From jb at FreeBSD.org Sun Jun 1 01:47:02 2008 From: jb at FreeBSD.org (John Birrell) Date: Sun Jun 1 01:47:05 2008 Subject: cvs commit: src/sys/cddl/contrib/opensolaris/uts/common/dtrace dtrace.c src/sys/cddl/contrib/opensolaris/uts/common/sys dtrace_impl.h Message-ID: <200806010147.m511l2QS035827@repoman.freebsd.org> jb 2008-06-01 01:47:02 UTC FreeBSD src repository Modified files: sys/cddl/contrib/opensolaris/uts/common/dtrace dtrace.c sys/cddl/contrib/opensolaris/uts/common/sys dtrace_impl.h Log: SVN rev 179469 on 2008-06-01 01:46:37Z by jb Merge a recent change from the OpenSolaris source tree. (Don't ask for a vendor import of this yet, we're in the early days of svn) Instead of using cyclic timers to call the state clean and deadman callbacks, use a callout on FreeBSD to avoid the deadlock on FreeBSD due to trying to send interprocessor interrupts with interrupts disabled. Reported by: ps, jhb, peter, thompsa Revision Changes Path 1.4 +89 -13 src/sys/cddl/contrib/opensolaris/uts/common/dtrace/dtrace.c 1.3 +7 -0 src/sys/cddl/contrib/opensolaris/uts/common/sys/dtrace_impl.h From alc at FreeBSD.org Sun Jun 1 07:37:18 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sun Jun 1 07:37:21 2008 Subject: cvs commit: src/sys/amd64/amd64 pmap.c Message-ID: <200806010737.m517bIff079854@repoman.freebsd.org> alc 2008-06-01 07:37:18 UTC FreeBSD src repository Modified files: sys/amd64/amd64 pmap.c Log: SVN rev 179471 on 2008-06-01 07:36:59Z by alc Correct an error in pmap_promote_pde() that may result in an errant promotion within the kernel's address space. Specifically, pmap_promote_pde() is only called when the page table page (PTP) that is referenced by the given PDE has a full "use count", i.e., its wire_count is 512. Although this guarantees for a user address space that all 512 PTEs in the PTP hold valid mappings, the same is not true of the kernel's address space. A kernel PTP always has a use count of 512 regardless of the state of the PTEs. Therefore, pmap_promote_pde() should not assume (or assert) that the first PTE in the PTP is valid. Revision Changes Path 1.619 +3 -5 src/sys/amd64/amd64/pmap.c From mav at FreeBSD.org Sun Jun 1 11:47:22 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 1 11:47:25 2008 Subject: cvs commit: src/sys/netinet/libalias alias.c Message-ID: <200806011147.m51BlLVc017957@repoman.freebsd.org> mav 2008-06-01 11:47:21 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias.c Log: SVN rev 179472 on 2008-06-01 11:47:04Z by mav Fix packet fragmentation support broken by copy/paste error in rev.1.60. ip_id should be u_short, but not u_char. Revision Changes Path 1.62 +2 -2 src/sys/netinet/libalias/alias.c From mav at FreeBSD.org Sun Jun 1 12:29:38 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 1 12:29:42 2008 Subject: cvs commit: src/sys/netinet ip_fw_nat.c Message-ID: <200806011229.m51CTcNl026391@repoman.freebsd.org> mav 2008-06-01 12:29:38 UTC FreeBSD src repository Modified files: sys/netinet ip_fw_nat.c Log: SVN rev 179473 on 2008-06-01 12:29:23Z by mav PKT_ALIAS_FOUND_HEADER_FRAGMENT result is not an error, so pass that packet. This fixes packet fragmentation handeling. Pass really available buffer size to libalias instead of MCLBYTES constant. MCLBYTES constant were used with believe that m_megapullup() always moves date into a fresh cluster that sometimes may become not so. Revision Changes Path 1.3 +4 -3 src/sys/netinet/ip_fw_nat.c From philip at FreeBSD.org Sun Jun 1 13:45:14 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sun Jun 1 13:45:20 2008 Subject: cvs commit: src/sys/dev/atkbdc psm.c Message-ID: <200806011345.m51DjEMQ043301@repoman.freebsd.org> philip 2008-06-01 13:45:14 UTC FreeBSD src repository Modified files: sys/dev/atkbdc psm.c Log: SVN rev 179474 on 2008-06-01 13:44:51Z by philip Try to detect a Synaptics touchpad before IntelliMouse. Some touchpads will pretend to be IntelliMouse (which have a few more features than generic mice) causing the IntelliMouse probe to work and the Synaptics code never to be called. This should not break "real" IntelliMouse because the Synaptics detection code is fairly specific. PR: kern/120833 Submitted by: Eygene Ryabinkin MFC after: 1 week Revision Changes Path 1.97 +2 -2 src/sys/dev/atkbdc/psm.c From ed at FreeBSD.org Sun Jun 1 14:03:01 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Sun Jun 1 14:03:05 2008 Subject: cvs commit: src/sys/fs/devfs devfs_vnops.c Message-ID: <200806011403.m51E30hN044784@repoman.freebsd.org> ed 2008-06-01 14:03:00 UTC FreeBSD src repository Modified files: sys/fs/devfs devfs_vnops.c Log: SVN rev 179475 on 2008-06-01 14:02:46Z by ed Revert the changes I made to devfs_setattr() in r179457. As discussed with Robert Watson and John Baldwin, it would be better if PTY's are created with proper permissions, turning grantpt() into a no-op. Bypassing security frameworks like MAC by passing NOCRED to VOP_SETATTR() will only make things more complex. Approved by: philip (mentor) Revision Changes Path 1.161 +3 -4 src/sys/fs/devfs/devfs_vnops.c From philip at FreeBSD.org Sun Jun 1 14:10:01 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sun Jun 1 14:10:05 2008 Subject: cvs commit: src/usr.sbin/bsnmpd/modules/snmp_pf pf_snmp.c Message-ID: <200806011410.m51EA19Y046588@repoman.freebsd.org> philip 2008-06-01 14:10:01 UTC FreeBSD src repository Modified files: usr.sbin/bsnmpd/modules/snmp_pf pf_snmp.c Log: SVN rev 179476 on 2008-06-01 14:09:54Z by philip Use INSERT_OBJECT_INT_LINK_INDEX macro instead of TAILQ_INSERT_TAIL when filling the table of ALTQ queues retrieved from the kernel. It is possible for the kernel to return the queues not by pa.altq.qid order. When this happens, pf_snmp would only partially fill its table. PR: bin/120974 Submitted by: Mykola Dzham MFC after: 3 days Revision Changes Path 1.9 +1 -1 src/usr.sbin/bsnmpd/modules/snmp_pf/pf_snmp.c From mav at FreeBSD.org Sun Jun 1 15:13:54 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 1 15:13:55 2008 Subject: cvs commit: src/sys/netgraph ng_nat.c Message-ID: <200806011513.m51FDsGx010741@repoman.freebsd.org> mav 2008-06-01 15:13:52 UTC FreeBSD src repository Modified files: sys/netgraph ng_nat.c Log: SVN rev 179477 on 2008-06-01 15:13:32Z by mav Pass really available buffer size to libalias instead of MCLBYTES constant. MCLBYTES constant were used with believe that m_megapullup() always moves date into a fresh cluster that may become not so. Revision Changes Path 1.12 +2 -2 src/sys/netgraph/ng_nat.c From mav at FreeBSD.org Sun Jun 1 17:52:56 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 1 17:52:59 2008 Subject: cvs commit: src/sys/netinet/libalias alias.c Message-ID: <200806011752.m51HquO4065483@repoman.freebsd.org> mav 2008-06-01 17:52:56 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias.c Log: SVN rev 179478 on 2008-06-01 17:52:40Z by mav Make m_megapullup() more intelligent: - to increase performance do not reallocate mbuf when possible, - to support up to 16K packets (was 2K max) use mbuf cluster of proper size. This change depends on recent ng_nat and ip_fw_nat changes. Revision Changes Path 1.63 +32 -12 src/sys/netinet/libalias/alias.c From marcel at FreeBSD.org Sun Jun 1 18:04:43 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Jun 1 18:04:45 2008 Subject: cvs commit: src/sys/ia64/ia64 machdep.c Message-ID: <200806011804.m51I4hl9069221@repoman.freebsd.org> marcel 2008-06-01 18:04:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ia64/ia64 machdep.c Log: SVN rev 179479 on 2008-06-01 18:04:29Z by marcel Merge rev 179173: We can call ia64_flush_dirty() when the corresponding process is locked or not. As such, use PROC_LOCKED() to determine which case it is and lock the process when not. This is a manual merge. No merge history is created. Revision Changes Path 1.225.2.6 +6 -3 src/sys/ia64/ia64/machdep.c From mav at FreeBSD.org Sun Jun 1 18:35:19 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 1 18:35:24 2008 Subject: cvs commit: src/sys/netinet/libalias alias_db.c alias_local.h Message-ID: <200806011835.m51IZJp5078001@repoman.freebsd.org> mav 2008-06-01 18:35:19 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias_db.c alias_local.h Log: SVN rev 179480 on 2008-06-01 18:34:58Z by mav Increase LINK_TABLE_OUT_SIZE from 101 to 4001 like LINK_TABLE_IN_SIZE to reduce performance degradation under heavy outgoing scan/flood. Scalability is now much more important then several kilobytes of RAM. Remove unneded TCP-specific expiration handeling. Before this connected TCP sessions could never expire. Now connected TCP sessions will expire after 24hours of inactivity. Simplify HouseKeeping() to avoid several mul/div-s per packet. Taking into account increased LINK_TABLE_OUT_SIZE, precision is still much more then required. Revision Changes Path 1.73 +22 -67 src/sys/netinet/libalias/alias_db.c 1.36 +1 -3 src/sys/netinet/libalias/alias_local.h From imp at FreeBSD.org Sun Jun 1 20:55:38 2008 From: imp at FreeBSD.org (Warner Losh) Date: Sun Jun 1 20:55:42 2008 Subject: cvs commit: src/sys/dev/pccard pccard_cis.c Message-ID: <200806012055.m51KtcKx008726@repoman.freebsd.org> imp 2008-06-01 20:55:38 UTC FreeBSD src repository Modified files: sys/dev/pccard pccard_cis.c Log: SVN rev 179483 on 2008-06-01 20:55:34Z by imp Slight simplification of the power parsing code, as well as using autoincrement in some places where it makes sense. This makes this .o about 180 bytes smaller on x86 and amd64 with no apparent functional changes. Revision Changes Path 1.41 +27 -45 src/sys/dev/pccard/pccard_cis.c From thompsa at FreeBSD.org Mon Jun 2 02:30:25 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Mon Jun 2 02:30:28 2008 Subject: cvs commit: src/sys/net bridgestp.c bridgestp.h Message-ID: <200806020230.m522UNfI053739@repoman.freebsd.org> thompsa 2008-06-02 02:30:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/net bridgestp.c bridgestp.h Log: SVN rev 179484 on 2008-06-02 02:30:02Z by thompsa MFC r177993 Fix spelling, retaining old define. Revision Changes Path 1.3.2.14 +7 -7 src/sys/net/bridgestp.c 1.12.2.4 +2 -1 src/sys/net/bridgestp.h From imp at FreeBSD.org Mon Jun 2 04:50:56 2008 From: imp at FreeBSD.org (Warner Losh) Date: Mon Jun 2 04:50:59 2008 Subject: cvs commit: src/usr.sbin/timed/timed timed.c Message-ID: <200806020450.m524ouNc076284@repoman.freebsd.org> imp 2008-06-02 04:50:56 UTC FreeBSD src repository Modified files: usr.sbin/timed/timed timed.c Log: SVN rev 179485 on 2008-06-02 04:50:47Z by imp getopt returns an int, not a char. Make sure that we store the variable in an int to avoid casting to an unsigned value which causes the comparison with -1 to fail. PR: 123807 Submitted by: Matthew Luckie Reviewed by: keramida@ MFC after: 1 week Revision Changes Path 1.12 +1 -1 src/usr.sbin/timed/timed/timed.c From ed at FreeBSD.org Mon Jun 2 08:40:21 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Mon Jun 2 08:40:24 2008 Subject: cvs commit: src/sys/compat/linux linux_stats.c linux_util.c Message-ID: <200806020840.m528eLha099850@repoman.freebsd.org> ed 2008-06-02 08:40:21 UTC FreeBSD src repository Modified files: sys/compat/linux linux_stats.c linux_util.c Log: SVN rev 179486 on 2008-06-02 08:40:06Z by ed Push down the major/minor conversion for pts/%u to improve consistency. In the mpsafetty branch, Linux sshd seems to work properly inside a jail. Some small modifications had to be made to the Linux compatibility layer. The Linux PTY routines always expect the device major number to be 136 or higher. Our code always set the major/minor number pair to 136:0. This makes routines like ttyname() and ptsname() fail, because we'll end up having ambiguous device numbers. The conversion was not performed on all *stat() routines, which meant in some cases the numbers didn't get transformed. By pushing the conversion into linux_driver_get_major_minor(), the transformation will take place on all calls. Approved by: philip (mentor), rdivacky Revision Changes Path 1.92 +4 -26 src/sys/compat/linux/linux_stats.c 1.35 +16 -0 src/sys/compat/linux/linux_util.c From rwatson at FreeBSD.org Mon Jun 2 14:20:44 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Jun 2 14:20:48 2008 Subject: cvs commit: src/sys/netinet tcp_subr.c tcp_timer.c Message-ID: <200806021420.m52EKiB2056043@repoman.freebsd.org> rwatson 2008-06-02 14:20:44 UTC FreeBSD src repository Modified files: sys/netinet tcp_subr.c tcp_timer.c Log: SVN rev 179487 on 2008-06-02 14:20:26Z by rwatson When allocating temporary storage to hold a TCP/IP packet header template, use an M_TEMP malloc(9) allocation rather than an mbuf with mtod(9) and dtom(9). This eliminates the last use of dtom(9) in TCP. MFC after: 3 weeks Revision Changes Path 1.308 +6 -10 src/sys/netinet/tcp_subr.c 1.101 +1 -1 src/sys/netinet/tcp_timer.c From jhb at FreeBSD.org Mon Jun 2 14:58:28 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Jun 2 14:58:36 2008 Subject: cvs commit: src/sys/ia64/ia64 machdep.c In-Reply-To: <200806011804.m51I4hl9069221@repoman.freebsd.org> References: <200806011804.m51I4hl9069221@repoman.freebsd.org> Message-ID: <200806021037.13943.jhb@FreeBSD.org> On Sunday 01 June 2008 02:04:29 pm Marcel Moolenaar wrote: > marcel 2008-06-01 18:04:43 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7) > sys/ia64/ia64 machdep.c > Log: > SVN rev 179479 on 2008-06-01 18:04:29Z by marcel > > Merge rev 179173: > > We can call ia64_flush_dirty() when the corresponding process is > locked or not. As such, use PROC_LOCKED() to determine which case > it is and lock the process when not. > > This is a manual merge. No merge history is created. proc_rwmem() can sleep (if it has to fault a page back in from swap), so you can't call it with the process locked. I think the only place where you have to worry about this is during a core dump from ia64/elf_machdep.c? In that case it should be fine to just drop the lock. You do still need to do PHOLD/PRELE to avoid an assertion failure and to prevent the process from being swapped out while it is sleeping. So maybe something more like: locked = PROC_LOCKED(); if (!locked) PROC_LOCK(); _PHOLD(); PROC_UNLOCK(); ... (proc_rwmem()) PROC_LOCK(); _PRELE(); if (!locked) PROC_UNLOCK(); -- John Baldwin From dfr at FreeBSD.org Mon Jun 2 15:59:22 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 2 15:59:24 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806021559.m52FxMN1066383@repoman.freebsd.org> dfr 2008-06-02 15:59:22 UTC FreeBSD src repository Modified files: sys/nlm nlm_prot_impl.c Log: SVN rev 179488 on 2008-06-02 15:59:10Z by dfr When attempting to use the NSM state number in a lock request to detect a client reboot, do this check before performing the lock otherwise we will trash the new lock along with any other old locks the client held before rebooting. Make sure nlm_check_idle always returns with nlm_global_lock held. MFC after: 1 week Revision Changes Path 1.8 +10 -8 src/sys/nlm/nlm_prot_impl.c From dfr at FreeBSD.org Mon Jun 2 16:01:09 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 2 16:01:17 2008 Subject: cvs commit: src/usr.sbin/rpc.lockd lockd.c Message-ID: <200806021601.m52G18QV066621@repoman.freebsd.org> dfr 2008-06-02 16:01:08 UTC FreeBSD src repository Modified files: usr.sbin/rpc.lockd lockd.c Log: SVN rev 179489 on 2008-06-02 16:00:49Z by dfr Add a missing call to init_nsm(). MFC after: 1 week Revision Changes Path 1.28 +1 -0 src/usr.sbin/rpc.lockd/lockd.c From mlaier at FreeBSD.org Mon Jun 2 18:58:24 2008 From: mlaier at FreeBSD.org (Max Laier) Date: Mon Jun 2 18:58:27 2008 Subject: cvs commit: src/sys/netinet ip_carp.c Message-ID: <200806021858.m52IwNZF094756@repoman.freebsd.org> mlaier 2008-06-02 18:58:23 UTC FreeBSD src repository Modified files: sys/netinet ip_carp.c Log: SVN rev 179490 on 2008-06-02 18:58:07Z by mlaier Sort IP addresses before hashing them for the signature. Otherwise carp is sensitive to address configuration order. PR: kern/121574 Reported by: Douglas K. Rand, Wouter de Jong Obtained from: OpenBSD (rev 1.114 + fixes) MFC after: 2 weeks Revision Changes Path 1.54 +39 -13 src/sys/netinet/ip_carp.c From jhb at FreeBSD.org Mon Jun 2 19:17:55 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Jun 2 19:18:31 2008 Subject: cvs commit: src/sys/dev/ie if_ie.c if_ie_isa.c if_ievar.h Message-ID: <200806021917.m52JHtuc097380@repoman.freebsd.org> jhb 2008-06-02 19:17:55 UTC FreeBSD src repository Modified files: sys/dev/ie if_ie.c if_ie_isa.c if_ievar.h Log: SVN rev 179491 on 2008-06-02 19:17:40Z by jhb Make ie(4) MPSAFE: - Add a mutex to the softc and use it to protect the softc and device. - Setup the interrupt handler in the common code instead of in each front end and do it after ether_ifattach(). - Use ie_stop() and ieinit_locked() in iereset() rather than frobbing IFF_UP and invoking ieioctl(). - Use DELAY() to implement a spin loop on a register with a timeout rather than scheduling a timeout and then doing a tight spin on the register. In the non-MPSAFE case this would never have worked because the spinning code held Giant and the timeout routine would have been blocked on Giant forever. The same approach would not worke in the MPSAFE case either for the same reason, hence use a loop around DELAY(). - Clear IFF_DRV_(RUNNING|OACTIVE) in ie_stop() rather than in callers. - Call ieinit_locked() directly rather than ieioctl(!) from ie_mc_reset(). - Don't leak the rx frame buffer on detach. Tested by: Thierry Herbelot thierry of herbelot.com Revision Changes Path 1.109 +69 -60 src/sys/dev/ie/if_ie.c 1.8 +0 -21 src/sys/dev/ie/if_ie_isa.c 1.5 +6 -0 src/sys/dev/ie/if_ievar.h From jhb at FreeBSD.org Mon Jun 2 19:43:41 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Jun 2 19:43:44 2008 Subject: cvs commit: src/sys/dev/xe if_xe.c if_xe_pccard.c if_xevar.h Message-ID: <200806021943.m52JhfFm099464@repoman.freebsd.org> jhb 2008-06-02 19:43:41 UTC FreeBSD src repository Modified files: sys/dev/xe if_xe.c if_xe_pccard.c if_xevar.h Log: SVN rev 179492 on 2008-06-02 19:43:24Z by jhb Add locking and make xe(4) MPSAFE: - Add a mutex to protect the softc and device hardware. - Use a callout rather than a callout_handle for the media timer. - Use a dedicated timer for managing the tx watchdog rather than if_timer. - Fix some resource leaks if xe_attach() fails. - Shutdown the device before detaching the driver. - Setup the interrupt handler after ether_ifattach(). Tested by: Ian FREISLICH ianf of clue.co.za Revision Changes Path 1.65 +89 -69 src/sys/dev/xe/if_xe.c 1.33 +7 -3 src/sys/dev/xe/if_xe_pccard.c 1.7 +8 -1 src/sys/dev/xe/if_xevar.h From jhb at FreeBSD.org Mon Jun 2 19:58:57 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Jun 2 19:59:00 2008 Subject: cvs commit: src/sys/dev/fe if_fe.c if_fe_pccard.c if_fevar.h Message-ID: <200806021958.m52Jwv1n000913@repoman.freebsd.org> jhb 2008-06-02 19:58:57 UTC FreeBSD src repository Modified files: sys/dev/fe if_fe.c if_fe_pccard.c if_fevar.h Log: SVN rev 179493 on 2008-06-02 19:58:48Z by jhb Make fe(4) MPSAFE: - Add a mutex to the softc to protect the softc and device hardware. - Don't leak bus resources if if_alloc() fails during attach. - Setup the interrupt handler after calling ether_ifattach(). - Use a private timer to manage the transmit watchdog. Tested by: WATANABE Kazuhiro CQG00620 of nifty.ne.jp Revision Changes Path 1.99 +77 -44 src/sys/dev/fe/if_fe.c 1.33 +4 -0 src/sys/dev/fe/if_fe_pccard.c 1.7 +5 -0 src/sys/dev/fe/if_fevar.h From jhb at freebsd.org Mon Jun 2 20:37:58 2008 From: jhb at freebsd.org (John Baldwin) Date: Mon Jun 2 20:38:00 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <200805151955.m4FJtRMb099378@repoman.freebsd.org> References: <200805151955.m4FJtRMb099378@repoman.freebsd.org> Message-ID: <200806021445.15974.jhb@freebsd.org> On Thursday 15 May 2008 03:55:27 pm Stefan Farfeleder wrote: > stefanf 2008-05-15 19:55:27 UTC > > FreeBSD src repository > > Modified files: > bin/sh expand.c parser.c parser.h > Log: > Expand $LINENO to the current line number. This is required by SUSv3's "User > Portability Utilities" option. > > Often configure scripts generated by the autotools test if $LINENO works and > refuse to use /bin/sh if not. > > Package test run by: pav This breaks the build of editors/openoffice-2 Specifically, the libxslt configure script has two statements like this: if test "1" == "1" then blah blah endif Specifically note the "==" passed to test(1). POSIX says this should be "=", and that's all our test(1) implements. The bash manpage for the builtin-test command says: string1 == string2 True if the strings are equal. = may be used in place of == for strict POSIX compliance. IOW, it encourages "==". I'm not sure if we want to force the use of bash for certain ports or if we want to just implement bash'isms in our tools as we encounter them (or patch the port?). In this case the patch is not complicated (just replace the two '==' with '=' in libxslt's configure script). -- John Baldwin From cokane at FreeBSD.org Mon Jun 2 20:54:24 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Mon Jun 2 20:54:30 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <200806021445.15974.jhb@freebsd.org> References: <200805151955.m4FJtRMb099378@repoman.freebsd.org> <200806021445.15974.jhb@freebsd.org> Message-ID: <1212439968.3115.53.camel@localhost> On Mon, 2008-06-02 at 14:45 -0400, John Baldwin wrote: > On Thursday 15 May 2008 03:55:27 pm Stefan Farfeleder wrote: > > stefanf 2008-05-15 19:55:27 UTC > > > > FreeBSD src repository > > > > Modified files: > > bin/sh expand.c parser.c parser.h > > Log: > > Expand $LINENO to the current line number. This is required by > SUSv3's "User > > Portability Utilities" option. > > > > Often configure scripts generated by the autotools test if $LINENO works > and > > refuse to use /bin/sh if not. > > > > Package test run by: pav > > This breaks the build of editors/openoffice-2 > > Specifically, the libxslt configure script has two statements like this: > > if test "1" == "1" > then > blah blah > endif > > Specifically note the "==" passed to test(1). POSIX says this should be "=", > and that's all our test(1) implements. The bash manpage for the builtin-test > command says: > > string1 == string2 > True if the strings are equal. = may be used in place of == for > strict POSIX compliance. > > IOW, it encourages "==". I'm not sure if we want to force the use of bash for > certain ports or if we want to just implement bash'isms in our tools as we > encounter them (or patch the port?). In this case the patch is not > complicated (just replace the two '==' with '=' in libxslt's configure > script). > This is annoying... I had to clean this behavior up once recently in someone else's script. POSIX "test" syntax has been "=" and not "==" for a long time. Bash is not C... so I don't understand why the attempt to document "==" as the "proper" operator. My thinking is the offending script should be fixed with a patch that gets forwarded upstream to the libxslt team (including a mention that /bin/sh and /bin/test are not documented to support "==" by POSIX). -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080602/fb537810/attachment.pgp From marcus at marcuscom.com Mon Jun 2 21:07:16 2008 From: marcus at marcuscom.com (Joe Marcus Clarke) Date: Mon Jun 2 21:07:19 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <1212439968.3115.53.camel@localhost> References: <200805151955.m4FJtRMb099378@repoman.freebsd.org> <200806021445.15974.jhb@freebsd.org> <1212439968.3115.53.camel@localhost> Message-ID: <1212440845.18384.49.camel@shumai.marcuscom.com> On Mon, 2008-06-02 at 16:52 -0400, Coleman Kane wrote: > On Mon, 2008-06-02 at 14:45 -0400, John Baldwin wrote: > > On Thursday 15 May 2008 03:55:27 pm Stefan Farfeleder wrote: > > > stefanf 2008-05-15 19:55:27 UTC > > > > > > FreeBSD src repository > > > > > > Modified files: > > > bin/sh expand.c parser.c parser.h > > > Log: > > > Expand $LINENO to the current line number. This is required by > > SUSv3's "User > > > Portability Utilities" option. > > > > > > Often configure scripts generated by the autotools test if $LINENO works > > and > > > refuse to use /bin/sh if not. > > > > > > Package test run by: pav > > > > This breaks the build of editors/openoffice-2 > > > > Specifically, the libxslt configure script has two statements like this: > > > > if test "1" == "1" > > then > > blah blah > > endif > > > > Specifically note the "==" passed to test(1). POSIX says this should be "=", > > and that's all our test(1) implements. The bash manpage for the builtin-test > > command says: > > > > string1 == string2 > > True if the strings are equal. = may be used in place of == for > > strict POSIX compliance. > > > > IOW, it encourages "==". I'm not sure if we want to force the use of bash for > > certain ports or if we want to just implement bash'isms in our tools as we > > encounter them (or patch the port?). In this case the patch is not > > complicated (just replace the two '==' with '=' in libxslt's configure > > script). > > > > This is annoying... I had to clean this behavior up once recently in > someone else's script. POSIX "test" syntax has been "=" and not "==" for > a long time. Bash is not C... so I don't understand why the attempt to > document "==" as the "proper" operator. My thinking is the offending > script should be fixed with a patch that gets forwarded upstream to the > libxslt team (including a mention that /bin/sh and /bin/test are not > documented to support "==" by POSIX). This is one of the most pervasive bashisms around. We (gnome@) typically fix the script to use "=" then forward the information upstream. Solaris is also bit by this, so it's usually not a big deal to get upstream vendors to fix their scripts. Joe > -- PGP Key : http://www.marcuscom.com/pgp.asc -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080602/a41ff94e/attachment.pgp From jkim at FreeBSD.org Mon Jun 2 23:37:49 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Mon Jun 2 23:37:51 2008 Subject: cvs commit: src/usr.sbin/tzsetup tzsetup.c Message-ID: <200806022337.m52NbnYD027684@repoman.freebsd.org> jkim 2008-06-02 23:37:49 UTC FreeBSD src repository Modified files: usr.sbin/tzsetup tzsetup.c Log: SVN rev 179497 on 2008-06-02 23:37:33Z by jkim Fix a crash when Arctic Ocean is selected. MFC after: 1 week Revision Changes Path 1.27 +2 -3 src/usr.sbin/tzsetup/tzsetup.c From cokane at FreeBSD.org Tue Jun 3 00:55:52 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Tue Jun 3 00:55:56 2008 Subject: cvs commit: src/sys/dev/if_ndis if_ndis.c if_ndisvar.h Message-ID: <200806030055.m530tqVF044140@repoman.freebsd.org> cokane 2008-06-03 00:55:52 UTC FreeBSD src repository Modified files: sys/dev/if_ndis if_ndis.c if_ndisvar.h Log: SVN rev 179498 on 2008-06-03 00:55:48Z by cokane Update if_ndis to remove the legacy if_watchdog interface and bring it more up to date. The watchdog timer, and its associated code, is all collapsed into the ndis_tick function that was implemented for the NDIS-subsystem watchdog. This implementation is similar to what numerous other drivers use to implement the watchdog. Reviewed by: thompsa, jhb MFC after: 2 weeks Revision Changes Path 1.141 +31 -44 src/sys/dev/if_ndis/if_ndis.c 1.32 +2 -0 src/sys/dev/if_ndis/if_ndisvar.h From imp at bsdimp.com Tue Jun 3 02:40:49 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Tue Jun 3 02:41:12 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <1212440845.18384.49.camel@shumai.marcuscom.com> References: <200806021445.15974.jhb@freebsd.org> <1212439968.3115.53.camel@localhost> <1212440845.18384.49.camel@shumai.marcuscom.com> Message-ID: <20080602.204014.-222576415.imp@bsdimp.com> In message: <1212440845.18384.49.camel@shumai.marcuscom.com> Joe Marcus Clarke writes: : On Mon, 2008-06-02 at 16:52 -0400, Coleman Kane wrote: : > On Mon, 2008-06-02 at 14:45 -0400, John Baldwin wrote: : > > On Thursday 15 May 2008 03:55:27 pm Stefan Farfeleder wrote: : > > > stefanf 2008-05-15 19:55:27 UTC : > > > : > > > FreeBSD src repository : > > > : > > > Modified files: : > > > bin/sh expand.c parser.c parser.h : > > > Log: : > > > Expand $LINENO to the current line number. This is required by : > > SUSv3's "User : > > > Portability Utilities" option. : > > > : > > > Often configure scripts generated by the autotools test if $LINENO works : > > and : > > > refuse to use /bin/sh if not. : > > > : > > > Package test run by: pav : > > : > > This breaks the build of editors/openoffice-2 : > > : > > Specifically, the libxslt configure script has two statements like this: : > > : > > if test "1" == "1" : > > then : > > blah blah : > > endif : > > : > > Specifically note the "==" passed to test(1). POSIX says this should be "=", : > > and that's all our test(1) implements. The bash manpage for the builtin-test : > > command says: : > > : > > string1 == string2 : > > True if the strings are equal. = may be used in place of == for : > > strict POSIX compliance. : > > : > > IOW, it encourages "==". I'm not sure if we want to force the use of bash for : > > certain ports or if we want to just implement bash'isms in our tools as we : > > encounter them (or patch the port?). In this case the patch is not : > > complicated (just replace the two '==' with '=' in libxslt's configure : > > script). : > > : > : > This is annoying... I had to clean this behavior up once recently in : > someone else's script. POSIX "test" syntax has been "=" and not "==" for : > a long time. Bash is not C... so I don't understand why the attempt to : > document "==" as the "proper" operator. My thinking is the offending : > script should be fixed with a patch that gets forwarded upstream to the : > libxslt team (including a mention that /bin/sh and /bin/test are not : > documented to support "==" by POSIX). : : This is one of the most pervasive bashisms around. We (gnome@) : typically fix the script to use "=" then forward the information : upstream. Solaris is also bit by this, so it's usually not a big deal : to get upstream vendors to fix their scripts. Maybe a 'grep ==' on all configure scripts should be SOP, eh? Warner From marcus at marcuscom.com Tue Jun 3 02:46:46 2008 From: marcus at marcuscom.com (Joe Marcus Clarke) Date: Tue Jun 3 02:46:48 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <20080602.204014.-222576415.imp@bsdimp.com> References: <200806021445.15974.jhb@freebsd.org> <1212439968.3115.53.camel@localhost> <1212440845.18384.49.camel@shumai.marcuscom.com> <20080602.204014.-222576415.imp@bsdimp.com> Message-ID: <1212461214.18384.83.camel@shumai.marcuscom.com> On Mon, 2008-06-02 at 20:40 -0600, M. Warner Losh wrote: > In message: <1212440845.18384.49.camel@shumai.marcuscom.com> > Joe Marcus Clarke writes: > : On Mon, 2008-06-02 at 16:52 -0400, Coleman Kane wrote: > : > On Mon, 2008-06-02 at 14:45 -0400, John Baldwin wrote: > : > > On Thursday 15 May 2008 03:55:27 pm Stefan Farfeleder wrote: > : > > > stefanf 2008-05-15 19:55:27 UTC > : > > > > : > > > FreeBSD src repository > : > > > > : > > > Modified files: > : > > > bin/sh expand.c parser.c parser.h > : > > > Log: > : > > > Expand $LINENO to the current line number. This is required by > : > > SUSv3's "User > : > > > Portability Utilities" option. > : > > > > : > > > Often configure scripts generated by the autotools test if $LINENO works > : > > and > : > > > refuse to use /bin/sh if not. > : > > > > : > > > Package test run by: pav > : > > > : > > This breaks the build of editors/openoffice-2 > : > > > : > > Specifically, the libxslt configure script has two statements like this: > : > > > : > > if test "1" == "1" > : > > then > : > > blah blah > : > > endif > : > > > : > > Specifically note the "==" passed to test(1). POSIX says this should be "=", > : > > and that's all our test(1) implements. The bash manpage for the builtin-test > : > > command says: > : > > > : > > string1 == string2 > : > > True if the strings are equal. = may be used in place of == for > : > > strict POSIX compliance. > : > > > : > > IOW, it encourages "==". I'm not sure if we want to force the use of bash for > : > > certain ports or if we want to just implement bash'isms in our tools as we > : > > encounter them (or patch the port?). In this case the patch is not > : > > complicated (just replace the two '==' with '=' in libxslt's configure > : > > script). > : > > > : > > : > This is annoying... I had to clean this behavior up once recently in > : > someone else's script. POSIX "test" syntax has been "=" and not "==" for > : > a long time. Bash is not C... so I don't understand why the attempt to > : > document "==" as the "proper" operator. My thinking is the offending > : > script should be fixed with a patch that gets forwarded upstream to the > : > libxslt team (including a mention that /bin/sh and /bin/test are not > : > documented to support "==" by POSIX). > : > : This is one of the most pervasive bashisms around. We (gnome@) > : typically fix the script to use "=" then forward the information > : upstream. Solaris is also bit by this, so it's usually not a big deal > : to get upstream vendors to fix their scripts. > > Maybe a 'grep ==' on all configure scripts should be SOP, eh? This will yield false positives as many (all?) contain embedded C code. We have been using one regexp that seems to work nicely: " == ". For example: @${FIND} ${WRKSRC} -name Makefile.in | ${XARGS} ${REINPLACE_CMD} -e \ 's|" == "|" = "|g' You don't typically find '"' on either side of a == in C. Joe -- PGP Key : http://www.marcuscom.com/pgp.asc -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080603/436edebb/attachment.pgp From kan at FreeBSD.org Tue Jun 3 03:14:13 2008 From: kan at FreeBSD.org (Alexander Kabaev) Date: Tue Jun 3 03:14:16 2008 Subject: cvs commit: src/sys/dev/usb ubsa.c usbdevs Message-ID: <200806030314.m533EDXl058950@repoman.freebsd.org> kan 2008-06-03 03:14:13 UTC FreeBSD src repository Modified files: sys/dev/usb ubsa.c usbdevs Log: SVN rev 179499 on 2008-06-03 03:13:57Z by kan Add vendor/device IDs for Novatel U740 repackaged by Dell. Submitted by: David Gilbert PR: 122803 Revision Changes Path 1.36 +2 -0 src/sys/dev/usb/ubsa.c 1.356 +1 -0 src/sys/dev/usb/usbdevs From imp at bsdimp.com Tue Jun 3 03:31:33 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Tue Jun 3 03:31:42 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <1212461214.18384.83.camel@shumai.marcuscom.com> References: <1212440845.18384.49.camel@shumai.marcuscom.com> <20080602.204014.-222576415.imp@bsdimp.com> <1212461214.18384.83.camel@shumai.marcuscom.com> Message-ID: <20080602.213239.-861051005.imp@bsdimp.com> In message: <1212461214.18384.83.camel@shumai.marcuscom.com> Joe Marcus Clarke writes: : On Mon, 2008-06-02 at 20:40 -0600, M. Warner Losh wrote: : > In message: <1212440845.18384.49.camel@shumai.marcuscom.com> : > Joe Marcus Clarke writes: : > : On Mon, 2008-06-02 at 16:52 -0400, Coleman Kane wrote: : > : > On Mon, 2008-06-02 at 14:45 -0400, John Baldwin wrote: : > : > > On Thursday 15 May 2008 03:55:27 pm Stefan Farfeleder wrote: : > : > > > stefanf 2008-05-15 19:55:27 UTC : > : > > > : > : > > > FreeBSD src repository : > : > > > : > : > > > Modified files: : > : > > > bin/sh expand.c parser.c parser.h : > : > > > Log: : > : > > > Expand $LINENO to the current line number. This is required by : > : > > SUSv3's "User : > : > > > Portability Utilities" option. : > : > > > : > : > > > Often configure scripts generated by the autotools test if $LINENO works : > : > > and : > : > > > refuse to use /bin/sh if not. : > : > > > : > : > > > Package test run by: pav : > : > > : > : > > This breaks the build of editors/openoffice-2 : > : > > : > : > > Specifically, the libxslt configure script has two statements like this: : > : > > : > : > > if test "1" == "1" : > : > > then : > : > > blah blah : > : > > endif : > : > > : > : > > Specifically note the "==" passed to test(1). POSIX says this should be "=", : > : > > and that's all our test(1) implements. The bash manpage for the builtin-test : > : > > command says: : > : > > : > : > > string1 == string2 : > : > > True if the strings are equal. = may be used in place of == for : > : > > strict POSIX compliance. : > : > > : > : > > IOW, it encourages "==". I'm not sure if we want to force the use of bash for : > : > > certain ports or if we want to just implement bash'isms in our tools as we : > : > > encounter them (or patch the port?). In this case the patch is not : > : > > complicated (just replace the two '==' with '=' in libxslt's configure : > : > > script). : > : > > : > : > : > : > This is annoying... I had to clean this behavior up once recently in : > : > someone else's script. POSIX "test" syntax has been "=" and not "==" for : > : > a long time. Bash is not C... so I don't understand why the attempt to : > : > document "==" as the "proper" operator. My thinking is the offending : > : > script should be fixed with a patch that gets forwarded upstream to the : > : > libxslt team (including a mention that /bin/sh and /bin/test are not : > : > documented to support "==" by POSIX). : > : : > : This is one of the most pervasive bashisms around. We (gnome@) : > : typically fix the script to use "=" then forward the information : > : upstream. Solaris is also bit by this, so it's usually not a big deal : > : to get upstream vendors to fix their scripts. : > : > Maybe a 'grep ==' on all configure scripts should be SOP, eh? : : This will yield false positives as many (all?) contain embedded C code. : We have been using one regexp that seems to work nicely: " == ". For : example: : : @${FIND} ${WRKSRC} -name Makefile.in | ${XARGS} ${REINPLACE_CMD} -e \ : 's|" == "|" = "|g' : : You don't typically find '"' on either side of a == in C. heh... Another option would be to add support for it to our test, maybe with a warning... :-) Warner From kan at FreeBSD.org Tue Jun 3 03:40:58 2008 From: kan at FreeBSD.org (Alexander Kabaev) Date: Tue Jun 3 03:40:59 2008 Subject: cvs commit: src/share/man/man4 ubsa.4 src/sys/dev/usb ubsa.c usbdevs Message-ID: <200806030340.m533evod061082@repoman.freebsd.org> kan 2008-06-03 03:40:57 UTC FreeBSD src repository Modified files: share/man/man4 ubsa.4 sys/dev/usb ubsa.c usbdevs Log: SVN rev 179500 on 2008-06-03 03:31:53Z by kan Add device ID for AnyDATA ADU-500A EV-DO modem. Submitted by: Oleksandr Tatmanyants PR: 118479 Revision Changes Path 1.12 +2 -0 src/share/man/man4/ubsa.4 1.37 +2 -0 src/sys/dev/usb/ubsa.c 1.357 +1 -0 src/sys/dev/usb/usbdevs From dougb at FreeBSD.org Tue Jun 3 05:39:07 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Tue Jun 3 05:39:11 2008 Subject: cvs commit: src/contrib/bind9 CHANGES COPYRIGHT FAQ FAQ.xml Makefile.in README configure.in version src/contrib/bind9/bin/check check-tool.c check-tool.h named-checkconf.8 named-checkconf.c named-checkconf.docbook ... Message-ID: <200806030539.m535d6ZY084605@repoman.freebsd.org> dougb 2008-06-03 05:39:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) contrib/bind9 CHANGES COPYRIGHT FAQ FAQ.xml Makefile.in README configure.in version contrib/bind9/bin/check check-tool.c check-tool.h named-checkconf.8 named-checkconf.c named-checkconf.docbook named-checkconf.html named-checkzone.8 named-checkzone.docbook named-checkzone.html contrib/bind9/bin/dig Makefile.in dig.1 dig.c dig.docbook dig.html dighost.c host.1 host.c host.docbook host.html nslookup.1 nslookup.c nslookup.docbook nslookup.html contrib/bind9/bin/dig/include/dig dig.h contrib/bind9/bin/dnssec Makefile.in dnssec-keygen.8 dnssec-keygen.c dnssec-keygen.docbook dnssec-keygen.html dnssec-signzone.8 dnssec-signzone.c dnssec-signzone.docbook dnssec-signzone.html contrib/bind9/bin/named Makefile.in aclconf.c client.c config.c control.c controlconf.c interfacemgr.c log.c logconf.c lwaddr.c lwdclient.c lwdgabn.c lwdgnba.c lwdgrbn.c lwdnoop.c lwresd.8 lwresd.docbook lwresd.html named.8 named.conf.5 named.conf.docbook named.conf.html named.docbook named.html query.c server.c sortlist.c tsigconf.c update.c contrib/bind9/bin/named/include/named builtin.h config.h interfacemgr.h log.h main.h query.h zoneconf.h contrib/bind9/bin/named/unix Makefile.in os.c contrib/bind9/bin/named/unix/include/named os.h contrib/bind9/bin/nsupdate Makefile.in nsupdate.8 nsupdate.c nsupdate.docbook nsupdate.html contrib/bind9/bin/rndc Makefile.in rndc-confgen.8 rndc-confgen.docbook rndc-confgen.html rndc.8 rndc.conf.5 rndc.conf.docbook rndc.conf.html rndc.docbook rndc.html contrib/bind9/bin/rndc/unix Makefile.in contrib/bind9/doc/arm Bv9ARM-book.xml Bv9ARM.ch01.html Bv9ARM.ch02.html Bv9ARM.ch03.html Bv9ARM.ch04.html Bv9ARM.ch05.html Bv9ARM.ch06.html Bv9ARM.ch07.html Bv9ARM.ch08.html Bv9ARM.ch09.html Bv9ARM.html Bv9ARM.pdf Makefile.in contrib/bind9/doc/misc Makefile.in dnssec format-options.pl migration options contrib/bind9/doc/rfc index contrib/bind9/lib Makefile.in contrib/bind9/lib/bind api configure.in port_after.h.in port_before.h.in contrib/bind9/lib/bind/dst dst_api.c hmac_link.c contrib/bind9/lib/bind/include Makefile.in contrib/bind9/lib/bind/include/isc eventlib.h contrib/bind9/lib/bind/inet inet_network.c contrib/bind9/lib/bind/irs dns_ho.c gai_strerror.c irp_ng.c irs_data.c contrib/bind9/lib/bind/isc ctl_clnt.c ctl_srvr.c contrib/bind9/lib/bind/make rules.in contrib/bind9/lib/bind/nameser ns_parse.c contrib/bind9/lib/bind/resolv res_data.c res_init.c res_send.c contrib/bind9/lib/bind9 Makefile.in api check.c getaddresses.c version.c contrib/bind9/lib/bind9/include Makefile.in contrib/bind9/lib/bind9/include/bind9 Makefile.in check.h getaddresses.h version.h contrib/bind9/lib/dns acl.c adb.c api dbtable.c dispatch.c dnssec.c dst_parse.c gen-unix.h journal.c keytable.c lib.c lookup.c master.c message.c name.c openssl_link.c openssldh_link.c openssldsa_link.c order.c rbt.c rbtdb.c request.c resolver.c rootns.c sdb.c tkey.c tsig.c ttl.c validator.c version.c view.c xfrin.c zone.c zt.c contrib/bind9/lib/dns/include/dns acl.h cache.h callbacks.h compress.h db.h diff.h dispatch.h dnssec.h events.h journal.h lib.h master.h masterdump.h ncache.h opcode.h order.h rbt.h rdataslab.h request.h sdb.h time.h tsig.h validator.h version.h zt.h contrib/bind9/lib/dns/rdata/generic dlv_32769.c ds_43.c gpos_27.c hinfo_13.c isdn_20.c minfo_14.c null_10.c nxt_30.h opt_41.c proforma.c rp_17.c soa_6.c txt_16.c unspec_103.c x25_19.c contrib/bind9/lib/dns/rdata/hs_4 a_1.c contrib/bind9/lib/dns/rdata/in_1 a_1.c aaaa_28.c apl_42.c apl_42.h nsap_22.c wks_11.c contrib/bind9/lib/isc api buffer.c event.c heap.c hmacmd5.c inet_aton.c inet_ntop.c lfsr.c lib.c mem.c mutexblock.c netaddr.c netscope.c quota.c ratelimiter.c region.c result.c symtab.c taskpool.c timer.c timer_p.h version.c contrib/bind9/lib/isc/include/isc buffer.h entropy.h event.h file.h ipv6.h lex.h lib.h list.h log.h mem.h netaddr.h netscope.h parseint.h platform.h.in quota.h ratelimiter.h region.h result.h socket.h string.h timer.h util.h version.h contrib/bind9/lib/isc/nls msgcat.c contrib/bind9/lib/isc/nothreads condition.c mutex.c contrib/bind9/lib/isc/pthreads condition.c mutex.c contrib/bind9/lib/isc/pthreads/include/isc mutex.h contrib/bind9/lib/isc/unix Makefile.in app.c dir.c entropy.c errno2result.c file.c ifiter_getifaddrs.c ifiter_ioctl.c keyboard.c net.c os.c resource.c socket.c stdtime.c strerror.c syslog.c contrib/bind9/lib/isc/unix/include/isc dir.h strerror.h time.h contrib/bind9/lib/isccc api cc.c lib.c sexpr.c symtab.c version.c contrib/bind9/lib/isccc/include/isccc Makefile.in lib.h version.h contrib/bind9/lib/isccfg api log.c namedconf.c version.c contrib/bind9/lib/isccfg/include/isccfg Makefile.in cfg.h log.h namedconf.h version.h contrib/bind9/lib/lwres Makefile.in api context.c gai_strerror.c getaddrinfo.c getipnode.c lwres_gabn.c lwres_gnba.c lwres_grbn.c version.c contrib/bind9/lib/lwres/include/lwres Makefile.in lwres.h platform.h.in version.h contrib/bind9/lib/lwres/man lwres.3 lwres.docbook lwres.html lwres_buffer.3 lwres_buffer.docbook lwres_buffer.html lwres_config.3 lwres_config.docbook lwres_config.html lwres_context.3 lwres_context.docbook lwres_context.html lwres_gabn.3 lwres_gabn.docbook lwres_gabn.html lwres_gai_strerror.3 lwres_gai_strerror.docbook lwres_gai_strerror.html lwres_getaddrinfo.3 lwres_getaddrinfo.docbook lwres_getaddrinfo.html lwres_gethostent.3 lwres_gethostent.docbook lwres_gethostent.html lwres_getipnode.3 lwres_getipnode.docbook lwres_getipnode.html lwres_getnameinfo.3 lwres_getnameinfo.docbook lwres_getnameinfo.html lwres_getrrsetbyname.3 lwres_getrrsetbyname.docbook lwres_getrrsetbyname.html lwres_gnba.3 lwres_gnba.docbook lwres_gnba.html lwres_hstrerror.3 lwres_hstrerror.docbook lwres_hstrerror.html lwres_inetntop.3 lwres_inetntop.docbook lwres_inetntop.html lwres_noop.3 lwres_noop.docbook lwres_noop.html lwres_packet.3 lwres_packet.docbook lwres_packet.html lwres_resutil.3 lwres_resutil.docbook lwres_resutil.html contrib/bind9/lib/lwres/unix/include/lwres net.h contrib/bind9/make includes.in rules.in Added files: (Branch: RELENG_6) contrib/bind9/doc/misc sort-options.pl contrib/bind9/doc/rfc rfc4193.txt rfc4255.txt rfc4343.txt rfc4367.txt rfc4398.txt rfc4408.txt rfc4431.txt rfc4470.txt rfc4634.txt rfc4641.txt contrib/bind9/lib/bind/include/isc platform.h.in Log: SVN rev 179502 on 2008-06-03 05:38:10Z by dougb Update to version 9.3.5. It contains the latest bug fixes, updates to root server addresses, and a fix for the vulnerability mentioned here: http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2008-0122 Users of BIND 9.3.x are strongly encouraged to upgrade to this version. Also, the 9.3.x branch is now in maintenance-only mode. Users are encouraged to investigate BIND 9.4.x or perhaps 9.5.x. http://www.isc.org/index.pl?/sw/bind/versions_and_support.php This udpate is being done by updating the files directly in this branch rather than an import + MFC because BIND in HEAD is 9.4.x. Revision Changes Path 1.1.1.3.2.7 +410 -6 src/contrib/bind9/CHANGES 1.1.1.2.2.2 +3 -3 src/contrib/bind9/COPYRIGHT 1.1.1.2.2.4 +499 -376 src/contrib/bind9/FAQ 1.1.1.1.2.4 +704 -538 src/contrib/bind9/FAQ.xml 1.1.1.1.4.2 +4 -4 src/contrib/bind9/Makefile.in 1.1.1.2.2.5 +0 -12 src/contrib/bind9/README 1.1.1.2.2.1 +5 -5 src/contrib/bind9/bin/check/check-tool.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/check/check-tool.h 1.1.1.1.4.3 +26 -12 src/contrib/bind9/bin/check/named-checkconf.8 1.1.1.2.2.2 +4 -4 src/contrib/bind9/bin/check/named-checkconf.c 1.1.1.1.4.2 +13 -8 src/contrib/bind9/bin/check/named-checkconf.docbook 1.1.1.1.4.3 +13 -12 src/contrib/bind9/bin/check/named-checkconf.html 1.1.1.1.4.3 +49 -19 src/contrib/bind9/bin/check/named-checkzone.8 1.1.1.1.4.3 +12 -7 src/contrib/bind9/bin/check/named-checkzone.docbook 1.1.1.1.4.3 +12 -11 src/contrib/bind9/bin/check/named-checkzone.html 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/dig/Makefile.in 1.1.1.1.4.3 +150 -59 src/contrib/bind9/bin/dig/dig.1 1.1.1.2.2.3 +52 -65 src/contrib/bind9/bin/dig/dig.c 1.1.1.1.4.2 +25 -16 src/contrib/bind9/bin/dig/dig.docbook 1.1.1.1.4.3 +29 -23 src/contrib/bind9/bin/dig/dig.html 1.1.1.2.2.3 +89 -61 src/contrib/bind9/bin/dig/dighost.c 1.1.1.1.4.3 +10 -7 src/contrib/bind9/bin/dig/host.1 1.1.1.2.2.3 +7 -5 src/contrib/bind9/bin/dig/host.c 1.1.1.1.4.2 +10 -8 src/contrib/bind9/bin/dig/host.docbook 1.1.1.1.4.3 +10 -10 src/contrib/bind9/bin/dig/host.html 1.1.1.2.2.3 +8 -3 src/contrib/bind9/bin/dig/include/dig/dig.h 1.1.1.1.4.3 +99 -42 src/contrib/bind9/bin/dig/nslookup.1 1.1.1.2.2.3 +6 -5 src/contrib/bind9/bin/dig/nslookup.c 1.1.1.1.4.3 +11 -12 src/contrib/bind9/bin/dig/nslookup.docbook 1.1.1.1.4.3 +15 -17 src/contrib/bind9/bin/dig/nslookup.html 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/dnssec/Makefile.in 1.1.1.1.4.3 +57 -28 src/contrib/bind9/bin/dnssec/dnssec-keygen.8 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/dnssec/dnssec-keygen.c 1.1.1.1.4.2 +11 -10 src/contrib/bind9/bin/dnssec/dnssec-keygen.docbook 1.1.1.1.4.3 +15 -15 src/contrib/bind9/bin/dnssec/dnssec-keygen.html 1.1.1.1.4.3 +94 -36 src/contrib/bind9/bin/dnssec/dnssec-signzone.8 1.1.1.2.2.3 +16 -53 src/contrib/bind9/bin/dnssec/dnssec-signzone.c 1.1.1.1.4.2 +33 -25 src/contrib/bind9/bin/dnssec/dnssec-signzone.docbook 1.1.1.1.4.3 +41 -34 src/contrib/bind9/bin/dnssec/dnssec-signzone.html 1.1.1.2.2.1 +4 -4 src/contrib/bind9/bin/named/Makefile.in 1.1.1.1.4.3 +4 -4 src/contrib/bind9/bin/named/aclconf.c 1.1.1.2.2.4 +18 -9 src/contrib/bind9/bin/named/client.c 1.1.1.2.2.2 +5 -5 src/contrib/bind9/bin/named/config.c 1.1.1.3.2.2 +4 -4 src/contrib/bind9/bin/named/control.c 1.1.1.1.4.2 +41 -50 src/contrib/bind9/bin/named/controlconf.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/include/named/builtin.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/include/named/config.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/include/named/interfacemgr.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/include/named/log.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/include/named/main.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/include/named/query.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/include/named/zoneconf.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/interfacemgr.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/log.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/logconf.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/lwaddr.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/lwdclient.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/lwdgabn.c 1.1.1.1.4.1 +4 -6 src/contrib/bind9/bin/named/lwdgnba.c 1.1.1.1.4.2 +1 -3 src/contrib/bind9/bin/named/lwdgrbn.c 1.1.1.1.4.1 +12 -13 src/contrib/bind9/bin/named/lwdnoop.c 1.1.1.1.4.3 +99 -21 src/contrib/bind9/bin/named/lwresd.8 1.1.1.1.4.2 +94 -28 src/contrib/bind9/bin/named/lwresd.docbook 1.1.1.1.4.3 +64 -29 src/contrib/bind9/bin/named/lwresd.html 1.1.1.1.4.3 +75 -27 src/contrib/bind9/bin/named/named.8 1.1.1.2.2.3 +18 -16 src/contrib/bind9/bin/named/named.conf.5 1.1.1.2.2.3 +21 -19 src/contrib/bind9/bin/named/named.conf.docbook 1.1.1.2.2.3 +23 -22 src/contrib/bind9/bin/named/named.conf.html 1.1.1.1.4.3 +34 -10 src/contrib/bind9/bin/named/named.docbook 1.1.1.1.4.3 +28 -17 src/contrib/bind9/bin/named/named.html 1.1.1.1.4.5 +73 -37 src/contrib/bind9/bin/named/query.c 1.1.1.2.2.3 +86 -57 src/contrib/bind9/bin/named/server.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/sortlist.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/bin/named/tsigconf.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/named/unix/Makefile.in 1.1.1.2.2.1 +4 -4 src/contrib/bind9/bin/named/unix/include/named/os.h 1.1.1.2.2.3 +14 -10 src/contrib/bind9/bin/named/unix/os.c 1.1.1.2.2.3 +20 -13 src/contrib/bind9/bin/named/update.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/nsupdate/Makefile.in 1.1.1.1.4.3 +86 -47 src/contrib/bind9/bin/nsupdate/nsupdate.8 1.1.1.2.2.3 +66 -31 src/contrib/bind9/bin/nsupdate/nsupdate.c 1.1.1.1.4.2 +16 -44 src/contrib/bind9/bin/nsupdate/nsupdate.docbook 1.1.1.1.4.3 +63 -34 src/contrib/bind9/bin/nsupdate/nsupdate.html 1.1.1.1.4.1 +6 -4 src/contrib/bind9/bin/rndc/Makefile.in 1.1.1.1.4.3 +37 -14 src/contrib/bind9/bin/rndc/rndc-confgen.8 1.1.1.1.4.2 +6 -5 src/contrib/bind9/bin/rndc/rndc-confgen.docbook 1.1.1.1.4.3 +9 -9 src/contrib/bind9/bin/rndc/rndc-confgen.html 1.1.1.1.4.3 +35 -18 src/contrib/bind9/bin/rndc/rndc.8 1.1.1.1.4.3 +9 -6 src/contrib/bind9/bin/rndc/rndc.conf.5 1.1.1.1.4.2 +7 -6 src/contrib/bind9/bin/rndc/rndc.conf.docbook 1.1.1.1.4.3 +10 -10 src/contrib/bind9/bin/rndc/rndc.conf.html 1.1.1.1.4.2 +23 -19 src/contrib/bind9/bin/rndc/rndc.docbook 1.1.1.1.4.3 +23 -23 src/contrib/bind9/bin/rndc/rndc.html 1.1.1.1.4.1 +4 -4 src/contrib/bind9/bin/rndc/unix/Makefile.in 1.1.1.2.2.4 +50 -26 src/contrib/bind9/configure.in 1.1.1.2.2.3 +116 -61 src/contrib/bind9/doc/arm/Bv9ARM-book.xml 1.1.1.1.4.3 +31 -31 src/contrib/bind9/doc/arm/Bv9ARM.ch01.html 1.1.1.1.4.3 +14 -14 src/contrib/bind9/doc/arm/Bv9ARM.ch02.html 1.1.1.2.2.3 +23 -17 src/contrib/bind9/doc/arm/Bv9ARM.ch03.html 1.1.1.2.2.3 +44 -40 src/contrib/bind9/doc/arm/Bv9ARM.ch04.html 1.1.1.2.2.3 +5 -5 src/contrib/bind9/doc/arm/Bv9ARM.ch05.html 1.1.1.2.2.3 +118 -99 src/contrib/bind9/doc/arm/Bv9ARM.ch06.html 1.1.1.2.2.3 +10 -10 src/contrib/bind9/doc/arm/Bv9ARM.ch07.html 1.1.1.2.2.3 +23 -21 src/contrib/bind9/doc/arm/Bv9ARM.ch08.html 1.1.1.2.2.3 +70 -53 src/contrib/bind9/doc/arm/Bv9ARM.ch09.html 1.1.1.2.2.3 +78 -77 src/contrib/bind9/doc/arm/Bv9ARM.html 1.1.1.1.2.3 +4654 -4641 src/contrib/bind9/doc/arm/Bv9ARM.pdf 1.1.1.1.4.2 +10 -10 src/contrib/bind9/doc/arm/Makefile.in 1.1.1.1.4.1 +20 -8 src/contrib/bind9/doc/misc/Makefile.in 1.1.1.1.4.1 +3 -3 src/contrib/bind9/doc/misc/dnssec 1.1.1.1.4.1 +21 -8 src/contrib/bind9/doc/misc/format-options.pl 1.1.1.2.2.1 +7 -5 src/contrib/bind9/doc/misc/migration 1.1.1.2.2.2 +262 -259 src/contrib/bind9/doc/misc/options 1.1.2.1 +50 -0 src/contrib/bind9/doc/misc/sort-options.pl (new) 1.1.1.1.4.2 +11 -0 src/contrib/bind9/doc/rfc/index 1.1.1.1.6.1 +899 -0 src/contrib/bind9/doc/rfc/rfc4193.txt (new) 1.1.1.1.6.1 +507 -0 src/contrib/bind9/doc/rfc/rfc4255.txt (new) 1.1.1.1.6.1 +563 -0 src/contrib/bind9/doc/rfc/rfc4343.txt (new) 1.1.1.1.6.1 +955 -0 src/contrib/bind9/doc/rfc/rfc4367.txt (new) 1.1.1.1.4.1 +955 -0 src/contrib/bind9/doc/rfc/rfc4398.txt (new) 1.1.1.1.4.1 +2691 -0 src/contrib/bind9/doc/rfc/rfc4408.txt (new) 1.1.1.1.6.1 +227 -0 src/contrib/bind9/doc/rfc/rfc4431.txt (new) 1.1.1.1.4.1 +451 -0 src/contrib/bind9/doc/rfc/rfc4470.txt (new) 1.1.1.1.4.1 +6051 -0 src/contrib/bind9/doc/rfc/rfc4634.txt (new) 1.1.1.1.4.1 +1963 -0 src/contrib/bind9/doc/rfc/rfc4641.txt (new) 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/Makefile.in 1.1.1.2.2.3 +1 -1 src/contrib/bind9/lib/bind/api 1.1.1.2.2.3 +31 -19 src/contrib/bind9/lib/bind/configure.in 1.1.1.2.2.3 +3 -2 src/contrib/bind9/lib/bind/dst/dst_api.c 1.1.1.1.4.3 +16 -10 src/contrib/bind9/lib/bind/dst/hmac_link.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/bind/include/Makefile.in 1.1.1.1.4.2 +3 -1 src/contrib/bind9/lib/bind/include/isc/eventlib.h 1.1.2.1 +36 -0 src/contrib/bind9/lib/bind/include/isc/platform.h.in (new) 1.1.1.1.4.1 +2 -2 src/contrib/bind9/lib/bind/inet/inet_network.c 1.1.1.1.4.3 +3 -3 src/contrib/bind9/lib/bind/irs/dns_ho.c 1.1.1.2.2.2 +3 -1 src/contrib/bind9/lib/bind/irs/gai_strerror.c 1.1.1.1.4.1 +3 -3 src/contrib/bind9/lib/bind/irs/irp_ng.c 1.1.1.2.2.2 +4 -2 src/contrib/bind9/lib/bind/irs/irs_data.c 1.1.1.1.4.1 +14 -1 src/contrib/bind9/lib/bind/isc/ctl_clnt.c 1.1.1.1.4.1 +4 -2 src/contrib/bind9/lib/bind/isc/ctl_srvr.c 1.1.1.2.2.1 +4 -4 src/contrib/bind9/lib/bind/make/rules.in 1.1.1.1.4.2 +2 -2 src/contrib/bind9/lib/bind/nameser/ns_parse.c 1.1.1.2.2.3 +7 -0 src/contrib/bind9/lib/bind/port_after.h.in 1.1.1.1.4.3 +10 -0 src/contrib/bind9/lib/bind/port_before.h.in 1.1.1.1.4.1 +6 -2 src/contrib/bind9/lib/bind/resolv/res_data.c 1.1.1.1.4.3 +24 -17 src/contrib/bind9/lib/bind/resolv/res_init.c 1.1.1.1.4.3 +21 -7 src/contrib/bind9/lib/bind/resolv/res_send.c 1.1.1.2.2.1 +4 -4 src/contrib/bind9/lib/bind9/Makefile.in 1.1.1.2.2.3 +1 -1 src/contrib/bind9/lib/bind9/api 1.1.1.2.2.3 +10 -9 src/contrib/bind9/lib/bind9/check.c 1.1.1.2.2.2 +4 -4 src/contrib/bind9/lib/bind9/getaddresses.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/bind9/include/Makefile.in 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/bind9/include/bind9/Makefile.in 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/bind9/include/bind9/check.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/bind9/include/bind9/getaddresses.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/bind9/include/bind9/version.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/bind9/version.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/acl.c 1.1.1.2.2.3 +7 -5 src/contrib/bind9/lib/dns/adb.c 1.1.1.2.2.4 +3 -3 src/contrib/bind9/lib/dns/api 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/dbtable.c 1.1.1.1.4.3 +10 -5 src/contrib/bind9/lib/dns/dispatch.c 1.1.1.1.4.2 +24 -16 src/contrib/bind9/lib/dns/dnssec.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/dst_parse.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/gen-unix.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/include/dns/acl.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/include/dns/cache.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/callbacks.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/include/dns/compress.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/db.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/diff.h 1.1.1.1.4.2 +3 -3 src/contrib/bind9/lib/dns/include/dns/dispatch.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/dnssec.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/events.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/journal.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/lib.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/master.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/include/dns/masterdump.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/ncache.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/opcode.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/order.h 1.1.1.2.2.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/rbt.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/rdataslab.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/request.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/sdb.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/time.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/tsig.h 1.1.1.1.4.4 +16 -3 src/contrib/bind9/lib/dns/include/dns/validator.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/version.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/include/dns/zt.h 1.1.1.2.2.2 +108 -64 src/contrib/bind9/lib/dns/journal.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/keytable.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/lib.c 1.1.1.1.4.2 +22 -10 src/contrib/bind9/lib/dns/lookup.c 1.1.1.1.4.1 +26 -26 src/contrib/bind9/lib/dns/master.c 1.1.1.1.4.3 +15 -6 src/contrib/bind9/lib/dns/message.c 1.1.1.1.4.3 +9 -11 src/contrib/bind9/lib/dns/name.c 1.1.1.1.4.2 +7 -5 src/contrib/bind9/lib/dns/openssl_link.c 1.1.1.1.4.2 +22 -77 src/contrib/bind9/lib/dns/openssldh_link.c 1.1.1.1.4.2 +22 -81 src/contrib/bind9/lib/dns/openssldsa_link.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/order.c 1.1.1.2.2.2 +13 -12 src/contrib/bind9/lib/dns/rbt.c 1.1.1.1.4.3 +50 -28 src/contrib/bind9/lib/dns/rbtdb.c 1.1.1.1.2.2 +35 -5 src/contrib/bind9/lib/dns/rdata/generic/dlv_32769.c 1.1.1.1.4.2 +34 -6 src/contrib/bind9/lib/dns/rdata/generic/ds_43.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/gpos_27.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/hinfo_13.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/isdn_20.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/minfo_14.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/null_10.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/nxt_30.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/opt_41.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/proforma.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/rp_17.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/soa_6.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/txt_16.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/unspec_103.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/generic/x25_19.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/hs_4/a_1.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/in_1/a_1.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/in_1/aaaa_28.c 1.1.1.1.4.1 +60 -9 src/contrib/bind9/lib/dns/rdata/in_1/apl_42.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/in_1/apl_42.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/rdata/in_1/nsap_22.c 1.1.1.2.2.1 +4 -4 src/contrib/bind9/lib/dns/rdata/in_1/wks_11.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/dns/request.c 1.1.1.2.2.7 +151 -71 src/contrib/bind9/lib/dns/resolver.c 1.1.1.1.4.1 +15 -11 src/contrib/bind9/lib/dns/rootns.c 1.1.1.1.4.1 +16 -10 src/contrib/bind9/lib/dns/sdb.c 1.1.1.1.4.3 +7 -7 src/contrib/bind9/lib/dns/tkey.c 1.1.1.1.4.3 +12 -8 src/contrib/bind9/lib/dns/tsig.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/ttl.c 1.1.1.2.2.4 +131 -72 src/contrib/bind9/lib/dns/validator.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/version.c 1.1.1.1.4.1 +4 -3 src/contrib/bind9/lib/dns/view.c 1.1.1.2.2.3 +14 -3 src/contrib/bind9/lib/dns/xfrin.c 1.1.1.2.2.3 +172 -14 src/contrib/bind9/lib/dns/zone.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/dns/zt.c 1.1.1.2.2.3 +1 -1 src/contrib/bind9/lib/isc/api 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/buffer.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/event.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/heap.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/hmacmd5.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/buffer.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/entropy.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/event.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/file.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/ipv6.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/lex.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/lib.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/include/isc/list.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/log.h 1.1.1.2.2.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/mem.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/include/isc/netaddr.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/netscope.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/parseint.h 1.1.1.1.4.1 +8 -3 src/contrib/bind9/lib/isc/include/isc/platform.h.in 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/include/isc/quota.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/ratelimiter.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/region.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/result.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/socket.h 1.1.1.1.4.1 +9 -5 src/contrib/bind9/lib/isc/include/isc/string.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/include/isc/timer.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/util.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/include/isc/version.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/inet_aton.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/inet_ntop.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/lfsr.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/lib.c 1.1.1.1.4.2 +24 -9 src/contrib/bind9/lib/isc/mem.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/mutexblock.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/netaddr.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/netscope.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/nls/msgcat.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/nothreads/condition.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/nothreads/mutex.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/pthreads/condition.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/pthreads/include/isc/mutex.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/pthreads/mutex.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/quota.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/ratelimiter.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/region.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/result.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/symtab.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/taskpool.c 1.1.1.1.4.3 +31 -20 src/contrib/bind9/lib/isc/timer.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/timer_p.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/Makefile.in 1.1.1.1.4.1 +10 -10 src/contrib/bind9/lib/isc/unix/app.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/dir.c 1.1.1.1.4.3 +1 -6 src/contrib/bind9/lib/isc/unix/entropy.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/errno2result.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/file.c 1.1.1.1.4.1 +6 -4 src/contrib/bind9/lib/isc/unix/ifiter_getifaddrs.c 1.1.1.2.2.3 +9 -4 src/contrib/bind9/lib/isc/unix/ifiter_ioctl.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/include/isc/dir.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/include/isc/strerror.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/include/isc/time.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/keyboard.c 1.1.1.1.4.2 +19 -10 src/contrib/bind9/lib/isc/unix/net.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/unix/os.c 1.1.1.1.4.1 +23 -54 src/contrib/bind9/lib/isc/unix/resource.c 1.1.1.2.2.3 +81 -18 src/contrib/bind9/lib/isc/unix/socket.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isc/unix/stdtime.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/unix/strerror.c 1.1.1.1.4.1 +5 -5 src/contrib/bind9/lib/isc/unix/syslog.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isc/version.c 1.1.1.1.4.2 +1 -1 src/contrib/bind9/lib/isccc/api 1.1.1.1.4.1 +14 -5 src/contrib/bind9/lib/isccc/cc.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccc/include/isccc/Makefile.in 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccc/include/isccc/lib.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccc/include/isccc/version.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccc/lib.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccc/sexpr.c 1.1.1.1.4.1 +5 -5 src/contrib/bind9/lib/isccc/symtab.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccc/version.c 1.1.1.2.2.2 +1 -1 src/contrib/bind9/lib/isccfg/api 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccfg/include/isccfg/Makefile.in 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/isccfg/include/isccfg/cfg.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccfg/include/isccfg/log.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccfg/include/isccfg/namedconf.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccfg/include/isccfg/version.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccfg/log.c 1.1.1.2.2.3 +28 -28 src/contrib/bind9/lib/isccfg/namedconf.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/isccfg/version.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/lwres/Makefile.in 1.1.1.2.2.3 +3 -3 src/contrib/bind9/lib/lwres/api 1.1.1.2.2.1 +23 -5 src/contrib/bind9/lib/lwres/context.c 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/lwres/gai_strerror.c 1.1.1.1.4.3 +5 -4 src/contrib/bind9/lib/lwres/getaddrinfo.c 1.1.1.1.4.2 +18 -4 src/contrib/bind9/lib/lwres/getipnode.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/lwres/include/lwres/Makefile.in 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/lwres/include/lwres/lwres.h 1.1.1.1.4.2 +4 -4 src/contrib/bind9/lib/lwres/include/lwres/platform.h.in 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/lwres/include/lwres/version.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/lwres/lwres_gabn.c 1.1.1.1.4.1 +4 -5 src/contrib/bind9/lib/lwres/lwres_gnba.c 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/lwres/lwres_grbn.c 1.1.1.1.4.3 +7 -4 src/contrib/bind9/lib/lwres/man/lwres.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres.docbook 1.1.1.1.4.3 +9 -9 src/contrib/bind9/lib/lwres/man/lwres.html 1.1.1.1.4.3 +8 -5 src/contrib/bind9/lib/lwres/man/lwres_buffer.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_buffer.docbook 1.1.1.1.4.3 +17 -115 src/contrib/bind9/lib/lwres/man/lwres_buffer.html 1.1.1.1.4.3 +7 -4 src/contrib/bind9/lib/lwres/man/lwres_config.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_config.docbook 1.1.1.1.4.3 +14 -48 src/contrib/bind9/lib/lwres/man/lwres_config.html 1.1.1.1.4.3 +7 -4 src/contrib/bind9/lib/lwres/man/lwres_context.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_context.docbook 1.1.1.1.4.3 +11 -52 src/contrib/bind9/lib/lwres/man/lwres_context.html 1.1.1.1.4.3 +8 -5 src/contrib/bind9/lib/lwres/man/lwres_gabn.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_gabn.docbook 1.1.1.1.4.3 +7 -37 src/contrib/bind9/lib/lwres/man/lwres_gabn.html 1.1.1.1.4.3 +40 -15 src/contrib/bind9/lib/lwres/man/lwres_gai_strerror.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_gai_strerror.docbook 1.1.1.1.4.3 +13 -8 src/contrib/bind9/lib/lwres/man/lwres_gai_strerror.html 1.1.1.1.4.3 +20 -9 src/contrib/bind9/lib/lwres/man/lwres_getaddrinfo.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_getaddrinfo.docbook 1.1.1.1.4.3 +9 -22 src/contrib/bind9/lib/lwres/man/lwres_getaddrinfo.html 1.1.1.1.4.3 +35 -14 src/contrib/bind9/lib/lwres/man/lwres_gethostent.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_gethostent.docbook 1.1.1.1.4.3 +45 -53 src/contrib/bind9/lib/lwres/man/lwres_gethostent.html 1.1.1.1.4.3 +47 -18 src/contrib/bind9/lib/lwres/man/lwres_getipnode.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_getipnode.docbook 1.1.1.1.4.3 +9 -27 src/contrib/bind9/lib/lwres/man/lwres_getipnode.html 1.1.1.1.4.3 +22 -9 src/contrib/bind9/lib/lwres/man/lwres_getnameinfo.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_getnameinfo.docbook 1.1.1.1.4.3 +8 -13 src/contrib/bind9/lib/lwres/man/lwres_getnameinfo.html 1.1.1.1.4.3 +26 -11 src/contrib/bind9/lib/lwres/man/lwres_getrrsetbyname.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_getrrsetbyname.docbook 1.1.1.1.4.3 +9 -22 src/contrib/bind9/lib/lwres/man/lwres_getrrsetbyname.html 1.1.1.1.4.3 +8 -5 src/contrib/bind9/lib/lwres/man/lwres_gnba.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_gnba.docbook 1.1.1.1.4.3 +11 -42 src/contrib/bind9/lib/lwres/man/lwres_gnba.html 1.1.1.1.4.3 +22 -9 src/contrib/bind9/lib/lwres/man/lwres_hstrerror.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_hstrerror.docbook 1.1.1.1.4.3 +21 -11 src/contrib/bind9/lib/lwres/man/lwres_hstrerror.html 1.1.1.1.4.3 +7 -4 src/contrib/bind9/lib/lwres/man/lwres_inetntop.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_inetntop.docbook 1.1.1.1.4.3 +7 -12 src/contrib/bind9/lib/lwres/man/lwres_inetntop.html 1.1.1.1.4.3 +8 -5 src/contrib/bind9/lib/lwres/man/lwres_noop.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_noop.docbook 1.1.1.1.4.3 +7 -37 src/contrib/bind9/lib/lwres/man/lwres_noop.html 1.1.1.1.4.3 +44 -17 src/contrib/bind9/lib/lwres/man/lwres_packet.