subject. Why is it lwpid_t anyway? We don't have >> lwps, we have threads. >> > lwpid_t is used for thread id, I thought marcel introduced this type, by the > way, gdb also understands lwpid_t, I have not found any problem with > it, it is just an integer. > > > From jroberson at chesapeake.net Tue Mar 4 02:40:30 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Tue Mar 4 02:40:35 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <47CCAF49.20903@freebsd.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CCAF49.20903@freebsd.org> Message-ID: <20080303164227.S920@desktop> On Tue, 4 Mar 2008, David Xu wrote: > Jeff Roberson wrote: > >>> One question is how I can determine the size of cpuset the kernel is >>> using ? >> >> I wrote it to tolerate user masks that were much larger than the kernel >> mask. I set the default CPU_SETSIZE in userspace to 128 and in kernel it's >> MAXCPU. So in practice an application shouldn't have to redefine >> CPU_SETSIZE. If your set is too small the kernel will return ERANGE >> however. Unfortunately, if your set is larger than the kernel's >> CPU_MAXSIZE it'll also return ERANGE. Maybe I should use different errnos >> for those cases. >> > > From my point, userland has to write some urgly code to guess what > kernel code wants, it is rather frustrate. You can use sysctl kern.smp.maxcpus to get the precise size. > > > > > > > > From davidxu at FreeBSD.org Tue Mar 4 03:03:25 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 4 03:03:27 2008 Subject: cvs commit: src/lib/libthr/thread thr_attr.c thr_create.c thr_init.c thr_private.h Message-ID: <200803040303.m2433P4a043236@repoman.freebsd.org> davidxu 2008-03-04 03:03:25 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_attr.c thr_create.c thr_init.c thr_private.h Log: implement pthread_attr_getaffinity_np and pthread_attr_setaffinity_np. Revision Changes Path 1.9 +58 -1 src/lib/libthr/thread/thr_attr.c 1.37 +5 -1 src/lib/libthr/thread/thr_create.c 1.48 +3 -1 src/lib/libthr/thread/thr_init.c 1.82 +2 -0 src/lib/libthr/thread/thr_private.h From marcel at FreeBSD.org Tue Mar 4 03:05:53 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Tue Mar 4 03:05:57 2008 Subject: cvs commit: src/sys/powerpc/conf NOTES Message-ID: <200803040305.m2435rcJ043338@repoman.freebsd.org> marcel 2008-03-04 03:05:53 UTC FreeBSD src repository Modified files: sys/powerpc/conf NOTES Log: Also comment-out options MPC85XX. We don't define CCSRBAR_* without E500. Revision Changes Path 1.16 +1 -1 src/sys/powerpc/conf/NOTES From davidxu at FreeBSD.org Tue Mar 4 03:08:30 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 4 03:08:32 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <20080303164227.S920@desktop> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CCAF49.20903@freebsd.org> <20080303164227.S920@desktop> Message-ID: <47CCBD78.5040708@freebsd.org> Jeff Roberson wrote: > > On Tue, 4 Mar 2008, David Xu wrote: > >> Jeff Roberson wrote: >> >>>> One question is how I can determine the size of cpuset the kernel is >>>> using ? >>> >>> I wrote it to tolerate user masks that were much larger than the >>> kernel mask. I set the default CPU_SETSIZE in userspace to 128 and >>> in kernel it's MAXCPU. So in practice an application shouldn't have >>> to redefine CPU_SETSIZE. If your set is too small the kernel will >>> return ERANGE however. Unfortunately, if your set is larger than the >>> kernel's CPU_MAXSIZE it'll also return ERANGE. Maybe I should use >>> different errnos for those cases. >>> >> >> From my point, userland has to write some urgly code to guess what >> kernel code wants, it is rather frustrate. > > You can use sysctl kern.smp.maxcpus to get the precise size. > if kern.smp.maxcpus is a stable ABI, I may use it, can it be guaranteed? I saw following code in kern_cpuset.c, obviously, maxcpus is not respected. if (uap->cpusetsize < CPU_SETSIZE || uap->cpusetsize > CPU_MAXSIZE) return (ERANGE); From davidxu at FreeBSD.org Tue Mar 4 03:45:12 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 4 03:45:14 2008 Subject: cvs commit: src/lib/libthr/thread thr_private.h Message-ID: <200803040345.m243jBLq044533@repoman.freebsd.org> davidxu 2008-03-04 03:45:11 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_private.h Log: Include cpuset.h, unbreak compiling. Revision Changes Path 1.83 +2 -0 src/lib/libthr/thread/thr_private.h From davidxu at FreeBSD.org Tue Mar 4 04:29:00 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 4 04:29:03 2008 Subject: cvs commit: src/lib/libthr/thread thr_create.c thr_private.h thr_sig.c Message-ID: <200803040428.m244SxOJ047369@repoman.freebsd.org> davidxu 2008-03-04 04:28:59 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_create.c thr_private.h thr_sig.c Log: If a new thread is created, it inherits current thread's signal masks, however if current thread is executing cancellation handler, signal SIGCANCEL may have already been blocked, this is unexpected, unblock the signal in new thread if this happens. MFC after: 1 week Revision Changes Path 1.38 +16 -1 src/lib/libthr/thread/thr_create.c 1.84 +6 -0 src/lib/libthr/thread/thr_private.h 1.26 +2 -0 src/lib/libthr/thread/thr_sig.c From imp at FreeBSD.org Tue Mar 4 05:35:28 2008 From: imp at FreeBSD.org (Warner Losh) Date: Tue Mar 4 05:35:35 2008 Subject: cvs commit: src/usr.bin/make Makefile.dist Message-ID: <200803040535.m245ZRIL059731@repoman.freebsd.org> imp 2008-03-04 05:35:27 UTC FreeBSD src repository Modified files: usr.bin/make Makefile.dist Log: Linux requires -D__dead2= and -D__unused= to get rid of the sys/cdef.h-isms in the make source. The variant of linux I tried it on doesn't have arc4random, so -Darc4random=random too. Revision Changes Path 1.6 +1 -1 src/usr.bin/make/Makefile.dist From brueffer at FreeBSD.org Tue Mar 4 06:08:20 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Tue Mar 4 06:08:22 2008 Subject: cvs commit: src/share/man/man4 ucycom.4 Message-ID: <200803040608.m2468Jw1061898@repoman.freebsd.org> brueffer 2008-03-04 06:08:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 ucycom.4 Log: MFC: rev. 1.5 Move the HARDWARE section below the DESCRIPTION section, like it is everywhere else. Revision Changes Path 1.4.2.1 +10 -10 src/share/man/man4/ucycom.4 From imp at FreeBSD.org Tue Mar 4 06:09:00 2008 From: imp at FreeBSD.org (Warner Losh) Date: Tue Mar 4 06:09:01 2008 Subject: cvs commit: src/usr.bin/make main.c Message-ID: <200803040608.m2468xUj061938@repoman.freebsd.org> imp 2008-03-04 06:08:59 UTC FreeBSD src repository Modified files: usr.bin/make main.c Log: Note 7.0 was the first version that FreeBSD/pc98 had a MACHINE of pc98 instead of i386. Revision Changes Path 1.165 +7 -8 src/usr.bin/make/main.c From brueffer at FreeBSD.org Tue Mar 4 06:09:14 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Tue Mar 4 06:09:20 2008 Subject: cvs commit: src/share/man/man4 ucycom.4 Message-ID: <200803040609.m2469DEu061969@repoman.freebsd.org> brueffer 2008-03-04 06:09:13 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 ucycom.4 Log: MFC: rev. 1.5 Move the HARDWARE section below the DESCRIPTION section, like it is everywhere else. Revision Changes Path 1.3.2.2 +10 -10 src/share/man/man4/ucycom.4 From jroberson at chesapeake.net Tue Mar 4 07:34:50 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Tue Mar 4 07:34:56 2008 Subject: cvs commit: src/sys/kern init_sysent.c syscalls.c systrace_args.c src/sys/sys syscall.h syscall.mk sysproto.h In-Reply-To: <47CCBD78.5040708@freebsd.org> References: <200803020741.m227fAoJ039644@repoman.freebsd.org> <47CB6FB0.9040602@freebsd.org> <20080302183513.P920@desktop> <47CCAF49.20903@freebsd.org> <20080303164227.S920@desktop> <47CCBD78.5040708@freebsd.org> Message-ID: <20080303213630.D920@desktop> On Tue, 4 Mar 2008, David Xu wrote: > Jeff Roberson wrote: >> >> On Tue, 4 Mar 2008, David Xu wrote: >> >>> Jeff Roberson wrote: >>> >>>>> One question is how I can determine the size of cpuset the kernel is >>>>> using ? >>>> >>>> I wrote it to tolerate user masks that were much larger than the kernel >>>> mask. I set the default CPU_SETSIZE in userspace to 128 and in kernel >>>> it's MAXCPU. So in practice an application shouldn't have to redefine >>>> CPU_SETSIZE. If your set is too small the kernel will return ERANGE >>>> however. Unfortunately, if your set is larger than the kernel's >>>> CPU_MAXSIZE it'll also return ERANGE. Maybe I should use different >>>> errnos for those cases. >>>> >>> >>> From my point, userland has to write some urgly code to guess what >>> kernel code wants, it is rather frustrate. >> >> You can use sysctl kern.smp.maxcpus to get the precise size. >> > > if kern.smp.maxcpus is a stable ABI, I may use it, can it be guaranteed? > I saw following code in kern_cpuset.c, obviously, maxcpus is not > respected. > > if (uap->cpusetsize < CPU_SETSIZE || uap->cpusetsize > CPU_MAXSIZE) > return (ERANGE); > > CPU_SETSIZE in the kernel is defined to MAXCPU. We can consider kern.smp.maxcpus a stable abi. I can make a note next to it and put it in the cpuset man page. From ru at FreeBSD.org Tue Mar 4 07:39:33 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Tue Mar 4 07:39:37 2008 Subject: cvs commit: src/sys/kern kern_shutdown.c Message-ID: <200803040739.m247dWQs066295@repoman.freebsd.org> ru 2008-03-04 07:39:32 UTC FreeBSD src repository Modified files: sys/kern kern_shutdown.c Log: Make it possible to continue working after calling doadump() manually from debugger. (This got broken in rev. 1.122.) Revision Changes Path 1.190 +1 -0 src/sys/kern/kern_shutdown.c From yar at FreeBSD.org Tue Mar 4 10:33:48 2008 From: yar at FreeBSD.org (Yar Tikhiy) Date: Tue Mar 4 10:33:55 2008 Subject: cvs commit: src/usr.bin/make make.1 Message-ID: <200803041033.m24AXggI084826@repoman.freebsd.org> yar 2008-03-04 10:33:42 UTC FreeBSD src repository Modified files: usr.bin/make make.1 Log: Split descriptions of .Ev MAKEFILE and .Va .MAKEFLAGS for clarity. Revision Changes Path 1.105 +2 -2 src/usr.bin/make/make.1 From kib at FreeBSD.org Tue Mar 4 11:05:04 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Tue Mar 4 11:05:17 2008 Subject: cvs commit: src/sys/nfsserver nfs_serv.c Message-ID: <200803041105.m24B54HU087844@repoman.freebsd.org> kib 2008-03-04 11:05:03 UTC FreeBSD src repository Modified files: sys/nfsserver nfs_serv.c Log: Fix the Giant leak in the nfsrv_remove(). Reported by: pluknet MFC after: 1 week Revision Changes Path 1.178 +6 -2 src/sys/nfsserver/nfs_serv.c From mav at FreeBSD.org Tue Mar 4 11:10:55 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 4 11:11:05 2008 Subject: cvs commit: src/sys/netgraph ng_nat.h Message-ID: <200803041110.m24BAtEU088078@repoman.freebsd.org> mav 2008-03-04 11:10:55 UTC FreeBSD src repository Modified files: sys/netgraph ng_nat.h Log: Fix incorrect field name. Revision Changes Path 1.4 +1 -1 src/sys/netgraph/ng_nat.h From yar at FreeBSD.org Tue Mar 4 11:25:24 2008 From: yar at FreeBSD.org (Yar Tikhiy) Date: Tue Mar 4 11:25:27 2008 Subject: cvs commit: src/usr.bin/make make.1 Message-ID: <200803041125.m24BPNPl088580@repoman.freebsd.org> yar 2008-03-04 11:25:23 UTC FreeBSD src repository Modified files: usr.bin/make make.1 Log: Revise the description of how .Ev MAKEFILE and .Va .MAKEFILE relate. The most important point is that -f option(s) are never copied from .Ev MAKEFILE to .Va .MAKEFILE by make(1), which is consistent with handling the command line. (-f silently sit in .Ev MAKEFILE and go to make's children unless overwritten via .Va .MAKEFILE) Bump .Dd. Revision Changes Path 1.106 +27 -21 src/usr.bin/make/make.1 From yar at FreeBSD.org Tue Mar 4 11:28:54 2008 From: yar at FreeBSD.org (Yar Tikhiy) Date: Tue Mar 4 11:29:05 2008 Subject: cvs commit: src/usr.bin/make make.1 Message-ID: <200803041128.m24BSsUg088678@repoman.freebsd.org> yar 2008-03-04 11:28:54 UTC FreeBSD src repository Modified files: usr.bin/make make.1 Log: Move a stray paragraph on .Ev MAKEFLAGS to where it belongs. Revision Changes Path 1.107 +14 -14 src/usr.bin/make/make.1 From piso at FreeBSD.org Tue Mar 4 12:07:35 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Tue Mar 4 12:07:41 2008 Subject: cvs commit: src/sbin/ipfw ipfw.8 ipfw2.c src/sys/netinet ip_fw2.c Message-ID: <200803041207.m24C7Ysx091435@repoman.freebsd.org> piso 2008-03-04 12:07:34 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/ipfw ipfw.8 ipfw2.c sys/netinet ip_fw2.c Log: MFC: Add table/tablearg support to ipfw's nat. Fix display of nat range. Fix display of multiple nat rules. Whitespace elimination. sbin/ipfw/ipfw2.c: rev. 1.115, 1.116 and 1.117 sbin/ipfw/ipfw.8: rev. 1.210 sys/netinet/ip_fw2.c: rev. 1.181 Revision Changes Path 1.203.2.2 +1 -1 src/sbin/ipfw/ipfw.8 1.108.2.3 +13 -19 src/sbin/ipfw/ipfw2.c 1.175.2.2 +4 -2 src/sys/netinet/ip_fw2.c From rwatson at FreeBSD.org Tue Mar 4 12:10:04 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Tue Mar 4 12:10:06 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_vfsops.c Message-ID: <200803041210.m24CA3Na091546@repoman.freebsd.org> rwatson 2008-03-04 12:10:03 UTC FreeBSD src repository Modified files: sys/ufs/ffs ffs_vfsops.c Log: Move setting of MNTK_MPSAFE flag before UFS1 extended attribute auto-start so that the flag is set before we start performing I/O in the auto-start routine. MFC after: 2 weeks Suggested by: kib Revision Changes Path 1.337 +3 -3 src/sys/ufs/ffs/ffs_vfsops.c From mav at FreeBSD.org Tue Mar 4 12:45:49 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 4 12:45:52 2008 Subject: cvs commit: src/sbin/natd natd.8 Message-ID: <200803041245.m24CjnOa093015@repoman.freebsd.org> mav 2008-03-04 12:45:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/natd natd.8 Log: MFC rev. 1.63-1.68 Document -instance and -globalport options. Note that the punch_fw option does not work in securelevel 3. Revision Changes Path 1.62.10.1 +189 -1 src/sbin/natd/natd.8 From rwatson at FreeBSD.org Tue Mar 4 12:50:12 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Tue Mar 4 12:50:16 2008 Subject: cvs commit: src/sys/ufs/ufs extattr.h ufs_extattr.c Message-ID: <200803041250.m24CoBFw002312@repoman.freebsd.org> rwatson 2008-03-04 12:50:11 UTC FreeBSD src repository Modified files: sys/ufs/ufs extattr.h ufs_extattr.c Log: Continue on-going campaign to replace lockmgr locks with sx locks where the specific semantics of ockmgr aren't required: update UFS1 extended attributes to protect its data structures using an sx lock. While here, update comments on lock granularity. MFC after: 2 weeks Revision Changes Path 1.22 +3 -2 src/sys/ufs/ufs/extattr.h 1.91 +11 -10 src/sys/ufs/ufs/ufs_extattr.c From mav at FreeBSD.org Tue Mar 4 12:52:35 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 4 12:52:37 2008 Subject: cvs commit: src/sbin/natd natd.8 Message-ID: <200803041252.m24CqY2O002412@repoman.freebsd.org> mav 2008-03-04 12:52:34 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sbin/natd natd.8 Log: MFC rev. 1.63-1.68 Document -instance and -globalport options. Note that the punch_fw option does not work in securelevel 3. Revision Changes Path 1.62.2.1 +189 -1 src/sbin/natd/natd.8 From imp at FreeBSD.org Tue Mar 4 15:56:18 2008 From: imp at FreeBSD.org (Warner Losh) Date: Tue Mar 4 15:56:25 2008 Subject: cvs commit: src/usr.bin/make arch.c Message-ID: <200803041556.m24FuIqw012139@repoman.freebsd.org> imp 2008-03-04 15:56:18 UTC FreeBSD src repository Modified files: usr.bin/make arch.c Log: is necessary for using INT_MIN, so included it here explicitly rather than relying on name space pollution to pull it in for us. NB: The usage of INT_MIN is somewhat bogus and suspect to my eye, but this commit doesn't address that issue. Revision Changes Path 1.67 +1 -0 src/usr.bin/make/arch.c From keramida at ceid.upatras.gr Tue Mar 4 16:10:59 2008 From: keramida at ceid.upatras.gr (Giorgos Keramidas) Date: Tue Mar 4 16:11:05 2008 Subject: cvs commit: src/usr.bin/make arch.c In-Reply-To: <200803041556.m24FuIqw012139@repoman.freebsd.org> References: <200803041556.m24FuIqw012139@repoman.freebsd.org> Message-ID: <20080304161026.GD95950@kobe.laptop> On 2008-03-04 15:56, Warner Losh wrote: > imp 2008-03-04 15:56:18 UTC > > FreeBSD src repository > > Modified files: > usr.bin/make arch.c > Log: > is necessary for using INT_MIN, so included it here > explicitly rather than relying on name space pollution to pull it in > for us. Thanks :-) From hartmut.brandt at dlr.de Tue Mar 4 16:15:49 2008 From: hartmut.brandt at dlr.de (Harti Brandt) Date: Tue Mar 4 16:15:53 2008 Subject: cvs commit: src/usr.bin/make arch.c In-Reply-To: <200803041556.m24FuIqw012139@repoman.freebsd.org> References: <200803041556.m24FuIqw012139@repoman.freebsd.org> Message-ID: <20080304171410.V893@knop-beagle.kn.op.dlr.de> On Tue, 4 Mar 2008, Warner Losh wrote: WL>imp 2008-03-04 15:56:18 UTC WL> WL> FreeBSD src repository WL> WL> Modified files: WL> usr.bin/make arch.c WL> Log: WL> is necessary for using INT_MIN, so included it here WL> explicitly rather than relying on name space pollution to pull it in WL> for us. WL> WL> NB: The usage of INT_MIN is somewhat bogus and suspect to my eye, but this WL> commit doesn't address that issue. Yes. This looks like a bug. ArchStatMember returns INT64_MIN if it cannot stat the archive member. So the caller must check for INT64_MIN. harti From imp at bsdimp.com Tue Mar 4 16:28:34 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Tue Mar 4 16:28:39 2008 Subject: cvs commit: src/usr.bin/make arch.c In-Reply-To: <20080304161026.GD95950@kobe.laptop> References: <200803041556.m24FuIqw012139@repoman.freebsd.org> <20080304161026.GD95950@kobe.laptop> Message-ID: <20080304.092741.-116096587.imp@bsdimp.com> In message: <20080304161026.GD95950@kobe.laptop> Giorgos Keramidas writes: : On 2008-03-04 15:56, Warner Losh wrote: : > imp 2008-03-04 15:56:18 UTC : > : > FreeBSD src repository : > : > Modified files: : > usr.bin/make arch.c : > Log: : > is necessary for using INT_MIN, so included it here : > explicitly rather than relying on name space pollution to pull it in : > for us. : : Thanks :-) You are welcome. Warner From jhb at FreeBSD.org Tue Mar 4 16:54:32 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 4 16:54:33 2008 Subject: cvs commit: src/sys/conf kmod.mk Message-ID: <200803041654.m24GsVKl024525@repoman.freebsd.org> jhb 2008-03-04 16:54:31 UTC FreeBSD src repository Modified files: sys/conf kmod.mk Log: Force an explicit dependency on opt_global.h for all module object files when building modules as part of a kernel build just as we do for kernel object files. MFC after: 1 week Reported by: kmacy, kris Reviewed by: ru Revision Changes Path 1.222 +7 -0 src/sys/conf/kmod.mk From emax at FreeBSD.org Tue Mar 4 17:41:59 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 4 17:42:04 2008 Subject: cvs commit: src/usr.bin/bluetooth/rfcomm_sppd rfcomm_sppd.1 rfcomm_sppd.c Message-ID: <200803041741.m24Hfwfx027278@repoman.freebsd.org> emax 2008-03-04 17:41:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/bluetooth/rfcomm_sppd rfcomm_sppd.1 rfcomm_sppd.c Log: MFC: add "server mode" to rfcomm_sppd(1) MFC: Check return value of listen() PR: bin/102523 Submitted by: Dave Eckhardt, bms, kevlo Tested by: Dave Eckhardt, Eric Anderson, bms Revision Changes Path 1.7.2.3 +53 -8 src/usr.bin/bluetooth/rfcomm_sppd/rfcomm_sppd.1 1.6.2.3 +100 -31 src/usr.bin/bluetooth/rfcomm_sppd/rfcomm_sppd.c From mav at FreeBSD.org Tue Mar 4 18:22:18 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 4 18:22:24 2008 Subject: cvs commit: src/sys/netgraph ng_base.c Message-ID: <200803041822.m24IMI22029994@repoman.freebsd.org> mav 2008-03-04 18:22:18 UTC FreeBSD src repository Modified files: sys/netgraph ng_base.c Log: Implement 128 items node name hash for faster name search. Increase node ID hash size from 32 to 128 items. Revision Changes Path 1.148 +70 -47 src/sys/netgraph/ng_base.c From alc at FreeBSD.org Tue Mar 4 18:50:16 2008 From: alc at FreeBSD.org (Alan Cox) Date: Tue Mar 4 18:50:23 2008 Subject: cvs commit: src/sys/amd64/amd64 pmap.c src/sys/amd64/include pmap.h Message-ID: <200803041850.m24IoGvO035970@repoman.freebsd.org> alc 2008-03-04 18:50:15 UTC FreeBSD src repository Modified files: sys/amd64/amd64 pmap.c sys/amd64/include pmap.h Log: Add support for automatic promotion of 4KB page mappings to 2MB page mappings. Automatic promotion can be enabled by setting the tunable "vm.pmap.pg_ps_enabled" to a non-zero value. By default, automatic promotion is disabled. (Expect this to change.) Reviewed by: ups Tested by: kris, Peter Holm Revision Changes Path 1.606 +1072 -104 src/sys/amd64/amd64/pmap.c 1.140 +11 -3 src/sys/amd64/include/pmap.h From obrien at FreeBSD.org Tue Mar 4 19:00:12 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Tue Mar 4 19:00:17 2008 Subject: cvs commit: src/share/mk sys.mk Message-ID: <200803041900.m24J0BK0036475@repoman.freebsd.org> obrien 2008-03-04 19:00:11 UTC FreeBSD src repository Modified files: share/mk sys.mk Log: Back out revision 1.97, which backed out part of revision 1.96. Change the default CFLAGS to match the simple defaults that the tinderboxes use. By using -fno-strict-aliasing by default we are choosing to ignore problems in code which had the potential to shoot ourselves in the foot. Revision Changes Path 1.98 +2 -2 src/share/mk/sys.mk From wilko at FreeBSD.org Tue Mar 4 19:02:34 2008 From: wilko at FreeBSD.org (Wilko Bulte) Date: Tue Mar 4 19:02:39 2008 Subject: cvs commit: CVSROOT access Message-ID: <200803041902.m24J2XfR036571@repoman.freebsd.org> wilko 2008-03-04 19:02:33 UTC FreeBSD src repository Modified files: . access Log: njl has asked core to notify the world that rpaulo is now free from mentorship Submitted by: njl Approved by: core (implicit) Revision Changes Path 1.863 +0 -0 CVSROOT/access From sgk at troutmask.apl.washington.edu Tue Mar 4 19:10:50 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Tue Mar 4 19:10:58 2008 Subject: cvs commit: src/share/mk sys.mk In-Reply-To: <200803041900.m24J0BK0036475@repoman.freebsd.org> References: <200803041900.m24J0BK0036475@repoman.freebsd.org> Message-ID: <20080304191015.GA67271@troutmask.apl.washington.edu> On Tue, Mar 04, 2008 at 07:00:11PM +0000, David E. O'Brien wrote: > obrien 2008-03-04 19:00:11 UTC > > FreeBSD src repository > > Modified files: > share/mk sys.mk > Log: > Back out revision 1.97, which backed out part of revision 1.96. > Change the default CFLAGS to match the simple defaults that the > tinderboxes use. By using -fno-strict-aliasing by default we are > choosing to ignore problems in code which had the potential to > shoot ourselves in the foot. > Does this change propagate into the ports collection make infrastructure? There are several ports that require -fno-strict-aliasing. -- Steve From rpaulo at FreeBSD.org Tue Mar 4 19:16:32 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Tue Mar 4 19:16:35 2008 Subject: cvs commit: src/sys/netinet in.h in_pcb.c Message-ID: <200803041916.m24JGMci038210@repoman.freebsd.org> rpaulo 2008-03-04 19:16:22 UTC FreeBSD src repository Modified files: sys/netinet in.h in_pcb.c Log: Change the default port range for outgoing connections by introducing IPPORT_EPHEMERALFIRST and IPPORT_EPHEMERALLAST with values 10000 and 65535 respectively. The rationale behind is that it makes the attacker's life more difficult if he/she wants to guess the ephemeral port range and also lowers the probability of a port colision (described in draft-ietf-tsvwg-port-randomization-01.txt). While there, remove code duplication in in_pcbbind_setup(). Submitted by: Fernando Gont Approved by: njl (mentor) Reviewed by: silby, bms Discussed on: freebsd-net Revision Changes Path 1.101 +8 -4 src/sys/netinet/in.h 1.199 +21 -40 src/sys/netinet/in_pcb.c From bms at FreeBSD.org Tue Mar 4 19:22:58 2008 From: bms at FreeBSD.org (Bruce M. Simpson) Date: Tue Mar 4 19:23:07 2008 Subject: cvs commit: src/sys/netinet in.h in_pcb.c In-Reply-To: <200803041916.m24JGMci038210@repoman.freebsd.org> References: <200803041916.m24JGMci038210@repoman.freebsd.org> Message-ID: <47CDA18F.7090007@FreeBSD.org> Rui Paulo wrote: > Submitted by: Fernando Gont > Approved by: njl (mentor) > Reviewed by: silby, bms > Disclaimer: I didn't review the code in this change, but just commented upon the intent behind it. BMS From obrien at FreeBSD.org Tue Mar 4 20:26:03 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Tue Mar 4 20:26:09 2008 Subject: cvs commit: src/share/mk sys.mk Message-ID: <200803042026.m24KQ3FP041999@repoman.freebsd.org> obrien 2008-03-04 20:26:03 UTC FreeBSD src repository Modified files: share/mk sys.mk Log: Temporarily back out revision 1.98 to give Portmgr some time to address PR ports/121363 (current day re-opening of PR ports/73797) to make ports CFLAGS more independent of src/'s CFLAGS WRT aliasing. Discussed with: brooks Revision Changes Path 1.99 +2 -2 src/share/mk/sys.mk From rpaulo at fnop.net Tue Mar 4 20:28:06 2008 From: rpaulo at fnop.net (Rui Paulo) Date: Tue Mar 4 20:28:12 2008 Subject: cvs commit: src/sys/netinet in.h in_pcb.c In-Reply-To: <47CDA18F.7090007@FreeBSD.org> References: <200803041916.m24JGMci038210@repoman.freebsd.org> <47CDA18F.7090007@FreeBSD.org> Message-ID: <20080304173217.GC1008@fnop.net> On Tue, Mar 04, 2008 at 07:22:55PM +0000, Bruce M. Simpson wrote: > Rui Paulo wrote: >> Submitted by: Fernando Gont >> Approved by: njl (mentor) >> Reviewed by: silby, bms >> > > Disclaimer: I didn't review the code in this change, but just commented > upon the intent behind it. Oh, I'm terribly sorry. My mistake. -- Rui Paulo From julian at elischer.org Tue Mar 4 21:21:35 2008 From: julian at elischer.org (Julian Elischer) Date: Tue Mar 4 21:21:39 2008 Subject: cvs commit: src/sys/netgraph ng_base.c In-Reply-To: <200803041822.m24IMI22029994@repoman.freebsd.org> References: <200803041822.m24IMI22029994@repoman.freebsd.org> Message-ID: <47CDBD5D.5020705@elischer.org> Alexander Motin wrote: > mav 2008-03-04 18:22:18 UTC > > FreeBSD src repository > > Modified files: > sys/netgraph ng_base.c > Log: > Implement 128 items node name hash for faster name search. > Increase node ID hash size from 32 to 128 items. > > Revision Changes Path > 1.148 +70 -47 src/sys/netgraph/ng_base.c I knew this day would come.. that's why I had a name lookup method... From obrien at FreeBSD.org Tue Mar 4 22:32:58 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Tue Mar 4 22:32:59 2008 Subject: cvs commit: src/usr.bin/make Makefile job.c Message-ID: <200803042232.m24MWwAQ058988@repoman.freebsd.org> obrien 2008-03-04 22:32:58 UTC FreeBSD src repository Modified files: usr.bin/make Makefile job.c Log: No need to tell make to DTRT with "make love", just do it. Also remove the 2002/08/31 bootstrapping aid for upgrades from year old (mid-2001) systems. Revision Changes Path 1.66 +0 -5 src/usr.bin/make/Makefile 1.127 +0 -6 src/usr.bin/make/job.c From obrien at FreeBSD.org Tue Mar 4 22:51:38 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Tue Mar 4 22:51:40 2008 Subject: cvs commit: src/usr.bin/make job.c Message-ID: <200803042251.m24MpbJ2059666@repoman.freebsd.org> obrien 2008-03-04 22:51:37 UTC FreeBSD src repository Modified files: usr.bin/make job.c Log: It seems some don't care for the anchient joke. Add WITHOUT_OLD_JOKE to your CFLAGS if you fall into this camp. Revision Changes Path 1.128 +2 -0 src/usr.bin/make/job.c From yongari at FreeBSD.org Wed Mar 5 01:15:10 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Wed Mar 5 01:15:18 2008 Subject: cvs commit: src/sys/dev/mii rgephy.c Message-ID: <200803050115.m251FAXh080712@repoman.freebsd.org> yongari 2008-03-05 01:15:10 UTC FreeBSD src repository Modified files: sys/dev/mii rgephy.c Log: Add detection of isolation state. PR: kern/76710 Revision Changes Path 1.19 +5 -0 src/sys/dev/mii/rgephy.c From simokawa at FreeBSD.org Wed Mar 5 01:30:49 2008 From: simokawa at FreeBSD.org (Hidetoshi Shimokawa) Date: Wed Mar 5 01:30:51 2008 Subject: cvs commit: src/usr.sbin/fwcontrol fwcontrol.8 fwcontrol.c Message-ID: <200803050130.m251UnP7081202@repoman.freebsd.org> simokawa 2008-03-05 01:30:49 UTC FreeBSD src repository Modified files: usr.sbin/fwcontrol fwcontrol.8 fwcontrol.c Log: - add '-f' option to force root node. - fix byte order in read_write_quad() - show hostnames in the list - fix typo in manpage MFC after: 1 week Revision Changes Path 1.22 +6 -1 src/usr.sbin/fwcontrol/fwcontrol.8 1.24 +17 -7 src/usr.sbin/fwcontrol/fwcontrol.c From jeff at FreeBSD.org Wed Mar 5 01:49:21 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 5 01:49:24 2008 Subject: cvs commit: src/usr.bin/cpuset cpuset.c src/sys/sys cpuset.h src/sys/kern kern_cpuset.c Message-ID: <200803050149.m251nL4S081921@repoman.freebsd.org> jeff 2008-03-05 01:49:20 UTC FreeBSD src repository Modified files: usr.bin/cpuset cpuset.c sys/sys cpuset.h sys/kern kern_cpuset.c Log: - Verify that when a user supplies a mask that is bigger than the kernel mask none of the upper bits are set. - Be more careful about enforcing the boundaries of masks and child sets. - Introduce a few more CPU_* macros for implementing these tests. - Change the cpusetsize argument to be bytes rather than bits to match other apis. Sponsored by: Nokia Revision Changes Path 1.2 +78 -26 src/sys/kern/kern_cpuset.c 1.2 +22 -0 src/sys/sys/cpuset.h 1.2 +7 -8 src/usr.bin/cpuset/cpuset.c From jeff at FreeBSD.org Wed Mar 5 02:10:44 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 5 02:10:46 2008 Subject: cvs commit: src/usr.bin/cpuset Makefile cpuset.1 cpuset.c Message-ID: <200803050210.m252AiQN083807@repoman.freebsd.org> jeff 2008-03-05 02:10:44 UTC FreeBSD src repository Modified files: usr.bin/cpuset Makefile cpuset.c Added files: usr.bin/cpuset cpuset.1 Log: - Remove the -i argument when running a command to simplify things a little bit and to prevent users from specifying a private mask that may later restrict other group changes. - Add a man page which brueffer generously contributed to. Sponsored by: Nokia Revision Changes Path 1.2 +0 -1 src/usr.bin/cpuset/Makefile 1.1 +171 -0 src/usr.bin/cpuset/cpuset.1 (new) 1.3 +6 -13 src/usr.bin/cpuset/cpuset.c From yongari at FreeBSD.org Wed Mar 5 05:36:10 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Wed Mar 5 05:36:11 2008 Subject: cvs commit: src/sys/dev/lge if_lge.c Message-ID: <200803050536.m255aALw004725@repoman.freebsd.org> yongari 2008-03-05 05:36:09 UTC FreeBSD src repository Modified files: sys/dev/lge if_lge.c Log: Plug memory leak in jumbo buffer allocation failure path. Patch in the PR was modified to check active jumbo buffers in use and other possible jumbo buffer leak. Jumbo buffer usage in lge(4) still wouldn't be reliable due to lack of driver lock in local jumbo buffer allocator. Either introduce a new lock to protect jumbo buffer or switch to UMA backed page allocator for jumbo frame is required. PR: kern/78072 Revision Changes Path 1.53 +12 -4 src/sys/dev/lge/if_lge.c From rodrigc at FreeBSD.org Wed Mar 5 06:24:43 2008 From: rodrigc at FreeBSD.org (Craig Rodrigues) Date: Wed Mar 5 06:24:46 2008 Subject: cvs commit: src/sbin/fsck_ffs main.c Message-ID: <200803050624.m256Og42007410@repoman.freebsd.org> rodrigc 2008-03-05 06:24:42 UTC FreeBSD src repository Modified files: sbin/fsck_ffs main.c Log: Remove hacks to filter out MNT_ROOTFS, since we now do that internally inside nmount() in revision 1.267 of vfs_mount.c. Revision Changes Path 1.48 +0 -10 src/sbin/fsck_ffs/main.c From davidxu at FreeBSD.org Wed Mar 5 06:55:49 2008 From: davidxu at FreeBSD.org (David Xu) Date: Wed Mar 5 06:55:52 2008 Subject: cvs commit: src/lib/libthr/thread thr_attr.c Message-ID: <200803050655.m256tn56009594@repoman.freebsd.org> davidxu 2008-03-05 06:55:49 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_attr.c Log: Check actual size of cpuset kernel is using and define underscore version of API. Revision Changes Path 1.10 +42 -7 src/lib/libthr/thread/thr_attr.c From davidxu at FreeBSD.org Wed Mar 5 06:56:36 2008 From: davidxu at FreeBSD.org (David Xu) Date: Wed Mar 5 06:56:39 2008 Subject: cvs commit: src/lib/libthr pthread.map Message-ID: <200803050656.m256uZX0009659@repoman.freebsd.org> davidxu 2008-03-05 06:56:35 UTC FreeBSD src repository Modified files: lib/libthr pthread.map Log: Add more cpu affinity function's symbols. Revision Changes Path 1.28 +6 -0 src/lib/libthr/pthread.map From davidxu at FreeBSD.org Wed Mar 5 07:01:22 2008 From: davidxu at FreeBSD.org (David Xu) Date: Wed Mar 5 07:01:27 2008 Subject: cvs commit: src/lib/libthr/thread thr_create.c thr_init.c thr_private.h Message-ID: <200803050701.m2571LNq009838@repoman.freebsd.org> davidxu 2008-03-05 07:01:21 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_create.c thr_init.c thr_private.h Log: Use cpuset defined in pthread_attr for newly created thread, for now, we set scheduling parameters and cpu binding fully in userland, and because default scheduling policy is SCHED_RR (time-sharing), we set default sched_inherit to PTHREAD_SCHED_INHERIT, this saves a system call. Revision Changes Path 1.39 +51 -20 src/lib/libthr/thread/thr_create.c 1.49 +1 -1 src/lib/libthr/thread/thr_init.c 1.85 +5 -0 src/lib/libthr/thread/thr_private.h From davidxu at FreeBSD.org Wed Mar 5 07:04:56 2008 From: davidxu at FreeBSD.org (David Xu) Date: Wed Mar 5 07:04:59 2008 Subject: cvs commit: src/lib/libthr/thread thr_sig.c Message-ID: <200803050704.m2574u56011099@repoman.freebsd.org> davidxu 2008-03-05 07:04:55 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_sig.c Log: Increase and decrease in_sigcancel_handler accordingly to avoid possible error caused by nested SIGCANCEL stack, it is a bit complex. Revision Changes Path 1.27 +2 -2 src/lib/libthr/thread/thr_sig.c From rodrigc at FreeBSD.org Wed Mar 5 07:55:09 2008 From: rodrigc at FreeBSD.org (Craig Rodrigues) Date: Wed Mar 5 07:55:11 2008 Subject: cvs commit: src/usr.sbin/mountd mountd.c Message-ID: <200803050755.m257t7jr014457@repoman.freebsd.org> rodrigc 2008-03-05 07:55:07 UTC FreeBSD src repository Modified files: usr.sbin/mountd mountd.c Log: Remove hacks which filter out MNT_ROOTFS. They are no longer needed now that we filter out MNT_ROOTFS inside the nmount() call in revision 1.267 of vfs_mount.c. Reviewed by: rink Revision Changes Path 1.97 +1 -15 src/usr.sbin/mountd/mountd.c From kevlo at FreeBSD.org Wed Mar 5 07:55:45 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Wed Mar 5 07:55:59 2008 Subject: cvs commit: src/sys/arm/conf AVILA Message-ID: <200803050755.m257tjww014501@repoman.freebsd.org> kevlo 2008-03-05 07:55:45 UTC FreeBSD src repository Modified files: sys/arm/conf AVILA Log: Add rl(4) support Revision Changes Path 1.8 +2 -0 src/sys/arm/conf/AVILA From jeff at FreeBSD.org Wed Mar 5 08:08:33 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 5 08:08:36 2008 Subject: cvs commit: src/sys/kern kern_cpuset.c Message-ID: <200803050808.m2588Wiw016070@repoman.freebsd.org> jeff 2008-03-05 08:08:32 UTC FreeBSD src repository Modified files: sys/kern kern_cpuset.c Log: - Don't overwrite the recently allocated 'nset' in cpuset_setthread() by passing it to cpuset_which(). Pass in 'set' instead. This argument is not used but for convenience cpuset_which() nulls all incoming parameters. Submitted by: davidxu Revision Changes Path 1.3 +1 -1 src/sys/kern/kern_cpuset.c From rodrigc at FreeBSD.org Wed Mar 5 08:25:51 2008 From: rodrigc at FreeBSD.org (Craig Rodrigues) Date: Wed Mar 5 08:25:53 2008 Subject: cvs commit: src/sbin/fsck_ffs main.c Message-ID: <200803050825.m258Ppv2016738@repoman.freebsd.org> rodrigc 2008-03-05 08:25:49 UTC FreeBSD src repository Modified files: sbin/fsck_ffs main.c Log: For a mounted file system which is read-only, when doing the MNT_RELOAD, pass in "ro" and "update" string mount options to nmount() instead of MNT_RDONLY and MNT_UPDATE flags. Due to the complexity of the mount parsing code especially with respect to the root file system, passing in MNT_RDONLY and MNT_UPDATE flags would do weird things and would cause fsck to convert the root file system from a read-only mount to read-write. To test: - boot into single user mode - show mounted file systems with: mount - root file system should be mounted read-only - fsck / - show mounted file systems with: mount - root file system should still be mounted read-only PR: 120319 MFC after: 1 month Reported by: yar Revision Changes Path 1.49 +3 -1 src/sbin/fsck_ffs/main.c From philip at FreeBSD.org Wed Mar 5 08:53:03 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Wed Mar 5 08:53:06 2008 Subject: cvs commit: CVSROOT access Message-ID: <200803050853.m258r2nQ027261@repoman.freebsd.org> philip 2008-03-05 08:53:02 UTC FreeBSD src repository Modified files: . access Log: Note on behalve of njl, who has turned in his commit bit for safe-keeping recently, that Rui Paolo (rpaulo) may now spread his wings free of mentorship. May his collection of pointy hats flourish! :-) Submitted by: njl (mentor - been a while since I typed that ;-)) Approved by: core (implicit) Revision Changes Path 1.864 +0 -0 CVSROOT/access From rodrigc at FreeBSD.org Wed Mar 5 09:41:24 2008 From: rodrigc at FreeBSD.org (Craig Rodrigues) Date: Wed Mar 5 09:41:26 2008 Subject: cvs commit: src/sys/nfsclient nfs_vfsops.c Message-ID: <200803050941.m259fNXF029941@repoman.freebsd.org> rodrigc 2008-03-05 09:41:22 UTC FreeBSD src repository Modified files: sys/nfsclient nfs_vfsops.c Log: In nfs_mount(), default initialize struct nfs_args the same way that it is default initialized in revision 1.77 of mount_nfs.c. Right now, this is a no-op, because currently we initialize struct nfs_args in mount_nfs in userspace, and pass it down into the kernel via nmount(), so we overwrite whatever we initialize here with the value passed in from userspace. However, this lays the groundwork for moving away from passing struct nfs_args from userspace to kernel via nmount(), so that we can instead pass string mount options via nmount() which can be parsed in the kernel. This will make it easier to add new NFS mount options. Revision Changes Path 1.201 +25 -1 src/sys/nfsclient/nfs_vfsops.c From rodrigc at FreeBSD.org Wed Mar 5 10:09:30 2008 From: rodrigc at FreeBSD.org (Craig Rodrigues) Date: Wed Mar 5 10:09:37 2008 Subject: cvs commit: src/sys/nfsclient nfs_vfsops.c Message-ID: <200803051009.m25A9U4f031955@repoman.freebsd.org> rodrigc 2008-03-05 10:09:29 UTC FreeBSD src repository Modified files: sys/nfsclient nfs_vfsops.c Log: Expand the nfs_opts array to include all possible string mount options that mount_nfs could pass down, if it passed down string mount options. Right now, mount_nfs jut passes down a single mount option named "nfs_args" with a fully initialized 'struct nfs_args'. In future commits, we will add code to the kernel for parsing stringified NFS mount options, so that we can convert mount_nfs to pass string options from userspace to kernel, instead of an initialized struct nfs_args. Revision Changes Path 1.202 +4 -1 src/sys/nfsclient/nfs_vfsops.c From bde at FreeBSD.org Wed Mar 5 11:11:54 2008 From: bde at FreeBSD.org (Bruce Evans) Date: Wed Mar 5 11:11:56 2008 Subject: cvs commit: src/sys/i386/include float.h Message-ID: <200803051111.m25BBreL035027@repoman.freebsd.org> bde 2008-03-05 11:11:53 UTC FreeBSD src repository Modified files: sys/i386/include float.h Log: Change float_t and double_t to long double on i386. All floating point expressions on i386 are evaluated in the range of the long double type, so this is wrong in a different but hopefully less worse way than before. Since expressions are evaluated in long double registers, there is no runtime cost to using long double instead of double to declare intermediate values (except in cases where this avoids compiler bugs), and by careful use of float_t or double_t it is possible to avoid some of the compiler bugs in this area, provided these types are declared as long double. I was going to change float.h to be less broken and more usable in combination with the change here (in particular, it is more necessary to know the effective number of bits in a double_t when double_t != double, since DBL_MANT_DIG no longer logically gives this, and LDBL_MANT_DIG doesn't give it either with FreeBSD-i386's default rounding precision. However, this was too hard for now. In particular, LDBL_MANT_DIG is used a lot in libm, so it cannot be changed. One thing that is completely broken now is LDBL_MAX. This may have sort of worked when it was changed from DBL_MAX in 2002 (adding 0 to it at runtime gave +Inf, but you could at least compare with it), but starting with gcc-3.3.1 in 2003, it is always +Inf due to evaluating it at compile time in the default rounding precision. Revision Changes Path 1.16 +5 -4 src/sys/i386/include/float.h From bde at FreeBSD.org Wed Mar 5 11:17:21 2008 From: bde at FreeBSD.org (Bruce Evans) Date: Wed Mar 5 11:17:30 2008 Subject: cvs commit: src/sys/i386/include float.h Message-ID: <200803051117.m25BHLMB035230@repoman.freebsd.org> bde 2008-03-05 11:17:21 UTC FreeBSD src repository Modified files: sys/i386/include float.h Log: Oops, back out previous commit since it was to the wrong file. Revision Changes Path 1.17 +4 -5 src/sys/i386/include/float.h From bde at FreeBSD.org Wed Mar 5 11:21:14 2008 From: bde at FreeBSD.org (Bruce Evans) Date: Wed Mar 5 11:21:22 2008 Subject: cvs commit: src/sys/i386/include _types.h Message-ID: <200803051121.m25BLE03035426@repoman.freebsd.org> bde 2008-03-05 11:21:14 UTC FreeBSD src repository Modified files: sys/i386/include _types.h Log: Change float_t and double_t to long double on i386. All floating point expressions on i386 are evaluated in the range of the long double type, so this is wrong in a different but hopefully less worse way than before. Since expressions are evaluated in long double registers, there is no runtime cost to using long double instead of double to declare intermediate values (except in cases where this avoids compiler bugs), and by careful use of float_t or double_t it is possible to avoid some of the compiler bugs in this area, provided these types are declared as long double. I was going to change float.h to be less broken and more usable in combination with the change here (in particular, it is more necessary to know the effective number of bits in a double_t when double_t != double, since DBL_MANT_DIG no longer logically gives this, and LDBL_MANT_DIG doesn't give it either with FreeBSD-i386's default rounding precision. However, this was too hard for now. In particular, LDBL_MANT_DIG is used a lot in libm, so it cannot be changed. One thing that is completely broken now is LDBL_MAX. This may have sort of worked when it was changed from DBL_MAX in 2002 (adding 0 to it at runtime gave +Inf, but you could at least compare with it), but starting with gcc-3.3.1 in 2003, it is always +Inf due to evaluating it at compile time in the default rounding precision. Revision Changes Path 1.15 +2 -2 src/sys/i386/include/_types.h From cperciva at freebsd.org Wed Mar 5 11:27:33 2008 From: cperciva at freebsd.org (Colin Percival) Date: Wed Mar 5 11:27:37 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <200803051121.m25BLE03035426@repoman.freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> Message-ID: <47CE8396.6020803@freebsd.org> Bruce Evans wrote: > Modified files: > sys/i386/include _types.h > Log: > Change float_t and double_t to long double on i386. Doesn't this have a rather severe performance impact on any code which uses double_t? Colin Percival From pjd at FreeBSD.org Wed Mar 5 11:51:14 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Wed Mar 5 11:51:15 2008 Subject: cvs commit: src/sbin/geom/core geom.8 Message-ID: <200803051151.m25BpD6R036364@repoman.freebsd.org> pjd 2008-03-05 11:51:13 UTC FreeBSD src repository Modified files: sbin/geom/core geom.8 Log: Add info about few missing GEOM classes that use geom(8). Revision Changes Path 1.22 +13 -1 src/sbin/geom/core/geom.8 From yar at comp.chem.msu.su Wed Mar 5 12:20:34 2008 From: yar at comp.chem.msu.su (Yar Tikhiy) Date: Wed Mar 5 12:20:43 2008 Subject: cvs commit: src/sbin/fsck_ffs main.c In-Reply-To: <200803050825.m258Ppv2016738@repoman.freebsd.org> References: <200803050825.m258Ppv2016738@repoman.freebsd.org> Message-ID: <20080305122029.GA7027@dg.local> On Wed, Mar 05, 2008 at 08:25:49AM +0000, Craig Rodrigues wrote: > rodrigc 2008-03-05 08:25:49 UTC > > FreeBSD src repository > > Modified files: > sbin/fsck_ffs main.c > Log: > For a mounted file system which is read-only, when > doing the MNT_RELOAD, pass in "ro" and "update" > string mount options to nmount() instead of MNT_RDONLY and MNT_UPDATE flags. > > Due to the complexity of the mount parsing code especially > with respect to the root file system, passing in MNT_RDONLY and MNT_UPDATE > flags would do weird things and would cause fsck to convert the root > file system from a read-only mount to read-write. Your analysis of the problem sounds not quite correct to me. In fact, the weird things happen because the "ro" string option isn't set on the root mount from the outset. After your patch, fsck just adds the missing "ro" to the root mount options `when it becomes safe to do so' -- the "ro" string option is dangerous to set on the root FS from the outset due to a bunch of bugs in the kernel code, as my experience with this issue showed to the public. The bugs are all the same: one piece of code passes bit flags and old-style mount() arguments while its counterpart wants string options or vice versa. Our current position in the middle of the way appears totally unacceptable as it leads to problems like this one. Either we should arrive at using string options exlusively, or abandon them at all. Also note that, were the string options implemented in full, there would be no need for fsck to pass "ro" along with "update" to preserve read-only status: existing string options go away from the mount point only if explicitly turned off through a "no" prefix, as in "noro". This is an evident design point of string options, and it's implemented already (modulo bugs in vfs_mergeopts().) Finally, it can be argued that "update" actually belongs to nmount() flags, not to string options, because essentially it is not a persistent mount option, it's a modifier for the current operation requested by this nmount() call. But this is a topic for a separate discussion on the desired nmount(2) semantics. I won't restart our back-out war, but please clearly mark your change as a temporary workaround so that it doesn't mislead people, including us when we revisit the code. Adding "ro" to the root mount options lately is an obvious hack, and it isn't a job for fsck at all. It'll be safe to set the "ro" string option on the root mount from the beginning as soon as all the root-capable filesystems and their mount_foo tools are converted to using string options properly. A crucial part of the task is your work on nfs_mount(), which I fully support. Thanks! > To test: > - boot into single user mode > - show mounted file systems with: mount > - root file system should be mounted read-only > - fsck / > - show mounted file systems with: mount > - root file system should still be mounted read-only > > PR: 120319 > MFC after: 1 month > Reported by: yar > > Revision Changes Path > 1.49 +3 -1 src/sbin/fsck_ffs/main.c -- Yar From rink at FreeBSD.org Wed Mar 5 14:13:31 2008 From: rink at FreeBSD.org (Rink Springer) Date: Wed Mar 5 14:13:37 2008 Subject: cvs commit: src/share/man/man4 Makefile uslcom.4 src/sys/amd64/conf GENERIC src/sys/conf NOTES files src/sys/dev/usb usbdevs uslcom.c src/sys/i386/conf GENERIC src/sys/modules Makefile src/sys/modules/uslcom Makefile Message-ID: <200803051413.m25EDU9o054744@repoman.freebsd.org> rink 2008-03-05 14:13:30 UTC FreeBSD src repository Modified files: share/man/man4 Makefile sys/amd64/conf GENERIC sys/conf NOTES files sys/dev/usb usbdevs sys/i386/conf GENERIC sys/modules Makefile Added files: share/man/man4 uslcom.4 sys/dev/usb uslcom.c sys/modules/uslcom Makefile Log: Import uslcom(4) from OpenBSD - this is a driver for Silicon Laboratories CP2101/CP2102 based USB serial adapters. Reviewed by: imp, emaste Obtained from: OpenBSD MFC after: 2 weeks Revision Changes Path 1.411 +1 -0 src/share/man/man4/Makefile 1.1 +422 -0 src/share/man/man4/uslcom.4 (new) 1.494 +1 -0 src/sys/amd64/conf/GENERIC 1.1473 +2 -0 src/sys/conf/NOTES 1.1277 +1 -0 src/sys/conf/files 1.341 +26 -0 src/sys/dev/usb/usbdevs 1.1 +2095 -0 src/sys/dev/usb/uslcom.c (new) 1.482 +1 -0 src/sys/i386/conf/GENERIC 1.557 +1 -0 src/sys/modules/Makefile 1.1 +10 -0 src/sys/modules/uslcom/Makefile (new) From rink at FreeBSD.org Wed Mar 5 14:18:30 2008 From: rink at FreeBSD.org (Rink Springer) Date: Wed Mar 5 14:18:35 2008 Subject: cvs commit: src/share/man/man4 uslcom.4 src/sys/dev/usb uslcom.c Message-ID: <200803051418.m25EITo2055479@repoman.freebsd.org> rink 2008-03-05 14:18:29 UTC FreeBSD src repository Modified files: share/man/man4 uslcom.4 sys/dev/usb uslcom.c Log: Oops, I accidently concatenated uslcom.[c4] multiple times before the commit :-/ Quickly fix before things get broken... Pointyhat to: me Revision Changes Path 1.2 +0 -336 src/share/man/man4/uslcom.4 1.2 +0 -1676 src/sys/dev/usb/uslcom.c From kib at FreeBSD.org Wed Mar 5 16:34:05 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Wed Mar 5 16:34:08 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_vfsops.c Message-ID: <200803051634.m25GY3Hp065730@repoman.freebsd.org> kib 2008-03-05 16:34:03 UTC FreeBSD src repository Modified files: sys/ufs/ffs ffs_vfsops.c Log: Initialize mnt_stat.f_iosize before autostarting UFS1 extattrs. It is normally initialized by ffs_statfs() after ffs_mount finished. The extattr autostart code calls the ufs_lookup(), that uses value above to iterate over the directory blocks, see bmask initialization in the ufs_lookup() and ufsdirhash. Having the filesystem with root directory spanning more then one block would result in reading a random kernel memory. PR: kern/120781 Test case provided by: rwatson MFC after: 1 week Revision Changes Path 1.338 +1 -0 src/sys/ufs/ffs/ffs_vfsops.c From marcel at FreeBSD.org Wed Mar 5 16:46:39 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Wed Mar 5 16:46:44 2008 Subject: cvs commit: src/sys/powerpc/mpc85xx pci_ocp.c Message-ID: <200803051646.m25GkceT066145@repoman.freebsd.org> marcel 2008-03-05 16:46:38 UTC FreeBSD src repository Modified files: sys/powerpc/mpc85xx pci_ocp.c Log: o Various fixes related to PCI Express: - Even for the PCI Express host controller we need to use bus 0 for configuration space accesses to devices directly on the host controller's bus. - Pass the maximum number of slots to pci_ocp_init() because the caller knows how many slots the bus has. Previously a PCI or PCI-X bus underneath a PCI Express host controller would not be enumerated properly. o Pull the interrupt routing logic out of pci_ocp_init() and into its own function. The logic is not quite right and is expected to be a bit more complex. o Fix/add support for PCI domains. The PCI domain is the unit number as per other PCI host controller drivers. As such, we can use logical bus numbers again and don't have to guarantee globally unique bus numbers. Remove pci_ocp_busnr. Return the highest bus number ito the caller of pci_ocp_init() now that we don't have a global variable anymore. o BAR programming fixes: - Non-type0 headers have at most 1 BAR, not 0. - First write ~0 to the BAR in question and then read back its size. Obtained from: Juniper Networks (mostly) Revision Changes Path 1.2 +65 -37 src/sys/powerpc/mpc85xx/pci_ocp.c From brooks at FreeBSD.org Wed Mar 5 17:51:07 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Wed Mar 5 17:51:10 2008 Subject: cvs commit: src/sbin/ddb ddb.8 ddb.c Message-ID: <200803051751.m25Hp7Bk079400@repoman.freebsd.org> brooks 2008-03-05 17:51:06 UTC FreeBSD src repository Modified files: sbin/ddb ddb.8 ddb.c Log: Add the ability to read a file of commands to ddb(8) modeled after the feature in ipfw(8). Revision Changes Path 1.3 +14 -0 src/sbin/ddb/ddb.8 1.2 +67 -5 src/sbin/ddb/ddb.c From brueffer at FreeBSD.org Wed Mar 5 18:04:21 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Wed Mar 5 18:04:27 2008 Subject: cvs commit: src/share/man/man4 uslcom.4 Message-ID: <200803051804.m25I4L2L081365@repoman.freebsd.org> brueffer 2008-03-05 18:04:21 UTC FreeBSD src repository Modified files: share/man/man4 uslcom.4 Log: mdoc changes to make this manpage consistent with our other section 4 manpages. Revision Changes Path 1.3 +19 -8 src/share/man/man4/uslcom.4 From kensmith at FreeBSD.org Wed Mar 5 18:15:05 2008 From: kensmith at FreeBSD.org (Ken Smith) Date: Wed Mar 5 18:15:08 2008 Subject: cvs commit: CVSROOT approvers Message-ID: <200803051815.m25IF5Il081907@repoman.freebsd.org> kensmith 2008-03-05 18:15:05 UTC FreeBSD src repository Modified files: . approvers Log: Turn RELENG_7_0 over to the Security Team. Approved by: core (implicit) Revision Changes Path 1.48 +1 -1 CVSROOT/approvers From peterjeremy at optushome.com.au Wed Mar 5 18:25:35 2008 From: peterjeremy at optushome.com.au (Peter Jeremy) Date: Wed Mar 5 18:25:43 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <200803051121.m25BLE03035426@repoman.freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> Message-ID: <20080305182531.GS68971@server.vk2pj.dyndns.org> On Wed, Mar 05, 2008 at 11:21:14AM +0000, Bruce Evans wrote: > Log: > Change float_t and double_t to long double on i386. All floating point > expressions on i386 are evaluated in the range of the long double type, npx.h currently defines __INITIAL_NPXCW__ as 0x127f which makes the x87 emulate double (53-bit) precision rather than its native 64-bit long double. Doesn't this change also need a corresponding change to npx.h to make the x87 run with 64-bit precision? -- Peter Jeremy Please excuse any delays as the result of my ISP's inability to implement an MTA that is either RFC2821-compliant or matches their claimed behaviour. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080305/531579d8/attachment.pgp From brooks at FreeBSD.org Wed Mar 5 18:32:59 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Wed Mar 5 18:33:04 2008 Subject: cvs commit: src/share/man/man5 rc.conf.5 src/etc Makefile ddb.conf src/etc/defaults rc.conf src/etc/rc.d Makefile ddb Message-ID: <200803051832.m25IWwmU082623@repoman.freebsd.org> brooks 2008-03-05 18:32:58 UTC FreeBSD src repository Modified files: share/man/man5 rc.conf.5 etc Makefile etc/defaults rc.conf etc/rc.d Makefile Added files: etc ddb.conf etc/rc.d ddb Log: Use the new command file feature of ddb(8) to support setting ddb(4) scripts at boot. This is currently disabled by default. /etc/ddb.conf contains some potentially reasonable default scripts. PR: conf/119995 Submitted by: Scot Hetzel (Earlier version) X-MFC after: textdumps Revision Changes Path 1.366 +1 -1 src/etc/Makefile 1.1 +15 -0 src/etc/ddb.conf (new) 1.327 +2 -0 src/etc/defaults/rc.conf 1.89 +1 -1 src/etc/rc.d/Makefile 1.1 +32 -0 src/etc/rc.d/ddb (new) 1.335 +13 -0 src/share/man/man5/rc.conf.5 From marcel at FreeBSD.org Wed Mar 5 18:39:02 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Wed Mar 5 18:39:08 2008 Subject: cvs commit: src/sys/powerpc/mpc85xx ocpbus.c Message-ID: <200803051839.m25Id18P082837@repoman.freebsd.org> marcel 2008-03-05 18:39:01 UTC FreeBSD src repository Modified files: sys/powerpc/mpc85xx ocpbus.c Log: o We don't have to keep track of the PIC, nor do we have to make sure it's probed first. The PowerPC platform code deals with everything. As such, probe devices in order of their location in the memory map. o Refactor the ocpbus_alloc_resource for readability and make sure we set the RID in the resource as per the new convention. Revision Changes Path 1.2 +39 -28 src/sys/powerpc/mpc85xx/ocpbus.c From jhb at FreeBSD.org Wed Mar 5 19:58:08 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 19:58:11 2008 Subject: cvs commit: src/sys/nfsclient nfs_socket.c nfs_subs.c nfsm_subs.h Message-ID: <200803051958.m25Jw8SG086517@repoman.freebsd.org> jhb 2008-03-05 19:58:08 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/nfsclient nfs_socket.c nfs_subs.c nfsm_subs.h Log: MFC: Consolidate the code to generate a new XID for a NFS request. Revision Changes Path 1.154.2.3 +1 -8 src/sys/nfsclient/nfs_socket.c 1.146.2.2 +23 -14 src/sys/nfsclient/nfs_subs.c 1.37.2.1 +1 -0 src/sys/nfsclient/nfsm_subs.h From jhb at FreeBSD.org Wed Mar 5 20:04:17 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 20:04:20 2008 Subject: cvs commit: src/sys/nfsclient nfs_socket.c nfs_subs.c nfsm_subs.h Message-ID: <200803052004.m25K4GSB087875@repoman.freebsd.org> jhb 2008-03-05 20:04:16 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/nfsclient nfs_socket.c nfs_subs.c nfsm_subs.h Log: MFC: Consolidate the code to generate a new XID for a NFS request. Revision Changes Path 1.125.2.19 +1 -5 src/sys/nfsclient/nfs_socket.c 1.140.2.4 +17 -11 src/sys/nfsclient/nfs_subs.c 1.36.2.2 +1 -0 src/sys/nfsclient/nfsm_subs.h From yar at FreeBSD.org Wed Mar 5 20:11:05 2008 From: yar at FreeBSD.org (Yar Tikhiy) Date: Wed Mar 5 20:11:08 2008 Subject: cvs commit: src/usr.bin/make main.c make.1 Message-ID: <200803052011.m25KB5kB088258@repoman.freebsd.org> yar 2008-03-05 20:11:05 UTC FreeBSD src repository Modified files: usr.bin/make main.c make.1 Log: Don't forget to set MAKEFLAGS in the childs' environment from the .MAKEFLAGS global variable even if it's empty or unset. This means setting MAKEFLAGS to just an empty string in the latter case. If not doing so, make(1) behaved inconsistently WRT MAKEFLAGS. In particular, it would let a `-f foo' option down to sub-makes if .MAKEFLAGS was unset. E.g., env MAKEFLAGS="-f mymakefile" make would pass `-f mymakefile' down to sub-makes via their environment (unless mymakefile added something to .MAKEFLAGS). But any additional options appearing would change this behaviour to not passing `-f mymakefile' to sub-makes, as in: env MAKEFLAGS="-f mymakefile" make -D DUMMY or env MAKEFLAGS="-f mymakefile -D DUMMY" make (unless mymakefile cleared .MAKEFLAGS). Also make(1) would leave MAKEFLAGS at its initial value if the makefile set .MAKEFLAGS to an empty value. I.e., it was impossible to override MAKEFLAGS with an empty value. (Note well that makefiles are not to touch MAKEFLAGS directly, they alter .MAKEFLAGS instead. So make(1) can filter out things such as -f when copying MAKEFLAGS to .MAKEFLAGS at startup. Direct modifications to MAKEFLAGS just go nowhere.) While the original intentions of the BSD make authors are somewhat unclear here, the bug proves that NOT passing -f options down is the settled behaviour because the opposite behaviour is totally unreliable in the presence of any other options. In addition, not passing down -f's found in the environment is consistent with doing so WRT the command line. Update the manpage accordingly and make the whole description of MAKEFLAGS and .MAKEFLAGS more consistent as this change indeed brings more consistency into the reliable behaviour of make(1). Submitted by: ru (main.c) Tested with: make world Revision Changes Path 1.166 +2 -0 src/usr.bin/make/main.c 1.108 +33 -31 src/usr.bin/make/make.1 From jhb at FreeBSD.org Wed Mar 5 20:17:13 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 20:17:23 2008 Subject: cvs commit: src/sys/nfs nfs_nqlease.c nfs_socket.c nfs_srvcache.c nfs_subs.c nfsm_subs.h Message-ID: <200803052017.m25KHCbB088495@repoman.freebsd.org> jhb 2008-03-05 20:17:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_4) sys/nfs nfs_nqlease.c nfs_socket.c nfs_srvcache.c nfs_subs.c nfsm_subs.h Log: MFC: Fix a problem with XID re-use when a server returns NFSERR_JUKEBOX. Revision Changes Path 1.50.2.1 +5 -5 src/sys/nfs/nfs_nqlease.c 1.60.2.8 +11 -3 src/sys/nfs/nfs_socket.c 1.21.2.1 +3 -1 src/sys/nfs/nfs_srvcache.c 1.90.2.5 +18 -12 src/sys/nfs/nfs_subs.c 1.27.2.3 +2 -1 src/sys/nfs/nfsm_subs.h From jhb at FreeBSD.org Wed Mar 5 20:19:01 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 20:19:07 2008 Subject: cvs commit: src/sys/nfs nfs_srvcache.c Message-ID: <200803052019.m25KJ09c088556@repoman.freebsd.org> jhb 2008-03-05 20:19:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_4) sys/nfs nfs_srvcache.c Log: Bah, extra hunk slipped in. Revision Changes Path 1.21.2.2 +0 -2 src/sys/nfs/nfs_srvcache.c From yar at FreeBSD.org Wed Mar 5 20:24:38 2008 From: yar at FreeBSD.org (Yar Tikhiy) Date: Wed Mar 5 20:24:47 2008 Subject: cvs commit: src/usr.bin/make main.c Message-ID: <200803052024.m25KOcO0088816@repoman.freebsd.org> yar 2008-03-05 20:24:38 UTC FreeBSD src repository Modified files: usr.bin/make main.c Log: The non-POSIX environment variable MAKE was superseded by MAKEFLAGS ages ago, so don't mention it in comments. Tested with: cmp(1) Revision Changes Path 1.167 +1 -1 src/usr.bin/make/main.c From jhb at FreeBSD.org Wed Mar 5 20:33:46 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 20:33:48 2008 Subject: cvs commit: src/sys/net route.c Message-ID: <200803052033.m25KXk7D089120@repoman.freebsd.org> jhb 2008-03-05 20:33:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/net route.c Log: MFC: Use RTFREE_LOCKED() instead of rtfree() when releasing a reference on the 'rt' route in rtredirect(). Revision Changes Path 1.120.2.3 +1 -1 src/sys/net/route.c From kris at FreeBSD.org Wed Mar 5 20:58:15 2008 From: kris at FreeBSD.org (Kris Kennaway) Date: Wed Mar 5 20:58:21 2008 Subject: cvs commit: src/secure/usr.bin/ssh Makefile src/secure/usr.sbin/sshd Makefile Message-ID: <200803052058.m25KwFP6099381@repoman.freebsd.org> kris 2008-03-05 20:58:15 UTC FreeBSD src repository Modified files: secure/usr.bin/ssh Makefile secure/usr.sbin/sshd Makefile Log: For users of FreeBSD <= 6.2 we recommend during the x.org 7.x upgrade that they add X11BASE=${LOCALBASE} to /etc/make.conf since X11BASE was hard-wired to the now-wrong location in old releases. However, both X11BASE and LOCALBASE have moved out of scope of src/ into ports/ now, which causes problems for upgraded users who have old make.conf files still containing the above setting. X11BASE becomes null and we instruct ssh and sshd to look for xauth in /bin/xauth where it is unlikely to be found. Instead, provide a copy of the default LOCALBASE?=/usr/local setting here. We also have to deal with the case where the user only overrides LOCALBASE and doesn't set an explicit X11BASE (in ports it will be set implicitly but not here), which will also move the location of xauth. MFC after: 3 days Reported by: rwatson Revision Changes Path 1.36 +10 -1 src/secure/usr.bin/ssh/Makefile 1.51 +9 -0 src/secure/usr.sbin/sshd/Makefile From jhb at FreeBSD.org Wed Mar 5 21:11:55 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 21:11:57 2008 Subject: cvs commit: src/sys/kern kern_mutex.c Message-ID: <200803052111.m25LBtON001250@repoman.freebsd.org> jhb 2008-03-05 21:11:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern kern_mutex.c Log: MFC: Add a couple of assertions and KTR logging to thread_lock_flags(). Revision Changes Path 1.198.2.2 +7 -1 src/sys/kern/kern_mutex.c From jhb at FreeBSD.org Wed Mar 5 21:12:50 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 21:12:52 2008 Subject: cvs commit: src/sys/kern subr_sleepqueue.c Message-ID: <200803052112.m25LCn8F001302@repoman.freebsd.org> jhb 2008-03-05 21:12:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern subr_sleepqueue.c Log: MFC: Mark sleepqueue chain spin mutexes as recursable. Revision Changes Path 1.39.2.2 +1 -1 src/sys/kern/subr_sleepqueue.c From jhb at FreeBSD.org Wed Mar 5 21:14:06 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 21:14:12 2008 Subject: cvs commit: src/sys/dev/syscons syscons.h Message-ID: <200803052114.m25LE6Ti001362@repoman.freebsd.org> jhb 2008-03-05 21:14:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/syscons syscons.h Log: MFC: Mark the syscons video spin mutex as recursable. Revision Changes Path 1.88.2.1 +2 -1 src/sys/dev/syscons/syscons.h From jhb at FreeBSD.org Wed Mar 5 21:17:59 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 21:18:02 2008 Subject: cvs commit: src/usr.bin/netstat route.c Message-ID: <200803052117.m25LHxPV001551@repoman.freebsd.org> jhb 2008-03-05 21:17:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/netstat route.c Log: MFC: Make netstat -rn more resilient to having the routing table change out from under it while running. Revision Changes Path 1.82.2.3 +43 -19 src/usr.bin/netstat/route.c From jhb at FreeBSD.org Wed Mar 5 21:19:26 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 5 21:19:28 2008 Subject: cvs commit: src/usr.bin/netstat route.c Message-ID: <200803052119.m25LJP6t001613@repoman.freebsd.org> jhb 2008-03-05 21:19:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/netstat route.c Log: MFC: Make netstat -rn more resilient to having the routing table change out from under it while running. Revision Changes Path 1.76.2.6 +43 -19 src/usr.bin/netstat/route.c From mav at FreeBSD.org Wed Mar 5 22:12:35 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Wed Mar 5 22:12:39 2008 Subject: cvs commit: src/sys/netgraph ng_base.c Message-ID: <200803052212.m25MCYBV005536@repoman.freebsd.org> mav 2008-03-05 22:12:34 UTC FreeBSD src repository Modified files: sys/netgraph ng_base.c Log: Increase default queue items allocation limit from 512 to 4096 items to avoid terrible unpredicted effects for netgraph operation of their exhaustion while allocating control messages. Add separate configurable 512 items limit for data items allocation for DoS/overload protection. Discussed with: julian Revision Changes Path 1.149 +16 -2 src/sys/netgraph/ng_base.c From rodrigc at crodrigues.org Wed Mar 5 22:24:02 2008 From: rodrigc at crodrigues.org (Craig Rodrigues) Date: Wed Mar 5 22:24:05 2008 Subject: cvs commit: src/sbin/fsck_ffs main.c In-Reply-To: <20080305122029.GA7027@dg.local> References: <200803050825.m258Ppv2016738@repoman.freebsd.org> <20080305122029.GA7027@dg.local> Message-ID: <20080305222402.GA80407@crodrigues.org> On Wed, Mar 05, 2008 at 03:20:29PM +0300, Yar Tikhiy wrote: > Your analysis of the problem sounds not quite correct to me. You make some interesting points in your e-mail. I suggest that you summarize the points and post them to arch@ for further review and discussion. There is how nmount() should work in an ideal world, and then there is the existing implentation, which is not exactly clean and perfect. It is important as we clean up the nmount() code in the tree, that we keep things working without introducing too many hacks. The problem with the mount code in FreeBSD is that a fix that appears "correct" in one area, may break things somewhere else, because that area has hacks/workarounds/improper coding/whatever that "just happened to work". So while you may disagree with the aesthetics of some of the mount patches.....I have been trying to slowly migrate things over and clean things up in the various mount binaries and file systems and still try to keep things working the way they have before. I've not been perfect and have made some mistakes along the way, but I try to clean things up. My complaints with your mode of operation have been: - you started committing in an area in which you don't have a lot of track record in working in - you started breaking things in configurations you were unable/unwilling/no time to fix or investigate - you post long e-mails, or commit messages complaining about how you don't like how the mount code is structured or works, but your posts tend to look like rants I agree with a lot of what you have posted, but you need to be more constructive, and post your suggestions to arch@ to get proper review and discussion going. I know you mean well, but you need to slow down and direct your focus more constructively....otherwise it looks like you are ranting and doing drive-by commits. Unfortunately, the mount code is core functionality, that when it doesn't work properly, people complain *A LOT*, so treading carefully is key. Cheers. -- Craig Rodrigues rodrigc@crodrigues.org From davidch at FreeBSD.org Wed Mar 5 22:58:03 2008 From: davidch at FreeBSD.org (David Christensen) Date: Wed Mar 5 22:58:06 2008 Subject: cvs commit: src/sys/dev/mii brgphy.c miidevs Message-ID: <200803052258.m25Mw27H011234@repoman.freebsd.org> davidch 2008-03-05 22:58:02 UTC FreeBSD src repository Modified files: sys/dev/mii brgphy.c miidevs Log: - Add PHY ID for BCM5709C 1000Base-T controllers. MFC after: 1 week Revision Changes Path 1.72 +4 -2 src/sys/dev/mii/brgphy.c 1.50 +2 -0 src/sys/dev/mii/miidevs From delphij at FreeBSD.org Wed Mar 5 23:17:20 2008 From: delphij at FreeBSD.org (Xin LI) Date: Wed Mar 5 23:17:23 2008 Subject: cvs commit: src/sbin/newfs mkfs.c Message-ID: <200803052317.m25NHJCC013131@repoman.freebsd.org> delphij 2008-03-05 23:17:19 UTC FreeBSD src repository Modified files: sbin/newfs mkfs.c Log: Use calloc(). Revision Changes Path 1.95 +1 -2 src/sbin/newfs/mkfs.c From delphij at FreeBSD.org Wed Mar 5 23:31:50 2008 From: delphij at FreeBSD.org (Xin LI) Date: Wed Mar 5 23:31:54 2008 Subject: cvs commit: src/sbin/geom Makefile src/sbin/geom/class/label geom_label.c src/sbin/geom/core geom.c Message-ID: <200803052331.m25NVoT8013683@repoman.freebsd.org> delphij 2008-03-05 23:31:49 UTC FreeBSD src repository Modified files: sbin/geom Makefile sbin/geom/class/label geom_label.c sbin/geom/core geom.c Log: Make it possible to build glabel into rescue geom(8) utility. Ok'ed by: marcel No objection: -current@ Revision Changes Path 1.3 +5 -2 src/sbin/geom/Makefile 1.11 +8 -3 src/sbin/geom/class/label/geom_label.c 1.35 +5 -0 src/sbin/geom/core/geom.c From delphij at FreeBSD.org Wed Mar 5 23:32:13 2008 From: delphij at FreeBSD.org (Xin LI) Date: Wed Mar 5 23:32:16 2008 Subject: cvs commit: src/rescue/rescue Makefile Message-ID: <200803052332.m25NWC6d013719@repoman.freebsd.org> delphij 2008-03-05 23:32:12 UTC FreeBSD src repository Modified files: rescue/rescue Makefile Log: Add an alias for glabel(8). Revision Changes Path 1.59 +1 -0 src/rescue/rescue/Makefile From stas at FreeBSD.org Thu Mar 6 00:02:08 2008 From: stas at FreeBSD.org (Stanislav Sedov) Date: Thu Mar 6 00:02:11 2008 Subject: cvs commit: src/sys/conf options.powerpc src/sys/dev/uart uart.h uart_bus_ocp.c uart_cpu_powerpc.c src/sys/kern subr_witness.c src/sys/powerpc/booke clock.c copyinout.c interrupt.c locore.S machdep.c pmap.c support.S swtch.S trap.c trap_subr.S ... In-Reply-To: <200803031717.m23HH18K068262@repoman.freebsd.org> References: <200803031717.m23HH18K068262@repoman.freebsd.org> Message-ID: <20080305234622.GC83691@dracon.ht-systems.ru> On Mon, Mar 03, 2008 at 05:17:01PM +0000 Rafal Jaworowski mentioned: > pci_ocp.c > Log: > Initial support for Freescale PowerQUICC III MPC85xx system-on-chip family. > > The PQ3 is a high performance integrated communications processing system > based on the e500 core, which is an embedded RISC processor that implements > the 32-bit Book E definition of the PowerPC architecture. For details refer > to: http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MPC8555E > Thank you all guys for this work! -- Stanislav Sedov ST4096-RIPE From jkim at FreeBSD.org Thu Mar 6 00:22:17 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Thu Mar 6 00:22:20 2008 Subject: cvs commit: src/usr.sbin/moused moused.c Message-ID: <200803060022.m260MHSx017085@repoman.freebsd.org> jkim 2008-03-06 00:22:17 UTC FreeBSD src repository Modified files: usr.sbin/moused moused.c Log: Prefer clock_gettime(2) over gettimeofday(2) and use CLOCK_MONOTONIC_FAST. It is only used to track elapsed time and it does not have to be precise. Revision Changes Path 1.80 +82 -63 src/usr.sbin/moused/moused.c From brde at optusnet.com.au Thu Mar 6 00:22:38 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Thu Mar 6 00:22:45 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080305182531.GS68971@server.vk2pj.dyndns.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> Message-ID: <20080306103025.A7605@delplex.bde.org> On Thu, 6 Mar 2008, Peter Jeremy wrote: > On Wed, Mar 05, 2008 at 11:21:14AM +0000, Bruce Evans wrote: >> Log: >> Change float_t and double_t to long double on i386. All floating point >> expressions on i386 are evaluated in the range of the long double type, > > npx.h currently defines __INITIAL_NPXCW__ as 0x127f which makes the x87 > emulate double (53-bit) precision rather than its native 64-bit long > double. Doesn't this change also need a corresponding change to npx.h > to make the x87 run with 64-bit precision? Not quite. The precision can be changed at runtime too, but much more than changing the control word is required for completely working 64-bit on precision x87s in C. It is also necessary to change gcc. Apart from the 20 year old bugs in the support for extra precision, it is necessary to not configure the almost-hard-coded support for FreeBSD's __INITIAL_NPXCW__ (this support mainly breaks increasing the precision at runtime by evaluating too much at compile time). Bruce From jkim at FreeBSD.org Thu Mar 6 00:24:11 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Thu Mar 6 00:24:18 2008 Subject: cvs commit: src/usr.sbin/moused moused.c Message-ID: <200803060024.m260OB8h017141@repoman.freebsd.org> jkim 2008-03-06 00:24:11 UTC FreeBSD src repository Modified files: usr.sbin/moused moused.c Log: Fix compiler warnings and style(9) bugs. Revision Changes Path 1.81 +202 -193 src/usr.sbin/moused/moused.c From jkim at FreeBSD.org Thu Mar 6 00:25:17 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Thu Mar 6 00:25:21 2008 Subject: cvs commit: src/usr.sbin/moused Makefile Message-ID: <200803060025.m260PGfF017248@repoman.freebsd.org> jkim 2008-03-06 00:25:16 UTC FreeBSD src repository Modified files: usr.sbin/moused Makefile Log: Mark moused(8) WARNS=6 clean. Revision Changes Path 1.10 +2 -0 src/usr.sbin/moused/Makefile From emax at FreeBSD.org Thu Mar 6 00:37:30 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Thu Mar 6 00:37:35 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c Message-ID: <200803060037.m260bTJY017717@repoman.freebsd.org> emax 2008-03-06 00:37:29 UTC FreeBSD src repository Modified files: usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c Log: Add an option to register DUN (Dial-Up Networking) service on the same RFCOMM channel if needed. There is really no good reason to not to support this. AT-command exchange can be faked with chat script in ppp.conf. MFC after: 1 week Revision Changes Path 1.15 +10 -1 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.8 1.7 +33 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c From delphij at FreeBSD.org Thu Mar 6 01:05:11 2008 From: delphij at FreeBSD.org (Xin LI) Date: Thu Mar 6 01:05:15 2008 Subject: cvs commit: src/lib/libc/resolv res_comp.c Message-ID: <200803060105.m2615Am3028953@repoman.freebsd.org> delphij 2008-03-06 01:05:10 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/resolv res_comp.c Log: MFC revision 1.5 date: 2008/02/16 00:16:49; author: delphij; state: Exp; lines: +2 -1 Allow underscore in domain names while resolving. While having underscore is a violation of RFC 1034 [STD 13], it is accepted by certain name servers as well as other popular operating systems' resolver library. Bugs are mine. Obtained from: ume Revision Changes Path 1.4.2.1 +2 -1 src/lib/libc/resolv/res_comp.c From delphij at FreeBSD.org Thu Mar 6 01:05:30 2008 From: delphij at FreeBSD.org (Xin LI) Date: Thu Mar 6 01:05:32 2008 Subject: cvs commit: src/lib/libc/resolv res_comp.c Message-ID: <200803060105.m2615UmH029018@repoman.freebsd.org> delphij 2008-03-06 01:05:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libc/resolv res_comp.c Log: MFC revision 1.5 date: 2008/02/16 00:16:49; author: delphij; state: Exp; lines: +2 -1 Allow underscore in domain names while resolving. While having underscore is a violation of RFC 1034 [STD 13], it is accepted by certain name servers as well as other popular operating systems' resolver library. Bugs are mine. Obtained from: ume Revision Changes Path 1.2.2.2 +2 -1 src/lib/libc/resolv/res_comp.c From yongari at FreeBSD.org Thu Mar 6 01:47:53 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Thu Mar 6 01:47:59 2008 Subject: cvs commit: src/sys/dev/nfe if_nfe.c if_nfereg.h if_nfevar.h Message-ID: <200803060147.m261lrNl030401@repoman.freebsd.org> yongari 2008-03-06 01:47:53 UTC FreeBSD src repository Modified files: sys/dev/nfe if_nfe.c if_nfereg.h if_nfevar.h Log: Nuke local jumbo allocator and switch to use of UMA backed page allocator for jumbo frame. Also remove unneeded jlist lock which is no longer required to protect jumbo buffers. With these changes jumbo frame performance of nfe(4) was slightly increased and users should not encounter jumbo buffer allocation failure anymore. Revision Changes Path 1.27 +6 -166 src/sys/dev/nfe/if_nfe.c 1.11 +0 -9 src/sys/dev/nfe/if_nfereg.h 1.6 +0 -10 src/sys/dev/nfe/if_nfevar.h From davidxu at FreeBSD.org Thu Mar 6 01:59:08 2008 From: davidxu at FreeBSD.org (David Xu) Date: Thu Mar 6 01:59:12 2008 Subject: cvs commit: src/lib/libthr/thread thr_create.c Message-ID: <200803060159.m261x8md030769@repoman.freebsd.org> davidxu 2008-03-06 01:59:08 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_create.c Log: Restore code setting new thread's scheduler parameters, I was thinking that there might be starvations, but because we have already locked the thread, the cpuset settings will always be done before the new thread does real-world work. Revision Changes Path 1.40 +11 -15 src/lib/libthr/thread/thr_create.c From davidxu at FreeBSD.org Thu Mar 6 02:07:18 2008 From: davidxu at FreeBSD.org (David Xu) Date: Thu Mar 6 02:07:21 2008 Subject: cvs commit: src/lib/libthr/thread thr_exit.c Message-ID: <200803060207.m2627I1e032229@repoman.freebsd.org> davidxu 2008-03-06 02:07:18 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_exit.c Log: Don't report death event to debugger if it is a forced exit. Revision Changes Path 1.25 +1 -1 src/lib/libthr/thread/thr_exit.c From das at FreeBSD.ORG Thu Mar 6 02:11:12 2008 From: das at FreeBSD.ORG (David Schultz) Date: Thu Mar 6 02:11:16 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080305182531.GS68971@server.vk2pj.dyndns.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> Message-ID: <20080306021222.GA46783@zim.MIT.EDU> On Wed, Mar 05, 2008, Colin Percival wrote: > Bruce Evans wrote: > > Modified files: > > sys/i386/include _types.h > > Log: > > Change float_t and double_t to long double on i386. > > Doesn't this have a rather severe performance impact on any code which > uses double_t? Yes, if the code is compiled with -msse2. Otherwise, doubles get fed to the i387 anyway. On Thu, Mar 06, 2008, Peter Jeremy wrote: > npx.h currently defines __INITIAL_NPXCW__ as 0x127f which makes the x87 > emulate double (53-bit) precision rather than its native 64-bit long > double. Doesn't this change also need a corresponding change to npx.h > to make the x87 run with 64-bit precision? Technically Bruce's change isn't wrong without changing the default precision, it's just rather pointless. (Why tell programs to store variables in a wider format if you're just going to evaluate them in a narrower format anyway?) So I guess I agree that if we're going to go down this path, we ought to just bite the bullet and change npx.h and contrib/gcc/config/i386/freebsd.h to use 64-bit precision by default on i386. From imp at bsdimp.com Thu Mar 6 02:13:24 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Thu Mar 6 02:13:31 2008 Subject: cvs commit: src/usr.sbin/moused Makefile In-Reply-To: <200803060025.m260PGfF017248@repoman.freebsd.org> References: <200803060025.m260PGfF017248@repoman.freebsd.org> Message-ID: <20080305.191223.-1615348076.imp@bsdimp.com> In message: <200803060025.m260PGfF017248@repoman.freebsd.org> Jung-uk Kim writes: : jkim 2008-03-06 00:25:16 UTC : : FreeBSD src repository : : Modified files: : usr.sbin/moused Makefile : Log: : Mark moused(8) WARNS=6 clean. NEVER SET WARNS to a higher level unless you've done a make universe. The arm tinderbox is failing: ===> usr.sbin/moused (all) cc -O -pipe -Wsystem-headers -Werror -Wall -Wno-format-y2k -W -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wreturn-type -Wcast-qual -Wwrite-strings -Wswitch -Wshadow -Wcast-align -Wunused-parameter -Wchar-subscripts -Winline -Wnested-externs -Wredundant-decls -Wno-pointer-sign -c /src/usr.sbin/moused/moused.c cc1: warnings being treated as errors /src/usr.sbin/moused/moused.c: In function 'r_timestamp': /src/usr.sbin/moused/moused.c:2550: warning: format '%ld' expects type 'long int', but argument 2 has type 'time_t' /src/usr.sbin/moused/moused.c:2562: warning: format '%ld' expects type 'long int', but argument 2 has type 'time_t' *** Error code 1 From imp at FreeBSD.org Thu Mar 6 02:13:53 2008 From: imp at FreeBSD.org (Warner Losh) Date: Thu Mar 6 02:13:55 2008 Subject: cvs commit: src/usr.sbin/moused Makefile Message-ID: <200803060213.m262Drud032448@repoman.freebsd.org> imp 2008-03-06 02:13:53 UTC FreeBSD src repository Modified files: usr.sbin/moused Makefile Log: This breaks the arm build, back it out until that's fixed Revision Changes Path 1.11 +1 -1 src/usr.sbin/moused/Makefile From jkim at FreeBSD.org Thu Mar 6 02:14:46 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Thu Mar 6 02:14:49 2008 Subject: cvs commit: src/usr.sbin/moused moused.c Message-ID: <200803060214.m262EjmG032525@repoman.freebsd.org> jkim 2008-03-06 02:14:45 UTC FreeBSD src repository Modified files: usr.sbin/moused moused.c Log: Cast time_t to long to mute tinderbox build failure. Revision Changes Path 1.82 +2 -2 src/usr.sbin/moused/moused.c From jkim at FreeBSD.org Thu Mar 6 02:26:54 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Thu Mar 6 02:26:57 2008 Subject: cvs commit: src/usr.sbin/moused Makefile In-Reply-To: <20080305.191223.-1615348076.imp@bsdimp.com> References: <200803060025.m260PGfF017248@repoman.freebsd.org> <20080305.191223.-1615348076.imp@bsdimp.com> Message-ID: <200803052126.45821.jkim@FreeBSD.org> On Wednesday 05 March 2008 09:12 pm, M. Warner Losh wrote: > In message: <200803060025.m260PGfF017248@repoman.freebsd.org> > > Jung-uk Kim writes: > : jkim 2008-03-06 00:25:16 UTC > : > : FreeBSD src repository > : > : Modified files: > : usr.sbin/moused Makefile > : Log: > : Mark moused(8) WARNS=6 clean. > > NEVER SET WARNS to a higher level unless you've done a make > universe. The arm tinderbox is failing: > > ===> usr.sbin/moused (all) > cc -O -pipe -Wsystem-headers -Werror -Wall -Wno-format-y2k -W > -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes > -Wpointer-arith -Wreturn-type -Wcast-qual -Wwrite-strings -Wswitch > -Wshadow -Wcast-align -Wunused-parameter -Wchar-subscripts -Winline > -Wnested-externs -Wredundant-decls -Wno-pointer-sign -c > /src/usr.sbin/moused/moused.c cc1: warnings being treated as errors > /src/usr.sbin/moused/moused.c: In function 'r_timestamp': > /src/usr.sbin/moused/moused.c:2550: warning: format '%ld' expects > type 'long int', but argument 2 has type 'time_t' > /src/usr.sbin/moused/moused.c:2562: warning: format '%ld' expects > type 'long int', but argument 2 has type 'time_t' *** Error code 1 It should be fixed now. Sorry. Jung-uk Kim From max at love2party.net Thu Mar 6 02:31:38 2008 From: max at love2party.net (Max Laier) Date: Thu Mar 6 02:31:44 2008 Subject: cvs commit: src/usr.sbin/moused moused.c In-Reply-To: <200803060214.m262EjmG032525@repoman.freebsd.org> References: <200803060214.m262EjmG032525@repoman.freebsd.org> Message-ID: <53687.192.168.4.151.1204770665.squirrel@router> Am Do, 6.03.2008, 03:14, schrieb Jung-uk Kim: > jkim 2008-03-06 02:14:45 UTC > > FreeBSD src repository > > Modified files: > usr.sbin/moused moused.c > Log: > Cast time_t to long to mute tinderbox build failure. doesn't this defeat the purpose of having time_t in the first place? I'd rather cast to intmax_t and print with %j ... though it probably doesn't matter. > Revision Changes Path > 1.82 +2 -2 src/usr.sbin/moused/moused.c > > -- /"\ Best regards, | mlaier@freebsd.org \ / Max Laier | ICQ #67774661 X http://pf4freebsd.love2party.net/ | mlaier@EFnet / \ ASCII Ribbon Campaign | Against HTML Mail and News From jkim at FreeBSD.org Thu Mar 6 02:43:58 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Thu Mar 6 02:44:00 2008 Subject: cvs commit: src/usr.sbin/moused moused.c In-Reply-To: <53687.192.168.4.151.1204770665.squirrel@router> References: <200803060214.m262EjmG032525@repoman.freebsd.org> <53687.192.168.4.151.1204770665.squirrel@router> Message-ID: <200803052143.48208.jkim@FreeBSD.org> On Wednesday 05 March 2008 09:31 pm, Max Laier wrote: > Am Do, 6.03.2008, 03:14, schrieb Jung-uk Kim: > > jkim 2008-03-06 02:14:45 UTC > > > > FreeBSD src repository > > > > Modified files: > > usr.sbin/moused moused.c > > Log: > > Cast time_t to long to mute tinderbox build failure. > > doesn't this defeat the purpose of having time_t in the first > place? I'd rather cast to intmax_t and print with %j ... though > it probably doesn't matter. Now the timestamp is uptime and the breakage was just debugging output, it shouldn't really matter at all but I will do that soon. Thanks, Jung-uk Kim From cperciva at freebsd.org Thu Mar 6 02:46:09 2008 From: cperciva at freebsd.org (Colin Percival) Date: Thu Mar 6 02:46:16 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306095645.V7605@delplex.bde.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <47CE8396.6020803@freebsd.org> <20080306095645.V7605@delplex.bde.org> Message-ID: <47CF5AE8.2090101@freebsd.org> Bruce Evans wrote: > On Wed, 5 Mar 2008, Colin Percival wrote: >> Bruce Evans wrote: >>> Change float_t and double_t to long double on i386. >> >> Doesn't this have a rather severe performance impact on any code which >> uses double_t? > > No. As mentioned in the commit message, this has no performance effect > except in cases where it avoids compiler bugs. [...] if you use long double > for memory variables then you get a severe performance impact and some > space loss for the load instruction, since loading long doubles is > much slower than loading doubles (about 4 times slower on Athlons). Either I'm misunderstanding something, or you seem to be disagreeing with yourself here... if I have the following code double_t foo, bar, foobar; foobar = foo + bar; then prior to this change the processor loads and stores doubles, while after this change the processor loads and stores long doubles, with the associated performance penalty. Colin Percival From cperciva at freebsd.org Thu Mar 6 02:56:32 2008 From: cperciva at freebsd.org (Colin Percival) Date: Thu Mar 6 02:56:33 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306021222.GA46783@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> Message-ID: <47CF5D19.3090100@freebsd.org> David Schultz wrote: > ... if we're going to go down this path, we ought to just bite > the bullet and change npx.h and contrib/gcc/config/i386/freebsd.h > to use 64-bit precision by default on i386. Why would we want to randomly and pointlessly break things? To quote Kahan, "if a programmer asks for IEEE double-precision arithmetic, of course that's what you should give him" -- double rounding is simply not acceptable. Colin Percival From brde at optusnet.com.au Thu Mar 6 03:15:41 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Thu Mar 6 03:15:44 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CE8396.6020803@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <47CE8396.6020803@freebsd.org> Message-ID: <20080306095645.V7605@delplex.bde.org> On Wed, 5 Mar 2008, Colin Percival wrote: > Bruce Evans wrote: >> Modified files: >> sys/i386/include _types.h >> Log: >> Change float_t and double_t to long double on i386. > > Doesn't this have a rather severe performance impact on any code which > uses double_t? No. As mentioned in the commit message, this has no performance effect except in cases where it avoids compiler bugs. In fact, I need it partly to improve performance in cases where other methods of avoiding the compiler bugs (including fixing the compiler) have a severe performance effect. Using a correctly defined double_t makes it possible for non-broken generated code to operate entirely on registers, while using plain double requires either broken generated code or lots of conversions via memory. E.g.: double x, y, z; z = x + y; Typical wrong generated i386 code for this: fadd %st(1),%st Non-broken code for this (get it by fixing the compiler or messing up the source code using something like STRICT_ASSIGN()): fadd %st(1),%st fstpl mem # or fstps/flds to convert to float fldl mem Using double_t == long double: double_t x, y, z; z = x + y; Typical generated i386 code for this: fadd %st(1),%st x and y would probably start out as doubles and be loaded into registers using fldl. This gives a long double whether you want it to or not (all FP registers are long double). Then the severe performance impact is from converting the long double back to "double z" on assignment as is required for C compilers. OTOH, if you use long double for memory variables then you get a severe performance impact and some space loss for the load instruction, since loading long doubles is much slower than loading doubles (about 4 times slower on Athlons). Bruce From davidxu at FreeBSD.org Thu Mar 6 03:24:04 2008 From: davidxu at FreeBSD.org (David Xu) Date: Thu Mar 6 03:24:05 2008 Subject: cvs commit: src/lib/libthr/thread thr_attr.c Message-ID: <200803060324.m263O3uN037715@repoman.freebsd.org> davidxu 2008-03-06 03:24:03 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_attr.c Log: Fix a bug when calculating remnant size. Revision Changes Path 1.11 +1 -1 src/lib/libthr/thread/thr_attr.c From das at FreeBSD.ORG Thu Mar 6 03:31:35 2008 From: das at FreeBSD.ORG (David Schultz) Date: Thu Mar 6 03:31:40 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF5D19.3090100@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> Message-ID: <20080306033246.GA47280@zim.MIT.EDU> On Wed, Mar 05, 2008, Colin Percival wrote: > David Schultz wrote: > > ... if we're going to go down this path, we ought to just bite > > the bullet and change npx.h and contrib/gcc/config/i386/freebsd.h > > to use 64-bit precision by default on i386. > > Why would we want to randomly and pointlessly break things? To quote > Kahan, "if a programmer asks for IEEE double-precision arithmetic, of > course that's what you should give him" -- double rounding is simply > not acceptable. gcc/i386 doesn't correctly support IEEE 754 floating point no matter which option we choose. It's really a question of whether we want unexpected behavior for doubles or for long doubles. First, what you say above isn't quite right. Even IEEE 754R says that implementations should be allowed to use wider precision when available, but there are four cases where values are required to be rounded to the precision of the type: assignments, casts, function arguments, and function return values. C99 has a subset of these requirements, but it's pretty clear that the C99 authors were confused about some of the details. gcc doesn't get any of this right. FreeBSD/i386 fixes the problem for /doubles/ by setting the i387 to use 53-bit precision. The downside is that this breaks long doubles. 15 years ago, this was clearly the right thing to do, because we had a double precision libm that would break when gcc got things wrong, and no long double support anyway. Now we have a bunch of long double math functions that have problems as a result of the reduced precision. Furthermore, since gcc is the 800-pound gorilla here, many programmers have grown accustomed to its shortcomings. From das at FreeBSD.ORG Thu Mar 6 03:36:55 2008 From: das at FreeBSD.ORG (David Schultz) Date: Thu Mar 6 03:36:57 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF5AE8.2090101@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <47CE8396.6020803@freebsd.org> <20080306095645.V7605@delplex.bde.org> <47CF5AE8.2090101@freebsd.org> Message-ID: <20080306033807.GB47280@zim.MIT.EDU> On Wed, Mar 05, 2008, Colin Percival wrote: > Bruce Evans wrote: > > On Wed, 5 Mar 2008, Colin Percival wrote: > >> Bruce Evans wrote: > >>> Change float_t and double_t to long double on i386. > >> > >> Doesn't this have a rather severe performance impact on any code which > >> uses double_t? > > > > No. As mentioned in the commit message, this has no performance effect > > except in cases where it avoids compiler bugs. [...] if you use long double > > for memory variables then you get a severe performance impact and some > > space loss for the load instruction, since loading long doubles is > > much slower than loading doubles (about 4 times slower on Athlons). > > Either I'm misunderstanding something, or you seem to be disagreeing with > yourself here... if I have the following code > > double_t foo, bar, foobar; > foobar = foo + bar; > > then prior to this change the processor loads and stores doubles, while > after this change the processor loads and stores long doubles, with the > associated performance penalty. If gcc actually implemented IEEE 754R / C99 / LIA1 correctly, then when compiling something like 'double x = y' would require it to store the value to memory and then reload it to force it to 53-bit precision. The whole point of double_t is to allow the programmer to ask for a variable that is "at least double precision", so that the compiler isn't forced to clip everything to double precision if it's inefficient to do so. However, gcc doesn't compile the above expression correctly, and FreeBSD works around the problem by setting the default rounding mode to 53 bits, which is why it's pointless to increase the size of double_t. From imp at bsdimp.com Thu Mar 6 04:07:32 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Thu Mar 6 04:07:36 2008 Subject: cvs commit: src/usr.sbin/moused Makefile In-Reply-To: <200803052126.45821.jkim@FreeBSD.org> References: <200803060025.m260PGfF017248@repoman.freebsd.org> <20080305.191223.-1615348076.imp@bsdimp.com> <200803052126.45821.jkim@FreeBSD.org> Message-ID: <20080305.210619.776506336.imp@bsdimp.com> In message: <200803052126.45821.jkim@FreeBSD.org> Jung-uk Kim writes: : On Wednesday 05 March 2008 09:12 pm, M. Warner Losh wrote: : > In message: <200803060025.m260PGfF017248@repoman.freebsd.org> : > : > Jung-uk Kim writes: : > : jkim 2008-03-06 00:25:16 UTC : > : : > : FreeBSD src repository : > : : > : Modified files: : > : usr.sbin/moused Makefile : > : Log: : > : Mark moused(8) WARNS=6 clean. : > : > NEVER SET WARNS to a higher level unless you've done a make : > universe. The arm tinderbox is failing: : > : > ===> usr.sbin/moused (all) : > cc -O -pipe -Wsystem-headers -Werror -Wall -Wno-format-y2k -W : > -Wno-unused-parameter -Wstrict-prototypes -Wmissing-prototypes : > -Wpointer-arith -Wreturn-type -Wcast-qual -Wwrite-strings -Wswitch : > -Wshadow -Wcast-align -Wunused-parameter -Wchar-subscripts -Winline : > -Wnested-externs -Wredundant-decls -Wno-pointer-sign -c : > /src/usr.sbin/moused/moused.c cc1: warnings being treated as errors : > /src/usr.sbin/moused/moused.c: In function 'r_timestamp': : > /src/usr.sbin/moused/moused.c:2550: warning: format '%ld' expects : > type 'long int', but argument 2 has type 'time_t' : > /src/usr.sbin/moused/moused.c:2562: warning: format '%ld' expects : > type 'long int', but argument 2 has type 'time_t' *** Error code 1 : : It should be fixed now. Sorry. Thanks! Warner From imp at FreeBSD.org Thu Mar 6 04:09:07 2008 From: imp at FreeBSD.org (Warner Losh) Date: Thu Mar 6 04:09:09 2008 Subject: cvs commit: src/usr.sbin/moused Makefile Message-ID: <200803060409.m26496Dm040342@repoman.freebsd.org> imp 2008-03-06 04:09:06 UTC FreeBSD src repository Modified files: usr.sbin/moused Makefile Log: The breakage from increased warns has been fixed, back out my back out Revision Changes Path 1.12 +1 -1 src/usr.sbin/moused/Makefile From silby at silby.com Thu Mar 6 05:13:36 2008 From: silby at silby.com (Mike Silbersack) Date: Thu Mar 6 05:13:41 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <200803051121.m25BLE03035426@repoman.freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> Message-ID: <20080305230954.X55190@odysseus.silby.com> On Wed, 5 Mar 2008, Bruce Evans wrote: > bde 2008-03-05 11:21:14 UTC > > FreeBSD src repository > > Modified files: > sys/i386/include _types.h > Log: > Change float_t and double_t to long double on i386. All floating point For those of us who are not floating point experts, can you explain a few things? 1) Does this really change every double to a long double in anything compiled? 2) How does this affect ABI compatibility? 3) How does this change the operation of various programs in the ports tree that use floating point, such as mplayer, mpg123, etc. Will this cause different behavior when these apps are used on FreeBSD vs other operating systems? As I've said in the past, I'd really, really, really like to see regression tests for any change to the floating point functions. The types of changes you've been making are not easy to verify just by looking at diffs. Thanks, -Mike From cperciva at freebsd.org Thu Mar 6 05:19:21 2008 From: cperciva at freebsd.org (Colin Percival) Date: Thu Mar 6 05:19:25 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306033807.GB47280@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <47CE8396.6020803@freebsd.org> <20080306095645.V7605@delplex.bde.org> <47CF5AE8.2090101@freebsd.org> <20080306033807.GB47280@zim.MIT.EDU> Message-ID: <47CF7EA6.90802@freebsd.org> David Schultz wrote: > If gcc actually implemented IEEE 754R / C99 / LIA1 correctly, then > when compiling something like 'double x = y' would require it to > store the value to memory and then reload it to force it to 53-bit > precision. Even this wouldn't work, since it would result in double rounding. > The whole point of double_t is to allow the programmer > to ask for a variable that is "at least double precision", so that > the compiler isn't forced to clip everything to double precision > if it's inefficient to do so. Right -- and that's why numerical analysts should use double (after proving that their code will produce the correct results using double-precision arithmetic) while people who don't understand floating-point math should always use double_t. > However, gcc doesn't compile the above > expression correctly, and FreeBSD works around the problem by > setting the default rounding mode to 53 bits, which is why it's > pointless to increase the size of double_t. I disagee with your characterization of setting the rounding mode as a "workaround", but your conclusion is correct: At the moment, increasing the size of double_t is pointless. Colin Percival From cperciva at freebsd.org Thu Mar 6 05:19:37 2008 From: cperciva at freebsd.org (Colin Percival) Date: Thu Mar 6 05:19:41 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306033246.GA47280@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> Message-ID: <47CF7EBF.6000009@freebsd.org> David Schultz wrote: > On Wed, Mar 05, 2008, Colin Percival wrote: >> David Schultz wrote: >>> ... if we're going to go down this path, we ought to just bite >>> the bullet and change npx.h and contrib/gcc/config/i386/freebsd.h >>> to use 64-bit precision by default on i386. >> Why would we want to randomly and pointlessly break things? To quote >> Kahan, "if a programmer asks for IEEE double-precision arithmetic, of >> course that's what you should give him" -- double rounding is simply >> not acceptable. > > gcc/i386 doesn't correctly support IEEE 754 floating point no > matter which option we choose. It's really a question of whether > we want unexpected behavior for doubles or for long doubles. As far as I can see, there's no question here. C99 requires that the "double" type matches the IEEE 754 64-bit double; but it allows the "long double" type to be an IEEE 754 80-bit extended, a 128-bit quadruple, a 64-bit double, or anything else which is at least as good as a 64-bit double. Setting the i387 FPU to 53-bit precision gives standards-compliant behaviour whether people are using "double" or "long double". Setting it to 64-bit precision gives standards-compliant behaviour when people are using long doubles, but not when people are using doubles. > First, what you say above isn't quite right. Even IEEE 754R says > that implementations should be allowed to use wider precision when > available, but there are four cases where values are required to > be rounded to the precision of the type: assignments, casts, > function arguments, and function return values. Yes and no. Double rounding isn't allowed; so performing operations with extended precision and then rounding to double precision later is not a valid option. The only way on the i387 to provide standards compliant floating-point arithmetic is to have the FPU set to not use the extra precision in the first place. > The downside is that this breaks long doubles. Except that it doesn't. Using 53-bit precision for long doubles is a perfectly valid and standards-compliant option. Colin Percival From cperciva at freebsd.org Thu Mar 6 05:30:15 2008 From: cperciva at freebsd.org (Colin Percival) Date: Thu Mar 6 05:30:22 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080305230954.X55190@odysseus.silby.com> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305230954.X55190@odysseus.silby.com> Message-ID: <47CF8161.1060202@freebsd.org> Mike Silbersack wrote: > On Wed, 5 Mar 2008, Bruce Evans wrote: >> Change float_t and double_t to long double on i386. All floating point > > 1) Does this really change every double to a long double in anything > compiled? No, it changes double_t (which is not the same as double). > 2) How does this affect ABI compatibility? This breaks ABI compatibility (when people use float_t or double_t). > 3) How does this change the operation of various programs in the ports > tree that use floating point, such as mplayer, mpg123, etc. Will this > cause different behavior when these apps are used on FreeBSD vs other > operating systems? Most code in the ports tree will be using double instead of double_t, and will consequently not be affected. Code which uses double_t will be slower due to the increased cost of loads and stores, and may produce different output if it changes the i387 precision flags. At the moment, FreeBSD behaves the same way as Microsoft Windows and C99, and differently from Linux: Linux sets the i387 by default to extended precision, which has the result of decreasing the average rounding error while increasing the maximum rounding error (due to double rounding when values which were internally rounded to extended precision are rounded to double precision later) and sometimes breaking code entirely (when the properties of floating-point rounding are used deliberately, e.g., to rapidly round a floating-point value to the nearest integer). I've answered queries from some mathematicians and scientists who were confused as to why they were seeing higher rounding errors on FreeBSD and Windows than they were on Linux; but when I've explained the behavioural differences to numerical analysts, the universal reaction I've received has been "dear god, Linux does WHAT?" -- oddly enough, mathematicians like to think that proving that their code is correct will be enough to ensure that it will produce correct output. Colin Percival From sgk at troutmask.apl.washington.edu Thu Mar 6 05:35:23 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Thu Mar 6 05:35:30 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080305230954.X55190@odysseus.silby.com> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305230954.X55190@odysseus.silby.com> Message-ID: <20080306053446.GA90967@troutmask.apl.washington.edu> On Wed, Mar 05, 2008 at 11:13:33PM -0600, Mike Silbersack wrote: > > 3) How does this change the operation of various programs in the ports > tree that use floating point, such as mplayer, mpg123, etc. Will this > cause different behavior when these apps are used on FreeBSD vs other > operating systems? IIRC, mplayer (as does several other ports) uses -ffast-math, which means mplayer doesn't care about the floating point. -- Steve From cperciva at FreeBSD.org Thu Mar 6 05:41:46 2008 From: cperciva at FreeBSD.org (Colin Percival) Date: Thu Mar 6 05:41:48 2008 Subject: cvs commit: src/sys/kern uipc_syscalls.c Message-ID: <200803060541.m265fjt7053744@repoman.freebsd.org> cperciva 2008-03-06 05:41:45 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern uipc_syscalls.c Log: MFC revision 1.269: Fix sending trailers in sendfile(2). Revision Changes Path 1.259.2.3 +3 -1 src/sys/kern/uipc_syscalls.c From peterjeremy at optushome.com.au Thu Mar 6 06:23:04 2008 From: peterjeremy at optushome.com.au (Peter Jeremy) Date: Thu Mar 6 06:23:10 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306033246.GA47280@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> Message-ID: <20080306062301.GX68971@server.vk2pj.dyndns.org> On Wed, Mar 05, 2008 at 10:32:46PM -0500, David Schultz wrote: >gcc doesn't get any of this right. FreeBSD/i386 fixes the problem >for /doubles/ by setting the i387 to use 53-bit precision. Note that FreeBSD/amd64 does not do this when running i386 binaries so you can get discrepancies between running the same program natively vs running it in compatibility mode. -- Peter Jeremy Please excuse any delays as the result of my ISP's inability to implement an MTA that is either RFC2821-compliant or matches their claimed behaviour. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080306/de2740b6/attachment.pgp From das at FreeBSD.ORG Thu Mar 6 06:27:21 2008 From: das at FreeBSD.ORG (David Schultz) Date: Thu Mar 6 06:27:25 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF7EA6.90802@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <47CE8396.6020803@freebsd.org> <20080306095645.V7605@delplex.bde.org> <47CF5AE8.2090101@freebsd.org> <20080306033807.GB47280@zim.MIT.EDU> <47CF7EA6.90802@freebsd.org> Message-ID: <20080306062834.GA48339@zim.MIT.EDU> On Wed, Mar 05, 2008, Colin Percival wrote: > David Schultz wrote: > > If gcc actually implemented IEEE 754R / C99 / LIA1 correctly, then > > when compiling something like 'double x = y' would require it to > > store the value to memory and then reload it to force it to 53-bit > > precision. > > Even this wouldn't work, since it would result in double rounding. Nevertheless, this is all that is guaranteed because it is often impractical to do better, and because these semantics are often sufficient. The usual formulas for splitting a number into hi + lo components work, for example. Furthermore, implementations that use extra guard bits in their FP registers can avoid the double rounding. > > The whole point of double_t is to allow the programmer > > to ask for a variable that is "at least double precision", so that > > the compiler isn't forced to clip everything to double precision > > if it's inefficient to do so. > > Right -- and that's why numerical analysts should use double (after > proving that their code will produce the correct results using > double-precision arithmetic) while people who don't understand > floating-point math should always use double_t. Not really. There are times when it matters and times when it doesn't. If you're evaluating a Chebyshev polynomial via Horner's formula, for example, imprecisions in rounding can generally be ignored until the last one or two multiplications. Similarly, if you're doing Newton iteration, only roundoff in the last iteration matters. At other times it really does matter, and numerical analysts know how to tell the difference. From das at FreeBSD.ORG Thu Mar 6 06:33:39 2008 From: das at FreeBSD.ORG (David Schultz) Date: Thu Mar 6 06:33:41 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF7EBF.6000009@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> Message-ID: <20080306063452.GB48339@zim.MIT.EDU> On Wed, Mar 05, 2008, Colin Percival wrote: > Setting the i387 FPU to 53-bit precision gives standards-compliant > behaviour whether people are using "double" or "long double". Not quite. > Yes and no. Double rounding isn't allowed; Yes, it is. > > The downside is that this breaks long doubles. > > Except that it doesn't. Using 53-bit precision for long doubles is > a perfectly valid and standards-compliant option. It's impossible to implement efficient library functions that produce correct long double results when the FPU is set to 64-bit mode and avoid double rounding and related problems when the FPU is set to 53-bit mode. (Many of these functions use bit fiddling, or range checks that are based on the expected width of the type, for example.) In practice, there are other impediments to making long doubles work correctly, such as constant folding and gcc's lack of support for FENV_ACCESS. From das at FreeBSD.ORG Thu Mar 6 06:36:33 2008 From: das at FreeBSD.ORG (David Schultz) Date: Thu Mar 6 06:36:35 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306062301.GX68971@server.vk2pj.dyndns.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <20080306062301.GX68971@server.vk2pj.dyndns.org> Message-ID: <20080306063746.GC48339@zim.MIT.EDU> On Thu, Mar 06, 2008, Peter Jeremy wrote: > On Wed, Mar 05, 2008 at 10:32:46PM -0500, David Schultz wrote: > >gcc doesn't get any of this right. FreeBSD/i386 fixes the problem > >for /doubles/ by setting the i387 to use 53-bit precision. > > Note that FreeBSD/amd64 does not do this when running i386 binaries > so you can get discrepancies between running the same program natively > vs running it in compatibility mode. Good point. I don't think i386 ever got fixed to do the appropriate mode change for Linux binaries, either, BTW. From cperciva at freebsd.org Thu Mar 6 06:57:14 2008 From: cperciva at freebsd.org (Colin Percival) Date: Thu Mar 6 06:57:19 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306063452.GB48339@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> Message-ID: <47CF9586.70707@freebsd.org> David Schultz wrote: > On Wed, Mar 05, 2008, Colin Percival wrote: >> Setting the i387 FPU to 53-bit precision gives standards-compliant >> behaviour whether people are using "double" or "long double". > > Not quite. How is it not standards-compliant? >> Yes and no. Double rounding isn't allowed; > > Yes, it is. No, it isn't. The following code double x, y, z; x = 1.0; y = 0x1.0p-53 + 0x1.0p-75; z = x + y; z = z - 1.0; if (z == 0.0) printf("BUGGY MATH\n"); else printf("IEEE754-COMPLIANT MATH\n"); should never output "BUGGY MATH", since x + y should round up to 1 + 2^(-52); but if you set the i387 to use extended precision, it gets the wrong answer. >>> The downside is that this breaks long doubles. >> Except that it doesn't. Using 53-bit precision for long doubles is >> a perfectly valid and standards-compliant option. > > It's impossible to implement efficient library functions that > produce correct long double results You could have stopped this sentence here -- for all practical purposes, correctly rounded trigonometric functions are not feasible. > when the FPU is set to 64-bit > mode and avoid double rounding and related problems when the FPU > is set to 53-bit mode. Fortunately, library functions aren't required to have any particular error bounds. It's always nice if they are correct to within 0.6ulp instead of 0.8ulp or suchlike -- but nobody should be proving code correctness based on assumed properties of library transcendental functions. People should, and do, prove code correct based on the assumption that double precision arithmetic behaves like double precision arithmetic, however (myself included). Colin Percival From brueffer at FreeBSD.org Thu Mar 6 07:19:51 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Thu Mar 6 07:19:53 2008 Subject: cvs commit: src/share/man/man4 fwohci.4 Message-ID: <200803060719.m267JoTe059334@repoman.freebsd.org> brueffer 2008-03-06 07:19:50 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 fwohci.4 Log: MFC: rev. 1.17 The Sony i.LINK (CXD1947) is _not_ supported. Revision Changes Path 1.16.2.1 +1 -3 src/share/man/man4/fwohci.4 From brueffer at FreeBSD.org Thu Mar 6 07:20:53 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Thu Mar 6 07:20:59 2008 Subject: cvs commit: src/share/man/man4 fwohci.4 Message-ID: <200803060720.m267Krl5059457@repoman.freebsd.org> brueffer 2008-03-06 07:20:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 fwohci.4 Log: MFC: rev. 1.17 The Sony i.LINK (CXD1947) is _not_ supported. Revision Changes Path 1.14.2.2 +1 -3 src/share/man/man4/fwohci.4 From das at FreeBSD.ORG Thu Mar 6 07:56:36 2008 From: das at FreeBSD.ORG (David Schultz) Date: Thu Mar 6 07:56:44 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF9586.70707@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> <47CF9586.70707@freebsd.org> Message-ID: <20080306075750.GA48995@zim.MIT.EDU> On Wed, Mar 05, 2008, Colin Percival wrote: > David Schultz wrote: > > On Wed, Mar 05, 2008, Colin Percival wrote: > >> Setting the i387 FPU to 53-bit precision gives standards-compliant > >> behaviour whether people are using "double" or "long double". > > > > Not quite. > > How is it not standards-compliant? First, because there are other issues, like constant folding and dynamic precision and rounding, that gcc doesn't get right. Second, because of the issues I mentioned with library functions, and because of ill-defined meanings of constants such as LDBL_MANT_DIG, etc. > >> Yes and no. Double rounding isn't allowed; > > > > Yes, it is. > > No, it isn't. The following code [...] > should never output "BUGGY MATH", since x + y should round up to > 1 + 2^(-52); but if you set the i387 to use extended precision, it > gets the wrong answer. True, but this is idealistic, and standards do not guarantee what you want because it is too hard to do efficiently in practice. Even the IEEE 754R draft basically says that double rounding is allowed as long as the language makes it clear what is going on: The last operation of many expressions is an assignment to an explicit final destination variable. As a part of expression evaluation rules, language standards shall specify when the next to last operation is performed by rounding at most once to the format of the explicit final destination, and when by rounding as many as two times, first to an implicit intermediate format, and then to the explicit final destination format. See the description of C99's FLT_EVAL_METHOD macro, which is part of C99's answer to the above. If you asked Kahan, he'd probably tell you that both strict and non-strict evaluation should be options. He was vehemently opposed to the idea that Java have only a strict mode (similar to the one you propose). > >>> The downside is that this breaks long doubles. > >> Except that it doesn't. Using 53-bit precision for long doubles is > >> a perfectly valid and standards-compliant option. > > > > It's impossible to implement efficient library functions that > > produce correct long double results > > You could have stopped this sentence here -- for all practical purposes, > correctly rounded trigonometric functions are not feasible. First, I said nothing about trigonometric anything, and second, your claim here isn't true anyway. FreeBSD provides trig functions that are correctly rounded most of the time (which is what most people call "correctly rounded"). Both the IBM Accurate Mathematical Library and Sun's crlibm provide trig functions that are correctly rounded all of the time. > > when the FPU is set to 64-bit > > mode and avoid double rounding and related problems when the FPU > > is set to 53-bit mode. > > Fortunately, library functions aren't required to have any particular > error bounds. It's always nice if they are correct to within 0.6ulp > instead of 0.8ulp or suchlike -- but nobody should be proving code > correctness based on assumed properties of library transcendental > functions. People should, and do, prove code correct based on the > assumption that double precision arithmetic behaves like double precision > arithmetic, however (myself included). No, you misunderstand my point. I'm not talking about a 1-ulp error, I'm talking about billions of ulps. In some long double functions, we actually have to do things like the following to get results that are even sane at all: #ifdef __i386__ if (fpgetprec() != FP_PE) return (double_precision_foo(x)); #endif From rink at FreeBSD.org Thu Mar 6 08:09:46 2008 From: rink at FreeBSD.org (Rink Springer) Date: Thu Mar 6 08:09:49 2008 Subject: cvs commit: src/share/man/man4 Makefile cmx.4 src/sys/conf NOTES files src/sys/dev/cmx cmx.c cmx_pccard.c cmxreg.h cmxvar.h src/sys/dev/pccard pccarddevs src/sys/modules Makefile src/sys/modules/cmx Makefile Message-ID: <200803060809.m2689ktf062200@repoman.freebsd.org> rink 2008-03-06 08:09:45 UTC FreeBSD src repository Modified files: share/man/man4 Makefile sys/conf NOTES files sys/dev/pccard pccarddevs sys/modules Makefile Added files: share/man/man4 cmx.4 sys/dev/cmx cmx.c cmx_pccard.c cmxreg.h cmxvar.h sys/modules/cmx Makefile Log: Commit cmx(4), a driver for Omnikey CardMan 4040 PCMCIA smartcard readers. PR: kern/114582 Submitted by: Daniel Roethlisberger Reviewed by: imp, myself Tested by: johans, myself MFC after: 2 weeks Revision Changes Path 1.412 +1 -0 src/share/man/man4/Makefile 1.1 +95 -0 src/share/man/man4/cmx.4 (new) 1.1474 +2 -0 src/sys/conf/NOTES 1.1278 +2 -0 src/sys/conf/files 1.1 +701 -0 src/sys/dev/cmx/cmx.c (new) 1.1 +115 -0 src/sys/dev/cmx/cmx_pccard.c (new) 1.1 +66 -0 src/sys/dev/cmx/cmxreg.h (new) 1.1 +99 -0 src/sys/dev/cmx/cmxvar.h (new) 1.131 +4 -0 src/sys/dev/pccard/pccarddevs 1.558 +4 -0 src/sys/modules/Makefile 1.1 +12 -0 src/sys/modules/cmx/Makefile (new) From yar at comp.chem.msu.su Thu Mar 6 08:43:57 2008 From: yar at comp.chem.msu.su (Yar Tikhiy) Date: Thu Mar 6 08:44:05 2008 Subject: cvs commit: src/sbin/fsck_ffs main.c In-Reply-To: <20080305222402.GA80407@crodrigues.org> References: <200803050825.m258Ppv2016738@repoman.freebsd.org> <20080305122029.GA7027@dg.local> <20080305222402.GA80407@crodrigues.org> Message-ID: <20080306084354.GA10064@dg.local> On Wed, Mar 05, 2008 at 10:24:02PM +0000, Craig Rodrigues wrote: > On Wed, Mar 05, 2008 at 03:20:29PM +0300, Yar Tikhiy wrote: > > Your analysis of the problem sounds not quite correct to me. > > You make some interesting points in your e-mail. > I suggest that you summarize the points and > post them to arch@ for further review and discussion. > > There is how nmount() should work in an ideal world, > and then there is the existing implentation, which is > not exactly clean and perfect. It is important as > we clean up the nmount() code in the tree, that we > keep things working without introducing too many hacks. > > The problem with the mount code in FreeBSD is that a fix that > appears "correct" in one area, may break things somewhere else, because > that area has hacks/workarounds/improper coding/whatever > that "just happened to work". So while you may disagree with > the aesthetics of some of the mount patches.....I have been trying > to slowly migrate things over and clean things up in the various > mount binaries and file systems and still try to keep things working > the way they have before. I've not been perfect and have made some > mistakes along the way, but I try to clean things up. > > My complaints with your mode of operation have been: > - you started committing in an area in which you don't have > a lot of track record in working in > - you started breaking things in configurations you were > unable/unwilling/no time to fix or investigate > - you post long e-mails, or commit messages complaining about how > you don't like how the mount code is structured or works, but your > posts tend to look like rants > > I agree with a lot of what you have posted, but you need to > be more constructive, and post your suggestions to arch@ to get > proper review and discussion going. > > I know you mean well, but you need to slow down and direct your > focus more constructively....otherwise it looks like you are > ranting and doing drive-by commits. > > Unfortunately, the mount code is core functionality, that when it > doesn't work properly, people complain *A LOT*, so treading carefully > is key. I'd appreciate if you respond to my technical criticism in equally _technical_ terms and refrain from diverting the discussion onto the moralization road. -- Yar From rink at FreeBSD.org Thu Mar 6 08:47:17 2008 From: rink at FreeBSD.org (Rink Springer) Date: Thu Mar 6 08:47:20 2008 Subject: cvs commit: src/share/man/man4 cmx.4 Message-ID: <200803060847.m268lGLX070935@repoman.freebsd.org> rink 2008-03-06 08:47:16 UTC FreeBSD src repository Modified files: share/man/man4 cmx.4 Log: Added missing BSD license text to the cmx.4 manpage. Noticed by: brueffer Revision Changes Path 1.2 +25 -0 src/share/man/man4/cmx.4 From rdivacky at FreeBSD.org Thu Mar 6 09:21:00 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Thu Mar 6 09:21:04 2008 Subject: cvs commit: src/usr.sbin/moused moused.c In-Reply-To: <200803060022.m260MHSx017085@repoman.freebsd.org> References: <200803060022.m260MHSx017085@repoman.freebsd.org> Message-ID: <20080306092045.GA54220@freebsd.org> On Thu, Mar 06, 2008 at 12:22:17AM +0000, Jung-uk Kim wrote: > jkim 2008-03-06 00:22:17 UTC > > FreeBSD src repository > > Modified files: > usr.sbin/moused moused.c > Log: > Prefer clock_gettime(2) over gettimeofday(2) and use CLOCK_MONOTONIC_FAST. > It is only used to track elapsed time and it does not have to be precise. what is the rationale behind this? what is the expected improvement? I am not questioning the change I am just curious.. thnx From yar at comp.chem.msu.su Thu Mar 6 09:45:17 2008 From: yar at comp.chem.msu.su (Yar Tikhiy) Date: Thu Mar 6 09:45:25 2008 Subject: cvs commit: src/usr.sbin/moused moused.c In-Reply-To: <200803052143.48208.jkim@FreeBSD.org> References: <200803060214.m262EjmG032525@repoman.freebsd.org> <53687.192.168.4.151.1204770665.squirrel@router> <200803052143.48208.jkim@FreeBSD.org> Message-ID: On Thu, Mar 6, 2008 at 5:43 AM, Jung-uk Kim wrote: > On Wednesday 05 March 2008 09:31 pm, Max Laier wrote: > > Am Do, 6.03.2008, 03:14, schrieb Jung-uk Kim: > > > jkim 2008-03-06 02:14:45 UTC > > > > > > FreeBSD src repository > > > > > > Modified files: > > > usr.sbin/moused moused.c > > > Log: > > > Cast time_t to long to mute tinderbox build failure. > > > > doesn't this defeat the purpose of having time_t in the first > > place? I'd rather cast to intmax_t and print with %j ... though > > it probably doesn't matter. > > Now the timestamp is uptime and the breakage was just debugging > output, it shouldn't really matter at all but I will do that soon. Yes it matters because it has to do with everyday coding practice. It is one of small yet important advances in C99 that we can now printf non-basic integral types _portably_[1] and _reliably_ through %j[du] and a cast to [u]intmax_t. [1] Well, almost portably as we need to guess the signedness of the type. -- Yar From peterjeremy at optushome.com.au Thu Mar 6 09:54:11 2008 From: peterjeremy at optushome.com.au (Peter Jeremy) Date: Thu Mar 6 09:54:18 2008 Subject: cvs commit: src/usr.sbin/moused moused.c In-Reply-To: <20080306092045.GA54220@freebsd.org> References: <200803060022.m260MHSx017085@repoman.freebsd.org> <20080306092045.GA54220@freebsd.org> Message-ID: <20080306095407.GS68971@server.vk2pj.dyndns.org> On Thu, Mar 06, 2008 at 10:20:45AM +0100, Roman Divacky wrote: >On Thu, Mar 06, 2008 at 12:22:17AM +0000, Jung-uk Kim wrote: >> Prefer clock_gettime(2) over gettimeofday(2) and use CLOCK_MONOTONIC_FAST. >> It is only used to track elapsed time and it does not have to be precise. > >what is the rationale behind this? what is the expected improvement? I am >not questioning the change I am just curious.. There's been a recent extensive thread about poor X11 performance and it turns out the the Xserver can issue multiple gettimeofday(2) calls every time it sees a mouse movement. gettimeofday(2) on FreeBSD is relatively expensive (the Project chose accuracy over speed) and clock_gettime(2) using CLOCK_MONOTONIC_FAST is a much cheaper alternative when accuracy isn't critical. The Xserver should be using clock_gettime(2) but there's a bug in its POSIX-compatibility selection. Applying a similar fix to moused is a logical extension. -- Peter Jeremy Please excuse any delays as the result of my ISP's inability to implement an MTA that is either RFC2821-compliant or matches their claimed behaviour. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080306/73b9e78e/attachment.pgp From rwatson at FreeBSD.org Thu Mar 6 10:10:44 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Thu Mar 6 10:10:53 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803061010.m26AAilh077521@repoman.freebsd.org> rwatson 2008-03-06 10:10:44 UTC FreeBSD src repository Modified files: sys/ddb db_input.c Log: When redrawing an input line, count backspaces to get to the beginning of the input field from the current cursor location, rather than the end of the input line, as the cursor may not be at the end of the line. Otherwise, we may overshoot, overwriting a bit of the previous line and failing to fully overwrite the current line. MFC after: 3 days PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.38 +1 -1 src/sys/ddb/db_input.c From rwatson at FreeBSD.org Thu Mar 6 10:29:25 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Thu Mar 6 10:29:32 2008 Subject: cvs commit: src/usr.sbin/moused moused.c In-Reply-To: <20080306095407.GS68971@server.vk2pj.dyndns.org> References: <200803060022.m260MHSx017085@repoman.freebsd.org> <20080306092045.GA54220@freebsd.org> <20080306095407.GS68971@server.vk2pj.dyndns.org> Message-ID: <20080306102751.J21309@fledge.watson.org> On Thu, 6 Mar 2008, Peter Jeremy wrote: > On Thu, Mar 06, 2008 at 10:20:45AM +0100, Roman Divacky wrote: >> On Thu, Mar 06, 2008 at 12:22:17AM +0000, Jung-uk Kim wrote: >>> Prefer clock_gettime(2) over gettimeofday(2) and use CLOCK_MONOTONIC_FAST. >>> It is only used to track elapsed time and it does not have to be >>> precise. >> >> what is the rationale behind this? what is the expected improvement? I am >> not questioning the change I am just curious.. > > There's been a recent extensive thread about poor X11 performance and it > turns out the the Xserver can issue multiple gettimeofday(2) calls every > time it sees a mouse movement. gettimeofday(2) on FreeBSD is relatively > expensive (the Project chose accuracy over speed) and clock_gettime(2) using > CLOCK_MONOTONIC_FAST is a much cheaper alternative when accuracy isn't > critical. > > The Xserver should be using clock_gettime(2) but there's a bug in its > POSIX-compatibility selection. Applying a similar fix to moused is a > logical extension. Many applications also assume that gettimeofday(2) returns a monotonically, and possibly consistently, increasing value. Sometimes this is fine (log timestamps) and other times this is not (time deltas used to calculate mouse acceleration, etc). Robert N M Watson Computer Laboratory University of Cambridge From mtm at FreeBSD.org Thu Mar 6 14:01:10 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Mar 6 14:01:17 2008 Subject: cvs commit: src/etc/defaults rc.conf Message-ID: <200803061401.m26E1Acf097471@repoman.freebsd.org> mtm 2008-03-06 14:01:10 UTC FreeBSD src repository Modified files: etc/defaults rc.conf Log: The rarpd(8) daemon must be instructed to start on all interfaces or a specific one. Instruct it to listen on all interfaces so that enabling it in rc.conf(5) works "out of the box." PR: conf/121406 Submited by: trasz MFC after: 1 week Revision Changes Path 1.328 +1 -1 src/etc/defaults/rc.conf From mtm at FreeBSD.org Thu Mar 6 14:10:57 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Mar 6 14:11:01 2008 Subject: cvs commit: src/usr.sbin/adduser adduser.8 adduser.sh Message-ID: <200803061410.m26EAuTQ098983@repoman.freebsd.org> mtm 2008-03-06 14:10:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/adduser adduser.8 adduser.sh Log: MFC: rev. 1.32 Add the -M command-line option, which will set home directory permissions. Works both in interactive or batch mode. This is a heavily modified version of the patch submitted in the PR. PR: bin/105060 Revision Changes Path 1.59.2.2 +5 -1 src/usr.sbin/adduser/adduser.8 1.30.2.2 +40 -1 src/usr.sbin/adduser/adduser.sh From mtm at FreeBSD.org Thu Mar 6 14:13:11 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Mar 6 14:13:17 2008 Subject: cvs commit: src/usr.sbin/adduser adduser.8 adduser.sh Message-ID: <200803061413.m26EDBHT099074@repoman.freebsd.org> mtm 2008-03-06 14:13:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/adduser adduser.8 adduser.sh Log: MFC: rev. 1.32 Add the -M command-line option, which will set home directory permissions. Works both in interactive or batch mode. This is a heavily modified version of the patch submitted in the PR. PR: bin/105060 Revision Changes Path 1.55.8.3 +5 -1 src/usr.sbin/adduser/adduser.8 1.26.2.5 +40 -1 src/usr.sbin/adduser/adduser.sh From mtm at FreeBSD.org Thu Mar 6 14:39:34 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Mar 6 14:39:47 2008 Subject: cvs commit: src/etc/rc.d mountcritlocal mountlate Message-ID: <200803061439.m26EdXP3000122@repoman.freebsd.org> mtm 2008-03-06 14:39:33 UTC FreeBSD src repository Modified files: etc/rc.d mountcritlocal mountlate Log: The check for errors from the mount command did not work as intended because another command (echo) is executed between the mount command and the check. Reported by: Sergey Baturov MFC after: 2 weeks Revision Changes Path 1.16 +4 -1 src/etc/rc.d/mountcritlocal 1.3 +4 -1 src/etc/rc.d/mountlate From ghelmer at FreeBSD.org Thu Mar 6 15:02:11 2008 From: ghelmer at FreeBSD.org (Guy Helmer) Date: Thu Mar 6 15:02:13 2008 Subject: cvs commit: src/usr.bin/uniq uniq.c Message-ID: <200803061502.m26F26wY005269@repoman.freebsd.org> ghelmer 2008-03-06 15:02:05 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/uniq uniq.c Log: MFH: 1.30 Fix truncation of lines at LINE_MAX characters by dynamically extending line buffers. PR: bin/76578 Revision Changes Path 1.29.2.1 +61 -15 src/usr.bin/uniq/uniq.c From ghelmer at FreeBSD.org Thu Mar 6 15:07:37 2008 From: ghelmer at FreeBSD.org (Guy Helmer) Date: Thu Mar 6 15:07:41 2008 Subject: cvs commit: src/usr.bin/uniq uniq.c Message-ID: <200803061507.m26F7bY1009627@repoman.freebsd.org> ghelmer 2008-03-06 15:07:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/uniq uniq.c Log: MFC: 1.30 Fix truncation of lines at LINE_MAX characters by dynamically extending line buffers. PR: bin/76578 Revision Changes Path 1.26.2.1 +61 -15 src/usr.bin/uniq/uniq.c From scf at FreeBSD.org Thu Mar 6 15:14:37 2008 From: scf at FreeBSD.org (Sean Farley) Date: Thu Mar 6 15:14:41 2008 Subject: cvs commit: src/lib/libc/stdlib getenv.c Message-ID: <200803061514.m26FEar3009851@repoman.freebsd.org> scf 2008-03-06 15:14:36 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/stdlib getenv.c Log: MFC: Replace the use of warnx() with direct output to stderr using _write(). This reduces the size of a statically-linked binary by approximately 100KB in a trivial "return (0)" test application. readelf -S was used to verify that the .text section was reduced and that using strlen() saved a few more bytes over using sizeof(). Since the section of code is only called when environ is corrupt (program bug), I went with fewer bytes over fewer cycles. I made minor edits to the submitted patch to make the output resemble warnx(). src/lib/libc/stdlib/getenv.c: rev 1.12 -> 1.13 Submitted by: kib bz Approved by: wes (mentor) Revision Changes Path 1.12.2.1 +36 -12 src/lib/libc/stdlib/getenv.c From marcel at FreeBSD.org Thu Mar 6 16:59:39 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Mar 6 16:59:47 2008 Subject: cvs commit: src/sys/powerpc/include pcpu.h Message-ID: <200803061659.m26Gxcgj024332@repoman.freebsd.org> marcel 2008-03-06 16:59:37 UTC FreeBSD src repository Modified files: sys/powerpc/include pcpu.h Log: Add a catch-all for PCPU_MD_FIELDS. While we expect this to be used in the kernel only (by virtue of checking for _KERNEL), ports like lsof (part of gtop) cheat. It sets _KERNEL, but does not set either AIM or E500. As such, PCPU_MD_FIELDS didn't get defined and the build broke. The catch-all is to define PCPU_MD_FIELDS with a dummy integer when at the end of line we ended up without a definition for it. Revision Changes Path 1.26 +7 -0 src/sys/powerpc/include/pcpu.h From marcel at FreeBSD.org Thu Mar 6 17:24:28 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Mar 6 17:24:35 2008 Subject: cvs commit: src/sys/sys pcpu.h Message-ID: <200803061724.m26HOSPO026401@repoman.freebsd.org> marcel 2008-03-06 17:24:28 UTC FreeBSD src repository Modified files: sys/sys pcpu.h Log: Move the PCPU_MD_FIELDS last in struct pcpu. While this header is private to the kernel, some ports define _KERNEL and include this header. While arguably this is wrong, it's also reality. By having the MD fields last, architectures that have CPU-specific variations of PCPU_MD_FIELDS will at least have the MI fields at a constant offset. Of course, having all MI fields first helps kernel debugging as well, so this is not a change without some benefits to us. This change does not result in an ABI breakage, because this header is not part of the ABI. Recompilation of lsof is required though :-) Revision Changes Path 1.26 +11 -1 src/sys/sys/pcpu.h From thompsa at FreeBSD.org Thu Mar 6 19:02:37 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Thu Mar 6 19:02:39 2008 Subject: cvs commit: src/sys/net if_gif.c Message-ID: <200803061902.m26J2bl8031933@repoman.freebsd.org> thompsa 2008-03-06 19:02:37 UTC FreeBSD src repository Modified files: sys/net if_gif.c Log: Improve EtherIP interaction with the bridge - Set M_BCAST|M_MCAST for incoming frames - Send the frame to a local interface if the bridge returns the mbuf Submitted by: Eugene Grosbein Tested by: Boris Kochergin Revision Changes Path 1.68 +30 -2 src/sys/net/if_gif.c From jeff at FreeBSD.org Thu Mar 6 20:11:25 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Thu Mar 6 20:11:30 2008 Subject: cvs commit: src/sys/kern kern_cpuset.c Message-ID: <200803062011.m26KBPIW036636@repoman.freebsd.org> jeff 2008-03-06 20:11:25 UTC FreeBSD src repository Modified files: sys/kern kern_cpuset.c Log: - Add a missing unlock to cpuset_setaffinity(CPU_LEVEL_CPUSET, CPU_WHICH_PID) Found by: gallatin Revision Changes Path 1.4 +1 -0 src/sys/kern/kern_cpuset.c From jhb at FreeBSD.org Thu Mar 6 21:42:49 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Mar 6 21:42:52 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h src/sys/dev/mii brgphy.c miidevs Message-ID: <200803062142.m26Lgn0i051581@repoman.freebsd.org> jhb 2008-03-06 21:42:48 UTC FreeBSD src repository Modified files: sys/dev/bge if_bge.c if_bgereg.h sys/dev/mii brgphy.c miidevs Log: Flesh out support for the BCM5722 by recognizing the phy on the 5722 and the specific ASIC revision. MFC after: 1 week Obtained from: OpenBSD (mii/phy bits) Revision Changes Path 1.203 +1 -0 src/sys/dev/bge/if_bge.c 1.76 +1 -0 src/sys/dev/bge/if_bgereg.h 1.73 +2 -1 src/sys/dev/mii/brgphy.c 1.51 +1 -0 src/sys/dev/mii/miidevs From jhb at FreeBSD.org Thu Mar 6 21:43:58 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Mar 6 21:44:03 2008 Subject: cvs commit: src/sys/boot/i386/pxeldr pxeldr.S Message-ID: <200803062143.m26LhvaG051629@repoman.freebsd.org> jhb 2008-03-06 21:43:56 UTC FreeBSD src repository Modified files: sys/boot/i386/pxeldr pxeldr.S Log: In the PROBE_KEYBOARD case, always enable multiple consoles and set the serial console as the primary console if the keyboard probe fails. MFC after: 1 week Revision Changes Path 1.16 +4 -2 src/sys/boot/i386/pxeldr/pxeldr.S From jhb at FreeBSD.org Thu Mar 6 21:48:34 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Mar 6 21:48:37 2008 Subject: cvs commit: src/sys/dev/bge if_bgereg.h Message-ID: <200803062148.m26LmYat051857@repoman.freebsd.org> jhb 2008-03-06 21:48:34 UTC FreeBSD src repository Modified files: sys/dev/bge if_bgereg.h Log: Style tweak to the 5722 chipid to match the rest of this file. Revision Changes Path 1.77 +1 -1 src/sys/dev/bge/if_bgereg.h From piso at FreeBSD.org Thu Mar 6 21:50:41 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Thu Mar 6 21:50:45 2008 Subject: cvs commit: src/sys/netinet/libalias alias.c alias_db.c alias_ftp.c alias_irc.c alias_local.h alias_proxy.c alias_smedia.c Message-ID: <200803062150.m26Lofth051979@repoman.freebsd.org> piso 2008-03-06 21:50:41 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias.c alias_db.c alias_ftp.c alias_irc.c alias_local.h alias_proxy.c alias_smedia.c Log: MFP4: restrict the utilization of direct pointers to the content of ip packet. These modifications are functionally nop()s thus can be merged with no side effects. Revision Changes Path 1.59 +29 -30 src/sys/netinet/libalias/alias.c 1.72 +11 -23 src/sys/netinet/libalias/alias_db.c 1.30 +4 -2 src/sys/netinet/libalias/alias_ftp.c 1.24 +4 -2 src/sys/netinet/libalias/alias_irc.c 1.35 +7 -5 src/sys/netinet/libalias/alias_local.h 1.32 +8 -13 src/sys/netinet/libalias/alias_proxy.c 1.18 +4 -2 src/sys/netinet/libalias/alias_smedia.c From cognet at FreeBSD.org Thu Mar 6 21:59:48 2008 From: cognet at FreeBSD.org (Olivier Houchard) Date: Thu Mar 6 21:59:57 2008 Subject: cvs commit: src/sys/arm/arm pmap.c src/sys/arm/include pmap.h Message-ID: <200803062159.m26LxlUc052363@repoman.freebsd.org> cognet 2008-03-06 21:59:47 UTC FreeBSD src repository Modified files: sys/arm/arm pmap.c sys/arm/include pmap.h Log: Remove unused pv_list_count from the vm_page, and pm_count from the struct pmap. Submitted by: Mark Tinguely Revision Changes Path 1.99 +0 -4 src/sys/arm/arm/pmap.c 1.29 +0 -2 src/sys/arm/include/pmap.h From cognet at FreeBSD.org Thu Mar 6 22:27:36 2008 From: cognet at FreeBSD.org (Olivier Houchard) Date: Thu Mar 6 22:27:37 2008 Subject: cvs commit: src/sys/arm/arm uio_machdep.c Message-ID: <200803062227.m26MRZVO055385@repoman.freebsd.org> cognet 2008-03-06 22:27:35 UTC FreeBSD src repository Modified files: sys/arm/arm uio_machdep.c Log: MFi386: revision 1.6 date: 2004/08/21 18:50:34; author: alc; state: Exp; lines: +3 -1 Properly free the temporary sf_buf in uiomove_fromphys() if a copyin or copyout fails. Obtained from: DragonFlyBSD Spotted out by: Mark Tinguely MFC After: 3 days Revision Changes Path 1.6 +3 -1 src/sys/arm/arm/uio_machdep.c From csjp at FreeBSD.org Thu Mar 6 22:57:04 2008 From: csjp at FreeBSD.org (Christian S.J. Peron) Date: Thu Mar 6 22:57:06 2008 Subject: cvs commit: src/sys/security/audit audit_syscalls.c Message-ID: <200803062257.m26Mv32F056409@repoman.freebsd.org> csjp 2008-03-06 22:57:03 UTC FreeBSD src repository Modified files: sys/security/audit audit_syscalls.c Log: Change auditon(2) so that if somebody supplies an invalid command, it returns EINVAL. Right now we return 0 or success for invalid commands, which could be quite problematic in certain conditions. MFC after: 1 week Discussed with: rwatson Revision Changes Path 1.25 +3 -0 src/sys/security/audit/audit_syscalls.c From brde at optusnet.com.au Thu Mar 6 23:05:16 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Thu Mar 6 23:05:23 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF9586.70707@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> <47CF9586.70707@freebsd.org> Message-ID: <20080307085031.P10724@delplex.bde.org> Eek, my mailbox filled up overnight. Trying to reply to only some details in LIFO order... On Wed, 5 Mar 2008, Colin Percival wrote: > David Schultz wrote: >> On Wed, Mar 05, 2008, Colin Percival wrote: >>> Setting the i387 FPU to 53-bit precision gives standards-compliant >>> behaviour whether people are using "double" or "long double". >> >> Not quite. > > How is it not standards-compliant? > >>> Yes and no. Double rounding isn't allowed; >> >> Yes, it is. > > No, it isn't. The following code Yes it is :-). Extra precision for intermediate values implies double rounding for the final results, and extra precision for intermediate values is a fundamental part of C (it used to be only that float expressions were (de-facto standard required to be) evaluated in double precision, but now it affects double expressions too, and C99 has vast complications to support i387'-ish extra precision. > double x, y, z; > > x = 1.0; > y = 0x1.0p-53 + 0x1.0p-75; > z = x + y; > z = z - 1.0; > if (z == 0.0) > printf("BUGGY MATH\n"); > else > printf("IEEE754-COMPLIANT MATH\n"); > > should never output "BUGGY MATH", since x + y should round up to > 1 + 2^(-52); but if you set the i387 to use extended precision, it > gets the wrong answer. In C, with FLT_EVAL_METHOD = 2 (evaluate all FP expressions in the range and precision of the long double type) and long doubles actually longer than doubles, this always gives double rounding: x + y is evaluated (and presumably rounded) in long double precision z = x + y rounds to double precision (but with gcc-i386, it doesn't round) (-O0 happens to work to fix this) (the -ffloat-store hack happens to work to fix this) z is now 1.0 except with gcc-i386 -O -fno-float-store when it is 1.0L + 0x1.0p-53L. z - 1.0 is evaluated in long double precision z = z - 1.0 rounds to double precision (except with gcc-i386...) z is now 0.0, except with gcc-i386 -O -fno-float-store it is 0x1.0p-53L if (z == 0.0) fails to detect the bug for gcc-i386 -O -fno-float-store since z is still long double and nonzero Another advantage of using double_t is that you can control when the double rounding occurs and not have to understand so many sub-cases from the compiler bugs. The result is then always 0x1.0p-53L (if double_t is long double, etc.), and if you want a double result then you can round as the last step. The result is then 0x1.0p-53. You don't get the double precision result of 0x1.0p-52, but you get a more accurate result. Extra precision is of course intended to deliver a more accurate result in this way. This assumes 64-bit precision. With 113-bit precision for long doubles, all intermediate values and the result would be exact; the result would be y. It seems to be impossible to choose y to demonstrate the problem for 113-bit long doubles. We see a similar phenomenon for floats vs doubles -- since the extra precision provided by doubles is actually (more than) double, double rounding never (?) occurs. Extended precision should have had more than 64 bits. I think it is mainly the result of Intel's transistor budget being limited 30 years ago. >>>> The downside is that this breaks long doubles. >>> Except that it doesn't. Using 53-bit precision for long doubles is >>> a perfectly valid and standards-compliant option. >> >> It's impossible to implement efficient library functions that >> produce correct long double results > > You could have stopped this sentence here -- for all practical purposes, > correctly rounded trigonometric functions are not feasible. Nah, it's quite feasible, especially for transcendental functions since transcendental numbers in general can be approximated more closely by rationals than algebraic numbers. The evaluation just needs to be done in extra precision, and at least for trig functions, "in general" happens in practice so relatively few cases require significant extra precision. With extra precision in hardware, all except the relatively few cases can even be implemented efficiently. >> when the FPU is set to 64-bit >> mode and avoid double rounding and related problems when the FPU >> is set to 53-bit mode. > > Fortunately, library functions aren't required to have any particular > error bounds. It's always nice if they are correct to within 0.6ulp > instead of 0.8ulp or suchlike -- but nobody should be proving code > correctness based on assumed properties of library transcendental > functions. People should, and do, prove code correct based on the > assumption that double precision arithmetic behaves like double precision > arithmetic, however (myself included). We require error bounds of < 1 ulp and aim for < 0.5[0]1 ulps. Error bounds are required for everything. E.g., libm uses expm1() to implement hyperbolic functions and needs good error bounds for expm1() and a nontrivial error analysis to justify using it (since the error can blow up and would blow up if exp(x)-1 were used instead of expm1(x)), yet it still has an error of > 1 ulp since expm1() and other things aren't accurate enough. Bruce From brde at optusnet.com.au Thu Mar 6 23:24:31 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Thu Mar 6 23:24:39 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306063746.GC48339@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <20080306062301.GX68971@server.vk2pj.dyndns.org> <20080306063746.GC48339@zim.MIT.EDU> Message-ID: <20080307100713.I10724@delplex.bde.org> On Thu, 6 Mar 2008, David Schultz wrote: > On Thu, Mar 06, 2008, Peter Jeremy wrote: >> On Wed, Mar 05, 2008 at 10:32:46PM -0500, David Schultz wrote: >>> gcc doesn't get any of this right. FreeBSD/i386 fixes the problem >>> for /doubles/ by setting the i387 to use 53-bit precision. >> Note that FreeBSD/amd64 does not do this when running i386 binaries >> so you can get discrepancies between running the same program natively >> vs running it in compatibility mode. On my list of things to fix. > Good point. I don't think i386 ever got fixed to do the > appropriate mode change for Linux binaries, either, BTW. Actually, someone named das fixed this on 2005/02/06 in rev.1.136 of i386/linux/linux_sysvec.c, but the fix is buggy -- it accesses the i387 directly and thus causes npxdna() traps which should cause panics since direct access to the i387 is not permitted (it should use npxinit(), which was originally for the corresponding initialization in setregs() but is no longer used there since the normal initialization is now virtual). Bruce From delphij at FreeBSD.org Fri Mar 7 00:01:19 2008 From: delphij at FreeBSD.org (Xin LI) Date: Fri Mar 7 00:01:22 2008 Subject: cvs commit: src/usr.bin/chat chat.c Message-ID: <200803070001.m2701Jdm059853@repoman.freebsd.org> delphij 2008-03-07 00:01:19 UTC FreeBSD src repository Modified files: usr.bin/chat chat.c Log: Merge revisions 1.10 and 1.11 from DragonFly: - Use real getopt() handling instead of the hand-rolled and IOCCC-worthy "Micro getopt()" macros, plus clean up to the option handling code: * Sort the options in the switch statement; * Plug piddling memory leaks when processing repeated options by freeing strings before allocating them for a second time; * Die with a fatal error if the requested report file cannot be opened for appending; * Don't call init() before usage() (to prevent the usage message being mangled by changes to the terminal settings;) - Clean up the usage message, both in usage() and in the main program comment, both stylistically (sort and combine options) and for accuracy (following the manual page, make note of the -s and -S flags, and use the term 'send' instead of 'say' to reduce confusion (SAY is the name of a command for output to the user, not the connection.)) Obtained from: DragonFly Revision Changes Path 1.22 +61 -68 src/usr.bin/chat/chat.c From jkim at FreeBSD.org Fri Mar 7 00:01:24 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Fri Mar 7 00:01:26 2008 Subject: cvs commit: src/usr.sbin/moused moused.c Message-ID: <200803070001.m2701Orc059880@repoman.freebsd.org> jkim 2008-03-07 00:01:24 UTC FreeBSD src repository Modified files: usr.sbin/moused moused.c Log: Use intmax_t instead of long when casting time_t. Suggested by: mlaier Tested with: make universe Revision Changes Path 1.83 +4 -3 src/usr.sbin/moused/moused.c From cperciva at freebsd.org Fri Mar 7 00:18:27 2008 From: cperciva at freebsd.org (Colin Percival) Date: Fri Mar 7 00:18:31 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080307085031.P10724@delplex.bde.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> <47CF9586.70707@freebsd.org> <20080307085031.P10724@delplex.bde.org> Message-ID: <47D089A4.4060208@freebsd.org> Bruce Evans wrote: > On Wed, 5 Mar 2008, Colin Percival wrote: >> David Schultz wrote: >>> On Wed, Mar 05, 2008, Colin Percival wrote: >>>> Yes and no. Double rounding isn't allowed; >>> >>> Yes, it is. >> >> No, it isn't. The following code > > Yes it is :-). Extra precision for intermediate values implies double > rounding for the final results, and extra precision for intermediate > values is a fundamental part of C (it used to be only that float > expressions were (de-facto standard required to be) evaluated in double > precision, but now it affects double expressions too, and C99 has vast > complications to support i387'-ish extra precision. Let me clarify: Double rounding isn't allowed by IEEE754. To quote from the October 5th draft of IEEE 754R (which I happen to have in front of me right now): 5.1 Overview All conforming implementations of this standard shall provide the operations listed in this clause for all supported floating-point formats available for arithmetic. Each of the computational operations that return a numeric result specified by this standard shall be performed as if it first produced an intermediate result correct to infinite precision and with unbounded range, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ and then rounded that intermediate result to fit in the destination?s format ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ This is clarified further in clause 10: 10.1 Expression evaluation rules Clause 5 of this standard specifies the result of a single arithmetic operation going to an explicit destination. Every operation has an explicit or implicit destination. One rounding occurs to fit the exact result into a ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ destination format. That result is reproducible in that the same operation ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ applied to the same operands under the same attributes produces the same ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ result on all conforming implementations in all languages. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Programming language standards might define syntax for expressions that combine one or more operations of this standard, producing a result to fit an explicit or implicit final destination. When a variable with a declared format ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ is a final destination, as in format conversion to a variable, that declared ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ format of that variable governs its rounding. The format of an implicit ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ destination, or of an explicit destination without a declared format, is defined by language standard expression evaluation rules. In other words, in the code double x, y, z; z = x * y + 1.0; it is perfectly legal for (x * y) to be computed with extra precision; but in the code double x, y, z; z = x * y; z = z - 1.0; both the multiplication and the subtraction must be rounded *once* to double precision. > In C, with FLT_EVAL_METHOD = 2 (evaluate all FP expressions in the range > and precision of the long double type) [...] As someone wrote earlier, the authors of C99 were a bit confused. I can only assume that they intended to be IEEE754-compliant and the rules concerning FLT_EVAL_METHOD apply to *implicit* destinations only. >>> It's impossible to implement efficient library functions that >>> produce correct long double results >> >> You could have stopped this sentence here -- for all practical purposes, >> correctly rounded trigonometric functions are not feasible. > > Nah, it's quite feasible, especially for transcendental functions since > transcendental numbers in general can be approximated more closely by > rationals than algebraic numbers. In fact, one of the common ways to prove that a real number is transcendental is to show that it can be approximated more closely by rationals than any algebraic function can be. > The evaluation just needs to be > done in extra precision How much extra precision do you need? If the exact value of a transcendental function is extremely close to halfway between two consecutive floating-point values, how do you decide which way to round it? Computing transcendental functions to within (0.5+x)ulp for small positive values of x is certainly feasible, but computing provably correctly rounded transcendental functions is very very hard. >> Fortunately, library functions aren't required to have any particular >> error bounds. > > We require error bounds of < 1 ulp and aim for < 0.5[0]1 ulps. What standard states that those bounds are required (or can be relied upon when proving that code is correct)? Intel FPUs over the years have been variously documented to compute transcendental functions to within 1.0, 2.5, or 3.5 ulp -- are you saying that the 8087 doesn't comply with the standard which it inspired? Colin Percival From brde at optusnet.com.au Fri Mar 7 01:21:52 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Fri Mar 7 01:21:57 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF8161.1060202@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305230954.X55190@odysseus.silby.com> <47CF8161.1060202@freebsd.org> Message-ID: <20080307104437.C11033@delplex.bde.org> On Wed, 5 Mar 2008, Colin Percival wrote: > Mike Silbersack wrote: >> On Wed, 5 Mar 2008, Bruce Evans wrote: >>> Change float_t and double_t to long double on i386. All floating point >> >> 1) Does this really change every double to a long double in anything >> compiled? > > No, it changes double_t (which is not the same as double). > >> 2) How does this affect ABI compatibility? > > This breaks ABI compatibility (when people use float_t or double_t). Only if float_t or double_t is actually used in an ABI. Such use is dubious since these types are like the int_fastN_t types they are the most efficient types that are at least as wide as float, double and intN_t, respectively. They aren't very suitable for anything except intermediate values in registers. I know of one useful use for them in ABIs: to work around the following bug in the C standard (C99 at least): a function returning double is permitted to return extra precision, and this is sometimes useful, but C perversely requires extra precision to be lost on return iff the function is implemented using double_t so as to ensure that extra precision is not lost internally. E.g.: double xp1(double x) { return x + 1.0; } This is permitted to evaluate x + 1.0 in extra precision and then is required (?) (at least permitted) to return the extra precision. double xp1(double x) { double_t tmp1 = x, tmp2 = 1.0; return tmp1 + tmp2; } If double_t has more precision than double, then (1) this evaluates x + 1.0 in extra precision. double_t is only used to emphasize this. In a more practical example. There would be many more temporary intermediate variables and we should use double_t for them all to prevent loss of precision on assignment. (2) This is required to lose the extra precision on return. To avoid the loss of precision on return, the simplest method is to change the ABI to return double_t instead of double. Since we really want to return the extra precision, this is not wrong. However, it may be prevented by API and ABI considerations. Note that on i386, all functions that are declared to return float or double actually return a long double, so declaring them as actually returning a long double is only an API change, but it will significantly affect the callers due to this alone (callers will actually know that they are getting a long double and may generate extra code to either discard or keep the extra precision which may or may not be present). Why did no one complain when I fixed float_t on amd64 and some other non-i386 arches? :-) This changed the ABI as a side effect. >> 3) How does this change the operation of various programs in the ports >> tree that use floating point, such as mplayer, mpg123, etc. Will this >> cause different behavior when these apps are used on FreeBSD vs other >> operating systems? > > Most code in the ports tree will be using double instead of double_t, and > will consequently not be affected. Code which uses double_t will be slower > due to the increased cost of loads and stores, and may produce different > output if it changes the i387 precision flags. I hope nothing actually uses float_t or double_t. They are esoteric and are still too broken in FreeBSD to trust. I'm changing them mainly because I want to be able to trust them for use in libm, without using extensive ifdefs which would amount to replacing them by non-broken versions for libm's internal use only. > At the moment, FreeBSD behaves the same way as Microsoft Windows and C99, > and differently from Linux: Linux sets the i387 by default to extended > precision, Does Windows and its compilers still default to non-extended precision and thus less than null support for long doubles? 20 years ago, this seemed to be normal for DOS compilers (though I never tested it on a Microsoft one), and I was happy to keep it in 386BSD in 1992 and FreeBSD later. In 386BSD, it was apparently inherited from ibcs2. Anyway, gcc didn't really support long doubles until 1993 or 1994. IIRC, Linux started in 1991 with the default of extended precision, and I was responsible for getting this changed to non-extended precision. Then when gcc started supporting long doubles, Linux changed the default back to extended precision, as is necessary but not sufficient for long doubles to work. Microsoft would have been more constrained by backwards compatibilty but should have changed by now. Maybe it is a compiler option (default = old in Windows?). > which has the result of decreasing the average rounding error > while increasing the maximum rounding error (due to double rounding when > values which were internally rounded to extended precision are rounded > to double precision later) and sometimes breaking code entirely (when > the properties of floating-point rounding are used deliberately, e.g., > to rapidly round a floating-point value to the nearest integer). I've only found double rounding to be a minor problem. It gives an error of at most 0.5 ulps, which isn't usually a problem. Breakage of the properties of floating point is mostly due to compiler bugs (assignments and casts don't work). However, fixing assignment requires large pessimizations. They used to be not so large, but now gcc's optimizer is reasonably good they are very large. So I now think that it is a bug in the C standard to require assignments and casts to discard extra precision. This should be implementation-defined, with some other standard method for discarding extra precision. One of the first things I need a non-broken float_t for is rapid rounding to the nearest [fraction of] an integer, without doing extra work to discard extra precision. E.g.: to round a non-huge float to the nearest integer, the fastest method is often: float f, result; result = f + 0x1.8pN - 0x1.8pN; where N is: FLT_MANT_DIG - 1 if float_t is float DBL_MANT_DIG - 1 if float_t is double LDBL_MANT_DIG - 1 if float_t is long double This is painful to configure if float_t is unknown or wrong. (It is still moderately painful with float_t only a type since it is hard to ifdef on a type. I would like to be able to write the N as FLT_T_MANT_DIG-1 and not have to use magic to build a literal out of it.) This depends on float_t being either float, or on float FLT_MANT_DIG being much smaller than FLT_T_MANT_DIG, else double rounding may be a problem. This method is already used, with double instead of float, for arg reduction in trig functions. There double rounding can occur but isn't a problem. It currently has N hard-coded as DBL_MANT_DIG-1. This assumes that double_t is never really long double (it still works when with my change to declare double_t as long double on i386, since long double isn't really long double -- N = DBL_MANT_DIG-1 still works, but N = LDBL_MANT_DIG-1 wouldn't work). > I've answered queries from some mathematicians and scientists who were > confused as to why they were seeing higher rounding errors on FreeBSD and > Windows than they were on Linux; but when I've explained the behavioural > differences to numerical analysts, the universal reaction I've received > has been "dear god, Linux does WHAT?" -- oddly enough, mathematicians like > to think that proving that their code is correct will be enough to ensure > that it will produce correct output. Most code isn't proved to be correct. It's impossible for numeric code written and proved by non-numerical-analysts and impractical for large code written by anyone. Extended precision is supposed to reduce the risks from this, and it seems to help in most cases. Unfortunately, extended precision isn't implemented in SSE, so even Linux on i386's now doesn't have it when the i386's are in amd64 mode, so extended precision no longer occurs automatically, and using it (by using long doubles explicitly) costs more than ever since it lives in a different register set. Bruce From brde at optusnet.com.au Fri Mar 7 01:39:53 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Fri Mar 7 01:39:58 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080305230954.X55190@odysseus.silby.com> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305230954.X55190@odysseus.silby.com> Message-ID: <20080307122252.Y11033@delplex.bde.org> On Wed, 5 Mar 2008, Mike Silbersack wrote: > On Wed, 5 Mar 2008, Bruce Evans wrote: >> Change float_t and double_t to long double on i386. All floating point > > For those of us who are not floating point experts, can you explain a few > things? Other points replied to separately. > As I've said in the past, I'd really, really, really like to see regression > tests for any change to the floating point functions. The types of changes > you've been making are not easy to verify just by looking at diffs. I run local regression tests of 4 billon to 64 billion cases per function or 1.3 trillion cases for 125 functions a 36-hour run on a 2.2HGz UP system. These are not well organized enough for commit. You will have to trust that they are done before commit (or after on some other machines) :-). I mostly use semi-exhaustive (exhaustive for 1-arg float precision functions) checks on machine-generated args. This seems to find problems more routinely than smarter tests, up to at least double precision. das@ committed some smarter tests. OTOH, I barely tested the changes to float_t and double_t. These types are so rarely used that they are never used in /usr/src, at least in my old src tree, except for my uncommitted changes in libm parts. Bruce From das at FreeBSD.ORG Fri Mar 7 02:22:54 2008 From: das at FreeBSD.ORG (David Schultz) Date: Fri Mar 7 02:22:59 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47D089A4.4060208@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> <47CF9586.70707@freebsd.org> <20080307085031.P10724@delplex.bde.org> <47D089A4.4060208@freebsd.org> Message-ID: <20080307022416.GA53988@zim.MIT.EDU> On Thu, Mar 06, 2008, Colin Percival wrote: > explicit or implicit final destination. When a variable with a declared format > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > is a final destination, as in format conversion to a variable, that declared > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > format of that variable governs its rounding. The format of an implicit > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > destination, or of an explicit destination without a declared format, is > defined by language standard expression evaluation rules. I think you missed the point about implicit destinations. See the part I quoted in the last email I sent you. The point is that in an expression like 'z = x * y', there are two operations: register <-- x * y z <-- register The register is an implicit destination. IEEE 754R explains this in more detail. All that is guaranteed in this case is that the wider register value is never substituted for z in subsequent operations. (This is part of what gcc gets wrong.) > > In C, with FLT_EVAL_METHOD = 2 (evaluate all FP expressions in the range > > and precision of the long double type) [...] > > As someone wrote earlier, the authors of C99 were a bit confused. I can only > assume that they intended to be IEEE754-compliant and the rules concerning > FLT_EVAL_METHOD apply to *implicit* destinations only. No, they got that part right. They got some of the stuff with function arguments and return values wrong in some pretty special and bizarre ways. Part of this is trying to cope with existing calling conventions, and part of it is just mistakes. > > We require error bounds of < 1 ulp and aim for < 0.5[0]1 ulps. > > What standard states that those bounds are required (or can be relied upon > when proving that code is correct)? Intel FPUs over the years have been > variously documented to compute transcendental functions to within 1.0, 2.5, > or 3.5 ulp -- are you saying that the 8087 doesn't comply with the standard > which it inspired? There's no standard requirements, but a max error < 0.50x ulps implies correct rounding nearly all of the time, and that's what libm attempts to provide. This is about the best you can do without resorting to multiprecision arithmetic in the hard cases, so it seems like a good tradeoff between accuracy and ease of implementation. When substantially faster and slightly less accurate MD implementations are available, that changes the tradeoff. From cperciva at freebsd.org Fri Mar 7 02:52:07 2008 From: cperciva at freebsd.org (Colin Percival) Date: Fri Mar 7 02:52:11 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080307022416.GA53988@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> <47CF9586.70707@freebsd.org> <20080307085031.P10724@delplex.bde.org> <47D089A4.4060208@freebsd.org> <20080307022416.GA53988@zim.MIT.EDU> Message-ID: <47D0ADCE.2070405@freebsd.org> David Schultz wrote: > On Thu, Mar 06, 2008, Colin Percival wrote: >> explicit or implicit final destination. When a variable with a declared format >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> is a final destination, as in format conversion to a variable, that declared >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> format of that variable governs its rounding. The format of an implicit >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> destination, or of an explicit destination without a declared format, is >> defined by language standard expression evaluation rules. > > I think you missed the point about implicit destinations. No. Another quote: 2.2.14 destination: The location for the result of an operation upon one or more operands. A destination might be either explicitly designated by the user or implicitly supplied by the system (for example, intermediate results in subexpressions or arguments for procedures). [...] Implicit destinations are things like the (x * y) in "x * y - 1.0" or in "sqrt(x * y)". > The register is an implicit destination. IEEE 754R explains this > in more detail. Really? I can't find any such statements... quite the contrary. > All that is guaranteed in this case is that the > wider register value is never substituted for z in subsequent > operations. (This is part of what gcc gets wrong.) If IEEE754R intended that "z = x + y" with x, y, and z all doubles did not guarantee reproducible results across compliant platforms, why does it say exactly the opposite? >> What standard states that those bounds are required (or can be relied upon >> when proving that code is correct)? [...] > > There's no standard requirements, but a max error < 0.50x ulps > implies correct rounding nearly all of the time, and that's what > libm attempts to provide. This is about the best you can do > without resorting to multiprecision arithmetic in the hard cases, > so it seems like a good tradeoff between accuracy and ease of > implementation. When substantially faster and slightly less > accurate MD implementations are available, that changes the > tradeoff. Sure. As I said before, more accurate transcendental functions are always nice to have, but they are not *required* by any standard. Correctly-rounded arithmetic operations *are* required. Colin Percival From brde at optusnet.com.au Fri Mar 7 02:57:16 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Fri Mar 7 02:57:20 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080306021222.GA46783@zim.MIT.EDU> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> Message-ID: <20080307125559.M11320@delplex.bde.org> On Wed, 5 Mar 2008, David Schultz wrote: > On Wed, Mar 05, 2008, Colin Percival wrote: >> Bruce Evans wrote: >>> Modified files: >>> sys/i386/include _types.h >>> Log: >>> Change float_t and double_t to long double on i386. >> >> Doesn't this have a rather severe performance impact on any code which >> uses double_t? > > Yes, if the code is compiled with -msse2. Otherwise, doubles get > fed to the i387 anyway. Not even then. Doubles get fed to the i387 anyway in many cases (in most cases I think). One exception is double to int conversion -- this uses SSE2 so as to avoid a slow mode switch (cvtsd2si always rounds towards 0, while the default i386 mode rounds to nearest; otherwise it would be the SSE code that is slower). Adds of doubles normally go through the i387 (I've never seen an exception to this). > Technically Bruce's change isn't wrong without changing the > default precision, it's just rather pointless. It isn't pointless, since the default is only the default. Changing it using fpsetprec() has been supported in FreeBSD since 1993, and before that it required 2 lines of asm. My change gives double_t a chance of working right when the default is changed at runtime. While such changes are not really supported, there is no need to force them to be broken. > (Why tell programs > to store variables in a wider format if you're just going to > evaluate them in a narrower format anyway?) After fpsetprec(FP_PE): (1) the evaluation is not in a narrower format. (2) I use double_t when I hope that the variable will stay in a register and not cost a store (3) when I store to a double, it is obviously narrower on i386, and the compiler is tricked into doing the store correctly without using something like STRICT_ASSIGN(). STRICT_ASSIGN() uses a volatile hack which makes it potentially slower than an ordinary assignment, and currently doesn't support this mode. Irrespective of fpsetprec(FP_PE): (1)' - (3)' for float_t vs float precision. float_t just needs to be double for this. Using long double costs little and is technically half required even with the default precision, since the range is almost that of a long double. (The different range affects at least overflow exceptions.) (3)'' I can write code that is more portable and don't need to pessimize all uses of STRICT_ALIGN() on doubles by supporting full long double precision everywhere. Code that supports extra precision (either by a runtime switch or a change in the default) can use double_t and code that doesn't care or doesn't understand the complications can keep using plain double. > So I guess I agree > that if we're going to go down this path, we ought to just bite > the bullet and change npx.h and contrib/gcc/config/i386/freebsd.h > to use 64-bit precision by default on i386. I hope to do that in a non-revolutionary way. Bruce From das at FreeBSD.ORG Fri Mar 7 03:41:16 2008 From: das at FreeBSD.ORG (David Schultz) Date: Fri Mar 7 03:41:22 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47D0ADCE.2070405@freebsd.org> References: <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> <47CF9586.70707@freebsd.org> <20080307085031.P10724@delplex.bde.org> <47D089A4.4060208@freebsd.org> <20080307022416.GA53988@zim.MIT.EDU> <47D0ADCE.2070405@freebsd.org> Message-ID: <20080307034239.GA54694@zim.MIT.EDU> On Thu, Mar 06, 2008, Colin Percival wrote: > David Schultz wrote: > > On Thu, Mar 06, 2008, Colin Percival wrote: > >> explicit or implicit final destination. When a variable with a declared format > >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > >> is a final destination, as in format conversion to a variable, that declared > >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > >> format of that variable governs its rounding. The format of an implicit > >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > >> destination, or of an explicit destination without a declared format, is > >> defined by language standard expression evaluation rules. > > > > I think you missed the point about implicit destinations. > > No. Another quote: > 2.2.14 destination: The location for the result of an operation upon one or > more operands. A destination might be either explicitly designated by the user > or implicitly supplied by the system (for example, intermediate results in > subexpressions or arguments for procedures). [...] > > Implicit destinations are things like the (x * y) in "x * y - 1.0" or in > "sqrt(x * y)". > > > The register is an implicit destination. IEEE 754R explains this > > in more detail. > > Really? I can't find any such statements... quite the contrary. When reading the above quotation, you need to keep in mind that in their terminology, 'z = x + y' is an expression and 'x + y' is a subexpression. What you're claiming doesn't even make sense. With your rules, the 'x * y' in 'k = x * y - 1.0' could be evaluated with extra precision, the '- 1.0' would be forced to machine precision, and you'd still have double rounding on the multiply. This really *isn't* what they meant. Look, please have a careful look at section 10.2. This is a really busy week for me, and I don't have the inclination to continue this discussion if you keep sniping. From brde at optusnet.com.au Fri Mar 7 04:26:40 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Fri Mar 7 04:26:46 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47CF5AE8.2090101@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <47CE8396.6020803@freebsd.org> <20080306095645.V7605@delplex.bde.org> <47CF5AE8.2090101@freebsd.org> Message-ID: <20080307135800.V11489@delplex.bde.org> On Wed, 5 Mar 2008, Colin Percival wrote: > Bruce Evans wrote: >> On Wed, 5 Mar 2008, Colin Percival wrote: >>> Bruce Evans wrote: >>>> Change float_t and double_t to long double on i386. >>> >>> Doesn't this have a rather severe performance impact on any code which >>> uses double_t? >> >> No. As mentioned in the commit message, this has no performance effect >> except in cases where it avoids compiler bugs. [...] if you use long double >> for memory variables then you get a severe performance impact and some >> space loss for the load instruction, since loading long doubles is >> much slower than loading doubles (about 4 times slower on Athlons). > > Either I'm misunderstanding something, or you seem to be disagreeing with > yourself here... if I have the following code > > double_t foo, bar, foobar; > foobar = foo + bar; > > then prior to this change the processor loads and stores doubles, while > after this change the processor loads and stores long doubles, with the > associated performance penalty. Low quality code might do that :-). Hmm, I thought that these types were intended to be used to avoid loss of precision in intermediate values, and thus should only be used in code that wants the extra precision at a possible cost in efficiency, but C99 (n869.txt draft) doesn't say exactly this, and a footnote (which is not part of the standard) says that they are trhe most efficient types that are wider than the basic types. It specifies them fully if FLT_EVAL_METHOD is 0, 1 or 2, but weird machines like i386 with 53-bit default precision plus compiler bugs must define FLT_EVAL_METHOD as -1. Then these types are implementation-defined and are not specifically required to be wider than the evaluation method types. I think they should be specifically required to be the most efficient types wider than the evaluation method types. Back to performance penalties... It is extremely unclear what are the most efficient types. Your example is too simple for there to be any stores if double_t is long double: > double_t foo, bar, foobar; > foobar = foo + bar; o Efficiency of double_t for foo and bar depends on where they came from. If they are long doubles in memory initially here, then loading them is slow. If they are in registers initially here, then there is no load to worry about and their type doesn't matter from now on, but their type may have affected previous loads and stores like subsequent use of foobar will (see below). o foo + bar is evaluated in registers (except i386 also allows adds from a value in memory to a value in a register if the memory type is not long double). It essentially has type long double, perhaps rounded to 53-bit or 24-bit precision. o Assignment to foobar requires a conversion from long double to double_t. If double_t is long double, then this conversion is null and foobar normally stays in the register. Otherwise, the conversion is not null, but it won't be done here since all the variables have type double_t and the compiler doesn't really know the type of result of the addition (it thinks that double + double gives double). A C compiler would know that the result is long double and always do the conversion if double_t is only double. o Here is an example of why a conversion is necessary on i386 unless all types are long double: double x = DBL_MAX, y = DBL_MAX, z; z = DBL_MAX + DBL_MAX; // should be +Inf/FE_OVERFLOW but is 2.0L*DBL_MAX z = z - DBL_MAX; // should be +Inf/FE_OVERFLOW but is 1.0L*DBL_MAX The extra range of a long double gives this. 53-bit precision doesn't affect it, and a conversion on assignment is still strictly needed to give the correct results for the intermediate and final z. z = DBL_MAX + DBL_MAX - DBL_MAX; Now both DBL_MAX and +Inf/FE_OVERFLOW are correct results, depending on where double expressions are evaluated with extra range. o Now we have a double_t foobar, reloaded into a register if that is useful, and can probably use it as either a double or a long double at no extra cost, but if double_t is double and the compiler is a C compiler, then converting foobar to a double always cost a lot and usually wasted its extra range if not its extra precision. We should probably have used double_t throughout to avoid this... o .. however, if we use double_t too much, then it will be used for memory variables and the loads and stores for these will cost more than conversions to double and reloads. o However2, someday when 64-bit precision is supported, using double_t will be especially important for memory variables, to prevent unintended loss of precision. Bruce From yongari at FreeBSD.org Fri Mar 7 04:55:59 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Fri Mar 7 04:56:01 2008 Subject: cvs commit: src/sys/dev/msk if_msk.c Message-ID: <200803070455.m274txUv094539@repoman.freebsd.org> yongari 2008-03-07 04:55:58 UTC FreeBSD src repository Modified files: sys/dev/msk if_msk.c Log: To overcome hardware checksum offload bug msk(4) used to compute TCP/UDP checksum in driver for short frames. For frames that requires hardware VLAN tag insertion, the checksum offload trick does not work due to changes of checksum offset in mbuf after the VLAN tag. Disable hardware checksum offload for VLAN interface to fix the bug. Reported by: Christopher Cowart < ccowart AT rescomp DOT berkeley DOT edu > Tested by: Christopher Cowart < ccowart AT rescomp DOT berkeley DOT edu > MFC after: 5 days Revision Changes Path 1.30 +6 -2 src/sys/dev/msk/if_msk.c From brde at optusnet.com.au Fri Mar 7 06:22:27 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Fri Mar 7 06:22:32 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <47D089A4.4060208@freebsd.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305182531.GS68971@server.vk2pj.dyndns.org> <20080306021222.GA46783@zim.MIT.EDU> <47CF5D19.3090100@freebsd.org> <20080306033246.GA47280@zim.MIT.EDU> <47CF7EBF.6000009@freebsd.org> <20080306063452.GB48339@zim.MIT.EDU> <47CF9586.70707@freebsd.org> <20080307085031.P10724@delplex.bde.org> <47D089A4.4060208@freebsd.org> Message-ID: <20080307163246.J22709@besplex.bde.org> On Thu, 6 Mar 2008, Colin Percival wrote: > Bruce Evans wrote: >> On Wed, 5 Mar 2008, Colin Percival wrote: >>> You could have stopped this sentence here -- for all practical purposes, >>> correctly rounded trigonometric functions are not feasible. >> >> Nah, it's quite feasible, especially for transcendental functions since >> transcendental numbers in general can be approximated more closely by >> rationals than algebraic numbers. > > In fact, one of the common ways to prove that a real number is transcendental > is to show that it can be approximated more closely by rationals than any > algebraic function can be. I've only studied a few pages related to this. >> The evaluation just needs to be >> done in extra precision > > How much extra precision do you need? If the exact value of a transcendental > function is extremely close to halfway between two consecutive floating-point > values, how do you decide which way to round it? Probably not much more than the same number of extra bits as there are in the original precision. Values of trig functions mod 2^N are apparently sort of uniformly distributed. Thus each extra bit of precision gives correct rounding for about half of the remaining cases and N extra bits gives correct rounding for about 2^N cases, which is all cases if there were only 2^N cases to start with. This works out in practice for float precision -- with about 24 bits of extra precision, all except about 10 cases are perfectly rounded, and it is easy to evaluate those cases in extra precision to see that only a few more bits are needed. If the exact value is very close to half-way then you may need many more bits but would be unlucky to need many more. AFAIK (not far), no one knows exactly how many bits are needed for the worst case for even double precision. The domain is just too large to search (almost 2^64 bits). However, the large size of the domain means that very-near- half-way cases are very unlikely to occur in practice, so calling a slow multi-precision library to handle such cases would be a good way to handle them and sending mail about them would be a not so good way. But detecting such cases is too costly to do routinely, at least without hardware support. > Computing transcendental functions to within (0.5+x)ulp for small positive > values of x is certainly feasible, but computing provably correctly rounded > transcendental functions is very very hard. I think it is only very hard globally and for general functions. Elementary transcendental functions like exp() have rational coefficients and well-known properties which make them easier to bound. >>> Fortunately, library functions aren't required to have any particular >>> error bounds. >> >> We require error bounds of < 1 ulp and aim for < 0.5[0]1 ulps. > > What standard states that those bounds are required (or can be relied upon > when proving that code is correct)? Just the principle of least error. > Intel FPUs over the years have been > variously documented to compute transcendental functions to within 1.0, 2.5, > or 3.5 ulp -- are you saying that the 8087 doesn't comply with the standard > which it inspired? I didn't say that before, but in fact the ix87 has always had multi-gulp errors where its docs indicate errors of < 1 ulp, mainly for trig functions. Its trig functions can't possibly have errors of < 1 ulp, since its approximation to pi has only 66-68 bits, but an approximation with thousands of bits is needed (see libm). Multi-gulp errors occur as early as for x near pi/2 in 64-bit precision, since subtracting a 64-bit x near pi/2 from a 66 or 68 bit approx to pi/2 gives about 2 or 4 bits of accuracy or about 62 or 60 bits of error. 