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_packet.docbook 1.1.1.1.4.3 +6 -16 src/contrib/bind9/lib/lwres/man/lwres_packet.html 1.1.1.1.4.3 +8 -5 src/contrib/bind9/lib/lwres/man/lwres_resutil.3 1.1.1.1.4.2 +6 -5 src/contrib/bind9/lib/lwres/man/lwres_resutil.docbook 1.1.1.1.4.3 +7 -27 src/contrib/bind9/lib/lwres/man/lwres_resutil.html 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/lwres/unix/include/lwres/net.h 1.1.1.1.4.1 +4 -4 src/contrib/bind9/lib/lwres/version.c 1.1.1.2.2.1 +4 -4 src/contrib/bind9/make/includes.in 1.1.1.1.4.3 +5 -5 src/contrib/bind9/make/rules.in 1.1.1.3.2.7 +4 -4 src/contrib/bind9/version From dougb at FreeBSD.org Tue Jun 3 05:40:55 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Tue Jun 3 05:40:57 2008 Subject: cvs commit: src/lib/bind/lwres Makefile Message-ID: <200806030540.m535esOJ085233@repoman.freebsd.org> dougb 2008-06-03 05:40:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/bind/lwres Makefile Log: SVN rev 179504 on 2008-06-03 05:40:47Z by dougb MFC versions 1.11 and 1.12, bmake glue updates that are also relevant to BIND 9.3.5. Revision Changes Path 1.7.2.2 +3 -0 src/lib/bind/lwres/Makefile From dougb at FreeBSD.org Tue Jun 3 05:43:41 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Tue Jun 3 05:43:44 2008 Subject: cvs commit: src/lib/bind config.h src/lib/bind/bind port_after.h port_before.h src/lib/bind/isc/isc platform.h Message-ID: <200806030543.m535hfjh085481@repoman.freebsd.org> dougb 2008-06-03 05:43:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/bind config.h lib/bind/bind port_after.h port_before.h lib/bind/isc/isc platform.h Log: SVN rev 179505 on 2008-06-03 05:43:27Z by dougb Changes to generated files related to the 9.3.5 update. Revision Changes Path 1.1.4.3 +7 -0 src/lib/bind/bind/port_after.h 1.1.4.3 +10 -0 src/lib/bind/bind/port_before.h 1.2.2.5 +8 -10 src/lib/bind/config.h 1.1.4.3 +8 -3 src/lib/bind/isc/isc/platform.h From imp at FreeBSD.org Tue Jun 3 05:47:42 2008 From: imp at FreeBSD.org (Warner Losh) Date: Tue Jun 3 05:47:44 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c Message-ID: <200806030547.m535lgdo085916@repoman.freebsd.org> imp 2008-06-03 05:47:42 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c Log: SVN rev 179507 on 2008-06-03 05:47:28Z by imp Correct logic error that would prevent cs pccards from working on systems where the CardBus bridge was connected to a APIC. The case where the probe routine is told to not setup the IRQ was mishandled but the error was masked in the case where the IRQ was a valid one for the card. MFC after: 1 week Revision Changes Path 1.46 +8 -8 src/sys/dev/cs/if_cs.c From dougb at FreeBSD.org Tue Jun 3 06:01:44 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Tue Jun 3 06:02:03 2008 Subject: cvs commit: src/etc network.subr Message-ID: <200806030601.m5361h45087195@repoman.freebsd.org> dougb 2008-06-03 06:01:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) etc network.subr Log: SVN rev 179508 on 2008-06-03 06:01:30Z by dougb MFC the change from version 1.189, add a space between a close quote and the closing bracket of a test. Revision Changes Path 1.180.2.3 +1 -1 src/etc/network.subr From dougb at FreeBSD.org Tue Jun 3 06:02:30 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Tue Jun 3 06:02:33 2008 Subject: cvs commit: src/etc network.subr Message-ID: <200806030602.m5362UQj087273@repoman.freebsd.org> dougb 2008-06-03 06:02:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) etc network.subr Log: SVN rev 179509 on 2008-06-03 06:02:27Z by dougb MFC the change from version 1.189, add a space between a close quote and the closing bracket of a test. Revision Changes Path 1.164.2.10 +1 -1 src/etc/network.subr From dfr at FreeBSD.org Tue Jun 3 08:11:06 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Tue Jun 3 08:11:13 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806030811.m538B6T2002383@repoman.freebsd.org> dfr 2008-06-03 08:11:06 UTC FreeBSD src repository Modified files: sys/nlm nlm_prot_impl.c Log: SVN rev 179510 on 2008-06-03 08:10:58Z by dfr Back out the nlm_global_lock part of the last change - I forgot that only exists in my perforce branch :( Pointy hat: dfr Revision Changes Path 1.9 +0 -1 src/sys/nlm/nlm_prot_impl.c From kib at FreeBSD.org Tue Jun 3 09:05:10 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Tue Jun 3 09:05:14 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200806030905.m53958GW017758@repoman.freebsd.org> kib 2008-06-03 09:05:08 UTC FreeBSD src repository Modified files: sbin/mount mount.8 Log: SVN rev 179513 on 2008-06-03 09:05:04Z by kib Add note about a reason to use mount(8) instead of mount_somefs. Reported and proof-readed by: pho Discussed with: rodrigc MFC after: 3 days Revision Changes Path 1.88 +11 -0 src/sbin/mount/mount.8 From ru at FreeBSD.org Tue Jun 3 09:43:40 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Tue Jun 3 09:43:45 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200806030943.m539hdha021119@repoman.freebsd.org> ru 2008-06-03 09:43:39 UTC FreeBSD src repository Modified files: sbin/mount mount.8 Log: SVN rev 179514 on 2008-06-03 09:43:28Z by ru Fix markup. Revision Changes Path 1.89 +6 -4 src/sbin/mount/mount.8 From bu7cher at yandex.ru Tue Jun 3 09:54:33 2008 From: bu7cher at yandex.ru (Andrey V. Elsukov) Date: Tue Jun 3 09:54:37 2008 Subject: cvs commit: src/sys/dev/ata ata-all.c ata-all.h ata-chipset.c ata-disk.c ata-dma.c ata-lowlevel.c ata-pci.c ata-queue.c ata-raid.c atapi-cam.c atapi-cd.c atapi-fd.c atapi-tape.c In-Reply-To: <20080417195338.257a1df2@fabiankeil.de> References: <200804171229.m3HCTaG4089203@repoman.freebsd.org> <20080417195338.257a1df2@fabiankeil.de> Message-ID: <484514A3.5060501@yandex.ru> Fabian Keil wrote: > With this commit, I'm getting a page fault in bus_dmamap_load() > reached through ata_dmaload(), right after ad1 is detected: > http://www.fabiankeil.de/tmp/freebsd/ata-bus-dmamap-load-page-fault-600x383.jpg > http://www.fabiankeil.de/tmp/freebsd/ata-bus-dmamap-load-page-fault-where-600x440.jpg I'm writing AHCI-capable driver which based on ata(4) and I have the same problem. Problem was introduced with this commit. AHCI driver is trying use DMA without ata_dmaalloc call. Problem fixed for me with attached patch. -- WBR, Andrey V. Elsukov -------------- next part -------------- --- ata.orig/ata-pci.c 2008-06-03 13:17:59.000000000 +0400 +++ ata/ata-pci.c 2008-06-03 13:33:36.000000000 +0400 @@ -617,14 +622,16 @@ ata_pcichannel_attach(device_t dev) if (ctlr->dmainit) ctlr->dmainit(dev); - if ((error = ctlr->allocate(dev))) - return error; + if (ch->dma.alloc) + ch->dma.alloc(dev); - if ((error = ata_attach(dev))) + if ((error = ctlr->allocate(dev))) { + if (ch->dma.free) + ch->dma.free(dev); return error; + } - ch->dma.alloc(dev); - return 0; + return ata_attach(dev); } static int From philip at FreeBSD.org Tue Jun 3 10:35:41 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Tue Jun 3 10:35:43 2008 Subject: cvs commit: src/share/syscons/keymaps INDEX.keymaps Makefile colemak.iso15.acc.kbd Message-ID: <200806031035.m53AZfe5026342@repoman.freebsd.org> philip 2008-06-03 10:35:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/syscons/keymaps INDEX.keymaps Makefile Added files: (Branch: RELENG_7) share/syscons/keymaps colemak.iso15.acc.kbd Log: SVN rev 179515 on 2008-06-03 10:35:23Z by philip MFC: Add support for the Colemak keyboard layout from: http://colemak.com/. Submitted by: Geert Hendrickx Revision Changes Path 1.56.2.1 +5 -3 src/share/syscons/keymaps/INDEX.keymaps 1.72.2.1 +1 -0 src/share/syscons/keymaps/Makefile 1.1.2.1 +149 -0 src/share/syscons/keymaps/colemak.iso15.acc.kbd (new) From philip at FreeBSD.org Tue Jun 3 10:55:29 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Tue Jun 3 10:55:33 2008 Subject: cvs commit: src/share/syscons/keymaps INDEX.keymaps Makefile colemak.iso15.acc.kbd Message-ID: <200806031055.m53AtSLe028082@repoman.freebsd.org> philip 2008-06-03 10:55:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/syscons/keymaps INDEX.keymaps Makefile Added files: (Branch: RELENG_6) share/syscons/keymaps colemak.iso15.acc.kbd Log: SVN rev 179516 on 2008-06-03 10:55:22Z by philip MFC: Add support for the Colemak keyboard layout from: http://colemak.com/. Submitted by: Geert Hendrickx Revision Changes Path 1.54.8.1 +5 -3 src/share/syscons/keymaps/INDEX.keymaps 1.70.2.1 +1 -0 src/share/syscons/keymaps/Makefile 1.1.4.1 +149 -0 src/share/syscons/keymaps/colemak.iso15.acc.kbd (new) From rwatson at FreeBSD.org Tue Jun 3 11:06:44 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Tue Jun 3 11:06:51 2008 Subject: cvs commit: src/sys/security/audit audit.c Message-ID: <200806031106.m53B6iOg030088@repoman.freebsd.org> rwatson 2008-06-03 11:06:44 UTC FreeBSD src repository Modified files: sys/security/audit audit.c Log: SVN rev 179517 on 2008-06-03 11:06:34Z by rwatson Add an XXX comment regarding a bug I introduced when modifying the behavior of audit log vnode rotation: on shutdown, we may not properly drain all pending records, which could lead to lost records during system shutdown. Revision Changes Path 1.42 +3 -0 src/sys/security/audit/audit.c From ru at FreeBSD.org Tue Jun 3 11:39:46 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Tue Jun 3 11:39:49 2008 Subject: cvs commit: src/etc/rc.d initrandom Message-ID: <200806031139.m53Bdjad032710@repoman.freebsd.org> ru 2008-06-03 11:39:45 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) etc/rc.d initrandom Log: SVN rev 179518 on 2008-06-03 11:39:39Z by ru MFC: Make it possible to disable sources of entropy harvesting. Revision Changes Path 1.7.2.1 +11 -9 src/etc/rc.d/initrandom From joerg at britannica.bec.de Tue Jun 3 11:58:25 2008 From: joerg at britannica.bec.de (Joerg Sonnenberger) Date: Tue Jun 3 11:58:28 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <20080602.204014.-222576415.imp@bsdimp.com> References: <200806021445.15974.jhb@freebsd.org> <1212439968.3115.53.camel@localhost> <1212440845.18384.49.camel@shumai.marcuscom.com> <20080602.204014.-222576415.imp@bsdimp.com> Message-ID: <20080603115821.GA1707@britannica.bec.de> On Mon, Jun 02, 2008 at 08:40:14PM -0600, M. Warner Losh wrote: > Maybe a 'grep ==' on all configure scripts should be SOP, eh? ...or look at the way the check is implemented in pkgsrc. Joerg From joerg at britannica.bec.de Tue Jun 3 11:59:11 2008 From: joerg at britannica.bec.de (Joerg Sonnenberger) Date: Tue Jun 3 11:59:13 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <200806021445.15974.jhb@freebsd.org> References: <200805151955.m4FJtRMb099378@repoman.freebsd.org> <200806021445.15974.jhb@freebsd.org> Message-ID: <20080603115909.GB1707@britannica.bec.de> On Mon, Jun 02, 2008 at 02:45:15PM -0400, John Baldwin wrote: > Specifically note the "==" passed to test(1). POSIX says this should be "=", > and that's all our test(1) implements. The bash manpage for the builtin-test > command says: > > string1 == string2 > True if the strings are equal. = may be used in place of == for > strict POSIX compliance. What is even more annoying about this is that it is a BASHism, not a GNUism. The test from coreutils doesn't understand == either. Joerg From ed at FreeBSD.org Tue Jun 3 12:38:18 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Tue Jun 3 12:38:24 2008 Subject: cvs commit: src/sys/kern tty_tty.c Message-ID: <200806031238.m53CcIVv039608@repoman.freebsd.org> ed 2008-06-03 12:38:18 UTC FreeBSD src repository Modified files: sys/kern tty_tty.c Log: SVN rev 179519 on 2008-06-03 12:38:00Z by ed Remove unneeded Giant locking of /dev/tty. The Giant lock is acquired in two places in tty_tty.c. In both places, it is unneeded. There is no reason to specify D_NEEDGIANT on this device node. The device node has only been designed to return ENXIO when opened. It doesn't make any sense to lock/unlock Giant, just to return this error. D_TTY is also unneeded. The unimplemented functions don't need to be patched by devfs. We don't need to lock Giant when we want to lookup the proper TTY vnode. s_ttyvp is already protected by proctree_lock (see devfs_vnops.c). Approved by: philip (mentor) Revision Changes Path 1.62 +0 -3 src/sys/kern/tty_tty.c From cokane at FreeBSD.org Tue Jun 3 14:07:34 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Tue Jun 3 14:07:44 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <20080602.213239.-861051005.imp@bsdimp.com> References: <1212440845.18384.49.camel@shumai.marcuscom.com> <20080602.204014.-222576415.imp@bsdimp.com> <1212461214.18384.83.camel@shumai.marcuscom.com> <20080602.213239.-861051005.imp@bsdimp.com> Message-ID: <1212501946.15220.1.camel@localhost> On Mon, 2008-06-02 at 21:32 -0600, M. Warner Losh wrote: > In message: <1212461214.18384.83.camel@shumai.marcuscom.com> > Joe Marcus Clarke writes: > : On Mon, 2008-06-02 at 20:40 -0600, M. Warner Losh wrote: > : > In message: <1212440845.18384.49.camel@shumai.marcuscom.com> > : > Joe Marcus Clarke writes: > : > : On Mon, 2008-06-02 at 16:52 -0400, Coleman Kane wrote: > : > : > On Mon, 2008-06-02 at 14:45 -0400, John Baldwin wrote: > : > : > > On Thursday 15 May 2008 03:55:27 pm Stefan Farfeleder wrote: > : > : > > > stefanf 2008-05-15 19:55:27 UTC > : > : > > > > : > : > > > FreeBSD src repository > : > : > > > > : > : > > > Modified files: > : > : > > > bin/sh expand.c parser.c parser.h > : > : > > > Log: > : > : > > > Expand $LINENO to the current line number. This is required by > : > : > > SUSv3's "User > : > : > > > Portability Utilities" option. > : > : > > > > : > : > > > Often configure scripts generated by the autotools test if $LINENO works > : > : > > and > : > : > > > refuse to use /bin/sh if not. > : > : > > > > : > : > > > Package test run by: pav > : > : > > > : > : > > This breaks the build of editors/openoffice-2 > : > : > > > : > : > > Specifically, the libxslt configure script has two statements like this: > : > : > > > : > : > > if test "1" == "1" > : > : > > then > : > : > > blah blah > : > : > > endif > : > : > > > : > : > > Specifically note the "==" passed to test(1). POSIX says this should be "=", > : > : > > and that's all our test(1) implements. The bash manpage for the builtin-test > : > : > > command says: > : > : > > > : > : > > string1 == string2 > : > : > > True if the strings are equal. = may be used in place of == for > : > : > > strict POSIX compliance. > : > : > > > : > : > > IOW, it encourages "==". I'm not sure if we want to force the use of bash for > : > : > > certain ports or if we want to just implement bash'isms in our tools as we > : > : > > encounter them (or patch the port?). In this case the patch is not > : > : > > complicated (just replace the two '==' with '=' in libxslt's configure > : > : > > script). > : > : > > > : > : > > : > : > This is annoying... I had to clean this behavior up once recently in > : > : > someone else's script. POSIX "test" syntax has been "=" and not "==" for > : > : > a long time. Bash is not C... so I don't understand why the attempt to > : > : > document "==" as the "proper" operator. My thinking is the offending > : > : > script should be fixed with a patch that gets forwarded upstream to the > : > : > libxslt team (including a mention that /bin/sh and /bin/test are not > : > : > documented to support "==" by POSIX). > : > : > : > : This is one of the most pervasive bashisms around. We (gnome@) > : > : typically fix the script to use "=" then forward the information > : > : upstream. Solaris is also bit by this, so it's usually not a big deal > : > : to get upstream vendors to fix their scripts. > : > > : > Maybe a 'grep ==' on all configure scripts should be SOP, eh? > : > : This will yield false positives as many (all?) contain embedded C code. > : We have been using one regexp that seems to work nicely: " == ". For > : example: > : > : @${FIND} ${WRKSRC} -name Makefile.in | ${XARGS} ${REINPLACE_CMD} -e \ > : 's|" == "|" = "|g' > : > : You don't typically find '"' on either side of a == in C. > > heh... > > Another option would be to add support for it to our test, maybe with > a warning... :-) > > Warner > If we chose to do that, then I'd also suggest a compile flag to turn on "strict POSIX compliance". Then, perhaps some ports-auto-build-box somewhere could generate a report of the configure scripts that do such things, and could be used to nag them into fixing this problem. Maybe someone should nag the bash developers... -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080603/99d9dfb3/attachment.pgp From kris at FreeBSD.org Tue Jun 3 14:15:23 2008 From: kris at FreeBSD.org (Kris Kennaway) Date: Tue Jun 3 14:15:30 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <1212501946.15220.1.camel@localhost> References: <1212440845.18384.49.camel@shumai.marcuscom.com> <20080602.204014.-222576415.imp@bsdimp.com> <1212461214.18384.83.camel@shumai.marcuscom.com> <20080602.213239.-861051005.imp@bsdimp.com> <1212501946.15220.1.camel@localhost> Message-ID: <484551F9.9070300@FreeBSD.org> Coleman Kane wrote: > If we chose to do that, then I'd also suggest a compile flag to turn on > "strict POSIX compliance". Then, perhaps some ports-auto-build-box > somewhere could generate a report of the configure scripts that do such > things, and could be used to nag them into fixing this problem. This could certainly be done, but the requirement is the set of developers to process the output. > Maybe someone should nag the bash developers... The bash developers have taken pride in how they have mangled their shell to support C-like syntax. Kris From flz at FreeBSD.org Tue Jun 3 14:48:21 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 3 14:48:31 2008 Subject: cvs commit: src/usr.sbin/pkg_install Makefile Message-ID: <200806031448.m53EmLiW064371@repoman.freebsd.org> flz 2008-06-03 14:48:21 UTC FreeBSD src repository Modified files: usr.sbin/pkg_install Makefile Log: SVN rev 179520 on 2008-06-03 14:48:16Z by flz Exclude .svn directories from generated distfile. Revision Changes Path 1.24 +2 -1 src/usr.sbin/pkg_install/Makefile From jhb at freebsd.org Tue Jun 3 14:52:28 2008 From: jhb at freebsd.org (John Baldwin) Date: Tue Jun 3 14:52:31 2008 Subject: cvs commit: src/sys/dev/if_ndis if_ndis.c if_ndisvar.h In-Reply-To: <200806030055.m530tqVF044140@repoman.freebsd.org> References: <200806030055.m530tqVF044140@repoman.freebsd.org> Message-ID: <200806031004.43883.jhb@freebsd.org> On Monday 02 June 2008 08:55:48 pm Coleman Kane wrote: > cokane 2008-06-03 00:55:52 UTC > > FreeBSD src repository > > Modified files: > sys/dev/if_ndis if_ndis.c if_ndisvar.h > Log: > SVN rev 179498 on 2008-06-03 00:55:48Z by cokane > > Update if_ndis to remove the legacy if_watchdog interface and > bring it more up to date. The watchdog timer, and its > associated code, is all collapsed into the ndis_tick function > that was implemented for the NDIS-subsystem watchdog. This > implementation is similar to what numerous other drivers use > to implement the watchdog. > > Reviewed by: thompsa, jhb > MFC after: 2 weeks You should make the callout routine use callout_init_mtx() instead as it is less race-prone. -- John Baldwin From flz at FreeBSD.org Tue Jun 3 14:59:19 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 3 14:59:22 2008 Subject: cvs commit: src/usr.sbin/pkg_install Makefile Makefile.inc src/usr.sbin/pkg_install/add perform.c src/usr.sbin/pkg_install/delete perform.c src/usr.sbin/pkg_install/lib lib.h match.c Message-ID: <200806031459.m53ExEq3065338@repoman.freebsd.org> flz 2008-06-03 14:59:13 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/pkg_install Makefile Makefile.inc usr.sbin/pkg_install/add perform.c usr.sbin/pkg_install/delete perform.c usr.sbin/pkg_install/lib lib.h match.c Log: SVN rev 179521 on 2008-06-03 14:58:51Z by flz MFC: distfile target, pkg registration optimization r178032,179377,179520: add distfile target r178103,178753-178754: optimize package registration Revision Changes Path 1.17.2.2 +17 -0 src/usr.sbin/pkg_install/Makefile 1.14.2.1 +0 -0 src/usr.sbin/pkg_install/Makefile.inc 1.80.2.2 +66 -24 src/usr.sbin/pkg_install/add/perform.c 1.41.18.2 +29 -10 src/usr.sbin/pkg_install/delete/perform.c 1.60.2.2 +1 -0 src/usr.sbin/pkg_install/lib/lib.h 1.20.2.2 +59 -18 src/usr.sbin/pkg_install/lib/match.c From flz at FreeBSD.org Tue Jun 3 15:02:21 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 3 15:02:27 2008 Subject: cvs commit: src/usr.sbin/pkg_install Makefile src/usr.sbin/pkg_install/add perform.c src/usr.sbin/pkg_install/delete perform.c src/usr.sbin/pkg_install/lib lib.h match.c Message-ID: <200806031502.m53F2KpB065703@repoman.freebsd.org> flz 2008-06-03 15:02:20 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/pkg_install Makefile usr.sbin/pkg_install/add perform.c usr.sbin/pkg_install/delete perform.c usr.sbin/pkg_install/lib lib.h match.c Log: SVN rev 179522 on 2008-06-03 15:02:04Z by flz MFC: distfile target, pkg registration optimization r178032,179377,179520: add distfile target r178103,178753-178754: optimize package registration Revision Changes Path 1.16.2.2 +17 -0 src/usr.sbin/pkg_install/Makefile 1.77.8.5 +66 -24 src/usr.sbin/pkg_install/add/perform.c 1.41.8.2 +29 -10 src/usr.sbin/pkg_install/delete/perform.c 1.56.2.4 +1 -0 src/usr.sbin/pkg_install/lib/lib.h 1.19.8.4 +59 -18 src/usr.sbin/pkg_install/lib/match.c From flz at FreeBSD.org Tue Jun 3 15:18:24 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 3 15:18:51 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <1212179252.1967.1.camel@localhost> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> Message-ID: On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: > On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: >> I am curious what is our policy on using long options in the base system >> (if any)? I believe that pkg_install is the first non-contributed base >> system utility to actually widely use it. For some reason I've got >> impression that use of getopt_long is considered "the Linux/GNU way", >> this API provided for compatibility purposes and its use in base system >> is discouraged. Quick grep through /use/src seemingly supports that. >> >> Can someone confirm/reject? > > I am not sure about policy, however I do appreciate the long options > sometimes. Primarily, I think they are useful (in a self-documenting > way) for use in shell scripts. I tend to prefer the single-char options > when I am doing the administration myself. I'm not aware of such policy. I think they're useful because as far as pkg_install is concerned, we are using single-char options that are hard to match to the action it's doing. Here are a couple examples: - pkg_create -h doesn't call usage() because it's already taken. - it's easy to confuse pkg_info -o and pkg_info -O. I'll back it out if general consensus is that long options should be avoided. -- Florent Thoumie flz@FreeBSD.org FreeBSD Committer From remko at elvandar.org Tue Jun 3 16:43:07 2008 From: remko at elvandar.org (Remko Lodder) Date: Tue Jun 3 16:43:10 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> Message-ID: <33dfad267556d52723e95461f5eb7c34.squirrel@galain.elvandar.org> On Tue, June 3, 2008 5:18 pm, Florent Thoumie wrote: > On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: >> On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: >>> I am curious what is our policy on using long options in the base >>> system >>> (if any)? I believe that pkg_install is the first non-contributed base >>> system utility to actually widely use it. For some reason I've got >>> impression that use of getopt_long is considered "the Linux/GNU way", >>> this API provided for compatibility purposes and its use in base system >>> is discouraged. Quick grep through /use/src seemingly supports that. >>> >>> Can someone confirm/reject? >> >> I am not sure about policy, however I do appreciate the long options >> sometimes. Primarily, I think they are useful (in a self-documenting >> way) for use in shell scripts. I tend to prefer the single-char options >> when I am doing the administration myself. > > I'm not aware of such policy. > > I think they're useful because as far as pkg_install is concerned, we > are using single-char options that are hard to match to the action > it's doing. Here are a couple examples: > > - pkg_create -h doesn't call usage() because it's already taken. > - it's easy to confuse pkg_info -o and pkg_info -O. > > I'll back it out if general consensus is that long options should be > avoided. > > -- > Florent Thoumie > flz@FreeBSD.org > FreeBSD Committer > I like the change (long opts). 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 marcus at freebsd.org Tue Jun 3 17:09:02 2008 From: marcus at freebsd.org (Joe Marcus Clarke) Date: Tue Jun 3 17:09:06 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <33dfad267556d52723e95461f5eb7c34.squirrel@galain.elvandar.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <33dfad267556d52723e95461f5eb7c34.squirrel@galain.elvandar.org> Message-ID: <48457574.20108@freebsd.org> Remko Lodder wrote: > On Tue, June 3, 2008 5:18 pm, Florent Thoumie wrote: >> On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: >>> On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: >>>> I am curious what is our policy on using long options in the base >>>> system >>>> (if any)? I believe that pkg_install is the first non-contributed base >>>> system utility to actually widely use it. For some reason I've got >>>> impression that use of getopt_long is considered "the Linux/GNU way", >>>> this API provided for compatibility purposes and its use in base system >>>> is discouraged. Quick grep through /use/src seemingly supports that. >>>> >>>> Can someone confirm/reject? >>> I am not sure about policy, however I do appreciate the long options >>> sometimes. Primarily, I think they are useful (in a self-documenting >>> way) for use in shell scripts. I tend to prefer the single-char options >>> when I am doing the administration myself. >> I'm not aware of such policy. >> >> I think they're useful because as far as pkg_install is concerned, we >> are using single-char options that are hard to match to the action >> it's doing. Here are a couple examples: >> >> - pkg_create -h doesn't call usage() because it's already taken. >> - it's easy to confuse pkg_info -o and pkg_info -O. >> >> I'll back it out if general consensus is that long options should be >> avoided. >> >> -- >> Florent Thoumie >> flz@FreeBSD.org >> FreeBSD Committer >> > > I like the change (long opts). +1 I don't see why we should abandon something that is convenient for our users just because Linux does it. Joe -- Joe Marcus Clarke FreeBSD GNOME Team :: gnome@FreeBSD.org FreeNode / #freebsd-gnome http://www.FreeBSD.org/gnome From cokane at FreeBSD.org Tue Jun 3 17:35:15 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Tue Jun 3 17:35:19 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <33dfad267556d52723e95461f5eb7c34.squirrel@galain.elvandar.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <33dfad267556d52723e95461f5eb7c34.squirrel@galain.elvandar.org> Message-ID: <1212514417.15220.30.camel@localhost> On Tue, 2008-06-03 at 18:45 +0200, Remko Lodder wrote: > On Tue, June 3, 2008 5:18 pm, Florent Thoumie wrote: > > On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: > >> On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: > >>> I am curious what is our policy on using long options in the base > >>> system > >>> (if any)? I believe that pkg_install is the first non-contributed base > >>> system utility to actually widely use it. For some reason I've got > >>> impression that use of getopt_long is considered "the Linux/GNU way", > >>> this API provided for compatibility purposes and its use in base system > >>> is discouraged. Quick grep through /use/src seemingly supports that. > >>> > >>> Can someone confirm/reject? > >> > >> I am not sure about policy, however I do appreciate the long options > >> sometimes. Primarily, I think they are useful (in a self-documenting > >> way) for use in shell scripts. I tend to prefer the single-char options > >> when I am doing the administration myself. > > > > I'm not aware of such policy. > > > > I think they're useful because as far as pkg_install is concerned, we > > are using single-char options that are hard to match to the action > > it's doing. Here are a couple examples: > > > > - pkg_create -h doesn't call usage() because it's already taken. > > - it's easy to confuse pkg_info -o and pkg_info -O. > > > > I'll back it out if general consensus is that long options should be > > avoided. > > > > -- > > Florent Thoumie > > flz@FreeBSD.org > > FreeBSD Committer > > > > I like the change (long opts). > > Cheers, > REmko > like++; -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080603/6d39a982/attachment.pgp From rdivacky at FreeBSD.org Tue Jun 3 17:55:50 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Tue Jun 3 17:55:53 2008 Subject: cvs commit: src UPDATING src/sys/compat/linux linux_mib.c Message-ID: <200806031755.m53Htos9094533@repoman.freebsd.org> rdivacky 2008-06-03 17:55:49 UTC FreeBSD src repository Modified files: . UPDATING sys/compat/linux linux_mib.c Log: SVN rev 179523 on 2008-06-03 17:50:13Z by rdivacky Switch to emulating Linux 2.6 on default. Approved by: kib (mentor) Revision Changes Path 1.530 +6 -0 src/UPDATING 1.30 +2 -2 src/sys/compat/linux/linux_mib.c From rdivacky at FreeBSD.org Tue Jun 3 18:13:12 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Tue Jun 3 18:13:15 2008 Subject: cvs commit: src UPDATING Message-ID: <200806031813.m53IDCBU097228@repoman.freebsd.org> rdivacky 2008-06-03 18:13:12 UTC FreeBSD src repository Modified files: . UPDATING Log: SVN rev 179524 on 2008-06-03 18:09:10Z by rdivacky Fix the date in the last commit. Approved by: kib (mentor) Revision Changes Path 1.531 +1 -1 src/UPDATING From jhb at FreeBSD.org Tue Jun 3 18:52:07 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 3 18:52:09 2008 Subject: cvs commit: src/sys/kern kern_clock.c Message-ID: <200806031852.m53Iq6A5000693@repoman.freebsd.org> jhb 2008-06-03 18:52:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/kern kern_clock.c Log: SVN rev 179525 on 2008-06-03 18:51:53Z by jhb Restore the global cp_time[] array that was removed when the per-CPU stats were MFC'd. This restores the ABI and allows some tools like vmstat to work on crash dumps again. Revision Changes Path 1.178.2.7 +16 -9 src/sys/kern/kern_clock.c From jhb at freebsd.org Tue Jun 3 20:12:05 2008 From: jhb at freebsd.org (John Baldwin) Date: Tue Jun 3 20:12:09 2008 Subject: cvs commit: src/lib/bind config.h src/lib/bind/bind port_after.h port_before.h src/lib/bind/isc/isc platform.h In-Reply-To: <200806030543.m535hfjh085481@repoman.freebsd.org> References: <200806030543.m535hfjh085481@repoman.freebsd.org> Message-ID: <200806031611.53492.jhb@freebsd.org> On Tuesday 03 June 2008 01:43:27 am Doug Barton wrote: > dougb 2008-06-03 05:43:41 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_6) > lib/bind config.h > lib/bind/bind port_after.h port_before.h > lib/bind/isc/isc platform.h > Log: > SVN rev 179505 on 2008-06-03 05:43:27Z by dougb > > Changes to generated files related to the 9.3.5 update. > > Revision Changes Path > 1.1.4.3 +7 -0 src/lib/bind/bind/port_after.h > 1.1.4.3 +10 -0 src/lib/bind/bind/port_before.h > 1.2.2.5 +8 -10 src/lib/bind/config.h > 1.1.4.3 +8 -3 src/lib/bind/isc/isc/platform.h Perhaps src/lib/bind/bind/config.h needs updating? It appears that HAVE_MEMMOVE isn't set, and so this new part of port_before.h is breaking the build: +#ifndef HAVE_MEMMOVE +#define memmove(a,b,c) bcopy(b,a,c) +#endif -- John Baldwin From jhb at freebsd.org Tue Jun 3 20:29:00 2008 From: jhb at freebsd.org (John Baldwin) Date: Tue Jun 3 20:29:03 2008 Subject: cvs commit: src/lib/bind config.h src/lib/bind/bind port_after.h port_before.h src/lib/bind/isc/isc platform.h In-Reply-To: <200806031611.53492.jhb@freebsd.org> References: <200806030543.m535hfjh085481@repoman.freebsd.org> <200806031611.53492.jhb@freebsd.org> Message-ID: <200806031628.16352.jhb@freebsd.org> On Tuesday 03 June 2008 04:11:53 pm John Baldwin wrote: > On Tuesday 03 June 2008 01:43:27 am Doug Barton wrote: > > dougb 2008-06-03 05:43:41 UTC > > > > FreeBSD src repository > > > > Modified files: (Branch: RELENG_6) > > lib/bind config.h > > lib/bind/bind port_after.h port_before.h > > lib/bind/isc/isc platform.h > > Log: > > SVN rev 179505 on 2008-06-03 05:43:27Z by dougb > > > > Changes to generated files related to the 9.3.5 update. > > > > Revision Changes Path > > 1.1.4.3 +7 -0 src/lib/bind/bind/port_after.h > > 1.1.4.3 +10 -0 src/lib/bind/bind/port_before.h > > 1.2.2.5 +8 -10 src/lib/bind/config.h > > 1.1.4.3 +8 -3 src/lib/bind/isc/isc/platform.h > > Perhaps src/lib/bind/bind/config.h needs updating? It appears that > HAVE_MEMMOVE isn't set, and so this new part of port_before.h is breaking the > build: > > +#ifndef HAVE_MEMMOVE > +#define memmove(a,b,c) bcopy(b,a,c) > +#endif I'm trying this patch to config.h based on comparing the changes to src/contrib/bind9/lib/bind/config.h.in: --- //depot/yahoo/ybsd_6/src/lib/bind/bind/config.h +++ /home/jhb/work/y/p4/ybsd_6/src/lib/bind/bind/config.h @@ -14,6 +14,8 @@ /* #undef POSIX_GETPWNAM_R */ /* #undef POSIX_GETGRGID_R */ /* #undef POSIX_GETGRNAM_R */ +#define HAVE_MEMMOVE +#define HAVE_MEMCHR /* #undef NEED_SETGROUPENT */ /* #undef NEED_GETGROUPLIST */ @@ -41,6 +43,7 @@ #define HAS_PW_CLASS 1 +/* #undef size_t */ /* #undef uintptr_t */ /* Shut up warnings about sputaux in stdio.h on BSD/OS pre-4.1 */ -- John Baldwin From jhb at FreeBSD.org Tue Jun 3 20:40:54 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 3 20:40:56 2008 Subject: cvs commit: src/sys/dev/xe if_xe.c if_xevar.h Message-ID: <200806032040.m53Ker4E015703@repoman.freebsd.org> jhb 2008-06-03 20:40:53 UTC FreeBSD src repository Modified files: sys/dev/xe if_xe.c if_xevar.h Log: SVN rev 179527 on 2008-06-03 20:40:33Z by jhb - Change the watchdog timer logic to match other drivers that use their own timer by keeping a once-a-second timer running that decrements a counter similar to if_timer and reset the chip if it gets down to zero via the decrement. - Use IFQ_SET_MAXLEN(). Revision Changes Path 1.66 +16 -10 src/sys/dev/xe/if_xe.c 1.8 +1 -0 src/sys/dev/xe/if_xevar.h From dougb at FreeBSD.org Tue Jun 3 20:50:02 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Tue Jun 3 20:50:10 2008 Subject: cvs commit: src/lib/bind config.h src/lib/bind/bind port_after.h port_before.h src/lib/bind/isc/isc platform.h In-Reply-To: <200806031628.16352.jhb@freebsd.org> References: <200806030543.m535hfjh085481@repoman.freebsd.org> <200806031611.53492.jhb@freebsd.org> <200806031628.16352.jhb@freebsd.org> Message-ID: <4845AE77.8040508@FreeBSD.org> John Baldwin wrote: > On Tuesday 03 June 2008 04:11:53 pm John Baldwin wrote: >> On Tuesday 03 June 2008 01:43:27 am Doug Barton wrote: >>> dougb 2008-06-03 05:43:41 UTC >>> >>> FreeBSD src repository >>> >>> Modified files: (Branch: RELENG_6) >>> lib/bind config.h >>> lib/bind/bind port_after.h port_before.h >>> lib/bind/isc/isc platform.h >>> Log: >>> SVN rev 179505 on 2008-06-03 05:43:27Z by dougb >>> >>> Changes to generated files related to the 9.3.5 update. >>> >>> Revision Changes Path >>> 1.1.4.3 +7 -0 src/lib/bind/bind/port_after.h >>> 1.1.4.3 +10 -0 src/lib/bind/bind/port_before.h >>> 1.2.2.5 +8 -10 src/lib/bind/config.h >>> 1.1.4.3 +8 -3 src/lib/bind/isc/isc/platform.h >> Perhaps src/lib/bind/bind/config.h needs updating? It appears that >> HAVE_MEMMOVE isn't set, and so this new part of port_before.h is breaking > the >> build: >> >> +#ifndef HAVE_MEMMOVE >> +#define memmove(a,b,c) bcopy(b,a,c) >> +#endif > > I'm trying this patch to config.h based on comparing the changes to > src/contrib/bind9/lib/bind/config.h.in: > > --- //depot/yahoo/ybsd_6/src/lib/bind/bind/config.h > +++ /home/jhb/work/y/p4/ybsd_6/src/lib/bind/bind/config.h > @@ -14,6 +14,8 @@ > /* #undef POSIX_GETPWNAM_R */ > /* #undef POSIX_GETGRGID_R */ > /* #undef POSIX_GETGRNAM_R */ > +#define HAVE_MEMMOVE > +#define HAVE_MEMCHR Yeah, I think that actually is the problem. The configure script for the port got that right, but when I ran configure with the sources in the base it missed it. I was trying to figure out why but I'm going to go ahead and test this patch now and investigate more while the world builds. Thanks for looking at this. My buildworld completed successfully last night before I committed, I'm also trying to figure out what was different between then and now, but I've got a sick little one at home today and that's not helping. Doug -- This .signature sanitized for your protection From gnn at FreeBSD.org Tue Jun 3 20:54:54 2008 From: gnn at FreeBSD.org (George V. Neville-Neil) Date: Tue Jun 3 20:55:11 2008 Subject: cvs commit: src/tools/tools/mctest mctest.1 mctest.cc mctest_run.sh Message-ID: <200806032054.m53Kss9G025957@repoman.freebsd.org> gnn 2008-06-03 20:54:54 UTC FreeBSD src repository Modified files: tools/tools/mctest mctest.1 mctest.cc mctest_run.sh Log: SVN rev 179528 on 2008-06-03 20:54:46Z by gnn Update mctest and its associated script to have a base port to listen on so that multiple copies can be run easily. Update documentation with group and port arguments which are optional. Revision Changes Path 1.3 +6 -0 src/tools/tools/mctest/mctest.1 1.5 +15 -9 src/tools/tools/mctest/mctest.cc 1.2 +7 -3 src/tools/tools/mctest/mctest_run.sh From sobomax at FreeBSD.org Tue Jun 3 21:05:50 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Tue Jun 3 21:05:54 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <48457574.20108@freebsd.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <33dfad267556d52723e95461f5eb7c34.squirrel@galain.elvandar.org> <48457574.20108@freebsd.org> Message-ID: <4845B220.5000300@FreeBSD.org> Joe Marcus Clarke wrote: > Remko Lodder wrote: >> On Tue, June 3, 2008 5:18 pm, Florent Thoumie wrote: >>> On Fri, May 30, 2008 at 9:27 PM, Coleman Kane >>> wrote: >>>> On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: >>>>> I am curious what is our policy on using long options in the base >>>>> system >>>>> (if any)? I believe that pkg_install is the first non-contributed base >>>>> system utility to actually widely use it. For some reason I've got >>>>> impression that use of getopt_long is considered "the Linux/GNU way", >>>>> this API provided for compatibility purposes and its use in base >>>>> system >>>>> is discouraged. Quick grep through /use/src seemingly supports that. >>>>> >>>>> Can someone confirm/reject? >>>> I am not sure about policy, however I do appreciate the long options >>>> sometimes. Primarily, I think they are useful (in a self-documenting >>>> way) for use in shell scripts. I tend to prefer the single-char options >>>> when I am doing the administration myself. >>> I'm not aware of such policy. >>> >>> I think they're useful because as far as pkg_install is concerned, we >>> are using single-char options that are hard to match to the action >>> it's doing. Here are a couple examples: >>> >>> - pkg_create -h doesn't call usage() because it's already taken. >>> - it's easy to confuse pkg_info -o and pkg_info -O. >>> >>> I'll back it out if general consensus is that long options should be >>> avoided. >> I like the change (long opts). > > +1 > > I don't see why we should abandon something that is convenient for our > users just because Linux does it. Apart from the BSD vs. GNU way, I think it's mistake that long "synonyms" have been added to existing options. The reason for that is because this is likely to promote creating superfluously incompatible scripts/software that relies on pkg_install, as developer who develops on say 8.0 may not be aware of the fact that in previous releases those long options were not existing. IMHO, long options is mostly for script writers (for whom it's harmful in this case according to the above). It adds no convenience for a CLI user. Except of maybe one or two options that one uses every day nobody can really remember what world to use without looking into the man page anyway (at which point short option wins since it's easier to type). Maybe I am old school, but for me for example remembering `-L' in wget it much easier than remembering `--relative'. -Maxim From dougb at FreeBSD.org Tue Jun 3 21:26:41 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Tue Jun 3 21:26:43 2008 Subject: cvs commit: src/lib/bind/bind config.h Message-ID: <200806032126.m53LQerq029639@repoman.freebsd.org> dougb 2008-06-03 21:26:40 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/bind/bind config.h Log: SVN rev 179529 on 2008-06-03 21:26:37Z by dougb This file should define HAVE_MEMMOVE and HAVE_MEMCHR, but for some reason configure did not pick it up. Add them here in the same way that it's done when running configure for the port. All of the other generated files are functionally equivalent to how they look when generated by the port. Revision Changes Path 1.1.4.5 +2 -0 src/lib/bind/bind/config.h From jkim at FreeBSD.org Tue Jun 3 22:35:04 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Tue Jun 3 22:35:08 2008 Subject: cvs commit: src/usr.sbin/tzsetup paths.h tzsetup.c Message-ID: <200806032234.m53MYxoS037304@repoman.freebsd.org> jkim 2008-06-03 22:34:59 UTC FreeBSD src repository Modified files: usr.sbin/tzsetup tzsetup.c Removed files: usr.sbin/tzsetup paths.h Log: SVN rev 179530 on 2008-06-03 22:34:52Z by jkim - Replace rcsid with __FBSDID. - Remove paths.h and embed it in the source. - Remove stale alpha support. - Clean up compiler warnings and fix style(9) bugs. Revision Changes Path 1.3 +0 -5 src/usr.sbin/tzsetup/paths.h (dead) 1.28 +260 -235 src/usr.sbin/tzsetup/tzsetup.c From gavin.atkinson at ury.york.ac.uk Tue Jun 3 22:37:02 2008 From: gavin.atkinson at ury.york.ac.uk (Gavin Atkinson) Date: Tue Jun 3 22:37:09 2008 Subject: cvs commit: src/sys/kern kern_clock.c In-Reply-To: <200806031852.m53Iq6A5000693@repoman.freebsd.org> References: <200806031852.m53Iq6A5000693@repoman.freebsd.org> Message-ID: <20080603224939.O94518@ury.york.ac.uk> On Tue, 3 Jun 2008, John Baldwin wrote: > jhb 2008-06-03 18:52:06 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_6) > sys/kern kern_clock.c > Log: > SVN rev 179525 on 2008-06-03 18:51:53Z by jhb > > Restore the global cp_time[] array that was removed when the per-CPU stats > were MFC'd. This restores the ABI and allows some tools like vmstat to > work on crash dumps again. Are you also planning on merging this to RELENG_7? It looks like cp_times was part of 7.0-RELEASE. As far as I can tell, this will fix PRs 123047, 123068 and 123146 on RELENG_6, and a merge to RELENG_7 would additionally fix PRs 122875 and 123014. PR 119923 is again caused by the removal of cp_times, but that user is on HEAD. Gavin From danfe at FreeBSD.org Wed Jun 4 04:18:15 2008 From: danfe at FreeBSD.org (Alexey Dokuchaev) Date: Wed Jun 4 04:18:18 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> Message-ID: <20080604041815.GA84027@FreeBSD.org> On Tue, Jun 03, 2008 at 04:18:23PM +0100, Florent Thoumie wrote: > On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: > > On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: > >> I am curious what is our policy on using long options in the base system > >> (if any)? I believe that pkg_install is the first non-contributed base > >> system utility to actually widely use it. For some reason I've got > >> impression that use of getopt_long is considered "the Linux/GNU way", > >> this API provided for compatibility purposes and its use in base system > >> is discouraged. Quick grep through /use/src seemingly supports that. > >> > >> Can someone confirm/reject? > > > > I am not sure about policy, however I do appreciate the long options > > sometimes. Primarily, I think they are useful (in a self-documenting > > way) for use in shell scripts. I tend to prefer the single-char options > > when I am doing the administration myself. > > I'm not aware of such policy. > > I think they're useful because as far as pkg_install is concerned, we > are using single-char options that are hard to match to the action > it's doing. Here are a couple examples: > > - pkg_create -h doesn't call usage() because it's already taken. > - it's easy to confuse pkg_info -o and pkg_info -O. > > I'll back it out if general consensus is that long options should be avoided. I'd rather avoid long options in *BSD utilities. They're hard to remember, easy to confuse, while not really gaining us anything useful (IMHO of course). ./danfe From sgk at troutmask.apl.washington.edu Wed Jun 4 04:40:35 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Wed Jun 4 04:40:38 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604041815.GA84027@FreeBSD.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> Message-ID: <20080604043955.GA38627@troutmask.apl.washington.edu> On Wed, Jun 04, 2008 at 04:18:15AM +0000, Alexey Dokuchaev wrote: > On Tue, Jun 03, 2008 at 04:18:23PM +0100, Florent Thoumie wrote: > > On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: > > > On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: > > >> I am curious what is our policy on using long options in the base system > > >> (if any)? I believe that pkg_install is the first non-contributed base > > >> system utility to actually widely use it. For some reason I've got > > >> impression that use of getopt_long is considered "the Linux/GNU way", > > >> this API provided for compatibility purposes and its use in base system > > >> is discouraged. Quick grep through /use/src seemingly supports that. > > >> > > >> Can someone confirm/reject? > > > > > > I am not sure about policy, however I do appreciate the long options > > > sometimes. Primarily, I think they are useful (in a self-documenting > > > way) for use in shell scripts. I tend to prefer the single-char options > > > when I am doing the administration myself. > > > > I'm not aware of such policy. > > > > I think they're useful because as far as pkg_install is concerned, we > > are using single-char options that are hard to match to the action > > it's doing. Here are a couple examples: > > > > - pkg_create -h doesn't call usage() because it's already taken. > > - it's easy to confuse pkg_info -o and pkg_info -O. > > > > I'll back it out if general consensus is that long options should be avoided. > > I'd rather avoid long options in *BSD utilities. They're hard to > remember, easy to confuse, while not really gaining us anything useful > (IMHO of course). > I agree. The argument that long options are self-documenting in a script is rather weak. If the command isn't obvious, then script writer should actually put a comment in her code. The argument that -h is already taken and can't be used is also weak, because one can always do 'man pkg_create'. I personally believe that commit should be backed out and core should establish a policy against adding long options to BSD utilities. Of course, my vote doesn't really matter in that most of my attempted contributions to FreeBSD live in limbo. -- Steve From lankfordandrew at charter.net Wed Jun 4 05:48:13 2008 From: lankfordandrew at charter.net (Andrew Lankford) Date: Wed Jun 4 05:48:42 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... Message-ID: <48462A1F.3070809@charter.net> $man find ...Ok, maybe "primaries" used with find don't count as command line options. Arguably, the output of "foo -h" is more light weight than "man foo". OTOH, if lots of long options make it necessary to do this instead: "foo -h 2>&1 | more -es", then I see where you're coming from. Why not just supply an one-letter option that is equivalent for every long option? Might make for a more verbose man page, but... Andrew Lankford From ru at FreeBSD.org Wed Jun 4 06:05:28 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Wed Jun 4 06:05:47 2008 Subject: cvs commit: src/gnu/usr.bin/groff/tmac mdoc.local Message-ID: <200806040605.m5465SPG013118@repoman.freebsd.org> ru 2008-06-04 06:05:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) gnu/usr.bin/groff/tmac mdoc.local Log: SVN rev 179531 on 2008-06-04 06:05:12Z by ru MFC: Introduce FreeBSD 8.0 to mdoc(7). Revision Changes Path 1.59.2.2 +1 -0 src/gnu/usr.bin/groff/tmac/mdoc.local From imp at FreeBSD.org Wed Jun 4 06:07:29 2008 From: imp at FreeBSD.org (Warner Losh) Date: Wed Jun 4 06:07:33 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c Message-ID: <200806040607.m5467TdL013296@repoman.freebsd.org> imp 2008-06-04 06:07:29 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c Log: SVN rev 179532 on 2008-06-04 06:07:13Z by imp o Improve the probe code dealing with interrupts. o When forced to be 10baseT, don't require that the 10baseT interface have link to succeed. Still require it for IFM_AUTO, however, since it appears that there's no way to tell if a specific type of interface worked. I'm doing a web search for a datasheet now to see if there's anything obvious. o Minor incidental formatting nits, including collapsing code of the form if (foo) { bar(); } else { if (baz) bing(); } into: if (foo) { bar(); } else if (baz) { bing(); } to save an indentation level. o Remove stray reference to 3.x config file syntax. # I believe John's patches still apply after this... Revision Changes Path 1.47 +54 -83 src/sys/dev/cs/if_cs.c From wb at freebie.xs4all.nl Wed Jun 4 06:36:37 2008 From: wb at freebie.xs4all.nl (Wilko Bulte) Date: Wed Jun 4 06:36:47 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604043955.GA38627@troutmask.apl.washington.edu> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> Message-ID: <20080604063631.GA28351@freebie.xs4all.nl> Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM -0700 .. > On Wed, Jun 04, 2008 at 04:18:15AM +0000, Alexey Dokuchaev wrote: > > On Tue, Jun 03, 2008 at 04:18:23PM +0100, Florent Thoumie wrote: > > > On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: > > > > On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: > > > >> I am curious what is our policy on using long options in the base system > > > >> (if any)? I believe that pkg_install is the first non-contributed base > > > >> system utility to actually widely use it. For some reason I've got > > > >> impression that use of getopt_long is considered "the Linux/GNU way", > > > >> this API provided for compatibility purposes and its use in base system > > > >> is discouraged. Quick grep through /use/src seemingly supports that. > > > >> > > > >> Can someone confirm/reject? > > > > > > > > I am not sure about policy, however I do appreciate the long options > > > > sometimes. Primarily, I think they are useful (in a self-documenting > > > > way) for use in shell scripts. I tend to prefer the single-char options > > > > when I am doing the administration myself. > > > > > > I'm not aware of such policy. > > > > > > I think they're useful because as far as pkg_install is concerned, we > > > are using single-char options that are hard to match to the action > > > it's doing. Here are a couple examples: > > > > > > - pkg_create -h doesn't call usage() because it's already taken. > > > - it's easy to confuse pkg_info -o and pkg_info -O. > > > > > > I'll back it out if general consensus is that long options should be avoided. > > > > I'd rather avoid long options in *BSD utilities. They're hard to > > remember, easy to confuse, while not really gaining us anything useful > > (IMHO of course). > > > > I agree. The argument that long options are self-documenting > in a script is rather weak. If the command isn't obvious, then > script writer should actually put a comment in her code. The > argument that -h is already taken and can't be used is also > weak, because one can always do 'man pkg_create'. > > I personally believe that commit should be backed out and core > should establish a policy against adding long options to BSD Gimme a break.. -- Wilko Bulte wilko@FreeBSD.org From grehan at FreeBSD.org Wed Jun 4 07:33:03 2008 From: grehan at FreeBSD.org (Peter Grehan) Date: Wed Jun 4 07:33:06 2008 Subject: cvs commit: src/sys/powerpc/aim trap.c Message-ID: <200806040733.m547X2xi021142@repoman.freebsd.org> grehan 2008-06-04 07:33:02 UTC FreeBSD src repository Modified files: sys/powerpc/aim trap.c Log: SVN rev 179533 on 2008-06-04 07:32:49Z by grehan Add link register to fatal trap printout to better diagnose NULL function pointer derefs. Revision Changes Path 1.73 +1 -0 src/sys/powerpc/aim/trap.c From rafan at FreeBSD.org Wed Jun 4 07:44:21 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Wed Jun 4 07:44:23 2008 Subject: cvs commit: src MAINTAINERS Message-ID: <200806040744.m547iLXX022063@repoman.freebsd.org> rafan 2008-06-04 07:44:20 UTC FreeBSD src repository Modified files: . MAINTAINERS Log: SVN rev 179534 on 2008-06-04 07:37:26Z by rafan - Associate myself with ncurses Revision Changes Path 1.149 +1 -0 src/MAINTAINERS From rafan at FreeBSD.org Wed Jun 4 07:51:05 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Wed Jun 4 07:51:07 2008 Subject: cvs commit: src MAINTAINERS Message-ID: <200806040751.m547p5Wt022670@repoman.freebsd.org> rafan 2008-06-04 07:51:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) . MAINTAINERS Log: SVN rev 179535 on 2008-06-04 07:37:39Z by rafan - Associate myself with ncurses Revision Changes Path 1.145.2.2 +1 -0 src/MAINTAINERS From rafan at FreeBSD.org Wed Jun 4 07:56:14 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Wed Jun 4 07:56:17 2008 Subject: cvs commit: src MAINTAINERS Message-ID: <200806040756.m547uDjK023082@repoman.freebsd.org> rafan 2008-06-04 07:56:13 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) . MAINTAINERS Log: SVN rev 179536 on 2008-06-04 07:37:56Z by rafan - Associate myself with ncurses Revision Changes Path 1.121.2.5 +1 -0 src/MAINTAINERS From rafan at FreeBSD.org Wed Jun 4 07:57:40 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Wed Jun 4 07:57:44 2008 Subject: cvs commit: src/contrib/ncurses FREEBSD-vendor Message-ID: <200806040757.m547vejo023190@repoman.freebsd.org> rafan 2008-06-04 07:57:40 UTC FreeBSD src repository Modified files: contrib/ncurses FREEBSD-vendor Log: SVN rev 179537 on 2008-06-04 07:38:38Z by rafan - update maintainer's info per src/MAINTAINERS Revision Changes Path 1.3 +1 -1 src/contrib/ncurses/FREEBSD-vendor From rafan at FreeBSD.org Wed Jun 4 07:57:57 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Wed Jun 4 07:58:03 2008 Subject: cvs commit: src/contrib/ncurses FREEBSD-upgrade INSTALL MANIFEST NEWS TO-DO aclocal.m4 config.guess config.sub configure configure.in dist.mk install-sh mk-1st.awk mk-hdr.awk mkdirs.sh mkinstalldirs src/contrib/ncurses/doc hackguide.doc ... Message-ID: <200806040757.m547vvJC023995@repoman.freebsd.org> rafan 2008-06-04 07:57:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) contrib/ncurses FREEBSD-upgrade INSTALL MANIFEST NEWS TO-DO aclocal.m4 config.guess config.sub configure configure.in dist.mk install-sh mk-1st.awk mkinstalldirs contrib/ncurses/doc hackguide.doc ncurses-intro.doc contrib/ncurses/doc/html ncurses-intro.html contrib/ncurses/form Makefile.in fld_def.c fld_dup.c fld_ftlink.c fld_link.c fld_newftyp.c frm_def.c frm_driver.c fty_alnum.c fty_alpha.c fty_enum.c fty_int.c fty_num.c fty_regex.c headers modules contrib/ncurses/include MKterm.h.awk.in curses.h.in curses.tail curses.wide headers nc_alloc.h ncurses_defs ncurses_dll.h tic.h contrib/ncurses/man Makefile.in captoinfo.1m clear.1 curs_add_wch.3x curs_addch.3x curs_attr.3x curs_border.3x curs_clear.3x curs_deleteln.3x curs_extend.3x curs_getcchar.3x curs_getyx.3x curs_insstr.3x curs_mouse.3x curs_outopts.3x curs_printw.3x curs_scanw.3x curs_scr_dump.3x curs_slk.3x curs_termattrs.3x curs_termcap.3x curs_terminfo.3x curs_trace.3x curs_util.3x default_colors.3x form.3x form_driver.3x form_field_new.3x form_field_opts.3x form_field_validation.3x form_hook.3x form_opts.3x infocmp.1m infotocap.1m legacy_coding.3x man_db.renames manlinks.sed menu_driver.3x menu_hook.3x menu_opts.3x mitem_opts.3x ncurses.3x panel.3x term.5 term.7 terminfo.head terminfo.tail tic.1m toe.1m tput.1 tset.1 contrib/ncurses/menu Makefile.in headers menu.h modules contrib/ncurses/misc Makefile.in ncurses-config.in shlib terminfo.src contrib/ncurses/ncurses Makefile.in curses.priv.h fifo_defs.h llib-lncurses llib-lncursesw modules contrib/ncurses/ncurses/base MKkeyname.awk MKlib_gen.sh MKunctrl.awk define_key.c key_defined.c keyok.c lib_addch.c lib_addstr.c lib_bkgd.c lib_color.c lib_delwin.c lib_freeall.c lib_getch.c lib_initscr.c lib_insch.c lib_instr.c lib_mouse.c lib_newterm.c lib_newwin.c lib_overlay.c lib_redrawln.c lib_refresh.c lib_restart.c lib_screen.c lib_set_term.c lib_slk.c lib_slkclear.c lib_slkinit.c lib_slkset.c lib_ungetch.c lib_window.c memmove.c resizeterm.c safe_sprintf.c tries.c wresize.c contrib/ncurses/ncurses/tinfo MKcaptab.awk MKnames.awk access.c add_tries.c comp_error.c comp_hash.c comp_parse.c db_iterator.c home_terminfo.c init_keytry.c lib_acs.c lib_baudrate.c lib_data.c lib_napms.c lib_options.c lib_raw.c lib_setup.c lib_termcap.c lib_tparm.c lib_tputs.c lib_ttyflags.c make_keys.c name_match.c parse_entry.c read_entry.c setbuf.c strings.c trim_sgr0.c write_entry.c contrib/ncurses/ncurses/trace lib_trace.c lib_traceatr.c lib_tracebits.c lib_tracechr.c lib_tracedmp.c lib_tracemse.c trace_buf.c trace_tries.c varargs.c visbuf.c contrib/ncurses/ncurses/tty hardscroll.c hashmap.c lib_mvcur.c lib_tstp.c lib_twait.c lib_vidattr.c tty_update.c contrib/ncurses/ncurses/widechar lib_cchar.c lib_get_wch.c lib_in_wchnstr.c lib_unget_wch.c lib_vid_attr.c lib_wunctrl.c contrib/ncurses/panel Makefile.in headers modules contrib/ncurses/progs Makefile.in clear.c dump_entry.c infocmp.c modules progs.priv.h tic.c toe.c tput.c tset.c contrib/ncurses/tack COPYING HISTORY Makefile.in README ansi.c charset.c color.c control.c crum.c edit.c fun.c init.c menu.c modes.c modules output.c pad.c scan.c sync.c sysdep.c tack.1 tack.c tack.h Added files: (Branch: RELENG_7) contrib/ncurses mk-hdr.awk mkdirs.sh contrib/ncurses/man curs_legacy.3x curs_opaque.3x curs_threads.3x contrib/ncurses/ncurses llib-lncursest contrib/ncurses/ncurses/base use_window.c contrib/ncurses/ncurses/tinfo MKcaptab.sh MKcodes.awk entries.c use_screen.c contrib/ncurses/ncurses/widechar lib_key_name.c Log: SVN rev 179538 on 2008-06-04 07:42:51Z by rafan MFC ncurses 5.6-20080503 rev 174993, 176187, 178866 Revision Changes Path 1.1.2.1 +7 -3 src/contrib/ncurses/FREEBSD-upgrade 1.1.1.7.2.1 +67 -2 src/contrib/ncurses/INSTALL 1.1.1.7.2.1 +25 -25 src/contrib/ncurses/MANIFEST 1.1.1.8.2.1 +644 -9 src/contrib/ncurses/NEWS 1.1.1.2.2.1 +10 -13 src/contrib/ncurses/TO-DO 1.1.1.7.2.1 +518 -197 src/contrib/ncurses/aclocal.m4 1.1.1.6.2.1 +113 -85 src/contrib/ncurses/config.guess 1.1.1.6.2.1 +77 -27 src/contrib/ncurses/config.sub 1.1.1.8.2.1 +2720 -1923 src/contrib/ncurses/configure 1.1.1.8.2.1 +255 -35 src/contrib/ncurses/configure.in 1.1.1.8.2.1 +4 -3 src/contrib/ncurses/dist.mk 1.1.1.2.2.1 +1 -1 src/contrib/ncurses/doc/hackguide.doc 1.1.1.3.2.1 +23 -15 src/contrib/ncurses/doc/html/ncurses-intro.html 1.1.1.2.2.1 +16 -17 src/contrib/ncurses/doc/ncurses-intro.doc 1.1.1.6.2.1 +5 -7 src/contrib/ncurses/form/Makefile.in 1.1.1.3.2.1 +7 -6 src/contrib/ncurses/form/fld_def.c 1.1.1.3.2.1 +4 -3 src/contrib/ncurses/form/fld_dup.c 1.1.1.3.2.1 +4 -3 src/contrib/ncurses/form/fld_ftlink.c 1.1.1.3.2.1 +4 -3 src/contrib/ncurses/form/fld_link.c 1.1.1.3.2.1 +4 -3 src/contrib/ncurses/form/fld_newftyp.c 1.1.1.3.2.1 +6 -4 src/contrib/ncurses/form/frm_def.c 1.1.1.3.2.1 +124 -9 src/contrib/ncurses/form/frm_driver.c 1.1.1.3.2.1 +12 -6 src/contrib/ncurses/form/fty_alnum.c 1.1.1.3.2.1 +12 -6 src/contrib/ncurses/form/fty_alpha.c 1.1.1.5.2.1 +9 -5 src/contrib/ncurses/form/fty_enum.c 1.1.1.3.2.1 +9 -5 src/contrib/ncurses/form/fty_int.c 1.1.1.3.2.1 +9 -5 src/contrib/ncurses/form/fty_num.c 1.1.1.3.2.1 +11 -8 src/contrib/ncurses/form/fty_regex.c 1.1.1.1.44.1 +2 -1 src/contrib/ncurses/form/headers 1.1.1.2.2.1 +5 -3 src/contrib/ncurses/form/modules 1.1.1.6.2.1 +26 -25 src/contrib/ncurses/include/MKterm.h.awk.in 1.1.1.9.2.1 +180 -71 src/contrib/ncurses/include/curses.h.in 1.1.1.2.2.1 +2 -1 src/contrib/ncurses/include/curses.tail 1.1.1.2.2.1 +6 -3 src/contrib/ncurses/include/curses.wide 1.1.1.2.24.1 +9 -2 src/contrib/ncurses/include/headers 1.1.1.4.2.1 +8 -2 src/contrib/ncurses/include/nc_alloc.h 1.1.1.4.2.1 +8 -7 src/contrib/ncurses/include/ncurses_defs 1.1.1.3.2.1 +9 -2 src/contrib/ncurses/include/ncurses_dll.h 1.1.1.8.2.1 +7 -22 src/contrib/ncurses/include/tic.h 1.1.1.3.2.1 +91 -48 src/contrib/ncurses/install-sh 1.1.1.5.2.1 +4 -4 src/contrib/ncurses/man/Makefile.in 1.1.1.4.2.1 +14 -14 src/contrib/ncurses/man/captoinfo.1m 1.1.1.3.2.1 +6 -6 src/contrib/ncurses/man/clear.1 1.1.1.2.2.1 +3 -3 src/contrib/ncurses/man/curs_add_wch.3x 1.1.1.3.2.1 +4 -4 src/contrib/ncurses/man/curs_addch.3x 1.1.1.5.2.1 +5 -2 src/contrib/ncurses/man/curs_attr.3x 1.1.1.4.2.1 +4 -4 src/contrib/ncurses/man/curs_border.3x 1.1.1.2.2.1 +8 -2 src/contrib/ncurses/man/curs_clear.3x 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/man/curs_deleteln.3x 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/man/curs_extend.3x 1.1.1.2.2.1 +4 -4 src/contrib/ncurses/man/curs_getcchar.3x 1.1.1.2.2.1 +21 -6 src/contrib/ncurses/man/curs_getyx.3x 1.1.1.4.2.1 +2 -2 src/contrib/ncurses/man/curs_insstr.3x 1.1.1.1.2.1 +93 -0 src/contrib/ncurses/man/curs_legacy.3x (new) 1.1.1.5.2.1 +26 -16 src/contrib/ncurses/man/curs_mouse.3x 1.1.1.1.2.1 +133 -0 src/contrib/ncurses/man/curs_opaque.3x (new) 1.1.1.5.2.1 +3 -3 src/contrib/ncurses/man/curs_outopts.3x 1.1.1.3.2.1 +4 -4 src/contrib/ncurses/man/curs_printw.3x 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/man/curs_scanw.3x 1.1.1.3.2.1 +2 -2 src/contrib/ncurses/man/curs_scr_dump.3x 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/man/curs_slk.3x 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/man/curs_termattrs.3x 1.1.1.6.2.1 +4 -4 src/contrib/ncurses/man/curs_termcap.3x 1.1.1.6.2.1 +18 -8 src/contrib/ncurses/man/curs_terminfo.3x 1.1.1.2.2.1 +603 -0 src/contrib/ncurses/man/curs_threads.3x (new) 1.1.1.3.2.1 +3 -3 src/contrib/ncurses/man/curs_trace.3x 1.1.1.3.2.1 +41 -6 src/contrib/ncurses/man/curs_util.3x 1.1.1.3.2.1 +3 -4 src/contrib/ncurses/man/default_colors.3x 1.1.1.3.2.1 +3 -3 src/contrib/ncurses/man/form.3x 1.1.1.3.2.1 +54 -3 src/contrib/ncurses/man/form_driver.3x 1.1.1.3.2.1 +4 -4 src/contrib/ncurses/man/form_field_new.3x 1.1.1.2.2.1 +6 -6 src/contrib/ncurses/man/form_field_opts.3x 1.1.1.3.2.1 +2 -2 src/contrib/ncurses/man/form_field_validation.3x 1.1.1.2.2.1 +10 -10 src/contrib/ncurses/man/form_hook.3x 1.1.1.2.2.1 +6 -6 src/contrib/ncurses/man/form_opts.3x 1.1.1.5.2.1 +15 -15 src/contrib/ncurses/man/infocmp.1m 1.1.1.3.2.1 +9 -9 src/contrib/ncurses/man/infotocap.1m 1.1.1.1.4.1 +2 -2 src/contrib/ncurses/man/legacy_coding.3x 1.1.1.6.2.1 +16 -10 src/contrib/ncurses/man/man_db.renames 1.1.1.3.2.1 +3 -2 src/contrib/ncurses/man/manlinks.sed 1.1.1.3.2.1 +46 -15 src/contrib/ncurses/man/menu_driver.3x 1.1.1.2.2.1 +10 -10 src/contrib/ncurses/man/menu_hook.3x 1.1.1.2.2.1 +6 -6 src/contrib/ncurses/man/menu_opts.3x 1.1.1.2.2.1 +6 -6 src/contrib/ncurses/man/mitem_opts.3x 1.1.1.7.2.1 +192 -77 src/contrib/ncurses/man/ncurses.3x 1.1.1.3.2.1 +21 -16 src/contrib/ncurses/man/panel.3x 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/man/term.5 1.1.1.4.2.1 +8 -8 src/contrib/ncurses/man/term.7 1.1.1.3.2.1 +7 -7 src/contrib/ncurses/man/terminfo.head 1.1.1.5.2.1 +14 -12 src/contrib/ncurses/man/terminfo.tail 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/man/tic.1m 1.1.1.3.2.1 +30 -16 src/contrib/ncurses/man/toe.1m 1.1.1.3.2.1 +38 -38 src/contrib/ncurses/man/tput.1 1.1.1.3.2.1 +2 -2 src/contrib/ncurses/man/tset.1 1.1.1.6.2.1 +5 -7 src/contrib/ncurses/menu/Makefile.in 1.1.1.1.44.1 +2 -1 src/contrib/ncurses/menu/headers 1.1.1.4.2.1 +6 -6 src/contrib/ncurses/menu/menu.h 1.1.1.2.2.1 +5 -3 src/contrib/ncurses/menu/modules 1.1.1.4.2.1 +9 -8 src/contrib/ncurses/misc/Makefile.in 1.1.1.1.4.1 +7 -2 src/contrib/ncurses/misc/ncurses-config.in 1.1.1.3.2.1 +2 -1 src/contrib/ncurses/misc/shlib 1.1.1.8.2.1 +460 -90 src/contrib/ncurses/misc/terminfo.src 1.1.1.6.2.1 +131 -52 src/contrib/ncurses/mk-1st.awk 1.1.1.1.2.1 +107 -0 src/contrib/ncurses/mk-hdr.awk (new) 1.1.1.1.2.1 +51 -0 src/contrib/ncurses/mkdirs.sh (new) 1.1.1.3.2.1 +0 -38 src/contrib/ncurses/mkinstalldirs 1.1.1.7.2.1 +22 -29 src/contrib/ncurses/ncurses/Makefile.in 1.1.1.5.2.1 +52 -31 src/contrib/ncurses/ncurses/base/MKkeyname.awk 1.1.1.5.2.1 +9 -4 src/contrib/ncurses/ncurses/base/MKlib_gen.sh 1.1.1.5.2.1 +119 -41 src/contrib/ncurses/ncurses/base/MKunctrl.awk 1.1.1.3.2.1 +9 -4 src/contrib/ncurses/ncurses/base/define_key.c 1.1.1.1.4.1 +12 -10 src/contrib/ncurses/ncurses/base/key_defined.c 1.1.1.3.2.1 +8 -6 src/contrib/ncurses/ncurses/base/keyok.c 1.1.1.5.2.1 +18 -14 src/contrib/ncurses/ncurses/base/lib_addch.c 1.1.1.6.2.1 +9 -6 src/contrib/ncurses/ncurses/base/lib_addstr.c 1.1.1.4.2.1 +3 -3 src/contrib/ncurses/ncurses/base/lib_bkgd.c 1.1.1.7.2.1 +35 -13 src/contrib/ncurses/ncurses/base/lib_color.c 1.1.1.3.24.1 +23 -12 src/contrib/ncurses/ncurses/base/lib_delwin.c 1.1.1.6.2.1 +31 -28 src/contrib/ncurses/ncurses/base/lib_freeall.c 1.1.1.7.2.1 +123 -89 src/contrib/ncurses/ncurses/base/lib_getch.c 1.1.1.4.2.1 +12 -6 src/contrib/ncurses/ncurses/base/lib_initscr.c 1.1.1.4.2.1 +20 -4 src/contrib/ncurses/ncurses/base/lib_insch.c 1.1.1.4.2.1 +4 -6 src/contrib/ncurses/ncurses/base/lib_instr.c 1.1.1.7.2.1 +14 -10 src/contrib/ncurses/ncurses/base/lib_mouse.c 1.1.1.6.2.1 +20 -19 src/contrib/ncurses/ncurses/base/lib_newterm.c 1.1.1.6.2.1 +63 -48 src/contrib/ncurses/ncurses/base/lib_newwin.c 1.1.1.4.2.1 +92 -71 src/contrib/ncurses/ncurses/base/lib_overlay.c 1.1.1.4.2.1 +7 -3 src/contrib/ncurses/ncurses/base/lib_redrawln.c 1.1.1.5.2.1 +121 -23 src/contrib/ncurses/ncurses/base/lib_refresh.c 1.1.1.4.2.1 +10 -10 src/contrib/ncurses/ncurses/base/lib_restart.c 1.1.1.4.2.1 +14 -5 src/contrib/ncurses/ncurses/base/lib_screen.c 1.1.1.7.2.1 +188 -129 src/contrib/ncurses/ncurses/base/lib_set_term.c 1.1.1.4.2.1 +11 -12 src/contrib/ncurses/ncurses/base/lib_slk.c 1.1.1.4.2.1 +19 -14 src/contrib/ncurses/ncurses/base/lib_slkclear.c 1.1.1.2.24.1 +10 -6 src/contrib/ncurses/ncurses/base/lib_slkinit.c 1.1.1.4.2.1 +9 -4 src/contrib/ncurses/ncurses/base/lib_slkset.c 1.1.1.4.2.1 +29 -18 src/contrib/ncurses/ncurses/base/lib_ungetch.c 1.1.1.4.2.1 +5 -3 src/contrib/ncurses/ncurses/base/lib_window.c 1.1.1.2.24.1 +3 -3 src/contrib/ncurses/ncurses/base/memmove.c 1.1.1.4.2.1 +176 -51 src/contrib/ncurses/ncurses/base/resizeterm.c 1.1.1.5.2.1 +29 -27 src/contrib/ncurses/ncurses/base/safe_sprintf.c 1.1.1.4.2.1 +12 -10 src/contrib/ncurses/ncurses/base/tries.c 1.1.1.3.2.1 +78 -0 src/contrib/ncurses/ncurses/base/use_window.c (new) 1.1.1.5.2.1 +13 -6 src/contrib/ncurses/ncurses/base/wresize.c 1.1.1.9.2.1 +389 -48 src/contrib/ncurses/ncurses/curses.priv.h 1.1.1.2.24.1 +5 -5 src/contrib/ncurses/ncurses/fifo_defs.h 1.1.1.7.2.1 +306 -243 src/contrib/ncurses/ncurses/llib-lncurses 1.1.1.1.2.1 +3493 -0 src/contrib/ncurses/ncurses/llib-lncursest (new) 1.1.1.2.2.1 +319 -254 src/contrib/ncurses/ncurses/llib-lncursesw 1.1.1.6.2.1 +20 -11 src/contrib/ncurses/ncurses/modules 1.1.1.3.2.1 +65 -69 src/contrib/ncurses/ncurses/tinfo/MKcaptab.awk 1.1.1.1.2.1 +149 -0 src/contrib/ncurses/ncurses/tinfo/MKcaptab.sh (new) 1.1.1.1.2.1 +161 -0 src/contrib/ncurses/ncurses/tinfo/MKcodes.awk (new) 1.1.1.3.2.1 +142 -84 src/contrib/ncurses/ncurses/tinfo/MKnames.awk 1.1.1.5.2.1 +5 -5 src/contrib/ncurses/ncurses/tinfo/access.c 1.1.1.4.2.1 +17 -22 src/contrib/ncurses/ncurses/tinfo/add_tries.c 1.1.1.5.2.1 +17 -17 src/contrib/ncurses/ncurses/tinfo/comp_error.c 1.1.1.4.2.1 +73 -38 src/contrib/ncurses/ncurses/tinfo/comp_hash.c 1.1.1.7.2.1 +25 -71 src/contrib/ncurses/ncurses/tinfo/comp_parse.c 1.1.1.1.4.1 +25 -27 src/contrib/ncurses/ncurses/tinfo/db_iterator.c 1.1.1.1.2.1 +144 -0 src/contrib/ncurses/ncurses/tinfo/entries.c (new) 1.1.1.4.2.1 +11 -10 src/contrib/ncurses/ncurses/tinfo/home_terminfo.c 1.1.1.5.2.1 +14 -14 src/contrib/ncurses/ncurses/tinfo/init_keytry.c 1.1.1.5.2.1 +10 -18 src/contrib/ncurses/ncurses/tinfo/lib_acs.c 1.2.2.1 +20 -7 src/contrib/ncurses/ncurses/tinfo/lib_baudrate.c 1.1.1.3.2.1 +207 -2 src/contrib/ncurses/ncurses/tinfo/lib_data.c 1.1.1.6.2.1 +3 -3 src/contrib/ncurses/ncurses/tinfo/lib_napms.c 1.1.1.7.2.1 +10 -10 src/contrib/ncurses/ncurses/tinfo/lib_options.c 1.7.2.1 +4 -4 src/contrib/ncurses/ncurses/tinfo/lib_raw.c 1.1.1.6.2.1 +88 -56 src/contrib/ncurses/ncurses/tinfo/lib_setup.c 1.12.2.1 +31 -31 src/contrib/ncurses/ncurses/tinfo/lib_termcap.c 1.1.1.5.2.1 +65 -88 src/contrib/ncurses/ncurses/tinfo/lib_tparm.c 1.1.1.7.2.1 +9 -8 src/contrib/ncurses/ncurses/tinfo/lib_tputs.c 1.1.1.6.2.1 +53 -31 src/contrib/ncurses/ncurses/tinfo/lib_ttyflags.c 1.1.1.5.2.1 +3 -3 src/contrib/ncurses/ncurses/tinfo/make_keys.c 1.1.1.4.2.1 +12 -12 src/contrib/ncurses/ncurses/tinfo/name_match.c 1.1.1.7.2.1 +11 -9 src/contrib/ncurses/ncurses/tinfo/parse_entry.c 1.1.1.5.2.1 +3 -3 src/contrib/ncurses/ncurses/tinfo/read_entry.c 1.1.1.3.2.1 +3 -3 src/contrib/ncurses/ncurses/tinfo/setbuf.c 1.1.1.3.2.1 +3 -3 src/contrib/ncurses/ncurses/tinfo/strings.c 1.1.1.1.4.1 +4 -6 src/contrib/ncurses/ncurses/tinfo/trim_sgr0.c 1.1.1.2.2.1 +60 -0 src/contrib/ncurses/ncurses/tinfo/use_screen.c (new) 1.1.1.6.2.1 +11 -11 src/contrib/ncurses/ncurses/tinfo/write_entry.c 1.1.1.6.2.1 +137 -36 src/contrib/ncurses/ncurses/trace/lib_trace.c 1.1.1.6.2.1 +28 -22 src/contrib/ncurses/ncurses/trace/lib_traceatr.c 1.1.1.7.2.1 +24 -3 src/contrib/ncurses/ncurses/trace/lib_tracebits.c 1.1.1.6.2.1 +8 -7 src/contrib/ncurses/ncurses/trace/lib_tracechr.c 1.1.1.4.2.1 +41 -16 src/contrib/ncurses/ncurses/trace/lib_tracedmp.c 1.1.1.4.2.1 +10 -17 src/contrib/ncurses/ncurses/trace/lib_tracemse.c 1.1.1.3.2.1 +19 -23 src/contrib/ncurses/ncurses/trace/trace_buf.c 1.1.1.3.24.1 +16 -14 src/contrib/ncurses/ncurses/trace/trace_tries.c 1.1.1.3.2.1 +14 -12 src/contrib/ncurses/ncurses/trace/varargs.c 1.1.1.2.2.1 +21 -11 src/contrib/ncurses/ncurses/trace/visbuf.c 1.1.1.2.24.1 +17 -17 src/contrib/ncurses/ncurses/tty/hardscroll.c 1.1.1.5.2.1 +26 -17 src/contrib/ncurses/ncurses/tty/hashmap.c 1.1.1.7.2.1 +6 -5 src/contrib/ncurses/ncurses/tty/lib_mvcur.c 1.1.1.7.2.1 +23 -29 src/contrib/ncurses/ncurses/tty/lib_tstp.c 1.1.1.7.2.1 +28 -29 src/contrib/ncurses/ncurses/tty/lib_twait.c 1.1.1.6.2.1 +22 -20 src/contrib/ncurses/ncurses/tty/lib_vidattr.c 1.1.1.7.2.1 +17 -10 src/contrib/ncurses/ncurses/tty/tty_update.c 1.1.1.2.2.1 +5 -4 src/contrib/ncurses/ncurses/widechar/lib_cchar.c 1.1.1.2.2.1 +3 -3 src/contrib/ncurses/ncurses/widechar/lib_get_wch.c 1.1.1.2.2.1 +17 -9 src/contrib/ncurses/ncurses/widechar/lib_in_wchnstr.c 1.1.1.1.2.1 +62 -0 src/contrib/ncurses/ncurses/widechar/lib_key_name.c (new) 1.1.1.2.2.1 +12 -8 src/contrib/ncurses/ncurses/widechar/lib_unget_wch.c 1.1.1.2.2.1 +3 -2 src/contrib/ncurses/ncurses/widechar/lib_vid_attr.c 1.1.1.2.2.1 +7 -9 src/contrib/ncurses/ncurses/widechar/lib_wunctrl.c 1.1.1.6.2.1 +5 -7 src/contrib/ncurses/panel/Makefile.in 1.1.1.1.44.1 +2 -1 src/contrib/ncurses/panel/headers 1.1.1.1.44.1 +4 -2 src/contrib/ncurses/panel/modules 1.1.1.6.2.1 +19 -16 src/contrib/ncurses/progs/Makefile.in 1.1.1.2.2.1 +3 -2 src/contrib/ncurses/progs/clear.c 1.1.1.7.2.1 +16 -9 src/contrib/ncurses/progs/dump_entry.c 1.1.1.6.2.1 +101 -62 src/contrib/ncurses/progs/infocmp.c 1.1.1.3.2.1 +5 -3 src/contrib/ncurses/progs/modules 1.1.1.6.2.1 +11 -2 src/contrib/ncurses/progs/progs.priv.h 1.1.1.8.2.1 +28 -10 src/contrib/ncurses/progs/tic.c 1.1.1.5.2.1 +11 -11 src/contrib/ncurses/progs/toe.c 1.1.1.6.2.1 +5 -3 src/contrib/ncurses/progs/tput.c 1.1.1.6.2.1 +20 -4 src/contrib/ncurses/progs/tset.c 1.1.1.2.2.1 +0 -340 src/contrib/ncurses/tack/COPYING 1.1.1.3.2.1 +0 -49 src/contrib/ncurses/tack/HISTORY 1.1.1.7.2.1 +0 -211 src/contrib/ncurses/tack/Makefile.in 1.1.1.2.24.1 +0 -14 src/contrib/ncurses/tack/README 1.1.1.5.2.1 +0 -854 src/contrib/ncurses/tack/ansi.c 1.1.1.5.2.1 +0 -711 src/contrib/ncurses/tack/charset.c 1.1.1.3.2.1 +0 -769 src/contrib/ncurses/tack/color.c 1.1.1.3.2.1 +0 -664 src/contrib/ncurses/tack/control.c 1.1.1.3.2.1 +0 -426 src/contrib/ncurses/tack/crum.c 1.1.1.5.2.1 +0 -992 src/contrib/ncurses/tack/edit.c 1.1.1.3.2.1 +0 -925 src/contrib/ncurses/tack/fun.c 1.1.1.4.2.1 +0 -316 src/contrib/ncurses/tack/init.c 1.1.1.3.2.1 +0 -421 src/contrib/ncurses/tack/menu.c 1.1.1.2.2.1 +0 -913 src/contrib/ncurses/tack/modes.c 1.1.1.2.2.1 +0 -45 src/contrib/ncurses/tack/modules 1.1.1.5.2.1 +0 -818 src/contrib/ncurses/tack/output.c 1.1.1.3.2.1 +0 -1958 src/contrib/ncurses/tack/pad.c 1.1.1.2.2.1 +0 -264 src/contrib/ncurses/tack/scan.c 1.1.1.4.2.1 +0 -436 src/contrib/ncurses/tack/sync.c 1.1.1.6.2.1 +0 -504 src/contrib/ncurses/tack/sysdep.c 1.1.1.2.2.1 +0 -340 src/contrib/ncurses/tack/tack.1 1.1.1.3.2.1 +0 -603 src/contrib/ncurses/tack/tack.c 1.1.1.5.2.1 +0 -428 src/contrib/ncurses/tack/tack.h From rafan at FreeBSD.org Wed Jun 4 07:57:58 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Wed Jun 4 07:58:05 2008 Subject: cvs commit: src/lib/ncurses/ncurses Makefile ncurses_cfg.h Message-ID: <200806040757.m547vwE5024026@repoman.freebsd.org> rafan 2008-06-04 07:57:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/ncurses/ncurses Makefile ncurses_cfg.h Log: SVN rev 179539 on 2008-06-04 07:43:58Z by rafan MFC ncurses 5.6-20080503 build glue rev 174999, 176190, 178869 Revision Changes Path 1.92.2.3 +56 -19 src/lib/ncurses/ncurses/Makefile 1.8.4.1 +8 -9 src/lib/ncurses/ncurses/ncurses_cfg.h From philip at FreeBSD.org Wed Jun 4 08:56:00 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Wed Jun 4 09:04:26 2008 Subject: cvs commit: src/usr.sbin/bsnmpd/modules/snmp_pf pf_snmp.c Message-ID: <200806040855.m548txHQ039473@repoman.freebsd.org> philip 2008-06-04 08:55:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/bsnmpd/modules/snmp_pf pf_snmp.c Log: SVN rev 179540 on 2008-06-04 08:55:49Z by philip MFC r179476: Use INSERT_OBJECT_INT_LINK_INDEX macro instead of TAILQ_INSERT_TAIL when filling the table of ALTQ queues retrieved from the kernel. It is possible for the kernel to return the queues not by pa.altq.qid order. When this happens, pf_snmp would only partially fill its table. PR: bin/120974 Submitted by: Mykola Dzham Revision Changes Path 1.8.2.1 +1 -1 src/usr.sbin/bsnmpd/modules/snmp_pf/pf_snmp.c From kevlo at FreeBSD.org Wed Jun 4 10:36:44 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Wed Jun 4 10:36:49 2008 Subject: cvs commit: src/sys/dev/usb usbdevs Message-ID: <200806041036.m54AahXe051310@repoman.freebsd.org> kevlo 2008-06-04 10:36:33 UTC FreeBSD src repository Modified files: sys/dev/usb usbdevs Log: SVN rev 179541 on 2008-06-04 10:36:24Z by kevlo Add device IDs for Ricoh R5U870-based OEM cameras Revision Changes Path 1.358 +8 -0 src/sys/dev/usb/usbdevs From cokane at FreeBSD.org Wed Jun 4 11:48:33 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Wed Jun 4 11:48:42 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604041815.GA84027@FreeBSD.