2^60 is about 1 exa-ulp (10^9 gulps). The errors are much larger for larger x, and give multi-gulp errors even for float precision for not very large x. Backwards compatibility has apparently resulted in these errors being perfectly preserved in all implementations of the ix87. Other ix87 errors are relatively minor AFAIK. E.g., exp(x) is not implemented directly in hardware. It requires several imprecise calculations like x*log2(e), so the hardware is good enough for double precision but not for long double precision where I guess it has an error of a few ulps. It has to switch the mode to long double precision to be good enough for double precision. log(x) is implemented more directly in hardware and I guess it has an error of about 1 ulp in long double precision. Bruce From sos at FreeBSD.org Fri Mar 7 09:29:21 2008 From: sos at FreeBSD.org (Sřren Schmidt) Date: Fri Mar 7 09:29:23 2008 Subject: cvs commit: src/sys/dev/ata ata-chipset.c ata-pci.h Message-ID: <200803070929.m279TLGM020793@repoman.freebsd.org> sos 2008-03-07 09:29:20 UTC FreeBSD src repository Modified files: sys/dev/ata ata-chipset.c ata-pci.h Log: Add proper support for the SATA/AHCI part of IXP[67]00 Revision Changes Path 1.212 +22 -12 src/sys/dev/ata/ata-chipset.c 1.83 +6 -0 src/sys/dev/ata/ata-pci.h From syrinx at FreeBSD.org Fri Mar 7 09:33:30 2008 From: syrinx at FreeBSD.org (Shteryana Shopova) Date: Fri Mar 7 09:33:38 2008 Subject: cvs commit: src/contrib/bsnmp - Imported sources Message-ID: <200803070933.m279XTD1021100@repoman.freebsd.org> syrinx 2008-03-07 09:33:29 UTC FreeBSD src repository src/contrib/bsnmp - Imported sources Update of /home/ncvs/src/contrib/bsnmp In directory repoman.freebsd.org:/tmp/cvs-serv20735 Log Message: Vendor patch: Prevent bsnmpd from dumping core when a module's init hook returns an error. Approved by: bz (mentor), harti Status: Vendor Tag: BEGEMOT Release Tags: syrinx_20080307_bugfix U src/contrib/bsnmp/Makefile.in U src/contrib/bsnmp/NEWS U src/contrib/bsnmp/README U src/contrib/bsnmp/TODO U src/contrib/bsnmp/VERSION U src/contrib/bsnmp/acinclude.m4 U src/contrib/bsnmp/aclocal.m4 U src/contrib/bsnmp/configure U src/contrib/bsnmp/configure.ac U src/contrib/bsnmp/oid-list U src/contrib/bsnmp/config/Makefile.build U src/contrib/bsnmp/config/Makefile.post U src/contrib/bsnmp/config/Makefile.pre U src/contrib/bsnmp/config/config.guess U src/contrib/bsnmp/config/config.sub U src/contrib/bsnmp/config/install.sh U src/contrib/bsnmp/config/ltmain.sh U src/contrib/bsnmp/config/mkinstalldirs U src/contrib/bsnmp/gensnmpdef/Makefile.in U src/contrib/bsnmp/gensnmpdef/gensnmpdef.1 U src/contrib/bsnmp/gensnmpdef/gensnmpdef.c U src/contrib/bsnmp/gensnmptree/Makefile.in U src/contrib/bsnmp/gensnmptree/gensnmptree.1 U src/contrib/bsnmp/gensnmptree/gensnmptree.c U src/contrib/bsnmp/lib/Makefile.in U src/contrib/bsnmp/lib/asn1.3 U src/contrib/bsnmp/lib/asn1.c U src/contrib/bsnmp/lib/asn1.h U src/contrib/bsnmp/lib/bsnmpagent.3 U src/contrib/bsnmp/lib/bsnmpclient.3 U src/contrib/bsnmp/lib/bsnmplib.3 U src/contrib/bsnmp/lib/snmp.c U src/contrib/bsnmp/lib/snmp.h U src/contrib/bsnmp/lib/snmpagent.c U src/contrib/bsnmp/lib/snmpagent.h U src/contrib/bsnmp/lib/snmpclient.c U src/contrib/bsnmp/lib/snmpclient.h U src/contrib/bsnmp/lib/snmppriv.h U src/contrib/bsnmp/lib/support.c U src/contrib/bsnmp/lib/support.h U src/contrib/bsnmp/snmp_mibII/BEGEMOT-IP-MIB.txt U src/contrib/bsnmp/snmp_mibII/BEGEMOT-MIB2-MIB.txt U src/contrib/bsnmp/snmp_mibII/mibII.c U src/contrib/bsnmp/snmp_mibII/mibII.h U src/contrib/bsnmp/snmp_mibII/mibII_begemot.c U src/contrib/bsnmp/snmp_mibII/mibII_ifmib.c U src/contrib/bsnmp/snmp_mibII/mibII_ifstack.c U src/contrib/bsnmp/snmp_mibII/mibII_interfaces.c U src/contrib/bsnmp/snmp_mibII/mibII_ip.c U src/contrib/bsnmp/snmp_mibII/mibII_ipaddr.c U src/contrib/bsnmp/snmp_mibII/mibII_nettomedia.c U src/contrib/bsnmp/snmp_mibII/mibII_rcvaddr.c U src/contrib/bsnmp/snmp_mibII/mibII_route.c U src/contrib/bsnmp/snmp_mibII/mibII_tcp.c U src/contrib/bsnmp/snmp_mibII/mibII_tree.def U src/contrib/bsnmp/snmp_mibII/mibII_udp.c U src/contrib/bsnmp/snmp_mibII/snmp_mibII.3 U src/contrib/bsnmp/snmp_mibII/snmp_mibII.h U src/contrib/bsnmp/snmp_ntp/BEGEMOT-NTP-MIB.txt U src/contrib/bsnmp/snmp_ntp/Makefile.in U src/contrib/bsnmp/snmp_ntp/NTP-MIB.txt U src/contrib/bsnmp/snmp_ntp/NTP-PROXY-MIB.txt U src/contrib/bsnmp/snmp_ntp/ntp_tree.def U src/contrib/bsnmp/snmp_ntp/snmp_ntp.c U src/contrib/bsnmp/snmpd/.gdbinit U src/contrib/bsnmp/snmpd/BEGEMOT-MIB.txt U src/contrib/bsnmp/snmpd/BEGEMOT-SNMPD.txt U src/contrib/bsnmp/snmpd/FOKUS-MIB.txt U src/contrib/bsnmp/snmpd/Makefile.in U src/contrib/bsnmp/snmpd/action.c U src/contrib/bsnmp/snmpd/bsnmpd.1 U src/contrib/bsnmp/snmpd/config.c U src/contrib/bsnmp/snmpd/export.c U src/contrib/bsnmp/snmpd/main.c U src/contrib/bsnmp/snmpd/snmpd.config U src/contrib/bsnmp/snmpd/snmpd.h U src/contrib/bsnmp/snmpd/snmpd.sh U src/contrib/bsnmp/snmpd/snmpmod.3 U src/contrib/bsnmp/snmpd/snmpmod.h U src/contrib/bsnmp/snmpd/trans_lsock.c U src/contrib/bsnmp/snmpd/trans_lsock.h U src/contrib/bsnmp/snmpd/trans_udp.c U src/contrib/bsnmp/snmpd/trans_udp.h U src/contrib/bsnmp/snmpd/trap.c U src/contrib/bsnmp/snmpd/tree.def No conflicts created by this import From rwatson at FreeBSD.org Fri Mar 7 13:13:18 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 13:13:20 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803071313.m27DDIhv044794@repoman.freebsd.org> rwatson 2008-03-07 13:13:18 UTC FreeBSD src repository Modified files: sys/ddb db_input.c Log: Reserve two bytes at the end of the DDB input line in db_readline() to hold the newline and nul terminator. Otherwise, there are cases where garbage may end up in the command history due to a lack of a nul terminator, or input may end up without room for a newline. MFC after: 3 days PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.39 +4 -1 src/sys/ddb/db_input.c From jhb at FreeBSD.org Fri Mar 7 13:36:39 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Mar 7 13:36:43 2008 Subject: cvs commit: src/sys/dev/agp agp_nvidia.c Message-ID: <200803071336.m27DadeQ045913@repoman.freebsd.org> jhb 2008-03-07 13:36:39 UTC FreeBSD src repository Modified files: sys/dev/agp agp_nvidia.c Log: Calculate the number of pages the GATT spans when reading from each page to flush the TLB instead of hardcoding a size of 33 pages. Apertures of 32MB and 64MB only use a 16 page GATT and an aperture of 128MB only uses a 32 page GATT, so without this the code could walk off the end of the pointer and cause a page fault if the next page was unmapped. Also, for aperture sizes > 128MB, not all of the pages would be read. The Linux driver has the same bug. MFC after: 1 week Tested by: Frédéric PRACA frederic.praca of freebsd-fr.org Revision Changes Path 1.14 +4 -3 src/sys/dev/agp/agp_nvidia.c From jhb at freebsd.org Fri Mar 7 14:19:06 2008 From: jhb at freebsd.org (John Baldwin) Date: Fri Mar 7 14:19:17 2008 Subject: cvs commit: src/sys/sys pcpu.h In-Reply-To: <200803061724.m26HOSPO026401@repoman.freebsd.org> References: <200803061724.m26HOSPO026401@repoman.freebsd.org> Message-ID: <200803070801.30460.jhb@freebsd.org> On Thursday 06 March 2008 12:24:28 pm Marcel Moolenaar wrote: > marcel 2008-03-06 17:24:28 UTC > > FreeBSD src repository > > Modified files: > sys/sys pcpu.h > Log: > Move the PCPU_MD_FIELDS last in struct pcpu. While this header is > private to the kernel, some ports define _KERNEL and include this > header. While arguably this is wrong, it's also reality. By having > the MD fields last, architectures that have CPU-specific variations > of PCPU_MD_FIELDS will at least have the MI fields at a constant > offset. Of course, having all MI fields first helps kernel debugging > as well, so this is not a change without some benefits to us. > > This change does not result in an ABI breakage, because this header > is not part of the ABI. Recompilation of lsof is required though :-) It actually started out being the last "field" in the structure but some new fields were incorrectly added after it rather than before it. It might warrant a comment to prevent similar regressions in the future. -- John Baldwin From rwatson at FreeBSD.org Fri Mar 7 14:26:31 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 14:26:37 2008 Subject: cvs commit: src/sys/ddb db_command.c Message-ID: <200803071426.m27EQVMp049638@repoman.freebsd.org> rwatson 2008-03-07 14:26:31 UTC FreeBSD src repository Modified files: sys/ddb db_command.c Log: When killing a user process from DDB, check that the requested signal is > 0 rather than >= 0, or we will panic when trying to deliver the signal. MFC after: 3 days PR: 100802 Submitted by: Valerio Daelli Revision Changes Path 1.77 +1 -1 src/sys/ddb/db_command.c From remko at FreeBSD.org Fri Mar 7 14:58:01 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 14:58:12 2008 Subject: cvs commit: src/sys/dev/drm drm_pciids.h Message-ID: <200803071458.m27Ew1l6050653@repoman.freebsd.org> remko 2008-03-07 14:58:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/drm drm_pciids.h Log: MFC rev 1.8 drm_pciids.h Add support for the 965GM. PR: kern/120978 Submitted by: Dominic Fandrey Approved by: imp (mentor, implicit), anholt (drm maintainer) MFC after: 3 days Approved by: imp (mentor, implicit) Revision Changes Path 1.7.2.1 +1 -0 src/sys/dev/drm/drm_pciids.h From remko at FreeBSD.org Fri Mar 7 14:58:50 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 14:59:02 2008 Subject: cvs commit: src/sys/dev/drm drm_pciids.h Message-ID: <200803071458.m27EwooY050700@repoman.freebsd.org> remko 2008-03-07 14:58:50 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/drm drm_pciids.h Log: MFC rev 1.8 drm_pciids.h Add support for the 965GM. PR: kern/120978 Submitted by: Dominic Fandrey Approved by: imp (mentor, implicit), anholt (drm maintainer) MFC after: 3 days Approved by: imp (mentor, implicit) Revision Changes Path 1.2.2.5 +1 -0 src/sys/dev/drm/drm_pciids.h From remko at FreeBSD.org Fri Mar 7 15:01:53 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 15:02:02 2008 Subject: cvs commit: src/lib/libfetch common.c fetch.3 ftp.c http.c Message-ID: <200803071501.m27F1rYD050861@repoman.freebsd.org> remko 2008-03-07 15:01:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libfetch common.c fetch.3 ftp.c http.c Log: MFC libfetch from HEAD to RELENG_7 Approved by: imp (mentor, implicit), DES Glanced at by: des Revision Changes Path 1.50.10.2 +1 -1 src/lib/libfetch/common.c 1.63.2.2 +1 -1 src/lib/libfetch/fetch.3 1.96.2.3 +4 -4 src/lib/libfetch/ftp.c 1.78.2.2 +4 -3 src/lib/libfetch/http.c From remko at FreeBSD.org Fri Mar 7 15:23:33 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 15:23:35 2008 Subject: cvs commit: src/sys/dev/usb usbdevs uscanner.c Message-ID: <200803071523.m27FNXZ9052931@repoman.freebsd.org> remko 2008-03-07 15:23:32 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/usb usbdevs uscanner.c Log: MFC rev 1.340 usbdevs MFC rev 1.90 uscanner.c Add support for the EPSON CX5400 scanner PR: usb/120980 Submitted by: Christophe Etcheverry MFC after: 3 days Approved by: imp (mentor, implicit for trivial changes) Approved by: imp (mentor, implicit) Revision Changes Path 1.232.2.26 +1 -0 src/sys/dev/usb/usbdevs 1.63.2.8 +1 -0 src/sys/dev/usb/uscanner.c From rwatson at FreeBSD.org Fri Mar 7 15:25:57 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 15:25:59 2008 Subject: cvs commit: src/lib/libc/posix1e mac.c mac_exec.c mac_get.c mac_set.c Message-ID: <200803071525.m27FPucC053111@repoman.freebsd.org> rwatson 2008-03-07 15:25:56 UTC FreeBSD src repository Modified files: lib/libc/posix1e mac.c mac_exec.c mac_get.c mac_set.c Log: Add __FBSDID() tags. MFC after: 3 days Revision Changes Path 1.13 +3 -2 src/lib/libc/posix1e/mac.c 1.3 +3 -2 src/lib/libc/posix1e/mac_exec.c 1.6 +3 -2 src/lib/libc/posix1e/mac_get.c 1.4 +3 -2 src/lib/libc/posix1e/mac_set.c From remko at FreeBSD.org Fri Mar 7 15:26:41 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 15:26:43 2008 Subject: cvs commit: src/sys/dev/usb usbdevs uscanner.c Message-ID: <200803071526.m27FQf4x053171@repoman.freebsd.org> remko 2008-03-07 15:26:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/usb usbdevs uscanner.c Log: MFC rev 1.340 usbdevs MFC rev 1.90 uscanner.c Add support for the EPSON CX5400 scanner PR: usb/120980 Submitted by: Christophe Etcheverry MFC after: 3 days Approved by: imp (mentor, implicit for trivial changes) Approved by: imp (mentor, implicit) Revision Changes Path 1.328.2.4 +1 -0 src/sys/dev/usb/usbdevs 1.89.2.1 +1 -0 src/sys/dev/usb/uscanner.c From rwatson at FreeBSD.org Fri Mar 7 15:27:09 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 15:27:14 2008 Subject: cvs commit: src/sys/kern kern_priv.c Message-ID: <200803071527.m27FR8ss053287@repoman.freebsd.org> rwatson 2008-03-07 15:27:08 UTC FreeBSD src repository Modified files: sys/kern kern_priv.c Log: Add __FBSDID() tag. MFC after: 3 days Pointed out by: antoine Revision Changes Path 1.5 +3 -2 src/sys/kern/kern_priv.c From remko at FreeBSD.org Fri Mar 7 15:54:42 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 15:54:47 2008 Subject: cvs commit: src/sys/dev/usb umass.c usbdevs Message-ID: <200803071554.m27FsgDs057345@repoman.freebsd.org> remko 2008-03-07 15:54:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/usb umass.c usbdevs Log: MFC rev 1.162 umass.c MFC rev 1.338 usbdevs Add support for the VIA USB2IDE bridge [1]. PR: usb/95173 Submitted by: Goegele Hannes Ben Kelly Tested by: blackend Approved by: imp (mentor, blanket for simple patches) X-MFC-After: 1 week Approved by: imp (mentor, implicit) Revision Changes Path 1.160.2.1 +19 -0 src/sys/dev/usb/umass.c 1.328.2.5 +3 -0 src/sys/dev/usb/usbdevs From xcllnt at mac.com Fri Mar 7 15:56:20 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Fri Mar 7 15:56:24 2008 Subject: cvs commit: src/sys/sys pcpu.h In-Reply-To: <200803070801.30460.jhb@freebsd.org> References: <200803061724.m26HOSPO026401@repoman.freebsd.org> <200803070801.30460.jhb@freebsd.org> Message-ID: <17C624F8-3233-49E6-8235-30F8B06C6513@mac.com> On Mar 7, 2008, at 5:01 AM, John Baldwin wrote: > On Thursday 06 March 2008 12:24:28 pm Marcel Moolenaar wrote: >> marcel 2008-03-06 17:24:28 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/sys pcpu.h >> Log: >> Move the PCPU_MD_FIELDS last in struct pcpu. While this header is >> private to the kernel, some ports define _KERNEL and include this >> header. While arguably this is wrong, it's also reality. By having >> the MD fields last, architectures that have CPU-specific variations >> of PCPU_MD_FIELDS will at least have the MI fields at a constant >> offset. Of course, having all MI fields first helps kernel >> debugging >> as well, so this is not a change without some benefits to us. >> >> This change does not result in an ABI breakage, because this header >> is not part of the ABI. Recompilation of lsof is required though :-) > > It actually started out being the last "field" in the structure but > some new > fields were incorrectly added after it rather than before it. It > might > warrant a comment to prevent similar regressions in the future. > > -- > John Baldwin -- Marcel Moolenaar xcllnt@mac.com From xcllnt at mac.com Fri Mar 7 15:58:18 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Fri Mar 7 15:58:22 2008 Subject: cvs commit: src/sys/sys pcpu.h In-Reply-To: <17C624F8-3233-49E6-8235-30F8B06C6513@mac.com> References: <200803061724.m26HOSPO026401@repoman.freebsd.org> <200803070801.30460.jhb@freebsd.org> <17C624F8-3233-49E6-8235-30F8B06C6513@mac.com> Message-ID: <195BED5E-661B-4AC4-8302-59BEC71708A6@mac.com> On Mar 7, 2008, at 7:56 AM, Marcel Moolenaar wrote: >> It actually started out being the last "field" in the structure but >> some new >> fields were incorrectly added after it rather than before it. It >> might >> warrant a comment to prevent similar regressions in the future. >> > > What I was about to say is that I agree about the comment and did add one as part of the change... -- Marcel Moolenaar xcllnt@mac.com From remko at FreeBSD.org Fri Mar 7 16:00:30 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 16:00:37 2008 Subject: cvs commit: src/sys/dev/usb umass.c usbdevs Message-ID: <200803071600.m27G0Shk057627@repoman.freebsd.org> remko 2008-03-07 16:00:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/usb umass.c usbdevs Log: MFC rev 1.162 umass.c MFC rev 1.338 usbdevs Add support for the VIA USB2IDE bridge [1]. PR: usb/95173 Submitted by: Goegele Hannes Ben Kelly Tested by: blackend Approved by: imp (mentor, blanket for simple patches) X-MFC-After: 1 week Approved by: imp (mentor, implicit) Revision Changes Path 1.122.2.13 +19 -0 src/sys/dev/usb/umass.c 1.232.2.27 +3 -0 src/sys/dev/usb/usbdevs From rwatson at FreeBSD.org Fri Mar 7 16:00:45 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 16:00:47 2008 Subject: cvs commit: src/sys/net if.c if_ethersubr.c Message-ID: <200803071600.m27G0iKq057677@repoman.freebsd.org> rwatson 2008-03-07 16:00:44 UTC FreeBSD src repository Modified files: sys/net if.c if_ethersubr.c Log: Move IFF_NEEDSGIANT warning from if_ethersubr.c to if.c so it is displayed for all network interfaces, not just ethernet-like ones. Upgrade it to a louder WARNING and be explicit that the flag is obsolete. Support for IFF_NEEDSGIANT will be removed in a few months (see arch@ for details) and will not appear in 8.0. Upgrade if_watchdog to a WARNING. Revision Changes Path 1.275 +5 -1 src/sys/net/if.c 1.241 +0 -2 src/sys/net/if_ethersubr.c From remko at FreeBSD.org Fri Mar 7 16:29:42 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Fri Mar 7 16:29:53 2008 Subject: cvs commit: src/sys/dev/ata ata-chipset.c ata-pci.h In-Reply-To: <200803070929.m279TLGM020793@repoman.freebsd.org> References: <200803070929.m279TLGM020793@repoman.freebsd.org> Message-ID: <47D16741.8070904@FreeBSD.org> S?ren Schmidt wrote: > sos 2008-03-07 09:29:20 UTC > > FreeBSD src repository > > Modified files: > sys/dev/ata ata-chipset.c ata-pci.h > Log: > Add proper support for the SATA/AHCI part of IXP[67]00 > > Revision Changes Path > 1.212 +22 -12 src/sys/dev/ata/ata-chipset.c > 1.83 +6 -0 src/sys/dev/ata/ata-pci.h Thanks! I am using a similiar patch for my production system, which works just fine on RELENG_7, please can you do the merge when possible? cheers Remko -- /"\ Best regards, | remko@FreeBSD.org \ / Remko Lodder | remko@EFnet X http://www.evilcoder.org/ | / \ ASCII Ribbon Campaign | Against HTML Mail and News From bmah at FreeBSD.org Fri Mar 7 16:32:44 2008 From: bmah at FreeBSD.org (Bruce A. Mah) Date: Fri Mar 7 16:32:46 2008 Subject: cvs commit: src/release/doc/en_US.ISO8859-1/errata article.sgml Message-ID: <200803071632.m27GWi4d059919@repoman.freebsd.org> bmah 2008-03-07 16:32:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) release/doc/en_US.ISO8859-1/errata article.sgml Log: New errata notes: SCTP needs INET6, kldxref(8) warnings on source upgrade, device crypto problems on powerpc. Revision Changes Path 1.73.4.3 +16 -0 src/release/doc/en_US.ISO8859-1/errata/article.sgml From bmah at FreeBSD.org Fri Mar 7 16:42:16 2008 From: bmah at FreeBSD.org (Bruce A. Mah) Date: Fri Mar 7 16:42:25 2008 Subject: cvs commit: src/release/doc/en_US.ISO8859-1/errata article.sgml Message-ID: <200803071642.m27GgGkk060277@repoman.freebsd.org> bmah 2008-03-07 16:42:16 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) release/doc/en_US.ISO8859-1/errata article.sgml Log: New errata note: sysinstall(8) X.org hazard. Revision Changes Path 1.73.2.55 +10 -0 src/release/doc/en_US.ISO8859-1/errata/article.sgml From jasone at FreeBSD.org Fri Mar 7 16:54:05 2008 From: jasone at FreeBSD.org (Jason Evans) Date: Fri Mar 7 16:54:11 2008 Subject: cvs commit: src/lib/libc/stdlib malloc.c Message-ID: <200803071654.m27Gs4iV070576@repoman.freebsd.org> jasone 2008-03-07 16:54:04 UTC FreeBSD src repository Modified files: lib/libc/stdlib malloc.c Log: Remove stale #include , which as needed by lazy deallocation. Revision Changes Path 1.167 +4 -4 src/lib/libc/stdlib/malloc.c From marcel at FreeBSD.org Fri Mar 7 16:55:25 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Fri Mar 7 16:55:46 2008 Subject: cvs commit: src/sys/dev/usb if_axe.c Message-ID: <200803071655.m27GtO6h070826@repoman.freebsd.org> marcel 2008-03-07 16:55:24 UTC FreeBSD src repository Modified files: sys/dev/usb if_axe.c Log: Apply le*toh() or htole*() to the variables of which we use the address as the buffer pointer in the call to axe_cmd(). This is needed to make the code work on big-endian machines. Ok'd: imp@ Revision Changes Path 1.57 +6 -1 src/sys/dev/usb/if_axe.c From rwatson at FreeBSD.org Fri Mar 7 17:02:34 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 17:02:40 2008 Subject: cvs commit: src/share/man/man9 bus_space.9 Message-ID: <200803071702.m27H2YSu071107@repoman.freebsd.org> rwatson 2008-03-07 17:02:34 UTC FreeBSD src repository Modified files: share/man/man9 bus_space.9 Log: Minor spelling/grammer fixes. MFC after: 3 days Revision Changes Path 1.5 +2 -2 src/share/man/man9/bus_space.9 From mav at FreeBSD.org Fri Mar 7 17:54:00 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Fri Mar 7 17:54:03 2008 Subject: cvs commit: src/sys/netgraph ng_nat.c ng_nat.h Message-ID: <200803071754.m27Hs07G074170@repoman.freebsd.org> mav 2008-03-07 17:53:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_nat.c ng_nat.h Log: MFC: Add support for the libalias redirect functionality. Submitted by: Vadim Goncharov Revision Changes Path 1.10.2.1 +431 -1 src/sys/netgraph/ng_nat.c 1.2.2.1 +127 -0 src/sys/netgraph/ng_nat.h From mav at FreeBSD.org Fri Mar 7 17:54:22 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Fri Mar 7 17:54:25 2008 Subject: cvs commit: src/share/man/man4 ng_nat.4 Message-ID: <200803071754.m27HsLTu074209@repoman.freebsd.org> mav 2008-03-07 17:54:21 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 ng_nat.4 Log: MFC: Add support for the libalias redirect functionality. Submitted by: Vadim Goncharov Revision Changes Path 1.3.2.1 +188 -2 src/share/man/man4/ng_nat.4 From mav at FreeBSD.org Fri Mar 7 17:58:50 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Fri Mar 7 17:58:52 2008 Subject: cvs commit: src/sys/netgraph ng_pppoe.c Message-ID: <200803071758.m27HwnZY074411@repoman.freebsd.org> mav 2008-03-07 17:58:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_pppoe.c Log: MFC rev. 1.92-1.93 Make session ID generator to use session ID hash. Make session ID generator thread-safe. Use more compact LIST instead of TAILQ for session hash. Add all listening hooks into LIST to simplify searches. Use ng_findhook() instead of own equal implementation. Revision Changes Path 1.84.2.5 +67 -100 src/sys/netgraph/ng_pppoe.c From rwatson at FreeBSD.org Fri Mar 7 18:09:08 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 18:09:18 2008 Subject: cvs commit: src/sys/ddb db_examine.c Message-ID: <200803071809.m27I97In075943@repoman.freebsd.org> rwatson 2008-03-07 18:09:07 UTC FreeBSD src repository Modified files: sys/ddb db_examine.c Log: Add a /S mode to DDB "ex" command, which interprets and prints the value at the requested address as a symbol. For example, "ex /S aio_swake" prints the name of the function currently registered in via aio_swake hook. The change as committed differs slightly from the patch in the PR, as I force the size of the retrieved value (and the automatic address increment) to be sizeof(void *). This seems to provide the most useful auto-increment behavior, and defaults using the default size (4), which is not sizeof(void *) on 64-bit platforms. MFC after: 3 days PR: 57976 Submitted by: Dan Strick Revision Changes Path 1.35 +6 -0 src/sys/ddb/db_examine.c From rwatson at FreeBSD.org Fri Mar 7 18:14:30 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 7 18:14:32 2008 Subject: cvs commit: src/share/man/man4 ddb.4 Message-ID: <200803071814.m27IEUqi076210@repoman.freebsd.org> rwatson 2008-03-07 18:14:30 UTC FreeBSD src repository Modified files: share/man/man4 ddb.4 Log: Update ddb.4 to reflect addition of /S (symbol name) printing mode for examine. MFC after: 3 days PR: 57976 Revision Changes Path 1.48 +2 -0 src/share/man/man4/ddb.4 From jhb at FreeBSD.org Fri Mar 7 20:19:24 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Mar 7 20:19:28 2008 Subject: cvs commit: src/sys/i386/i386 identcpu.c initcpu.c Message-ID: <200803072019.m27KJO55084604@repoman.freebsd.org> jhb 2008-03-07 20:19:24 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/i386/i386 identcpu.c initcpu.c Log: MFC: VIA Padlock changes: - Read the brand string from VIA/IDT CPUs. - Add a VIA Padlock feature line in dmesg. - Support the newer Via C7 core (0x6d0). Revision Changes Path 1.180.2.1 +43 -23 src/sys/i386/i386/identcpu.c 1.56.2.1 +1 -0 src/sys/i386/i386/initcpu.c From jhb at FreeBSD.org Fri Mar 7 20:19:34 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Mar 7 20:19:36 2008 Subject: cvs commit: src/sys/i386/i386 identcpu.c initcpu.c Message-ID: <200803072019.m27KJXYY084637@repoman.freebsd.org> jhb 2008-03-07 20:19:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/i386/i386 identcpu.c initcpu.c Log: MFC: VIA Padlock changes: - Read the brand string from VIA/IDT CPUs. - Add a VIA Padlock feature line in dmesg. - Support the newer Via C7 core (0x6d0). Revision Changes Path 1.145.2.7 +41 -26 src/sys/i386/i386/identcpu.c 1.52.2.4 +1 -0 src/sys/i386/i386/initcpu.c From jhb at freebsd.org Fri Mar 7 20:33:50 2008 From: jhb at freebsd.org (John Baldwin) Date: Fri Mar 7 20:33:57 2008 Subject: cvs commit: src/sys/sys pcpu.h In-Reply-To: <195BED5E-661B-4AC4-8302-59BEC71708A6@mac.com> References: <200803061724.m26HOSPO026401@repoman.freebsd.org> <17C624F8-3233-49E6-8235-30F8B06C6513@mac.com> <195BED5E-661B-4AC4-8302-59BEC71708A6@mac.com> Message-ID: <200803071459.43636.jhb@freebsd.org> On Friday 07 March 2008 10:58:12 am Marcel Moolenaar wrote: > > On Mar 7, 2008, at 7:56 AM, Marcel Moolenaar wrote: > > >> It actually started out being the last "field" in the structure but > >> some new > >> fields were incorrectly added after it rather than before it. It > >> might > >> warrant a comment to prevent similar regressions in the future. > >> > > > > > > > What I was about to say is that I agree about the comment and did add > one > as part of the change... Oh, my bad then. :) -- John Baldwin From mav at FreeBSD.org Fri Mar 7 21:12:56 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Fri Mar 7 21:12:59 2008 Subject: cvs commit: src/sys/netgraph ng_ksocket.c Message-ID: <200803072112.m27LCuGQ096881@repoman.freebsd.org> mav 2008-03-07 21:12:56 UTC FreeBSD src repository Modified files: sys/netgraph ng_ksocket.c Log: Send only one incoming notification at a time to reduce queue trashing and improve performance. Remove waitflag argument from ng_ksocket_incoming2(), it means nothing as function call was queued by netgraph. Remove node validity check, as node validity guarantied by netgraph. Update comments. Revision Changes Path 1.61 +26 -32 src/sys/netgraph/ng_ksocket.c From marcel at FreeBSD.org Fri Mar 7 22:08:43 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Fri Mar 7 22:08:46 2008 Subject: cvs commit: src/sys/powerpc/include intr_machdep.h openpicvar.h src/sys/powerpc/mpc85xx ocpbus.c opic.c src/sys/powerpc/powermac openpic_macio.c src/sys/powerpc/powerpc intr_machdep.c openpic.c pic_if.m src/sys/powerpc/psim openpic_iobus.c Message-ID: <200803072208.m27M8h8r001749@repoman.freebsd.org> marcel 2008-03-07 22:08:43 UTC FreeBSD src repository Modified files: sys/powerpc/include intr_machdep.h openpicvar.h sys/powerpc/mpc85xx ocpbus.c opic.c sys/powerpc/powermac openpic_macio.c sys/powerpc/powerpc intr_machdep.c openpic.c pic_if.m sys/powerpc/psim openpic_iobus.c Log: Add support for the BUS_CONFIG_INTR() method to the platform and to openpic(4). Make use of it in ocpbus(4). On the MPC85xxCDS, IRQ0:4 are active-low. Revision Changes Path 1.11 +1 -0 src/sys/powerpc/include/intr_machdep.h 1.8 +1 -0 src/sys/powerpc/include/openpicvar.h 1.3 +16 -0 src/sys/powerpc/mpc85xx/ocpbus.c 1.3 +1 -0 src/sys/powerpc/mpc85xx/opic.c 1.12 +1 -0 src/sys/powerpc/powermac/openpic_macio.c 1.16 +88 -42 src/sys/powerpc/powerpc/intr_machdep.c 1.18 +20 -0 src/sys/powerpc/powerpc/openpic.c 1.6 +7 -0 src/sys/powerpc/powerpc/pic_if.m 1.9 +1 -0 src/sys/powerpc/psim/openpic_iobus.c From marcel at FreeBSD.org Fri Mar 7 22:27:11 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Fri Mar 7 22:27:14 2008 Subject: cvs commit: src/sys/powerpc/aim locore.S machdep.c src/sys/powerpc/include md_var.h src/sys/powerpc/powerpc cpu.c Message-ID: <200803072227.m27MRA7r002440@repoman.freebsd.org> marcel 2008-03-07 22:27:06 UTC FreeBSD src repository Modified files: sys/powerpc/aim locore.S machdep.c sys/powerpc/include md_var.h sys/powerpc/powerpc cpu.c Log: For AIM, have cpu_idle() set MSR_POW when the powerpc_pow_enabled variable is set. On my Mac Mini this puts the CPU in NAP mode when the kernel is idle and, any technical or environmental reasons aside, avoids that I have to listen to the fan all day :-) Revision Changes Path 1.25 +0 -3 src/sys/powerpc/aim/locore.S 1.110 +10 -2 src/sys/powerpc/aim/machdep.c 1.32 +2 -0 src/sys/powerpc/include/md_var.h 1.11 +4 -6 src/sys/powerpc/powerpc/cpu.c From jasone at FreeBSD.org Fri Mar 7 22:29:54 2008 From: jasone at FreeBSD.org (Jason Evans) Date: Fri Mar 7 22:29:58 2008 Subject: cvs commit: src/share/man/man3 Makefile Message-ID: <200803072229.m27MTrpt002503@repoman.freebsd.org> jasone 2008-03-07 22:29:53 UTC FreeBSD src repository Modified files: share/man/man3 Makefile Log: Install manpage links for RB_PREV and RB_FOREACH_REVERSE. Revision Changes Path 1.46 +2 -0 src/share/man/man3/Makefile From jasone at FreeBSD.org Fri Mar 7 22:34:17 2008 From: jasone at FreeBSD.org (Jason Evans) Date: Fri Mar 7 22:34:24 2008 Subject: cvs commit: src/share/man/man3 Makefile tree.3 src/sys/sys tree.h Message-ID: <200803072234.m27MYHno002766@repoman.freebsd.org> jasone 2008-03-07 22:34:17 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man3 tree.3 Makefile sys/sys tree.h Log: MFC: Implement the RB_FOREACH_REVERSE() and RB_PREV() macros. RB_PREV is needed by malloc(3). Revision Changes Path 1.45.2.1 +2 -0 src/share/man/man3/Makefile 1.6.2.1 +10 -2 src/share/man/man3/tree.3 1.6.2.1 +29 -0 src/sys/sys/tree.h From jasone at FreeBSD.org Fri Mar 7 22:39:40 2008 From: jasone at FreeBSD.org (Jason Evans) Date: Fri Mar 7 22:39:43 2008 Subject: cvs commit: src/lib/libc/gen Symbol.map src/lib/libc/stdlib malloc.3 malloc.c src/lib/libkse kse.map src/lib/libkse/sys lock.c lock.h src/lib/libkse/thread thr_cond.c thr_init.c thr_kern.c thr_mutex.c thr_rtld.c src/lib/libthr pthread.map ... Message-ID: <200803072239.m27MddwJ003085@repoman.freebsd.org> jasone 2008-03-07 22:39:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/gen Symbol.map lib/libc/stdlib malloc.3 malloc.c lib/libkse kse.map lib/libkse/sys lock.c lock.h lib/libkse/thread thr_cond.c thr_init.c thr_kern.c thr_mutex.c thr_rtld.c lib/libthr pthread.map lib/libthr/thread thr_mutex.c Log: MFC: Merge malloc(3) improvements and fixes. The highlights are: * Avoid re-zeroing memory in calloc() when possible. * Use pthread mutexes where possible instead of libc "spinlocks", and actually spin some during contention before blocking. * Implement dynamic load balancing of thread-->arena mapping. * Avoid floating point math in order to avoid increased context switch overhead for applications that otherwise would not use floating point math. * Restructure how sbrk() and mmap() are used to acquire memory mappings. This provides a way to force malloc to only use sbrk(), which can be useful in the context of resource limits. * Reduce the number of mmap() calls typically necessary when allocating a chunk. * Track dirty unused pages so that they can be purged if they exceed a threshold. * Try to realloc() large objects in place. * Manage page runs with trees instead of chunk maps, which allows logarithmic-time run allocation. Revision Changes Path 1.6.2.1 +1 -0 src/lib/libc/gen/Symbol.map 1.73.2.1 +95 -36 src/lib/libc/stdlib/malloc.3 1.147.2.2 +2283 -1341 src/lib/libc/stdlib/malloc.c 1.1.2.1 +1 -0 src/lib/libkse/kse.map 1.11.2.2 +4 -3 src/lib/libkse/sys/lock.c 1.8.2.1 +1 -1 src/lib/libkse/sys/lock.h 1.57.2.1 +1 -1 src/lib/libkse/thread/thr_cond.c 1.76.2.1 +4 -4 src/lib/libkse/thread/thr_init.c 1.125.2.2 +5 -5 src/lib/libkse/thread/thr_kern.c 1.53.2.2 +30 -7 src/lib/libkse/thread/thr_mutex.c 1.7.2.1 +1 -1 src/lib/libkse/thread/thr_rtld.c 1.18.2.1 +1 -0 src/lib/libthr/pthread.map 1.54.2.3 +23 -6 src/lib/libthr/thread/thr_mutex.c From delphij at FreeBSD.org Fri Mar 7 23:09:11 2008 From: delphij at FreeBSD.org (Xin LI) Date: Fri Mar 7 23:09:14 2008 Subject: cvs commit: src/sys/amd64/conf GENERIC src/sys/i386/conf GENERIC src/sys/powerpc/conf GENERIC Message-ID: <200803072309.m27N9AAI005293@repoman.freebsd.org> delphij 2008-03-07 23:09:10 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/amd64/conf GENERIC sys/i386/conf GENERIC sys/powerpc/conf GENERIC Log: MFC: Switch to use ULE as default scheduler on amd64, i386 and powerpc architectures. Revision Changes Path 1.484.2.4 +1 -1 src/sys/amd64/conf/GENERIC 1.474.2.4 +1 -1 src/sys/i386/conf/GENERIC 1.71.2.3 +1 -1 src/sys/powerpc/conf/GENERIC From delphij at FreeBSD.org Sat Mar 8 00:04:59 2008 From: delphij at FreeBSD.org (Xin LI) Date: Sat Mar 8 00:05:03 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c Message-ID: <200803080004.m2804xiC011923@repoman.freebsd.org> delphij 2008-03-08 00:04:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/pkg_install/add main.c Log: Merge from RELENG_7_0: Teach pkg_add(1) about 6.3-RELEASE, 7.0-RELEASE and 7-STABLE. Revision Changes Path 1.72.2.1 +3 -1 src/usr.sbin/pkg_install/add/main.c From doconnor at gsoft.com.au Sat Mar 8 00:16:17 2008 From: doconnor at gsoft.com.au (Daniel O'Connor) Date: Sat Mar 8 00:16:21 2008 Subject: cvs commit: src/share/man/man4 Makefile uslcom.4 src/sys/amd64/conf GENERIC src/sys/conf NOTES files src/sys/dev/usb usbdevs uslcom.c src/sys/i386/conf GENERIC src/sys/modules Makefile src/sys/modules/uslcom Makefile In-Reply-To: <200803051413.m25EDU9o054744@repoman.freebsd.org> References: <200803051413.m25EDU9o054744@repoman.freebsd.org> Message-ID: <200803081033.42393.doconnor@gsoft.com.au> On Thu, 6 Mar 2008, Rink Springer wrote: > Log: > Import uslcom(4) from OpenBSD - this is a driver for Silicon > Laboratories CP2101/CP2102 based USB serial adapters. Hah, I ported this too.. Mine has issues though - I have the CP2102 in a a AVR programmer/debugger tool and while the programmer part works the JTAG debugger part times out.. It seemed to me it was some subtle flow control issue to me but I haven't been able to debug it. The code you have appears to be very similar to mine though (hardly surprising given they have the same source). I will try and find time to see if your works with the JTAG debugger I have (fingers crossed :) -- Daniel O'Connor software and network engineer for Genesis Software - http://www.gsoft.com.au "The nice thing about standards is that there are so many of them to choose from." -- Andrew Tanenbaum GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: This is a digitally signed message part. Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080308/048bd9e2/attachment.pgp From delphij at FreeBSD.org Sat Mar 8 00:53:54 2008 From: delphij at FreeBSD.org (Xin LI) Date: Sat Mar 8 00:53:58 2008 Subject: cvs commit: src/sbin/ffsinfo ffsinfo.c Message-ID: <200803080053.m280rsnP022617@repoman.freebsd.org> delphij 2008-03-08 00:53:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/ffsinfo ffsinfo.c Log: MFC revision 1.14 date: 2008/01/29 00:20:00; author: delphij; state: Exp; lines: +1 -1 Don't coredump when executed with -o. Revision Changes Path 1.13.2.1 +1 -1 src/sbin/ffsinfo/ffsinfo.c From delphij at FreeBSD.org Sat Mar 8 00:54:28 2008 From: delphij at FreeBSD.org (Xin LI) Date: Sat Mar 8 00:54:31 2008 Subject: cvs commit: src/sbin/ffsinfo ffsinfo.c Message-ID: <200803080054.m280sR5d022661@repoman.freebsd.org> delphij 2008-03-08 00:54:27 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sbin/ffsinfo ffsinfo.c Log: MFC revision 1.14 date: 2008/01/29 00:20:00; author: delphij; state: Exp; lines: +1 -1 Don't coredump when executed with -o. Revision Changes Path 1.10.2.3 +1 -1 src/sbin/ffsinfo/ffsinfo.c From delphij at FreeBSD.org Sat Mar 8 01:00:26 2008 From: delphij at FreeBSD.org (Xin LI) Date: Sat Mar 8 01:00:28 2008 Subject: cvs commit: src/usr.sbin/newsyslog newsyslog.c Message-ID: <200803080100.m2810Pp8022916@repoman.freebsd.org> delphij 2008-03-08 01:00:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/newsyslog newsyslog.c Log: MFC revision 1.108 date: 2008/01/30 22:11:59; author: delphij; state: Exp; lines: +2 -2 Fix two typos. Revision Changes Path 1.107.2.1 +2 -2 src/usr.sbin/newsyslog/newsyslog.c From delphij at FreeBSD.org Sat Mar 8 01:00:40 2008 From: delphij at FreeBSD.org (Xin LI) Date: Sat Mar 8 01:00:42 2008 Subject: cvs commit: src/usr.sbin/newsyslog newsyslog.c Message-ID: <200803080100.m2810d4M022949@repoman.freebsd.org> delphij 2008-03-08 01:00:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/newsyslog newsyslog.c Log: MFC revision 1.108 date: 2008/01/30 22:11:59; author: delphij; state: Exp; lines: +2 -2 Fix two typos. Revision Changes Path 1.101.2.5 +2 -2 src/usr.sbin/newsyslog/newsyslog.c From lippe at FreeBSD.org Sat Mar 8 03:54:31 2008 From: lippe at FreeBSD.org (Felippe de Meirelles Motta) Date: Sat Mar 8 03:54:34 2008 Subject: cvs commit: src/usr.bin/calendar/calendars calendar.freebsd Message-ID: <200803080354.m283sURd032550@repoman.freebsd.org> lippe 2008-03-08 03:54:30 UTC FreeBSD src repository (ports committer) Modified files: usr.bin/calendar/calendars calendar.freebsd Log: - Add my birthday entry Approved by: gabor (mentor) Revision Changes Path 1.228 +1 -0 src/usr.bin/calendar/calendars/calendar.freebsd From marcel at FreeBSD.org Sat Mar 8 05:36:26 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sat Mar 8 05:37:03 2008 Subject: cvs commit: src/sys/powerpc/booke machdep.c Message-ID: <200803080536.m285aPN7047604@repoman.freebsd.org> marcel 2008-03-08 05:36:25 UTC FreeBSD src repository Modified files: sys/powerpc/booke machdep.c Log: Enable the D-cache and I-cache when not already enabled. It so happens that U-Boot disables the D-cache when booting an ELF image, so this change makes sure we run with the D-cache enabled from now on. It shows too... While here, remove the duplicate definition of the hw.model sysctl. Revision Changes Path 1.2 +28 -3 src/sys/powerpc/booke/machdep.c From rafan at FreeBSD.org Sat Mar 8 05:42:09 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Sat Mar 8 05:42:12 2008 Subject: cvs commit: src/share/termcap termcap.src Message-ID: <200803080542.m285g9Co047824@repoman.freebsd.org> rafan 2008-03-08 05:42:09 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/termcap termcap.src Log: MFC termcap changes for the END/ENTER keys Log: - Remove kH and *6 from xterm. They are defined to the same key as @7 (kp_end). As ncurses has the limitation that it returns the first matched key symbol, you can not use END in ncurses based program, like mutt, with xterm. - Add @8 (kp_enter) definition for xterm so you can use ENTER in xterm with ncurses based program. I also found that NetBSD's xterm does the same thing. PR: 100150 Reported by: Arseny Nasokin Discussed with: Thomas Dickey, Ulrich Spoerlein Reviewed by: freebsd-arch@ Revision Changes Path 1.148.2.1 +2 -2 src/share/termcap/termcap.src From rafan at FreeBSD.org Sat Mar 8 05:42:53 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Sat Mar 8 05:42:57 2008 Subject: cvs commit: src/share/termcap termcap.src Message-ID: <200803080542.m285gq2s047868@repoman.freebsd.org> rafan 2008-03-08 05:42:52 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/termcap termcap.src Log: MFC termcap changes for the END/ENTER keys Log: - Remove kH and *6 from xterm. They are defined to the same key as @7 (kp_end) As ncurses has the limitation that it returns the first matched key symbol, you can not use END in ncurses based program, like mutt, with xterm. - Add @8 (kp_enter) definition for xterm so you can use ENTER in xterm with ncurses based program. I also found that NetBSD's xterm does the same thing. PR: 100150 Reported by: Arseny Nasokin Discussed with: Thomas Dickey, Ulrich Spoerlein Reviewed by: freebsd-arch@ Revision Changes Path 1.144.8.1 +2 -2 src/share/termcap/termcap.src From peter at wemm.org Sat Mar 8 09:29:55 2008 From: peter at wemm.org (Peter Wemm) Date: Sat Mar 8 09:30:02 2008 Subject: cvs commit: src/sys/amd64/conf GENERIC src/sys/i386/conf GENERIC src/sys/powerpc/conf GENERIC In-Reply-To: <200803072309.m27N9AAI005293@repoman.freebsd.org> References: <200803072309.m27N9AAI005293@repoman.freebsd.org> Message-ID: On Fri, Mar 7, 2008 at 3:09 PM, Xin LI wrote: > delphij 2008-03-07 23:09:10 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7) > sys/amd64/conf GENERIC > sys/i386/conf GENERIC > sys/powerpc/conf GENERIC > Log: > MFC: Switch to use ULE as default scheduler on amd64, i386 and powerpc > architectures. You broke powerpc tinderbox: #if !defined(__i386__) && !defined(__amd64__) && !defined(__arm__) #error "This architecture is not currently compatible with ULE" #endif -- Peter Wemm - peter@wemm.org; peter@FreeBSD.org; peter@yahoo-inc.com "All of this is for nothing if we don't go to the stars" - JMS/B5 "If Java had true garbage collection, most programs would delete themselves upon execution." -- Robert Sewell From delphij at FreeBSD.org Sat Mar 8 09:38:19 2008 From: delphij at FreeBSD.org (Xin LI) Date: Sat Mar 8 09:38:20 2008 Subject: cvs commit: src/sys/powerpc/conf GENERIC Message-ID: <200803080938.m289cIYI069066@repoman.freebsd.org> delphij 2008-03-08 09:38:18 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/powerpc/conf GENERIC Log: Revert powerpc scheduler change for now. This change should be done after necessary ULE changes to be MFC'ed. Pointed out by: peter Revision Changes Path 1.71.2.4 +1 -1 src/sys/powerpc/conf/GENERIC From delphij at delphij.net Sat Mar 8 09:40:27 2008 From: delphij at delphij.net (LI Xin) Date: Sat Mar 8 09:40:30 2008 Subject: cvs commit: src/sys/amd64/conf GENERIC src/sys/i386/conf GENERIC src/sys/powerpc/conf GENERIC In-Reply-To: References: <200803072309.m27N9AAI005293@repoman.freebsd.org> Message-ID: <47D25EC9.5090704@delphij.net> Peter Wemm wrote: > On Fri, Mar 7, 2008 at 3:09 PM, Xin LI wrote: >> delphij 2008-03-07 23:09:10 UTC >> >> FreeBSD src repository >> >> Modified files: (Branch: RELENG_7) >> sys/amd64/conf GENERIC >> sys/i386/conf GENERIC >> sys/powerpc/conf GENERIC >> Log: >> MFC: Switch to use ULE as default scheduler on amd64, i386 and powerpc >> architectures. > > You broke powerpc tinderbox: > > #if !defined(__i386__) && !defined(__amd64__) && !defined(__arm__) > #error "This architecture is not currently compatible with ULE" > #endif Sorry for that, I have backed out the change. Cheers, -- Xin LI http://www.delphij.net/ FreeBSD - The Power to Serve! -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 249 bytes Desc: OpenPGP digital signature Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080308/9484b73d/signature-0001.pgp From raj at semihalf.com Sat Mar 8 10:30:32 2008 From: raj at semihalf.com (Rafal Jaworowski) Date: Sat Mar 8 10:30:37 2008 Subject: cvs commit: src/sys/powerpc/booke machdep.c In-Reply-To: <200803080536.m285aPN7047604@repoman.freebsd.org> References: <200803080536.m285aPN7047604@repoman.freebsd.org> Message-ID: <47D2681F.2070309@semihalf.com> Marcel Moolenaar wrote: > marcel 2008-03-08 05:36:25 UTC > > FreeBSD src repository > > Modified files: > sys/powerpc/booke machdep.c > Log: > Enable the D-cache and I-cache when not already enabled. > It so happens that U-Boot disables the D-cache when booting > an ELF image, so this change makes sure we run with the > D-cache enabled from now on. It shows too... > Thanks for hooking this up, Marcel; it was [far] on my list too :) BTW: currently U-Boot is broken for most of PPC variations (including 85xx) in the following way: d-cache is not being flushed before disabling and strange things happen as you imagine. For a quick workaround apply something like this when using ELF images: http://people.freebsd.org/~raj/patches/misc/uboot-dcache-workaround.diff Rafal From brueffer at FreeBSD.org Sat Mar 8 12:13:02 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sat Mar 8 12:13:05 2008 Subject: cvs commit: src/sbin/geom/class/multipath gmultipath.8 Message-ID: <200803081213.m28CD0mI083279@repoman.freebsd.org> brueffer 2008-03-08 12:13:00 UTC FreeBSD src repository Modified files: sbin/geom/class/multipath gmultipath.8 Log: Fix typos. PR: 121486 Submitted by: Anatoly Borodin MFC after: 3 days Revision Changes Path 1.5 +2 -2 src/sbin/geom/class/multipath/gmultipath.8 From rpaulo at FreeBSD.org Sat Mar 8 14:12:36 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Sat Mar 8 14:12:40 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803081412.m28ECZqO008903@repoman.freebsd.org> rpaulo 2008-03-08 14:12:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/i386/cpufreq est.c Log: MFC r1.12: Validate the id16 values gathered from ACPI (previously a TODO item). Style changes by me and njl. Submitted by: Takeharu KATO PR: 119350 Revision Changes Path 1.11.2.1 +59 -17 src/sys/i386/cpufreq/est.c From rpaulo at FreeBSD.org Sat Mar 8 14:14:35 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Sat Mar 8 14:14:42 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803081414.m28EEZtK008977@repoman.freebsd.org> rpaulo 2008-03-08 14:14:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/i386/cpufreq est.c Log: MFC r1.12: Validate the id16 values gathered from ACPI (previously a TODO item). Style changes by me and njl. Submitted by: Takeharu KATO PR: 119350 Revision Changes Path 1.7.2.2 +59 -17 src/sys/i386/cpufreq/est.c From wearabnet at yahoo.ca Sat Mar 8 14:21:12 2008 From: wearabnet at yahoo.ca (Abdullah Ibn Hamad Al-Marri) Date: Sat Mar 8 14:21:17 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <943547.71624.qm@web33701.mail.mud.yahoo.com> ----- Original Message ---- > From: Rui Paulo > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; cvs-all@FreeBSD.org > Sent: Saturday, March 8, 2008 5:12:35 PM > Subject: cvs commit: src/sys/i386/cpufreq est.c > > rpaulo 2008-03-08 14:12:35 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7) > sys/i386/cpufreq est.c > Log: > MFC r1.12: > Validate the id16 values gathered from ACPI (previously a TODO > item). > Style changes by me and njl. > > Submitted by: Takeharu KATO > PR: 119350 > > Revision Changes Path > 1.11.2.1 +59 -17 src/sys/i386/cpufreq/est.c Thanks, how about AMD64? Regards, -Abdullah Ibn Hamad Al-Marri Arab Portal http://www.WeArab.Net/ ____________________________________________________________________________________ Looking for last minute shopping deals? Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping From rpaulo at FreeBSD.org Sat Mar 8 15:38:11 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Sat Mar 8 15:38:44 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <943547.71624.qm@web33701.mail.mud.yahoo.com> References: <943547.71624.qm@web33701.mail.mud.yahoo.com> Message-ID: On Sat, Mar 8, 2008 at 2:21 PM, Abdullah Ibn Hamad Al-Marri wrote: > > ----- Original Message ---- > > > From: Rui Paulo > > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; cvs-all@FreeBSD.org > > Sent: Saturday, March 8, 2008 5:12:35 PM > > Subject: cvs commit: src/sys/i386/cpufreq est.c > > > > rpaulo 2008-03-08 14:12:35 UTC > > > > FreeBSD src repository > > > > Modified files: (Branch: RELENG_7) > > > sys/i386/cpufreq est.c > > Log: > > MFC r1.12: > > Validate the id16 values gathered from ACPI (previously a TODO > > item). > > Style changes by me and njl. > > > > Submitted by: Takeharu KATO > > PR: 119350 > > > > Revision Changes Path > > 1.11.2.1 +59 -17 src/sys/i386/cpufreq/est.c > > Thanks, how about AMD64? This directory is shared by i386 and amd64. Regards. -- Rui Paulo From remko at FreeBSD.org Sat Mar 8 15:41:07 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Sat Mar 8 15:41:10 2008 Subject: cvs commit: src/tools/tools/tinybsd tinybsd Message-ID: <200803081541.m28Ff6JM012925@repoman.freebsd.org> remko 2008-03-08 15:41:06 UTC FreeBSD src repository Modified files: tools/tools/tinybsd tinybsd Log: Improve handling of the tinybsd application wrt ports. Ports that have dependencies or need to install a new user/group, are not problematic anymore. PR: 121367 Submitted by: Richard Arends < richard at unixguru dot nl > MFC after: 3 days Revision Changes Path 1.8 +41 -53 src/tools/tools/tinybsd/tinybsd From remko at FreeBSD.org Sat Mar 8 15:42:24 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Sat Mar 8 15:42:31 2008 Subject: cvs commit: src/tools/tools/tinybsd tinybsd In-Reply-To: <200803081541.m28Ff6JM012925@repoman.freebsd.org> References: <200803081541.m28Ff6JM012925@repoman.freebsd.org> Message-ID: <47D2B3DF.5000409@FreeBSD.org> Remko Lodder wrote: > remko 2008-03-08 15:41:06 UTC > > FreeBSD src repository > > Modified files: > tools/tools/tinybsd tinybsd > Log: > Improve handling of the tinybsd application wrt ports. Ports that > have dependencies or need to install a new user/group, are not > problematic anymore. > > PR: 121367 > Submitted by: Richard Arends < richard at unixguru dot nl > > MFC after: 3 days Approved by: imp (mentor, implicit for trivial changes) -- /"\ Best regards, | remko@FreeBSD.org \ / Remko Lodder | remko@EFnet X http://www.evilcoder.org/ | / \ ASCII Ribbon Campaign | Against HTML Mail and News From rwatson at FreeBSD.org Sat Mar 8 15:48:06 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 8 15:48:09 2008 Subject: cvs commit: src/sys/kern kern_sig.c Message-ID: <200803081548.m28Fm6L2013117@repoman.freebsd.org> rwatson 2008-03-08 15:48:06 UTC FreeBSD src repository Modified files: sys/kern kern_sig.c Log: Unlock the process lock when expand_name() fails, or we may leak the process lock leading to a hang. This bug was introduced in kern_sig.c:1.351, when the call to expand_name() was moved earlier bit this particular error case was not updated. Revision Changes Path 1.356 +1 -0 src/sys/kern/kern_sig.c From rwatson at FreeBSD.org Sat Mar 8 16:31:32 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 8 16:31:35 2008 Subject: cvs commit: src/sys/kern kern_sig.c Message-ID: <200803081631.m28GVVZd015835@repoman.freebsd.org> rwatson 2008-03-08 16:31:29 UTC FreeBSD src repository Modified files: sys/kern kern_sig.c Log: Use sbuf routines to construct core dump filenames rather than custom string buffer handling, making the code both easier to read and more robust against string-handling bugs. MFC after: 1 week Revision Changes Path 1.357 +22 -27 src/sys/kern/kern_sig.c From wearabnet at yahoo.ca Sat Mar 8 16:33:14 2008 From: wearabnet at yahoo.ca (Abdullah Ibn Hamad Al-Marri) Date: Sat Mar 8 16:33:19 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <484742.73871.qm@web33703.mail.mud.yahoo.com> ----- Original Message ---- > From: Rui Paulo > To: Abdullah Ibn Hamad Al-Marri > Cc: src-committers@freebsd.org; cvs-src@freebsd.org; cvs-all@freebsd.org > Sent: Saturday, March 8, 2008 6:38:10 PM > Subject: Re: cvs commit: src/sys/i386/cpufreq est.c > > On Sat, Mar 8, 2008 at 2:21 PM, Abdullah Ibn Hamad Al-Marri > wrote: > > > > ----- Original Message ---- > > > > > From: Rui Paulo > > > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; cvs-all@FreeBSD.org > > > Sent: Saturday, March 8, 2008 5:12:35 PM > > > Subject: cvs commit: src/sys/i386/cpufreq est.c > > > > > > rpaulo 2008-03-08 14:12:35 UTC > > > > > > FreeBSD src repository > > > > > > Modified files: (Branch: RELENG_7) > > > > > sys/i386/cpufreq est.c > > > Log: > > > MFC r1.12: > > > Validate the id16 values gathered from ACPI (previously a TODO > > > item). > > > Style changes by me and njl. > > > > > > Submitted by: Takeharu KATO > > > PR: 119350 > > > > > > Revision Changes Path > > > 1.11.2.1 +59 -17 src/sys/i386/cpufreq/est.c > > > > Thanks, how about AMD64? > > This directory is shared by i386 and amd64. > > Regards. > -- > Rui Paulo > Thank you, Here what I got. FreeBSD 7.0-STABLE #0: Sat Mar 8 16:16:54 UTC 2008 arabian@MX1.WeArab.Net:/usr/obj/usr/src/sys/MX1 Timecounter "i8254" frequency 1193182 Hz quality 0 CPU: Intel(R) Core(TM)2 Duo CPU E6750 @ 2.66GHz (2666.62-MHz K8-class CPU) Origin = "GenuineIntel" Id = 0x6fb Stepping = 11 Features=0xbfebfbff Features2=0xe3fd AMD Features=0x20100800 AMD Features2=0x1 Cores per package: 2 usable memory = 4276043776 (4077 MB) avail memory = 4106977280 (3916 MB) ACPI APIC Table: FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs cpu0 (BSP): APIC ID: 0 cpu1 (AP): APIC ID: 1 ioapic0 irqs 0-23 on motherboard acpi0: on motherboard acpi0: [ITHREAD] acpi0: Power Button (fixed) Timecounter "ACPI-fast" frequency 3579545 Hz quality 1000 acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 acpi_hpet0: iomem 0xfed00000-0xfed003ff on acpi0 Timecounter "HPET" frequency 14318180 Hz quality 900 cpu0: on acpi0 est0: on cpu0 est: CPU supports Enhanced Speedstep, but is not recognized. est: cpu_vendor GenuineIntel, msr 82a082a0600082a device_attach: est0 attach returned 6 p4tcc0: on cpu0 cpu1: on acpi0 est1: on cpu1 est: CPU supports Enhanced Speedstep, but is not recognized. est: cpu_vendor GenuineIntel, msr 82a082a0600082a device_attach: est1 attach returned 6 p4tcc1: on cpu1 Regards, -Abdullah Ibn Hamad Al-Marri Arab Portal http://www.WeArab.Net/ ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs From csjp at FreeBSD.org Sat Mar 8 16:51:47 2008 From: csjp at FreeBSD.org (Christian S.J. Peron) Date: Sat Mar 8 16:51:51 2008 Subject: cvs commit: src/sys/kern kern_sig.c In-Reply-To: <200803081548.m28Fm6L2013117@repoman.freebsd.org> References: <200803081548.m28Fm6L2013117@repoman.freebsd.org> Message-ID: <20080308165146.GA29675@sub.vaned.net> On Sat, Mar 08, 2008 at 03:48:06PM +0000, Robert Watson wrote: > rwatson 2008-03-08 15:48:06 UTC > > FreeBSD src repository > > Modified files: > sys/kern kern_sig.c > Log: > Unlock the process lock when expand_name() fails, or we may leak the > process lock leading to a hang. This bug was introduced in > kern_sig.c:1.351, when the call to expand_name() was moved earlier > bit this particular error case was not updated. > Pointy hat to: csjp Thanks for catching this! From bz at FreeBSD.org Sat Mar 8 16:58:22 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 16:58:24 2008 Subject: cvs commit: src/sys/netipsec key.c key_debug.c Message-ID: <200803081658.m28GwLKt025769@repoman.freebsd.org> bz 2008-03-08 16:58:21 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netipsec key.c key_debug.c Log: MFC: rev. 1.29 key.c, 1.6 key_debug.c Fix bugs when allocating and passing information of current lifetime and soft lifetime [1] introduced in rev. 1.21 of key.c. Along with that, fix a related problem in key_debug printing the correct data. While there replace a printf by panic in a sanity check. PR: 120751 Submitted by: Kazuaki ODA (kazuaki aliceblue.jp) [1] Revision Changes Path 1.28.2.1 +7 -2 src/sys/netipsec/key.c 1.5.2.1 +22 -4 src/sys/netipsec/key_debug.c From bz at FreeBSD.org Sat Mar 8 17:01:27 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 17:01:30 2008 Subject: cvs commit: src/sys/netinet tcp_input.c Message-ID: <200803081701.m28H1RaN025974@repoman.freebsd.org> bz 2008-03-08 17:01:27 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet tcp_input.c Log: MFC: rev. 1.372 Fix indentation (whitespace changes only). rev. 1.373 Some "cleanup" of tcp_mss(): - Move the assigment of the socket down before we first need it. No need to do it at the beginning and then drop out the function by one of the returns before using it 100 lines further down. - Use t_maxopd which was assigned the "tcp_mssdflt" for the corrrect AF already instead of another #ifdef ? : #endif block doing the same. - Remove an unneeded (duplicate) assignment of mss to t_maxseg just before we possibly change mss and re-do the assignment without using t_maxseg in between. Revision Changes Path 1.370.2.1 +8 -12 src/sys/netinet/tcp_input.c From scottl at FreeBSD.org Sat Mar 8 18:06:49 2008 From: scottl at FreeBSD.org (Scott Long) Date: Sat Mar 8 18:06:53 2008 Subject: cvs commit: src/sys/dev/hptrr hptrr_os_bsd.c hptrr_osm_bsd.c ldm.h osm.h Message-ID: <200803081806.m28I6m3K033883@repoman.freebsd.org> scottl 2008-03-08 18:06:48 UTC FreeBSD src repository Modified files: sys/dev/hptrr hptrr_os_bsd.c hptrr_osm_bsd.c ldm.h osm.h Log: Fix a mistake made during the import of the driver. Previous versions of HPT drivers would sometimes test the value of a preprocessor definition but not always make sure that the definition existed in the first place, leading to warnings on newer compilers. I blindly assumed the same with this driver, and it turned out to be wrong and to enable some code that doesn't work. Revision Changes Path 1.4 +3 -3 src/sys/dev/hptrr/hptrr_os_bsd.c 1.4 +2 -2 src/sys/dev/hptrr/hptrr_osm_bsd.c 1.3 +1 -1 src/sys/dev/hptrr/ldm.h 1.4 +1 -1 src/sys/dev/hptrr/osm.h From scottl at FreeBSD.org Sat Mar 8 18:16:33 2008 From: scottl at FreeBSD.org (Scott Long) Date: Sat Mar 8 18:16:35 2008 Subject: cvs commit: src/sys/dev/hptrr amd64-elf.hptrr_lib.o.uu array.h him.h himfuncs.h hptintf.h hptrr_config.c hptrr_config.h hptrr_os_bsd.c hptrr_osm_bsd.c i386-elf.hptrr_lib.o.uu ldm.h list.h os_bsd.h osm.h Message-ID: <200803081816.m28IGXY8034513@repoman.freebsd.org> scottl 2008-03-08 18:16:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/hptrr amd64-elf.hptrr_lib.o.uu array.h him.h himfuncs.h hptintf.h hptrr_config.c hptrr_config.h hptrr_os_bsd.c hptrr_osm_bsd.c i386-elf.hptrr_lib.o.uu ldm.h list.h os_bsd.h osm.h Log: Re-update to rev 1.2 of the hptrr driver. Revision Changes Path 1.1.2.4 +13261 -12830 src/sys/dev/hptrr/amd64-elf.hptrr_lib.o.uu 1.1.2.4 +8 -5 src/sys/dev/hptrr/array.h 1.1.2.4 +134 -117 src/sys/dev/hptrr/him.h 1.1.2.3 +0 -0 src/sys/dev/hptrr/himfuncs.h 1.1.2.4 +76 -20 src/sys/dev/hptrr/hptintf.h 1.1.2.4 +5 -2 src/sys/dev/hptrr/hptrr_config.c 1.1.2.4 +8 -3 src/sys/dev/hptrr/hptrr_config.h 1.1.2.5 +4 -4 src/sys/dev/hptrr/hptrr_os_bsd.c 1.1.2.4 +9 -10 src/sys/dev/hptrr/hptrr_osm_bsd.c 1.1.2.4 +8768 -8363 src/sys/dev/hptrr/i386-elf.hptrr_lib.o.uu 1.1.2.4 +57 -32 src/sys/dev/hptrr/ldm.h 1.1.2.3 +0 -0 src/sys/dev/hptrr/list.h 1.1.2.4 +4 -0 src/sys/dev/hptrr/os_bsd.h 1.1.2.5 +11 -9 src/sys/dev/hptrr/osm.h From bz at FreeBSD.org Sat Mar 8 18:17:26 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:17:27 2008 Subject: cvs commit: src/sys/netipsec ipsec.c Message-ID: <200803081817.m28IHP7V034584@repoman.freebsd.org> bz 2008-03-08 18:17:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec ipsec.c Log: "MFC" (code is different) rev. 1.20 ipsec.c s,#if INET*,#ifdef INET*, Revision Changes Path 1.12.2.1 +2 -2 src/sys/netipsec/ipsec.c From bz at FreeBSD.org Sat Mar 8 18:19:50 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:19:54 2008 Subject: cvs commit: src/sys/netipsec key.c Message-ID: <200803081819.m28IJoYa034657@repoman.freebsd.org> bz 2008-03-08 18:19:50 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec key.c Log: MFC rev. 1.23 key.c Add missing break; so when comparing AF_INET6 addresses, scope and ports we do not run into the default case and return 'no match' instead of 'match'. Revision Changes Path 1.20.2.2 +1 -0 src/sys/netipsec/key.c From bz at FreeBSD.org Sat Mar 8 18:22:21 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:22:23 2008 Subject: cvs commit: src/sys/netipsec ipsec.c Message-ID: <200803081822.m28IMLFU034836@repoman.freebsd.org> bz 2008-03-08 18:22:20 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec ipsec.c Log: MFC rev. 1.23 ipsec.c Implement ICMPv6 support in ipsec6_get_ulp(). This is needed to make security policies work correctly if ICMPv6 type and/or code are given. See setkey(8) 'upperspec' para. for details. Revision Changes Path 1.12.2.2 +10 -0 src/sys/netipsec/ipsec.c From bz at FreeBSD.org Sat Mar 8 18:23:56 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:24:02 2008 Subject: cvs commit: src/sys/netipsec ipsec_output.c Message-ID: <200803081823.m28INtfS034875@repoman.freebsd.org> bz 2008-03-08 18:23:55 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec ipsec_output.c Log: MFC rev. 1.13 ipsec_output.c fix typo: s,applyed,applied,g Revision Changes Path 1.10.8.2 +2 -2 src/sys/netipsec/ipsec_output.c From bz at FreeBSD.org Sat Mar 8 18:25:58 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:26:01 2008 Subject: cvs commit: src/sys/netipsec ipsec_input.c Message-ID: <200803081825.m28IPwru034978@repoman.freebsd.org> bz 2008-03-08 18:25:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec ipsec_input.c Log: MFC rev. 1.15 ipsec_input.c Though we are only called for the three security protocols we can handle, document those sprotos using an IPSEC_ASSERT so that it will be clear that 'spi' will always be initialized when used the first time. Found with: Coverity Prevent(tm) CID: 2533 Revision Changes Path 1.9.2.3 +4 -0 src/sys/netipsec/ipsec_input.c From bz at FreeBSD.org Sat Mar 8 18:27:28 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:27:29 2008 Subject: cvs commit: src/sys/netipsec ipsec_input.c Message-ID: <200803081827.m28IRRj8035077@repoman.freebsd.org> bz 2008-03-08 18:27:27 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec ipsec_input.c Log: MFC rev. 1.16 ipsec_input.c s,#,*, in a multi-line comment. This is C. No functional change. Revision Changes Path 1.9.2.4 +1 -1 src/sys/netipsec/ipsec_input.c From bz at FreeBSD.org Sat Mar 8 18:29:39 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:29:42 2008 Subject: cvs commit: src/sys/netipsec ipsec_input.c Message-ID: <200803081829.m28ITdZo035157@repoman.freebsd.org> bz 2008-03-08 18:29:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec ipsec_input.c Log: MFC rev. 1.17 ipsec_input.c Looking at {ah,esp}_input_cb it seems we might be able to end up without an mtag in ipsec4_common_input_cb. So in case of !IPCOMP (AH,ESP) only change the m_tag_id if an mtag was passed to ipsec4_common_input_cb. Found with: Coverity Prevent(tm) CID: 2523 Revision Changes Path 1.9.2.5 +1 -1 src/sys/netipsec/ipsec_input.c From bz at FreeBSD.org Sat Mar 8 18:34:18 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:34:20 2008 Subject: cvs commit: src/sys/netipsec key.c Message-ID: <200803081834.m28IYHB6035877@repoman.freebsd.org> bz 2008-03-08 18:34:17 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec key.c Log: MFC rev. 1.25 Add a missing return so that we drop out in case of an error and do not continue with a NULL pointer. [1] While here change the return of the error handling code path above. I cannot see why we should always return 0 there. Neither does KAME nor do we in here for the similar check in all the other functions. Found with: Coverity Prevent(tm) [1] CID: 2521 Revision Changes Path 1.20.2.3 +2 -3 src/sys/netipsec/key.c From bz at FreeBSD.org Sat Mar 8 18:37:03 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:37:05 2008 Subject: cvs commit: src/sys/netipsec key.c Message-ID: <200803081837.m28Ib3C3036414@repoman.freebsd.org> bz 2008-03-08 18:37:03 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec key.c Log: MFC rev. 1.26 key.c In case of failure we can directly return ENOBUFS because 'result' is still NULL and we do not need to free anything. That allows us to gc the entire goto parts and a now unused variable. Found with: Coverity Prevent(tm) CID: 2519 Revision Changes Path 1.20.2.4 +3 -10 src/sys/netipsec/key.c From bz at FreeBSD.org Sat Mar 8 18:38:27 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:38:30 2008 Subject: cvs commit: src/sys/netipsec key.c Message-ID: <200803081838.m28IcQfP036460@repoman.freebsd.org> bz 2008-03-08 18:38:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec key.c Log: MFC rev. 1.27 key.c 'spi' and the return value of ntohl are unsigned. Remove the extra >=0 check which was always true. Document the special meaning of spi values of 0 and 1-255 with a comment. Found with: Coverity Prevent(tm) CID: 2047 Revision Changes Path 1.20.2.5 +6 -1 src/sys/netipsec/key.c From bz at FreeBSD.org Sat Mar 8 18:55:08 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:55:13 2008 Subject: cvs commit: src/sys/netipsec key.c In-Reply-To: <200803081834.m28IYHB6035877@repoman.freebsd.org> References: <200803081834.m28IYHB6035877@repoman.freebsd.org> Message-ID: <20080308185230.Q50685@maildrop.int.zabbadoz.net> On Sat, 8 Mar 2008, Bjoern A. Zeeb wrote: > bz 2008-03-08 18:34:17 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_6) > sys/netipsec key.c > Log: > MFC rev. 1.25 > > Add a missing return so that we drop out in case of an error and > do not continue with a NULL pointer. [1] > > While here change the return of the error handling code path above. > I cannot see why we should always return 0 there. Neither does KAME > nor do we in here for the similar check in all the other functions. > > Found with: Coverity Prevent(tm) [1] > CID: 2521 > > Revision Changes Path > 1.20.2.3 +2 -3 src/sys/netipsec/key.c The problem (which was a missing MFC) was Reported by: Matthew Grooms mgrooms shrew.net PR: 121140 -- Bjoern A. Zeeb bzeeb at Zabbadoz dot NeT Software is harder than hardware so better get it right the first time. From bz at FreeBSD.org Sat Mar 8 18:56:28 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 18:56:32 2008 Subject: cvs commit: src/sys/netipsec ipsec_input.c ipsec_output.c xform_ipip.c Message-ID: <200803081856.m28IuMsD037735@repoman.freebsd.org> bz 2008-03-08 18:56:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netipsec ipsec_input.c ipsec_output.c xform_ipip.c Log: MFC rev.s 1.18 ipsec_input.c, 1.16 ipsec_output.c, 1.15 xform_ipip.c Replace hard coded options by their defined PFIL_{IN,OUT} names. Revision Changes Path 1.9.2.6 +2 -1 src/sys/netipsec/ipsec_input.c 1.10.8.3 +2 -1 src/sys/netipsec/ipsec_output.c 1.11.2.3 +2 -1 src/sys/netipsec/xform_ipip.c From xcllnt at mac.com Sat Mar 8 18:59:48 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Sat Mar 8 18:59:53 2008 Subject: cvs commit: src/sys/powerpc/booke machdep.c In-Reply-To: <47D2681F.2070309@semihalf.com> References: <200803080536.m285aPN7047604@repoman.freebsd.org> <47D2681F.2070309@semihalf.com> Message-ID: <82F5E125-8914-46CB-8522-8FA39F05DC4A@mac.com> On Mar 8, 2008, at 2:19 AM, Rafal Jaworowski wrote: > Marcel Moolenaar wrote: >> marcel 2008-03-08 05:36:25 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/powerpc/booke machdep.c >> Log: >> Enable the D-cache and I-cache when not already enabled. >> It so happens that U-Boot disables the D-cache when booting >> an ELF image, so this change makes sure we run with the >> D-cache enabled from now on. It shows too... >> > > Thanks for hooking this up, Marcel; it was [far] on my list too :) I just wanted my kernel builds to be faster :-) > BTW: currently U-Boot is broken for most of PPC variations > (including 85xx) in > the following way: d-cache is not being flushed before disabling and > strange > things happen as you imagine. For a quick workaround apply something > like this > when using ELF images: > http://people.freebsd.org/~raj/patches/misc/uboot-dcache-workaround.diff Yup. That's what we do at Juniper. I don't see any bad effects other than having to invalidate the cache before (re-)enabling it so I don't have that patch applied myself... The problem I'm having with U-Boot is with the TSEC driver as you've seen on the U-Boot mailing list. I have one MPC8555CDS sitting idle after the U-Boot upgrade because I can't netboot :-/ -- Marcel Moolenaar xcllnt@mac.com From obrien at FreeBSD.org Sat Mar 8 19:14:44 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sat Mar 8 19:14:46 2008 Subject: cvs commit: src/sys/modules/zfs Makefile Message-ID: <200803081914.m28JEhtd039746@repoman.freebsd.org> obrien 2008-03-08 19:14:43 UTC FreeBSD src repository Modified files: sys/modules/zfs Makefile Log: "root" the include path so there is less duplication. Revision Changes Path 1.9 +18 -16 src/sys/modules/zfs/Makefile From raj at semihalf.com Sat Mar 8 19:21:31 2008 From: raj at semihalf.com (Rafal Jaworowski) Date: Sat Mar 8 19:21:37 2008 Subject: cvs commit: src/sys/powerpc/booke machdep.c In-Reply-To: <82F5E125-8914-46CB-8522-8FA39F05DC4A@mac.com> References: <200803080536.m285aPN7047604@repoman.freebsd.org> <47D2681F.2070309@semihalf.com> <82F5E125-8914-46CB-8522-8FA39F05DC4A@mac.com> Message-ID: <47D2E72E.1000107@semihalf.com> Marcel Moolenaar wrote: >> >> Thanks for hooking this up, Marcel; it was [far] on my list too :) > > I just wanted my kernel builds to be faster :-) > Ah, self-hosted builds you mean? Over NFS, right? (this is how we usually test this) >> BTW: currently U-Boot is broken for most of PPC variations (including >> 85xx) in >> the following way: d-cache is not being flushed before disabling and >> strange >> things happen as you imagine. For a quick workaround apply something >> like this >> when using ELF images: >> http://people.freebsd.org/~raj/patches/misc/uboot-dcache-workaround.diff > > Yup. That's what we do at Juniper. I don't see any bad effects other > than having to invalidate the cache before (re-)enabling it so I don't > have that patch applied myself... > I have seen strange things, including hangs, and also TSEC non-functional effects due to this issue... :) > The problem I'm having with U-Boot is with the TSEC driver as you've > seen on the U-Boot mailing list. I have one MPC8555CDS sitting idle > after the U-Boot upgrade because I can't netboot :-/ > Yes, I saw. Let me re-flash my 8555 with the latest image and try, but when I'm not remote i.e. Monday. Don't know if you know this tip, but there used to be a second flash bank on the MPC8555CDS that you can access while running from the boot one, so you can recover in such scenario without a h/w debugger by copying over that backup flash bank, provided you have a working image there... Rafal From marius at FreeBSD.org Sat Mar 8 19:33:35 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sat Mar 8 19:33:37 2008 Subject: cvs commit: src/contrib/gcc/config/sparc freebsd.h Message-ID: <200803081933.m28JXZ0d046051@repoman.freebsd.org> marius 2008-03-08 19:33:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) contrib/gcc/config/sparc freebsd.h Log: MFC: 1.11 Allow the targeted CPU type to be set via -mcpu instead of limiting to the hardcoded equivalent of -mcpu=ultrasparc. PR: 109510 Revision Changes Path 1.10.2.1 +8 -6 src/contrib/gcc/config/sparc/freebsd.h From antoine at FreeBSD.org Sat Mar 8 21:56:00 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Sat Mar 8 21:56:06 2008 Subject: cvs commit: src/lib/libutil humanize_number.c Message-ID: <200803082156.m28Lu0cQ070344@repoman.freebsd.org> antoine 2008-03-08 21:55:59 UTC FreeBSD src repository Modified files: lib/libutil humanize_number.c Log: Merge changes from NetBSD on humanize_number.c, 1.8 -> 1.13 Significant changes: - rev. 1.11: Use PRId64 instead of a cast to long long and %lld to print an int64_t. - rev. 1.12: Fix a bug that humanize_number() produces "1000" where it should be "1.0G" or "1.0M". The bug reported by Greg Troxel. PR: 118461 PR: 102694 Approved by: rwatson (mentor) Obtained from: NetBSD MFC after: 1 month Revision Changes Path 1.3 +10 -5 src/lib/libutil/humanize_number.c From antoine at FreeBSD.org Sat Mar 8 21:57:36 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Sat Mar 8 21:57:39 2008 Subject: cvs commit: src/tools/build/mk OptionalObsoleteFiles.inc Message-ID: <200803082157.m28LvZCq070411@repoman.freebsd.org> antoine 2008-03-08 21:57:35 UTC FreeBSD src repository Modified files: tools/build/mk OptionalObsoleteFiles.inc Log: Add some checks for amd64 target around usr/lib32 libraries, missed in previous commit. Approved by: rwatson (mentor) X-MFC: with revision 1.13 Revision Changes Path 1.14 +4 -0 src/tools/build/mk/OptionalObsoleteFiles.inc From antoine at FreeBSD.org Sat Mar 8 21:59:44 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Sat Mar 8 21:59:46 2008 Subject: cvs commit: src ObsoleteFiles.inc Message-ID: <200803082159.m28LxidP070488@repoman.freebsd.org> antoine 2008-03-08 21:59:44 UTC FreeBSD src repository Modified files: . ObsoleteFiles.inc Log: Add some obsolete usr/lib32 libraries. PR: 117189 Submitted by: Scot Hetzel Approved by: rwatson (mentor) MFC after: 1 month Revision Changes Path 1.132 +50 -0 src/ObsoleteFiles.inc From antoine at FreeBSD.org Sat Mar 8 22:02:22 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Sat Mar 8 22:02:23 2008 Subject: cvs commit: src/lib/libc/sys fcntl.2 src/sys/kern kern_descrip.c src/sys/sys fcntl.h src/tools/regression/file/dup dup.c Message-ID: <200803082202.m28M2Lxu070689@repoman.freebsd.org> antoine 2008-03-08 22:02:21 UTC FreeBSD src repository Modified files: lib/libc/sys fcntl.2 sys/kern kern_descrip.c sys/sys fcntl.h tools/regression/file/dup dup.c Log: Introduce a new F_DUP2FD command to fcntl(2), for compatibility with Solaris and AIX. fcntl(fd, F_DUP2FD, arg) and dup2(fd, arg) are functionnaly equivalent. Document it. Add some regression tests (identical to the dup2(2) regression tests). PR: 120233 Submitted by: Jukka Ukkonen Approved by: rwaston (mentor) MFC after: 1 month Revision Changes Path 1.46 +37 -1 src/lib/libc/sys/fcntl.2 1.325 +6 -1 src/sys/kern/kern_descrip.c 1.18 +1 -0 src/sys/sys/fcntl.h 1.3 +71 -2 src/tools/regression/file/dup/dup.c From antoine at FreeBSD.org Sat Mar 8 22:17:15 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Sat Mar 8 22:17:17 2008 Subject: cvs commit: src/sys/sys param.h Message-ID: <200803082217.m28MHFHD072326@repoman.freebsd.org> antoine 2008-03-08 22:17:14 UTC FreeBSD src repository Modified files: sys/sys param.h Log: Bump __FreeBSD_version for F_DUP2FD command to fcntl(2) Requested by: Craig Rodrigues Approved by: rwatson (mentor) Revision Changes Path 1.341 +1 -1 src/sys/sys/param.h From bz at FreeBSD.org Sat Mar 8 23:36:26 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 23:36:28 2008 Subject: cvs commit: src/sys/boot/i386/libi386 biosdisk.c Message-ID: <200803082336.m28NaPjo076007@repoman.freebsd.org> bz 2008-03-08 23:36:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/i386/libi386 biosdisk.c Log: MFC rev. 1.54 Though we are currently not interested in the EDD3 flag, Enhanced Disk Drive Specification Ver 3.0 defines that the version of extension in AH would be 30h. Correct the check for that to be >=30h instead of >3h. Revision Changes Path 1.48.2.5 +1 -1 src/sys/boot/i386/libi386/biosdisk.c From bz at FreeBSD.org Sat Mar 8 23:41:48 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 23:41:51 2008 Subject: cvs commit: src/sys/netinet in.c Message-ID: <200803082341.m28Nfmae076162@repoman.freebsd.org> bz 2008-03-08 23:41:47 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet in.c Log: MFC rev. 1.103 in.c Differentiate between addifaddr and delifaddr for the privilege check. Revision Changes Path 1.102.2.1 +2 -1 src/sys/netinet/in.c From bz at FreeBSD.org Sat Mar 8 23:45:40 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sat Mar 8 23:45:42 2008 Subject: cvs commit: src/sys/netinet6 in6_src.c Message-ID: <200803082345.m28NjdXI076351@repoman.freebsd.org> bz 2008-03-08 23:45:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet6 in6_src.c Log: MFC rev. 1.49 in6_src.c Correct the commented out debugging printf()s in REPLACE and NEXT macros. ip6_sprintf() needs a buffer as first argument these days. Revision Changes Path 1.46.2.1 +8 -2 src/sys/netinet6/in6_src.c From mav at FreeBSD.org Sat Mar 8 23:55:30 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sat Mar 8 23:55:34 2008 Subject: cvs commit: src/sys/netgraph ng_l2tp.c Message-ID: <200803082355.m28NtU6Y076702@repoman.freebsd.org> mav 2008-03-08 23:55:29 UTC FreeBSD src repository Modified files: sys/netgraph ng_l2tp.c Log: To avoid control data losses do not acknowledge recieving of control packet if netgraph reported error while delivering to destination. Reset 'next send' counter to the last requested by peer on ack timeout to resend all subsequest packets after lost one again without additional hints. Revision Changes Path 1.23 +44 -54 src/sys/netgraph/ng_l2tp.c From bz at FreeBSD.org Sun Mar 9 00:04:46 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Mar 9 00:04:48 2008 Subject: cvs commit: src/sys/kern uipc_debug.c Message-ID: <200803090004.m2904koY078111@repoman.freebsd.org> bz 2008-03-09 00:04:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern uipc_debug.c Log: MFC rev. 1.3 uipc_debug.c Add missing sb_sndptr* fields to db_print_sockbuf(). While here change %d to %u for u_ints. Revision Changes Path 1.2.2.1 +9 -5 src/sys/kern/uipc_debug.c From marcel at FreeBSD.org Sun Mar 9 02:29:22 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Mar 9 02:29:26 2008 Subject: cvs commit: src/sys/powerpc/mpc85xx ocpbus.c Message-ID: <200803090229.m292TKvL099172@repoman.freebsd.org> marcel 2008-03-09 02:29:19 UTC FreeBSD src repository Modified files: sys/powerpc/mpc85xx ocpbus.c Log: Don't use in32() and out32() when writing to the CCSRBAR. The in*() and out*() primitives should not be used, other than by ISA drivers. In this case they were used for memory-mapped I/O and were not even used in the spirit of the primitives. Revision Changes Path 1.4 +26 -9 src/sys/powerpc/mpc85xx/ocpbus.c From imp at FreeBSD.org Sun Mar 9 05:10:24 2008 From: imp at FreeBSD.org (Warner Losh) Date: Sun Mar 9 05:10:26 2008 Subject: cvs commit: src/sys/kern subr_bus.c src/sys/sys bus.h Message-ID: <200803090510.m295ANVj049932@repoman.freebsd.org> imp 2008-03-09 05:10:22 UTC FreeBSD src repository Modified files: sys/kern subr_bus.c sys/sys bus.h Log: Any driver that relies on its parent to set the devclass has no way to know if has siblings that need an actual probe. Introduce a specail return value called BUS_PROBE_NOOWILDCARD. If the driver returns this, the probe is only successful for devices that have had a specific devclass set for them. Reviewed by: current@, jhb@, grehan@ Revision Changes Path 1.205 +8 -0 src/sys/kern/subr_bus.c 1.80 +5 -1 src/sys/sys/bus.h From imp at FreeBSD.org Sun Mar 9 05:31:33 2008 From: imp at FreeBSD.org (Warner Losh) Date: Sun Mar 9 05:31:36 2008 Subject: cvs commit: src/share/man/man9 DEVICE_PROBE.9 Message-ID: <200803090531.m295VWii050565@repoman.freebsd.org> imp 2008-03-09 05:31:32 UTC FreeBSD src repository Modified files: share/man/man9 DEVICE_PROBE.9 Log: Document all the BUS_PROBE_* values. Note to the mdoc gods: This may need some teaks. Revision Changes Path 1.22 +45 -1 src/share/man/man9/DEVICE_PROBE.9 From das at FreeBSD.ORG Sun Mar 9 06:04:46 2008 From: das at FreeBSD.ORG (David Schultz) Date: Sun Mar 9 06:04:53 2008 Subject: cvs commit: src/share/man/man9 DEVICE_PROBE.9 In-Reply-To: <200803090531.m295VWii050565@repoman.freebsd.org> References: <200803090531.m295VWii050565@repoman.freebsd.org> Message-ID: <20080309060633.GA68839@zim.MIT.EDU> On Sun, Mar 09, 2008, Warner Losh wrote: > imp 2008-03-09 05:31:32 UTC > > FreeBSD src repository > > Modified files: > share/man/man9 DEVICE_PROBE.9 > Log: > Document all the BUS_PROBE_* values. > > Note to the mdoc gods: This may need some teaks. ^^^^^ Woodn't be surprising. From alc at FreeBSD.org Sun Mar 9 06:08:58 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sun Mar 9 06:09:00 2008 Subject: cvs commit: src/sys/vm vm_fault.c Message-ID: <200803090608.m2968wWC052872@repoman.freebsd.org> alc 2008-03-09 06:08:58 UTC FreeBSD src repository Modified files: sys/vm vm_fault.c Log: Eliminate an unnecessary test from vm_fault's delete-behind heuristic. Specifically, since the delete-behind heuristic is never applied to a device-backed object, there is no point in checking whether each of the object's pages is fictitious. (Only device-backed objects have fictitious pages.) Revision Changes Path 1.240 +1 -1 src/sys/vm/vm_fault.c From edwin at FreeBSD.org Sun Mar 9 06:25:01 2008 From: edwin at FreeBSD.org (Edwin Groothuis) Date: Sun Mar 9 06:25:06 2008 Subject: cvs commit: src/share/zoneinfo - Imported sources Message-ID: <200803090625.m296P1BD053455@repoman.freebsd.org> edwin 2008-03-09 06:25:00 UTC FreeBSD src repository src/share/zoneinfo - Imported sources Update of /home/ncvs/src/share/zoneinfo In directory repoman.freebsd.org:/tmp/cvs-serv53413 Log Message: Vendor import of tzdata version 2008a Status: Vendor Tag: ADO Release Tags: tzdata2008a U src/share/zoneinfo/africa U src/share/zoneinfo/antarctica U src/share/zoneinfo/asia U src/share/zoneinfo/australasia U src/share/zoneinfo/europe U src/share/zoneinfo/northamerica C src/share/zoneinfo/southamerica U src/share/zoneinfo/pacificnew U src/share/zoneinfo/etcetera U src/share/zoneinfo/factory U src/share/zoneinfo/backward U src/share/zoneinfo/systemv U src/share/zoneinfo/solar87 U src/share/zoneinfo/solar88 U src/share/zoneinfo/solar89 C src/share/zoneinfo/iso3166.tab U src/share/zoneinfo/zone.tab C src/share/zoneinfo/leapseconds U src/share/zoneinfo/yearistype.sh 3 conflicts created by this import. Use the following command to help the merge: cvs checkout -jADO:yesterday -jADO src/share/zoneinfo From rwatson at FreeBSD.org Sun Mar 9 11:16:42 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 11:16:46 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c Message-ID: <200803091116.m29BGgcE078982@repoman.freebsd.org> rwatson 2008-03-09 11:16:42 UTC FreeBSD src repository Modified files: usr.bin/gzip gzip.c Log: When gzipping to a target file, only set the source file flags on the target file after the timestamp has been set; otherwise setting the timestamp will fail if the flags don't permit it (i.e., uchg). MFC after: 1 week PR: 120208 Submitted by: Ighighi Revision Changes Path 1.5 +4 -4 src/usr.bin/gzip/gzip.c From mav at FreeBSD.org Sun Mar 9 11:17:01 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 9 11:17:03 2008 Subject: cvs commit: src/sys/netgraph ng_l2tp.c Message-ID: <200803091117.m29BH0JC079017@repoman.freebsd.org> mav 2008-03-09 11:17:00 UTC FreeBSD src repository Modified files: sys/netgraph ng_l2tp.c Log: Addition to the previous commit. Release inproc in case of memory error. Revision Changes Path 1.24 +1 -0 src/sys/netgraph/ng_l2tp.c From rpaulo at FreeBSD.org Sun Mar 9 11:19:03 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Sun Mar 9 11:19:07 2008 Subject: cvs commit: src/sys/dev/acpica acpi_cpu.c Message-ID: <200803091119.m29BJ3OJ079115@repoman.freebsd.org> rpaulo 2008-03-09 11:19:03 UTC FreeBSD src repository Modified files: sys/dev/acpica acpi_cpu.c Log: Some PIIX4 chipsets need to be told to generate Stop Breaks by setting the appropriate bit in the DEVACTB register. This change allows the C2 state on those systems to work as expected. Reviewed by: njl Submitted by: Andriy Gapon MFC after: 1 week Revision Changes Path 1.71 +20 -0 src/sys/dev/acpica/acpi_cpu.c From edwin at FreeBSD.org Sun Mar 9 11:19:16 2008 From: edwin at FreeBSD.org (Edwin Groothuis) Date: Sun Mar 9 11:19:17 2008 Subject: cvs commit: src/share/zoneinfo leapseconds southamerica Message-ID: <200803091119.m29BJGYs079150@repoman.freebsd.org> edwin 2008-03-09 11:19:16 UTC FreeBSD src repository Modified files: share/zoneinfo leapseconds southamerica Log: MFV: tzdata2008a Notable changes: - Chilis DST has been extended for 3 weeks. - No leapsecond in 2008 Approved by: grog@ (mentor) Revision Changes Path 1.21 +7 -6 src/share/zoneinfo/leapseconds 1.34 +26 -2 src/share/zoneinfo/southamerica From edwin at FreeBSD.org Sun Mar 9 11:21:37 2008 From: edwin at FreeBSD.org (Edwin Groothuis) Date: Sun Mar 9 11:21:39 2008 Subject: cvs commit: src/share/zoneinfo leapseconds southamerica Message-ID: <200803091121.m29BLbFl079365@repoman.freebsd.org> edwin 2008-03-09 11:21:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/zoneinfo leapseconds southamerica Log: MFC: tzdata2008a Notable changes: - Chilis DST has been extended for 3 weeks. - No leapsecond in 2008 Approved by: grog@ (mentor) Revision Changes Path 1.19.2.1 +7 -6 src/share/zoneinfo/leapseconds 1.31.2.2 +26 -2 src/share/zoneinfo/southamerica From edwin at FreeBSD.org Sun Mar 9 11:22:31 2008 From: edwin at FreeBSD.org (Edwin Groothuis) Date: Sun Mar 9 11:22:34 2008 Subject: cvs commit: src/share/zoneinfo leapseconds southamerica Message-ID: <200803091122.m29BMUmQ079431@repoman.freebsd.org> edwin 2008-03-09 11:22:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/zoneinfo leapseconds southamerica Log: MFV: tzdata2008a Notable changes: - Chilis DST has been extended for 3 weeks. - No leapsecond in 2008 Approved by: grog@ (mentor) Revision Changes Path 1.13.2.5 +7 -6 src/share/zoneinfo/leapseconds 1.24.2.8 +26 -2 src/share/zoneinfo/southamerica From edwin at FreeBSD.org Sun Mar 9 11:23:09 2008 From: edwin at FreeBSD.org (Edwin Groothuis) Date: Sun Mar 9 11:23:13 2008 Subject: cvs commit: src/share/zoneinfo leapseconds southamerica Message-ID: <200803091123.m29BN8dE079488@repoman.freebsd.org> edwin 2008-03-09 11:23:08 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) share/zoneinfo leapseconds southamerica Log: MFV: tzdata2008a Notable changes: - Chilis DST has been extended for 3 weeks. - No leapsecond in 2008 Approved by: grog@ (mentor) Revision Changes Path 1.10.4.6 +7 -6 src/share/zoneinfo/leapseconds 1.22.4.11 +26 -2 src/share/zoneinfo/southamerica From rwatson at FreeBSD.org Sun Mar 9 11:30:00 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 11:30:02 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200803091129.m29BTxFj079650@repoman.freebsd.org> rwatson 2008-03-09 11:29:59 UTC FreeBSD src repository Modified files: sys/conf NOTES Log: HZ now defaults to 1000 on many architectures, so update NOTES to reflect that. MFC after: 3 days PR: 113670 Submitted by: Ighighi Revision Changes Path 1.1475 +9 -9 src/sys/conf/NOTES From rwatson at FreeBSD.org Sun Mar 9 12:10:25 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 12:10:29 2008 Subject: cvs commit: src/bin/chflags chflags.1 chflags.c Message-ID: <200803091210.m29CAPQ6082385@repoman.freebsd.org> rwatson 2008-03-09 12:10:24 UTC FreeBSD src repository Modified files: bin/chflags chflags.1 chflags.c Log: Make chflags(1) more chmod(1)-like (and more feature complete): - Add -v to print file names as they are processed; -vv prints the flags change as well. - Add -f to ignore failures with the same semantics as chflags(1), neither printing an error nor affecting the return code. - Don't try to set the flags if they won't change. I made minor cosmetic tweaks to the code in the patch. MFC after: 1 week PR: 112827 Submitted by: Ighighi Revision Changes Path 1.30 +15 -1 src/bin/chflags/chflags.1 1.24 +28 -16 src/bin/chflags/chflags.c From rwatson at FreeBSD.org Sun Mar 9 12:46:41 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 12:46:43 2008 Subject: cvs commit: src/bin/realpath realpath.1 realpath.c Message-ID: <200803091246.m29Ckdvs083549@repoman.freebsd.org> rwatson 2008-03-09 12:46:39 UTC FreeBSD src repository Modified files: bin/realpath realpath.1 realpath.c Log: Enhance realpath(1) in a number of ways: - Allow realpath to accept multiple paths on the command line. - Add -q to suppress warnings if some paths can't be processed, and use getopt(3) to process flags. - Print the path being requested rather than a possibly partially processed path when a failure occurs so that you can tell which of several passed paths did fail. MFC after: 1 week PR: 112920 Submitted by: Ighighi Revision Changes Path 1.11 +8 -0 src/bin/realpath/realpath.1 1.7 +26 -7 src/bin/realpath/realpath.c From bz at FreeBSD.org Sun Mar 9 13:26:51 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Mar 9 13:26:54 2008 Subject: cvs commit: src/sys/netinet tcp_output.c Message-ID: <200803091326.m29DQoCI095152@repoman.freebsd.org> bz 2008-03-09 13:26:50 UTC FreeBSD src repository Modified files: sys/netinet tcp_output.c Log: Padding after EOL option must be zeros according to RFC793 but the NOPs used are 0x01. While we could simply pad with EOLs (which are 0x00), rather use an explicit 0x00 constant there to not confuse poeple with 'EOL padding'. Put in a comment saying just that. Problem discussed on: src-committers with andre, silby, dwhite as follow up to the rev. 1.161 commit of tcp_var.h. MFC after: 11 days Revision Changes Path 1.146 +10 -2 src/sys/netinet/tcp_output.c From lippe at FreeBSD.org Sun Mar 9 13:33:45 2008 From: lippe at FreeBSD.org (Felippe de Meirelles Motta) Date: Sun Mar 9 13:33:47 2008 Subject: cvs commit: src/share/misc committers-ports.dot Message-ID: <200803091333.m29DXif0097783@repoman.freebsd.org> lippe 2008-03-09 13:33:44 UTC FreeBSD src repository (ports committer) Modified files: share/misc committers-ports.dot Log: - Add myself. Approved by: gabor (mentor) Revision Changes Path 1.68 +5 -0 src/share/misc/committers-ports.dot From rwatson at FreeBSD.org Sun Mar 9 13:49:10 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 13:49:12 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c Message-ID: <200803091349.m29Dn91v003746@repoman.freebsd.org> rwatson 2008-03-09 13:49:09 UTC FreeBSD src repository Modified files: usr.bin/gzip gzip.c Log: Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar to its existing understanding that .tgz uncompresses to .tar. MFC after: 3 days PR: 121519 Submitted by: Ighighi Revision Changes Path 1.6 +2 -0 src/usr.bin/gzip/gzip.c From kib at FreeBSD.org Sun Mar 9 13:55:52 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sun Mar 9 13:55:57 2008 Subject: cvs commit: src/sys/nfsserver nfs_serv.c Message-ID: <200803091355.m29Dtpw5006511@repoman.freebsd.org> kib 2008-03-09 13:55:51 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/nfsserver nfs_serv.c Log: MFC rev. 1.178: Fix the Giant leak in the nfsrv_remove(). Revision Changes Path 1.174.2.2 +6 -2 src/sys/nfsserver/nfs_serv.c From kib at FreeBSD.org Sun Mar 9 13:57:48 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sun Mar 9 13:57:52 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_vfsops.c Message-ID: <200803091357.m29DvmGe007341@repoman.freebsd.org> kib 2008-03-09 13:57:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ufs/ffs ffs_vfsops.c Log: MFC rev. 1.338: Initialize mnt_stat.f_iosize before autostarting UFS1 extattrs. Revision Changes Path 1.329.2.1 +1 -0 src/sys/ufs/ffs/ffs_vfsops.c From kib at FreeBSD.org Sun Mar 9 13:59:20 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sun Mar 9 13:59:22 2008 Subject: cvs commit: src/sys/compat/linux linux_futex.c Message-ID: <200803091359.m29DxJC6007935@repoman.freebsd.org> kib 2008-03-09 13:59:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/compat/linux linux_futex.c Log: MFC rev. 1.11: Return ENOSYS instead of 0 for the unknown futex operations. Revision Changes Path 1.10.2.1 +1 -1 src/sys/compat/linux/linux_futex.c From marius at FreeBSD.org Sun Mar 9 14:10:24 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Mar 9 14:10:29 2008 Subject: cvs commit: src/contrib/gcc/config/sparc freebsd.h Message-ID: <200803091410.m29EANvk011997@repoman.freebsd.org> marius 2008-03-09 14:10:23 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) contrib/gcc/config/sparc freebsd.h Log: MFC: 1.11 Allow the targeted CPU type to be set via -mcpu instead of limiting to the hardcoded equivalent of -mcpu=ultrasparc. PR: 109510 Revision Changes Path 1.9.8.1 +8 -6 src/contrib/gcc/config/sparc/freebsd.h From marius at FreeBSD.org Sun Mar 9 14:25:02 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Mar 9 14:25:05 2008 Subject: cvs commit: src/sys/conf files.sparc64 src/sys/sparc64/include md_var.h src/sys/sparc64/sparc64 ata_machdep.c Message-ID: <200803091425.m29EP2Jx016375@repoman.freebsd.org> marius 2008-03-09 14:25:02 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf files.sparc64 sys/sparc64/include md_var.h Added files: (Branch: RELENG_7) sys/sparc64/sparc64 ata_machdep.c Log: MFC: sys/conf/files.sparc64 1.94; sys/sparc64/include/md_var.h 1.17; sys/sparc64/sparc64/ata_machdep.c 1.1 The Sun disk label only uses 16-bit fields for cylinders, heads and sectors so the geometry of large IDE disks has to be adjusted. Revision Changes Path 1.92.2.1 +1 -0 src/sys/conf/files.sparc64 1.16.10.1 +13 -0 src/sys/sparc64/include/md_var.h 1.1.2.1 +52 -0 src/sys/sparc64/sparc64/ata_machdep.c (new) From marius at FreeBSD.org Sun Mar 9 14:30:50 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Mar 9 14:30:51 2008 Subject: cvs commit: src/sys/dev/ata ata-disk.c Message-ID: <200803091430.m29EUnUd017875@repoman.freebsd.org> marius 2008-03-09 14:30:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/ata ata-disk.c Log: MFC: 1.198 Add a hook that allows the geometry to be overridden on certain platforms. Revision Changes Path 1.189.2.8 +9 -0 src/sys/dev/ata/ata-disk.c From marius at FreeBSD.org Sun Mar 9 14:34:26 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Mar 9 14:34:31 2008 Subject: cvs commit: src/sys/conf files.sparc64 src/sys/sparc64/include md_var.h src/sys/sparc64/sparc64 ata_machdep.c Message-ID: <200803091434.m29EYQif018808@repoman.freebsd.org> marius 2008-03-09 14:34:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/conf files.sparc64 sys/sparc64/include md_var.h Added files: (Branch: RELENG_6) sys/sparc64/sparc64 ata_machdep.c Log: MFC: sys/conf/files.sparc64 1.94; sys/sparc64/include/md_var.h 1.17; sys/sparc64/sparc64/ata_machdep.c 1.1 The Sun disk label only uses 16-bit fields for cylinders, heads and sectors so the geometry of large IDE disks has to be adjusted. Revision Changes Path 1.73.2.5 +1 -0 src/sys/conf/files.sparc64 1.16.2.1 +13 -0 src/sys/sparc64/include/md_var.h 1.1.4.1 +55 -0 src/sys/sparc64/sparc64/ata_machdep.c (new) From rwatson at FreeBSD.org Sun Mar 9 14:49:40 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 14:49:43 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803091449.m29End5B022551@repoman.freebsd.org> rwatson 2008-03-09 14:49:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ddb db_input.c Log: Merge db_input.c:1.38 from HEAD to RELENG_7: When redrawing an input line, count backspaces to get to the beginning of the input field from the current cursor location, rather than the end of the input line, as the cursor may not be at the end of the line. Otherwise, we may overshoot, overwriting a bit of the previous line and failing to fully overwrite the current line. PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.36.10.1 +1 -1 src/sys/ddb/db_input.c From rwatson at FreeBSD.org Sun Mar 9 14:50:29 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 14:50:31 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803091450.m29EoSIO022864@repoman.freebsd.org> rwatson 2008-03-09 14:50:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/ddb db_input.c Log: Merge db_input.c:1.38 from HEAD to RELENG_6: When redrawing an input line, count backspaces to get to the beginning of the input field from the current cursor location, rather than the end of the input line, as the cursor may not be at the end of the line. Otherwise, we may overshoot, overwriting a bit of the previous line and failing to fully overwrite the current line. PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.36.2.1 +1 -1 src/sys/ddb/db_input.c From rwatson at FreeBSD.org Sun Mar 9 14:55:00 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 14:55:04 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803091455.m29Et05M023809@repoman.freebsd.org> rwatson 2008-03-09 14:55:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) sys/ddb db_input.c Log: Merge db_input.c:1.38 from HEAD to RELENG_5: When redrawing an input line, count backspaces to get to the beginning of the input field from the current cursor location, rather than the end of the input line, as the cursor may not be at the end of the line. Otherwise, we may overshoot, overwriting a bit of the previous line and failing to fully overwrite the current line. PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.35.4.2 +1 -1 src/sys/ddb/db_input.c From rwatson at FreeBSD.org Sun Mar 9 15:06:13 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 15:06:15 2008 Subject: cvs commit: src/sys/fs/coda coda_vfsops.c Message-ID: <200803091506.m29F6Cx3025449@repoman.freebsd.org> rwatson 2008-03-09 15:06:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vfsops.c Log: Merge coda_vfsops.c:1.73 from HEAD to RELENG_7: Remove undefined coda excluded by #if 1 #else, which previously protected vget() calls using inode numbers to query the root of /coda, which is not needed since we now cache the root vnode with the mountpoint. Revision Changes Path 1.67.2.2 +0 -12 src/sys/fs/coda/coda_vfsops.c From rwatson at FreeBSD.org Sun Mar 9 15:09:23 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 15:09:24 2008 Subject: cvs commit: src/sys/fs/coda coda_vnops.c Message-ID: <200803091509.m29F9MVf025566@repoman.freebsd.org> rwatson 2008-03-09 15:09:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vnops.c Log: Merge coda_vnops.c:1.86 from HEAD to RELENG_7: Before invoking vnode operations on cache vnodes, acquire the vnode locks of those vnodes. Probably, Coda should do the same lock sharing/ pass-through that is done for nullfs, but in the mean time this ensures that locks are adequately held to prevent corruption of data structures in the cache file system. Assuming most operations came from the top layer of Coda and weren't performed directly on the cache vnodes, in practice this corruption was relatively unlikely as the Coda vnode locks were ensuring exclusive access for most consumers. This causes WITNESS to squeal like a pig immediately when Coda is used, rather than waiting until file close; I noticed these problems because of the lack of said squealing. NB: Because WITNESS in RELENG_7 doesn't yet support lockmgr, said squealing doesn't yet happen in RELENG_7, but will do if/when lockmgr grows support for WITNESS there. Revision Changes Path 1.76.2.2 +9 -1 src/sys/fs/coda/coda_vnops.c From rwatson at FreeBSD.org Sun Mar 9 15:12:12 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 15:12:15 2008 Subject: cvs commit: src/sys/fs/coda coda_psdev.c coda_subr.c coda_venus.c coda_vfsops.c coda_vnops.c coda_vnops.h Message-ID: <200803091512.m29FCChT025736@repoman.freebsd.org> rwatson 2008-03-09 15:12:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_psdev.c coda_subr.c coda_venus.c coda_vfsops.c coda_vnops.c coda_vnops.h Log: Merge coda_psdev.c:1.41, coda_subr.c:1.45, coda_venus.c:1.28, coda_vfsops.c:1.74, coda_vnops.c:1.87, coda_vnops.h:1.20 from HEAD to RELENG_7: The Coda module was originally ported to NetBSD from Mach by rvb, and then later to FreeBSD. Update various NetBSD-related comments: in some cases delete them because they don't appply, in others update to say FreeBSD as they still apply but in FreeBSD (and might for that matter no longer apply on NetBSD), and flag one case where I'm not sure whether it applies. Revision Changes Path 1.39.2.2 +1 -1 src/sys/fs/coda/coda_psdev.c 1.33.2.1 +2 -2 src/sys/fs/coda/coda_subr.c 1.27.2.1 +1 -1 src/sys/fs/coda/coda_venus.c 1.67.2.3 +1 -1 src/sys/fs/coda/coda_vfsops.c 1.76.2.3 +10 -7 src/sys/fs/coda/coda_vnops.c 1.19.2.1 +1 -1 src/sys/fs/coda/coda_vnops.h From marius at FreeBSD.org Sun Mar 9 15:53:35 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Mar 9 15:53:36 2008 Subject: cvs commit: src/sys/sparc64/include tlb.h src/sys/sparc64/sparc64 cache.c cheetah.c mp_machdep.c pmap.c spitfire.c tlb.c Message-ID: <200803091553.m29FrYZS028376@repoman.freebsd.org> marius 2008-03-09 15:53:34 UTC FreeBSD src repository Modified files: sys/sparc64/include tlb.h sys/sparc64/sparc64 cache.c cheetah.c mp_machdep.c pmap.c spitfire.c tlb.c Log: - Do as the comment in pmap_bootstrap() suggests and flush all non-locked TLB entries possibly left over by the firmware and also do so while bootstrapping APs. - Use __FBSDID. MFC after: 1 month Revision Changes Path 1.27 +8 -0 src/sys/sparc64/include/tlb.h 1.21 +5 -2 src/sys/sparc64/sparc64/cache.c 1.6 +20 -2 src/sys/sparc64/sparc64/cheetah.c 1.38 +6 -1 src/sys/sparc64/sparc64/mp_machdep.c 1.171 +12 -5 src/sys/sparc64/sparc64/pmap.c 1.6 +28 -5 src/sys/sparc64/sparc64/spitfire.c 1.9 +6 -4 src/sys/sparc64/sparc64/tlb.c From marius at FreeBSD.org Sun Mar 9 17:09:17 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Mar 9 17:09:20 2008 Subject: cvs commit: src/sys/sparc64/sparc64 iommu.c Message-ID: <200803091709.m29H9GTX043136@repoman.freebsd.org> marius 2008-03-09 17:09:16 UTC FreeBSD src repository Modified files: sys/sparc64/sparc64 iommu.c Log: - Fix some style bugs. - Replace hard-coded functions names missed in rev. 1.44 with __func__. MFC after: 1 week Revision Changes Path 1.46 +54 -48 src/sys/sparc64/sparc64/iommu.c From marius at FreeBSD.org Sun Mar 9 17:55:20 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Mar 9 17:55:22 2008 Subject: cvs commit: src/sys/dev/gem if_gem.c Message-ID: <200803091755.m29HtKDv056080@repoman.freebsd.org> marius 2008-03-09 17:55:20 UTC FreeBSD src repository Modified files: sys/dev/gem if_gem.c Log: - Fix some style bugs and remove another banal comment missed in rev. 1.46. - Move the KASSERT on gem_add_rxbuf() to the right spot and add an equivalent one to gem_disable_tx(). Revision Changes Path 1.48 +11 -10 src/sys/dev/gem/if_gem.c From eng at prowip.net.br Sun Mar 9 18:30:59 2008 From: eng at prowip.net.br (HM Eng.Prowip) Date: Sun Mar 9 18:31:26 2008 Subject: cvs commit: src/sys/conf NOTES files src/sys/modules/ipfw_nat Makefile src/sys/netinet ip_fw.h ip_fw2.c ip_fw_nat.c In-Reply-To: <200802292227.m1TMRJwb013420@repoman.freebsd.org> References: <200802292227.m1TMRJwb013420@repoman.freebsd.org> Message-ID: <200803091416.45296.eng@prowip.net.br> On Friday 29 February 2008 19:27:19 Paolo Pisati wrote: > piso 2008-02-29 22:27:19 UTC > > FreeBSD src repository > > Modified files: > sys/conf NOTES files > sys/netinet ip_fw.h ip_fw2.c > Added files: > sys/modules/ipfw_nat Makefile > sys/netinet ip_fw_nat.c > Log: > Move ipfw's nat code into its own kld: ipfw_nat. > > Revision Changes Path > 1.1469 +1 -2 src/sys/conf/NOTES > 1.1272 +1 -0 src/sys/conf/files > 1.1 +8 -0 src/sys/modules/ipfw_nat/Makefile (new) > 1.112 +32 -0 src/sys/netinet/ip_fw.h > 1.182 +54 -552 src/sys/netinet/ip_fw2.c > 1.1 +646 -0 src/sys/netinet/ip_fw_nat.c (new) > _______________________________________________ since this update I get constant Mar 9 14:06:27 gw kernel: ipfw: install_state: entry already present, done Mar 9 14:06:30 gw kernel: ipfw: install_state: entry already present, done in my log I do not use nat what does this mean? -- Atenciosamente Prowip Telecom Ltda. http://prowip.net.br Cx.Postal 101 S?o Jos? do Rio Preto SP 15010-970 A mensagem foi scaneada pelo sistema de e-mail e pode ser considerada segura. Service fornecido pelo Datacenter Matik https://datacenter.matik.com.br From bz at FreeBSD.org Sun Mar 9 19:01:52 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Mar 9 19:01:53 2008 Subject: cvs commit: src/sys/kern kern_jail.c src/sys/netinet ip_options.c src/sys/netinet6 in6.c ip6_output.c ip6_var.h raw_ip6.c udp6_usrreq.c Message-ID: <200803091901.m29J1pe6059778@repoman.freebsd.org> bz 2008-03-09 19:01:51 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern kern_jail.c sys/netinet ip_options.c sys/netinet6 in6.c ip6_output.c ip6_var.h raw_ip6.c udp6_usrreq.c Log: MFC: 1.75 sys/kern/kern_jail.c 1.8 sys/netinet/ip_options.c 1.78 sys/netinet6/in6.c 1.113 sys/netinet6/ip6_output.c 1.41 sys/netinet6/ip6_var.h 1.76 sys/netinet6/raw_ip6.c 1.85 sys/netinet6/udp6_usrreq.c [ previously MFCed by rwatson 1.18 sys/sys/priv.h belonging to this change ] Replace the last susers calls in netinet6/ with privilege checks. Introduce a new privilege allowing to set certain IP header options (hop-by-hop, routing headers). Leave a few comments that we would like to address later. Discussed with: rwatson Revision Changes Path 1.70.2.1 +6 -0 src/sys/kern/kern_jail.c 1.6.2.1 +1 -0 src/sys/netinet/ip_options.c 1.73.2.1 +10 -10 src/sys/netinet6/in6.c 1.109.2.2 +75 -44 src/sys/netinet6/ip6_output.c 1.39.2.1 +1 -1 src/sys/netinet6/ip6_var.h 1.73.2.1 +4 -7 src/sys/netinet6/raw_ip6.c 1.81.2.1 +1 -6 src/sys/netinet6/udp6_usrreq.c From phk at FreeBSD.org Sun Mar 9 19:14:38 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Sun Mar 9 19:14:39 2008 Subject: cvs commit: src/usr.sbin/fifolog Makefile Makefile.inc flint.lnt src/usr.sbin/fifolog/fifolog_create Makefile fifolog.1 fifolog_create.c src/usr.sbin/fifolog/fifolog_reader Makefile fifolog_reader.c src/usr.sbin/fifolog/fifolog_writer Makefile ... Message-ID: <200803091914.m29JEbw1061867@repoman.freebsd.org> phk 2008-03-09 19:14:37 UTC FreeBSD src repository Added files: usr.sbin/fifolog Makefile Makefile.inc flint.lnt usr.sbin/fifolog/fifolog_create Makefile fifolog.1 fifolog_create.c usr.sbin/fifolog/fifolog_reader Makefile fifolog_reader.c usr.sbin/fifolog/fifolog_writer Makefile fifolog_writer.c usr.sbin/fifolog/lib Makefile fifolog.h fifolog_create.c fifolog_int.c fifolog_reader.c fifolog_write.h fifolog_write_poll.c getdate.y libfifolog.h libfifolog_int.h miniobj.h Log: Add the fifolog tools to FreeBSD. Quoth the man-page: Fifologs provide a compact round-robin circular storage for recording text and binary information to permanent storage in a bounded and pre- dictable fashion, time and space wise. Not yet connected to the build, but feel free to test & review. Revision Changes Path 1.1 +10 -0 src/usr.sbin/fifolog/Makefile (new) 1.1 +13 -0 src/usr.sbin/fifolog/Makefile.inc (new) 1.1 +22 -0 src/usr.sbin/fifolog/fifolog_create/Makefile (new) 1.1 +210 -0 src/usr.sbin/fifolog/fifolog_create/fifolog.1 (new) 1.1 +105 -0 src/usr.sbin/fifolog/fifolog_create/fifolog_create.c (new) 1.1 +21 -0 src/usr.sbin/fifolog/fifolog_reader/Makefile (new) 1.1 +170 -0 src/usr.sbin/fifolog/fifolog_reader/fifolog_reader.c (new) 1.1 +16 -0 src/usr.sbin/fifolog/fifolog_writer/Makefile (new) 1.1 +105 -0 src/usr.sbin/fifolog/fifolog_writer/fifolog_writer.c (new) 1.1 +49 -0 src/usr.sbin/fifolog/flint.lnt (new) 1.1 +17 -0 src/usr.sbin/fifolog/lib/Makefile (new) 1.1 +138 -0 src/usr.sbin/fifolog/lib/fifolog.h (new) 1.1 +122 -0 src/usr.sbin/fifolog/lib/fifolog_create.c (new) 1.1 +276 -0 src/usr.sbin/fifolog/lib/fifolog_int.c (new) 1.1 +315 -0 src/usr.sbin/fifolog/lib/fifolog_reader.c (new) 1.1 +65 -0 src/usr.sbin/fifolog/lib/fifolog_write.h (new) 1.1 +416 -0 src/usr.sbin/fifolog/lib/fifolog_write_poll.c (new) 1.1 +889 -0 src/usr.sbin/fifolog/lib/getdate.y (new) 1.1 +62 -0 src/usr.sbin/fifolog/lib/libfifolog.h (new) 1.1 +45 -0 src/usr.sbin/fifolog/lib/libfifolog_int.h (new) 1.1 +66 -0 src/usr.sbin/fifolog/lib/miniobj.h (new) From mav at FreeBSD.org Sun Mar 9 20:02:35 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 9 20:02:37 2008 Subject: cvs commit: src/share/man/man4 ng_nat.4 src/sys/netgraph ng_nat.c ng_nat.h Message-ID: <200803092002.m29K2ZAf063737@repoman.freebsd.org> mav 2008-03-09 20:02:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 ng_nat.4 sys/netgraph ng_nat.c ng_nat.h Log: MFC: Add support for the libalias redirect functionality. Submitted by: Vadim Goncharov Revision Changes Path 1.2.2.2 +188 -2 src/share/man/man4/ng_nat.4 1.4.2.3 +431 -1 src/sys/netgraph/ng_nat.c 1.1.2.2 +127 -0 src/sys/netgraph/ng_nat.h From mav at FreeBSD.org Sun Mar 9 20:05:45 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 9 20:05:46 2008 Subject: cvs commit: src/sys/netgraph ng_pppoe.c Message-ID: <200803092005.m29K5ehV065022@repoman.freebsd.org> mav 2008-03-09 20:05:40 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netgraph ng_pppoe.c Log: MFC rev. 1.93-1.94 Make session ID generator to use session ID hash. Make session ID generator thread-safe. Use more compact LIST instead of TAILQ for session hash. Add all listening hooks into LIST to simplify searches. Use ng_findhook() instead of own equal implementation. Revision Changes Path 1.74.2.10 +67 -100 src/sys/netgraph/ng_pppoe.c From mav at FreeBSD.org Sun Mar 9 20:10:55 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 9 20:10:58 2008 Subject: cvs commit: src/sys/netgraph ng_base.c Message-ID: <200803092010.m29KAs3q065275@repoman.freebsd.org> mav 2008-03-09 20:10:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_base.c Log: MFC: rev. 1.148-1.149 Implement 128 items node name hash for faster name search. Increase node ID hash size from 32 to 128 items. Increase default queue items allocation limit from 512 to 4096 items to avoid terrible unpredicted effects for netgraph operation of their exhaustion while allocating control messages. Add separate configurable 512 items limit for data items allocation for DoS/overload protection. Revision Changes Path 1.135.2.5 +86 -49 src/sys/netgraph/ng_base.c From piso at freebsd.org Sun Mar 9 20:16:30 2008 From: piso at freebsd.org (Paolo Pisati) Date: Sun Mar 9 20:16:34 2008 Subject: cvs commit: src/sys/conf NOTES files src/sys/modules/ipfw_nat Makefile src/sys/netinet ip_fw.h ip_fw2.c ip_fw_nat.c In-Reply-To: <200803091416.45296.eng@prowip.net.br> References: <200802292227.m1TMRJwb013420@repoman.freebsd.org> <200803091416.45296.eng@prowip.net.br> Message-ID: <20080309201955.GA51679@tin.it> On Sun, Mar 09, 2008 at 02:16:45PM -0300, HM Eng.Prowip wrote: > > Move ipfw's nat code into its own kld: ipfw_nat. > > > > Revision Changes Path > > 1.1469 +1 -2 src/sys/conf/NOTES > > 1.1272 +1 -0 src/sys/conf/files > > 1.1 +8 -0 src/sys/modules/ipfw_nat/Makefile (new) > > 1.112 +32 -0 src/sys/netinet/ip_fw.h > > 1.182 +54 -552 src/sys/netinet/ip_fw2.c > > 1.1 +646 -0 src/sys/netinet/ip_fw_nat.c (new) > > _______________________________________________ > > > > since this update I get constant > > Mar 9 14:06:27 gw kernel: ipfw: install_state: entry already present, done > Mar 9 14:06:30 gw kernel: ipfw: install_state: entry already present, done > > in my log > I do not use nat > > what does this mean? uhm... i dont' see how this could happen: ip_fw2::install_state() q = lookup_dyn_rule_locked(&args->f_id, NULL, NULL); if (q != NULL) { /* should never occur */ if (last_log != time_uptime) { last_log = time_uptime; printf("ipfw: %s: entry already present, done\n", __func__); } IPFW_DYN_UNLOCK(); return (0); } show me your ipfw ruleset/kernel config and i'll try to reproduce it. Is it possible to have access to that box? -- bye, P. From wpaul at FreeBSD.ORG Sun Mar 9 20:26:33 2008 From: wpaul at FreeBSD.ORG (Bill Paul) Date: Sun Mar 9 20:26:37 2008 Subject: cvs commit: src/sys/powerpc/booke machdep.c In-Reply-To: <47D2E72E.1000107@semihalf.com> from Rafal Jaworowski at "Mar 8, 2008 08:21:18 pm" Message-ID: <20080309202633.1DEC11065672@hub.freebsd.org> > Marcel Moolenaar wrote: > >> > >> Thanks for hooking this up, Marcel; it was [far] on my list too :) > > > > I just wanted my kernel builds to be faster :-) > > > > Ah, self-hosted builds you mean? Over NFS, right? (this is how we usually test > this) > > >> BTW: currently U-Boot is broken for most of PPC variations (including > >> 85xx) in > >> the following way: d-cache is not being flushed before disabling and > >> strange > >> things happen as you imagine. For a quick workaround apply something > >> like this > >> when using ELF images: > >> http://people.freebsd.org/~raj/patches/misc/uboot-dcache-workaround.diff > > > > Yup. That's what we do at Juniper. I don't see any bad effects other > > than having to invalidate the cache before (re-)enabling it so I don't > > have that patch applied myself... > > > > I have seen strange things, including hangs, and also TSEC non-functional > effects due to this issue... :) > > > The problem I'm having with U-Boot is with the TSEC driver as you've > > seen on the U-Boot mailing list. I have one MPC8555CDS sitting idle > > after the U-Boot upgrade because I can't netboot :-/ > > > > Yes, I saw. Let me re-flash my 8555 with the latest image and try, but when > I'm not remote i.e. Monday. Don't know if you know this tip, but there used to > be a second flash bank on the MPC8555CDS that you can access while running > from the boot one, so you can recover in such scenario without a h/w debugger > by copying over that backup flash bank, provided you have a working image there... > > Rafal > I am not on the U-Boot mailing list, so I'm not familiar with the TSEC problems you're having, but I'm curious: are you saying that when you try to boot a FreeBSD kernel image with the cache disabled, U-Boot can download and start the kernel, but FreeBSD fails to mount its rootfs correctly because of problems with the TSEC? If that is what you're saying, then something about this is very confusing: the FreeBSD TSEC driver appears to be enabling the TSEC's RX and TX snooping featurss, so in theory (if I understand things correctly, which I might not), the cache behavior really shouldn't affect it. With VxWorks, the bootrom typically runs with the data cache disabled and then the VxWorks image enables it once it's loaded, but the same TSEC driver works properly in both scenarios. Some questions that I have: - Is bus_dma aware of the fact that the TSEC is snooped? Because if it is, it seems to me that all the sync operations should reduce to no-ops. - In the encap routine, the driver always sets the R bit immediately on the first descriptor of every outbound frame. (The Ready bit is basically the ownership bit that indicates to the TSEC that the descriptor is ready for processing.) The VxWorks driver always leaves the R bit in first descriptor cleared until all subsequent descriptors have been set up. This is to prevent the TSEC from potentially racing with the encap routine and processing the first few descriptors of an outbound frame before all descriptors are ready. It might be a good idea to have the encap routine in the FreeBSD driver work the same way. I know the idea is to set up all the descriptors first and then initiate transmission by setting the THLT bit in the TSTAT register, but this doesn't account for the possibility that the transmitter is already busy with a previous transmit request when you go to set up new descriptors. - Has anyone actually tried to debug the TSEC issue, such as by setting up tcpdump to capture traffic from the 8555 or instrumenting the driver? It might be worthwhile to actually find the problem rather than just saying "it works if we hack U-Boot" since this might be a symptom of a larger problem that needs to be fixed. -Bill From bz at FreeBSD.org Sun Mar 9 21:04:56 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Mar 9 21:04:57 2008 Subject: cvs commit: src/sys/netinet ip_output.c src/sys/netinet6 ip6_output.c src/sys/netipsec ipsec.c ipsec.h ipsec6.h Message-ID: <200803092104.m29L4tL8077415@repoman.freebsd.org> bz 2008-03-09 21:04:55 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet ip_output.c sys/netinet6 ip6_output.c sys/netipsec ipsec.c ipsec.h ipsec6.h Log: MFC 1.278 sys/netinet/ip_output.c 1.114 sys/netinet6/ip6_output.c 1.26 sys/netipsec/ipsec.c 1.15 sys/netipsec/ipsec.h 1.4 sys/netipsec/ipsec6.h Rather than passing around a cached 'priv', pass in an ucred to ipsec*_set_policy and do the privilege check only if needed. Try to assimilate both ip*_ctloutput code blocks calling ipsec*_set_policy. Revision Changes Path 1.276.2.1 +5 -20 src/sys/netinet/ip_output.c 1.109.2.3 +9 -26 src/sys/netinet6/ip6_output.c 1.24.2.1 +14 -11 src/sys/netipsec/ipsec.c 1.13.2.1 +1 -1 src/sys/netipsec/ipsec.h 1.3.2.1 +1 -1 src/sys/netipsec/ipsec6.h From rwatson at FreeBSD.org Sun Mar 9 21:13:44 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 9 21:13:48 2008 Subject: cvs commit: src/sys/net bpf_filter.c Message-ID: <200803092113.m29LDhAv077859@repoman.freebsd.org> rwatson 2008-03-09 21:13:43 UTC FreeBSD src repository Modified files: sys/net bpf_filter.c Log: Improve convergence of bpf_filter.c toward style(9). MFC after: 3 weeks Submitted by: csjp Revision Changes Path 1.30 +29 -40 src/sys/net/bpf_filter.c From raj at semihalf.com Sun Mar 9 21:46:09 2008 From: raj at semihalf.com (Rafal Jaworowski) Date: Sun Mar 9 21:46:16 2008 Subject: cvs commit: src/sys/powerpc/booke machdep.c In-Reply-To: <20080309202633.1DEC11065672@hub.freebsd.org> References: <20080309202633.1DEC11065672@hub.freebsd.org> Message-ID: <47D45A91.3080602@semihalf.com> Bill Paul wrote: >>>> BTW: currently U-Boot is broken for most of PPC variations (including >>>> 85xx) in >>>> the following way: d-cache is not being flushed before disabling and >>>> strange >>>> things happen as you imagine. For a quick workaround apply something >>>> like this >>>> when using ELF images: >>>> http://people.freebsd.org/~raj/patches/misc/uboot-dcache-workaround.diff >>> Yup. That's what we do at Juniper. I don't see any bad effects other >>> than having to invalidate the cache before (re-)enabling it so I don't >>> have that patch applied myself... >>> >> I have seen strange things, including hangs, and also TSEC non-functional >> effects due to this issue... :) >> >>> The problem I'm having with U-Boot is with the TSEC driver as you've >>> seen on the U-Boot mailing list. I have one MPC8555CDS sitting idle >>> after the U-Boot upgrade because I can't netboot :-/ >>> >> Yes, I saw. Let me re-flash my 8555 with the latest image and try, but when >> I'm not remote i.e. Monday. Don't know if you know this tip, but there used to >> be a second flash bank on the MPC8555CDS that you can access while running >> from the boot one, so you can recover in such scenario without a h/w debugger >> by copying over that backup flash bank, provided you have a working image there... >> >> Rafal >> > > I am not on the U-Boot mailing list, so I'm not familiar with the TSEC > problems you're having, but I'm curious: are you saying that when you > try to boot a FreeBSD kernel image with the cache disabled, U-Boot can > download and start the kernel, but FreeBSD fails to mount its rootfs > correctly because of problems with the TSEC? > I guess I didn't say anything like this :) There are two issues, both local to U-Boot and not affecting kernel TSEC driver operation: 1. What Marcel is suffering from is some basic problem with TSEC not working at all (link failure) after U-Boot update on one of his systems. 2. I referred to a U-Boot bug present on most PPC variations, which might surface when running standalone apps enveloped in ELF: d-cache is disabled before passing control, without flushing it first, so depending on the data which didn't get flushed there might be corruptions or undesired effects within the standalone app. One of the problems I was observing was that TSEC stopped working after return from a standalone app to U-Boot, also other corrupt data happend, hangs etc. FreeBSD loader(8) for U-Boot environment is such an external app and when this issue hit the loader wouldn't usually be able to load the kernel via network at all. The pure binary produced from this same ELF image works without ill effects ('go' command doesn't do cache trickery before passing control). Please note the workaround patch for handling ELF images does not disable cache, just the opposite: it prevents from doing so. > If that is what you're saying, then something about this is very confusing: > the FreeBSD TSEC driver appears to be enabling the TSEC's RX and TX snooping > featurss, so in theory (if I understand things correctly, which I might not), > the cache behavior really shouldn't affect it. With VxWorks, the bootrom > typically runs with the data cache disabled and then the VxWorks image > enables it once it's loaded, but the same TSEC driver works properly in > both scenarios. > Our TSEC driver works fine with caches enabled. > Some questions that I have: > > - Is bus_dma aware of the fact that the TSEC is snooped? Because if it is, > it seems to me that all the sync operations should reduce to no-ops. > _bus_dmamap_sync() for PowerPC is a no-op. TSEC driver uses DMA sync ops per convention (it's also possible it could be integrated with something else than PowerPC core one day, so we are prepared :) > - In the encap routine, the driver always sets the R bit immediately on the > first descriptor of every outbound frame. (The Ready bit is basically > the ownership bit that indicates to the TSEC that the descriptor is > ready for processing.) The VxWorks driver always leaves the R bit in > first descriptor cleared until all subsequent descriptors have been > set up. This is to prevent the TSEC from potentially racing with the > encap routine and processing the first few descriptors of an outbound > frame before all descriptors are ready. It might be a good idea to have > the encap routine in the FreeBSD driver work the same way. I know the > idea is to set up all the descriptors first and then initiate transmission > by setting the THLT bit in the TSTAT register, but this doesn't account > for the possibility that the transmitter is already busy with a previous > transmit request when you go to set up new descriptors. > Thanks for pointing this out. I'll have a look; there are more changes and improvements to TSEC in my queue, hope you don't mind reviewing these too? > - Has anyone actually tried to debug the TSEC issue, such as by setting up > tcpdump to capture traffic from the 8555 or instrumenting the driver? > It might be worthwhile to actually find the problem rather than just > saying "it works if we hack U-Boot" since this might be a symptom of > a larger problem that needs to be fixed. > As above, the hack is preventing from some other problems, that need to be addressed in U-Boot code, and not quite related to the Ethernet controller and its driver. Rafal From jeff at FreeBSD.org Mon Mar 10 01:30:35 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 01:30:37 2008 Subject: cvs commit: src/sys/sys sched.h src/sys/kern sched_4bsd.c Message-ID: <200803100130.m2A1UZFK007409@repoman.freebsd.org> jeff 2008-03-10 01:30:35 UTC FreeBSD src repository Modified files: sys/sys sched.h sys/kern sched_4bsd.c Log: - Add a sched_preempt() routine to be called by md code after IPI_PREEMPT is delivered. - Add a simple implementation to 4bsd. Revision Changes Path 1.115 +11 -0 src/sys/kern/sched_4bsd.c 1.35 +1 -0 src/sys/sys/sched.h From jeff at FreeBSD.org Mon Mar 10 01:32:03 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 01:32:06 2008 Subject: cvs commit: src/sys/kern sched_ule.c Message-ID: <200803100132.m2A1W1wb007462@repoman.freebsd.org> jeff 2008-03-10 01:32:01 UTC FreeBSD src repository Modified files: sys/kern sched_ule.c Log: - Add an implementation of sched_preempt() that avoids excessive IPIs. - Normalize the preemption/ipi setting code by introducing sched_shouldpreempt() so the logical is identical and not repeated between tdq_notify() and sched_setpreempt(). - In tdq_notify() don't set NEEDRESCHED as we may not actually own the thread lock this could have caused us to lose td_flags settings. - Garbage collect some tunables that are no longer relevant. Revision Changes Path 1.227 +72 -62 src/sys/kern/sched_ule.c From jeff at FreeBSD.org Mon Mar 10 01:32:49 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 01:32:50 2008 Subject: cvs commit: src/sys/amd64/amd64 mp_machdep.c src/sys/i386/i386 mp_machdep.c src/sys/sun4v/sun4v mp_machdep.c Message-ID: <200803100132.m2A1WmW3007539@repoman.freebsd.org> jeff 2008-03-10 01:32:48 UTC FreeBSD src repository Modified files: sys/amd64/amd64 mp_machdep.c sys/i386/i386 mp_machdep.c sys/sun4v/sun4v mp_machdep.c Log: - Rather than repeating the same preemption code everywhere call the scheduler specific sched_preempt() routine. Revision Changes Path 1.289 +2 -9 src/sys/amd64/amd64/mp_machdep.c 1.284 +1 -7 src/sys/i386/i386/mp_machdep.c 1.10 +1 -8 src/sys/sun4v/sun4v/mp_machdep.c From jeff at FreeBSD.org Mon Mar 10 01:38:54 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 01:38:58 2008 Subject: cvs commit: src/sys/kern subr_smp.c Message-ID: <200803100138.m2A1cssW007788@repoman.freebsd.org> jeff 2008-03-10 01:38:54 UTC FreeBSD src repository Modified files: sys/kern subr_smp.c Log: - Add the missing '2' case to the switch table for kern.smp.topology and assign it to create the flat 'none' topology where all cpus are scheduled as if they are equal and unrelated. Revision Changes Path 1.205 +4 -0 src/sys/kern/subr_smp.c From imp at FreeBSD.org Mon Mar 10 01:48:25 2008 From: imp at FreeBSD.org (Warner Losh) Date: Mon Mar 10 01:48:28 2008 Subject: cvs commit: src/sys/kern subr_bus.c Message-ID: <200803100148.m2A1mP5A008148@repoman.freebsd.org> imp 2008-03-10 01:48:25 UTC FreeBSD src repository Modified files: sys/kern subr_bus.c Log: Tiny bit of KNF to make bus_setup_intr() look like the rest of this function. Revision Changes Path 1.206 +18 -19 src/sys/kern/subr_bus.c From jeff at FreeBSD.org Mon Mar 10 03:15:20 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 03:15:21 2008 Subject: cvs commit: src/sys/kern sched_ule.c Message-ID: <200803100315.m2A3FKc6019027@repoman.freebsd.org> jeff 2008-03-10 03:15:19 UTC FreeBSD src repository Modified files: sys/kern sched_ule.c Log: Reduce ULE context switch time by over 25%. - Only calculate timeshare priorities once per tick or when a thread is woken from sleeping. - Keep the ts_runq pointer valid after all priority changes. - Call tdq_runq_add() directly from sched_switch() without passing in via tdq_add(). We don't need to adjust loads or runqs anymore. - Sort tdq and ts_sched according to utilization to improve cache behavior. Sponsored by: Nokia Revision Changes Path 1.228 +52 -52 src/sys/kern/sched_ule.c From jeff at FreeBSD.org Mon Mar 10 03:16:52 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 03:16:54 2008 Subject: cvs commit: src/sys/kern kern_synch.c Message-ID: <200803100316.m2A3GpPC019073@repoman.freebsd.org> jeff 2008-03-10 03:16:51 UTC FreeBSD src repository Modified files: sys/kern kern_synch.c Log: - Handle kdb switch panics outside of mi_switch() to remove some instructions from the common path and make the code more clear. Whether this has any impact on performance may depend on optimization levels. Sponsored by: Nokia Revision Changes Path 1.306 +11 -6 src/sys/kern/kern_synch.c From silby at silby.com Mon Mar 10 06:55:29 2008 From: silby at silby.com (Mike Silbersack) Date: Mon Mar 10 06:55:33 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080307122252.Y11033@delplex.bde.org> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305230954.X55190@odysseus.silby.com> <20080307122252.Y11033@delplex.bde.org> Message-ID: <20080310015118.L43534@odysseus.silby.com> On Fri, 7 Mar 2008, Bruce Evans wrote: > On Wed, 5 Mar 2008, Mike Silbersack wrote: >> As I've said in the past, I'd really, really, really like to see regression >> tests for any change to the floating point functions. The types of changes >> you've been making are not easy to verify just by looking at diffs. > > I run local regression tests of 4 billon to 64 billion cases per > function or 1.3 trillion cases for 125 functions a 36-hour run on a > 2.2HGz UP system. These are not well organized enough for commit. > You will have to trust that they are done before commit (or after on > some other machines) :-). I mostly use semi-exhaustive (exhaustive > for 1-arg float precision functions) checks on machine-generated args. > This seems to find problems more routinely than smarter tests, up to > at least double precision. das@ committed some smarter tests. So clean them up and commit them. I'm not saying that anyone will actually run them, but at least they will be there if one chooses to. > OTOH, I barely tested the changes to float_t and double_t. These types > are so rarely used that they are never used in /usr/src, at least in > my old src tree, except for my uncommitted changes in libm parts. > > Bruce The subsequent thread has taught me that I don't care enough about floating point to participate in this thread beyond the regression test point. :) -Mike From jeff at FreeBSD.org Mon Mar 10 09:50:31 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 09:50:33 2008 Subject: cvs commit: src/sys/kern sched_ule.c Message-ID: <200803100950.m2A9oUC9069165@repoman.freebsd.org> jeff 2008-03-10 09:50:30 UTC FreeBSD src repository Modified files: sys/kern sched_ule.c Log: - Don't rely on a side effect of sched_prio() to set the initial ts_runq for thread0. Set it directly in sched_setup(). This fixes traps on boot seen on some machines. Reported by: phk Revision Changes Path 1.229 +1 -0 src/sys/kern/sched_ule.c From rwatson at FreeBSD.org Mon Mar 10 09:54:13 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 09:54:15 2008 Subject: cvs commit: src/lib/libc/sys Makefile.inc kldunload.2 Message-ID: <200803100954.m2A9sDHZ069286@repoman.freebsd.org> rwatson 2008-03-10 09:54:13 UTC FreeBSD src repository Modified files: lib/libc/sys Makefile.inc kldunload.2 Log: Add reference to kldunloadf system call, which was previously not mentioned in the kldunload(2) man page. MFC after: 3 days Spotted by: rink Revision Changes Path 1.129 +1 -0 src/lib/libc/sys/Makefile.inc 1.19 +14 -1 src/lib/libc/sys/kldunload.2 From rwatson at FreeBSD.org Mon Mar 10 10:13:49 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:13:51 2008 Subject: cvs commit: src/sys/fs/coda coda_vfsops.c coda_vnops.c Message-ID: <200803101013.m2AADnbG071070@repoman.freebsd.org> rwatson 2008-03-10 10:13:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vfsops.c coda_vnops.c Log: Merge coda_vfsops.c:1.75, coda_vnops.c:1.88 from HEAD to RELENG_7: Make all calls to vn_lock() in Coda, including recently added ones, use LK_RETRY, since failure is undesirable (and not handled). Pointed out by: kib Revision Changes Path 1.67.2.4 +3 -3 src/sys/fs/coda/coda_vfsops.c 1.76.2.4 +4 -4 src/sys/fs/coda/coda_vnops.c From rwatson at FreeBSD.org Mon Mar 10 10:14:52 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:14:57 2008 Subject: cvs commit: src/sys/fs/coda coda_vnops.c Message-ID: <200803101014.m2AAEqt2071126@repoman.freebsd.org> rwatson 2008-03-10 10:14:52 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vnops.c Log: Merge coda_vnops.c:1.89 from HEAD to RELENG_7: Lock cache vnode when VOP_FSYNC() is called on a Coda vnode. Revision Changes Path 1.76.2.5 +5 -2 src/sys/fs/coda/coda_vnops.c From rwatson at FreeBSD.org Mon Mar 10 10:16:09 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:16:13 2008 Subject: cvs commit: src/sys/fs/coda coda_vnops.c coda_vnops.h Message-ID: <200803101016.m2AAG9fG071221@repoman.freebsd.org> rwatson 2008-03-10 10:16:09 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vnops.c coda_vnops.h Log: Merge coda_vnops.c:1.90, coda_vnops.h:1.21 from HEAD to RELENG_7: Since Coda is effectively a stacked file system, use VOP_EOPNOTSUPP for vop_bmap; delete the existing stub that returned either EINVAL or EOPNOTSUPP, and had unreachable calls to VOP_BMAP on the cache vnode. Revision Changes Path 1.76.2.6 +1 -37 src/sys/fs/coda/coda_vnops.c 1.19.2.2 +0 -1 src/sys/fs/coda/coda_vnops.h From rwatson at FreeBSD.org Mon Mar 10 10:18:07 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:18:08 2008 Subject: cvs commit: src/sys/fs/coda cnode.h coda_vfsops.c Message-ID: <200803101018.m2AAI6Vq071298@repoman.freebsd.org> rwatson 2008-03-10 10:18:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda cnode.h coda_vfsops.c Log: Merge cnode.h:1.23, coda_vfsops.c:1.76 from HEAD to RELENG_7: Remove unused devtomp(), which exploited UFS-specific knowledge to find the mountpoint for a specific device. This was implemented incorrectly, a bad idea in a fundamental sense, and also never used, so presumably a long-idle debugging function. Revision Changes Path 1.21.2.2 +0 -3 src/sys/fs/coda/cnode.h 1.67.2.5 +0 -21 src/sys/fs/coda/coda_vfsops.c From rwatson at FreeBSD.org Mon Mar 10 10:18:48 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:18:54 2008 Subject: cvs commit: src/sys/fs/coda coda_vfsops.c Message-ID: <200803101018.m2AAIm17071352@repoman.freebsd.org> rwatson 2008-03-10 10:18:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vfsops.c Log: Merge coda_vfsops.c:1.77 from HEAD to RELENG_7: No reason not to maintain stats on statfs in Coda, as it's done for other VFS operations, so uncomment the existing statistics gathering. Revision Changes Path 1.67.2.6 +3 -3 src/sys/fs/coda/coda_vfsops.c From rwatson at FreeBSD.org Mon Mar 10 10:22:23 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:22:27 2008 Subject: cvs commit: src/sys/fs/coda cnode.h coda_fbsd.c coda_namecache.c coda_psdev.c coda_psdev.h coda_venus.c coda_vfsops.c coda_vfsops.h coda_vnops.c coda_vnops.h Message-ID: <200803101022.m2AAMNT5071574@repoman.freebsd.org> rwatson 2008-03-10 10:22:23 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda cnode.h coda_fbsd.c coda_namecache.c coda_psdev.c coda_psdev.h coda_venus.c coda_vfsops.c coda_vfsops.h coda_vnops.c coda_vnops.h Log: Merge cnode.h:1.24, coda_fbsd.c:1.47, coda_namecache.c:1.24, coda_psdev.c:1.42, coda_psdev.h:1.7, coda_venus.c:1.29, coda_vfsops.c:1.78, coda_vfsops.h:1.13, coda_vnops.c:1.91, coda_vnops.h:1.22 from HEAD to RELENG_7: Various further non-functional cleanups to coda: - Rename print_vattr to coda_print_vattr and make static, rename print_cred to coda_print_cred. - Remove unused coda_vop_nop. - Add XXX comment because coda_readdir forwards to the cache vnode's readdir rather than venus_readdir, and annotate venus_readdir as unused. - Rename vc_nb_* to vc_*. - Use d_open_t, d_close_t, d_read_t, d_write_t, d_ioctl_t and d_poll_t for prototyping vc_* as that is the intent, don't use our own definitions. - Rename coda_nb_statfs to coda_statfs, rename NB_SFS_SIZ to CODA_SFS_SIZ. - Replace one more OBE reference to NetBSD with a reference to FreeBSD. - Tidy up a little vertical whitespace here and there. - Annotate coda_nc_zapvnode as unused. - Remove unused vcodattach. - Annotate VM_INTR as unused. - Annotate that coda_fhtovp is unused and doesn't match the FreeBSD prototype, so isn't hooked up to vfs_fhtovp. If we want NFS export of Coda to work someday, this needs to be fixed. - Remove unused getNewVnode. - Remove unused coda_vget, coda_init, coda_quotactl prototypes. Revision Changes Path 1.21.2.3 +1 -2 src/sys/fs/coda/cnode.h 1.46.2.1 +7 -7 src/sys/fs/coda/coda_fbsd.c 1.23.2.1 +4 -6 src/sys/fs/coda/coda_namecache.c 1.39.2.3 +9 -17 src/sys/fs/coda/coda_psdev.c 1.6.2.1 +6 -6 src/sys/fs/coda/coda_psdev.h 1.27.2.2 +3 -0 src/sys/fs/coda/coda_venus.c 1.67.2.7 +14 -38 src/sys/fs/coda/coda_vfsops.c 1.12.2.1 +1 -6 src/sys/fs/coda/coda_vfsops.h 1.76.2.7 +10 -19 src/sys/fs/coda/coda_vnops.c 1.19.2.3 +1 -3 src/sys/fs/coda/coda_vnops.h From rwatson at FreeBSD.org Mon Mar 10 10:23:03 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:23:06 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803101023.m2AAN3J8071611@repoman.freebsd.org> rwatson 2008-03-10 10:23:03 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ddb db_input.c Log: Merge db_input.c:1.39 from HEAD to RELENG_7: Reserve two bytes at the end of the DDB input line in db_readline() to hold the newline and nul terminator. Otherwise, there are cases where garbage may end up in the command history due to a lack of a nul terminator, or input may end up without room for a newline. PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.36.10.2 +4 -1 src/sys/ddb/db_input.c From rwatson at FreeBSD.org Mon Mar 10 10:23:38 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:23:39 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803101023.m2AANbwn071648@repoman.freebsd.org> rwatson 2008-03-10 10:23:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/ddb db_input.c Log: Merge db_input.c:1.39 from HEAD to RELENG_6: Reserve two bytes at the end of the DDB input line in db_readline() to hold the newline and nul terminator. Otherwise, there are cases where garbage may end up in the command history due to a lack of a nul terminator, or input may end up without room for a newline. PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.36.2.2 +4 -1 src/sys/ddb/db_input.c From rwatson at FreeBSD.org Mon Mar 10 10:24:12 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:24:14 2008 Subject: cvs commit: src/sys/ddb db_input.c Message-ID: <200803101024.m2AAOCkR071682@repoman.freebsd.org> rwatson 2008-03-10 10:24:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) sys/ddb db_input.c Log: Merge db_input.c:1.39 from HEAD to RELENG_5: Reserve two bytes at the end of the DDB input line in db_readline() to hold the newline and nul terminator. Otherwise, there are cases where garbage may end up in the command history due to a lack of a nul terminator, or input may end up without room for a newline. PR: 119079 Submitted by: Michael Plass Revision Changes Path 1.35.4.3 +4 -1 src/sys/ddb/db_input.c From rwatson at FreeBSD.org Mon Mar 10 10:26:04 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:26:11 2008 Subject: cvs commit: src/sys/ddb db_command.c Message-ID: <200803101026.m2AAQ4lw071801@repoman.freebsd.org> rwatson 2008-03-10 10:26:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ddb db_command.c Log: Merge db_command.c:1.77 from HEAD to RELENG_7: When killing a user process from DDB, check that the requested signal is > 0 rather than >= 0, or we will panic when trying to deliver the signal. PR: 100802 Submitted by: Valerio Daelli Revision Changes Path 1.73.2.1 +1 -1 src/sys/ddb/db_command.c From rwatson at FreeBSD.org Mon Mar 10 10:40:01 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 10:40:06 2008 Subject: cvs commit: src/sys/ddb db_command.c Message-ID: <200803101040.m2AAe0tx072423@repoman.freebsd.org> rwatson 2008-03-10 10:40:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/ddb db_command.c Log: Merge db_command.c from HEAD to RELENG_6: When killing a user process from DDB, check that the requested signal is > 0 rather than >= 0, or we will panic when trying to deliver the signal. PR: 100802 Submitted by: Valerio Daelli Revision Changes Path 1.60.2.5 +1 -1 src/sys/ddb/db_command.c From rwatson at FreeBSD.org Mon Mar 10 11:38:58 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 11:38:59 2008 Subject: cvs commit: src/sys/ddb db_command.c Message-ID: <200803101138.m2ABcvGw076064@repoman.freebsd.org> rwatson 2008-03-10 11:38:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) sys/ddb db_command.c Log: Merge db_command.c:1.77 from HEAD to RELENG_5: When killing a user process from DDB, check that the requested signal is > 0 rather than >= 0, or we will panic when trying to deliver the signal. PR: 100802 Submitted by: Valerio Daelli Revision Changes Path 1.57.2.3 +1 -1 src/sys/ddb/db_command.c From nyan at FreeBSD.org Mon Mar 10 12:25:05 2008 From: nyan at FreeBSD.org (Takahashi Yoshihiro) Date: Mon Mar 10 12:25:11 2008 Subject: cvs commit: src/sys/pc98/conf GENERIC Message-ID: <200803101225.m2ACP4Aa078731@repoman.freebsd.org> nyan 2008-03-10 12:25:04 UTC FreeBSD src repository Modified files: sys/pc98/conf GENERIC Log: MFi386: revision 1.482. Import uslcom(4) from OpenBSD - this is a driver for Silicon Laboratories CP2101/CP2102 based USB serial adapters. Revision Changes Path 1.301 +1 -0 src/sys/pc98/conf/GENERIC From nyan at FreeBSD.org Mon Mar 10 12:41:51 2008 From: nyan at FreeBSD.org (Takahashi Yoshihiro) Date: Mon Mar 10 12:41:52 2008 Subject: cvs commit: src/sys/pc98/pc98 machdep.c Message-ID: <200803101241.m2ACfonA079285@repoman.freebsd.org> nyan 2008-03-10 12:41:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/pc98/pc98 machdep.c Log: MFC: revision 1.400 Add "show sysregs" command to ddb. On i386, this gives gdt, idt, ldt, cr0-4, etc. Support should be added for other platforms that have a different set of registers for system use. Revision Changes Path 1.396.2.2 +19 -0 src/sys/pc98/pc98/machdep.c From phk at FreeBSD.org Mon Mar 10 12:53:43 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Mar 10 12:53:44 2008 Subject: cvs commit: src/usr.sbin Makefile Message-ID: <200803101253.m2ACrgeW088842@repoman.freebsd.org> phk 2008-03-10 12:53:41 UTC FreeBSD src repository Modified files: usr.sbin Makefile Log: Hook fifolog tools up to the build. To the extent make universe works, it hasn't found any fault with this. Revision Changes Path 1.382 +1 -0 src/usr.sbin/Makefile From nyan at FreeBSD.org Mon Mar 10 12:57:50 2008 From: nyan at FreeBSD.org (Takahashi Yoshihiro) Date: Mon Mar 10 12:58:16 2008 Subject: cvs commit: src/sys/pc98/include md_var.h src/sys/pc98/pc98 pc98_machdep.c Message-ID: <200803101257.m2ACvnNd089059@repoman.freebsd.org> nyan 2008-03-10 12:57:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/pc98/include md_var.h sys/pc98/pc98 pc98_machdep.c Log: MFC: sys/pc98/include/md_var.h 1.3; sys/pc98/pc98/pc98_machdep.c 1.28 Add pc98 specific code to adjust the firmware geometry when it differs from the actual geometry. This enables support of disks larger than ~120GB on pc98 boxes. Revision Changes Path 1.2.2.1 +12 -0 src/sys/pc98/include/md_var.h 1.25.2.3 +38 -0 src/sys/pc98/pc98/pc98_machdep.c From das at FreeBSD.ORG Mon Mar 10 13:22:24 2008 From: das at FreeBSD.ORG (David Schultz) Date: Mon Mar 10 13:22:31 2008 Subject: cvs commit: src/sys/i386/include _types.h In-Reply-To: <20080310015118.L43534@odysseus.silby.com> References: <200803051121.m25BLE03035426@repoman.freebsd.org> <20080305230954.X55190@odysseus.silby.com> <20080307122252.Y11033@delplex.bde.org> <20080310015118.L43534@odysseus.silby.com> Message-ID: <20080310132414.GA2698@zim.MIT.EDU> On Mon, Mar 10, 2008, Mike Silbersack wrote: > > On Fri, 7 Mar 2008, Bruce Evans wrote: > > >On Wed, 5 Mar 2008, Mike Silbersack wrote: > >>As I've said in the past, I'd really, really, really like to see > >>regression tests for any change to the floating point functions. The > >>types of changes you've been making are not easy to verify just by > >>looking at diffs. > > > >I run local regression tests of 4 billon to 64 billion cases per > >function or 1.3 trillion cases for 125 functions a 36-hour run on a > >2.2HGz UP system. These are not well organized enough for commit. > >You will have to trust that they are done before commit (or after on > >some other machines) :-). I mostly use semi-exhaustive (exhaustive > >for 1-arg float precision functions) checks on machine-generated args. > >This seems to find problems more routinely than smarter tests, up to > >at least double precision. das@ committed some smarter tests. > > So clean them up and commit them. I'm not saying that anyone will > actually run them, but at least they will be there if one chooses to. > > >OTOH, I barely tested the changes to float_t and double_t. These types > >are so rarely used that they are never used in /usr/src, at least in > >my old src tree, except for my uncommitted changes in libm parts. > > > >Bruce > > The subsequent thread has taught me that I don't care enough about > floating point to participate in this thread beyond the regression test > point. :) I don't necessarily agree with this change, but in Bruce's defense, it's a one-line change, and there aren't many useful ways you could test it. The only thing it would break is buggy third-party apps. As for the argument that ``The types of changes [Bruce has] making are not easy to verify just by looking at diffs,'' well, yes---if you don't care about floating point by your own admission, that's certainly the case. ;-) I have checked in a bunch of tests for various libm functions. This works great for things like ilogb that are always supposed to return the exact answer, okay for things like fe* that are often limited by architectural differences, and not so great for functions like exp() and sin() that are not always exact. The best way to test the functions in the last category is to look for the worst error via random or systematic search, and make sure that error is below some threshold. But this requires tests that run for a long time, and the tests themselves won't work without a reference implementation that's more accurate than (or at least different from) libm's implementation. I have many tests that use mpfr or SunOS 5.10 the math library as a reference, but none of that can be committed. Hence, for these functions, I generally only commit tests for corner cases. On top of all that, tests often fail on Alpha, IA64, powerpc, and sparc64, since these architectures rely partly on software emulation of some features, which is incomplete in FreeBSD, and I don't have time to fix it. From yar at comp.chem.msu.su Mon Mar 10 13:43:59 2008 From: yar at comp.chem.msu.su (Yar Tikhiy) Date: Mon Mar 10 13:44:08 2008 Subject: cvs commit: src/sbin/fsck_ffs main.c In-Reply-To: <20080305222402.GA80407@crodrigues.org> References: <200803050825.m258Ppv2016738@repoman.freebsd.org> <20080305122029.GA7027@dg.local> <20080305222402.GA80407@crodrigues.org> Message-ID: <20080310134410.GA47195@dg.local> On Wed, Mar 05, 2008 at 10:24:02PM +0000, Craig Rodrigues wrote: > On Wed, Mar 05, 2008 at 03:20:29PM +0300, Yar Tikhiy wrote: > > Your analysis of the problem sounds not quite correct to me. > > You make some interesting points in your e-mail. > I suggest that you summarize the points and > post them to arch@ for further review and discussion. My points are directly related to your commit. To illustrate them clearlier, I instrumented vfs_mount.c with the patch shown in the attachment and investigated how fsck_ffs now affects the root mount options. My findings agree with my prognosis; they are as follows. When fsck_ffs just invokes nmount(2), the initial mount options on the root are: fstype, fspath, from. Just before nmount() returns, the mount options on the root become: fstype, from, fspath, errmsg, update, ro. Prior to your commit the final options were: fstype, from, fspath, errmsg. Thus, due to your commit, fsck_ffs now sets two additional string options on the root mount, "ro" and "update". Both are permanent, i.e., they remain there after fsck_ffs exited. I maintain that: 1) Setting "ro" from fsck_ffs is but a hack needed because the option cannot be set on the root mount at boot time due to bugs. The hack is tolerable, but you failed to mark it clearly as a temporary workaround, e.g., using a reasonable XXX-style comment providing enough details. 2) Setting "update" as a _permanent_ option on the mount point is just wrong because "update" isn't one. The nmount(2) code cannot cope with "update" as a string option yet; it still needs to be spelled MNT_UPDATE. And, to the best of my knowledge, there is no problem with still using that bit flag, although you hint at some `weird things'. This part of the change should be reverted. (Another approach could be to fix vfs_donmount(), but that would require the architectural decision be made on whether nmount(2) modifiers are spelled as string options or bit flags. Currently no FS code tries to check for "update", and "reload" isn't there yet, so the issue is still pending. Feel free to conduct a thread on -arch.) Now I expect you either prove my points wrong using technical facts, or react to them by the respective commit(s) to fsck_ffs. Otherwise I'll be forced to request fsck_ffs/main.c rev. 1.49 be backed out and PR 120319 be reopened. Thanks! -- Yar P.S. Here's how I instrumented vfs_mount.c: --- vfs_mount.c.orig 2008-02-19 20:04:36.000000000 +0300 +++ vfs_mount.c 2008-03-10 12:29:30.000000000 +0300 @@ -324,6 +324,19 @@ return (error); } +static void +vfs_printopts(struct vfsoptlist *opts) +{ + struct vfsopt *opt; + int first = 1; + + TAILQ_FOREACH(opt, opts, link) { + printf("%s%s", first ? "" : " ", opt->name); + if (first) + first = 0; + } +} + /* * Merge the old mount options with the new ones passed * in the MNT_UPDATE case. @@ -948,7 +961,16 @@ MNT_IUNLOCK(mp); VOP_UNLOCK(vp, 0); mp->mnt_optnew = fsdata; + printf(">>> opts before: "); + vfs_printopts(mp->mnt_opt); + printf("\n"); + printf(">>> opts new: "); + vfs_printopts(mp->mnt_optnew); + printf("\n"); vfs_mergeopts(mp->mnt_optnew, mp->mnt_opt); + printf(">>> opts merged: "); + vfs_printopts(mp->mnt_optnew); + printf("\n"); } else { /* * If the user is not root, ensure that they own the directory @@ -1027,6 +1049,9 @@ if (mp->mnt_opt != NULL) vfs_freeopts(mp->mnt_opt); mp->mnt_opt = mp->mnt_optnew; + printf(">>> opts final: "); + vfs_printopts(mp->mnt_opt); + printf("\n"); (void)VFS_STATFS(mp, &mp->mnt_stat, td); } /* From jkoshy at FreeBSD.org Mon Mar 10 14:44:41 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Mon Mar 10 14:44:43 2008 Subject: cvs commit: src/lib/libarchive archive_entry.3 archive_util.3 archive_write.3 archive_write_disk.3 Message-ID: <200803101444.m2AEifrL098300@repoman.freebsd.org> jkoshy 2008-03-10 14:44:41 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry.3 archive_util.3 archive_write.3 archive_write_disk.3 Log: Use .Fo/.Fc and .Xo/.Xc to bring the line widths below 79 columns. Reviewed by: ru Revision Changes Path 1.16 +45 -7 src/lib/libarchive/archive_entry.3 1.8 +6 -1 src/lib/libarchive/archive_util.3 1.23 +36 -6 src/lib/libarchive/archive_write.3 1.2 +16 -3 src/lib/libarchive/archive_write_disk.3 From jkoshy at FreeBSD.org Mon Mar 10 14:45:30 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Mon Mar 10 14:45:37 2008 Subject: cvs commit: src/lib/libarchive archive_read.3 Message-ID: <200803101445.m2AEjT1k098393@repoman.freebsd.org> jkoshy 2008-03-10 14:45:29 UTC FreeBSD src repository Modified files: lib/libarchive archive_read.3 Log: Use .Fo/.Fc and .Xo/.Xc to bring the line widths below 79 columns. Correct a typo [a misplaced comma]. Reviewed by: ru Revision Changes Path 1.36 +71 -13 src/lib/libarchive/archive_read.3 From rwatson at FreeBSD.org Mon Mar 10 14:55:36 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 14:55:40 2008 Subject: cvs commit: src/sys/netsmb smb_conn.h Message-ID: <200803101455.m2AEtaTJ001288@repoman.freebsd.org> rwatson 2008-03-10 14:55:34 UTC FreeBSD src repository Modified files: sys/netsmb smb_conn.h Log: Remove unused vc_tnode field from struct smb_vc. MFC after: 3 days Revision Changes Path 1.13 +0 -1 src/sys/netsmb/smb_conn.h From jhb at FreeBSD.org Mon Mar 10 16:49:44 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 16:49:54 2008 Subject: cvs commit: src/share/zoneinfo leapseconds southamerica Message-ID: <200803101649.m2AGnh13019095@repoman.freebsd.org> jhb 2008-03-10 16:49:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_4) share/zoneinfo leapseconds southamerica Log: MFC: tzdata2008a Revision Changes Path 1.4.2.11 +7 -6 src/share/zoneinfo/leapseconds 1.12.2.14 +26 -2 src/share/zoneinfo/southamerica From jhb at FreeBSD.org Mon Mar 10 16:57:24 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 16:57:38 2008 Subject: cvs commit: src/sys/sparc64/include tlb.h src/sys/sparc64/sparc64 cache.c cheetah.c mp_machdep.c pmap.c spitfire.c tlb.c In-Reply-To: <200803091553.m29FrYZS028376@repoman.freebsd.org> References: <200803091553.m29FrYZS028376@repoman.freebsd.org> Message-ID: <200803101239.37335.jhb@freebsd.org> On Sunday 09 March 2008 11:53:34 am Marius Strobl wrote: > marius 2008-03-09 15:53:34 UTC > > FreeBSD src repository > > Modified files: > sys/sparc64/include tlb.h > sys/sparc64/sparc64 cache.c cheetah.c mp_machdep.c pmap.c > spitfire.c tlb.c > Log: > - Do as the comment in pmap_bootstrap() suggests and flush all non-locked > TLB entries possibly left over by the firmware and also do so while > bootstrapping APs. > - Use __FBSDID. Does this fix the panics when switching to a smaller kernel in the loader? -- John Baldwin From jhb at FreeBSD.org Mon Mar 10 16:57:29 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 16:57:39 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c In-Reply-To: <200803091349.m29Dn91v003746@repoman.freebsd.org> References: <200803091349.m29Dn91v003746@repoman.freebsd.org> Message-ID: <200803101244.52073.jhb@freebsd.org> On Sunday 09 March 2008 09:49:09 am Robert Watson wrote: > rwatson 2008-03-09 13:49:09 UTC > > FreeBSD src repository > > Modified files: > usr.bin/gzip gzip.c > Log: > Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar to > its existing understanding that .tgz uncompresses to .tar. > > MFC after: 3 days > PR: 121519 > Submitted by: Ighighi > > Revision Changes Path > 1.6 +2 -0 src/usr.bin/gzip/gzip.c So are we creating packages for ports that are compressed with gzip but using bzip names? Otherwise, gunzip should probably fail on a '.tbz' because it isn't a gzipped file (unless gunzip auto-invokes bunzip2 or some such feature that I'm not aware of). -- John Baldwin From jhb at FreeBSD.org Mon Mar 10 16:57:33 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 16:57:47 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <484742.73871.qm@web33703.mail.mud.yahoo.com> References: <484742.73871.qm@web33703.mail.mud.yahoo.com> Message-ID: <200803101256.56550.jhb@freebsd.org> On Saturday 08 March 2008 11:33:13 am Abdullah Ibn Hamad Al-Marri wrote: > ----- Original Message ---- > > From: Rui Paulo > > To: Abdullah Ibn Hamad Al-Marri > > Cc: src-committers@freebsd.org; cvs-src@freebsd.org; cvs-all@freebsd.org > > Sent: Saturday, March 8, 2008 6:38:10 PM > > Subject: Re: cvs commit: src/sys/i386/cpufreq est.c > > > > On Sat, Mar 8, 2008 at 2:21 PM, Abdullah Ibn Hamad Al-Marri > > wrote: > > > > > > ----- Original Message ---- > > > > > > > From: Rui Paulo > > > > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; cvs-all@FreeBSD.org > > > > Sent: Saturday, March 8, 2008 5:12:35 PM > > > > Subject: cvs commit: src/sys/i386/cpufreq est.c > > > > > > > > rpaulo 2008-03-08 14:12:35 UTC > > > > > > > > FreeBSD src repository > > > > > > > > Modified files: (Branch: RELENG_7) > > > > > > > sys/i386/cpufreq est.c > > > > Log: > > > > MFC r1.12: > > > > Validate the id16 values gathered from ACPI (previously a TODO > > > > item). > > > > Style changes by me and njl. > > > > > > > > Submitted by: Takeharu KATO > > > > PR: 119350 > > > > > > > > Revision Changes Path > > > > 1.11.2.1 +59 -17 src/sys/i386/cpufreq/est.c > > > > > > Thanks, how about AMD64? > > > > This directory is shared by i386 and amd64. > > > > Regards. > > -- > > Rui Paulo > > > > > Thank you, > > Here what I got. > FreeBSD 7.0-STABLE #0: Sat Mar 8 16:16:54 UTC 2008 > arabian@MX1.WeArab.Net:/usr/obj/usr/src/sys/MX1 > Timecounter "i8254" frequency 1193182 Hz quality 0 > CPU: Intel(R) Core(TM)2 Duo CPU E6750 @ 2.66GHz (2666.62-MHz K8-class CPU) > Origin = "GenuineIntel" Id = 0x6fb Stepping = 11 > Features=0xbfebfbff > Features2=0xe3fd > AMD Features=0x20100800 > AMD Features2=0x1 > Cores per package: 2 > usable memory = 4276043776 (4077 MB) > avail memory = 4106977280 (3916 MB) > ACPI APIC Table: > FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs > cpu0 (BSP): APIC ID: 0 > cpu1 (AP): APIC ID: 1 > ioapic0 irqs 0-23 on motherboard > acpi0: on motherboard > acpi0: [ITHREAD] > acpi0: Power Button (fixed) > Timecounter "ACPI-fast" frequency 3579545 Hz quality 1000 > acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 > acpi_hpet0: iomem 0xfed00000-0xfed003ff on acpi0 > Timecounter "HPET" frequency 14318180 Hz quality 900 > cpu0: on acpi0 > est0: on cpu0 > est: CPU supports Enhanced Speedstep, but is not recognized. > est: cpu_vendor GenuineIntel, msr 82a082a0600082a > device_attach: est0 attach returned 6 > p4tcc0: on cpu0 > cpu1: on acpi0 > est1: on cpu1 > est: CPU supports Enhanced Speedstep, but is not recognized. > est: cpu_vendor GenuineIntel, msr 82a082a0600082a > device_attach: est1 attach returned 6 > p4tcc1: on cpu1 Try www.freebsd.org/~jhb/patches/est_msr.patch -- John Baldwin From rwatson at FreeBSD.org Mon Mar 10 17:16:30 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 17:16:33 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c In-Reply-To: <200803101244.52073.jhb@freebsd.org> References: <200803091349.m29Dn91v003746@repoman.freebsd.org> <200803101244.52073.jhb@freebsd.org> Message-ID: <20080310171551.Y70549@fledge.watson.org> On Mon, 10 Mar 2008, John Baldwin wrote: >> usr.bin/gzip gzip.c >> Log: >> Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar > to >> its existing understanding that .tgz uncompresses to .tar. >> >> MFC after: 3 days >> PR: 121519 >> Submitted by: Ighighi >> >> Revision Changes Path >> 1.6 +2 -0 src/usr.bin/gzip/gzip.c > > So are we creating packages for ports that are compressed with gzip but > using bzip names? Otherwise, gunzip should probably fail on a '.tbz' > because it isn't a gzipped file (unless gunzip auto-invokes bunzip2 or some > such feature that I'm not aware of). gunzip is smarter than you think. :-) Robert N M Watson Computer Laboratory University of Cambridge From jhb at freebsd.org Mon Mar 10 17:50:32 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 10 17:50:40 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c In-Reply-To: <20080310171551.Y70549@fledge.watson.org> References: <200803091349.m29Dn91v003746@repoman.freebsd.org> <200803101244.52073.jhb@freebsd.org> <20080310171551.Y70549@fledge.watson.org> Message-ID: <200803101348.47724.jhb@freebsd.org> On Monday 10 March 2008 01:16:29 pm Robert Watson wrote: > > On Mon, 10 Mar 2008, John Baldwin wrote: > > >> usr.bin/gzip gzip.c > >> Log: > >> Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar > > to > >> its existing understanding that .tgz uncompresses to .tar. > >> > >> MFC after: 3 days > >> PR: 121519 > >> Submitted by: Ighighi > >> > >> Revision Changes Path > >> 1.6 +2 -0 src/usr.bin/gzip/gzip.c > > > > So are we creating packages for ports that are compressed with gzip but > > using bzip names? Otherwise, gunzip should probably fail on a '.tbz' > > because it isn't a gzipped file (unless gunzip auto-invokes bunzip2 or some > > such feature that I'm not aware of). > > gunzip is smarter than you think. :-) Hmm. I guess I would consider it a user error to use gunzip on a bzip'd file (round peg, square hole sort of deal). However, that has more to do with the functionality (which I find surprising and perhaps a bit non-UNIX-ish) rather than this specific patch. -- John Baldwin From rwatson at FreeBSD.org Mon Mar 10 18:11:10 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 18:11:19 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c In-Reply-To: <200803101348.47724.jhb@freebsd.org> References: <200803091349.m29Dn91v003746@repoman.freebsd.org> <200803101244.52073.jhb@freebsd.org> <20080310171551.Y70549@fledge.watson.org> <200803101348.47724.jhb@freebsd.org> Message-ID: <20080310180939.V70549@fledge.watson.org> On Mon, 10 Mar 2008, John Baldwin wrote: >>> So are we creating packages for ports that are compressed with gzip but >>> using bzip names? Otherwise, gunzip should probably fail on a '.tbz' >>> because it isn't a gzipped file (unless gunzip auto-invokes bunzip2 or >>> some such feature that I'm not aware of). >> >> gunzip is smarter than you think. :-) > > Hmm. I guess I would consider it a user error to use gunzip on a bzip'd > file (round peg, square hole sort of deal). However, that has more to do > with the functionality (which I find surprising and perhaps a bit > non-UNIX-ish) rather than this specific patch. Yeah -- this change was really just about teaching gunzip that unzipping a .tbz or .tbz2 file is like unzipping a .tgz or .tgz2 file as was the case for .tgz: rename it to .tar when it's unzipped. The functionality to unzip bz2 files was already present, you just end up with a different file extension when done. I can see arguments for and against that, but nothing to do with the patch I committed. :-) Robert N M Watson Computer Laboratory University of Cambridge From rwatson at FreeBSD.org Mon Mar 10 18:15:03 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 10 18:15:11 2008 Subject: cvs commit: src/sys/security/audit audit_worker.c Message-ID: <200803101815.m2AIF3CD029650@repoman.freebsd.org> rwatson 2008-03-10 18:15:03 UTC FreeBSD src repository Modified files: sys/security/audit audit_worker.c Log: Remove XXX to remind me to check the free space calculation, which to my eyes appears right following a check. MFC after: 3 days Revision Changes Path 1.22 +0 -3 src/sys/security/audit/audit_worker.c From emaste at FreeBSD.org Mon Mar 10 18:44:07 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Mon Mar 10 18:44:11 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_softdep.c Message-ID: <200803101844.m2AIi7OF030653@repoman.freebsd.org> emaste 2008-03-10 18:44:07 UTC FreeBSD src repository Modified files: sys/ufs/ffs ffs_softdep.c Log: Remove include of opt_quota.h; as of revision 1.205 there is no longer any #ifdef QUOTA conditional code. Revision Changes Path 1.218 +0 -1 src/sys/ufs/ffs/ffs_softdep.c From tmclaugh at FreeBSD.org Mon Mar 10 19:03:56 2008 From: tmclaugh at FreeBSD.org (Tom McLaughlin) Date: Mon Mar 10 19:04:05 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200803101903.m2AJ3tYM032584@repoman.freebsd.org> tmclaugh 2008-03-10 19:03:55 UTC FreeBSD src repository (ports committer) Modified files: sbin/mount mount.8 Log: - unionfs -b option is deprecated in favor of '-o below' as per mount_unionfs(8). Approved by: remko Revision Changes Path 1.85 +2 -2 src/sbin/mount/mount.8 From marius at alchemy.franken.de Mon Mar 10 19:44:46 2008 From: marius at alchemy.franken.de (Marius Strobl) Date: Mon Mar 10 19:44:54 2008 Subject: cvs commit: src/sys/sparc64/include tlb.h src/sys/sparc64/sparc64 cache.c cheetah.c mp_machdep.c pmap.c spitfire.c tlb.c In-Reply-To: <200803101239.37335.jhb@freebsd.org> References: <200803091553.m29FrYZS028376@repoman.freebsd.org> <200803101239.37335.jhb@freebsd.org> Message-ID: <20080310194443.GA771@alchemy.franken.de> On Mon, Mar 10, 2008 at 12:39:37PM -0400, John Baldwin wrote: > On Sunday 09 March 2008 11:53:34 am Marius Strobl wrote: > > marius 2008-03-09 15:53:34 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/sparc64/include tlb.h > > sys/sparc64/sparc64 cache.c cheetah.c mp_machdep.c pmap.c > > spitfire.c tlb.c > > Log: > > - Do as the comment in pmap_bootstrap() suggests and flush all non-locked > > TLB entries possibly left over by the firmware and also do so while > > bootstrapping APs. > > - Use __FBSDID. > > Does this fix the panics when switching to a smaller kernel in the loader? > No, that's a different problem. Marius From jhb at FreeBSD.org Mon Mar 10 19:58:42 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 19:58:43 2008 Subject: cvs commit: src/bin/cp cp.1 cp.c utils.c Message-ID: <200803101958.m2AJwfmm037050@repoman.freebsd.org> jhb 2008-03-10 19:58:41 UTC FreeBSD src repository Modified files: bin/cp cp.1 cp.c utils.c Log: Add an '-a' option which is identical to specifying '-RpP' for compatibility with other implementations. MFC after: 1 month Revision Changes Path 1.40 +6 -2 src/bin/cp/cp.1 1.60 +7 -1 src/bin/cp/cp.c 1.53 +2 -2 src/bin/cp/utils.c From mav at FreeBSD.org Mon Mar 10 20:38:57 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Mon Mar 10 20:39:04 2008 Subject: cvs commit: src/sys/netgraph ng_base.c Message-ID: <200803102038.m2AKcuTf039676@repoman.freebsd.org> mav 2008-03-10 20:38:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netgraph ng_base.c Log: MFC: rev. 1.148-1.149 Implement 128 items node name hash for faster name search. Increase node ID hash size from 32 to 128 items. Increase default queue items allocation limit from 512 to 4096 items to avoid terrible unpredicted effects for netgraph operation of their exhaustion while allocating control messages. Add separate configurable 512 items limit for data items allocation for DoS/overload protection. Revision Changes Path 1.102.2.17 +86 -49 src/sys/netgraph/ng_base.c From imp at bsdimp.com Mon Mar 10 20:44:07 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Mon Mar 10 20:44:10 2008 Subject: cvs commit: src/bin/cp cp.1 cp.c utils.c In-Reply-To: <200803101958.m2AJwfmm037050@repoman.freebsd.org> References: <200803101958.m2AJwfmm037050@repoman.freebsd.org> Message-ID: <20080310.144327.-856141340.imp@bsdimp.com> In message: <200803101958.m2AJwfmm037050@repoman.freebsd.org> John Baldwin writes: : jhb 2008-03-10 19:58:41 UTC : : FreeBSD src repository : : Modified files: : bin/cp cp.1 cp.c utils.c : Log: : Add an '-a' option which is identical to specifying '-RpP' for : compatibility with other implementations. : : MFC after: 1 month Submitted by: David Frascone at Cisco right? Warner From tmclaugh at FreeBSD.org Mon Mar 10 20:44:28 2008 From: tmclaugh at FreeBSD.org (Tom McLaughlin) Date: Mon Mar 10 20:44:29 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200803102044.m2AKiR9r040052@repoman.freebsd.org> tmclaugh 2008-03-10 20:44:27 UTC FreeBSD src repository (ports committer) Modified files: sbin/mount mount.8 Log: - Also change the /sbin/mount_unionfs line I managed to miss just two lines down to '-o below'. Approved by: remko Noticed by: rodrigc Pointyhat by: me Revision Changes Path 1.86 +1 -1 src/sbin/mount/mount.8 From remko at FreeBSD.org Mon Mar 10 20:49:08 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Mon Mar 10 20:49:15 2008 Subject: cvs commit: src/sbin/mount mount.8 In-Reply-To: <200803102044.m2AKiR9r040052@repoman.freebsd.org> References: <200803102044.m2AKiR9r040052@repoman.freebsd.org> Message-ID: <47D59EC2.7060409@FreeBSD.org> Tom McLaughlin wrote: > tmclaugh 2008-03-10 20:44:27 UTC > > FreeBSD src repository (ports committer) > > Modified files: > sbin/mount mount.8 > Log: > - Also change the /sbin/mount_unionfs line I managed to miss just two > lines down to '-o below'. > > Approved by: remko > Noticed by: rodrigc > Pointyhat by: me ^^^ & remko (For not properly reviewing the patch) Thanks Craig for noticing! -- /"\ Best regards, | remko@FreeBSD.org \ / Remko Lodder | remko@EFnet X http://www.evilcoder.org/ | / \ ASCII Ribbon Campaign | Against HTML Mail and News From wearabnet at yahoo.ca Mon Mar 10 20:53:58 2008 From: wearabnet at yahoo.ca (Abdullah Ibn Hamad Al-Marri) Date: Mon Mar 10 20:54:03 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <703678.84353.qm@web33702.mail.mud.yahoo.com> ----- Original Message ---- > From: John Baldwin > To: Abdullah Ibn Hamad Al-Marri > Cc: cvs-src@FreeBSD.org; src-committers@FreeBSD.org; Rui Paulo > Sent: Monday, March 10, 2008 7:56:56 PM > Subject: Re: cvs commit: src/sys/i386/cpufreq est.c > > On Saturday 08 March 2008 11:33:13 am Abdullah Ibn Hamad Al-Marri wrote: > > ----- Original Message ---- > > > From: Rui Paulo > > > To: Abdullah Ibn Hamad Al-Marri > > > Cc: src-committers@freebsd.org; cvs-src@freebsd.org; cvs-all@freebsd.org > > > Sent: Saturday, March 8, 2008 6:38:10 PM > > > Subject: Re: cvs commit: src/sys/i386/cpufreq est.c > > > > > > On Sat, Mar 8, 2008 at 2:21 PM, Abdullah Ibn Hamad Al-Marri > > > wrote: > > > > > > > > ----- Original Message ---- > > > > > > > > > From: Rui Paulo > > > > > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; > cvs-all@FreeBSD.org > > > > > Sent: Saturday, March 8, 2008 5:12:35 PM > > > > > Subject: cvs commit: src/sys/i386/cpufreq est.c > > > > > > > > > > rpaulo 2008-03-08 14:12:35 UTC > > > > > > > > > > FreeBSD src repository > > > > > > > > > > Modified files: (Branch: RELENG_7) > > > > > > > > > sys/i386/cpufreq est.c > > > > > Log: > > > > > MFC r1.12: > > > > > Validate the id16 values gathered from ACPI (previously a TODO > > > > > item). > > > > > Style changes by me and njl. > > > > > > > > > > Submitted by: Takeharu KATO > > > > > PR: 119350 > > > > > > > > > > Revision Changes Path > > > > > 1.11.2.1 +59 -17 src/sys/i386/cpufreq/est.c > > > > > > > > Thanks, how about AMD64? > > > > > > This directory is shared by i386 and amd64. > > > > > > Regards. > > > -- > > > Rui Paulo > > > > > > > > > Thank you, > > > > Here what I got. > > FreeBSD 7.0-STABLE #0: Sat Mar 8 16:16:54 UTC 2008 > > arabian@MX1.WeArab.Net:/usr/obj/usr/src/sys/MX1 > > Timecounter "i8254" frequency 1193182 Hz quality 0 > > CPU: Intel(R) Core(TM)2 Duo CPU E6750 @ 2.66GHz (2666.62-MHz K8-class > CPU) > > Origin = "GenuineIntel" Id = 0x6fb Stepping = 11 > > > Features=0xbfebfbff > > > Features2=0xe3fd > > AMD Features=0x20100800 > > AMD Features2=0x1 > > Cores per package: 2 > > usable memory = 4276043776 (4077 MB) > > avail memory = 4106977280 (3916 MB) > > ACPI APIC Table: > > FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs > > cpu0 (BSP): APIC ID: 0 > > cpu1 (AP): APIC ID: 1 > > ioapic0 irqs 0-23 on motherboard > > acpi0: on motherboard > > acpi0: [ITHREAD] > > acpi0: Power Button (fixed) > > Timecounter "ACPI-fast" frequency 3579545 Hz quality 1000 > > acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 > > acpi_hpet0: iomem 0xfed00000-0xfed003ff on > acpi0 > > Timecounter "HPET" frequency 14318180 Hz quality 900 > > cpu0: on acpi0 > > est0: on cpu0 > > est: CPU supports Enhanced Speedstep, but is not recognized. > > est: cpu_vendor GenuineIntel, msr 82a082a0600082a > > device_attach: est0 attach returned 6 > > p4tcc0: on cpu0 > > cpu1: on acpi0 > > est1: on cpu1 > > est: CPU supports Enhanced Speedstep, but is not recognized. > > est: cpu_vendor GenuineIntel, msr 82a082a0600082a > > device_attach: est1 attach returned 6 > > p4tcc1: on cpu1 > > Try www.freebsd.org/~jhb/patches/est_msr.patch > > -- > John Baldwin Hey John FreeBSD is a registered trademark of The FreeBSD Foundation. FreeBSD 7.0-STABLE #1: Mon Mar 10 20:43:50 UTC 2008 arabian@MX1.WeArab.Net:/usr/obj/usr/src/sys/MX1 Timecounter "i8254" frequency 1193182 Hz quality 0 CPU: Intel(R) Core(TM)2 Duo CPU E6750 @ 2.66GHz (2666.62-MHz K8-class CPU) Origin = "GenuineIntel" Id = 0x6fb Stepping = 11 Features=0xbfebfbff Features2=0xe3fd AMD Features=0x20100800 AMD Features2=0x1 Cores per package: 2 usable memory = 4276043776 (4077 MB) avail memory = 4106977280 (3916 MB) ACPI APIC Table: FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs cpu0 (BSP): APIC ID: 0 cpu1 (AP): APIC ID: 1 ioapic0 irqs 0-23 on motherboard acpi0: on motherboard acpi0: [ITHREAD] acpi0: Power Button (fixed) Timecounter "ACPI-fast" frequency 3579545 Hz quality 1000 acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 acpi_hpet0: iomem 0xfed00000-0xfed003ff on acpi0 Timecounter "HPET" frequency 14318180 Hz quality 900 cpu0: on acpi0 est0: on cpu0 est0: Guessed bus clock (high) of 333 MHz ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ est0: Guessed bus clock (low) of 333 MHz ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ est: CPU supports Enhanced Speedstep, but is not recognized. est: cpu_vendor GenuineIntel, msr 82a082a0600082a ^^^^^^^^^^^^^^^^^^^^^^^^^^ device_attach: est0 attach returned 6 p4tcc0: on cpu0 Regards, -Abdullah Ibn Hamad Al-Marri Arab Portal http://www.WeArab.Net/ ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs From jhb at FreeBSD.org Mon Mar 10 21:43:31 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 21:43:32 2008 Subject: cvs commit: src/sys/boot/i386/btx/btx btx.S Message-ID: <200803102143.m2ALhVbF052196@repoman.freebsd.org> jhb 2008-03-10 21:43:31 UTC FreeBSD src repository Modified files: sys/boot/i386/btx/btx btx.S Log: Change the BTX kernel to drop all the way out to real mode to invoke BIOS routines (V86 requests from the client and hardware interrupt handlers): - Install trampoline real mode interrupt handlers at IDT vectors 0x20-0x2f to handle hardware interrupts by invoking the appropriate vector (0x8-0xf or 0x70-0x78). This allows the 8259As to use vectors 0x20-0x2f in real mode as well as protected mode will ensuring that the master 8259A doesn't share IDT space with CPU exceptions in protected mode. - Since we don't need to reserve space for page tables and a page directory anymore since dropping paging support, move the TSS and protected mode IDT up by 16k. Grow the ring 1 link stack by 16k as a result. - Repurpose the ring 1 link stack to be used as a real mode stack when invoking real mode routines either via a V86 request or a hardware interrupts. This simplifies a few things as we avoid disturbing the original user stack. - Add some more block comments to explain how the code interacts with the V86 structure as this wasn't immediately obvious from the prior comments (e.g. that we explicitly copy the seg regs for real mode out of the V86 struct onto the stack to be popped off when going into real mode, etc.). Also, document some of the stack frames we create going to real mode and back. - Remove all of the virtual 86 related code including having to simulate various instructions and BIOS calls on a trap from virtual 86 mode. - Explicitly panic if a user client attempts to perform a V86 CALL request that isn't a far call. - Bump version to 1.2. Assuming this works ok this should fix some of the long standing issues with USB booting as well as etherboot. MFC after: 2 weeks Submitted by: kib (some parts from his original real mode patch) Revision Changes Path 1.46 +317 -402 src/sys/boot/i386/btx/btx/btx.S From jhb at FreeBSD.org Mon Mar 10 22:00:37 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 22:00:40 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803102200.m2AM0asg052806@repoman.freebsd.org> jhb 2008-03-10 22:00:36 UTC FreeBSD src repository Modified files: sys/i386/cpufreq est.c Log: - Don't execute cpuid to fetch the features. We already have the features present in cpu_feature2. Also, use CPUID2_EST rather than a magic number. - Don't free the ACPI settings list in detach if we are going to fail the request. Otherwise an attempt to kldunload est would free the array but the driver would keep trying to use it. MFC after: 1 week Revision Changes Path 1.14 +5 -5 src/sys/i386/cpufreq/est.c From jhb at FreeBSD.org Mon Mar 10 22:02:56 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 22:03:02 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <703678.84353.qm@web33702.mail.mud.yahoo.com> References: <703678.84353.qm@web33702.mail.mud.yahoo.com> Message-ID: <200803101756.23487.jhb@freebsd.org> On Monday 10 March 2008 04:53:57 pm Abdullah Ibn Hamad Al-Marri wrote: > ----- Original Message ---- > > From: John Baldwin > > To: Abdullah Ibn Hamad Al-Marri > > Cc: cvs-src@FreeBSD.org; src-committers@FreeBSD.org; Rui Paulo > > Sent: Monday, March 10, 2008 7:56:56 PM > > Subject: Re: cvs commit: src/sys/i386/cpufreq est.c > > > > On Saturday 08 March 2008 11:33:13 am Abdullah Ibn Hamad Al-Marri wrote: > > > ----- Original Message ---- > > > > From: Rui Paulo > > > > To: Abdullah Ibn Hamad Al-Marri > > > > Cc: src-committers@freebsd.org; cvs-src@freebsd.org; cvs-all@freebsd.org > > > > Sent: Saturday, March 8, 2008 6:38:10 PM > > > > Subject: Re: cvs commit: src/sys/i386/cpufreq est.c > > > > > > > > On Sat, Mar 8, 2008 at 2:21 PM, Abdullah Ibn Hamad Al-Marri > > > > wrote: > > > > > > > > > > ----- Original Message ---- > > > > > > > > > > > From: Rui Paulo > > > > > > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; > > cvs-all@FreeBSD.org > > > > > > Sent: Saturday, March 8, 2008 5:12:35 PM > > > > > > Subject: cvs commit: src/sys/i386/cpufreq est.c > > > > > > > > > > > > rpaulo 2008-03-08 14:12:35 UTC > > > > > > > > > > > > FreeBSD src repository > > > > > > > > > > > > Modified files: (Branch: RELENG_7) > > > > > > > > > > > sys/i386/cpufreq est.c > > > > > > Log: > > > > > > MFC r1.12: > > > > > > Validate the id16 values gathered from ACPI (previously a TODO > > > > > > item). > > > > > > Style changes by me and njl. > > > > > > > > > > > > Submitted by: Takeharu KATO > > > > > > PR: 119350 > > > > > > > > > > > > Revision Changes Path > > > > > > 1.11.2.1 +59 -17 src/sys/i386/cpufreq/est.c > > > > > > > > > > Thanks, how about AMD64? > > > > > > > > This directory is shared by i386 and amd64. > > > > > > > > Regards. > > > > -- > > > > Rui Paulo > > > > > > > > > > > > > Thank you, > > > > > > Here what I got. > > > FreeBSD 7.0-STABLE #0: Sat Mar 8 16:16:54 UTC 2008 > > > arabian@MX1.WeArab.Net:/usr/obj/usr/src/sys/MX1 > > > Timecounter "i8254" frequency 1193182 Hz quality 0 > > > CPU: Intel(R) Core(TM)2 Duo CPU E6750 @ 2.66GHz (2666.62-MHz K8-class > > CPU) > > > Origin = "GenuineIntel" Id = 0x6fb Stepping = 11 > > > > > Features=0xbfebfbff > > > > > Features2=0xe3fd > > > AMD Features=0x20100800 > > > AMD Features2=0x1 > > > Cores per package: 2 > > > usable memory = 4276043776 (4077 MB) > > > avail memory = 4106977280 (3916 MB) > > > ACPI APIC Table: > > > FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs > > > cpu0 (BSP): APIC ID: 0 > > > cpu1 (AP): APIC ID: 1 > > > ioapic0 irqs 0-23 on motherboard > > > acpi0: on motherboard > > > acpi0: [ITHREAD] > > > acpi0: Power Button (fixed) > > > Timecounter "ACPI-fast" frequency 3579545 Hz quality 1000 > > > acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 > > > acpi_hpet0: iomem 0xfed00000-0xfed003ff on > > acpi0 > > > Timecounter "HPET" frequency 14318180 Hz quality 900 > > > cpu0: on acpi0 > > > est0: on cpu0 > > > est: CPU supports Enhanced Speedstep, but is not recognized. > > > est: cpu_vendor GenuineIntel, msr 82a082a0600082a > > > device_attach: est0 attach returned 6 > > > p4tcc0: on cpu0 > > > cpu1: on acpi0 > > > est1: on cpu1 > > > est: CPU supports Enhanced Speedstep, but is not recognized. > > > est: cpu_vendor GenuineIntel, msr 82a082a0600082a > > > device_attach: est1 attach returned 6 > > > p4tcc1: on cpu1 > > > > Try www.freebsd.org/~jhb/patches/est_msr.patch > > > > -- > > John Baldwin > > Hey John > > FreeBSD is a registered trademark of The FreeBSD Foundation. > FreeBSD 7.0-STABLE #1: Mon Mar 10 20:43:50 UTC 2008 > arabian@MX1.WeArab.Net:/usr/obj/usr/src/sys/MX1 > Timecounter "i8254" frequency 1193182 Hz quality 0 > CPU: Intel(R) Core(TM)2 Duo CPU E6750 @ 2.66GHz (2666.62-MHz K8-class CPU) > Origin = "GenuineIntel" Id = 0x6fb Stepping = 11 > Features=0xbfebfbff HTT,TM,PBE> > Features2=0xe3fd > AMD Features=0x20100800 > AMD Features2=0x1 > Cores per package: 2 > usable memory = 4276043776 (4077 MB) > avail memory = 4106977280 (3916 MB) > ACPI APIC Table: > FreeBSD/SMP: Multiprocessor System Detected: 2 CPUs > cpu0 (BSP): APIC ID: 0 > cpu1 (AP): APIC ID: 1 > ioapic0 irqs 0-23 on motherboard > acpi0: on motherboard > acpi0: [ITHREAD] > acpi0: Power Button (fixed) > Timecounter "ACPI-fast" frequency 3579545 Hz quality 1000 > acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 > acpi_hpet0: iomem 0xfed00000-0xfed003ff on acpi0 > Timecounter "HPET" frequency 14318180 Hz quality 900 > cpu0: on acpi0 > est0: on cpu0 > est0: Guessed bus clock (high) of 333 MHz > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > est0: Guessed bus clock (low) of 333 MHz > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > est: CPU supports Enhanced Speedstep, but is not recognized. > est: cpu_vendor GenuineIntel, msr 82a082a0600082a > ^^^^^^^^^^^^^^^^^^^^^^^^^^ > device_attach: est0 attach returned 6 > p4tcc0: on cpu0 Unless you really have a bus clock of 333 (in which case the test needs to be updated) the MSR's don't add up. You can try changing the statement that tests for 100 or 133 to accept 333 as well. -- John Baldwin From jhb at FreeBSD.org Mon Mar 10 22:02:57 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 22:03:10 2008 Subject: cvs commit: src/bin/cp cp.1 cp.c utils.c In-Reply-To: <20080310.144327.-856141340.imp@bsdimp.com> References: <200803101958.m2AJwfmm037050@repoman.freebsd.org> <20080310.144327.-856141340.imp@bsdimp.com> Message-ID: <200803101747.51354.jhb@freebsd.org> On Monday 10 March 2008 04:43:27 pm M. Warner Losh wrote: > In message: <200803101958.m2AJwfmm037050@repoman.freebsd.org> > John Baldwin writes: > : jhb 2008-03-10 19:58:41 UTC > : > : FreeBSD src repository > : > : Modified files: > : bin/cp cp.1 cp.c utils.c > : Log: > : Add an '-a' option which is identical to specifying '-RpP' for > : compatibility with other implementations. > : > : MFC after: 1 month > > Submitted by: David Frascone at Cisco > > right? Don't think so. I've had this patch for almost 2 years now on my laptop from the first time this came up (back when -l was added IIRC). The recent thread just reminded me to commit it. :) The recent thread also had a patch that used -rpP whereas -RpP is more correct IMO. -- John Baldwin From jhb at FreeBSD.org Mon Mar 10 22:18:07 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 10 22:18:11 2008 Subject: cvs commit: src/sys/amd64/amd64 legacy.c src/sys/dev/acpica acpi.c src/sys/dev/cpufreq ichss.c src/sys/i386/i386 legacy.c Message-ID: <200803102218.m2AMI7pM054603@repoman.freebsd.org> jhb 2008-03-10 22:18:07 UTC FreeBSD src repository Modified files: sys/amd64/amd64 legacy.c sys/dev/acpica acpi.c sys/dev/cpufreq ichss.c sys/i386/i386 legacy.c Log: Probe CPUs after the PCI hierarchy on i386, amd64, and ia64. This allows the cpufreq drivers to reliably use properties of PCI devices for quirks, etc. - For the legacy drivers, add CPU devices via an identify routine in the CPU driver itself rather than in the legacy driver's attach routine. - Add CPU devices after Host-PCI bridges in the acpi bus driver. - Change the ichss(4) driver to use pci_find_bsf() to locate the ICH and check its device ID rather than having a bogus PCI attachment that only checked for the ID in probe and always failed. As a side effect, you can now kldload ichss after boot. - Fix the ichss(4) driver to use the correct device_t for the ICH (and not for ichss0) when doing PCI config space operations to enable SpeedStep. MFC after: 2 weeks Reviewed by: njl, Andriy Gapon avg of icyb.net.ua Revision Changes Path 1.62 +23 -12 src/sys/amd64/amd64/legacy.c 1.246 +22 -6 src/sys/dev/acpica/acpi.c 1.11 +50 -61 src/sys/dev/cpufreq/ichss.c 1.64 +23 -12 src/sys/i386/i386/legacy.c From jeff at FreeBSD.org Mon Mar 10 22:48:27 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Mon Mar 10 22:48:29 2008 Subject: cvs commit: src/sys/kern sched_ule.c Message-ID: <200803102248.m2AMmRFv055635@repoman.freebsd.org> jeff 2008-03-10 22:48:27 UTC FreeBSD src repository Modified files: sys/kern sched_ule.c Log: - Fix the invalid priority panics people are seeing by forcing tdq_runq_add to select the runq rather than hoping we set it properly when we adjusted the priority. This involves the same number of branches as before so should perform identically without the extra fragility. Tested by: bz Reviewed by: bz Revision Changes Path 1.230 +10 -25 src/sys/kern/sched_ule.c From thompsa at FreeBSD.org Mon Mar 10 23:16:48 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Mon Mar 10 23:17:07 2008 Subject: cvs commit: src/sys/dev/wpi if_wpi.c if_wpireg.h if_wpivar.h Message-ID: <200803102316.m2ANGmT8058136@repoman.freebsd.org> thompsa 2008-03-10 23:16:48 UTC FreeBSD src repository Modified files: sys/dev/wpi if_wpi.c if_wpireg.h if_wpivar.h Log: Update wpi(4) with stability fixes - remove second taskqueue - busdma 16k alignment workaround - use busdma instead of external mbuf storage on Rx - locking fixes - net80211 state change fixes - improve scanning reliability - improve radio hw switch interaction - consolidate callouts Parts obtained from: benjsc, sam Tested by: many Revision Changes Path 1.9 +480 -553 src/sys/dev/wpi/if_wpi.c 1.3 +2 -0 src/sys/dev/wpi/if_wpireg.h 1.2 +31 -53 src/sys/dev/wpi/if_wpivar.h From emax at FreeBSD.org Tue Mar 11 00:08:40 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 11 00:08:42 2008 Subject: cvs commit: src/lib/libsdp sdp.h Message-ID: <200803110008.m2B08e5W060893@repoman.freebsd.org> emax 2008-03-11 00:08:40 UTC FreeBSD src repository Modified files: lib/libsdp sdp.h Log: Add structures to hold SDP parameters for the NAP, GN and PANU profiles. It should be mentioned that a somewhat similar patch was submitted by Rako < rako29 at gmail dot com > MFC after: 1 week Revision Changes Path 1.8 +17 -0 src/lib/libsdp/sdp.h From aryeh.friedman at gmail.com Tue Mar 11 00:47:03 2008 From: aryeh.friedman at gmail.com (Aryeh M. Friedman) Date: Tue Mar 11 00:47:08 2008 Subject: cvs commit: src/sys/kern sched_ule.c In-Reply-To: <200803102248.m2AMmRFv055635@repoman.freebsd.org> References: <200803102248.m2AMmRFv055635@repoman.freebsd.org> Message-ID: <47D5D680.9000302@gmail.com> Jeff Roberson wrote: > jeff 2008-03-10 22:48:27 UTC > > FreeBSD src repository > > Modified files: > sys/kern sched_ule.c > Log: > - Fix the invalid priority panics people are seeing by forcing > tdq_runq_add to select the runq rather than hoping we set it properly > when we adjusted the priority. This involves the same number of > branches as before so should perform identically without the extra > fragility. > Does this fix the diablo panic I reported to you last week? From jroberson at chesapeake.net Tue Mar 11 01:33:15 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Tue Mar 11 01:33:19 2008 Subject: cvs commit: src/sys/kern sched_ule.c In-Reply-To: <47D5D680.9000302@gmail.com> References: <200803102248.m2AMmRFv055635@repoman.freebsd.org> <47D5D680.9000302@gmail.com> Message-ID: <20080310153347.X1091@desktop> On Mon, 10 Mar 2008, Aryeh M. Friedman wrote: > Jeff Roberson wrote: >> jeff 2008-03-10 22:48:27 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/kern sched_ule.c Log: >> - Fix the invalid priority panics people are seeing by forcing >> tdq_runq_add to select the runq rather than hoping we set it properly >> when we adjusted the priority. This involves the same number of >> branches as before so should perform identically without the extra >> fragility. >> > > Does this fix the diablo panic I reported to you last week? > I'm not 100% certain. I don't remember the details. You can always try and find out. :-) From doconnor at gsoft.com.au Tue Mar 11 01:48:28 2008 From: doconnor at gsoft.com.au (Daniel O'Connor) Date: Tue Mar 11 01:48:32 2008 Subject: cvs commit: src/sys/boot/i386/btx/btx btx.S In-Reply-To: <200803102143.m2ALhVbF052196@repoman.freebsd.org> References: <200803102143.m2ALhVbF052196@repoman.freebsd.org> Message-ID: <200803111218.21976.doconnor@gsoft.com.au> On Tue, 11 Mar 2008, John Baldwin wrote: > - Remove all of the virtual 86 related code including having to > simulate various instructions and BIOS calls on a trap from virtual > 86 mode. - Explicitly panic if a user client attempts to perform a > V86 CALL request that isn't a far call. > - Bump version to 1.2. > > Assuming this works ok this should fix some of the long standing > issues with USB booting as well as etherboot. Wohoo thanks John! This gets USB booting working on the Supermicro C2SBA's I am using here. FreeBSD flash disk installer here I come! -- Daniel O'Connor software and network engineer for Genesis Software - http://www.gsoft.com.au "The nice thing about standards is that there are so many of them to choose from." -- Andrew Tanenbaum GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: This is a digitally signed message part. Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080311/a7781b2c/attachment.pgp From yongari at FreeBSD.org Tue Mar 11 02:39:58 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 11 02:40:02 2008 Subject: cvs commit: src/sys/dev/msk if_msk.c Message-ID: <200803110239.m2B2dqEt079240@repoman.freebsd.org> yongari 2008-03-11 02:39:52 UTC FreeBSD src repository Modified files: sys/dev/msk if_msk.c Log: Move comments block 1 line up to remark on the setting if_capabilities. This would make comments clear. Suggested by: yar Revision Changes Path 1.31 +2 -2 src/sys/dev/msk/if_msk.c From bms at FreeBSD.org Tue Mar 11 03:18:16 2008 From: bms at FreeBSD.org (Bruce M. Simpson) Date: Tue Mar 11 03:18:21 2008 Subject: cvs commit: src/sys/boot/i386/btx/btx btx.S In-Reply-To: <200803102143.m2ALhVbF052196@repoman.freebsd.org> References: <200803102143.m2ALhVbF052196@repoman.freebsd.org> Message-ID: <47D5F9F4.4090600@FreeBSD.org> John Baldwin wrote: > Log: > Change the BTX kernel to drop all the way out to real mode to invoke BIOS > routines (V86 requests from the client and hardware interrupt handlers): > Excellent work. I am very pleased indeed to see this going in. BMS From davidxu at FreeBSD.org Tue Mar 11 03:26:48 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 11 03:26:57 2008 Subject: cvs commit: src/lib/libthr/thread thr_sem.c Message-ID: <200803110326.m2B3Ql6I082078@repoman.freebsd.org> davidxu 2008-03-11 03:26:47 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_sem.c Log: If a thread is cancelled, it may have already consumed a umtx_wake, check waiter and semphore counter to see if we may wake up next thread. Revision Changes Path 1.14 +2 -0 src/lib/libthr/thread/thr_sem.c From yongari at FreeBSD.org Tue Mar 11 03:44:48 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 11 03:44:51 2008 Subject: cvs commit: src/sys/dev/vr if_vr.c if_vrreg.h Message-ID: <200803110344.m2B3ilaw082692@repoman.freebsd.org> yongari 2008-03-11 03:44:47 UTC FreeBSD src repository Modified files: sys/dev/vr if_vr.c if_vrreg.h Log: Forced commit to note that vr(4) was repocopied from sys/pci and modified for its new location. Revision Changes Path 1.129 +1 -1 src/sys/dev/vr/if_vr.c 1.35 +0 -0 src/sys/dev/vr/if_vrreg.h From yongari at FreeBSD.org Tue Mar 11 03:50:57 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 11 03:50:59 2008 Subject: cvs commit: src/sys/conf files src/sys/modules/vr Makefile Message-ID: <200803110350.m2B3ovd1083212@repoman.freebsd.org> yongari 2008-03-11 03:50:57 UTC FreeBSD src repository Modified files: sys/conf files sys/modules/vr Makefile Log: Update file list and Makefile after repocopying vr(4) from src/sys/pci to src/sys/dev. Revision Changes Path 1.1279 +1 -1 src/sys/conf/files 1.14 +1 -1 src/sys/modules/vr/Makefile From yongari at FreeBSD.org Tue Mar 11 03:53:54 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 11 03:53:55 2008 Subject: cvs commit: src/sys/pci if_vr.c if_vrreg.h Message-ID: <200803110353.m2B3rrcQ083296@repoman.freebsd.org> yongari 2008-03-11 03:53:53 UTC FreeBSD src repository Removed files: sys/pci if_vr.c if_vrreg.h Log: vr(4) was repocopied to src/sys/dev/vr. Revision Changes Path 1.129 +0 -1605 src/sys/pci/if_vr.c (dead) 1.35 +0 -520 src/sys/pci/if_vrreg.h (dead) From aryeh.friedman at gmail.com Tue Mar 11 04:44:34 2008 From: aryeh.friedman at gmail.com (Aryeh M. Friedman) Date: Tue Mar 11 04:44:43 2008 Subject: cvs commit: src/sys/kern sched_ule.c In-Reply-To: <20080310153347.X1091@desktop> References: <200803102248.m2AMmRFv055635@repoman.freebsd.org> <47D5D680.9000302@gmail.com> <20080310153347.X1091@desktop> Message-ID: <47D60DFC.90502@gmail.com> Jeff Roberson wrote: > On Mon, 10 Mar 2008, Aryeh M. Friedman wrote: > >> Jeff Roberson wrote: >>> jeff 2008-03-10 22:48:27 UTC >>> >>> FreeBSD src repository >>> >>> Modified files: >>> sys/kern sched_ule.c Log: >>> - Fix the invalid priority panics people are seeing by forcing >>> tdq_runq_add to select the runq rather than hoping we set it >>> properly >>> when we adjusted the priority. This involves the same number of >>> branches as before so should perform identically without the extra >>> fragility. >>> >> >> Does this fix the diablo panic I reported to you last week? >> > > I'm not 100% certain. I don't remember the details. You can always > try and find out. :-) > No it doesn't and as far I can tell cpio is what actually causes the panic From yongari at FreeBSD.org Tue Mar 11 04:51:23 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 11 04:51:25 2008 Subject: cvs commit: src/sys/dev/vr if_vr.c if_vrreg.h Message-ID: <200803110451.m2B4pMLB095467@repoman.freebsd.org> yongari 2008-03-11 04:51:22 UTC FreeBSD src repository Modified files: sys/dev/vr if_vr.c if_vrreg.h Log: Teach vr(4) to use bus_dma(9) and major overhauling to handle link state change and reliable error recovery. o Moved vr_softc structure and relevant macros to header file. o Use PCIR_BAR macro to get BARs. o Implemented suspend/resume methods. o Implemented automatic Tx threshold configuration which will be activated when it suffers from Tx underrun. Also Tx underrun will try to restart only Tx path and resort to previous full-reset(both Rx/Tx) operation if restarting Tx path have failed. o Removed old bit-banging MII interface. Rhine provides simple and efficient MII interface. While I'm here show PHY address and PHY register number when its read/write operation was failed. o Define VR_MII_TIMEOUT constant and use it in MII access routines. o Always honor link up/down state reported by mii layers. The link state information is used in vr_start() to determine whether we got a valid link. o Removed vr_setcfg() which is now handled in vr_link_task(), link state taskqueue handler. When mii layer reports link state changes the taskqueue handler reprograms MAC to reflect negotiated duplex settings. Flow-control changes are not handled yet and it should be revisited when mii layer knows the notion of flow-control. o Added a new sysctl interface to get statistics of an instance of the driver.(sysctl dev.vr.0.stats=1) o Chip name was renamed to reflect the official name of the chips described in VIA Rhine I/II/III datasheet. REV_ID_3065_A -> REV_ID_VT6102_A REV_ID_3065_B -> REV_ID_VT6102_B REV_ID_3065_C -> REV_ID_VT6102_C REV_ID_3106_J -> REV_ID_VT6105_A0 REV_ID_3106_S -> REV_ID_VT6105M_A0 The following chip revisions were added. #define REV_ID_VT6105_B0 0x83 #define REV_ID_VT6105_LOM 0x8A #define REV_ID_VT6107_A0 0x8C #define REV_ID_VT6107_A1 0x8D #define REV_ID_VT6105M_B1 0x94 o Always show chip revision number in device attach. This shall help identifying revision specific issues. o Check whether EEPROM reloading is complete by inspecting the state of VR_EECSR_LOAD bit. This bit is self-cleared after the EEPROM reloading. Previously vr(4) blindly spins for 200us which may/may not enough to complete the EEPROM reload. o Removed if_mtu setup. It's done in ether_ifattach(). o Use our own callout to drive watchdog timer. o In vr_attach disable further interrupts after reset. For VT6102 or newer hardwares, diable MII state change interrupt as well because mii state handling is done by mii layer. o Add more sane register initialization for VT6102 or newer chips. - Have NIC report error instead of retrying forever. - Let hardware detect MII coding error. - Enable MODE10T mode. - Enable memory-read-multiple for VT6107. o PHY address for VT6105 or newer chips is located at fixed address 1. For older chips the PHY address is stored in VR_PHYADDR register. Armed with these information, there is no need to re-read VR_PHYADDR register in miibus handler to get PHY address. This saves one register access cycle for each MII access. o Don't reprogram VR_PHYADDR register whenever access to a register located at a PHY address is made. Rhine fmaily allows reprogramming PHY address location via VR_PHYADDR register depending on VR_MIISTAT_PHYOPT bit of VR_MIISTAT register. This used to lead numerous phantom PHYs attached to miibus during phy probe phase and driver used to limit allowable PHY address in mii register accessors for certain chip revisions. This removes one more register access cycle for each MII access. o Correctly set VLAN header length. o bus_dma(9) conversion. - Limit DMA access to be in range of 32bit address space. Hardware doesn't support DAC. - Apply descriptor ring alignment requirements(16 bytes alignment) - Apply Rx buffer address alignment requirements(4 bytes alignment) - Apply Tx buffer address alignment requirements(4 bytes alignment) for Rhine I chip. Rhine II or III has no Tx buffer address alignment restrictions, though. - Reduce number of allowable number of DMA segments to 8. - Removed the atomic(9) used in descriptor ownership managements as it's job of bus_dmamap_sync(9). With these change vr(4) should work on all platforms. o Rhine uses two separated 8bits command registers to control Tx/Rx MAC. So don't access it as a single 16bit register. o For non-strict alignment architectures vr(4) no longer require time-consuming copy operation for received frames to align IP header. This greatly improves Rx performance on i386/amd64 platforms. However the alignment is still necessary for strict-alignment platforms(e.g. sparc64). The alignment is handled in new fuction vr_fixup_rx(). o vr_rxeof() now rejects multiple-segmented(fragmented) frames as vr(4) is not ready to handle this situation. Datasheet said nothing about the reason when/why it happens. o In vr_newbuf() don't set VR_RXSTAT_FIRSTFRAG/VR_RXSTAT_LASTFRAG bits as it's set by hardware. o Don't pass checksum offload information to upper layer for fragmented frames. The hardware assisted checksum is valid only when the frame is non-fragmented IP frames. Also mark the checksum is valid for corrupted frames such that upper layers doesn't need to recompute the checksum with software routine. o Removed vr_rxeoc(). RxDMA doesn't seem to need to be idle before sending VR_CMD_RX_GO command. Previously it used to stop RxDMA first which in turn resulted in long delays in Rx error recovery. o Rewrote Tx completion handler. - Always check VR_TXSTAT_OWN bit in status word prior to inspecting other status bits in the status word. - Collision counter updates were corrected as VT3071 or newer ones use different bits to notify collisions. - Unlike other chip revisions, VT86C100A uses different bit to indicate Tx underrun. For VT3071 or newer ones, check both VR_TXSTAT_TBUFF and VR_TXSTAT_UDF bits to see whether Tx underrun was happend. In case of Tx underrun requeue the failed frame and restart stalled Tx SM. Also double Tx DMA threshold size on each failure to mitigate future Tx underruns. - Disarm watchdog timer only if we have no queued packets, otherwise don't touch watchdog timer. o Rewrote interrupt handler. - status word in Tx/Rx descriptors indicates more detailed error state required to recover from the specific error. There is no need to rely on interrupt status word to recover from Tx/Rx error except PCI bus error. Other event notifications like statistics counter overflows or link state events will be handled in main interrupt handler. - Don't touch VR_IMR register if we are in suspend mode. Touching the register may hang the hardware if we are in suspended state. Previously it seems that touching VR_IMR register in interrupt handler was to work-around panic occurred in system shutdown stage on SMP systems. I think that work-around would hide root-cause of the panic and I couldn't reproduce the panic with multiple attempts on my box. o While padding space to meet minimum frame size, zero the pad data in order to avoid possibly leaking sensitive data. o Rewrote vr_start_locked(). - Don't try to queue packets if number of available Tx descriptors are short than that of required one. o Don't reinitialize hardware whenever media configuration is changed. Media/link state changes are reported from mii layer if this happens and vr_link_task() will perform necessary changes. o Don't reinitialize hardware if only PROMISC bit was changed. Just toggle the PROMISC bit in hardware is sufficient to reflect the request. o Rearrganed the IFCAP_POLLING/IFCAP_HWCSUM handling in vr_ioctl(). o Generate Tx completion interrupts for every VR_TX_INTR_THRESH-th frames. This reduces Tx completion interrupts under heavy network loads. o Since vr(4) doesn't request Tx interrupts for every queued frames, reclaim any pending descriptors not handled in Tx completion handler before actually firing up watchdog timeouts. o Added vr_tx_stop()/vr_rx_stop() to wait for the end of active TxDMA/RxDMA cycles(draining). These routines are used in vr_stop() to ensure sane state of MAC before releasing allocated Tx/Rx buffers. vr_link_task() also takes advantage of these functions to get to idle state prior to restarting Tx/Rx. o Added vr_tx_start()/vr_rx_start() to restart Rx/Tx. By separating Rx operation from Tx operation vr(4) no longer need to full-reset the hardware in case of Tx/Rx error recovery. o Implemented WOL. o Added VT6105M specific register definitions. VT6105M has the following hardware capabilities. - Tx/Rx IP/TCP/UDP checksum offload. - VLAN hardware tag insertion/extraction. Due to lack of information for getting extracted VLAN tag in Rx path, VLAN hardware support was not implemented yet. - CAM(Content Addressable Memory) based 32 entry perfect multicast/ VLAN filtering. - 8 priority queues. o Implemented CAM based 32 entry perfect multicast filtering for VT6105M. If number of multicast entry is greater than 32, vr(4) uses traditional hash based filtering. o Reflect real Tx/Rx descriptor structure. Previously vr(4) used to embed other driver (private) data into these structure. This type of embedding make it hard to work on LP64 systems. o Removed unused vr_mii_frame structure and MII bit-baning definitions. o Added new PCI configuration registers that controls mii operation and mode selection. o Reduced number of Tx/Rx descriptors to 128 from 256. From my testing, increasing number of descriptors above than 64 didn't help increasing performance at all. Experimentations show 128 Rx descriptors seems to help a lot reducing Rx FIFO overruns under high system loads. It seems the poor Tx performance of Rhine hardwares comes from the limitation of hardware. You wouldn't satuarte the link with vr(4) no matter how fast CPU/large number of descriptors are used. o Added vr_statistics structure to hold various counter values. No regression was reported but one variant of Rhine III(VT6105M) found on RouterBOARD 44 does not work yet(Reported by Milan Obuch). I hope this would be resolved in near future. I'd like to say big thanks to Mike Tancsa who kindly donated a Rhine hardware to me. Without his enthusiastic testing and feedbacks overhauling vr(4) never have been possible. Also thanks to Masayuki Murayama who provided some good comments on the hardware's internals. This driver is result of combined effort of many users who provided many feedbacks so I'd like to say special thanks to them. Hardware donated by: Mike Tancsa (mike AT sentex dot net) Reviewed by: remko (initial version) Tested by: Mike Tancsa(x86), JoaoBR ( joao AT matik DOT com DOT br ) Marcin Wisnicki ( mwisnicki+freebsd AT gmail DOT com ) Stefan Ehmann ( shoesoft AT gmx DOT net ) Florian Smeets ( flo AT kasimir DOT com ) Phil Oleson ( oz AT nixil DOT net ) Larry Baird ( lab AT gta DOT com ) Milan Obuch ( freebsd-current AT dino DOT sk ) remko (initial version) Revision Changes Path 1.130 +1699 -714 src/sys/dev/vr/if_vr.c 1.36 +330 -98 src/sys/dev/vr/if_vrreg.h From yongari at FreeBSD.org Tue Mar 11 05:05:56 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 11 05:05:58 2008 Subject: cvs commit: src/share/man/man4 vr.4 Message-ID: <200803110505.m2B55uUH097120@repoman.freebsd.org> yongari 2008-03-11 05:05:56 UTC FreeBSD src repository Modified files: share/man/man4 vr.4 Log: Document a new sysctl variable, dev.vr.