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> Message-ID: <1212580012.15220.58.camel@localhost> On Wed, 2008-06-04 at 04:18 +0000, Alexey Dokuchaev wrote: > On Tue, Jun 03, 2008 at 04:18:23PM +0100, Florent Thoumie wrote: > > On Fri, May 30, 2008 at 9:27 PM, Coleman Kane wrote: > > > On Fri, 2008-05-30 at 12:58 -0700, Maxim Sobolev wrote: > > >> I am curious what is our policy on using long options in the base system > > >> (if any)? I believe that pkg_install is the first non-contributed base > > >> system utility to actually widely use it. For some reason I've got > > >> impression that use of getopt_long is considered "the Linux/GNU way", > > >> this API provided for compatibility purposes and its use in base system > > >> is discouraged. Quick grep through /use/src seemingly supports that. > > >> > > >> Can someone confirm/reject? > > > > > > I am not sure about policy, however I do appreciate the long options > > > sometimes. Primarily, I think they are useful (in a self-documenting > > > way) for use in shell scripts. I tend to prefer the single-char options > > > when I am doing the administration myself. > > > > I'm not aware of such policy. > > > > I think they're useful because as far as pkg_install is concerned, we > > are using single-char options that are hard to match to the action > > it's doing. Here are a couple examples: > > > > - pkg_create -h doesn't call usage() because it's already taken. > > - it's easy to confuse pkg_info -o and pkg_info -O. > > > > I'll back it out if general consensus is that long options should be avoided. > > I'd rather avoid long options in *BSD utilities. They're hard to > remember, easy to confuse, while not really gaining us anything useful > (IMHO of course). > > ./danfe > Short options aren't going away, nor should they. So if you can't remember them, then you can rest assured that you won't have to. -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080604/ac24af66/attachment.pgp From philip at FreeBSD.org Wed Jun 4 12:22:18 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Wed Jun 4 12:22:23 2008 Subject: cvs commit: src/share/syscons/keymaps colemak.iso15.acc.kbd Message-ID: <200806041222.m54CMGUb061943@repoman.freebsd.org> philip 2008-06-04 12:22:16 UTC FreeBSD src repository Modified files: share/syscons/keymaps colemak.iso15.acc.kbd Log: SVN rev 179542 on 2008-06-04 12:22:03Z by philip Recode the colemak keymap to ISO-8859. I accidently converted the file to Unicode before committing it and kbdcontrol doesn't like that very much. Spotted by: ed Pointy hat to: philip MFC after: 1 day Revision Changes Path 1.2 +28 -28 src/share/syscons/keymaps/colemak.iso15.acc.kbd From sgk at troutmask.apl.washington.edu Wed Jun 4 15:00:56 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Wed Jun 4 15:01:02 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604063631.GA28351@freebie.xs4all.nl> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> Message-ID: <20080604150013.GA44358@troutmask.apl.washington.edu> On Wed, Jun 04, 2008 at 08:36:31AM +0200, Wilko Bulte wrote: > Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM -0700 .. > > > > I personally believe that commit should be backed out and core > > should establish a policy against adding long options to BSD > > Gimme a break.. > Note I wrote "I personally believe". You don't have to agree with me. -- Steve From jhb at FreeBSD.org Wed Jun 4 15:20:08 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Jun 4 15:20:25 2008 Subject: cvs commit: src/sys/dev/xe if_xe.c if_xevar.h Message-ID: <200806041520.m54FK61R092143@repoman.freebsd.org> jhb 2008-06-04 15:20:06 UTC FreeBSD src repository Modified files: sys/dev/xe if_xe.c if_xevar.h Log: SVN rev 179543 on 2008-06-04 15:19:58Z by jhb - Split the interrupt handler up into separate subroutines for rx, tx, and MAC events. - Use bus_*() rather than bus_space_*() and remove the bus space tag and handle from the softc. - Retire unused macros for examining CIS tuples. Revision Changes Path 1.67 +75 -47 src/sys/dev/xe/if_xe.c 1.9 +4 -14 src/sys/dev/xe/if_xevar.h From csjp at FreeBSD.org Wed Jun 4 15:38:37 2008 From: csjp at FreeBSD.org (Christian S.J. Peron) Date: Wed Jun 4 15:38:47 2008 Subject: cvs commit: src/sys/security/mac mac_inet.c Message-ID: <200806041538.m54FcamC093646@repoman.freebsd.org> csjp 2008-06-04 15:38:36 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/security/mac mac_inet.c Log: SVN rev 179544 on 2008-06-04 15:38:28Z by csjp MFC cvs revision 1.19 Plug a memory leak Revision Changes Path 1.11.2.1 +4 -0 src/sys/security/mac/mac_inet.c From csjp at FreeBSD.org Wed Jun 4 15:59:08 2008 From: csjp at FreeBSD.org (Christian S.J. Peron) Date: Wed Jun 4 15:59:14 2008 Subject: cvs commit: src/sys/security/mac mac_inet.c Message-ID: <200806041559.m54Fx7jg095254@repoman.freebsd.org> csjp 2008-06-04 15:59:07 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/security/mac mac_inet.c Log: SVN rev 179545 on 2008-06-04 15:58:52Z by csjp Back out previous commit. These entry point names are not valid in this branch. Revision Changes Path 1.11.2.2 +0 -4 src/sys/security/mac/mac_inet.c From csjp at FreeBSD.org Wed Jun 4 16:13:26 2008 From: csjp at FreeBSD.org (Christian S.J. Peron) Date: Wed Jun 4 16:13:28 2008 Subject: cvs commit: src/sys/security/mac mac_inet.c Message-ID: <200806041613.m54GDP2d097551@repoman.freebsd.org> csjp 2008-06-04 16:13:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/security/mac mac_inet.c Log: SVN rev 179546 on 2008-06-04 16:13:12Z by csjp MFC revision 1.19 with some tweaks to the entry point naming. Revision Changes Path 1.11.2.3 +4 -0 src/sys/security/mac/mac_inet.c From danfe at FreeBSD.org Wed Jun 4 17:30:59 2008 From: danfe at FreeBSD.org (Alexey Dokuchaev) Date: Wed Jun 4 17:31:10 2008 Subject: cvs commit: src/bin/sh expand.c parser.c parser.h In-Reply-To: <20080602.213239.-861051005.imp@bsdimp.com> References: <1212440845.18384.49.camel@shumai.marcuscom.com> <20080602.204014.-222576415.imp@bsdimp.com> <1212461214.18384.83.camel@shumai.marcuscom.com> <20080602.213239.-861051005.imp@bsdimp.com> Message-ID: <20080604173058.GA4863@FreeBSD.org> On Mon, Jun 02, 2008 at 09:32:39PM -0600, M. Warner Losh wrote: > Another option would be to add support for it to our test, maybe with > a warning... :-) Glad you put that smiley there! ./danfe From jhb at freebsd.org Wed Jun 4 18:13:14 2008 From: jhb at freebsd.org (John Baldwin) Date: Wed Jun 4 18:13:31 2008 Subject: cvs commit: src/sys/kern kern_clock.c In-Reply-To: <20080603224939.O94518@ury.york.ac.uk> References: <200806031852.m53Iq6A5000693@repoman.freebsd.org> <20080603224939.O94518@ury.york.ac.uk> Message-ID: <200806041118.01307.jhb@freebsd.org> On Tuesday 03 June 2008 06:05:50 pm Gavin Atkinson wrote: > On Tue, 3 Jun 2008, John Baldwin wrote: > > > jhb 2008-06-03 18:52:06 UTC > > > > FreeBSD src repository > > > > Modified files: (Branch: RELENG_6) > > sys/kern kern_clock.c > > Log: > > SVN rev 179525 on 2008-06-03 18:51:53Z by jhb > > > > Restore the global cp_time[] array that was removed when the per-CPU stats > > were MFC'd. This restores the ABI and allows some tools like vmstat to > > work on crash dumps again. > > Are you also planning on merging this to RELENG_7? It looks like > cp_times was part of 7.0-RELEASE. > > As far as I can tell, this will fix PRs 123047, 123068 and 123146 on > RELENG_6, and a merge to RELENG_7 would additionally fix PRs 122875 and > 123014. > > PR 119923 is again caused by the removal of cp_times, but that user is on > HEAD. I want to fix the various kvm consumers in HEAD and 7.x using the new per-cpu stuff instead of this. As far as the PR's, ports running agains the live system should be using the sysctl's instead of direct kvm access anyway so they will work on 8.0, etc. So those PR's should stay open including 123047, 123068, 123146, and 119923. PR 122875/123014 (123014 has a patch) should be fixed in the tree and I'll work on that. -- John Baldwin From wb at freebie.xs4all.nl Wed Jun 4 19:13:46 2008 From: wb at freebie.xs4all.nl (Wilko Bulte) Date: Wed Jun 4 19:13:53 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604150013.GA44358@troutmask.apl.washington.edu> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> Message-ID: <20080604191339.GA31570@freebie.xs4all.nl> Quoting Steve Kargl, who wrote on Wed, Jun 04, 2008 at 08:00:13AM -0700 .. > On Wed, Jun 04, 2008 at 08:36:31AM +0200, Wilko Bulte wrote: > > Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM -0700 .. > > > > > > I personally believe that commit should be backed out and core > > > should establish a policy against adding long options to BSD > > > > Gimme a break.. > > > > Note I wrote "I personally believe". You don't have to agree > with me. Well I indeed do not agree. Aren't the developers old enough to make this kind of judgement call themselves, without all sorts of written policies? I would think they all are! cheers, Wilko -- Wilko Bulte wilko@FreeBSD.org From dwmalone at FreeBSD.org Wed Jun 4 19:17:13 2008 From: dwmalone at FreeBSD.org (David Malone) Date: Wed Jun 4 19:17:16 2008 Subject: cvs commit: src/usr.bin/su su.c Message-ID: <200806041917.m54JHDYS033239@repoman.freebsd.org> dwmalone 2008-06-04 19:17:13 UTC FreeBSD src repository Modified files: usr.bin/su su.c Log: SVN rev 179547 on 2008-06-04 19:16:54Z by dwmalone Fix a strict aliasing warning - I think it is really telling us that the way char * and void * pointers may not be stored in the same way. Revision Changes Path 1.88 +3 -2 src/usr.bin/su/su.c From sgk at troutmask.apl.washington.edu Wed Jun 4 19:30:40 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Wed Jun 4 19:30:43 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604191339.GA31570@freebie.xs4all.nl> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> Message-ID: <20080604192955.GA46284@troutmask.apl.washington.edu> On Wed, Jun 04, 2008 at 09:13:39PM +0200, Wilko Bulte wrote: > Quoting Steve Kargl, who wrote on Wed, Jun 04, 2008 at 08:00:13AM -0700 .. > > On Wed, Jun 04, 2008 at 08:36:31AM +0200, Wilko Bulte wrote: > > > Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM -0700 .. > > > > > > > > I personally believe that commit should be backed out and core > > > > should establish a policy against adding long options to BSD > > > > > > Gimme a break.. > > > > > > > Note I wrote "I personally believe". You don't have to agree > > with me. > > Well I indeed do not agree. Aren't the developers old enough > to make this kind of judgement call themselves, without all sorts of written > policies? Apparently, not. See the recent commit to pkg_create. > > I would think they all are! > Where do we stop? Should we add long options to all /usr/bin utilities? Why stop at /usr/bin, let's add long options to /usr/sbin, /bin, /sbin, /rescue, etc. -- Steve From jhb at FreeBSD.org Wed Jun 4 19:33:30 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Jun 4 19:33:36 2008 Subject: cvs commit: src/sys/dev/xe if_xe.c Message-ID: <200806041933.m54JXT8Y034645@repoman.freebsd.org> jhb 2008-06-04 19:33:29 UTC FreeBSD src repository Modified files: sys/dev/xe if_xe.c Log: SVN rev 179548 on 2008-06-04 19:33:23Z by jhb Gah, always reset the timer for the watchdog check. Revision Changes Path 1.68 +1 -1 src/sys/dev/xe/if_xe.c From remko at FreeBSD.org Wed Jun 4 19:37:59 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Wed Jun 4 19:38:04 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604192955.GA46284@troutmask.apl.washington.edu> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> Message-ID: <4846EF10.1020803@FreeBSD.org> Steve Kargl wrote: > On Wed, Jun 04, 2008 at 09:13:39PM +0200, Wilko Bulte wrote: >> Quoting Steve Kargl, who wrote on Wed, Jun 04, 2008 at 08:00:13AM -0700 .. >>> On Wed, Jun 04, 2008 at 08:36:31AM +0200, Wilko Bulte wrote: >>>> Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM -0700 .. >>>>> I personally believe that commit should be backed out and core >>>>> should establish a policy against adding long options to BSD >>>> Gimme a break.. >>>> >>> Note I wrote "I personally believe". You don't have to agree >>> with me. >> Well I indeed do not agree. Aren't the developers old enough >> to make this kind of judgement call themselves, without all sorts of written >> policies? > > Apparently, not. See the recent commit to pkg_create. The developer that did the commit is mature enough to make the decision. That others perhaps do not agree doesn't reflect the maturity of the developer / maintainer. > >> I would think they all are! >> > > Where do we stop? Should we add long options to all > /usr/bin utilities? Why stop at /usr/bin, let's add > long options to /usr/sbin, /bin, /sbin, /rescue, etc. > That is not your call. If a maintainer wants to add all options he can consider, he is free to do so. Though others might not appreciate that as much as he does. It can be discussed ofcourse, but to a certain extend. Also note that the time wasted by these interesting discussions is time better spend on different areas (y'all go close a PR or two orso ;-)). Thanks, move on, nothing to see here. 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 cokane at FreeBSD.org Wed Jun 4 19:44:31 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Wed Jun 4 19:44:40 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604192955.GA46284@troutmask.apl.washington.edu> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> Message-ID: <1212608575.15220.109.camel@localhost> On Wed, 2008-06-04 at 12:29 -0700, Steve Kargl wrote: > On Wed, Jun 04, 2008 at 09:13:39PM +0200, Wilko Bulte wrote: > > Quoting Steve Kargl, who wrote on Wed, Jun 04, 2008 at 08:00:13AM -0700 .. > > > On Wed, Jun 04, 2008 at 08:36:31AM +0200, Wilko Bulte wrote: > > > > Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM -0700 .. > > > > > > > > > > I personally believe that commit should be backed out and core > > > > > should establish a policy against adding long options to BSD > > > > > > > > Gimme a break.. > > > > > > > > > > Note I wrote "I personally believe". You don't have to agree > > > with me. > > > > Well I indeed do not agree. Aren't the developers old enough > > to make this kind of judgement call themselves, without all sorts of written > > policies? > > Apparently, not. See the recent commit to pkg_create. > I believe the comment was meant to convey that this sort of call should be left to the person doing the work. Making up a book of obtuse policy rules such as this, for purposes that aren't very concrete, doesn't seem to serve anybody well. The whole reason long options exist is because a significant group of users actually find them helpful. It's really counterproductive (not to mention narrow-minded) to dismiss that. I support the efforts of flz on this one, he took the time to do the work and wanted to commit it. I actually find your above comment to be pretty offensive, and dismissive of the hard word that flz has done. > > > > I would think they all are! > > > > Where do we stop? Should we add long options to all > /usr/bin utilities? Why stop at /usr/bin, let's add > long options to /usr/sbin, /bin, /sbin, /rescue, etc. > I'm sure if someone has some "add long options to /bin/cp, etc..." patches, we can surely discuss them on this list as adults and respect the decision to add new features without deprecating any existing features, even if we won't be making use of those new features. -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080604/b0725e02/attachment.pgp From dwmalone at FreeBSD.org Wed Jun 4 19:50:47 2008 From: dwmalone at FreeBSD.org (David Malone) Date: Wed Jun 4 19:50:48 2008 Subject: cvs commit: src/usr.bin/lex ccl.c dfa.c ecs.c gen.c initscan.c misc.c nfa.c sym.c tblcmp.c Message-ID: <200806041950.m54JokwB036339@repoman.freebsd.org> dwmalone 2008-06-04 19:50:46 UTC FreeBSD src repository Modified files: usr.bin/lex ccl.c dfa.c ecs.c gen.c initscan.c misc.c nfa.c sym.c tblcmp.c Log: SVN rev 179549 on 2008-06-04 19:50:34Z by dwmalone De-register declarations. Revision Changes Path 1.7 +2 -2 src/usr.bin/lex/ccl.c 1.7 +10 -10 src/usr.bin/lex/dfa.c 1.7 +1 -1 src/usr.bin/lex/ecs.c 1.8 +13 -13 src/usr.bin/lex/gen.c 1.10 +22 -22 src/usr.bin/lex/initscan.c 1.7 +14 -14 src/usr.bin/lex/misc.c 1.7 +1 -1 src/usr.bin/lex/nfa.c 1.7 +10 -10 src/usr.bin/lex/sym.c 1.7 +13 -13 src/usr.bin/lex/tblcmp.c From sobomax at FreeBSD.org Wed Jun 4 19:55:02 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Wed Jun 4 19:55:06 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <4846EF10.1020803@FreeBSD.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <4846EF10.1020803@FreeBSD.org> Message-ID: <4846F30A.5070204@FreeBSD.org> Remko Lodder wrote: >> Where do we stop? Should we add long options to all >> /usr/bin utilities? Why stop at /usr/bin, let's add >> long options to /usr/sbin, /bin, /sbin, /rescue, etc. >> > > That is not your call. If a maintainer wants to add all options he can > consider, he is free to do so. Though others might not appreciate that > as much as he does. It can be discussed ofcourse, but to a certain extend. It's not your call either. We have style(9), which says: For consistency, getopt(3) should be used to parse options. Options should be sorted in the getopt(3) call and the switch statement, unless parts of the switch cascade. Elements in a switch statement that cascade should have a FALLTHROUGH comment. Numerical arguments should be checked for accuracy. Code that cannot be reached should have a NOTREACHED com- ment. There is nothing about getopt_long(3) being acceptable replacement/addition to the getopt(3). -Maxim From sgk at troutmask.apl.washington.edu Wed Jun 4 19:58:32 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Wed Jun 4 19:58:38 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <1212608575.15220.109.camel@localhost> References: <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> Message-ID: <20080604195747.GA46922@troutmask.apl.washington.edu> On Wed, Jun 04, 2008 at 03:42:55PM -0400, Coleman Kane wrote: > On Wed, 2008-06-04 at 12:29 -0700, Steve Kargl wrote: >>> >>> I would think they all are! >>> >> >> Where do we stop? Should we add long options to all >> /usr/bin utilities? Why stop at /usr/bin, let's add >> long options to /usr/sbin, /bin, /sbin, /rescue, etc. >> > > I'm sure if someone has some "add long options to /bin/cp, etc..." > patches, we can surely discuss them on this list as adults and respect > the decision to add new features without deprecating any existing > features, even if we won't be making use of those new features. Would you commit the changes before or after the discussion? -- Steve From sobomax at FreeBSD.org Wed Jun 4 20:03:54 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Wed Jun 4 20:03:57 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <1212608575.15220.109.camel@localhost> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> Message-ID: <4846F520.6040400@FreeBSD.org> Coleman Kane wrote: >> Where do we stop? Should we add long options to all >> /usr/bin utilities? Why stop at /usr/bin, let's add >> long options to /usr/sbin, /bin, /sbin, /rescue, etc. >> > > I'm sure if someone has some "add long options to /bin/cp, etc..." > patches, we can surely discuss them on this list as adults and respect > the decision to add new features without deprecating any existing > features, even if we won't be making use of those new features. Please don't. Idea to add "long" options to existing "short" ones in base system utilities is very short-sighted. It could look like a cool pet project for somebody learning how to use C ("gee, ma, look, I have made a huge improvement in FreeBSD cp(1) in less than 10 minutes of work"), but in the long run it will hurt us all since sooner or later you will find yourself struggling with scripts that don't work on release X just because it was created on release X+N and therefore uses those cool new long options. -Maxim From marcel at FreeBSD.org Wed Jun 4 20:08:19 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Wed Jun 4 20:08:24 2008 Subject: cvs commit: src/sbin/geom Makefile src/sbin/geom/class/label geom_label.c src/sbin/geom/class/part geom_part.c src/sbin/geom/core geom.c Message-ID: <200806042008.m54K8JA6039038@repoman.freebsd.org> marcel 2008-06-04 20:08:19 UTC FreeBSD src repository Modified files: sbin/geom Makefile sbin/geom/class/label geom_label.c sbin/geom/class/part geom_part.c sbin/geom/core geom.c Log: SVN rev 179550 on 2008-06-04 20:07:59Z by marcel Replace checks for RESCUE in sources with checks for STATIC_GEOM_CLASSES and define STATIC_GEOM_CLASSES when building the rescue binary. This way geom can more easily be part of other crunched binaries, as it requires only a Makefile change. Revision Changes Path 1.4 +1 -1 src/sbin/geom/Makefile 1.12 +1 -1 src/sbin/geom/class/label/geom_label.c 1.6 +1 -1 src/sbin/geom/class/part/geom_part.c 1.36 +5 -5 src/sbin/geom/core/geom.c From remko at FreeBSD.org Wed Jun 4 20:12:58 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Wed Jun 4 20:13:03 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <4846F30A.5070204@FreeBSD.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <4846EF10.1020803@FreeBSD.org> <4846F30A.5070204@FreeBSD.org> Message-ID: <4846F743.3040903@FreeBSD.org> Maxim Sobolev wrote: > Remko Lodder wrote: >>> Where do we stop? Should we add long options to all >>> /usr/bin utilities? Why stop at /usr/bin, let's add >>> long options to /usr/sbin, /bin, /sbin, /rescue, etc. >>> >> >> That is not your call. If a maintainer wants to add all options he can >> consider, he is free to do so. Though others might not appreciate that >> as much as he does. It can be discussed ofcourse, but to a certain >> extend. > > It's not your call either. We have style(9), which says: > > For consistency, getopt(3) should be used to parse options. Options > should be sorted in the getopt(3) call and the switch statement, > unless > parts of the switch cascade. Elements in a switch statement that > cascade > should have a FALLTHROUGH comment. Numerical arguments should be > checked > for accuracy. Code that cannot be reached should have a NOTREACHED > com- > ment. > > There is nothing about getopt_long(3) being acceptable > replacement/addition to the getopt(3). > getopt(3) is implemented, and it's expanded by getopt_long(3) in this case. The requirement is fullfilled and made more readable (in my eyes) then before. Not everyone agrees, too bad, the world is not perfect :-). (I'll end discussing this with this email). 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 jhb at FreeBSD.org Wed Jun 4 20:27:06 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Jun 4 20:27:17 2008 Subject: cvs commit: src/sys/dev/xe if_xe.c if_xe_pccard.c if_xevar.h Message-ID: <200806042027.m54KR61S041968@repoman.freebsd.org> jhb 2008-06-04 20:27:06 UTC FreeBSD src repository Modified files: sys/dev/xe if_xe.c if_xe_pccard.c if_xevar.h Log: SVN rev 179551 on 2008-06-04 20:26:57Z by jhb Style(9) xe(4). The MD5 sums are different, but comparing the dissassemblies the only changes are from gcc moving some global variables around slightly. Revision Changes Path 1.69 +1534 -1478 src/sys/dev/xe/if_xe.c 1.34 +24 -25 src/sys/dev/xe/if_xe_pccard.c 1.10 +43 -43 src/sys/dev/xe/if_xevar.h From wb at freebie.xs4all.nl Wed Jun 4 20:29:45 2008 From: wb at freebie.xs4all.nl (Wilko Bulte) Date: Wed Jun 4 20:29:54 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <1212608575.15220.109.camel@localhost> References: <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> Message-ID: <20080604202939.GB32282@freebie.xs4all.nl> Quoting Coleman Kane, who wrote on Wed, Jun 04, 2008 at 03:42:55PM -0400 .. > On Wed, 2008-06-04 at 12:29 -0700, Steve Kargl wrote: > > On Wed, Jun 04, 2008 at 09:13:39PM +0200, Wilko Bulte wrote: > > > Quoting Steve Kargl, who wrote on Wed, Jun 04, 2008 at 08:00:13AM -0700 .. > > > > On Wed, Jun 04, 2008 at 08:36:31AM +0200, Wilko Bulte wrote: > > > > > Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM -0700 .. > > > > > > > > > > > > I personally believe that commit should be backed out and core > > > > > > should establish a policy against adding long options to BSD > > > > > > > > > > Gimme a break.. > > > > > > > > > > > > > Note I wrote "I personally believe". You don't have to agree > > > > with me. > > > > > > Well I indeed do not agree. Aren't the developers old enough > > > to make this kind of judgement call themselves, without all sorts of written > > > policies? > > > > Apparently, not. See the recent commit to pkg_create. > > > > I believe the comment was meant to convey that this sort of call should > be left to the person doing the work. Making up a book of obtuse policy > rules such as this, for purposes that aren't very concrete, doesn't seem Well Coleman, I do apologise for being slightly terse but indeed, that is exactly what I meant to say. I prefer common sense over a library of rule books any day. > to serve anybody well. The whole reason long options exist is because a > significant group of users actually find them helpful. It's really > counterproductive (not to mention narrow-minded) to dismiss that. > > I support the efforts of flz on this one, he took the time to do the > work and wanted to commit it. I actually find your above comment to be > pretty offensive, and dismissive of the hard word that flz has done. > > > > > > > I would think they all are! > > > > > > > Where do we stop? Should we add long options to all > > /usr/bin utilities? Why stop at /usr/bin, let's add > > long options to /usr/sbin, /bin, /sbin, /rescue, etc. > > > > I'm sure if someone has some "add long options to /bin/cp, etc..." > patches, we can surely discuss them on this list as adults and respect > the decision to add new features without deprecating any existing > features, even if we won't be making use of those new features. Amen. > -- > Coleman Kane --- end of quoted text --- -- Wilko Bulte wilko@FreeBSD.org From brooks at FreeBSD.org Wed Jun 4 20:22:19 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Wed Jun 4 20:50:01 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <4846F30A.5070204@FreeBSD.org> References: <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <4846EF10.1020803@FreeBSD.org> <4846F30A.5070204@FreeBSD.org> Message-ID: <20080604202227.GA79834@lor.one-eyed-alien.net> On Wed, Jun 04, 2008 at 12:54:50PM -0700, Maxim Sobolev wrote: > Remko Lodder wrote: >>> Where do we stop? Should we add long options to all >>> /usr/bin utilities? Why stop at /usr/bin, let's add >>> long options to /usr/sbin, /bin, /sbin, /rescue, etc. The existence of programs with long options clearly does not imply that all programs will grow them. If people start adding long options to yes(1) we'll have something to discuss, but we're a long way from that point. :) >> That is not your call. If a maintainer wants to add all options he can >> consider, he is free to do so. Though others might not appreciate that as >> much as he does. It can be discussed ofcourse, but to a certain extend. > > It's not your call either. We have style(9), which says: > > For consistency, getopt(3) should be used to parse options. Options > should be sorted in the getopt(3) call and the switch statement, unless > parts of the switch cascade. Elements in a switch statement that cascade > should have a FALLTHROUGH comment. Numerical arguments should be checked > for accuracy. Code that cannot be reached should have a NOTREACHED com- > ment. > > There is nothing about getopt_long(3) being acceptable replacement/addition > to the getopt(3). style(9) is and will remain a guideline, not a stick with which to beat your fellow developers. In this case there is clearly an implicit "if your application's arguments are parsable with getopt(3)". For an application with long options, getopt_long(3) is obviously the right solution. -- Brooks -------------- 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/20080604/e023472f/attachment.pgp From cokane at FreeBSD.org Wed Jun 4 21:26:02 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Wed Jun 4 21:26:07 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <4846F520.6040400@FreeBSD.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> <4846F520.6040400@FreeBSD.org> Message-ID: <1212614663.15220.136.camel@localhost> On Wed, 2008-06-04 at 13:03 -0700, Maxim Sobolev wrote: > Coleman Kane wrote: > >> Where do we stop? Should we add long options to all > >> /usr/bin utilities? Why stop at /usr/bin, let's add > >> long options to /usr/sbin, /bin, /sbin, /rescue, etc. > >> > > > > I'm sure if someone has some "add long options to /bin/cp, etc..." > > patches, we can surely discuss them on this list as adults and respect > > the decision to add new features without deprecating any existing > > features, even if we won't be making use of those new features. > > Please don't. Idea to add "long" options to existing "short" ones in > base system utilities is very short-sighted. It could look like a cool > pet project for somebody learning how to use C ("gee, ma, look, I have > made a huge improvement in FreeBSD cp(1) in less than 10 minutes of > work"), but in the long run it will hurt us all since sooner or later > you will find yourself struggling with scripts that don't work on > release X just because it was created on release X+N and therefore uses > those cool new long options. > > -Maxim > It's probably premature to begin shooting that down. I merely picked "cp" as the first tool that came to mind which lives in /bin. Maybe I should have picked on /bin/pax or /usr/bin/lex instead, or something else with 20+ getopt args :P. My point was that it was pretty absurd to start treating this like it will turn into a giant mad party of getopt_long conversions. I don't really have the inclination to go out and add long options to any tool myself, because I have better things to do with my time. As I said earlier, I think it's a better policy to discuss this when the patch/commit comes along, and see what people's feelings are on the matter then. I don't think it is anybody's place at the moment (except maybe core@) to try and deter a committer from making what some feel is a beneficial change to software. Using your argument above about scripts written for release X+1 not working under release X means that the addition of "-a" to /bin/cp should also be reverted, as well as numerous other changes that have happened to CLI tools like ifconfig. There are probably at least 100-or-so different things that change with each .0 release, this is why you have those monstrous case statements in configure scripts. I don't see how it is fair to cherry-pick the getopt_long changes here for these reasons, and chastise someone who's doing good work. Let's just calm down a bit here on this, I seem to have counted at least 6 people (including myself) who either liked it or at least felt that it is up to flz to make that call. Only 2 specifically wanted the commit yanked, and 1 person stated that they didn't like it but didn't state if they wanted it yanked. That sounds like a pretty good start to the decision making process. Remember, this is supposed to be fun. -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080604/38d46e60/attachment.pgp From sgk at troutmask.apl.washington.edu Wed Jun 4 22:30:18 2008 From: sgk at troutmask.apl.washington.edu (Steve Kargl) Date: Wed Jun 4 22:30:22 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <1212614663.15220.136.camel@localhost> References: <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> <4846F520.6040400@FreeBSD.org> <1212614663.15220.136.camel@localhost> Message-ID: <20080604222931.GA50114@troutmask.apl.washington.edu> On Wed, Jun 04, 2008 at 05:24:23PM -0400, Coleman Kane wrote: > It's probably premature to begin shooting that down. I merely picked > "cp" as the first tool that came to mind which lives in /bin. Maybe I > should have picked on /bin/pax or /usr/bin/lex instead, or something > else with 20+ getopt args :P. > > My point was that it was pretty absurd to start treating this like it > will turn into a giant mad party of getopt_long conversions. I don't > really have the inclination to go out and add long options to any tool > myself, because I have better things to do with my time. As I said > earlier, I think it's a better policy to discuss this when the > patch/commit comes along, and see what people's feelings are on the > matter then. I don't think it is anybody's place at the moment (except > maybe core@) to try and deter a committer from making what some feel is > a beneficial change to software. > Please commit. The patch gives conformance to Posix (ie., IEEE Std 1003.1, 2004 Edition). As a bonus, I've included long options for compatibility with GNU wc. Index: wc.1 =================================================================== RCS file: /home/ncvs/src/usr.bin/wc/wc.1,v retrieving revision 1.25 diff -u -p -r1.25 wc.1 --- wc.1 21 Dec 2006 22:59:07 -0000 1.25 +++ wc.1 4 Jun 2008 22:27:15 -0000 @@ -71,25 +71,19 @@ the last file. .Pp The following options are available: .Bl -tag -width indent -.It Fl c +.It Fl c , -bytes The number of bytes in each input file is written to the standard output. -This will cancel out any prior usage of the -.Fl m -option. -.It Fl l +.It Fl l , -lines The number of lines in each input file is written to the standard output. -.It Fl m +.It Fl m , -chars The number of characters in each input file is written to the standard output. If the current locale does not support multibyte characters, this is equivalent to the .Fl c option. -This will cancel out any prior usage of the -.Fl c -option. -.It Fl w +.It Fl w , -words The number of words in each input file is written to the standard output. .El Index: wc.c =================================================================== RCS file: /home/ncvs/src/usr.bin/wc/wc.c,v retrieving revision 1.21 diff -u -p -r1.21 wc.c --- wc.c 27 Dec 2004 22:27:56 -0000 1.21 +++ wc.c 4 Jun 2008 22:27:15 -0000 @@ -53,6 +53,7 @@ __FBSDID("$FreeBSD: src/usr.bin/wc/wc.c, #include #include #include +#include #include #include #include @@ -68,6 +69,15 @@ int doline, doword, dochar, domulti; static int cnt(const char *); static void usage(void); +static char opts[] = "clmw"; +static struct option longopts[] = { + { "bytes", no_argument, NULL, 'c' }, + { "lines", no_argument, NULL, 'l' }, + { "chars", no_argument, NULL, 'm' }, + { "words", no_argument, NULL, 'w' }, + { NULL, 0, NULL, 0 }, +}; + int main(int argc, char *argv[]) { @@ -75,7 +85,7 @@ main(int argc, char *argv[]) (void) setlocale(LC_CTYPE, ""); - while ((ch = getopt(argc, argv, "clmw")) != -1) + while ((ch = getopt_long(argc, argv, opts, longopts, NULL)) != -1) switch((char)ch) { case 'l': doline = 1; @@ -85,11 +95,13 @@ main(int argc, char *argv[]) break; case 'c': dochar = 1; - domulti = 0; + if (domulti) + errx(1, "-c conflicts with -m"); break; case 'm': domulti = 1; - dochar = 0; + if (dochar) + errx(1, "-m conflicts with -c"); break; case '?': default: @@ -263,6 +275,6 @@ word: gotsp = 1; static void usage() { - (void)fprintf(stderr, "usage: wc [-clmw] [file ...]\n"); + (void)fprintf(stderr, "usage: wc [-c|-m][-lw] [file ...]\n"); exit(1); } From mlaier at FreeBSD.org Wed Jun 4 23:32:12 2008 From: mlaier at FreeBSD.org (Max Laier) Date: Wed Jun 4 23:32:14 2008 Subject: cvs commit: src/usr.sbin/setfib setfib.c Message-ID: <200806042332.m54NWCCv072882@repoman.freebsd.org> mlaier 2008-06-04 23:32:12 UTC FreeBSD src repository Modified files: usr.sbin/setfib setfib.c Log: SVN rev 179552 on 2008-06-04 23:31:53Z by mlaier Add a newline after usage. Revision Changes Path 1.3 +1 -1 src/usr.sbin/setfib/setfib.c From ihteshamali at gmail.com Thu Jun 5 01:39:27 2008 From: ihteshamali at gmail.com (Ihtesham Ali) Date: Thu Jun 5 01:39:30 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604041815.GA84027@FreeBSD.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> Message-ID: <83fc7c9e0806041812o2a249bf8w9e63d8f38f6ae39e@mail.gmail.com> On 04/06/2008, Alexey Dokuchaev wrote: > I'd rather avoid long options in *BSD utilities. They're hard to > remember, easy to confuse, while not really gaining us anything useful Exactly! -1 Florent Thoumie's bringing up a possibility and considering to go by the general consensus is healthy though, and I completely appreciate it. On 05/06/2008, Wilko Bulte wrote: > Well I indeed do not agree. Aren't the developers old enough > to make this kind of judgement call themselves, without all sorts of written > policies? Policies in such projects are generally not against anyone but for everyone. It's nice of the committer that they in a way have sought your opinion. I'm sure you don't want FreeBSD to look like a haywire Linux distro anytime soon. I hope FreeBSD doesn't become a soup. P.S. I am not against the Linux culture, in fact I am happy that they are happy with it. I merely referred to FreeBSD's distinction of being an excellent advanced operating system project from every aspect. On 05/06/2008, Coleman Kane wrote: > Making up a book of obtuse policy > rules such as this, for purposes that aren't very concrete, doesn't seem > to serve anybody well. Anything that can affect habits of a larger group of BSDists is at least worth a second thought. Long options are not just about me but a lot more out there who may not even know of the discussion, so I'd try to be more considerate. -- Ali From danfe at FreeBSD.org Thu Jun 5 03:17:29 2008 From: danfe at FreeBSD.org (Alexey Dokuchaev) Date: Thu Jun 5 03:17:32 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <4846F743.3040903@FreeBSD.org> References: <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <4846EF10.1020803@FreeBSD.org> <4846F30A.5070204@FreeBSD.org> <4846F743.3040903@FreeBSD.org> Message-ID: <20080605031728.GA94057@FreeBSD.org> On Wed, Jun 04, 2008 at 10:12:51PM +0200, Remko Lodder wrote: > Maxim Sobolev wrote: > > > > For consistency, getopt(3) should be used to parse options. Options > > should be sorted in the getopt(3) call and the switch statement, > >unless > > parts of the switch cascade. Elements in a switch statement that > >cascade > > should have a FALLTHROUGH comment. Numerical arguments should be > >checked > > for accuracy. Code that cannot be reached should have a NOTREACHED > >com- > > ment. > > > >There is nothing about getopt_long(3) being acceptable > >replacement/addition to the getopt(3). > > > > getopt(3) is implemented, and it's expanded by getopt_long(3) in this > case. The requirement is fullfilled and made more readable (in my > eyes) then before. It breaks *consistency* across vast majority of BSD utilities. Also, being readable in your eyes does not automagically make them so in others'. ./danfe From imp at FreeBSD.org Thu Jun 5 05:51:35 2008 From: imp at FreeBSD.org (Warner Losh) Date: Thu Jun 5 05:51:39 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c Message-ID: <200806050551.m555pZdd036944@repoman.freebsd.org> imp 2008-06-05 05:51:35 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c Log: SVN rev 179553 on 2008-06-05 05:51:19Z by imp Fix the media auto code by breaking it :-). Auto now just means 'use 10BaseT' since it required 10BaseT to have carrier to switch to it. This chip makes it hard to do proper auto, so we don't do it. We can't test carrier on things easily. Don't insist on carrier when we set the media. Don't report failures. Remove a 1s! delay that appears to not be needed. With these patches, and John Baldwin's patches, I'm able to pass packets on my IBM EtherJet card again. Revision Changes Path 1.48 +8 -19 src/sys/dev/cs/if_cs.c From yanefbsd at gmail.com Thu Jun 5 06:17:32 2008 From: yanefbsd at gmail.com (Garrett Cooper) Date: Thu Jun 5 06:17:37 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604150013.GA44358@troutmask.apl.washington.edu> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> Message-ID: <103F0EDE-6609-4B57-ABB0-43F4CCDD8A62@gmail.com> On Jun 4, 2008, at 8:00 AM, Steve Kargl wrote: > On Wed, Jun 04, 2008 at 08:36:31AM +0200, Wilko Bulte wrote: >> Quoting Steve Kargl, who wrote on Tue, Jun 03, 2008 at 09:39:55PM >> -0700 .. >>> >>> I personally believe that commit should be backed out and core >>> should establish a policy against adding long options to BSD >> >> Gimme a break.. >> > > Note I wrote "I personally believe". You don't have to agree > with me. I don't see establishing this point as another issue, because many things along these lines are already covered by style(9). I personally believe that the issue with the short options isn't the fact that they're short, but the fact that many of the options tend to vary, between pkg_* applications. Before I ventured off and was working on other projects for school and work I actually wrote down a matrix of all of the given options that were in pkg_* -- as well as options which I thought would be helpful in new commands I was going to add to pkg_*, and I was able to draw more points of consistency between the current pkg_* commands and options. So, while long options may be better in some places, I don't think they will do more other than lengthen manpages and "help messages" -- a GNU~ish behavior that I don't think we will benefit from here because the help message will span more than one terminal screen, unless you pipe the output to less or more. Thus, that's why I vote (for the unfortunate non-POLA change) to make the options consistent amongst all pkg_* apps. I've attached the matrix I made a few months ago, just for reference. Some options may not exist anymore due to the phasing out of options within pkg_add as well as pkg_check/pkg_sign going out the window as well... Thanks, -Garrett -------------- next part -------------- add create delete info version a Delete all packages Show all installed packages A b Use blocksize var B c Fetch packing list (- or file) Do one-liner for each pkg C Chrootdir Set conflicts d Fetch desc (- or file) Prune directories Do long desc for each pkg D Set display file Don't exec deinstall script Display install file for each pkg e E Use extended regex Exact match only f Force install Fetch the packing list Force deinstall Show pkg list instructions for each pkg F Don't report already installed pkgs as errors g Show pkgs w/ non-matching checksums G Match exact pkg name Don't expand glob patterns Don't expand glob patterns h Make tar follow symlinks Print help msg H i Set preinstall script Request for confirmation Show install script for each pkg I Don't exec install scripts Set postinstall script Show index line for each pkg Use INDEX for determining versioning info j Use bzip2 compression Show reqs script J k Set deinstall script Show deinstall script K Keep pkg Set postdeinstall script Keep downloaded pkg l Prefix each header w/ str Positive limit to expression char L Show all files in pkg Negative limit to expression char m Show mtree file M n Don't install pkg Don't deinstall pkg N Default -> no o Set ports origin Show origin path Show origin instead of pkg name O Use packing list mode for creating packing lists from ports Show only pkgs by origin p Prefix to install to Prefix to create from Prefix to delete from Show install prefix P Same as -p, but applies to leaf packages (-r only) Set package dependencies to the list provided Show revision of pkg tools q Be quiet, but print out non-human compat printouts Be quiet Q Do -q, but also prefix with pkg name r Remote fetch pkgs Set reqs script Recursive removal Show list of dependent pkgs R Don't record pkg install Do the same as -r, recursively s Set src dir Show total size of all pkg files Limit output to provided str S Set base dir t Use arg for mkstemp(3) Use arg for mkstemp(3) Use arg for mkstemp(3) Compare pair of expressions T Do same as -t, but return appropriate exit code u U v Verbose Verbose Verbose Verbose Verbose V Packing list format version w W Determine what pkg a file belongs to x Use basic regex Use basic regex Use basic regex X Exclude file Use extended regex Use extended regex Use extended regex y Y Default -> yes z Use gzip compression Z -------------- next part -------------- From sobomax at FreeBSD.org Thu Jun 5 06:47:29 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Thu Jun 5 06:47:36 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <1212614663.15220.136.camel@localhost> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> <4846F520.6040400@FreeBSD.org> <1212614663.15220.136.camel@localhost> Message-ID: <48478BF6.6030309@FreeBSD.org> Coleman Kane wrote: > Using your argument above about scripts written for release X+1 not > working under release X means that the addition of "-a" to /bin/cp > should also be reverted, as well as numerous other changes that have > happened to CLI tools like ifconfig. There are probably at least Coleman, You either don't understand what I have written or trying to play trick with my words. I am not talking about adding new options for new functionality, which is good and fine, but rather about massive addition of long options as a synonyms to existing short ones, which we have seen in this case. That's why I largely see this change as a superfluous and potentially harmful. -Maxim From kib at FreeBSD.org Thu Jun 5 09:15:57 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Thu Jun 5 09:15:59 2008 Subject: cvs commit: src/sys/fs/devfs devfs_vnops.c Message-ID: <200806050915.m559FuGG078928@repoman.freebsd.org> kib 2008-06-05 09:15:56 UTC FreeBSD src repository Modified files: sys/fs/devfs devfs_vnops.c Log: SVN rev 179554 on 2008-06-05 09:15:47Z by kib When devfs_allocv() committed to create new vnode, since de_vnode is NULL, the dm_lock is held while the newly allocated vnode is locked. Since no other threads may try to lock the new vnode yet, the LOR there cannot result in the deadlock. Shut down the witness warning to note this fact. Tested by: pho Prodded by: attilio Revision Changes Path 1.162 +1 -1 src/sys/fs/devfs/devfs_vnops.c From philip at FreeBSD.org Thu Jun 5 10:50:29 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Thu Jun 5 10:50:31 2008 Subject: cvs commit: src/share/syscons/keymaps colemak.iso15.acc.kbd Message-ID: <200806051050.m55AoSCK089268@repoman.freebsd.org> philip 2008-06-05 10:50:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/syscons/keymaps colemak.iso15.acc.kbd Log: SVN rev 179555 on 2008-06-05 10:50:22Z by philip MFC: Recode the colemak keymap to ISO-8859. Revision Changes Path 1.1.2.2 +28 -28 src/share/syscons/keymaps/colemak.iso15.acc.kbd From philip at FreeBSD.org Thu Jun 5 10:53:14 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Thu Jun 5 10:53:17 2008 Subject: cvs commit: src/share/syscons/keymaps colemak.iso15.acc.kbd Message-ID: <200806051053.m55ArElu089471@repoman.freebsd.org> philip 2008-06-05 10:53:14 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/syscons/keymaps colemak.iso15.acc.kbd Log: SVN rev 179556 on 2008-06-05 10:53:01Z by philip MFC: Recode the colemak keymap to ISO-8859. Revision Changes Path 1.1.4.2 +28 -28 src/share/syscons/keymaps/colemak.iso15.acc.kbd From cokane at FreeBSD.org Thu Jun 5 12:17:27 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Thu Jun 5 12:17:57 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080604222931.GA50114@troutmask.apl.washington.edu> References: <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> <4846F520.6040400@FreeBSD.org> <1212614663.15220.136.camel@localhost> <20080604222931.GA50114@troutmask.apl.washington.edu> Message-ID: <1212668141.1855.90.camel@localhost> On Wed, 2008-06-04 at 15:29 -0700, Steve Kargl wrote: > On Wed, Jun 04, 2008 at 05:24:23PM -0400, Coleman Kane wrote: > > It's probably premature to begin shooting that down. I merely picked > > "cp" as the first tool that came to mind which lives in /bin. Maybe I > > should have picked on /bin/pax or /usr/bin/lex instead, or something > > else with 20+ getopt args :P. > > > > My point was that it was pretty absurd to start treating this like it > > will turn into a giant mad party of getopt_long conversions. I don't > > really have the inclination to go out and add long options to any tool > > myself, because I have better things to do with my time. As I said > > earlier, I think it's a better policy to discuss this when the > > patch/commit comes along, and see what people's feelings are on the > > matter then. I don't think it is anybody's place at the moment (except > > maybe core@) to try and deter a committer from making what some feel is > > a beneficial change to software. > > > > Please commit. The patch gives conformance to Posix > (ie., IEEE Std 1003.1, 2004 Edition). > > As a bonus, I've included long options for compatibility > with GNU wc. > > Index: wc.1 > =================================================================== > RCS file: /home/ncvs/src/usr.bin/wc/wc.1,v > retrieving revision 1.25 > diff -u -p -r1.25 wc.1 > --- wc.1 21 Dec 2006 22:59:07 -0000 1.25 > +++ wc.1 4 Jun 2008 22:27:15 -0000 > @@ -71,25 +71,19 @@ the last file. > .Pp > The following options are available: > .Bl -tag -width indent > -.It Fl c > +.It Fl c , -bytes > The number of bytes in each input file > is written to the standard output. > -This will cancel out any prior usage of the > -.Fl m > -option. > -.It Fl l > +.It Fl l , -lines > The number of lines in each input file > is written to the standard output. > -.It Fl m > +.It Fl m , -chars > The number of characters in each input file is written to the standard output. > If the current locale does not support multibyte characters, this > is equivalent to the > .Fl c > option. > -This will cancel out any prior usage of the > -.Fl c > -option. > -.It Fl w > +.It Fl w , -words > The number of words in each input file > is written to the standard output. > .El > Index: wc.c > =================================================================== > RCS file: /home/ncvs/src/usr.bin/wc/wc.c,v > retrieving revision 1.21 > diff -u -p -r1.21 wc.c > --- wc.c 27 Dec 2004 22:27:56 -0000 1.21 > +++ wc.c 4 Jun 2008 22:27:15 -0000 > @@ -53,6 +53,7 @@ __FBSDID("$FreeBSD: src/usr.bin/wc/wc.c, > #include > #include > #include > +#include > #include > #include > #include > @@ -68,6 +69,15 @@ int doline, doword, dochar, domulti; > static int cnt(const char *); > static void usage(void); > > +static char opts[] = "clmw"; > +static struct option longopts[] = { > + { "bytes", no_argument, NULL, 'c' }, > + { "lines", no_argument, NULL, 'l' }, > + { "chars", no_argument, NULL, 'm' }, > + { "words", no_argument, NULL, 'w' }, > + { NULL, 0, NULL, 0 }, > +}; > + > int > main(int argc, char *argv[]) > { > @@ -75,7 +85,7 @@ main(int argc, char *argv[]) > > (void) setlocale(LC_CTYPE, ""); > > - while ((ch = getopt(argc, argv, "clmw")) != -1) > + while ((ch = getopt_long(argc, argv, opts, longopts, NULL)) != -1) > switch((char)ch) { > case 'l': > doline = 1; > @@ -85,11 +95,13 @@ main(int argc, char *argv[]) > break; > case 'c': > dochar = 1; > - domulti = 0; > + if (domulti) > + errx(1, "-c conflicts with -m"); > break; > case 'm': > domulti = 1; > - dochar = 0; > + if (dochar) > + errx(1, "-m conflicts with -c"); > break; > case '?': > default: > @@ -263,6 +275,6 @@ word: gotsp = 1; > static void > usage() > { > - (void)fprintf(stderr, "usage: wc [-clmw] [file ...]\n"); > + (void)fprintf(stderr, "usage: wc [-c|-m][-lw] [file ...]\n"); > exit(1); > } > First of all, your patch removes functionality in our wc that deviates from POSIX in that the last -c or -m option takes precedence over the priors. This functionality enables there to be something like a "wc -c" alias to "wc", and then if the user executes "wc -m" they will still get what they want. Yes, the POSIX manual says that you cannot specify these two options together. No, *I* am not going to "fix" our behavior, because I don't care about that particular case. If someone else does (with a commit bit who is willing to take the fall), let them speak up. I also don't personally care about adding long options to all of the core tools, especially the ones that only have 4 or 5 options. Second (as is obvious), you are merely posting this out of spite, and not because you actually have any interest in improving the tools through adding strict POSIX conformance and long options. Don't send me any more patches. Surely you have something better to do with your time. -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080605/4f1f8602/attachment.pgp From cokane at FreeBSD.org Thu Jun 5 12:31:36 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Thu Jun 5 12:31:43 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <48478BF6.6030309@FreeBSD.org> References: <200805301426.m4UEQ92d025434@repoman.freebsd.org> <48405C4B.3050603@FreeBSD.org> <1212179252.1967.1.camel@localhost> <20080604041815.GA84027@FreeBSD.org> <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> <4846F520.6040400@FreeBSD.org> <1212614663.15220.136.camel@localhost> <48478BF6.6030309@FreeBSD.org> Message-ID: <1212667313.1855.81.camel@localhost> On Wed, 2008-06-04 at 23:47 -0700, Maxim Sobolev wrote: > Coleman Kane wrote: > > Using your argument above about scripts written for release X+1 not > > working under release X means that the addition of "-a" to /bin/cp > > should also be reverted, as well as numerous other changes that have > > happened to CLI tools like ifconfig. There are probably at least > > Coleman, > > You either don't understand what I have written or trying to play trick > with my words. I am not talking about adding new options for new > functionality, which is good and fine, but rather about massive addition > of long options as a synonyms to existing short ones, which we have seen > in this case. That's why I largely see this change as a superfluous and > potentially harmful. > > -Maxim > The cp "-a" option is a synonym (for -R -p -P), which many consider to be superfluous, and is why our cp lasted so long without it. The option was added to GNU file-utils cp, and is also used by rsync and probably other tools that I am forgetting. So, no new functionality. It doesn't bother me personally, I just wanted to point it out. -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080605/e14152cc/attachment.pgp From danfe at FreeBSD.org Thu Jun 5 13:05:13 2008 From: danfe at FreeBSD.org (Alexey Dokuchaev) Date: Thu Jun 5 13:05:19 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <1212668141.1855.90.camel@localhost> References: <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> <4846F520.6040400@FreeBSD.org> <1212614663.15220.136.camel@localhost> <20080604222931.GA50114@troutmask.apl.washington.edu> <1212668141.1855.90.camel@localhost> Message-ID: <20080605130513.GA44988@FreeBSD.org> On Thu, Jun 05, 2008 at 08:15:41AM -0400, Coleman Kane wrote: > On Wed, 2008-06-04 at 15:29 -0700, Steve Kargl wrote: > > Please commit. The patch gives conformance to Posix > > (ie., IEEE Std 1003.1, 2004 Edition). > > > > As a bonus, I've included long options for compatibility > > with GNU wc. > > > > First of all, your patch removes functionality in our wc that deviates > from POSIX in that the last -c or -m option takes precedence over the > priors. This functionality enables there to be something like a "wc -c" > alias to "wc", and then if the user executes "wc -m" they will still get > what they want. > > Yes, the POSIX manual says that you cannot specify these two options > together. No, *I* am not going to "fix" our behavior, because I don't > care about that particular case. If someone else does (with a commit bit > who is willing to take the fall), let them speak up. I also don't > personally care about adding long options to all of the core tools, > especially the ones that only have 4 or 5 options. > > Second (as is obvious), you are merely posting this out of spite, and > not because you actually have any interest in improving the tools > through adding strict POSIX conformance and long options. Don't send me > any more patches. Geez, I guess this was just a mere joke. ./danfe From cokane at FreeBSD.org Thu Jun 5 13:36:09 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Thu Jun 5 13:36:15 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <20080605130513.GA44988@FreeBSD.org> References: <20080604043955.GA38627@troutmask.apl.washington.edu> <20080604063631.GA28351@freebie.xs4all.nl> <20080604150013.GA44358@troutmask.apl.washington.edu> <20080604191339.GA31570@freebie.xs4all.nl> <20080604192955.GA46284@troutmask.apl.washington.edu> <1212608575.15220.109.camel@localhost> <4846F520.6040400@FreeBSD.org> <1212614663.15220.136.camel@localhost> <20080604222931.GA50114@troutmask.apl.washington.edu> <1212668141.1855.90.camel@localhost> <20080605130513.GA44988@FreeBSD.org> Message-ID: <1212672868.1855.95.camel@localhost> On Thu, 2008-06-05 at 13:05 +0000, Alexey Dokuchaev wrote: > On Thu, Jun 05, 2008 at 08:15:41AM -0400, Coleman Kane wrote: > > On Wed, 2008-06-04 at 15:29 -0700, Steve Kargl wrote: > > > Please commit. The patch gives conformance to Posix > > > (ie., IEEE Std 1003.1, 2004 Edition). > > > > > > As a bonus, I've included long options for compatibility > > > with GNU wc. > > > > > > > First of all, your patch removes functionality in our wc that deviates > > from POSIX in that the last -c or -m option takes precedence over the > > priors. This functionality enables there to be something like a "wc -c" > > alias to "wc", and then if the user executes "wc -m" they will still get > > what they want. > > > > Yes, the POSIX manual says that you cannot specify these two options > > together. No, *I* am not going to "fix" our behavior, because I don't > > care about that particular case. If someone else does (with a commit bit > > who is willing to take the fall), let them speak up. I also don't > > personally care about adding long options to all of the core tools, > > especially the ones that only have 4 or 5 options. > > > > Second (as is obvious), you are merely posting this out of spite, and > > not because you actually have any interest in improving the tools > > through adding strict POSIX conformance and long options. Don't send me > > any more patches. > > Geez, I guess this was just a mere joke. > > ./danfe > Sorry, sense of humor kicks in with my morning coffee. I praise the attention to detail by supplying the man-page patch along too. -- Coleman Kane -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 195 bytes Desc: This is a digitally signed message part Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080605/75967e22/attachment.pgp From rafan at FreeBSD.org Thu Jun 5 13:37:48 2008 From: rafan at FreeBSD.org (Rong-En Fan) Date: Thu Jun 5 13:37:55 2008 Subject: cvs commit: src/release/doc/en_US.ISO8859-1/relnotes article.sgml Message-ID: <200806051337.m55DbmOh019926@repoman.freebsd.org> rafan 2008-06-05 13:37:47 UTC FreeBSD src repository Modified files: release/doc/en_US.ISO8859-1/relnotes article.sgml Log: SVN rev 179557 on 2008-06-05 13:37:40Z by rafan - ncurses has been updated to 5.6-20080503 Revision Changes Path 1.1090 +1 -1 src/release/doc/en_US.ISO8859-1/relnotes/article.sgml From jhb at FreeBSD.org Thu Jun 5 14:44:07 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 5 14:44:11 2008 Subject: cvs commit: src/sys/dev/ie if_ie.c Message-ID: <200806051444.m55Ei4pd026480@repoman.freebsd.org> jhb 2008-06-05 14:44:04 UTC FreeBSD src repository Modified files: sys/dev/ie if_ie.c Log: SVN rev 179558 on 2008-06-05 14:43:55Z by jhb - Fix two calls to ieinit() in ieioctl() to call ieinit_locked() instead to avoid recursing on the lock. - Use IFQ_SET_MAXLEN(). Revision Changes Path 1.110 +3 -3 src/sys/dev/ie/if_ie.c From jhb at FreeBSD.org Thu Jun 5 14:45:51 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 5 14:45:57 2008 Subject: cvs commit: src/sys/dev/ie if_ie_isa.c Message-ID: <200806051445.m55EjoAF026721@repoman.freebsd.org> jhb 2008-06-05 14:45:50 UTC FreeBSD src repository Modified files: sys/dev/ie if_ie_isa.c Log: SVN rev 179559 on 2008-06-05 14:45:32Z by jhb Rename the ie_EE16 module to just 'ie' so that at least one of the modules for this driver is called 'ie'. Otherwise, ifconfig(8) doesn't recognize any of the modules as being the ie(4) driver and will always try to kldload the driver even when it is already present in the kernel. Reported by: Thierry Herbelot Revision Changes Path 1.9 +7 -1 src/sys/dev/ie/if_ie_isa.c From jhb at FreeBSD.org Thu Jun 5 14:49:51 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 5 14:50:04 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c if_cs_isa.c if_cs_pccard.c if_csvar.h Message-ID: <200806051449.m55EnpCF027042@repoman.freebsd.org> jhb 2008-06-05 14:49:51 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c if_cs_isa.c if_cs_pccard.c if_csvar.h Log: SVN rev 179560 on 2008-06-05 14:49:35Z by jhb Make the cs(4) driver MPSAFE: - Add a mutex to the softc to protect the softc and the device hardware. - Add a private timer to manage transmit watchdogs rather than using if_timer/if_watchdog. - Setup the interrupt handler after ether_ifattach(). Tested by: imp Revision Changes Path 1.49 +112 -56 src/sys/dev/cs/if_cs.c 1.8 +0 -8 src/sys/dev/cs/if_cs_isa.c 1.20 +0 -4 src/sys/dev/cs/if_cs_pccard.c 1.10 +7 -0 src/sys/dev/cs/if_csvar.h From des at des.no Thu Jun 5 17:56:51 2008 From: des at des.no (=?utf-8?Q?Dag-Erling_Sm=C3=B8rgrav?=) Date: Thu Jun 5 17:56:58 2008 Subject: cvs commit: src/share/man/man3 Makefile queue.3 src/sys/sys queue.h In-Reply-To: <4835AA65.6070708@FreeBSD.org> (Maxim Sobolev's message of "Thu\, 22 May 2008 10\:16\:21 -0700") References: <200805221440.m4MEe32W039559@repoman.freebsd.org> <4835A619.3010502@FreeBSD.org> <4835A71E.3030703@freebsd.org> <4835AA65.6070708@FreeBSD.org> Message-ID: <86y75jdbjz.fsf@ds4.des.no> Maxim Sobolev writes: > This is local use convenience vs. easy of porting tradeoff. On one > hand API is used heavily in the kernel on the other I don't think we > have lot of kernel code being ported from OpenBSD nowadays. So that > local use convinience should win a big. a bit late, but - is also used (quite extensively) in userland. DES -- Dag-Erling Sm?rgrav - des@des.no From peter at FreeBSD.org Thu Jun 5 18:05:51 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Thu Jun 5 18:05:53 2008 Subject: cvs commit: CVSROOT commit_prep.pl Message-ID: <200806051805.m55I5pXO056576@repoman.freebsd.org> peter 2008-06-05 18:05:50 UTC FreeBSD src repository Modified files: . commit_prep.pl Log: Turn off $keyword$ checks. Leave the $keyword: foo $ collapse code active. Revision Changes Path 1.67 +2 -2 CVSROOT/commit_prep.pl From obrien at FreeBSD.org Thu Jun 5 18:06:05 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Thu Jun 5 18:06:11 2008 Subject: cvs commit: src/contrib/cvs/src classify.c Message-ID: <200806051806.m55I65TJ056634@repoman.freebsd.org> obrien 2008-06-05 18:06:05 UTC FreeBSD src repository Modified files: contrib/cvs/src classify.c Log: SVN rev 179561 on 2008-06-05 15:09:53Z by obrien Optimize the CVS bug #17168 fix by relaxing the conditions for update. Revision Changes Path 1.2 +211 -234 src/contrib/cvs/src/classify.c From marcel at FreeBSD.org Thu Jun 5 18:06:06 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 5 18:06:22 2008 Subject: cvs commit: src/sbin/geom Makefile Message-ID: <200806051806.m55I66oh056665@repoman.freebsd.org> marcel 2008-06-05 18:06:06 UTC FreeBSD src repository Modified files: sbin/geom Makefile Log: SVN rev 179562 on 2008-06-05 15:24:10Z by marcel Allow building a static geom(8) for release related crunched binaries. In particular, this allows geom to be added to the boot_crunch binary on ia64. Revision Changes Path 1.5 +1 -1 src/sbin/geom/Makefile From marcel at FreeBSD.org Thu Jun 5 18:06:07 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 5 18:06:22 2008 Subject: cvs commit: src/release/ia64 boot_crunch.conf Message-ID: <200806051806.m55I678w056692@repoman.freebsd.org> marcel 2008-06-05 18:06:07 UTC FreeBSD src repository Modified files: release/ia64 boot_crunch.conf Log: SVN rev 179563 on 2008-06-05 15:32:05Z by marcel Obsolete gpt(8) by replacing it with geom(8). Add the glabel and gpart symlinks for easy access. Revision Changes Path 1.14 +4 -2 src/release/ia64/boot_crunch.conf From attilio at FreeBSD.org Thu Jun 5 18:06:18 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Thu Jun 5 18:06:33 2008 Subject: cvs commit: src/sys/dev/usb ubsa.c usbdevs Message-ID: <200806051806.m55I68CF056722@repoman.freebsd.org> attilio 2008-06-05 18:06:08 UTC FreeBSD src repository Modified files: sys/dev/usb ubsa.c usbdevs Log: SVN rev 179564 on 2008-06-05 16:56:56Z by attilio Add the support for the Globetrotter Max 3.6 HSDPA Modem. PR: usb/118374 Submitted by: Greg Rivers Revision Changes Path 1.38 +2 -0 src/sys/dev/usb/ubsa.c 1.359 +1 -0 src/sys/dev/usb/usbdevs From attilio at FreeBSD.org Thu Jun 5 18:06:19 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Thu Jun 5 18:06:44 2008 Subject: cvs commit: src/share/man/man9 VOP_LOCK.9 Message-ID: <200806051806.m55I6Jsv056758@repoman.freebsd.org> attilio 2008-06-05 18:06:19 UTC FreeBSD src repository Modified files: share/man/man9 VOP_LOCK.9 Log: SVN rev 179565 on 2008-06-05 17:07:33Z by attilio Add the missing option LK_NOWITNESS for vn_lock / VOP_LOCK operations. Revision Changes Path 1.24 +4 -0 src/share/man/man9/VOP_LOCK.9 From brooks at FreeBSD.org Thu Jun 5 18:06:21 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Thu Jun 5 18:06:45 2008 Subject: cvs commit: src/etc/rc.d Makefile NETWORKING defaultroute routing Message-ID: <200806051806.m55I6Kh9056796@repoman.freebsd.org> brooks 2008-06-05 18:06:20 UTC FreeBSD src repository Modified files: etc/rc.d Makefile NETWORKING routing Added files: etc/rc.d defaultroute Log: SVN rev 179566 on 2008-06-05 17:26:47Z by brooks Fix the wait for default route change I made a few weeks ago by creating a new defaultroute script that just does the wait. The previous attempt created a circular dependency through network_ipv6. Pointy hat to: brooks Revision Changes Path 1.92 +1 -1 src/etc/rc.d/Makefile 1.16 +1 -1 src/etc/rc.d/NETWORKING 1.1 +48 -0 src/etc/rc.d/defaultroute (new) 1.145 +1 -26 src/etc/rc.d/routing From ed at FreeBSD.org Thu Jun 5 18:06:24 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Thu Jun 5 18:06:50 2008 Subject: cvs commit: src/sys/sys ttydefaults.h Message-ID: <200806051806.m55I6LMF056823@repoman.freebsd.org> ed 2008-06-05 18:06:21 UTC FreeBSD src repository Modified files: sys/sys ttydefaults.h Log: SVN rev 179567 on 2008-06-05 17:44:18Z by ed Fix faulty character to control-character conversion for CTRL(). The CTRL() macro seems to perform character to control-character conversion (i.e. 