%d.stats. Touch Dd. Revision Changes Path 1.31 +9 -1 src/share/man/man4/vr.4 From yongari at FreeBSD.org Tue Mar 11 05:09:03 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 11 05:09:07 2008 Subject: cvs commit: src/sys/sparc64/conf GENERIC Message-ID: <200803110509.m2B593Zg097205@repoman.freebsd.org> yongari 2008-03-11 05:09:03 UTC FreeBSD src repository Modified files: sys/sparc64/conf GENERIC Log: Uncomment vr(4), vr(4) should work on all architectures. Revision Changes Path 1.132 +1 -1 src/sys/sparc64/conf/GENERIC From silby at FreeBSD.org Tue Mar 11 06:34:29 2008 From: silby at FreeBSD.org (Mike Silbersack) Date: Tue Mar 11 06:34:31 2008 Subject: cvs commit: src/sys/netinet tcp_var.h Message-ID: <200803110634.m2B6YSlt043810@repoman.freebsd.org> silby 2008-03-11 06:34:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet tcp_var.h Log: MFC rev 1.160; Change tcp options order back to how it was in FreeBSD 6 to improve compatibility. Revision Changes Path 1.157.2.2 +5 -2 src/sys/netinet/tcp_var.h From silby at silby.com Tue Mar 11 07:01:46 2008 From: silby at silby.com (Mike Silbersack) Date: Tue Mar 11 07:01:52 2008 Subject: cvs commit: src/sys/dev/vr if_vr.c if_vrreg.h In-Reply-To: <200803110451.m2B4pMLB095467@repoman.freebsd.org> References: <200803110451.m2B4pMLB095467@repoman.freebsd.org> Message-ID: <20080311015929.Q48057@odysseus.silby.com> On Tue, 11 Mar 2008, Pyun YongHyeon wrote: > yongari 2008-03-11 04:51:22 UTC > > FreeBSD src repository > > Modified files: > sys/dev/vr if_vr.c if_vrreg.h > Log: > Teach vr(4) to use bus_dma(9) and major overhauling to handle link > state change and reliable error recovery. Wow, 180 lines of changes. I think this commit message has more content than most of the *code* that I commit! -Mike From patrickchan at hangsengbank.com Tue Mar 11 08:07:45 2008 From: patrickchan at hangsengbank.com (Patrick Chan) Date: Tue Mar 11 08:07:47 2008 Subject: Reply Me ASAP!!! Message-ID: <294050559062376@mx.freenet.de> I have a business of US$24.5million for you.Please contact me for details. Patrick Chan. E-mail: chanpatrick_hs@yahoo.com.hk From brueffer at FreeBSD.org Tue Mar 11 08:28:45 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Tue Mar 11 08:28:47 2008 Subject: cvs commit: src/release/doc/share/misc dev.archlist.txt Message-ID: <200803110828.m2B8Si0f050014@repoman.freebsd.org> brueffer 2008-03-11 08:28:44 UTC FreeBSD src repository Modified files: release/doc/share/misc dev.archlist.txt Log: vr(4) should work on all archs now, remove it from here. Revision Changes Path 1.104 +0 -1 src/release/doc/share/misc/dev.archlist.txt From rwatson at FreeBSD.org Tue Mar 11 10:14:48 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Tue Mar 11 10:14:55 2008 Subject: cvs commit: src/sys/fs/coda cnode.h coda_fbsd.c coda_io.h coda_kernel.h coda_namecache.c coda_namecache.h coda_opstats.h coda_pioctl.h coda_psdev.c coda_psdev.h coda_subr.c coda_subr.h coda_venus.c coda_venus.h coda_vfsops.c coda_vfsops.h coda_vnops.c ... Message-ID: <200803111014.m2BAEmJT064770@repoman.freebsd.org> rwatson 2008-03-11 10:14:47 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda cnode.h coda_fbsd.c coda_io.h coda_kernel.h coda_namecache.c coda_namecache.h coda_opstats.h coda_pioctl.h coda_psdev.c coda_psdev.h coda_subr.c coda_subr.h coda_venus.c coda_venus.h coda_vfsops.c coda_vfsops.h coda_vnops.c coda_vnops.h Log: Merge cnode.h:1.25, coda_fbsd.c:1.48, coda_io.h:1.10, coda_kernel.h:1.7, coda_namecache.c:1.25, coda_namecache.h:1.14, coda_opstats.h:1.8, coda_pioctl.h:1.10, coda_psdev.c:1.43, coda_psdev.h:1.8, coda_subr.c:1.35, coda_subr.h:1.11, coda_venus.c:1.30, coda_venus.h:1.11, coda_vfsops.c:1.79, coda_vfsops.h:1.14, coda_vnops.c:1.92, coda_vnops.h:1.23 from HEAD to RELENG_7: Since we're now actively maintaining the Coda module in the FreeBSD source tree, restyle everything but coda.h (which is more explicitly shared across systems) into a closer approximation to style(9). Remove a few more unused function prototypes. Add or clarify some comments. Minor differences required in the MFC due to diverging VFS locking in HEAD vs RELENG_7. Revision Changes Path 1.21.2.4 +85 -78 src/sys/fs/coda/cnode.h 1.46.2.2 +40 -33 src/sys/fs/coda/coda_fbsd.c 1.9.2.1 +30 -16 src/sys/fs/coda/coda_io.h 1.6.2.1 +31 -30 src/sys/fs/coda/coda_kernel.h 1.23.2.2 +383 -439 src/sys/fs/coda/coda_namecache.c 1.11.2.2 +127 -118 src/sys/fs/coda/coda_namecache.h 1.7.2.1 +60 -57 src/sys/fs/coda/coda_opstats.h 1.9.2.1 +25 -26 src/sys/fs/coda/coda_pioctl.h 1.39.2.4 +494 -484 src/sys/fs/coda/coda_psdev.c 1.6.2.2 +14 -7 src/sys/fs/coda/coda_psdev.h 1.33.2.2 +325 -327 src/sys/fs/coda/coda_subr.c 1.10.2.1 +24 -20 src/sys/fs/coda/coda_subr.h 1.27.2.3 +494 -477 src/sys/fs/coda/coda_venus.c 1.10.2.1 +55 -105 src/sys/fs/coda/coda_venus.h 1.67.2.8 +308 -318 src/sys/fs/coda/coda_vfsops.c 1.12.2.2 +21 -18 src/sys/fs/coda/coda_vfsops.h 1.76.2.8 +1450 -1364 src/sys/fs/coda/coda_vnops.c 1.19.2.4 +45 -44 src/sys/fs/coda/coda_vnops.h From anderson at freebsd.org Tue Mar 11 13:10:55 2008 From: anderson at freebsd.org (Eric Anderson) Date: Tue Mar 11 13:11:04 2008 Subject: cvs commit: src/bin/cp cp.1 cp.c utils.c In-Reply-To: <200803101747.51354.jhb@freebsd.org> References: <200803101958.m2AJwfmm037050@repoman.freebsd.org> <20080310.144327.-856141340.imp@bsdimp.com> <200803101747.51354.jhb@freebsd.org> Message-ID: <47D67D14.8010106@freebsd.org> John Baldwin wrote: > On Monday 10 March 2008 04:43:27 pm M. Warner Losh wrote: >> In message: <200803101958.m2AJwfmm037050@repoman.freebsd.org> >> John Baldwin writes: >> : jhb 2008-03-10 19:58:41 UTC >> : >> : FreeBSD src repository >> : >> : Modified files: >> : bin/cp cp.1 cp.c utils.c >> : Log: >> : Add an '-a' option which is identical to specifying '-RpP' for >> : compatibility with other implementations. >> : >> : MFC after: 1 month >> >> Submitted by: David Frascone at Cisco >> >> right? > > Don't think so. I've had this patch for almost 2 years now on my laptop from > the first time this came up (back when -l was added IIRC). The recent thread > just reminded me to commit it. :) The recent thread also had a patch that > used -rpP whereas -RpP is more correct IMO. > Thanks John for committing this!! This makes life easier for rsnapshot users too. Eric p.s. - I think I was the original submitter along with the -l patch. From joel at FreeBSD.org Tue Mar 11 14:48:59 2008 From: joel at FreeBSD.org (Joel Dahl) Date: Tue Mar 11 14:49:26 2008 Subject: cvs commit: src/usr.bin/cpuset cpuset.1 Message-ID: <200803111448.m2BEmx1j041601@repoman.freebsd.org> joel 2008-03-11 14:48:59 UTC FreeBSD src repository (doc committer) Modified files: usr.bin/cpuset cpuset.1 Log: Fix minor typo. Revision Changes Path 1.2 +2 -2 src/usr.bin/cpuset/cpuset.1 From jhb at FreeBSD.org Tue Mar 11 15:05:55 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 11 15:06:01 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c Message-ID: <200803111505.m2BF5tJ4043633@repoman.freebsd.org> jhb 2008-03-11 15:05:55 UTC FreeBSD src repository Modified files: sys/dev/bge if_bge.c Log: Don't enable the workaround for the jitter bug on the 5722. Obtained from: Linux tg3 driver Revision Changes Path 1.204 +4 -3 src/sys/dev/bge/if_bge.c From emaste at FreeBSD.org Tue Mar 11 15:14:32 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Tue Mar 11 15:14:33 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_softdep.c Message-ID: <200803111514.m2BFEV93044346@repoman.freebsd.org> emaste 2008-03-11 15:14:31 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/ufs/ffs ffs_softdep.c Log: MFC minor cleanups: ffs_softdep.c 1.210 Add a newline to the printf message. ffs_softdep.c 1.214 (part) Update comment to match the DIAGNOSTIC -> INVARIANTS change. Revision Changes Path 1.181.2.23 +2 -2 src/sys/ufs/ffs/ffs_softdep.c From emax at FreeBSD.org Tue Mar 11 16:51:10 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 11 16:51:13 2008 Subject: cvs commit: src/usr.sbin/bluetooth/sdpd Makefile gn.c nap.c panu.c profile.c profile.h Message-ID: <200803111651.m2BGp8Cq096329@repoman.freebsd.org> emax 2008-03-11 16:51:07 UTC FreeBSD src repository Modified files: usr.sbin/bluetooth/sdpd Makefile profile.c profile.h Added files: usr.sbin/bluetooth/sdpd gn.c nap.c panu.c Log: Add support for the NAP, GN and PANU profiles to the sdpd(8). It should be mentioned that a somewhat similar patch was submitted by Rako < rako29 at gmail dot com > MFC after: 1 week Revision Changes Path 1.3 +3 -3 src/usr.sbin/bluetooth/sdpd/Makefile 1.1 +146 -0 src/usr.sbin/bluetooth/sdpd/gn.c (new) 1.1 +183 -0 src/usr.sbin/bluetooth/sdpd/nap.c (new) 1.1 +131 -0 src/usr.sbin/bluetooth/sdpd/panu.c (new) 1.3 +93 -2 src/usr.sbin/bluetooth/sdpd/profile.c 1.2 +6 -1 src/usr.sbin/bluetooth/sdpd/profile.h From delphij at FreeBSD.org Tue Mar 11 17:20:35 2008 From: delphij at FreeBSD.org (Xin LI) Date: Tue Mar 11 17:20:37 2008 Subject: cvs commit: src/etc/rc.d mountlate Message-ID: <200803111720.m2BHKZee098554@repoman.freebsd.org> delphij 2008-03-11 17:20:34 UTC FreeBSD src repository Modified files: etc/rc.d mountlate Log: root Revision Changes Path 1.4 +6 -0 src/etc/rc.d/mountlate From delphij at FreeBSD.org Tue Mar 11 17:21:14 2008 From: delphij at FreeBSD.org (Xin LI) Date: Tue Mar 11 17:21:17 2008 Subject: cvs commit: src/etc/rc.d root Message-ID: <200803111721.m2BHLED7098609@repoman.freebsd.org> delphij 2008-03-11 17:21:14 UTC FreeBSD src repository Modified files: etc/rc.d root Log: Do nextboot -D twice during boot. The first time in rc.d/root which ensures that we can remove the file as early as possible, but shut up nextboot at this moment if the operation is failed, because /boot is not necessarily a part of /; the newly added second run is placed in rc.d/mountlate after all filesystems were mounted. Discussed at: -rc@ Suggestions from: brooks, mtm MFC after: 1 month Revision Changes Path 1.17 +1 -1 src/etc/rc.d/root From delphij at FreeBSD.org Tue Mar 11 17:21:56 2008 From: delphij at FreeBSD.org (Xin LI) Date: Tue Mar 11 17:22:03 2008 Subject: cvs commit: src/etc/rc.d mountlate Message-ID: <200803111721.m2BHLt6H098656@repoman.freebsd.org> delphij 2008-03-11 17:21:55 UTC FreeBSD src repository Modified files: etc/rc.d mountlate Log: Forced commit to correct previous commit message: Do nextboot -D twice during boot. The first time in rc.d/root which ensures that we can remove the file as early as possible, but shut up nextboot at this moment if the operation is failed, because /boot is not necessarily a part of /; the newly added second run is placed in rc.d/mountlate after all filesystems were mounted. Discussed at: -rc@ Suggestions from: brooks, mtm MFC after: 1 month Revision Changes Path 1.5 +0 -0 src/etc/rc.d/mountlate From wearabnet at yahoo.ca Tue Mar 11 17:56:17 2008 From: wearabnet at yahoo.ca (Abdullah Ibn Hamad Al-Marri) Date: Tue Mar 11 17:56:22 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <151782.81617.qm@web33708.mail.mud.yahoo.com> ----- Original Message ---- > From: John Baldwin > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; cvs-all@FreeBSD.org > Sent: Tuesday, March 11, 2008 1:00:36 AM > Subject: cvs commit: src/sys/i386/cpufreq est.c > > jhb 2008-03-10 22:00:36 UTC > > FreeBSD src repository > > Modified files: > sys/i386/cpufreq est.c > Log: > - Don't execute cpuid to fetch the features. We already have the features > present in cpu_feature2. Also, use CPUID2_EST rather than a magic > number. > - Don't free the ACPI settings list in detach if we are going to fail the > request. Otherwise an attempt to kldunload est would free the array > but the driver would keep trying to use it. > > MFC after: 1 week > > Revision Changes Path > 1.14 +5 -5 src/sys/i386/cpufreq/est.c Hey John, Could I have a aatch to try it with Core 2 Quad and other cpus please? Regards, -Abdullah Ibn Hamad Al-Marri Arab Portal http://www.WeArab.Net/ ____________________________________________________________________________________ Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ From kaiw at FreeBSD.org Tue Mar 11 18:35:52 2008 From: kaiw at FreeBSD.org (Kai Wang) Date: Tue Mar 11 18:35:58 2008 Subject: cvs commit: src/usr.bin/ar ar.1 ar.c ar.h write.c Message-ID: <200803111835.m2BIZq7c007409@repoman.freebsd.org> kaiw 2008-03-11 18:35:52 UTC FreeBSD src repository Modified files: usr.bin/ar ar.1 ar.c ar.h write.c Log: GNU ar did NOT implment option -q as a synonym of -r as the manual page stated, thus BSD ar(1) option -q, which was implemented based on the GNU ar manual page, turns out to be incompatible with GNU ar -q. This change will make BSD ar(1) -q a *REAL* GNU ar -q: 1. It will update symbol table. (same as unfixed version) 2. It will NOT compare new members spcified in the command line args with existing members, instead, append them directly. Reported by: Johannes 5 Joemann Reported by: Timothy Bourke Tested by: Johannes 5 Joemann Reviewed by: jkoshy Approved by: jkoshy (mentor) Revision Changes Path 1.20 +2 -2 src/usr.bin/ar/ar.1 1.18 +2 -1 src/usr.bin/ar/ar.c 1.3 +1 -0 src/usr.bin/ar/ar.h 1.5 +20 -3 src/usr.bin/ar/write.c From uspoerlein at gmail.com Tue Mar 11 18:58:09 2008 From: uspoerlein at gmail.com (Ulrich Spoerlein) Date: Tue Mar 11 18:58:16 2008 Subject: cvs commit: src/sys/dev/vr if_vr.c if_vrreg.h In-Reply-To: <200803110451.m2B4pMLB095467@repoman.freebsd.org> References: <200803110451.m2B4pMLB095467@repoman.freebsd.org> Message-ID: <20080311182559.GC1535@roadrunner.spoerlein.net> On Tue, 11.03.2008 at 04:51:22 +0000, Pyun YongHyeon wrote: > yongari 2008-03-11 04:51:22 UTC > > FreeBSD src repository > > Modified files: > sys/dev/vr if_vr.c if_vrreg.h Awesome! Cheers, Ulrich Spoerlein -- It is better to remain silent and be thought a fool, than to speak, and remove all doubt. From mav at FreeBSD.org Tue Mar 11 19:03:35 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 11 19:03:38 2008 Subject: cvs commit: src/sys/netgraph ng_ksocket.c Message-ID: <200803111903.m2BJ3ZnK011001@repoman.freebsd.org> mav 2008-03-11 19:03:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_ksocket.c Log: MFC rev. 1.61 Send only one incoming notification at a time to reduce queue trashing and improve performance. Remove waitflag argument from ng_ksocket_incoming2(), it means nothing as function call was queued by netgraph. Remove node validity check, as node validity guarantied by netgraph. Update comments. Revision Changes Path 1.60.2.1 +26 -32 src/sys/netgraph/ng_ksocket.c From mav at FreeBSD.org Tue Mar 11 19:04:42 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 11 19:04:48 2008 Subject: cvs commit: src/sys/netgraph ng_ksocket.c Message-ID: <200803111904.m2BJ4g1K011053@repoman.freebsd.org> mav 2008-03-11 19:04:42 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netgraph ng_ksocket.c Log: MFC rev. 1.61 Send only one incoming notification at a time to reduce queue trashing and improve performance. Remove waitflag argument from ng_ksocket_incoming2(), it means nothing as function call was queued by netgraph. Remove node validity check, as node validity guarantied by netgraph. Update comments. Revision Changes Path 1.55.2.5 +26 -32 src/sys/netgraph/ng_ksocket.c From mav at FreeBSD.org Tue Mar 11 19:09:57 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 11 19:10:01 2008 Subject: cvs commit: src/sys/netgraph ng_l2tp.c Message-ID: <200803111909.m2BJ9ucU011246@repoman.freebsd.org> mav 2008-03-11 19:09:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_l2tp.c Log: MFC rev. 1.23-1.24 To avoid control data losses, do not acknowledge recieving of control packet if netgraph reported error while delivering to destination. Reset 'next send' counter to the last requested by peer on ack timeout, to resend all subsequest packets after lost one again without additional hints. Revision Changes Path 1.17.2.3 +45 -54 src/sys/netgraph/ng_l2tp.c From mav at FreeBSD.org Tue Mar 11 19:14:10 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 11 19:14:18 2008 Subject: cvs commit: src/sys/netgraph ng_l2tp.c Message-ID: <200803111914.m2BJEAEA011533@repoman.freebsd.org> mav 2008-03-11 19:14:10 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netgraph ng_l2tp.c Log: MFC rev. 1.23-1.24 To avoid control data losses, do not acknowledge recieving of control packet if netgraph reported error while delivering to destination. Reset 'next send' counter to the last requested by peer on ack timeout, to resend all subsequest packets after lost one again without additional hints. Revision Changes Path 1.14.2.4 +45 -54 src/sys/netgraph/ng_l2tp.c From remko at FreeBSD.org Tue Mar 11 19:19:15 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Tue Mar 11 19:19:16 2008 Subject: cvs commit: src/tools/tools/tinybsd tinybsd Message-ID: <200803111919.m2BJJEd8011834@repoman.freebsd.org> remko 2008-03-11 19:19:14 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) tools/tools/tinybsd tinybsd Log: MFC rev 1.8 tinybsd Improve handling of the tinybsd application wrt ports. Ports that have dependencies or need to install a new user/group, are not problematic anymore. PR: 121367 Submitted by: Richard Arends < richard at unixguru dot nl > MFC after: 3 days Approved by: imp (mentor, implicit) Revision Changes Path 1.7.2.1 +41 -53 src/tools/tools/tinybsd/tinybsd From marcel at FreeBSD.org Tue Mar 11 19:58:53 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Tue Mar 11 19:58:55 2008 Subject: cvs commit: src/sys/powerpc/powerpc intr_machdep.c Message-ID: <200803111958.m2BJwqre013258@repoman.freebsd.org> marcel 2008-03-11 19:58:52 UTC FreeBSD src repository Modified files: sys/powerpc/powerpc intr_machdep.c Log: In intr_lookup(), when adding an IRQ to powerpc_intrs[], also set a default name. If the IRQ is added as a consequence of configurating the IRQ without there ever being a handler assigned to it, we will not have a name. This breaks the fragile intrcnt/intrnames logic. Revision Changes Path 1.17 +10 -7 src/sys/powerpc/powerpc/intr_machdep.c From jhb at FreeBSD.org Tue Mar 11 20:10:38 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 11 20:10:43 2008 Subject: cvs commit: src/sys/amd64/amd64 amd64_mem.c src/sys/amd64/include specialreg.h src/sys/i386/i386 i686_mem.c src/sys/i386/include specialreg.h Message-ID: <200803112010.m2BKAbut015748@repoman.freebsd.org> jhb 2008-03-11 20:10:37 UTC FreeBSD src repository Modified files: sys/amd64/amd64 amd64_mem.c sys/amd64/include specialreg.h sys/i386/i386 i686_mem.c sys/i386/include specialreg.h Log: Add constants for the various fields in MTRR registers. MFC after: 1 week Verified by: md5(1) Revision Changes Path 1.26 +12 -11 src/sys/amd64/amd64/amd64_mem.c 1.44 +15 -0 src/sys/amd64/include/specialreg.h 1.25 +12 -11 src/sys/i386/i386/i686_mem.c 1.43 +15 -0 src/sys/i386/include/specialreg.h From wilko at FreeBSD.org Tue Mar 11 20:21:16 2008 From: wilko at FreeBSD.org (Wilko Bulte) Date: Tue Mar 11 20:21:22 2008 Subject: cvs commit: CVSROOT access Message-ID: <200803112021.m2BKLGe5016306@repoman.freebsd.org> wilko 2008-03-11 20:21:15 UTC FreeBSD src repository Modified files: . access Log: reduce email during absence Approved by: core (me) Revision Changes Path 1.865 +1 -1 CVSROOT/access From jeremie at le-hen.org Tue Mar 11 21:32:17 2008 From: jeremie at le-hen.org (Jeremie Le Hen) Date: Tue Mar 11 21:32:21 2008 Subject: cvs commit [HEAD] src/sys/boot/i386/btx/btx btx.S In-Reply-To: <200803102143.m2ALhVbF052196@repoman.freebsd.org> References: <200803102143.m2ALhVbF052196@repoman.freebsd.org> Message-ID: <20080311211059.GJ39998@obiwan.tataz.chchile.org> Hi, On Mon, Mar 10, 2008 at 09:43:31PM +0000, John Baldwin wrote: > jhb 2008-03-10 21:43:31 UTC > > FreeBSD src repository > > Modified files: > sys/boot/i386/btx/btx btx.S > Log: > Change the BTX kernel to drop all the way out to real mode to invoke BIOS > routines (V86 requests from the client and hardware interrupt handlers): > - Install trampoline real mode interrupt handlers at IDT vectors 0x20-0x2f > to handle hardware interrupts by invoking the appropriate vector (0x8-0xf > or 0x70-0x78). This allows the 8259As to use vectors 0x20-0x2f in real > mode as well as protected mode will ensuring that the master 8259A > doesn't share IDT space with CPU exceptions in protected mode. > - Since we don't need to reserve space for page tables and a page directory > anymore since dropping paging support, move the TSS and protected mode > IDT up by 16k. Grow the ring 1 link stack by 16k as a result. > - Repurpose the ring 1 link stack to be used as a real mode stack when > invoking real mode routines either via a V86 request or a hardware > interrupts. This simplifies a few things as we avoid disturbing the > original user stack. > - Add some more block comments to explain how the code interacts with the > V86 structure as this wasn't immediately obvious from the prior comments > (e.g. that we explicitly copy the seg regs for real mode out of the V86 > struct onto the stack to be popped off when going into real mode, etc.). > Also, document some of the stack frames we create going to real mode and > back. > - Remove all of the virtual 86 related code including having to simulate > various instructions and BIOS calls on a trap from virtual 86 mode. > - Explicitly panic if a user client attempts to perform a V86 CALL > request that isn't a far call. > - Bump version to 1.2. I am not able to test this right now, but I wonder if with this change FreeBSD will be able to be booted in a fully virtualized Xen DomU. At the time I tested it, about 2 month ago, ISTR FreeBSD stopped with "BTX halted". If someone is being able to test this (I suppose the easiest way is to wait from the next monthly -CURRENT snapshot), I'd be glad to here about the result. Thanks. Regards, -- Jeremie Le Hen < jeremie at le-hen dot org >< ttz at chchile dot org > From jhb at FreeBSD.org Tue Mar 11 21:41:37 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 11 21:41:48 2008 Subject: cvs commit: src/sys/amd64/amd64 amd64_mem.c src/sys/i386/i386 i686_mem.c k6_mem.c Message-ID: <200803112141.m2BLfahF036840@repoman.freebsd.org> jhb 2008-03-11 21:41:36 UTC FreeBSD src repository Modified files: sys/amd64/amd64 amd64_mem.c sys/i386/i386 i686_mem.c k6_mem.c Log: Style(9) these files. No changes in the compiled code. (Verified by diff'ing objdump -d output). Revision Changes Path 1.27 +451 -413 src/sys/amd64/amd64/amd64_mem.c 1.26 +466 -423 src/sys/i386/i386/i686_mem.c 1.12 +36 -35 src/sys/i386/i386/k6_mem.c From mav at FreeBSD.org Tue Mar 11 21:58:49 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 11 21:58:51 2008 Subject: cvs commit: src/sys/netgraph netgraph.h ng_base.c ng_socket.c Message-ID: <200803112158.m2BLwnLJ039871@repoman.freebsd.org> mav 2008-03-11 21:58:49 UTC FreeBSD src repository Modified files: sys/netgraph netgraph.h ng_base.c ng_socket.c Log: Improve apply callback error reporting: Before this patch callback returned result of the last finished call chain. Now it returns last nonzero result from all call chain results in this request. As soon as this improvement gives reliable error reporting, it is now possible to remove dirty workaround in ng_socket, made to return ENOBUFS error statuses of request-response operations. That workaround was responsible for returning ENOBUFS errors to completely unrelated requests working at the same time on socket. Revision Changes Path 1.72 +2 -1 src/sys/netgraph/netgraph.h 1.150 +25 -8 src/sys/netgraph/ng_base.c 1.85 +1 -1 src/sys/netgraph/ng_socket.c From delphij at FreeBSD.org Tue Mar 11 22:12:59 2008 From: delphij at FreeBSD.org (Xin LI) Date: Tue Mar 11 22:13:03 2008 Subject: cvs commit: src/contrib/tcsh FREEBSD-Xlist FREEBSD-upgrade Fixes Imakefile MAKEDIFFS MAKESHAR Makefile.in Makefile.std Makefile.vms README complete.tcsh config.h.in config_f.h configure configure.in ed.chared.c ed.decls.h ed.defns.c ed.h ed.init.c ed.inputl.c ... Message-ID: <200803112212.m2BMCwtl042902@repoman.freebsd.org> delphij 2008-03-11 22:12:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) contrib/tcsh Fixes Imakefile MAKEDIFFS MAKESHAR Makefile.in Makefile.std Makefile.vms README complete.tcsh config.h.in config_f.h configure configure.in ed.chared.c ed.decls.h ed.defns.c ed.h ed.init.c ed.inputl.c ed.refresh.c ed.screen.c ed.term.c ed.term.h ed.xmap.c eight-bit.me gethost.c glob.c glob.h host.defs imake.config ma.setp.c mi.termios.c mi.varargs.h patchlevel.h pathnames.h sh.c sh.char.c sh.char.h sh.decls.h sh.dir.c sh.dir.h sh.dol.c sh.err.c sh.exec.c sh.exp.c sh.file.c sh.func.c sh.glob.c sh.h sh.hist.c sh.init.c sh.lex.c sh.misc.c sh.parse.c sh.print.c sh.proc.c sh.proc.h sh.sem.c sh.set.c sh.time.c sh.types.h snames.h tc.alloc.c tc.bind.c tc.const.c tc.decls.h tc.disc.c tc.func.c tc.h tc.nls.c tc.nls.h tc.os.c tc.os.h tc.printf.c tc.prompt.c tc.sched.c tc.sig.c tc.sig.h tc.str.c tc.vers.c tc.wait.h tc.who.c tcsh.man tcsh.man2html termcap.vms tw.color.c tw.comp.c tw.decls.h tw.h tw.help.c tw.init.c tw.parse.c tw.spell.c vms.termcap.c contrib/tcsh/config bsd4.4 contrib/tcsh/nls Makefile contrib/tcsh/nls/C set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/et set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/finnish set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/french set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/german set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/greek set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/italian set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/ja set1 set10 set11 set12 set13 set15 set16 set17 set18 set2 set21 set24 set29 set3 set30 set4 set5 set6 set7 set8 contrib/tcsh/nls/pl set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/russian set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/spanish set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 contrib/tcsh/nls/ukrainian set1 set10 set11 set12 set13 set14 set15 set16 set17 set18 set19 set2 set20 set21 set22 set23 set24 set25 set26 set27 set29 set3 set30 set31 set4 set5 set6 set7 set8 set9 bin/csh Makefile config.h config_p.h Removed files: (Branch: RELENG_6) contrib/tcsh FREEBSD-Xlist FREEBSD-upgrade Log: MFC: tcsh 6.15.00. Revision Changes Path 1.36.2.2 +1 -1 src/bin/csh/Makefile 1.12.2.2 +41 -10 src/bin/csh/config.h 1.11.2.2 +1 -15 src/bin/csh/config_p.h 1.2.8.2 +0 -8 src/contrib/tcsh/FREEBSD-Xlist (dead) 1.1.14.1 +0 -13 src/contrib/tcsh/FREEBSD-upgrade (dead) 1.1.1.7.2.1 +99 -0 src/contrib/tcsh/Fixes 1.1.1.7.2.1 +1 -1 src/contrib/tcsh/Imakefile 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/MAKEDIFFS 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/MAKESHAR 1.1.1.7.2.1 +79 -37 src/contrib/tcsh/Makefile.in 1.1.1.6.2.1 +1 -1 src/contrib/tcsh/Makefile.std 1.1.1.6.2.1 +1 -1 src/contrib/tcsh/Makefile.vms 1.1.1.7.2.1 +13 -19 src/contrib/tcsh/README 1.1.1.7.2.1 +1 -1 src/contrib/tcsh/complete.tcsh 1.1.1.2.2.1 +38 -10 src/contrib/tcsh/config.h.in 1.1.1.5.2.1 +1 -14 src/contrib/tcsh/config/bsd4.4 1.1.1.5.2.1 +6 -12 src/contrib/tcsh/config_f.h 1.1.1.5.2.1 +1533 -384 src/contrib/tcsh/configure 1.1.1.5.2.1 +46 -16 src/contrib/tcsh/configure.in 1.1.1.6.2.1 +530 -571 src/contrib/tcsh/ed.chared.c 1.1.1.5.2.1 +202 -197 src/contrib/tcsh/ed.decls.h 1.1.1.5.2.1 +28 -14 src/contrib/tcsh/ed.defns.c 1.1.1.5.2.1 +30 -28 src/contrib/tcsh/ed.h 1.1.1.6.2.1 +42 -80 src/contrib/tcsh/ed.init.c 1.1.1.5.2.1 +112 -102 src/contrib/tcsh/ed.inputl.c 1.1.1.6.2.1 +85 -124 src/contrib/tcsh/ed.refresh.c 1.1.1.5.2.1 +233 -263 src/contrib/tcsh/ed.screen.c 1.1.1.6.2.1 +40 -52 src/contrib/tcsh/ed.term.c 1.1.1.3.2.1 +1 -1 src/contrib/tcsh/ed.term.h 1.1.1.5.2.1 +114 -197 src/contrib/tcsh/ed.xmap.c 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/eight-bit.me 1.1.1.4.2.1 +8 -13 src/contrib/tcsh/gethost.c 1.1.1.6.2.1 +140 -220 src/contrib/tcsh/glob.c 1.1.1.5.2.1 +9 -11 src/contrib/tcsh/glob.h 1.1.1.7.2.1 +7 -9 src/contrib/tcsh/host.defs 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/imake.config 1.1.1.3.2.1 +42 -66 src/contrib/tcsh/ma.setp.c 1.1.1.2.14.1 +2 -2 src/contrib/tcsh/mi.termios.c 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/mi.varargs.h 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set10 1.1.1.2.2.1 +1 -1 src/contrib/tcsh/nls/C/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set16 1.1.1.3.8.1 +1 -1 src/contrib/tcsh/nls/C/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set2 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set29 1.1.1.2.14.1 +3 -1 src/contrib/tcsh/nls/C/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/C/set9 1.1.1.4.8.1 +2 -1 src/contrib/tcsh/nls/Makefile 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/et/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set2 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set29 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/et/set9 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/finnish/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set13 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/finnish/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/finnish/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set2 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/finnish/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set21 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/finnish/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/finnish/set3 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/finnish/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/finnish/set9 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set10 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/french/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/french/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set2 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/french/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/french/set9 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set1 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/german/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/german/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set2 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/german/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/german/set9 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/greek/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set2 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/greek/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/greek/set9 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/italian/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set2 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/italian/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/italian/set9 1.1.1.2.2.1 +1 -1 src/contrib/tcsh/nls/ja/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/ja/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set2 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set29 1.1.1.1.2.1 +1 -1 src/contrib/tcsh/nls/ja/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set30 1.1.1.1.2.1 +10 -10 src/contrib/tcsh/nls/ja/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set6 1.1.1.1.2.1 +1 -1 src/contrib/tcsh/nls/ja/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/ja/set8 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set12 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set16 1.1.1.3.8.1 +1 -1 src/contrib/tcsh/nls/pl/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set18 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set19 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set2 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set20 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set21 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set22 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set25 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set3 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set31 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set6 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/pl/set8 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/pl/set9 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set1 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set12 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set16 1.1.1.3.8.1 +1 -1 src/contrib/tcsh/nls/russian/set17 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set18 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set19 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set2 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set21 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set22 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set25 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set26 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set3 1.1.1.3.8.1 +1 -1 src/contrib/tcsh/nls/russian/set30 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set31 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set5 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set6 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/russian/set8 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/russian/set9 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set1 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set10 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set11 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set12 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set13 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set14 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set15 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set16 1.1.1.2.8.1 +1 -1 src/contrib/tcsh/nls/spanish/set17 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set18 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set19 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set2 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/spanish/set20 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set21 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set22 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set23 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set24 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set25 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set26 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set27 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set29 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/spanish/set3 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set30 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set31 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set4 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set5 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set6 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set7 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set8 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/nls/spanish/set9 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set1 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set10 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set11 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set12 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set13 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set14 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set15 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set16 1.1.1.3.8.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set17 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set18 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set19 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set2 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set20 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set21 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set22 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set23 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set24 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set25 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set26 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set27 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set29 1.1.1.3.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set3 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set30 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set31 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set4 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set5 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set6 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set7 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set8 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/nls/ukrainian/set9 1.1.1.7.2.1 +3 -3 src/contrib/tcsh/patchlevel.h 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/pathnames.h 1.1.1.8.2.1 +296 -469 src/contrib/tcsh/sh.c 1.1.1.4.14.1 +2 -2 src/contrib/tcsh/sh.char.c 1.1.1.5.2.1 +28 -33 src/contrib/tcsh/sh.char.h 1.1.1.5.2.1 +280 -242 src/contrib/tcsh/sh.decls.h 1.1.1.7.2.1 +312 -312 src/contrib/tcsh/sh.dir.c 1.1.1.2.14.1 +1 -1 src/contrib/tcsh/sh.dir.h 1.1.1.9.2.1 +293 -407 src/contrib/tcsh/sh.dol.c 1.1.1.4.2.1 +181 -52 src/contrib/tcsh/sh.err.c 1.1.1.6.2.1 +203 -241 src/contrib/tcsh/sh.exec.c 1.1.1.3.2.1 +98 -184 src/contrib/tcsh/sh.exp.c 1.1.1.5.2.1 +155 -253 src/contrib/tcsh/sh.file.c 1.1.1.8.2.1 +427 -515 src/contrib/tcsh/sh.func.c 1.1.1.7.2.1 +263 -430 src/contrib/tcsh/sh.glob.c 1.1.1.8.2.1 +79 -155 src/contrib/tcsh/sh.h 1.1.1.4.2.1 +79 -104 src/contrib/tcsh/sh.hist.c 1.1.1.4.2.1 +17 -20 src/contrib/tcsh/sh.init.c 1.1.1.6.2.1 +283 -426 src/contrib/tcsh/sh.lex.c 1.1.1.4.2.1 +281 -128 src/contrib/tcsh/sh.misc.c 1.1.1.3.2.1 +97 -109 src/contrib/tcsh/sh.parse.c 1.1.1.5.2.1 +36 -41 src/contrib/tcsh/sh.print.c 1.1.1.6.2.1 +233 -435 src/contrib/tcsh/sh.proc.c 1.1.1.3.2.1 +1 -6 src/contrib/tcsh/sh.proc.h 1.1.1.6.2.1 +138 -174 src/contrib/tcsh/sh.sem.c 1.1.1.8.2.1 +150 -192 src/contrib/tcsh/sh.set.c 1.1.1.6.2.1 +23 -51 src/contrib/tcsh/sh.time.c 1.1.1.6.2.1 +3 -26 src/contrib/tcsh/sh.types.h 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/snames.h 1.1.1.4.2.1 +63 -82 src/contrib/tcsh/tc.alloc.c 1.1.1.4.2.1 +109 -648 src/contrib/tcsh/tc.bind.c 1.1.1.8.2.1 +9 -3 src/contrib/tcsh/tc.const.c 1.1.1.6.2.1 +158 -169 src/contrib/tcsh/tc.decls.h 1.1.1.3.2.1 +4 -6 src/contrib/tcsh/tc.disc.c 1.1.1.7.2.1 +413 -640 src/contrib/tcsh/tc.func.c 1.1.1.3.2.1 +31 -2 src/contrib/tcsh/tc.h 1.1.1.1.2.1 +36 -221 src/contrib/tcsh/tc.nls.c 1.1.1.1.2.1 +7 -46 src/contrib/tcsh/tc.nls.h 1.1.1.5.2.1 +196 -285 src/contrib/tcsh/tc.os.c 1.1.1.5.2.1 +29 -50 src/contrib/tcsh/tc.os.h 1.1.1.5.2.1 +82 -89 src/contrib/tcsh/tc.printf.c 1.1.1.9.2.1 +191 -219 src/contrib/tcsh/tc.prompt.c 1.1.1.4.2.1 +42 -68 src/contrib/tcsh/tc.sched.c 1.1.1.4.2.1 +67 -336 src/contrib/tcsh/tc.sig.c 1.1.1.4.2.1 +58 -113 src/contrib/tcsh/tc.sig.h 1.1.1.4.2.1 +196 -119 src/contrib/tcsh/tc.str.c 1.1.1.3.2.1 +15 -12 src/contrib/tcsh/tc.vers.c 1.1.1.4.2.1 +1 -1 src/contrib/tcsh/tc.wait.h 1.1.1.6.2.1 +67 -93 src/contrib/tcsh/tc.who.c 1.1.1.8.2.1 +47 -22 src/contrib/tcsh/tcsh.man 1.1.1.4.2.1 +1 -1 src/contrib/tcsh/tcsh.man2html 1.1.1.1.16.1 +1 -1 src/contrib/tcsh/termcap.vms 1.1.1.7.2.1 +45 -49 src/contrib/tcsh/tw.color.c 1.1.1.4.2.1 +110 -106 src/contrib/tcsh/tw.comp.c 1.1.1.4.2.1 +68 -62 src/contrib/tcsh/tw.decls.h 1.1.1.4.2.1 +1 -3 src/contrib/tcsh/tw.h 1.1.1.5.2.1 +62 -53 src/contrib/tcsh/tw.help.c 1.1.1.4.2.1 +208 -264 src/contrib/tcsh/tw.init.c 1.1.1.7.2.1 +663 -617 src/contrib/tcsh/tw.parse.c 1.1.1.3.2.1 +40 -42 src/contrib/tcsh/tw.spell.c 1.1.1.3.2.1 +11 -24 src/contrib/tcsh/vms.termcap.c From delphij at FreeBSD.org Tue Mar 11 22:40:04 2008 From: delphij at FreeBSD.org (Xin LI) Date: Tue Mar 11 22:40:07 2008 Subject: cvs commit: src/release/doc/zh_CN.GB2312/errata article.sgml Message-ID: <200803112240.m2BMe46x045493@repoman.freebsd.org> delphij 2008-03-11 22:40:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) release/doc/zh_CN.GB2312/errata article.sgml Log: MFen 1.73 -> 1.73.4.3 Revision Changes Path 1.5.2.1 +50 -29 src/release/doc/zh_CN.GB2312/errata/article.sgml From peter at FreeBSD.org Wed Mar 12 00:13:50 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Mar 12 00:13:53 2008 Subject: cvs commit: src/usr.bin/ministat ministat.1 Message-ID: <200803120013.m2C0DnDF053478@repoman.freebsd.org> peter 2008-03-12 00:13:49 UTC FreeBSD src repository Modified files: usr.bin/ministat ministat.1 Log: Fix apparent typo. The permitted confidence values include 95%, not 85%. Revision Changes Path 1.2 +1 -1 src/usr.bin/ministat/ministat.1 From citrin at citrin.ru Wed Mar 12 01:42:54 2008 From: citrin at citrin.ru (Anton Yuzhaninov) Date: Wed Mar 12 01:43:00 2008 Subject: cvs commit: src/contrib/tcsh FREEBSD-Xlist FREEBSD-upgrade Fixes Imakefile MAKEDIFFS MAKESHAR Makefile.in Makefile.std Makefile.vms README complete.tcsh config.h.in config_f.h configure configure.in ed.chared.c ed.decls.h ed.defns.c ed.h ed.init.c ed.inputl.c ... In-Reply-To: <200803112212.m2BMCwtl042902@repoman.freebsd.org> References: <200803112212.m2BMCwtl042902@repoman.freebsd.org> Message-ID: <47D7325E.1000009@citrin.ru> On 12.03.2008 1:12, Xin LI wrote: > delphij 2008-03-11 22:12:58 UTC > > ... > > Log: > MFC: tcsh 6.15.00. > > Revision Changes Path > ... > 1.1.1.6.2.1 +283 -426 src/contrib/tcsh/sh.lex.c > ... Now tcsh postcmd broken also in RELENG_6. for more info see my messages in current@: http://docs.freebsd.org/cgi/mid.cgi?47BB4E5D.7010505 -- WBR, Anton Yuzhaninov From delphij at FreeBSD.org Wed Mar 12 01:52:48 2008 From: delphij at FreeBSD.org (Xin LI) Date: Wed Mar 12 01:52:52 2008 Subject: cvs commit: src/contrib/tcsh sh.lex.c Message-ID: <200803120152.m2C1qmwY073481@repoman.freebsd.org> delphij 2008-03-12 01:52:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) contrib/tcsh sh.lex.c Log: Merge revision 1.1.1.8 date: 2007/04/03 15:51:53; author: mp; state: Exp; lines: +2 -1 Import vendor patch to fix postcmd regression in tcsh-6.15.00. Revision Changes Path 1.1.1.6.2.2 +2 -1 src/contrib/tcsh/sh.lex.c From tmclaugh at FreeBSD.org Wed Mar 12 02:09:23 2008 From: tmclaugh at FreeBSD.org (Tom McLaughlin) Date: Wed Mar 12 02:09:27 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200803120209.m2C29NUY075183@repoman.freebsd.org> tmclaugh 2008-03-12 02:09:22 UTC FreeBSD src repository (ports committer) Modified files: sbin/mount mount.8 Log: - Update with a better example which shows that options specific to a file system may be passed using -o. Approved by: remko, rodrigc Revision Changes Path 1.87 +3 -3 src/sbin/mount/mount.8 From delphij at delphij.net Wed Mar 12 02:14:27 2008 From: delphij at delphij.net (Xin LI) Date: Wed Mar 12 02:14:31 2008 Subject: cvs commit: src/contrib/tcsh FREEBSD-Xlist FREEBSD-upgrade Fixes Imakefile MAKEDIFFS MAKESHAR Makefile.in Makefile.std Makefile.vms README complete.tcsh config.h.in config_f.h configure configure.in ed.chared.c ed.decls.h ed.defns.c ed.h ed.init.c ed.inputl.c ... In-Reply-To: <47D7325E.1000009@citrin.ru> References: <200803112212.m2BMCwtl042902@repoman.freebsd.org> <47D7325E.1000009@citrin.ru> Message-ID: <47D73C73.8030400@delphij.net> Anton Yuzhaninov wrote: > On 12.03.2008 1:12, Xin LI wrote: >> delphij 2008-03-11 22:12:58 UTC >> >> ... >> >> Log: >> MFC: tcsh 6.15.00. >> Revision Changes Path > > ... >> 1.1.1.6.2.1 +283 -426 src/contrib/tcsh/sh.lex.c >> ... > > Now tcsh postcmd broken also in RELENG_6. > > for more info see my messages in current@: > http://docs.freebsd.org/cgi/mid.cgi?47BB4E5D.7010505 Having checked with the original tcsh 6.15.00 patchlevel 1 I believe that you are correct. I have added the fix. Cheers, -- Xin LI http://www.delphij.net/ FreeBSD - The Power to Serve! From davidxu at FreeBSD.org Wed Mar 12 02:33:18 2008 From: davidxu at FreeBSD.org (David Xu) Date: Wed Mar 12 02:33:19 2008 Subject: cvs commit: src/lib/libc/gen Makefile.inc sem_timedwait.3 Message-ID: <200803120233.m2C2XHef078516@repoman.freebsd.org> davidxu 2008-03-12 02:33:17 UTC FreeBSD src repository Modified files: lib/libc/gen Makefile.inc Added files: lib/libc/gen sem_timedwait.3 Log: Add manual for function sem_timedwait(). Reviewed by: ru, deischen Revision Changes Path 1.132 +1 -1 src/lib/libc/gen/Makefile.inc 1.1 +122 -0 src/lib/libc/gen/sem_timedwait.3 (new) From davidxu at FreeBSD.org Wed Mar 12 02:37:32 2008 From: davidxu at FreeBSD.org (David Xu) Date: Wed Mar 12 02:37:33 2008 Subject: cvs commit: src/lib/libc/gen sem_timedwait.3 Message-ID: <200803120237.m2C2bVNq079427@repoman.freebsd.org> davidxu 2008-03-12 02:37:31 UTC FreeBSD src repository Modified files: lib/libc/gen sem_timedwait.3 Log: Add missing comma. Revision Changes Path 1.2 +1 -1 src/lib/libc/gen/sem_timedwait.3 From kientzle at FreeBSD.org Wed Mar 12 04:20:31 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Wed Mar 12 04:20:33 2008 Subject: cvs commit: src/lib/libarchive archive_endian.h archive_read_open_filename.c archive_read_support_compression_bzip2.c archive_read_support_compression_gzip.c archive_read_support_format_ar.c archive_read_support_format_mtree.c archive_read_support_format_zip.c archive_write_disk.c archive_write_open_filename.c archive_write_set_compression_gzip.c config_freebsd.h ... Message-ID: <200803120420.m2C4KUus085266@repoman.freebsd.org> kientzle 2008-03-12 04:20:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libarchive archive_endian.h archive_read_open_filename.c archive_read_support_compression_bzip2.c archive_read_support_compression_gzip.c archive_read_support_format_ar.c archive_read_support_format_mtree.c archive_read_support_format_zip.c archive_write_disk.c archive_write_open_filename.c archive_write_set_compression_gzip.c config_freebsd.h lib/libarchive/test main.c test.h Log: MFC a number of minor edits and portability fixups. This is mostly just to catch 7-STABLE up with -CURRENT before the next big round of changes. Revision Changes Path 1.1.2.2 +18 -18 src/lib/libarchive/archive_endian.h 1.20.2.1 +5 -1 src/lib/libarchive/archive_read_open_filename.c 1.16.2.1 +1 -1 src/lib/libarchive/archive_read_support_compression_bzip2.c 1.15.2.1 +1 -1 src/lib/libarchive/archive_read_support_compression_gzip.c 1.6.4.2 +2 -4 src/lib/libarchive/archive_read_support_format_ar.c 1.1.2.2 +8 -2 src/lib/libarchive/archive_read_support_format_mtree.c 1.14.2.4 +28 -35 src/lib/libarchive/archive_read_support_format_zip.c 1.17.2.2 +67 -24 src/lib/libarchive/archive_write_disk.c 1.19.4.1 +5 -1 src/lib/libarchive/archive_write_open_filename.c 1.14.2.2 +19 -0 src/lib/libarchive/archive_write_set_compression_gzip.c 1.5.2.1 +7 -0 src/lib/libarchive/config_freebsd.h 1.8.2.2 +7 -7 src/lib/libarchive/test/main.c 1.6.2.2 +5 -2 src/lib/libarchive/test/test.h From kientzle at FreeBSD.org Wed Mar 12 04:26:45 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Wed Mar 12 04:27:04 2008 Subject: cvs commit: src/lib/libarchive archive_entry.c Message-ID: <200803120426.m2C4Qi7X085485@repoman.freebsd.org> kientzle 2008-03-12 04:26:44 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry.c Log: Typo, thanks to: Jeff Johnson. MFC after: 3 days Revision Changes Path 1.46 +1 -1 src/lib/libarchive/archive_entry.c From kientzle at FreeBSD.org Wed Mar 12 04:47:38 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Wed Mar 12 04:47:43 2008 Subject: cvs commit: src/lib/libarchive archive_entry.c src/lib/libarchive/test test_entry.c Message-ID: <200803120447.m2C4lcrE095216@repoman.freebsd.org> kientzle 2008-03-12 04:47:38 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry.c lib/libarchive/test test_entry.c Log: Tighten up the semantics of acl_next() and xattr_next() when you hit the end of the ACL or xattr list. Thanks to: Jeff Johnson for pointing out the obvious typo Revision Changes Path 1.47 +5 -0 src/lib/libarchive/archive_entry.c 1.3 +22 -0 src/lib/libarchive/test/test_entry.c From kientzle at FreeBSD.org Wed Mar 12 04:58:36 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Wed Mar 12 04:58:37 2008 Subject: cvs commit: src/lib/libarchive archive_read.c archive_read_private.h Message-ID: <200803120458.m2C4wXx3095570@repoman.freebsd.org> kientzle 2008-03-12 04:58:32 UTC FreeBSD src repository Modified files: lib/libarchive archive_read.c archive_read_private.h Log: Remove some unused fields from the private archive_read structure (left over from when the unified read/write structure was copied to form separate read and write structures) and eliminate the pointless initialization of a couple of the unused fields. Revision Changes Path 1.38 +0 -15 src/lib/libarchive/archive_read.c 1.5 +0 -37 src/lib/libarchive/archive_read_private.h From jeff at FreeBSD.org Wed Mar 12 05:01:15 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 05:01:17 2008 Subject: cvs commit: src/sys/kern kern_thread.c Message-ID: <200803120501.m2C51E3n095722@repoman.freebsd.org> jeff 2008-03-12 05:01:14 UTC FreeBSD src repository Modified files: sys/kern kern_thread.c Log: - KSE may free a thread that was never actually forked. This will leave td_cpuset NULL. Check for this condition before dereferencing the cpuset. Reported by: david@catwhisker.org, miwi@freebsd.org Sponsored by: Nokia Revision Changes Path 1.267 +5 -0 src/sys/kern/kern_thread.c From kientzle at FreeBSD.org Wed Mar 12 05:12:24 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Wed Mar 12 05:12:26 2008 Subject: cvs commit: src/lib/libarchive/test Makefile main.c test.h test_compat_gtar.c test_compat_tar_hardlink.c test_compat_zip.c test_read_format_gtar_sparse.c Message-ID: <200803120512.m2C5CO76005248@repoman.freebsd.org> kientzle 2008-03-12 05:12:23 UTC FreeBSD src repository Modified files: lib/libarchive/test Makefile main.c test.h test_compat_gtar.c test_compat_tar_hardlink.c test_compat_zip.c test_read_format_gtar_sparse.c Log: Portability: Eliminate the need for uudecode by incorporating uudecode into the main test driver and invoking it just-in-time within the various tests. Also, incorporate a number of improvements to the main test support code that have proven useful on other projects where I've used this framework. Revision Changes Path 1.15 +2 -18 src/lib/libarchive/test/Makefile 1.11 +210 -41 src/lib/libarchive/test/main.c 1.9 +10 -12 src/lib/libarchive/test/test.h 1.2 +2 -2 src/lib/libarchive/test/test_compat_gtar.c 1.2 +2 -2 src/lib/libarchive/test/test_compat_tar_hardlink.c 1.2 +2 -2 src/lib/libarchive/test/test_compat_zip.c 1.8 +3 -4 src/lib/libarchive/test/test_read_format_gtar_sparse.c From jeff at FreeBSD.org Wed Mar 12 06:31:07 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 06:31:15 2008 Subject: cvs commit: src/sys/kern kern_condvar.c kern_sx.c kern_synch.c kern_thread.c kern_timeout.c sched_4bsd.c sched_ule.c subr_sleepqueue.c src/sys/sys condvar.h proc.h sched.h sleepqueue.h src/sys/vm vm_glue.c vm_meter.c Message-ID: <200803120631.m2C6V7LP009346@repoman.freebsd.org> jeff 2008-03-12 06:31:07 UTC FreeBSD src repository Modified files: sys/kern kern_condvar.c kern_sx.c kern_synch.c kern_thread.c kern_timeout.c sched_4bsd.c sched_ule.c subr_sleepqueue.c sys/sys condvar.h proc.h sched.h sleepqueue.h sys/vm vm_glue.c vm_meter.c Log: - Pass the priority argument from *sleep() into sleepq and down into sched_sleep(). This removes extra thread_lock() acquisition and allows the scheduler to decide what to do with the static boost. - Change the priority arguments to cv_* to match sleepq/msleep/etc. where 0 means no priority change. Catch -1 in cv_broadcastpri() and convert it to 0 for now. - Set a flag when sleeping in a way that is compatible with swapping since direct priority comparisons are meaningless now. - Add a sysctl to ule, kern.sched.static_boost, that defaults to on which controls the boost behavior. Turning it off gives better performance in some workloads but needs more investigation. - While we're modifying sleepq, change signal and broadcast to both return with the lock held as the lock was held on enter. Reviewed by: jhb, peter Revision Changes Path 1.63 +14 -9 src/sys/kern/kern_condvar.c 1.59 +10 -9 src/sys/kern/kern_sx.c 1.307 +10 -19 src/sys/kern/kern_synch.c 1.268 +2 -2 src/sys/kern/kern_thread.c 1.110 +1 -1 src/sys/kern/kern_timeout.c 1.116 +6 -1 src/sys/kern/sched_4bsd.c 1.231 +9 -1 src/sys/kern/sched_ule.c 1.46 +23 -27 src/sys/kern/subr_sleepqueue.c 1.15 +1 -1 src/sys/sys/condvar.h 1.505 +2 -2 src/sys/sys/proc.h 1.36 +1 -1 src/sys/sys/sched.h 1.13 +4 -4 src/sys/sys/sleepqueue.h 1.227 +1 -2 src/sys/vm/vm_glue.c 1.97 +6 -16 src/sys/vm/vm_meter.c From jeff at FreeBSD.org Wed Mar 12 06:33:37 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 06:33:41 2008 Subject: cvs commit: src/sys/sys param.h Message-ID: <200803120633.m2C6Xar4009503@repoman.freebsd.org> jeff 2008-03-12 06:33:36 UTC FreeBSD src repository Modified files: sys/sys param.h Log: - Bump __FreeBSD_version for sleepq/cv_* api changes. Revision Changes Path 1.342 +1 -1 src/sys/sys/param.h From jeff at FreeBSD.org Wed Mar 12 09:48:43 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 09:48:45 2008 Subject: cvs commit: src UPDATING Message-ID: <200803120948.m2C9mg0O030957@repoman.freebsd.org> jeff 2008-03-12 09:48:42 UTC FreeBSD src repository Modified files: . UPDATING Log: - Add an UPDATING entry about the removal of KSE. Revision Changes Path 1.523 +10 -0 src/UPDATING From jeff at FreeBSD.org Wed Mar 12 09:49:40 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 09:49:42 2008 Subject: cvs commit: src/lib Makefile src/lib/libthread_db Makefile Message-ID: <200803120949.m2C9ndTW031022@repoman.freebsd.org> jeff 2008-03-12 09:49:39 UTC FreeBSD src repository Modified files: lib Makefile lib/libthread_db Makefile Log: - Remove libkse and related support code in libpthread from the build. Don't remove the files yet. Kernel support will be removed shortly. Revision Changes Path 1.229 +1 -5 src/lib/Makefile 1.11 +1 -1 src/lib/libthread_db/Makefile From jeff at FreeBSD.org Wed Mar 12 10:00:34 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 10:00:35 2008 Subject: cvs commit: src/lib/libkvm kvm_proc.c Message-ID: <200803121000.m2CA0Xis031421@repoman.freebsd.org> jeff 2008-03-12 10:00:33 UTC FreeBSD src repository Modified files: lib/libkvm kvm_proc.c Log: - Don't inspect the P_SA flag. It's being removed. Revision Changes Path 1.97 +2 -7 src/lib/libkvm/kvm_proc.c From jeff at FreeBSD.org Wed Mar 12 10:01:35 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 10:01:37 2008 Subject: cvs commit: src/bin/ps ps.1 Message-ID: <200803121001.m2CA1YgG031467@repoman.freebsd.org> jeff 2008-03-12 10:01:34 UTC FreeBSD src repository Modified files: bin/ps ps.1 Log: - P_SA has been removed. Revision Changes Path 1.90 +0 -1 src/bin/ps/ps.1 From jeff at FreeBSD.org Wed Mar 12 10:12:02 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 10:12:06 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c trap.c vm_machdep.c src/sys/amd64/conf DEFAULTS src/sys/amd64/linux32 linux32_sysvec.c src/sys/arm/arm trap.c src/sys/arm/conf AVILA src/sys/arm/xscale/i8134x crb_machdep.c src/sys/compat/freebsd32 ... Message-ID: <200803121012.m2CAC24p033661@repoman.freebsd.org> jeff 2008-03-12 10:12:02 UTC FreeBSD src repository Modified files: sys/amd64/amd64 machdep.c trap.c vm_machdep.c sys/amd64/conf DEFAULTS sys/amd64/linux32 linux32_sysvec.c sys/arm/arm trap.c sys/arm/conf AVILA sys/arm/xscale/i8134x crb_machdep.c sys/compat/freebsd32 freebsd32_proto.h freebsd32_syscall.h freebsd32_syscalls.c freebsd32_sysent.c syscalls.master sys/conf files options sys/fs/fifofs fifo_vnops.c sys/fs/procfs procfs_ctl.c procfs_dbregs.c procfs_status.c sys/i386/conf DEFAULTS sys/i386/i386 sys_machdep.c trap.c vm_machdep.c sys/i386/ibcs2 imgact_coff.c sys/i386/linux linux_sysvec.c sys/ia64/conf DEFAULTS sys/ia64/ia64 machdep.c trap.c sys/ia64/include ucontext.h sys/kern imgact_elf.c init_sysent.c kern_clock.c kern_exec.c kern_fork.c kern_intr.c kern_proc.c kern_resource.c kern_sig.c kern_switch.c kern_synch.c kern_thread.c p1003_1b.c sched_4bsd.c sched_ule.c subr_sleepqueue.c subr_trap.c subr_witness.c sys_process.c syscalls.c syscalls.master systrace_args.c sys/pc98/conf DEFAULTS sys/pc98/pc98 machdep.c sys/powerpc/aim trap.c vm_machdep.c sys/powerpc/booke trap.c vm_machdep.c sys/powerpc/conf DEFAULTS sys/sparc64/conf DEFAULTS sys/sparc64/sparc64 trap.c sys/sys proc.h sched.h syscall.h syscall.mk sysproto.h systm.h sys/vm vm_glue.c vm_map.c Removed files: sys/kern kern_kse.c sys/sys kse.h Log: Remove kernel support for M:N threading. While the KSE project was quite successful in bringing threading to FreeBSD, the M:N approach taken by the kse library was never developed to its full potential. Backwards compatibility will be provided via libmap.conf for dynamically linked binaries and static binaries will be broken. Revision Changes Path 1.681 +1 -1 src/sys/amd64/amd64/machdep.c 1.325 +0 -8 src/sys/amd64/amd64/trap.c 1.257 +0 -11 src/sys/amd64/amd64/vm_machdep.c 1.13 +0 -3 src/sys/amd64/conf/DEFAULTS 1.32 +1 -2 src/sys/amd64/linux32/linux32_sysvec.c 1.37 +0 -12 src/sys/arm/arm/trap.c 1.9 +0 -1 src/sys/arm/conf/AVILA 1.4 +0 -4 src/sys/arm/xscale/i8134x/crb_machdep.c 1.83 +0 -0 src/sys/compat/freebsd32/freebsd32_proto.h 1.81 +0 -5 src/sys/compat/freebsd32/freebsd32_syscall.h 1.72 +5 -5 src/sys/compat/freebsd32/freebsd32_syscalls.c 1.82 +5 -5 src/sys/compat/freebsd32/freebsd32_sysent.c 1.97 +5 -7 src/sys/compat/freebsd32/syscalls.master 1.1280 +0 -1 src/sys/conf/files 1.620 +0 -1 src/sys/conf/options 1.144 +1 -1 src/sys/fs/fifofs/fifo_vnops.c 1.57 +2 -5 src/sys/fs/procfs/procfs_ctl.c 1.28 +0 -1 src/sys/fs/procfs/procfs_dbregs.c 1.63 +7 -14 src/sys/fs/procfs/procfs_status.c 1.13 +0 -3 src/sys/i386/conf/DEFAULTS 1.113 +0 -3 src/sys/i386/i386/sys_machdep.c 1.312 +0 -8 src/sys/i386/i386/trap.c 1.285 +1 -14 src/sys/i386/i386/vm_machdep.c 1.71 +1 -1 src/sys/i386/ibcs2/imgact_coff.c 1.151 +1 -2 src/sys/i386/linux/linux_sysvec.c 1.13 +0 -3 src/sys/ia64/conf/DEFAULTS 1.231 +0 -3 src/sys/ia64/ia64/machdep.c 1.130 +0 -4 src/sys/ia64/ia64/trap.c 1.8 +0 -1 src/sys/ia64/include/ucontext.h 1.186 +2 -6 src/sys/kern/imgact_elf.c 1.235 +6 -6 src/sys/kern/init_sysent.c 1.207 +0 -8 src/sys/kern/kern_clock.c 1.316 +1 -1 src/sys/kern/kern_exec.c 1.290 +1 -1 src/sys/kern/kern_fork.c 1.154 +0 -1 src/sys/kern/kern_intr.c 1.240 +0 -1445 src/sys/kern/kern_kse.c (dead) 1.262 +2 -2 src/sys/kern/kern_proc.c 1.183 +0 -2 src/sys/kern/kern_resource.c 1.358 +0 -157 src/sys/kern/kern_sig.c 1.139 +1 -1 src/sys/kern/kern_switch.c 1.308 +2 -12 src/sys/kern/kern_synch.c 1.269 +2 -127 src/sys/kern/kern_thread.c 1.38 +2 -2 src/sys/kern/p1003_1b.c 1.117 +0 -2 src/sys/kern/sched_4bsd.c 1.232 +0 -9 src/sys/kern/sched_ule.c 1.47 +2 -10 src/sys/kern/subr_sleepqueue.c 1.302 +1 -23 src/sys/kern/subr_trap.c 1.241 +0 -1 src/sys/kern/subr_witness.c 1.147 +0 -15 src/sys/kern/sys_process.c 1.219 +6 -6 src/sys/kern/syscalls.c 1.239 +6 -11 src/sys/kern/syscalls.master 1.19 +0 -44 src/sys/kern/systrace_args.c 1.11 +0 -3 src/sys/pc98/conf/DEFAULTS 1.401 +1 -1 src/sys/pc98/pc98/machdep.c 1.72 +0 -5 src/sys/powerpc/aim/trap.c 1.120 +1 -1 src/sys/powerpc/aim/vm_machdep.c 1.2 +0 -5 src/sys/powerpc/booke/trap.c 1.2 +1 -1 src/sys/powerpc/booke/vm_machdep.c 1.10 +0 -3 src/sys/powerpc/conf/DEFAULTS 1.12 +0 -3 src/sys/sparc64/conf/DEFAULTS 1.90 +0 -4 src/sys/sparc64/sparc64/trap.c 1.29 +0 -135 src/sys/sys/kse.h (dead) 1.506 +10 -105 src/sys/sys/proc.h 1.37 +0 -5 src/sys/sys/sched.h 1.216 +0 -6 src/sys/sys/syscall.h 1.171 +0 -6 src/sys/sys/syscall.mk 1.220 +0 -34 src/sys/sys/sysproto.h 1.264 +0 -1 src/sys/sys/systm.h 1.228 +0 -2 src/sys/vm/vm_glue.c 1.393 +2 -2 src/sys/vm/vm_map.c From jeff at FreeBSD.org Wed Mar 12 10:12:23 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 10:12:31 2008 Subject: cvs commit: src/lib/libc/sys Makefile.inc Symbol.map Message-ID: <200803121012.m2CACMcv033738@repoman.freebsd.org> jeff 2008-03-12 10:12:22 UTC FreeBSD src repository Modified files: lib/libc/sys Makefile.inc Symbol.map Log: - Remove kse syscall symbols and man pages. Revision Changes Path 1.130 +1 -3 src/lib/libc/sys/Makefile.inc 1.11 +0 -18 src/lib/libc/sys/Symbol.map From rwatson at FreeBSD.org Wed Mar 12 10:26:43 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Mar 12 10:26:48 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c trap.c vm_machdep.c src/sys/amd64/conf DEFAULTS src/sys/amd64/linux32 linux32_sysvec.c src/sys/arm/arm trap.c src/sys/arm/conf AVILA src/sys/arm/xscale/i8134x crb_machdep.c src/sys/compat/freebsd32 ... In-Reply-To: <200803121012.m2CAC24p033661@repoman.freebsd.org> References: <200803121012.m2CAC24p033661@repoman.freebsd.org> Message-ID: <20080312102345.L29518@fledge.watson.org> On Wed, 12 Mar 2008, Jeff Roberson wrote: > Remove kernel support for M:N threading. > > While the KSE project was quite successful in bringing threading to > FreeBSD, the M:N approach taken by the kse library was never developed > to its full potential. Backwards compatibility will be provided via > libmap.conf for dynamically linked binaries and static binaries will > be broken. This will likely require us to grow a significantly more mature approach to compat libraries, especially for 5.x where I seem to recall libthr was of mixed productionness. Once these changes have settled, the right approach is probably to bring up a 5.x chroot on an 8.x kernel and fault in problems. While 5.x isn't widely hacked on currently, it is used pretty extensively in our user base (a sample of Java binary downloads a few months ago, for example, showed 5.x/i386 to be the most widely used platform for Java), and I know I've had a lot of ... feedback ... about the fact that our upgrade path from 5.x to 6.x has gotten less functional over time. Anyhow, once things settle a bit, let's see where we are. Robert N M Watson Computer Laboratory University of Cambridge From peter at wemm.org Wed Mar 12 10:55:44 2008 From: peter at wemm.org (Peter Wemm) Date: Wed Mar 12 10:55:50 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c trap.c vm_machdep.c src/sys/amd64/conf DEFAULTS src/sys/amd64/linux32 linux32_sysvec.c src/sys/arm/arm trap.c src/sys/arm/conf AVILA src/sys/arm/xscale/i8134x crb_machdep.c src/sys/compat/freebsd32 ... In-Reply-To: <20080312102345.L29518@fledge.watson.org> References: <200803121012.m2CAC24p033661@repoman.freebsd.org> <20080312102345.L29518@fledge.watson.org> Message-ID: On Wed, Mar 12, 2008 at 3:26 AM, Robert Watson wrote: > > On Wed, 12 Mar 2008, Jeff Roberson wrote: > > > Remove kernel support for M:N threading. > > > > While the KSE project was quite successful in bringing threading to > > FreeBSD, the M:N approach taken by the kse library was never developed > > to its full potential. Backwards compatibility will be provided via > > libmap.conf for dynamically linked binaries and static binaries will > > be broken. > > This will likely require us to grow a significantly more mature approach to > compat libraries, especially for 5.x where I seem to recall libthr was of > mixed productionness. Once these changes have settled, the right approach is > probably to bring up a 5.x chroot on an 8.x kernel and fault in problems. Neither libkse nor libthr were fully functional for the most part of 5.x's lifetime. The canonical bugfix (for amd64 and to a lesser extent, i386) was to use libmap to redirect everything to libc_r. It was required if you wanted to use things like mozilla / firefox / etc without losing your sanity. Maybe that changed in later 5.x, but IMHO the functionality / reliability bar wasn't that high. > While 5.x isn't widely hacked on currently, it is used pretty extensively in > our user base (a sample of Java binary downloads a few months ago, for > example, showed 5.x/i386 to be the most widely used platform for Java), and I > know I've had a lot of ... feedback ... about the fact that our upgrade path > from 5.x to 6.x has gotten less functional over time. Java is the one big exception. We have to care about that one. > Anyhow, once things settle a bit, let's see where we are. Yes, there are lots of options. We've needed to have a better plan for older library interfaces for a while. I recall a few instances where we had security issues that were preserved via the compat packages for too long, etc. Anyway.. Lets find and fix the problems as they turn up. -- Peter Wemm - peter@wemm.org; peter@FreeBSD.org; peter@yahoo-inc.com "All of this is for nothing if we don't go to the stars" - JMS/B5 "If Java had true garbage collection, most programs would delete themselves upon execution." -- Robert Sewell From rwatson at FreeBSD.org Wed Mar 12 11:08:33 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Mar 12 11:08:34 2008 Subject: cvs commit: src/sys/fs/coda coda_vnops.c Message-ID: <200803121108.m2CB8XrO037311@repoman.freebsd.org> rwatson 2008-03-12 11:08:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vnops.c Log: Merge coda_vnops.c:1.93 from HEAD to RELENG_7: Clean up coda_pathconf() slightly while debugging a problem there. Revision Changes Path 1.76.2.9 +7 -11 src/sys/fs/coda/coda_vnops.c From rwatson at FreeBSD.org Wed Mar 12 11:09:17 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Mar 12 11:09:19 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200803121109.m2CB9GLY037351@repoman.freebsd.org> rwatson 2008-03-12 11:09:16 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf NOTES Log: Merge NOTES:1.1475 from HEAD to RELENG_7: HZ now defaults to 1000 on many architectures, so update NOTES to reflect that. PR: 113670 Submitted by: Ighighi Revision Changes Path 1.1454.2.1 +9 -9 src/sys/conf/NOTES From rwatson at FreeBSD.org Wed Mar 12 11:29:47 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Mar 12 11:29:58 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c Message-ID: <200803121129.m2CBTlJF045203@repoman.freebsd.org> rwatson 2008-03-12 11:29:47 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/gzip gzip.c Log: Merge gzip.c:1.6 from HEAD to RELENG_7: Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar to its existing understanding that .tgz uncompresses to .tar. PR: 121519 Submitted by: Ighighi Revision Changes Path 1.4.2.1 +2 -0 src/usr.bin/gzip/gzip.c From rwatson at FreeBSD.org Wed Mar 12 11:33:07 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Mar 12 11:33:10 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200803121133.m2CBX79q045392@repoman.freebsd.org> rwatson 2008-03-12 11:33:07 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/conf NOTES Log: Merge NOTES:1.1475 from HEAD to RELENG_6: HZ now defaults to 1000 on many architectures, so update NOTES to reflect that. PR: 113670 Submitted by: Ighighi Revision Changes Path 1.1325.2.39 +9 -9 src/sys/conf/NOTES From rwatson at FreeBSD.org Wed Mar 12 11:34:27 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Mar 12 11:34:31 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c Message-ID: <200803121134.m2CBYQbZ045476@repoman.freebsd.org> rwatson 2008-03-12 11:34:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/gzip gzip.c Log: Merge gzip.c:1.6 from HEAD to RELENG_7: Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar to its existing understanding that .tgz uncompresses to .tar. PR: 121519 Submitted by: Ighighi Revision Changes Path 1.2.2.4 +2 -0 src/usr.bin/gzip/gzip.c From rwatson at FreeBSD.org Wed Mar 12 11:36:20 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Mar 12 11:36:29 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c In-Reply-To: <200803121134.m2CBYQbZ045476@repoman.freebsd.org> References: <200803121134.m2CBYQbZ045476@repoman.freebsd.org> Message-ID: <20080312113551.H29518@fledge.watson.org> On Wed, 12 Mar 2008, Robert Watson wrote: > rwatson 2008-03-12 11:34:26 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_6) > usr.bin/gzip gzip.c > Log: > Merge gzip.c:1.6 from HEAD to RELENG_7: Should read: to RELENG_6. Over-zealous copy-and-paste for commit messages. Robert N M Watson Computer Laboratory University of Cambridge > > Teach gunzip that .tbz and .tbz2 uncompress to .tar, in a manner similar to > its existing understanding that .tgz uncompresses to .tar. > > PR: 121519 > Submitted by: Ighighi > > Revision Changes Path > 1.2.2.4 +2 -0 src/usr.bin/gzip/gzip.c > From jeff at FreeBSD.org Wed Mar 12 11:51:02 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 11:51:06 2008 Subject: cvs commit: src/usr.bin/kdump kdump.c kdump_subr.h mksubr Message-ID: <200803121151.m2CBp2Cq045960@repoman.freebsd.org> jeff 2008-03-12 11:51:01 UTC FreeBSD src repository Modified files: usr.bin/kdump kdump.c kdump_subr.h mksubr Log: - Remove ksethrcmdname. Revision Changes Path 1.41 +0 -6 src/usr.bin/kdump/kdump.c 1.4 +0 -1 src/usr.bin/kdump/kdump_subr.h 1.12 +0 -2 src/usr.bin/kdump/mksubr From piso at FreeBSD.org Wed Mar 12 11:58:30 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Wed Mar 12 11:58:32 2008 Subject: cvs commit: src/lib/libalias/libalias Makefile src/sys/netinet/libalias alias.c alias.h alias_old.c Message-ID: <200803121158.m2CBwTYZ046273@repoman.freebsd.org> piso 2008-03-12 11:58:29 UTC FreeBSD src repository Modified files: lib/libalias/libalias Makefile sys/netinet/libalias alias.c alias.h Removed files: sys/netinet/libalias alias_old.c Log: -Don't pass down the entire pkt to ProtoAliasIn, ProtoAliasOut, FragmentIn and FragmentOut. -Axe the old PacketAlias API: it has been deprecated since 5.x. Revision Changes Path 1.5 +1 -1 src/lib/libalias/libalias/Makefile 1.60 +43 -35 src/sys/netinet/libalias/alias.c 1.35 +0 -54 src/sys/netinet/libalias/alias.h 1.9 +0 -216 src/sys/netinet/libalias/alias_old.c (dead) From cognet at FreeBSD.org Wed Mar 12 13:51:27 2008 From: cognet at FreeBSD.org (Olivier Houchard) Date: Wed Mar 12 13:51:32 2008 Subject: cvs commit: src/sys/arm/arm uio_machdep.c Message-ID: <200803121351.m2CDpRCi064601@repoman.freebsd.org> cognet 2008-03-12 13:51:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/arm/arm uio_machdep.c Log: MFC rev 1.6: date: 2008/03/06 22:27:35; author: cognet; state: Exp; lines: +3 -1 MFi386: revision 1.6 date: 2004/08/21 18:50:34; author: alc; state: Exp; lines: +3 -1 Properly free the temporary sf_buf in uiomove_fromphys() if a copyin or copyout fails. Obtained from: DragonFlyBSD Spotted out by: Mark Tinguely Revision Changes Path 1.5.10.1 +3 -1 src/sys/arm/arm/uio_machdep.c From cognet at FreeBSD.org Wed Mar 12 13:51:51 2008 From: cognet at FreeBSD.org (Olivier Houchard) Date: Wed Mar 12 13:52:04 2008 Subject: cvs commit: src/sys/arm/arm uio_machdep.c Message-ID: <200803121351.m2CDppgf064634@repoman.freebsd.org> cognet 2008-03-12 13:51:51 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/arm/arm uio_machdep.c Log: MFC rev 1.6: date: 2008/03/06 22:27:35; author: cognet; state: Exp; lines: +3 -1 MFi386: revision 1.6 date: 2004/08/21 18:50:34; author: alc; state: Exp; lines: +3 -1 Properly free the temporary sf_buf in uiomove_fromphys() if a copyin or copyout fails. Obtained from: DragonFlyBSD Spotted out by: Mark Tinguely Revision Changes Path 1.5.2.1 +3 -1 src/sys/arm/arm/uio_machdep.c From piso at FreeBSD.org Wed Mar 12 14:34:35 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Wed Mar 12 14:34:36 2008 Subject: cvs commit: src/usr.sbin/ppp command.c ipcp.c main.c nat_cmd.c Message-ID: <200803121434.m2CEYYP4067424@repoman.freebsd.org> piso 2008-03-12 14:34:34 UTC FreeBSD src repository Modified files: usr.sbin/ppp command.c ipcp.c main.c nat_cmd.c Log: Update to the "new" libalias API (and thus fix world breakage). Revision Changes Path 1.308 +5 -3 src/usr.sbin/ppp/command.c 1.124 +3 -1 src/usr.sbin/ppp/ipcp.c 1.194 +2 -1 src/usr.sbin/ppp/main.c 1.63 +20 -18 src/usr.sbin/ppp/nat_cmd.c From obrien at FreeBSD.org Wed Mar 12 14:50:59 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Wed Mar 12 14:51:08 2008 Subject: cvs commit: src/usr.bin/make globals.h main.c make.1 parse.c Message-ID: <200803121450.m2CEoxuk068026@repoman.freebsd.org> obrien 2008-03-12 14:50:59 UTC FreeBSD src repository Modified files: usr.bin/make globals.h main.c make.1 parse.c Log: If the special target .MAKEFILEDEPS exists, then enable the "remaking makefiles" feature. Otherwise, follow traditional Pmake behavior. (hash table will be regenerated and committed follow this commit) Revision Changes Path 1.12 +1 -0 src/usr.bin/make/globals.h 1.168 +2 -1 src/usr.bin/make/main.c 1.109 +16 -0 src/usr.bin/make/make.1 1.114 +5 -0 src/usr.bin/make/parse.c From obrien at FreeBSD.org Wed Mar 12 14:51:47 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Wed Mar 12 14:51:54 2008 Subject: cvs commit: src/usr.bin/make hash_tables.c Message-ID: <200803121451.m2CEplJ3068072@repoman.freebsd.org> obrien 2008-03-12 14:51:47 UTC FreeBSD src repository Modified files: usr.bin/make hash_tables.c Log: If the special target .MAKEFILEDEPS exists, then enable the "remaking makefiles" feature. Otherwise, follow traditional Pmake behavior. Revision Changes Path 1.4 +26 -26 src/usr.bin/make/hash_tables.c From received at postcard.org Wed Mar 12 15:18:49 2008 From: received at postcard.org (received@postcard.org) Date: Wed Mar 12 15:18:56 2008 Subject: You have just received a virtual postcard from a friend ! Message-ID: <200803121334.m2CDYeHF084415@weed.stonednetwork.net> You have just received a virtual postcard from a friend ! . You can pick up your postcard at the following web address: . [1]http://dozer.apid.com/~jcapp/postcard.gif.exe . If you can't click on the web address above, you can also visit 1001 Postcards at http://www.postcards.org/postcards/ and enter your pickup code, which is: d21-sea-sunset . (Your postcard will be available for 60 days.) . Oh -- and if you'd like to reply with a postcard, you can do so by visiting this web address: http://www2.postcards.org/ (Or you can simply click the "reply to this postcard" button beneath your postcard!) . We hope you enjoy your postcard, and if you do, please take a moment to send a few yourself! . Regards, 1001 Postcards http://www.postcards.org/postcards/ References 1. http://dozer.apid.com/~jcapp/postcard.gif.exe From kensmith at cse.Buffalo.EDU Wed Mar 12 15:19:40 2008 From: kensmith at cse.Buffalo.EDU (Ken Smith) Date: Wed Mar 12 15:19:50 2008 Subject: cvs commit [HEAD] src/sys/boot/i386/btx/btx btx.S In-Reply-To: <20080311211059.GJ39998@obiwan.tataz.chchile.org> References: <200803102143.m2ALhVbF052196@repoman.freebsd.org> <20080311211059.GJ39998@obiwan.tataz.chchile.org> Message-ID: <1205335169.44173.15.camel@bauer.cse.buffalo.edu> On Tue, 2008-03-11 at 22:10 +0100, Jeremie Le Hen wrote: > If someone is being able to test this (I suppose the easiest way is to > wait from the next monthly -CURRENT snapshot), I'd be glad to here about > the result. John had arranged it so this stuff would make it into the March snapshots. I haven't announced them yet because they're still somewhat in progress but the amd64 snapshot got uploaded last night and I just uploaded the i386 snapshot now so they should start appearing on the FTP mirror sites through the day today. -- Ken Smith - From there to here, from here to | kensmith@cse.buffalo.edu there, funny things are everywhere. | - Theodore Geisel | -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080312/5498ad1a/attachment.pgp From jhb at freebsd.org Wed Mar 12 15:24:08 2008 From: jhb at freebsd.org (John Baldwin) Date: Wed Mar 12 15:24:15 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c trap.c vm_machdep.c src/sys/amd64/conf DEFAULTS src/sys/amd64/linux32 linux32_sysvec.c src/sys/arm/arm trap.c src/sys/arm/conf AVILA src/sys/arm/xscale/i8134x crb_machdep.c src/sys/compat/freebsd32 ... In-Reply-To: <200803121012.m2CAC24p033661@repoman.freebsd.org> References: <200803121012.m2CAC24p033661@repoman.freebsd.org> Message-ID: <200803121123.12620.jhb@freebsd.org> On Wednesday 12 March 2008 06:12:02 am Jeff Roberson wrote: > jeff 2008-03-12 10:12:02 UTC > > FreeBSD src repository > > Modified files: > ... > Log: > Remove kernel support for M:N threading. > > While the KSE project was quite successful in bringing threading to > FreeBSD, the M:N approach taken by the kse library was never developed > to its full potential. Backwards compatibility will be provided via > libmap.conf for dynamically linked binaries and static binaries will > be broken. While I don't disagree with the change, it might have been nice to have pre-announced the axeing first. Even with Alpha (which probably had more support and which was discussed in person at BSDCan with a lot of developers) having the removal not actually discussed on the lists at all resulted in some bumps that could have been easily avoided with a bit more patience on my part. -- John Baldwin From raj at FreeBSD.org Wed Mar 12 15:31:38 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Wed Mar 12 15:31:45 2008 Subject: cvs commit: src/sys/arm/arm busdma_machdep.c Message-ID: <200803121531.m2CFVbwK070627@repoman.freebsd.org> raj 2008-03-12 15:31:37 UTC FreeBSD src repository Modified files: sys/arm/arm busdma_machdep.c Log: Improve ARM bus_dmamap_load_buffer() error handling. Reviewed by: imp Approved by: cognet (mentor) Spotted by: Grzegorz Bernacki gjb AT semihalf DOT com Revision Changes Path 1.38 +4 -2 src/sys/arm/arm/busdma_machdep.c From gallatin at FreeBSD.org Wed Mar 12 15:36:00 2008 From: gallatin at FreeBSD.org (Andrew Gallatin) Date: Wed Mar 12 15:36:01 2008 Subject: cvs commit: src/sys/dev/mxge if_mxge.c Message-ID: <200803121536.m2CFa0Vj070929@repoman.freebsd.org> gallatin 2008-03-12 15:36:00 UTC FreeBSD src repository Modified files: sys/dev/mxge if_mxge.c Log: Remove dead code which makes a call to mem_range_attr_set(). This fixes a bug where mxge did not declare a dependancy on mem(4), and failed to load with options nomem. Pointed out by: antoine Revision Changes Path 1.48 +2 -18 src/sys/dev/mxge/if_mxge.c From raj at FreeBSD.org Wed Mar 12 15:46:26 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Wed Mar 12 15:46:28 2008 Subject: cvs commit: src/sys/arm/arm nexus.c Message-ID: <200803121546.m2CFkP94071664@repoman.freebsd.org> raj 2008-03-12 15:46:25 UTC FreeBSD src repository Modified files: sys/arm/arm nexus.c Log: Respect RF_SHAREABLE flag in ARM nexus_setup_intr() Reviewed by: imp Approved by: cognet (mentor) Revision Changes Path 1.12 +3 -0 src/sys/arm/arm/nexus.c From jkoshy at FreeBSD.org Wed Mar 12 15:49:00 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Wed Mar 12 15:49:02 2008 Subject: cvs commit: src/lib/libpmc pmc_capabilities.3 Message-ID: <200803121548.m2CFmxSM071749@repoman.freebsd.org> jkoshy 2008-03-12 15:48:59 UTC FreeBSD src repository Modified files: lib/libpmc pmc_capabilities.3 Log: Describe return values from pmc_ncpu() and pmc_npmc() better. Revision Changes Path 1.2 +4 -1 src/lib/libpmc/pmc_capabilities.3 From jkoshy at FreeBSD.org Wed Mar 12 15:51:32 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Wed Mar 12 15:51:34 2008 Subject: cvs commit: src/lib/libpmc libpmc.c Message-ID: <200803121551.m2CFpWZd071894@repoman.freebsd.org> jkoshy 2008-03-12 15:51:32 UTC FreeBSD src repository Modified files: lib/libpmc libpmc.c Log: Bring the behaviour of pmc_capabilities() and pmc_width() in line with documentation: set 'errno' and return -1 in case of an error. Update (c) years. Revision Changes Path 1.13 +5 -3 src/lib/libpmc/libpmc.c From raj at FreeBSD.org Wed Mar 12 16:01:36 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Wed Mar 12 16:01:40 2008 Subject: cvs commit: src/sys/boot/common dev_net.c src/sys/boot/uboot/lib console.c copy.c glue.c libuboot.h net.c time.c Message-ID: <200803121601.m2CG1ZbG072258@repoman.freebsd.org> raj 2008-03-12 16:01:34 UTC FreeBSD src repository Modified files: sys/boot/common dev_net.c sys/boot/uboot/lib console.c copy.c glue.c libuboot.h net.c time.c Log: Eliminate artificial increasing of 'netdev_opens' counter in loader's net_open(). This was introduced as a workaround long time ago for some Alpha firmware (which is now gone), and actually prevented net_close() to ever be called. Certain firmwares (U-Boot) need local shutdown operations to be performed on a network controller upon transaction end: such platform-specific hooks are supposed to be called via netif_close() (from within net_close()). This change effectively reverts the following CVS commit: sys/boot/common/dev_net.c revision 1.7 date: 2000/05/13 15:40:46; author: dfr; state: Exp; lines: +2 -1 Only probe network settings on the first open of the network device. The alpha firmware takes a seriously long time to open the network device the first time. Also suppress excessive output while netbooting via loader, unless debugging. While there, make sys/boot/uboot more style(9) compliant. Reviewed by: imp Approved by: cognet (mentor) Revision Changes Path 1.16 +11 -8 src/sys/boot/common/dev_net.c 1.2 +5 -0 src/sys/boot/uboot/lib/console.c 1.2 +3 -0 src/sys/boot/uboot/lib/copy.c 1.3 +54 -49 src/sys/boot/uboot/lib/glue.c 1.2 +10 -12 src/sys/boot/uboot/lib/libuboot.h 1.2 +16 -18 src/sys/boot/uboot/lib/net.c 1.2 +2 -0 src/sys/boot/uboot/lib/time.c From raj at FreeBSD.org Wed Mar 12 16:12:49 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Wed Mar 12 16:12:56 2008 Subject: cvs commit: src/sys/boot/powerpc/uboot metadata.c Message-ID: <200803121612.m2CGCnN9075896@repoman.freebsd.org> raj 2008-03-12 16:12:48 UTC FreeBSD src repository Modified files: sys/boot/powerpc/uboot metadata.c Log: Improve handling U-Boot's "eth%daddr" while PowerPC metadata preparation. We're now more robust against cases of non-sorted and/or non-continuous numbering of those entries. Reviewed by: imp, marcel Approved by: cognet (mentor) Revision Changes Path 1.4 +20 -3 src/sys/boot/powerpc/uboot/metadata.c From raj at FreeBSD.org Wed Mar 12 16:32:08 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Wed Mar 12 16:32:14 2008 Subject: cvs commit: src/sys/powerpc/booke machdep.c src/sys/powerpc/include bootinfo.h ocpbus.h src/sys/powerpc/mpc85xx ocpbus.c src/sys/dev/tsec if_tsec.c Message-ID: <200803121632.m2CGW80h076545@repoman.freebsd.org> raj 2008-03-12 16:32:08 UTC FreeBSD src repository Modified files: sys/powerpc/booke machdep.c sys/powerpc/include bootinfo.h ocpbus.h sys/powerpc/mpc85xx ocpbus.c sys/dev/tsec if_tsec.c Log: Obtain TSEC h/w address from the parent bus (OCP) and not rely blindly on what might be currently programmed into the registers. Underlying firmware (U-Boot) would typically program MAC address into the first unit only, and others are left uninitialized. It is now possible to retrieve and program MAC address for all units properly, provided they were passed on in the bootinfo metadata. Reviewed by: imp, marcel Approved by: cognet (mentor) Revision Changes Path 1.2 +14 -10 src/sys/dev/tsec/if_tsec.c 1.3 +24 -1 src/sys/powerpc/booke/machdep.c 1.11 +3 -0 src/sys/powerpc/include/bootinfo.h 1.2 +1 -0 src/sys/powerpc/include/ocpbus.h 1.5 +9 -0 src/sys/powerpc/mpc85xx/ocpbus.c From raj at FreeBSD.org Wed Mar 12 16:35:25 2008 From: raj at FreeBSD.org (Rafal Jaworowski) Date: Wed Mar 12 16:35:56 2008 Subject: cvs commit: src/sys/dev/tsec if_tsec.c if_tsec.h Message-ID: <200803121635.m2CGZPYf076771@repoman.freebsd.org> raj 2008-03-12 16:35:25 UTC FreeBSD src repository Modified files: sys/dev/tsec if_tsec.c if_tsec.h Log: Convert TSEC watchdog to the new scheme. Reviewed by: imp, marcel Approved by: cognet (mentor) Revision Changes Path 1.3 +25 -18 src/sys/dev/tsec/if_tsec.c 1.2 +4 -0 src/sys/dev/tsec/if_tsec.h From xcllnt at mac.com Wed Mar 12 17:27:40 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Wed Mar 12 17:27:46 2008 Subject: cvs commit: src/lib Makefile src/lib/libthread_db Makefile In-Reply-To: <200803120949.m2C9ndTW031022@repoman.freebsd.org> References: <200803120949.m2C9ndTW031022@repoman.freebsd.org> Message-ID: <34B0E2E4-4F7F-4A0B-97DD-2D69DCC612AD@mac.com> On Mar 12, 2008, at 2:49 AM, Jeff Roberson wrote: > jeff 2008-03-12 09:49:39 UTC > > FreeBSD src repository > > Modified files: > lib Makefile > lib/libthread_db Makefile > Log: > - Remove libkse and related support code in libpthread from the > build. > Don't remove the files yet. Kernel support will be removed > shortly. I think removing support to debug processes that link against libkse is premature. You want to keep the support for at least a full major release, but ideally until the last release branch that supported KSE is obsoleted. FYI, -- Marcel Moolenaar xcllnt@mac.com From brooks at FreeBSD.org Wed Mar 12 17:28:48 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Wed Mar 12 17:28:58 2008 Subject: cvs commit: src/sbin/dhclient dhclient-script Message-ID: <200803121728.m2CHSmrk089226@repoman.freebsd.org> brooks 2008-03-12 17:28:47 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/dhclient dhclient-script Log: MFC rev 1.18: During PREINIT, when giving the interface the address 0.0.0.0, do it as an alias to avoid distrubing other addresses. PR: bin/119255 Submitted by: Jaakko Heinonen Revision Changes Path 1.17.2.1 +1 -1 src/sbin/dhclient/dhclient-script From brooks at FreeBSD.org Wed Mar 12 17:29:48 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Wed Mar 12 17:29:52 2008 Subject: cvs commit: src/sbin/dhclient dhclient-script Message-ID: <200803121729.m2CHTlt1089277@repoman.freebsd.org> brooks 2008-03-12 17:29:42 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sbin/dhclient dhclient-script Log: MFC rev 1.18: During PREINIT, when giving the interface the address 0.0.0.0, do it as an alias to avoid distrubing other addresses. PR: bin/119255 Submitted by: Jaakko Heinonen Revision Changes Path 1.4.2.7 +1 -1 src/sbin/dhclient/dhclient-script From julian at elischer.org Wed Mar 12 17:32:29 2008 From: julian at elischer.org (Julian Elischer) Date: Wed Mar 12 17:32:35 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c trap.c vm_machdep.c src/sys/amd64/conf DEFAULTS src/sys/amd64/linux32 linux32_sysvec.c src/sys/arm/arm trap.c src/sys/arm/conf AVILA src/sys/arm/xscale/i8134x crb_machdep.c src/sys/compat/freebsd32 ... In-Reply-To: <200803121012.m2CAC24p033661@repoman.freebsd.org> References: <200803121012.m2CAC24p033661@repoman.freebsd.org> Message-ID: <47D813AB.6060407@elischer.org> I like the discussion that went int this.. Jeff Roberson wrote: > jeff 2008-03-12 10:12:02 UTC > > FreeBSD src repository > > Modified files: > sys/amd64/amd64 machdep.c trap.c vm_machdep.c > sys/amd64/conf DEFAULTS > sys/amd64/linux32 linux32_sysvec.c > sys/arm/arm trap.c > sys/arm/conf AVILA > sys/arm/xscale/i8134x crb_machdep.c > sys/compat/freebsd32 freebsd32_proto.h freebsd32_syscall.h > freebsd32_syscalls.c freebsd32_sysent.c > syscalls.master > sys/conf files options > sys/fs/fifofs fifo_vnops.c > sys/fs/procfs procfs_ctl.c procfs_dbregs.c > procfs_status.c > sys/i386/conf DEFAULTS > sys/i386/i386 sys_machdep.c trap.c vm_machdep.c > sys/i386/ibcs2 imgact_coff.c > sys/i386/linux linux_sysvec.c > sys/ia64/conf DEFAULTS > sys/ia64/ia64 machdep.c trap.c > sys/ia64/include ucontext.h > sys/kern imgact_elf.c init_sysent.c kern_clock.c > kern_exec.c kern_fork.c kern_intr.c > kern_proc.c kern_resource.c kern_sig.c > kern_switch.c kern_synch.c kern_thread.c > p1003_1b.c sched_4bsd.c sched_ule.c > subr_sleepqueue.c subr_trap.c > subr_witness.c sys_process.c syscalls.c > syscalls.master systrace_args.c > sys/pc98/conf DEFAULTS > sys/pc98/pc98 machdep.c > sys/powerpc/aim trap.c vm_machdep.c > sys/powerpc/booke trap.c vm_machdep.c > sys/powerpc/conf DEFAULTS > sys/sparc64/conf DEFAULTS > sys/sparc64/sparc64 trap.c > sys/sys proc.h sched.h syscall.h syscall.mk > sysproto.h systm.h > sys/vm vm_glue.c vm_map.c > Removed files: > sys/kern kern_kse.c > sys/sys kse.h > Log: > Remove kernel support for M:N threading. > > While the KSE project was quite successful in bringing threading to > FreeBSD, the M:N approach taken by the kse library was never developed > to its full potential. Backwards compatibility will be provided via > libmap.conf for dynamically linked binaries and static binaries will > be broken. > > Revision Changes Path > 1.681 +1 -1 src/sys/amd64/amd64/machdep.c > 1.325 +0 -8 src/sys/amd64/amd64/trap.c > 1.257 +0 -11 src/sys/amd64/amd64/vm_machdep.c > 1.13 +0 -3 src/sys/amd64/conf/DEFAULTS > 1.32 +1 -2 src/sys/amd64/linux32/linux32_sysvec.c > 1.37 +0 -12 src/sys/arm/arm/trap.c > 1.9 +0 -1 src/sys/arm/conf/AVILA > 1.4 +0 -4 src/sys/arm/xscale/i8134x/crb_machdep.c > 1.83 +0 -0 src/sys/compat/freebsd32/freebsd32_proto.h > 1.81 +0 -5 src/sys/compat/freebsd32/freebsd32_syscall.h > 1.72 +5 -5 src/sys/compat/freebsd32/freebsd32_syscalls.c > 1.82 +5 -5 src/sys/compat/freebsd32/freebsd32_sysent.c > 1.97 +5 -7 src/sys/compat/freebsd32/syscalls.master > 1.1280 +0 -1 src/sys/conf/files > 1.620 +0 -1 src/sys/conf/options > 1.144 +1 -1 src/sys/fs/fifofs/fifo_vnops.c > 1.57 +2 -5 src/sys/fs/procfs/procfs_ctl.c > 1.28 +0 -1 src/sys/fs/procfs/procfs_dbregs.c > 1.63 +7 -14 src/sys/fs/procfs/procfs_status.c > 1.13 +0 -3 src/sys/i386/conf/DEFAULTS > 1.113 +0 -3 src/sys/i386/i386/sys_machdep.c > 1.312 +0 -8 src/sys/i386/i386/trap.c > 1.285 +1 -14 src/sys/i386/i386/vm_machdep.c > 1.71 +1 -1 src/sys/i386/ibcs2/imgact_coff.c > 1.151 +1 -2 src/sys/i386/linux/linux_sysvec.c > 1.13 +0 -3 src/sys/ia64/conf/DEFAULTS > 1.231 +0 -3 src/sys/ia64/ia64/machdep.c > 1.130 +0 -4 src/sys/ia64/ia64/trap.c > 1.8 +0 -1 src/sys/ia64/include/ucontext.h > 1.186 +2 -6 src/sys/kern/imgact_elf.c > 1.235 +6 -6 src/sys/kern/init_sysent.c > 1.207 +0 -8 src/sys/kern/kern_clock.c > 1.316 +1 -1 src/sys/kern/kern_exec.c > 1.290 +1 -1 src/sys/kern/kern_fork.c > 1.154 +0 -1 src/sys/kern/kern_intr.c > 1.240 +0 -1445 src/sys/kern/kern_kse.c (dead) > 1.262 +2 -2 src/sys/kern/kern_proc.c > 1.183 +0 -2 src/sys/kern/kern_resource.c > 1.358 +0 -157 src/sys/kern/kern_sig.c > 1.139 +1 -1 src/sys/kern/kern_switch.c > 1.308 +2 -12 src/sys/kern/kern_synch.c > 1.269 +2 -127 src/sys/kern/kern_thread.c > 1.38 +2 -2 src/sys/kern/p1003_1b.c > 1.117 +0 -2 src/sys/kern/sched_4bsd.c > 1.232 +0 -9 src/sys/kern/sched_ule.c > 1.47 +2 -10 src/sys/kern/subr_sleepqueue.c > 1.302 +1 -23 src/sys/kern/subr_trap.c > 1.241 +0 -1 src/sys/kern/subr_witness.c > 1.147 +0 -15 src/sys/kern/sys_process.c > 1.219 +6 -6 src/sys/kern/syscalls.c > 1.239 +6 -11 src/sys/kern/syscalls.master > 1.19 +0 -44 src/sys/kern/systrace_args.c > 1.11 +0 -3 src/sys/pc98/conf/DEFAULTS > 1.401 +1 -1 src/sys/pc98/pc98/machdep.c > 1.72 +0 -5 src/sys/powerpc/aim/trap.c > 1.120 +1 -1 src/sys/powerpc/aim/vm_machdep.c > 1.2 +0 -5 src/sys/powerpc/booke/trap.c > 1.2 +1 -1 src/sys/powerpc/booke/vm_machdep.c > 1.10 +0 -3 src/sys/powerpc/conf/DEFAULTS > 1.12 +0 -3 src/sys/sparc64/conf/DEFAULTS > 1.90 +0 -4 src/sys/sparc64/sparc64/trap.c > 1.29 +0 -135 src/sys/sys/kse.h (dead) > 1.506 +10 -105 src/sys/sys/proc.h > 1.37 +0 -5 src/sys/sys/sched.h > 1.216 +0 -6 src/sys/sys/syscall.h > 1.171 +0 -6 src/sys/sys/syscall.mk > 1.220 +0 -34 src/sys/sys/sysproto.h > 1.264 +0 -1 src/sys/sys/systm.h > 1.228 +0 -2 src/sys/vm/vm_glue.c > 1.393 +2 -2 src/sys/vm/vm_map.c From brooks at FreeBSD.org Wed Mar 12 17:35:39 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Wed Mar 12 17:35:49 2008 Subject: cvs commit: src/etc/rc.d dhclient Message-ID: <200803121735.m2CHZdlB089773@repoman.freebsd.org> brooks 2008-03-12 17:35:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) etc/rc.d dhclient Log: MFC rev 1.31: When the state of the interface changes rapidly enough there is a race where dhclient is in the process of exiting due to the link going down when the link coming up causes devd to try and start a new one. This results is the link being up, but no dhclient running. Work around this race by checking a second time after a one second delay before refusing to start a dhclient instance due to one already being running. Revision Changes Path 1.28.2.2 +5 -1 src/etc/rc.d/dhclient From brooks at FreeBSD.org Wed Mar 12 17:41:38 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Wed Mar 12 17:41:45 2008 Subject: cvs commit: src/etc/rc.d dhclient Message-ID: <200803121741.m2CHfcrr090299@repoman.freebsd.org> brooks 2008-03-12 17:41:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) etc/rc.d dhclient Log: Sync with current merging revs 1.27 and 1.31. 1.27: Use get_if_var() to retrieve interface specific values of dhclient_flags and background_dhclient. This allows interfaces who's names are not valid parts of shell variables and shortens the code. 1.31: When the state of the interface changes rapidly enough there is a race where dhclient is in the process of exiting due to the link going down when the link coming up causes devd to try and start a new one. This results is the link being up, but no dhclient running. Work around this race by checking a second time after a one second delay before refusing to start a dhclient instance due to one already being running. Revision Changes Path 1.20.2.6 +8 -8 src/etc/rc.d/dhclient From remko at FreeBSD.org Wed Mar 12 18:23:41 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Wed Mar 12 18:23:46 2008 Subject: cvs commit: src/sys/dev/agp agp_i810.c Message-ID: <200803121823.m2CINdaX093482@repoman.freebsd.org> remko 2008-03-12 18:23:39 UTC FreeBSD src repository Modified files: sys/dev/agp agp_i810.c Log: Add resume support to the agp_i810 family. Submitted by: "Robert Noland" Reviewed by: anholt Approved by: anholt, imp (mentor) MFC after: 1 week Revision Changes Path 1.45 +17 -0 src/sys/dev/agp/agp_i810.c From remko at FreeBSD.org Wed Mar 12 18:25:48 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Wed Mar 12 18:25:57 2008 Subject: cvs commit: src/sys/sys disklabel.h Message-ID: <200803121825.m2CIPlJC093583@repoman.freebsd.org> remko 2008-03-12 18:25:47 UTC FreeBSD src repository Modified files: sys/sys disklabel.h Log: Add missing comma. PR: bin/121645 Submitted by: OISHI Masakuni Approved by: imp (mentor, implicit for trivial changes) MFC after: 3 days Revision Changes Path 1.111 +1 -1 src/sys/sys/disklabel.h From sam at FreeBSD.org Wed Mar 12 19:09:21 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Wed Mar 12 19:09:22 2008 Subject: cvs commit: src/sys/dev/uart uart.h uart_dev_ns8250.c Message-ID: <200803121909.m2CJ9KIn097769@repoman.freebsd.org> sam 2008-03-12 19:09:20 UTC FreeBSD src repository Modified files: sys/dev/uart uart.h uart_dev_ns8250.c Log: add device hints to control the rx FIFO interrupt level on 16550A parts PR: kern/121421 Submitted by: UEMURA Tetsuya Reviewed by: marcel MFC after: 2 weeks Revision Changes Path 1.8 +4 -0 src/sys/dev/uart/uart.h 1.28 +14 -1 src/sys/dev/uart/uart_dev_ns8250.c From sam at FreeBSD.org Wed Mar 12 19:09:59 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Wed Mar 12 19:10:00 2008 Subject: cvs commit: src/share/man/man4 uart.4 Message-ID: <200803121909.m2CJ9wPe097811@repoman.freebsd.org> sam 2008-03-12 19:09:58 UTC FreeBSD src repository Modified files: share/man/man4 uart.4 Log: document device hints including new ones to control rx fifo trigger MFC after: 2 weeks Revision Changes Path 1.12 +28 -1 src/share/man/man4/uart.4 From sam at FreeBSD.org Wed Mar 12 20:03:32 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Wed Mar 12 20:03:37 2008 Subject: cvs commit: src/sys/dev/wpi if_wpi.c Message-ID: <200803122003.m2CK3VQe001207@repoman.freebsd.org> sam 2008-03-12 20:03:31 UTC FreeBSD src repository Modified files: sys/dev/wpi if_wpi.c Log: fix inverted test that disabled ACK's on xmit Revision Changes Path 1.10 +1 -1 src/sys/dev/wpi/if_wpi.c From jhb at FreeBSD.org Wed Mar 12 20:20:36 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 12 20:20:43 2008 Subject: cvs commit: src/sys/dev/usb ums.c Message-ID: <200803122020.m2CKKaJu001992@repoman.freebsd.org> jhb 2008-03-12 20:20:36 UTC FreeBSD src repository Modified files: sys/dev/usb ums.c Log: Partially revert 1.95. It changed the probe for a mouse device to only accept a mouse using the boot subclass. Instead, restore the original hid_is_collection() test and fallback to testing the interface class, subclass, and protocol if that fails. MFC after: 1 week PR: usb/118670 Revision Changes Path 1.98 +4 -1 src/sys/dev/usb/ums.c From jhb at FreeBSD.org Wed Mar 12 20:57:18 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 12 20:57:23 2008 Subject: cvs commit: src/sys/dev/usb ehci_pci.c Message-ID: <200803122057.m2CKvITs012197@repoman.freebsd.org> jhb 2008-03-12 20:57:17 UTC FreeBSD src repository Modified files: sys/dev/usb ehci_pci.c Log: Relax the BIOS/OS sempahore handoff code to workaround different hard hangs (one at boot, one at shutdown) in recent machines. First, only try to take ownership of the EHCI controller if the BIOS currently owns the controller. On a HP DL160 G5, the machine hangs when we try to take ownership. Second, don't bother trying to give up ownership of the controller during shutdown. It's not strictly required and a Dell DCS S29 hangs on shutdown after the config write. Both of these changes match the behavior of the Linux EHCI driver. I also think both of these hangs are caused by bugs in the BIOS' SMM handler causing it to get stuck in an infinite loop in SMM. MFC after: 1 week Revision Changes Path 1.30 +4 -1 src/sys/dev/usb/ehci_pci.c From kaiw at FreeBSD.org Wed Mar 12 21:10:26 2008 From: kaiw at FreeBSD.org (Kai Wang) Date: Wed Mar 12 21:10:28 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_ar.c src/lib/libarchive/test test_read_format_ar.c test_write_format_ar.c Message-ID: <200803122110.m2CLAQPC013916@repoman.freebsd.org> kaiw 2008-03-12 21:10:26 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_ar.c lib/libarchive/test test_read_format_ar.c test_write_format_ar.c Log: Current 'ar' read support in libarchive can only handle a GNU/SVR4 filename table whose size is less than 65536 bytes. The original intention was to not consume the filename table, so the client will have a chance to look at it. To achieve that, the library call decompressor->read_ahead to read(look ahead) but do not call decompressor->consume to consume the data, thus a limit was raised since read_ahead call can only look ahead at most BUFFER_SIZE(65536) bytes at the moment, and you can not "look any further" before you consume what you already "saw". This commit will turn GNU/SVR4 filename table into "archive format data", i.e., filename table will be consumed by libarchive, so the 65536-bytes limit will be gone, but client can no longer have access to the content of filename table. 'ar' support test suite is changed accordingly. BSD ar(1) is not affected by this change since it doesn't look at the filename table. Reported by: erwin Discussed with: jkoshy, kientzle Reviewed by: jkoshy, kientzle Approved by: jkoshy(mentor), kientzle Revision Changes Path 1.9 +39 -36 src/lib/libarchive/archive_read_support_format_ar.c 1.5 +1 -3 src/lib/libarchive/test/test_read_format_ar.c 1.6 +1 -3 src/lib/libarchive/test/test_write_format_ar.c From jhb at FreeBSD.org Wed Mar 12 21:43:51 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 12 21:43:55 2008 Subject: cvs commit: src/sys/amd64/amd64 amd64_mem.c mem.c Message-ID: <200803122143.m2CLhoVu017039@repoman.freebsd.org> jhb 2008-03-12 21:43:50 UTC FreeBSD src repository Modified files: sys/amd64/amd64 amd64_mem.c mem.c Log: Minimize diffs with i686_mem.c: - A few whitespace changes I missed in the style(9) changes. - Move M_MEMDESC to mem.c. Revision Changes Path 1.28 +4 -6 src/sys/amd64/amd64/amd64_mem.c 1.122 +5 -0 src/sys/amd64/amd64/mem.c From jhb at FreeBSD.org Wed Mar 12 21:44:47 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 12 21:44:48 2008 Subject: cvs commit: src/sys/i386/i386 i686_mem.c k6_mem.c Message-ID: <200803122144.m2CLilCr017118@repoman.freebsd.org> jhb 2008-03-12 21:44:47 UTC FreeBSD src repository Modified files: sys/i386/i386 i686_mem.c k6_mem.c Log: MFamd64: Break up the probe logic in the mem_drvinit routines so it's a bit easier to parse. Revision Changes Path 1.27 +10 -7 src/sys/i386/i386/i686_mem.c 1.13 +8 -6 src/sys/i386/i386/k6_mem.c From jroberson at chesapeake.net Wed Mar 12 21:54:21 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Wed Mar 12 21:54:25 2008 Subject: cvs commit: src/lib Makefile src/lib/libthread_db Makefile In-Reply-To: <34B0E2E4-4F7F-4A0B-97DD-2D69DCC612AD@mac.com> References: <200803120949.m2C9ndTW031022@repoman.freebsd.org> <34B0E2E4-4F7F-4A0B-97DD-2D69DCC612AD@mac.com> Message-ID: <20080312115424.U1091@desktop> On Wed, 12 Mar 2008, Marcel Moolenaar wrote: > > On Mar 12, 2008, at 2:49 AM, Jeff Roberson wrote: > >> jeff 2008-03-12 09:49:39 UTC >> >> FreeBSD src repository >> >> Modified files: >> lib Makefile >> lib/libthread_db Makefile >> Log: >> - Remove libkse and related support code in libpthread from the build. >> Don't remove the files yet. Kernel support will be removed shortly. > > I think removing support to debug processes that link > against libkse is premature. You want to keep the > support for at least a full major release, but ideally > until the last release branch that supported KSE is > obsoleted. Hey Marcel, I thought about that but I'm not sure how it would work. Any debugging would have to happen on a kernel that supported KSE. That kernel will have to be 7.x or prior. Thanks, Jeff > > FYI, > > -- > Marcel Moolenaar > xcllnt@mac.com > From xcllnt at mac.com Wed Mar 12 21:57:54 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Wed Mar 12 21:57:58 2008 Subject: cvs commit: src/lib Makefile src/lib/libthread_db Makefile In-Reply-To: <20080312115424.U1091@desktop> References: <200803120949.m2C9ndTW031022@repoman.freebsd.org> <34B0E2E4-4F7F-4A0B-97DD-2D69DCC612AD@mac.com> <20080312115424.U1091@desktop> Message-ID: On Mar 12, 2008, at 2:55 PM, Jeff Roberson wrote: > > On Wed, 12 Mar 2008, Marcel Moolenaar wrote: > >> >> On Mar 12, 2008, at 2:49 AM, Jeff Roberson wrote: >> >>> jeff 2008-03-12 09:49:39 UTC >>> FreeBSD src repository >>> Modified files: >>> lib Makefile >>> lib/libthread_db Makefile >>> Log: >>> - Remove libkse and related support code in libpthread from the >>> build. >>> Don't remove the files yet. Kernel support will be removed >>> shortly. >> >> I think removing support to debug processes that link >> against libkse is premature. You want to keep the >> support for at least a full major release, but ideally >> until the last release branch that supported KSE is >> obsoleted. > > Hey Marcel, > > I thought about that but I'm not sure how it would work. Any > debugging would have to happen on a kernel that supported KSE. That > kernel will have to be 7.x or prior. I was thinking about core files. There's no need for kernel support in that case, but we do need suport in libthread_db in order to extract thread state from the core file. FYI, -- Marcel Moolenaar xcllnt@mac.com From jhb at FreeBSD.org Wed Mar 12 22:09:20 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Mar 12 22:09:24 2008 Subject: cvs commit: src/sys/amd64/amd64 amd64_mem.c src/sys/amd64/include specialreg.h src/sys/i386/i386 i686_mem.c src/sys/i386/include specialreg.h Message-ID: <200803122209.m2CM9JjL019437@repoman.freebsd.org> jhb 2008-03-12 22:09:19 UTC FreeBSD src repository Modified files: sys/amd64/amd64 amd64_mem.c sys/amd64/include specialreg.h sys/i386/i386 i686_mem.c sys/i386/include specialreg.h Log: The variable MTRR registers actually have variable-sized PhysBase and PhysMask fields based on the number of physical address bits supported by the current CPU. The old code assumed 36 bits on i386 and 40 bits on amd64. In truth, all Intel CPUs up until recently used 36 bits (a newer Intel CPU uses 38 bits) and all the Opteron CPUs used 40 bits. In at least one case (the new Intel CPU) having the size of the mask field wrong resulted in writing questionable values into the MTRR registers on the application processors (BSP as well if you modify the MTRRs via memcontrol or running X, etc.). The result of the questionable physmask was that all of memory was apparently treated as uncached rather than write-back resulting in a very significant performance hit. Fix this by constructing a run-time mask for the PhysBase and PhysMask fields based on the number of physical address bits supported by the CPU. All 64-bit capable CPUs provide a count of PA bits supported via the 0x80000008 extended CPUID feature, so use that if it is available. If that feature is not available, then assume 36 PA bits. While I'm here, expand the (now-unused) macros for the PhysBase and PhysMask fields to the current largest possible value (52 PA bits). MFC after: 1 week PR: i386/120516 Reported by: Nokia Revision Changes Path 1.29 +24 -6 src/sys/amd64/amd64/amd64_mem.c 1.45 +2 -2 src/sys/amd64/include/specialreg.h 1.28 +24 -6 src/sys/i386/i386/i686_mem.c 1.44 +2 -2 src/sys/i386/include/specialreg.h From jeff at FreeBSD.org Wed Mar 12 22:14:15 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 22:14:17 2008 Subject: cvs commit: src/sys/ia64/ia64 machdep.c Message-ID: <200803122214.m2CMEFZn019644@repoman.freebsd.org> jeff 2008-03-12 22:14:15 UTC FreeBSD src repository Modified files: sys/ia64/ia64 machdep.c Log: - Fix build breakage; there was a reference to a removed syscall in a KASSERT(). Attempt to cleanup the comment to reflect reality. Revision Changes Path 1.232 +3 -4 src/sys/ia64/ia64/machdep.c From jeff at FreeBSD.org Wed Mar 12 22:17:06 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 22:17:08 2008 Subject: cvs commit: src/sys/compat/svr4 svr4_sysvec.c Message-ID: <200803122217.m2CMH6kl019744@repoman.freebsd.org> jeff 2008-03-12 22:17:06 UTC FreeBSD src repository Modified files: sys/compat/svr4 svr4_sysvec.c Log: - The P_SA flag has been removed. Don't reference it in a KASSERT. Revision Changes Path 1.43 +1 -2 src/sys/compat/svr4/svr4_sysvec.c From jroberson at chesapeake.net Wed Mar 12 22:33:20 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Wed Mar 12 22:33:23 2008 Subject: cvs commit: src/lib Makefile src/lib/libthread_db Makefile In-Reply-To: References: <200803120949.m2C9ndTW031022@repoman.freebsd.org> <34B0E2E4-4F7F-4A0B-97DD-2D69DCC612AD@mac.com> <20080312115424.U1091@desktop> Message-ID: <20080312123331.T1091@desktop> On Wed, 12 Mar 2008, Marcel Moolenaar wrote: > > On Mar 12, 2008, at 2:55 PM, Jeff Roberson wrote: > >> >> On Wed, 12 Mar 2008, Marcel Moolenaar wrote: >> >>> >>> On Mar 12, 2008, at 2:49 AM, Jeff Roberson wrote: >>> >>>> jeff 2008-03-12 09:49:39 UTC >>>> FreeBSD src repository >>>> Modified files: >>>> lib Makefile >>>> lib/libthread_db Makefile >>>> Log: >>>> - Remove libkse and related support code in libpthread from the build. >>>> Don't remove the files yet. Kernel support will be removed shortly. >>> >>> I think removing support to debug processes that link >>> against libkse is premature. You want to keep the >>> support for at least a full major release, but ideally >>> until the last release branch that supported KSE is >>> obsoleted. >> >> Hey Marcel, >> >> I thought about that but I'm not sure how it would work. Any debugging >> would have to happen on a kernel that supported KSE. That kernel will have >> to be 7.x or prior. > > I was thinking about core files. There's no need for kernel > support in that case, but we do need suport in libthread_db > in order to extract thread state from the core file. > FYI, Oh excellent point. I'll restore that. I can move the kse.h file from the Attic into libthread_db. I hadn't deleted the library or the thread_db yet incase something like this came up. Thanks, Jeff > > -- > Marcel Moolenaar > xcllnt@mac.com > From xcllnt at mac.com Wed Mar 12 22:42:50 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Wed Mar 12 22:42:58 2008 Subject: cvs commit: src/lib Makefile src/lib/libthread_db Makefile In-Reply-To: <20080312123331.T1091@desktop> References: <200803120949.m2C9ndTW031022@repoman.freebsd.org> <34B0E2E4-4F7F-4A0B-97DD-2D69DCC612AD@mac.com> <20080312115424.U1091@desktop> <20080312123331.T1091@desktop> Message-ID: <675E8F7C-16B2-410C-8239-E9E396A6F6BC@mac.com> On Mar 12, 2008, at 3:34 PM, Jeff Roberson wrote: > > On Wed, 12 Mar 2008, Marcel Moolenaar wrote: > >> >> On Mar 12, 2008, at 2:55 PM, Jeff Roberson wrote: >> >>> On Wed, 12 Mar 2008, Marcel Moolenaar wrote: >>>> On Mar 12, 2008, at 2:49 AM, Jeff Roberson wrote: >>>>> jeff 2008-03-12 09:49:39 UTC >>>>> FreeBSD src repository >>>>> Modified files: >>>>> lib Makefile >>>>> lib/libthread_db Makefile >>>>> Log: >>>>> - Remove libkse and related support code in libpthread from the >>>>> build. >>>>> Don't remove the files yet. Kernel support will be removed >>>>> shortly. >>>> I think removing support to debug processes that link >>>> against libkse is premature. You want to keep the >>>> support for at least a full major release, but ideally >>>> until the last release branch that supported KSE is >>>> obsoleted. >>> Hey Marcel, >>> I thought about that but I'm not sure how it would work. Any >>> debugging would have to happen on a kernel that supported KSE. >>> That kernel will have to be 7.x or prior. >> >> I was thinking about core files. There's no need for kernel >> support in that case, but we do need suport in libthread_db >> in order to extract thread state from the core file. >> FYI, > > Oh excellent point. I'll restore that. I can move the kse.h file > from the Attic into libthread_db. I hadn't deleted the library or > the thread_db yet incase something like this came up. Thanks; and thanks for taking care of the ia64 tinderbox failure :-) -- Marcel Moolenaar xcllnt@mac.com From delphij at FreeBSD.org Wed Mar 12 23:01:41 2008 From: delphij at FreeBSD.org (Xin LI) Date: Wed Mar 12 23:01:47 2008 Subject: cvs commit: src/contrib/tcsh - Imported sources Message-ID: <200803122301.m2CN1fvG026873@repoman.freebsd.org> delphij 2008-03-12 23:01:41 UTC FreeBSD src repository src/contrib/tcsh - Imported sources Update of /home/ncvs/src/contrib/tcsh In directory repoman.freebsd.org:/tmp/cvs-serv26837 Log Message: Reapply a bugfix that was done as sh.lex.c,v 1.1.1.8 Status: Vendor Tag: ZOULAS Release Tags: tcsh_6_15b U src/contrib/tcsh/tc.os.c U src/contrib/tcsh/sh.sem.c U src/contrib/tcsh/Makefile.ADMIN U src/contrib/tcsh/tc.str.c U src/contrib/tcsh/tw.color.c U src/contrib/tcsh/sh.exec.c U src/contrib/tcsh/FAQ U src/contrib/tcsh/tc.alloc.c U src/contrib/tcsh/sh.proc.c U src/contrib/tcsh/MAKESHAR U src/contrib/tcsh/tw.parse.c U src/contrib/tcsh/tc.printf.c U src/contrib/tcsh/Makefile.vms U src/contrib/tcsh/mi.termios.c U src/contrib/tcsh/tw.init.c U src/contrib/tcsh/gethost.c U src/contrib/tcsh/ed.init.c U src/contrib/tcsh/tw.comp.c U src/contrib/tcsh/termcap.vms U src/contrib/tcsh/ed.term.h U src/contrib/tcsh/glob.c U src/contrib/tcsh/sh.file.c U src/contrib/tcsh/NewThings U src/contrib/tcsh/glob.3 U src/contrib/tcsh/sh.types.h U src/contrib/tcsh/tcsh.man2html U src/contrib/tcsh/sh.init.c U src/contrib/tcsh/sh.dir.c U src/contrib/tcsh/tc.who.c U src/contrib/tcsh/Ported U src/contrib/tcsh/tc.func.c U src/contrib/tcsh/tw.spell.c U src/contrib/tcsh/sh.dol.c U src/contrib/tcsh/sh.exp.c U src/contrib/tcsh/ed.refresh.c U src/contrib/tcsh/ed.h U src/contrib/tcsh/tc.disc.c U src/contrib/tcsh/sh.print.c U src/contrib/tcsh/ed.decls.h U src/contrib/tcsh/config.guess U src/contrib/tcsh/sh.char.h U src/contrib/tcsh/tw.help.c U src/contrib/tcsh/MAKEDIFFS U src/contrib/tcsh/mi.wait.h U src/contrib/tcsh/Fixes U src/contrib/tcsh/tc.nls.h U src/contrib/tcsh/tc.sig.h U src/contrib/tcsh/ed.defns.c U src/contrib/tcsh/sh.glob.c U src/contrib/tcsh/Y2K U src/contrib/tcsh/sh.c U src/contrib/tcsh/tcsh.man U src/contrib/tcsh/install-sh U src/contrib/tcsh/config.h.in U src/contrib/tcsh/README.imake U src/contrib/tcsh/sh.decls.h U src/contrib/tcsh/configure.in U src/contrib/tcsh/sh.set.c U src/contrib/tcsh/tw.h U src/contrib/tcsh/sh.dir.h U src/contrib/tcsh/ma.setp.c U src/contrib/tcsh/RELEASE-PROCEDURE U src/contrib/tcsh/sh.func.c U src/contrib/tcsh/WishList U src/contrib/tcsh/csh-mode.el U src/contrib/tcsh/glob.h U src/contrib/tcsh/ed.term.c U src/contrib/tcsh/sh.parse.c U src/contrib/tcsh/sh.proc.h U src/contrib/tcsh/tc.h U src/contrib/tcsh/README U src/contrib/tcsh/sh.time.c U src/contrib/tcsh/configure U src/contrib/tcsh/complete.tcsh U src/contrib/tcsh/vms.termcap.c U src/contrib/tcsh/snames.h U src/contrib/tcsh/eight-bit.me U src/contrib/tcsh/patchlevel.h U src/contrib/tcsh/tc.wait.h U src/contrib/tcsh/tc.os.h U src/contrib/tcsh/ed.screen.c U src/contrib/tcsh/ed.chared.c U src/contrib/tcsh/tc.vers.c U src/contrib/tcsh/Makefile.in U src/contrib/tcsh/src.desc U src/contrib/tcsh/ed.inputl.c U src/contrib/tcsh/config.sub U src/contrib/tcsh/tc.sched.c U src/contrib/tcsh/sh.hist.c U src/contrib/tcsh/host.defs U src/contrib/tcsh/tc.nls.c U src/contrib/tcsh/tc.sig.c U src/contrib/tcsh/sh.misc.c U src/contrib/tcsh/Imakefile U src/contrib/tcsh/sh.h U src/contrib/tcsh/mi.varargs.h U src/contrib/tcsh/sh.err.c U src/contrib/tcsh/sh.lex.c U src/contrib/tcsh/ed.xmap.c U src/contrib/tcsh/Copyright U src/contrib/tcsh/sh.char.c U src/contrib/tcsh/tc.const.c U src/contrib/tcsh/tc.bind.c U src/contrib/tcsh/imake.config U src/contrib/tcsh/tw.decls.h U src/contrib/tcsh/config_f.h U src/contrib/tcsh/tc.prompt.c U src/contrib/tcsh/pathnames.h U src/contrib/tcsh/tc.decls.h U src/contrib/tcsh/Makefile.std U src/contrib/tcsh/nls/Makefile U src/contrib/tcsh/nls/pl/set24 U src/contrib/tcsh/nls/pl/set1 U src/contrib/tcsh/nls/pl/set16 U src/contrib/tcsh/nls/pl/set11 U src/contrib/tcsh/nls/pl/set23 U src/contrib/tcsh/nls/pl/set6 U src/contrib/tcsh/nls/pl/set29 U src/contrib/tcsh/nls/pl/set30 U src/contrib/tcsh/nls/pl/set2 U src/contrib/tcsh/nls/pl/set27 U src/contrib/tcsh/nls/pl/set8 U src/contrib/tcsh/nls/pl/set15 U src/contrib/tcsh/nls/pl/set18 U src/contrib/tcsh/nls/pl/set12 U src/contrib/tcsh/nls/pl/set5 U src/contrib/tcsh/nls/pl/set20 U src/contrib/tcsh/nls/pl/charset U src/contrib/tcsh/nls/pl/set31 U src/contrib/tcsh/nls/pl/set10 U src/contrib/tcsh/nls/pl/set7 U src/contrib/tcsh/nls/pl/set22 U src/contrib/tcsh/nls/pl/README U src/contrib/tcsh/nls/pl/set25 U src/contrib/tcsh/nls/pl/set17 U src/contrib/tcsh/nls/pl/set13 U src/contrib/tcsh/nls/pl/set19 U src/contrib/tcsh/nls/pl/set21 U src/contrib/tcsh/nls/pl/set4 U src/contrib/tcsh/nls/pl/set26 U src/contrib/tcsh/nls/pl/set9 U src/contrib/tcsh/nls/pl/set3 U src/contrib/tcsh/nls/pl/set14 U src/contrib/tcsh/nls/greek/set9 U src/contrib/tcsh/nls/greek/set3 U src/contrib/tcsh/nls/greek/set14 U src/contrib/tcsh/nls/greek/set26 U src/contrib/tcsh/nls/greek/set21 U src/contrib/tcsh/nls/greek/set19 U src/contrib/tcsh/nls/greek/charset U src/contrib/tcsh/nls/greek/set13 U src/contrib/tcsh/nls/greek/set4 U src/contrib/tcsh/nls/greek/set17 U src/contrib/tcsh/nls/greek/set25 U src/contrib/tcsh/nls/greek/set22 U src/contrib/tcsh/nls/greek/set10 U src/contrib/tcsh/nls/greek/set7 U src/contrib/tcsh/nls/greek/set31 U src/contrib/tcsh/nls/greek/set5 U src/contrib/tcsh/nls/greek/set20 U src/contrib/tcsh/nls/greek/set12 U src/contrib/tcsh/nls/greek/set18 U src/contrib/tcsh/nls/greek/set15 U src/contrib/tcsh/nls/greek/set27 U src/contrib/tcsh/nls/greek/set2 U src/contrib/tcsh/nls/greek/set8 U src/contrib/tcsh/nls/greek/set6 U src/contrib/tcsh/nls/greek/set30 U src/contrib/tcsh/nls/greek/set29 U src/contrib/tcsh/nls/greek/set23 U src/contrib/tcsh/nls/greek/set11 U src/contrib/tcsh/nls/greek/set16 U src/contrib/tcsh/nls/greek/set24 U src/contrib/tcsh/nls/greek/set1 U src/contrib/tcsh/nls/spanish/set25 U src/contrib/tcsh/nls/spanish/set17 U src/contrib/tcsh/nls/spanish/set7 U src/contrib/tcsh/nls/spanish/charset U src/contrib/tcsh/nls/spanish/set31 U src/contrib/tcsh/nls/spanish/set10 U src/contrib/tcsh/nls/spanish/set22 U src/contrib/tcsh/nls/spanish/set26 U src/contrib/tcsh/nls/spanish/set14 U src/contrib/tcsh/nls/spanish/set4 U src/contrib/tcsh/nls/spanish/set3 U src/contrib/tcsh/nls/spanish/set9 U src/contrib/tcsh/nls/spanish/set13 U src/contrib/tcsh/nls/spanish/set19 U src/contrib/tcsh/nls/spanish/set21 U src/contrib/tcsh/nls/spanish/set11 U src/contrib/tcsh/nls/spanish/set23 U src/contrib/tcsh/nls/spanish/set29 U src/contrib/tcsh/nls/spanish/set1 U src/contrib/tcsh/nls/spanish/set30 U src/contrib/tcsh/nls/spanish/set6 U src/contrib/tcsh/nls/spanish/set24 U src/contrib/tcsh/nls/spanish/set16 U src/contrib/tcsh/nls/spanish/set18 U src/contrib/tcsh/nls/spanish/set12 U src/contrib/tcsh/nls/spanish/set20 U src/contrib/tcsh/nls/spanish/set8 U src/contrib/tcsh/nls/spanish/set2 U src/contrib/tcsh/nls/spanish/set5 U src/contrib/tcsh/nls/spanish/set27 U src/contrib/tcsh/nls/spanish/set15 U src/contrib/tcsh/nls/et/set22 U src/contrib/tcsh/nls/et/set10 U src/contrib/tcsh/nls/et/set3 U src/contrib/tcsh/nls/et/set9 U src/contrib/tcsh/nls/et/set31 U src/contrib/tcsh/nls/et/set4 U src/contrib/tcsh/nls/et/set17 U src/contrib/tcsh/nls/et/set25 U src/contrib/tcsh/nls/et/set21 U src/contrib/tcsh/nls/et/set19 U src/contrib/tcsh/nls/et/set13 U src/contrib/tcsh/nls/et/set7 U src/contrib/tcsh/nls/et/set14 U src/contrib/tcsh/nls/et/set26 U src/contrib/tcsh/nls/et/set16 U src/contrib/tcsh/nls/et/set24 U src/contrib/tcsh/nls/et/set5 U src/contrib/tcsh/nls/et/set8 U src/contrib/tcsh/nls/et/set30 U src/contrib/tcsh/nls/et/set2 U src/contrib/tcsh/nls/et/set29 U src/contrib/tcsh/nls/et/set23 U src/contrib/tcsh/nls/et/set11 U src/contrib/tcsh/nls/et/set15 U src/contrib/tcsh/nls/et/set27 U src/contrib/tcsh/nls/et/charset U src/contrib/tcsh/nls/et/set6 U src/contrib/tcsh/nls/et/set1 U src/contrib/tcsh/nls/et/set20 U src/contrib/tcsh/nls/et/set12 U src/contrib/tcsh/nls/et/set18 U src/contrib/tcsh/nls/finnish/set9 U src/contrib/tcsh/nls/finnish/set3 U src/contrib/tcsh/nls/finnish/set12 U src/contrib/tcsh/nls/finnish/set18 U src/contrib/tcsh/nls/finnish/set20 U src/contrib/tcsh/nls/finnish/set27 U src/contrib/tcsh/nls/finnish/set15 U src/contrib/tcsh/nls/finnish/set4 U src/contrib/tcsh/nls/finnish/set30 U src/contrib/tcsh/nls/finnish/set11 U src/contrib/tcsh/nls/finnish/set29 U src/contrib/tcsh/nls/finnish/set23 U src/contrib/tcsh/nls/finnish/set24 U src/contrib/tcsh/nls/finnish/charset U src/contrib/tcsh/nls/finnish/set16 U src/contrib/tcsh/nls/finnish/set7 U src/contrib/tcsh/nls/finnish/set5 U src/contrib/tcsh/nls/finnish/set26 U src/contrib/tcsh/nls/finnish/set14 U src/contrib/tcsh/nls/finnish/set19 U src/contrib/tcsh/nls/finnish/set13 U src/contrib/tcsh/nls/finnish/set21 U src/contrib/tcsh/nls/finnish/set2 U src/contrib/tcsh/nls/finnish/set8 U src/contrib/tcsh/nls/finnish/set6 U src/contrib/tcsh/nls/finnish/set25 U src/contrib/tcsh/nls/finnish/set17 U src/contrib/tcsh/nls/finnish/set10 U src/contrib/tcsh/nls/finnish/set22 U src/contrib/tcsh/nls/finnish/set1 U src/contrib/tcsh/nls/finnish/set31 U src/contrib/tcsh/nls/italian/set6 U src/contrib/tcsh/nls/italian/set14 U src/contrib/tcsh/nls/italian/set26 U src/contrib/tcsh/nls/italian/set21 U src/contrib/tcsh/nls/italian/set13 U src/contrib/tcsh/nls/italian/set19 U src/contrib/tcsh/nls/italian/set1 U src/contrib/tcsh/nls/italian/set5 U src/contrib/tcsh/nls/italian/set17 U src/contrib/tcsh/nls/italian/set25 U src/contrib/tcsh/nls/italian/set22 U src/contrib/tcsh/nls/italian/set10 U src/contrib/tcsh/nls/italian/set31 U src/contrib/tcsh/nls/italian/set2 U src/contrib/tcsh/nls/italian/set8 U src/contrib/tcsh/nls/italian/set20 U src/contrib/tcsh/nls/italian/set18 U src/contrib/tcsh/nls/italian/set12 U src/contrib/tcsh/nls/italian/set15 U src/contrib/tcsh/nls/italian/set27 U src/contrib/tcsh/nls/italian/set7 U src/contrib/tcsh/nls/italian/set30 U src/contrib/tcsh/nls/italian/set9 U src/contrib/tcsh/nls/italian/set3 U src/contrib/tcsh/nls/italian/set23 U src/contrib/tcsh/nls/italian/set29 U src/contrib/tcsh/nls/italian/set11 U src/contrib/tcsh/nls/italian/set16 U src/contrib/tcsh/nls/italian/charset U src/contrib/tcsh/nls/italian/set24 U src/contrib/tcsh/nls/italian/set4 U src/contrib/tcsh/nls/ukrainian/set24 U src/contrib/tcsh/nls/ukrainian/charset U src/contrib/tcsh/nls/ukrainian/set16 U src/contrib/tcsh/nls/ukrainian/set5 U src/contrib/tcsh/nls/ukrainian/set30 U src/contrib/tcsh/nls/ukrainian/set2 U src/contrib/tcsh/nls/ukrainian/set8 U src/contrib/tcsh/nls/ukrainian/set11 U src/contrib/tcsh/nls/ukrainian/set29 U src/contrib/tcsh/nls/ukrainian/set23 U src/contrib/tcsh/nls/ukrainian/set27 U src/contrib/tcsh/nls/ukrainian/set15 U src/contrib/tcsh/nls/ukrainian/set6 U src/contrib/tcsh/nls/ukrainian/set1 U src/contrib/tcsh/nls/ukrainian/set12 U src/contrib/tcsh/nls/ukrainian/set18 U src/contrib/tcsh/nls/ukrainian/set20 U src/contrib/tcsh/nls/ukrainian/set10 U src/contrib/tcsh/nls/ukrainian/set22 U src/contrib/tcsh/nls/ukrainian/set31 U src/contrib/tcsh/nls/ukrainian/set9 U src/contrib/tcsh/nls/ukrainian/set3 U src/contrib/tcsh/nls/ukrainian/set4 U src/contrib/tcsh/nls/ukrainian/set25 U src/contrib/tcsh/nls/ukrainian/set17 U src/contrib/tcsh/nls/ukrainian/set19 U src/contrib/tcsh/nls/ukrainian/set13 U src/contrib/tcsh/nls/ukrainian/set21 U src/contrib/tcsh/nls/ukrainian/set7 U src/contrib/tcsh/nls/ukrainian/set26 U src/contrib/tcsh/nls/ukrainian/set14 U src/contrib/tcsh/nls/french/set10 U src/contrib/tcsh/nls/french/set22 U src/contrib/tcsh/nls/french/set9 U src/contrib/tcsh/nls/french/set3 U src/contrib/tcsh/nls/french/set31 U src/contrib/tcsh/nls/french/set4 U src/contrib/tcsh/nls/french/set25 U src/contrib/tcsh/nls/french/set17 U src/contrib/tcsh/nls/french/set19 U src/contrib/tcsh/nls/french/set13 U src/contrib/tcsh/nls/french/set21 U src/contrib/tcsh/nls/french/set26 U src/contrib/tcsh/nls/french/set7 U src/contrib/tcsh/nls/french/set14 U src/contrib/tcsh/nls/french/set5 U src/contrib/tcsh/nls/french/set24 U src/contrib/tcsh/nls/french/set16 U src/contrib/tcsh/nls/french/charset U src/contrib/tcsh/nls/french/set30 U src/contrib/tcsh/nls/french/set11 U src/contrib/tcsh/nls/french/set29 U src/contrib/tcsh/nls/french/set23 U src/contrib/tcsh/nls/french/set2 U src/contrib/tcsh/nls/french/set8 U src/contrib/tcsh/nls/french/set27 U src/contrib/tcsh/nls/french/set6 U src/contrib/tcsh/nls/french/set15 U src/contrib/tcsh/nls/french/set12 U src/contrib/tcsh/nls/french/set18 U src/contrib/tcsh/nls/french/set1 U src/contrib/tcsh/nls/french/set20 U src/contrib/tcsh/nls/C/set10 U src/contrib/tcsh/nls/C/set4 U src/contrib/tcsh/nls/C/set22 U src/contrib/tcsh/nls/C/set31 U src/contrib/tcsh/nls/C/set25 U src/contrib/tcsh/nls/C/set17 U src/contrib/tcsh/nls/C/set3 U src/contrib/tcsh/nls/C/set9 U src/contrib/tcsh/nls/C/set19 U src/contrib/tcsh/nls/C/set7 U src/contrib/tcsh/nls/C/set13 U src/contrib/tcsh/nls/C/set21 U src/contrib/tcsh/nls/C/set26 U src/contrib/tcsh/nls/C/set14 U src/contrib/tcsh/nls/C/set24 U src/contrib/tcsh/nls/C/set8 U src/contrib/tcsh/nls/C/set2 U src/contrib/tcsh/nls/C/set16 U src/contrib/tcsh/nls/C/charset U src/contrib/tcsh/nls/C/set30 U src/contrib/tcsh/nls/C/set5 U src/contrib/tcsh/nls/C/set11 U src/contrib/tcsh/nls/C/set29 U src/contrib/tcsh/nls/C/set23 U src/contrib/tcsh/nls/C/set27 U src/contrib/tcsh/nls/C/set15 U src/contrib/tcsh/nls/C/set1 U src/contrib/tcsh/nls/C/set12 U src/contrib/tcsh/nls/C/set6 U src/contrib/tcsh/nls/C/set18 U src/contrib/tcsh/nls/C/set20 U src/contrib/tcsh/nls/russian/set14 U src/contrib/tcsh/nls/russian/set1 U src/contrib/tcsh/nls/russian/set26 U src/contrib/tcsh/nls/russian/set6 U src/contrib/tcsh/nls/russian/set21 U src/contrib/tcsh/nls/russian/set19 U src/contrib/tcsh/nls/russian/set13 U src/contrib/tcsh/nls/russian/set17 U src/contrib/tcsh/nls/russian/set25 U src/contrib/tcsh/nls/russian/set8 U src/contrib/tcsh/nls/russian/set2 U src/contrib/tcsh/nls/russian/set31 U src/contrib/tcsh/nls/russian/set22 U src/contrib/tcsh/nls/russian/set5 U src/contrib/tcsh/nls/russian/set10 U src/contrib/tcsh/nls/russian/set20 U src/contrib/tcsh/nls/russian/set7 U src/contrib/tcsh/nls/russian/set12 U src/contrib/tcsh/nls/russian/set18 U src/contrib/tcsh/nls/russian/set15 U src/contrib/tcsh/nls/russian/set27 U src/contrib/tcsh/nls/russian/set29 U src/contrib/tcsh/nls/russian/set4 U src/contrib/tcsh/nls/russian/set23 U src/contrib/tcsh/nls/russian/set11 U src/contrib/tcsh/nls/russian/charset U src/contrib/tcsh/nls/russian/set30 U src/contrib/tcsh/nls/russian/set16 U src/contrib/tcsh/nls/russian/set3 U src/contrib/tcsh/nls/russian/set9 U src/contrib/tcsh/nls/russian/set24 U src/contrib/tcsh/nls/ja/set11 U src/contrib/tcsh/nls/ja/set29 U src/contrib/tcsh/nls/ja/set4 U src/contrib/tcsh/nls/ja/set30 U src/contrib/tcsh/nls/ja/set3 U src/contrib/tcsh/nls/ja/set24 U src/contrib/tcsh/nls/ja/set16 U src/contrib/tcsh/nls/ja/charset U src/contrib/tcsh/nls/ja/set18 U src/contrib/tcsh/nls/ja/set12 U src/contrib/tcsh/nls/ja/set7 U src/contrib/tcsh/nls/ja/set15 U src/contrib/tcsh/nls/ja/set17 U src/contrib/tcsh/nls/ja/set2 U src/contrib/tcsh/nls/ja/set8 U src/contrib/tcsh/nls/ja/set5 U src/contrib/tcsh/nls/ja/set10 U src/contrib/tcsh/nls/ja/set1 U src/contrib/tcsh/nls/ja/set6 U src/contrib/tcsh/nls/ja/set13 U src/contrib/tcsh/nls/ja/set21 U src/contrib/tcsh/nls/german/set20 U src/contrib/tcsh/nls/german/set12 U src/contrib/tcsh/nls/german/set18 U src/contrib/tcsh/nls/german/set7 U src/contrib/tcsh/nls/german/set15 U src/contrib/tcsh/nls/german/set27 U src/contrib/tcsh/nls/german/set29 U src/contrib/tcsh/nls/german/set23 U src/contrib/tcsh/nls/german/set11 U src/contrib/tcsh/nls/german/set30 U src/contrib/tcsh/nls/german/set4 U src/contrib/tcsh/nls/german/set3 U src/contrib/tcsh/nls/german/set9 U src/contrib/tcsh/nls/german/set16 U src/contrib/tcsh/nls/german/set24 U src/contrib/tcsh/nls/german/set14 U src/contrib/tcsh/nls/german/set26 U src/contrib/tcsh/nls/german/set1 U src/contrib/tcsh/nls/german/set6 U src/contrib/tcsh/nls/german/set21 U src/contrib/tcsh/nls/german/set19 U src/contrib/tcsh/nls/german/set13 U src/contrib/tcsh/nls/german/set17 U src/contrib/tcsh/nls/german/set25 U src/contrib/tcsh/nls/german/set8 U src/contrib/tcsh/nls/german/set2 U src/contrib/tcsh/nls/german/set31 U src/contrib/tcsh/nls/german/charset U src/contrib/tcsh/nls/german/set5 U src/contrib/tcsh/nls/german/set22 U src/contrib/tcsh/nls/german/set10 U src/contrib/tcsh/config/bsd4.4 No conflicts created by this import From jeff at FreeBSD.org Wed Mar 12 23:54:40 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Wed Mar 12 23:54:42 2008 Subject: cvs commit: src/usr.bin/cpuset cpuset.1 cpuset.c Message-ID: <200803122354.m2CNse7E033853@repoman.freebsd.org> jeff 2008-03-12 23:54:40 UTC FreeBSD src repository Modified files: usr.bin/cpuset cpuset.1 cpuset.c Log: - When running a new command cause cpuset to operate on the per-thread mask by default rather than the setmask. This is consistent with the linux tool and more consistent with the notion that the default level is the process level. The cpuset mask can still be modified by specifying the -c option. You can not set the per-thread and cpuset mask in a single command. - Update the man page to reflect this change. Contributed by: gallatin Revision Changes Path 1.3 +6 -2 src/usr.bin/cpuset/cpuset.1 1.4 +2 -2 src/usr.bin/cpuset/cpuset.c From jeff at FreeBSD.org Thu Mar 13 00:46:13 2008 From: jeff at FreeBSD.org (Jeff Roberson) Date: Thu Mar 13 00:46:16 2008 Subject: cvs commit: src/sys/kern subr_sleepqueue.c Message-ID: <200803130046.m2D0kDbx036689@repoman.freebsd.org> jeff 2008-03-13 00:46:12 UTC FreeBSD src repository Modified files: sys/kern subr_sleepqueue.c Log: PR 117603 - Close a sleepqueue signal race by interlocking with the per-process spinlock. This was mistakenly omitted from the thread_lock patch and has been a race since. MFC After: 1 week PR: bin/117603 Reported by: Danny Braniss Revision Changes Path 1.48 +5 -2 src/sys/kern/subr_sleepqueue.c From davidxu at FreeBSD.org Thu Mar 13 01:53:29 2008 From: davidxu at FreeBSD.org (David Xu) Date: Thu Mar 13 01:53:31 2008 Subject: cvs commit: src/lib/libc/gen sem_timedwait.3 sem_wait.3 Message-ID: <200803130153.m2D1rSO6049937@repoman.freebsd.org> davidxu 2008-03-13 01:53:28 UTC FreeBSD src repository Modified files: lib/libc/gen sem_timedwait.3 sem_wait.3 Log: In file sem_timewait.3, remove reference to SYSV semphore in SEE ALSO section, sync it with sem_wait.3. Revision Changes Path 1.3 +2 -4 src/lib/libc/gen/sem_timedwait.3 1.14 +1 -0 src/lib/libc/gen/sem_wait.3 From wollman at FreeBSD.org Thu Mar 13 02:28:31 2008 From: wollman at FreeBSD.org (Garrett Wollman) Date: Thu Mar 13 02:28:35 2008 Subject: cvs commit: src/lib/libc/gen sysconf.c Message-ID: <200803130228.m2D2SUiV052202@repoman.freebsd.org> wollman 2008-03-13 02:28:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/gen sysconf.c Log: MFC rev. 1.26: stdio is currently limited to file descriptors not greater than {SHRT_MAX}, so {STREAM_MAX} should be no greater than that. (This does not exactly meet the letter of POSIX but comes reasonably close to it in spirit.) Revision Changes Path 1.25.2.1 +19 -1 src/lib/libc/gen/sysconf.c From wollman at FreeBSD.org Thu Mar 13 02:31:38 2008 From: wollman at FreeBSD.org (Garrett Wollman) Date: Thu Mar 13 02:31:44 2008 Subject: cvs commit: src/lib/libc/gen sysconf.c Message-ID: <200803130231.m2D2VcYk052361@repoman.freebsd.org> wollman 2008-03-13 02:31:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libc/gen sysconf.c Log: MFC rev. 1.26: stdio is currently limited to file descriptors not greater than {SHRT_MAX}, so {STREAM_MAX} should be no greater than that. (This does not exactly meet the letter of POSIX but comes reasonably close to it in spirit.) Revision Changes Path 1.20.14.1 +19 -1 src/lib/libc/gen/sysconf.c From wollman at FreeBSD.org Thu Mar 13 02:33:21 2008 From: wollman at FreeBSD.org (Garrett Wollman) Date: Thu Mar 13 02:33:25 2008 Subject: cvs commit: src/lib/libc/gen sysconf.c Message-ID: <200803130233.m2D2XLXj052477@repoman.freebsd.org> wollman 2008-03-13 02:33:21 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) lib/libc/gen sysconf.c Log: MFC rev. 1.26: stdio is currently limited to file descriptors not greater than {SHRT_MAX}, so {STREAM_MAX} should be no greater than that. (This does not exactly meet the letter of POSIX but comes reasonably close to it in spirit.) Revision Changes Path 1.20.8.1 +19 -1 src/lib/libc/gen/sysconf.c From davidxu at FreeBSD.org Thu Mar 13 02:56:12 2008 From: davidxu at FreeBSD.org (David Xu) Date: Thu Mar 13 02:56:14 2008 Subject: cvs commit: src/sys/sys cpuset.h Message-ID: <200803130256.m2D2uCrj056124@repoman.freebsd.org> davidxu 2008-03-13 02:56:11 UTC FreeBSD src repository Modified files: sys/sys cpuset.h Log: Add const qualifier to cpuset mask's pointer, since the cpuset mask should be not changed by the system call. Revision Changes Path 1.3 +1 -1 src/sys/sys/cpuset.h From kip.macy at gmail.com Thu Mar 13 07:26:58 2008 From: kip.macy at gmail.com (Kip Macy) Date: Thu Mar 13 07:27:04 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c trap.c vm_machdep.c src/sys/amd64/conf DEFAULTS src/sys/amd64/linux32 linux32_sysvec.c src/sys/arm/arm trap.c src/sys/arm/conf AVILA src/sys/arm/xscale/i8134x crb_machdep.c src/sys/compat/freebsd32 ... In-Reply-To: <200803121123.12620.jhb@freebsd.org> References: <200803121012.m2CAC24p033661@repoman.freebsd.org> <200803121123.12620.jhb@freebsd.org> Message-ID: It has in fact been discussed at great length. The only question was the timing. Most of us would agree that JeffR's time would be much better spent on lock pushdown than arguing about the specifics of KSE removal. -Kip On Wed, Mar 12, 2008 at 8:23 AM, John Baldwin wrote: > On Wednesday 12 March 2008 06:12:02 am Jeff Roberson wrote: > > jeff 2008-03-12 10:12:02 UTC > > > > FreeBSD src repository > > > > Modified files: > > ... > > > Log: > > Remove kernel support for M:N threading. > > > > While the KSE project was quite successful in bringing threading to > > FreeBSD, the M:N approach taken by the kse library was never developed > > to its full potential. Backwards compatibility will be provided via > > libmap.conf for dynamically linked binaries and static binaries will > > be broken. > > While I don't disagree with the change, it might have been nice to have > pre-announced the axeing first. Even with Alpha (which probably had more > support and which was discussed in person at BSDCan with a lot of developers) > having the removal not actually discussed on the lists at all resulted in > some bumps that could have been easily avoided with a bit more patience on my > part. > > -- > John Baldwin > From yongari at FreeBSD.org Thu Mar 13 07:50:49 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Thu Mar 13 07:50:53 2008 Subject: cvs commit: src/sys/dev/bce if_bce.c src/sys/dev/bfe if_bfe.c src/sys/dev/dc if_dc.c src/sys/dev/lge if_lge.c src/sys/dev/msk if_msk.c src/sys/dev/nfe if_nfe.c src/sys/dev/nge if_nge.c src/sys/dev/nve if_nve.c src/sys/dev/re if_re.c src/sys/dev/sk ... Message-ID: <200803130750.m2D7on6c003023@repoman.freebsd.org> yongari 2008-03-13 07:50:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/bce if_bce.c sys/dev/bfe if_bfe.c sys/dev/dc if_dc.c sys/dev/lge if_lge.c sys/dev/msk if_msk.c sys/dev/nfe if_nfe.c sys/dev/nge if_nge.c sys/dev/nve if_nve.c sys/dev/re if_re.c sys/dev/sk if_sk.c sys/dev/stge if_stge.c sys/dev/ti if_ti.c sys/dev/vge if_vge.c sys/dev/vx if_vx_pci.c sys/pci if_rl.c if_sf.c if_ste.c if_vr.c if_xl.c Log: MFC: Fix function prototype for device_shutdown method. Revision Changes Path 1.34.2.1 +5 -3 src/sys/dev/bce/if_bce.c 1.42.2.1 +4 -3 src/sys/dev/bfe/if_bfe.c 1.192.2.1 +4 -2 src/sys/dev/dc/if_dc.c 1.50.2.1 +3 -3 src/sys/dev/lge/if_lge.c 1.18.2.8 +3 -2 src/sys/dev/msk/if_msk.c 1.21.2.3 +4 -2 src/sys/dev/nfe/if_nfe.c 1.91.2.1 +3 -3 src/sys/dev/nge/if_nge.c 1.28.2.1 +4 -2 src/sys/dev/nve/if_nve.c 1.95.2.5 +4 -2 src/sys/dev/re/if_re.c 1.137.2.1 +3 -3 src/sys/dev/sk/if_sk.c 1.9.2.1 +4 -2 src/sys/dev/stge/if_stge.c 1.128.2.2 +4 -2 src/sys/dev/ti/if_ti.c 1.31.2.1 +4 -2 src/sys/dev/vge/if_vge.c 1.37.2.1 +4 -2 src/sys/dev/vx/if_vx_pci.c 1.170.2.2 +4 -2 src/sys/pci/if_rl.c 1.95.2.1 +4 -2 src/sys/pci/if_sf.c 1.99.2.1 +3 -3 src/sys/pci/if_ste.c 1.126.2.2 +4 -2 src/sys/pci/if_vr.c 1.210.2.1 +4 -2 src/sys/pci/if_xl.c From kib at FreeBSD.org Thu Mar 13 09:52:49 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Thu Mar 13 09:52:50 2008 Subject: cvs commit: src/sys/i386/i386 k6_mem.c Message-ID: <200803130952.m2D9qmP9024492@repoman.freebsd.org> kib 2008-03-13 09:52:48 UTC FreeBSD src repository Modified files: sys/i386/i386 k6_mem.c Log: Add missed parentheses Revision Changes Path 1.14 +1 -1 src/sys/i386/i386/k6_mem.c From rwatson at FreeBSD.org Thu Mar 13 09:58:08 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Thu Mar 13 09:58:11 2008 Subject: cvs commit: src/lib/libc/sys Makefile.inc kldunload.2 Message-ID: <200803130958.m2D9w740024752@repoman.freebsd.org> rwatson 2008-03-13 09:58:07 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/sys Makefile.inc kldunload.2 Log: Merge Makefile.inc:1.129, kldunload.2:1.19 from HEAD to RELENG_7: Add reference to kldunloadf system call, which was previously not mentioned in the kldunload(2) man page. Spotted by: rink Revision Changes Path 1.127.2.1 +1 -0 src/lib/libc/sys/Makefile.inc 1.18.10.1 +14 -1 src/lib/libc/sys/kldunload.2 From bz at FreeBSD.org Thu Mar 13 10:09:12 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Thu Mar 13 10:09:18 2008 Subject: cvs commit: src/sys/netinet tcp_output.c Message-ID: <200803131009.m2DA9C4j026379@repoman.freebsd.org> bz 2008-03-13 10:09:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet tcp_output.c Log: MFC rev. 1.146 tcp_output.c Padding after EOL option must be zeros according to RFC793 but the NOPs used are 0x01. While we could simply pad with EOLs (which are 0x00), rather use an explicit 0x00 constant there to not confuse poeple with 'EOL padding'. Put in a comment saying just that. Early MFC requested by: silby because of more people reporting problems on net@ Revision Changes Path 1.141.2.4 +10 -2 src/sys/netinet/tcp_output.c From ru at FreeBSD.org Thu Mar 13 10:25:31 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Thu Mar 13 10:25:37 2008 Subject: cvs commit: src/lib/libc/gen sem_wait.3 Message-ID: <200803131025.m2DAPVal027038@repoman.freebsd.org> ru 2008-03-13 10:25:30 UTC FreeBSD src repository Modified files: lib/libc/gen sem_wait.3 Log: Add missing section number. Revision Changes Path 1.15 +1 -1 src/lib/libc/gen/sem_wait.3 From ru at FreeBSD.org Thu Mar 13 10:26:18 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Thu Mar 13 10:26:26 2008 Subject: cvs commit: src/lib/libc/gen sem_timedwait.3 Message-ID: <200803131026.m2DAQHnq027093@repoman.freebsd.org> ru 2008-03-13 10:26:17 UTC FreeBSD src repository Modified files: lib/libc/gen sem_timedwait.3 Log: Remove trailing whitespace. Revision Changes Path 1.4 +1 -1 src/lib/libc/gen/sem_timedwait.3 From rwatson at FreeBSD.org Thu Mar 13 10:29:36 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Thu Mar 13 10:29:43 2008 Subject: cvs commit: src/sys/netsmb smb_conn.h Message-ID: <200803131029.m2DATZfT027172@repoman.freebsd.org> rwatson 2008-03-13 10:29:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netsmb smb_conn.h Log: Merge smb_conn.h:1.13 from HEAD to RELENG_7: Remove unused vc_tnode field from struct smb_vc. Revision Changes Path 1.11.10.1 +0 -1 src/sys/netsmb/smb_conn.h From rwatson at FreeBSD.org Thu Mar 13 10:30:07 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Thu Mar 13 10:30:10 2008 Subject: cvs commit: src/sys/security/audit audit_worker.c Message-ID: <200803131030.m2DAU6Yo027247@repoman.freebsd.org> rwatson 2008-03-13 10:30:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/security/audit audit_worker.c Log: Merge audit_worker.c:1.22 from HEAD to RELENG_7: Remove XXX to remind me to check the free space calculation, which to my eyes appears right following a check. Revision Changes Path 1.16.2.1 +0 -3 src/sys/security/audit/audit_worker.c From ru at FreeBSD.org Thu Mar 13 10:33:25 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Thu Mar 13 10:33:27 2008 Subject: cvs commit: src/lib/libc/sys kldunload.2 Message-ID: <200803131033.m2DAXOtU027603@repoman.freebsd.org> ru 2008-03-13 10:33:24 UTC FreeBSD src repository Modified files: lib/libc/sys kldunload.2 Log: Fix bugs in previous revision (missing comma, misspelled syscall name). Revision Changes Path 1.20 +2 -3 src/lib/libc/sys/kldunload.2 From kib at FreeBSD.org Thu Mar 13 10:54:38 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Thu Mar 13 10:54:40 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c src/sys/amd64/ia32 ia32_signal.c src/sys/amd64/linux32 linux32_sysvec.c src/sys/i386/i386 machdep.c src/sys/i386/linux linux_sysvec.c src/sys/i386/svr4 svr4_machdep.c src/sys/pc98/pc98 machdep.c Message-ID: <200803131054.m2DAscBo028803@repoman.freebsd.org> kib 2008-03-13 10:54:38 UTC FreeBSD src repository Modified files: sys/amd64/amd64 machdep.c sys/amd64/ia32 ia32_signal.c sys/amd64/linux32 linux32_sysvec.c sys/i386/i386 machdep.c sys/i386/linux linux_sysvec.c sys/i386/svr4 svr4_machdep.c sys/pc98/pc98 machdep.c Log: Since version 4.3, gcc changed its behaviour concerning the i386/amd64 ABI and the direction flag, that is it now assumes that the direction flag is cleared at the entry of a function and it doesn't clear once more if needed. This new behaviour conforms to the i386/amd64 ABI. Modify the signal handler frame setup code to clear the DF {e,r}flags bit on the amd64/i386 for the signal handlers. jhb@ noted that it might break old apps if they assumed DF == 1 would be preserved in the signal handlers, but that such apps should be rare and that older versions of gcc would not generate such apps. Submitted by: Aurelien Jarno PR: 121422 Reviewed by: jhb MFC after: 2 weeks Revision Changes Path 1.682 +1 -1 src/sys/amd64/amd64/machdep.c 1.16 +2 -2 src/sys/amd64/ia32/ia32_signal.c 1.33 +2 -2 src/sys/amd64/linux32/linux32_sysvec.c 1.666 +3 -3 src/sys/i386/i386/machdep.c 1.152 +2 -2 src/sys/i386/linux/linux_sysvec.c 1.39 +2 -2 src/sys/i386/svr4/svr4_machdep.c 1.402 +3 -3 src/sys/pc98/pc98/machdep.c From csjp at FreeBSD.org Thu Mar 13 13:52:30 2008 From: csjp at FreeBSD.org (Christian S.J. Peron) Date: Thu Mar 13 13:52:40 2008 Subject: cvs commit: src/sys/security/audit audit_syscalls.c Message-ID: <200803131352.m2DDqSbX048962@repoman.freebsd.org> csjp 2008-03-13 13:52:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/security/audit audit_syscalls.c Log: MFC revision 1.25 -- Make sure we return EINVAL for invalid commands Revision Changes Path 1.21.2.1 +3 -0 src/sys/security/audit/audit_syscalls.c From csjp at FreeBSD.org Thu Mar 13 13:53:21 2008 From: csjp at FreeBSD.org (Christian S.J. Peron) Date: Thu Mar 13 13:53:22 2008 Subject: cvs commit: src/sys/security/audit audit_syscalls.c Message-ID: <200803131353.m2DDrK8d049010@repoman.freebsd.org> csjp 2008-03-13 13:53:20 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/security/audit audit_syscalls.c Log: MFC revision 1.25 -- Make sure we return EINVAL for invalid commands Revision Changes Path 1.1.2.13 +3 -0 src/sys/security/audit/audit_syscalls.c From cokane at FreeBSD.org Thu Mar 13 14:08:42 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Thu Mar 13 14:08:47 2008 Subject: cvs commit: src/sys/dev/tdfx tdfx_pci.c Message-ID: <200803131408.m2DE8fi5051051@repoman.freebsd.org> cokane 2008-03-13 14:08:41 UTC FreeBSD src repository Modified files: sys/dev/tdfx tdfx_pci.c Log: Add the module dependency on the mem(4) module. This will fix the module failing to load on a kernel that has "nodevice mem" in the config. It will now properly bring in the mem(4) module. Submitted by: antoine Reviewed by: imp MFC after: 1 week Revision Changes Path 1.40 +1 -0 src/sys/dev/tdfx/tdfx_pci.c From brueffer at FreeBSD.org Thu Mar 13 15:15:40 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Thu Mar 13 15:15:42 2008 Subject: cvs commit: src/sbin/geom/class/multipath gmultipath.8 Message-ID: <200803131515.m2DFFd5Y070518@repoman.freebsd.org> brueffer 2008-03-13 15:15:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/geom/class/multipath gmultipath.8 Log: MFC: rev. 1.5 Fix typos. Revision Changes Path 1.4.2.1 +2 -2 src/sbin/geom/class/multipath/gmultipath.8 From kensmith at FreeBSD.org Thu Mar 13 15:28:18 2008 From: kensmith at FreeBSD.org (Ken Smith) Date: Thu Mar 13 15:28:23 2008 Subject: cvs commit: src/usr.sbin/sysinstall dist.c dist.h menus.c sysinstall.h Message-ID: <200803131528.m2DFSHeH071099@repoman.freebsd.org> kensmith 2008-03-13 15:28:17 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/sysinstall dist.c dist.h menus.c sysinstall.h Log: MFC: dist.c v1.254, dist.h v1.72, menus.c v1.429, sysinstall.h v1.275 > Change sysinstall's handling of X11 stuff. Doing it in pieces was > probably the right thing to do a while ago but xorg has progressed > to the point that for novice users (who are the ones expected to think > installing X11 during an install...) it's best to just install the > whole x11/xorg metaport for them. This removes the X11 sub-menus > and sets it up so you just select whether or not you want X11. While > here garbage collect an X11 configuration menu I missed removing when > I removed support for attempting xorg configuration from inside sysinstall > a while ago. Revision Changes Path 1.239.2.9 +3 -37 src/usr.sbin/sysinstall/dist.c 1.65.2.5 +1 -27 src/usr.sbin/sysinstall/dist.h 1.410.2.14 +1 -196 src/usr.sbin/sysinstall/menus.c 1.264.2.7 +0 -8 src/usr.sbin/sysinstall/sysinstall.h From brueffer at FreeBSD.org Thu Mar 13 15:37:02 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Thu Mar 13 15:37:08 2008 Subject: cvs commit: src/sbin/geom/class/multipath gmultipath.8 Message-ID: <200803131537.m2DFb2tV071465@repoman.freebsd.org> brueffer 2008-03-13 15:37:02 UTC FreeBSD src repository Modified files: sbin/geom/class/multipath gmultipath.8 Log: - Use an uppercase provider name in the example, to make the name change after labeling the provider more obvious. (1) - Correct nomenclature usage PR: 121487 (1) Submitted by: Anatoly Borodin MFC after: 3 days Revision Changes Path 1.6 +7 -7 src/sbin/geom/class/multipath/gmultipath.8 From kib at FreeBSD.org Thu Mar 13 16:45:42 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Thu Mar 13 16:45:44 2008 Subject: cvs commit: CVSROOT access Message-ID: <200803131645.m2DGjg0I075184@repoman.freebsd.org> kib 2008-03-13 16:45:42 UTC FreeBSD src repository Modified files: . access Log: Please, welcome Roman Divacky as src commiter. He successfully finished two Google Summers of Code (2006 and 2007) working on linuxolator. Also Roman submitted a lot of kernel code cleanup patches. I expect Roman to continue his work and expand both the area of interest and knowledge. I will mentor him. Approved by: core Revision Changes Path 1.866 +1 -0 CVSROOT/access From remko at elvandar.org Thu Mar 13 17:30:02 2008 From: remko at elvandar.org (Remko Lodder) Date: Thu Mar 13 17:30:09 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <200803131645.m2DGjg0I075184@repoman.freebsd.org> References: <200803131645.m2DGjg0I075184@repoman.freebsd.org> Message-ID: <59806.195.64.94.120.1205426831.squirrel@galain.elvandar.org> On Thu, March 13, 2008 5:45 pm, Konstantin Belousov wrote: > kib 2008-03-13 16:45:42 UTC > > FreeBSD src repository > > Modified files: > . access > Log: > Please, welcome Roman Divacky as src commiter. He successfully finished > two Google Summers of Code (2006 and 2007) working on linuxolator. > Also Roman submitted a lot of kernel code cleanup patches. > > I expect Roman to continue his work and expand both the area of interest > and knowledge. I will mentor him. > > Approved by: core > > Revision Changes Path > 1.866 +1 -0 CVSROOT/access > Yay! Finally! :) Welcome Roman! -- /"\ Best regards, | remko@FreeBSD.org \ / Remko Lodder | remko@EFnet X http://www.evilcoder.org/ | / \ ASCII Ribbon Campaign | Against HTML Mail and News From attilio at freebsd.org Thu Mar 13 17:34:36 2008 From: attilio at freebsd.org (Attilio Rao) Date: Thu Mar 13 17:34:38 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <200803131645.m2DGjg0I075184@repoman.freebsd.org> References: <200803131645.m2DGjg0I075184@repoman.freebsd.org> Message-ID: <3bbf2fe10803131034v717cd538s3e71729df2621658@mail.gmail.com> 2008/3/13, Konstantin Belousov : > kib 2008-03-13 16:45:42 UTC > > FreeBSD src repository > > Modified files: > . access > Log: > Please, welcome Roman Divacky as src commiter. He successfully finished > two Google Summers of Code (2006 and 2007) working on linuxolator. > Also Roman submitted a lot of kernel code cleanup patches. hear hear... Welcome and congratulations! Attilio -- Peace can only be achieved by understanding - A. Einstein From obrien at FreeBSD.org Thu Mar 13 17:38:07 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Thu Mar 13 17:38:08 2008 Subject: cvs commit: src/usr.bin Makefile src/bin Makefile src/bin/uuidgen Makefile src/usr.bin/uuidgen Makefile uuidgen.1 uuidgen.c Message-ID: <200803131738.m2DHc6Z9087224@repoman.freebsd.org> obrien 2008-03-13 17:38:06 UTC FreeBSD src repository Modified files: usr.bin Makefile bin Makefile bin/uuidgen Makefile Removed files: usr.bin/uuidgen Makefile uuidgen.1 uuidgen.c Log: uuidgen has been repo-copied from usr.bin/ to bin/ to match its "new" (2007/04/09) installation location. Revision Changes Path 1.27 +2 -1 src/bin/Makefile 1.4 +0 -2 src/bin/uuidgen/Makefile 1.312 +0 -1 src/usr.bin/Makefile 1.4 +0 -8 src/usr.bin/uuidgen/Makefile (dead) 1.8 +0 -80 src/usr.bin/uuidgen/uuidgen.1 (dead) 1.4 +0 -113 src/usr.bin/uuidgen/uuidgen.c (dead) From delphij at delphij.net Thu Mar 13 17:50:54 2008 From: delphij at delphij.net (Xin LI) Date: Thu Mar 13 17:50:55 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <200803131645.m2DGjg0I075184@repoman.freebsd.org> References: <200803131645.m2DGjg0I075184@repoman.freebsd.org> Message-ID: <47D96970.2000308@delphij.net> Konstantin Belousov wrote: > kib 2008-03-13 16:45:42 UTC > > FreeBSD src repository > > Modified files: > . access > Log: > Please, welcome Roman Divacky as src commiter. He successfully finished > two Google Summers of Code (2006 and 2007) working on linuxolator. > Also Roman submitted a lot of kernel code cleanup patches. > > I expect Roman to continue his work and expand both the area of interest > and knowledge. I will mentor him. > > Approved by: core > > Revision Changes Path > 1.866 +1 -0 CVSROOT/access Woo! Welcome! =-) -- Xin LI http://www.delphij.net/ FreeBSD - The Power to Serve! From obrien at FreeBSD.org Thu Mar 13 17:54:22 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Thu Mar 13 17:54:29 2008 Subject: cvs commit: src/sys/boot/uboot Makefile src/sys/boot/uboot/common main.c src/sys/boot/uboot/lib Makefile api_public.h copy.c devicename.c disk.c elf_freebsd.c glue.c glue.h libuboot.h module.c net.c reboot.c Message-ID: <200803131754.m2DHsMpE088696@repoman.freebsd.org> obrien 2008-03-13 17:54:21 UTC FreeBSD src repository Modified files: sys/boot/uboot Makefile sys/boot/uboot/common main.c sys/boot/uboot/lib Makefile api_public.h copy.c devicename.c disk.c elf_freebsd.c glue.c glue.h libuboot.h module.c net.c reboot.c Log: style(9) & style.Makefile(9) Reviewed by: raj Revision Changes Path 1.2 +2 -2 src/sys/boot/uboot/Makefile 1.3 +8 -9 src/sys/boot/uboot/common/main.c 1.3 +2 -3 src/sys/boot/uboot/lib/Makefile 1.3 +29 -29 src/sys/boot/uboot/lib/api_public.h 1.3 +2 -2 src/sys/boot/uboot/lib/copy.c 1.2 +156 -149 src/sys/boot/uboot/lib/devicename.c 1.2 +8 -2 src/sys/boot/uboot/lib/disk.c 1.2 +3 -4 src/sys/boot/uboot/lib/elf_freebsd.c 1.4 +58 -30 src/sys/boot/uboot/lib/glue.c 1.2 +10 -10 src/sys/boot/uboot/lib/glue.h 1.3 +10 -10 src/sys/boot/uboot/lib/libuboot.h 1.2 +1 -0 src/sys/boot/uboot/lib/module.c 1.3 +9 -10 src/sys/boot/uboot/lib/net.c 1.2 +1 -0 src/sys/boot/uboot/lib/reboot.c From brueffer at FreeBSD.org Thu Mar 13 18:17:48 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Thu Mar 13 18:17:49 2008 Subject: cvs commit: src ObsoleteFiles.inc Message-ID: <200803131817.m2DIHl45093548@repoman.freebsd.org> brueffer 2008-03-13 18:17:47 UTC FreeBSD src repository Modified files: . ObsoleteFiles.inc Log: Add an include file from the removed XRPU driver. PR: 121654 Submitted by: pluknet Approved by: rwatson (mentor) Revision Changes Path 1.133 +2 -0 src/ObsoleteFiles.inc From obrien at FreeBSD.org Thu Mar 13 18:52:47 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Thu Mar 13 18:52:55 2008 Subject: cvs commit: src/tools/tools/editing freebsd.vim Message-ID: <200803131852.m2DIqkw3095857@repoman.freebsd.org> obrien 2008-03-13 18:52:46 UTC FreeBSD src repository Added files: tools/tools/editing freebsd.vim Log: Vim function to style(9). Issue ':call FreeBSD_Style()' while editing a file. Submitted by: scf Revision Changes Path 1.1 +68 -0 src/tools/tools/editing/freebsd.vim (new) From jhb at FreeBSD.org Thu Mar 13 18:56:53 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Mar 13 18:56:55 2008 Subject: cvs commit: src/sys/i386/i386 machdep.c Message-ID: <200803131856.m2DIurjS096021@repoman.freebsd.org> jhb 2008-03-13 18:56:53 UTC FreeBSD src repository Modified files: sys/i386/i386 machdep.c Log: Use the SMAP data from the loader if it is provided instead of using virtual 86 mode to query the BIOS directly. This is needed for certain HP machines whose BIOS only provide an SMAP when invoked from real mode. On such machines the loader will be able to query the SMAP successfully due to the recent BTX changes, but the kernel will not. One thing I'm not sure of is if we can skip the INT 12h probe altogether if we have the SMAP from the loader as it seems that we do the INT 12h probe to setup enough state so we can use vm86 to call the BIOS. MFC after: 1 week Revision Changes Path 1.667 +99 -60 src/sys/i386/i386/machdep.c From stas at FreeBSD.org Thu Mar 13 20:15:33 2008 From: stas at FreeBSD.org (Stanislav Sedov) Date: Thu Mar 13 20:15:41 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <200803131645.m2DGjg0I075184@repoman.freebsd.org> References: <200803131645.m2DGjg0I075184@repoman.freebsd.org> Message-ID: <20080313201529.GE68662@dracon.ht-systems.ru> On Thu, Mar 13, 2008 at 04:45:42PM +0000 Konstantin Belousov mentioned: > kib 2008-03-13 16:45:42 UTC > > FreeBSD src repository > > Modified files: > . access > Log: > Please, welcome Roman Divacky as src commiter. He successfully finished > two Google Summers of Code (2006 and 2007) working on linuxolator. > Also Roman submitted a lot of kernel code cleanup patches. > > I expect Roman to continue his work and expand both the area of interest > and knowledge. I will mentor him. > Welcome! -- Stanislav Sedov ST4096-RIPE From cokane at FreeBSD.org Thu Mar 13 20:15:49 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Thu Mar 13 20:15:56 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_softdep.c Message-ID: <200803132015.m2DKFndQ008412@repoman.freebsd.org> cokane 2008-03-13 20:15:49 UTC FreeBSD src repository Modified files: sys/ufs/ffs ffs_softdep.c Log: Replace the non-MPSAFE timeout(9) API in ffs_softdep.c with the MPSAFE callout_* API (e.g. callout_init_mtx(9)). This was one of the numerous items on the http://wiki.freebsd.org/SMPTODO list. Reviewed by: imp, obrien, jhb MFC after: 1 week Revision Changes Path 1.219 +15 -8 src/sys/ufs/ffs/ffs_softdep.c From rink at FreeBSD.org Thu Mar 13 20:16:28 2008 From: rink at FreeBSD.org (Rink Springer) Date: Thu Mar 13 20:16:33 2008 Subject: cvs commit: src/sys/i386/i386 machdep.c In-Reply-To: <200803131856.m2DIurjS096021@repoman.freebsd.org> References: <200803131856.m2DIurjS096021@repoman.freebsd.org> Message-ID: <20080313195116.GJ77132@rink.nu> On Thu, Mar 13, 2008 at 06:56:53PM +0000, John Baldwin wrote: > jhb 2008-03-13 18:56:53 UTC > > FreeBSD src repository > > Modified files: > sys/i386/i386 machdep.c > Log: > Use the SMAP data from the loader if it is provided instead of using > virtual 86 mode to query the BIOS directly. This is needed for certain > HP machines whose BIOS only provide an SMAP when invoked from real mode. > On such machines the loader will be able to query the SMAP successfully > due to the recent BTX changes, but the kernel will not. Thanks! I was thinking about doing this myself - seems you beat me to it. > One thing I'm not sure of is if we can skip the INT 12h probe altogether > if we have the SMAP from the loader as it seems that we do the INT 12h > probe to setup enough state so we can use vm86 to call the BIOS. Hmm, can't we just have loader(8) pass this information as well? It seems it already does in the bootinfo structure... -- Rink P.W. Springer - http://rink.nu "Anyway boys, this is America. Just because you get more votes doesn't mean you win." - Fox Mulder From cokane at FreeBSD.org Thu Mar 13 20:19:37 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Thu Mar 13 20:19:42 2008 Subject: cvs commit: src/tools/tools/editing freebsd.vim In-Reply-To: <200803131852.m2DIqkw3095857@repoman.freebsd.org> References: <200803131852.m2DIqkw3095857@repoman.freebsd.org> Message-ID: <47D987BA.5060802@FreeBSD.org> David E. O'Brien wrote: > obrien 2008-03-13 18:52:46 UTC > > FreeBSD src repository > > Added files: > tools/tools/editing freebsd.vim > Log: > Vim function to style(9). > Issue ':call FreeBSD_Style()' while editing a file. > > Submitted by: scf > > Revision Changes Path > 1.1 +68 -0 src/tools/tools/editing/freebsd.vim (new) > Most excellent! -- coleman From jhb at freebsd.org Thu Mar 13 20:38:28 2008 From: jhb at freebsd.org (John Baldwin) Date: Thu Mar 13 20:38:33 2008 Subject: cvs commit: src/sys/i386/i386 machdep.c In-Reply-To: <20080313195116.GJ77132@rink.nu> References: <200803131856.m2DIurjS096021@repoman.freebsd.org> <20080313195116.GJ77132@rink.nu> Message-ID: <200803131556.24388.jhb@freebsd.org> On Thursday 13 March 2008 03:51:16 pm Rink Springer wrote: > On Thu, Mar 13, 2008 at 06:56:53PM +0000, John Baldwin wrote: > > jhb 2008-03-13 18:56:53 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/i386/i386 machdep.c > > Log: > > Use the SMAP data from the loader if it is provided instead of using > > virtual 86 mode to query the BIOS directly. This is needed for certain > > HP machines whose BIOS only provide an SMAP when invoked from real mode. > > On such machines the loader will be able to query the SMAP successfully > > due to the recent BTX changes, but the kernel will not. > > Thanks! I was thinking about doing this myself - seems you beat me to > it. > > > One thing I'm not sure of is if we can skip the INT 12h probe altogether > > if we have the SMAP from the loader as it seems that we do the INT 12h > > probe to setup enough state so we can use vm86 to call the BIOS. > > Hmm, can't we just have loader(8) pass this information as well? It > seems it already does in the bootinfo structure... I think people still want to be able to boot kernels from boot2/gptboot rather than loader in which case we'd always have to have the vm86 calls that are there now. -- John Baldwin From jhb at FreeBSD.org Thu Mar 13 20:39:05 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Mar 13 20:39:07 2008 Subject: cvs commit: src/sys/amd64/acpica acpi_machdep.c src/sys/amd64/amd64 legacy.c nexus.c src/sys/amd64/include nexusvar.h src/sys/dev/acpica acpi.c acpivar.h src/sys/i386/acpica acpi_machdep.c src/sys/i386/i386 legacy.c nexus.c ... Message-ID: <200803132039.m2DKd4BK009423@repoman.freebsd.org> jhb 2008-03-13 20:39:04 UTC FreeBSD src repository Modified files: sys/amd64/acpica acpi_machdep.c sys/amd64/amd64 legacy.c nexus.c sys/dev/acpica acpi.c acpivar.h sys/i386/acpica acpi_machdep.c sys/i386/i386 legacy.c nexus.c sys/ia64/ia64 nexus.c Added files: sys/amd64/include nexusvar.h sys/i386/include nexusvar.h Log: Rework how the nexus(4) device works on x86 to better handle the idea of different "platforms" on x86 machines. The existing code already handles having two platforms: ACPI and legacy. However, the existing approach was rather hardcoded and difficult to extend. These changes take the approach that each x86 hardware platform should provide its own nexus(4) driver (it can inherit most of its behavior from the default legacy nexus(4) driver) which is responsible for probing for the platform and performing appropriate platform-specific setup during attach (such as adding a platform-specific bus device). This does mean changing the x86 platform busses to no longer use an identify routine for probing, but to move that logic into their matching nexus(4) driver instead. - Make the default nexus(4) driver in nexus.c on i386 and amd64 handle the legacy platform. It's probe routine now returns BUS_PROBE_GENERIC so it can be overriden. - Expose a nexus_init_resources() routine which initializes the various resource managers so that subclassed nexus(4) drivers can invoke it from their attach routine. - The legacy nexus(4) driver explicitly adds a legacy0 device in its attach routine. - The ACPI driver no longer contains an new-bus identify method. Instead it exposes a public function (acpi_identify()) which is a probe routine that the MD nexus(4) drivers can use to probe for ACPI. All of the probe logic in acpi_probe() is now moved into acpi_identify() and acpi_probe() is just a stub. - On i386 and amd64, an ACPI-specific nexus(4) driver checks for ACPI via acpi_identify() and claims the nexus0 device if the probe succeeds. It then explicitly adds an acpi0 device in its attach routine. - The legacy(4) driver no longer knows anything about the acpi0 device. - On ia64 if acpi_identify() fails you basically end up with no devices. This matches the previous behavior where the old acpi_identify() would fail to add an acpi0 device again leaving you with no devices. Discussed with: imp Silence on: arch@ Revision Changes Path 1.18 +44 -0 src/sys/amd64/acpica/acpi_machdep.c 1.63 +0 -21 src/sys/amd64/amd64/legacy.c 1.79 +24 -17 src/sys/amd64/amd64/nexus.c 1.1 +45 -0 src/sys/amd64/include/nexusvar.h (new) 1.247 +38 -45 src/sys/dev/acpica/acpi.c 1.109 +1 -0 src/sys/dev/acpica/acpivar.h 1.38 +43 -0 src/sys/i386/acpica/acpi_machdep.c 1.65 +0 -21 src/sys/i386/i386/legacy.c 1.74 +24 -17 src/sys/i386/i386/nexus.c 1.4 +45 -0 src/sys/i386/include/nexusvar.h (new) 1.19 +6 -0 src/sys/ia64/ia64/nexus.c From grehan at freebsd.org Thu Mar 13 20:46:36 2008 From: grehan at freebsd.org (Peter Grehan) Date: Thu Mar 13 20:46:46 2008 Subject: cvs commit: src/sys/amd64/acpica acpi_machdep.c src/sys/amd64/amd64 legacy.c nexus.c src/sys/amd64/include nexusvar.h src/sys/dev/acpica acpi.c acpivar.h src/sys/i386/acpica acpi_machdep.c src/sys/i386/i386 legacy.c nexus.c ... In-Reply-To: <200803132039.m2DKd4BK009423@repoman.freebsd.org> References: <200803132039.m2DKd4BK009423@repoman.freebsd.org> Message-ID: <47D9928F.4020308@freebsd.org> John Baldwin wrote: > jhb 2008-03-13 20:39:04 UTC > > FreeBSD src repository > > Modified files: > sys/amd64/acpica acpi_machdep.c > sys/amd64/amd64 legacy.c nexus.c > sys/dev/acpica acpi.c acpivar.h > sys/i386/acpica acpi_machdep.c > sys/i386/i386 legacy.c nexus.c > sys/ia64/ia64 nexus.c > Added files: > sys/amd64/include nexusvar.h > sys/i386/include nexusvar.h > Log: > Rework how the nexus(4) device works on x86 to better handle the idea of > different "platforms" on x86 machines. Thank you very much for this ! It will eliminate a lot of nasty hacks that have been done in the NetApp codebase to support non-ACPI Opteron systems. > Silence on: arch@ Interpret my silence as enthusiastic support :) later, Peter. From rink at freebsd.org Thu Mar 13 21:02:15 2008 From: rink at freebsd.org (Rink Springer) Date: Thu Mar 13 21:02:21 2008 Subject: cvs commit: src/sys/i386/i386 machdep.c In-Reply-To: <200803131556.24388.jhb@freebsd.org> References: <200803131856.m2DIurjS096021@repoman.freebsd.org> <20080313195116.GJ77132@rink.nu> <200803131556.24388.jhb@freebsd.org> Message-ID: <20080313210202.GL77132@rink.nu> On Thu, Mar 13, 2008 at 03:56:24PM -0400, John Baldwin wrote: > I think people still want to be able to boot kernels from boot2/gptboot rather > than loader in which case we'd always have to have the vm86 calls that are > there now. My suggestion is to always use SMAP / bootinfo if it's there; if it's not there, use the old-fashioned way. Otherwise, things like GRUB will die too ... -- Rink P.W. Springer - http://rink.nu "Anyway boys, this is America. Just because you get more votes doesn't mean you win." - Fox Mulder From imp at bsdimp.com Thu Mar 13 21:22:36 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Thu Mar 13 21:22:41 2008 Subject: cvs commit: src/sys/amd64/amd64 machdep.c trap.c vm_machdep.c src/sys/amd64/conf DEFAULTS src/sys/amd64/linux32 linux32_sysvec.c src/sys/arm/arm trap.c src/sys/arm/conf AVILA src/sys/arm/xscale/i8134x crb_machdep.c src/sys/compat/freebsd32 ... In-Reply-To: References: <200803121012.m2CAC24p033661@repoman.freebsd.org> <200803121123.12620.jhb@freebsd.org> Message-ID: <20080313.152100.1530737473.imp@bsdimp.com> In message: "Kip Macy" writes: : It has in fact been discussed at great length. The only question was : the timing. : : Most of us would agree that JeffR's time would be much better spent on : lock pushdown than arguing about the specifics of KSE removal. However, he should have sent out a heads up that this was going to happen to make sure there were no lingering issues or objections. In a large project like FreeBSD, you can't go "cowboy" on the tree like Jeff did. Warner : On Wed, Mar 12, 2008 at 8:23 AM, John Baldwin wrote: : > On Wednesday 12 March 2008 06:12:02 am Jeff Roberson wrote: : > > jeff 2008-03-12 10:12:02 UTC : > > : > > FreeBSD src repository : > > : > > Modified files: : > > ... : > : > > Log: : > > Remove kernel support for M:N threading. : > > : > > While the KSE project was quite successful in bringing threading to : > > FreeBSD, the M:N approach taken by the kse library was never developed : > > to its full potential. Backwards compatibility will be provided via : > > libmap.conf for dynamically linked binaries and static binaries will : > > be broken. : > : > While I don't disagree with the change, it might have been nice to have : > pre-announced the axeing first. Even with Alpha (which probably had more : > support and which was discussed in person at BSDCan with a lot of developers) : > having the removal not actually discussed on the lists at all resulted in : > some bumps that could have been easily avoided with a bit more patience on my : > part. : > : > -- : > John Baldwin : > : From rwatson at FreeBSD.org Thu Mar 13 21:56:27 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Thu Mar 13 21:56:34 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_softdep.c In-Reply-To: <200803132015.m2DKFndQ008412@repoman.freebsd.org> References: <200803132015.m2DKFndQ008412@repoman.freebsd.org> Message-ID: <20080313215520.X83983@fledge.watson.org> On Thu, 13 Mar 2008, Coleman Kane wrote: > cokane 2008-03-13 20:15:49 UTC > > FreeBSD src repository > > Modified files: > sys/ufs/ffs ffs_softdep.c > Log: > Replace the non-MPSAFE timeout(9) API in ffs_softdep.c with the MPSAFE > callout_* API (e.g. callout_init_mtx(9)). This was one of the numerous > items on the http://wiki.freebsd.org/SMPTODO list. > > Reviewed by: imp, obrien, jhb > MFC after: 1 week I might give this one a month to MFC rather than a week -- it can take a while for bugs in this sort of thing to get diagnosed, especially with lots of other changes in the tree at the same time relating to schedulers, etc. Robert N M Watson Computer Laboratory University of Cambridge From gabor at kovesdan.org Thu Mar 13 22:22:12 2008 From: gabor at kovesdan.org (Gabor Kovesdan) Date: Thu Mar 13 22:22:16 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <200803131645.m2DGjg0I075184@repoman.freebsd.org> References: <200803131645.m2DGjg0I075184@repoman.freebsd.org> Message-ID: <47D9A540.70201@kovesdan.org> Konstantin Belousov escribi?: > kib 2008-03-13 16:45:42 UTC > > FreeBSD src repository > > Modified files: > . access > Log: > Please, welcome Roman Divacky as src commiter. He successfully finished > two Google Summers of Code (2006 and 2007) working on linuxolator. > Also Roman submitted a lot of kernel code cleanup patches. > > Finally! :) Welcome! G?bor K?vesd?n From edwin at FreeBSD.org Thu Mar 13 22:36:03 2008 From: edwin at FreeBSD.org (Edwin Groothuis) Date: Thu Mar 13 22:36:04 2008 Subject: cvs commit: src/share/examples/cvsup cvs-supfile Message-ID: <200803132236.m2DMa2Gu031397@repoman.freebsd.org> edwin 2008-03-13 22:36:02 UTC FreeBSD src repository Modified files: share/examples/cvsup cvs-supfile Log: Add projects-all collection to cvs-supfile PR: misc/121680 Submitted by: "Philip M. Gollucci" Approved by: grog@ (mentor) MFC after: 1 week Revision Changes Path 1.48 +6 -0 src/share/examples/cvsup/cvs-supfile From scf at FreeBSD.org Thu Mar 13 23:21:27 2008 From: scf at FreeBSD.org (Sean C. Farley) Date: Thu Mar 13 23:21:29 2008 Subject: cvs commit: src/tools/tools/editing freebsd.vim In-Reply-To: <200803131852.m2DIqkw3095857@repoman.freebsd.org> References: <200803131852.m2DIqkw3095857@repoman.freebsd.org> Message-ID: On Thu, 13 Mar 2008, David E. O'Brien wrote: > obrien 2008-03-13 18:52:46 UTC > > FreeBSD src repository > > Added files: > tools/tools/editing freebsd.vim > Log: > Vim function to style(9). > Issue ':call FreeBSD_Style()' while editing a file. > > Submitted by: scf > > Revision Changes Path > 1.1 +68 -0 src/tools/tools/editing/freebsd.vim (new) Should this file (and the Emacs file? Nah! :)) be documented somewhere (maybe the committer's guide or style(9)) to make it easier to find? Sean -- scf@FreeBSD.org From sgk at troutmask.apl.washington.edu Thu Mar 13 23:51:17 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Thu Mar 13 23:51:22 2008 Subject: cvs commit: src/tools/tools/editing freebsd.vim In-Reply-To: References: <200803131852.m2DIqkw3095857@repoman.freebsd.org> Message-ID: <20080313235028.GA84074@troutmask.apl.washington.edu> On Thu, Mar 13, 2008 at 06:02:44PM -0500, Sean C. Farley wrote: > On Thu, 13 Mar 2008, David E. O'Brien wrote: > > >obrien 2008-03-13 18:52:46 UTC > > > > FreeBSD src repository > > > > Added files: > > tools/tools/editing freebsd.vim > > Log: > > Vim function to style(9). > > Issue ':call FreeBSD_Style()' while editing a file. > > > > Submitted by: scf > > > > Revision Changes Path > > 1.1 +68 -0 src/tools/tools/editing/freebsd.vim (new) > > Should this file (and the Emacs file? Nah! :)) be documented somewhere > (maybe the committer's guide or style(9)) to make it easier to find? > IMHO. It does not belong in style(9) unless you are prepared to add this info for all other editors that may eventual have a similar macro. It should also be noted that the items in src/tools/\ tools/editing are not installed by a "make installworld", so if you don't have /usr/src, you don't have these available. I vote committers or developer's guide. -- Steve From obrien at FreeBSD.org Fri Mar 14 00:32:06 2008 From: obrien at FreeBSD.org (David O'Brien) Date: Fri Mar 14 00:32:17 2008 Subject: cvs commit: src/tools/tools/editing freebsd.vim In-Reply-To: References: <200803131852.m2DIqkw3095857@repoman.freebsd.org> Message-ID: <20080314003205.GA94057@dragon.NUXI.org> On Thu, Mar 13, 2008 at 06:02:44PM -0500, Sean C. Farley wrote: >> Vim function to style(9). >> Issue ':call FreeBSD_Style()' while editing a file. > > Should this file (and the Emacs file? Nah! :)) be documented somewhere > (maybe the committer's guide or style(9)) to make it easier to find? Sounds like a good idea - but I thin maybe something wider than committers guide. Maybe the Developer's Handbook - given the dot files are public and available to all Handbook readers. -- -- David (obrien@FreeBSD.org) From cokane at FreeBSD.org Fri Mar 14 00:56:49 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Fri Mar 14 00:56:54 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_softdep.c In-Reply-To: <20080313215520.X83983@fledge.watson.org> References: <200803132015.m2DKFndQ008412@repoman.freebsd.org> <20080313215520.X83983@fledge.watson.org> Message-ID: <47D9CC76.9020000@FreeBSD.org> Robert Watson wrote: > > On Thu, 13 Mar 2008, Coleman Kane wrote: > >> cokane 2008-03-13 20:15:49 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/ufs/ffs ffs_softdep.c >> Log: >> Replace the non-MPSAFE timeout(9) API in ffs_softdep.c with the MPSAFE >> callout_* API (e.g. callout_init_mtx(9)). This was one of the numerous >> items on the http://wiki.freebsd.org/SMPTODO list. >> >> Reviewed by: imp, obrien, jhb >> MFC after: 1 week > > I might give this one a month to MFC rather than a week -- it can take > a while for bugs in this sort of thing to get diagnosed, especially > with lots of other changes in the tree at the same time relating to > schedulers, etc. > > Robert N M Watson > Computer Laboratory > University of Cambridge > Thanks, that's a good idea. I'll hold onto it for a month before it goes into RELENG_7. -- Coleman From gibbs at FreeBSD.org Fri Mar 14 02:57:47 2008 From: gibbs at FreeBSD.org (Justin T. Gibbs) Date: Fri Mar 14 02:57:51 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803140257.m2E2vkiu075095@repoman.freebsd.org> gibbs 2008-03-14 02:57:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/i386/cpufreq est.c Log: MFC: In est_acpi_info(), initialize count before passing its pointer to CPUFREQ_DRV_SETTINGS(). Revision Changes Path 1.11.2.2 +1 -0 src/sys/i386/cpufreq/est.c From jhb at FreeBSD.org Fri Mar 14 03:44:44 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Mar 14 03:44:48 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/i386/i386 intr_machdep.c Message-ID: <200803140344.m2E3ihbA078122@repoman.freebsd.org> jhb 2008-03-14 03:44:42 UTC FreeBSD src repository Modified files: sys/amd64/amd64 intr_machdep.c sys/i386/i386 intr_machdep.c Log: Fix a silly bogon which prevented all the CPUs that are tagged as interrupt receivers from being given interrupts if any CPUs in the system were not tagged as interrupt receivers that I introduced when switching the x86 interrupt code to track CPUs via FreeBSD CPU IDs rather than local APIC IDs. In practice this only affects systems with Hyperthreading (though disabling HTT in the BIOS would workaround the issue) as that is the only case currently where one can have CPUs that aren't tagged as interrupt receivers. On a Dell SC1425 test box with 2 x Xeon w/ HTT (so 4 logical CPUs of which 2 were interrupt receivers) the result was that all device interrupts were sent to CPU 0. MFC after: 1 week Pointy hat to: jhb Revision Changes Path 1.36 +4 -4 src/sys/amd64/amd64/intr_machdep.c 1.31 +4 -4 src/sys/i386/i386/intr_machdep.c From jkoshy at FreeBSD.org Fri Mar 14 06:16:19 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Fri Mar 14 06:16:21 2008 Subject: cvs commit: src/sys/dev/hwpmc pmc_events.h Message-ID: <200803140616.m2E6GIHw099522@repoman.freebsd.org> jkoshy 2008-03-14 06:16:18 UTC FreeBSD src repository Modified files: sys/dev/hwpmc pmc_events.h Log: Correct a typo. Revision Changes Path 1.2 +6 -2 src/sys/dev/hwpmc/pmc_events.h From jkoshy at FreeBSD.org Fri Mar 14 06:22:07 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Fri Mar 14 06:22:09 2008 Subject: cvs commit: src/lib/libpmc pmc.3 Message-ID: <200803140622.m2E6M6iG099727@repoman.freebsd.org> jkoshy 2008-03-14 06:22:03 UTC FreeBSD src repository Modified files: lib/libpmc pmc.3 Log: - Document Pentium and Pentium MMX events. - Update (c) years and the manual page's date. Revision Changes Path 1.17 +338 -2 src/lib/libpmc/pmc.3 From nyan at FreeBSD.org Fri Mar 14 09:00:05 2008 From: nyan at FreeBSD.org (Takahashi Yoshihiro) Date: Fri Mar 14 09:00:06 2008 Subject: cvs commit: src/sys/pc98/include nexusvar.h Message-ID: <200803140900.m2E904W1017484@repoman.freebsd.org> nyan 2008-03-14 09:00:04 UTC FreeBSD src repository Added files: sys/pc98/include nexusvar.h Log: Add stub for pc98. Revision Changes Path 1.1 +6 -0 src/sys/pc98/include/nexusvar.h (new) From piso at FreeBSD.org Fri Mar 14 09:41:47 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Fri Mar 14 09:41:50 2008 Subject: cvs commit: src/sbin/ipfw ipfw2.c Message-ID: <200803140941.m2E9fktM020563@repoman.freebsd.org> piso 2008-03-14 09:41:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/ipfw ipfw2.c Log: Fix showing nat rules. Bug spotted by: Gael Roualland PR: bin/121683 Revision Changes Path 1.108.2.4 +1 -0 src/sbin/ipfw/ipfw2.c From bz at FreeBSD.org Fri Mar 14 11:09:12 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Fri Mar 14 11:09:17 2008 Subject: cvs commit: src/sys/netinet6 ip6_output.c Message-ID: <200803141109.m2EB9BiB037870@repoman.freebsd.org> bz 2008-03-14 11:09:11 UTC FreeBSD src repository Modified files: sys/netinet6 ip6_output.c Log: Replace the function name in two identical printfs by __func__, __LINE__ so we can distinguish them when people report a problem. PR: 121373 MFC after: 5 days Revision Changes Path 1.115 +4 -2 src/sys/netinet6/ip6_output.c From bz at FreeBSD.org Fri Mar 14 11:44:30 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Fri Mar 14 11:44:40 2008 Subject: cvs commit: src/sys/netinet6 ip6_ipsec.c ip6_ipsec.h Message-ID: <200803141144.m2EBiU6O042092@repoman.freebsd.org> bz 2008-03-14 11:44:30 UTC FreeBSD src repository Modified files: sys/netinet6 ip6_ipsec.c ip6_ipsec.h Log: #if 0 out a currently unsued (and incomplete) function: ip6_ipsec_mtu(). No need to compile 'dead' code. I am leaving it in because we will have to review the concept and should use the common function in various places. MFC after: 5 days Revision Changes Path 1.8 +5 -5 src/sys/netinet6/ip6_ipsec.c 1.4 +3 -1 src/sys/netinet6/ip6_ipsec.h From bz at FreeBSD.org Fri Mar 14 11:55:06 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Fri Mar 14 11:55:11 2008 Subject: cvs commit: src/sys/netinet6 ip6_ipsec.c ip6_output.c Message-ID: <200803141155.m2EBt5aB042527@repoman.freebsd.org> bz 2008-03-14 11:55:05 UTC FreeBSD src repository Modified files: sys/netinet6 ip6_ipsec.c ip6_output.c Log: Correct reference counting on the SP for outgoing IPv6 IPsec connections. PR: 121374 Reported by: Cyrus Rahman (crahman gmail.com) Tested by: Cyrus Rahman (crahman gmail.com) MFC after: 5 days Revision Changes Path 1.9 +1 -7 src/sys/netinet6/ip6_ipsec.c 1.116 +4 -0 src/sys/netinet6/ip6_output.c From wearabnet at yahoo.ca Fri Mar 14 12:45:02 2008 From: wearabnet at yahoo.ca (Abdullah Ibn Hamad Al-Marri) Date: Fri Mar 14 12:45:05 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <565127.65545.qm@web33705.mail.mud.yahoo.com> ----- Original Message ---- > From: John Baldwin > To: src-committers@FreeBSD.org; cvs-src@FreeBSD.org; cvs-all@FreeBSD.org > Sent: Tuesday, March 11, 2008 1:00:36 AM > Subject: cvs commit: src/sys/i386/cpufreq est.c > > jhb 2008-03-10 22:00:36 UTC > > FreeBSD src repository > > Modified files: > sys/i386/cpufreq est.c > Log: > - Don't execute cpuid to fetch the features. We already have the features > present in cpu_feature2. Also, use CPUID2_EST rather than a magic > number. > - Don't free the ACPI settings list in detach if we are going to fail the > request. Otherwise an attempt to kldunload est would free the array > but the driver would keep trying to use it. > > MFC after: 1 week > > Revision Changes Path > 1.14 +5 -5 src/sys/i386/cpufreq/est.c Hello John Could you please MFC it to RELENG_7 now? :) -- Regards, -Abdullah Ibn Hamad Al-Marri Arab Portal http://www.WeArab.Net/ ____________________________________________________________________________________ Looking for last minute shopping deals? Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping From rwatson at FreeBSD.org Fri Mar 14 13:13:39 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 14 13:13:44 2008 Subject: cvs commit: src/share/man/man9 Makefile stack.9 Message-ID: <200803141313.m2EDDc1l058638@repoman.freebsd.org> rwatson 2008-03-14 13:13:37 UTC FreeBSD src repository Modified files: share/man/man9 Makefile stack.9 Log: Add stack_print_ddb() to the synposis of stack(9), add mlink for it; fix stack_print(9) mlink. Spotted by: attilio Revision Changes Path 1.332 +2 -1 src/share/man/man9/Makefile 1.6 +2 -0 src/share/man/man9/stack.9 From jhb at FreeBSD.org Fri Mar 14 15:22:39 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Mar 14 15:22:43 2008 Subject: cvs commit: src/sys/kern sched_ule.c Message-ID: <200803141522.m2EFMcjj070233@repoman.freebsd.org> jhb 2008-03-14 15:22:38 UTC FreeBSD src repository Modified files: sys/kern sched_ule.c Log: Make the function prototype for cpu_search() match the declaration so that this still compiles with gcc3. Revision Changes Path 1.233 +2 -2 src/sys/kern/sched_ule.c From emaste at FreeBSD.org Fri Mar 14 15:25:44 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Fri Mar 14 15:25:50 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_softdep.c Message-ID: <200803141525.m2EFPioL070382@repoman.freebsd.org> emaste 2008-03-14 15:25:44 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/ufs/ffs ffs_softdep.c Log: Honour mpsafe_vfs=0 in softdep_flush. This is basically a MFC of revision 1.206, except that Giant remains unconditionally acquired in the #ifdef QUOTA case here (as QUOTA-enabled UFS on RELENG_6 is not MPSAFE). Reviewed by: kib Revision Changes Path 1.181.2.24 +4 -0 src/sys/ufs/ffs/ffs_softdep.c From jkim at FreeBSD.org Fri Mar 14 15:59:33 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Fri Mar 14 15:59:38 2008 Subject: cvs commit: src/sys/dev/usb usb_quirks.c usbdevs Message-ID: <200803141559.m2EFxWs1074538@repoman.freebsd.org> jkim 2008-03-14 15:59:30 UTC FreeBSD src repository Modified files: sys/dev/usb usb_quirks.c usbdevs Log: Add a quirk to ignore ASUS LCM display found on some ASUS laptops. Revision Changes Path 1.65 +2 -0 src/sys/dev/usb/usb_quirks.c 1.342 +1 -0 src/sys/dev/usb/usbdevs From emax at FreeBSD.org Fri Mar 14 16:07:10 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Fri Mar 14 16:07:22 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c Message-ID: <200803141607.m2EG7ApX076463@repoman.freebsd.org> emax 2008-03-14 16:07:10 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c Log: MFC: Add an option to register DUN (Dial-Up Networking) service on the same RFCOMM channel if needed. There is really no good reason to not to support this. AT-command exchange can be faked with chat script in ppp.conf. Revision Changes Path 1.14.2.1 +10 -1 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.8 1.6.2.1 +33 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c From emax at FreeBSD.org Fri Mar 14 16:11:21 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Fri Mar 14 16:11:31 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c Message-ID: <200803141611.m2EGBLwF076696@repoman.freebsd.org> emax 2008-03-14 16:11:20 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c Log: MFC: Add an option to register DUN (Dial-Up Networking) service on the same RFCOMM channel if needed. There is really no good reason to not to support this. AT-command exchange can be faked with chat script in ppp.conf. Revision Changes Path 1.10.2.5 +10 -1 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.8 1.5.8.2 +33 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c From bz at FreeBSD.org Fri Mar 14 16:25:40 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Fri Mar 14 16:25:42 2008 Subject: cvs commit: src/sys/netipsec key.c Message-ID: <200803141625.m2EGPesH077308@repoman.freebsd.org> bz 2008-03-14 16:25:40 UTC FreeBSD src repository Modified files: sys/netipsec key.c Log: Remove the "Fast " from the "Fast IPsec: Initialized Security Association Processing." printf. People kept asking questions about this after the IPsec shuffle. This still is the Fast IPsec implementation so no worries that it would be any slower now. There are no functional changes. Discussed with: sam MFC after: 4 days Revision Changes Path 1.30 +1 -1 src/sys/netipsec/key.c From emax at FreeBSD.org Fri Mar 14 16:29:26 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Fri Mar 14 16:29:38 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.c Message-ID: <200803141629.m2EGTQUi078232@repoman.freebsd.org> emax 2008-03-14 16:29:26 UTC FreeBSD src repository Modified files: usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.c Log: Update usage() - add new -D option. While I'm here, update copyright information and license formatting MFC after: 1 day Revision Changes Path 1.8 +5 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c From bz at FreeBSD.org Fri Mar 14 16:38:11 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Fri Mar 14 16:38:14 2008 Subject: cvs commit: src/sys/netinet ip_ipsec.c src/sys/netinet6 ip6_output.c src/sys/netipsec ipsec_output.c Message-ID: <200803141638.m2EGcB99078942@repoman.freebsd.org> bz 2008-03-14 16:38:11 UTC FreeBSD src repository Modified files: sys/netinet ip_ipsec.c sys/netinet6 ip6_output.c sys/netipsec ipsec_output.c Log: Correct IPsec behaviour with a 'use' level in SP but no SA available. In that case return an continue processing the packet without IPsec. PR: 121384 MFC after: 5 days Reported by: Cyrus Rahman (crahman gmail.com) Tested by: Cyrus Rahman (crahman gmail.com) [slightly older version] Revision Changes Path 1.9 +12 -0 src/sys/netinet/ip_ipsec.c 1.117 +16 -2 src/sys/netinet6/ip6_output.c 1.18 +33 -22 src/sys/netipsec/ipsec_output.c From rwatson at FreeBSD.org Fri Mar 14 17:12:41 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 14 17:12:45 2008 Subject: cvs commit: src/sys/fs/coda cnode.h coda_namecache.c coda_namecache.h coda_psdev.c coda_subr.c coda_subr.h coda_vnops.c coda_vnops.h src/sys/modules/coda Makefile Message-ID: <200803141712.m2EHCfoD006679@repoman.freebsd.org> rwatson 2008-03-14 17:12:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda cnode.h coda_psdev.c coda_subr.c coda_subr.h coda_vnops.c coda_vnops.h sys/modules/coda Makefile Removed files: (Branch: RELENG_7) sys/fs/coda coda_namecache.c coda_namecache.h Log: Merge cnode.h:1.26, coda_psdev.c:1.44, coda_subr.c:1.36, coda_subr.h:1.12, coda_vnops.c:1.94, coda_vnops.h:1.24, coda_namecache.c:1.26, coda_namecache.h:1.15, Makefile:1.18, from HEAD to RELENG_7: Rather than having the Coda module use its own namecache, use the global VFS namecache, as is done by the Coda module on Linux. Unlike the Coda namecache, the global VFS namecache isn't tagged by credential, so use ore conservative flushing behavior (for now) when CODA_PURGEUSER is issued by Venus. This improves overall integration with the FreeBSD VFS, including allowing __getcwd() to work better, procfs/procstat monitoring, and so on. This improves shell behavior in many cases, and improves ".." handling. It may lead to some slowdown until we've implemented a specific access cache, which should net improve performance, but in the mean time, lookup access control now always goes to Venus, whereas previously it didn't. Revision Changes Path 1.21.2.5 +2 -1 src/sys/fs/coda/cnode.h 1.23.2.3 +0 -700 src/sys/fs/coda/coda_namecache.c (dead) 1.11.2.3 +0 -205 src/sys/fs/coda/coda_namecache.h (dead) 1.39.2.5 +5 -20 src/sys/fs/coda/coda_psdev.c 1.33.2.3 +30 -17 src/sys/fs/coda/coda_subr.c 1.10.2.2 +1 -1 src/sys/fs/coda/coda_subr.h 1.76.2.10 +75 -127 src/sys/fs/coda/coda_vnops.c 1.19.2.5 +1 -1 src/sys/fs/coda/coda_vnops.h 1.17.2.1 +2 -3 src/sys/modules/coda/Makefile From rwatson at FreeBSD.org Fri Mar 14 17:14:42 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 14 17:14:49 2008 Subject: cvs commit: src/sys/conf files Message-ID: <200803141714.m2EHEgU3011504@repoman.freebsd.org> rwatson 2008-03-14 17:14:42 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf files Log: Merge files:1.1268 from HEAD to RELENG_7: Remove coda_namecache from "options vcoda", it is no longer required. Revision Changes Path 1.1243.2.5 +0 -1 src/sys/conf/files From rwatson at FreeBSD.org Fri Mar 14 17:15:12 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 14 17:15:22 2008 Subject: cvs commit: src/sys/modules/coda5 Makefile Message-ID: <200803141715.m2EHFCwX011579@repoman.freebsd.org> rwatson 2008-03-14 17:15:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/modules/coda5 Makefile Log: Merge Makefile:1.9 from HEAD to RELENG_7: Remove coda_namecache from coda5 as well. We should probably GC coda5 entirely at this point as coda6 is considered the supported branch. Revision Changes Path 1.8.2.1 +1 -1 src/sys/modules/coda5/Makefile From rwatson at FreeBSD.org Fri Mar 14 17:17:01 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 14 17:17:02 2008 Subject: cvs commit: src/sys/fs/coda cnode.h coda_subr.c coda_vnops.c Message-ID: <200803141717.m2EHH1QA011642@repoman.freebsd.org> rwatson 2008-03-14 17:17:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda cnode.h coda_subr.c coda_vnops.c Log: Merge cnode.h:1.27, coda_subr.c:1.37, coda_vnops.c:1.95 from HEAD to RELENG_7: Implement a rudimentary access cache for the Coda kernel module, modeled on the access cache found in NFS, smbfs, and the Linux coda module. This is a positive access cache of a single entry per file, tracking recently granted rights, but unlike NFS and smbfs, supporting explicit invalidation by the distributed file system. For each cnode, maintain a C_ACCCACHE flag indicating the validity of the cache, and a cached uid and mode tracking recently granted positive access control decisions. Prefer the cache to venus_access() in VOP_ACCESS() if it is valid, and when we must fall back to venus_access(), update the cache. Allow Venus to clear the access cache, either the whole cache on CODA_FLUSH, or just entries for a specific uid on CODA_PURGEUSER. Unlike the Coda module on Linux, we don't flush all entries on a user purge using a generation number, we instead walk present cnodes and clear only entries for the specific user, meaning it is somewhat more expensive but won't hit all users. Since the Coda module is agressive about not keeping around unopened cnodes, the utility of the cache is somewhat limited for files, but works will for directories. We should make Coda less agressive about GCing cnodes in VOP_INACTIVE() in order to improve the effectiveness of in-kernel caching of attributes and access rights. Revision Changes Path 1.21.2.6 +4 -0 src/sys/fs/coda/cnode.h 1.33.2.4 +63 -8 src/sys/fs/coda/coda_subr.c 1.76.2.11 +51 -21 src/sys/fs/coda/coda_vnops.c From rwatson at FreeBSD.org Fri Mar 14 17:17:49 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Mar 14 17:17:56 2008 Subject: cvs commit: src/sys/fs/coda coda_subr.c Message-ID: <200803141717.m2EHHnZS011696@repoman.freebsd.org> rwatson 2008-03-14 17:17:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_subr.c Log: Merge coda_subr.c:1.38 from HEAD to RELENG_7: Update cache flushing behavior in light of recent namecache and access cache improvements: - Flush just access control state on CODA_PURGEUSER, not the full namecache for /coda. - When replacing a fid on a cnode as a result of, e.g., reintegration after offline operation, we no longer need to purge the namecache entries associated with its vnode. Revision Changes Path 1.33.2.5 +0 -7 src/sys/fs/coda/coda_subr.c From jhb at freebsd.org Fri Mar 14 18:32:15 2008 From: jhb at freebsd.org (John Baldwin) Date: Fri Mar 14 18:32:18 2008 Subject: cvs commit: src/sys/pc98/include nexusvar.h In-Reply-To: <200803140900.m2E904W1017484@repoman.freebsd.org> References: <200803140900.m2E904W1017484@repoman.freebsd.org> Message-ID: <200803141110.10748.jhb@freebsd.org> On Friday 14 March 2008 05:00:04 am Takahashi Yoshihiro wrote: > nyan 2008-03-14 09:00:04 UTC > > FreeBSD src repository > > Added files: > sys/pc98/include nexusvar.h > Log: > Add stub for pc98. Sorry. :( -- John Baldwin From jhb at FreeBSD.org Fri Mar 14 19:41:48 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Mar 14 19:41:50 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... Message-ID: <200803141941.m2EJfmL2020579@repoman.freebsd.org> jhb 2008-03-14 19:41:48 UTC FreeBSD src repository Modified files: sys/amd64/amd64 intr_machdep.c sys/amd64/include intr_machdep.h sys/arm/arm intr.c sys/i386/i386 intr_machdep.c sys/i386/include intr_machdep.h sys/ia64/ia64 interrupt.c sys/kern kern_intr.c sys/powerpc/powerpc intr_machdep.c sys/sparc64/sparc64 intr_machdep.c sys/sun4v/sun4v intr_machdep.c sys/sys interrupt.h Log: Add preliminary support for binding interrupts to CPUs: - Add a new intr_event method ie_assign_cpu() that is invoked when the MI code wishes to bind an interrupt source to an individual CPU. The MD code may reject the binding with an error. If an assign_cpu function is not provided, then the kernel assumes the platform does not support binding interrupts to CPUs and fails all requests to do so. - Bind ithreads to CPUs on their next execution loop once an interrupt event is bound to a CPU. Only shared ithreads are bound. We currently leave private ithreads for drivers using filters + ithreads in the INTR_FILTER case unbound. - A new intr_event_bind() routine is used to bind an interrupt event to a CPU. - Implement binding on amd64 and i386 by way of the existing pic_assign_cpu PIC method. - For x86, provide a 'intr_bind(IRQ, cpu)' wrapper routine that looks up an interrupt source and binds its interrupt event to the specified CPU. MI code can currently (ab)use this by doing: intr_bind(rman_get_start(irq_res), cpu); however, I plan to add a truly MI interface (probably a bus_bind_intr(9)) where the implementation in the x86 nexus(4) driver would end up calling intr_bind() internally. Requested by: kmacy, gallatin, jeff Tested on: {amd64, i386} x {regular, INTR_FILTER} Revision Changes Path 1.37 +52 -9 src/sys/amd64/amd64/intr_machdep.c 1.19 +3 -0 src/sys/amd64/include/intr_machdep.h 1.18 +2 -2 src/sys/arm/arm/intr.c 1.32 +52 -9 src/sys/i386/i386/intr_machdep.c 1.21 +3 -0 src/sys/i386/include/intr_machdep.h 1.63 +1 -1 src/sys/ia64/ia64/interrupt.c 1.155 +92 -4 src/sys/kern/kern_intr.c 1.18 +1 -1 src/sys/powerpc/powerpc/intr_machdep.c 1.29 +2 -2 src/sys/sparc64/sparc64/intr_machdep.c 1.8 +1 -1 src/sys/sun4v/sun4v/intr_machdep.c 1.38 +9 -4 src/sys/sys/interrupt.h From kientzle at FreeBSD.org Fri Mar 14 20:32:21 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 20:32:28 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_tar.c Message-ID: <200803142032.m2EKWLOt023447@repoman.freebsd.org> kientzle 2008-03-14 20:32:21 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_tar.c Log: A block in a tar file is 512 bytes. Period. Remove the entirely pointless symbolic constant and sizeof(unsigned char). (The constant here is doubly wrong, since not only does it obscure a basic format constant, it was never intended to be a tar-specific value, so could conceivably be changed at some point in the future.) Revision Changes Path 1.66 +1 -1 src/lib/libarchive/archive_read_support_format_tar.c From kientzle at FreeBSD.org Fri Mar 14 20:35:38 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 20:35:43 2008 Subject: cvs commit: src/lib/libarchive Makefile archive.h.in archive_write_set_compression_compress.c src/lib/libarchive/test Makefile test_write_compress.c Message-ID: <200803142035.m2EKZc4f023670@repoman.freebsd.org> kientzle 2008-03-14 20:35:38 UTC FreeBSD src repository Modified files: lib/libarchive Makefile archive.h.in lib/libarchive/test Makefile Added files: lib/libarchive archive_write_set_compression_compress.c lib/libarchive/test test_write_compress.c Log: Support for writing 'compress' format, thanks to Joerg Sonnenberger. Revision Changes Path 1.81 +1 -0 src/lib/libarchive/Makefile 1.48 +1 -0 src/lib/libarchive/archive.h.in 1.1 +494 -0 src/lib/libarchive/archive_write_set_compression_compress.c (new) 1.16 +1 -0 src/lib/libarchive/test/Makefile 1.1 +99 -0 src/lib/libarchive/test/test_write_compress.c (new) From emaste at FreeBSD.org Fri Mar 14 21:59:12 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Fri Mar 14 21:59:14 2008 Subject: cvs commit: src/sys/dev/aac aac.c Message-ID: <200803142159.m2ELxBnj037256@repoman.freebsd.org> emaste 2008-03-14 21:59:11 UTC FreeBSD src repository Modified files: sys/dev/aac aac.c Log: Change spelling and eliminate a typo in comments to reduce diffs with Adaptec's vendor driver. I have some fixes to bring in and this makes ongoing review of the FreeBSD-Adaptec driver diffs easier. Revision Changes Path 1.128 +11 -11 src/sys/dev/aac/aac.c From kientzle at FreeBSD.org Fri Mar 14 22:00:09 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 22:00:10 2008 Subject: cvs commit: src/lib/libarchive archive_string.h archive_string_sprintf.c Message-ID: <200803142200.m2EM09NU037421@repoman.freebsd.org> kientzle 2008-03-14 22:00:09 UTC FreeBSD src repository Modified files: lib/libarchive archive_string.h archive_string_sprintf.c Log: Add a useful sprintf()-style wrapper around archive_string_vsprintf(). (Which is built on top of libarchive's internal resizable string support.) Revision Changes Path 1.10 +3 -0 src/lib/libarchive/archive_string.h 1.10 +10 -0 src/lib/libarchive/archive_string_sprintf.c From kientzle at FreeBSD.org Fri Mar 14 22:19:51 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 22:20:10 2008 Subject: cvs commit: src/lib/libarchive archive.h.in Message-ID: <200803142219.m2EMJoE3039219@repoman.freebsd.org> kientzle 2008-03-14 22:19:50 UTC FreeBSD src repository Modified files: lib/libarchive archive.h.in Log: Rework the versioning information, hopefully for the last time. * There are now only two public version identifiers: "number" is a single integer that combines Major/minor/release in a single value of the form Mmmmrrr. This is easy to compare against for checking feature support. "string" is a displayable text string of the form "libarchive M.mm.rr". * The number is present both as a macro (version of the installed header) and a function (version of the shared library). The string form is available only as a function. * Retain the older version definitions for now, but mark them all as deprecated, to disappear in libarchive 3.0 (whenever that happens). * Rework the various deprecation conditionals to use ARCHIVE_VERSION_NUMBER. An ancillary goal is to reduce the number of @...