'A' to 0x01) to lowercase characters. This is actually not valid. If we use lowercase characters, conversions such as CTRL('\\') and CTRL('?') will result to invalid conversions. Because we must still support old source code that uses CTRL() (bad!), we make CTRL() accept both forms. When the character is a lowercase character, we perform the old style conversion. Approved by: philip (mentor) Revision Changes Path 1.20 +23 -16 src/sys/sys/ttydefaults.h From ed at FreeBSD.org Thu Jun 5 18:06:24 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Thu Jun 5 18:06:52 2008 Subject: cvs commit: src/usr.sbin/ppp physical.c Message-ID: <200806051806.m55I6Mdd056850@repoman.freebsd.org> ed 2008-06-05 18:06:22 UTC FreeBSD src repository Modified files: usr.sbin/ppp physical.c Log: SVN rev 179568 on 2008-06-05 17:46:32Z by ed Make ppp use , not . ppp's physical.c is filled with calls to termios. For some reason, it includes , not . Even though this works with the current version of FreeBSD, we'd better follow the standards. Approved by: philip (mentor) Revision Changes Path 1.61 +1 -1 src/usr.sbin/ppp/physical.c From peter at FreeBSD.org Thu Jun 5 18:16:11 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Thu Jun 5 18:16:12 2008 Subject: cvs commit: CVSROOT log_accum.pl Message-ID: <200806051816.m55IGAXU057822@repoman.freebsd.org> peter 2008-06-05 18:16:10 UTC FreeBSD src repository Modified files: . log_accum.pl Log: Use forged timestamp for first line of commit messages. Revision Changes Path 1.130 +9 -1 CVSROOT/log_accum.pl From mlaier at FreeBSD.org Thu Jun 5 19:30:25 2008 From: mlaier at FreeBSD.org (Max Laier) Date: Thu Jun 5 19:30:27 2008 Subject: cvs commit: src/sys/contrib/pf/net pf_ioctl.c Message-ID: <200806051930.m55JUOGe065268@repoman.freebsd.org> mlaier 2008-06-05 19:30:20 UTC FreeBSD src repository Modified files: sys/contrib/pf/net pf_ioctl.c Log: SVN rev 179570 on 2008-06-05 19:30:20Z by mlaier Fix range check for rtable id. Revision Changes Path 1.33 +2 -2 src/sys/contrib/pf/net/pf_ioctl.c From peter at FreeBSD.org Thu Jun 5 19:54:20 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Thu Jun 5 19:54:25 2008 Subject: cvs commit: src LOCKS Message-ID: <200806051954.m55JsJJu067591@repoman.freebsd.org> peter 2008-06-05 19:47:58 UTC FreeBSD src repository Modified files: . LOCKS Log: SVN rev 179574 on 2008-06-05 19:47:58Z by peter Update LOCKS syntax. Approved by: core (implicit) Revision Changes Path 1.5 +4 -4 src/LOCKS From brooks at FreeBSD.org Fri Jun 6 00:06:45 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Fri Jun 6 00:06:47 2008 Subject: cvs commit: src/usr.bin/cksum cksum.c Message-ID: <200806060006.m5606j2S005540@repoman.freebsd.org> brooks 2008-06-05 22:35:07 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/cksum cksum.c Log: SVN rev 179583 on 2008-06-05 22:35:07Z by brooks MFC r179024: Change a use of u_int32_t to uint32_t. PR: bin/93172 Revision Changes Path 1.17.22.1 +1 -1 src/usr.bin/cksum/cksum.c From brooks at FreeBSD.org Fri Jun 6 00:06:47 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Fri Jun 6 00:06:51 2008 Subject: cvs commit: src/usr.bin/cksum cksum.c Message-ID: <200806060006.m5606klx005570@repoman.freebsd.org> brooks 2008-06-05 22:35:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/cksum cksum.c Log: SVN rev 179584 on 2008-06-05 22:35:58Z by brooks MFC r179024: Change a use of u_int32_t to uint32_t. PR: bin/93172 Revision Changes Path 1.17.12.1 +1 -1 src/usr.bin/cksum/cksum.c From peter at FreeBSD.org Fri Jun 6 00:06:48 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Fri Jun 6 00:06:52 2008 Subject: cvs commit: src/usr.sbin/config Makefile Message-ID: <200806060006.m5606lCu005597@repoman.freebsd.org> peter 2008-06-06 00:01:22 UTC FreeBSD src repository Modified files: usr.sbin/config Makefile Log: SVN rev 179587 on 2008-06-06 00:01:22Z by peter Remove versioned property cvs2svn:cvs-rev (test-commit) Revision Changes Path 1.39 +0 -0 src/usr.sbin/config/Makefile From peter at FreeBSD.org Fri Jun 6 03:22:08 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Fri Jun 6 03:22:10 2008 Subject: cvs commit: src/sys/dev/si si.c si.h Message-ID: <200806060322.m563M7uO034335@repoman.freebsd.org> peter 2008-06-06 03:21:59 UTC FreeBSD src repository Modified files: sys/dev/si si.c si.h Log: SVN rev 179589 on 2008-06-06 03:21:59Z by peter Checkpoint what I've been running for the last year. Tidy up a bunch of loose ends that "can't happen" any more, if they ever could. Revision Changes Path 1.140 +81 -70 src/sys/dev/si/si.c 1.21 +1 -3 src/sys/dev/si/si.h From peter at FreeBSD.org Fri Jun 6 03:23:40 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Fri Jun 6 03:23:43 2008 Subject: cvs commit: src/usr.sbin/sicontrol Makefile sicontrol.c Message-ID: <200806060323.m563NcYr034454@repoman.freebsd.org> peter 2008-06-06 03:23:34 UTC FreeBSD src repository Modified files: usr.sbin/sicontrol Makefile sicontrol.c Log: SVN rev 179590 on 2008-06-06 03:23:34Z by peter Add more card state dump routines. Decode more bitfields etc. Revision Changes Path 1.8 +2 -1 src/usr.sbin/sicontrol/Makefile 1.17 +253 -37 src/usr.sbin/sicontrol/sicontrol.c From imp at FreeBSD.org Fri Jun 6 04:56:41 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 6 04:56:48 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c if_cs_isa.c if_cs_pccard.c if_csreg.h if_csvar.h Message-ID: <200806060456.m564uetG052182@repoman.freebsd.org> imp 2008-06-06 04:56:27 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c if_cs_isa.c if_cs_pccard.c if_csreg.h if_csvar.h Log: SVN rev 179591 on 2008-06-06 04:56:27Z by imp Remove unused fields in softc. If they are ever really needed again, they can re-added. Remove CS_NAME. Don't whine when there's an ignored checksum error: User has said STFU, so we should S the FU. (remove mandated properties). Revision Changes Path 1.50 +3 -17 src/sys/dev/cs/if_cs.c 1.9 +0 -3 src/sys/dev/cs/if_cs_isa.c 1.21 +0 -0 src/sys/dev/cs/if_cs_pccard.c 1.7 +0 -3 src/sys/dev/cs/if_csreg.h 1.11 +0 -4 src/sys/dev/cs/if_csvar.h From benno at FreeBSD.org Fri Jun 6 05:02:59 2008 From: benno at FreeBSD.org (Benno Rice) Date: Fri Jun 6 05:03:03 2008 Subject: cvs commit: src/sys/conf files src/sys/dev/mii miidevs smcphy.c src/sys/dev/smc if_smc.c if_smcreg.h if_smcvar.h Message-ID: <200806060502.m5652xMl052760@repoman.freebsd.org> benno 2008-06-06 05:00:49 UTC FreeBSD src repository Modified files: sys/conf files sys/dev/mii miidevs Added files: sys/dev/mii smcphy.c sys/dev/smc if_smc.c if_smcreg.h if_smcvar.h Log: SVN rev 179592 on 2008-06-06 05:00:49Z by benno This is a rewritten driver for the SMSC LAN91C111. It's based in part on the sn(4) driver and also looking at newer drivers. The reason for the rewrite is to support MII and to try and resolve some performance issues found when trying to use the sn(4) driver on the Gumstix network boards. For reference, the SMSC LAN91C111 is a non-PCI ethernet part whose lineage dates back to Ye Olde Days of ISA. It seems to get some use in the embedded space these days on parts lacking on-board MACs or on-board PCI controllers, such as the XScale PXA line of ARM CPUs. This also includes a driver for the SMSC LAN83C183 10/100 PHY. Man page to follow. Revision Changes Path 1.1307 +2 -0 src/sys/conf/files 1.57 +4 -0 src/sys/dev/mii/miidevs 1.1 +265 -0 src/sys/dev/mii/smcphy.c (new) 1.1 +1312 -0 src/sys/dev/smc/if_smc.c (new) 1.1 +261 -0 src/sys/dev/smc/if_smcreg.h (new) 1.1 +87 -0 src/sys/dev/smc/if_smcvar.h (new) From imp at FreeBSD.org Fri Jun 6 05:03:32 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 6 05:03:35 2008 Subject: cvs commit: src/sys/dev/cs if_cs_isa.c Message-ID: <200806060503.m5653VrJ053927@repoman.freebsd.org> imp 2008-06-06 05:02:01 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs_isa.c Log: SVN rev 179593 on 2008-06-06 05:02:01Z by imp cs has detach, remove bogus ifdef. Remove dedundant initialization of error to 0. Revision Changes Path 1.10 +1 -3 src/sys/dev/cs/if_cs_isa.c From imp at FreeBSD.org Fri Jun 6 05:03:32 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 6 05:03:36 2008 Subject: cvs commit: src/sys/dev/cs if_cs_pccard.c Message-ID: <200806060503.m5653W13053954@repoman.freebsd.org> imp 2008-06-06 05:02:36 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs_pccard.c Log: SVN rev 179594 on 2008-06-06 05:02:36Z by imp Simplify error checking when reading the function.... Revision Changes Path 1.22 +1 -3 src/sys/dev/cs/if_cs_pccard.c From benno at FreeBSD.org Fri Jun 6 05:09:12 2008 From: benno at FreeBSD.org (Benno Rice) Date: Fri Jun 6 05:09:14 2008 Subject: cvs commit: src/sys/arm/conf GUMSTIX GUMSTIX.hints src/sys/arm/include intr.h src/sys/arm/xscale/pxa files.pxa if_smc_smi.c pxa_gpio.c pxa_icu.c pxa_machdep.c pxa_obio.c pxa_smi.c pxa_space.c pxa_timer.c pxareg.h pxavar.h std.pxa ... Message-ID: <200806060509.m5659Cgk054736@repoman.freebsd.org> benno 2008-06-06 05:08:09 UTC FreeBSD src repository Modified files: sys/arm/include intr.h sys/conf options.arm Added files: sys/arm/conf GUMSTIX GUMSTIX.hints sys/arm/xscale/pxa files.pxa if_smc_smi.c pxa_gpio.c pxa_icu.c pxa_machdep.c pxa_obio.c pxa_smi.c pxa_space.c pxa_timer.c pxareg.h pxavar.h std.pxa uart_bus_pxa.c uart_cpu_pxa.c Log: SVN rev 179595 on 2008-06-06 05:08:09Z by benno Support for the XScale PXA255 SoC as found on the Gumstix Basix and Connex boards. This is enough to net-boot to multiuser. Also supported is the SMSC LAN91C111 parts used on the netCF, netDUO and netMMC add-on boards. I'll be putting some instructions on how to boot this on the Gumstix boards online soon. This is still fairly rough and will be refined over time but I felt it was better to get this out there where other people can help out. Revision Changes Path 1.1 +93 -0 src/sys/arm/conf/GUMSTIX (new) 1.1 +19 -0 src/sys/arm/conf/GUMSTIX.hints (new) 1.9 +3 -0 src/sys/arm/include/intr.h 1.1 +17 -0 src/sys/arm/xscale/pxa/files.pxa (new) 1.1 +124 -0 src/sys/arm/xscale/pxa/if_smc_smi.c (new) 1.1 +358 -0 src/sys/arm/xscale/pxa/pxa_gpio.c (new) 1.1 +258 -0 src/sys/arm/xscale/pxa/pxa_icu.c (new) 1.1 +576 -0 src/sys/arm/xscale/pxa/pxa_machdep.c (new) 1.1 +388 -0 src/sys/arm/xscale/pxa/pxa_obio.c (new) 1.1 +355 -0 src/sys/arm/xscale/pxa/pxa_smi.c (new) 1.1 +329 -0 src/sys/arm/xscale/pxa/pxa_space.c (new) 1.1 +317 -0 src/sys/arm/xscale/pxa/pxa_timer.c (new) 1.1 +759 -0 src/sys/arm/xscale/pxa/pxareg.h (new) 1.1 +112 -0 src/sys/arm/xscale/pxa/pxavar.h (new) 1.1 +9 -0 src/sys/arm/xscale/pxa/std.pxa (new) 1.1 +87 -0 src/sys/arm/xscale/pxa/uart_bus_pxa.c (new) 1.1 +69 -0 src/sys/arm/xscale/pxa/uart_cpu_pxa.c (new) 1.20 +1 -0 src/sys/conf/options.arm From marcel at FreeBSD.org Fri Jun 6 05:14:31 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Fri Jun 6 05:14:33 2008 Subject: cvs commit: src/sbin/geom/class/part gpart.8 Message-ID: <200806060514.m565EVYo055209@repoman.freebsd.org> marcel 2008-06-06 05:14:16 UTC FreeBSD src repository Modified files: sbin/geom/class/part gpart.8 Log: SVN rev 179596 on 2008-06-06 05:14:16Z by marcel Update the manpage to reflect reality: o The BSD, PC98 and VTOC8 schemes are supported. o The bootcode command was added to allow installing bootstrap code into the scheme's metadata. Revision Changes Path 1.3 +32 -3 src/sbin/geom/class/part/gpart.8 From imp at FreeBSD.org Fri Jun 6 05:25:32 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 6 05:25:34 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c if_cs_isa.c if_cs_pccard.c if_csvar.h Message-ID: <200806060525.m565PVm5056157@repoman.freebsd.org> imp 2008-06-06 05:25:24 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c if_cs_isa.c if_cs_pccard.c if_csvar.h Log: SVN rev 179597 on 2008-06-06 05:25:24Z by imp Minor clean up to shave about 1.5k off the size of the driver: o remove unused fields from softc and args from cs_alloc_irq o remove some commented code that will never be implemented. o Don't try to send a packet and see if it worked. We don't need this anymore, and it doesn't add any value. o tweaks for BNC and AUI. o limit possible time hung in the kernel to 4s rather than 40s. Revision Changes Path 1.51 +6 -84 src/sys/dev/cs/if_cs.c 1.11 +5 -7 src/sys/dev/cs/if_cs_isa.c 1.23 +0 -1 src/sys/dev/cs/if_cs_pccard.c 1.12 +1 -3 src/sys/dev/cs/if_csvar.h From keramida at FreeBSD.org Fri Jun 6 07:17:26 2008 From: keramida at FreeBSD.org (Giorgos Keramidas) Date: Fri Jun 6 07:17:27 2008 Subject: cvs commit: src/etc rc.firewall Message-ID: <200806060717.m567HQkZ067471@repoman.freebsd.org> keramida 2008-06-06 07:17:04 UTC FreeBSD src repository Modified files: etc rc.firewall Log: SVN rev 179598 on 2008-06-06 07:17:04Z by keramida Tweak rc.firewall to allow incoming limited broadcast traffic, when configured to run in 'client' mode. PR: conf/15010 Submitted by: Bill Trost, trost at cloud.rain.com Reviewed by: bz MFC after: 2 weeks Revision Changes Path 1.55 +3 -0 src/etc/rc.firewall From keramida at FreeBSD.org Fri Jun 6 08:28:13 2008 From: keramida at FreeBSD.org (Giorgos Keramidas) Date: Fri Jun 6 08:28:16 2008 Subject: cvs commit: src/bin/ln ln.1 ln.c Message-ID: <200806060828.m568SDK1075838@repoman.freebsd.org> keramida 2008-06-06 08:27:59 UTC FreeBSD src repository Modified files: bin/ln ln.1 ln.c Log: SVN rev 179603 on 2008-06-06 08:27:59Z by keramida Add a -w warning flag to ln(1). When the -w option is enabled, ln(1) checks to see if the source of a symlink, i.e. the file it should point to actually exists. The default is the old ln behavior, that does not check, to avoid surprising people who may be using ln(1) in scripts or other non-interactive places. PR: bin/7265 Submitted by: Joel Ray Holveck, detlev!joelh at mail.camalott.com MFC after: 2 weeks Revision Changes Path 1.33 +10 -5 src/bin/ln/ln.1 1.37 +37 -2 src/bin/ln/ln.c From dfr at FreeBSD.org Fri Jun 6 08:54:29 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Fri Jun 6 08:54:31 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806060854.m568sTJ5087163@repoman.freebsd.org> dfr 2008-06-06 08:54:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/nlm nlm_prot_impl.c Log: SVN rev 179605 on 2008-06-06 08:54:19Z by dfr MFC: r179425 - cope better when client NLM port changes Revision Changes Path 1.4.2.4 +17 -6 src/sys/nlm/nlm_prot_impl.c From dfr at FreeBSD.org Fri Jun 6 08:55:00 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Fri Jun 6 08:55:08 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806060855.m568t0at087238@repoman.freebsd.org> dfr 2008-06-06 08:54:51 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/nlm nlm_prot_impl.c Log: SVN rev 179606 on 2008-06-06 08:54:51Z by dfr MFC: r179425 - cope better when client NLM port changes Revision Changes Path 1.6.2.2 +17 -6 src/sys/nlm/nlm_prot_impl.c From brian at FreeBSD.org Fri Jun 6 09:00:16 2008 From: brian at FreeBSD.org (Brian Somers) Date: Fri Jun 6 09:00:21 2008 Subject: cvs commit: src/sbin/natd natd.c Message-ID: <200806060900.m5690Gn4087721@repoman.freebsd.org> brian 2008-06-06 08:59:55 UTC FreeBSD src repository Modified files: sbin/natd natd.c Log: SVN rev 179607 on 2008-06-06 08:59:55Z by brian Don't abend if we get ENOMEM from sysctl(3). The data returned is sufficient MFC after: 2 weeks Revision Changes Path 1.51 +1 -1 src/sbin/natd/natd.c From dfr at FreeBSD.org Fri Jun 6 09:09:55 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Fri Jun 6 09:10:02 2008 Subject: cvs commit: src/lib/libc/sys Symbol.map fcntl.c src/lib/libc_r/uthread uthread_fcntl.c src/lib/libkse/thread thr_fcntl.c src/lib/libthr/thread thr_syscalls.c Message-ID: <200806060909.m5699tnh089622@repoman.freebsd.org> dfr 2008-06-06 09:09:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/sys Symbol.map fcntl.c lib/libc_r/uthread uthread_fcntl.c lib/libkse/thread thr_fcntl.c lib/libthr/thread thr_syscalls.c Log: SVN rev 179608 on 2008-06-06 09:09:37Z by dfr MFC: r179358,179434 - make libc's fcntl symbol weak and extend the limited forward ABI compatibility for fcntl to the thread libraries. Revision Changes Path 1.9.2.2 +1 -0 src/lib/libc/sys/Symbol.map 1.1.2.2 +3 -1 src/lib/libc/sys/fcntl.c 1.16.2.1 +3 -1 src/lib/libc_r/uthread/uthread_fcntl.c 1.22.2.1 +2 -1 src/lib/libkse/thread/thr_fcntl.c 1.16.2.1 +2 -1 src/lib/libthr/thread/thr_syscalls.c From dfr at FreeBSD.org Fri Jun 6 09:14:05 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Fri Jun 6 09:14:08 2008 Subject: cvs commit: src/lib/libc/sys fcntl.c src/lib/libc_r/uthread uthread_fcntl.c src/lib/libpthread/thread thr_fcntl.c src/lib/libthr/thread thr_syscalls.c Message-ID: <200806060914.m569E5DZ090093@repoman.freebsd.org> dfr 2008-06-06 09:13:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libc/sys fcntl.c lib/libc_r/uthread uthread_fcntl.c lib/libpthread/thread thr_fcntl.c lib/libthr/thread thr_syscalls.c Log: SVN rev 179609 on 2008-06-06 09:13:43Z by dfr MFC: r179358,179434 - make libc's fcntl symbol weak and extend the limited forward ABI compatibility for fcntl to the thread libraries. Revision Changes Path 1.1.4.2 +3 -1 src/lib/libc/sys/fcntl.c 1.15.14.1 +3 -1 src/lib/libc_r/uthread/uthread_fcntl.c 1.19.8.1 +3 -1 src/lib/libpthread/thread/thr_fcntl.c 1.9.2.3 +2 -1 src/lib/libthr/thread/thr_syscalls.c From brian at Awfulhak.org Fri Jun 6 09:17:18 2008 From: brian at Awfulhak.org (Brian Somers) Date: Fri Jun 6 09:17:25 2008 Subject: cvs commit: src/usr.sbin/config Makefile In-Reply-To: <200806060006.m5606lCu005597@repoman.freebsd.org> References: <200806060006.m5606lCu005597@repoman.freebsd.org> Message-ID: <20080606021637.588d5fb0@dev.lan.Awfulhak.org> On Fri, 6 Jun 2008 00:01:22 +0000 (UTC) Peter Wemm wrote: > peter 2008-06-06 00:01:22 UTC > > FreeBSD src repository > > Modified files: > usr.sbin/config Makefile > Log: > SVN rev 179587 on 2008-06-06 00:01:22Z by peter > > Remove versioned property cvs2svn:cvs-rev (test-commit) > > Revision Changes Path > 1.39 +0 -0 src/usr.sbin/config/Makefile Do you have any words of wisdom about "versioned properties"? I did a 'svn propdel cvs2svn:cvs-rev natd.c' after my attempted commit said: Sending natd.c Transmitting file data .svn: Commit failed (details follow): svn: Commit blocked by pre-commit hook (exit code 1) with output: Path "head/sbin/natd/natd.c" needs to have "cvs2svn:cvs-rev" removed with "svn propdel". == Pre-commit problem count: 1 The commit worked after the 'propdel'... Cheers. -- Brian Somers Don't _EVER_ lose your sense of humour ! From stas at FreeBSD.org Fri Jun 6 09:35:22 2008 From: stas at FreeBSD.org (Stanislav Sedov) Date: Fri Jun 6 09:35:25 2008 Subject: cvs commit: src/sys/arm/at91 if_ate.c Message-ID: <200806060935.m569ZLbW091883@repoman.freebsd.org> stas 2008-06-06 09:35:08 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/arm/at91 if_ate.c Log: SVN rev 179610 on 2008-06-06 09:35:08Z by stas MFC rev. 1.22. - Don't return 0xffff if PHY id isn't equal 0. Revision Changes Path 1.20.2.2 +0 -2 src/sys/arm/at91/if_ate.c From stas at FreeBSD.org Fri Jun 6 09:59:41 2008 From: stas at FreeBSD.org (Stanislav Sedov) Date: Fri Jun 6 09:59:47 2008 Subject: cvs commit: src/sys/arm/at91 kb920x_machdep.c Message-ID: <200806060959.m569xeJt093718@repoman.freebsd.org> stas 2008-06-06 09:59:34 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/arm/at91 kb920x_machdep.c Log: SVN rev 179611 on 2008-06-06 09:59:34Z by stas MFC rev. 1.28. - Fix a typo in comment. Approved by: kib Revision Changes Path 1.23.2.3 +1 -1 src/sys/arm/at91/kb920x_machdep.c From stas at FreeBSD.org Fri Jun 6 10:02:56 2008 From: stas at FreeBSD.org (Stanislav Sedov) Date: Fri Jun 6 10:03:04 2008 Subject: cvs commit: src/sys/dev/usb uplcom.c Message-ID: <200806061002.m56A2uiN094027@repoman.freebsd.org> stas 2008-06-06 10:02:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/usb uplcom.c Log: SVN rev 179612 on 2008-06-06 10:02:43Z by stas MFC rev. 1.51 - Add quirk for the TrendNet TU-S9 adapter, which uses new version of PL2303. Approved by: kib Revision Changes Path 1.50.2.1 +3 -0 src/sys/dev/usb/uplcom.c From stas at FreeBSD.org Fri Jun 6 10:12:50 2008 From: stas at FreeBSD.org (Stanislav Sedov) Date: Fri Jun 6 10:13:18 2008 Subject: cvs commit: src/sys/arm/at91 if_ate.c In-Reply-To: <200806060935.m569ZLbW091883@repoman.freebsd.org> References: <200806060935.m569ZLbW091883@repoman.freebsd.org> Message-ID: <20080606135554.ef27d033.stas@FreeBSD.org> On Fri, 6 Jun 2008 09:35:08 +0000 (UTC) Stanislav Sedov mentioned: > stas 2008-06-06 09:35:08 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7) > sys/arm/at91 if_ate.c > Log: > SVN rev 179610 on 2008-06-06 09:35:08Z by stas > > MFC rev. 1.22. > > - Don't return 0xffff if PHY id isn't equal 0. This commit was approved by kib@. It seems that svn doesn't support forced commits:-( -- Stanislav Sedov ST4096-RIPE -------------- 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/20080606/379fec3e/attachment-0001.pgp From kib at FreeBSD.org Fri Jun 6 11:20:00 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Fri Jun 6 11:20:03 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200806061119.m56BJxmq002386@repoman.freebsd.org> kib 2008-06-06 11:19:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/mount mount.8 Log: SVN rev 179613 on 2008-06-06 11:19:54Z by kib MFC rev. 179513. Add note about a reason to use mount(8) instead of mount_somefs. Revision Changes Path 1.82.2.2 +11 -0 src/sbin/mount/mount.8 From kib at FreeBSD.org Fri Jun 6 11:25:16 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Fri Jun 6 11:25:22 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200806061125.m56BPF8Q002899@repoman.freebsd.org> kib 2008-06-06 11:25:03 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/mount mount.8 Log: SVN rev 179614 on 2008-06-06 11:25:03Z by kib MFC rev. 179514 (committed by ru). Fix markup after the rev. 179513. Revision Changes Path 1.82.2.3 +6 -4 src/sbin/mount/mount.8 From kib at FreeBSD.org Fri Jun 6 12:04:47 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Fri Jun 6 12:04:51 2008 Subject: cvs commit: src/sys/kern sys_pipe.c Message-ID: <200806061204.m56C4kEP007481@repoman.freebsd.org> kib 2008-06-06 12:04:31 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern sys_pipe.c Log: SVN rev 179615 on 2008-06-06 12:04:31Z by kib MFC rev. 1.197 AKA r179242: The pipe_peer that has the knote for write shall be present. Ignore the pipe_present value for EVFILT_WRITE in filt_pipedetach(). Revision Changes Path 1.191.2.2 +1 -6 src/sys/kern/sys_pipe.c From kib at FreeBSD.org Fri Jun 6 12:18:33 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Fri Jun 6 12:18:41 2008 Subject: cvs commit: src/sys/kern sys_pipe.c src/sys/sys pipe.h Message-ID: <200806061218.m56CIXJX013760@repoman.freebsd.org> kib 2008-06-06 12:17:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern sys_pipe.c sys/sys pipe.h Log: SVN rev 179616 on 2008-06-06 12:17:28Z by kib MFC rev. 1.198 of sys/kern/sys_pipe.c, rev. 1.30 of sys/sys/pipe.h AKA r179243. Another problem caused by the knlist_cleardel() potentially dropping PIPE_MTX(). Since the pipe_present is cleared before (potentially) sleeping, the second thread may enter the pipeclose() for the reciprocal pipe end. The test at the end of the pipeclose() for the pipe_present == 0 would succeed, allowing the second thread to free the pipe memory. First threads then accesses the freed memory after being woken up. Properly track the closing state of the pipe in the pipe_present. Introduce the intermediate state that marks the pipe as mostly dismantled but might be sleeping waiting for the knote list to be cleared. Free the pipe pair memory only when both ends pass that point. Revision Changes Path 1.191.2.3 +22 -11 src/sys/kern/sys_pipe.c 1.29.10.1 +7 -0 src/sys/sys/pipe.h From kib at FreeBSD.org Fri Jun 6 12:28:49 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Fri Jun 6 12:28:57 2008 Subject: cvs commit: src/sbin/mount mount.8 Message-ID: <200806061228.m56CSnSH018650@repoman.freebsd.org> kib 2008-06-06 12:28:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/mount mount.8 Log: SVN rev 179617 on 2008-06-06 12:28:33Z by kib Put the svn:mergeinfo on the directory instead of the mount.8. Revision Changes Path 1.82.2.4 +0 -0 src/sbin/mount/mount.8 From imp at bsdimp.com Fri Jun 6 14:28:41 2008 From: imp at bsdimp.com (M. Warner Losh) Date: Fri Jun 6 14:41:23 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create main.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... In-Reply-To: <4846F30A.5070204@FreeBSD.org> References: <20080604192955.GA46284@troutmask.apl.washington.edu> <4846EF10.1020803@FreeBSD.org> <4846F30A.5070204@FreeBSD.org> Message-ID: <20080606.082706.-345494353.imp@bsdimp.com> : It's not your call either. We have style(9), which says: Thus quote he from the holy book of style, section nine, verses 127-134 : For consistency, getopt(3) should be used to parse options. Options : should be sorted in the getopt(3) call and the switch statement, : unless : parts of the switch cascade. Elements in a switch statement that : cascade : should have a FALLTHROUGH comment. Numerical arguments should be : checked : for accuracy. Code that cannot be reached should have a : NOTREACHED com- : ment. : : There is nothing about getopt_long(3) being acceptable : replacement/addition to the getopt(3). Style(9) is about documenting the normative style of the project. While we prefer people to use it, it isn't religious dogma. getopt_long() is in the spirit of style(9) which is trying to say "don't be a dumb-ass and write your own argument parsing routine." Warner From imp at FreeBSD.org Fri Jun 6 16:10:20 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 6 16:10:22 2008 Subject: cvs commit: src/sys/dev/cs if_cs_isa.c if_cs_pccard.c Message-ID: <200806061610.m56GADVv047389@repoman.freebsd.org> imp 2008-06-06 16:10:10 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs_isa.c if_cs_pccard.c Log: SVN rev 179618 on 2008-06-06 16:10:10Z by imp Forgot to commit these files too :-( Revision Changes Path 1.12 +2 -2 src/sys/dev/cs/if_cs_isa.c 1.24 +1 -1 src/sys/dev/cs/if_cs_pccard.c From obrien at FreeBSD.org Fri Jun 6 16:17:01 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Fri Jun 6 16:17:03 2008 Subject: cvs commit: src/contrib/cvs/src classify.c Message-ID: <200806061617.m56GH1SE048015@repoman.freebsd.org> obrien 2008-06-06 16:16:44 UTC FreeBSD src repository Modified files: contrib/cvs/src classify.c Log: SVN rev 179619 on 2008-06-06 16:16:44Z by obrien Optimize the CVS bug #17168 fix by relaxing the conditions for update with respect to sticky options. (r179561 was WRT sticky and branch tags) Consulted with: Mark Baushke Revision Changes Path 1.3 +1 -0 src/contrib/cvs/src/classify.c From imp at FreeBSD.org Fri Jun 6 17:22:18 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 6 17:22:21 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c Message-ID: <200806061722.m56HMIKx067342@repoman.freebsd.org> imp 2008-06-06 17:22:07 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c Log: SVN rev 179620 on 2008-06-06 17:22:07Z by imp o Remove unused cs_debug tunable. I think I added it and then nothing with it. Bad imp. Removing us dips us under 10,000 in size too. o Replace an unconditional 30ms DELAY (yes, busy wait) with a check of the SIBUSY bit in the SelfST register before accessing the eeprom. This changes the time to read the EEPROM from 2 * 20 * 30ms (1.2s) to < 20*25us (.0005s) and make the attach of the card tolerable when ethernet media is present. Include data from the datasheet about why this works. While this is a 2500x speed increase, it doesn't really matter at all once the card is probed... o set dev earlier in softc. Revision Changes Path 1.52 +22 -10 src/sys/dev/cs/if_cs.c From imp at FreeBSD.org Fri Jun 6 17:27:34 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 6 17:27:37 2008 Subject: cvs commit: src/sys/dev/cs if_cs.c Message-ID: <200806061727.m56HRYFg067763@repoman.freebsd.org> imp 2008-06-06 17:27:19 UTC FreeBSD src repository Modified files: sys/dev/cs if_cs.c Log: SVN rev 179621 on 2008-06-06 17:27:19Z by imp Tweak a comment and a constant to restore old 30ms upper bound. cs_readreg takes ~2us not ~1us to run. Revision Changes Path 1.53 +3 -3 src/sys/dev/cs/if_cs.c From jhb at FreeBSD.org Fri Jun 6 18:31:25 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 6 18:31:27 2008 Subject: cvs commit: src/sys/dev/ichsmb ichsmb.c ichsmb_pci.c ichsmb_var.h src/sys/pci amdpm.c amdsmb.c intpm.c nfsmb.c viapm.c Message-ID: <200806061831.m56IVPCC074008@repoman.freebsd.org> jhb 2008-06-06 18:29:56 UTC FreeBSD src repository Modified files: sys/dev/ichsmb ichsmb.c ichsmb_pci.c ichsmb_var.h sys/pci amdpm.c amdsmb.c intpm.c nfsmb.c viapm.c Log: SVN rev 179622 on 2008-06-06 18:29:56Z by jhb - Use bus_foo() rather than bus_space_foo() and remove bus space tag/handle from softc. - Mark interrupt handlers MPSAFE as these drivers have been locked for a while. Revision Changes Path 1.19 +62 -62 src/sys/dev/ichsmb/ichsmb.c 1.21 +1 -3 src/sys/dev/ichsmb/ichsmb_pci.c 1.5 +0 -2 src/sys/dev/ichsmb/ichsmb_var.h 1.22 +4 -8 src/sys/pci/amdpm.c 1.6 +2 -6 src/sys/pci/amdsmb.c 1.42 +2 -2 src/sys/pci/intpm.c 1.9 +2 -8 src/sys/pci/nfsmb.c 1.19 +3 -9 src/sys/pci/viapm.c From alc at FreeBSD.org Fri Jun 6 18:38:56 2008 From: alc at FreeBSD.org (Alan Cox) Date: Fri Jun 6 18:39:00 2008 Subject: cvs commit: src/sys/vm vm_page.c Message-ID: <200806061838.m56Icu2E074642@repoman.freebsd.org> alc 2008-06-06 18:38:43 UTC FreeBSD src repository Modified files: sys/vm vm_page.c Log: SVN rev 179623 on 2008-06-06 18:38:43Z by alc Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(..., MADV_FREE) work. (Moreover, I don't believe that they have ever worked as intended.) The explanation is fairly simple. Both MADV_DONTNEED and MADV_FREE perform vm_page_dontneed() on each page within the range given to madvise(). This function moves the page to the inactive queue. Specifically, if the page is clean, it is moved to the head of the inactive queue where it is first in line for processing by the page daemon. On the other hand, if it is dirty, it is placed at the tail. Let's further examine the case in which the page is clean. Recall that the page is at the head of the line for processing by the page daemon. The expectation of vm_page_dontneed()'s author was that the page would be transferred from the inactive queue to the cache queue by the page daemon. (Once the page is in the cache queue, it is, in effect, free, that is, it can be reallocated to a new vm object by vm_page_alloc() if it isn't reactivated quickly enough by a user of the old vm object.) The trouble is that nowhere in the execution of either MADV_DONTNEED or MADV_FREE is either the machine-independent reference flag (PG_REFERENCED) or the reference bit in any page table entry (PTE) mapping the page cleared. Consequently, the immediate reaction of the page daemon is to reactivate the page because it is referenced. In effect, the madvise() was for naught. The case in which the page was dirty is not too different. Instead of being laundered, the page is reactivated. Note: The essential difference between MADV_DONTNEED and MADV_FREE is that MADV_FREE clears a page's dirty field. So, MADV_FREE is always executing the clean case above. This revision changes vm_page_dontneed() to clear both the machine- independent reference flag (PG_REFERENCED) and the reference bit in all PTEs mapping the page. MFC after: 6 weeks Revision Changes Path 1.371 +7 -0 src/sys/vm/vm_page.c From jhb at FreeBSD.org Fri Jun 6 18:41:07 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 6 18:41:09 2008 Subject: cvs commit: src/sys/dev/bktr bktr_i2c.c src/sys/dev/iicbus iicsmb.c Message-ID: <200806061841.m56If6Y7074934@repoman.freebsd.org> jhb 2008-06-06 18:40:38 UTC FreeBSD src repository Modified files: sys/dev/bktr bktr_i2c.c sys/dev/iicbus iicsmb.c Log: SVN rev 179624 on 2008-06-06 18:40:38Z by jhb Explicitly lock Giant in smbus_if methods in the bktr_i2c and iicsmb drivers for now. This can be replaced with driver locks when these drivers are locked. Revision Changes Path 1.30 +26 -1 src/sys/dev/bktr/bktr_i2c.c 1.15 +41 -8 src/sys/dev/iicbus/iicsmb.c From jhb at FreeBSD.org Fri Jun 6 18:45:54 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 6 18:45:56 2008 Subject: cvs commit: src/sys/dev/smbus smb.c Message-ID: <200806061845.m56IjrsU075326@repoman.freebsd.org> jhb 2008-06-06 18:45:32 UTC FreeBSD src repository Modified files: sys/dev/smbus smb.c Log: SVN rev 179625 on 2008-06-06 18:45:32Z by jhb - Store the device_t of the smbX device in the softc. - Store the softc of the device in the 'si_drv1' of the cdev. - Lookup the softc via 'si_drv1' in cdev methods rather than using the minor number as a unit for devclass_get_softc(). - Lookup the device_t via the softc field in cdev methods rather than using the minor number as a unit for devclass_get_device(). - Add a mutex to the softc to protect 'sc_opened'. - Remove D_NEEDGIANT as all the smbus drivers are now MPSAFE and this driver is now MPSAFE. - Remove some checks for NULL softc pointers that can't happen and don't bzero the softc during attach. Revision Changes Path 1.39 +22 -37 src/sys/dev/smbus/smb.c From phk at phk.freebsd.dk Fri Jun 6 19:11:24 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Fri Jun 6 19:11:31 2008 Subject: cvs commit: src/sys/vm vm_page.c In-Reply-To: Your message of "Fri, 06 Jun 2008 18:38:43 GMT." <200806061838.m56Icu2E074642@repoman.freebsd.org> Message-ID: <1809.1212779481@critter.freebsd.dk> In message <200806061838.m56Icu2E074642@repoman.freebsd.org>, Alan Cox writes: >alc 2008-06-06 18:38:43 UTC > > FreeBSD src repository > > Modified files: > sys/vm vm_page.c > Log: > SVN rev 179623 on 2008-06-06 18:38:43Z by alc > > Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(..., MADV_FREE) > work. (Moreover, I don't believe that they have ever worked as intended.) I remember they did have a measurable but minor impact on performance in systems which paged a lot, back when they were introduced, sort of 1995-ish. But with the subsequent plunge in RAM prices they never saw much use. Thanks for fixing them. -- 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 jhb at freebsd.org Fri Jun 6 20:56:47 2008 From: jhb at freebsd.org (John Baldwin) Date: Fri Jun 6 20:56:52 2008 Subject: cvs commit: src/sys/dev/smbus smb.c In-Reply-To: <200806061845.m56IjrsU075326@repoman.freebsd.org> References: <200806061845.m56IjrsU075326@repoman.freebsd.org> Message-ID: <200806061524.57417.jhb@freebsd.org> On Friday 06 June 2008 02:45:32 pm John Baldwin wrote: > jhb 2008-06-06 18:45:32 UTC > > FreeBSD src repository > > Modified files: > sys/dev/smbus smb.c > Log: > SVN rev 179625 on 2008-06-06 18:45:32Z by jhb > > - Store the device_t of the smbX device in the softc. > - Store the softc of the device in the 'si_drv1' of the cdev. > - Lookup the softc via 'si_drv1' in cdev methods rather than using the > minor number as a unit for devclass_get_softc(). > - Lookup the device_t via the softc field in cdev methods rather than > using the minor number as a unit for devclass_get_device(). > - Add a mutex to the softc to protect 'sc_opened'. > - Remove D_NEEDGIANT as all the smbus drivers are now MPSAFE and this driver > is now MPSAFE. > - Remove some checks for NULL softc pointers that can't happen and don't > bzero the softc during attach. All the smbus(4) stuff is MPSAFE now. I do plan to look at iic(4) soonish as it is similar in design to smbus(4). Once it is locked then if_ic(4) can be locked. -- John Baldwin From peter at FreeBSD.org Fri Jun 6 21:37:13 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Fri Jun 6 21:37:16 2008 Subject: cvs commit: src MAINTAINERS Message-ID: <200806062137.m56LbD88001542@repoman.freebsd.org> peter 2008-06-06 21:32:01 UTC FreeBSD src repository Modified files: . MAINTAINERS Log: SVN rev 179626 on 2008-06-06 21:32:01Z by peter Update description text Revision Changes Path 1.150 +1 -1 src/MAINTAINERS From thompsa at FreeBSD.org Fri Jun 6 21:57:34 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Fri Jun 6 21:57:37 2008 Subject: cvs commit: src/sys/net if_bridge.c Message-ID: <200806062157.m56LvXBN003158@repoman.freebsd.org> thompsa 2008-06-06 21:57:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/net if_bridge.c Log: SVN rev 179627 on 2008-06-06 21:57:28Z by thompsa MFC r174749, r175396, r175419, r175432 - Simplify the error handling - Nuke licence clause 3 & 4 - Drop frames to any of the reserved multicast addresses - Remove duplicated code for testing local packets. Revision Changes Path 1.11.2.56 +65 -91 src/sys/net/if_bridge.c From marcel at FreeBSD.org Fri Jun 6 23:22:48 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Fri Jun 6 23:22:53 2008 Subject: cvs commit: src/sbin/geom/misc subr.c subr.h Message-ID: <200806062322.m56NMmk3077256@repoman.freebsd.org> marcel 2008-06-06 22:44:03 UTC FreeBSD src repository Modified files: sbin/geom/misc subr.c subr.h Log: SVN rev 179628 on 2008-06-06 22:44:03Z by marcel Add two support functions: o gctl_delete_param() -- intended for parameters that are consumed by geom(8) itself and which should not be passed to the kernel. o gctl_has_param() -- intended to check if optional parameters are present. Both are needed by gpart(8) to process the optional parameter for writing bootcode to a partition (as part of the bootcode verb). However, the kernel is itself not involved in this matter and the parameter needs to be removed from the request destined for the kernel. Revision Changes Path 1.8 +44 -0 src/sbin/geom/misc/subr.c 1.9 +3 -0 src/sbin/geom/misc/subr.h From alc at cs.rice.edu Fri Jun 6 23:25:46 2008 From: alc at cs.rice.edu (Alan Cox) Date: Fri Jun 6 23:25:53 2008 Subject: cvs commit: src/sys/vm vm_page.c In-Reply-To: <1809.1212779481@critter.freebsd.dk> References: <1809.1212779481@critter.freebsd.dk> Message-ID: <4849C2D0.9060704@cs.rice.edu> Poul-Henning Kamp wrote: > In message <200806061838.m56Icu2E074642@repoman.freebsd.org>, Alan Cox writes: > >> alc 2008-06-06 18:38:43 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/vm vm_page.c >> Log: >> SVN rev 179623 on 2008-06-06 18:38:43Z by alc >> >> Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(..., MADV_FREE) >> work. (Moreover, I don't believe that they have ever worked as intended.) >> > > I remember they did have a measurable but minor impact on performance > in systems which paged a lot, back when they were introduced, sort > of 1995-ish. But with the subsequent plunge in RAM prices they > never saw much use. > > After more thorough digging through CVS history, I now believe that MADV_DONTNEED worked properly between revision 1.70 (May 1996) and 1.76 (June 1996) of vm/vm_object.c. :-) It's more difficult to say for MADV_FREE. It may have been effective until 1999. Alan From phk at phk.freebsd.dk Fri Jun 6 23:27:50 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Fri Jun 6 23:27:54 2008 Subject: cvs commit: src/sys/vm vm_page.c In-Reply-To: Your message of "Fri, 06 Jun 2008 18:05:52 EST." <4849C2D0.9060704@cs.rice.edu> Message-ID: <3171.1212794867@critter.freebsd.dk> In message <4849C2D0.9060704@cs.rice.edu>, Alan Cox writes: >Poul-Henning Kamp wrote: >>> Essentially, neither madvise(..., MADV_DONTNEED) nor madvise(..., MADV_FREE) >>> work. (Moreover, I don't believe that they have ever worked as intended.) >> >> I remember they did have a measurable but minor impact on performance >> in systems which paged a lot, back when they were introduced, sort >> of 1995-ish. But with the subsequent plunge in RAM prices they >> never saw much use. > >After more thorough digging through CVS history, I now believe that >MADV_DONTNEED worked properly between revision 1.70 (May 1996) and 1.76 >(June 1996) of vm/vm_object.c. :-) It's more difficult to say for >MADV_FREE. It may have been effective until 1999. Sounds quite probable. -- 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 marcel at FreeBSD.org Fri Jun 6 23:58:39 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Fri Jun 6 23:58:42 2008 Subject: cvs commit: src/sbin/geom/class/part geom_part.c Message-ID: <200806062358.m56Nwchx080028@repoman.freebsd.org> marcel 2008-06-06 23:58:29 UTC FreeBSD src repository Modified files: sbin/geom/class/part geom_part.c Log: SVN rev 179629 on 2008-06-06 23:58:29Z by marcel Enhance the bootcode command to also allow bootcode to be written to a partition. This avoids that users need to use dd(1) to install boot code (as is needed for VTOC8 and booting GPT on PCs). Revision Changes Path 1.7 +143 -25 src/sbin/geom/class/part/geom_part.c From marcel at FreeBSD.org Sat Jun 7 00:14:12 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sat Jun 7 00:14:15 2008 Subject: cvs commit: src/sbin/geom/class/part gpart.8 Message-ID: <200806070014.m570E9kK082335@repoman.freebsd.org> marcel 2008-06-07 00:14:06 UTC FreeBSD src repository Modified files: sbin/geom/class/part gpart.8 Log: SVN rev 179630 on 2008-06-07 00:14:06Z by marcel Document the new -p and -i options for writing bootstrap code into a partition. Make it clear that the -b option embeds boot code in the meta-data. Revision Changes Path 1.4 +24 -8 src/sbin/geom/class/part/gpart.8 From jhb at FreeBSD.org Sat Jun 7 03:07:45 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Sat Jun 7 03:07:47 2008 Subject: cvs commit: src/sys/boot/i386/libi386 biossmap.c Message-ID: <200806070307.m5737jno010251@repoman.freebsd.org> jhb 2008-06-07 03:07:32 UTC FreeBSD src repository Modified files: sys/boot/i386/libi386 biossmap.c Log: SVN rev 179631 on 2008-06-07 03:07:32Z by jhb Workaround a bug in the BIOS of Dell R900 machines. Specifically, each entry in the SMAP is a 20 byte structure and they are queried from the BIOS via sucessive BIOS calls. Due to an apparent bug in the R900's BIOS, for some SMAP requests the BIOS overflows the 20 byte buffer trashing a few bytes of memory immediately after the SMAP structure. As a workaround, add 8 bytes of padding after the SMAP structure used in the loader for SMAP queries. PR: i386/122668 Submitted by: Mike Hibler mike flux.utah.edu, silby MFC after: 3 days Revision Changes Path 1.8 +4 -1 src/sys/boot/i386/libi386/biossmap.c From jhb at freebsd.org Sat Jun 7 03:26:09 2008 From: jhb at freebsd.org (John Baldwin) Date: Sat Jun 7 03:26:11 2008 Subject: cvs commit: src/sys/boot/i386/libi386 biossmap.c In-Reply-To: <200806070307.m5737jno010251@repoman.freebsd.org> References: <200806070307.m5737jno010251@repoman.freebsd.org> Message-ID: <200806062311.17987.jhb@freebsd.org> On Friday 06 June 2008 11:07:32 pm John Baldwin wrote: > jhb 2008-06-07 03:07:32 UTC > > FreeBSD src repository > > Modified files: > sys/boot/i386/libi386 biossmap.c > Log: > SVN rev 179631 on 2008-06-07 03:07:32Z by jhb > > Workaround a bug in the BIOS of Dell R900 machines. Specifically, each > entry in the SMAP is a 20 byte structure and they are queried from the > BIOS via sucessive BIOS calls. Due to an apparent bug in the R900's > BIOS, for some SMAP requests the BIOS overflows the 20 byte buffer > trashing a few bytes of memory immediately after the SMAP structure. As > a workaround, add 8 bytes of padding after the SMAP structure used in > the loader for SMAP queries. > > PR: i386/122668 > Submitted by: Mike Hibler mike flux.utah.edu, silby > MFC after: 3 days The code to parse the SMAP in the kernel from vm86 mode is unaffected as it allocates an entire page for use when fetching the SMAP entry. -- John Baldwin From kib at FreeBSD.org Sat Jun 7 05:42:33 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sat Jun 7 05:42:36 2008 Subject: cvs commit: src/sys/boot/common ufsread.c Message-ID: <200806070542.m575gXue035280@repoman.freebsd.org> kib 2008-06-07 05:42:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/common ufsread.c Log: SVN rev 179632 on 2008-06-07 05:42:22Z by kib Fix the incorrect calculation of a block address within a single indirect block. PR: 108215 Submitted by: Yuichiro Goto, y7goto gmail com MFC after: 2 weeks Revision Changes Path 1.14.10.2 +1 -1 src/sys/boot/common/ufsread.c From kib at FreeBSD.org Sat Jun 7 05:47:34 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sat Jun 7 05:47:38 2008 Subject: cvs commit: src/sys/boot/common ufsread.c Message-ID: <200806070547.m575lY6H035693@repoman.freebsd.org> kib 2008-06-07 05:47:24 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/common ufsread.c Log: SVN rev 179633 on 2008-06-07 05:47:24Z by kib Revert the r179632, it was committed to the wrong branch. Revision Changes Path 1.14.10.3 +1 -1 src/sys/boot/common/ufsread.c From kib at FreeBSD.org Sat Jun 7 05:49:35 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sat Jun 7 05:49:39 2008 Subject: cvs commit: src/sys/boot/common ufsread.c Message-ID: <200806070549.m575nZno035842@repoman.freebsd.org> kib 2008-06-07 05:49:24 UTC FreeBSD src repository Modified files: sys/boot/common ufsread.c Log: SVN rev 179634 on 2008-06-07 05:49:24Z by kib Fix the incorrect calculation of a block address within a single indirect block. PR: 108215 Submitted by: Yuichiro Goto, y7goto gmail com MFC after: 2 weeks Revision Changes Path 1.18 +1 -1 src/sys/boot/common/ufsread.c From adrian at FreeBSD.org Sat Jun 7 10:08:18 2008 From: adrian at FreeBSD.org (Adrian Chadd) Date: Sat Jun 7 10:08:27 2008 Subject: cvs commit: src/sys/i386/conf PAE Message-ID: <200806071008.m57A8HRE071115@repoman.freebsd.org> adrian 2008-06-07 09:35:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/i386/conf PAE Log: SVN rev 179635 on 2008-06-07 09:35:41Z by adrian USB is disabled in RELENG_7 PAE; don't try to build USB related devices. PR: kern/123619 Revision Changes Path 1.25.2.2 +10 -0 src/sys/i386/conf/PAE From adrian at FreeBSD.org Sat Jun 7 10:13:01 2008 From: adrian at FreeBSD.org (Adrian Chadd) Date: Sat Jun 7 10:13:04 2008 Subject: cvs commit: src/sys/i386/conf PAE Message-ID: <200806071013.m57AD0Mf071666@repoman.freebsd.org> adrian 2008-06-07 09:35:41 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/i386/conf PAE Log: SVN rev 179635 on 2008-06-07 09:35:41Z by adrian USB is disabled in RELENG_7 PAE; don't try to build USB related devices. PR: kern/123619 Revision Changes Path 1.25.2.3 +0 -0 src/sys/i386/conf/PAE From keramida at FreeBSD.org Sat Jun 7 10:13:02 2008 From: keramida at FreeBSD.org (Giorgos Keramidas) Date: Sat Jun 7 10:13:08 2008 Subject: cvs commit: src/bin/ln ln.c Message-ID: <200806071013.m57AD1dO071696@repoman.freebsd.org> keramida 2008-06-07 09:37:30 UTC FreeBSD src repository Modified files: bin/ln ln.c Log: SVN rev 179636 on 2008-06-07 09:37:30Z by keramida Unbreak last commit to ln for amd64. Cast string precision to `int'. amd64 systems warn about the field precision being `long int' if we don't, and pathnames are normally short enough to fit in an `int'. Noticed by: pav Revision Changes Path 1.38 +1 -1 src/bin/ln/ln.c From phk at FreeBSD.org Sat Jun 7 10:13:03 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Sat Jun 7 10:13:08 2008 Subject: cvs commit: src/sys/conf newvers.sh Message-ID: <200806071013.m57AD2DD071723@repoman.freebsd.org> phk 2008-06-07 09:49:57 UTC FreeBSD src repository Modified files: sys/conf newvers.sh Log: SVN rev 179637 on 2008-06-07 09:49:57Z by phk If we can find it, include SVN version number in kernel version strings. See also: http://www.bikeshed.org/ Revision Changes Path 1.74 +9 -2 src/sys/conf/newvers.sh From jhb at freebsd.org Sat Jun 7 13:48:10 2008 From: jhb at freebsd.org (John Baldwin) Date: Sat Jun 7 13:48:12 2008 Subject: cvs commit: src/sys/i386/conf PAE In-Reply-To: <200806071008.m57A8HRE071115@repoman.freebsd.org> References: <200806071008.m57A8HRE071115@repoman.freebsd.org> Message-ID: <200806070909.10727.jhb@freebsd.org> On Saturday 07 June 2008 05:35:41 am Adrian Chadd wrote: > adrian 2008-06-07 09:35:41 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7) > sys/i386/conf PAE > Log: > SVN rev 179635 on 2008-06-07 09:35:41Z by adrian > > USB is disabled in RELENG_7 PAE; don't try to build USB > related devices. > > PR: kern/123619 > > Revision Changes Path > 1.25.2.2 +10 -0 src/sys/i386/conf/PAE > Actually, I think it is safe to enable USB in 7.x PAE. -- John Baldwin From adrian at freebsd.org Sat Jun 7 15:17:58 2008 From: adrian at freebsd.org (Adrian Chadd) Date: Sat Jun 7 15:18:01 2008 Subject: cvs commit: src/sys/i386/conf PAE In-Reply-To: <200806070909.10727.jhb@freebsd.org> References: <200806071008.m57A8HRE071115@repoman.freebsd.org> <200806070909.10727.jhb@freebsd.org> Message-ID: 2008/6/7 John Baldwin : > > Actually, I think it is safe to enable USB in 7.x PAE. I was wondering if that were the case; this commit was just to fix the broken PAE kernel build. Adrian -- Adrian Chadd - adrian@freebsd.org From rse at FreeBSD.org Sat Jun 7 17:54:37 2008 From: rse at FreeBSD.org (Ralf S. Engelschall) Date: Sat Jun 7 17:54:45 2008 Subject: cvs commit: src/bin/sh alias.c Message-ID: <200806071754.m57Hsbud036702@repoman.freebsd.org> rse 2008-06-07 16:19:28 UTC FreeBSD src repository Modified files: bin/sh alias.c Log: SVN rev 179638 on 2008-06-07 16:19:28Z by rse remove an unnecessary include Revision Changes Path 1.21 +0 -1 src/bin/sh/alias.c From rse at FreeBSD.org Sat Jun 7 17:54:40 2008 From: rse at FreeBSD.org (Ralf S. Engelschall) Date: Sat Jun 7 17:54:49 2008 Subject: cvs commit: src/bin/sh alias.c Message-ID: <200806071754.m57HscUY036732@repoman.freebsd.org> rse 2008-06-07 16:28:20 UTC FreeBSD src repository Modified files: bin/sh alias.c Log: SVN rev 179639 on 2008-06-07 16:28:20Z by rse use 'const' for the parameters of the two static functions unalias() and hashalias() Revision Changes Path 1.22 +4 -4 src/bin/sh/alias.c From sam at FreeBSD.org Sat Jun 7 17:54:40 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Sat Jun 7 17:54:52 2008 Subject: cvs commit: src/sys/net80211 ieee80211_node.c Message-ID: <200806071754.m57Hsdd2036759@repoman.freebsd.org> sam 2008-06-07 17:43:41 UTC FreeBSD src repository Modified files: sys/net80211 ieee80211_node.c Log: SVN rev 179640 on 2008-06-07 17:43:41Z by sam correct code returned in AsssocResp; the previous code is spec'd for use only in deauth+disassoc frames Submitted by: Chris Zimmermann MFC after: 1 month Revision Changes Path 1.96 +1 -2 src/sys/net80211/ieee80211_node.c From sam at FreeBSD.org Sat Jun 7 17:54:41 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Sat Jun 7 17:54:58 2008 Subject: cvs commit: src/sys/net80211 ieee80211_node.c Message-ID: <200806071754.m57HseJD036786@repoman.freebsd.org> sam 2008-06-07 17:50:24 UTC FreeBSD src repository Modified files: sys/net80211 ieee80211_node.c Log: SVN rev 179641 on 2008-06-07 17:50:24Z by sam Fix node cleanup issues when the last reference is held by the driver (e.g. in the tx q); ni_vap may not be valid, we must carefully check before using it. Revision Changes Path 1.97 +27 -16 src/sys/net80211/ieee80211_node.c From sam at FreeBSD.org Sat Jun 7 17:54:42 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Sat Jun 7 17:54:59 2008 Subject: cvs commit: src/sys/net80211 ieee80211_node.c Message-ID: <200806071754.m57Hsfjx036813@repoman.freebsd.org> sam 2008-06-07 17:51:41 UTC FreeBSD src repository Modified files: sys/net80211 ieee80211_node.c Log: SVN rev 179642 on 2008-06-07 17:51:41Z by sam Split ieee80211_notify_erp into locked and unlocked variants and use the locked version to handle the hostap input path case where the com lock is not already held. Noticed by: Jared Go Revision Changes Path 1.98 +12 -4 src/sys/net80211/ieee80211_node.c From sam at FreeBSD.org Sat Jun 7 18:39:08 2008 From: sam at FreeBSD.org (Sam Leffler) Date: Sat Jun 7 18:39:10 2008 Subject: cvs commit: src/sys/dev/ath if_ath.c src/sys/dev/iwi if_iwi.c src/sys/dev/iwn if_iwn.c src/sys/dev/ral rt2560.c rt2661.c src/sys/dev/usb if_rum.c if_ural.c if_zyd.c src/sys/dev/wpi if_wpi.c src/sys/net80211 ieee80211_node.c ieee80211_var.h Message-ID: <200806071839.m57Id8b0043440@repoman.freebsd.org> sam 2008-06-07 18:38:02 UTC FreeBSD src repository Modified files: sys/dev/ath if_ath.c sys/dev/iwi if_iwi.c sys/dev/iwn if_iwn.c sys/dev/ral rt2560.c rt2661.c sys/dev/usb if_rum.c if_ural.c if_zyd.c sys/dev/wpi if_wpi.c sys/net80211 ieee80211_node.c ieee80211_var.h Log: SVN rev 179643 on 2008-06-07 18:38:02Z by sam Change the calling convention for ic_node_alloc to deal with some longstanding issues: o pass the vap since it's now the "coin of the realm" and required to do things like set initial tx parameters in private node state for use prior to association o pass the mac address as cards that maintain outboard station tables require this to create an entry (e.g. in ibss mode) o remove the node table reference, we only have one node table and it's unlikely this will change so this is not needed to find the com structure Revision Changes Path 1.191 +4 -3 src/sys/dev/ath/if_ath.c 1.63 +4 -3 src/sys/dev/iwi/if_iwi.c 1.7 +3 -2 src/sys/dev/iwn/if_iwn.c 1.26 +4 -3 src/sys/dev/ral/rt2560.c 1.26 +4 -3 src/sys/dev/ral/rt2661.c 1.23 +4 -2 src/sys/dev/usb/if_rum.c 1.74 +4 -2 src/sys/dev/usb/if_ural.c 1.18 +4 -2 src/sys/dev/usb/if_zyd.c 1.15 +4 -2 src/sys/dev/wpi/if_wpi.c 1.99 +5 -4 src/sys/net80211/ieee80211_node.c 1.59 +2 -1 src/sys/net80211/ieee80211_var.h From marcel at FreeBSD.org Sat Jun 7 21:57:31 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sat Jun 7 21:57:34 2008 Subject: cvs commit: src/sys/conf files.powerpc src/sys/powerpc/include dbdma.h src/sys/powerpc/powermac dbdma.c dbdmavar.h Message-ID: <200806072157.m57LvVZ2076384@repoman.freebsd.org> marcel 2008-06-07 21:56:48 UTC FreeBSD src repository Modified files: sys/conf files.powerpc Added files: sys/powerpc/include dbdma.h sys/powerpc/powermac dbdma.c dbdmavar.h Log: SVN rev 179644 on 2008-06-07 21:56:48Z by marcel Add support for Apple's Descriptor-Based DMA (DBDMA) engine. The DMA engine is usful to various existing drivers, such as ata(4) and scc(4), and is used bhy the soon to be added bm(4). Submitted by: Nathan Whitehorn Revision Changes Path 1.78 +1 -0 src/sys/conf/files.powerpc 1.1 +118 -0 src/sys/powerpc/include/dbdma.h (new) 1.1 +331 -0 src/sys/powerpc/powermac/dbdma.c (new) 1.1 +102 -0 src/sys/powerpc/powermac/dbdmavar.h (new) From marcel at FreeBSD.org Sat Jun 7 22:59:48 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sat Jun 7 22:59:50 2008 Subject: cvs commit: src/sys/conf NOTES files.powerpc src/sys/dev/bm if_bm.c if_bmreg.h if_bmvar.h src/sys/dev/mii lxtphy.c src/sys/modules Makefile src/sys/modules/bm Makefile src/sys/powerpc/conf GENERIC Message-ID: <200806072259.m57Mxmdo083847@repoman.freebsd.org> marcel 2008-06-07 22:58:32 UTC FreeBSD src repository Modified files: sys/conf NOTES files.powerpc sys/dev/mii lxtphy.c sys/modules Makefile sys/powerpc/conf GENERIC Added files: sys/dev/bm if_bm.c if_bmreg.h if_bmvar.h sys/modules/bm Makefile Log: SVN rev 179645 on 2008-06-07 22:58:32Z by marcel Add support for the Apple Big Mac (BMAC) Ethernet controller, found on various Apple G3 models. Submitted by: Nathan Whitehorn Revision Changes Path 1.1493 +2 -0 src/sys/conf/NOTES 1.79 +1 -0 src/sys/conf/files.powerpc 1.1 +1452 -0 src/sys/dev/bm/if_bm.c (new) 1.1 +176 -0 src/sys/dev/bm/if_bmreg.h (new) 1.1 +127 -0 src/sys/dev/bm/if_bmvar.h (new) 1.20 +10 -0 src/sys/dev/mii/lxtphy.c 1.572 +2 -0 src/sys/modules/Makefile 1.1 +9 -0 src/sys/modules/bm/Makefile (new) 1.79 +1 -0 src/sys/powerpc/conf/GENERIC From marcel at FreeBSD.org Sun Jun 8 01:59:03 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Sun Jun 8 01:59:06 2008 Subject: cvs commit: src/sys/conf NOTES src/sys/powerpc/conf NOTES Message-ID: <200806080159.m581x39b010644@repoman.freebsd.org> marcel 2008-06-08 01:58:11 UTC FreeBSD src repository Modified files: sys/conf NOTES sys/powerpc/conf NOTES Log: SVN rev 179646 on 2008-06-08 01:58:11Z by marcel Move bm(4) from the sys/conf/NOTES to sys/powerpc/conf/NOTES. The driver applies to PowerPC only. Revision Changes Path 1.1494 +0 -2 src/sys/conf/NOTES 1.18 +1 -0 src/sys/powerpc/conf/NOTES From yongari at FreeBSD.org Sun Jun 8 02:52:36 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Sun Jun 8 02:52:40 2008 Subject: cvs commit: src/sys/dev/dc if_dc.c Message-ID: <200806080252.m582qaB6015926@repoman.freebsd.org> yongari 2008-06-08 02:52:26 UTC FreeBSD src repository Modified files: sys/dev/dc if_dc.c Log: SVN rev 179647 on 2008-06-08 02:52:26Z by yongari Ethernet hardware address stored in DC_AL_PAR0/DC_AL_PAR1 register is in little endian form. Likewise setting DC_AL_PAR0/DC_AL_PAR1 register expect the address to be in little endian form. For big endian architectures the address should be swapped to get correct one. Change setting/getting ethernet hardware address to big endian architecture frendly. Reported by: Robert Murillo ( billypilgrim782001 at yahoo dot com ) Tested by: Robert Murillo ( billypilgrim782001 at yahoo dot com ) Revision Changes Path 1.195 +14 -6 src/sys/dev/dc/if_dc.c From jhb at freebsd.org Sun Jun 8 04:57:17 2008 From: jhb at freebsd.org (John Baldwin) Date: Sun Jun 8 04:57:32 2008 Subject: cvs commit: src/sys/conf NOTES files.powerpc src/sys/dev/bm if_bm.c if_bmreg.h if_bmvar.h src/sys/dev/mii lxtphy.c src/sys/modules Makefile src/sys/modules/bm Makefile src/sys/powerpc/conf GENERIC In-Reply-To: <200806072259.m57Mxmdo083847@repoman.freebsd.org> References: <200806072259.m57Mxmdo083847@repoman.freebsd.org> Message-ID: <200806080052.25362.jhb@freebsd.org> On Saturday 07 June 2008 06:58:32 pm Marcel Moolenaar wrote: > marcel 2008-06-07 22:58:32 UTC > > FreeBSD src repository > > Modified files: > sys/conf NOTES files.powerpc > sys/dev/mii lxtphy.c > sys/modules Makefile > sys/powerpc/conf GENERIC > Added files: > sys/dev/bm if_bm.c if_bmreg.h if_bmvar.h > sys/modules/bm Makefile > Log: > SVN rev 179645 on 2008-06-07 22:58:32Z by marcel > > Add support for the Apple Big Mac (BMAC) Ethernet controller, > found on various Apple G3 models. > > Submitted by: Nathan Whitehorn A few notes from a quick perusal: You should not need locking in the mii read/write routines as the lock is already held at the higher levels via the ifmedia handlers and the timer routine. Once that is done you can remove MTX_RECURSE from the mutex. New drivers should probably just use bus_foo() (e.g. bus_read_1()) rather than bus_space_foo(). This would mean not having the bus space tag/handle in the softc anymore. ether_ifattach() already sets if_data.ifi_hdrlen to the correct size for ethernet so no need to do that in the attach routine. In detach you should only hold the lock over the call to bm_stop(). Right now it is subjec to deadlocks by holding it over bus_teardown_intr(), for example. The callout_drain() must be moved after the bm_stop(). It is also missing a call to ether_ifdetach(). The first part of detach shound look like: { BM_LOCK(sc); bm_stop(sc); BM_UNLOCK(sc); callout_drain(...) ether_ifdetach(...) bus_teardown_intr(...) Then you can free DMA memory and I/O resources and finally destroy the mutex. The detach routine is also missing a call to if_free() (near the bottom usually, but anytime after the above code sequence is fine). bm_shutdown() is missing locking. bm_stop() should do a callout_stop() of the timer. There isn't a need to check for link status or tx timeouts when the device is marked down. You really don't need to have the EJUSTRETURN magic in the watchdog routine. The callout code doesn't mind if you schedule the callout twice. -- John Baldwin From jeremie at le-hen.org Sun Jun 8 07:56:45 2008 From: jeremie at le-hen.org (Jeremie Le Hen) Date: Sun Jun 8 07:56:51 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <200805261721.m4QHLCF4070787@repoman.freebsd.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> Message-ID: <20080608075256.GA35560@obiwan.tataz.chchile.org> Hi Tim, First of all, thanks for this work. It greatly contributes to the cleanness of the overall source tree :-). On Mon, May 26, 2008 at 05:21:12PM +0000, Tim Kientzle wrote: > kientzle 2008-05-26 17:21:12 UTC > > FreeBSD src repository > > Modified files: > gnu/usr.bin/cpio Makefile > Log: > Install GNU cpio as 'gcpio', symlink to 'cpio' unless WITH_BSDCPIO is > defined. There have been much work to glean all build flags over the source tree into a single place, namely src/share/bsd.own.mk. See [1]. I think it would be better to stick to this scheme and not spread those knobs back in the tree. [1] http://www.freebsd.org/cgi/cvsweb.cgi/src/share/mk/bsd.own.mk#rev1.44 Thanks. Regards, -- Jeremie Le Hen < jeremie at le-hen dot org >< ttz at chchile dot org > From wkoszek at FreeBSD.org Sun Jun 8 08:56:56 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Sun Jun 8 08:56:58 2008 Subject: cvs commit: src/sys/mips/include pte.h src/sys/mips/mips trap.c Message-ID: <200806080856.m588uuQ9079902@repoman.freebsd.org> wkoszek 2008-06-08 08:56:46 UTC FreeBSD src repository Modified files: sys/mips/include pte.h sys/mips/mips trap.c Log: SVN rev 179648 on 2008-06-08 08:56:46Z by wkoszek Fix spelling of "virtual". There should be no visible change. Reviewed by: rink Revision Changes Path 1.2 +1 -1 src/sys/mips/include/pte.h 1.2 +4 -4 src/sys/mips/mips/trap.c From philip at FreeBSD.org Sun Jun 8 09:06:57 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sun Jun 8 09:07:02 2008 Subject: cvs commit: src/sys/dev/atkbdc psm.c Message-ID: <200806080906.m5896vKs081874@repoman.freebsd.org> philip 2008-06-08 09:06:45 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/atkbdc psm.c Log: SVN rev 179649 on 2008-06-08 09:06:45Z by philip MFC r179474: Try to detect a Synaptics touchpad before IntelliMouse. Some touchpads will pretend to be IntelliMouse (which have a few more features than generic mice) causing the IntelliMouse probe to work and the Synaptics code never to be called. This should not break "real" IntelliMouse because the Synaptics detection code is fairly specific. PR: kern/120833 Submitted by: Eygene Ryabinkin Revision Changes Path 1.93.2.3 +2 -2 src/sys/dev/atkbdc/psm.c From wkoszek at FreeBSD.org Sun Jun 8 09:09:14 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Sun Jun 8 09:09:21 2008 Subject: cvs commit: src/share/man/man4 ddb.4 Message-ID: <200806080909.m5899DDg082043@repoman.freebsd.org> wkoszek 2008-06-08 09:09:08 UTC FreeBSD src repository Modified files: share/man/man4 ddb.4 Log: SVN rev 179650 on 2008-06-08 09:09:08Z by wkoszek Bring more commands to the ddb(4) manual page. This should be considered "complete" list, but some commands might be still missing. Reviewed by: julian Revision Changes Path 1.51 +495 -33 src/share/man/man4/ddb.4 From rdivacky at FreeBSD.org Sun Jun 8 11:09:31 2008 From: rdivacky at FreeBSD.org (Roman Divacky) Date: Sun Jun 8 11:09:34 2008 Subject: cvs commit: src/sys/compat/linux linux_file.c Message-ID: <200806081109.m58B9U9r095057@repoman.freebsd.org> rdivacky 2008-06-08 11:09:25 UTC FreeBSD src repository Modified files: sys/compat/linux linux_file.c Log: SVN rev 179651 on 2008-06-08 11:09:25Z by rdivacky d_ino member of linux_dirent structure should be unsigned long. Submitted by: Chagin Dmitry Approved by: kib (mentor) Revision Changes Path 1.114 +2 -2 src/sys/compat/linux/linux_file.c From wkoszek at FreeBSD.org Sun Jun 8 12:43:17 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Sun Jun 8 12:43:21 2008 Subject: cvs commit: src/usr.bin/rev rev.c Message-ID: <200806081243.m58ChHlk007910@repoman.freebsd.org> wkoszek 2008-06-08 12:43:02 UTC FreeBSD src repository Modified files: usr.bin/rev rev.c Log: SVN rev 179652 on 2008-06-08 12:43:02Z by wkoszek Make usage() 'static'. Revision Changes Path 1.10 +1 -1 src/usr.bin/rev/rev.c From rpaulo at FreeBSD.org Sun Jun 8 14:42:51 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Sun Jun 8 14:42:55 2008 Subject: cvs commit: src/sys/dev/age if_age.c Message-ID: <200806081442.m58Ego69031955@repoman.freebsd.org> rpaulo 2008-06-08 14:42:43 UTC FreeBSD src repository Modified files: sys/dev/age if_age.c Log: SVN rev 179653 on 2008-06-08 14:42:43Z by rpaulo Fix typo in comment. Revision Changes Path 1.2 +1 -1 src/sys/dev/age/if_age.c From jeremie at le-hen.org Sun Jun 8 15:00:17 2008 From: jeremie at le-hen.org (Jeremie Le Hen) Date: Sun Jun 8 15:00:21 2008 Subject: cvs commit: src/include string.h src/lib/libc/string Makefile.inc memchr.3 memrchr.c src/sys/sys param.h In-Reply-To: <20080527234335.GA24995@owl.midgard.homeip.net> References: <200805272004.m4RK4SZt029194@repoman.freebsd.org> <483C7FF2.6000607@FreeBSD.org> <20080527234335.GA24995@owl.midgard.homeip.net> Message-ID: <20080608145627.GE35560@obiwan.tataz.chchile.org> Hi Erik, On Wed, May 28, 2008 at 01:43:35AM +0200, Erik Trulsson wrote: > Compatibility in the other direction (that everything that works on 6.x+1 > will also work on 6.x), while obviously somewhat desirable, has AFAIK never > been required and certainly not guaranteed. Such compatibility has almost > certainly already been broken several times on all the -STABLE branches by > now. It has been never required, but binary packages are built on the latest -STABLE AFAIK, so not being careful about foreward compat incidentally requires users to chase -STABLE as well. This may be a problem for -RELEASE users, though this is somewhat alleviated by the fact the latters surely don't upgrade their 3rd party packages very often. every so often. Regards, -- Jeremie Le Hen < jeremie at le-hen dot org >< ttz at chchile dot org > From marius at alchemy.franken.de Sun Jun 8 15:13:31 2008 From: marius at alchemy.franken.de (Marius Strobl) Date: Sun Jun 8 15:13:36 2008 Subject: cvs commit: src/sys/conf kern.post.mk kern.pre.mk In-Reply-To: <200805230353.m4N3rnsC025686@repoman.freebsd.org> References: <200805230353.m4N3rnsC025686@repoman.freebsd.org> Message-ID: <20080608151312.GB37620@alchemy.franken.de> On Fri, May 23, 2008 at 03:53:49AM +0000, John Birrell wrote: > jb 2008-05-23 03:53:49 UTC > > FreeBSD src repository > > Modified files: > sys/conf kern.post.mk kern.pre.mk > Log: > Add support for generating CTF data for the kernel. > > Revision Changes Path > 1.102 +3 -0 src/sys/conf/kern.post.mk > 1.98 +15 -0 src/sys/conf/kern.pre.mk Is there a particular reason you disabled optimizations for sparc64 along with this? Disabling them completely reveals unaligned accesses throughout the kernel. Marius From scf at FreeBSD.org Sun Jun 8 19:35:56 2008 From: scf at FreeBSD.org (Sean Farley) Date: Sun Jun 8 19:35:57 2008 Subject: cvs commit: src/games/morse morse.6 morse.c Message-ID: <200806081935.m58JZt3j075741@repoman.freebsd.org> scf 2008-06-08 19:35:47 UTC FreeBSD src repository Modified files: games/morse morse.6 morse.c Log: SVN rev 179654 on 2008-06-08 19:35:47Z by scf Fixed the output grammar to properly speak non-terminal dits. Updated Lyndon Nerenberg's radio callsign and E-mail address. PR: bin/7868 Submitted by: Lyndon Nerenberg MFC after: 1 week Revision Changes Path 1.18 +3 -2 src/games/morse/morse.6 1.22 +4 -3 src/games/morse/morse.c From dougb at FreeBSD.org Sun Jun 8 19:46:27 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Sun Jun 8 19:46:31 2008 Subject: cvs commit: src/sys/conf newvers.sh Message-ID: <200806081946.m58JkQLR076605@repoman.freebsd.org> dougb 2008-06-08 19:46:23 UTC FreeBSD src repository Modified files: sys/conf newvers.sh Log: SVN rev 179655 on 2008-06-08 19:46:23Z by dougb The change to add subversion ID has two problems. The first is that when newvers.sh is run pwd is actually the obj directory, so "../../.svn" doesn't exist and the test always fails. The second is that buildkernel is executed with a restrictive PATH, so unless you have svnversion in /bin or /usr/bin it can't run. Fix this by looking for svnversion in /bin, /usr/bin, and /usr/local/bin in that order. If found, store the location and derive the value of the source directory. Then run svnversion in the appropriate directory. There is one possible refinement which would be to add a test for LOCALBASE!