@ substitutions that are required. Someday, I might even be able to avoid build-time processing of archive.h entirely. Revision Changes Path 1.49 +37 -41 src/lib/libarchive/archive.h.in From kientzle at FreeBSD.org Fri Mar 14 22:31:58 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 22:32:03 2008 Subject: cvs commit: src/lib/libarchive archive_util.c src/lib/libarchive/test test_archive_api_feature.c Message-ID: <200803142231.m2EMVvuP042059@repoman.freebsd.org> kientzle 2008-03-14 22:31:57 UTC FreeBSD src repository Modified files: lib/libarchive archive_util.c lib/libarchive/test test_archive_api_feature.c Log: Rework the versioning implementation and test to match the new interface. Mark the functions that are going away in libarchive 3.0. In particular, archive_version_string() now computes the string rather than assuming that it will be created by the build infrastructure. Eventually, this will allow some simplification of the build infrastructure. Revision Changes Path 1.17 +44 -2 src/lib/libarchive/archive_util.c 1.4 +14 -0 src/lib/libarchive/test/test_archive_api_feature.c From kientzle at FreeBSD.org Fri Mar 14 22:40:36 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 22:40:52 2008 Subject: cvs commit: src/lib/libarchive archive_entry.c Message-ID: <200803142240.m2EMeaA1042823@repoman.freebsd.org> kientzle 2008-03-14 22:40:36 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry.c Log: Let archive_entry_clear() accept a NULL pointer and simply do nothing. In particular, this allows archive_entry_free() to work correctly for a NULL pointer, which makes it parallel with free(3). Revision Changes Path 1.48 +2 -0 src/lib/libarchive/archive_entry.c From kientzle at FreeBSD.org Fri Mar 14 22:44:07 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 22:44:09 2008 Subject: cvs commit: src/lib/libarchive archive_entry.c Message-ID: <200803142244.m2EMi71N043021@repoman.freebsd.org> kientzle 2008-03-14 22:44:07 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry.c Log: Since "length" computes the length of a string and is used as an argument to malloc(3), it should be size_t, not int. Revision Changes Path 1.49 +1 -1 src/lib/libarchive/archive_entry.c From kientzle at FreeBSD.org Fri Mar 14 22:47:39 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 22:47:44 2008 Subject: cvs commit: src/lib/libarchive archive_entry.h Message-ID: <200803142247.m2EMldaA043266@repoman.freebsd.org> kientzle 2008-03-14 22:47:39 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry.h Log: Update some comments, comment out argument names to guard against namespace problems. Revision Changes Path 1.25 +50 -16 src/lib/libarchive/archive_entry.h From kientzle at FreeBSD.org Fri Mar 14 23:00:54 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 23:00:58 2008 Subject: cvs commit: src/lib/libarchive Makefile archive_entry.3 archive_entry.c archive_entry.h src/lib/libarchive/test test_entry.c Message-ID: <200803142300.m2EN0sEF043849@repoman.freebsd.org> kientzle 2008-03-14 23:00:53 UTC FreeBSD src repository Modified files: lib/libarchive Makefile archive_entry.3 archive_entry.c archive_entry.h lib/libarchive/test test_entry.c Log: New public functions archive_entry_copy_link() and archive_entry_copy_link_w() override the currently set link value, whether that's a hardlink or a symlink. Plus documentation update and tests. Revision Changes Path 1.82 +2 -0 src/lib/libarchive/Makefile 1.17 +2 -0 src/lib/libarchive/archive_entry.3 1.50 +22 -0 src/lib/libarchive/archive_entry.c 1.26 +2 -0 src/lib/libarchive/archive_entry.h 1.4 +30 -1 src/lib/libarchive/test/test_entry.c From kientzle at FreeBSD.org Fri Mar 14 23:09:03 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 23:09:07 2008 Subject: cvs commit: src/lib/libarchive archive_write.c Message-ID: <200803142309.m2EN92Mg045381@repoman.freebsd.org> kientzle 2008-03-14 23:09:02 UTC FreeBSD src repository Modified files: lib/libarchive archive_write.c Log: Don't advertise the default block size as a constant; don't rely on a deprecated value to set the default. This is also related to a longer-term goal of setting the default block size based on format and possibly other factors, which makes it a bad idea to tie this to a published constant. Revision Changes Path 1.27 +6 -1 src/lib/libarchive/archive_write.c From kientzle at FreeBSD.org Fri Mar 14 23:19:47 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Fri Mar 14 23:19:49 2008 Subject: cvs commit: src/lib/libarchive archive_entry.c src/lib/libarchive/test test_entry.c Message-ID: <200803142319.m2ENJkdX045820@repoman.freebsd.org> kientzle 2008-03-14 23:19:46 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry.c lib/libarchive/test test_entry.c Log: Don't lie. If a string can't be converted to a wide (Unicode) string, return a NULL instead of an incomplete string. Expand the test coverage to verify the correct behavior here. Revision Changes Path 1.51 +8 -1 src/lib/libarchive/archive_entry.c 1.5 +35 -0 src/lib/libarchive/test/test_entry.c From kientzle at FreeBSD.org Sat Mar 15 00:52:24 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 00:52:47 2008 Subject: cvs commit: src/lib/libarchive/test .cvsignore Message-ID: <200803150052.m2F0qNOt059615@repoman.freebsd.org> kientzle 2008-03-15 00:52:23 UTC FreeBSD src repository Added files: lib/libarchive/test .cvsignore Log: Ignore some built files. Revision Changes Path 1.1 +4 -0 src/lib/libarchive/test/.cvsignore (new) From julian at elischer.org Sat Mar 15 00:56:43 2008 From: julian at elischer.org (Julian Elischer) Date: Sat Mar 15 00:56:55 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c In-Reply-To: <200803141611.m2EGBLwF076696@repoman.freebsd.org> References: <200803141611.m2EGBLwF076696@repoman.freebsd.org> Message-ID: <47DB1EC8.2040100@elischer.org> Maksim Yevmenkin wrote: > emax 2008-03-14 16:11:20 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_6) > usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c > Log: > MFC: > > Add an option to register DUN (Dial-Up Networking) service on the same > RFCOMM channel if needed. There is really no good reason to not to support > this. AT-command exchange can be faked with chat script in ppp.conf. > > Revision Changes Path > 1.10.2.5 +10 -1 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.8 > 1.5.8.2 +33 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c Hi max.. we were having a beer induced discussion here.. and I thought you might know about the answer to one question that came up as ajoke.. how many barriers are there from stopping a computer from doing "drive-by" downloads and connections from random passing cell phones.. obviously they need to associate, but how many layers of security are there? From jkim at FreeBSD.org Sat Mar 15 01:09:48 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Sat Mar 15 01:09:54 2008 Subject: cvs commit: src/usr.sbin/mixer mixer.c Message-ID: <200803150109.m2F19lsV061536@repoman.freebsd.org> jkim 2008-03-15 01:09:47 UTC FreeBSD src repository Modified files: usr.sbin/mixer mixer.c Log: - Check device masks and bypass printing headers if capable device does not exist. /etc/rc.d/mixer tried to restore non-existent recording source because /var/var/db/mixer*-state contains '=rec' for example. - Remove hardcoded mixer2 and mixer3 and do the right thing. - Replace getopt(3) with hand-rolled option parser. It was not able to handle 'mixer -rec mic' although it is a valid command. - Make '-s' and '-S' mutualy exclusive as mixer(8) says. - Do not re-read recording source unless it has been modified. Revision Changes Path 1.24 +72 -47 src/usr.sbin/mixer/mixer.c From jkim at FreeBSD.org Sat Mar 15 01:15:16 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Sat Mar 15 01:15:21 2008 Subject: cvs commit: src/usr.sbin/mixer mixer.8 Message-ID: <200803150115.m2F1FGBv061762@repoman.freebsd.org> jkim 2008-03-15 01:15:16 UTC FreeBSD src repository Modified files: usr.sbin/mixer mixer.8 Log: Update list of possible mixer devices. These devices were added to soundcard.h more than ten years ago. Revision Changes Path 1.26 +3 -2 src/usr.sbin/mixer/mixer.8 From jkim at FreeBSD.org Sat Mar 15 01:20:06 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Sat Mar 15 01:20:08 2008 Subject: cvs commit: src/usr.sbin/mixer mixer.c Message-ID: <200803150120.m2F1K57Z061916@repoman.freebsd.org> jkim 2008-03-15 01:20:05 UTC FreeBSD src repository Modified files: usr.sbin/mixer mixer.c Log: Clean up code and fix style(9) nits. Revision Changes Path 1.25 +101 -99 src/usr.sbin/mixer/mixer.c From jkim at FreeBSD.org Sat Mar 15 01:36:21 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Sat Mar 15 01:36:23 2008 Subject: cvs commit: src/usr.sbin/mixer mixer.c Message-ID: <200803150136.m2F1aLDa062449@repoman.freebsd.org> jkim 2008-03-15 01:36:20 UTC FreeBSD src repository Modified files: usr.sbin/mixer mixer.c Log: Fix a typo. Revision Changes Path 1.26 +1 -1 src/usr.sbin/mixer/mixer.c From kientzle at FreeBSD.org Sat Mar 15 01:43:59 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 01:44:05 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_tar.c archive_write_set_format_pax.c src/lib/libarchive/test Makefile test_pax_filename_encoding.c test_pax_filename_encoding.tar.gz.uu Message-ID: <200803150143.m2F1hxt7062719@repoman.freebsd.org> kientzle 2008-03-15 01:43:59 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_tar.c archive_write_set_format_pax.c lib/libarchive/test Makefile Added files: lib/libarchive/test test_pax_filename_encoding.c test_pax_filename_encoding.tar.gz.uu Log: A subtle point: "pax interchange format" mandates that all strings (including pathname, gname, uname) be stored in UTF-8. This usually doesn't cause problems on FreeBSD because the "C" locale on FreeBSD can convert any byte to Unicode/wchar_t and from there to UTF-8. In other locales (including the "C" locale on Linux which is really ASCII), you can get into trouble with pathnames that cannot be converted to UTF-8. Libarchive's pax writer truncated pathnames and other strings at the first nonconvertible character. (ouch!) Other archivers have worked around this by storing unconvertible pathnames as raw binary, a practice which has been sanctioned by the Austin group. However, libarchive's pax reader would segfault reading headers that weren't proper UTF-8. (ouch!) Since bsdtar defaults to pax format, this affects bsdtar rather heavily. To correctly support the new "hdrcharset" header that is going into SUS and to handle conversion failures in general, libarchive's pax reader and writer have been overhauled fairly extensively. They used to do most of the pax header processing using wchar_t (Unicode); they now do most of it using char so that common logic applies to either UTF-8 or "binary" strings. As a bonus, a number of extraneous conversions to/from wchar_t have been eliminated, which should speed things up just a tad. Thanks to: Bjoern Jacke for originally reporting this to me Thanks to: Joerg Sonnenberger for noting a bad typo in my first draft of this Thanks to: Gunnar Ritter for getting the standard fixed MFC after: 5 days Revision Changes Path 1.67 +240 -209 src/lib/libarchive/archive_read_support_format_tar.c 1.43 +126 -50 src/lib/libarchive/archive_write_set_format_pax.c 1.17 +1 -0 src/lib/libarchive/test/Makefile 1.1 +161 -0 src/lib/libarchive/test/test_pax_filename_encoding.c (new) 1.1 +10 -0 src/lib/libarchive/test/test_pax_filename_encoding.tar.gz.uu (new) From kientzle at FreeBSD.org Sat Mar 15 02:22:09 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 02:22:11 2008 Subject: cvs commit: src/lib/libarchive/test Makefile Message-ID: <200803150222.m2F2M8tI065661@repoman.freebsd.org> kientzle 2008-03-15 02:22:08 UTC FreeBSD src repository Modified files: lib/libarchive/test Makefile Log: GC a reference to the defunct TESTFILES variable. Revision Changes Path 1.18 +1 -1 src/lib/libarchive/test/Makefile From kientzle at FreeBSD.org Sat Mar 15 02:30:43 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 02:30:47 2008 Subject: cvs commit: src/lib/libarchive archive_write_set_format_pax.c config_freebsd.h Message-ID: <200803150230.m2F2Uhab065955@repoman.freebsd.org> kientzle 2008-03-15 02:30:42 UTC FreeBSD src repository Modified files: lib/libarchive archive_write_set_format_pax.c config_freebsd.h Log: Resolve a minor nit in SUS compliance by including the PID in the fake directory name used for pax extended headers. Revision Changes Path 1.44 +26 -12 src/lib/libarchive/archive_write_set_format_pax.c 1.7 +1 -0 src/lib/libarchive/config_freebsd.h From kientzle at FreeBSD.org Sat Mar 15 02:31:29 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 02:31:31 2008 Subject: cvs commit: src/lib/libarchive/test .cvsignore Message-ID: <200803150231.m2F2VSAK065996@repoman.freebsd.org> kientzle 2008-03-15 02:31:28 UTC FreeBSD src repository Modified files: lib/libarchive/test .cvsignore Log: Ignore a few more common files. Revision Changes Path 1.2 +7 -1 src/lib/libarchive/test/.cvsignore From emax at FreeBSD.org Sat Mar 15 02:35:28 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Sat Mar 15 02:35:33 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.c Message-ID: <200803150235.m2F2ZR8c066210@repoman.freebsd.org> emax 2008-03-15 02:35:27 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.c Log: MFC: Update usage() - add new -D option. While I'm here, update copyright information and license formatting Revision Changes Path 1.6.2.2 +5 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c From emax at FreeBSD.org Sat Mar 15 02:36:44 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Sat Mar 15 02:36:49 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.c Message-ID: <200803150236.m2F2aW9b066314@repoman.freebsd.org> emax 2008-03-15 02:36:32 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.c Log: MFC: Update usage() - add new -D option. While I'm here, update copyright information and license formatting Revision Changes Path 1.5.8.3 +5 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c From kientzle at FreeBSD.org Sat Mar 15 02:41:44 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 02:41:46 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.1 bsdtar.c write.c Message-ID: <200803150241.m2F2fibu066679@repoman.freebsd.org> kientzle 2008-03-15 02:41:44 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.1 bsdtar.c write.c Log: Enable and document the -Z option for selecting "compress" compression, now that libarchive supports it. Revision Changes Path 1.38 +10 -0 src/usr.bin/tar/bsdtar.1 1.80 +0 -5 src/usr.bin/tar/bsdtar.c 1.65 +3 -0 src/usr.bin/tar/write.c From kientzle at FreeBSD.org Sat Mar 15 02:45:58 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 02:46:02 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.c Message-ID: <200803150245.m2F2jv1b066887@repoman.freebsd.org> kientzle 2008-03-15 02:45:57 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.c Log: Support --compress as a synonym for -Z. Revision Changes Path 1.81 +1 -0 src/usr.bin/tar/bsdtar.c From kientzle at FreeBSD.org Sat Mar 15 03:06:47 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 03:06:49 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.c bsdtar.h config_freebsd.h read.c Message-ID: <200803150306.m2F36lHW069167@repoman.freebsd.org> kientzle 2008-03-15 03:06:46 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.c bsdtar.h config_freebsd.h read.c Log: --chroot implementation thanks to Joerg Sonnenberger. Revision Changes Path 1.82 +5 -0 src/usr.bin/tar/bsdtar.c 1.30 +1 -0 src/usr.bin/tar/bsdtar.h 1.3 +1 -0 src/usr.bin/tar/config_freebsd.h 1.36 +11 -0 src/usr.bin/tar/read.c From kientzle at FreeBSD.org Sat Mar 15 03:11:13 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 03:11:16 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.c Message-ID: <200803150311.m2F3BDlk069380@repoman.freebsd.org> kientzle 2008-03-15 03:11:13 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.c Log: Also accept --uncompress as a synonym for -Z. Thanks to: Joerg Sonnenberger Revision Changes Path 1.83 +1 -0 src/usr.bin/tar/bsdtar.c From kientzle at FreeBSD.org Sat Mar 15 03:18:25 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 03:18:30 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.1 bsdtar.c Message-ID: <200803150318.m2F3IPWY069594@repoman.freebsd.org> kientzle 2008-03-15 03:18:24 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.1 bsdtar.c Log: Support -q (from NetBSD) as a synonym for --fast-read (the old FreeBSD extension to GNU tar). Thanks to: Joerg Sonnenberger Revision Changes Path 1.39 +9 -9 src/usr.bin/tar/bsdtar.1 1.84 +4 -5 src/usr.bin/tar/bsdtar.c From alfred at FreeBSD.org Sat Mar 15 03:20:53 2008 From: alfred at FreeBSD.org (Alfred Perlstein) Date: Sat Mar 15 03:20:58 2008 Subject: cvs commit: src/sys/kern kern_resource.c Message-ID: <200803150320.m2F3Krn1069706@repoman.freebsd.org> alfred 2008-03-15 03:20:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/kern kern_resource.c Log: In calcru, we make sure the thread's state is correct, however we do not emit the thread or proc information making debugging this assertion difficult. Emit the thread/proc that caused the KASSERT to fail to ease debugging. Note: this assert does not exist in later branches of FreeBSD. Revision Changes Path 1.148.2.4 +2 -1 src/sys/kern/kern_resource.c From kientzle at FreeBSD.org Sat Mar 15 03:25:27 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 03:25:32 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.1 Message-ID: <200803150325.m2F3PRm6069914@repoman.freebsd.org> kientzle 2008-03-15 03:25:26 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.1 Log: Document --chroot. Revision Changes Path 1.40 +6 -0 src/usr.bin/tar/bsdtar.1 From kientzle at FreeBSD.org Sat Mar 15 03:28:39 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 03:28:43 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.c Message-ID: <200803150328.m2F3SdEb070034@repoman.freebsd.org> kientzle 2008-03-15 03:28:39 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.c Log: Support NetBSD's --insecure as a synonym for -P. Revision Changes Path 1.85 +1 -0 src/usr.bin/tar/bsdtar.c From kientzle at FreeBSD.org Sat Mar 15 03:49:19 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 03:49:23 2008 Subject: cvs commit: src/lib/libarchive archive_write_set_format_pax.c Message-ID: <200803150349.m2F3nIiB070742@repoman.freebsd.org> kientzle 2008-03-15 03:49:18 UTC FreeBSD src repository Modified files: lib/libarchive archive_write_set_format_pax.c Log: Testability is more important than standards conformance. Disable the use of PaxHeader. for the fake pax extension pathname until I can make the name here settable. Otherwise, tests that try to compare output to static pre-generated reference files break. Revision Changes Path 1.45 +8 -3 src/lib/libarchive/archive_write_set_format_pax.c From kientzle at FreeBSD.org Sat Mar 15 04:20:52 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 04:20:58 2008 Subject: cvs commit: src/lib/libarchive archive_write_disk.c config_freebsd.h Message-ID: <200803150420.m2F4Kq3b073088@repoman.freebsd.org> kientzle 2008-03-15 04:20:50 UTC FreeBSD src repository Modified files: lib/libarchive archive_write_disk.c config_freebsd.h Log: FreeBSD does have fstat(). Correct the nasty typo this uncovers. Revision Changes Path 1.24 +1 -1 src/lib/libarchive/archive_write_disk.c 1.8 +1 -0 src/lib/libarchive/config_freebsd.h From kientzle at FreeBSD.org Sat Mar 15 05:08:21 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 05:08:26 2008 Subject: cvs commit: src/usr.bin/tar bsdtar.c Message-ID: <200803150508.m2F58Lm7084928@repoman.freebsd.org> kientzle 2008-03-15 05:08:21 UTC FreeBSD src repository Modified files: usr.bin/tar bsdtar.c Log: style(9) fix to add whitespace around an '=' sign and restore a comment that got accidentally dropped. Revision Changes Path 1.86 +2 -1 src/usr.bin/tar/bsdtar.c From imp at FreeBSD.org Sat Mar 15 06:44:46 2008 From: imp at FreeBSD.org (Warner Losh) Date: Sat Mar 15 06:44:47 2008 Subject: cvs commit: src/sys/ia64/isa isa_dma.c src/sys/sys bus_dma.h Message-ID: <200803150644.m2F6ij8S089013@repoman.freebsd.org> imp 2008-03-15 06:44:45 UTC FreeBSD src repository Modified files: sys/ia64/isa isa_dma.c sys/sys bus_dma.h Log: BUS_DMA_ISA is left over from Alpha, and is not used in the tree at all. The reference in ia64 code is due to cutNpaste in its history and can safely be removed. Revired by: cognet, raj, marcel, jhb and maybe one other whom I'm forgetting Revision Changes Path 1.11 +1 -1 src/sys/ia64/isa/isa_dma.c 1.31 +0 -1 src/sys/sys/bus_dma.h From bmah at FreeBSD.org Sat Mar 15 07:06:13 2008 From: bmah at FreeBSD.org (Bruce A. Mah) Date: Sat Mar 15 07:06:15 2008 Subject: cvs commit: src/release/doc/en_US.ISO8859-1/relnotes article.sgml Message-ID: <200803150706.m2F76CX7090985@repoman.freebsd.org> bmah 2008-03-15 07:06:12 UTC FreeBSD src repository Modified files: release/doc/en_US.ISO8859-1/relnotes article.sgml Log: New release notes: cpuset, BTX real mode, gptboot (+MFC), BSD ar(1)/ranlib(1), traceroute(8) -a/-A, traceroute6(8) -U. Updated release note: tzdata2008a (+MFC). Revision Changes Path 1.1080 +29 -3 src/release/doc/en_US.ISO8859-1/relnotes/article.sgml From kientzle at FreeBSD.org Sat Mar 15 07:10:25 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 07:10:27 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_mtree.c Message-ID: <200803150710.m2F7AO5l091266@repoman.freebsd.org> kientzle 2008-03-15 07:10:24 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_mtree.c Log: Refactor the mtree code a bit to make the layering clearer: Each "file" is described by multiple "lines" each possibly containing multiple "keywords." Incorporate some additions from Joerg Sonnenberger to handle linked files and correctly deal with backing files on disk. Revision Changes Path 1.3 +173 -107 src/lib/libarchive/archive_read_support_format_mtree.c From kientzle at FreeBSD.org Sat Mar 15 11:02:48 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 11:02:49 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_mtree.c Message-ID: <200803151102.m2FB2lGT011395@repoman.freebsd.org> kientzle 2008-03-15 11:02:47 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_mtree.c Log: Correct a sign mismatch that only showed up on 64-bit systems. Pointy hat: me Revision Changes Path 1.4 +2 -2 src/lib/libarchive/archive_read_support_format_mtree.c From rwatson at FreeBSD.org Sat Mar 15 11:03:57 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 15 11:04:00 2008 Subject: cvs commit: src/sys/fs/coda coda_subr.c Message-ID: <200803151103.m2FB3u2Q012640@repoman.freebsd.org> rwatson 2008-03-15 11:03:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_subr.c Log: Merge coda_subr.c:1.39 from HEAD to RELENG_7: In Coda, flush the attribute cache for a cnode when its fid is changed, as its synthesized inode number may have changed and we want stat(2) to pick up the new inode number. Revision Changes Path 1.33.2.6 +4 -1 src/sys/fs/coda/coda_subr.c From kientzle at FreeBSD.org Sat Mar 15 11:04:46 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 11:04:50 2008 Subject: cvs commit: src/lib/libarchive archive_write_private.h archive_write_set_format_ar.c archive_write_set_format_cpio.c archive_write_set_format_cpio_newc.c archive_write_set_format_pax.c archive_write_set_format_shar.c archive_write_set_format_ustar.c Message-ID: <200803151104.m2FB4jcX012699@repoman.freebsd.org> kientzle 2008-03-15 11:04:45 UTC FreeBSD src repository Modified files: lib/libarchive archive_write_private.h archive_write_set_format_ar.c archive_write_set_format_cpio.c archive_write_set_format_cpio_newc.c archive_write_set_format_pax.c archive_write_set_format_shar.c archive_write_set_format_ustar.c Log: Remove the duplicate "archive_format" and "archive_format_name" fields from the private archive_write structure and fix up all writers to use the format fields in the base "archive" structure. This error made it impossible to query the format after setting up a writer because the write format was stored in an inaccessible place. Revision Changes Path 1.3 +0 -7 src/lib/libarchive/archive_write_private.h 1.6 +6 -6 src/lib/libarchive/archive_write_set_format_ar.c 1.14 +2 -2 src/lib/libarchive/archive_write_set_format_cpio.c 1.4 +2 -2 src/lib/libarchive/archive_write_set_format_cpio_newc.c 1.46 +6 -6 src/lib/libarchive/archive_write_set_format_pax.c 1.19 +4 -4 src/lib/libarchive/archive_write_set_format_shar.c 1.26 +2 -2 src/lib/libarchive/archive_write_set_format_ustar.c From kientzle at FreeBSD.org Sat Mar 15 11:05:49 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 11:05:54 2008 Subject: cvs commit: src/lib/libarchive/test test_write_compress.c Message-ID: <200803151105.m2FB5njr012822@repoman.freebsd.org> kientzle 2008-03-15 11:05:49 UTC FreeBSD src repository Modified files: lib/libarchive/test test_write_compress.c Log: Release the buffers used for exercising the compress code. Revision Changes Path 1.2 +3 -0 src/lib/libarchive/test/test_write_compress.c From kientzle at FreeBSD.org Sat Mar 15 11:06:16 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 11:06:18 2008 Subject: cvs commit: src/lib/libarchive/test test_empty_write.c Message-ID: <200803151106.m2FB6Fdc012880@repoman.freebsd.org> kientzle 2008-03-15 11:06:15 UTC FreeBSD src repository Modified files: lib/libarchive/test test_empty_write.c Log: Free up the entry objects allocated during this test. Revision Changes Path 1.2 +3 -0 src/lib/libarchive/test/test_empty_write.c From rwatson at FreeBSD.org Sat Mar 15 11:06:57 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 15 11:06:59 2008 Subject: cvs commit: src/sys/fs/coda coda_psdev.c coda_vnops.c Message-ID: <200803151106.m2FB6usO012925@repoman.freebsd.org> rwatson 2008-03-15 11:06:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_psdev.c coda_vnops.c Log: Merge coda_psdev.c:1.45 and coda_vnops.c:1.96 from HEAD to RELENG_7: Remove debugging code under OLD_DIAGNOSTIC; this is all >10 years old and hasn't been used in that time. Revision Changes Path 1.39.2.6 +0 -4 src/sys/fs/coda/coda_psdev.c 1.76.2.12 +3 -28 src/sys/fs/coda/coda_vnops.c From rwatson at FreeBSD.org Sat Mar 15 11:08:38 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Mar 15 11:08:39 2008 Subject: cvs commit: src/sys/kern kern_sig.c Message-ID: <200803151108.m2FB8cqF012982@repoman.freebsd.org> rwatson 2008-03-15 11:08:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern kern_sig.c Log: Merge kern_sig.c:1.357 from HEAD to RELENG_7: Use sbuf routines to construct core dump filenames rather than custom string buffer handling, making the code both easier to read and more robust against string-handling bugs. Revision Changes Path 1.349.2.3 +22 -27 src/sys/kern/kern_sig.c From kientzle at FreeBSD.org Sat Mar 15 11:09:16 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Mar 15 11:09:20 2008 Subject: cvs commit: src/lib/libarchive archive_read_private.h Message-ID: <200803151109.m2FB9Gpp013018@repoman.freebsd.org> kientzle 2008-03-15 11:09:16 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_private.h Log: Update a comment: the format bid only runs once per archive; it no longer runs once per entry. Revision Changes Path 1.6 +2 -9 src/lib/libarchive/archive_read_private.h From remko at FreeBSD.org Sat Mar 15 14:41:27 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Sat Mar 15 14:41:29 2008 Subject: cvs commit: src/sys/sys disklabel.h Message-ID: <200803151441.m2FEfRvY037357@repoman.freebsd.org> remko 2008-03-15 14:41:27 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/sys disklabel.h Log: MFC rev 1.111 disklabel.h Add missing comma. PR: bin/121645 Submitted by: OISHI Masakuni Approved by: imp (mentor, implicit for trivial changes) MFC after: 3 days Approved by: imp (mentor, implicit) Revision Changes Path 1.108.2.2 +1 -1 src/sys/sys/disklabel.h From ru at FreeBSD.org Sat Mar 15 17:40:19 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Sat Mar 15 17:40:21 2008 Subject: cvs commit: src/sys/kern link_elf.c link_elf_obj.c Message-ID: <200803151740.m2FHeIcU057488@repoman.freebsd.org> ru 2008-03-15 17:40:18 UTC FreeBSD src repository Modified files: sys/kern link_elf.c link_elf_obj.c Log: Fix panic on e.g. "kldload /dev/null". PR: kern/121427 Reviewed by: sem MFC after: 3 days Revision Changes Path 1.96 +5 -0 src/sys/kern/link_elf.c 1.98 +4 -0 src/sys/kern/link_elf_obj.c From gshapiro at FreeBSD.org Sat Mar 15 20:54:09 2008 From: gshapiro at FreeBSD.org (Gregory Neil Shapiro) Date: Sat Mar 15 20:54:13 2008 Subject: cvs commit: src/lib/libmilter Makefile Message-ID: <200803152054.m2FKs8Y5079338@repoman.freebsd.org> gshapiro 2008-03-15 20:54:08 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libmilter Makefile Log: MFC: Switch libmilter from select(2) to poll(2) so milters are not limited by the size of FD_SETSIZE. PR: 118824 Submitted by: vsevolod Revision Changes Path 1.7.2.1 +1 -0 src/lib/libmilter/Makefile From gshapiro at FreeBSD.org Sat Mar 15 20:55:37 2008 From: gshapiro at FreeBSD.org (Gregory Neil Shapiro) Date: Sat Mar 15 20:55:39 2008 Subject: cvs commit: src/lib/libmilter Makefile Message-ID: <200803152055.m2FKtbvu079470@repoman.freebsd.org> gshapiro 2008-03-15 20:55:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libmilter Makefile Log: MFC: Switch libmilter from select(2) to poll(2) so milters are not limited by the size of FD_SETSIZE. PR: 118824 Submitted by: vsevolod Revision Changes Path 1.5.2.2 +1 -0 src/lib/libmilter/Makefile From pjd at FreeBSD.org Sat Mar 15 23:23:04 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sat Mar 15 23:23:10 2008 Subject: cvs commit: src/sys/contrib/opensolaris/uts/common/fs/zfs zfs_vnops.c Message-ID: <200803152323.m2FNN4TH088266@repoman.freebsd.org> pjd 2008-03-15 23:23:04 UTC FreeBSD src repository Modified files: sys/contrib/opensolaris/uts/common/fs/zfs zfs_vnops.c Log: Fix mmap(2) on ZFS after some changes in VM subsystem. Submitted by: alc Reported by: kris (originally) and many others Tested with: fsx MFC after: 1 week Revision Changes Path 1.28 +4 -0 src/sys/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c From davidxu at FreeBSD.org Sun Mar 16 03:22:40 2008 From: davidxu at FreeBSD.org (David Xu) Date: Sun Mar 16 03:22:45 2008 Subject: cvs commit: src/lib/libthr/thread thr_create.c Message-ID: <200803160322.m2G3MdN5016379@repoman.freebsd.org> davidxu 2008-03-16 03:22:38 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_create.c Log: Actually delete SIGCANCEL mask for suspended thread, so the signal will not be masked when it is resumed. Revision Changes Path 1.41 +2 -3 src/lib/libthr/thread/thr_create.c From sobomax at FreeBSD.org Sun Mar 16 06:21:30 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Sun Mar 16 06:21:35 2008 Subject: cvs commit: src/sys/sys eventhandler.h src/sys/kern kern_descrip.c uipc_socket.c Message-ID: <200803160621.m2G6LUlE034766@repoman.freebsd.org> sobomax 2008-03-16 06:21:30 UTC FreeBSD src repository Modified files: sys/sys eventhandler.h sys/kern kern_descrip.c uipc_socket.c Log: Properly set size of the file_zone to match kern.maxfiles parameter. Otherwise the parameter is no-op, since zone by default limits number of descriptors to some 12K entries. Attempt to allocate more ends up sleeping on zonelimit. MFC after: 2 weeks Revision Changes Path 1.326 +36 -3 src/sys/kern/kern_descrip.c 1.306 +1 -0 src/sys/kern/uipc_socket.c 1.39 +1 -0 src/sys/sys/eventhandler.h From obrien at FreeBSD.org Sun Mar 16 06:53:41 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 06:53:47 2008 Subject: cvs commit: src/sbin/md5 md5.c Message-ID: <200803160653.m2G6rf2x035868@repoman.freebsd.org> obrien 2008-03-16 06:53:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/md5 md5.c Log: MFC: 1.36: style(9) Revision Changes Path 1.35.2.1 +13 -14 src/sbin/md5/md5.c From obrien at FreeBSD.org Sun Mar 16 06:55:14 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 06:55:19 2008 Subject: cvs commit: src/lib/libthr/arch/arm/include pthread_md.h src/lib/libthr/arch/sparc64/include pthread_md.h Message-ID: <200803160655.m2G6tDFF035993@repoman.freebsd.org> obrien 2008-03-16 06:55:13 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libthr/arch/arm/include pthread_md.h lib/libthr/arch/sparc64/include pthread_md.h Log: MFC: style(9) arm/include/pthread_md.h rev 1.5 sparc64/include/pthread_md.h rev 1.3 Revision Changes Path 1.4.2.1 +3 -3 src/lib/libthr/arch/arm/include/pthread_md.h 1.2.2.1 +3 -3 src/lib/libthr/arch/sparc64/include/pthread_md.h From obrien at FreeBSD.org Sun Mar 16 06:56:23 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 06:56:32 2008 Subject: cvs commit: src/lib/libthr/arch/amd64 Makefile.inc src/lib/libthr/arch/arm Makefile.inc src/lib/libthr/arch/i386 Makefile.inc src/lib/libthr/arch/ia64 Makefile.inc src/lib/libthr/arch/powerpc Makefile.inc src/lib/libthr/arch/sparc64 Makefile.inc Message-ID: <200803160656.m2G6uM4m036161@repoman.freebsd.org> obrien 2008-03-16 06:56:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libthr/arch/amd64 Makefile.inc lib/libthr/arch/arm Makefile.inc lib/libthr/arch/i386 Makefile.inc lib/libthr/arch/ia64 Makefile.inc lib/libthr/arch/powerpc Makefile.inc lib/libthr/arch/sparc64 Makefile.inc Log: MFC: style.Makefile(5) Revision Changes Path 1.2.10.1 +1 -1 src/lib/libthr/arch/amd64/Makefile.inc 1.2.10.1 +1 -1 src/lib/libthr/arch/arm/Makefile.inc 1.2.10.1 +1 -1 src/lib/libthr/arch/i386/Makefile.inc 1.2.10.1 +1 -1 src/lib/libthr/arch/ia64/Makefile.inc 1.2.10.1 +1 -1 src/lib/libthr/arch/powerpc/Makefile.inc 1.3.10.1 +1 -1 src/lib/libthr/arch/sparc64/Makefile.inc From obrien at FreeBSD.org Sun Mar 16 06:59:50 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 06:59:52 2008 Subject: cvs commit: src/usr.bin/make Makefile job.c Message-ID: <200803160659.m2G6xoEN036243@repoman.freebsd.org> obrien 2008-03-16 06:59:50 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/make Makefile job.c Log: MFC: "make love" and GC bootstrapping aid for old (mid-2001) systems. Revision Changes Path 1.65.2.1 +0 -5 src/usr.bin/make/Makefile 1.126.2.1 +1 -5 src/usr.bin/make/job.c From obrien at FreeBSD.org Sun Mar 16 07:00:52 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:01:00 2008 Subject: cvs commit: src/tools/tools/editing freebsd.vim Message-ID: <200803160700.m2G70qAc036367@repoman.freebsd.org> obrien 2008-03-16 07:00:52 UTC FreeBSD src repository Added files: (Branch: RELENG_7) tools/tools/editing freebsd.vim Log: MFC: rev 1.1. Revision Changes Path 1.1.2.1 +68 -0 src/tools/tools/editing/freebsd.vim (new) From obrien at FreeBSD.org Sun Mar 16 07:04:04 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:04:08 2008 Subject: cvs commit: src/bin Makefile src/usr.bin Makefile src/bin/uuidgen Makefile uuidgen.1 uuidgen.c src/usr.bin/uuidgen Makefile uuidgen.1 uuidgen.c Message-ID: <200803160704.m2G7444n037709@repoman.freebsd.org> obrien 2008-03-16 07:04:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) bin Makefile usr.bin Makefile Added files: (Branch: RELENG_7) bin/uuidgen Makefile uuidgen.1 uuidgen.c Removed files: (Branch: RELENG_7) usr.bin/uuidgen Makefile uuidgen.1 uuidgen.c Log: MFC: repcopy from usr.bin/ to bin/ Revision Changes Path 1.26.2.1 +2 -1 src/bin/Makefile 1.4.2.1 +6 -0 src/bin/uuidgen/Makefile (new) 1.7.6.1 +80 -0 src/bin/uuidgen/uuidgen.1 (new) 1.3.28.1 +113 -0 src/bin/uuidgen/uuidgen.c (new) 1.303.2.1 +0 -1 src/usr.bin/Makefile 1.3.2.1 +0 -8 src/usr.bin/uuidgen/Makefile (dead) 1.7.2.1 +0 -80 src/usr.bin/uuidgen/uuidgen.1 (dead) 1.3.22.1 +0 -113 src/usr.bin/uuidgen/uuidgen.c (dead) From obrien at FreeBSD.org Sun Mar 16 07:09:17 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:09:19 2008 Subject: cvs commit: src/share/man/man9 style.9 Message-ID: <200803160709.m2G79HM6037882@repoman.freebsd.org> obrien 2008-03-16 07:09:16 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man9 style.9 Log: MFC: rev 1.125: add to history Revision Changes Path 1.123.2.1 +5 -1 src/share/man/man9/style.9 From obrien at FreeBSD.org Sun Mar 16 07:12:48 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:12:50 2008 Subject: cvs commit: src/usr.bin/top machine.c src/contrib/top top.c Message-ID: <200803160712.m2G7CmgD038060@repoman.freebsd.org> obrien 2008-03-16 07:12:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/top machine.c contrib/top top.c Log: MFC: improve -u (limit uid lookups) behavior. machine.c rev 1.84 & top.c rev 1.24 Revision Changes Path 1.23.2.1 +2 -2 src/contrib/top/top.c 1.82.2.1 +8 -2 src/usr.bin/top/machine.c From obrien at FreeBSD.org Sun Mar 16 07:20:20 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:20:24 2008 Subject: cvs commit: src/usr.bin/netstat atalk.c bpf.c if.c inet.c inet6.c ipsec.c ipx.c main.c mbuf.c mcast.c mroute.c mroute6.c netgraph.c netstat.h pfkey.c route.c sctp.c unix.c Message-ID: <200803160720.m2G7KJQP038407@repoman.freebsd.org> obrien 2008-03-16 07:20:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/netstat atalk.c bpf.c if.c inet.c inet6.c ipsec.c ipx.c main.c mbuf.c mcast.c mroute.c mroute6.c netgraph.c netstat.h pfkey.c route.c sctp.c unix.c Log: MFC: style(9) Revision Changes Path 1.26.2.1 +2 -2 src/usr.bin/netstat/atalk.c 1.9.2.1 +4 -3 src/usr.bin/netstat/bpf.c 1.69.2.1 +19 -18 src/usr.bin/netstat/if.c 1.78.2.2 +102 -101 src/usr.bin/netstat/inet.c 1.29.2.1 +57 -57 src/usr.bin/netstat/inet6.c 1.17.2.1 +12 -15 src/usr.bin/netstat/ipsec.c 1.27.2.1 +7 -7 src/usr.bin/netstat/ipx.c 1.87.2.1 +28 -28 src/usr.bin/netstat/main.c 1.53.2.1 +2 -2 src/usr.bin/netstat/mbuf.c 1.9.2.1 +2 -2 src/usr.bin/netstat/mcast.c 1.30.2.1 +4 -4 src/usr.bin/netstat/mroute.c 1.21.2.1 +2 -3 src/usr.bin/netstat/mroute6.c 1.13.2.1 +3 -3 src/usr.bin/netstat/netgraph.c 1.51.2.2 +6 -6 src/usr.bin/netstat/netstat.h 1.5.2.1 +5 -11 src/usr.bin/netstat/pfkey.c 1.82.2.4 +19 -19 src/usr.bin/netstat/route.c 1.7.2.2 +36 -37 src/usr.bin/netstat/sctp.c 1.20.2.1 +3 -3 src/usr.bin/netstat/unix.c From obrien at FreeBSD.org Sun Mar 16 07:21:02 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:21:07 2008 Subject: cvs commit: src/usr.bin/netstat inet.c Message-ID: <200803160721.m2G7L2nV038465@repoman.freebsd.org> obrien 2008-03-16 07:21:02 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/netstat inet.c Log: MFC: rev 1.82: make a few messages more consistant Revision Changes Path 1.78.2.3 +2 -2 src/usr.bin/netstat/inet.c From obrien at FreeBSD.org Sun Mar 16 07:22:57 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:23:01 2008 Subject: cvs commit: src/usr.sbin/sysinstall media.c system.c Message-ID: <200803160722.m2G7MvfC038543@repoman.freebsd.org> obrien 2008-03-16 07:22:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/sysinstall media.c system.c Log: MFC: use UNZIPPER everywhere. Revision Changes Path 1.123.2.1 +2 -4 src/usr.sbin/sysinstall/media.c 1.129.2.1 +3 -2 src/usr.sbin/sysinstall/system.c From obrien at FreeBSD.org Sun Mar 16 07:23:47 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:23:49 2008 Subject: cvs commit: src/usr.sbin/syslogd syslogd.c Message-ID: <200803160723.m2G7NkJU038581@repoman.freebsd.org> obrien 2008-03-16 07:23:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/syslogd syslogd.c Log: MFC: rev 1.153: style(9) Revision Changes Path 1.152.2.1 +15 -15 src/usr.sbin/syslogd/syslogd.c From obrien at FreeBSD.org Sun Mar 16 07:52:23 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 07:52:27 2008 Subject: cvs commit: src/sbin/init init.c Message-ID: <200803160752.m2G7qMOU039915@repoman.freebsd.org> obrien 2008-03-16 07:52:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/init init.c Log: MFC: revs 1.64-65: style(9) Revision Changes Path 1.63.2.1 +99 -98 src/sbin/init/init.c From obrien at FreeBSD.org Sun Mar 16 08:00:50 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 08:00:51 2008 Subject: cvs commit: src/share/man/man3 queue.3 Message-ID: <200803160800.m2G80ndV040213@repoman.freebsd.org> obrien 2008-03-16 08:00:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man3 queue.3 Log: MFC: rev 1.41: don't imply O(n) removal for the doubly linked data structures. Revision Changes Path 1.40.2.1 +8 -2 src/share/man/man3/queue.3 From obrien at FreeBSD.org Sun Mar 16 08:01:39 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 08:01:41 2008 Subject: cvs commit: src/sbin/route route.8 Message-ID: <200803160801.m2G81dPQ040258@repoman.freebsd.org> obrien 2008-03-16 08:01:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/route route.8 Log: MFC: 1.47: document -t Revision Changes Path 1.46.2.1 +4 -0 src/sbin/route/route.8 From obrien at FreeBSD.org Sun Mar 16 08:02:38 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 08:02:43 2008 Subject: cvs commit: src/usr.sbin/sade sade.8 Message-ID: <200803160802.m2G82btS040308@repoman.freebsd.org> obrien 2008-03-16 08:02:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/sade sade.8 Log: MFC: 1.78: adjust history Revision Changes Path 1.77.4.1 +1 -1 src/usr.sbin/sade/sade.8 From obrien at FreeBSD.org Sun Mar 16 08:03:12 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 08:03:16 2008 Subject: cvs commit: src/gnu/usr.bin/sort Makefile Message-ID: <200803160803.m2G83B39041538@repoman.freebsd.org> obrien 2008-03-16 08:03:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) gnu/usr.bin/sort Makefile Log: MFC: rev 1.13: style.Makefile(5) Revision Changes Path 1.12.18.1 +2 -2 src/gnu/usr.bin/sort/Makefile From obrien at FreeBSD.org Sun Mar 16 08:03:49 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Mar 16 08:03:54 2008 Subject: cvs commit: src/gnu/usr.bin/diff Makefile Message-ID: <200803160803.m2G83nVB041572@repoman.freebsd.org> obrien 2008-03-16 08:03:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) gnu/usr.bin/diff Makefile Log: MFC: rev 1.18: style.Makefile(5) Revision Changes Path 1.17.2.1 +2 -2 src/gnu/usr.bin/diff/Makefile From jkim at FreeBSD.org Sun Mar 16 08:06:36 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Sun Mar 16 08:06:39 2008 Subject: cvs commit: src/usr.sbin/mixer Makefile Message-ID: <200803160806.m2G86ai3041718@repoman.freebsd.org> jkim 2008-03-16 08:06:36 UTC FreeBSD src repository Modified files: usr.sbin/mixer Makefile Log: mixer(8) is WARNS=6 clean since 1.25. Tested with: 'make universe' Revision Changes Path 1.11 +2 -2 src/usr.sbin/mixer/Makefile From brueffer at FreeBSD.org Sun Mar 16 08:18:05 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sun Mar 16 08:18:06 2008 Subject: cvs commit: src/sbin/geom/class/multipath gmultipath.8 Message-ID: <200803160818.m2G8I4pU042036@repoman.freebsd.org> brueffer 2008-03-16 08:18:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/geom/class/multipath gmultipath.8 Log: MFC: rev. 1.6 - Use an uppercase provider name in the example, to make the name change after labeling the provider more obvious. - Correct nomenclature usage Revision Changes Path 1.4.2.2 +7 -7 src/sbin/geom/class/multipath/gmultipath.8 From rwatson at FreeBSD.org Sun Mar 16 10:29:23 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 10:29:26 2008 Subject: cvs commit: src/sys/sys eventhandler.h src/sys/kern kern_descrip.c uipc_socket.c In-Reply-To: <200803160621.m2G6LUlE034766@repoman.freebsd.org> References: <200803160621.m2G6LUlE034766@repoman.freebsd.org> Message-ID: <20080316100836.F44049@fledge.watson.org> On Sun, 16 Mar 2008, Maxim Sobolev wrote: > sobomax 2008-03-16 06:21:30 UTC > > FreeBSD src repository > > Modified files: > sys/sys eventhandler.h > sys/kern kern_descrip.c uipc_socket.c > Log: > Properly set size of the file_zone to match kern.maxfiles parameter. > Otherwise the parameter is no-op, since zone by default limits number > of descriptors to some 12K entries. Attempt to allocate more ends up > sleeping on zonelimit. Hmm. Could you be more specific about the nature of the problem you're seeing here? UMA shouldn't impose any default zone limit on the file_zone if none is set. Are you sure it's the struct file zonelimit you're running into, and not, say, the socket zone limit, etc? Also, don't use SI_SUB_EVENTHANDLER+1 -- add a new event type, perhaps SI_SUB_FILE, or use SI_ORDER_LAST. Robert N M Watson Computer Laboratory University of Cambridge From jroberson at chesapeake.net Sun Mar 16 10:39:17 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Sun Mar 16 10:39:25 2008 Subject: cvs commit: src/sys/sys eventhandler.h src/sys/kern kern_descrip.c uipc_socket.c In-Reply-To: <200803160621.m2G6LUlE034766@repoman.freebsd.org> References: <200803160621.m2G6LUlE034766@repoman.freebsd.org> Message-ID: <20080316003849.L910@desktop> On Sun, 16 Mar 2008, Maxim Sobolev wrote: > sobomax 2008-03-16 06:21:30 UTC > > FreeBSD src repository > > Modified files: > sys/sys eventhandler.h > sys/kern kern_descrip.c uipc_socket.c > Log: > Properly set size of the file_zone to match kern.maxfiles parameter. > Otherwise the parameter is no-op, since zone by default limits number > of descriptors to some 12K entries. Attempt to allocate more ends up > sleeping on zonelimit. Why do you need to set a limit on the uma_zone when falloc() takes care of this for us by checking openfiles vs maxfiles? Thanks, Jeff > > MFC after: 2 weeks > > Revision Changes Path > 1.326 +36 -3 src/sys/kern/kern_descrip.c > 1.306 +1 -0 src/sys/kern/uipc_socket.c > 1.39 +1 -0 src/sys/sys/eventhandler.h > From rwatson at FreeBSD.org Sun Mar 16 10:58:10 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 10:58:15 2008 Subject: cvs commit: src/sys/sys eventhandler.h module.h taskqueue.h src/sys/amd64/acpica madt.c src/sys/amd64/amd64 intr_machdep.c local_apic.c machdep.c mptable.c src/sys/amd64/isa atpic.c src/sys/arm/arm machdep.c vm_machdep.c ... Message-ID: <200803161058.m2GAw9t3059676@repoman.freebsd.org> rwatson 2008-03-16 10:58:09 UTC FreeBSD src repository Modified files: sys/sys eventhandler.h module.h taskqueue.h sys/amd64/acpica madt.c sys/amd64/amd64 intr_machdep.c local_apic.c machdep.c mptable.c sys/amd64/isa atpic.c sys/arm/arm machdep.c vm_machdep.c sys/contrib/opensolaris/uts/common/os callb.c taskq.c sys/dev/ofw ofw_console.c sys/dev/syscons sysmouse.c sys/i386/acpica acpi_wakeup.c madt.c sys/i386/i386 intr_machdep.c local_apic.c machdep.c mp_machdep.c mptable.c vm_machdep.c sys/i386/isa atpic.c sys/i386/xbox xbox.c sys/ia64/ia64 machdep.c sscdisk.c sys/kern init_main.c kern_alq.c kern_clock.c kern_descrip.c kern_event.c kern_idle.c kern_intr.c kern_linker.c kern_malloc.c kern_mbuf.c kern_module.c kern_ntptime.c kern_poll.c kern_shutdown.c kern_switch.c kern_synch.c kern_tc.c sched_4bsd.c sched_ule.c subr_autoconf.c subr_eventhandler.c subr_log.c subr_prof.c subr_smp.c subr_witness.c tty_cons.c tty_pts.c tty_pty.c tty_subr.c tty_tty.c uipc_domain.c vfs_acl.c vfs_bio.c vfs_cache.c vfs_hash.c vfs_lookup.c vfs_subr.c sys/net bpf.c if.c netisr.c rtsock.c sys/netatm/ipatm ipatm_load.c sys/netatm/sigpvc sigpvc_if.c sys/netatm/spans spans_if.c sys/netatm/uni uni_load.c sys/netipsec ipsec.c sys/nfsclient nfs_diskless.c sys/pc98/pc98 machdep.c sys/powerpc/aim machdep.c sys/powerpc/booke machdep.c vm_machdep.c sys/security/audit audit.c sys/sparc64/sparc64 vm_machdep.c sys/ufs/ffs ffs_softdep.c sys/vm vm_glue.c vm_init.c vm_mmap.c vm_pageout.c vm_zeroidle.c Log: In keeping with style(9)'s recommendations on macros, use a ';' after each SYSINIT() macro invocation. This makes a number of lightweight C parsers much happier with the FreeBSD kernel source, including cflow's prcc and lxr. MFC after: 1 month Discussed with: imp, rink Revision Changes Path 1.26 +2 -2 src/sys/amd64/acpica/madt.c 1.38 +3 -2 src/sys/amd64/amd64/intr_machdep.c 1.44 +4 -3 src/sys/amd64/amd64/local_apic.c 1.683 +1 -1 src/sys/amd64/amd64/machdep.c 1.241 +1 -1 src/sys/amd64/amd64/mptable.c 1.23 +1 -1 src/sys/amd64/isa/atpic.c 1.29 +1 -1 src/sys/arm/arm/machdep.c 1.36 +1 -1 src/sys/arm/arm/vm_machdep.c 1.2 +1 -1 src/sys/contrib/opensolaris/uts/common/os/callb.c 1.3 +1 -1 src/sys/contrib/opensolaris/uts/common/os/taskq.c 1.38 +1 -1 src/sys/dev/ofw/ofw_console.c 1.30 +1 -1 src/sys/dev/syscons/sysmouse.c 1.47 +1 -1 src/sys/i386/acpica/acpi_wakeup.c 1.30 +2 -2 src/sys/i386/acpica/madt.c 1.33 +3 -2 src/sys/i386/i386/intr_machdep.c 1.45 +2 -2 src/sys/i386/i386/local_apic.c 1.668 +2 -2 src/sys/i386/i386/machdep.c 1.285 +1 -1 src/sys/i386/i386/mp_machdep.c 1.246 +1 -1 src/sys/i386/i386/mptable.c 1.286 +1 -1 src/sys/i386/i386/vm_machdep.c 1.29 +1 -1 src/sys/i386/isa/atpic.c 1.5 +1 -1 src/sys/i386/xbox/xbox.c 1.233 +1 -1 src/sys/ia64/ia64/machdep.c 1.34 +1 -1 src/sys/ia64/ia64/sscdisk.c 1.292 +15 -13 src/sys/kern/init_main.c 1.23 +2 -2 src/sys/kern/kern_alq.c 1.208 +1 -1 src/sys/kern/kern_clock.c 1.327 +2 -2 src/sys/kern/kern_descrip.c 1.117 +1 -1 src/sys/kern/kern_event.c 1.54 +1 -1 src/sys/kern/kern_idle.c 1.156 +2 -1 src/sys/kern/kern_intr.c 1.155 +5 -4 src/sys/kern/kern_linker.c 1.163 +1 -1 src/sys/kern/kern_malloc.c 1.39 +1 -1 src/sys/kern/kern_mbuf.c 1.55 +1 -1 src/sys/kern/kern_module.c 1.65 +1 -1 src/sys/kern/kern_ntptime.c 1.32 +3 -2 src/sys/kern/kern_poll.c 1.191 +1 -1 src/sys/kern/kern_shutdown.c 1.140 +1 -1 src/sys/kern/kern_switch.c 1.309 +2 -1 src/sys/kern/kern_synch.c 1.181 +1 -1 src/sys/kern/kern_tc.c 1.118 +3 -2 src/sys/kern/sched_4bsd.c 1.234 +3 -2 src/sys/kern/sched_ule.c 1.25 +1 -1 src/sys/kern/subr_autoconf.c 1.24 +1 -1 src/sys/kern/subr_eventhandler.c 1.65 +1 -1 src/sys/kern/subr_log.c 1.80 +2 -2 src/sys/kern/subr_prof.c 1.206 +3 -3 src/sys/kern/subr_smp.c 1.242 +2 -1 src/sys/kern/subr_witness.c 1.142 +1 -1 src/sys/kern/tty_cons.c 1.17 +1 -1 src/sys/kern/tty_pts.c 1.158 +1 -1 src/sys/kern/tty_pty.c 1.44 +1 -1 src/sys/kern/tty_subr.c 1.61 +1 -1 src/sys/kern/tty_tty.c 1.52 +2 -2 src/sys/kern/uipc_domain.c 1.57 +1 -1 src/sys/kern/vfs_acl.c 1.538 +1 -1 src/sys/kern/vfs_bio.c 1.119 +1 -1 src/sys/kern/vfs_cache.c 1.14 +1 -1 src/sys/kern/vfs_hash.c 1.108 +1 -1 src/sys/kern/vfs_lookup.c 1.721 +4 -3 src/sys/kern/vfs_subr.c 1.187 +1 -1 src/sys/net/bpf.c 1.276 +2 -2 src/sys/net/if.c 1.21 +1 -1 src/sys/net/netisr.c 1.144 +1 -1 src/sys/net/rtsock.c 1.24 +1 -1 src/sys/netatm/ipatm/ipatm_load.c 1.19 +1 -1 src/sys/netatm/sigpvc/sigpvc_if.c 1.17 +1 -1 src/sys/netatm/spans/spans_if.c 1.13 +1 -1 src/sys/netatm/uni/uni_load.c 1.27 +1 -1 src/sys/netipsec/ipsec.c 1.20 +1 -1 src/sys/nfsclient/nfs_diskless.c 1.403 +2 -2 src/sys/pc98/pc98/machdep.c 1.111 +1 -1 src/sys/powerpc/aim/machdep.c 1.4 +1 -1 src/sys/powerpc/booke/machdep.c 1.3 +1 -1 src/sys/powerpc/booke/vm_machdep.c 1.40 +1 -1 src/sys/security/audit/audit.c 1.78 +1 -1 src/sys/sparc64/sparc64/vm_machdep.c 1.40 +1 -1 src/sys/sys/eventhandler.h 1.24 +1 -1 src/sys/sys/module.h 1.17 +2 -2 src/sys/sys/taskqueue.h 1.220 +2 -1 src/sys/ufs/ffs/ffs_softdep.c 1.229 +2 -2 src/sys/vm/vm_glue.c 1.47 +1 -1 src/sys/vm/vm_init.c 1.217 +2 -1 src/sys/vm/vm_mmap.c 1.295 +3 -2 src/sys/vm/vm_pageout.c 1.51 +1 -1 src/sys/vm/vm_zeroidle.c From rwatson at FreeBSD.org Sun Mar 16 11:01:33 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 11:01:35 2008 Subject: cvs commit: src/sys/sys kernel.h Message-ID: <200803161101.m2GB1X8j059957@repoman.freebsd.org> rwatson 2008-03-16 11:01:33 UTC FreeBSD src repository Modified files: sys/sys kernel.h Log: Remove trailing ';' from C_SYSINIT() macro definition, in keeping with style(9) recommendation that macros not contain the terminating ';', leaving that to the invoker. All SYSINIT() consumers must now provide a trailing ';'. Unlike the change to remove the ';'s from callers, this change shouldn't be MFC'd unless we don't mind requiring source changes to third party modules that might still depend on SYSINIT() providing its own ';'. Revision Changes Path 1.139 +1 -1 src/sys/sys/kernel.h From rwatson at FreeBSD.org Sun Mar 16 13:02:05 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 13:02:10 2008 Subject: cvs commit: src/sys/gdb gdb_cons.c Message-ID: <200803161302.m2GD24j6077512@repoman.freebsd.org> rwatson 2008-03-16 13:02:04 UTC FreeBSD src repository Modified files: sys/gdb gdb_cons.c Log: Commit SYSINIT() ;-adding patch missed in previous pass. MFC after: 1 month Caught by: tinderbox Revision Changes Path 1.3 +1 -1 src/sys/gdb/gdb_cons.c From das at FreeBSD.org Sun Mar 16 15:15:33 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Mar 16 15:15:39 2008 Subject: cvs commit: src/usr.bin/split split.1 split.c Message-ID: <200803161515.m2GFFXxv085074@repoman.freebsd.org> das 2008-03-16 15:15:32 UTC FreeBSD src repository Modified files: usr.bin/split split.1 split.c Log: Add a -n option to split(1) to split files into N chunks rather than having to specify the right number of bytes. Obtained from: NetBSD Submitted by: Jan Schaumann PR: 113175 Revision Changes Path 1.22 +9 -1 src/usr.bin/split/split.1 1.20 +69 -11 src/usr.bin/split/split.c From rdivacky at FreeBSD.org Sun Mar 16 16:24:21 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Sun Mar 16 16:24:24 2008 Subject: cvs commit: src/sys/amd64/linux32 syscalls.master src/sys/compat/linux linux_misc.c Message-ID: <200803161624.m2GGOK9J088747@repoman.freebsd.org> rdivacky 2008-03-16 16:24:20 UTC FreeBSD src repository Modified files: sys/amd64/linux32 syscalls.master sys/compat/linux linux_misc.c Log: Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor) Revision Changes Path 1.31 +2 -1 src/sys/amd64/linux32/syscalls.master 1.218 +40 -9 src/sys/compat/linux/linux_misc.c From rdivacky at FreeBSD.org Sun Mar 16 16:27:44 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Sun Mar 16 16:27:47 2008 Subject: cvs commit: src/sys/i386/linux syscalls.master Message-ID: <200803161627.m2GGRi1I088907@repoman.freebsd.org> rdivacky 2008-03-16 16:27:44 UTC FreeBSD src repository Modified files: sys/i386/linux syscalls.master Log: Implement sched_setaffinity and get_setaffinity using real cpu affinity setting primitives. Reviewed by: jeff Approved by: kib (mentor) Revision Changes Path 1.90 +2 -1 src/sys/i386/linux/syscalls.master From rdivacky at FreeBSD.org Sun Mar 16 16:29:38 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Sun Mar 16 16:29:40 2008 Subject: cvs commit: src/sys/amd64/linux32 linux32_proto.h linux32_syscall.h linux32_sysent.c src/sys/i386/linux linux_proto.h linux_syscall.h linux_sysent.c Message-ID: <200803161629.m2GGTbda089007@repoman.freebsd.org> rdivacky 2008-03-16 16:29:37 UTC FreeBSD src repository Modified files: sys/amd64/linux32 linux32_sysent.c linux32_syscall.h linux32_proto.h sys/i386/linux linux_proto.h linux_syscall.h linux_sysent.c Log: Regen. Revision Changes Path 1.34 +8 -1 src/sys/amd64/linux32/linux32_proto.h 1.34 +2 -1 src/sys/amd64/linux32/linux32_syscall.h 1.34 +2 -2 src/sys/amd64/linux32/linux32_sysent.c 1.96 +8 -1 src/sys/i386/linux/linux_proto.h 1.88 +2 -1 src/sys/i386/linux/linux_syscall.h 1.95 +2 -2 src/sys/i386/linux/linux_sysent.c From rdivacky at FreeBSD.org Sun Mar 16 16:31:04 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Sun Mar 16 16:31:10 2008 Subject: cvs commit: src/sys/i386/linux syscalls.master In-Reply-To: <200803161627.m2GGRi1I088907@repoman.freebsd.org> References: <200803161627.m2GGRi1I088907@repoman.freebsd.org> Message-ID: <20080316163045.GA66295@freebsd.org> On Sun, Mar 16, 2008 at 04:27:44PM +0000, Roman Divacky wrote: > rdivacky 2008-03-16 16:27:44 UTC > > FreeBSD src repository > > Modified files: > sys/i386/linux syscalls.master > Log: > Implement sched_setaffinity and get_setaffinity using > real cpu affinity setting primitives. > > Reviewed by: jeff > Approved by: kib (mentor) > > Revision Changes Path > 1.90 +2 -1 src/sys/i386/linux/syscalls.master sorry for omitting this in the first commit From pjd at FreeBSD.org Sun Mar 16 17:06:05 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sun Mar 16 17:06:10 2008 Subject: cvs commit: src/share/man/man9 rwlock.9 Message-ID: <200803161706.m2GH65fM000642@repoman.freebsd.org> pjd 2008-03-16 17:06:05 UTC FreeBSD src repository Modified files: share/man/man9 rwlock.9 Log: Add missing .El. Revision Changes Path 1.13 +1 -0 src/share/man/man9/rwlock.9 From pjd at FreeBSD.org Sun Mar 16 17:10:53 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sun Mar 16 17:10:58 2008 Subject: cvs commit: src/sys/sys rwlock.h src/share/man/man9 rwlock.9 Message-ID: <200803161710.m2GHAqmW000896@repoman.freebsd.org> pjd 2008-03-16 17:10:52 UTC FreeBSD src repository Modified files: sys/sys rwlock.h share/man/man9 rwlock.9 Log: Implement soon-to-be-used rw_unlock() macro. Revision Changes Path 1.14 +9 -1 src/share/man/man9/rwlock.9 1.16 +6 -0 src/sys/sys/rwlock.h From alc at FreeBSD.org Sun Mar 16 17:37:21 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sun Mar 16 17:37:26 2008 Subject: cvs commit: src/sys/vm vm_fault.c Message-ID: <200803161737.m2GHbK0V003422@repoman.freebsd.org> alc 2008-03-16 17:37:20 UTC FreeBSD src repository Modified files: sys/vm vm_fault.c Log: Simplify the inner loop of vm_fault()'s delete-behind heuristic. Instead of checking each page for PG_UNMANAGED, perform a one-time check whether the object is OBJT_PHYS. (PG_UNMANAGED pages only belong to OBJT_PHYS objects.) Revision Changes Path 1.241 +2 -2 src/sys/vm/vm_fault.c From phk at FreeBSD.org Sun Mar 16 17:54:55 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Sun Mar 16 17:54:57 2008 Subject: cvs commit: src/sbin/atacontrol atacontrol.c Message-ID: <200803161754.m2GHstl2003966@repoman.freebsd.org> phk 2008-03-16 17:54:55 UTC FreeBSD src repository Modified files: sbin/atacontrol atacontrol.c Log: Un-cut&paste argument processing, fix things lint found. Revision Changes Path 1.46 +67 -104 src/sbin/atacontrol/atacontrol.c From pjd at FreeBSD.org Sun Mar 16 17:55:07 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sun Mar 16 17:55:08 2008 Subject: cvs commit: src/sys/kern kern_resource.c Message-ID: <200803161755.m2GHt64v004045@repoman.freebsd.org> pjd 2008-03-16 17:55:06 UTC FreeBSD src repository Modified files: sys/kern kern_resource.c Log: Fix information leak. We can find PIDs of running processes from within a jail, etc. by simply calling setpriority(PRIO_PROCESS, , 0) and checking the return value: 0 means that the process exists and -1 that it doesn't exist. Reviewed by: rwatson MFC after: 1 week Revision Changes Path 1.184 +2 -1 src/sys/kern/kern_resource.c From pjd at FreeBSD.org Sun Mar 16 18:27:00 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sun Mar 16 18:27:02 2008 Subject: cvs commit: src/sys/kern kern_resource.c Message-ID: <200803161826.m2GIQxgN006347@repoman.freebsd.org> pjd 2008-03-16 18:26:59 UTC FreeBSD src repository Modified files: sys/kern kern_resource.c Log: Style fixes. Revision Changes Path 1.185 +7 -11 src/sys/kern/kern_resource.c From rwatson at FreeBSD.org Sun Mar 16 18:27:58 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:28:00 2008 Subject: cvs commit: src/sys/fs/coda coda_vnops.c Message-ID: <200803161827.m2GIRv3J006396@repoman.freebsd.org> rwatson 2008-03-16 18:27:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_vnops.c Log: Merge coda_vnops.c:1.97 from HEAD to RELENG_7: Reorder and clean up make_coda_node(), annotate weaknesses in the implementation. Revision Changes Path 1.76.2.13 +25 -20 src/sys/fs/coda/coda_vnops.c From rwatson at FreeBSD.org Sun Mar 16 18:28:38 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:28:43 2008 Subject: cvs commit: src/sys/fs/coda cnode.h coda_psdev.c Message-ID: <200803161828.m2GISbdH006438@repoman.freebsd.org> rwatson 2008-03-16 18:28:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda cnode.h coda_psdev.c Log: Merge cnode.h:1.28, coda_psdev.c:1.46 from HEAD to RELENG_7: Spell replys as replies. Revision Changes Path 1.21.2.7 +1 -1 src/sys/fs/coda/cnode.h 1.39.2.7 +7 -7 src/sys/fs/coda/coda_psdev.c From rwatson at FreeBSD.org Sun Mar 16 18:29:03 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:29:07 2008 Subject: cvs commit: src/sys/fs/coda coda_psdev.c Message-ID: <200803161829.m2GIT2pW006472@repoman.freebsd.org> rwatson 2008-03-16 18:29:02 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda coda_psdev.c Log: Merge coda_psdev.c:1.47 from HEAD to RELENG_7: The possibly interruptible msleep in coda_call() means well, but is fundamentally fairly confused about how signals work and when it is appropriate for upcalls to be interrupted. In particular, we should be exempting certain upcalls from interruption, we should not always eventually time out sleeping on a upcall, and we should not be interrupting the sleep for certain signals that we currently are (including SIGINFO). This code needs to be reworked in the style of NFS interruptible mounts. Revision Changes Path 1.39.2.8 +1 -1 src/sys/fs/coda/coda_psdev.c From rwatson at FreeBSD.org Sun Mar 16 18:52:20 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:52:24 2008 Subject: cvs commit: src/sys/ufs/ufs ufs_extattr.c Message-ID: <200803161852.m2GIqJPt007377@repoman.freebsd.org> rwatson 2008-03-16 18:52:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ufs/ufs ufs_extattr.c Log: Merge ufs_extattr.c:1.90 from HEAD to RELENG_7: Don't auto-start or allow extattrctl for UFS2 file systems, as UFS2 has native extended attributes. This didn't interfere with the operation of UFS2 extended attributes, but the code shouldn't be running for UFS2. Revision Changes Path 1.86.2.1 +18 -0 src/sys/ufs/ufs/ufs_extattr.c From rwatson at FreeBSD.org Sun Mar 16 18:53:18 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:53:19 2008 Subject: cvs commit: src/usr.bin/gzip gzip.c Message-ID: <200803161853.m2GIrH9P007420@repoman.freebsd.org> rwatson 2008-03-16 18:53:17 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/gzip gzip.c Log: Merge gzip.c:1.5 from HEAD to RELENG_7: When gzipping to a target file, only set the source file flags on the target file after the timestamp has been set; otherwise setting the timestamp will fail if the flags don't permit it (i.e., uchg). PR: 120208 Submitted by: Ighighi Revision Changes Path 1.4.2.2 +4 -4 src/usr.bin/gzip/gzip.c From gallatin at cs.duke.edu Sun Mar 16 18:53:29 2008 From: gallatin at cs.duke.edu (Andrew Gallatin) Date: Sun Mar 16 18:53:34 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... In-Reply-To: <200803141941.m2EJfmL2020579@repoman.freebsd.org>; from jhb@FreeBSD.org on Fri, Mar 14, 2008 at 07:41:25PM +0000 References: <200803141941.m2EJfmL2020579@repoman.freebsd.org> Message-ID: <20080316145259.A37148@grasshopper.cs.duke.edu> John Baldwin [jhb@FreeBSD.org] wrote: > MI code can currently (ab)use this by doing: > > intr_bind(rman_get_start(irq_res), cpu); > > however, I plan to add a truly MI interface (probably a bus_bind_intr(9)) Thank you very much for this! Do you plan to add a generic adminstrative interface to bind interrupts, or may I add a driver specific sysctl to bind mxge's interrupts in mxge? If you plan to add a generic administrative interface, I think we also need to add a way for drivers to label their interrupts so that an administrator can differentiate between the different MSI-X vectors. Thanks again, Drew From rwatson at FreeBSD.org Sun Mar 16 18:54:54 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:54:56 2008 Subject: cvs commit: src/bin/chflags chflags.1 chflags.c Message-ID: <200803161854.m2GIsrQr007462@repoman.freebsd.org> rwatson 2008-03-16 18:54:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) bin/chflags chflags.1 chflags.c Log: Merge chflags.1:1.30, chflags.c:1.24 from HEAD to RELENG_7: Make chflags(1) more chmod(1)-like (and more feature complete): - Add -v to print file names as they are processed; -vv prints the flags change as well. - Add -f to ignore failures with the same semantics as chflags(1), neither printing an error nor affecting the return code. - Don't try to set the flags if they won't change. I made minor cosmetic tweaks to the code in the patch. PR: 112827 Submitted by: Ighighi Revision Changes Path 1.29.2.1 +15 -1 src/bin/chflags/chflags.1 1.23.10.1 +28 -16 src/bin/chflags/chflags.c From rwatson at FreeBSD.org Sun Mar 16 18:55:57 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:56:02 2008 Subject: cvs commit: src/bin/realpath realpath.1 realpath.c Message-ID: <200803161855.m2GItuKO007586@repoman.freebsd.org> rwatson 2008-03-16 18:55:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) bin/realpath realpath.1 realpath.c Log: Merge realpath.1:1.11, realpath.c:1.7 from HEAD to RELENG_7: Enhance realpath(1) in a number of ways: - Allow realpath to accept multiple paths on the command line. - Add -q to suppress warnings if some paths can't be processed, and use getopt(3) to process flags. - Print the path being requested rather than a possibly partially processed path when a failure occurs so that you can tell which of several passed paths did fail. PR: 112920 Submitted by: Ighighi Revision Changes Path 1.10.10.1 +8 -0 src/bin/realpath/realpath.1 1.6.18.1 +26 -7 src/bin/realpath/realpath.c From rwatson at FreeBSD.org Sun Mar 16 18:59:22 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Mar 16 18:59:27 2008 Subject: cvs commit: src/sys/kern kern_synch.c Message-ID: <200803161859.m2GIxLaf007680@repoman.freebsd.org> rwatson 2008-03-16 18:59:21 UTC FreeBSD src repository Modified files: sys/kern kern_synch.c Log: Consistently use ANSI C declarationsfor all functions in kern_synch.c. Revision Changes Path 1.310 +7 -19 src/sys/kern/kern_synch.c From das at FreeBSD.org Sun Mar 16 19:08:54 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Mar 16 19:08:59 2008 Subject: cvs commit: src/lib/libc/gen scandir.c Message-ID: <200803161908.m2GJ8rvp009244@repoman.freebsd.org> das 2008-03-16 19:08:53 UTC FreeBSD src repository Modified files: lib/libc/gen scandir.c Log: scandir(3) previously used st_size to obtain an initial estimate of the array length needed to store all the directory entries. Although BSD has historically guaranteed that st_size is the size of the directory file, POSIX does not, and more to the point, some recent filesystems such as ZFS use st_size to mean something else. The fix is to not stat the directory at all, set the initial array size to 32 entries, and realloc it in powers of 2 if that proves insufficient. PR: 113668 Revision Changes Path 1.9 +3 -13 src/lib/libc/gen/scandir.c From thompsa at FreeBSD.org Sun Mar 16 19:25:31 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Sun Mar 16 19:25:32 2008 Subject: cvs commit: src/sys/net ieee8023ad_lacp.c ieee8023ad_lacp.h if_lagg.c Message-ID: <200803161925.m2GJPUNP013565@repoman.freebsd.org> thompsa 2008-03-16 19:25:30 UTC FreeBSD src repository Modified files: sys/net ieee8023ad_lacp.c ieee8023ad_lacp.h if_lagg.c Log: Switch the LACP state machine over to its own mutex to protect the internals, this means that it no longer grabs the lagg rwlock. Use two port table arrays which list the active ports for Tx and switch between them with an atomic op. Now the lagg rwlock is only exclusively locked for management (ioctls) and queuing of lacp control frames isnt needed. Revision Changes Path 1.15 +103 -124 src/sys/net/ieee8023ad_lacp.c 1.10 +47 -3 src/sys/net/ieee8023ad_lacp.h 1.27 +6 -3 src/sys/net/if_lagg.c From sam at freebsd.org Sun Mar 16 19:37:08 2008 From: sam at freebsd.org (Sam Leffler) Date: Sun Mar 16 19:37:15 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... In-Reply-To: <20080316145259.A37148@grasshopper.cs.duke.edu> References: <200803141941.m2EJfmL2020579@repoman.freebsd.org> <20080316145259.A37148@grasshopper.cs.duke.edu> Message-ID: <47DD7188.8000109@freebsd.org> Andrew Gallatin wrote: > John Baldwin [jhb@FreeBSD.org] wrote: > > >> MI code can currently (ab)use this by doing: >> >> intr_bind(rman_get_start(irq_res), cpu); >> >> however, I plan to add a truly MI interface (probably a bus_bind_intr(9)) >> > > Thank you very much for this! > > Do you plan to add a generic adminstrative interface to bind > interrupts, or may I add a driver specific sysctl to bind mxge's > interrupts in mxge? If you plan to add a generic administrative > interface, I think we also need to add a way for drivers to label > their interrupts so that an administrator can differentiate between > the different MSI-X vectors. > Any idea where this should go? Might be time to grow a tool that grok's the newbus hierarchy and pushes requests to devices. I've wanted functionality like netbsd has recently added to control power to devices and this would be seem to be similar. Not sure if we can do something that'd unify programs like atacontrol and camcontrol (or whether this is a good idea). Sam From kib at FreeBSD.org Sun Mar 16 19:55:17 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sun Mar 16 19:55:19 2008 Subject: cvs commit: src/sys/fs/msdosfs msdosfs_vfsops.c Message-ID: <200803161955.m2GJtHiO017666@repoman.freebsd.org> kib 2008-03-16 19:55:16 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/msdosfs msdosfs_vfsops.c Log: MFC rev. 1.186: Don't check the bpbSecPerTrack and bpbHeads fields of the BPB. Approved by: marcel Revision Changes Path 1.174.2.1 +7 -8 src/sys/fs/msdosfs/msdosfs_vfsops.c From pjd at FreeBSD.org Sun Mar 16 21:20:50 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sun Mar 16 21:20:55 2008 Subject: cvs commit: src/sys/amd64/include atomic.h src/sys/i386/include atomic.h src/sys/ia64/include atomic.h src/sys/powerpc/include atomic.h src/sys/sparc64/include atomic.h src/sys/sun4v/include atomic.h src/share/man/man9 atomic.9 Message-ID: <200803162120.m2GLKoCT063722@repoman.freebsd.org> pjd 2008-03-16 21:20:50 UTC FreeBSD src repository Modified files: sys/amd64/include atomic.h sys/i386/include atomic.h sys/ia64/include atomic.h sys/powerpc/include atomic.h sys/sparc64/include atomic.h sys/sun4v/include atomic.h share/man/man9 atomic.9 Log: Implement atomic_fetchadd_long() for all architectures and document it. Reviewed by: attilio, jhb, jeff, kris (as a part of the uidinfo_waitfree.patch) Revision Changes Path 1.14 +2 -1 src/share/man/man9/atomic.9 1.45 +20 -0 src/sys/amd64/include/atomic.h 1.48 +7 -0 src/sys/i386/include/atomic.h 1.13 +11 -0 src/sys/ia64/include/atomic.h 1.19 +2 -0 src/sys/powerpc/include/atomic.h 1.11 +2 -0 src/sys/sparc64/include/atomic.h 1.2 +2 -0 src/sys/sun4v/include/atomic.h From pjd at FreeBSD.org Sun Mar 16 21:29:03 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sun Mar 16 21:29:04 2008 Subject: cvs commit: src/sys/sys resourcevar.h src/sys/kern kern_resource.c Message-ID: <200803162129.m2GLT2FH064162@repoman.freebsd.org> pjd 2008-03-16 21:29:02 UTC FreeBSD src repository Modified files: sys/sys resourcevar.h sys/kern kern_resource.c Log: - Use wait-free method to manage ui_sbsize and ui_proccnt fields in the uidinfo structure. This entirely removes contention observed on the ui_mtxp mutex (as it is now gone). - Convert the uihashtbl_mtx mutex to a rwlock, as most of the time we just need to read-lock it. Reviewed by: jhb, jeff, kris & others Tested by: kris Revision Changes Path 1.186 +50 -60 src/sys/kern/kern_resource.c 1.53 +3 -7 src/sys/sys/resourcevar.h From xcllnt at mac.com Sun Mar 16 21:31:00 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Sun Mar 16 21:31:09 2008 Subject: cvs commit: src/sys/fs/msdosfs msdosfs_vfsops.c In-Reply-To: <200803161955.m2GJtHiO017666@repoman.freebsd.org> References: <200803161955.m2GJtHiO017666@repoman.freebsd.org> Message-ID: <065F5CD3-C05A-4AE9-8305-CE5D66D73301@mac.com> On Mar 16, 2008, at 12:55 PM, Konstantin Belousov wrote: > kib 2008-03-16 19:55:16 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7) > sys/fs/msdosfs msdosfs_vfsops.c > Log: > MFC rev. 1.186: > Don't check the bpbSecPerTrack and bpbHeads fields of the BPB. Thanks! -- Marcel Moolenaar xcllnt@mac.com From pjd at FreeBSD.org Sun Mar 16 21:32:21 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Sun Mar 16 21:32:26 2008 Subject: cvs commit: src/sys/kern kern_resource.c Message-ID: <200803162132.m2GLWLc1064464@repoman.freebsd.org> pjd 2008-03-16 21:32:21 UTC FreeBSD src repository Modified files: sys/kern kern_resource.c Log: Whitespace cleanups. Revision Changes Path 1.187 +7 -7 src/sys/kern/kern_resource.c From mav at FreeBSD.org Sun Mar 16 21:33:13 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 16 21:33:15 2008 Subject: cvs commit: src/sys/netgraph ng_l2tp.c Message-ID: <200803162133.m2GLXD0T064580@repoman.freebsd.org> mav 2008-03-16 21:33:12 UTC FreeBSD src repository Modified files: sys/netgraph ng_l2tp.c Log: Add session ID hashing to speedup incoming packets dispatch in case of many connections working via the same tunnel. For example, in case of full "client <-> LAC <-> LNS" setup. Revision Changes Path 1.25 +47 -38 src/sys/netgraph/ng_l2tp.c From brueffer at FreeBSD.org Sun Mar 16 21:36:06 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sun Mar 16 21:36:10 2008 Subject: cvs commit: src/usr.sbin/adduser adduser.8 Message-ID: <200803162136.m2GLa543064894@repoman.freebsd.org> brueffer 2008-03-16 21:36:05 UTC FreeBSD src repository Modified files: usr.sbin/adduser adduser.8 Log: In the description of the password field, -w was meant, not the nonexistant -p flag. PR: 120122 Submitted by: Andy Kosela MFC after: 3 days Revision Changes Path 1.62 +2 -2 src/usr.sbin/adduser/adduser.8 From mav at FreeBSD.org Sun Mar 16 22:06:07 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 16 22:06:09 2008 Subject: cvs commit: src/sys/netgraph netgraph.h ng_base.c ng_socket.c Message-ID: <200803162206.m2GM67RC070936@repoman.freebsd.org> mav 2008-03-16 22:06:07 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph netgraph.h ng_base.c ng_socket.c Log: MFC: Improve apply callback error reporting: Before this patch callback returned result of the last finished call chain. Now it returns last nonzero result from all call chain results in this request. As soon as this improvement gives reliable error reporting, it is now possible to remove dirty workaround in ng_socket, made to return ENOBUFS error statuses of request-response operations. That workaround was responsible for returning ENOBUFS errors to completely unrelated requests working at the same time on socket. Revision Changes Path 1.68.2.4 +2 -1 src/sys/netgraph/netgraph.h 1.135.2.6 +25 -8 src/sys/netgraph/ng_base.c 1.82.2.2 +1 -1 src/sys/netgraph/ng_socket.c From brueffer at FreeBSD.org Sun Mar 16 22:26:18 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sun Mar 16 22:26:20 2008 Subject: cvs commit: src/share/man/man4/man4.i386 ar.4 Message-ID: <200803162226.m2GMQHxY071733@repoman.freebsd.org> brueffer 2008-03-16 22:26:17 UTC FreeBSD src repository Modified files: share/man/man4/man4.i386 ar.4 Log: Device counts are long gone, also remove the outdated FILES section. MFC after: 3 days Revision Changes Path 1.29 +2 -11 src/share/man/man4/man4.i386/ar.4 From marcel at FreeBSD.org Sun Mar 16 22:51:29 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Mar 16 22:51:34 2008 Subject: cvs commit: src/sys/conf NOTES files options src/sys/geom geom_bsd_enc.c src/sys/geom/part g_part.c g_part.h g_part_bsd.c g_part_mbr.c g_part_vtoc8.c src/sys/sys vtoc.h Message-ID: <200803162251.m2GMpTAY072945@repoman.freebsd.org> marcel 2008-03-16 22:51:29 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf NOTES files options sys/geom geom_bsd_enc.c sys/geom/part g_part.c g_part.h g_part_mbr.c Added files: (Branch: RELENG_7) sys/geom/part g_part_bsd.c g_part_vtoc8.c sys/sys vtoc.h Log: Sync gpart with HEAD: o Add support for BSD disklabels (option GEOM_PART_BSD) o Add support for VTOC8 disklabels (option GEOM_PART_VTOC8) o Avoid false positives detecting MBRs Revision Changes Path 1.1454.2.2 +2 -0 src/sys/conf/NOTES 1.1243.2.6 +2 -0 src/sys/conf/files 1.608.2.1 +2 -0 src/sys/conf/options 1.6.10.1 +1 -1 src/sys/geom/geom_bsd_enc.c 1.9.2.3 +51 -19 src/sys/geom/part/g_part.c 1.3.2.3 +8 -0 src/sys/geom/part/g_part.h 1.3.2.1 +431 -0 src/sys/geom/part/g_part_bsd.c (new) 1.2.2.1 +53 -6 src/sys/geom/part/g_part_mbr.c 1.1.2.1 +480 -0 src/sys/geom/part/g_part_vtoc8.c (new) 1.1.2.1 +106 -0 src/sys/sys/vtoc.h (new) From brueffer at FreeBSD.org Sun Mar 16 22:51:31 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Sun Mar 16 22:51:36 2008 Subject: cvs commit: src/share/man/man4/man4.i386 pnp.4 Message-ID: <200803162251.m2GMpU35072972@repoman.freebsd.org> brueffer 2008-03-16 22:51:30 UTC FreeBSD src repository Modified files: share/man/man4/man4.i386 pnp.4 Log: Fix AUTHORS formatting. MFC after: 3 days Revision Changes Path 1.18 +1 -0 src/share/man/man4/man4.i386/pnp.4 From phk at FreeBSD.org Sun Mar 16 23:00:44 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Sun Mar 16 23:00:47 2008 Subject: cvs commit: src/usr.sbin/fifolog/lib miniobj.h Message-ID: <200803162300.m2GN0ikq074355@repoman.freebsd.org> phk 2008-03-16 23:00:43 UTC FreeBSD src repository Modified files: usr.sbin/fifolog/lib miniobj.h Log: Remove trailing ';' on macros. Spotted by: antoine Revision Changes Path 1.2 +2 -2 src/usr.sbin/fifolog/lib/miniobj.h From imp at bsdimp.com Sun Mar 16 23:10:26 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Sun Mar 16 23:10:33 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... In-Reply-To: <47DD7188.8000109@freebsd.org> References: <200803141941.m2EJfmL2020579@repoman.freebsd.org> <20080316145259.A37148@grasshopper.cs.duke.edu> <47DD7188.8000109@freebsd.org> Message-ID: <20080316.170935.-713546699.imp@bsdimp.com> In message: <47DD7188.8000109@freebsd.org> Sam Leffler writes: : Andrew Gallatin wrote: : > John Baldwin [jhb@FreeBSD.org] wrote: : > : > : >> MI code can currently (ab)use this by doing: : >> : >> intr_bind(rman_get_start(irq_res), cpu); : >> : >> however, I plan to add a truly MI interface (probably a bus_bind_intr(9)) : >> : > : > Thank you very much for this! : > : > Do you plan to add a generic adminstrative interface to bind : > interrupts, or may I add a driver specific sysctl to bind mxge's : > interrupts in mxge? If you plan to add a generic administrative : > interface, I think we also need to add a way for drivers to label : > their interrupts so that an administrator can differentiate between : > the different MSI-X vectors. : > : : Any idea where this should go? Might be time to grow a tool that grok's : the newbus hierarchy and pushes requests to devices. I've wanted : functionality like netbsd has recently added to control power to devices : and this would be seem to be similar. Not sure if we can do something : that'd unify programs like atacontrol and camcontrol (or whether this is : a good idea). I've wanted to maek devctl grow this sort of thing for a long time, but never had time to implement this. It wouldn't be hard, just a bit tedious. Also, there's two types of requests that would need to be handled. One is where you tell a device to do something directly. The other is where you tell its parent to do it to the child (and maybe do some cleanup afterwards). Warner From mav at FreeBSD.org Sun Mar 16 23:12:18 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Mar 16 23:12:20 2008 Subject: cvs commit: src/sys/netgraph ng_base.c Message-ID: <200803162312.m2GNCHWk077683@repoman.freebsd.org> mav 2008-03-16 23:12:17 UTC FreeBSD src repository Modified files: sys/netgraph ng_base.c Log: Remove impossible (hk_peer == NULL) check from ng_address_hook(). Valid hook can't have NULL peer. Even invalid one can't, as it is resets to deadhook, but not NULL. Revision Changes Path 1.151 +0 -1 src/sys/netgraph/ng_base.c From phk at FreeBSD.org Mon Mar 17 00:36:17 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Mar 17 00:36:19 2008 Subject: cvs commit: src/sys/i386/cpufreq p4tcc.c Message-ID: <200803170036.m2H0aG9S082520@repoman.freebsd.org> phk 2008-03-17 00:36:16 UTC FreeBSD src repository Modified files: sys/i386/cpufreq p4tcc.c Log: Use correct bitmask for identifying chip family. Revision Changes Path 1.13 +1 -1 src/sys/i386/cpufreq/p4tcc.c From marcel at FreeBSD.org Mon Mar 17 00:46:54 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Mar 17 00:46:57 2008 Subject: cvs commit: src/sys/powerpc/include gdb_machdep.h src/sys/powerpc/powerpc gdb_machdep.c Message-ID: <200803170046.m2H0kr7L082880@repoman.freebsd.org> marcel 2008-03-17 00:46:52 UTC FreeBSD src repository Modified files: sys/powerpc/include gdb_machdep.h sys/powerpc/powerpc gdb_machdep.c Log: Make remote GDB work for AIM processors. For BookE, the kernel will have a special section, named .PPC.EMB.apuinfo, which will tell GDB that a BookE processor is targeted and which will result in GDB using a different register definition. In order to support remote GDB for BookE, we need the GDB stub in the kernel look for that section and use the BookE definitions. Revision Changes Path 1.2 +15 -3 src/sys/powerpc/include/gdb_machdep.h 1.3 +12 -40 src/sys/powerpc/powerpc/gdb_machdep.c From thompsa at FreeBSD.org Mon Mar 17 01:26:45 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Mon Mar 17 01:26:50 2008 Subject: cvs commit: src/sys/net ieee8023ad_lacp.h Message-ID: <200803170126.m2H1QjcS094361@repoman.freebsd.org> thompsa 2008-03-17 01:26:45 UTC FreeBSD src repository Modified files: sys/net ieee8023ad_lacp.h Log: Remove extra semicolons. Pointed out by: antoine Revision Changes Path 1.11 +2 -2 src/sys/net/ieee8023ad_lacp.h From weongyo at FreeBSD.org Mon Mar 17 02:30:14 2008 From: weongyo at FreeBSD.org (Weongyo Jeong) Date: Mon Mar 17 02:30:18 2008 Subject: cvs commit: src/sys/dev/usb if_zyd.c Message-ID: <200803170230.m2H2UDq9000864@repoman.freebsd.org> weongyo 2008-03-17 02:30:13 UTC FreeBSD src repository Modified files: sys/dev/usb if_zyd.c Log: don't set sniffer mode to ON when the driver is running with the monitor mode. This solves a problem that sometimes mangled frames are passed. Submitted by: Werner Backes Tested by: Werner Backes PR: kern/121608 Approved by: thompsa (mentor) Revision Changes Path 1.13 +1 -2 src/sys/dev/usb/if_zyd.c From sobomax at FreeBSD.org Mon Mar 17 04:01:47 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Mon Mar 17 04:01:52 2008 Subject: cvs commit: src/sys/sys eventhandler.h src/sys/kern kern_descrip.c uipc_socket.c In-Reply-To: <20080316100836.F44049@fledge.watson.org> References: <200803160621.m2G6LUlE034766@repoman.freebsd.org> <20080316100836.F44049@fledge.watson.org> Message-ID: <47DDED1F.8000707@FreeBSD.org> Robert Watson wrote: > On Sun, 16 Mar 2008, Maxim Sobolev wrote: > >> sobomax 2008-03-16 06:21:30 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/sys eventhandler.h >> sys/kern kern_descrip.c uipc_socket.c >> Log: >> Properly set size of the file_zone to match kern.maxfiles parameter. >> Otherwise the parameter is no-op, since zone by default limits number >> of descriptors to some 12K entries. Attempt to allocate more ends up >> sleeping on zonelimit. > > Hmm. Could you be more specific about the nature of the problem you're > seeing here? UMA shouldn't impose any default zone limit on the > file_zone if none is set. Are you sure it's the struct file zonelimit > you're running into, and not, say, the socket zone limit, etc? Yes, you are right, further investigation revealed that in my case I was hitting maxsockets limit, not maxfiles limit. Not sure if the backout is in order or having files zone limit has some value, so that possible leaks can be detected and UMA zone can potentially do size-based optimizations in the future? > Also, don't use SI_SUB_EVENTHANDLER+1 -- add a new event type, perhaps > SI_SUB_FILE, or use SI_ORDER_LAST. OK, what about maxsockets code? Should it be changed to? -Maxim From obrien at FreeBSD.org Mon Mar 17 07:57:39 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Mon Mar 17 07:57:43 2008 Subject: cvs commit: src/contrib/cvs/lib getdate Message-ID: <200803170757.m2H7vc7v031682@repoman.freebsd.org> obrien 2008-03-17 07:57:38 UTC FreeBSD src repository Removed files: contrib/cvs/lib getdate Log: egad! The cvs 1.11.22 distribution had a Linux binary in it. Yuck. Revision Changes Path 1.2 +0 -50 src/contrib/cvs/lib/getdate (dead) From brueffer at FreeBSD.org Mon Mar 17 08:06:42 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Mon Mar 17 08:06:44 2008 Subject: cvs commit: src/share/man/man4/man4.i386 scd.4 Message-ID: <200803170806.m2H86gEY033332@repoman.freebsd.org> brueffer 2008-03-17 08:06:42 UTC FreeBSD src repository Modified files: share/man/man4/man4.i386 scd.4 Log: Device counts are long gone. MFC after: 3 days Revision Changes Path 1.23 +2 -2 src/share/man/man4/man4.i386/scd.4 From brueffer at FreeBSD.org Mon Mar 17 08:17:05 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Mon Mar 17 08:17:38 2008 Subject: cvs commit: src/share/man/man4/man4.i386 linux.4 streams.4 svr4.4 Message-ID: <200803170817.m2H8H4HV033680@repoman.freebsd.org> brueffer 2008-03-17 08:17:04 UTC FreeBSD src repository Modified files: share/man/man4/man4.i386 linux.4 streams.4 svr4.4 Log: Use a standard section 4 SYNOPSIS. MFC after: 3 days Revision Changes Path 1.13 +13 -4 src/share/man/man4/man4.i386/linux.4 1.8 +14 -8 src/share/man/man4/man4.i386/streams.4 1.13 +14 -5 src/share/man/man4/man4.i386/svr4.4 From phk at FreeBSD.org Mon Mar 17 08:38:39 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Mar 17 08:38:40 2008 Subject: cvs commit: src/sys/i386/cpufreq p4tcc.c Message-ID: <200803170838.m2H8ccTs034442@repoman.freebsd.org> phk 2008-03-17 08:38:38 UTC FreeBSD src repository Modified files: sys/i386/cpufreq p4tcc.c Log: Increase time we wait for things to settle to 1 millisecond, 10 microseconds is too short. Always set the cpu to the highest frequency so that we get through boot and don't handicap cpus where powerd(8) is not used. Revision Changes Path 1.14 +26 -5 src/sys/i386/cpufreq/p4tcc.c From phk at FreeBSD.org Mon Mar 17 09:01:00 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Mar 17 09:01:24 2008 Subject: cvs commit: src/sys/i386/cpufreq p4tcc.c Message-ID: <200803170900.m2H90xfu044568@repoman.freebsd.org> phk 2008-03-17 09:00:59 UTC FreeBSD src repository Modified files: sys/i386/cpufreq p4tcc.c Log: Revert last commit and stop committing before morning tea. Revision Changes Path 1.15 +5 -26 src/sys/i386/cpufreq/p4tcc.c From phk at FreeBSD.org Mon Mar 17 09:01:43 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Mar 17 09:01:46 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803170901.m2H91hCC044632@repoman.freebsd.org> phk 2008-03-17 09:01:43 UTC FreeBSD src repository Modified files: sys/i386/cpufreq est.c Log: Increase time we wait for things to settle to 1 millisecond, 10 microseconds is too short. Always set the cpu to the highest frequency so that we get through boot and don't handicap cpus where powerd(8) is not used. Revision Changes Path 1.15 +16 -2 src/sys/i386/cpufreq/est.c From phk at FreeBSD.org Mon Mar 17 09:05:16 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Mar 17 09:05:22 2008 Subject: cvs commit: src/sys/i386/cpufreq p4tcc.c Message-ID: <200803170905.m2H95F8f046020@repoman.freebsd.org> phk 2008-03-17 09:05:15 UTC FreeBSD src repository Modified files: sys/i386/cpufreq p4tcc.c Log: A cautionary XXX comment about seemingly bogus errata checks. Revision Changes Path 1.16 +8 -0 src/sys/i386/cpufreq/p4tcc.c From phk at FreeBSD.org Mon Mar 17 10:33:24 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Mar 17 10:33:32 2008 Subject: cvs commit: src/sys/sys ata.h src/sbin/atacontrol atacontrol.8 atacontrol.c src/sys/dev/ata ata-all.c ata-all.h ata-disk.c ata-disk.h Message-ID: <200803171033.m2HAXOeN055116@repoman.freebsd.org> phk 2008-03-17 10:33:23 UTC FreeBSD src repository Modified files: sys/sys ata.h sbin/atacontrol atacontrol.8 atacontrol.c sys/dev/ata ata-all.c ata-all.h ata-disk.c ata-disk.h Log: Add a "spindown" facility to ata-disks: If no requests have been received for a configurable number of seconds, spin the disk down. Spin it back up on the next request. Notice that the timeout is only armed by a request, so to spin down a disk you may have to do: atacontrol spindown ad10 5 dd if=/dev/ad10 of=/dev/null count=1 To disable spindown, set timeout to zero: atacontrol spindown ad10 0 In order to debug any trouble caused, this code is somewhat noisy on the console. Enabling spindown on a disk containing / or /var/log/messages is not going to do anything sensible. Spinning a disk up and down all the time will wear it out, use sensibly. Approved by: sos Revision Changes Path 1.30 +8 -0 src/sbin/atacontrol/atacontrol.8 1.47 +27 -0 src/sbin/atacontrol/atacontrol.c 1.281 +6 -0 src/sys/dev/ata/ata-all.c 1.128 +3 -0 src/sys/dev/ata/ata-all.h 1.207 +49 -1 src/sys/dev/ata/ata-disk.c 1.53 +1 -1 src/sys/dev/ata/ata-disk.h 1.39 +3 -0 src/sys/sys/ata.h From pjd at FreeBSD.org Mon Mar 17 11:48:41 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Mon Mar 17 11:48:45 2008 Subject: cvs commit: src/sys/kern subr_witness.c Message-ID: <200803171148.m2HBmeWP061368@repoman.freebsd.org> pjd 2008-03-17 11:48:40 UTC FreeBSD src repository Modified files: sys/kern subr_witness.c Log: - There is no more "uidinfo struct" mutex. - The "uidinfo hash" lock is now a rwlock. Reminded by: kib Revision Changes Path 1.243 +1 -2 src/sys/kern/subr_witness.c From jroberson at chesapeake.net Mon Mar 17 12:00:47 2008 From: jroberson at chesapeake.net (Jeff Roberson) Date: Mon Mar 17 12:00:58 2008 Subject: cvs commit: src/sys/sys eventhandler.h src/sys/kern kern_descrip.c uipc_socket.c In-Reply-To: <47DDED1F.8000707@FreeBSD.org> References: <200803160621.m2G6LUlE034766@repoman.freebsd.org> <20080316100836.F44049@fledge.watson.org> <47DDED1F.8000707@FreeBSD.org> Message-ID: <20080317020036.M910@desktop> On Sun, 16 Mar 2008, Maxim Sobolev wrote: > Robert Watson wrote: >> On Sun, 16 Mar 2008, Maxim Sobolev wrote: >> >>> sobomax 2008-03-16 06:21:30 UTC >>> >>> FreeBSD src repository >>> >>> Modified files: >>> sys/sys eventhandler.h >>> sys/kern kern_descrip.c uipc_socket.c >>> Log: >>> Properly set size of the file_zone to match kern.maxfiles parameter. >>> Otherwise the parameter is no-op, since zone by default limits number >>> of descriptors to some 12K entries. Attempt to allocate more ends up >>> sleeping on zonelimit. >> >> Hmm. Could you be more specific about the nature of the problem you're >> seeing here? UMA shouldn't impose any default zone limit on the file_zone >> if none is set. Are you sure it's the struct file zonelimit you're running >> into, and not, say, the socket zone limit, etc? > > Yes, you are right, further investigation revealed that in my case I was > hitting maxsockets limit, not maxfiles limit. Not sure if the backout is in > order or having files zone limit has some value, so that possible leaks can > be detected and UMA zone can potentially do size-based optimizations in the > future? The UMA max is complicated due to per-cpu caches. Really the limit code in uma should be rewritten or all consumers should handle it themselves the way struct file does. It'd probably be best if you backed this out. Thanks, Jeff > >> Also, don't use SI_SUB_EVENTHANDLER+1 -- add a new event type, perhaps >> SI_SUB_FILE, or use SI_ORDER_LAST. > > OK, what about maxsockets code? Should it be changed to? > > -Maxim > From rwatson at FreeBSD.org Mon Mar 17 13:04:56 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 17 13:05:05 2008 Subject: cvs commit: src/sys/netinet in_pcb.c Message-ID: <200803171304.m2HD4uDQ078506@repoman.freebsd.org> rwatson 2008-03-17 13:04:56 UTC FreeBSD src repository Modified files: sys/netinet in_pcb.c Log: Fix indentation for a closing brace in in_pcballoc(). MFC after: 3 days Revision Changes Path 1.200 +1 -1 src/sys/netinet/in_pcb.c From kib at FreeBSD.org Mon Mar 17 13:17:10 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Mon Mar 17 13:17:16 2008 Subject: cvs commit: src/sys/kern kern_conf.c src/sys/sys conf.h Message-ID: <200803171317.m2HDHAZ9079210@repoman.freebsd.org> kib 2008-03-17 13:17:10 UTC FreeBSD src repository Modified files: sys/kern kern_conf.c sys/sys conf.h Log: Fix two races in the handling of the d_gianttrick for the D_NEEDGIANT drivers. In the giant_XXX wrappers for the device methods of the D_NEEDGIANT drivers, do not dereference the cdev->si_devsw. It is racing with the destroy_devl() clearing of the si_devsw. Instead, use the dev_refthread() and return ENXIO for the destroyed device. [1] The check for the D_INIT in the prep_cdevsw() was not synchronized with the call of the fini_cdevsw() in destroy_devl(), that under rapid device creation/destruction may result in the use of uninitialized cdevsw [2]. Change the protocol for the prep_cdevsw(), requiring it to be called under dev_mtx, where the check for D_INIT is done. Do not free the memory allocated for the gianttrick cdevsw while holding the dev_mtx, put it into the free list to be freed later. Reuse the d_gianttrick pointer to keep the size and layout of the struct cdevsw (requested by phk). Free the memory in the dev_unlock_and_free(), and do all the free after the dev_mtx is dropped (suggested by jhb). Reported by: bsdimp + many [1], pho [2] Reviewed by: phk, jhb Tested by: pho MFC after: 1 week Revision Changes Path 1.210 +121 -35 src/sys/kern/kern_conf.c 1.236 +6 -1 src/sys/sys/conf.h From jhb at freebsd.org Mon Mar 17 13:55:30 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 17 13:55:43 2008 Subject: cvs commit: src/sys/sys rwlock.h src/share/man/man9 rwlock.9 In-Reply-To: <200803161710.m2GHAqmW000896@repoman.freebsd.org> References: <200803161710.m2GHAqmW000896@repoman.freebsd.org> Message-ID: <200803170931.10170.jhb@freebsd.org> On Sunday 16 March 2008 01:10:52 pm Pawel Jakub Dawidek wrote: > pjd 2008-03-16 17:10:52 UTC > > FreeBSD src repository > > Modified files: > sys/sys rwlock.h > share/man/man9 rwlock.9 > Log: > Implement soon-to-be-used rw_unlock() macro. Which I hate. :( Programmers should know if a lock is read-locked or write-locked. -- John Baldwin From jhb at freebsd.org Mon Mar 17 13:55:31 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 17 13:55:45 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <200803170901.m2H91hCC044632@repoman.freebsd.org> References: <200803170901.m2H91hCC044632@repoman.freebsd.org> Message-ID: <200803170933.48212.jhb@freebsd.org> On Monday 17 March 2008 05:01:43 am Poul-Henning Kamp wrote: > phk 2008-03-17 09:01:43 UTC > > FreeBSD src repository > > Modified files: > sys/i386/cpufreq est.c > Log: > Increase time we wait for things to settle to 1 millisecond, > 10 microseconds is too short. > > Always set the cpu to the highest frequency so that we get through > boot and don't handicap cpus where powerd(8) is not used. Hmm, I actually consider this a feature when I'm not running powerd to use less battery. I think we should only bump up the CPU on battery power when using powerd so that it can be lowered again to save battery power when the CPU is idle. -- John Baldwin From jhb at freebsd.org Mon Mar 17 13:55:33 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 17 13:55:56 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... In-Reply-To: <20080316145259.A37148@grasshopper.cs.duke.edu> References: <200803141941.m2EJfmL2020579@repoman.freebsd.org> <20080316145259.A37148@grasshopper.cs.duke.edu> Message-ID: <200803170952.42262.jhb@freebsd.org> On Sunday 16 March 2008 02:52:59 pm Andrew Gallatin wrote: > John Baldwin [jhb@FreeBSD.org] wrote: > > MI code can currently (ab)use this by doing: > > > > intr_bind(rman_get_start(irq_res), cpu); > > > > however, I plan to add a truly MI interface (probably a > > bus_bind_intr(9)) > > Thank you very much for this! > > Do you plan to add a generic adminstrative interface to bind > interrupts, or may I add a driver specific sysctl to bind mxge's > interrupts in mxge? If you plan to add a generic administrative > interface, I think we also need to add a way for drivers to label > their interrupts so that an administrator can differentiate between > the different MSI-X vectors. I already have in my tree an amd64/i386-specific sysarch request to bind an IRQ to a CPU, but it's sort of a pain to use since we don't expose interrupt info to userland very well so the sysadmin can't tell which CPU an IRQ is already connected to w/o a verbose dmesg. The first thing I plan to implement is the aforementioned bus_bind_intr(9) for device drivers to use. I'm not sure what the sysadmin interface will look like. Probably the ideal is to make the UI do what my existing little tool does 'ibind '. However, the MI code doesn't actually know anything about IRQ values. Probably would need some sort of MD helper routine to that takes the requested 'IRQ' value and maps it to an interrupt event. Also, the code does not currently differentiate between a userland (administrative) bind request vs. a driver bind request. My guess is that the administrative requests should have precedence over the driver request. I also think driver requests need to fail with EBUSY like they do now if the interrupt is already bound whereas administrative requests should maybe fail with EBUSY if the driver has bound it unless the sysadmin uses a force option (ibind -f or some such). -- John Baldwin From phk at phk.freebsd.dk Mon Mar 17 14:02:16 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Mon Mar 17 14:02:19 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: Your message of "Mon, 17 Mar 2008 09:33:47 -0400." <200803170933.48212.jhb@freebsd.org> Message-ID: <3709.1205762533@critter.freebsd.dk> In message <200803170933.48212.jhb@freebsd.org>, John Baldwin writes: >On Monday 17 March 2008 05:01:43 am Poul-Henning Kamp wrote: >> phk 2008-03-17 09:01:43 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/i386/cpufreq est.c >> Log: >> Increase time we wait for things to settle to 1 millisecond, >> 10 microseconds is too short. >> >> Always set the cpu to the highest frequency so that we get through >> boot and don't handicap cpus where powerd(8) is not used. > >Hmm, I actually consider this a feature when I'm not running powerd to use >less battery. I think we should only bump up the CPU on battery power when >using powerd so that it can be lowered again to save battery power when the >CPU is idle. We have cpufreq enabled by default now, badly configured machines run at 50% of rated CPU power because people don't know that they need to enable powerd(8) on servers. This is only going to get worse when more EnergyStar compliant servers hit the channel. I think setting full speed is the correct choice, if people care about powersaving, they need to configured it, if they don't they should get their moneys worth out of their hardware. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. From rwatson at FreeBSD.org Mon Mar 17 14:18:49 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Mar 17 14:18:53 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <3709.1205762533@critter.freebsd.dk> References: <3709.1205762533@critter.freebsd.dk> Message-ID: <20080317141717.U3253@fledge.watson.org> On Mon, 17 Mar 2008, Poul-Henning Kamp wrote: > In message <200803170933.48212.jhb@freebsd.org>, John Baldwin writes: > >> Hmm, I actually consider this a feature when I'm not running powerd to use >> less battery. I think we should only bump up the CPU on battery power when >> using powerd so that it can be lowered again to save battery power when the >> CPU is idle. > > We have cpufreq enabled by default now, badly configured machines run at 50% > of rated CPU power because people don't know that they need to enable > powerd(8) on servers. > > This is only going to get worse when more EnergyStar compliant servers hit > the channel. > > I think setting full speed is the correct choice, if people care about > powersaving, they need to configured it, if they don't they should get their > moneys worth out of their hardware. If cpufreq is going to be enabled by default, should we be enabling powerd by default, or at least having a powerd_enable="AUTO" that detects the appropriate frobs and feedback sources and turns on powerd if it's going to be useful? There might be a reasonable argument to be made that in two of the three computing environments of choice for FreeBSD (notebooks, servers in colos), power management is a basic assumption and we should turn on the necessary bits to deal with it. Robert N M Watson Computer Laboratory University of Cambridge From gallatin at cs.duke.edu Mon Mar 17 14:34:07 2008 From: gallatin at cs.duke.edu (Andrew Gallatin) Date: Mon Mar 17 14:34:12 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... In-Reply-To: <200803170952.42262.jhb@freebsd.org> References: <200803141941.m2EJfmL2020579@repoman.freebsd.org> <20080316145259.A37148@grasshopper.cs.duke.edu> <200803170952.42262.jhb@freebsd.org> Message-ID: <18398.33113.488394.494142@grasshopper.cs.duke.edu> John Baldwin writes: > > I already have in my tree an amd64/i386-specific sysarch request to bind an > IRQ to a CPU, but it's sort of a pain to use since we don't expose interrupt > info to userland very well so the sysadmin can't tell which CPU an IRQ is > already connected to w/o a verbose dmesg. The first thing I plan to > implement is the aforementioned bus_bind_intr(9) for device drivers to use. More than that, an admin cannot tell what MSI-X vector corresponds to what functionality. For example, on linux, the driver can attach a custom label to the irq information seen from cat /proc/irq. Hence my linux 10GbE driver attaches meaninful labels to each MSI-X vector it uses. Eg, for 2 "slices" (what we call qsets), we have this on Linux: % grep eth0 /proc/interrupts 510: 0 0 PCI-MSI-edge eth0:slice-1 511: 0 5 PCI-MSI-edge eth0:slice-0 But on FreeBSD: % vmstat -i | grep mxge irq256: mxge0 318254 74 irq257: mxge0 311469 73 So I'd like the driver to be able to append a custom string (":slice-%d") to the label used by vmstat -i so the adminstrator can know what he's binding. > I'm not sure what the sysadmin interface will look like. Probably the ideal > is to make the UI do what my existing little tool does 'ibind '. > However, the MI code doesn't actually know anything about IRQ values. > Probably would need some sort of MD helper routine to that takes the > requested 'IRQ' value and maps it to an interrupt event. Also, the code does > not currently differentiate between a userland (administrative) bind request > vs. a driver bind request. My guess is that the administrative requests > should have precedence over the driver request. I also think driver requests > need to fail with EBUSY like they do now if the interrupt is already bound > whereas administrative requests should maybe fail with EBUSY if the driver > has bound it unless the sysadmin uses a force option (ibind -f or some such). This sounds reasonable. Thanks again, Drew From pjd at FreeBSD.org Mon Mar 17 14:46:31 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Mon Mar 17 14:47:06 2008 Subject: cvs commit: src/tools/regression/bin/mv regress.sh In-Reply-To: <20071218100355.GR16982@elvis.mu.org> References: <200712180849.lBI8nmEi088947@repoman.freebsd.org> <20071218100355.GR16982@elvis.mu.org> Message-ID: <20080314223652.GA20470@garage.freebsd.pl> On Tue, Dec 18, 2007 at 02:03:55AM -0800, Alfred Perlstein wrote: > * Diomidis Spinellis [071218 00:48] wrote: > > dds 2007-12-18 08:49:47 UTC > > > > FreeBSD src repository > > > > Modified files: > > tools/regression/bin/mv regress.sh > > Log: > > Add more tests. All rename(2)-based tests now succeed. > > The performance of the cross-device equivalents is under investigation. > > Diomidis, > > Does 'mv' spawn off tar or something to do the copy part? > > You can _usually_ get much better cross device performance > by hooking two processes together like so: > > tar -cf - -C /path/to/source . | tar -xf - -C /path/to/dst > > This will keep the disks a lot busier, but this can make > things worse on cross device moves that happen to be > on the same disk. I was thinking about adding two options to cp(1) (-1 and -2) to give cp(1) a hint if the copy is done inside one disk or between separate disks. In -1 case cp(1) will read as large blocks as possible and then write them, in -2 case it will spawn two threads: one reader and one writer working in parallel. Performance improvements are very visible from what I tested. -- Pawel Jakub Dawidek http://www.wheel.pl pjd@FreeBSD.org http://www.FreeBSD.org FreeBSD committer Am I Evil? Yes, I Am! -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 187 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080317/5cb8c614/attachment.pgp From kientzle at freebsd.org Mon Mar 17 15:19:11 2008 From: kientzle at freebsd.org (Tim Kientzle) Date: Mon Mar 17 15:19:15 2008 Subject: cvs commit: src/tools/regression/bin/mv regress.sh In-Reply-To: <20080314223652.GA20470@garage.freebsd.pl> References: <200712180849.lBI8nmEi088947@repoman.freebsd.org> <20071218100355.GR16982@elvis.mu.org> <20080314223652.GA20470@garage.freebsd.pl> Message-ID: <47DE8BEE.7040405@freebsd.org> >>> FreeBSD src repository >>> >>> Modified files: >>> tools/regression/bin/mv regress.sh >>> Log: >>> Add more tests. All rename(2)-based tests now succeed. >>> The performance of the cross-device equivalents is under investigation. > > I was thinking about adding two options to cp(1) (-1 and -2) to give > cp(1) a hint if the copy is done inside one disk or between separate > disks. Will anyone ever use such a flag? I'm skeptical. If there were a way to determine this programmatically, it would be a big win. You'd need to look at the first file copied to determine if the destination was on the same (physical) disk or not. Tim From dds at aueb.gr Mon Mar 17 16:12:42 2008 From: dds at aueb.gr (Diomidis Spinellis) Date: Mon Mar 17 16:12:55 2008 Subject: cvs commit: src/tools/regression/bin/mv regress.sh In-Reply-To: <20080314223652.GA20470@garage.freebsd.pl> References: <200712180849.lBI8nmEi088947@repoman.freebsd.org> <20071218100355.GR16982@elvis.mu.org> <20080314223652.GA20470@garage.freebsd.pl> Message-ID: <47DE8D10.9090201@aueb.gr> Pawel Jakub Dawidek wrote: > On Tue, Dec 18, 2007 at 02:03:55AM -0800, Alfred Perlstein wrote: >> * Diomidis Spinellis [071218 00:48] wrote: >>> dds 2007-12-18 08:49:47 UTC >>> >>> FreeBSD src repository >>> >>> Modified files: >>> tools/regression/bin/mv regress.sh >>> Log: >>> Add more tests. All rename(2)-based tests now succeed. >>> The performance of the cross-device equivalents is under investigation. >> Diomidis, >> >> Does 'mv' spawn off tar or something to do the copy part? >> >> You can _usually_ get much better cross device performance >> by hooking two processes together like so: >> >> tar -cf - -C /path/to/source . | tar -xf - -C /path/to/dst >> >> This will keep the disks a lot busier, but this can make >> things worse on cross device moves that happen to be >> on the same disk. > > I was thinking about adding two options to cp(1) (-1 and -2) to give > cp(1) a hint if the copy is done inside one disk or between separate > disks. In -1 case cp(1) will read as large blocks as possible and then > write them, in -2 case it will spawn two threads: one reader and one > writer working in parallel. Performance improvements are very visible > from what I tested. > Adding hints to cp(1) hinders portability. Even on the same OS, if two systems have their disks differently configured, a script may end up with an inappropriate hint on one of the two. Why not use fstat(2) to automatically determine if the file resides on different disks, and act accordingly? -- Diomidis Spinellis - http://www.spinellis.gr From xcllnt at mac.com Mon Mar 17 16:28:36 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Mon Mar 17 16:28:41 2008 Subject: cvs commit: src/sys/sys ata.h src/sbin/atacontrol atacontrol.8 atacontrol.c src/sys/dev/ata ata-all.c ata-all.h ata-disk.c ata-disk.h In-Reply-To: <200803171033.m2HAXOeN055116@repoman.freebsd.org> References: <200803171033.m2HAXOeN055116@repoman.freebsd.org> Message-ID: <4033FB26-0458-4EAE-A43E-4AB17B78B37D@mac.com> On Mar 17, 2008, at 3:33 AM, Poul-Henning Kamp wrote: > phk 2008-03-17 10:33:23 UTC > > FreeBSD src repository > > Modified files: > sys/sys ata.h > sbin/atacontrol atacontrol.8 atacontrol.c > sys/dev/ata ata-all.c ata-all.h ata-disk.c ata-disk.h > Log: > Add a "spindown" facility to ata-disks: If no requests have been > received > for a configurable number of seconds, spin the disk down. Spin it > back > up on the next request. Nice! -- Marcel Moolenaar xcllnt@mac.com From brueffer at FreeBSD.org Mon Mar 17 16:33:35 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Mon Mar 17 16:33:40 2008 Subject: cvs commit: src/share/man/man9 ieee80211_radiotap.9 Message-ID: <200803171633.m2HGXY0t008072@repoman.freebsd.org> brueffer 2008-03-17 16:33:34 UTC FreeBSD src repository Modified files: share/man/man9 ieee80211_radiotap.9 Log: Correct bpf data-link type. PR: 121477 Submitted by: Sam Banks MFC after: 3 days Revision Changes Path 1.4 +2 -2 src/share/man/man9/ieee80211_radiotap.9 From brueffer at FreeBSD.org Mon Mar 17 16:37:36 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Mon Mar 17 16:37:41 2008 Subject: cvs commit: src/share/man/man9 ieee80211_radiotap.9 Message-ID: <200803171637.m2HGbZ86009188@repoman.freebsd.org> brueffer 2008-03-17 16:37:35 UTC FreeBSD src repository Modified files: share/man/man9 ieee80211_radiotap.9 Log: Bah, missed one instance in the previous commit. Correct bpf data-link type here as well. PR: 121477 Revision Changes Path 1.5 +1 -1 src/share/man/man9/ieee80211_radiotap.9 From jhb at freebsd.org Mon Mar 17 16:44:30 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 17 16:44:35 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <3709.1205762533@critter.freebsd.dk> References: <3709.1205762533@critter.freebsd.dk> Message-ID: <200803171200.45273.jhb@freebsd.org> On Monday 17 March 2008 10:02:13 am Poul-Henning Kamp wrote: > In message <200803170933.48212.jhb@freebsd.org>, John Baldwin writes: > >On Monday 17 March 2008 05:01:43 am Poul-Henning Kamp wrote: > >> phk 2008-03-17 09:01:43 UTC > >> > >> FreeBSD src repository > >> > >> Modified files: > >> sys/i386/cpufreq est.c > >> Log: > >> Increase time we wait for things to settle to 1 millisecond, > >> 10 microseconds is too short. > >> > >> Always set the cpu to the highest frequency so that we get through > >> boot and don't handicap cpus where powerd(8) is not used. > > > >Hmm, I actually consider this a feature when I'm not running powerd to use > >less battery. I think we should only bump up the CPU on battery power when > >using powerd so that it can be lowered again to save battery power when the > >CPU is idle. > > We have cpufreq enabled by default now, badly configured machines run > at 50% of rated CPU power because people don't know that they need to > enable powerd(8) on servers. > > This is only going to get worse when more EnergyStar compliant servers > hit the channel. > > I think setting full speed is the correct choice, if people care about > powersaving, they need to configured it, if they don't they should get > their moneys worth out of their hardware. You have servers that default to half-speed when not on battery? That is very odd, but you can just run powerd with '-a max' and it will just set them to run full speed on startup which will work fine for you. I think the vast majority of machines that don't run at full speed on startup are machines on battery power and that the previous default was a more useful policy. We don't have any servers at work that have this bizzare behavior (and we have a lot of machines at work). -- John Baldwin From jhb at freebsd.org Mon Mar 17 16:44:39 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 17 16:44:43 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... In-Reply-To: <18398.33113.488394.494142@grasshopper.cs.duke.edu> References: <200803141941.m2EJfmL2020579@repoman.freebsd.org> <200803170952.42262.jhb@freebsd.org> <18398.33113.488394.494142@grasshopper.cs.duke.edu> Message-ID: <200803171157.54453.jhb@freebsd.org> On Monday 17 March 2008 10:33:38 am Andrew Gallatin wrote: > > John Baldwin writes: > > > > I already have in my tree an amd64/i386-specific sysarch request to bind an > > IRQ to a CPU, but it's sort of a pain to use since we don't expose interrupt > > info to userland very well so the sysadmin can't tell which CPU an IRQ is > > already connected to w/o a verbose dmesg. The first thing I plan to > > implement is the aforementioned bus_bind_intr(9) for device drivers to use. > > More than that, an admin cannot tell what MSI-X vector corresponds to > what functionality. For example, on linux, the driver can attach a > custom label to the irq information seen from cat /proc/irq. Hence my > linux 10GbE driver attaches meaninful labels to each MSI-X vector it > uses. Eg, for 2 "slices" (what we call qsets), we have this on Linux: > > % grep eth0 /proc/interrupts > 510: 0 0 PCI-MSI-edge eth0:slice-1 > 511: 0 5 PCI-MSI-edge eth0:slice-0 > > But on FreeBSD: > > % vmstat -i | grep mxge > irq256: mxge0 318254 74 > irq257: mxge0 311469 73 > > So I'd like the driver to be able to append a custom string > (":slice-%d") to the label used by vmstat -i so the adminstrator > can know what he's binding. Hmm, I've thought about the ability to name MSIs (or at least include the MSI/MSI-X index in the name). Maybe default to index name and let the driver override? -- John Baldwin From phk at phk.freebsd.dk Mon Mar 17 16:52:55 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Mon Mar 17 16:53:01 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/amd64/include intr_machdep.h src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/i386/include intr_machdep.h src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c ... In-Reply-To: Your message of "Mon, 17 Mar 2008 11:57:54 -0400." <200803171157.54453.jhb@freebsd.org> Message-ID: <4564.1205772771@critter.freebsd.dk> In message <200803171157.54453.jhb@freebsd.org>, John Baldwin writes: >Hmm, I've thought about the ability to name MSIs (or at least include the >MSI/MSI-X index in the name). Maybe default to index name and let the driver >override? I'm adding a "const char *name" to the new callout() api, and I think we should take that clue from msleep() whereever relevant. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. From phk at phk.freebsd.dk Mon Mar 17 16:53:48 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Mon Mar 17 16:53:52 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: Your message of "Mon, 17 Mar 2008 12:00:44 -0400." <200803171200.45273.jhb@freebsd.org> Message-ID: <4578.1205772826@critter.freebsd.dk> >You have servers that default to half-speed when not on battery? That is very >odd, but you can just run powerd [...] Yes, but you need to know that you should run powerd in the first place. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. From scottl at FreeBSD.org Mon Mar 17 17:18:17 2008 From: scottl at FreeBSD.org (Scott Long) Date: Mon Mar 17 17:18:23 2008 Subject: cvs commit: src/sys/cam/scsi scsi_ses.c Message-ID: <200803171718.m2HHIHQM022939@repoman.freebsd.org> scottl 2008-03-17 17:18:16 UTC FreeBSD src repository Modified files: sys/cam/scsi scsi_ses.c Log: Locking in the ses_ioctl handler doesn't have to be so strict because the referenced data is only obtained/changed in the device open handler, and the ioctl handler can only run after the open handler. Also fix a few nearby style issues. Submitted by: Matt Jacob Revision Changes Path 1.37 +25 -21 src/sys/cam/scsi/scsi_ses.c From jhb at FreeBSD.org Mon Mar 17 17:28:22 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 17:28:32 2008 Subject: cvs commit: src/sys/amd64/amd64 busdma_machdep.c src/sys/arm/arm busdma_machdep.c src/sys/i386/i386 busdma_machdep.c src/sys/ia64/ia64 busdma_machdep.c Message-ID: <200803171728.m2HHSLCh025008@repoman.freebsd.org> jhb 2008-03-17 17:28:21 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/amd64/amd64 busdma_machdep.c sys/arm/arm busdma_machdep.c sys/i386/i386 busdma_machdep.c sys/ia64/ia64 busdma_machdep.c Log: MFC: Remove the 'needbounce' variable from _bus_dmamap_load_buffer(). Revision Changes Path 1.83.2.1 +2 -5 src/sys/amd64/amd64/busdma_machdep.c 1.35.2.1 +4 -10 src/sys/arm/arm/busdma_machdep.c 1.89.2.1 +4 -9 src/sys/i386/i386/busdma_machdep.c 1.45.2.1 +2 -5 src/sys/ia64/ia64/busdma_machdep.c From jhb at FreeBSD.org Mon Mar 17 17:33:32 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 17:33:35 2008 Subject: cvs commit: src/sys/amd64/amd64 busdma_machdep.c src/sys/arm/arm busdma_machdep.c src/sys/i386/i386 busdma_machdep.c src/sys/ia64/ia64 busdma_machdep.c Message-ID: <200803171733.m2HHXWVt025403@repoman.freebsd.org> jhb 2008-03-17 17:33:32 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/amd64/amd64 busdma_machdep.c sys/arm/arm busdma_machdep.c sys/i386/i386 busdma_machdep.c sys/ia64/ia64 busdma_machdep.c Log: MFC: Remove the 'needbounce' variable from _bus_dmamap_load_buffer(). Revision Changes Path 1.70.2.7 +2 -5 src/sys/amd64/amd64/busdma_machdep.c 1.22.2.6 +4 -10 src/sys/arm/arm/busdma_machdep.c 1.74.2.6 +3 -5 src/sys/i386/i386/busdma_machdep.c 1.39.2.1 +2 -5 src/sys/ia64/ia64/busdma_machdep.c From jhb at FreeBSD.org Mon Mar 17 17:38:48 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 17:38:59 2008 Subject: cvs commit: src/lib/libc/stdio fdopen.c fopen.c freopen.c Message-ID: <200803171738.m2HHcmZg027886@repoman.freebsd.org> jhb 2008-03-17 17:38:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/stdio fdopen.c fopen.c freopen.c Log: MFC: Fail attempts to use file descriptors larger than SHRT_MAX. Revision Changes Path 1.8.2.1 +13 -0 src/lib/libc/stdio/fdopen.c 1.11.2.1 +14 -0 src/lib/libc/stdio/fopen.c 1.18.2.1 +15 -0 src/lib/libc/stdio/freopen.c From jhb at FreeBSD.org Mon Mar 17 17:39:01 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 17:39:02 2008 Subject: cvs commit: src/lib/libc/stdio fdopen.c fopen.c freopen.c Message-ID: <200803171739.m2HHd0g1027929@repoman.freebsd.org> jhb 2008-03-17 17:39:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libc/stdio fdopen.c fopen.c freopen.c Log: MFC: Fail attempts to use file descriptors larger than SHRT_MAX. Revision Changes Path 1.7.14.1 +12 -0 src/lib/libc/stdio/fdopen.c 1.10.14.1 +13 -0 src/lib/libc/stdio/fopen.c 1.13.8.2 +14 -0 src/lib/libc/stdio/freopen.c From jhb at FreeBSD.org Mon Mar 17 17:42:27 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 17:42:31 2008 Subject: cvs commit: src/lib/libc/stdio fdopen.c fopen.c freopen.c Message-ID: <200803171742.m2HHgQKf029255@repoman.freebsd.org> jhb 2008-03-17 17:42:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_4) lib/libc/stdio fdopen.c fopen.c freopen.c Log: MFC: Fail attempts to use file descriptors larger than SHRT_MAX. Revision Changes Path 1.3.2.1 +14 -1 src/lib/libc/stdio/fdopen.c 1.3.2.2 +15 -1 src/lib/libc/stdio/fopen.c 1.5.2.2 +16 -1 src/lib/libc/stdio/freopen.c From dougb at FreeBSD.org Mon Mar 17 17:54:27 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Mar 17 17:54:33 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <3709.1205762533@critter.freebsd.dk> References: <3709.1205762533@critter.freebsd.dk> Message-ID: <47DEB04F.3000808@FreeBSD.org> Poul-Henning Kamp wrote: > I think setting full speed is the correct choice, if people care about > powersaving, they need to configured it, if they don't they should get > their moneys worth out of their hardware. Trying to think as a user for a minute, I find it counterintuitive to think that my cpu will run at a lower speed unless I pump in some magic smoke (that I may or may not know exists). I agree that power saving is a very important feature, and I think Robert's idea is a good one, but _until it actually exists_, full speed is the appropriate default. Doug -- This .signature sanitized for your protection From jhb at FreeBSD.org Mon Mar 17 18:04:41 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:04:43 2008 Subject: cvs commit: src/sys/boot/i386/boot2 boot2.c src/sys/boot/i386/gptboot gptboot.c Message-ID: <200803171804.m2HI4fUd040488@repoman.freebsd.org> jhb 2008-03-17 18:04:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/i386/boot2 boot2.c sys/boot/i386/gptboot gptboot.c Log: MFC: Display the correct /boot.config file contents after parsing it. Revision Changes Path 1.83.2.3 +3 -2 src/sys/boot/i386/boot2/boot2.c 1.86.2.2 +3 -2 src/sys/boot/i386/gptboot/gptboot.c From jhb at FreeBSD.org Mon Mar 17 18:05:00 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:05:07 2008 Subject: cvs commit: src/sys/boot/i386/boot2 boot2.c src/sys/boot/i386/gptboot gptboot.c Message-ID: <200803171804.m2HI4xl7040538@repoman.freebsd.org> jhb 2008-03-17 18:04:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/boot/i386/boot2 boot2.c sys/boot/i386/gptboot gptboot.c Log: MFC: Display the correct /boot.config file contents after parsing it. Revision Changes Path 1.72.2.8 +3 -2 src/sys/boot/i386/boot2/boot2.c 1.86.4.2 +3 -2 src/sys/boot/i386/gptboot/gptboot.c From jhb at FreeBSD.org Mon Mar 17 18:07:33 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:07:38 2008 Subject: cvs commit: src/share/mk bsd.cpu.mk Message-ID: <200803171807.m2HI7Wxx040699@repoman.freebsd.org> jhb 2008-03-17 18:07:32 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/mk bsd.cpu.mk Log: MFC: Add a 'c7' CPUTYPE for VIA C7 CPUs. Revision Changes Path 1.62.2.2 +2 -0 src/share/mk/bsd.cpu.mk From jhb at FreeBSD.org Mon Mar 17 18:08:46 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:08:51 2008 Subject: cvs commit: src/share/mk bsd.cpu.mk Message-ID: <200803171808.m2HI8jIn041942@repoman.freebsd.org> jhb 2008-03-17 18:08:45 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/mk bsd.cpu.mk Log: MFC: Add a 'c7' CPUTYPE for VIA C7 CPUs. Revision Changes Path 1.48.2.5 +2 -0 src/share/mk/bsd.cpu.mk From jhb at FreeBSD.org Mon Mar 17 18:10:02 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:10:20 2008 Subject: cvs commit: src/sys/conf kmod.mk Message-ID: <200803171810.m2HIA1Nc042027@repoman.freebsd.org> jhb 2008-03-17 18:10:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf kmod.mk Log: MFC: Force an explicit dependency on opt_global.h for all module object files. Revision Changes Path 1.219.2.1 +7 -0 src/sys/conf/kmod.mk From jhb at FreeBSD.org Mon Mar 17 18:12:13 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:12:18 2008 Subject: cvs commit: src/sys/conf kmod.mk Message-ID: <200803171812.m2HICDE5043398@repoman.freebsd.org> jhb 2008-03-17 18:12:13 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/conf kmod.mk Log: MFC: Force an explicit dependency on opt_global.h for all module object files. Revision Changes Path 1.192.2.6 +7 -0 src/sys/conf/kmod.mk From antoine at FreeBSD.org Mon Mar 17 18:22:24 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Mon Mar 17 18:22:31 2008 Subject: cvs commit: src/lib/libc/locale wctype.c Message-ID: <200803171822.m2HIMNLb049522@repoman.freebsd.org> antoine 2008-03-17 18:22:23 UTC FreeBSD src repository Modified files: lib/libc/locale wctype.c Log: Don't allocate the constant array "props" on the stack in wctype. PR: 74743 Submitted by: knut st. osmundsen Approved by: rwatson (mentor) MFC after: 1 month Revision Changes Path 1.4 +1 -1 src/lib/libc/locale/wctype.c From jhb at FreeBSD.org Mon Mar 17 18:23:50 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:23:59 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h src/sys/dev/mii brgphy.c miidevs Message-ID: <200803171823.m2HINiml051920@repoman.freebsd.org> jhb 2008-03-17 18:23:44 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/bge if_bge.c if_bgereg.h sys/dev/mii brgphy.c miidevs Log: MFC: Add support for the BCM5722. Revision Changes Path 1.198.2.4 +6 -3 src/sys/dev/bge/if_bge.c 1.73.2.1 +2 -0 src/sys/dev/bge/if_bgereg.h 1.70.2.1 +1 -0 src/sys/dev/mii/brgphy.c 1.46.2.3 +1 -0 src/sys/dev/mii/miidevs From jhb at FreeBSD.org Mon Mar 17 18:24:05 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:24:10 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h src/sys/dev/mii brgphy.c miidevs Message-ID: <200803171824.m2HIO4Cv052012@repoman.freebsd.org> jhb 2008-03-17 18:24:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/bge if_bge.c if_bgereg.h sys/dev/mii brgphy.c miidevs Log: MFC: Add support for the BCM5722. Revision Changes Path 1.91.2.26 +6 -3 src/sys/dev/bge/if_bge.c 1.36.2.11 +2 -0 src/sys/dev/bge/if_bgereg.h 1.34.2.16 +1 -0 src/sys/dev/mii/brgphy.c 1.30.2.13 +1 -0 src/sys/dev/mii/miidevs From jhb at FreeBSD.org Mon Mar 17 18:27:13 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:27:17 2008 Subject: cvs commit: src/sys/pci agp_nvidia.c Message-ID: <200803171827.m2HIRC9S054491@repoman.freebsd.org> jhb 2008-03-17 18:27:12 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/pci agp_nvidia.c Log: MFC: Calculate the number of pages the GATT spans instead of assuming it always spans 33 pages. Revision Changes Path 1.11.2.2 +4 -3 src/sys/pci/agp_nvidia.c From jhb at FreeBSD.org Mon Mar 17 18:27:24 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:27:33 2008 Subject: cvs commit: src/sys/pci agp_nvidia.c Message-ID: <200803171827.m2HIRNoI055256@repoman.freebsd.org> jhb 2008-03-17 18:27:23 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/pci agp_nvidia.c Log: MFC: Calculate the number of pages the GATT spans instead of assuming it always spans 33 pages. Revision Changes Path 1.9.2.3 +4 -3 src/sys/pci/agp_nvidia.c From antoine at FreeBSD.org Mon Mar 17 18:27:29 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Mon Mar 17 18:27:36 2008 Subject: cvs commit: src/sys/compat/svr4 svr4_fcntl.c Message-ID: <200803171827.m2HIRSn3055693@repoman.freebsd.org> antoine 2008-03-17 18:27:28 UTC FreeBSD src repository Modified files: sys/compat/svr4 svr4_fcntl.c Log: Simplify fcntl(SVR4_F_DUP2FD) code now that FreeBSD has F_DUP2FD. Approved by: rwatson (mentor) Revision Changes Path 1.44 +3 -13 src/sys/compat/svr4/svr4_fcntl.c From jhb at FreeBSD.org Mon Mar 17 18:30:38 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:30:43 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803171830.m2HIUbpw058152@repoman.freebsd.org> jhb 2008-03-17 18:30:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/i386/cpufreq est.c Log: MFC: Use cpu_feature2 and don't free anything in detach. Revision Changes Path 1.11.2.3 +5 -5 src/sys/i386/cpufreq/est.c From jhb at FreeBSD.org Mon Mar 17 18:31:07 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 18:31:11 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200803171831.m2HIV673058202@repoman.freebsd.org> jhb 2008-03-17 18:31:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/i386/cpufreq est.c Log: MFC: Use cpu_feature2 and don't free anything in detach. Revision Changes Path 1.7.2.3 +5 -5 src/sys/i386/cpufreq/est.c From antoine at FreeBSD.org Mon Mar 17 18:31:44 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Mon Mar 17 18:31:45 2008 Subject: cvs commit: src/usr.bin/ruptime ruptime.c Message-ID: <200803171831.m2HIVhfD058241@repoman.freebsd.org> antoine 2008-03-17 18:31:43 UTC FreeBSD src repository Modified files: usr.bin/ruptime ruptime.c Log: Initialize variable "now" before using it (it is used by the LEFTEARTH macro) PR: 121418 Submitted by: Atsuo Ohki Approved by: rwatson (mentor) MFC after: 1 month Revision Changes Path 1.19 +1 -1 src/usr.bin/ruptime/ruptime.c From sobomax at FreeBSD.org Mon Mar 17 18:40:35 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Mon Mar 17 18:40:42 2008 Subject: cvs commit: src/sys/sys eventhandler.h src/sys/kern kern_descrip.c uipc_socket.c In-Reply-To: <20080317020036.M910@desktop> References: <200803160621.m2G6LUlE034766@repoman.freebsd.org> <20080316100836.F44049@fledge.watson.org> <47DDED1F.8000707@FreeBSD.org> <20080317020036.M910@desktop> Message-ID: <47DEBB0D.8090304@FreeBSD.org> Jeff Roberson wrote: > On Sun, 16 Mar 2008, Maxim Sobolev wrote: > >> Robert Watson wrote: >>> On Sun, 16 Mar 2008, Maxim Sobolev wrote: >>> >>>> sobomax 2008-03-16 06:21:30 UTC >>>> >>>> FreeBSD src repository >>>> >>>> Modified files: >>>> sys/sys eventhandler.h >>>> sys/kern kern_descrip.c uipc_socket.c >>>> Log: >>>> Properly set size of the file_zone to match kern.maxfiles parameter. >>>> Otherwise the parameter is no-op, since zone by default limits number >>>> of descriptors to some 12K entries. Attempt to allocate more ends up >>>> sleeping on zonelimit. >>> >>> Hmm. Could you be more specific about the nature of the problem >>> you're seeing here? UMA shouldn't impose any default zone limit on >>> the file_zone if none is set. Are you sure it's the struct file >>> zonelimit you're running into, and not, say, the socket zone limit, etc? >> >> Yes, you are right, further investigation revealed that in my case I >> was hitting maxsockets limit, not maxfiles limit. Not sure if the >> backout is in order or having files zone limit has some value, so that >> possible leaks can be detected and UMA zone can potentially do >> size-based optimizations in the future? > > The UMA max is complicated due to per-cpu caches. Really the limit code > in uma should be rewritten or all consumers should handle it themselves > the way struct file does. It'd probably be best if you backed this out. OK, no problem, but I have two question: 1. What to do with SI_SUB_EVENTHANDLER+1 in the socket code and SI_SUB_LOCK in the old file descriptor code? Neither looks good. 2. Also I wonder why hitting sockets limit works differently than file descriptors limit. In the latter case the program gets the proper error code (ENFILE) instantly, while in the former the program just hangs sleeping on the UMA zone limit. Perhaps socket(2) should fail in such case with either ENFILE or ENOBUFS? -Maxim From sobomax at FreeBSD.org Mon Mar 17 18:44:31 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Mon Mar 17 18:44:38 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <3709.1205762533@critter.freebsd.dk> References: <3709.1205762533@critter.freebsd.dk> Message-ID: <47DEBC02.8060700@FreeBSD.org> Poul-Henning Kamp wrote: > In message <200803170933.48212.jhb@freebsd.org>, John Baldwin writes: >> On Monday 17 March 2008 05:01:43 am Poul-Henning Kamp wrote: >>> phk 2008-03-17 09:01:43 UTC >>> >>> FreeBSD src repository >>> >>> Modified files: >>> sys/i386/cpufreq est.c >>> Log: >>> Increase time we wait for things to settle to 1 millisecond, >>> 10 microseconds is too short. >>> >>> Always set the cpu to the highest frequency so that we get through >>> boot and don't handicap cpus where powerd(8) is not used. >> Hmm, I actually consider this a feature when I'm not running powerd to use >> less battery. I think we should only bump up the CPU on battery power when >> using powerd so that it can be lowered again to save battery power when the >> CPU is idle. > > We have cpufreq enabled by default now, badly configured machines run > at 50% of rated CPU power because people don't know that they need to > enable powerd(8) on servers. > > This is only going to get worse when more EnergyStar compliant servers > hit the channel. > > I think setting full speed is the correct choice, if people care about > powersaving, they need to configured it, if they don't they should get > their moneys worth out of their hardware. Perhaps make it a kernel option? But I agree that CPU should run at full speed by default, otherwise it would break POLA for users upgrading from previous releases. -Maxim From jhb at freebsd.org Mon Mar 17 18:53:07 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 17 18:53:14 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <4578.1205772826@critter.freebsd.dk> References: <4578.1205772826@critter.freebsd.dk> Message-ID: <200803171317.15812.jhb@freebsd.org> On Monday 17 March 2008 12:53:46 pm Poul-Henning Kamp wrote: > > >You have servers that default to half-speed when not on battery? That is very > >odd, but you can just run powerd [...] > > Yes, but you need to know that you should run powerd in the first place. Err, I would just always run it. You are now saying that laptops have to run powerd to avoid full-speed CPUs on boot (even though the BIOS throttles it down for you) but you can't run powerd on the server? -- John Baldwin From antoine at FreeBSD.org Mon Mar 17 18:57:08 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Mon Mar 17 18:57:14 2008 Subject: cvs commit: src/sys/kern subr_autoconf.c Message-ID: <200803171857.m2HIv7RF082731@repoman.freebsd.org> antoine 2008-03-17 18:57:07 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern subr_autoconf.c Log: MFC to RELENG_7 Remove a superfluous line in run_interrupt_driven_config_hooks(), next_entry is already initialized during TAILQ_FOREACH_SAFE(). PR: kern/119604 Approved by: rwatson (mentor) MFC after: 1 month Revision Changes Path 1.23.2.1 +0 -1 src/sys/kern/subr_autoconf.c From phk at phk.freebsd.dk Mon Mar 17 18:59:26 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Mon Mar 17 18:59:33 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: Your message of "Mon, 17 Mar 2008 13:17:15 -0400." <200803171317.15812.jhb@freebsd.org> Message-ID: <5293.1205780363@critter.freebsd.dk> In message <200803171317.15812.jhb@freebsd.org>, John Baldwin writes: >On Monday 17 March 2008 12:53:46 pm Poul-Henning Kamp wrote: >> >> >You have servers that default to half-speed when not on battery? That is >very >> >odd, but you can just run powerd [...] >> >> Yes, but you need to know that you should run powerd in the first place. > >Err, I would just always run it. You are now saying that laptops have to run >powerd to avoid full-speed CPUs on boot (even though the BIOS throttles it >down for you) but you can't run powerd on the server? We don't enable powerd by default, as Robert said, maybe we should. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. From antoine at FreeBSD.org Mon Mar 17 19:05:37 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Mon Mar 17 19:05:42 2008 Subject: cvs commit: src/lib/libdisk disk.c Message-ID: <200803171905.m2HJ5a6m092337@repoman.freebsd.org> antoine 2008-03-17 19:05:36 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libdisk disk.c Log: MFC to RELENG_7 - Make Disk_Names() behave as documented in libdisk(3): return an array of disk names, where you must free each pointer, as well as the array by hand. [1] - Destaticize "disks" in Disk_Names, it has no reasons to be static. PR: kern/96077 [1] PR: kern/114110 [1] MFC after: 1 month Approved by: rwatson (mentor) Revision Changes Path 1.127.2.1 +14 -5 src/lib/libdisk/disk.c From antoine at FreeBSD.org Mon Mar 17 19:08:33 2008 From: antoine at FreeBSD.org (Antoine Brodin) Date: Mon Mar 17 19:08:35 2008 Subject: cvs commit: src/lib/libdisk disk.c Message-ID: <200803171908.m2HJ8X4E095454@repoman.freebsd.org> antoine 2008-03-17 19:08:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libdisk disk.c Log: MFC to RELENG_6 - Make Disk_Names() behave as documented in libdisk(3): return an array of disk names, where you must free each pointer, as well as the array by hand. [1] - Destaticize "disks" in Disk_Names, it has no reasons to be static. PR: kern/96077 [1] PR: kern/114110 [1] MFC after: 1 month Approved by: rwatson (mentor) Revision Changes Path 1.125.2.2 +14 -5 src/lib/libdisk/disk.c From rpaulo at FreeBSD.org Mon Mar 17 19:56:23 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Mon Mar 17 19:56:27 2008 Subject: cvs commit: src/sys/dev/acpica acpi_cpu.c Message-ID: <200803171956.m2HJuMU7045702@repoman.freebsd.org> rpaulo 2008-03-17 19:56:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/acpica acpi_cpu.c Log: MFC r1.71: Some PIIX4 chipsets need to be told to generate Stop Breaks by setting the appropriate bit in the DEVACTB register. This change allows the C2 state on those systems to work as expected. Reviewed by: njl Submitted by: Andriy Gapon Revision Changes Path 1.67.2.4 +20 -0 src/sys/dev/acpica/acpi_cpu.c From rpaulo at FreeBSD.org Mon Mar 17 19:57:25 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Mon Mar 17 19:57:27 2008 Subject: cvs commit: src/sys/dev/acpica acpi_cpu.c Message-ID: <200803171957.m2HJvLnX046180@repoman.freebsd.org> rpaulo 2008-03-17 19:57:21 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/acpica acpi_cpu.c Log: MFC r1.71: Some PIIX4 chipsets need to be told to generate Stop Breaks by setting the appropriate bit in the DEVACTB register. This change allows the C2 state on those systems to work as expected. Reviewed by: njl Submitted by: Andriy Gapon Revision Changes Path 1.57.2.8 +20 -0 src/sys/dev/acpica/acpi_cpu.c From jhb at freebsd.org Mon Mar 17 20:36:15 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Mar 17 20:36:22 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <47DEBC02.8060700@FreeBSD.org> References: <3709.1205762533@critter.freebsd.dk> <47DEBC02.8060700@FreeBSD.org> Message-ID: <200803171635.51748.jhb@freebsd.org> On Monday 17 March 2008 02:44:18 pm Maxim Sobolev wrote: > Poul-Henning Kamp wrote: > > In message <200803170933.48212.jhb@freebsd.org>, John Baldwin writes: > >> On Monday 17 March 2008 05:01:43 am Poul-Henning Kamp wrote: > >>> phk 2008-03-17 09:01:43 UTC > >>> > >>> FreeBSD src repository > >>> > >>> Modified files: > >>> sys/i386/cpufreq est.c > >>> Log: > >>> Increase time we wait for things to settle to 1 millisecond, > >>> 10 microseconds is too short. > >>> > >>> Always set the cpu to the highest frequency so that we get through > >>> boot and don't handicap cpus where powerd(8) is not used. > >> Hmm, I actually consider this a feature when I'm not running powerd to use > >> less battery. I think we should only bump up the CPU on battery power when > >> using powerd so that it can be lowered again to save battery power when the > >> CPU is idle. > > > > We have cpufreq enabled by default now, badly configured machines run > > at 50% of rated CPU power because people don't know that they need to > > enable powerd(8) on servers. > > > > This is only going to get worse when more EnergyStar compliant servers > > hit the channel. > > > > I think setting full speed is the correct choice, if people care about > > powersaving, they need to configured it, if they don't they should get > > their moneys worth out of their hardware. > > Perhaps make it a kernel option? But I agree that CPU should run at full > speed by default, otherwise it would break POLA for users upgrading from > previous releases. Err, actually, it's the other way around. On previous releases the BIOS is going to set your CPU to a lower speed on boot to save battery (on a laptop not connected to A/C during boot). The CPU will stay in the lower speed unless you both enable cpufreq and powerd in 6.x and 7.x. Now in HEAD just enabling the cpufreq driver causes the CPU to run at the full speed (and drain the battery faster) unless you also run powerd to slow it back down when it is idle. -- John Baldwin From phk at phk.freebsd.dk Mon Mar 17 20:40:22 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Mon Mar 17 20:40:27 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: Your message of "Mon, 17 Mar 2008 16:35:51 -0400." <200803171635.51748.jhb@freebsd.org> Message-ID: <5836.1205786419@critter.freebsd.dk> In message <200803171635.51748.jhb@freebsd.org>, John Baldwin writes: >Err, actually, it's the other way around. On previous releases the BIOS is >going to set your CPU to a lower speed on boot to save battery (on a laptop >not connected to A/C during boot). John, you are right about laptops, but people with laptops know power management is part of the job. Nobody would ever dream of having to run powerd on a server, in particular not to get _full_ performance. If you can find some way to detect if we are on a server or a laptop, I'm fine with divergent defaults. But given that suspend/resume has barely worked for any new laptop for 18 months, I can't really justify hurting the servers for the laptops on this issue. Please also note that p4tcc.c driver has done the same thing all along, no matter what the bios hat set, it would set 100%. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. From bzeeb-lists at lists.zabbadoz.net Mon Mar 17 20:44:17 2008 From: bzeeb-lists at lists.zabbadoz.net (Bjoern A. Zeeb) Date: Mon Mar 17 20:44:24 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <5293.1205780363@critter.freebsd.dk> References: <5293.1205780363@critter.freebsd.dk> Message-ID: <20080317202023.X50685@maildrop.int.zabbadoz.net> On Mon, 17 Mar 2008, Poul-Henning Kamp wrote: > In message <200803171317.15812.jhb@freebsd.org>, John Baldwin writes: >> On Monday 17 March 2008 12:53:46 pm Poul-Henning Kamp wrote: >>> >>>> You have servers that default to half-speed when not on battery? That is >> very >>>> odd, but you can just run powerd [...] >>> >>> Yes, but you need to know that you should run powerd in the first place. >> >> Err, I would just always run it. You are now saying that laptops have to run >> powerd to avoid full-speed CPUs on boot (even though the BIOS throttles it >> down for you) but you can't run powerd on the server? > > We don't enable powerd by default, as Robert said, maybe we should. in the past I had machines that after enabling powerd brought the machine down needing a full power cycle to restore. this had always been machines in production unfornately where pople had forgotten to do so during install and thus almost impossible to debug. I am fine with having this on by default for "install" (whatever that means) but changing that policy with a system update mergemastering might not be a good idea. -- Bjoern A. Zeeb bzeeb at Zabbadoz dot NeT Software is harder than hardware so better get it right the first time. From marius at FreeBSD.org Mon Mar 17 20:52:13 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Mon Mar 17 20:52:21 2008 Subject: cvs commit: src/sys/dev/gem if_gem.c if_gem_pci.c if_gemreg.h if_gemvar.h Message-ID: <200803172052.m2HKqBbp095355@repoman.freebsd.org> marius 2008-03-17 20:52:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/gem if_gem.c if_gem_pci.c if_gemreg.h if_gemvar.h Log: MFC: if_gem.c 1.46; if_gem_pci.c 1.24; if_gemreg.h 1.6; if_gemvar.h 1.16 o In preparation for basing a new driver on this one: - ANSIfy - try to bring closer to style(9) - remove banal comments. o Add my copyright for having done lots of fixes and improvements. Revision Changes Path 1.44.2.2 +336 -415 src/sys/dev/gem/if_gem.c 1.23.2.1 +32 -32 src/sys/dev/gem/if_gem_pci.c 1.5.2.1 +9 -9 src/sys/dev/gem/if_gemreg.h 1.15.2.1 +46 -58 src/sys/dev/gem/if_gemvar.h From marius at FreeBSD.org Mon Mar 17 20:53:23 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Mon Mar 17 20:53:31 2008 Subject: cvs commit: src/sys/sparc64/sparc64 iommu.c Message-ID: <200803172053.m2HKrN8W095419@repoman.freebsd.org> marius 2008-03-17 20:53:23 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/sparc64/sparc64 iommu.c Log: MFC: 1.46 - Fix some style bugs. - Replace hard-coded functions names missed in rev. 1.44 with __func__. Revision Changes Path 1.45.2.1 +54 -48 src/sys/sparc64/sparc64/iommu.c From h.schmalzbauer at omnisec.de Mon Mar 17 21:35:39 2008 From: h.schmalzbauer at omnisec.de (Harald Schmalzbauer) Date: Mon Mar 17 21:35:42 2008 Subject: cvs commit: src/sys/sys ata.h src/sbin/atacontrol atacontrol.8 atacontrol.c src/sys/dev/ata ata-all.c ata-all.h ata-disk.c ata-disk.h In-Reply-To: <4033FB26-0458-4EAE-A43E-4AB17B78B37D@mac.com> References: <200803171033.m2HAXOeN055116@repoman.freebsd.org> <4033FB26-0458-4EAE-A43E-4AB17B78B37D@mac.com> Message-ID: <47DEE428.70300@omnisec.de> Marcel Moolenaar wrote am 17.03.2008 17:28 (localtime): > > On Mar 17, 2008, at 3:33 AM, Poul-Henning Kamp wrote: > >> phk 2008-03-17 10:33:23 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/sys ata.h >> sbin/atacontrol atacontrol.8 atacontrol.c >> sys/dev/ata ata-all.c ata-all.h ata-disk.c ata-disk.h >> Log: >> Add a "spindown" facility to ata-disks: If no requests have been >> received >> for a configurable number of seconds, spin the disk down. Spin it back >> up on the next request. > > Nice! Very nice indeed, I've been hoping for this. Thanks Poul-Henning! For those who want to try it against -STABLE, you'll find the cumulated diff at http://www.schmalzbauer.de/downloads/atacontrol-spindown_RELENG7.diff Copy to /usr/src and 'patch -i atacontrol-spindown_RELENG7.diff' Best regards, -Harry -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 195 bytes Desc: OpenPGP digital signature Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080317/b883933a/signature.pgp From piso at FreeBSD.org Mon Mar 17 22:08:32 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Mon Mar 17 22:08:38 2008 Subject: cvs commit: src/sys/netinet/libalias alias_irc.c Message-ID: <200803172208.m2HM8VvP011208@repoman.freebsd.org> piso 2008-03-17 22:08:31 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias_irc.c Log: Don't abuse stack space while in kernel land, use heap instead. Revision Changes Path 1.25 +10 -5 src/sys/netinet/libalias/alias_irc.c From piso at FreeBSD.org Mon Mar 17 22:16:20 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Mon Mar 17 22:16:23 2008 Subject: cvs commit: src/sys/netinet/libalias alias_irc.c Message-ID: <200803172216.m2HMGJKj011742@repoman.freebsd.org> piso 2008-03-17 22:16:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet/libalias alias_irc.c Log: MFC revision 1.25: Don't abuse stack space while in kernel land, use heap instead. PR: kern/121693 Revision Changes Path 1.23.2.1 +10 -5 src/sys/netinet/libalias/alias_irc.c From jhb at FreeBSD.org Mon Mar 17 22:42:02 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Mar 17 22:42:07 2008 Subject: cvs commit: src/sys/amd64/amd64 intr_machdep.c src/sys/arm/arm intr.c src/sys/i386/i386 intr_machdep.c src/sys/ia64/ia64 interrupt.c src/sys/kern kern_intr.c src/sys/powerpc/powerpc intr_machdep.c src/sys/sparc64/sparc64 intr_machdep.c src/sys/sys ... Message-ID: <200803172242.m2HMg1Cd015714@repoman.freebsd.org> jhb 2008-03-17 22:42:01 UTC FreeBSD src repository Modified files: sys/amd64/amd64 intr_machdep.c sys/arm/arm intr.c sys/i386/i386 intr_machdep.c sys/ia64/ia64 interrupt.c sys/kern kern_intr.c sys/powerpc/powerpc intr_machdep.c sys/sparc64/sparc64 intr_machdep.c sys/sys interrupt.h Log: Simplify the interrupt code a bit: - Always include the ie_disable and ie_eoi methods in 'struct intr_event' and collapse down to one intr_event_create() routine. The disable and eoi hooks simply aren't used currently in the !INTR_FILTER case. - Expand 'disab' to 'disable' in a few places. - Use function casts for arm and i386:intr_eoi_src() instead of wrapper routines since to trim one extra indirection. Compiled on: {arm,amd64,i386,ia64,ppc,sparc64} x {FILTER, !FILTER} Tested on: {amd64,i386} x {FILTER, !FILTER} Revision Changes Path 1.39 +15 -32 src/sys/amd64/amd64/intr_machdep.c 1.19 +4 -28 src/sys/arm/arm/intr.c 1.34 +15 -32 src/sys/i386/i386/intr_machdep.c 1.64 +1 -4 src/sys/ia64/ia64/interrupt.c 1.157 +6 -46 src/sys/kern/kern_intr.c 1.19 +1 -6 src/sys/powerpc/powerpc/intr_machdep.c 1.30 +2 -6 src/sys/sparc64/sparc64/intr_machdep.c 1.39 +4 -13 src/sys/sys/interrupt.h From piso at FreeBSD.org Mon Mar 17 23:02:56 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Mon Mar 17 23:03:04 2008 Subject: cvs commit: src/sys/netinet ip_fw2.c Message-ID: <200803172302.m2HN2ukI024265@repoman.freebsd.org> piso 2008-03-17 23:02:56 UTC FreeBSD src repository Modified files: sys/netinet ip_fw2.c Log: Don't cache ptr to nat rule in case of tablearg argument. Bug spotted by: Dyadchenko Mihail Revision Changes Path 1.183 +2 -1 src/sys/netinet/ip_fw2.c From piso at FreeBSD.org Mon Mar 17 23:08:42 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Mon Mar 17 23:08:44 2008 Subject: cvs commit: src/sys/netinet ip_fw2.c Message-ID: <200803172308.m2HN8gv6026710@repoman.freebsd.org> piso 2008-03-17 23:08:42 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet ip_fw2.c Log: MFC revision 1.182: Don't cache ptr to nat rule in case of tablearg argument. Bug spotted by: Dyadchenko Mihail Revision Changes Path 1.175.2.3 +2 -1 src/sys/netinet/ip_fw2.c From emax at FreeBSD.org Tue Mar 18 00:25:27 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 18 00:25:29 2008 Subject: cvs commit: src/lib/libsdp sdp.h Message-ID: <200803180025.m2I0PQ6d049770@repoman.freebsd.org> emax 2008-03-18 00:25:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libsdp sdp.h Log: MFC: Add structures to hold SDP parameters for the NAP, GN and PANU profiles. It should be mentioned that a somewhat similar patch was submitted by Rako < rako29 at gmail dot com > Revision Changes Path 1.7.2.1 +17 -0 src/lib/libsdp/sdp.h From emax at FreeBSD.org Tue Mar 18 00:26:01 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 18 00:26:09 2008 Subject: cvs commit: src/lib/libsdp sdp.h Message-ID: <200803180026.m2I0Q15Z049816@repoman.freebsd.org> emax 2008-03-18 00:26:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libsdp sdp.h Log: MFC: Add structures to hold SDP parameters for the NAP, GN and PANU profiles. It should be mentioned that a somewhat similar patch was submitted by Rako < rako29 at gmail dot com > Revision Changes Path 1.5.2.3 +17 -0 src/lib/libsdp/sdp.h From emax at FreeBSD.org Tue Mar 18 00:27:48 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 18 00:27:51 2008 Subject: cvs commit: src/usr.sbin/bluetooth/sdpd Makefile gn.c nap.c panu.c profile.c profile.h Message-ID: <200803180027.m2I0Rmdf049894@repoman.freebsd.org> emax 2008-03-18 00:27:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/bluetooth/sdpd Makefile profile.c profile.h Added files: (Branch: RELENG_7) usr.sbin/bluetooth/sdpd gn.c nap.c panu.c Log: MFC: Add support for the NAP, GN and PANU profiles to the sdpd(8). It should be mentioned that a somewhat similar patch was submitted by Rako < rako29 at gmail dot com > Revision Changes Path 1.2.10.1 +3 -3 src/usr.sbin/bluetooth/sdpd/Makefile 1.1.2.1 +146 -0 src/usr.sbin/bluetooth/sdpd/gn.c (new) 1.1.2.1 +183 -0 src/usr.sbin/bluetooth/sdpd/nap.c (new) 1.1.2.1 +131 -0 src/usr.sbin/bluetooth/sdpd/panu.c (new) 1.2.18.1 +93 -2 src/usr.sbin/bluetooth/sdpd/profile.c 1.1.18.1 +6 -1 src/usr.sbin/bluetooth/sdpd/profile.h From emax at FreeBSD.org Tue Mar 18 00:29:26 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 18 00:29:28 2008 Subject: cvs commit: src/usr.sbin/bluetooth/sdpd Makefile gn.c nap.c panu.c profile.c profile.h Message-ID: <200803180029.m2I0TQl2049964@repoman.freebsd.org> emax 2008-03-18 00:29:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/bluetooth/sdpd Makefile profile.c profile.h Added files: (Branch: RELENG_6) usr.sbin/bluetooth/sdpd gn.c nap.c panu.c Log: MFC: Add support for the NAP, GN and PANU profiles to the sdpd(8). It should be mentioned that a somewhat similar patch was submitted by Rako < rako29 at gmail dot com > Revision Changes Path 1.2.2.1 +3 -3 src/usr.sbin/bluetooth/sdpd/Makefile 1.1.4.1 +146 -0 src/usr.sbin/bluetooth/sdpd/gn.c (new) 1.1.4.1 +183 -0 src/usr.sbin/bluetooth/sdpd/nap.c (new) 1.1.4.1 +131 -0 src/usr.sbin/bluetooth/sdpd/panu.c (new) 1.2.8.1 +93 -2 src/usr.sbin/bluetooth/sdpd/profile.c 1.1.8.1 +6 -1 src/usr.sbin/bluetooth/sdpd/profile.h From yongari at FreeBSD.org Tue Mar 18 00:42:27 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 00:42:29 2008 Subject: cvs commit: src/sys/arm/xscale/ixp425 if_npe.c src/sys/dev/gem if_gem.c src/sys/dev/msk if_msk.c src/sys/dev/nfe if_nfe.c src/sys/dev/wpi if_wpi.c Message-ID: <200803180042.m2I0gRqF054007@repoman.freebsd.org> yongari 2008-03-18 00:42:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/arm/xscale/ixp425 if_npe.c sys/dev/gem if_gem.c sys/dev/msk if_msk.c sys/dev/nfe if_nfe.c sys/dev/wpi if_wpi.c Log: MFC: if_npe.c 1.8, if_gem.c 1.47, if_msk.c 1.27, if_nfe.c 1.25, if_wpi.c 1.6 - Retire npe_defrag(), gem_defrag(), msk_defrag(), nfe_defrag(), and re_defrag() and use m_collapse() instead. - Replace a reference to ath_defrag() in a comment in if_wpi.c with m_collapse(). Revision Changes Path 1.6.2.1 +1 -85 src/sys/arm/xscale/ixp425/if_npe.c 1.44.2.3 +1 -88 src/sys/dev/gem/if_gem.c 1.18.2.9 +1 -88 src/sys/dev/msk/if_msk.c 1.21.2.4 +1 -89 src/sys/dev/nfe/if_nfe.c 1.5.2.3 +1 -1 src/sys/dev/wpi/if_wpi.c From yongari at FreeBSD.org Tue Mar 18 01:01:19 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 01:01:20 2008 Subject: cvs commit: src/sys/dev/msk if_msk.c if_mskreg.h Message-ID: <200803180101.m2I11Ic0070111@repoman.freebsd.org> yongari 2008-03-18 01:01:18 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/msk if_msk.c if_mskreg.h Log: MFC: if_msk.c 1.29, if_mskreg.h 1.12 to RELENG_7 Workaround GMAC hardware hang of Yukon II on the receipt of pause frames. This bug seems to happen on certain hardware model/revision (e.g. 88E8053) but it's not identified which hardwares are affected. Revision 1.4 of if_mskreg.h was not enough to workaround the bug. To workaround it, inrease GMAC FIFO threshold by one FIFO word to flush received pause frames. Revision Changes Path 1.18.2.10 +5 -2 src/sys/dev/msk/if_msk.c 1.6.2.6 +1 -0 src/sys/dev/msk/if_mskreg.h From yongari at FreeBSD.org Tue Mar 18 01:27:16 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 01:27:29 2008 Subject: cvs commit: src/sys/dev/msk if_msk.c if_mskreg.h Message-ID: <200803180127.m2I1RFSm078598@repoman.freebsd.org> yongari 2008-03-18 01:27:15 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/msk if_msk.c if_mskreg.h Log: MFC: if_msk.c 1.29, if_mskreg.h 1.12 to RELENG_6 Workaround GMAC hardware hang of Yukon II on the receipt of pause frames. This bug seems to happen on certain hardware model/revision (e.g. 88E8053) but it's not identified which hardwares are affected. Revision 1.4 of if_mskreg.h was not enough to workaround the bug. To workaround it, inrease GMAC FIFO threshold by one FIFO word to flush received pause frames. Revision Changes Path 1.11.2.16 +5 -2 src/sys/dev/msk/if_msk.c 1.3.2.10 +1 -0 src/sys/dev/msk/if_mskreg.h From yongari at FreeBSD.org Tue Mar 18 01:31:38 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 01:31:41 2008 Subject: cvs commit: src/sys/dev/msk if_msk.c Message-ID: <200803180131.m2I1VbuF078819@repoman.freebsd.org> yongari 2008-03-18 01:31:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/msk if_msk.c Log: MFC: if_msk.c 1.30, if_msk.c 1.31 to RELENG_7 To overcome hardware checksum offload bug msk(4) used to compute TCP/UDP checksum in driver for short frames. For frames that requires hardware VLAN tag insertion, the checksum offload trick does not work due to changes of checksum offset in mbuf after the VLAN tag. Disable hardware checksum offload for VLAN interface to fix the bug. Revision Changes Path 1.18.2.11 +8 -4 src/sys/dev/msk/if_msk.c From yongari at FreeBSD.org Tue Mar 18 01:33:48 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 01:33:54 2008 Subject: cvs commit: src/sys/dev/msk if_msk.c Message-ID: <200803180133.m2I1XlgF078970@repoman.freebsd.org> yongari 2008-03-18 01:33:47 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/msk if_msk.c Log: MFC: if_msk.c 1.30, if_msk.c 1.31 to RELENG_6 To overcome hardware checksum offload bug msk(4) used to compute TCP/UDP checksum in driver for short frames. For frames that requires hardware VLAN tag insertion, the checksum offload trick does not work due to changes of checksum offset in mbuf after the VLAN tag. Disable hardware checksum offload for VLAN interface to fix the bug. Revision Changes Path 1.11.2.17 +8 -4 src/sys/dev/msk/if_msk.c From yongari at FreeBSD.org Tue Mar 18 01:53:47 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 01:53:52 2008 Subject: cvs commit: src/sys/dev/lge if_lge.c Message-ID: <200803180153.m2I1rknD080974@repoman.freebsd.org> yongari 2008-03-18 01:53:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/lge if_lge.c Log: MFC if_lge.c rev 1.53 to RELENG_7: Plug memory leak in jumbo buffer allocation failure path. Patch in the PR was modified to check active jumbo buffers in use and other possible jumbo buffer leak. Jumbo buffer usage in lge(4) still wouldn't be reliable due to lack of driver lock in local jumbo buffer allocator. Either introduce a new lock to protect jumbo buffer or switch to UMA backed page allocator for jumbo frame is required. PR: kern/78072 Revision Changes Path 1.50.2.2 +12 -4 src/sys/dev/lge/if_lge.c From yongari at FreeBSD.org Tue Mar 18 02:01:24 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 02:01:26 2008 Subject: cvs commit: src/sys/dev/lge if_lge.c Message-ID: <200803180201.m2I21OBW082473@repoman.freebsd.org> yongari 2008-03-18 02:01:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/lge if_lge.c Log: MFC if_lge.c rev 1.53 to RELENG_6: Plug memory leak in jumbo buffer allocation failure path. Patch in the PR was modified to check active jumbo buffers in use and other possible jumbo buffer leak. Jumbo buffer usage in lge(4) still wouldn't be reliable due to lack of driver lock in local jumbo buffer allocator. Either introduce a new lock to protect jumbo buffer or switch to UMA backed page allocator for jumbo frame is required. I've removed informational device_printf in lge_free_jumbo_mem as lge_dev member variable in softc does not exist on RELENG_6. if_printf can't be used as sc->lge_ifp could be NULL. PR: kern/78072 Revision Changes Path 1.39.2.6 +10 -4 src/sys/dev/lge/if_lge.c From davidxu at FreeBSD.org Tue Mar 18 02:06:52 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Mar 18 02:06:57 2008 Subject: cvs commit: src/lib/libthr/thread thr_create.c thr_exit.c thr_sig.c Message-ID: <200803180206.m2I26pwt085040@repoman.freebsd.org> davidxu 2008-03-18 02:06:51 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_create.c thr_exit.c thr_sig.c Log: - Copy signal mask out before THR_UNLOCK(), because THR_UNLOCK() may call _thr_suspend_check() which messes sigmask saved in thread structure. - Don't suspend a thread has force_exit set. - In pthread_exit(), if there is a suspension flag set, wake up waiting- thread after setting PS_DEAD, this causes waiting-thread to break loop in suspend_common(). Revision Changes Path 1.42 +10 -4 src/lib/libthr/thread/thr_create.c 1.26 +4 -0 src/lib/libthr/thread/thr_exit.c 1.28 +3 -0 src/lib/libthr/thread/thr_sig.c From yongari at FreeBSD.org Tue Mar 18 02:11:41 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 02:11:43 2008 Subject: cvs commit: src/sys/dev/bfe if_bfe.c if_bfereg.h Message-ID: <200803180211.m2I2Beka087599@repoman.freebsd.org> yongari 2008-03-18 02:11:40 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/bfe if_bfe.c if_bfereg.h Log: MFC if_bfe.c rev 1.44, if_bfereg.h rev 1.11 to RELENG_7: Fix link state handling in bfe(4). o conversion to callout(9) API. o add a missing driver lock in bfe_ifmedia_sts(). o use our callout to drive watchdog timer. o restart Tx routine if pending queued packets are present in watchdog handler. o unarm watchdog timer only if there are no queued packets. o don't blindly reset phy and let phy driver handle link change request in bfe_init_locked(). o return the status of mii_mediachg() to caller in bfe_ifmedia_upd(). Previously it always returned 0 to caller. o add check for IFF_DRV_RUNNING flag as well as IFF_DRV_OACTIVE in bfe_start_locked(). o implement miibus_statchg method that keeps track of current link state changes as well as negotiated speed/duplex/ flow-control configuration. Reprogram MAC to appropriate duplex state. Flow-control configuration was also implemented but commented out at the moment. The flow-control configuration will be enabled again after we have general flow-control framework in mii layer. Revision Changes Path 1.42.2.2 +82 -47 src/sys/dev/bfe/if_bfe.c 1.10.2.1 +6 -1 src/sys/dev/bfe/if_bfereg.h From yongari at FreeBSD.org Tue Mar 18 02:17:38 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Tue Mar 18 02:17:44 2008 Subject: cvs commit: src/sys/dev/bfe if_bfe.c if_bfereg.h Message-ID: <200803180217.m2I2Hb36089606@repoman.freebsd.org> yongari 2008-03-18 02:17:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/bfe if_bfe.c if_bfereg.h Log: MFC if_bfe.c rev 1.44, if_bfereg.h rev 1.11 to RELENG_6: Fix link state handling in bfe(4). o conversion to callout(9) API. o add a missing driver lock in bfe_ifmedia_sts(). o use our callout to drive watchdog timer. o restart Tx routine if pending queued packets are present in watchdog handler. o unarm watchdog timer only if there are no queued packets. o don't blindly reset phy and let phy driver handle link change request in bfe_init_locked(). o return the status of mii_mediachg() to caller in bfe_ifmedia_upd(). Previously it always returned 0 to caller. o add check for IFF_DRV_RUNNING flag as well as IFF_DRV_OACTIVE in bfe_start_locked(). o implement miibus_statchg method that keeps track of current link state changes as well as negotiated speed/duplex/ flow-control configuration. Reprogram MAC to appropriate duplex state. Flow-control configuration was also implemented but commented out at the moment. The flow-control configuration will be enabled again after we have general flow-control framework in mii layer. Revision Changes Path 1.25.2.8 +82 -47 src/sys/dev/bfe/if_bfe.c 1.7.2.2 +6 -1 src/sys/dev/bfe/if_bfereg.h From kmacy at FreeBSD.org Tue Mar 18 03:55:13 2008 From: kmacy at FreeBSD.org (Kip Macy) Date: Tue Mar 18 03:58:31 2008 Subject: cvs commit: src/sys/dev/cxgb cxgb_ioctl.h cxgb_main.c src/sys/dev/cxgb/common cxgb_ael1002.c cxgb_common.h cxgb_version.h src/sys/dev/cxgb/ulp/tom cxgb_cpl_io.c cxgb_cpl_socket.c cxgb_ddp.c cxgb_defs.h Message-ID: <200803180355.m2I3tCnu037804@repoman.freebsd.org> kmacy 2008-03-18 03:55:12 UTC FreeBSD src repository Modified files: sys/dev/cxgb cxgb_ioctl.h cxgb_main.c sys/dev/cxgb/common cxgb_ael1002.c cxgb_common.h cxgb_version.h sys/dev/cxgb/ulp/tom cxgb_cpl_io.c cxgb_cpl_socket.c cxgb_ddp.c cxgb_defs.h Log: - Integrate 1.133 vendor driver changes - update some copyrights - add improved support for delayed ack - fix issue with fec Revision Changes Path 1.5 +2 -2 src/sys/dev/cxgb/common/cxgb_ael1002.c 1.9 +3 -3 src/sys/dev/cxgb/common/cxgb_common.h 1.6 +2 -2 src/sys/dev/cxgb/common/cxgb_version.h 1.7 +1 -1 src/sys/dev/cxgb/cxgb_ioctl.h 1.51 +45 -16 src/sys/dev/cxgb/cxgb_main.c 1.6 +52 -29 src/sys/dev/cxgb/ulp/tom/cxgb_cpl_io.c 1.6 +2 -2 src/sys/dev/cxgb/ulp/tom/cxgb_cpl_socket.c 1.2 +2 -1 src/sys/dev/cxgb/ulp/tom/cxgb_ddp.c 1.3 +1 -0 src/sys/dev/cxgb/ulp/tom/cxgb_defs.h From dds at aueb.gr Tue Mar 18 05:43:05 2008 From: dds at aueb.gr (Diomidis Spinellis) Date: Tue Mar 18 05:43:14 2008 Subject: cvs commit: src/tools/regression/bin/mv regress.sh In-Reply-To: <20080317235900.GA25950@garage.freebsd.pl> References: <200712180849.lBI8nmEi088947@repoman.freebsd.org> <20071218100355.GR16982@elvis.mu.org> <20080314223652.GA20470@garage.freebsd.pl> <47DE8D10.9090201@aueb.gr> <20080317235900.GA25950@garage.freebsd.pl> Message-ID: <47DF5658.7060700@aueb.gr> Pawel Jakub Dawidek wrote: > On Mon, Mar 17, 2008 at 05:24:00PM +0200, Diomidis Spinellis wrote: >> Pawel Jakub Dawidek wrote: >>> I was thinking about adding two options to cp(1) (-1 and -2) to give >>> cp(1) a hint if the copy is done inside one disk or between separate >>> disks. In -1 case cp(1) will read as large blocks as possible and then >>> write them, in -2 case it will spawn two threads: one reader and one >>> writer working in parallel. Performance improvements are very visible >> >from what I tested. >> Adding hints to cp(1) hinders portability. Even on the same OS, if two >> systems have their disks differently configured, a script may end up >> with an inappropriate hint on one of the two. Why not use fstat(2) to >> automatically determine if the file resides on different disks, and act >> accordingly? > > You can't say if this is the same disk or not. Think about more complex > situations like partitions on RAIDs, etc. (Or, even worse, volumes spanning multiple disks). What if cp(1) always worked in multithread mode for large files. Is that a pessimization for copies on the same disk? Don't disk accesses get appropriately reordered lower down? -- Diomidis Spinellis - http://www.spinellis.gr From kientzle at FreeBSD.org Tue Mar 18 06:18:50 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Tue Mar 18 06:18:56 2008 Subject: cvs commit: src/usr.bin/tar Makefile matching.c Message-ID: <200803180618.m2I6InlQ082018@repoman.freebsd.org> kientzle 2008-03-18 06:18:49 UTC FreeBSD src repository Modified files: usr.bin/tar Makefile matching.c Log: Fix --fast-read by decrementing the remaining patterns to be matched, instead of incrementing. Pointy hat: me Thanks to: Tomasz Przygoda MFC after: 3 days Revision Changes Path 1.34 +1 -1 src/usr.bin/tar/Makefile 1.12 +1 -1 src/usr.bin/tar/matching.c From alc at FreeBSD.org Tue Mar 18 06:52:16 2008 From: alc at FreeBSD.org (Alan Cox) Date: Tue Mar 18 06:52:22 2008 Subject: cvs commit: src/sys/conf files src/sys/vm vm_page.c vm_page.h vm_pageq.c Message-ID: <200803180652.m2I6qF1J090102@repoman.freebsd.org> alc 2008-03-18 06:52:15 UTC FreeBSD src repository Modified files: sys/conf files sys/vm vm_page.c vm_page.h Removed files: sys/vm vm_pageq.c Log: Almost seven years ago, vm/vm_page.c was split into three parts: vm/vm_contig.c, vm/vm_page.c, and vm/vm_pageq.c. Today, vm/vm_pageq.c has withered to the point that it contains only four short functions, two of which are only used by vm/vm_page.c. Since I can't foresee any reason for vm/vm_pageq.c to grow, it is time to fold the remaining contents of vm/vm_pageq.c back into vm/vm_page.c. Add some comments. Rename one of the functions, vm_pageq_enqueue(), that is now static within vm/vm_page.c to vm_page_enqueue(). Eliminate PQ_MAXCOUNT as it no longer serves any purpose. Revision Changes Path 1.1281 +0 -1 src/sys/conf/files 1.367 +74 -5 src/sys/vm/vm_page.c 1.153 +1 -4 src/sys/vm/vm_page.h 1.36 +0 -115 src/sys/vm/vm_pageq.c (dead) From peterjeremy at optushome.com.au Tue Mar 18 06:53:20 2008 From: peterjeremy at optushome.com.au (Peter Jeremy) Date: Tue Mar 18 06:53:27 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c In-Reply-To: <200803171317.15812.jhb@freebsd.org> References: <4578.1205772826@critter.freebsd.dk> <200803171317.15812.jhb@freebsd.org> Message-ID: <20080318065315.GR44676@server.vk2pj.dyndns.org> On Mon, Mar 17, 2008 at 01:17:15PM -0400, John Baldwin wrote: >On Monday 17 March 2008 12:53:46 pm Poul-Henning Kamp wrote: >> >> >You have servers that default to half-speed when not on battery? That is >very >> >odd, but you can just run powerd [...] >> >> Yes, but you need to know that you should run powerd in the first place. > >Err, I would just always run it. You are now saying that laptops have to run >powerd to avoid full-speed CPUs on boot (even though the BIOS throttles it >down for you) but you can't run powerd on the server? In general, I agree that powerd should be enabled but actually implementing this may have some adverse side-effects. I know using acpi_throttle with powerd on my HP nx6125 causes it to wedge randomly. -- Peter Jeremy Please excuse any delays as the result of my ISP's inability to implement an MTA that is either RFC2821-compliant or matches their claimed behaviour. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080318/6b18f21d/attachment.pgp From adrian at FreeBSD.org Tue Mar 18 08:36:20 2008 From: adrian at FreeBSD.org (Adrian Chadd) Date: Tue Mar 18 08:36:22 2008 Subject: cvs commit: src/sys/dev/hwpmc hwpmc_amd.c Message-ID: <200803180836.m2I8aJtb070823@repoman.freebsd.org> adrian 2008-03-18 08:36:19 UTC FreeBSD src repository Modified files: sys/dev/hwpmc hwpmc_amd.c Log: Fix the debugging output - the '0x' was duplicated from the %p option. Revision Changes Path 1.15 +1 -1 src/sys/dev/hwpmc/hwpmc_amd.c From adrian at FreeBSD.org Tue Mar 18 08:39:12 2008 From: adrian at FreeBSD.org (Adrian Chadd) Date: Tue Mar 18 08:39:14 2008 Subject: cvs commit: src/sys/dev/hwpmc hwpmc_amd.c Message-ID: <200803180839.m2I8dCgH073280@repoman.freebsd.org> adrian 2008-03-18 08:39:12 UTC FreeBSD src repository Modified files: sys/dev/hwpmc hwpmc_amd.c Log: Sign-extend the 48-bit AMD PMC counter before treating it to a 64-bit 2's compliment. The 2's compliment transform is done so a "count down" sampling interval can be converted into a "count up" PMC value. a 2's complimented 'count down' value is written to the PMC counter; then the read-back counter is reverted via another 2's compliment. PR: kern/121660 Reviewed by: jkoshy Approved by: jkoshy MFC after: 1 week Revision Changes Path 1.16 +8 -5 src/sys/dev/hwpmc/hwpmc_amd.c From lulf at FreeBSD.org Tue Mar 18 08:48:52 2008 From: lulf at FreeBSD.org (Ulf Lilleengen) Date: Tue Mar 18 08:48:56 2008 Subject: cvs commit: src/sys/geom/vinum geom_vinum_drive.c Message-ID: <200803180848.m2I8mpwD089411@repoman.freebsd.org> lulf 2008-03-18 08:48:51 UTC FreeBSD src repository Modified files: sys/geom/vinum geom_vinum_drive.c Log: - Fix a memory leak when re-discovering a gvinum configuration. Approved by: pjd (mentor) MFC after: 1 week Revision Changes Path 1.27 +2 -0 src/sys/geom/vinum/geom_vinum_drive.c From rwatson at FreeBSD.org Tue Mar 18 11:37:58 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Tue Mar 18 11:38:03 2008 Subject: cvs commit: src/sys/fs/coda TODO Message-ID: <200803181137.m2IBbwca049324@repoman.freebsd.org> rwatson 2008-03-18 11:37:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda TODO Log: Merge TODO:1.4 from HEAD to RELENG_7: Remove namecache performance-tuning todo for Coda: we now use the FreeBSD name cache. Revision Changes Path 1.3.2.2 +0 -1 src/sys/fs/coda/TODO From rwatson at FreeBSD.org Tue Mar 18 11:40:49 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Tue Mar 18 11:40:51 2008 Subject: cvs commit: src/sys/fs/coda TODO cnode.h coda_kernel.h coda_psdev.c Message-ID: <200803181140.m2IBenKL050596@repoman.freebsd.org> rwatson 2008-03-18 11:40:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/coda TODO cnode.h coda_psdev.c Removed files: (Branch: RELENG_7) sys/fs/coda coda_kernel.h Log: Merge TODO:1.5, cnode.h:1.29, coda_kernel.h:1.8, coda_psdev.c:1.48 from HEAD to RELENG_7: Remove custom queue macros in Coda, replacing them with queue(9) tailq macros. The only semantic change was the need to add a vc_opened field to struct vcomm since we can no longer use the request queue returning to an uninitialized state to hold whether or not the device is open. Revision Changes Path 1.3.2.3 +0 -3 src/sys/fs/coda/TODO 1.21.2.8 +8 -12 src/sys/fs/coda/cnode.h 1.6.2.2 +0 -67 src/sys/fs/coda/coda_kernel.h (dead) 1.39.2.9 +22 -23 src/sys/fs/coda/coda_psdev.c From rwatson at FreeBSD.org Tue Mar 18 11:45:30 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Tue Mar 18 11:45:36 2008 Subject: cvs commit: src/sys/ufs/ffs ffs_vfsops.c Message-ID: <200803181145.m2IBjUJm050736@repoman.freebsd.org> rwatson 2008-03-18 11:45:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ufs/ffs ffs_vfsops.c Log: Merge ffs_vfsops.c:1.337 from HEAD to RELENG_7: Move setting of MNTK_MPSAFE flag before UFS1 extended attribute auto-start so that the flag is set before we start performing I/O in the auto-start routine. Suggested by: kib Revision Changes Path 1.329.2.2 +3 -3 src/sys/ufs/ffs/ffs_vfsops.c From ru at FreeBSD.org Tue Mar 18 11:49:11 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Tue Mar 18 11:49:17 2008 Subject: cvs commit: src/sys/kern link_elf.c link_elf_obj.c Message-ID: <200803181149.m2IBnBed051595@repoman.freebsd.org> ru 2008-03-18 11:49:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern link_elf.c link_elf_obj.c Log: MFC: Fix panic on e.g. "kldload /dev/null". PR: kern/121427 Revision Changes Path 1.93.2.1 +5 -0 src/sys/kern/link_elf.c 1.95.2.1 +4 -0 src/sys/kern/link_elf_obj.c From jhb at FreeBSD.org Tue Mar 18 13:31:46 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 18 13:31:47 2008 Subject: cvs commit: src/sys/sun4v/sun4v intr_machdep.c Message-ID: <200803181331.m2IDVjTP047636@repoman.freebsd.org> jhb 2008-03-18 13:31:45 UTC FreeBSD src repository Modified files: sys/sun4v/sun4v intr_machdep.c Log: Catch up to intr_event_create() prototype change. Pointy hat: jhb Revision Changes Path 1.9 +1 -1 src/sys/sun4v/sun4v/intr_machdep.c From mav at FreeBSD.org Tue Mar 18 14:08:07 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 18 14:08:16 2008 Subject: cvs commit: src/usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c In-Reply-To: <47DB1EC8.2040100@elischer.org> References: <200803141611.m2EGBLwF076696@repoman.freebsd.org> <47DB1EC8.2040100@elischer.org> Message-ID: <47DFBEB2.7080500@FreeBSD.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Julian Elischer wrote: > Maksim Yevmenkin wrote: >> emax 2008-03-14 16:11:20 UTC >> >> FreeBSD src repository >> >> Modified files: (Branch: RELENG_6) >> usr.sbin/bluetooth/rfcomm_pppd rfcomm_pppd.8 rfcomm_pppd.c Log: >> MFC: >> Add an option to register DUN (Dial-Up Networking) service on the >> same >> RFCOMM channel if needed. There is really no good reason to not to >> support >> this. AT-command exchange can be faked with chat script in ppp.conf. >> Revision Changes Path >> 1.10.2.5 +10 -1 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.8 >> 1.5.8.2 +33 -2 src/usr.sbin/bluetooth/rfcomm_pppd/rfcomm_pppd.c Thanks, Maksim! At last I have got the way to connect my WM6 Pocket to the internet via my FreeBSD machine! After some time on writing fake chat script it has got working. PS: Waiting to play with PAN. ;) > how many barriers are there from stopping a computer from doing > "drive-by" downloads and connections from random passing cell phones.. > obviously they need to associate, but how many layers of security are > there? AFAIK the first barrier will be the bluetooth device pairing PIN code and the second optional one is usual PPP authentication. - -- Alexander Motin -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (FreeBSD) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFH376x0kCgngV3usoRAp5QAKC7CswREm1NKFi1yl6iTkEbOW7LnACg0SUa u1SXc9jDCkgFWHYT/dIQf+Y= =iiEf -----END PGP SIGNATURE----- From piso at FreeBSD.org Tue Mar 18 14:39:06 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Tue Mar 18 14:39:12 2008 Subject: cvs commit: src/sys/netinet/libalias alias_db.c alias_irc.c Message-ID: <200803181439.m2IEd51u084153@repoman.freebsd.org> piso 2008-03-18 14:39:05 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netinet/libalias alias_db.c alias_irc.c Log: MFC Revision 1.25: Don't abuse stack space while in kernel land, use heap instead. PR: kern/118432 Revision Changes Path 1.67.2.2 +10 -0 src/sys/netinet/libalias/alias_db.c 1.21.2.1 +7 -2 src/sys/netinet/libalias/alias_irc.c From ume at FreeBSD.org Tue Mar 18 15:04:07 2008 From: ume at FreeBSD.org (Hajimu UMEMOTO) Date: Tue Mar 18 15:04:08 2008 Subject: cvs commit: src/usr.bin/netstat inet6.c Message-ID: <200803181504.m2IF46AG097060@repoman.freebsd.org> ume 2008-03-18 15:04:06 UTC FreeBSD src repository Modified files: usr.bin/netstat inet6.c Log: Change .8s port name restriction to .15s. This change corresponds to inet.c 1.13. MFC after: 1 week Revision Changes Path 1.31 +1 -1 src/usr.bin/netstat/inet6.c From mav at FreeBSD.org Tue Mar 18 17:31:12 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 18 17:31:18 2008 Subject: cvs commit: src/sys/netgraph ng_l2tp.c Message-ID: <200803181731.m2IHVBxU004005@repoman.freebsd.org> mav 2008-03-18 17:31:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_l2tp.c Log: MFC rev. 1.25 Add session ID hashing to speedup incoming packets dispatch in case of many connections working via the same tunnel. For example, in case of full "client <-> LAC <-> LNS" setup. Revision Changes Path 1.17.2.4 +47 -38 src/sys/netgraph/ng_l2tp.c From piso at FreeBSD.org Tue Mar 18 17:32:12 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Tue Mar 18 17:32:13 2008 Subject: cvs commit: src/sys/conf NOTES files src/sys/modules Makefile src/sys/modules/ipfw_nat Makefile src/sys/netinet ip_fw.h ip_fw2.c ip_fw_nat.c Message-ID: <200803181732.m2IHWBsW004129@repoman.freebsd.org> piso 2008-03-18 17:32:10 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf NOTES files sys/modules Makefile sys/netinet ip_fw.h ip_fw2.c Added files: (Branch: RELENG_7) sys/modules/ipfw_nat Makefile sys/netinet ip_fw_nat.c Log: MFC: -Move ipfw's nat code into its own kld: ipfw_nat. -Raise a bit ipfw kld priority to avoid race on ipfw_nat load. Revision Changes Path 1.1454.2.3 +1 -2 src/sys/conf/NOTES 1.1243.2.7 +1 -0 src/sys/conf/files 1.540.2.7 +1 -0 src/sys/modules/Makefile 1.1.2.1 +8 -0 src/sys/modules/ipfw_nat/Makefile (new) 1.110.2.2 +32 -0 src/sys/netinet/ip_fw.h 1.175.2.4 +52 -550 src/sys/netinet/ip_fw2.c 1.2.2.1 +653 -0 src/sys/netinet/ip_fw_nat.c (new) From mav at FreeBSD.org Tue Mar 18 17:33:03 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 18 17:33:06 2008 Subject: cvs commit: src/sys/netgraph ng_l2tp.c Message-ID: <200803181733.m2IHX3VD004429@repoman.freebsd.org> mav 2008-03-18 17:33:03 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netgraph ng_l2tp.c Log: MFC rev. 1.25 Add session ID hashing to speedup incoming packets dispatch in case of many connections working via the same tunnel. For example, in case of full "client <-> LAC <-> LNS" setup. Revision Changes Path 1.14.2.5 +47 -38 src/sys/netgraph/ng_l2tp.c From mav at FreeBSD.org Tue Mar 18 17:39:43 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Tue Mar 18 17:39:44 2008 Subject: cvs commit: src/sys/netgraph netgraph.h ng_base.c ng_socket.c Message-ID: <200803181739.m2IHdg8g008529@repoman.freebsd.org> mav 2008-03-18 17:39:42 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/netgraph netgraph.h ng_base.c ng_socket.c Log: MFC: Improve apply callback error reporting: Before this patch callback returned result of the last finished call chain. Now it returns last nonzero result from all call chain results in this request. As soon as this improvement gives reliable error reporting, it is now possible to remove dirty workaround in ng_socket, made to return ENOBUFS error statuses of request-response operations. That workaround was responsible for returning ENOBUFS errors to completely unrelated requests working at the same time on socket. Revision Changes Path 1.55.2.12 +2 -1 src/sys/netgraph/netgraph.h 1.102.2.18 +25 -8 src/sys/netgraph/ng_base.c 1.67.2.8 +1 -1 src/sys/netgraph/ng_socket.c From jhb at FreeBSD.org Tue Mar 18 17:43:37 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 18 17:43:43 2008 Subject: cvs commit: src/sys/boot/i386/btx/btx Makefile btx.S Message-ID: <200803181743.m2IHhaS2008826@repoman.freebsd.org> jhb 2008-03-18 17:43:36 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/i386/btx/btx Makefile btx.S Log: MFC: Use real mode to invoke BIOS routines rather than virtual 86 mode. Revision Changes Path 1.19.10.1 +0 -4 src/sys/boot/i386/btx/btx/Makefile 1.44.2.1 +318 -454 src/sys/boot/i386/btx/btx/btx.S From jhb at FreeBSD.org Tue Mar 18 17:44:29 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 18 17:44:30 2008 Subject: cvs commit: src/sys/boot/i386/btx/btx Makefile btx.S Message-ID: <200803181744.m2IHiSA4008941@repoman.freebsd.org> jhb 2008-03-18 17:44:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/boot/i386/btx/btx Makefile btx.S Log: MFC: Use real moder to invoke BIOS routines rather than virtual 86 mode. Revision Changes Path 1.19.2.1 +0 -4 src/sys/boot/i386/btx/btx/Makefile 1.38.2.3 +318 -454 src/sys/boot/i386/btx/btx/btx.S From emax at FreeBSD.org Tue Mar 18 18:21:39 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Tue Mar 18 18:21:45 2008 Subject: cvs commit: src/lib/libsdp sdp.h src/usr.sbin/bluetooth/sdpd gn.c lan.c nap.c panu.c profile.c profile.h Message-ID: <200803181821.m2IILdGC033601@repoman.freebsd.org> emax 2008-03-18 18:21:39 UTC FreeBSD src repository Modified files: lib/libsdp sdp.h usr.sbin/bluetooth/sdpd gn.c lan.c nap.c panu.c profile.c profile.h Log: Add PSM and Load Factor SDP parameters to the BNEP based profiles (NAP, GN and PANU). No reason to not to support them. Separate SDP parameters data structures for the BNEP based profiles. Generalize Service Availability SDP parameter creation. Requested by: Iain Hibbert < plunky at rya-online dot net > MFC after: 3 days Revision Changes Path 1.9 +25 -10 src/lib/libsdp/sdp.h 1.2 +28 -2 src/usr.sbin/bluetooth/sdpd/gn.c 1.2 +2 -7 src/usr.sbin/bluetooth/sdpd/lan.c 1.2 +28 -2 src/usr.sbin/bluetooth/sdpd/nap.c 1.2 +28 -2 src/usr.sbin/bluetooth/sdpd/panu.c 1.4 +23 -2 src/usr.sbin/bluetooth/sdpd/profile.c 1.3 +1 -0 src/usr.sbin/bluetooth/sdpd/profile.h From jhb at FreeBSD.org Tue Mar 18 18:28:10 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Mar 18 18:28:20 2008 Subject: cvs commit: src/sys/boot/i386/btx/btx Makefile btx.s Message-ID: <200803181828.m2IIS9XL038440@repoman.freebsd.org> jhb 2008-03-18 18:28:09 UTC FreeBSD src repository Modified files: (Branch: RELENG_4) sys/boot/i386/btx/btx Makefile btx.s Log: MFC: Use real moder to invoke BIOS routines rather than virtual 86 mode. Revision Changes Path 1.7.2.3 +0 -4 src/sys/boot/i386/btx/btx/Makefile 1.15.2.5 +346 -473 src/sys/boot/i386/btx/btx/btx.s From dfr at FreeBSD.org Tue Mar 18 18:43:00 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Tue Mar 18 18:43:02 2008 Subject: cvs commit: src/sys/nfsserver nfs_srvsock.c Message-ID: <200803181843.m2IIh0kM048215@repoman.freebsd.org> dfr 2008-03-18 18:42:59 UTC FreeBSD src repository Modified files: sys/nfsserver nfs_srvsock.c Log: Don't call nfs_realign while holding locks. Reviewed by: kib Revision Changes Path 1.106 +5 -4 src/sys/nfsserver/nfs_srvsock.c From thompsa at FreeBSD.org Tue Mar 18 18:52:53 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Tue Mar 18 18:52:56 2008 Subject: cvs commit: src/sys/dev/wpi if_wpi.c if_wpireg.h if_wpivar.h Message-ID: <200803181852.m2IIqrsN056899@repoman.freebsd.org> thompsa 2008-03-18 18:52:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/wpi if_wpi.c if_wpireg.h if_wpivar.h Log: MFC if_wpi.c r1.9-10, if_wpireg.h r1.3, if_wpivar.h r1.2 Update wpi(4) with stability fixes - remove second taskqueue - busdma 16k alignment workaround - use busdma instead of external mbuf storage on Rx - locking fixes - net80211 state change fixes - improve scanning reliability - improve radio hw switch interaction - consolidate callouts Revision Changes Path 1.5.2.4 +480 -553 src/sys/dev/wpi/if_wpi.c 1.2.2.2 +2 -0 src/sys/dev/wpi/if_wpireg.h 1.1.2.2 +31 -53 src/sys/dev/wpi/if_wpivar.h From vadim_nuclight at mail.ru Tue Mar 18 19:38:11 2008 From: vadim_nuclight at mail.ru (Vadim Goncharov) Date: Tue Mar 18 19:38:14 2008 Subject: cvs commit: src/sys/netinet/libalias alias_irc.c In-Reply-To: <200803172216.m2HMGJKj011742@repoman.freebsd.org> References: <200803172216.m2HMGJKj011742@repoman.freebsd.org> Message-ID: <200803181920.m2IJKcoD038230@hostel.avtf.net> Hi Paolo Pisati! On Mon, 17 Mar 2008 22:16:19 +0000 (UTC); Paolo Pisati wrote: > piso 2008-03-17 22:16:19 UTC > FreeBSD src repository > Modified files: (Branch: RELENG_7) > sys/netinet/libalias alias_irc.c > Log: > MFC revision 1.25: > Don't abuse stack space while in kernel land, use heap instead. > > PR: kern/121693 > > Revision Changes Path > 1.23.2.1 +10 -5 src/sys/netinet/libalias/alias_irc.c Paolo, this violates rules 6 and 10 from committers-guide. The original commit log in HEAD did not contained "MFC After:", so the code was not supposed to be MFCed to STABLE. Moreover, minimum MFC period is 3 days, and you've done MFCs to both 6-STABLE and 7-STABLE in 8 minutes after commit to HEAD. Then, PR must not be closed so fast, as originator has not yet confirmed that patch has fixed his problem. This could be non-critical, but that rules exists to guard against bugs in the code, so - let's look at the code. You are changing "char newpacket[65536]" to "char *newpacket", and there are places there like: if (iCopy + 4 > sizeof(newpacket)) goto lPACKET_DONE; which you hasn't changed, while "iCopy" is "unsigned int". This will break behaviour, of course, in 7-STABLE - in 6-STABLE it is still an array. It is questionable whether silent change in size from 65536 to IP_MAXPACKET (which is one less) is acceptable, but there is also another clear point: this will be allocated with M_NOWAIT on every packet. It is better to prealloc that memory on module load, the way you've done in 6-STABLE's MFC variant. -- WBR, Vadim Goncharov. ICQ#166852181 mailto:vadim_nuclight@mail.ru [Moderator of RU.ANTI-ECOLOGY][FreeBSD][http://antigreen.org][LJ:/nuclight] From kostikbel at gmail.com Tue Mar 18 21:05:12 2008 From: kostikbel at gmail.com (Kostik Belousov) Date: Tue Mar 18 21:05:16 2008 Subject: cvs commit: src/sys/nfsserver nfs_srvsock.c In-Reply-To: <200803181843.m2IIh0kM048215@repoman.freebsd.org> References: <200803181843.m2IIh0kM048215@repoman.freebsd.org> Message-ID: <20080318210505.GF10374@deviant.kiev.zoral.com.ua> On Tue, Mar 18, 2008 at 06:42:59PM +0000, Doug Rabson wrote: > dfr 2008-03-18 18:42:59 UTC > > FreeBSD src repository > > Modified files: > sys/nfsserver nfs_srvsock.c > Log: > Don't call nfs_realign while holding locks. > > Reviewed by: kib > > Revision Changes Path > 1.106 +5 -4 src/sys/nfsserver/nfs_srvsock.c Any plans for the MFC ? -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080318/f084abdf/attachment.pgp From sam at FreeBSD.org Tue Mar 18 21:45:28 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Tue Mar 18 21:45:30 2008 Subject: cvs commit: src/usr.sbin/arp arp.8 arp.c Message-ID: <200803182145.m2ILjRsM020076@repoman.freebsd.org> sam 2008-03-18 21:45:27 UTC FreeBSD src repository Modified files: usr.sbin/arp arp.8 arp.c Log: add reject+blackhole keywords to install entries with RTF_BLACKHOLE and RTF_REJECT, respectively PR: bin/79228 Submitted by: Dan Lukes MFC after: 2 weeks Revision Changes Path 1.26 +16 -0 src/usr.sbin/arp/arp.8 1.67 +7 -4 src/usr.sbin/arp/arp.c From piso at FreeBSD.org Tue Mar 18 23:32:46 2008 From: piso at FreeBSD.org (Paolo Pisati) Date: Tue Mar 18 23:32:47 2008 Subject: cvs commit: src/sys/netinet ip_fw2.c Message-ID: <200803182332.m2INWjFe012193@repoman.freebsd.org> piso 2008-03-18 23:32:45 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet ip_fw2.c Log: Fix printf args. Revision Changes Path 1.175.2.5 +1 -1 src/sys/netinet/ip_fw2.c From emax at FreeBSD.org Wed Mar 19 00:06:33 2008 From: emax at FreeBSD.org (Maksim Yevmenkin) Date: Wed Mar 19 00:06:34 2008 Subject: cvs commit: src/lib/libsdp sdp.h src/usr.sbin/bluetooth/sdpd panu.c Message-ID: <200803190006.m2J06VJc042811@repoman.freebsd.org> emax 2008-03-19 00:06:30 UTC FreeBSD src repository Modified files: lib/libsdp sdp.h usr.sbin/bluetooth/sdpd panu.c Log: Add mandatory "security description" SDP parameter to the PANU profile Pointed-out by: Iain Hibbert < plunky at rya-online dot net > MFC after: 3 days Revision Changes