=/usr/local if we don't find svnversion the first time, but IMO that's not necessary at this time. Revision Changes Path 1.75 +11 -2 src/sys/conf/newvers.sh From stas at FreeBSD.org Sun Jun 8 19:48:58 2008 From: stas at FreeBSD.org (Stanislav Sedov) Date: Sun Jun 8 19:49:17 2008 Subject: cvs commit: src/sbin/fsck_ffs main.c Message-ID: <200806081948.m58Jmw36076776@repoman.freebsd.org> stas 2008-06-08 19:48:41 UTC FreeBSD src repository Modified files: sbin/fsck_ffs main.c Log: SVN rev 179656 on 2008-06-08 19:48:41Z by stas - Display '-C' option in usage(). Approved by: kib Revision Changes Path 1.52 +1 -1 src/sbin/fsck_ffs/main.c From dwmalone at FreeBSD.org Sun Jun 8 19:59:30 2008 From: dwmalone at FreeBSD.org (David Malone) Date: Sun Jun 8 19:59:32 2008 Subject: cvs commit: src/usr.bin/lex flex.skl flexdef.h lex.1 parse.y scan.l Message-ID: <200806081959.m58JxTlF077719@repoman.freebsd.org> dwmalone 2008-06-08 19:59:15 UTC FreeBSD src repository Modified files: usr.bin/lex flex.skl flexdef.h lex.1 parse.y scan.l Log: SVN rev 179657 on 2008-06-08 19:59:15Z by dwmalone I missed some "register"s in non-dot-C files. Revision Changes Path 1.9 +18 -18 src/usr.bin/lex/flex.skl 1.7 +7 -7 src/usr.bin/lex/flexdef.h 1.19 +1 -1 src/usr.bin/lex/lex.1 1.4 +2 -2 src/usr.bin/lex/parse.y 1.6 +2 -2 src/usr.bin/lex/scan.l From wkoszek at FreeBSD.org Sun Jun 8 20:43:46 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Sun Jun 8 20:43:48 2008 Subject: cvs commit: src/sys/kern vfs_mount.c Message-ID: <200806082043.m58Khjno082296@repoman.freebsd.org> wkoszek 2008-06-08 20:43:27 UTC FreeBSD src repository Modified files: sys/kern vfs_mount.c Log: SVN rev 179658 on 2008-06-08 20:43:27Z by wkoszek Remove checks against DDB, which isn't used in this file. My intention is to bring no functional change. Discussion on: IRC Reviewed by: ed, kan, rink, Revision Changes Path 1.283 +0 -5 src/sys/kern/vfs_mount.c From dwmalone at FreeBSD.org Sun Jun 8 21:00:33 2008 From: dwmalone at FreeBSD.org (David Malone) Date: Sun Jun 8 21:00:34 2008 Subject: cvs commit: src/sbin/restore dirs.c interactive.c tape.c Message-ID: <200806082100.m58L0WgX093046@repoman.freebsd.org> dwmalone 2008-06-08 21:00:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/restore dirs.c interactive.c tape.c Log: SVN rev 179659 on 2008-06-08 21:00:25Z by dwmalone MFC: Sync restore in RELENG_7 with -current. Fix remaining nits from PR bin/39905. Check for write failure while building temp files. Correctly set file group when restore is run by a user other than root. Use safer string handling. Fixes the "getfile: lost data" panic when restoring older dumps. Fix spurious "Header with wrong dumpdate" message. Revision Changes Path 1.31.2.1 +81 -33 src/sbin/restore/dirs.c 1.18.2.1 +4 -6 src/sbin/restore/interactive.c 1.49.2.1 +16 -17 src/sbin/restore/tape.c From wkoszek at FreeBSD.org Sun Jun 8 21:01:49 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Sun Jun 8 21:01:53 2008 Subject: cvs commit: src/share/man/man4 ddb.4 Message-ID: <200806082101.m58L1mnt093150@repoman.freebsd.org> wkoszek 2008-06-08 21:01:39 UTC FreeBSD src repository Modified files: share/man/man4 ddb.4 Log: SVN rev 179660 on 2008-06-08 21:01:39Z by wkoszek Fix the way the date must be specified (leading 0 is redundant). Fix several spelling mistakes brought by my earlier commit. Trim whitespace. Submitted by: maxim Revision Changes Path 1.52 +4 -4 src/share/man/man4/ddb.4 From wkoszek at FreeBSD.org Sun Jun 8 21:08:35 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Sun Jun 8 21:08:36 2008 Subject: cvs commit: src/share/man/man4 ddb.4 Message-ID: <200806082108.m58L8YcB094819@repoman.freebsd.org> wkoszek 2008-06-08 21:08:20 UTC FreeBSD src repository Modified files: share/man/man4 ddb.4 Log: SVN rev 179661 on 2008-06-08 21:08:20Z by wkoszek Change header file references from .h to path relative to /usr/include. It looks much better anyway. Instead of referencing "socket", which is bogus, reference 'sys/socket.h', which is what should be placed here from the beggining. Suggested by: maxim Revision Changes Path 1.53 +10 -10 src/share/man/man4/ddb.4 From jb at what-creek.com Sun Jun 8 22:25:49 2008 From: jb at what-creek.com (John Birrell) Date: Sun Jun 8 22:25:52 2008 Subject: cvs commit: src/sys/conf kern.post.mk kern.pre.mk In-Reply-To: <20080608151312.GB37620@alchemy.franken.de> References: <200805230353.m4N3rnsC025686@repoman.freebsd.org> <20080608151312.GB37620@alchemy.franken.de> Message-ID: <20080608222547.GA29976@what-creek.com> On Sun, Jun 08, 2008 at 05:13:13PM +0200, Marius Strobl wrote: > On Fri, May 23, 2008 at 03:53:49AM +0000, John Birrell wrote: > > jb 2008-05-23 03:53:49 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/conf kern.post.mk kern.pre.mk > > Log: > > Add support for generating CTF data for the kernel. > > > > Revision Changes Path > > 1.102 +3 -0 src/sys/conf/kern.post.mk > > 1.98 +15 -0 src/sys/conf/kern.pre.mk > > Is there a particular reason you disabled optimizations for > sparc64 along with this? Disabling them completely reveals > unaligned accesses throughout the kernel. That looks like a mistake on my part. I'm not sure where it came from. I'll take it out. -- John Birrell From davidxu at FreeBSD.org Mon Jun 9 01:18:53 2008 From: davidxu at FreeBSD.org (David Xu) Date: Mon Jun 9 01:18:56 2008 Subject: cvs commit: src/include pthread.h src/lib/libc/include namespace.h un-namespace.h src/lib/libthr pthread.map src/lib/libthr/thread thr_clean.c thr_private.h Message-ID: <200806090118.m591IqkN039870@repoman.freebsd.org> davidxu 2008-06-09 01:14:10 UTC FreeBSD src repository Modified files: include pthread.h lib/libc/include namespace.h un-namespace.h lib/libthr pthread.map lib/libthr/thread thr_clean.c thr_private.h Log: SVN rev 179662 on 2008-06-09 01:14:10Z by davidxu Make pthread_cleanup_push() and pthread_cleanup_pop() as a pair of macros, use stack space to keep cleanup information, this eliminates overhead of calling malloc() and free() in thread library. Discussed on: thread@ Revision Changes Path 1.43 +21 -2 src/include/pthread.h 1.24 +0 -2 src/lib/libc/include/namespace.h 1.21 +0 -2 src/lib/libc/include/un-namespace.h 1.30 +2 -0 src/lib/libthr/pthread.map 1.6 +39 -16 src/lib/libthr/thread/thr_clean.c 1.95 +9 -6 src/lib/libthr/thread/thr_private.h From kientzle at freebsd.org Mon Jun 9 05:01:28 2008 From: kientzle at freebsd.org (Tim Kientzle) Date: Mon Jun 9 05:01:35 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <20080608075256.GA35560@obiwan.tataz.chchile.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> Message-ID: <484CB927.3050101@freebsd.org> This case is a little unusual: by default, I want to install both 'cpio' programs and symlink the common name to one of them. How about this: WITH_GCPIO (default) - installs gcpio, symlinks 'cpio' unless WITH_BSDCPIO defined WITHOUT_GCPIO - does not install gcpio at all WITH_BSDCPIO (default) - installs bsdcpio, symlinks 'cpio' WITHOUT_BSDCPIO - does not install bsdcpio at all This way, WITHOUT_BSDCPIO gives you the old behavior, WITHOUT_GCPIO gives you the proposed new behavior; the default gives you the transitional behavior. Tim Kientzle Jeremie Le Hen wrote: > Hi Tim, > > First of all, thanks for this work. It greatly contributes to the > cleanness of the overall source tree :-). > > On Mon, May 26, 2008 at 05:21:12PM +0000, Tim Kientzle wrote: > >>kientzle 2008-05-26 17:21:12 UTC >> >> FreeBSD src repository >> >> Modified files: >> gnu/usr.bin/cpio Makefile >> Log: >> Install GNU cpio as 'gcpio', symlink to 'cpio' unless WITH_BSDCPIO is >> defined. > > > There have been much work to glean all build flags over the source tree > into a single place, namely src/share/bsd.own.mk. See [1]. I think it > would be better to stick to this scheme and not spread those knobs back > in the tree. > > [1] http://www.freebsd.org/cgi/cvsweb.cgi/src/share/mk/bsd.own.mk#rev1.44 > > Thanks. > Regards, From kevlo at FreeBSD.org Mon Jun 9 05:50:49 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Mon Jun 9 05:50:52 2008 Subject: cvs commit: src/sys/arm/xscale/pxa pxa_machdep.c Message-ID: <200806090550.m595omNm075824@repoman.freebsd.org> kevlo 2008-06-09 05:50:42 UTC FreeBSD src repository Modified files: sys/arm/xscale/pxa pxa_machdep.c Log: SVN rev 179663 on 2008-06-09 05:50:42Z by kevlo Unify arminit() and clean up Revision Changes Path 1.2 +1 -58 src/sys/arm/xscale/pxa/pxa_machdep.c From kevlo at FreeBSD.org Mon Jun 9 05:53:20 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Mon Jun 9 05:53:24 2008 Subject: cvs commit: src/sys/arm/xscale/i8134x crb_machdep.c Message-ID: <200806090553.m595rJoZ076032@repoman.freebsd.org> kevlo 2008-06-09 05:53:04 UTC FreeBSD src repository Modified files: sys/arm/xscale/i8134x crb_machdep.c Log: SVN rev 179664 on 2008-06-09 05:53:04Z by kevlo Remove sa1_cache_clean_addr Revision Changes Path 1.7 +0 -2 src/sys/arm/xscale/i8134x/crb_machdep.c From jb at FreeBSD.org Mon Jun 9 06:31:22 2008 From: jb at FreeBSD.org (John Birrell) Date: Mon Jun 9 06:31:25 2008 Subject: cvs commit: src/sys/conf kern.pre.mk Message-ID: <200806090631.m596VLDo080578@repoman.freebsd.org> jb 2008-06-09 06:31:17 UTC FreeBSD src repository Modified files: sys/conf kern.pre.mk Log: SVN rev 179665 on 2008-06-09 06:31:17Z by jb Remove some sparc-specific stuff from my earlier sun4v work in p4. It never belonged in current. Pointed out by: marius Revision Changes Path 1.99 +0 -2 src/sys/conf/kern.pre.mk From jb at FreeBSD.org Mon Jun 9 06:33:38 2008 From: jb at FreeBSD.org (John Birrell) Date: Mon Jun 9 06:33:40 2008 Subject: cvs commit: src/usr.sbin/config mkmakefile.c Message-ID: <200806090633.m596XbJG080799@repoman.freebsd.org> jb 2008-06-09 06:33:26 UTC FreeBSD src repository Modified files: usr.sbin/config mkmakefile.c Log: SVN rev 179666 on 2008-06-09 06:33:26Z by jb Change the CTF conversion makefile code to use a new line to avoid spawning another shell. Requested by: Ed Schouten M config/mkmakefile.c Revision Changes Path 1.93 +1 -1 src/usr.sbin/config/mkmakefile.c From dwmalone at FreeBSD.org Mon Jun 9 07:53:25 2008 From: dwmalone at FreeBSD.org (David Malone) Date: Mon Jun 9 07:53:28 2008 Subject: cvs commit: src/sbin/restore dirs.c interactive.c main.c restore.8 restore.c restore.h tape.c Message-ID: <200806090753.m597rPsi097430@repoman.freebsd.org> dwmalone 2008-06-09 07:53:16 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sbin/restore dirs.c interactive.c main.c restore.8 restore.c restore.h tape.c Log: SVN rev 179667 on 2008-06-09 07:53:16Z by dwmalone MFC: Sync restore with the version in -current. Fix remaining nits from PR bin/39905. Check for write failure while building temp files. Correctly set file group when restore is run by a user other than root. Use safer string handling. Fixes the "getfile: lost data" panic when restoring older dumps. Fix spurious "Header with wrong dumpdate" message. Add -D for degraded mode, which tries harder to read corrupted backups. Remove a diagnostic message from man page that is no longer printed. Revision Changes Path 1.30.2.2 +81 -33 src/sbin/restore/dirs.c 1.17.2.2 +4 -6 src/sbin/restore/interactive.c 1.26.2.1 +5 -2 src/sbin/restore/main.c 1.50.2.2 +11 -8 src/sbin/restore/restore.8 1.18.2.1 +12 -0 src/sbin/restore/restore.c 1.10.2.2 +1 -0 src/sbin/restore/restore.h 1.44.2.3 +90 -31 src/sbin/restore/tape.c From ed at FreeBSD.org Mon Jun 9 08:47:50 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Mon Jun 9 08:47:56 2008 Subject: cvs commit: src/sys/dev/si si.c si.h src/usr.sbin/sicontrol sicontrol.8 sicontrol.c Message-ID: <200806090847.m598ln6m012259@repoman.freebsd.org> ed 2008-06-09 08:43:27 UTC FreeBSD src repository Modified files: sys/dev/si si.c si.h usr.sbin/sicontrol sicontrol.8 sicontrol.c Log: SVN rev 179668 on 2008-06-09 08:43:27Z by ed Remove sicontrol(8)'s "ttystat". In the FreeBSD base system, there are only two utilities that use struct tty, namely pstat and sicontrol. The sicontrol utility calls the TCSI_TTY ioctl(), which copies struct tty back to userspace. sicontrol should not have this functionality. The same data is already provided by pstat. If we really want to be able to export these numbers through a file descriptor to userspace, we can export struct xtty, which should provide a better abstraction. The ttystat option was only used as a debugging aid. This makes sicontrol compile in the mpsafetty branch. Reviewed by: peter Approved by: philip (mentor) Revision Changes Path 1.141 +0 -5 src/sys/dev/si/si.c 1.22 +0 -1 src/sys/dev/si/si.h 1.21 +0 -3 src/usr.sbin/sicontrol/sicontrol.8 1.18 +1 -64 src/usr.sbin/sicontrol/sicontrol.c From mtm at FreeBSD.org Mon Jun 9 09:08:05 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 9 09:08:12 2008 Subject: cvs commit: src/sbin/rcorder rcorder.8 Message-ID: <200806090908.m59984a2015239@repoman.freebsd.org> mtm 2008-06-09 09:07:58 UTC FreeBSD src repository Modified files: sbin/rcorder rcorder.8 Log: SVN rev 179669 on 2008-06-09 09:07:58Z by mtm Document the misleading nature of the REQUIRE line. The patch in the PR has been heavily edited for style(9) and clarity. Mistakes are mine. PR: bin/124251 Revision Changes Path 1.5 +20 -1 src/sbin/rcorder/rcorder.8 From kib at FreeBSD.org Mon Jun 9 10:32:23 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Mon Jun 9 10:32:26 2008 Subject: cvs commit: src/sys/kern vfs_export.c vfs_mount.c src/sys/sys mount.h Message-ID: <200806091032.m59AWNrd023418@repoman.freebsd.org> kib 2008-06-09 10:31:38 UTC FreeBSD src repository Modified files: sys/kern vfs_export.c vfs_mount.c sys/sys mount.h Log: SVN rev 179670 on 2008-06-09 10:31:38Z by kib Provide the mutual exclusion between the nfs export list modifications and nfs requests processing. Lockmgr lock provides the shared locking for nfs requests, while exclusive mode is used for modifications. The writer starvation is handled by lockmgr too. Reported by: kris, pho, many Based on the submission by: mohan Tested by: pho MFC after: 2 weeks Revision Changes Path 1.343 +4 -0 src/sys/kern/vfs_export.c 1.284 +2 -0 src/sys/kern/vfs_mount.c 1.231 +1 -0 src/sys/sys/mount.h From dfr at FreeBSD.org Mon Jun 9 10:51:57 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 9 10:52:08 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806091051.m59ApvRH025001@repoman.freebsd.org> dfr 2008-06-09 10:51:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/nlm nlm_prot_impl.c Log: SVN rev 179671 on 2008-06-09 10:51:43Z by dfr MFC: r179488 - check for client reboot before executing the lock request Revision Changes Path 1.4.2.5 +10 -8 src/sys/nlm/nlm_prot_impl.c From dfr at FreeBSD.org Mon Jun 9 10:52:59 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 9 10:53:00 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806091052.m59Aqwa9025089@repoman.freebsd.org> dfr 2008-06-09 10:52:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/nlm nlm_prot_impl.c Log: SVN rev 179672 on 2008-06-09 10:52:43Z by dfr MFC: r179488 - check for client reboot before executing the lock request Revision Changes Path 1.6.2.3 +10 -8 src/sys/nlm/nlm_prot_impl.c From dfr at FreeBSD.org Mon Jun 9 10:55:20 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 9 10:55:22 2008 Subject: cvs commit: src/usr.sbin/rpc.lockd lockd.c Message-ID: <200806091055.m59AtJS6025304@repoman.freebsd.org> dfr 2008-06-09 10:54:55 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/rpc.lockd lockd.c Log: SVN rev 179673 on 2008-06-09 10:54:55Z by dfr MFC: r179489 - add a missing call to init_nsm(). Revision Changes Path 1.20.2.4 +1 -0 src/usr.sbin/rpc.lockd/lockd.c From dfr at FreeBSD.org Mon Jun 9 10:56:06 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 9 10:56:09 2008 Subject: cvs commit: src/usr.sbin/rpc.lockd lockd.c Message-ID: <200806091056.m59Au5NF025412@repoman.freebsd.org> dfr 2008-06-09 10:55:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/rpc.lockd lockd.c Log: SVN rev 179674 on 2008-06-09 10:55:59Z by dfr MFC: r179489 - add a missing call to init_nsm(). Revision Changes Path 1.18.8.4 +1 -0 src/usr.sbin/rpc.lockd/lockd.c From dfr at FreeBSD.org Mon Jun 9 12:10:54 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 9 12:10:56 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806091210.m59CArKY033673@repoman.freebsd.org> dfr 2008-06-09 12:10:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/nlm nlm_prot_impl.c Log: SVN rev 179675 on 2008-06-09 12:10:43Z by dfr MFC: r179510 - nlm_global_lock doesn't exist (yet) Pointed out by: kib Revision Changes Path 1.4.2.6 +0 -1 src/sys/nlm/nlm_prot_impl.c From dfr at FreeBSD.org Mon Jun 9 12:11:39 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Mon Jun 9 12:11:47 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806091211.m59CBd1e033793@repoman.freebsd.org> dfr 2008-06-09 12:11:27 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/nlm nlm_prot_impl.c Log: SVN rev 179676 on 2008-06-09 12:11:27Z by dfr MFC: r179510 - nlm_global_lock doesn't exist (yet) Pointed out by: kib Revision Changes Path 1.6.2.4 +0 -1 src/sys/nlm/nlm_prot_impl.c From obrien at FreeBSD.org Mon Jun 9 14:02:05 2008 From: obrien at FreeBSD.org (David O'Brien) Date: Mon Jun 9 14:02:14 2008 Subject: cvs commit: src/sys/conf kern.pre.mk In-Reply-To: <200806090631.m596VLDo080578@repoman.freebsd.org> References: <200806090631.m596VLDo080578@repoman.freebsd.org> Message-ID: <20080609140204.GA70228@dragon.NUXI.org> On Mon, Jun 09, 2008 at 06:31:17AM +0000, John Birrell wrote: > jb 2008-06-09 06:31:17 UTC > FreeBSD src repository > > Modified files: > sys/conf kern.pre.mk > Log: > SVN rev 179665 on 2008-06-09 06:31:17Z by jb > > Remove some sparc-specific stuff from my earlier sun4v work in p4. > It never belonged in current. For those having trouble building a non-SMP sparc64 kernel - this is what broke the build. -- -- David (obrien@FreeBSD.org) From cperciva at FreeBSD.org Mon Jun 9 14:04:11 2008 From: cperciva at FreeBSD.org (Colin Percival) Date: Mon Jun 9 14:04:13 2008 Subject: cvs commit: src/usr.bin/tar util.c Message-ID: <200806091404.m59E4BfQ062819@repoman.freebsd.org> cperciva 2008-06-09 14:03:55 UTC FreeBSD src repository Modified files: usr.bin/tar util.c Log: SVN rev 179677 on 2008-06-09 14:03:55Z by cperciva Rework code to avoid using a pointer after freeing it. Aside from the possibility of memory becoming undereferenceable when it is freed, this change should have no effect on bsdtar behaviour. Found by: Coverity Prevent Revision Changes Path 1.20 +4 -2 src/usr.bin/tar/util.c From cperciva at FreeBSD.org Mon Jun 9 14:41:43 2008 From: cperciva at FreeBSD.org (Colin Percival) Date: Mon Jun 9 14:41:48 2008 Subject: cvs commit: src/usr.bin/tar subst.c Message-ID: <200806091441.m59Efhj5067292@repoman.freebsd.org> cperciva 2008-06-09 14:41:28 UTC FreeBSD src repository Modified files: usr.bin/tar subst.c Log: SVN rev 179678 on 2008-06-09 14:41:28Z by cperciva Make one-bit fields unsigned instead of signed. This has no effect, since they are only tested for zero/nonzero; but it's arguably a bad idea to set a {-1, 0} variable to 1 (as happens in this code). Found by: Coverity Prevent Revision Changes Path 1.3 +1 -1 src/usr.bin/tar/subst.c From cperciva at FreeBSD.org Mon Jun 9 14:45:15 2008 From: cperciva at FreeBSD.org (Colin Percival) Date: Mon Jun 9 14:45:19 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_mtree.c Message-ID: <200806091445.m59EjEO9067576@repoman.freebsd.org> cperciva 2008-06-09 14:44:56 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_mtree.c Log: SVN rev 179679 on 2008-06-09 14:44:56Z by cperciva Don't store a pointer in "next" if we're never going to use it. Fix typo in checking results of strchr. Found by: Coverity Prevent Revision Changes Path 1.7 +2 -2 src/lib/libarchive/archive_read_support_format_mtree.c From dougb at FreeBSD.org Mon Jun 9 16:15:25 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 9 16:15:28 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <484CB927.3050101@freebsd.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> Message-ID: <484D5719.1070907@FreeBSD.org> Tim Kientzle wrote: > This case is a little unusual: by default, I want to > install both 'cpio' programs and symlink the common name > to one of them. I'm not sure how that relates to Jeremie's (valid) point that these flags need to be part of the system in bsd.own.mk. I admit that getting stuff in there was a little intimidating to me at first, but Ruslan has created a nice automated system, and he'll be glad to help you if you need it. > How about this: > WITH_GCPIO (default) - installs gcpio, symlinks 'cpio' unless > WITH_BSDCPIO defined > WITHOUT_GCPIO - does not install gcpio at all > WITH_BSDCPIO (default) - installs bsdcpio, symlinks 'cpio' > WITHOUT_BSDCPIO - does not install bsdcpio at all If by "not install" you mean "build or install" then I'm with you, but I think it might be a little complicated. I'm not sure why you want to build and install both by default (except maybe to ease people's concerns in the interim period) but if it were me, in HEAD I would do this: WITH_BSDCPIO (default) build, install, and symlink to cpio WITHOUT_GCPIO (default) do not build or install Then if WITHOUT_BSDCPIO is defined you don't build or install it, and if WITH_GCPIO is defined you build, install, and symlink it, regardless of the state of the BSDCPIO knob. When this gets MFC'ed I would just flip the defaults. FWIW, Doug -- This .signature sanitized for your protection From yanefbsd at gmail.com Mon Jun 9 16:17:56 2008 From: yanefbsd at gmail.com (Garrett Cooper) Date: Mon Jun 9 16:18:01 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <484D5719.1070907@FreeBSD.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> <484D5719.1070907@FreeBSD.org> Message-ID: <7d6fde3d0806090917h2ed0799dv6b7044a959e039af@mail.gmail.com> On Mon, Jun 9, 2008 at 9:15 AM, Doug Barton wrote: > Tim Kientzle wrote: >> >> This case is a little unusual: by default, I want to >> install both 'cpio' programs and symlink the common name >> to one of them. > > I'm not sure how that relates to Jeremie's (valid) point that these flags > need to be part of the system in bsd.own.mk. I admit that getting stuff in > there was a little intimidating to me at first, but Ruslan has created a > nice automated system, and he'll be glad to help you if you need it. > >> How about this: >> WITH_GCPIO (default) - installs gcpio, symlinks 'cpio' unless >> WITH_BSDCPIO defined >> WITHOUT_GCPIO - does not install gcpio at all >> WITH_BSDCPIO (default) - installs bsdcpio, symlinks 'cpio' >> WITHOUT_BSDCPIO - does not install bsdcpio at all > > If by "not install" you mean "build or install" then I'm with you, but I > think it might be a little complicated. I'm not sure why you want to build > and install both by default (except maybe to ease people's concerns in the > interim period) but if it were me, in HEAD I would do this: > > WITH_BSDCPIO (default) build, install, and symlink to cpio > WITHOUT_GCPIO (default) do not build or install > > Then if WITHOUT_BSDCPIO is defined you don't build or install it, and if > WITH_GCPIO is defined you build, install, and symlink it, regardless of the > state of the BSDCPIO knob. > > When this gets MFC'ed I would just flip the defaults. > > > FWIW, > > Doug I agree completely with Doug. -Garrett From dougb at FreeBSD.org Mon Jun 9 16:29:28 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 9 16:29:40 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <7d6fde3d0806090917h2ed0799dv6b7044a959e039af@mail.gmail.com> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> <484D5719.1070907@FreeBSD.org> <7d6fde3d0806090917h2ed0799dv6b7044a959e039af@mail.gmail.com> Message-ID: <484D5A65.2060603@FreeBSD.org> Garrett Cooper wrote: > I agree completely with Doug. Wow! No one EVER says that. I think I'm going to print this one out and frame it. :) Doug -- This .signature sanitized for your protection From marcel at FreeBSD.org Mon Jun 9 16:43:22 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 9 16:43:29 2008 Subject: cvs commit: src/sbin Makefile Message-ID: <200806091643.m59GhLmP079564@repoman.freebsd.org> marcel 2008-06-09 16:42:55 UTC FreeBSD src repository Modified files: sbin Makefile Log: SVN rev 179680 on 2008-06-09 16:42:55Z by marcel Disconnect gpt(8) from the build. gpt(8) is replaced by gpart(8). Revision Changes Path 1.170 +0 -1 src/sbin/Makefile From marcel at FreeBSD.org Mon Jun 9 17:59:53 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 9 18:00:01 2008 Subject: cvs commit: src/sys/geom geom.h geom_subr.c Message-ID: <200806091759.m59Hxrb5097189@repoman.freebsd.org> marcel 2008-06-09 17:59:44 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/geom geom.h geom_subr.c Log: SVN rev 179681 on 2008-06-09 17:59:44Z by marcel MFC rev 177509: Add g_retaste(). Revision Changes Path 1.100.2.1 +1 -0 src/sys/geom/geom.h 1.91.2.1 +60 -0 src/sys/geom/geom_subr.c From marcel at FreeBSD.org Mon Jun 9 18:03:55 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 9 18:04:06 2008 Subject: cvs commit: src/sys/geom geom_subr.c Message-ID: <200806091803.m59I3tUw098644@repoman.freebsd.org> marcel 2008-06-09 18:03:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/geom geom_subr.c Log: SVN rev 179682 on 2008-06-09 18:03:48Z by marcel MFC rev 177681: Wither GEOMs while retasting. Revision Changes Path 1.91.2.2 +7 -1 src/sys/geom/geom_subr.c From jkim at FreeBSD.org Mon Jun 9 18:41:42 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Mon Jun 9 18:41:45 2008 Subject: cvs commit: src/usr.sbin/tzsetup tzsetup.c Message-ID: <200806091841.m59Iff2k005637@repoman.freebsd.org> jkim 2008-06-09 18:41:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/tzsetup tzsetup.c Log: SVN rev 179685 on 2008-06-09 18:41:37Z by jkim MFC: r179497 Fix a crash when Arctic Ocean is selected. Revision Changes Path 1.26.18.1 +2 -3 src/usr.sbin/tzsetup/tzsetup.c From jkim at FreeBSD.org Mon Jun 9 18:45:27 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Mon Jun 9 18:45:31 2008 Subject: cvs commit: src/usr.sbin/tzsetup tzsetup.c Message-ID: <200806091845.m59IjR3L006409@repoman.freebsd.org> jkim 2008-06-09 18:45:21 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/tzsetup tzsetup.c Log: SVN rev 179686 on 2008-06-09 18:45:21Z by jkim MFC: r179497 Fix a crash when Arctic Ocean is selected. Revision Changes Path 1.26.8.1 +2 -3 src/usr.sbin/tzsetup/tzsetup.c From marcel at FreeBSD.org Mon Jun 9 18:51:44 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 9 18:51:47 2008 Subject: cvs commit: src/sys/geom/part g_part.c g_part.h g_part_apm.c g_part_bsd.c g_part_gpt.c g_part_if.m g_part_mbr.c g_part_pc98.c g_part_vtoc8.c Message-ID: <200806091851.m59IphNH007327@repoman.freebsd.org> marcel 2008-06-09 18:51:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/geom/part g_part.c g_part.h g_part_apm.c g_part_bsd.c g_part_gpt.c g_part_if.m g_part_mbr.c g_part_vtoc8.c Added files: (Branch: RELENG_7) sys/geom/part g_part_pc98.c Log: SVN rev 179687 on 2008-06-09 18:51:30Z by marcel MFC r177223 through r179684: o Redefine G_PART_SCHEME_DECLARE() from populating a private linker set to declaring a proper module. o Add support for PC-9800 partition tables. o Change the order from SI_ORDER_FIRST to SI_ORDER_ANY (within SI_SUB_DRIVERS) to avoid loading schemes before all the GEOM classes have been loaded and initialized. o Add the bootcode verb for installing boot code. o Implement the G_PART_DUMPCONF method for all 6 schemes. Revision Changes Path 1.9.2.4 +181 -23 src/sys/geom/part/g_part.c 1.3.2.4 +19 -1 src/sys/geom/part/g_part.h 1.3.2.2 +18 -1 src/sys/geom/part/g_part_apm.c 1.3.2.2 +18 -1 src/sys/geom/part/g_part_bsd.c 1.3.2.3 +64 -22 src/sys/geom/part/g_part_gpt.c 1.2.2.1 +6 -0 src/sys/geom/part/g_part_if.m 1.2.2.2 +31 -1 src/sys/geom/part/g_part_mbr.c 1.3.2.1 +459 -0 src/sys/geom/part/g_part_pc98.c (new) 1.1.2.2 +20 -1 src/sys/geom/part/g_part_vtoc8.c From marcel at FreeBSD.org Mon Jun 9 19:25:44 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 9 19:25:47 2008 Subject: cvs commit: src/sys/modules/geom Makefile src/sys/modules/geom/geom_part Makefile src/sys/modules/geom/geom_part/geom_part_apm Makefile src/sys/modules/geom/geom_part/geom_part_bsd Makefile src/sys/modules/geom/geom_part/geom_part_gpt Makefile src/sys/modules/geom/geom_part/geom_part_mbr Makefile ... Message-ID: <200806091925.m59JPiWb012802@repoman.freebsd.org> marcel 2008-06-09 19:25:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/modules/geom Makefile Added files: (Branch: RELENG_7) sys/modules/geom/geom_part Makefile sys/modules/geom/geom_part/geom_part_apm Makefile sys/modules/geom/geom_part/geom_part_bsd Makefile sys/modules/geom/geom_part/geom_part_gpt Makefile sys/modules/geom/geom_part/geom_part_mbr Makefile sys/modules/geom/geom_part/geom_part_pc98 Makefile sys/modules/geom/geom_part/geom_part_vtoc8 Makefile Log: SVN rev 179688 on 2008-06-09 19:25:04Z by marcel MFC r176724 through r179687 (geom_part): Module support Revision Changes Path 1.19.2.3 +1 -0 src/sys/modules/geom/Makefile 1.3.2.1 +10 -0 src/sys/modules/geom/geom_part/Makefile (new) 1.1.2.1 +12 -0 src/sys/modules/geom/geom_part/geom_part_apm/Makefile (new) 1.1.2.1 +12 -0 src/sys/modules/geom/geom_part/geom_part_bsd/Makefile (new) 1.1.2.1 +12 -0 src/sys/modules/geom/geom_part/geom_part_gpt/Makefile (new) 1.1.2.1 +12 -0 src/sys/modules/geom/geom_part/geom_part_mbr/Makefile (new) 1.1.2.1 +12 -0 src/sys/modules/geom/geom_part/geom_part_pc98/Makefile (new) 1.1.2.1 +12 -0 src/sys/modules/geom/geom_part/geom_part_vtoc8/Makefile (new) From brooks at FreeBSD.org Mon Jun 9 20:03:46 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Mon Jun 9 20:03:48 2008 Subject: cvs commit: src/sbin/dhclient dhclient-script Message-ID: <200806092003.m59K3jKl016913@repoman.freebsd.org> brooks 2008-06-09 20:03:35 UTC FreeBSD src repository Modified files: sbin/dhclient dhclient-script Log: SVN rev 179689 on 2008-06-09 20:03:35Z by brooks Use the -n flag to route(8) when calling "route get". Otherwise we hang for a long time if we get a lease, but DNS isn't working. MFC after: 1 week Revision Changes Path 1.20 +1 -1 src/sbin/dhclient/dhclient-script From marcel at FreeBSD.org Mon Jun 9 21:39:01 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 9 21:39:03 2008 Subject: cvs commit: src UPDATING Message-ID: <200806092139.m59Ld1bY045617@repoman.freebsd.org> marcel 2008-06-09 21:33:57 UTC FreeBSD src repository Modified files: . UPDATING Log: SVN rev 179691 on 2008-06-09 21:33:57Z by marcel Note removal of gpt(8). Revision Changes Path 1.532 +4 -0 src/UPDATING From marcel at FreeBSD.org Mon Jun 9 21:44:14 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 9 21:44:17 2008 Subject: cvs commit: src ObsoleteFiles.inc Message-ID: <200806092144.m59LiDpO046049@repoman.freebsd.org> marcel 2008-06-09 21:36:46 UTC FreeBSD src repository Modified files: . ObsoleteFiles.inc Log: SVN rev 179692 on 2008-06-09 21:36:46Z by marcel Removal of gpt(8) and manpage. Revision Changes Path 1.142 +3 -0 src/ObsoleteFiles.inc From wkoszek at FreeBSD.org Mon Jun 9 21:52:00 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Mon Jun 9 21:52:06 2008 Subject: cvs commit: src/sys/arm/at91 if_ate.c Message-ID: <200806092151.m59LpxvH046662@repoman.freebsd.org> wkoszek 2008-06-09 21:51:48 UTC FreeBSD src repository Modified files: sys/arm/at91 if_ate.c Log: SVN rev 179693 on 2008-06-09 21:51:48Z by wkoszek Since we create a DMA tag "mtag" for TX map with bus_dmamap_create(), we must synchronize such a map against "mtag" with bus_dmamap_sync(), not the tag designated for RX map. Fix it. Approved by: cognet Revision Changes Path 1.24 +2 -2 src/sys/arm/at91/if_ate.c From wkoszek at FreeBSD.org Mon Jun 9 22:13:32 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Mon Jun 9 22:13:36 2008 Subject: cvs commit: src/share/man/man9 bus_dma.9 Message-ID: <200806092213.m59MDWZA049560@repoman.freebsd.org> wkoszek 2008-06-09 22:13:14 UTC FreeBSD src repository Modified files: share/man/man9 bus_dma.9 Log: SVN rev 179694 on 2008-06-09 22:13:14Z by wkoszek Mark "highaddr" and "lowaddr" appropriately. Change "dma" to "DMA". Fix spelling mistage found by rodrigc@. Reviewed by: rodrigc, danger Revision Changes Path 1.41 +18 -9 src/share/man/man9/bus_dma.9 From davidch at FreeBSD.org Tue Jun 10 02:19:22 2008 From: davidch at FreeBSD.org (David Christensen) Date: Tue Jun 10 02:19:24 2008 Subject: cvs commit: src/sys/dev/bce if_bce.c if_bcereg.h Message-ID: <200806100219.m5A2JM9H083748@repoman.freebsd.org> davidch 2008-06-10 02:19:11 UTC FreeBSD src repository Modified files: sys/dev/bce if_bce.c if_bcereg.h Log: SVN rev 179695 on 2008-06-10 02:19:11Z by davidch - Fixed kern/123696 by increasing firmware timeout value from 100 to 1000. - Fixed a problem on i386 architecture when using split header/jumbo frame firmware caused by hardware alignment requirements. - Added #define BCE_USE_SPLIT_HEADER to allow the feature to be enabled/ disabled. Enabled by default. PR: kern/123696 MFC after: 2 weeks Revision Changes Path 1.41 +185 -49 src/sys/dev/bce/if_bce.c 1.19 +8 -5 src/sys/dev/bce/if_bcereg.h From kevlo at FreeBSD.org Tue Jun 10 02:41:54 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Tue Jun 10 02:41:57 2008 Subject: cvs commit: src/sys/dev/usb ehci.c ehcivar.h Message-ID: <200806100241.m5A2frsA085664@repoman.freebsd.org> kevlo 2008-06-10 02:41:36 UTC FreeBSD src repository Modified files: sys/dev/usb ehci.c ehcivar.h Log: SVN rev 179696 on 2008-06-10 02:41:36Z by kevlo >From NetBSD: Remove the code which disables port status change interrupts for 1s when one occured -- this makes that events get lost or delayed until the next change. Obtained from: NetBSD Revision Changes Path 1.64 +0 -45 src/sys/dev/usb/ehci.c 1.18 +0 -1 src/sys/dev/usb/ehcivar.h From marcel at FreeBSD.org Tue Jun 10 03:03:10 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Tue Jun 10 03:03:14 2008 Subject: cvs commit: src/sbin/geom/misc subr.c subr.h Message-ID: <200806100303.m5A33A3o088708@repoman.freebsd.org> marcel 2008-06-10 03:02:51 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/geom/misc subr.c subr.h Log: SVN rev 179697 on 2008-06-10 03:02:51Z by marcel MFC rev 179628: Add gctl_delete_param() and gctl_has_param(). Revision Changes Path 1.7.2.1 +44 -0 src/sbin/geom/misc/subr.c 1.8.2.1 +3 -0 src/sbin/geom/misc/subr.h From marcel at FreeBSD.org Tue Jun 10 03:19:42 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Tue Jun 10 03:19:45 2008 Subject: cvs commit: src/sbin/geom/class/part geom_part.c gpart.8 Message-ID: <200806100319.m5A3Jfq9090000@repoman.freebsd.org> marcel 2008-06-10 03:19:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sbin/geom/class/part geom_part.c gpart.8 Log: SVN rev 179699 on 2008-06-10 03:19:33Z by marcel MFC revisions 172857 through 179688: o Allow building a rescue version (not applicable yet) o Add the bootcode verb for installing bootcode. o Various manpage updates. Revision Changes Path 1.1.2.2 +177 -9 src/sbin/geom/class/part/geom_part.c 1.1.2.2 +48 -3 src/sbin/geom/class/part/gpart.8 From kevlo at FreeBSD.org Tue Jun 10 03:44:28 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Tue Jun 10 03:44:31 2008 Subject: cvs commit: src/sys/arm/xscale/pxa files.pxa pxa_space.c Message-ID: <200806100344.m5A3iSkk092053@repoman.freebsd.org> kevlo 2008-06-10 03:44:14 UTC FreeBSD src repository Modified files: sys/arm/xscale/pxa files.pxa pxa_space.c Log: SVN rev 179700 on 2008-06-10 03:44:14Z by kevlo Pull all the code to deal with bus space methods into a shared set of routines. Revision Changes Path 1.2 +1 -0 src/sys/arm/xscale/pxa/files.pxa 1.2 +6 -102 src/sys/arm/xscale/pxa/pxa_space.c From kevlo at FreeBSD.org Tue Jun 10 06:06:31 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Tue Jun 10 06:06:33 2008 Subject: cvs commit: src/sys/arm/xscale/pxa pxa_obio.c Message-ID: <200806100606.m5A66Vn4017231@repoman.freebsd.org> kevlo 2008-06-10 06:06:15 UTC FreeBSD src repository Modified files: sys/arm/xscale/pxa pxa_obio.c Log: SVN rev 179701 on 2008-06-10 06:06:15Z by kevlo Add the pxa_teardown_intr() bus method function to de-associate the interrupt handler Revision Changes Path 1.2 +20 -13 src/sys/arm/xscale/pxa/pxa_obio.c From flz at FreeBSD.org Tue Jun 10 09:55:38 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 10 09:55:41 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create create.h main.c perform.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... Message-ID: <200806100955.m5A9tcJT049291@repoman.freebsd.org> flz 2008-06-10 09:55:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/pkg_install/add main.c pkg_add.1 usr.sbin/pkg_install/create create.h main.c perform.c pkg_create.1 usr.sbin/pkg_install/delete main.c pkg_delete.1 usr.sbin/pkg_install/info main.c pkg_info.1 usr.sbin/pkg_install/lib lib.h usr.sbin/pkg_install/updating main.c pkg_updating.1 usr.sbin/pkg_install/version main.c pkg_version.1 Log: SVN rev 179702 on 2008-06-10 09:55:25Z by flz Synchronize pkg_install with HEAD (20080530) r179352,179366: don't regenerate packages r179433: add long options Revision Changes Path 1.72.2.3 +24 -5 src/usr.sbin/pkg_install/add/main.c 1.71.2.3 +14 -14 src/usr.sbin/pkg_install/add/pkg_add.1 1.26.2.1 +1 -0 src/usr.sbin/pkg_install/create/create.h 1.40.2.2 +30 -6 src/usr.sbin/pkg_install/create/main.c 1.82.2.1 +13 -0 src/usr.sbin/pkg_install/create/perform.c 1.70.2.2 +33 -15 src/usr.sbin/pkg_install/create/pkg_create.1 1.28.2.2 +22 -4 src/usr.sbin/pkg_install/delete/main.c 1.31.10.2 +13 -13 src/usr.sbin/pkg_install/delete/pkg_delete.1 1.50.2.3 +23 -5 src/usr.sbin/pkg_install/info/main.c 1.60.2.2 +17 -15 src/usr.sbin/pkg_install/info/pkg_info.1 1.60.2.3 +1 -1 src/usr.sbin/pkg_install/lib/lib.h 1.2.2.2 +12 -3 src/usr.sbin/pkg_install/updating/main.c 1.3.2.2 +11 -5 src/usr.sbin/pkg_install/updating/pkg_updating.1 1.7.2.2 +20 -5 src/usr.sbin/pkg_install/version/main.c 1.36.2.3 +10 -10 src/usr.sbin/pkg_install/version/pkg_version.1 From flz at FreeBSD.org Tue Jun 10 09:56:26 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 10 09:56:30 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c pkg_add.1 src/usr.sbin/pkg_install/create create.h main.c perform.c pkg_create.1 src/usr.sbin/pkg_install/delete main.c pkg_delete.1 src/usr.sbin/pkg_install/info main.c pkg_info.1 ... Message-ID: <200806100956.m5A9uQlh049515@repoman.freebsd.org> flz 2008-06-10 09:56:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/pkg_install/add main.c pkg_add.1 usr.sbin/pkg_install/create create.h main.c perform.c pkg_create.1 usr.sbin/pkg_install/delete main.c pkg_delete.1 usr.sbin/pkg_install/info main.c pkg_info.1 usr.sbin/pkg_install/lib lib.h usr.sbin/pkg_install/updating main.c pkg_updating.1 usr.sbin/pkg_install/version main.c pkg_version.1 Log: SVN rev 179703 on 2008-06-10 09:56:11Z by flz Synchronize pkg_install with HEAD (20080530) r179352,179366: don't regenerate packages r179433: add long options Revision Changes Path 1.61.2.12 +24 -5 src/usr.sbin/pkg_install/add/main.c 1.66.2.6 +14 -14 src/usr.sbin/pkg_install/add/pkg_add.1 1.25.2.3 +1 -0 src/usr.sbin/pkg_install/create/create.h 1.36.2.4 +30 -6 src/usr.sbin/pkg_install/create/main.c 1.80.2.4 +13 -0 src/usr.sbin/pkg_install/create/perform.c 1.68.2.4 +33 -15 src/usr.sbin/pkg_install/create/pkg_create.1 1.26.8.3 +22 -4 src/usr.sbin/pkg_install/delete/main.c 1.31.2.2 +13 -13 src/usr.sbin/pkg_install/delete/pkg_delete.1 1.47.2.5 +23 -5 src/usr.sbin/pkg_install/info/main.c 1.54.2.4 +17 -15 src/usr.sbin/pkg_install/info/pkg_info.1 1.56.2.5 +1 -1 src/usr.sbin/pkg_install/lib/lib.h 1.2.4.2 +12 -3 src/usr.sbin/pkg_install/updating/main.c 1.3.4.2 +11 -5 src/usr.sbin/pkg_install/updating/pkg_updating.1 1.5.2.3 +20 -5 src/usr.sbin/pkg_install/version/main.c 1.34.2.3 +10 -10 src/usr.sbin/pkg_install/version/pkg_version.1 From jasone at FreeBSD.org Tue Jun 10 15:46:33 2008 From: jasone at FreeBSD.org (Jason Evans) Date: Tue Jun 10 15:46:37 2008 Subject: cvs commit: src/lib/libc/stdlib malloc.c Message-ID: <200806101546.m5AFkX0U094880@repoman.freebsd.org> jasone 2008-06-10 15:46:18 UTC FreeBSD src repository Modified files: lib/libc/stdlib malloc.c Log: SVN rev 179704 on 2008-06-10 15:46:18Z by jasone In the error path through base_alloc(), release base_mtx [1]. Fix bit vector initialization for run headers. Submitted by: [1] Mike Schuster Revision Changes Path 1.172 +7 -3 src/lib/libc/stdlib/malloc.c From ps at FreeBSD.org Tue Jun 10 17:56:32 2008 From: ps at FreeBSD.org (Paul Saab) Date: Tue Jun 10 17:56:37 2008 Subject: cvs commit: src/share/man/man4 ciss.4 src/sys/dev/ciss ciss.c Message-ID: <200806101756.m5AHuWbK016841@repoman.freebsd.org> ps 2008-06-10 17:51:51 UTC FreeBSD src repository Modified files: share/man/man4 ciss.4 sys/dev/ciss ciss.c Log: SVN rev 179705 on 2008-06-10 17:51:51Z by ps Add support for the P212, P410, P410i, P411, and P812 HP Smart Array controllers. Submitted by: Scott Benesh at HP Revision Changes Path 1.16 +10 -0 src/share/man/man4/ciss.4 1.89 +5 -0 src/sys/dev/ciss/ciss.c From jhb at FreeBSD.org Tue Jun 10 17:59:51 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 10 17:59:54 2008 Subject: cvs commit: src/sys/dev/tx if_tx.c if_txvar.h Message-ID: <200806101759.m5AHxpUv017103@repoman.freebsd.org> jhb 2008-06-10 17:59:43 UTC FreeBSD src repository Modified files: sys/dev/tx if_tx.c if_txvar.h Log: SVN rev 179706 on 2008-06-10 17:59:43Z by jhb Make tx(4) MPSAFE - Add a mutex to the softc to protect the softc and device hardware. - Use a private timer to implement a watchdog for tx timeouts and drive the timer for auto negotiation. - Use bus_foo() rather than bus_space_foo() and remove the bus space tag & handle from the softc. - Call bus_setup_intr() after ether_ifattach(). Tested by: Florian Smeets flo of kasimir.com Revision Changes Path 1.98 +121 -107 src/sys/dev/tx/if_tx.c 1.20 +13 -10 src/sys/dev/tx/if_txvar.h From ps at FreeBSD.org Tue Jun 10 18:16:40 2008 From: ps at FreeBSD.org (Paul Saab) Date: Tue Jun 10 18:16:42 2008 Subject: cvs commit: src/share/man/man4 ciss.4 src/sys/dev/ciss ciss.c Message-ID: <200806101816.m5AIGe4v030321@repoman.freebsd.org> ps 2008-06-10 18:11:14 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 ciss.4 sys/dev/ciss ciss.c Log: SVN rev 179707 on 2008-06-10 18:11:14Z by ps MFC: r179705 Add support for the P212, P410, P410i, P411, and P812 HP Smart Array controllers. Revision Changes Path 1.14.2.2 +10 -0 src/share/man/man4/ciss.4 1.81.2.5 +5 -0 src/sys/dev/ciss/ciss.c From ps at FreeBSD.org Tue Jun 10 18:34:44 2008 From: ps at FreeBSD.org (Paul Saab) Date: Tue Jun 10 18:34:46 2008 Subject: cvs commit: src/share/man/man4 ciss.4 src/sys/dev/ciss ciss.c Message-ID: <200806101834.m5AIYipc031794@repoman.freebsd.org> ps 2008-06-10 18:31:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 ciss.4 sys/dev/ciss ciss.c Log: SVN rev 179708 on 2008-06-10 18:31:19Z by ps MFC: r179705 Add support for the P212, P410, P410i, P411, and P812 HP Smart Array controllers. Revision Changes Path 1.11.2.4 +10 -0 src/share/man/man4/ciss.4 1.64.2.6 +5 -0 src/sys/dev/ciss/ciss.c From ps at FreeBSD.org Tue Jun 10 18:44:33 2008 From: ps at FreeBSD.org (Paul Saab) Date: Tue Jun 10 18:44:37 2008 Subject: cvs commit: src/share/man/man4 ciss.4 src/sys/dev/ciss ciss.c Message-ID: <200806101844.m5AIiXE2033980@repoman.freebsd.org> ps 2008-06-10 18:40:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) share/man/man4 ciss.4 sys/dev/ciss ciss.c Log: SVN rev 179709 on 2008-06-10 18:40:38Z by ps MFC: r179705 Add support for the P212, P410, P410i, P411, and P812 HP Smart Array controllers. Revision Changes Path 1.10.2.2 +17 -1 src/share/man/man4/ciss.4 1.56.2.7 +5 -0 src/sys/dev/ciss/ciss.c From jhb at FreeBSD.org Tue Jun 10 18:47:58 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 10 18:48:00 2008 Subject: cvs commit: src/libexec/rpc.rstatd Makefile rstat_proc.c Message-ID: <200806101847.m5AIlwAZ034271@repoman.freebsd.org> jhb 2008-06-10 18:47:43 UTC FreeBSD src repository Modified files: libexec/rpc.rstatd Makefile rstat_proc.c Log: SVN rev 179710 on 2008-06-10 18:47:43Z by jhb Use sysctl to fetch stats from the kernel instead of reading variables directly via libkvm. PR: kern/122875, bin/123014 Tested by: Danny Braniss danny cs.huji.ac.il MFC after: 1 week Revision Changes Path 1.9 +2 -2 src/libexec/rpc.rstatd/Makefile 1.26 +17 -46 src/libexec/rpc.rstatd/rstat_proc.c From ps at FreeBSD.org Tue Jun 10 18:50:30 2008 From: ps at FreeBSD.org (Paul Saab) Date: Tue Jun 10 18:50:31 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c Message-ID: <200806101850.m5AIoTOq034525@repoman.freebsd.org> ps 2008-06-10 18:50:14 UTC FreeBSD src repository Modified files: (Branch: RELENG_4) sys/dev/ciss ciss.c Log: SVN rev 179711 on 2008-06-10 18:50:14Z by ps Add the missing support for the P800 Revision Changes Path 1.2.2.24 +1 -0 src/sys/dev/ciss/ciss.c From ps at FreeBSD.org Tue Jun 10 18:57:22 2008 From: ps at FreeBSD.org (Paul Saab) Date: Tue Jun 10 18:57:24 2008 Subject: cvs commit: src/share/man/man4 ciss.4 src/sys/dev/ciss ciss.c Message-ID: <200806101857.m5AIvMP3035053@repoman.freebsd.org> ps 2008-06-10 18:51:05 UTC FreeBSD src repository Modified files: (Branch: RELENG_4) share/man/man4 ciss.4 sys/dev/ciss ciss.c Log: SVN rev 179712 on 2008-06-10 18:51:05Z by ps MFC: r179705 Add support for the P212, P410, P410i, P411, and P812 HP Smart Array controllers. Revision Changes Path 1.2.2.6 +26 -0 src/share/man/man4/ciss.4 1.2.2.25 +5 -0 src/sys/dev/ciss/ciss.c From imp at FreeBSD.org Tue Jun 10 19:31:28 2008 From: imp at FreeBSD.org (Warner Losh) Date: Tue Jun 10 19:31:34 2008 Subject: cvs commit: src/sys/dev/usb uplcom.c Message-ID: <200806101931.m5AJVSwa038959@repoman.freebsd.org> imp 2008-06-10 19:31:09 UTC FreeBSD src repository Modified files: sys/dev/usb uplcom.c Log: SVN rev 179713 on 2008-06-10 19:31:09Z by imp Add Belkin F5U257 to the mix. Submitted by: wilko@ Revision Changes Path 1.52 +2 -0 src/sys/dev/usb/uplcom.c From imp at FreeBSD.org Tue Jun 10 19:33:45 2008 From: imp at FreeBSD.org (Warner Losh) Date: Tue Jun 10 19:33:47 2008 Subject: cvs commit: src/share/man/man4 uplcom.4 Message-ID: <200806101933.m5AJXjU2039176@repoman.freebsd.org> imp 2008-06-10 19:33:31 UTC FreeBSD src repository Modified files: share/man/man4 uplcom.4 Log: SVN rev 179714 on 2008-06-10 19:33:31Z by imp Add Belkin F5U257 to the mix. Revision Changes Path 1.17 +2 -0 src/share/man/man4/uplcom.4 From simon at FreeBSD.org Tue Jun 10 20:05:47 2008 From: simon at FreeBSD.org (Simon L. Nielsen) Date: Tue Jun 10 20:05:50 2008 Subject: cvs commit: src/sys/security/audit audit_worker.c Message-ID: <200806102005.m5AK5k3e042876@repoman.freebsd.org> simon 2008-06-10 20:05:32 UTC FreeBSD src repository Modified files: sys/security/audit audit_worker.c Log: SVN rev 179715 on 2008-06-10 20:05:32Z by simon When the file-system containing the audit log file is running low on disk space a warning is printed. Make this warning a bit more informative. Approved by: rwatson Revision Changes Path 1.26 +3 -1 src/sys/security/audit/audit_worker.c From gonzo at FreeBSD.org Tue Jun 10 20:55:19 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Tue Jun 10 20:55:22 2008 Subject: cvs commit: src/sys/kern uipc_sem.c Message-ID: <200806102055.m5AKtJVT056104@repoman.freebsd.org> gonzo 2008-06-10 20:55:10 UTC FreeBSD src repository Modified files: sys/kern uipc_sem.c Log: SVN rev 179716 on 2008-06-10 20:55:10Z by gonzo Keep proper track of nsegs counter: sem_free is called for all allocated semaphores, so it's wrong to increase it conditionally, in this case for every over-the-limit semaphore nsegs is decreased without being previously increased. PR: kern/123685 Approved by: cognet (mentor) Revision Changes Path 1.32 +3 -4 src/sys/kern/uipc_sem.c From sos at FreeBSD.org Wed Jun 11 06:45:15 2008 From: sos at FreeBSD.org (Søren Schmidt) Date: Wed Jun 11 06:45:18 2008 Subject: cvs commit: src/sys/dev/ata ata-all.c ata-pci.c Message-ID: <200806110645.m5B6jE7Q050794@repoman.freebsd.org> sos 2008-06-11 06:44:58 UTC FreeBSD src repository Modified files: sys/dev/ata ata-all.c ata-pci.c Log: SVN rev 179717 on 2008-06-11 06:44:58Z by sos Rearrange how to call dma.alloc() so that we have resources alloc'd when need but also late enough to know how many to create. Revision Changes Path 1.289 +4 -0 src/sys/dev/ata/ata-all.c 1.128 +1 -6 src/sys/dev/ata/ata-pci.c From benno at FreeBSD.org Wed Jun 11 06:54:00 2008 From: benno at FreeBSD.org (Benno Rice) Date: Wed Jun 11 06:54:03 2008 Subject: cvs commit: src/sys/dev/smc if_smc.c if_smcvar.h Message-ID: <200806110654.m5B6s08e051439@repoman.freebsd.org> benno 2008-06-11 06:53:55 UTC FreeBSD src repository Modified files: sys/dev/smc if_smc.c if_smcvar.h Log: SVN rev 179718 on 2008-06-11 06:53:55Z by benno Convert bus_space_{read,write}_* calls to bus_{read,write}_* calls. Suggested by: jhb Revision Changes Path 1.2 +17 -25 src/sys/dev/smc/if_smc.c 1.2 +1 -5 src/sys/dev/smc/if_smcvar.h From benno at FreeBSD.org Wed Jun 11 07:26:16 2008 From: benno at FreeBSD.org (Benno Rice) Date: Wed Jun 11 07:26:20 2008 Subject: cvs commit: src/sys/dev/smc if_smc.c if_smcvar.h Message-ID: <200806110726.m5B7QGVH055137@repoman.freebsd.org> benno 2008-06-11 07:26:02 UTC FreeBSD src repository Modified files: sys/dev/smc if_smc.c if_smcvar.h Log: SVN rev 179719 on 2008-06-11 07:26:02Z by benno Switch to using a normal mutex instead of a spin mutex. We still use the interrupt filter due to performance problems that show up if we don't. The main problem seen is that, due to the interrupt being edge triggered, we occasionally miss interrupts which leads us to not notice that we can transmit more packets. Using the new approach, which just schedules a task on a taskqueue, we are guaranteed to have the task run even if the interrupt arrived while we were already executing. If we were to use an ithread the system would mask the interrupt while the handler is run and we'd miss interrupts. Revision Changes Path 1.3 +29 -15 src/sys/dev/smc/if_smc.c 1.3 +1 -7 src/sys/dev/smc/if_smcvar.h From weongyo at FreeBSD.org Wed Jun 11 07:55:19 2008 From: weongyo at FreeBSD.org (Weongyo Jeong) Date: Wed Jun 11 07:55:24 2008 Subject: cvs commit: src/sys/compat/ndis subr_ndis.c Message-ID: <200806110755.m5B7tIYC057463@repoman.freebsd.org> weongyo 2008-06-11 07:55:07 UTC FreeBSD src repository Modified files: sys/compat/ndis subr_ndis.c Log: SVN rev 179720 on 2008-06-11 07:55:07Z by weongyo fix a page fault that it occurred during ifp is NULL. This bug happens when NDIS driver's initialization is failed and NDIS driver's trying to call NdisWriteErrorLogEntry(). Revision Changes Path 1.113 +2 -2 src/sys/compat/ndis/subr_ndis.c From sos at FreeBSD.org Wed Jun 11 08:48:36 2008 From: sos at FreeBSD.org (Søren Schmidt) Date: Wed Jun 11 08:48:40 2008 Subject: cvs commit: src/sys/dev/ata ata-chipset.c Message-ID: <200806110848.m5B8ma7C071993@repoman.freebsd.org> sos 2008-06-11 08:48:25 UTC FreeBSD src repository Modified files: sys/dev/ata ata-chipset.c Log: SVN rev 179721 on 2008-06-11 08:48:25Z by sos Wait up to 1S for the TFD data to signal un-busy before fetching the signature. This at least helps a few slow devices out there. Submitted by: Andrey V. Elsukov Revision Changes Path 1.220 +3 -3 src/sys/dev/ata/ata-chipset.c From h.schmalzbauer at omnisec.de Wed Jun 11 11:22:37 2008 From: h.schmalzbauer at omnisec.de (Harald Schmalzbauer) Date: Wed Jun 11 11:22:42 2008 Subject: cvs commit: src/sbin Makefile In-Reply-To: <200806091643.m59GhLmP079564@repoman.freebsd.org> References: <200806091643.m59GhLmP079564@repoman.freebsd.org> Message-ID: <484FB191.3020907@omnisec.de> Marcel Moolenaar schrieb am 09.06.2008 18:42 (localtime): > marcel 2008-06-09 16:42:55 UTC > > FreeBSD src repository > > Modified files: > sbin Makefile > Log: > SVN rev 179680 on 2008-06-09 16:42:55Z by marcel > > Disconnect gpt(8) from the build. gpt(8) is replaced by gpart(8). I highly appreciate the gpart work, but what I'm missing (in RELENG_7) is support for "labels" (gpt -r show -l ad99). Was that added to HEAD? Any plans to MFC? Thanks a lot, -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/20080611/5fcf0ba1/signature.pgp From kib at FreeBSD.org Wed Jun 11 12:46:25 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Wed Jun 11 12:46:30 2008 Subject: cvs commit: src/sys/fs/cd9660 cd9660_vnops.c Message-ID: <200806111246.m5BCkP4x009855@repoman.freebsd.org> kib 2008-06-11 12:46:09 UTC FreeBSD src repository Modified files: sys/fs/cd9660 cd9660_vnops.c Log: SVN rev 179722 on 2008-06-11 12:46:09Z by kib In cd9660_readdir vop, always initialize the idp->uio_off member. The while loop that is assumed to initialize the uio_off later, may be not entered at all, causing uninitialized value to be returned in uio->uio_offset. PR: 122925 Submitted by: Jaakko Heinonen MFC after: 1 weeks Revision Changes Path 1.114 +1 -0 src/sys/fs/cd9660/cd9660_vnops.c From xcllnt at mac.com Wed Jun 11 13:06:10 2008 From: xcllnt at mac.com (Marcel Moolenaar) Date: Wed Jun 11 13:06:15 2008 Subject: cvs commit: src/sbin Makefile In-Reply-To: <484FB191.3020907@omnisec.de> References: <200806091643.m59GhLmP079564@repoman.freebsd.org> <484FB191.3020907@omnisec.de> Message-ID: <63F38830-DCFC-4120-B8CE-92D95A492FA2@mac.com> On Jun 11, 2008, at 4:05 AM, Harald Schmalzbauer wrote: > Marcel Moolenaar schrieb am 09.06.2008 18:42 (localtime): >> marcel 2008-06-09 16:42:55 UTC >> FreeBSD src repository >> Modified files: >> sbin Makefile Log: >> SVN rev 179680 on 2008-06-09 16:42:55Z by marcel >> Disconnect gpt(8) from the build. gpt(8) is replaced by gpart(8). > > I highly appreciate the gpart work, but what I'm missing (in > RELENG_7) is support for "labels" (gpt -r show -l ad99). > Was that added to HEAD? Any plans to MFC? No, that's not in HEAD either. I'll see about adding it. -- Marcel Moolenaar xcllnt@mac.com From cokane at FreeBSD.org Wed Jun 11 13:40:45 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Wed Jun 11 13:40:56 2008 Subject: cvs commit: src/sys/dev/if_ndis if_ndis.c if_ndisvar.h Message-ID: <200806111340.m5BDejje027411@repoman.freebsd.org> cokane 2008-06-11 13:40:15 UTC FreeBSD src repository Modified files: sys/dev/if_ndis if_ndis.c if_ndisvar.h Log: SVN rev 179723 on 2008-06-11 13:40:15Z by cokane Convert ndis_spinlock to ndis_mtx and start using the sleepable mtx interface for NDIS_LOCK/UNLOCK. This should result in less CPU utilization on behalf of the ndis driver. Additionally, this commit also fixes a potential LOR in the ndis_tick code, by not locking inside the ndis_tick function, but instead delegating that work to the helpers called through IoQueueWorkItem. The way that this is currently set up for NDIS prevents us from simply implementing a callout_init_mtx mechanism. However, the helper functions that handle the various timeout cases implement fine-grained locking using the spinlocks provided by the NDIS-compat layer, and using the mtx that is added with this commit. This leaves the following ndis_softc members operated on in ndis_tick in an unlocked context: * ndis_hang_timer - Only modified outside of ndis_tick once, before the first callout_reset to schedule ndis_tick * ifp->if_oerrors - Only incremented in two places, which should be an atomic op * ndis_tx_timer - Assigned to 5 (when guaranteed to be 0) or 0 (in txeof), to indicate to ndis_tick what to do. This is the only member of which I was suspicious for needing the NDIS_LOCK here. My testing (and another's) have been fine so far. * ndis_stat_callout - Only uses a simple set of callout routines, callout_reset only called by ndis_tick after the initial reset, and then callout_drain is used exactly once in shutdown code. The benefit is that ndis_tick doesn't acquire NDIS_LOCK unless one of the timeout conditions is flagged, and it still obeys the locking order semantics that are dictated by the NDIS layer at the moment. I have been investigating a more thorough s/spinlock/mtx/ of the NDIS layer, but the simplest naive approach (replace KeAcquireSpinLock with an mtx implementation) has anti-succeeded for me so far. This is a good first step though. Tested by: onemda@gmail.com Reviewed by: current@, jhb, thompsa Proposed by: jhb Revision Changes Path 1.142 +2 -3 src/sys/dev/if_ndis/if_ndis.c 1.33 +3 -5 src/sys/dev/if_ndis/if_ndisvar.h From jfv at FreeBSD.org Wed Jun 11 18:50:05 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 18:50:08 2008 Subject: cvs commit: src/share/man/man4 ixgbe.4 Message-ID: <200806111850.m5BIo5jc068800@repoman.freebsd.org> jfv 2008-06-11 18:49:55 UTC FreeBSD src repository Added files: share/man/man4 ixgbe.4 Log: SVN rev 179724 on 2008-06-11 18:49:55Z by jfv Add man page for ixgbe driver Revision Changes Path 1.1 +121 -0 src/share/man/man4/ixgbe.4 (new) From dougb at FreeBSD.org Wed Jun 11 18:54:22 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Wed Jun 11 18:54:26 2008 Subject: cvs commit: src/usr.sbin/mergemaster mergemaster.8 Message-ID: <200806111854.m5BIsL6W069132@repoman.freebsd.org> dougb 2008-06-11 18:54:06 UTC FreeBSD src repository Modified files: usr.sbin/mergemaster mergemaster.8 Log: SVN rev 179725 on 2008-06-11 18:54:06Z by dougb Document the AUTO_UPGRADE (-U) knob for .mergemasterrc Submitted by: mezz No Cookie For: gordon :) Revision Changes Path 1.37 +3 -0 src/usr.sbin/mergemaster/mergemaster.8 From ed at FreeBSD.org Wed Jun 11 18:57:25 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Wed Jun 11 18:57:27 2008 Subject: cvs commit: src/sys/cddl/contrib/opensolaris/uts/common/dtrace dtrace.c src/sys/dev/nmdm nmdm.c src/sys/dev/snp snp.c src/sys/dev/vkbd vkbd.c src/sys/fs/ntfs ntfs_ihash.c src/sys/i386/acpica acpi_machdep.c src/sys/kern kern_conf.c src/sys/net if_tap.c if_tun.c ... Message-ID: <200806111857.m5BIvOVE069657@repoman.freebsd.org> ed 2008-06-11 18:55:19 UTC FreeBSD src repository Modified files: sys/cddl/contrib/opensolaris/uts/common/dtrace dtrace.c sys/dev/nmdm nmdm.c sys/dev/snp snp.c sys/dev/vkbd vkbd.c sys/fs/ntfs ntfs_ihash.c sys/i386/acpica acpi_machdep.c sys/kern kern_conf.c sys/net if_tap.c if_tun.c sys/security/audit audit_pipe.c sys/sys conf.h Log: SVN rev 179726 on 2008-06-11 18:55:19Z by ed Don't enforce unique device minor number policy anymore. Except for the case where we use the cloner library (clone_create() and friends), there is no reason to enforce a unique device minor number policy. There are various drivers in the source tree that allocate unr pools and such to provide minor numbers, without using them themselves. Because we still need to support unique device minor numbers for the cloner library, introduce a new flag called D_NEEDMINOR. All cdevsw's that are used in combination with the cloner library should be marked with this flag to make the cloning work. This means drivers can now freely use si_drv0 to store their own flags and state, making it effectively the same as si_drv1 and si_drv2. We still keep the minor() and dev2unit() routines around to make drivers happy. The NTFS code also used the minor number in its hash table. We should not do this anymore. If the si_drv0 field would be changed, it would no longer end up in the same list. Approved by: philip (mentor) Revision Changes Path 1.5 +1 -0 src/sys/cddl/contrib/opensolaris/uts/common/dtrace/dtrace.c 1.40 +1 -1 src/sys/dev/nmdm/nmdm.c 1.108 +1 -1 src/sys/dev/snp/snp.c 1.13 +1 -1 src/sys/dev/vkbd/vkbd.c 1.24 +3 -3 src/sys/fs/ntfs/ntfs_ihash.c 1.39 +1 -1 src/sys/i386/acpica/acpi_machdep.c 1.216 +9 -4 src/sys/kern/kern_conf.c 1.73 +1 -1 src/sys/net/if_tap.c 1.165 +1 -1 src/sys/net/if_tun.c 1.16 +1 -1 src/sys/security/audit/audit_pipe.c 1.240 +2 -1 src/sys/sys/conf.h From marius at FreeBSD.org Wed Jun 11 19:48:52 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 11 19:48:57 2008 Subject: cvs commit: src/sys/dev/hme if_hme.c if_hme_pci.c if_hme_sbus.c if_hmereg.h if_hmevar.h Message-ID: <200806111948.m5BJmqvu075039@repoman.freebsd.org> marius 2008-06-11 19:48:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/hme if_hme.c if_hme_pci.c if_hme_sbus.c if_hmereg.h if_hmevar.h Log: SVN rev 179727 on 2008-06-11 19:48:35Z by marius MFC: r178470 o Disable HMEDEBUG by default. o Add CTASSERTs ensuring that HME_NRXDESC and HME_NTXDESC are set to legal values. o Use appropriate maxsize, nsegments and maxsegsize parameters when creating DMA tags and correct some comments related to them. o The FreeBSD bus_dmamap_sync(9) supports ored together flags for quite some time now so collapse calls accordingly. o Add missing BUS_DMASYNC_PREREAD when syncing the control DMA maps in hme_rint() and hme_start_locked(). o Keep state of the link state and use it to enable or disable the MAC in hme_mii_statchg() accordingly as well as to return early from hme_start_locked() in case the link is down. o Introduce a sc_flags and use it to replace individual members like sc_pci. o Add bus_barrier(9) calls to hme_mac_bitflip(), hme_mii_readreg(), hme_mii_writereg() and hme_stop() to ensure the respective bit has been written before we starting polling on it and for the right bits to change. o Rather just returning in case hme_mac_bitflip() fails and leaving us in an undefined state report the problem and move on; chances are the requested configuration will become active shortly after. o Don't call hme_start_locked() in hme_init_locked() unconditionally but only after calls to hme_init_locked() when it's appropriate, i.e. in hme_watchdog(). o Add a KASSERT which asserts nsegs is valid also to hme_load_txmbuf(). o In hme_load_txmbuf(): - use a maximum of the newly introduced HME_NTXSEGS segments instead of the incorrect HME_NTXQ, which reflects the maximum TX queue length, for loading the mbufs and put the DMA segments back onto the stack instead of the softc as 16 should be ok there. - use the common errno(2) return values instead of homegrown ones, - given that hme_load_txmbuf() is allowed to fail resulting in a packet drop for quite some time now implement the functionality of hme_txcksum() by means of m_pullup(9), which de-obfuscates the code and allows to always retrieve the correct length of the IP header, [1] - also add a KASSERT which asserts nsegs is valid, - take advantage of m_collapse(9) instead of m_defrag(9) for performance reasons. o Don't bother to check whether the interface is running or whether its queue is empty before calling hme_start_locked() in hme_tint(), the former will check these anyway. o In hme_intr() call hme_rint() before hme_tint() as gem_tint() may take quite a while to return when it calls hme_start_locked(). o Get rid of sc_debug and just check if_flags for IFF_DEBUG directly. o Add a shadow sc_ifflags so we don't reset the chip when unnecessary. o Handle IFF_ALLMULTI correctly. [2] o Use PCIR_BAR instead of a homegrown macro. o Replace sc_enaddr[6] with sc_enaddr[ETHER_ADDR_LEN]. o Use the maximum of 256 TX descriptors for better performance as using all of them has no additional static cost rather than using just half of them. Reported by: rwatson [2] Suggested by: yongari [1] Reviewed by: yongari Revision Changes Path 1.53.2.1 +214 -183 src/sys/dev/hme/if_hme.c 1.23.2.1 +2 -2 src/sys/dev/hme/if_hme_pci.c 1.18.2.1 +1 -1 src/sys/dev/hme/if_hme_sbus.c 1.7.2.1 +0 -2 src/sys/dev/hme/if_hmereg.h 1.10.2.1 +11 -11 src/sys/dev/hme/if_hmevar.h From marius at FreeBSD.org Wed Jun 11 19:49:23 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 11 19:49:28 2008 Subject: cvs commit: src/sys/dev/hme if_hme.c if_hme_pci.c if_hme_sbus.c if_hmereg.h if_hmevar.h Message-ID: <200806111949.m5BJnNA4075116@repoman.freebsd.org> marius 2008-06-11 19:49:10 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/hme if_hme.c if_hme_pci.c if_hme_sbus.c if_hmereg.h if_hmevar.h Log: SVN rev 179728 on 2008-06-11 19:49:10Z by marius MFC: r178470 o Disable HMEDEBUG by default. o Add CTASSERTs ensuring that HME_NRXDESC and HME_NTXDESC are set to legal values. o Use appropriate maxsize, nsegments and maxsegsize parameters when creating DMA tags and correct some comments related to them. o The FreeBSD bus_dmamap_sync(9) supports ored together flags for quite some time now so collapse calls accordingly. o Add missing BUS_DMASYNC_PREREAD when syncing the control DMA maps in hme_rint() and hme_start_locked(). o Keep state of the link state and use it to enable or disable the MAC in hme_mii_statchg() accordingly as well as to return early from hme_start_locked() in case the link is down. o Introduce a sc_flags and use it to replace individual members like sc_pci. o Add bus_barrier(9) calls to hme_mac_bitflip(), hme_mii_readreg(), hme_mii_writereg() and hme_stop() to ensure the respective bit has been written before we starting polling on it and for the right bits to change. o Rather just returning in case hme_mac_bitflip() fails and leaving us in an undefined state report the problem and move on; chances are the requested configuration will become active shortly after. o Don't call hme_start_locked() in hme_init_locked() unconditionally but only after calls to hme_init_locked() when it's appropriate, i.e. in hme_watchdog(). o Add a KASSERT which asserts nsegs is valid also to hme_load_txmbuf(). o In hme_load_txmbuf(): - use a maximum of the newly introduced HME_NTXSEGS segments instead of the incorrect HME_NTXQ, which reflects the maximum TX queue length, for loading the mbufs and put the DMA segments back onto the stack instead of the softc as 16 should be ok there. - use the common errno(2) return values instead of homegrown ones, - given that hme_load_txmbuf() is allowed to fail resulting in a packet drop for quite some time now implement the functionality of hme_txcksum() by means of m_pullup(9), which de-obfuscates the code and allows to always retrieve the correct length of the IP header, [1] - also add a KASSERT which asserts nsegs is valid, - take advantage of m_collapse(9) instead of m_defrag(9) for performance reasons. o Don't bother to check whether the interface is running or whether its queue is empty before calling hme_start_locked() in hme_tint(), the former will check these anyway. o In hme_intr() call hme_rint() before hme_tint() as gem_tint() may take quite a while to return when it calls hme_start_locked(). o Get rid of sc_debug and just check if_flags for IFF_DEBUG directly. o Add a shadow sc_ifflags so we don't reset the chip when unnecessary. o Handle IFF_ALLMULTI correctly. [2] o Use PCIR_BAR instead of a homegrown macro. o Replace sc_enaddr[6] with sc_enaddr[ETHER_ADDR_LEN]. o Use the maximum of 256 TX descriptors for better performance as using all of them has no additional static cost rather than using just half of them. Reported by: rwatson [2] Suggested by: yongari [1] Reviewed by: yongari Revision Changes Path 1.37.2.11 +214 -183 src/sys/dev/hme/if_hme.c 1.22.2.1 +2 -2 src/sys/dev/hme/if_hme_pci.c 1.17.2.1 +1 -1 src/sys/dev/hme/if_hme_sbus.c 1.6.8.2 +0 -2 src/sys/dev/hme/if_hmereg.h 1.8.2.3 +11 -11 src/sys/dev/hme/if_hmevar.h From wkoszek at FreeBSD.org Wed Jun 11 20:05:41 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Wed Jun 11 20:05:44 2008 Subject: cvs commit: src/sys/powerpc/booke pmap.c Message-ID: <200806112005.m5BK5eqE082243@repoman.freebsd.org> wkoszek 2008-06-11 20:05:25 UTC FreeBSD src repository Modified files: sys/powerpc/booke pmap.c Log: SVN rev 179729 on 2008-06-11 20:05:25Z by wkoszek Fix a typo in a comment. Revision Changes Path 1.4 +1 -1 src/sys/powerpc/booke/pmap.c From marius at FreeBSD.org Wed Jun 11 20:38:27 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 11 20:38:29 2008 Subject: cvs commit: src/sys/dev/hme if_hme.c Message-ID: <200806112038.m5BKcQ1s085052@repoman.freebsd.org> marius 2008-06-11 20:38:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/hme if_hme.c Log: SVN rev 179730 on 2008-06-11 20:38:06Z by marius Revert the part of r179728 which switched from m_defrag(9) to m_collapse(9) as the latter doesn't exist in FreeBSD 6. Revision Changes Path 1.37.2.12 +1 -1 src/sys/dev/hme/if_hme.c From marius at FreeBSD.org Wed Jun 11 20:52:42 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 11 20:52:46 2008 Subject: cvs commit: src/sys/dev/hme if_hme_pci.c if_hme_sbus.c Message-ID: <200806112052.m5BKqggF095527@repoman.freebsd.org> marius 2008-06-11 20:52:34 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/hme if_hme_pci.c if_hme_sbus.c Log: SVN rev 179731 on 2008-06-11 20:52:34Z by marius MFC: r178589 - Use the revamped code from the gem(4) PCI front-end, which doesn't require parts of the Expansion ROM to be copied around, for obtaining the MAC address on !OFW platforms. - Don't unnecessarily cache bus space tag and handle nor RIDs in the softcs of the front-ends. - Don't use function calls in initializers. - Let the SBus front-end depend on sbus(4). Revision Changes Path 1.23.2.2 +119 -107 src/sys/dev/hme/if_hme_pci.c 1.18.2.2 +55 -54 src/sys/dev/hme/if_hme_sbus.c From marius at FreeBSD.org Wed Jun 11 20:59:14 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 11 20:59:16 2008 Subject: cvs commit: src/sys/dev/hme if_hme_pci.c if_hme_sbus.c Message-ID: <200806112059.m5BKxDFK096030@repoman.freebsd.org> marius 2008-06-11 20:58:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/hme if_hme_pci.c if_hme_sbus.c Log: SVN rev 179732 on 2008-06-11 20:58:53Z by marius MFC: r178589 - Use the revamped code from the gem(4) PCI front-end, which doesn't require parts of the Expansion ROM to be copied around, for obtaining the MAC address on !OFW platforms. - Don't unnecessarily cache bus space tag and handle nor RIDs in the softcs of the front-ends. - Don't use function calls in initializers. - Let the SBus front-end depend on sbus(4). Revision Changes Path 1.22.2.2 +119 -107 src/sys/dev/hme/if_hme_pci.c 1.17.2.2 +55 -54 src/sys/dev/hme/if_hme_sbus.c From marius at FreeBSD.org Wed Jun 11 21:10:15 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 11 21:10:17 2008 Subject: cvs commit: src/sys/dev/gem if_gem.c if_gem_pci.c if_gemreg.h if_gemvar.h Message-ID: <200806112110.m5BLAFYn098148@repoman.freebsd.org> marius 2008-06-11 21:10:06 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: SVN rev 179733 on 2008-06-11 21:10:06Z by marius MFC: r176996, r177560 - 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(). - Split the registers into two halves in preparation for SBus support. [1] Obtained from: NetBSD (loosely) [1] Revision Changes Path 1.44.2.4 +192 -199 src/sys/dev/gem/if_gem.c 1.23.2.2 +37 -15 src/sys/dev/gem/if_gem_pci.c 1.5.2.2 +24 -6 src/sys/dev/gem/if_gemreg.h 1.15.2.2 +42 -1 src/sys/dev/gem/if_gemvar.h From jfv at FreeBSD.org Wed Jun 11 22:00:33 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 22:00:36 2008 Subject: cvs commit: src/sys/net tcp_lro.c tcp_lro.h Message-ID: <200806112200.m5BM0WKY002592@repoman.freebsd.org> jfv 2008-06-11 22:00:29 UTC FreeBSD src repository Added files: sys/net tcp_lro.c tcp_lro.h Log: SVN rev 179734 on 2008-06-11 22:00:29Z by jfv Add generic TCP LRO code, moved from the ixgbe driver into net Revision Changes Path 1.1 +379 -0 src/sys/net/tcp_lro.c (new) 1.1 +85 -0 src/sys/net/tcp_lro.h (new) From jfv at FreeBSD.org Wed Jun 11 22:08:20 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 22:08:21 2008 Subject: cvs commit: src/sys/net tcp_lro.c tcp_lro.h Message-ID: <200806112208.m5BM8Jn2004318@repoman.freebsd.org> jfv 2008-06-11 22:08:13 UTC FreeBSD src repository Removed files: sys/net tcp_lro.c tcp_lro.h Log: SVN rev 179735 on 2008-06-11 22:08:13Z by jfv Duh, wrong directory, needed to be in netinet Revision Changes Path 1.2 +0 -379 src/sys/net/tcp_lro.c (dead) 1.2 +0 -85 src/sys/net/tcp_lro.h (dead) From jfv at FreeBSD.org Wed Jun 11 22:10:21 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 22:10:24 2008 Subject: cvs commit: src/sys/conf files Message-ID: <200806112210.m5BMAKcM004550@repoman.freebsd.org> jfv 2008-06-11 22:10:10 UTC FreeBSD src repository Modified files: sys/conf files Log: SVN rev 179736 on 2008-06-11 22:10:10Z by jfv Add LRO into kernel build Revision Changes Path 1.1308 +1 -0 src/sys/conf/files From jfv at FreeBSD.org Wed Jun 11 22:13:07 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 22:13:11 2008 Subject: cvs commit: src/sys/netinet tcp_lro.c tcp_lro.h Message-ID: <200806112213.m5BMD68f004858@repoman.freebsd.org> jfv 2008-06-11 22:12:50 UTC FreeBSD src repository Added files: sys/netinet tcp_lro.c tcp_lro.h Log: SVN rev 179737 on 2008-06-11 22:12:50Z by jfv Add generic TCP LOR into netinet Revision Changes Path 1.1 +379 -0 src/sys/netinet/tcp_lro.c (new) 1.1 +85 -0 src/sys/netinet/tcp_lro.h (new) From jfv at FreeBSD.org Wed Jun 11 22:19:07 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 22:19:11 2008 Subject: cvs commit: src/sys/modules/ixgbe Makefile Message-ID: <200806112219.m5BMJ7iC005346@repoman.freebsd.org> jfv 2008-06-11 22:18:50 UTC FreeBSD src repository Modified files: sys/modules/ixgbe Makefile Log: SVN rev 179738 on 2008-06-11 22:18:50Z by jfv Remove compile of tcp_lro since its now in netinet Revision Changes Path 1.4 +1 -1 src/sys/modules/ixgbe/Makefile From jfv at FreeBSD.org Wed Jun 11 22:26:39 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 22:26:40 2008 Subject: cvs commit: src/sys/dev/ixgbe tcp_lro.c tcp_lro.h Message-ID: <200806112226.m5BMQcBK006009@repoman.freebsd.org> jfv 2008-06-11 22:26:27 UTC FreeBSD src repository Removed files: sys/dev/ixgbe tcp_lro.c tcp_lro.h Log: SVN rev 179739 on 2008-06-11 22:26:27Z by jfv Remove the tcp_lro.[ch] files as they are now in netinet. Revision Changes Path 1.2 +0 -380 src/sys/dev/ixgbe/tcp_lro.c (dead) 1.2 +0 -85 src/sys/dev/ixgbe/tcp_lro.h (dead) From jfv at FreeBSD.org Wed Jun 11 22:29:54 2008 From: jfv at FreeBSD.org (Jack F Vogel) Date: Wed Jun 11 22:29:59 2008 Subject: cvs commit: src/sys/dev/ixgbe ixgbe.h Message-ID: <200806112229.m5BMTseK006237@repoman.freebsd.org> jfv 2008-06-11 22:29:47 UTC FreeBSD src repository Modified files: sys/dev/ixgbe ixgbe.h Log: SVN rev 179740 on 2008-06-11 22:29:47Z by jfv Change include to reflect tcp_lro.h move. Revision Changes Path 1.5 +1 -1 src/sys/dev/ixgbe/ixgbe.h From peter at FreeBSD.org Wed Jun 11 23:03:53 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 11 23:03:57 2008 Subject: cvs commit: CVSROOT access Message-ID: <200806112303.m5BN3qj9010083@repoman.freebsd.org> peter 2008-06-11 23:03:45 UTC FreeBSD src repository Modified files: . access Log: SVN rev 179741 on 2008-06-11 23:03:45Z by peter NOP commit to svnadmin/conf/access, to test cvs export. Approved by: core (implicit) Revision Changes Path 1.879 +0 -2 CVSROOT/access From mlaier at FreeBSD.org Wed Jun 11 23:36:24 2008 From: mlaier at FreeBSD.org (Max Laier) Date: Wed Jun 11 23:36:26 2008 Subject: cvs commit: CVSROOT access Message-ID: <200806112336.m5BNaNWM012888@repoman.freebsd.org> mlaier 2008-06-11 23:36:15 UTC FreeBSD src repository Modified files: . access Log: SVN rev 179743 on 2008-06-11 23:36:15Z by mlaier Please welcome Ermal Luçi (eri@) as a new src committer. He is a very active developer at pfSense and will be working with me to integrate some of his additional pf-features into FreeBSD. thompsa@, also involved with the pfSense project, will co-mentor. Approved by: core Revision Changes Path 1.880 +1 -0 CVSROOT/access From benno at FreeBSD.org Thu Jun 12 00:39:11 2008 From: benno at FreeBSD.org (Benno Rice) Date: Thu Jun 12 00:39:12 2008 Subject: cvs commit: src/sys/dev/smc if_smc.c if_smcvar.h Message-ID: <200806120039.m5C0dAfp019588@repoman.freebsd.org> benno 2008-06-12 00:38:54 UTC FreeBSD src repository Modified files: sys/dev/smc if_smc.c if_smcvar.h Log: SVN rev 179744 on 2008-06-12 00:38:54Z by benno Make sure we drain our taskqueues and stop our callouts in detach. Revision Changes Path 1.4 +36 -7 src/sys/dev/smc/if_smc.c 1.4 +1 -0 src/sys/dev/smc/if_smcvar.h From kevlo at FreeBSD.org Thu Jun 12 01:46:13 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Thu Jun 12 01:46:21 2008 Subject: cvs commit: src/sys/arm/xscale/i8134x i81342_pci.c Message-ID: <200806120146.m5C1kDRm037459@repoman.freebsd.org> kevlo 2008-06-12 01:46:06 UTC FreeBSD src repository Modified files: sys/arm/xscale/i8134x i81342_pci.c Log: SVN rev 179745 on 2008-06-12 01:46:06Z by kevlo Fix a typo: i80321_pci_probe -> i81342_pci_probe Revision Changes Path 1.3 +1 -1 src/sys/arm/xscale/i8134x/i81342_pci.c From kevlo at FreeBSD.org Thu Jun 12 02:29:15 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Thu Jun 12 02:29:23 2008 Subject: cvs commit: src/sys/powerpc/powermac grackle.c macio.c uninorth.c Message-ID: <200806120229.m5C2TEXT041854@repoman.freebsd.org> kevlo 2008-06-12 02:28:58 UTC FreeBSD src repository Modified files: sys/powerpc/powermac grackle.c macio.c uninorth.c Log: SVN rev 179746 on 2008-06-12 02:28:58Z by kevlo Return an error code rather than ENXIO when both rman_init() and rman_manage_region() failed. Reviewed by: marcel Revision Changes Path 1.13 +12 -12 src/sys/powerpc/powermac/grackle.c 1.25 +10 -5 src/sys/powerpc/powermac/macio.c 1.20 +11 -11 src/sys/powerpc/powermac/uninorth.c From marcel at FreeBSD.org Thu Jun 12 04:37:48 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 04:37:50 2008 Subject: cvs commit: src/sys/sys apm.h Message-ID: <200806120437.m5C4blwQ054168@repoman.freebsd.org> marcel 2008-06-12 04:37:37 UTC FreeBSD src repository Modified files: sys/sys apm.h Log: SVN rev 179747 on 2008-06-12 04:37:37Z by marcel Define APM_ENT_NAMELEN and APM_ENT_TYPELEN for general use. Revision Changes Path 1.3 +5 -2 src/sys/sys/apm.h From marcel at FreeBSD.org Thu Jun 12 04:43:48 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 04:43:50 2008 Subject: cvs commit: src/sys/geom/part g_part_apm.c Message-ID: <200806120443.m5C4hmUA054667@repoman.freebsd.org> marcel 2008-06-12 04:43:34 UTC FreeBSD src repository Modified files: sys/geom/part g_part_apm.c Log: SVN rev 179748 on 2008-06-12 04:43:34Z by marcel Add the partition label and the raw partition type to the XML. Revision Changes Path 1.7 +19 -5 src/sys/geom/part/g_part_apm.c From alc at FreeBSD.org Thu Jun 12 05:18:20 2008 From: alc at FreeBSD.org (Alan Cox) Date: Thu Jun 12 05:18:24 2008 Subject: cvs commit: src/sys/amd64/amd64 pmap.c Message-ID: <200806120518.m5C5IJcb067691@repoman.freebsd.org> alc 2008-06-12 05:18:09 UTC FreeBSD src repository Modified files: sys/amd64/amd64 pmap.c Log: SVN rev 179749 on 2008-06-12 05:18:09Z by alc Reverse the direction of pmap_promote_pde()'s traversal over the specified page table page. The direction of the traversal can matter if pmap_promote_pde() has to remove write access (PG_RW) from a PTE that hasn't been modified (PG_M). In general, if there are two or more such PTEs to choose among, it is better to write protect the one nearer the high end of the page table page rather than the low end. This is because most programs access memory in an ascending direction. The net result of this change is a sometimes significant reduction in the number of failed promotion attempts and the number of pages that are write protected by pmap_promote_pde(). Revision Changes Path 1.620 +28 -12 src/sys/amd64/amd64/pmap.c From marcel at FreeBSD.org Thu Jun 12 05:27:36 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 05:27:39 2008 Subject: cvs commit: src/sys/geom/part g_part_bsd.c Message-ID: <200806120527.m5C5RZTt068459@repoman.freebsd.org> marcel 2008-06-12 05:27:23 UTC FreeBSD src repository Modified files: sys/geom/part g_part_bsd.c Log: SVN rev 179750 on 2008-06-12 05:27:23Z by marcel Add the raw partition type to the XML. Revision Changes Path 1.6 +10 -4 src/sys/geom/part/g_part_bsd.c From marcel at FreeBSD.org Thu Jun 12 05:28:53 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 05:28:57 2008 Subject: cvs commit: src/sys/geom/part g_part_gpt.c Message-ID: <200806120528.m5C5Spmu068561@repoman.freebsd.org> marcel 2008-06-12 05:28:47 UTC FreeBSD src repository Modified files: sys/geom/part g_part_gpt.c Log: SVN rev 179751 on 2008-06-12 05:28:47Z by marcel Add the raw partiton type to the XML. Revision Changes Path 1.9 +14 -6 src/sys/geom/part/g_part_gpt.c From marcel at FreeBSD.org Thu Jun 12 05:56:09 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 05:56:12 2008 Subject: cvs commit: src/sys/geom/part g_part_mbr.c Message-ID: <200806120556.m5C5u7GL070778@repoman.freebsd.org> marcel 2008-06-12 05:56:03 UTC FreeBSD src repository Modified files: sys/geom/part g_part_mbr.c Log: SVN rev 179752 on 2008-06-12 05:56:03Z by marcel Add the raw partition type to the XML. Revision Changes Path 1.8 +10 -4 src/sys/geom/part/g_part_mbr.c From kientzle at FreeBSD.org Thu Jun 12 06:17:24 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Thu Jun 12 06:17:27 2008 Subject: cvs commit: src/usr.bin/tar matching.c Message-ID: <200806120617.m5C6HNTP073658@repoman.freebsd.org> kientzle 2008-06-12 06:17:15 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/tar matching.c Log: SVN rev 179753 on 2008-06-12 06:17:15Z by kientzle MFC r177341: decrement matched patterns instead of incrementing. In particular, this fixes --exit-fast behavior. Revision Changes Path 1.11.2.1 +1 -1 src/usr.bin/tar/matching.c From kientzle at FreeBSD.org Thu Jun 12 06:21:09 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Thu Jun 12 06:21:16 2008 Subject: cvs commit: src/usr.bin/tar Makefile matching.c Message-ID: <200806120621.m5C6L9C5074005@repoman.freebsd.org> kientzle 2008-06-12 06:20:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/tar Makefile matching.c Log: SVN rev 179754 on 2008-06-12 06:20:49Z by kientzle MFC 177341: Decrement matched patterns so that --extract-fast works. Revision Changes Path 1.24.2.5 +1 -1 src/usr.bin/tar/Makefile 1.9.2.3 +1 -1 src/usr.bin/tar/matching.c From marcel at FreeBSD.org Thu Jun 12 06:26:42 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 06:26:46 2008 Subject: cvs commit: src/sys/geom/part g_part_pc98.c Message-ID: <200806120626.m5C6QeMJ074435@repoman.freebsd.org> marcel 2008-06-12 06:26:36 UTC FreeBSD src repository Modified files: sys/geom/part g_part_pc98.c Log: SVN rev 179755 on 2008-06-12 06:26:36Z by marcel Add the raw partition type to the XML. Revision Changes Path 1.4 +11 -4 src/sys/geom/part/g_part_pc98.c From marcel at FreeBSD.org Thu Jun 12 06:34:26 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 06:34:28 2008 Subject: cvs commit: src/sys/geom/part g_part_vtoc8.c Message-ID: <200806120634.m5C6YQUr075069@repoman.freebsd.org> marcel 2008-06-12 06:34:14 UTC FreeBSD src repository Modified files: sys/geom/part g_part_vtoc8.c Log: SVN rev 179756 on 2008-06-12 06:34:14Z by marcel Add the raw partition type to the XML. Revision Changes Path 1.4 +12 -6 src/sys/geom/part/g_part_vtoc8.c From ed at FreeBSD.org Thu Jun 12 08:31:57 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Thu Jun 12 08:32:04 2008 Subject: cvs commit: src/sys/cddl/contrib/opensolaris/uts/common/fs/zfs zfs_znode.c src/sys/kern kern_conf.c src/sys/sys conf.h systm.h Message-ID: <200806120831.m5C8VvKB089149@repoman.freebsd.org> ed 2008-06-12 08:30:54 UTC FreeBSD src repository Modified files: sys/cddl/contrib/opensolaris/uts/common/fs/zfs zfs_znode.c sys/kern kern_conf.c sys/sys conf.h systm.h Log: SVN rev 179757 on 2008-06-12 08:30:54Z by ed Turn dev2unit(), minor(), unit2minor() and minor2unit() into macro's. Now that we got rid of the minor-to-unit conversion and the constraints on device minor numbers, we can convert the functions that operate on minor and unit numbers to simple macro's. The unit2minor() and minor2unit() macro's are now no-ops. The ZFS code als defined a macro named `minor'. Change the ZFS code to use umajor() and uminor() here, as it is the correct approach to do this. Also add $FreeBSD$ to keep SVN happy. Approved by: philip (mentor), pjd Revision Changes Path 1.16 +3 -7 src/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_znode.c 1.217 +0 -35 src/sys/kern/kern_conf.c 1.241 +5 -3 src/sys/sys/conf.h 1.267 +0 -1 src/sys/sys/systm.h From ed at FreeBSD.org Thu Jun 12 08:53:59 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Thu Jun 12 08:54:02 2008 Subject: cvs commit: src/sys/cddl/contrib/opensolaris/uts/common/fs/zfs zfs_znode.c Message-ID: <200806120853.m5C8rwc9000462@repoman.freebsd.org> ed 2008-06-12 08:53:54 UTC FreeBSD src repository Modified files: sys/cddl/contrib/opensolaris/uts/common/fs/zfs zfs_znode.c Log: SVN rev 179758 on 2008-06-12 08:53:54Z by ed Remove the $FreeBSD$ tag again, now I know fbsd:nokeywords exists. Requested by: pjd Approved by: philip (mentor) Revision Changes Path 1.17 +0 -2 src/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_znode.c From pjd at FreeBSD.org Thu Jun 12 09:47:54 2008 From: pjd at FreeBSD.org (Pawel Jakub Dawidek) Date: Thu Jun 12 09:48:00 2008 Subject: cvs commit: src/sys/cddl/contrib/opensolaris/uts/common/fs/zfs zfs_znode.c In-Reply-To: <200806120853.m5C8rwc9000462@repoman.freebsd.org> References: <200806120853.m5C8rwc9000462@repoman.freebsd.org> Message-ID: <20080612094752.GC4757@garage.freebsd.pl> On Thu, Jun 12, 2008 at 08:53:54AM +0000, Ed Schouten wrote: > ed 2008-06-12 08:53:54 UTC > > FreeBSD src repository > > Modified files: > sys/cddl/contrib/opensolaris/uts/common/fs/zfs zfs_znode.c > Log: > SVN rev 179758 on 2008-06-12 08:53:54Z by ed > > Remove the $FreeBSD$ tag again, now I know fbsd:nokeywords exists. Thank you. -- 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/20080612/2c5213b9/attachment.pgp From ed at FreeBSD.org Thu Jun 12 10:15:31 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Thu Jun 12 10:15:43 2008 Subject: cvs commit: src/usr.bin/fstat fstat.c Message-ID: <200806121015.m5CAFVMM009314@repoman.freebsd.org> ed 2008-06-12 10:15:14 UTC FreeBSD src repository Modified files: usr.bin/fstat fstat.c Log: SVN rev 179759 on 2008-06-12 10:15:14Z by ed Fix build of fstat after minor() changes. Even though I ran a `make universe' to see whether the changes to the device minor number macro's broke the build, I was not expecting `make universe' to silently continue if build errors occured, thus causing me to overlook the build error. Approved by: philip (mentor) Pointyhat to: me Revision Changes Path 1.68 +1 -1 src/usr.bin/fstat/fstat.c From flz at FreeBSD.org Thu Jun 12 15:21:23 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Thu Jun 12 15:21:25 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add perform.c src/usr.sbin/pkg_install/lib lib.h src/usr.sbin/pkg_install/updating main.c Message-ID: <200806121521.m5CFLNJo062612@repoman.freebsd.org> flz 2008-06-12 15:21:13 UTC FreeBSD src repository Modified files: usr.sbin/pkg_install/add perform.c usr.sbin/pkg_install/lib lib.h usr.sbin/pkg_install/updating main.c Log: SVN rev 179760 on 2008-06-12 15:21:13Z by flz - add: Keep dependent packages too if -K is specified. - updating: terminating '\n' is not part of the package origin. - bump PKG_INSTALL_VERSION to 20080612. PR: bin/119368 [1], bin/124459 [2] Submitted by: gcooper [1], Beat Gatzi [2] MFC after: 3 days Revision Changes Path 1.85 +1 -1 src/usr.sbin/pkg_install/add/perform.c 1.68 +1 -1 src/usr.sbin/pkg_install/lib/lib.h 1.4 +2 -0 src/usr.sbin/pkg_install/updating/main.c From kaiw at FreeBSD.org Thu Jun 12 16:45:56 2008 From: kaiw at FreeBSD.org (Kai Wang) Date: Thu Jun 12 16:45:58 2008 Subject: cvs commit: src/sys/dev/usb ums.c usbdevs Message-ID: <200806121645.m5CGjucY070673@repoman.freebsd.org> kaiw 2008-06-12 16:45:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/usb ums.c usbdevs Log: SVN rev 179761 on 2008-06-12 16:45:46Z by kaiw MFC r178697: Add support for Microsoft Notebook Optical Mouse 3000 Model 1049. PR: usb/121052 Revision Changes Path 1.96.2.3 +20 -0 src/sys/dev/usb/ums.c 1.328.2.12 +1 -0 src/sys/dev/usb/usbdevs From remko at FreeBSD.org Thu Jun 12 18:49:29 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Thu Jun 12 18:49:30 2008 Subject: cvs commit: src/share/man/man5 tmpfs.5 Message-ID: <200806121849.m5CInScX092654@repoman.freebsd.org> remko 2008-06-12 18:49:24 UTC FreeBSD src repository Modified files: share/man/man5 tmpfs.5 Log: SVN rev 179762 on 2008-06-12 18:49:24Z by remko Document more tmpfs mount options that are already available. PR: 124488 Submitted by: gahr Revision Changes Path 1.4 +17 -1 src/share/man/man5/tmpfs.5 From marcel at FreeBSD.org Thu Jun 12 19:34:16 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Thu Jun 12 19:34:19 2008 Subject: cvs commit: src/sys/geom/part g_part_gpt.c Message-ID: <200806121934.m5CJYF9o097486@repoman.freebsd.org> marcel 2008-06-12 19:34:07 UTC FreeBSD src repository Modified files: sys/geom/part g_part_gpt.c Log: SVN rev 179763 on 2008-06-12 19:34:07Z by marcel Finish the support for partition labels and add it to the XML. Revision Changes Path 1.10 +86 -10 src/sys/geom/part/g_part_gpt.c From gnn at FreeBSD.org Thu Jun 12 19:47:17 2008 From: gnn at FreeBSD.org (George V. Neville-Neil) Date: Thu Jun 12 19:47:19 2008 Subject: cvs commit: CVSROOT access Message-ID: <200806121947.m5CJlGZg098571@repoman.freebsd.org> gnn 2008-06-12 19:46:57 UTC FreeBSD src repository Modified files: . access Log: SVN rev 179764 on 2008-06-12 19:46:57Z by gnn Add Ivan Voras as a src committer. Approved by: core Revision Changes Path 1.881 +1 -0 CVSROOT/access From ups at FreeBSD.org Thu Jun 12 20:47:05 2008 From: ups at FreeBSD.org (Stephan Uphoff) Date: Thu Jun 12 20:47:09 2008 Subject: cvs commit: src/sys/vm vnode_pager.c Message-ID: <200806122047.m5CKl4IQ006890@repoman.freebsd.org> ups 2008-06-12 20:46:47 UTC FreeBSD src repository Modified files: sys/vm vnode_pager.c Log: SVN rev 179765 on 2008-06-12 20:46:47Z by ups Fix vm object creation locking to allow SHARED vnode locking for vnode_create_vobject. (Not currently used) Noticed by: kib@ Revision Changes Path 1.242 +7 -6 src/sys/vm/vnode_pager.c From ivoras at FreeBSD.org Thu Jun 12 22:52:22 2008 From: ivoras at FreeBSD.org (Ivan Voras) Date: Thu Jun 12 22:52:25 2008 Subject: cvs commit: src/usr.bin/calendar/calendars calendar.freebsd Message-ID: <200806122252.m5CMqLN4027317@repoman.freebsd.org> ivoras 2008-06-12 22:52:11 UTC FreeBSD src repository Modified files: usr.bin/calendar/calendars calendar.freebsd Log: SVN rev 179766 on 2008-06-12 22:52:11Z by ivoras Add myself to the calendar. Approved by: gnn (mentor) Revision Changes Path 1.238 +1 -0 src/usr.bin/calendar/calendars/calendar.freebsd From ivoras at FreeBSD.org Thu Jun 12 22:53:38 2008 From: ivoras at FreeBSD.org (Ivan Voras) Date: Thu Jun 12 22:53:42 2008 Subject: cvs commit: src/share/misc committers-src.dot Message-ID: <200806122253.m5CMrbTk027416@repoman.freebsd.org> ivoras 2008-06-12 22:53:27 UTC FreeBSD src repository Modified files: share/misc committers-src.dot Log: SVN rev 179767 on 2008-06-12 22:53:27Z by ivoras Record my place in the universe. Approved by: gnn (mentor) Revision Changes Path 1.74 +2 -0 src/share/misc/committers-src.dot From wkoszek at FreeBSD.org Thu Jun 12 22:58:53 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Thu Jun 12 22:58:57 2008 Subject: cvs commit: src/lib/libc/sys getsockopt.2 Message-ID: <200806122258.m5CMwrfb027834@repoman.freebsd.org> wkoszek 2008-06-12 22:58:35 UTC FreeBSD src repository Modified files: lib/libc/sys getsockopt.2 Log: SVN rev 179768 on 2008-06-12 22:58:35Z by wkoszek Bring missing getsockopt(2) options: SO_LABEL SO_PEERLABEL SO_LISTENQLIMIT SO_LISTENQLEN SO_LISTENINCQLEN to the manual page. Till now those were only present in sys/socket.h file. Reviewed by: rwatson, gnn, keramida (with mdoc hat) Revision Changes Path 1.40 +29 -2 src/lib/libc/sys/getsockopt.2 From marcel at FreeBSD.org Fri Jun 13 00:04:25 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Fri Jun 13 00:04:27 2008 Subject: cvs commit: src/sbin/geom/class/part geom_part.c Message-ID: <200806130004.m5D04OlE035856@repoman.freebsd.org> marcel 2008-06-13 00:04:10 UTC FreeBSD src repository Modified files: sbin/geom/class/part geom_part.c Log: SVN rev 179769 on 2008-06-13 00:04:10Z by marcel Implement the -l and -r options for gpart show. The -l option changes the output to show the partition label, if applicable and when present. The -r option changes the output to show the raw (i.e. scheme-specific) partition types. Revision Changes Path 1.8 +32 -6 src/sbin/geom/class/part/geom_part.c From benno at FreeBSD.org Fri Jun 13 00:48:26 2008 From: benno at FreeBSD.org (Benno Rice) Date: Fri Jun 13 00:48:28 2008 Subject: cvs commit: src/sys/dev/smc if_smc.c if_smcvar.h Message-ID: <200806130048.m5D0mPfl051050@repoman.freebsd.org> benno 2008-06-13 00:48:09 UTC FreeBSD src repository Modified files: sys/dev/smc if_smc.c if_smcvar.h Log: SVN rev 179770 on 2008-06-13 00:48:09Z by benno Use callout_init_mtx to simplify locking somewhat. While we're here, rearrange some operations in smc_detach to remove the need for the smc_shutdown variable. Suggested by: jhb Revision Changes Path 1.5 +18 -34 src/sys/dev/smc/if_smc.c 1.5 +0 -1 src/sys/dev/smc/if_smcvar.h From davidch at FreeBSD.org Fri Jun 13 01:16:57 2008 From: davidch at FreeBSD.org (David Christensen) Date: Fri Jun 13 01:17:02 2008 Subject: cvs commit: src/sys/dev/bce if_bce.c if_bcefw.h if_bcereg.h Message-ID: <200806130116.m5D1GvkX054537@repoman.freebsd.org> davidch 2008-06-13 01:16:37 UTC FreeBSD src repository Modified files: sys/dev/bce if_bce.c if_bcefw.h if_bcereg.h Log: SVN rev 179771 on 2008-06-13 01:16:37Z by davidch - Added support for BCM5709 and BCM5716 controllers. MFC after: 2 weeks Revision Changes Path 1.42 +2565 -1209 src/sys/dev/bce/if_bce.c 1.7 +11461 -4244 src/sys/dev/bce/if_bcefw.h 1.20 +2043 -613 src/sys/dev/bce/if_bcereg.h From davidch at FreeBSD.org Fri Jun 13 01:20:43 2008 From: davidch at FreeBSD.org (David Christensen) Date: Fri Jun 13 01:20:47 2008 Subject: cvs commit: src/sys/dev/mii brgphy.c Message-ID: <200806130120.m5D1Khbw054908@repoman.freebsd.org> davidch 2008-06-13 01:20:29 UTC FreeBSD src repository Modified files: sys/dev/mii brgphy.c Log: SVN rev 179772 on 2008-06-13 01:20:29Z by davidch - Added support for BCM5709 and BCM5716. MFC after: 2 weeks Revision Changes Path 1.75 +59 -9 src/sys/dev/mii/brgphy.c From kevlo at FreeBSD.org Fri Jun 13 02:02:30 2008 From: kevlo at FreeBSD.org (Kevin Lo) Date: Fri Jun 13 02:02:32 2008 Subject: cvs commit: src/sys/dev/usb if_rum.c usbdevs Message-ID: <200806130202.m5D22UOv059905@repoman.freebsd.org> kevlo 2008-06-13 02:02:21 UTC FreeBSD src repository Modified files: sys/dev/usb if_rum.c usbdevs Log: SVN rev 179773 on 2008-06-13 02:02:21Z by kevlo Add the D-Link DWA-111 Revision Changes Path 1.24 +1 -0 src/sys/dev/usb/if_rum.c 1.360 +1 -0 src/sys/dev/usb/usbdevs From fjoe at FreeBSD.org Fri Jun 13 06:54:35 2008 From: fjoe at FreeBSD.org (Max Khon) Date: Fri Jun 13 06:54:39 2008 Subject: cvs commit: src/usr.bin/make make.1 var.c Message-ID: <200806130654.m5D6sZE3000695@repoman.freebsd.org> fjoe 2008-06-13 06:54:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/make make.1 var.c Log: SVN rev 179774 on 2008-06-13 06:54:28Z by fjoe MFC: :u (unique) var modifier Requested by: kris Revision Changes Path 1.89.2.5 +3 -0 src/usr.bin/make/make.1 1.159.2.1 +40 -0 src/usr.bin/make/var.c From jhb at FreeBSD.org Fri Jun 13 12:14:26 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 13 12:14:30 2008 Subject: cvs commit: src/sys/dev/ex if_ex.c if_ex_isa.c if_ex_pccard.c if_exvar.h Message-ID: <200806131214.m5DCEPbA043245@repoman.freebsd.org> jhb 2008-06-13 12:14:22 UTC FreeBSD src repository Modified files: sys/dev/ex if_ex.c if_ex_isa.c if_ex_pccard.c if_exvar.h Log: SVN rev 179775 on 2008-06-13 12:14:22Z by jhb Make ex(4) MPSAFE: - Add a mutex to the softc to protect the softc and device hardware. - Use a private watchdog timer. - Setup interrupt handler after ether_ifattach(). - Use bus_foo() rather than bus_space_foo() and remove bus space tag and handle from softc. Tested by: imp Revision Changes Path 1.63 +77 -47 src/sys/dev/ex/if_ex.c 1.13 +7 -15 src/sys/dev/ex/if_ex_isa.c 1.19 +0 -7 src/sys/dev/ex/if_ex_pccard.c 1.8 +16 -20 src/sys/dev/ex/if_exvar.h From ambrisko at FreeBSD.org Fri Jun 13 15:46:01 2008 From: ambrisko at FreeBSD.org (Doug Ambrisko) Date: Fri Jun 13 15:46:04 2008 Subject: cvs commit: src/sys/dev/mfi mfi.c mfi_ioctl.h mfi_pci.c Message-ID: <200806131546.m5DFk0k7073614@repoman.freebsd.org> ambrisko 2008-06-13 15:45:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/mfi mfi.c mfi_ioctl.h mfi_pci.c Log: SVN rev 179776 on 2008-06-13 15:45:48Z by ambrisko MFC: All changes in current to 7.X. - Native 32bit compat modes for amd64 - Limit max commands to 128 - Add in compat mode for using mfi0 to talk to any card Revision Changes Path 1.33.2.3 +162 -15 src/sys/dev/mfi/mfi.c 1.5.2.1 +38 -0 src/sys/dev/mfi/mfi_ioctl.h 1.9.2.1 +1 -0 src/sys/dev/mfi/mfi_pci.c From joel.dahl at gmail.com Fri Jun 13 15:51:46 2008 From: joel.dahl at gmail.com (Joel) Date: Fri Jun 13 15:51:51 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <200806121947.m5CJlGZg098571@repoman.freebsd.org> References: <200806121947.m5CJlGZg098571@repoman.freebsd.org> Message-ID: <485290EF.5010601@FreeBSD.org> George V. Neville-Neil skrev: > gnn 2008-06-12 19:46:57 UTC > > FreeBSD src repository > > Modified files: > . access > Log: > SVN rev 179764 on 2008-06-12 19:46:57Z by gnn > > Add Ivan Voras as a src committer. You failed to include two essential pieces of information: - Who the mentor is. - What Ivan will be working on. Not everyone has access to the introduction emails on developers@ like you and me have. -- Joel From obrien at FreeBSD.org Fri Jun 13 16:05:33 2008 From: obrien at FreeBSD.org (David O'Brien) Date: Fri Jun 13 16:05:37 2008 Subject: cvs commit: CVSROOT access In-Reply-To: <485290EF.5010601@FreeBSD.org> References: <200806121947.m5CJlGZg098571@repoman.freebsd.org> <485290EF.5010601@FreeBSD.org> Message-ID: <20080613160532.GC74595@dragon.NUXI.org> On Fri, Jun 13, 2008 at 05:23:27PM +0200, Joel wrote: > George V. Neville-Neil skrev: >> gnn 2008-06-12 19:46:57 UTC >> FreeBSD src repository >> Modified files: >> . access Log: >> SVN rev 179764 on 2008-06-12 19:46:57Z by gnn >> Add Ivan Voras as a src committer. > > You failed to include two essential pieces of information: > > - Who the mentor is. > - What Ivan will be working on. Its also nice to add the freefall login. Add Ivan Voras (ivoras) as a src committer. Ivan will be mentored by me (gnn). ... It is good to have all the information in one place about new folks. From alc at FreeBSD.org Fri Jun 13 19:34:05 2008 From: alc at FreeBSD.org (Alan Cox) Date: Fri Jun 13 19:34:08 2008 Subject: cvs commit: src/sys/amd64/amd64 pmap.c Message-ID: <200806131934.m5DJY4Xc010670@repoman.freebsd.org> alc 2008-06-13 19:33:56 UTC FreeBSD src repository Modified files: sys/amd64/amd64 pmap.c Log: SVN rev 179777 on 2008-06-13 19:33:56Z by alc Tweak the promotion test in pmap_promote_pde(). Specifically, test PG_A before PG_M. This sometimes prevents unnecessary removal of write access from a PTE. Overall, the net result is fewer demotions and promotion failures. Revision Changes Path 1.621 +4 -5 src/sys/amd64/amd64/pmap.c From gonzo at FreeBSD.org Fri Jun 13 19:35:21 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Fri Jun 13 19:35:24 2008 Subject: cvs commit: src/share/man/man9 taskqueue.9 Message-ID: <200806131935.m5DJZKmr010856@repoman.freebsd.org> gonzo 2008-06-13 19:35:17 UTC FreeBSD src repository Modified files: share/man/man9 taskqueue.9 Log: SVN rev 179778 on 2008-06-13 19:35:17Z by gonzo Add "fast" versions of macros and functions to taskqueue(9) API description: TASKQUEUE_FAST_DEFINE(9), TASKQUEUE_FAST_DEFINE_THREAD(9), taskqueue_create_fast(9). They deal with taskqueues intended for use in fast interrupt handlers. Approved by: cognet (mentor) Reviewed by: keramida Revision Changes Path 1.23 +22 -5 src/share/man/man9/taskqueue.9 From gonzo at FreeBSD.org Fri Jun 13 19:45:22 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Fri Jun 13 19:45:24 2008 Subject: cvs commit: src/share/man/man9 taskqueue.9 Message-ID: <200806131945.m5DJjLRD011739@repoman.freebsd.org> gonzo 2008-06-13 19:45:12 UTC FreeBSD src repository Modified files: share/man/man9 taskqueue.9 Log: SVN rev 179779 on 2008-06-13 19:45:12Z by gonzo It's Friday the 13th, not the 10th and I can't travel in time. Yet. Revision Changes Path 1.24 +1 -1 src/share/man/man9/taskqueue.9 From thompsa at FreeBSD.org Fri Jun 13 21:49:27 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Fri Jun 13 21:49:29 2008 Subject: cvs commit: src/lib/libgeom libgeom.3 Message-ID: <200806132149.m5DLnRYi034979@repoman.freebsd.org> thompsa 2008-06-13 21:49:21 UTC FreeBSD src repository Modified files: lib/libgeom libgeom.3 Log: SVN rev 179780 on 2008-06-13 21:49:21Z by thompsa Fix spelling. PR: docs/124561 Submitted by: Sam Banks Revision Changes Path 1.10 +1 -1 src/lib/libgeom/libgeom.3 From rwatson at FreeBSD.org Fri Jun 13 22:14:32 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Fri Jun 13 22:14:34 2008 Subject: cvs commit: src/sys/security/mac mac_framework.h mac_inet.c mac_policy.h src/sys/security/mac_biba mac_biba.c src/sys/security/mac_lomac mac_lomac.c src/sys/security/mac_mls mac_mls.c src/sys/security/mac_stub mac_stub.c ... Message-ID: <200806132214.m5DMEWnE038627@repoman.freebsd.org> rwatson 2008-06-13 22:14:15 UTC FreeBSD src repository Modified files: sys/security/mac mac_framework.h mac_inet.c mac_policy.h sys/security/mac_biba mac_biba.c sys/security/mac_lomac mac_lomac.c sys/security/mac_mls mac_mls.c sys/security/mac_stub mac_stub.c sys/security/mac_test mac_test.c Log: SVN rev 179781 on 2008-06-13 22:14:15Z by rwatson The TrustedBSD MAC Framework named struct ipq instances 'ipq', which is the same as the global variable defined in ip_input.c. Instead, adopt the name 'q' as found in about 1/2 of uses in ip_input.c, preventing a collision on the name. This is non-harmful, but means that search and replace on the global works less well (as in the virtualization work), as well as indexing tools. MFC after: 1 week Reported by: julian Revision Changes Path 1.98 +6 -6 src/sys/security/mac/mac_framework.h 1.20 +14 -14 src/sys/security/mac/mac_inet.c 1.108 +5 -6 src/sys/security/mac/mac_policy.h 1.118 +10 -10 src/sys/security/mac_biba/mac_biba.c 1.63 +11 -11 src/sys/security/mac_lomac/mac_lomac.c 1.100 +10 -10 src/sys/security/mac_mls/mac_mls.c 1.81 +8 -8 src/sys/security/mac_stub/mac_stub.c 1.95 +9 -9 src/sys/security/mac_test/mac_test.c From brooks at FreeBSD.org Fri Jun 13 22:42:18 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Fri Jun 13 22:42:22 2008 Subject: cvs commit: CVSROOT access Message-ID: <200806132242.m5DMgI0O040850@repoman.freebsd.org> brooks 2008-06-13 22:42:09 UTC FreeBSD src repository Modified files: . access Log: SVN rev 179782 on 2008-06-13 22:42:09Z by brooks Take jkh's commit bit into safe keeping per his request. Approved by: core Revision Changes Path 1.882 +0 -1 CVSROOT/access From rrs at FreeBSD.org Sat Jun 14 07:59:07 2008 From: rrs at FreeBSD.org (Randall Stewart) Date: Sat Jun 14 07:59:11 2008 Subject: cvs commit: src/sys/netinet sctp.h sctp_asconf.c sctp_auth.c sctp_bsd_addr.c sctp_bsd_addr.h sctp_cc_functions.c sctp_header.h sctp_indata.c sctp_input.c sctp_lock_bsd.h sctp_os_bsd.h sctp_output.c sctp_pcb.c sctp_pcb.h sctp_peeloff.c sctp_sysctl.c ... Message-ID: <200806140759.m5E7x7cc038498@repoman.freebsd.org> rrs 2008-06-14 07:58:05 UTC FreeBSD src repository Modified files: sys/netinet sctp.h sctp_asconf.c sctp_auth.c sctp_bsd_addr.c sctp_bsd_addr.h sctp_cc_functions.c sctp_header.h sctp_indata.c sctp_input.c sctp_lock_bsd.h sctp_os_bsd.h sctp_output.c sctp_pcb.c sctp_pcb.h sctp_peeloff.c sctp_sysctl.c sctp_sysctl.h sctp_timer.c sctp_uio.h sctp_usrreq.c sctp_var.h sctputil.c sys/netinet6 sctp6_usrreq.c Log: SVN rev 179783 on 2008-06-14 07:58:05Z by rrs - Macro-izes the packed declaration in all headers. - Vimage prep - these are major restructures to move all global variables to be accessed via a macro or two. The variables all go into a single structure. - Asconf address addition tweaks (add_or_del Interfaces) - Fix rwnd calcualtion to be more conservative. - Support SACK_IMMEDIATE flag to skip delayed sack by demand of peer. - Comment updates in the sack mapping calculations - Invarients panic added. - Pre-support for UDP tunneling (we can do this on MAC but will need added support from UDP to get a "pipe" of UDP packets in. - clear trace buffer sysctl added when local tracing on. Note the majority of this huge patch is all the vimage prep stuff :-) Revision Changes Path 1.25 +61 -65 src/sys/netinet/sctp.h 1.37 +11 -11 src/sys/netinet/sctp_asconf.c 1.20 +3 -3 src/sys/netinet/sctp_auth.c 1.19 +72 -75 src/sys/netinet/sctp_bsd_addr.c 1.8 +2 -0 src/sys/netinet/sctp_bsd_addr.h 1.6 +63 -60 src/sys/netinet/sctp_cc_functions.c 1.7 +2 -0 src/sys/netinet/sctp_header.h 1.51 +122 -176 src/sys/netinet/sctp_indata.c 1.72 +59 -44 src/sys/netinet/sctp_input.c 1.12 +46 -46 src/sys/netinet/sctp_lock_bsd.h 1.36 +13 -6 src/sys/netinet/sctp_os_bsd.h 1.72 +193 -130 src/sys/netinet/sctp_output.c 1.69 +209 -187 src/sys/netinet/sctp_pcb.c 1.35 +25 -1 src/sys/netinet/sctp_pcb.h 1.18 +1 -1 src/sys/netinet/sctp_peeloff.c 1.19 +231 -191 src/sys/netinet/sctp_sysctl.c 1.15 +86 -78 src/sys/netinet/sctp_sysctl.h 1.35 +26 -26 src/sys/netinet/sctp_timer.c 1.31 +2 -2 src/sys/netinet/sctp_uio.h 1.56 +41 -26 src/sys/netinet/sctp_usrreq.c 1.25 +13 -12 src/sys/netinet/sctp_var.h 1.78 +50 -50 src/sys/netinet/sctputil.c 1.45 +20 -6 src/sys/netinet6/sctp6_usrreq.c From ed at FreeBSD.org Sat Jun 14 10:45:47 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Sat Jun 14 10:45:55 2008 Subject: cvs commit: src ObsoleteFiles.inc src/include sgtty.h src/lib/libcompat Makefile src/lib/libcompat/4.1 gtty.c stty.3 stty.c src/sys/sys ioctl.h ioctl_compat.h param.h ttychars.h ttydev.h Message-ID: <200806141045.m5EAjlNw067030@repoman.freebsd.org> ed 2008-06-14 10:42:18 UTC FreeBSD src repository Modified files: . ObsoleteFiles.inc include sgtty.h lib/libcompat Makefile sys/sys ioctl.h ioctl_compat.h param.h Removed files: lib/libcompat/4.1 gtty.c stty.3 stty.c sys/sys ttychars.h ttydev.h Log: SVN rev 179784 on 2008-06-14 10:42:18Z by ed Turn sgtty into a binary-only compatibility interface. sgtty was the original interface to configure terminal attributes on my UNIX-like operating systems. It has been deprecated by the POSIX termios interface, which is implemented in almost any modern system. An advantage of turning this into a binary compatibility interface, is that we can now eventually remove the COMPAT_43TTY switch from kernel configurations. This removes many ioctl()'s from the TTY layer. While there, increase the __FreeBSD_version, which may be useful for the people working on the Ports tree. Reviewed by: kib Approved by: philip (mentor) Revision Changes Path 1.143 +5 -0 src/ObsoleteFiles.inc 1.2 +3 -39 src/include/sgtty.h 1.6 +0 -52 src/lib/libcompat/4.1/gtty.c (dead) 1.11 +0 -94 src/lib/libcompat/4.1/stty.3 (dead) 1.6 +0 -52 src/lib/libcompat/4.1/stty.c (dead) 1.25 +2 -3 src/lib/libcompat/Makefile 1.16 +0 -11 src/sys/sys/ioctl.h 1.11 +7 -15 src/sys/sys/ioctl_compat.h 1.356 +1 -1 src/sys/sys/param.h 1.11 +0 -62 src/sys/sys/ttychars.h (dead) 1.13 +0 -62 src/sys/sys/ttydev.h (dead) From wkoszek at FreeBSD.org Sat Jun 14 12:52:30 2008 From: wkoszek at FreeBSD.org (Wojciech A. Koszek) Date: Sat Jun 14 12:52:44 2008 Subject: cvs commit: src/sys/conf files.i386 files.pc98 options.i386 options.pc98 src/sys/i386/conf NOTES src/sys/i386/include pecoff_machdep.h src/sys/modules Makefile src/sys/pc98/conf NOTES src/sys/pc98/include pecoff_machdep.h Message-ID: <200806141252.m5ECqUGh093173@repoman.freebsd.org> wkoszek 2008-06-14 12:51:44 UTC FreeBSD src repository Modified files: sys/conf files.i386 files.pc98 options.i386 options.pc98 sys/i386/conf NOTES sys/modules Makefile sys/pc98/conf NOTES Removed files: sys/i386/include pecoff_machdep.h sys/pc98/include pecoff_machdep.h Log: SVN rev 179785 on 2008-06-14 12:51:44Z by wkoszek Remove obselete PECOFF image activator support. PRs assigned at the time of removal: kern/80742 Discussed on: freebsd-current (silence), IRC Tested by: make universe Approved by: cognet (mentor) Revision Changes Path 1.597 +0 -1 src/sys/conf/files.i386 1.364 +0 -1 src/sys/conf/files.pc98 1.242 +0 -2 src/sys/conf/options.i386 1.204 +0 -2 src/sys/conf/options.pc98 1.1255 +0 -4 src/sys/i386/conf/NOTES 1.2 +0 -60 src/sys/i386/include/pecoff_machdep.h (dead) 1.573 +1 -3 src/sys/modules/Makefile 1.89 +0 -4 src/sys/pc98/conf/NOTES 1.2 +0 -6 src/sys/pc98/include/pecoff_machdep.h (dead) From rrs at FreeBSD.org Sat Jun 14 13:25:01 2008 From: rrs at FreeBSD.org (Randall Stewart) Date: Sat Jun 14 13:25:06 2008 Subject: cvs commit: src/sys/netinet sctp_lock_bsd.h sctp_os_bsd.h Message-ID: <200806141325.m5EDP1ML096873@repoman.freebsd.org> rrs 2008-06-14 13:24:49 UTC FreeBSD src repository Modified files: sys/netinet sctp_lock_bsd.h sctp_os_bsd.h Log: SVN rev 179786 on 2008-06-14 13:24:49Z by rrs - Fixes foobar on my part. Some missing virtualization macros from specific logging cases. Revision Changes Path 1.13 +4 -4 src/sys/netinet/sctp_lock_bsd.h 1.37 +3 -3 src/sys/netinet/sctp_os_bsd.h From rwatson at FreeBSD.org Sat Jun 14 15:17:07 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sat Jun 14 15:17:16 2008 Subject: cvs commit: src/sys/conf files Message-ID: <200806141517.m5EFH6Rx012717@repoman.freebsd.org> rwatson 2008-06-14 15:17:02 UTC FreeBSD src repository Modified files: sys/conf files Log: SVN rev 179787 on 2008-06-14 15:17:02Z by rwatson When NETATALK is compiled into the kernel, at_rmx.c is required regardless of whether NETATALKDEBUG is enabled, so make building it conditional on NETATALK instead. This problem appears to have been present from the time that the netatalk implementation was imported. PR: 124456 Submitted by: Nathan Whitehorn MFC after: 3 days Revision Changes Path 1.1309 +1 -1 src/sys/conf/files From kientzle at FreeBSD.org Sun Jun 15 04:32:03 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 04:32:07 2008 Subject: cvs commit: src/lib/libarchive archive_entry_link_resolver.c src/lib/libarchive/test test_link_resolver.c Message-ID: <200806150432.m5F4W3vs033608@repoman.freebsd.org> kientzle 2008-06-15 04:31:43 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry_link_resolver.c lib/libarchive/test test_link_resolver.c Log: SVN rev 179789 on 2008-06-15 04:31:43Z by kientzle Fix the new generic link resolver in libarchive to never match dirs as hardlinks. In particular, this fixes some recent ports build failures. Thanks to: Kris Kennaway Revision Changes Path 1.3 +3 -0 src/lib/libarchive/archive_entry_link_resolver.c 1.2 +18 -0 src/lib/libarchive/test/test_link_resolver.c From kientzle at FreeBSD.org Sun Jun 15 05:06:06 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 05:06:10 2008 Subject: cvs commit: src/lib/libarchive Makefile archive.h archive.h.in src/lib/libarchive/test Makefile Message-ID: <200806150506.m5F565e8046910@repoman.freebsd.org> kientzle 2008-06-15 05:05:53 UTC FreeBSD src repository Modified files: lib/libarchive Makefile lib/libarchive/test Makefile Added files: lib/libarchive archive.h Removed files: lib/libarchive archive.h.in Log: SVN rev 179790 on 2008-06-15 05:05:53Z by kientzle archive.h is no longer constructed from archive.h.in, so we can rename it and drop some no-longer-necessary build magic from the Makefile. Revision Changes Path 1.85 +1 -9 src/lib/libarchive/Makefile 1.17 +592 -0 src/lib/libarchive/archive.h (new) 1.51 +0 -592 src/lib/libarchive/archive.h.in (dead) 1.21 +0 -6 src/lib/libarchive/test/Makefile From kientzle at FreeBSD.org Sun Jun 15 05:11:21 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 05:11:23 2008 Subject: cvs commit: src/lib/libarchive archive_string.h Message-ID: <200806150511.m5F5BLB6047351@repoman.freebsd.org> kientzle 2008-06-15 05:11:08 UTC FreeBSD src repository Modified files: lib/libarchive archive_string.h Log: SVN rev 179791 on 2008-06-15 05:11:08Z by kientzle Portability: wchar_t is defined in stdlib.h on some systems. Revision Changes Path 1.12 +3 -0 src/lib/libarchive/archive_string.h From kientzle at FreeBSD.org Sun Jun 15 05:12:52 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 05:12:55 2008 Subject: cvs commit: src/lib/libarchive archive_string.c config_freebsd.h Message-ID: <200806150512.m5F5CqbJ047471@repoman.freebsd.org> kientzle 2008-06-15 05:12:47 UTC FreeBSD src repository Modified files: lib/libarchive archive_string.c config_freebsd.h Log: SVN rev 179792 on 2008-06-15 05:12:47Z by kientzle Portability: We can get away with the older and better-supported wctomb() here; we don't need wcrtomb(). In particular, this fixes libarchive building on FreeBSD 4. Revision Changes Path 1.13 +23 -33 src/lib/libarchive/archive_string.c 1.10 +1 -0 src/lib/libarchive/config_freebsd.h From kientzle at FreeBSD.org Sun Jun 15 05:14:08 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 05:14:11 2008 Subject: cvs commit: src/lib/libarchive archive_entry_strmode.c Message-ID: <200806150514.m5F5E8mD047588@repoman.freebsd.org> kientzle 2008-06-15 05:14:01 UTC FreeBSD src repository Modified files: lib/libarchive archive_entry_strmode.c Log: SVN rev 179793 on 2008-06-15 05:14:01Z by kientzle Mark hardlinks that lack any other interesting filetype information with an 'h'. Revision Changes Path 1.4 +5 -0 src/lib/libarchive/archive_entry_strmode.c From kientzle at FreeBSD.org Sun Jun 15 05:16:09 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 05:16:12 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_zip.c Message-ID: <200806150516.m5F5G9pr047794@repoman.freebsd.org> kientzle 2008-06-15 05:15:53 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_zip.c Log: SVN rev 179794 on 2008-06-15 05:15:53Z by kientzle Fix reading TOC from zip archives with unsupported compression. We can't read the body, but we shouldn't try to skip the body twice. Revision Changes Path 1.24 +6 -2 src/lib/libarchive/archive_read_support_format_zip.c From dougb at FreeBSD.org Sun Jun 15 06:39:26 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Sun Jun 15 06:39:29 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <484D5719.1070907@FreeBSD.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> <484D5719.1070907@FreeBSD.org> Message-ID: <4854B91A.30403@FreeBSD.org> Doug Barton wrote: > Tim Kientzle wrote: >> This case is a little unusual: by default, I want to >> install both 'cpio' programs and symlink the common name >> to one of them. > > I'm not sure how that relates to Jeremie's (valid) point that these > flags need to be part of the system in bsd.own.mk. I admit that getting > stuff in there was a little intimidating to me at first, but Ruslan has > created a nice automated system, and he'll be glad to help you if you > need it. > >> How about this: >> WITH_GCPIO (default) - installs gcpio, symlinks 'cpio' unless >> WITH_BSDCPIO defined >> WITHOUT_GCPIO - does not install gcpio at all >> WITH_BSDCPIO (default) - installs bsdcpio, symlinks 'cpio' >> WITHOUT_BSDCPIO - does not install bsdcpio at all > > If by "not install" you mean "build or install" then I'm with you, but I > think it might be a little complicated. I'm not sure why you want to > build and install both by default (except maybe to ease people's > concerns in the interim period) but if it were me, in HEAD I would do this: > > WITH_BSDCPIO (default) build, install, and symlink to cpio > WITHOUT_GCPIO (default) do not build or install > > Then if WITHOUT_BSDCPIO is defined you don't build or install it, and if > WITH_GCPIO is defined you build, install, and symlink it, regardless of > the state of the BSDCPIO knob. > > When this gets MFC'ed I would just flip the defaults. I've attached a patch which does what I described above for CPIO, and adds a knob to turn off building of GNU grep. When we get to the point where bsd grep is imported into the base system, I would do the same in HEAD/8-current for bsd grep as I am proposing we do for cpio. But for now it is a step forward to be able to disable the grep build to allow for easier testing. If this stuff makes anyone happy I'm glad to commit it. Note that after you add the options files you have to run the makeman script in that directory, and redirect the output to src/share/man/man5/src.conf.5 then commit the changes there too. Side note to gabor, I noticed right away when I set up the test tonight that anything which expects there to be a grep in /usr/bin only fails, and that the base grep has 9 links, but your port only installs 6. The difference is that the base version installs links for {bze|bzf|bz}grep too. If I were you I would offer an option to create symlinks in /usr/bin, and install the bz* stuff as well. hth, Doug -- This .signature sanitized for your protection -------------- next part -------------- Index: usr.bin/Makefile =================================================================== --- usr.bin/Makefile (revision 179788) +++ usr.bin/Makefile (working copy) @@ -43,7 +43,7 @@ comm \ compile_et \ compress \ - cpio \ + ${_cpio} \ cpuset \ csplit \ ${_csup} \ @@ -257,6 +257,10 @@ _bluetooth= bluetooth .endif +.if ${MK_BSD_CPIO} != "no" +_cpio= cpio +.endif + .if ${MK_CALENDAR} != "no" _calendar= calendar .endif Index: usr.bin/cpio/Makefile =================================================================== --- usr.bin/cpio/Makefile (revision 179788) +++ usr.bin/cpio/Makefile (working copy) @@ -9,7 +9,7 @@ CFLAGS+= -DPLATFORM_CONFIG_H=\"config_freebsd.h\" LDADD+= -larchive -lz -lbz2 -.if defined(WITH_BSDCPIO) +.if ${MK_GNU_CPIO} != "yes" SYMLINKS=bsdcpio ${BINDIR}/cpio MLINKS= bsdcpio.1 cpio.1 .endif Index: gnu/usr.bin/cpio/Makefile =================================================================== --- gnu/usr.bin/cpio/Makefile (revision 179788) +++ gnu/usr.bin/cpio/Makefile (working copy) @@ -62,9 +62,7 @@ gcpio.1: ${CPIODIR}/doc/cpio.1 cat ${CPIODIR}/doc/cpio.1 >gcpio.1 -.if !defined(WITH_BSDCPIO) SYMLINKS=gcpio ${BINDIR}/cpio MLINKS=gcpio.1 cpio.1 -.endif .include Index: gnu/usr.bin/Makefile =================================================================== --- gnu/usr.bin/Makefile (revision 179788) +++ gnu/usr.bin/Makefile (working copy) @@ -5,7 +5,7 @@ SUBDIR= bc \ ${_binutils} \ ${_cc} \ - cpio \ + ${_cpio} \ ${_cvs} \ dc \ dialog \ @@ -13,7 +13,7 @@ diff3 \ ${_gdb} \ ${_gperf} \ - grep \ + ${_grep} \ ${_groff} \ ${_man} \ patch \ @@ -30,10 +30,18 @@ .endif .endif +.if ${MK_GNU_CPIO} == "yes" +_cpio= cpio +.endif + .if ${MK_CVS} != "no" _cvs= cvs .endif +.if ${MK_GNU_GREP} != "no" +_grep= grep +.endif + .if ${MK_MAN} != "no" _man= man .endif Index: share/mk/bsd.own.mk =================================================================== --- share/mk/bsd.own.mk (revision 179788) +++ share/mk/bsd.own.mk (working copy) @@ -287,6 +287,7 @@ ATM \ AUDIT \ AUTHPF \ + BSD_CPIO \ BIND \ BIND_DNSSEC \ BIND_ETC \ @@ -312,6 +313,7 @@ GCOV \ GDB \ GNU \ + GNU_GREP \ GPIB \ GROFF \ HTML \ @@ -374,6 +376,7 @@ # .for var in \ BIND_LIBS \ + GNU_CPIO \ HESIOD \ IDEA .if defined(WITH_${var}) && defined(WITHOUT_${var}) From kientzle at freebsd.org Sun Jun 15 09:34:41 2008 From: kientzle at freebsd.org (Tim Kientzle) Date: Sun Jun 15 09:34:45 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <4854B91A.30403@FreeBSD.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> <484D5719.1070907@FreeBSD.org> <4854B91A.30403@FreeBSD.org> Message-ID: <4854E230.5020004@freebsd.org> >> .... in HEAD I would do this: >> >> WITH_BSDCPIO (default) build, install, and symlink to cpio >> WITHOUT_GCPIO (default) do not build or install >> >> Then if WITHOUT_BSDCPIO is defined you don't build or install it, and >> if WITH_GCPIO is defined you build, install, and symlink it, >> regardless of the state of the BSDCPIO knob. >> >> When this gets MFC'ed I would just flip the defaults. > > I've attached a patch which does what I described above for CPIO, and > adds a knob to turn off building of GNU grep. ... > > If this stuff makes anyone happy I'm glad to commit it. Looks good to me. Go for it. Tim From kientzle at FreeBSD.org Sun Jun 15 10:08:15 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 10:08:17 2008 Subject: cvs commit: src/usr.bin/tar/test main.c test.h Message-ID: <200806151008.m5FA8Exv087438@repoman.freebsd.org> kientzle 2008-06-15 10:07:54 UTC FreeBSD src repository Modified files: usr.bin/tar/test main.c test.h Log: SVN rev 179795 on 2008-06-15 10:07:54Z by kientzle MfP4: test harness cleanup. Revision Changes Path 1.3 +6 -3 src/usr.bin/tar/test/main.c 1.3 +15 -16 src/usr.bin/tar/test/test.h From kientzle at FreeBSD.org Sun Jun 15 10:08:31 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 10:08:32 2008 Subject: cvs commit: src/usr.bin/tar subst.c Message-ID: <200806151008.m5FA8UVd087487@repoman.freebsd.org> kientzle 2008-06-15 10:08:16 UTC FreeBSD src repository Modified files: usr.bin/tar subst.c Log: SVN rev 179796 on 2008-06-15 10:08:16Z by kientzle MfP4: Minor portability fix. Revision Changes Path 1.4 +9 -1 src/usr.bin/tar/subst.c From kientzle at FreeBSD.org Sun Jun 15 10:26:17 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 10:26:21 2008 Subject: cvs commit: src/lib/libarchive archive_string.c Message-ID: <200806151026.m5FAQGBf088901@repoman.freebsd.org> kientzle 2008-06-15 10:26:08 UTC FreeBSD src repository Modified files: lib/libarchive archive_string.c Log: SVN rev 179797 on 2008-06-15 10:26:08Z by kientzle Fix a bad cast. Revision Changes Path 1.14 +1 -1 src/lib/libarchive/archive_string.c From kientzle at FreeBSD.org Sun Jun 15 10:35:33 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 10:35:35 2008 Subject: cvs commit: src/lib/libarchive/test main.c test.h test_read_extract.c test_write_disk.c test_write_disk_perms.c test_write_disk_secure.c Message-ID: <200806151035.m5FAZWC7089749@repoman.freebsd.org> kientzle 2008-06-15 10:35:22 UTC FreeBSD src repository Modified files: lib/libarchive/test main.c test.h test_read_extract.c test_write_disk.c test_write_disk_perms.c test_write_disk_secure.c Log: SVN rev 179798 on 2008-06-15 10:35:22Z by kientzle MfP4: test harness improvements. Revision Changes Path 1.13 +49 -4 src/lib/libarchive/test/main.c 1.10 +23 -16 src/lib/libarchive/test/test.h 1.4 +18 -9 src/lib/libarchive/test/test_read_extract.c 1.9 +5 -1 src/lib/libarchive/test/test_write_disk.c 1.9 +3 -3 src/lib/libarchive/test/test_write_disk_perms.c 1.4 +2 -2 src/lib/libarchive/test/test_write_disk_secure.c From kientzle at FreeBSD.org Sun Jun 15 10:44:19 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 10:44:21 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_mtree.c Message-ID: <200806151044.m5FAiIUJ090505@repoman.freebsd.org> kientzle 2008-06-15 10:43:59 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_mtree.c Log: SVN rev 179799 on 2008-06-15 10:43:59Z by kientzle Be a little more careful about closing file descriptors that may not exist. Revision Changes Path 1.8 +2 -1 src/lib/libarchive/archive_read_support_format_mtree.c From kientzle at FreeBSD.org Sun Jun 15 10:46:04 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 10:46:06 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_compression_program.c archive_write_set_compression_program.c filter_fork.c Message-ID: <200806151046.m5FAk4WB090705@repoman.freebsd.org> kientzle 2008-06-15 10:45:57 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_compression_program.c archive_write_set_compression_program.c filter_fork.c Log: SVN rev 179800 on 2008-06-15 10:45:57Z by kientzle A number of minor corrections to the support for external compression programs: * Support platforms that have fork() but not vfork() * Don't write(), select(), or poll() against closed file descriptors Revision Changes Path 1.4 +9 -2 src/lib/libarchive/archive_read_support_compression_program.c 1.3 +9 -1 src/lib/libarchive/archive_write_set_compression_program.c 1.4 +27 -10 src/lib/libarchive/filter_fork.c From kientzle at FreeBSD.org Sun Jun 15 11:15:22 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 11:15:29 2008 Subject: cvs commit: src/lib/libarchive archive_string.c Message-ID: <200806151115.m5FBFM1S094134@repoman.freebsd.org> kientzle 2008-06-15 11:15:12 UTC FreeBSD src repository Modified files: lib/libarchive archive_string.c Log: SVN rev 179801 on 2008-06-15 11:15:12Z by kientzle Since wctomb() returns int, temporaries should be int, not size_t. Revision Changes Path 1.15 +2 -2 src/lib/libarchive/archive_string.c From kientzle at FreeBSD.org Sun Jun 15 11:29:08 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sun Jun 15 11:29:16 2008 Subject: cvs commit: src/lib/libarchive archive_string.c Message-ID: <200806151129.m5FBT8gn095258@repoman.freebsd.org> kientzle 2008-06-15 11:28:56 UTC FreeBSD src repository Modified files: lib/libarchive archive_string.c Log: SVN rev 179802 on 2008-06-15 11:28:56Z by kientzle Rework the my_mbtowc_utf8() support function to fully match the mbtowc() calling convention, not the mbrtowc() convention. Revision Changes Path 1.16 +14 -14 src/lib/libarchive/archive_string.c From rrs at FreeBSD.org Sun Jun 15 12:31:40 2008 From: rrs at FreeBSD.org (Randall Stewart) Date: Sun Jun 15 12:31:41 2008 Subject: cvs commit: src/sys/netinet sctp_pcb.c sctp_pcb.h sctp_usrreq.c sctp_var.h Message-ID: <200806151231.m5FCVecW001691@repoman.freebsd.org> rrs 2008-06-15 12:31:23 UTC FreeBSD src repository Modified files: sys/netinet sctp_pcb.c sctp_pcb.h sctp_usrreq.c sctp_var.h Log: SVN rev 179803 on 2008-06-15 12:31:23Z by rrs More prep for Vimage: - only one functino to destroy an SCTP stack sctp_finish() - Make it so this function also arranges for any threads created by the image to do a kthread_exit() Revision Changes Path 1.70 +16 -0 src/sys/netinet/sctp_pcb.c 1.36 +1 -1 src/sys/netinet/sctp_pcb.h 1.57 +0 -17 src/sys/netinet/sctp_usrreq.c 1.26 +0 -1 src/sys/netinet/sctp_var.h From kmacy at FreeBSD.org Sun Jun 15 13:25:42 2008 From: kmacy at FreeBSD.org (Kip Macy) Date: Sun Jun 15 13:25:46 2008 Subject: cvs commit: src/etc devd.conf Message-ID: <200806151325.m5FDPfNp020181@repoman.freebsd.org> kmacy 2008-06-15 13:25:23 UTC FreeBSD src repository Modified files: etc devd.conf Log: SVN rev 179804 on 2008-06-15 13:25:23Z by kmacy Stop moused on a detach event. Remove incorrect comment. This fixes frequent problems with usb mice and kvm switches caused by moused hanging around. Suggested by: Matthew Dodd MFC after: 2 weeks Revision Changes Path 1.40 +5 -0 src/etc/devd.conf From kmacy at FreeBSD.org Sun Jun 15 13:26:42 2008 From: kmacy at FreeBSD.org (Kip Macy) Date: Sun Jun 15 13:26:47 2008 Subject: cvs commit: src/etc devd.conf Message-ID: <200806151326.m5FDQgD5020277@repoman.freebsd.org> kmacy 2008-06-15 13:26:25 UTC FreeBSD src repository Modified files: etc devd.conf Log: SVN rev 179805 on 2008-06-15 13:26:25Z by kmacy remove incorrect comment that I missed in my last change Revision Changes Path 1.41 +0 -2 src/etc/devd.conf From cokane at FreeBSD.org Sun Jun 15 13:37:44 2008 From: cokane at FreeBSD.org (Coleman Kane) Date: Sun Jun 15 13:37:46 2008 Subject: cvs commit: src/sys/compat/ndis subr_ntoskrnl.c Message-ID: <200806151337.m5FDbhOZ021194@repoman.freebsd.org> cokane 2008-06-15 13:37:29 UTC FreeBSD src repository Modified files: sys/compat/ndis subr_ntoskrnl.c Log: SVN rev 179806 on 2008-06-15 13:37:29Z by cokane Silence warning about missing IoGetDeviceObjectPointer by implementing a simple stub that always returns STATUS_SUCCESS. Submitted by: Paul B. Mahol Reviewed by: thompsa MFC after: 1 week Revision Changes Path 1.96 +13 -0 src/sys/compat/ndis/subr_ntoskrnl.c From keramida at FreeBSD.org Sun Jun 15 19:07:00 2008 From: keramida at FreeBSD.org (Giorgos Keramidas) Date: Sun Jun 15 19:07:12 2008 Subject: cvs commit: src/usr.bin/shar shar.sh Message-ID: <200806151906.m5FJ6xJq065797@repoman.freebsd.org> keramida 2008-06-15 14:36:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/shar shar.sh Log: SVN rev 179807 on 2008-06-15 14:36:01Z by keramida MFC -r 179355 on behalf of remko. % r179355 | remko | 2008-05-27 12:45:18 +0300 (Tue, 27 May 2008) | 11 lines % % Limit the EOF marker length to a maximum of 79 characters. [1] % % Add $FreeBSD$ tag so that I can actually commit this. % % PR: bin/118782 % Reported by: Bjoern Koenig % Patch by: edwin, Jaakko Heinonen (not used patch) % MFC after: 1 week % Approved by: imp (mentor, implicit) Revision Changes Path 1.3.48.1 +5 -2 src/usr.bin/shar/shar.sh From kib at FreeBSD.org Sun Jun 15 19:07:00 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sun Jun 15 19:07:16 2008 Subject: cvs commit: src/sys/fs/tmpfs tmpfs_subr.c Message-ID: <200806151907.m5FJ70r3065824@repoman.freebsd.org> kib 2008-06-15 18:40:58 UTC FreeBSD src repository Modified files: sys/fs/tmpfs tmpfs_subr.c Log: SVN rev 179808 on 2008-06-15 18:40:58Z by kib Do not redo the vnode tear-down work already done by insmntque() when vnode cannot be put on the vnode list for mount. Reported and tested by: marck Guilty party: me MFC after: 3 days Revision Changes Path 1.17 +1 -4 src/sys/fs/tmpfs/tmpfs_subr.c From remko at FreeBSD.org Sun Jun 15 20:12:23 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Sun Jun 15 20:12:25 2008 Subject: cvs commit: src/share/sendmail Makefile Message-ID: <200806152012.m5FKCNUW072160@repoman.freebsd.org> remko 2008-06-15 20:12:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/sendmail Makefile Log: SVN rev 179809 on 2008-06-15 20:12:11Z by remko MFC rev 179460 Dont install .svn metadata with the sendmail install. Submitted by: marcel Approved by: imp (mentor, implicit) Revision Changes Path 1.10.10.1 +2 -2 src/share/sendmail/Makefile From fjoe at FreeBSD.org Sun Jun 15 20:26:25 2008 From: fjoe at FreeBSD.org (Max Khon) Date: Sun Jun 15 20:26:26 2008 Subject: cvs commit: src/sys/sys param.h Message-ID: <200806152026.m5FKQOpD073415@repoman.freebsd.org> fjoe 2008-06-15 20:26:09 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/sys param.h Log: SVN rev 179810 on 2008-06-15 20:26:09Z by fjoe Bump __FreeBSD_version to 603104 for make(1) :u var modifier. Revision Changes Path 1.244.2.38 +1 -1 src/sys/sys/param.h From jb at FreeBSD.org Mon Jun 16 04:44:46 2008 From: jb at FreeBSD.org (John Birrell) Date: Mon Jun 16 04:44:49 2008 Subject: cvs commit: src/sys/kern kern_dtrace.c Message-ID: <200806160444.m5G4ijVM044147@repoman.freebsd.org> jb 2008-06-16 04:44:29 UTC FreeBSD src repository Modified files: sys/kern kern_dtrace.c Log: SVN rev 179811 on 2008-06-16 04:44:29Z by jb Remove code that isn't required. It actually breaks the case where KDTRACE_HOOKS is defined and KDB isn't. This is the case that it was intended for. Revision Changes Path 1.2 +0 -12 src/sys/kern/kern_dtrace.c From dougb at FreeBSD.org Mon Jun 16 05:33:32 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 16 05:33:35 2008 Subject: cvs commit: src/share/man/man5 src.conf.5 Message-ID: <200806160533.m5G5XWiT058276@repoman.freebsd.org> dougb 2008-06-16 05:33:16 UTC FreeBSD src repository Modified files: share/man/man5 src.conf.5 Log: SVN rev 179812 on 2008-06-16 05:33:16Z by dougb Regenerate using svn $FreeBSD$ tags. There is no change in the actual man page text other than the date. Revision Changes Path 1.27 +87 -87 src/share/man/man5/src.conf.5 From dougb at FreeBSD.org Mon Jun 16 05:53:10 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 16 05:53:12 2008 Subject: cvs commit: src/gnu/usr.bin Makefile src/gnu/usr.bin/cpio Makefile src/share/mk bsd.own.mk src/tools/build/options WITHOUT_BSD_CPIO WITHOUT_GNU_GREP WITH_GNU_CPIO src/usr.bin Makefile src/usr.bin/cpio Makefile Message-ID: <200806160553.m5G5r4Ub060117@repoman.freebsd.org> dougb 2008-06-16 05:48:15 UTC FreeBSD src repository Modified files: gnu/usr.bin Makefile gnu/usr.bin/cpio Makefile share/mk bsd.own.mk usr.bin Makefile usr.bin/cpio Makefile Added files: tools/build/options WITHOUT_BSD_CPIO WITHOUT_GNU_GREP WITH_GNU_CPIO Log: SVN rev 179813 on 2008-06-16 05:48:15Z by dougb 1. Make the BSD version of cpio the default [1] a. The BSD version will be built and installed unless WITHOUT_BSD_CPIO is defined. b. The GNU version will not be built or installed unless WITH_GNU_CPIO is defined. If this is defined, the symlink in /usr/bin will be to the GNU version whether the BSD version is present or not. When these changes are MFCed the defaults should be flipped. 2. Add a knob to disable the building of GNU grep. This will make it easier for those that want to test the BSD version in the ports. Approved by: kientzle [1] Revision Changes Path 1.97 +10 -2 src/gnu/usr.bin/Makefile 1.14 +0 -2 src/gnu/usr.bin/cpio/Makefile 1.76 +3 -0 src/share/mk/bsd.own.mk 1.1 +3 -0 src/tools/build/options/WITHOUT_BSD_CPIO (new) 1.1 +2 -0 src/tools/build/options/WITHOUT_GNU_GREP (new) 1.1 +6 -0 src/tools/build/options/WITH_GNU_CPIO (new) 1.315 +5 -1 src/usr.bin/Makefile 1.3 +1 -1 src/usr.bin/cpio/Makefile From dougb at FreeBSD.org Mon Jun 16 05:54:18 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 16 05:54:22 2008 Subject: cvs commit: src/share/man/man5 src.conf.5 Message-ID: <200806160554.m5G5sIFI060214@repoman.freebsd.org> dougb 2008-06-16 05:50:21 UTC FreeBSD src repository Modified files: share/man/man5 src.conf.5 Log: SVN rev 179814 on 2008-06-16 05:50:21Z by dougb Regenerate to add the definitions of WITHOUT_BSD_CPIO WITH_GNU_CPIO and WITHOUT_GNU_GREP Revision Changes Path 1.28 +14 -0 src/share/man/man5/src.conf.5 From dougb at FreeBSD.org Mon Jun 16 07:23:20 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 16 07:23:22 2008 Subject: cvs commit: src/share/mk bsd.own.mk Message-ID: <200806160723.m5G7NKwX071003@repoman.freebsd.org> dougb 2008-06-16 07:23:12 UTC FreeBSD src repository Modified files: share/mk bsd.own.mk Log: SVN rev 179815 on 2008-06-16 07:23:12Z by dougb Properly alphabetize the BSD_CPIO option Revision Changes Path 1.77 +1 -1 src/share/mk/bsd.own.mk From dougb at FreeBSD.org Mon Jun 16 07:24:21 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 16 07:24:30 2008 Subject: cvs commit: src/usr.bin/cpio Makefile Message-ID: <200806160724.m5G7OLi8071091@repoman.freebsd.org> dougb 2008-06-16 07:24:05 UTC FreeBSD src repository Modified files: usr.bin/cpio Makefile Log: SVN rev 179816 on 2008-06-16 07:24:05Z by dougb Include bsd.own.mk to pick up the definition of MK_GNU_CPIO Revision Changes Path 1.4 +2 -0 src/usr.bin/cpio/Makefile From flz at FreeBSD.org Mon Jun 16 09:03:09 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Mon Jun 16 09:03:19 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add perform.c src/usr.sbin/pkg_install/lib lib.h src/usr.sbin/pkg_install/updating main.c Message-ID: <200806160903.m5G9389v090499@repoman.freebsd.org> flz 2008-06-16 09:02:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/pkg_install/add perform.c usr.sbin/pkg_install/lib lib.h usr.sbin/pkg_install/updating main.c Log: SVN rev 179817 on 2008-06-16 09:02:59Z by flz Synchronize pkg_install with HEAD (20080612) r179760: fix recursive -K (add), '\n' not part of pkg origin (updating) Revision Changes Path 1.80.2.3 +1 -1 src/usr.sbin/pkg_install/add/perform.c 1.60.2.4 +1 -1 src/usr.sbin/pkg_install/lib/lib.h 1.2.2.3 +2 -0 src/usr.sbin/pkg_install/updating/main.c From flz at FreeBSD.org Mon Jun 16 09:04:11 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Mon Jun 16 09:04:13 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add perform.c src/usr.sbin/pkg_install/lib lib.h src/usr.sbin/pkg_install/updating main.c Message-ID: <200806160904.m5G94BLP090630@repoman.freebsd.org> flz 2008-06-16 09:03:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/pkg_install/add perform.c usr.sbin/pkg_install/lib lib.h usr.sbin/pkg_install/updating main.c Log: SVN rev 179818 on 2008-06-16 09:03:57Z by flz Synchronize pkg_install with HEAD (20080612) r179760: fix recursive -K (add), '\n' not part of pkg origin (updating) Revision Changes Path 1.77.8.6 +1 -1 src/usr.sbin/pkg_install/add/perform.c 1.56.2.6 +1 -1 src/usr.sbin/pkg_install/lib/lib.h 1.2.4.3 +2 -0 src/usr.sbin/pkg_install/updating/main.c From flz at FreeBSD.org Mon Jun 16 09:15:44 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Mon Jun 16 09:15:59 2008 Subject: cvs commit: src/usr.sbin/pkg_install Makefile Message-ID: <200806160915.m5G9FhXe091644@repoman.freebsd.org> flz 2008-06-16 09:15:27 UTC FreeBSD src repository Modified files: usr.sbin/pkg_install Makefile Log: SVN rev 179819 on 2008-06-16 09:15:27Z by flz Style fix (use naked commands). Reported by: obrien Revision Changes Path 1.25 +3 -7 src/usr.sbin/pkg_install/Makefile From flz at FreeBSD.org Mon Jun 16 09:18:00 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Mon Jun 16 09:18:09 2008 Subject: cvs commit: src/usr.sbin/pkg_install Makefile Message-ID: <200806160917.m5G9Hx42091804@repoman.freebsd.org> flz 2008-06-16 09:17:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/pkg_install Makefile Log: SVN rev 179820 on 2008-06-16 09:17:46Z by flz MFC: style fixes (naked commands, r179819) Revision Changes Path 1.16.2.3 +3 -7 src/usr.sbin/pkg_install/Makefile From flz at FreeBSD.org Mon Jun 16 09:18:01 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Mon Jun 16 09:18:10 2008 Subject: cvs commit: src/usr.sbin/pkg_install Makefile Message-ID: <200806160918.m5G9I0DV091838@repoman.freebsd.org> flz 2008-06-16 09:17:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/pkg_install Makefile Log: SVN rev 179821 on 2008-06-16 09:17:59Z by flz MFC: style fixes (naked commands, r179819) Revision Changes Path 1.17.2.3 +3 -7 src/usr.sbin/pkg_install/Makefile From obrien at FreeBSD.org Mon Jun 16 14:18:56 2008 From: obrien at FreeBSD.org (David O'Brien) Date: Mon Jun 16 14:19:04 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <4854B91A.30403@FreeBSD.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> <484D5719.1070907@FreeBSD.org> <4854B91A.30403@FreeBSD.org> Message-ID: <20080616141855.GG74595@dragon.NUXI.org> On Sat, Jun 14, 2008 at 11:39:22PM -0700, Doug Barton wrote: > I've attached a patch which does what I described above for CPIO, and > adds a knob to turn off building of GNU grep. When we get to the point > where bsd grep is imported into the base system, I would do the same in > HEAD/8-current for bsd grep as I am proposing we do for cpio. Doug, Before changing the default grep away from GNU grep, please have a full and wide discussion. cpio is one thing, but GNU grep is faster than any alternative so far (educated me if other greps are finally faster). Speed of grep is a tier-1 feature. -- -- David (obrien@FreeBSD.org) From gnn at FreeBSD.org Mon Jun 16 14:34:07 2008 From: gnn at FreeBSD.org (George V. Neville-Neil) Date: Mon Jun 16 14:34:13 2008 Subject: cvs commit: src/share/man/man4 crypto.4 Message-ID: <200806161434.m5GEY7Uo036443@repoman.freebsd.org> gnn 2008-06-16 14:33:54 UTC FreeBSD src repository Modified files: share/man/man4 crypto.4 Log: SVN rev 179822 on 2008-06-16 14:33:54Z by gnn Update to include the Camellia algorithm which is in the code but which was accidentally left undocumented in the manual page. Revision Changes Path 1.8 +1 -0 src/share/man/man4/crypto.4 From fanf at FreeBSD.org Mon Jun 16 14:50:38 2008 From: fanf at FreeBSD.org (Tony Finch) Date: Mon Jun 16 14:50:44 2008 Subject: cvs commit: src/lib/libc/sys setgroups.2 Message-ID: <200806161450.m5GEobfn037786@repoman.freebsd.org> fanf 2008-06-16 14:50:21 UTC FreeBSD src repository Modified files: lib/libc/sys setgroups.2 Log: SVN rev 179823 on 2008-06-16 14:50:21Z by fanf Make it clearer that privilege is needed to reduce as well as increase group membership. Revision Changes Path 1.15 +1 -1 src/lib/libc/sys/setgroups.2 From dougb at FreeBSD.org Mon Jun 16 15:42:12 2008 From: dougb at FreeBSD.org (Doug Barton) Date: Mon Jun 16 15:42:20 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <20080616141855.GG74595@dragon.NUXI.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> <484D5719.1070907@FreeBSD.org> <4854B91A.30403@FreeBSD.org> <20080616141855.GG74595@dragon.NUXI.org> Message-ID: <485689D0.2000009@FreeBSD.org> David O'Brien wrote: > On Sat, Jun 14, 2008 at 11:39:22PM -0700, Doug Barton wrote: >> I've attached a patch which does what I described above for CPIO, and >> adds a knob to turn off building of GNU grep. When we get to the point >> where bsd grep is imported into the base system, I would do the same in >> HEAD/8-current for bsd grep as I am proposing we do for cpio. > > Doug, > Before changing the default grep away from GNU grep, please have a full > and wide discussion. FWIW, I'm not the driving force for changing cpio, I'm just helping with the knobs. > cpio is one thing, but GNU grep is faster than any > alternative so far (educated me if other greps are finally faster). According to Gabor his latest version compares favorably. However, when I did some quick tests of his new version the first two tests I did were unsuccessful based on compatibility issues, so I'm not in any way suggesting that we're ready to move forward on that. I added a knob to NOT build gnu grep which will make further testing easier for those who are interested. > Speed of grep is a tier-1 feature. I agree with you on that, but keep in mind that _if_ we want to ship 8.0-RELEASE with a BSD licensed grep the default would have to be changed in HEAD sooner rather than later. That's the only way it'll get any kind of useful testing. Default settings to build old versions of critical tools can always be reversed prior to a -RELEASE. Doug -- This .signature sanitized for your protection From obrien at FreeBSD.org Mon Jun 16 16:26:24 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Mon Jun 16 16:26:33 2008 Subject: cvs commit: src/share/man/man4 ciss.4 src/sys/dev/ciss ciss.c src/sys/vm vm_kern.c src/usr.bin/cksum cksum.c Message-ID: <200806161626.m5GGQOqg051639@repoman.freebsd.org> obrien 2008-06-16 16:23:09 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 ciss.4 sys/dev/ciss ciss.c sys/vm vm_kern.c usr.bin/cksum cksum.c Log: SVN rev 179824 on 2008-06-16 16:23:09Z by obrien MFC rev 1.126 / r165854: Declare the map entry created by kmem_init() for the range from VM_MIN_KERNEL_ADDRESS to the end of the kernel's bootstrap data as MAP_NOFAULT. Revision Changes Path 1.11.2.5 +0 -0 src/share/man/man4/ciss.4 1.64.2.7 +0 -0 src/sys/dev/ciss/ciss.c 1.122.2.1 +2 -1 src/sys/vm/vm_kern.c 1.17.12.2 +0 -0 src/usr.bin/cksum/cksum.c From olli at FreeBSD.org Mon Jun 16 17:04:23 2008 From: olli at FreeBSD.org (Oliver Fromme) Date: Mon Jun 16 17:04:31 2008 Subject: cvs commit: src/sys/boot/i386/libi386 time.c Message-ID: <200806161704.m5GH4Mms066706@repoman.freebsd.org> olli 2008-06-16 17:04:04 UTC FreeBSD src repository Modified files: sys/boot/i386/libi386 time.c Log: SVN rev 179825 on 2008-06-16 17:04:04Z by olli Implement a workaround for a long-standing problem in libi386's time(), caused by a qemu bug. The bug might be present in other BIOSes, too. qemu either does not simulate the AT RTC correctly or has a broken BIOS 1A/02 implementation, and will return an incorrect value if the RTC is read while it is being updated. The effect is worsened by the fact that qemu's INT 15/86 function ("wait" a.k.a. usleep) is non-implmeneted or broken and returns immediately, causing beastie.4th to spin in a tight loop calling the "read RTC" function millions of times, triggering the problem quickly. Therefore, we keep reading the BIOS value until we get the same result twice. This change fixes beastie.4th's countdown under qemu. Approved by: des (mentor) Revision Changes Path 1.6 +32 -8 src/sys/boot/i386/libi386/time.c From obrien at FreeBSD.org Mon Jun 16 17:06:23 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Mon Jun 16 17:06:28 2008 Subject: cvs commit: src/contrib/cvs/src classify.c Message-ID: <200806161706.m5GH6NoK066970@repoman.freebsd.org> obrien 2008-06-16 17:06:17 UTC FreeBSD src repository Modified files: contrib/cvs/src classify.c Log: SVN rev 179826 on 2008-06-16 17:06:17Z by obrien Add $FreeBSD$ since we now have local changes. Revision Changes Path 1.4 +3 -0 src/contrib/cvs/src/classify.c From kib at FreeBSD.org Mon Jun 16 17:25:40 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Mon Jun 16 17:25:47 2008 Subject: cvs commit: src/sys/sys param.h Message-ID: <200806161725.m5GHPdUD073531@repoman.freebsd.org> kib 2008-06-16 17:25:20 UTC FreeBSD src repository Modified files: sys/sys param.h Log: SVN rev 179827 on 2008-06-16 17:25:20Z by kib Add the member2struct() macro, that returns pointer to the containing structure given pointer to some structure member. MFC after: 2 weeks Revision Changes Path 1.357 +7 -0 src/sys/sys/param.h From obrien at FreeBSD.org Mon Jun 16 17:34:30 2008 From: obrien at FreeBSD.org (David O'Brien) Date: Mon Jun 16 17:34:37 2008 Subject: cvs commit [HEAD] src/gnu/usr.bin/cpio Makefile In-Reply-To: <485689D0.2000009@FreeBSD.org> References: <200805261721.m4QHLCF4070787@repoman.freebsd.org> <20080608075256.GA35560@obiwan.tataz.chchile.org> <484CB927.3050101@freebsd.org> <484D5719.1070907@FreeBSD.org> <4854B91A.30403@FreeBSD.org> <20080616141855.GG74595@dragon.NUXI.org> <485689D0.2000009@FreeBSD.org> Message-ID: <20080616173428.GA38024@dragon.NUXI.org> On Mon, Jun 16, 2008 at 08:42:08AM -0700, Doug Barton wrote: >> Speed of grep is a tier-1 feature. > > I agree with you on that, but keep in mind that _if_ we want to ship > 8.0-RELEASE with a BSD licensed grep the default would have to be changed > in HEAD sooner rather than later. That's the only way it'll get any kind of > useful testing. Default settings to build old versions of critical tools > can always be reversed prior to a -RELEASE. I've no problems with the knob - and would like to see a BSDL grep'ed brought in and at least have available [to live beside GNU grep]. My only request was to not throw the /usr/bin/grep != GNU grep switch without much thought. -- -- David (obrien@FreeBSD.org) From kib at FreeBSD.org Mon Jun 16 17:42:03 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Mon Jun 16 17:42:07 2008 Subject: cvs commit: src/sys/fs/devfs devfs_devs.c devfs_int.h devfs_vnops.c src/sys/kern kern_conf.c src/sys/sys conf.h src/usr.bin/fstat fstat.c Message-ID: <200806161742.m5GHg2jj075120@repoman.freebsd.org> kib 2008-06-16 17:34:59 UTC FreeBSD src repository Modified files: sys/fs/devfs devfs_devs.c devfs_int.h devfs_vnops.c sys/kern kern_conf.c sys/sys conf.h usr.bin/fstat fstat.c Log: SVN rev 179828 on 2008-06-16 17:34:59Z by kib Struct cdev is always the member of the struct cdev_priv. When devfs needed to promote cdev to cdev_priv, the si_priv pointer was followed. Use member2struct() to calculate address of the wrapping cdev_priv. Rename si_priv to __si_reserved. Tested by: pho Reviewed by: ed MFC after: 2 weeks Revision Changes Path 1.54 +3 -4 src/sys/fs/devfs/devfs_devs.c 1.6 +2 -0 src/sys/fs/devfs/devfs_int.h 1.163 +5 -5 src/sys/fs/devfs/devfs_vnops.c 1.218 +6 -6 src/sys/kern/kern_conf.c 1.242 +1 -1 src/sys/sys/conf.h 1.69 +2 -3 src/usr.bin/fstat/fstat.c From eri at FreeBSD.org Mon Jun 16 17:43:36 2008 From: eri at FreeBSD.org (Ermal Luçi) Date: Mon Jun 16 17:43:43 2008 Subject: cvs commit: src/usr.bin/calendar/calendars calendar.freebsd Message-ID: <200806161743.m5GHhZ4S075271@repoman.freebsd.org> eri 2008-06-16 17:35:34 UTC FreeBSD src repository Modified files: usr.bin/calendar/calendars calendar.freebsd Log: SVN rev 179829 on 2008-06-16 17:35:34Z by eri Add my birthday to the calendar. Approved by: mlaier (mentor) Revision Changes Path 1.239 +1 -0 src/usr.bin/calendar/calendars/calendar.freebsd From eri at FreeBSD.org Mon Jun 16 17:43:36 2008 From: eri at FreeBSD.org (Ermal Luçi) Date: Mon Jun 16 17:43:44 2008 Subject: cvs commit: src/share/misc committers-src.dot Message-ID: <200806161743.m5GHhasF075299@repoman.freebsd.org> eri 2008-06-16 17:38:48 UTC FreeBSD src repository Modified files: share/misc committers-src.dot Log: SVN rev 179830 on 2008-06-16 17:38:48Z by eri Record my roots. Reviewed by: mlaier (mentor) Revision Changes Path 1.75 +3 -0 src/share/misc/committers-src.dot From remko at FreeBSD.org Mon Jun 16 18:32:38 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Mon Jun 16 18:32:40 2008 Subject: cvs commit: src/sys/pci if_rl.c if_rlreg.h Message-ID: <200806161832.m5GIWbQH080784@repoman.freebsd.org> remko 2008-06-16 18:32:20 UTC FreeBSD src repository Modified files: sys/pci if_rl.c if_rlreg.h Log: SVN rev 179831 on 2008-06-16 18:32:20Z by remko Add another 8139D variant. PR: 124622 Submitted by: Evgeny Zhirnov Approved by: imp (mentor, implicit) MFC after: 3 days Revision Changes Path 1.175 +2 -0 src/sys/pci/if_rl.c 1.75 +1 -0 src/sys/pci/if_rlreg.h From ups at FreeBSD.org Mon Jun 16 19:57:09 2008 From: ups at FreeBSD.org (Stephan Uphoff) Date: Mon Jun 16 19:57:11 2008 Subject: cvs commit: src/sys/netinet tcp_syncache.c Message-ID: <200806161957.m5GJv9mR089112@repoman.freebsd.org> ups 2008-06-16 19:56:59 UTC FreeBSD src repository Modified files: sys/netinet tcp_syncache.c Log: SVN rev 179832 on 2008-06-16 19:56:59Z by ups Fix a check in SYN cache expansion (syncache_expand()) to accept packets that arrive in the receive window instead of just on the left edge of the receive window. This is needed for correct behavior when packets are lost or reordered. PR: kern/123950 Reviewed by: andre@, silby@ Reported by: Yahoo!, Wang Jin MFC after: 1 week Revision Changes Path 1.146 +7 -4 src/sys/netinet/tcp_syncache.c From ups at FreeBSD.org Mon Jun 16 20:08:41 2008 From: ups at FreeBSD.org (Stephan Uphoff) Date: Mon Jun 16 20:08:45 2008 Subject: cvs commit: src/sys/netinet tcp_syncache.c Message-ID: <200806162008.m5GK8eeN091141@repoman.freebsd.org> ups 2008-06-16 20:08:22 UTC FreeBSD src repository Modified files: sys/netinet tcp_syncache.c Log: SVN rev 179833 on 2008-06-16 20:08:22Z by ups Change incorrect stale cookie detection in syncookie_lookup() that prematurely declared a cookie as expired. Reviewed by: andre@, silby@ Reported by: Yahoo! Revision Changes Path 1.147 +1 -1 src/sys/netinet/tcp_syncache.c From kan at FreeBSD.org Mon Jun 16 22:49:43 2008 From: kan at FreeBSD.org (Alexander Kabaev) Date: Mon Jun 16 22:49:47 2008 Subject: cvs commit: src/gnu/lib/libstdc++ Makefile Message-ID: <200806162249.m5GMngDb018760@repoman.freebsd.org> kan 2008-06-16 22:49:30 UTC FreeBSD src repository Modified files: gnu/lib/libstdc++ Makefile Log: SVN rev 179834 on 2008-06-16 22:49:30Z by kan Install extra include files that were forgotten in original GCC 4.2.x import changes. PR: 124647 Submitted by: Vlad GALU MFC after: 2 days Revision Changes Path 1.62 +15 -1 src/gnu/lib/libstdc++/Makefile From flz at FreeBSD.org Mon Jun 16 23:41:29 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Mon Jun 16 23:41:31 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c perform.c src/usr.sbin/pkg_install/create main.c src/usr.sbin/pkg_install/lib file.c url.c Message-ID: <200806162341.m5GNfTSf024576@repoman.freebsd.org> flz 2008-06-16 23:41:11 UTC FreeBSD src repository Modified files: usr.sbin/pkg_install/add main.c perform.c usr.sbin/pkg_install/create main.c usr.sbin/pkg_install/lib file.c url.c Log: SVN rev 179835 on 2008-06-16 23:41:11Z by flz Remove support for RELENG_4 (__FreeBSD_version < 500039). MFC after: 1 day Revision Changes Path 1.78 +1 -6 src/usr.sbin/pkg_install/add/main.c 1.86 +0 -4 src/usr.sbin/pkg_install/add/perform.c 1.45 +0 -4 src/usr.sbin/pkg_install/create/main.c 1.69 +0 -4 src/usr.sbin/pkg_install/lib/file.c 1.10 +0 -12 src/usr.sbin/pkg_install/lib/url.c From jasone at FreeBSD.org Mon Jun 16 23:42:16 2008 From: jasone at FreeBSD.org (Jason Evans) Date: Mon Jun 16 23:42:18 2008 Subject: cvs commit: src/lib/libc/stdlib malloc.c rb.h Message-ID: <200806162342.m5GNgFTU024655@repoman.freebsd.org> jasone 2008-06-16 23:42:05 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/stdlib malloc.c Added files: (Branch: RELENG_7) lib/libc/stdlib rb.h Log: SVN rev 179836 on 2008-06-16 23:42:05Z by jasone MFC allocator improvements and fixes: * Implement more compact red-black trees, thus reducing memory usage by ~0.5-1%. * Add a separate tree to track dirty-page-containing chunks, thus improving worst case allocation performance. * Fix a deadlock in base_alloc() for the error (OOM) path. * Catch integer overflow for huge allocations when using sbrk(2). * Fix bit vector initialization for run headers. This fix has no practical impact for correct programs. Incorrect programs will potentially experience allocation failures rather than memory corruption, both of which are "undefined behavior". Revision Changes Path 1.147.2.3 +209 -162 src/lib/libc/stdlib/malloc.c 1.4.2.1 +947 -0 src/lib/libc/stdlib/rb.h (new) From benno at FreeBSD.org Tue Jun 17 05:48:49 2008 From: benno at FreeBSD.org (Benno Rice) Date: Tue Jun 17 05:48:52 2008 Subject: cvs commit: src/sys/dev/smc if_smc.c Message-ID: <200806170548.m5H5mnnP079768@repoman.freebsd.org> benno 2008-06-17 05:48:42 UTC FreeBSD src repository Modified files: sys/dev/smc if_smc.c Log: SVN rev 179837 on 2008-06-17 05:48:42Z by benno - Move ether_ifdetach earlier. - Drain callouts after ether_ifdetach. Suggested by: jhb Revision Changes Path 1.6 +7 -2 src/sys/dev/smc/if_smc.c From davidxu at FreeBSD.org Tue Jun 17 06:33:22 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Jun 17 06:33:25 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c Message-ID: <200806170633.m5H6XMJH084600@repoman.freebsd.org> davidxu 2008-06-17 06:26:29 UTC FreeBSD src repository Modified files: include Makefile unistd.h lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c Added files: include spawn.h lib/libc/gen posix_spawn.c Log: SVN rev 179838 on 2008-06-17 06:26:29Z by davidxu Add POSIX routines called posix_spawn() and posix_spawnp(), which can be used as replacements for exec/fork in a lot of cases. This change also added execvpe() which allows environment variable PATH to be used for searching executable file, it is used for implementing posix_spawnp(). PR: standards/122051 Revision Changes Path 1.280 +1 -1 src/include/Makefile 1.1 +115 -0 src/include/spawn.h (new) 1.88 +1 -0 src/include/unistd.h 1.136 +2 -2 src/lib/libc/gen/Makefile.inc 1.11 +22 -0 src/lib/libc/gen/Symbol.map 1.27 +14 -3 src/lib/libc/gen/exec.3 1.24 +22 -9 src/lib/libc/gen/exec.c 1.1 +472 -0 src/lib/libc/gen/posix_spawn.c (new) From peter at FreeBSD.org Tue Jun 17 06:35:00 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Tue Jun 17 06:35:04 2008 Subject: cvs commit: src/share/man/man4 ciss.4 Message-ID: <200806170634.m5H6YxxY084734@repoman.freebsd.org> peter 2008-06-17 06:34:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 ciss.4 Log: SVN rev 179839 on 2008-06-17 06:34:48Z by peter Fix a bad mergeinfo path Revision Changes Path 1.14.2.3 +0 -0 src/share/man/man4/ciss.4 From ed at FreeBSD.org Tue Jun 17 07:15:13 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Tue Jun 17 07:15:16 2008 Subject: cvs commit: src/include spawn.h src/lib/libc/gen posix_spawn.c Message-ID: <200806170715.m5H7FCcY089045@repoman.freebsd.org> ed 2008-06-17 07:09:58 UTC FreeBSD src repository Modified files: include spawn.h lib/libc/gen posix_spawn.c Log: SVN rev 179840 on 2008-06-17 07:09:58Z by ed Change my email address to the one from the FreeBSD project. Approved by: philip (mentor, implicit), davidxu Revision Changes Path 1.2 +1 -1 src/include/spawn.h 1.2 +1 -1 src/lib/libc/gen/posix_spawn.c From davidxu at FreeBSD.org Tue Jun 17 08:24:05 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Jun 17 08:24:10 2008 Subject: cvs commit: src/lib/libc/gen posix_spawn.c Message-ID: <200806170824.m5H8O5gO095582@repoman.freebsd.org> davidxu 2008-06-17 08:23:45 UTC FreeBSD src repository Modified files: lib/libc/gen posix_spawn.c Log: SVN rev 179841 on 2008-06-17 08:23:45Z by davidxu Style fix. Revision Changes Path 1.3 +2 -1 src/lib/libc/gen/posix_spawn.c From bz at FreeBSD.org Tue Jun 17 09:11:50 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Tue Jun 17 09:11:51 2008 Subject: cvs commit: src Makefile Message-ID: <200806170911.m5H9BncF010061@repoman.freebsd.org> bz 2008-06-17 09:06:46 UTC FreeBSD src repository Modified files: . Makefile Log: SVN rev 179842 on 2008-06-17 09:06:46Z by bz For make universe, in addition to make.conf, also ignore a src.conf with possibe non-default options. Reviewed by: ru MFC after: 10 days Revision Changes Path 1.352 +2 -2 src/Makefile From phk at phk.freebsd.dk Tue Jun 17 10:23:00 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Tue Jun 17 10:23:05 2008 Subject: cvs commit: src Makefile In-Reply-To: Your message of "Tue, 17 Jun 2008 09:06:46 GMT." <200806170911.m5H9BncF010061@repoman.freebsd.org> Message-ID: <75185.1213698178@critter.freebsd.dk> In message <200806170911.m5H9BncF010061@repoman.freebsd.org>, "Bjoern A. Zeeb" writes: >bz 2008-06-17 09:06:46 UTC > > FreeBSD src repository > > Modified files: > . Makefile > Log: > SVN rev 179842 on 2008-06-17 09:06:46Z by bz > > For make universe, in addition to make.conf, also ignore a > src.conf with possibe non-default options. Actually, that was deliberately not ignored to make it possible to test nonstandard trees with universe also. -- 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 flz at FreeBSD.org Tue Jun 17 10:47:43 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 17 10:47:47 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c perform.c src/usr.sbin/pkg_install/create main.c src/usr.sbin/pkg_install/lib file.c url.c Message-ID: <200806171047.m5HAlh4b018788@repoman.freebsd.org> flz 2008-06-17 10:47:24 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/pkg_install/add main.c perform.c usr.sbin/pkg_install/create main.c usr.sbin/pkg_install/lib file.c url.c Log: SVN rev 179843 on 2008-06-17 10:47:24Z by flz MFC: remove 4.x support (r179835) Revision Changes Path 1.72.2.4 +1 -6 src/usr.sbin/pkg_install/add/main.c 1.80.2.4 +0 -4 src/usr.sbin/pkg_install/add/perform.c 1.40.2.3 +0 -4 src/usr.sbin/pkg_install/create/main.c 1.68.18.1 +0 -4 src/usr.sbin/pkg_install/lib/file.c 1.9.2.1 +0 -12 src/usr.sbin/pkg_install/lib/url.c From flz at FreeBSD.org Tue Jun 17 10:49:00 2008 From: flz at FreeBSD.org (Florent Thoumie) Date: Tue Jun 17 10:49:02 2008 Subject: cvs commit: src/usr.sbin/pkg_install/add main.c perform.c src/usr.sbin/pkg_install/create main.c src/usr.sbin/pkg_install/lib file.c url.c Message-ID: <200806171049.m5HAn0Up018934@repoman.freebsd.org> flz 2008-06-17 10:48:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/pkg_install/add main.c perform.c usr.sbin/pkg_install/create main.c usr.sbin/pkg_install/lib file.c url.c Log: SVN rev 179844 on 2008-06-17 10:48:46Z by flz MFC: remove 4.x support (r179835) Revision Changes Path 1.61.2.13 +1 -6 src/usr.sbin/pkg_install/add/main.c 1.77.8.7 +0 -4 src/usr.sbin/pkg_install/add/perform.c 1.36.2.5 +0 -4 src/usr.sbin/pkg_install/create/main.c 1.68.8.1 +0 -4 src/usr.sbin/pkg_install/lib/file.c 1.4.10.4 +0 -12 src/usr.sbin/pkg_install/lib/url.c From bz at FreeBSD.org Tue Jun 17 10:50:07 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Tue Jun 17 10:50:13 2008 Subject: cvs commit: src Makefile In-Reply-To: <75185.1213698178@critter.freebsd.dk> References: <75185.1213698178@critter.freebsd.dk> Message-ID: <20080617103603.W83875@maildrop.int.zabbadoz.net> On Tue, 17 Jun 2008, Poul-Henning Kamp wrote: > In message <200806170911.m5H9BncF010061@repoman.freebsd.org>, "Bjoern A. Zeeb" > writes: >> bz 2008-06-17 09:06:46 UTC >> >> FreeBSD src repository >> >> Modified files: >> . Makefile >> Log: >> SVN rev 179842 on 2008-06-17 09:06:46Z by bz >> >> For make universe, in addition to make.conf, also ignore a >> src.conf with possibe non-default options. > > Actually, that was deliberately not ignored to make it possible > to test nonstandard trees with universe also. Okay. I am not sure how it worked in the old make.conf only days. I usually forget that I do have a per machine /etc/src.conf until the next morning, thus nightly universe do not help me before committing because it doesn't give me the builds that everyone else/the tb is building. And we are still setting __MAKE_CONF to /dev/null thus any private CFLAGS etc. cannot be tested this way either. Maybe we need a better overall solution? I have no idea if we can implement it only for universe but what about something like this? - do not pre-define __MAKE_CONF/SRCCONF for universe - if __MAKE_CONF/SRCCONF is not defined manually (env) we'll ingore it (it's not set) - if it is defined (env) use it -- Bjoern A. Zeeb Stop bit received. Insert coin for new game. From phk at phk.freebsd.dk Tue Jun 17 10:52:41 2008 From: phk at phk.freebsd.dk (Poul-Henning Kamp) Date: Tue Jun 17 10:52:47 2008 Subject: cvs commit: src Makefile In-Reply-To: Your message of "Tue, 17 Jun 2008 10:47:31 GMT." <20080617103603.W83875@maildrop.int.zabbadoz.net> Message-ID: <75336.1213699959@critter.freebsd.dk> In message <20080617103603.W83875@maildrop.int.zabbadoz.net>, "Bjoern A. Zeeb" writes: >>> For make universe, in addition to make.conf, also ignore a >>> src.conf with possibe non-default options. >> >> Actually, that was deliberately not ignored to make it possible >> to test nonstandard trees with universe also. > >Maybe we need a better overall solution? Dunno, I'm not particular religious about it, I merely pointed out that it was not a random oversight. -- 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 bz at FreeBSD.org Tue Jun 17 11:12:53 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Tue Jun 17 11:13:04 2008 Subject: cvs commit: src Makefile Message-ID: <200806171112.m5HBCrmE022303@repoman.freebsd.org> bz 2008-06-17 11:08:49 UTC FreeBSD src repository Modified files: . Makefile Log: SVN rev 179845 on 2008-06-17 11:08:49Z by bz Back out rev. 1.352 (SVN rev 179842) as phk pointed out that SRCCONF was omitted here to be able to build non standard trees with universe as well. Revision Changes Path 1.353 +2 -2 src/Makefile From bz at FreeBSD.org Tue Jun 17 11:13:08 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Tue Jun 17 11:13:11 2008 Subject: cvs commit: src Makefile In-Reply-To: <75336.1213699959@critter.freebsd.dk> References: <75336.1213699959@critter.freebsd.dk> Message-ID: <20080617111035.W83875@maildrop.int.zabbadoz.net> On Tue, 17 Jun 2008, Poul-Henning Kamp wrote: > In message <20080617103603.W83875@maildrop.int.zabbadoz.net>, "Bjoern A. Zeeb" > writes: > >>>> For make universe, in addition to make.conf, also ignore a >>>> src.conf with possibe non-default options. >>> >>> Actually, that was deliberately not ignored to make it possible >>> to test nonstandard trees with universe also. >> >> Maybe we need a better overall solution? > > Dunno, I'm not particular religious about it, I merely pointed > out that it was not a random oversight. Ok, old way of how things worked restored as I can see that it might be usefull to various setups. This will also give us enough time to come up with a better solution instead of trying to find a fix asap. -- Bjoern A. Zeeb Stop bit received. Insert coin for new game. From das at FreeBSD.ORG Tue Jun 17 13:47:44 2008 From: das at FreeBSD.ORG (David Schultz) Date: Tue Jun 17 13:47:47 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <200806170633.m5H6XMJH084600@repoman.freebsd.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> Message-ID: <20080617134828.GA30076@zim.MIT.EDU> On Tue, Jun 17, 2008, David Xu wrote: > davidxu 2008-06-17 06:26:29 UTC > > FreeBSD src repository > > Modified files: > include Makefile unistd.h > lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c > Added files: > include spawn.h > lib/libc/gen posix_spawn.c > Log: > SVN rev 179838 on 2008-06-17 06:26:29Z by davidxu > > Add POSIX routines called posix_spawn() and posix_spawnp(), which > can be used as replacements for exec/fork in a lot of cases. This > change also added execvpe() which allows environment variable > PATH to be used for searching executable file, it is used for > implementing posix_spawnp(). > > PR: standards/122051 I have no objections to this, but doesn't it defeat the whole purpose to implement posix_spawn() as a library function that just calls fork/exec? From ed at FreeBSD.org Tue Jun 17 14:05:13 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Tue Jun 17 14:05:18 2008 Subject: cvs commit: src/lib/libc/stdlib Symbol.map grantpt.c Message-ID: <200806171405.m5HE5Cv0051044@repoman.freebsd.org> ed 2008-06-17 14:05:03 UTC FreeBSD src repository Modified files: lib/libc/stdlib Symbol.map grantpt.c Log: SVN rev 179846 on 2008-06-17 14:05:03Z by ed Don't export the unused __use_pts() routine. The __use_pts() routine was once probably used by libutil to determine if we are using BSD or UNIX98 style PTY device names. It doesn't seem to be used outside grantpt.c, which means we can make it static and remove it from the Symbol.map. Reviewed by: cognet, kib Approved by: philip (mentor) Revision Changes Path 1.7 +0 -1 src/lib/libc/stdlib/Symbol.map 1.14 +1 -1 src/lib/libc/stdlib/grantpt.c From ed at 80386.nl Tue Jun 17 14:11:23 2008 From: ed at 80386.nl (Ed Schouten) Date: Tue Jun 17 14:11:29 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <20080617134828.GA30076@zim.MIT.EDU> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> Message-ID: <20080617140600.GE1176@hoeg.nl> * David Schultz wrote: > I have no objections to this, but doesn't it defeat the whole > purpose to implement posix_spawn() as a library function that just > calls fork/exec? When (if?) applications start to use posix_spawn() we may decide to move it into the kernel at any time. It should be okay for now. -- Ed Schouten WWW: http://80386.nl/ -------------- 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/20080617/070a1f93/attachment.pgp From kostikbel at gmail.com Tue Jun 17 14:26:51 2008 From: kostikbel at gmail.com (Kostik Belousov) Date: Tue Jun 17 14:26:59 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <20080617134828.GA30076@zim.MIT.EDU> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> Message-ID: <20080617140507.GM82830@deviant.kiev.zoral.com.ua> On Tue, Jun 17, 2008 at 09:48:28AM -0400, David Schultz wrote: > On Tue, Jun 17, 2008, David Xu wrote: > > davidxu 2008-06-17 06:26:29 UTC > > > > FreeBSD src repository > > > > Modified files: > > include Makefile unistd.h > > lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c > > Added files: > > include spawn.h > > lib/libc/gen posix_spawn.c > > Log: > > SVN rev 179838 on 2008-06-17 06:26:29Z by davidxu > > > > Add POSIX routines called posix_spawn() and posix_spawnp(), which > > can be used as replacements for exec/fork in a lot of cases. This > > change also added execvpe() which allows environment variable > > PATH to be used for searching executable file, it is used for > > implementing posix_spawnp(). > > > > PR: standards/122051 > > I have no objections to this, but doesn't it defeat the whole > purpose to implement posix_spawn() as a library function that just > calls fork/exec? How it can be ? Unless there is some problem with implementation, it shall be ok. Also, SUSv3 explicitely stated that design of the posix_spawn allows it to be implemented as fork/exec. -------------- 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/20080617/652f2691/attachment.pgp From sobomax at FreeBSD.org Tue Jun 17 15:15:44 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Tue Jun 17 15:15:47 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <20080617140600.GE1176@hoeg.nl> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> Message-ID: <4857D508.8070907@FreeBSD.org> Ed Schouten wrote: > * David Schultz wrote: >> I have no objections to this, but doesn't it defeat the whole >> purpose to implement posix_spawn() as a library function that just >> calls fork/exec? > > When (if?) applications start to use posix_spawn() we may decide to move > it into the kernel at any time. It should be okay for now. Are there any benefits of doing it in the kernel vs. doing it via fork+exec? -Maxim From ed at 80386.nl Tue Jun 17 16:09:17 2008 From: ed at 80386.nl (Ed Schouten) Date: Tue Jun 17 16:09:28 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <4857D508.8070907@FreeBSD.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> Message-ID: <20080617160352.GG1176@hoeg.nl> * Maxim Sobolev wrote: > Are there any benefits of doing it in the kernel vs. doing it via fork+exec? In theory there should be performance benefits, because there is no need to fork an entire process. You would only need to handcraft a new one. -- Ed Schouten WWW: http://80386.nl/ -------------- 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/20080617/5dc91ac1/attachment.pgp From joerg at britannica.bec.de Tue Jun 17 16:11:16 2008 From: joerg at britannica.bec.de (Joerg Sonnenberger) Date: Tue Jun 17 16:12:29 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <4857D508.8070907@FreeBSD.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> Message-ID: <20080617160103.GA13041@britannica.bec.de> On Tue, Jun 17, 2008 at 08:15:20AM -0700, Maxim Sobolev wrote: > Ed Schouten wrote: >> When (if?) applications start to use posix_spawn() we may decide to move >> it into the kernel at any time. It should be okay for now. > > Are there any benefits of doing it in the kernel vs. doing it via fork+exec? You get the same advantages as vfork() offers without introduces most of the rope. Joerg From ambrisko at FreeBSD.org Tue Jun 17 17:04:50 2008 From: ambrisko at FreeBSD.org (Doug Ambrisko) Date: Tue Jun 17 17:04:52 2008 Subject: cvs commit: src/sys/dev/mfi mfi.c mfi_ioctl.h mfi_pci.c Message-ID: <200806171704.m5HH4oTN078055@repoman.freebsd.org> ambrisko 2008-06-17 17:04:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/mfi mfi.c mfi_ioctl.h mfi_pci.c Log: SVN rev 179847 on 2008-06-17 17:04:38Z by ambrisko MFC: All changes in current/7 minus cam - Native 32bit compat modes for amd64 - Limit max commands to 128 - Add in compat mode for using mfi0 to talk to any card Revision Changes Path 1.3.2.12 +164 -21 src/sys/dev/mfi/mfi.c 1.1.2.4 +38 -0 src/sys/dev/mfi/mfi_ioctl.h 1.1.2.8 +1 -0 src/sys/dev/mfi/mfi_pci.c From das at FreeBSD.ORG Tue Jun 17 17:07:27 2008 From: das at FreeBSD.ORG (David Schultz) Date: Tue Jun 17 17:07:31 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <4857D508.8070907@FreeBSD.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> Message-ID: <20080617170755.GA30958@zim.MIT.EDU> On Tue, Jun 17, 2008, Maxim Sobolev wrote: > Ed Schouten wrote: > >* David Schultz wrote: > >>I have no objections to this, but doesn't it defeat the whole > >>purpose to implement posix_spawn() as a library function that just > >>calls fork/exec? > > > >When (if?) applications start to use posix_spawn() we may decide to move > >it into the kernel at any time. It should be okay for now. > > Are there any benefits of doing it in the kernel vs. doing it via fork+exec? The only reason spawn exists is to better support platforms where fork is slow, so implementing it in terms of fork/exec defeats the purpose and potentially tricks configure scripts into making incorrect assumptions about performance tradeoffs. However, maybe spawn would still be useful if misguided application writers used it for other reasons (e.g., to make it easier to port Win32 apps), and I'm guessing that's why it was added. Implementing it in the kernel has disadvantages, too; in particular, it would add a lot of complexity for gains that are likely to be minimal in FreeBSD. From jhb at freebsd.org Tue Jun 17 17:59:16 2008 From: jhb at freebsd.org (John Baldwin) Date: Tue Jun 17 17:59:30 2008 Subject: cvs commit: src/sys/dev/smc if_smc.c In-Reply-To: <200806170548.m5H5mnnP079768@repoman.freebsd.org> References: <200806170548.m5H5mnnP079768@repoman.freebsd.org> Message-ID: <200806170953.00350.jhb@freebsd.org> On Tuesday 17 June 2008 01:48:42 am Benno Rice wrote: > benno 2008-06-17 05:48:42 UTC > > FreeBSD src repository > > Modified files: > sys/dev/smc if_smc.c > Log: > SVN rev 179837 on 2008-06-17 05:48:42Z by benno > > - Move ether_ifdetach earlier. > - Drain callouts after ether_ifdetach. > > Suggested by: jhb Thanks. -- John Baldwin From jhb at freebsd.org Tue Jun 17 17:59:24 2008 From: jhb at freebsd.org (John Baldwin) Date: Tue Jun 17 17:59:31 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <4857D508.8070907@FreeBSD.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> Message-ID: <200806171122.41340.jhb@freebsd.org> On Tuesday 17 June 2008 11:15:20 am Maxim Sobolev wrote: > Ed Schouten wrote: > > * David Schultz wrote: > >> I have no objections to this, but doesn't it defeat the whole > >> purpose to implement posix_spawn() as a library function that just > >> calls fork/exec? > > > > When (if?) applications start to use posix_spawn() we may decide to move > > it into the kernel at any time. It should be okay for now. > > Are there any benefits of doing it in the kernel vs. doing it via fork+exec? Speed. You don't have to go mark all your pages as COW or some such only to turn around and throw the new mappings away and undo that. -- John Baldwin From ceri at submonkey.net Tue Jun 17 18:36:07 2008 From: ceri at submonkey.net (Ceri Davies) Date: Tue Jun 17 18:36:09 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <20080617170755.GA30958@zim.MIT.EDU> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> <20080617170755.GA30958@zim.MIT.EDU> Message-ID: <20080617183603.GD24940@submonkey.net> On Tue, Jun 17, 2008 at 01:07:55PM -0400, David Schultz wrote: > On Tue, Jun 17, 2008, Maxim Sobolev wrote: > > Ed Schouten wrote: > > >* David Schultz wrote: > > >>I have no objections to this, but doesn't it defeat the whole > > >>purpose to implement posix_spawn() as a library function that just > > >>calls fork/exec? > > > > > >When (if?) applications start to use posix_spawn() we may decide to move > > >it into the kernel at any time. It should be okay for now. > > > > Are there any benefits of doing it in the kernel vs. doing it via fork+exec? > > The only reason spawn exists is to better support platforms where > fork is slow, so implementing it in terms of fork/exec defeats the > purpose and potentially tricks configure scripts into making > incorrect assumptions about performance tradeoffs. It also helps on platforms like Solaris which refuse to overcommit, where a large process, say a 4GB JVM, would otherwise need another 4GB of swap free in order for fork to succeed. 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/20080617/65cc7f13/attachment.pgp From remko at FreeBSD.org Tue Jun 17 18:56:11 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Tue Jun 17 18:56:16 2008 Subject: cvs commit: src/usr.bin/shar shar.sh Message-ID: <200806171856.m5HIuAei088958@repoman.freebsd.org> remko 2008-06-17 18:56:04 UTC FreeBSD src repository Modified files: usr.bin/shar shar.sh Log: SVN rev 179848 on 2008-06-17 18:56:04Z by remko Remove superfluous eofmarker. Requested by: Jaakko Heinonen Discussed with: Jaakko, edwin Approved by: imp (mentor, implicit) Revision Changes Path 1.5 +0 -1 src/usr.bin/shar/shar.sh From remko at FreeBSD.org Tue Jun 17 19:04:26 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Tue Jun 17 19:04:28 2008 Subject: cvs commit: src/usr.bin/shar shar.sh Message-ID: <200806171904.m5HJ4QpU090945@repoman.freebsd.org> remko 2008-06-17 19:04:17 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/shar shar.sh Log: SVN rev 179849 on 2008-06-17 19:04:17Z by remko MFC rev r179848 shar.sh Remove superfluous eofmarker. Requested by: Jaakko Heinonen Discussed with: Jaakko, edwin Approved by: imp (mentor, implicit) Approved by: imp (mentor, implicit) Revision Changes Path 1.3.48.2 +0 -1 src/usr.bin/shar/shar.sh From remko at FreeBSD.org Tue Jun 17 19:10:59 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Tue Jun 17 19:11:00 2008 Subject: cvs commit: src/usr.bin/shar shar.sh Message-ID: <200806171910.m5HJAwRV092329@repoman.freebsd.org> remko 2008-06-17 19:10:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/shar shar.sh Log: SVN rev 179850 on 2008-06-17 19:10:38Z by remko MFC rev 179355 and 179848 shar.sh r179355 Limit the EOF marker length to a maximum of 79 characters. [1] Add $FreeBSD$ tag so that I can actually commit this. PR: bin/118782 Reported by: Bjoern Koenig Patch by: edwin, Jaakko Heinonen (not used patch) MFC after: 1 week Approved by: imp (mentor, implicit) r179848: Remove superfluous eofmarker. Requested by: Jaakko Heinonen Discussed with: Jaakko, edwin Revision Changes Path 1.3.38.1 +4 -2 src/usr.bin/shar/shar.sh From remko at FreeBSD.org Tue Jun 17 19:15:47 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Tue Jun 17 19:15:59 2008 Subject: cvs commit: src/share/sendmail Makefile Message-ID: <200806171915.m5HJFkno092761@repoman.freebsd.org> remko 2008-06-17 19:15:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/sendmail Makefile Log: SVN rev 179851 on 2008-06-17 19:15:38Z by remko MFC rev 179460 Makefile Dont install .svn metadata with the sendmail install. Submitted by: marcel Approved by: imp (mentor, implicit) MFC after: 3 days Approved by: imp (mentor, implicit) Revision Changes Path 1.10.2.1 +2 -2 src/share/sendmail/Makefile From kostikbel at gmail.com Tue Jun 17 20:19:42 2008 From: kostikbel at gmail.com (Kostik Belousov) Date: Tue Jun 17 20:19:46 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <200806171122.41340.jhb@freebsd.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> <200806171122.41340.jhb@freebsd.org> Message-ID: <20080617201919.GQ82830@deviant.kiev.zoral.com.ua> On Tue, Jun 17, 2008 at 11:22:40AM -0400, John Baldwin wrote: > On Tuesday 17 June 2008 11:15:20 am Maxim Sobolev wrote: > > Ed Schouten wrote: > > > * David Schultz wrote: > > >> I have no objections to this, but doesn't it defeat the whole > > >> purpose to implement posix_spawn() as a library function that just > > >> calls fork/exec? > > > > > > When (if?) applications start to use posix_spawn() we may decide to move > > > it into the kernel at any time. It should be okay for now. > > > > Are there any benefits of doing it in the kernel vs. doing it via fork+exec? > > Speed. You don't have to go mark all your pages as COW or some such only to > turn around and throw the new mappings away and undo that. The committed implementation uses vfork(). No COW is needed. On the other hand, after single-threading of the parent process have been removed from fork1(9), vfork(2) could have interesting consequences when done from the multithreaded process. -------------- 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/20080617/b1c74f13/attachment.pgp From sobomax at FreeBSD.org Tue Jun 17 20:54:53 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Tue Jun 17 20:55:00 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <200806171122.41340.jhb@freebsd.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> <200806171122.41340.jhb@freebsd.org> Message-ID: <48582484.3040606@FreeBSD.org> John Baldwin wrote: > On Tuesday 17 June 2008 11:15:20 am Maxim Sobolev wrote: >> Ed Schouten wrote: >>> * David Schultz wrote: >>>> I have no objections to this, but doesn't it defeat the whole >>>> purpose to implement posix_spawn() as a library function that just >>>> calls fork/exec? >>> When (if?) applications start to use posix_spawn() we may decide to move >>> it into the kernel at any time. It should be okay for now. >> Are there any benefits of doing it in the kernel vs. doing it via fork+exec? > > Speed. You don't have to go mark all your pages as COW or some such only to > turn around and throw the new mappings away and undo that. Don't we have vfork() for that? -Maxim From brueffer at FreeBSD.org Tue Jun 17 21:14:23 2008 From: brueffer at FreeBSD.org (Christian Brueffer) Date: Tue Jun 17 21:14:28 2008 Subject: cvs commit: src/share/man/man4 ixgbe.4 Message-ID: <200806172114.m5HLEMFG016165@repoman.freebsd.org> brueffer 2008-06-17 21:14:02 UTC FreeBSD src repository Modified files: share/man/man4 ixgbe.4 Log: SVN rev 179852 on 2008-06-17 21:14:02Z by brueffer Bring this up to date with regard to our other section 4 and Intel manpages, also fixes a few mdoc bugs. Revision Changes Path 1.2 +18 -14 src/share/man/man4/ixgbe.4 From marcel at FreeBSD.org Wed Jun 18 01:13:47 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Wed Jun 18 01:13:52 2008 Subject: cvs commit: src/sys/geom/part g_part.c g_part.h g_part_if.m g_part_mbr.c Message-ID: <200806180113.m5I1DliY050397@repoman.freebsd.org> marcel 2008-06-18 01:13:34 UTC FreeBSD src repository Modified files: sys/geom/part g_part.c g_part.h g_part_if.m g_part_mbr.c Log: SVN rev 179853 on 2008-06-18 01:13:34Z by marcel Add the set and unset verbs used to set and clear attributes for partition entries. Implement the setunset method for the MBR scheme to control the active flag. Revision Changes Path 1.19 +75 -2 src/sys/geom/part/g_part.c 1.10 +3 -1 src/sys/geom/part/g_part.h 1.4 +9 -1 src/sys/geom/part/g_part_if.m 1.9 +43 -1 src/sys/geom/part/g_part_mbr.c From davidxu at freebsd.org Wed Jun 18 01:33:40 2008 From: davidxu at freebsd.org (David Xu) Date: Wed Jun 18 01:33:42 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <4857D508.8070907@FreeBSD.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> Message-ID: <48586652.2070204@freebsd.org> Maxim Sobolev wrote: > Ed Schouten wrote: >> * David Schultz wrote: >>> I have no objections to this, but doesn't it defeat the whole >>> purpose to implement posix_spawn() as a library function that just >>> calls fork/exec? >> >> When (if?) applications start to use posix_spawn() we may decide to move >> it into the kernel at any time. It should be okay for now. > > Are there any benefits of doing it in the kernel vs. doing it via > fork+exec? > > -Maxim > I just want to make some programs to work, since Linux and Solaris both have it now. Doing it in kernel is OK, but I don't want to put it into the kernel until it is necessary since kernel code can not be swapped out. ;-) Doing in kernel may have advantages, for example, you may get ride of zoombie process recycling by hacking kernel, a library version using vfork will need parent to handle SIGCHLD and recycle zoombie process like you use popen() and system() etcs, another advantage is it won't involve rtld if a symbol needs to resolved, because vfork shares same vmspace, it might block other threads in parent process if the child process holding a rtld's reader lock, I don't know if writer lock will be held in child, I guess it won't be. if the child process get killed wrongly by someone, it might hold reader lock forever, this is the race condition. a kernel version may have trouble if something can not be easily done in kernel, though I didn't find the one can not be done in kernel at present. Regards, David Xu From davidxu at freebsd.org Wed Jun 18 01:43:10 2008 From: davidxu at freebsd.org (David Xu) Date: Wed Jun 18 01:43:12 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <48586652.2070204@freebsd.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> <48586652.2070204@freebsd.org> Message-ID: <4858688E.2030404@freebsd.org> David Xu wrote: > I just want to make some programs to work, since Linux and Solaris > both have it now. Doing it in kernel is OK, but I don't want to put > it into the kernel until it is necessary since kernel code can not > be swapped out. ;-) > > Doing in kernel may have advantages, for example, you may get ride > of zoombie process recycling by hacking kernel, a library version > using vfork will need parent to handle SIGCHLD and recycle zoombie > process like you use popen() and system() etcs, Oops, I rethinked it, zoombie process is not a correct point. only rtld is the problem. > another advantage > is it won't involve rtld if a symbol needs to resolved, because > vfork shares same vmspace, it might block other threads in > parent process if the child process holding a rtld's > reader lock, I don't know if writer lock will be held in child, > I guess it won't be. if the child process get killed wrongly > by someone, it might hold reader lock forever, this is the race > condition. a kernel version may have trouble if something can not > be easily done in kernel, though I didn't find the one can not > be done in kernel at present. > > Regards, > David Xu > > From marcel at FreeBSD.org Wed Jun 18 01:46:49 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Wed Jun 18 01:46:53 2008 Subject: cvs commit: src/sbin/geom/class/part geom_part.c gpart.8 Message-ID: <200806180146.m5I1kmho053025@repoman.freebsd.org> marcel 2008-06-18 01:46:32 UTC FreeBSD src repository Modified files: sbin/geom/class/part geom_part.c gpart.8 Log: SVN rev 179854 on 2008-06-18 01:46:32Z by marcel Implement the set and unset verbs. While here, have the manpage catch up with various changes. Revision Changes Path 1.9 +31 -4 src/sbin/geom/class/part/geom_part.c 1.5 +53 -7 src/sbin/geom/class/part/gpart.8 From weongyo at FreeBSD.org Wed Jun 18 06:35:57 2008 From: weongyo at FreeBSD.org (Weongyo Jeong) Date: Wed Jun 18 06:36:01 2008 Subject: cvs commit: src/usr.sbin/ndiscvt inf.c Message-ID: <200806180635.m5I6Zutg090702@repoman.freebsd.org> weongyo 2008-06-18 06:35:37 UTC FreeBSD src repository Modified files: usr.sbin/ndiscvt inf.c Log: SVN rev 179855 on 2008-06-18 06:35:37Z by weongyo handle .INF files for PCMCIA correctly that specify multiple entries in their [Manufacturer] sections and prevent a case that NDIS_PCI_DEV_TABLE definition was always emitted that it's only emitted once if a .INF file is for PCI. Revision Changes Path 1.18 +13 -7 src/usr.sbin/ndiscvt/inf.c From rwatson at FreeBSD.org Wed Jun 18 08:16:42 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Jun 18 08:16:51 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <20080617170755.GA30958@zim.MIT.EDU> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> <20080617170755.GA30958@zim.MIT.EDU> Message-ID: <20080618091320.F18654@fledge.watson.org> On Tue, 17 Jun 2008, David Schultz wrote: > On Tue, Jun 17, 2008, Maxim Sobolev wrote: >> Ed Schouten wrote: >>> * David Schultz wrote: >>>> I have no objections to this, but doesn't it defeat the whole purpose to >>>> implement posix_spawn() as a library function that just calls fork/exec? >>> >>> When (if?) applications start to use posix_spawn() we may decide to move >>> it into the kernel at any time. It should be okay for now. >> >> Are there any benefits of doing it in the kernel vs. doing it via >> fork+exec? > > The only reason spawn exists is to better support platforms where fork is > slow, so implementing it in terms of fork/exec defeats the purpose and > potentially tricks configure scripts into making incorrect assumptions about > performance tradeoffs. However, maybe spawn would still be useful if > misguided application writers used it for other reasons (e.g., to make it > easier to port Win32 apps), and I'm guessing that's why it was added. > Implementing it in the kernel has disadvantages, too; in particular, it > would add a lot of complexity for gains that are likely to be minimal in > FreeBSD. Apple's experience has been somewhat to the contrary -- while the architecture varies some by OS release, one of the persisting performance problems they were seeing was the cost of fork()+execve() from applications with very large numbers of shared libraries, plugins, memory mappings, etc. Currently, they address this by having a process launch applications "by proxy" as a result of IPC requests instead of forking and execing, but you might reasonably argue that the problem is with the fork()+execve() model. Robert N M Watson Computer Laboratory University of Cambridge From peter at wemm.org Wed Jun 18 09:11:40 2008 From: peter at wemm.org (Peter Wemm) Date: Wed Jun 18 09:11:46 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <20080618091320.F18654@fledge.watson.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> <20080617170755.GA30958@zim.MIT.EDU> <20080618091320.F18654@fledge.watson.org> Message-ID: On Wed, Jun 18, 2008 at 1:16 AM, Robert Watson wrote: > On Tue, 17 Jun 2008, David Schultz wrote: > >> On Tue, Jun 17, 2008, Maxim Sobolev wrote: >>> >>> Ed Schouten wrote: >>>> >>>> * David Schultz wrote: >>>>> >>>>> I have no objections to this, but doesn't it defeat the whole purpose >>>>> to implement posix_spawn() as a library function that just calls fork/exec? >>>> >>>> When (if?) applications start to use posix_spawn() we may decide to move >>>> it into the kernel at any time. It should be okay for now. >>> >>> Are there any benefits of doing it in the kernel vs. doing it via >>> fork+exec? >> >> The only reason spawn exists is to better support platforms where fork is >> slow, so implementing it in terms of fork/exec defeats the purpose and >> potentially tricks configure scripts into making incorrect assumptions about >> performance tradeoffs. However, maybe spawn would still be useful if >> misguided application writers used it for other reasons (e.g., to make it >> easier to port Win32 apps), and I'm guessing that's why it was added. >> Implementing it in the kernel has disadvantages, too; in particular, it >> would add a lot of complexity for gains that are likely to be minimal in >> FreeBSD. > > Apple's experience has been somewhat to the contrary -- while the > architecture varies some by OS release, one of the persisting performance > problems they were seeing was the cost of fork()+execve() from applications > with very large numbers of shared libraries, plugins, memory mappings, etc. > Currently, they address this by having a process launch applications "by > proxy" as a result of IPC requests instead of forking and execing, but you > might reasonably argue that the problem is with the fork()+execve() model. It gets significant for FreeBSD too. Here's some food for thought. Imagine % cc -static -o forkspeed forkspeed.c % time ./forkspeed 10000 0.020u 0.592s 0:00.55 110.9% 170+2652k 0+0io 0pf+0w % cc -o forkspeed forkspeed.c % time ./forkspeed 10000 0.090u 1.239s 0:01.25 105.6% 5+196k 0+0io 0pf+0w % cc -o forkspeed forkspeed.c -lncurses -lutil -lcrypto -lssl % time ./forkspeed 10000 0.070u 1.785s 0:02.09 88.5% 4+154k 0+0io 0pf+0w -static vs dynamic: 2.27 times slower. -static vs extra dynamic libs: 3.80 times slower. % cat forkspeed.c #include #include #include #include int main(int ac, char *av[]) { int n = atoi(av[1]); int i; for (i = 0; i < n; i++) { if (fork() == 0) _exit(0); wait(0); } return (0); } No exec(), just measuring the overhead of fork(). Just for amusement, the same with vfork(): peter@overcee[1:48am]~-130> cc -static -o forkspeed forkspeed.c peter@overcee[1:48am]~-131> time ./forkspeed 10000 0.007u 0.097s 0:00.11 81.8% 161+2503k 0+0io 0pf+0w peter@overcee[1:48am]~-132> cc -o forkspeed forkspeed.c peter@overcee[1:49am]~-133> time ./forkspeed 10000 0.008u 0.170s 0:00.11 100.0% 4+139k 0+0io 0pf+0w peter@overcee[1:49am]~-134> cc -o forkspeed forkspeed.c -lncurses -lutil -lcrypto -lssl peter@overcee[1:49am]~-135> time ./forkspeed 10000 0.000u 0.106s 0:00.10 100.0% 4+158k 0+0io 0pf+0w Essentially the same regardless of libraries. vfork is 5 times faster for -static, 11 times faster for regular dynamic, and 20 times faster for extra libraries. So.. if something auto-detects posix_spawn(), which uses vfork(), it would be a win compared to the usual fork()/exec(). A small win, but still a win. It would have to do a lot of iterations to add up. Incidently, this is why /usr/bin/make and /usr/bin/gcc are statically linked. /bin/sh used to be, but isn't so that ~user can use nsswitch. For amusement, think of kde and gnome with all their libraries. -- 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 Jun 18 09:20:37 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Wed Jun 18 09:20:44 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> <20080617134828.GA30076@zim.MIT.EDU> <20080617140600.GE1176@hoeg.nl> <4857D508.8070907@FreeBSD.org> <20080617170755.GA30958@zim.MIT.EDU> <20080618091320.F18654@fledge.watson.org> Message-ID: <20080618101830.S18654@fledge.watson.org> On Wed, 18 Jun 2008, Peter Wemm wrote: >> Apple's experience has been somewhat to the contrary -- while the >> architecture varies some by OS release, one of the persisting performance >> problems they were seeing was the cost of fork()+execve() from applications >> with very large numbers of shared libraries, plugins, memory mappings, etc. >> Currently, they address this by having a process launch applications "by >> proxy" as a result of IPC requests instead of forking and execing, but you >> might reasonably argue that the problem is with the fork()+execve() model. ... > Essentially the same regardless of libraries. vfork is 5 times faster for > -static, 11 times faster for regular dynamic, and 20 times faster for extra > libraries. > > So.. if something auto-detects posix_spawn(), which uses vfork(), it would > be a win compared to the usual fork()/exec(). A small win, but still a win. > It would have to do a lot of iterations to add up. > > Incidently, this is why /usr/bin/make and /usr/bin/gcc are statically > linked. /bin/sh used to be, but isn't so that ~user can use nsswitch. > > For amusement, think of kde and gnome with all their libraries. Well, kdeinit already performs pre-linking to avoid repeated runtime linker costs -- kdeinit is basically a template process waiting to be filled in with the specifics of any particular application, but with all the shared libraries already mapped and linked. I've never benchmarked it, but one might reasonably assume that the technique works or they wouldn't ship with it. Seems like some systemic benchmarking and profiling is required to decide how much blame to point at the cost of forking complex address spaces, run-time linking, exec overhead, etc in order to decide how much to pin the blame on each. Robert N M Watson Computer Laboratory University of Cambridge From kib at FreeBSD.org Wed Jun 18 09:25:40 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Wed Jun 18 09:25:42 2008 Subject: cvs commit: src/sys/fs/tmpfs tmpfs_subr.c Message-ID: <200806180925.m5I9Pe2P018504@repoman.freebsd.org> kib 2008-06-18 09:25:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/tmpfs tmpfs_subr.c Log: SVN rev 179856 on 2008-06-18 09:25:26Z by kib MFC r179808: Do not redo the vnode tear-down work already done by insmntque() when vnode cannot be put on the vnode list for mount. Revision Changes Path 1.12.2.2 +1 -4 src/sys/fs/tmpfs/tmpfs_subr.c From kib at FreeBSD.org Wed Jun 18 09:32:11 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Wed Jun 18 09:32:18 2008 Subject: cvs commit: src/sys/fs/cd9660 cd9660_vnops.c Message-ID: <200806180932.m5I9WBlA018987@repoman.freebsd.org> kib 2008-06-18 09:31:56 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/cd9660 cd9660_vnops.c Log: SVN rev 179857 on 2008-06-18 09:31:56Z by kib MFC r179722: In cd9660_readdir vop, always initialize the idp->uio_off member. PR: 122925 Revision Changes Path 1.113.2.1 +1 -0 src/sys/fs/cd9660/cd9660_vnops.c From jhb at freebsd.org Wed Jun 18 13:19:32 2008 From: jhb at freebsd.org (John Baldwin) Date: Wed Jun 18 13:19:38 2008 Subject: cvs commit: src/sys/conf files.i386 files.pc98 options.i386 options.pc98 src/sys/i386/conf NOTES src/sys/i386/include pecoff_machdep.h src/sys/modules Makefile src/sys/pc98/conf NOTES src/sys/pc98/include pecoff_machdep.h In-Reply-To: <200806141252.m5ECqUGh093173@repoman.freebsd.org> References: <200806141252.m5ECqUGh093173@repoman.freebsd.org> Message-ID: <200806171814.12174.jhb@freebsd.org> On Saturday 14 June 2008 08:51:44 am Wojciech A. Koszek wrote: > wkoszek 2008-06-14 12:51:44 UTC > > FreeBSD src repository > > Modified files: > sys/conf files.i386 files.pc98 options.i386 > options.pc98 > sys/i386/conf NOTES > sys/modules Makefile > sys/pc98/conf NOTES > Removed files: > sys/i386/include pecoff_machdep.h > sys/pc98/include pecoff_machdep.h > Log: > SVN rev 179785 on 2008-06-14 12:51:44Z by wkoszek > > Remove obselete PECOFF image activator support. > > PRs assigned at the time of removal: kern/80742 > > Discussed on: freebsd-current (silence), IRC > Tested by: make universe > Approved by: cognet (mentor) Looks like the SVN -> CVS importer doesn't handle directory deletions. It didn't delete the files underneath the sys/modules/pecoff and sys/compat/pecoff directories. Original SVN log excerpt: Deleted: head/sys/compat/pecoff/ head/sys/i386/include/pecoff_machdep.h head/sys/modules/pecoff/ head/sys/pc98/include/pecoff_machdep.h Modified: head/sys/conf/files.i386 head/sys/conf/files.pc98 head/sys/conf/options.i386 head/sys/conf/options.pc98 head/sys/i386/conf/NOTES head/sys/modules/Makefile head/sys/pc98/conf/NOTES -- John Baldwin From obrien at FreeBSD.org Wed Jun 18 13:58:40 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Wed Jun 18 13:58:47 2008 Subject: cvs commit: src Makefile.inc1 Message-ID: <200806181358.m5IDwe9p055272@repoman.freebsd.org> obrien 2008-06-18 13:52:58 UTC FreeBSD src repository Modified files: . Makefile.inc1 Log: SVN rev 179858 on 2008-06-18 13:52:58Z by obrien Add MIPS to the list of known arches. Revision Changes Path 1.606 +1 -1 src/Makefile.inc1 From jhb at FreeBSD.org Wed Jun 18 14:23:43 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Wed Jun 18 14:23:49 2008 Subject: cvs commit: src/gnu/usr.bin/gdb/kgdb trgt_mips.c Message-ID: <200806181423.m5IENgco058461@repoman.freebsd.org> jhb 2008-06-18 14:23:28 UTC FreeBSD src repository Modified files: gnu/usr.bin/gdb/kgdb trgt_mips.c Log: SVN rev 179859 on 2008-06-18 14:23:28Z by jhb Catch up to recentish kgdb changes: - Use ptid_get_pid() rather than ptid_get_tid() (part of the changes to let 'tid' work for remote kgdb). - Add a stub kgdb_trgt_new_objfile() hook. Silence from: obrien, mips@ Revision Changes Path 1.3 +6 -1 src/gnu/usr.bin/gdb/kgdb/trgt_mips.c From joerg at FreeBSD.org Wed Jun 18 20:40:18 2008 From: joerg at FreeBSD.org (Joerg Wunsch) Date: Wed Jun 18 20:40:20 2008 Subject: cvs commit: src/sys/pci nfsmb.c Message-ID: <200806182040.m5IKeI9n006178@repoman.freebsd.org> joerg 2008-06-18 20:39:56 UTC FreeBSD src repository Modified files: sys/pci nfsmb.c Log: SVN rev 179860 on 2008-06-18 20:39:56Z by joerg Add the SMB functionality for the MCP65 chipset I happen to have in my new motherboard. Revision Changes Path 1.10 +3 -0 src/sys/pci/nfsmb.c From attilio at FreeBSD.org Wed Jun 18 20:42:23 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Wed Jun 18 20:43:44 2008 Subject: cvs commit: src/sys/ddb db_ps.c Message-ID: <200806182042.m5IKgMQ0006343@repoman.freebsd.org> attilio 2008-06-18 20:42:01 UTC FreeBSD src repository Modified files: sys/ddb db_ps.c Log: SVN rev 179861 on 2008-06-18 20:42:01Z by attilio Print out the container lock when showing the thread state in DDB. Tested by: benjsc Revision Changes Path 1.69 +3 -0 src/sys/ddb/db_ps.c From attilio at FreeBSD.org Wed Jun 18 20:50:41 2008 From: attilio at FreeBSD.org (Attilio Rao) Date: Wed Jun 18 20:50:46 2008 Subject: cvs commit: src/share/man/man4 ubsa.4 Message-ID: <200806182050.m5IKoeIY016404@repoman.freebsd.org> attilio 2008-06-18 20:50:30 UTC FreeBSD src repository Modified files: share/man/man4 ubsa.4 Log: SVN rev 179862 on 2008-06-18 20:50:30Z by attilio Add the option stub for the Globetrotter Max 3.6 Modem. Submitted by: Greg Rivers Revision Changes Path 1.13 +2 -0 src/share/man/man4/ubsa.4 From marius at FreeBSD.org Wed Jun 18 21:20:59 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 18 21:21:01 2008 Subject: cvs commit: src/kerberos5/tools/slc Makefile Message-ID: <200806182120.m5ILKwXX021274@repoman.freebsd.org> marius 2008-06-18 21:20:50 UTC FreeBSD src repository Modified files: kerberos5/tools/slc Makefile Log: SVN rev 179863 on 2008-06-18 21:20:50Z by marius Add roken.h to SRCS. This fixes the compilation of slc during a buildworld on a host running a world built with WITHOUT_KERBEROS defined. Revision Changes Path 1.2 +1 -0 src/kerberos5/tools/slc/Makefile From ambrisko at FreeBSD.org Wed Jun 18 21:39:15 2008 From: ambrisko at FreeBSD.org (Doug Ambrisko) Date: Wed Jun 18 21:39:17 2008 Subject: cvs commit: src/sys/boot/forth support.4th Message-ID: <200806182139.m5ILdF5B026906@repoman.freebsd.org> ambrisko 2008-06-18 21:39:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/forth support.4th Log: SVN rev 179864 on 2008-06-18 21:39:00Z by ambrisko MFC: Allow negative values to be specified in the loader. Revision Changes Path 1.16.2.1 +2 -0 src/sys/boot/forth/support.4th From yongari at FreeBSD.org Thu Jun 19 01:36:44 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Thu Jun 19 01:36:49 2008 Subject: cvs commit: src/sys/dev/dc if_dc.c Message-ID: <200806190136.m5J1aihj085525@repoman.freebsd.org> yongari 2008-06-19 01:36:31 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/dc if_dc.c Log: SVN rev 179865 on 2008-06-19 01:36:31Z by yongari MFC: r179647 Ethernet hardware address stored in DC_AL_PAR0/DC_AL_PAR1 register is in little endian form. Likewise setting DC_AL_PAR0/DC_AL_PAR1 register expect the address to be in little endian form. For big endian architectures the address should be swapped to get correct one. Change setting/getting ethernet hardware address to big endian architecture frendly. Reported by: Robert Murillo ( billypilgrim782001 at yahoo dot com ) Tested by: Robert Murillo ( billypilgrim782001 at yahoo dot com ) Revision Changes Path 1.192.2.3 +14 -6 src/sys/dev/dc/if_dc.c From davidxu at FreeBSD.org Thu Jun 19 02:42:53 2008 From: davidxu at FreeBSD.org (David Xu) Date: Thu Jun 19 02:42:55 2008 Subject: cvs commit: src/lib/libc/gen posix_spawn.c Message-ID: <200806190242.m5J2gqvT092092@repoman.freebsd.org> davidxu 2008-06-19 02:42:50 UTC FreeBSD src repository Modified files: lib/libc/gen posix_spawn.c Log: SVN rev 179866 on 2008-06-19 02:42:50Z by davidxu Process spawn attributes in POSIX document order. Revision Changes Path 1.4 +22 -21 src/lib/libc/gen/posix_spawn.c From yongari at FreeBSD.org Thu Jun 19 03:38:05 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Thu Jun 19 03:38:21 2008 Subject: cvs commit: src/sys/pci if_dc.c Message-ID: <200806190337.m5J3btHQ000732@repoman.freebsd.org> yongari 2008-06-19 03:37:40 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/pci if_dc.c Log: SVN rev 179867 on 2008-06-19 03:37:40Z by yongari MFC: r179647 Ethernet hardware address stored in DC_AL_PAR0/DC_AL_PAR1 register is in little endian form. Likewise setting DC_AL_PAR0/DC_AL_PAR1 register expect the address to be in little endian form. For big endian architectures the address should be swapped to get correct one. Change setting/getting ethernet hardware address to big endian architecture frendly. Reported by: Robert Murillo ( billypilgrim782001 at yahoo dot com ) Tested by: Robert Murillo ( billypilgrim782001 at yahoo dot com ) Revision Changes Path 1.160.2.15 +15 -5 src/sys/pci/if_dc.c From yongari at FreeBSD.org Thu Jun 19 03:57:08 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Thu Jun 19 03:57:26 2008 Subject: cvs commit: src/sys/dev/mii atphy.c atphyreg.h miidevs Message-ID: <200806190356.m5J3uuD5002374@repoman.freebsd.org> yongari 2008-06-19 03:56:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/mii miidevs Added files: (Branch: RELENG_7) sys/dev/mii atphy.c atphyreg.h Log: SVN rev 179868 on 2008-06-19 03:56:39Z by yongari MFC: r179098 Add Attansic/Atheros F1 PHY driver. Revision Changes Path 1.1.2.1 +413 -0 src/sys/dev/mii/atphy.c (new) 1.1.2.1 +63 -0 src/sys/dev/mii/atphyreg.h (new) 1.46.2.5 +4 -0 src/sys/dev/mii/miidevs From yongari at FreeBSD.org Thu Jun 19 04:25:01 2008 From: yongari at FreeBSD.org (Pyun YongHyeon) Date: Thu Jun 19 04:25:41 2008 Subject: cvs commit: src/sys/conf files src/sys/modules/mii Makefile Message-ID: <200806190424.m5J4OvG2005691@repoman.freebsd.org> yongari 2008-06-19 04:23:26 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf files sys/modules/mii Makefile Log: SVN rev 179869 on 2008-06-19 04:23:26Z by yongari MFC: r179099 Connect atphy(4) to the build. Revision Changes Path 1.1243.2.23 +1 -0 src/sys/conf/files 1.30.2.3 +1 -1 src/sys/modules/mii/Makefile From mtm at FreeBSD.org Thu Jun 19 06:11:53 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Jun 19 06:11:59 2008 Subject: cvs commit: src/etc rc.subr Message-ID: <200806190611.m5J6Br4i032103@repoman.freebsd.org> mtm 2008-06-19 06:11:34 UTC FreeBSD src repository Modified files: etc rc.subr Log: SVN rev 179870 on 2008-06-19 06:11:34Z by mtm Move the check for enabled knobs further down in run_rc_command() so that bogus commands cause usage information to be printed instead of diagnostics about enabling the knob. Revision Changes Path 1.84 +14 -14 src/etc/rc.subr From cperciva at FreeBSD.org Thu Jun 19 06:45:56 2008 From: cperciva at FreeBSD.org (Colin Percival) Date: Thu Jun 19 06:46:00 2008 Subject: cvs commit: src UPDATING src/sys/conf newvers.sh src/sys/netinet tcp.h tcp_output.c Message-ID: <200806190645.m5J6jtov039730@repoman.freebsd.org> cperciva 2008-06-19 06:36:10 UTC FreeBSD src repository Modified files: (Branch: RELENG_7_0) . UPDATING sys/conf newvers.sh sys/netinet tcp.h tcp_output.c Log: SVN rev 179871 on 2008-06-19 06:36:10Z by cperciva Fix errors in the padding of TCP options. Errata: FreeBSD-EN-08:02.tcp Approved by: so (cperciva) Revision Changes Path 1.507.2.3.2.6 +3 -0 src/UPDATING 1.72.2.5.2.6 +1 -1 src/sys/conf/newvers.sh 1.40.4.1 +2 -0 src/sys/netinet/tcp.h 1.141.2.3.2.1 +19 -3 src/sys/netinet/tcp_output.c From mtm at FreeBSD.org Thu Jun 19 07:06:20 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Jun 19 07:06:26 2008 Subject: cvs commit: src/etc rc.subr src/etc/defaults rc.conf src/etc/rc.d quota Message-ID: <200806190706.m5J76KQb042653@repoman.freebsd.org> mtm 2008-06-19 07:06:11 UTC FreeBSD src repository Modified files: etc rc.subr etc/defaults rc.conf etc/rc.d quota Log: SVN rev 179872 on 2008-06-19 07:06:11Z by mtm Make quota knob conform to other rc(8) knobs. Keep older knob for compatibility. Requested by: Volker Revision Changes Path 1.337 +1 -1 src/etc/defaults/rc.conf 1.8 +1 -1 src/etc/rc.d/quota 1.85 +4 -0 src/etc/rc.subr From ed at FreeBSD.org Thu Jun 19 07:30:52 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Thu Jun 19 07:30:57 2008 Subject: cvs commit: src/include spawn.h Message-ID: <200806190730.m5J7UpWM044567@repoman.freebsd.org> ed 2008-06-19 07:30:32 UTC FreeBSD src repository Modified files: include spawn.h Log: SVN rev 179873 on 2008-06-19 07:30:32Z by ed Remove __restrict keywords from array arguments to make GCC's -std=c99 work. When GCC is invoked with -std=c99, the following errors are displayed when including : /usr/include/spawn.h:69: error: static or type qualifiers in abstract declarator /usr/include/spawn.h:69: error: static or type qualifiers in abstract declarator /usr/include/spawn.h:72: error: static or type qualifiers in abstract declarator /usr/include/spawn.h:72: error: static or type qualifiers in abstract declarator We'd better remove the __restrict keywords here. The same is also done in . Submitted by: Andrzej Tobola Reviewed by: davidxu Approved by: philip (mentor, implicit) Revision Changes Path 1.3 +5 -2 src/include/spawn.h From bz at FreeBSD.org Thu Jun 19 07:35:07 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Thu Jun 19 07:35:11 2008 Subject: cvs commit: src UPDATING src/sys/conf newvers.sh src/sys/netinet tcp.h tcp_output.c In-Reply-To: <200806190645.m5J6jtov039730@repoman.freebsd.org> References: <200806190645.m5J6jtov039730@repoman.freebsd.org> Message-ID: <20080619072808.O83875@maildrop.int.zabbadoz.net> On Thu, 19 Jun 2008, Colin Percival wrote: > cperciva 2008-06-19 06:36:10 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7_0) > . UPDATING > sys/conf newvers.sh > sys/netinet tcp.h tcp_output.c > Log: > SVN rev 179871 on 2008-06-19 06:36:10Z by cperciva > > Fix errors in the padding of TCP options. > > Errata: FreeBSD-EN-08:02.tcp > Approved by: so (cperciva) And again thanks to the following people who had helped to debug various setups: anders@, s3raphi, Matt Reimer Doug Hardie and Randy Rose, John Mayer, Susan Guzzardi And special thanks to dwhite@ and BitGravity for running a large scale test to give me enough data to analyse and helping me with the analysing. > Revision Changes Path > 1.507.2.3.2.6 +3 -0 src/UPDATING > 1.72.2.5.2.6 +1 -1 src/sys/conf/newvers.sh > 1.40.4.1 +2 -0 src/sys/netinet/tcp.h > 1.141.2.3.2.1 +19 -3 src/sys/netinet/tcp_output.c > -- Bjoern A. Zeeb Stop bit received. Insert coin for new game. From joerg at FreeBSD.org Thu Jun 19 07:35:23 2008 From: joerg at FreeBSD.org (Joerg Wunsch) Date: Thu Jun 19 07:35:28 2008 Subject: cvs commit: src/sys/pci nfsmb.c Message-ID: <200806190735.m5J7ZMtJ045284@repoman.freebsd.org> joerg 2008-06-19 07:35:08 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/pci nfsmb.c Log: SVN rev 179874 on 2008-06-19 07:35:08Z by joerg MFC: SMBus detection for the MCP65 chipset. Revision Changes Path 1.6.2.3 +3 -0 src/sys/pci/nfsmb.c From joerg at FreeBSD.org Thu Jun 19 08:27:28 2008 From: joerg at FreeBSD.org (Joerg Wunsch) Date: Thu Jun 19 08:27:32 2008 Subject: cvs commit: CVSROOT access Message-ID: <200806190827.m5J8ROlJ050510@repoman.freebsd.org> joerg 2008-06-19 08:27:21 UTC FreeBSD src repository Modified files: . access Log: SVN rev 179875 on 2008-06-19 08:27:21Z by joerg Change my commit email address. Approved by: core (implicitly) Revision Changes Path 1.883 +1 -1 CVSROOT/access From attilio at freebsd.org Thu Jun 19 15:42:04 2008 From: attilio at freebsd.org (Attilio Rao) Date: Thu Jun 19 15:42:07 2008 Subject: cvs commit: src/sys/amd64/amd64 cpu_switch.S In-Reply-To: <200803232309.m2NN96Qa080896@repoman.freebsd.org> References: <200803232309.m2NN96Qa080896@repoman.freebsd.org> Message-ID: <3bbf2fe10806190842s381611del5c5dc27d2dd22a7e@mail.gmail.com> 2008/3/24, Peter Wemm : > peter 2008-03-23 23:09:06 UTC > > FreeBSD src repository > > Modified files: > sys/amd64/amd64 cpu_switch.S > Log: > First pass at (possibly futile) microoptimizing of cpu_switch. Results > are mixed. Some pure context switch microbenchmarks show up to 29% > improvement. Pipe based context switch microbenchmarks show up to 7% > improvement. Real world tests are far less impressive as they are > dominated more by actual work than switch overheads, but depending on > the machine in question, workload, kernel options, phase of moon, etc, a > few percent gain might be seen. > > Summary of changes: > - don't reload MSR_[FG]SBASE registers when context switching between > non-threaded userland apps. These typically cost 120 clock cycles each > on an AMD cpu (less on Barcelona/Phenom). Intel cores are probably no > faster on this. > - The above change only helps unthreaded userland apps that tend to use > the same value for gsbase. Threaded apps will get no benefit from this. > - reorder things like accessing the pcb to be in memory order, to give > prefetching a better chance of working. Operations are now in increasing > memory address order, rather than reverse or random. > - Push some lesser used code out of the main code paths. Hopefully > allowing better code density in cache lines. This is probably futile. > - (part 2 of previous item) Reorder code so that branches have a more > realistic static branch prediction hint. Both Intel and AMD cpus > default to predicting branches to lower memory addresses as being > taken, and to higher memory addresses as not being taken. This is > overridden by the limited dynamic branch prediction subsystem. A trip > through userland might overflow this. > - Futule attempt at spreading the use of the results of previous operations > in new operations. Hopefully this will allow the cpus to execute in > parallel better. > - stop wasting 16 bytes at the top of kernel stack, below the PCB. > - Never load the userland fs/gsbase registers for kthreads, but preserve > curpcb->pcb_[fg]sbase as caches for the cpu. (Thanks Jeff!) > > Microbenchmarking this code seems to be really sensitive to things like > scheduling luck, timing, cache behavior, tlb behavior, kernel options, > other random code changes, etc. > > While it doesn't help heavy userland workloads much, it does help high > context switch loads a little, and should help those that involve > switching via kthreads a bit more. > > A special thanks to Kris for the testing and reality checks, and Jeff for > tormenting me into doing this. :) > > This is still work-in-progress. It looks like this patch introduces a regression. In particular, this chunk: @@ -181,82 +166,138 @@ sw1: cmpq %rcx, %rdx pause je 1b - lfence #endif is not totally right as we want to enforce an acq -- Peace can only be achieved by understanding - A. Einstein From attilio at freebsd.org Thu Jun 19 15:45:58 2008 From: attilio at freebsd.org (Attilio Rao) Date: Thu Jun 19 15:46:06 2008 Subject: cvs commit: src/sys/amd64/amd64 cpu_switch.S In-Reply-To: <3bbf2fe10806190842s381611del5c5dc27d2dd22a7e@mail.gmail.com> References: <200803232309.m2NN96Qa080896@repoman.freebsd.org> <3bbf2fe10806190842s381611del5c5dc27d2dd22a7e@mail.gmail.com> Message-ID: <3bbf2fe10806190845p15e0758cre88cd83ec0bd975d@mail.gmail.com> 2008/6/19, Attilio Rao : > 2008/3/24, Peter Wemm : > > > peter 2008-03-23 23:09:06 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/amd64/amd64 cpu_switch.S > > Log: > > First pass at (possibly futile) microoptimizing of cpu_switch. Results > > are mixed. Some pure context switch microbenchmarks show up to 29% > > improvement. Pipe based context switch microbenchmarks show up to 7% > > improvement. Real world tests are far less impressive as they are > > dominated more by actual work than switch overheads, but depending on > > the machine in question, workload, kernel options, phase of moon, etc, a > > few percent gain might be seen. > > > > Summary of changes: > > - don't reload MSR_[FG]SBASE registers when context switching between > > non-threaded userland apps. These typically cost 120 clock cycles each > > on an AMD cpu (less on Barcelona/Phenom). Intel cores are probably no > > faster on this. > > - The above change only helps unthreaded userland apps that tend to use > > the same value for gsbase. Threaded apps will get no benefit from this. > > - reorder things like accessing the pcb to be in memory order, to give > > prefetching a better chance of working. Operations are now in increasing > > memory address order, rather than reverse or random. > > - Push some lesser used code out of the main code paths. Hopefully > > allowing better code density in cache lines. This is probably futile. > > - (part 2 of previous item) Reorder code so that branches have a more > > realistic static branch prediction hint. Both Intel and AMD cpus > > default to predicting branches to lower memory addresses as being > > taken, and to higher memory addresses as not being taken. This is > > overridden by the limited dynamic branch prediction subsystem. A trip > > through userland might overflow this. > > - Futule attempt at spreading the use of the results of previous operations > > in new operations. Hopefully this will allow the cpus to execute in > > parallel better. > > - stop wasting 16 bytes at the top of kernel stack, below the PCB. > > - Never load the userland fs/gsbase registers for kthreads, but preserve > > curpcb->pcb_[fg]sbase as caches for the cpu. (Thanks Jeff!) > > > > Microbenchmarking this code seems to be really sensitive to things like > > scheduling luck, timing, cache behavior, tlb behavior, kernel options, > > other random code changes, etc. > > > > While it doesn't help heavy userland workloads much, it does help high > > context switch loads a little, and should help those that involve > > switching via kthreads a bit more. > > > > A special thanks to Kris for the testing and reality checks, and Jeff for > > tormenting me into doing this. :) > > > > This is still work-in-progress. > > > It looks like this patch introduces a regression. > In particular, this chunk: > > @@ -181,82 +166,138 @@ sw1: > cmpq %rcx, %rdx > pause > je 1b > - lfence > #endif > > is not totally right as we want to enforce an acq ...an acq memory barrier in order to handle correctly an eventual thread migration. We could use this approach, that is what I implemented on ia32 in order to solve the same problem: #define BLOCK_SPIN(reg) \ movl $blocked_lock,%eax ; \ 100: ; \ lock ; \ cmpxchgl %eax,TD_LOCK(reg) ; \ jne 101f ; \ pause ; \ jmp 100b ; \ 101: Thanks, Attilio [Sorry if I pushed "send" wrongly] -- Peace can only be achieved by understanding - A. Einstein From amdmi3 at FreeBSD.org Thu Jun 19 16:29:50 2008 From: amdmi3 at FreeBSD.org (Dmitry Marakasov) Date: Thu Jun 19 16:29:56 2008 Subject: cvs commit: src/usr.bin/calendar/calendars calendar.freebsd Message-ID: <200806191629.m5JGToAh019853@repoman.freebsd.org> amdmi3 2008-06-19 16:29:37 UTC FreeBSD src repository Modified files: usr.bin/calendar/calendars calendar.freebsd Log: SVN rev 179876 on 2008-06-19 16:29:37Z by amdmi3 Add myself. Approved by: miwi (mentor) Revision Changes Path 1.240 +1 -0 src/usr.bin/calendar/calendars/calendar.freebsd From amdmi3 at FreeBSD.org Thu Jun 19 16:30:21 2008 From: amdmi3 at FreeBSD.org (Dmitry Marakasov) Date: Thu Jun 19 16:30:26 2008 Subject: cvs commit: src/share/misc committers-ports.dot Message-ID: <200806191630.m5JGULxK019985@repoman.freebsd.org> amdmi3 2008-06-19 16:30:06 UTC FreeBSD src repository Modified files: share/misc committers-ports.dot Log: SVN rev 179877 on 2008-06-19 16:30:06Z by amdmi3 Add myself. Approved by: miwi (mentor) Revision Changes Path 1.69 +2 -0 src/share/misc/committers-ports.dot From mav at FreeBSD.org Thu Jun 19 17:10:24 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Thu Jun 19 17:10:30 2008 Subject: cvs commit: src/usr.bin/calendar/calendars calendar.freebsd Message-ID: <200806191710.m5JHAOZR033588@repoman.freebsd.org> mav 2008-06-19 17:10:05 UTC FreeBSD src repository Modified files: usr.bin/calendar/calendars calendar.freebsd Log: SVN rev 179878 on 2008-06-19 17:10:05Z by mav Add myself. Better late then never. Revision Changes Path 1.241 +1 -0 src/usr.bin/calendar/calendars/calendar.freebsd From gonzo at FreeBSD.org Thu Jun 19 18:08:56 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Thu Jun 19 18:09:01 2008 Subject: cvs commit: src/sys/kern sysv_sem.c Message-ID: <200806191808.m5JI8u4W043450@repoman.freebsd.org> gonzo 2008-06-19 18:08:42 UTC FreeBSD src repository Modified files: sys/kern sysv_sem.c Log: SVN rev 179879 on 2008-06-19 18:08:42Z by gonzo Renew semaphore's pointer after wakeup since during msleep sem_base may have been modified by destroying one of semaphores and semptr would not be valid in this case. PR: kern/123731 Revision Changes Path 1.91 +7 -0 src/sys/kern/sysv_sem.c From remko at FreeBSD.org Thu Jun 19 18:33:58 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Thu Jun 19 18:34:03 2008 Subject: cvs commit: src/share/man/man9 zone.9 Message-ID: <200806191833.m5JIXvwH046959@repoman.freebsd.org> remko 2008-06-19 18:33:38 UTC FreeBSD src repository Modified files: share/man/man9 zone.9 Log: SVN rev 179880 on 2008-06-19 18:33:38Z by remko Document the _arg versions of the uma_zalloc and uma_zfree functions. PR: docs/120357 Submitted by: gahr MFC after: 3 days Revision Changes Path 1.29 +18 -1 src/share/man/man9/zone.9 From delphij at FreeBSD.org Thu Jun 19 21:43:17 2008 From: delphij at FreeBSD.org (Xin LI) Date: Thu Jun 19 21:43:22 2008 Subject: cvs commit: src/sys/kern kern_jail.c src/sys/sys jail.h Message-ID: <200806192143.m5JLhHmE078610@repoman.freebsd.org> delphij 2008-06-19 21:41:57 UTC FreeBSD src repository Modified files: sys/kern kern_jail.c sys/sys jail.h Log: SVN rev 179881 on 2008-06-19 21:41:57Z by delphij Revert rev. 178124 as requested by kris@. Having jail id not being reused too frequently is useful for script controlled environment. Revision Changes Path 1.78 +24 -18 src/sys/kern/kern_jail.c 1.31 +2 -0 src/sys/sys/jail.h From das at FreeBSD.org Thu Jun 19 22:40:04 2008 From: das at FreeBSD.org (David Schultz) Date: Thu Jun 19 22:40:06 2008 Subject: cvs commit: src/lib/msun Makefile Symbol.map src/lib/msun/man fmod.3 src/lib/msun/src e_fmodl.c math.h Message-ID: <200806192240.m5JMe4Ol084499@repoman.freebsd.org> das 2008-06-19 22:39:53 UTC FreeBSD src repository Modified files: lib/msun Makefile Symbol.map lib/msun/man fmod.3 lib/msun/src math.h Added files: lib/msun/src e_fmodl.c Log: SVN rev 179882 on 2008-06-19 22:39:53Z by das Implement fmodl. Document fmodl and fix some errors in the fmod manpage. Revision Changes Path 1.94 +2 -2 src/lib/msun/Makefile 1.19 +1 -0 src/lib/msun/Symbol.map 1.10 +25 -21 src/lib/msun/man/fmod.3 1.1 +149 -0 src/lib/msun/src/e_fmodl.c (new) 1.73 +0 -2 src/lib/msun/src/math.h From scf at FreeBSD.org Thu Jun 19 23:25:21 2008 From: scf at FreeBSD.org (Sean Farley) Date: Thu Jun 19 23:25:23 2008 Subject: cvs commit: src/games/morse morse.6 morse.c Message-ID: <200806192325.m5JNPLjE089233@repoman.freebsd.org> scf 2008-06-19 23:25:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) games/morse morse.6 morse.c Log: SVN rev 179883 on 2008-06-19 23:25:00Z by scf MFC revisions: 179654 Fixed the output grammar to properly speak non-terminal dits. Updated Lyndon Nerenberg's radio callsign and E-mail address. PR: bin/7868 Submitted by: Lyndon Nerenberg Revision Changes Path 1.17.10.1 +3 -2 src/games/morse/morse.6 1.21.2.1 +4 -3 src/games/morse/morse.c From scf at FreeBSD.org Thu Jun 19 23:40:55 2008 From: scf at FreeBSD.org (Sean Farley) Date: Thu Jun 19 23:41:17 2008 Subject: cvs commit: src/games/morse morse.6 morse.c Message-ID: <200806192340.m5JNetDT090483@repoman.freebsd.org> scf 2008-06-19 23:40:32 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) games/morse morse.6 morse.c Log: SVN rev 179884 on 2008-06-19 23:40:32Z by scf MFC revisions: 179654 Fixed the output grammar to properly speak non-terminal dits. Updated Lyndon Nerenberg's radio callsign and E-mail address. PR: bin/7868 Submitted by: Lyndon Nerenberg Revision Changes Path 1.17.2.1 +3 -2 src/games/morse/morse.6 1.20.2.2 +4 -3 src/games/morse/morse.c From kan at FreeBSD.org Fri Jun 20 00:13:41 2008 From: kan at FreeBSD.org (Alexander Kabaev) Date: Fri Jun 20 00:13:44 2008 Subject: cvs commit: src/gnu/lib/libstdc++ Makefile Message-ID: <200806200013.m5K0Df4x094235@repoman.freebsd.org> kan 2008-06-20 00:13:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) gnu/lib/libstdc++ Makefile Log: SVN rev 179885 on 2008-06-20 00:13:37Z by kan MFC svn rev 179834. Install extra include files that were forgotten in original GCC 4.2.x import changes. PR: 124647 Submitted by: Vlad GALU Revision Changes Path 1.61.2.1 +15 -1 src/gnu/lib/libstdc++/Makefile From alc at FreeBSD.org Fri Jun 20 05:22:33 2008 From: alc at FreeBSD.org (Alan Cox) Date: Fri Jun 20 05:22:37 2008 Subject: cvs commit: src/sys/amd64/amd64 mem.c pmap.c trap.c src/sys/amd64/include pmc_mdep.h stack.h Message-ID: <200806200522.m5K5MWYQ043206@repoman.freebsd.org> alc 2008-06-20 05:22:09 UTC FreeBSD src repository Modified files: sys/amd64/amd64 mem.c pmap.c trap.c sys/amd64/include pmc_mdep.h stack.h Log: SVN rev 179886 on 2008-06-20 05:22:09Z by alc Make preparations for increasing the size of the kernel virtual address space on the amd64 architecture. The amd64 architecture requires kernel code and global variables to reside in the highest 2GB of the 64-bit virtual address space. Thus, KERNBASE cannot change. However, KERNBASE is sometimes used as the start of the kernel virtual address space. Henceforth, VM_MIN_KERNEL_ADDRESS should be used instead. Since KERNBASE and VM_MIN_KERNEL_ADDRESS are still the same address, there should be no visible effect from this change (yet). Revision Changes Path 1.123 +1 -1 src/sys/amd64/amd64/mem.c 1.622 +2 -2 src/sys/amd64/amd64/pmap.c 1.327 +1 -1 src/sys/amd64/amd64/trap.c 1.5 +2 -2 src/sys/amd64/include/pmc_mdep.h 1.2 +1 -1 src/sys/amd64/include/stack.h From alc at FreeBSD.org Fri Jun 20 06:24:49 2008 From: alc at FreeBSD.org (Alan Cox) Date: Fri Jun 20 06:24:55 2008 Subject: cvs commit: src/sys/kern link_elf_obj.c Message-ID: <200806200624.m5K6OnQs049677@repoman.freebsd.org> alc 2008-06-20 06:24:34 UTC FreeBSD src repository Modified files: sys/kern link_elf_obj.c Log: SVN rev 179887 on 2008-06-20 06:24:34Z by alc Enforce the mapping of kernel loadable modules in the uppermost 2GB of the kernel virtual address space on amd64. Revision Changes Path 1.100 +7 -0 src/sys/kern/link_elf_obj.c From joerg at FreeBSD.org Fri Jun 20 08:39:52 2008 From: joerg at FreeBSD.org (Joerg Wunsch) Date: Fri Jun 20 08:39:57 2008 Subject: cvs commit: src/usr.bin/whereis whereis.c Message-ID: <200806200839.m5K8dp1c063576@repoman.freebsd.org> joerg 2008-06-20 08:39:42 UTC FreeBSD src repository Modified files: usr.bin/whereis whereis.c Log: SVN rev 179888 on 2008-06-20 08:39:42Z by joerg Make the search for sources in PATH_PORTS more accurate. I only noticed that a "whereis -qs qemu" matched the distfiles subdir of qemu rather than /usr/ports/emulators/qemu. It now ignores all dot entries in /usr/ports, plus all entries starting with a capital letter (maintenance stuff like Templates, but also includes subdir CVS), plus /usr/ports/distfiles which is simply a magic name in that respect. Revision Changes Path 1.16 +21 -2 src/usr.bin/whereis/whereis.c From joerg at FreeBSD.org Fri Jun 20 08:48:53 2008 From: joerg at FreeBSD.org (Joerg Wunsch) Date: Fri Jun 20 08:48:55 2008 Subject: cvs commit: src/usr.bin/whereis whereis.c Message-ID: <200806200848.m5K8mqlD073549@repoman.freebsd.org> joerg 2008-06-20 08:48:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/whereis whereis.c Log: SVN rev 179889 on 2008-06-20 08:48:33Z by joerg MFC: more intelligent handling of /usr/ports subdirectories. Revision Changes Path 1.15.10.1 +21 -2 src/usr.bin/whereis/whereis.c From joerg at FreeBSD.org Fri Jun 20 08:50:09 2008 From: joerg at FreeBSD.org (Joerg Wunsch) Date: Fri Jun 20 08:50:11 2008 Subject: cvs commit: src/usr.bin/whereis whereis.c Message-ID: <200806200850.m5K8o8iV073696@repoman.freebsd.org> joerg 2008-06-20 08:49:58 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/whereis whereis.c Log: SVN rev 179890 on 2008-06-20 08:49:58Z by joerg MFC: more intelligent handling of /usr/ports subdirectories. Revision Changes Path 1.15.2.1 +21 -2 src/usr.bin/whereis/whereis.c From brooks at FreeBSD.org Fri Jun 20 12:47:12 2008 From: brooks at FreeBSD.org (Brooks Davis) Date: Fri Jun 20 12:47:14 2008 Subject: cvs commit: CVSROOT access Message-ID: <200806201247.m5KClCZk004908@repoman.freebsd.org> brooks 2008-06-20 12:47:06 UTC FreeBSD src repository Modified files: . access Log: SVN rev 179891 on 2008-06-20 12:47:06Z by brooks Please welcome Erik Cederstrand to the ranks of src committers. He will be working on things related to his automated performance tracking system. kris and I will mentor. Approved by: core Revision Changes Path 1.884 +1 -0 CVSROOT/access From phk at FreeBSD.org Fri Jun 20 14:47:14 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Fri Jun 20 14:47:22 2008 Subject: cvs commit: src/lib/libc/sys accept.2 Message-ID: <200806201447.m5KElEYI024580@repoman.freebsd.org> phk 2008-06-20 14:47:06 UTC FreeBSD src repository Modified files: lib/libc/sys accept.2 Log: SVN rev 179892 on 2008-06-20 14:47:06Z by phk Add Xr to getsockname(2) Revision Changes Path 1.33 +1 -0 src/lib/libc/sys/accept.2 From imp at FreeBSD.org Fri Jun 20 16:59:22 2008 From: imp at FreeBSD.org (Warner Losh) Date: Fri Jun 20 16:59:24 2008 Subject: cvs commit: src/sys/kern subr_bus.c src/sys/sys bus.h Message-ID: <200806201659.m5KGxLmq047885@repoman.freebsd.org> imp 2008-06-20 16:58:15 UTC FreeBSD src repository Modified files: sys/kern subr_bus.c sys/sys bus.h Log: SVN rev 179893 on 2008-06-20 16:58:15Z by imp Split out the probing magic of device_probe_and_attach into device_probe() so that it can be used by busses that may wish to do additional processing between probe and attach. Reviewed by: dfr@ Revision Changes Path 1.209 +25 -6 src/sys/kern/subr_bus.c 1.82 +1 -0 src/sys/sys/bus.h From thompsa at FreeBSD.org Fri Jun 20 17:30:10 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Fri Jun 20 17:30:15 2008 Subject: cvs commit: src/sbin/ifconfig ifconfig.8 ifconfig.c src/share/man/man4 gre.4 src/sys/net if_gre.c if_gre.h Message-ID: <200806201730.m5KHUAbT051544@repoman.freebsd.org> thompsa 2008-06-20 17:26:34 UTC FreeBSD src repository Modified files: sbin/ifconfig ifconfig.8 ifconfig.c share/man/man4 gre.4 sys/net if_gre.c if_gre.h Log: SVN rev 179894 on 2008-06-20 17:26:34Z by thompsa Add support for the optional key in the GRE header. PR: kern/114714 Submitted by: Cristian KLEIN Revision Changes Path 1.148 +11 -1 src/sbin/ifconfig/ifconfig.8 1.137 +20 -0 src/sbin/ifconfig/ifconfig.c 1.8 +14 -2 src/share/man/man4/gre.4 1.49 +44 -3 src/sys/net/if_gre.c 1.15 +7 -0 src/sys/net/if_gre.h From delphij at FreeBSD.org Fri Jun 20 19:32:39 2008 From: delphij at FreeBSD.org (Xin LI) Date: Fri Jun 20 19:32:41 2008 Subject: cvs commit: src/sys/amd64/conf GENERIC src/sys/conf NOTES files src/sys/dev/et if_et.c if_etreg.h if_etvar.h src/sys/dev/mii miidevs src/sys/i386/conf GENERIC src/sys/modules Makefile src/sys/modules/et Makefile src/sys/modules/mii Makefile Message-ID: <200806201932.m5KJWdGJ068730@repoman.freebsd.org> delphij 2008-06-20 19:28:33 UTC FreeBSD src repository Modified files: sys/amd64/conf GENERIC sys/conf NOTES files sys/dev/mii miidevs sys/i386/conf GENERIC sys/modules Makefile sys/modules/mii Makefile Added files: sys/dev/et if_et.c if_etreg.h if_etvar.h sys/modules/et Makefile Log: SVN rev 179895 on 2008-06-20 19:28:33Z by delphij Add et(4), a port of DragonFly's Agere ET1310 10/100/Gigabit Ethernet device driver, written by sephe@ Obtained from: DragonFly Sponsored by: iXsystems MFC after: 2 weeks Revision Changes Path 1.502 +1 -0 src/sys/amd64/conf/GENERIC 1.1495 +1 -0 src/sys/conf/NOTES 1.1310 +2 -0 src/sys/conf/files 1.1 +2389 -0 src/sys/dev/et/if_et.c (new) 1.1 +436 -0 src/sys/dev/et/if_etreg.h (new) 1.1 +279 -0 src/sys/dev/et/if_etvar.h (new) 1.58 +4 -0 src/sys/dev/mii/miidevs 1.492 +1 -0 src/sys/i386/conf/GENERIC 1.574 +3 -0 src/sys/modules/Makefile 1.1 +10 -0 src/sys/modules/et/Makefile (new) 1.35 +2 -1 src/sys/modules/mii/Makefile From delphij at FreeBSD.org Fri Jun 20 19:33:08 2008 From: delphij at FreeBSD.org (Xin LI) Date: Fri Jun 20 19:33:10 2008 Subject: cvs commit: src/sys/dev/mii truephy.c truephyreg.h Message-ID: <200806201933.m5KJX7RE068811@repoman.freebsd.org> delphij 2008-06-20 19:30:44 UTC FreeBSD src repository Added files: sys/dev/mii truephy.c truephyreg.h Log: SVN rev 179896 on 2008-06-20 19:30:44Z by delphij Add et(4), a port of DragonFly's Agere ET1310 10/100/Gigabit Ethernet device driver, written by sephe@ Obtained from: DragonFly Sponsored by: iXsystems MFC after: 2 weeks Revision Changes Path 1.1 +355 -0 src/sys/dev/mii/truephy.c (new) 1.1 +64 -0 src/sys/dev/mii/truephyreg.h (new) From lulf at FreeBSD.org Fri Jun 20 19:48:24 2008 From: lulf at FreeBSD.org (Ulf Lilleengen) Date: Fri Jun 20 19:48:29 2008 Subject: cvs commit: src/sys/geom/journal g_journal.c Message-ID: <200806201948.m5KJmNBc070059@repoman.freebsd.org> lulf 2008-06-20 19:48:18 UTC FreeBSD src repository Modified files: sys/geom/journal g_journal.c Log: SVN rev 179897 on 2008-06-20 19:48:18Z by lulf - Fix spelling errors. Approved by: kib (mentor) PR: kern/124788 Submitted by: Hywel Mallett Revision Changes Path 1.15 +5 -5 src/sys/geom/journal/g_journal.c From alc at FreeBSD.org Fri Jun 20 20:59:40 2008 From: alc at FreeBSD.org (Alan Cox) Date: Fri Jun 20 20:59:43 2008 Subject: cvs commit: src/sys/amd64/amd64 minidump_machdep.c Message-ID: <200806202059.m5KKxetE086520@repoman.freebsd.org> alc 2008-06-20 20:59:31 UTC FreeBSD src repository Modified files: sys/amd64/amd64 minidump_machdep.c Log: SVN rev 179898 on 2008-06-20 20:59:31Z by alc Make preparations for increasing the size of the kernel virtual address space on the amd64 architecture. The amd64 architecture requires kernel code and global variables to reside in the highest 2GB of the 64-bit virtual address space. Thus, KERNBASE cannot change. However, KERNBASE is sometimes used as the start of the kernel virtual address space. Henceforth, VM_MIN_KERNEL_ADDRESS should be used instead. Since KERNBASE and VM_MIN_KERNEL_ADDRESS are still the same address, there should be no visible effect from this change (yet). That said, kris@ has tested crash dumps under the full patch that increases the kernel virtual address space on amd64 to 6GB. Tested by: kris@ Revision Changes Path 1.5 +3 -3 src/sys/amd64/amd64/minidump_machdep.c From lulf at FreeBSD.org Fri Jun 20 21:17:43 2008 From: lulf at FreeBSD.org (Ulf Lilleengen) Date: Fri Jun 20 21:17:48 2008 Subject: cvs commit: src/sys/geom/journal g_journal.c Message-ID: <200806202117.m5KLHfIY089244@repoman.freebsd.org> lulf 2008-06-20 21:17:40 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/geom/journal g_journal.c Log: SVN rev 179899 on 2008-06-20 21:17:40Z by lulf MFC r179897: - Fix spelling errors. Approved by: kib (mentor) PR: kern/124788 Submitted by: Hywel Mallett Revision Changes Path 1.13.2.1 +5 -5 src/sys/geom/journal/g_journal.c From gonzo at FreeBSD.org Fri Jun 20 21:41:58 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Fri Jun 20 21:42:03 2008 Subject: cvs commit: src/sbin/ggate/ggated ggated.c Message-ID: <200806202141.m5KLfvHU091337@repoman.freebsd.org> gonzo 2008-06-20 21:41:44 UTC FreeBSD src repository Modified files: sbin/ggate/ggated ggated.c Log: SVN rev 179900 on 2008-06-20 21:41:44Z by gonzo Fix spelling PR: kern/124723 Event: Bugathon#5 Revision Changes Path 1.10 +2 -2 src/sbin/ggate/ggated/ggated.c From gonzo at FreeBSD.org Fri Jun 20 22:23:59 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Fri Jun 20 22:24:01 2008 Subject: cvs commit: src/sys/dev/agp agp_i810.c Message-ID: <200806202223.m5KMNwvp097058@repoman.freebsd.org> gonzo 2008-06-20 22:23:41 UTC FreeBSD src repository Modified files: sys/dev/agp agp_i810.c Log: SVN rev 179901 on 2008-06-20 22:23:41Z by gonzo Fix PCI id for 945GME Express Integrated Graphics Controller: set to 8086:27AE PR: kern/124782 Event: Bugathon#5 Revision Changes Path 1.46 +1 -1 src/sys/dev/agp/agp_i810.c From danger at FreeBSD.org Fri Jun 20 22:52:45 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Fri Jun 20 22:52:47 2008 Subject: cvs commit: src/share/man/man4 blackhole.4 Message-ID: <200806202252.m5KMqi5i099471@repoman.freebsd.org> danger 2008-06-20 22:52:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 blackhole.4 Log: SVN rev 179902 on 2008-06-20 22:52:33Z by danger MFC: Do not claim that ipfw is the only firewall package available in FreeBSD. PR: docs/106315 Revision Changes Path 1.12.2.1 +8 -7 src/share/man/man4/blackhole.4 From danger at FreeBSD.org Fri Jun 20 23:04:46 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Fri Jun 20 23:04:49 2008 Subject: cvs commit: src/share/man/man4 cpufreq.4 Message-ID: <200806202304.m5KN4kdE001812@repoman.freebsd.org> danger 2008-06-20 23:04:34 UTC FreeBSD src repository Modified files: share/man/man4 cpufreq.4 Log: SVN rev 179903 on 2008-06-20 23:04:34Z by danger - link powerd(8) man page PR: docs/123111 MFC after: 3 days Revision Changes Path 1.16 +1 -0 src/share/man/man4/cpufreq.4 From pgollucci at p6m7g8.com Sat Jun 21 00:52:38 2008 From: pgollucci at p6m7g8.com (Philip M. Gollucci) Date: Sat Jun 21 00:52:42 2008 Subject: cvs commit: src/usr.bin/whereis whereis.c In-Reply-To: <200806200839.m5K8dp1c063576@repoman.freebsd.org> References: <200806200839.m5K8dp1c063576@repoman.freebsd.org> Message-ID: <485C4E76.2000704@p6m7g8.com> Joerg Wunsch wrote: > joerg 2008-06-20 08:39:42 UTC > > FreeBSD src repository > > Modified files: > usr.bin/whereis whereis.c > Log: > SVN rev 179888 on 2008-06-20 08:39:42Z by joerg > > Make the search for sources in PATH_PORTS more accurate. I only > noticed that a "whereis -qs qemu" matched the distfiles subdir of qemu > rather than /usr/ports/emulators/qemu. > > It now ignores all dot entries in /usr/ports, plus all entries > starting with a capital letter (maintenance stuff like Templates, but > also includes subdir CVS), plus /usr/ports/distfiles which is simply a > magic name in that respect. While your at it, what about .svn dirs .... I guess thats covered by 'dot' entries. > | if (dirp->d_name[0] == '.' || > | - strcmp(dirp->d_name, "CVS") == 0) > | - /* ignore dot entries and CVS subdir */ -- ------------------------------------------------------------------------ Philip M. Gollucci (philip@ridecharge.com) o:703.549.2050x206 Senior System Admin - Riderway, Inc. http://riderway.com / http://ridecharge.com 1024D/DB9B8C1C B90B FBC3 A3A1 C71A 8E70 3F8C 75B8 8FFB DB9B 8C1C Work like you don't need the money, love like you'll never get hurt, and dance like nobody's watching. From kientzle at FreeBSD.org Sat Jun 21 02:17:35 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Jun 21 02:17:40 2008 Subject: cvs commit: src/usr.bin/cpio/test main.c test.h Message-ID: <200806210217.m5L2HYVd041140@repoman.freebsd.org> kientzle 2008-06-21 02:17:18 UTC FreeBSD src repository Modified files: usr.bin/cpio/test main.c test.h Log: SVN rev 179904 on 2008-06-21 02:17:18Z by kientzle MfP4: test improvements, mostly for portability. Revision Changes Path 1.2 +48 -3 src/usr.bin/cpio/test/main.c 1.2 +20 -17 src/usr.bin/cpio/test/test.h From kientzle at FreeBSD.org Sat Jun 21 02:19:06 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Jun 21 02:19:10 2008 Subject: cvs commit: src/usr.bin/cpio cmdline.c Message-ID: <200806210219.m5L2J5p4041316@repoman.freebsd.org> kientzle 2008-06-21 02:18:52 UTC FreeBSD src repository Modified files: usr.bin/cpio cmdline.c Log: SVN rev 179905 on 2008-06-21 02:18:52Z by kientzle Various long options for GNU cpio compat. Revision Changes Path 1.2 +8 -0 src/usr.bin/cpio/cmdline.c From kientzle at FreeBSD.org Sat Jun 21 02:20:37 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Jun 21 02:20:43 2008 Subject: cvs commit: src/usr.bin/cpio cmdline.c cpio.c cpio.h matching.c Message-ID: <200806210220.m5L2KacJ041585@repoman.freebsd.org> kientzle 2008-06-21 02:20:20 UTC FreeBSD src repository Modified files: usr.bin/cpio cmdline.c cpio.c cpio.h matching.c Log: SVN rev 179906 on 2008-06-21 02:20:20Z by kientzle Rework line-processing framework to add support for --null and to eliminate a callback. Revision Changes Path 1.3 +2 -1 src/usr.bin/cpio/cmdline.c 1.2 +111 -57 src/usr.bin/cpio/cpio.c 1.2 +7 -3 src/usr.bin/cpio/cpio.h 1.2 +10 -1 src/usr.bin/cpio/matching.c From kib at FreeBSD.org Sat Jun 21 08:51:26 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Sat Jun 21 08:51:31 2008 Subject: cvs commit: src/sys/boot/common ufsread.c Message-ID: <200806210851.m5L8pPge002662@repoman.freebsd.org> kib 2008-06-21 08:51:21 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/common ufsread.c Log: SVN rev 179907 on 2008-06-21 08:51:21Z by kib MFC r179634: Fix the incorrect calculation of a block address within a single indirect block. PR: 108215 Revision Changes Path 1.14.10.4 +1 -1 src/sys/boot/common/ufsread.c From gonzo at FreeBSD.org Sat Jun 21 11:34:43 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Sat Jun 21 11:34:48 2008 Subject: cvs commit: src/sys/kern vfs_aio.c Message-ID: <200806211134.m5LBYhPw021232@repoman.freebsd.org> gonzo 2008-06-21 11:34:34 UTC FreeBSD src repository Modified files: sys/kern vfs_aio.c Log: SVN rev 179908 on 2008-06-21 11:34:34Z by gonzo Use minimum of max_aio_procs and target_aio_procs when spawning new aiod since there should be no more then max_aio_procs processes. Revision Changes Path 1.239 +1 -1 src/sys/kern/vfs_aio.c From simon at FreeBSD.org Sat Jun 21 13:53:45 2008 From: simon at FreeBSD.org (Simon L. Nielsen) Date: Sat Jun 21 13:53:46 2008 Subject: cvs commit: src/share/man/man5 rc.conf.5 Message-ID: <200806211353.m5LDriI3044681@repoman.freebsd.org> simon 2008-06-21 13:53:35 UTC FreeBSD src repository Modified files: share/man/man5 rc.conf.5 Log: SVN rev 179909 on 2008-06-21 13:53:35Z by simon To catch up with rev 179872: rename enable_quotas to quota_enable. Revision Changes Path 1.343 +5 -5 src/share/man/man5/rc.conf.5 From marck at FreeBSD.org Sat Jun 21 15:05:01 2008 From: marck at FreeBSD.org (Dmitry Morozovsky) Date: Sat Jun 21 15:05:03 2008 Subject: cvs commit: src/sbin/mdconfig mdconfig.8 mdconfig.c Message-ID: <200806211505.m5LF51hi052645@repoman.freebsd.org> marck 2008-06-21 15:04:42 UTC FreeBSD src repository Modified files: sbin/mdconfig mdconfig.8 mdconfig.c Log: SVN rev 179910 on 2008-06-21 15:04:42Z by marck Add -v (verbose) option to -l command, to show size and backing store of all md devices at one time. Approved by: phk MFC after: 2 weeks Revision Changes Path 1.43 +5 -1 src/sbin/mdconfig/mdconfig.8 1.56 +12 -6 src/sbin/mdconfig/mdconfig.c From ru at FreeBSD.org Sat Jun 21 15:48:19 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Sat Jun 21 15:48:27 2008 Subject: cvs commit: src/contrib/top top.X top.c Message-ID: <200806211548.m5LFmHMX056296@repoman.freebsd.org> ru 2008-06-21 15:48:16 UTC FreeBSD src repository Modified files: contrib/top top.X top.c Log: SVN rev 179911 on 2008-06-21 15:48:16Z by ru Removed the no-op -p; documented -P. MFC after: 3 days Revision Changes Path 1.21 +4 -1 src/contrib/top/top.X 1.26 +2 -6 src/contrib/top/top.c From mav at FreeBSD.org Sat Jun 21 16:23:04 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sat Jun 21 16:23:09 2008 Subject: cvs commit: src/sys/netinet/libalias alias_ftp.c Message-ID: <200806211623.m5LGN3bV060272@repoman.freebsd.org> mav 2008-06-21 16:22:56 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias_ftp.c Log: SVN rev 179912 on 2008-06-21 16:22:56Z by mav Add support for PORT/EPRT FTP commands in lowercase. Use strncasecmp() instead of huge local implementation to reduce code size. Check space presence after command/code. PR: kern/73034 Revision Changes Path 1.31 +21 -96 src/sys/netinet/libalias/alias_ftp.c From kientzle at FreeBSD.org Sat Jun 21 17:48:06 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Jun 21 17:48:13 2008 Subject: cvs commit: src/usr.bin/cpio cpio.c Message-ID: <200806211748.m5LHm6Pn077709@repoman.freebsd.org> kientzle 2008-06-21 17:47:56 UTC FreeBSD src repository Modified files: usr.bin/cpio cpio.c Log: SVN rev 179913 on 2008-06-21 17:47:56Z by kientzle If we're using -l and can't hardlink the file because of a cross-device link, just ignore the -l option and copy the file instead. In particular, this should fix the COPYTREE_* macros used in the ports infrastructure which use -l to preserve space but often get used for cross-device copies. Revision Changes Path 1.3 +17 -3 src/usr.bin/cpio/cpio.c From kientzle at FreeBSD.org Sat Jun 21 19:05:38 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Jun 21 19:05:43 2008 Subject: cvs commit: src/lib/libarchive archive_write_disk.c Message-ID: <200806211905.m5LJ5cf0087506@repoman.freebsd.org> kientzle 2008-06-21 19:05:29 UTC FreeBSD src repository Modified files: lib/libarchive archive_write_disk.c Log: SVN rev 179914 on 2008-06-21 19:05:29Z by kientzle A security-check failure here should be ARCHIVE_FAILED (cannot continue this operation) and not ARCHIVE_WARN, since we don't actually open the file. Both bsdtar and bsdcpio will try to copy file contents after an ARCHIVE_WARN, which will fail loudly. Revision Changes Path 1.26 +3 -3 src/lib/libarchive/archive_write_disk.c From kientzle at FreeBSD.org Sat Jun 21 19:06:54 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Jun 21 19:06:56 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_mtree.c Message-ID: <200806211906.m5LJ6sve087606@repoman.freebsd.org> kientzle 2008-06-21 19:06:37 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_mtree.c Log: SVN rev 179915 on 2008-06-21 19:06:37Z by kientzle MfP4: Joerg Sonnenberg's extensions to the mtree support for more complete quoting. In particular, this handles escaped newlines and common C-style escapes. Revision Changes Path 1.9 +79 -12 src/lib/libarchive/archive_read_support_format_mtree.c From kientzle at FreeBSD.org Sat Jun 21 19:11:55 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Sat Jun 21 19:11:57 2008 Subject: cvs commit: src/lib/libarchive/test test_read_format_isorr_bz2.c test_read_format_isorr_bz2.iso.bz2.uu Message-ID: <200806211911.m5LJBtFL088046@repoman.freebsd.org> kientzle 2008-06-21 19:11:51 UTC FreeBSD src repository Modified files: lib/libarchive/test test_read_format_isorr_bz2.c Added files: lib/libarchive/test test_read_format_isorr_bz2.iso.bz2.uu Log: SVN rev 179916 on 2008-06-21 19:11:51Z by kientzle Refactor one of the ISO extraction tests: Move the reference file into a separate file (instead of embedding it in the C code) and use later timestamps (timestamps too close to the Epoch fail predictably on systems that lack timegm(), whose mktime() doesn't support dates before the Epoch and which are running in timezones with negative offsets from GMT). The goal here is to test the ISO extraction, not the local platform's time support. Revision Changes Path 1.4 +52 -94 src/lib/libarchive/test/test_read_format_isorr_bz2.c 1.1 +24 -0 src/lib/libarchive/test/test_read_format_isorr_bz2.iso.bz2.uu (new) From alc at FreeBSD.org Sat Jun 21 19:19:29 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sat Jun 21 19:19:35 2008 Subject: cvs commit: src/sys/amd64/amd64 pmap.c Message-ID: <200806211919.m5LJJTvE088600@repoman.freebsd.org> alc 2008-06-21 19:19:09 UTC FreeBSD src repository Modified files: sys/amd64/amd64 pmap.c Log: SVN rev 179917 on 2008-06-21 19:19:09Z by alc Prepare for a larger kernel virtual address space. Specifically, once KERNBASE and VM_MIN_KERNEL_ADDRESS are no longer the same, the physical memory allocated during bootstrap will be offset from the low-end of the kernel's page table. Revision Changes Path 1.623 +8 -4 src/sys/amd64/amd64/pmap.c From das at FreeBSD.org Sat Jun 21 19:28:01 2008 From: das at FreeBSD.org (David Schultz) Date: Sat Jun 21 19:28:09 2008 Subject: cvs commit: src/contrib/gdtoa gethex.c Message-ID: <200806211928.m5LJS1Iw089370@repoman.freebsd.org> das 2008-06-21 19:27:54 UTC FreeBSD src repository Modified files: contrib/gdtoa gethex.c Log: SVN rev 179918 on 2008-06-21 19:27:54Z by das Bring in the vendor's fix for a bug in strtod() whereby strtod("0xyz", &endp) resulted in endp pointing to "0xyz" instead of "xyz". Reported by: Tony Finch MFC after: 1 week Revision Changes Path 1.2 +39 -31 src/contrib/gdtoa/gethex.c From das at FreeBSD.org Sat Jun 21 19:28:47 2008 From: das at FreeBSD.org (David Schultz) Date: Sat Jun 21 19:28:50 2008 Subject: cvs commit: src/tools/regression/lib/libc/stdio test-scanfloat.c Message-ID: <200806211928.m5LJSkA8089439@repoman.freebsd.org> das 2008-06-21 19:28:26 UTC FreeBSD src repository Modified files: tools/regression/lib/libc/stdio test-scanfloat.c Log: SVN rev 179919 on 2008-06-21 19:28:26Z by das Regression test for a recently fixed strtod bug. Revision Changes Path 1.10 +11 -1 src/tools/regression/lib/libc/stdio/test-scanfloat.c From mav at FreeBSD.org Sat Jun 21 20:19:03 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sat Jun 21 20:19:05 2008 Subject: cvs commit: src/sys/netinet/libalias alias.c Message-ID: <200806212019.m5LKJ2tT095444@repoman.freebsd.org> mav 2008-06-21 20:18:57 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias.c Log: SVN rev 179920 on 2008-06-21 20:18:57Z by mav Implement UDP transparent proxy support. PR: bin/54274 Submitted by: Nicolai Petri Revision Changes Path 1.64 +76 -10 src/sys/netinet/libalias/alias.c From alc at FreeBSD.org Sat Jun 21 21:02:19 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sat Jun 21 21:02:20 2008 Subject: cvs commit: src/sys/vm vm_map.c Message-ID: <200806212102.m5LL2IgR008366@repoman.freebsd.org> alc 2008-06-21 21:02:13 UTC FreeBSD src repository Modified files: sys/vm vm_map.c Log: SVN rev 179921 on 2008-06-21 21:02:13Z by alc KERNBASE is not necessarily an address within the kernel map, e.g., PowerPC/AIM. Consequently, it should not be used to determine the maximum number of kernel map entries. Intead, use VM_MIN_KERNEL_ADDRESS, which marks the start of the kernel map on all architectures. Tested by: marcel@ (PowerPC/AIM) Revision Changes Path 1.397 +1 -1 src/sys/vm/vm_map.c From peter at FreeBSD.org Sun Jun 22 00:13:38 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Sun Jun 22 00:13:43 2008 Subject: cvs commit: src/gnu/usr.bin/cvs/contrib Makefile Message-ID: <200806220013.m5M0Db7Q028259@repoman.freebsd.org> peter 2008-06-22 00:13:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) gnu/usr.bin/cvs/contrib Makefile Log: SVN rev 179922 on 2008-06-22 00:13:19Z by peter MFC: revs 175291,175384. Add .POSIX: to break timestamp-sensitive problems between Makefile and Makefile.in that was disturbed by changes in src/usr.bin/make/main.c rev 167330 (1.161). Revision Changes Path 1.18.10.1 +5 -3 src/gnu/usr.bin/cvs/contrib/Makefile From alc at FreeBSD.org Sun Jun 22 04:54:41 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sun Jun 22 04:54:44 2008 Subject: cvs commit: src/sys/vm vm_kern.c Message-ID: <200806220454.m5M4seVF077085@repoman.freebsd.org> alc 2008-06-22 04:54:27 UTC FreeBSD src repository Modified files: sys/vm vm_kern.c Log: SVN rev 179923 on 2008-06-22 04:54:27Z by alc Make preparations for increasing the size of the kernel virtual address space on the amd64 architecture. The amd64 architecture requires kernel code and global variables to reside in the highest 2GB of the 64-bit virtual address space. Thus, the memory allocated during bootstrap, before the call to kmem_init(), starts at KERNBASE, which is not necessarily the same as VM_MIN_KERNEL_ADDRESS on amd64. Revision Changes Path 1.135 +6 -2 src/sys/vm/vm_kern.c From mav at FreeBSD.org Sun Jun 22 11:39:46 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 22 11:39:51 2008 Subject: cvs commit: src/sys/netinet/libalias alias_db.c Message-ID: <200806221139.m5MBdimu026572@repoman.freebsd.org> mav 2008-06-22 11:39:42 UTC FreeBSD src repository Modified files: sys/netinet/libalias alias_db.c Log: SVN rev 179924 on 2008-06-22 11:39:42Z by mav Partially revert previous commit. DeleteLink() does not deletes permanent links so we should be aware of it and try to delete every link only once or we will loop forever. Revision Changes Path 1.74 +5 -1 src/sys/netinet/libalias/alias_db.c From marius at FreeBSD.org Sun Jun 22 13:55:02 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Sun Jun 22 13:55:04 2008 Subject: cvs commit: src/sys/dev/gem if_gem.c Message-ID: <200806221355.m5MDt1nB049274@repoman.freebsd.org> marius 2008-06-22 13:54:51 UTC FreeBSD src repository Modified files: sys/dev/gem if_gem.c Log: SVN rev 179925 on 2008-06-22 13:54:51Z by marius o The FreeBSD bus_dmamap_sync(9) supports ored together flags for quite some time now so collapse calls accordingly. o Given that gem_load_txmbuf() is allowed to fail resulting in a packet drop also for quite some time now implement the functionality of gem_txcksum() by means of m_pullup(9), which de-obfuscates the code and allows to always retrieve the correct length of the IP header. o Add missing BUS_DMASYNC_PREREAD when syncing the control DMA maps in gem_rint() and gem_start_locked(). o Correct some bus_barrier(9) calls to do a read/write barrier as we do a read after a write. Add some missing ones in gem_mii_readreg() and gem_mii_writereg(). o According to the Apple GMAC driver, the GEM ASIC specification and the OpenSolaris eri(7D) the TX FIFO threshold has to be set to 0x4ff for the Gigabit variants and 0x100 for the ERI in order do avoid TX underruns. o In gem_init_locked(): - be conservative and enable the RX and TX MACs, - don't clear GEM_LINK otherwise we don't ever mark the link as up again if gem_init_locked() is called from gem_watchdog(), - remove superfluous setting of sc_ifflags. o Don't bother to check whether the interface is running or whether its queue is empty before calling gem_start_locked() in gem_tint(), the former will check these anyway. o Call gem_start_locked() in gem_watchdog() in order to try to get some more packets going. o In gem_mii_writereg() after reseting the PCS restore its configuration. GMAC testing: grehan, marcel MFC after: 2 weeks Revision Changes Path 1.50 +90 -90 src/sys/dev/gem/if_gem.c From gonzo at FreeBSD.org Sun Jun 22 14:34:48 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Sun Jun 22 14:34:54 2008 Subject: cvs commit: src/sys/fs/devfs devfs_rule.c Message-ID: <200806221434.m5MEYlTr053683@repoman.freebsd.org> gonzo 2008-06-22 14:34:38 UTC FreeBSD src repository Modified files: sys/fs/devfs devfs_rule.c Log: SVN rev 179926 on 2008-06-22 14:34:38Z by gonzo Get pointer to devfs_ruleset struct after garbage collection has been performed. Otherwise if ruleset is used by given mountpoint and is empty it's freed by devfs_ruleset_reap and pointer becomes bogus. Submitted by: Mateusz Guzik PR: kern/124853 Revision Changes Path 1.25 +3 -3 src/sys/fs/devfs/devfs_rule.c From mtm at FreeBSD.org Sun Jun 22 15:34:49 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Sun Jun 22 15:34:51 2008 Subject: cvs commit: src/etc/rc.d localpkg Message-ID: <200806221534.m5MFYmcv059486@repoman.freebsd.org> mtm 2008-06-22 15:34:40 UTC FreeBSD src repository Modified files: etc/rc.d localpkg Log: SVN rev 179927 on 2008-06-22 15:34:40Z by mtm Don't say we're going to [start|stop] local packages unless there actually are local (pre rc.d) scripts to run. Revision Changes Path 1.9 +16 -4 src/etc/rc.d/localpkg From mtm at FreeBSD.org Sun Jun 22 15:40:44 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Sun Jun 22 15:40:50 2008 Subject: cvs commit: src/etc/rc.d mountcritremote mountlate Message-ID: <200806221540.m5MFei4j060090@repoman.freebsd.org> mtm 2008-06-22 15:40:19 UTC FreeBSD src repository Modified files: etc/rc.d mountcritremote mountlate Log: SVN rev 179928 on 2008-06-22 15:40:19Z by mtm Don't say we're going to mount filesystems of a certain type unless there actually are filesystems of that type to mount. Revision Changes Path 1.19 +9 -3 src/etc/rc.d/mountcritremote 1.6 +16 -5 src/etc/rc.d/mountlate From mtm at FreeBSD.org Sun Jun 22 15:58:00 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Sun Jun 22 15:58:07 2008 Subject: cvs commit: src/etc/rc.d archdep Message-ID: <200806221558.m5MFw0s3061405@repoman.freebsd.org> mtm 2008-06-22 15:57:50 UTC FreeBSD src repository Modified files: etc/rc.d archdep Log: SVN rev 179929 on 2008-06-22 15:57:50Z by mtm Simplify this script with the added bonus that the bit about i386 initialization doesn't get printed unless ibcs2_enable is set. Revision Changes Path 1.15 +17 -22 src/etc/rc.d/archdep From mtm at FreeBSD.org Sun Jun 22 16:20:02 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Sun Jun 22 16:20:07 2008 Subject: cvs commit: src/etc/rc.d abi Message-ID: <200806221620.m5MGK1rF064277@repoman.freebsd.org> mtm 2008-06-22 16:19:50 UTC FreeBSD src repository Modified files: etc/rc.d abi Log: SVN rev 179930 on 2008-06-22 16:19:50Z by mtm Do not print anything unless at least one of the abi emulators is enabled. Revision Changes Path 1.12 +8 -2 src/etc/rc.d/abi From mtm at FreeBSD.org Sun Jun 22 16:23:47 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Sun Jun 22 16:23:53 2008 Subject: cvs commit: src/etc/rc.d local Message-ID: <200806221623.m5MGNl0S064609@repoman.freebsd.org> mtm 2008-06-22 16:23:39 UTC FreeBSD src repository Modified files: etc/rc.d local Log: SVN rev 179931 on 2008-06-22 16:23:39Z by mtm Output information only if /etc/rc.local exists. Revision Changes Path 1.8 +4 -4 src/etc/rc.d/local From obrien at FreeBSD.org Sun Jun 22 17:53:04 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sun Jun 22 17:53:06 2008 Subject: cvs commit: src/usr.sbin/powerd powerd.8 powerd.c Message-ID: <200806221753.m5MHr4rL083811@repoman.freebsd.org> obrien 2008-06-22 17:52:57 UTC FreeBSD src repository Modified files: usr.sbin/powerd powerd.8 powerd.c Log: SVN rev 179932 on 2008-06-22 17:52:57Z by obrien Add an abbreviation for adaptive mode, and document all the abreviations. Revision Changes Path 1.11 +1 -0 src/usr.sbin/powerd/powerd.8 1.23 +1 -1 src/usr.sbin/powerd/powerd.c From gonzo at FreeBSD.org Sun Jun 22 20:53:37 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Sun Jun 22 20:53:43 2008 Subject: cvs commit: src/sys/pci viapm.c Message-ID: <200806222053.m5MKrb0K013832@repoman.freebsd.org> gonzo 2008-06-22 20:53:29 UTC FreeBSD src repository Modified files: sys/pci viapm.c Log: SVN rev 179933 on 2008-06-22 20:53:29Z by gonzo Add support for VT8237 ISA bridge. PR: kern/120714 Event: Bugathon#5 Revision Changes Path 1.20 +7 -0 src/sys/pci/viapm.c From rwatson at FreeBSD.org Sun Jun 22 21:03:38 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Jun 22 21:03:44 2008 Subject: cvs commit: src/tools/regression/fifo/fifo_create fifo_create.c Message-ID: <200806222103.m5ML3cPw017053@repoman.freebsd.org> rwatson 2008-06-22 21:03:26 UTC FreeBSD src repository Modified files: tools/regression/fifo/fifo_create fifo_create.c Log: SVN rev 179934 on 2008-06-22 21:03:26Z by rwatson Teach fifo_create regression test to also try to use mknod(2) to create fifos, as this is required by the Single UNIX Specification, although not currently implemented on FreeBSD. While here, fix a bug in the directory timestamp checking test by sleeping after querying the starting timestamp, rather than before. Revision Changes Path 1.2 +84 -36 src/tools/regression/fifo/fifo_create/fifo_create.c From mav at FreeBSD.org Sun Jun 22 21:22:40 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 22 21:22:41 2008 Subject: cvs commit: src/sbin/natd natd.c Message-ID: <200806222122.m5MLMdN5018616@repoman.freebsd.org> mav 2008-06-22 21:22:25 UTC FreeBSD src repository Modified files: sbin/natd natd.c Log: SVN rev 179935 on 2008-06-22 21:22:25Z by mav Use strdup() instead of static buffer allocation to avoid 128 bytes limit on -redirect_XXX arguments length. PR: bin/86647 Submitted by: Stephen Hurd Revision Changes Path 1.52 +18 -6 src/sbin/natd/natd.c From rwatson at FreeBSD.org Sun Jun 22 21:51:41 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Sun Jun 22 21:51:42 2008 Subject: cvs commit: src/sys/kern vfs_syscalls.c Message-ID: <200806222151.m5MLpeZ2021214@repoman.freebsd.org> rwatson 2008-06-22 21:51:32 UTC FreeBSD src repository Modified files: sys/kern vfs_syscalls.c Log: SVN rev 179936 on 2008-06-22 21:51:32Z by rwatson If S_IFIFO is passed to mknod(2), invoke kern_mkfifoat(9) to create a FIFO, as required by SUSv3. No specific privilege check is performed in this case, as FIFOs may be created by unprivileged processes (subject to the normal file system name space restrictions that may be in place). Unlike the Apple implementation, we reject requests to create a FIFO using mknod(2) if there is a non-zero dev argument to the system call, which is permitted by the Open Group specification ("... undefined ..."). We might want to revise this if we find it causes compatibility problems for applications in practice. PR: kern/74242, kern/68459 Obtained from: Apple, Inc. MFC after: 3 weeks Revision Changes Path 1.454 +4 -0 src/sys/kern/vfs_syscalls.c From mav at FreeBSD.org Sun Jun 22 22:14:12 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Sun Jun 22 22:14:13 2008 Subject: cvs commit: src/sbin/natd natd.8 natd.c natd.h Message-ID: <200806222214.m5MMEBkF024405@repoman.freebsd.org> mav 2008-06-22 22:14:02 UTC FreeBSD src repository Modified files: sbin/natd natd.8 natd.c natd.h Log: SVN rev 179937 on 2008-06-22 22:14:02Z by mav Add exit_delay parameter to control daemon exit delay after signal. PR: bin/58696 Submitted by: sp@alkor.ru Revision Changes Path 1.69 +6 -1 src/sbin/natd/natd.8 1.53 +22 -3 src/sbin/natd/natd.c 1.6 +3 -0 src/sbin/natd/natd.h From ru at freebsd.org Sun Jun 22 23:54:46 2008 From: ru at freebsd.org (Ruslan Ermilov) Date: Sun Jun 22 23:54:56 2008 Subject: cvs commit: src/include Makefile spawn.h unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c In-Reply-To: <200806170633.m5H6XMJH084600@repoman.freebsd.org> References: <200806170633.m5H6XMJH084600@repoman.freebsd.org> Message-ID: <20080622233801.GA43637@team.vega.ru> On Tue, Jun 17, 2008 at 06:26:29AM +0000, David Xu wrote: > davidxu 2008-06-17 06:26:29 UTC > > FreeBSD src repository > > Modified files: > include Makefile unistd.h > lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c > Added files: > include spawn.h > lib/libc/gen posix_spawn.c > Log: > SVN rev 179838 on 2008-06-17 06:26:29Z by davidxu > > Add POSIX routines called posix_spawn() and posix_spawnp(), which > can be used as replacements for exec/fork in a lot of cases. This > change also added execvpe() which allows environment variable > PATH to be used for searching executable file, it is used for > implementing posix_spawnp(). > > PR: standards/122051 > As you should already know, adding execvpe() broke many ports. Amongst them is jdk-1.6. Please fix. Cheers, -- Ruslan Ermilov ru@FreeBSD.org FreeBSD committer From thompsa at FreeBSD.org Mon Jun 23 00:51:44 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Mon Jun 23 00:51:45 2008 Subject: cvs commit: src/sys/modules/ath_rate_amrr Makefile Message-ID: <200806230051.m5N0phZK050833@repoman.freebsd.org> thompsa 2008-06-23 00:51:34 UTC FreeBSD src repository Modified files: sys/modules/ath_rate_amrr Makefile Log: SVN rev 179938 on 2008-06-23 00:51:34Z by thompsa Always create opt_ah.h regardless of KERNBUILDDIR. Revision Changes Path 1.7 +1 -1 src/sys/modules/ath_rate_amrr/Makefile From mtm at FreeBSD.org Mon Jun 23 03:49:47 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 03:49:53 2008 Subject: cvs commit: src/etc/rc.d localpkg Message-ID: <200806230349.m5N3nlsv069531@repoman.freebsd.org> mtm 2008-06-23 03:49:30 UTC FreeBSD src repository Modified files: etc/rc.d localpkg Log: SVN rev 179939 on 2008-06-23 03:49:30Z by mtm s/daemon processes/local packages/ for consisitency. Revision Changes Path 1.10 +1 -1 src/etc/rc.d/localpkg From mtm at FreeBSD.org Mon Jun 23 04:01:04 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 04:01:06 2008 Subject: cvs commit: src/etc/rc.d netoptions routing Message-ID: <200806230401.m5N413dC070424@repoman.freebsd.org> mtm 2008-06-23 04:00:45 UTC FreeBSD src repository Modified files: etc/rc.d netoptions routing Log: SVN rev 179940 on 2008-06-23 04:00:45Z by mtm Do not print anything unless one of the net/routing options is set. Revision Changes Path 1.145 +17 -2 src/etc/rc.d/netoptions 1.146 +18 -2 src/etc/rc.d/routing From mtm at FreeBSD.org Mon Jun 23 04:05:50 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 04:05:52 2008 Subject: cvs commit: src/etc/rc.d lockd statd Message-ID: <200806230405.m5N45nsO071936@repoman.freebsd.org> mtm 2008-06-23 04:05:39 UTC FreeBSD src repository Modified files: etc/rc.d lockd statd Log: SVN rev 179941 on 2008-06-23 04:05:39Z by mtm Argh! s/nfs_client_enable/nfsclient_enable/g Revision Changes Path 1.20 +2 -2 src/etc/rc.d/lockd 1.19 +2 -2 src/etc/rc.d/statd From mtm at FreeBSD.org Mon Jun 23 04:18:37 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 04:18:42 2008 Subject: cvs commit: src/etc rc Message-ID: <200806230418.m5N4IasY073061@repoman.freebsd.org> mtm 2008-06-23 04:18:22 UTC FreeBSD src repository Modified files: etc rc Log: SVN rev 179942 on 2008-06-23 04:18:22Z by mtm Remove pointless informational message. Revision Changes Path 1.342 +0 -1 src/etc/rc From mtm at FreeBSD.org Mon Jun 23 04:39:52 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 04:39:55 2008 Subject: cvs commit: src/etc/rc.d dumpon Message-ID: <200806230439.m5N4dqcB074741@repoman.freebsd.org> mtm 2008-06-23 04:39:36 UTC FreeBSD src repository Modified files: etc/rc.d dumpon Log: SVN rev 179943 on 2008-06-23 04:39:36Z by mtm Remove the -v flag from the command line to dumpon(8), and instead print diagnostic ouput only if the command fails. Revision Changes Path 1.12 +2 -1 src/etc/rc.d/dumpon From mtm at FreeBSD.org Mon Jun 23 04:43:08 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 04:43:18 2008 Subject: cvs commit: src/etc/rc.d cleartmp Message-ID: <200806230443.m5N4h8uE075058@repoman.freebsd.org> mtm 2008-06-23 04:42:58 UTC FreeBSD src repository Modified files: etc/rc.d cleartmp Log: SVN rev 179944 on 2008-06-23 04:42:58Z by mtm Align the script more with rc.d/cleanvar (which doesn't output any diagnostics). Instead, move output behind $rc_quiet. Revision Changes Path 1.19 +2 -2 src/etc/rc.d/cleartmp From mtm at FreeBSD.org Mon Jun 23 04:47:10 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 04:47:15 2008 Subject: cvs commit: src/etc/rc.d bgfsck fsck hostid hostname ldconfig motd moused newsyslog savecore Message-ID: <200806230447.m5N4lA0p079966@repoman.freebsd.org> mtm 2008-06-23 04:46:54 UTC FreeBSD src repository Modified files: etc/rc.d bgfsck fsck hostid hostname ldconfig motd moused newsyslog savecore Log: SVN rev 179945 on 2008-06-23 04:46:54Z by mtm Move a lot of diagnostic output behind $rc_quiet in scripts that implement their own start command. Revision Changes Path 1.9 +1 -1 src/etc/rc.d/bgfsck 1.13 +1 -1 src/etc/rc.d/fsck 1.10 +4 -2 src/etc/rc.d/hostid 1.13 +1 -1 src/etc/rc.d/hostname 1.20 +5 -3 src/etc/rc.d/ldconfig 1.11 +3 -3 src/etc/rc.d/motd 1.13 +1 -3 src/etc/rc.d/moused 1.7 +2 -2 src/etc/rc.d/newsyslog 1.12 +1 -1 src/etc/rc.d/savecore From mtm at FreeBSD.org Mon Jun 23 05:09:14 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 05:09:16 2008 Subject: cvs commit: src/etc rc.subr Message-ID: <200806230509.m5N59DcR088256@repoman.freebsd.org> mtm 2008-06-23 05:09:10 UTC FreeBSD src repository Modified files: etc rc.subr Log: SVN rev 179946 on 2008-06-23 05:09:10Z by mtm Move the diagnostic output when the rc.subr(8) glue automatically starts a service behind $rc_quiet. Instead, output a warning if the pre-command routine or the command itself failed. Arguably, it's more useful to know when a command failed to start than it is to have an endless list of "Starting ...." lines[1]. [1] - This change actually helped me to discover a bug in rc.d/{lockd,statd} (fixed in r179941) that used to fail silently before. Revision Changes Path 1.86 +9 -3 src/etc/rc.subr From ed at FreeBSD.org Mon Jun 23 05:26:26 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Mon Jun 23 05:26:32 2008 Subject: cvs commit: src/include unistd.h src/lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c src/lib/libc/include libc_private.h Message-ID: <200806230526.m5N5QQdS091438@repoman.freebsd.org> ed 2008-06-23 05:22:06 UTC FreeBSD src repository Modified files: include unistd.h lib/libc/gen Makefile.inc Symbol.map exec.3 exec.c posix_spawn.c lib/libc/include libc_private.h Log: SVN rev 179947 on 2008-06-23 05:22:06Z by ed Turn execvpe() into an internal libc routine. Adding exevpe() has caused some ports to break. Even though execvpe() is a useful routine, it does not conform to any standards. This patch is a little bit different from the patch sent to the mailing list. I forgot to remove execvpe from the Symbol.map (which does not seem to miscompile libc, though). Reviewed by: davidxu Approved by: philip Revision Changes Path 1.89 +0 -1 src/include/unistd.h 1.137 +1 -1 src/lib/libc/gen/Makefile.inc 1.12 +0 -1 src/lib/libc/gen/Symbol.map 1.28 +3 -14 src/lib/libc/gen/exec.3 1.25 +3 -2 src/lib/libc/gen/exec.c 1.5 +2 -1 src/lib/libc/gen/posix_spawn.c 1.19 +3 -0 src/lib/libc/include/libc_private.h From mtm at wubethiopia.com Mon Jun 23 05:43:52 2008 From: mtm at wubethiopia.com (Mike Makonnen) Date: Mon Jun 23 05:43:59 2008 Subject: svn commit: r179946 - head/etc In-Reply-To: <200806230509.m5N599qK060278@svn.freebsd.org> References: <200806230509.m5N599qK060278@svn.freebsd.org> Message-ID: <485F34D0.1080002@wubethiopia.com> Mike Makonnen wrote: > Author: mtm > Date: Mon Jun 23 05:09:09 2008 > New Revision: 179946 > URL: http://svn.freebsd.org/changeset/base/179946 > > Log: > Move the diagnostic output when the rc.subr(8) glue automatically starts a > service behind $rc_quiet. Instead, output a warning if the pre-command > routine or the command itself failed. Arguably, it's more useful to know when > a command failed to start than it is to have an endless list of > "Starting ...." lines[1]. > > [1] - This change actually helped me to discover a bug in rc.d/{lockd,statd} > (fixed in r179941) that used to fail silently before. > > Modified: > head/etc/rc.subr > This ends the bulk of the commits to reduce console-spamming by rc.d during bootup. It should greatly improve the signal-to-noise ratio on your console :-). I've been running these changes for a while without problems, but if I've missed something or you think a script should be more (or less) verbose, please let me know. Cheers. -- Mike Makonnen | GPG-KEY: http://people.freebsd.org/~mtm/mtm.asc mtm @ FreeBSD.Org | AC7B 5672 2D11 F4D0 EBF8 5279 5359 2B82 7CD4 1F55 FreeBSD | http://www.freebsd.org From danger at FreeBSD.org Mon Jun 23 08:02:11 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 23 08:02:13 2008 Subject: cvs commit: src/share/man/man4 ciss.4 cpufreq.4 Message-ID: <200806230802.m5N82AVO008284@repoman.freebsd.org> danger 2008-06-23 08:01:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 ciss.4 cpufreq.4 Log: SVN rev 179948 on 2008-06-23 08:01:53Z by danger MFC rev. 179903: link powerd(8) man page PR: docs/123111 Revision Changes Path 1.14.2.4 +0 -0 src/share/man/man4/ciss.4 1.15.2.1 +1 -0 src/share/man/man4/cpufreq.4 From danger at FreeBSD.org Mon Jun 23 08:07:12 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 23 08:07:14 2008 Subject: cvs commit: src/share/man/man4 ciss.4 cpufreq.4 Message-ID: <200806230807.m5N87Bi8009820@repoman.freebsd.org> danger 2008-06-23 08:06:55 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 ciss.4 cpufreq.4 Log: SVN rev 179949 on 2008-06-23 08:06:55Z by danger - MFC rev 179903: link powerd(8) man page PR: docs/123111 Revision Changes Path 1.11.2.6 +0 -0 src/share/man/man4/ciss.4 1.10.2.5 +1 -0 src/share/man/man4/cpufreq.4 From kib at FreeBSD.org Mon Jun 23 11:03:48 2008 From: kib at FreeBSD.org (Konstantin Belousov) Date: Mon Jun 23 11:03:55 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/kern vfs_export.c vfs_mount.c src/sys/pci nfsmb.c src/sys/sys mount.h Message-ID: <200806231103.m5NB3mO2037400@repoman.freebsd.org> kib 2008-06-23 11:02:40 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/ciss ciss.c sys/kern vfs_export.c vfs_mount.c sys/pci nfsmb.c sys/sys mount.h Log: SVN rev 179950 on 2008-06-23 11:02:40Z by kib MFC r179670: Provide the mutual exclusion between the nfs export list modifications and nfs requests processing. Revision Changes Path 1.81.2.6 +0 -0 src/sys/dev/ciss/ciss.c 1.341.2.1 +4 -0 src/sys/kern/vfs_export.c 1.265.2.4 +2 -0 src/sys/kern/vfs_mount.c 1.6.2.4 +0 -0 src/sys/pci/nfsmb.c 1.228.2.2 +1 -0 src/sys/sys/mount.h From mtm at FreeBSD.org Mon Jun 23 12:06:50 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 12:06:52 2008 Subject: cvs commit: src/etc/rc.d netoptions Message-ID: <200806231206.m5NC6nkv044015@repoman.freebsd.org> mtm 2008-06-23 12:06:35 UTC FreeBSD src repository Modified files: etc/rc.d netoptions Log: SVN rev 179951 on 2008-06-23 12:06:35Z by mtm Set the sysctl(8) value in the same shell, not a subshell. This was causing calls to netoptions_init() to not properly set a global variable, which ended up being in the parent shell. Revision Changes Path 1.146 +2 -2 src/etc/rc.d/netoptions From mav at FreeBSD.org Mon Jun 23 14:16:06 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Mon Jun 23 14:16:15 2008 Subject: cvs commit: src/sys/netinet ip_fw_nat.c Message-ID: <200806231416.m5NEG6lQ065890@repoman.freebsd.org> mav 2008-06-23 14:15:53 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet ip_fw_nat.c Log: SVN rev 179952 on 2008-06-23 14:15:53Z by mav MFC r179473: PKT_ALIAS_FOUND_HEADER_FRAGMENT result is not an error, so pass that packet. This fixes packet fragmentation handeling. Pass really available buffer size to libalias instead of MCLBYTES constant. MCLBYTES constant were used with believe that m_megapullup() always moves date into a fresh cluster that sometimes may become not so. Revision Changes Path 1.2.2.2 +4 -3 src/sys/netinet/ip_fw_nat.c From mav at FreeBSD.org Mon Jun 23 14:19:53 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Mon Jun 23 14:19:55 2008 Subject: cvs commit: src/sys/netgraph ng_nat.c Message-ID: <200806231419.m5NEJqOv066285@repoman.freebsd.org> mav 2008-06-23 14:19:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_nat.c Log: SVN rev 179953 on 2008-06-23 14:19:48Z by mav MFC r179477: Pass really available buffer size to libalias instead of MCLBYTES constant. MCLBYTES constant were used with believe that m_megapullup() always moves date into a fresh cluster that may become not so. Revision Changes Path 1.10.2.2 +2 -2 src/sys/netgraph/ng_nat.c From mav at FreeBSD.org Mon Jun 23 14:28:40 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Mon Jun 23 14:28:42 2008 Subject: cvs commit: src/sys/netinet/libalias alias_db.c alias_local.h Message-ID: <200806231428.m5NEScqn067014@repoman.freebsd.org> mav 2008-06-23 14:28:32 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet/libalias alias_db.c alias_local.h Log: SVN rev 179954 on 2008-06-23 14:28:32Z by mav MFC r179480, r179924: Increase LINK_TABLE_OUT_SIZE from 101 to 4001 like LINK_TABLE_IN_SIZE to reduce performance degradation under heavy outgoing scan/flood. Scalability is now much more important then several kilobytes of RAM. Remove unneded TCP-specific expiration handeling. Before this connected TCP sessions could never expire. Now connected TCP sessions will expire after 24hours of inactivity. Simplify HouseKeeping() to avoid several mul/div-s per packet. Taking into account increased LINK_TABLE_OUT_SIZE, precision is still much more then required. Revision Changes Path 1.71.2.1 +22 -63 src/sys/netinet/libalias/alias_db.c 1.34.2.1 +1 -3 src/sys/netinet/libalias/alias_local.h From mav at FreeBSD.org Mon Jun 23 14:38:39 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Mon Jun 23 14:38:41 2008 Subject: cvs commit: src/sys/netinet/libalias alias.c Message-ID: <200806231438.m5NEcdOH067876@repoman.freebsd.org> mav 2008-06-23 14:38:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netinet/libalias alias.c Log: SVN rev 179955 on 2008-06-23 14:38:33Z by mav MFC r179478: Make m_megapullup() more intelligent: - to increase performance do not reallocate mbuf when possible, - to support up to 16K packets (was 2K max) use mbuf cluster of proper size. This change depends on recent ng_nat and ip_fw_nat changes. Revision Changes Path 1.58.2.1 +32 -12 src/sys/netinet/libalias/alias.c From alc at FreeBSD.org Mon Jun 23 15:23:11 2008 From: alc at FreeBSD.org (Alan Cox) Date: Mon Jun 23 15:23:20 2008 Subject: cvs commit: src/sys/amd64/include vmparam.h Message-ID: <200806231523.m5NFNAIO073960@repoman.freebsd.org> alc 2008-06-23 15:22:53 UTC FreeBSD src repository Modified files: sys/amd64/include vmparam.h Log: SVN rev 179956 on 2008-06-23 15:22:53Z by alc Ensure that KERNBASE is no less than the virtual address -2GB. Revision Changes Path 1.51 +1 -1 src/sys/amd64/include/vmparam.h From thompsa at FreeBSD.org Mon Jun 23 15:41:12 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Mon Jun 23 15:41:18 2008 Subject: cvs commit: src/sys/dev/wpi if_wpi.c Message-ID: <200806231541.m5NFfBZd075493@repoman.freebsd.org> thompsa 2008-06-23 15:40:56 UTC FreeBSD src repository Modified files: sys/dev/wpi if_wpi.c Log: SVN rev 179957 on 2008-06-23 15:40:56Z by thompsa - Fix compile if WPI_DEBUG is not defined - Allow debug.wpi to be set from a tunable - Put ring reset messages back under debug - Add more debug output around channel init Revision Changes Path 1.16 +39 -27 src/sys/dev/wpi/if_wpi.c From thompsa at FreeBSD.org Mon Jun 23 16:08:59 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Mon Jun 23 16:09:08 2008 Subject: cvs commit: src/sbin/ifconfig ifieee80211.c Message-ID: <200806231608.m5NG8wYj078771@repoman.freebsd.org> thompsa 2008-06-23 16:08:40 UTC FreeBSD src repository Modified files: sbin/ifconfig ifieee80211.c Log: SVN rev 179958 on 2008-06-23 16:08:40Z by thompsa Ensure the channel is a number and not a range, the 'channel' command can be easily mixed up with 'chanlist' and would give unexpected results by fixing the channel on the first number in the range. Revision Changes Path 1.52 +4 -1 src/sbin/ifconfig/ifieee80211.c From oberman at es.net Mon Jun 23 16:13:06 2008 From: oberman at es.net (Kevin Oberman) Date: Mon Jun 23 16:13:11 2008 Subject: svn commit: r179946 - head/etc In-Reply-To: Your message of "Mon, 23 Jun 2008 08:29:52 +0300." <485F34D0.1080002@wubethiopia.com> Message-ID: <20080623160255.CB6B645010@ptavv.es.net> > Date: Mon, 23 Jun 2008 08:29:52 +0300 > From: Mike Makonnen > Sender: owner-cvs-all@freebsd.org > > Mike Makonnen wrote: > > Author: mtm > > Date: Mon Jun 23 05:09:09 2008 > > New Revision: 179946 > > URL: http://svn.freebsd.org/changeset/base/179946 > > > > Log: > > Move the diagnostic output when the rc.subr(8) glue automatically starts a > > service behind $rc_quiet. Instead, output a warning if the pre-command > > routine or the command itself failed. Arguably, it's more useful to know when > > a command failed to start than it is to have an endless list of > > "Starting ...." lines[1]. > > > > [1] - This change actually helped me to discover a bug in rc.d/{lockd,statd} > > (fixed in r179941) that used to fail silently before. > > > > Modified: > > head/etc/rc.subr > > > > This ends the bulk of the commits to reduce console-spamming by rc.d > during bootup. It should greatly improve the signal-to-noise ratio on > your console :-). I've been running these changes for a while without > problems, but if I've missed something or you think a script should be > more (or less) verbose, please let me know. Thanks, Mike! This will help a lot. If I missed it, I apologize, but are there plans to MFC these changes to RELENG_7 or RELENG6? -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman@es.net Phone: +1 510 486-8634 Key fingerprint:059B 2DDF 031C 9BA3 14A4 EADA 927D EBB3 987B 3751 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 224 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080623/5fd8d629/attachment.pgp From jhb at FreeBSD.org Mon Jun 23 18:16:45 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Jun 23 18:16:47 2008 Subject: cvs commit: src/sys/dev/fe if_fe.c if_fe_cbus.c if_fe_isa.c if_fe_pccard.c if_fevar.h Message-ID: <200806231816.m5NIGjsg001328@repoman.freebsd.org> jhb 2008-06-23 18:16:25 UTC FreeBSD src repository Modified files: sys/dev/fe if_fe.c if_fe_cbus.c if_fe_isa.c if_fe_pccard.c if_fevar.h Log: SVN rev 179959 on 2008-06-23 18:16:25Z by jhb - Use bus_foo() rather than bus_space_foo() and retire the bus tag and handle from the softc. - Rework the watchdog timer to match other NIC drivers: - Start a timer in fe_init() that runs once a second and checks a counter in the softc that is identical to the deprecated 'if_timer'. - Just adjust the softc tx timeout value when sending packets instead of scheduling the timer. - Use IFQ_SET_MAXLEN(). Tested by: WATANABE Kazuhiro Revision Changes Path 1.100 +15 -27 src/sys/dev/fe/if_fe.c 1.14 +0 -3 src/sys/dev/fe/if_fe_cbus.c 1.10 +1 -1 src/sys/dev/fe/if_fe_isa.c 1.34 +0 -1 src/sys/dev/fe/if_fe_pccard.c 1.8 +11 -12 src/sys/dev/fe/if_fevar.h From phk at FreeBSD.org Mon Jun 23 20:27:49 2008 From: phk at FreeBSD.org (Poul-Henning Kamp) Date: Mon Jun 23 20:27:54 2008 Subject: cvs commit: src/games/fortune/datfiles fortunes Message-ID: <200806232027.m5NKRmdn013926@repoman.freebsd.org> phk 2008-06-23 20:27:37 UTC FreeBSD src repository Modified files: games/fortune/datfiles fortunes Log: SVN rev 179960 on 2008-06-23 20:27:37Z by phk +Computers can't cruise. Meandering is a foreign concept to them. +The computer assumes that all behavior is in pursuit of an ultimate +goal. Whenever a motorist changes his or her mind and veers off +course, the GPS lady issues that snippy announcement: "Recalculating!" + -- Joel Achenbach (www.slate.com, 20 jun 2008) Revision Changes Path 1.264 +6 -0 src/games/fortune/datfiles/fortunes From mtm at FreeBSD.org Mon Jun 23 20:50:20 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 20:50:27 2008 Subject: cvs commit: src/etc network.subr pccard_ether src/etc/rc.d netif Message-ID: <200806232050.m5NKoKtI025106@repoman.freebsd.org> mtm 2008-06-23 20:50:11 UTC FreeBSD src repository Modified files: etc network.subr pccard_ether etc/rc.d netif Log: SVN rev 179961 on 2008-06-23 20:50:11Z by mtm Implement a "quiet" mode for rc.d/netif, which only outputs the interface name of interfaces that were configured. This change has the added benefit that ifn_start() and ifn_stop() in network.subr no longer write to standard output. Whether to output and what to output is now handled entirely in rc.d/netif. Revision Changes Path 1.190 +0 -8 src/etc/network.subr 1.55 +2 -2 src/etc/pccard_ether 1.29 +21 -5 src/etc/rc.d/netif From mtm at FreeBSD.org Mon Jun 23 20:54:37 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 20:54:42 2008 Subject: cvs commit: src/etc/rc.d savecore Message-ID: <200806232054.m5NKsaRS025388@repoman.freebsd.org> mtm 2008-06-23 20:54:32 UTC FreeBSD src repository Modified files: etc/rc.d savecore Log: SVN rev 179962 on 2008-06-23 20:54:32Z by mtm Run savecore(8) only if there is a core dump to save. If there is no core dump hide the message to that effect behind $rc_quiet. Revision Changes Path 1.13 +5 -2 src/etc/rc.d/savecore From mtm at wubethiopia.com Mon Jun 23 21:14:58 2008 From: mtm at wubethiopia.com (Mike Makonnen) Date: Mon Jun 23 21:15:07 2008 Subject: svn commit: r179946 - head/etc In-Reply-To: <20080623160255.CB6B645010@ptavv.es.net> References: <20080623160255.CB6B645010@ptavv.es.net> Message-ID: <4860139A.6060309@wubethiopia.com> Kevin Oberman wrote: >> This ends the bulk of the commits to reduce console-spamming by rc.d >> during bootup. It should greatly improve the signal-to-noise ratio on >> your console :-). I've been running these changes for a while without >> problems, but if I've missed something or you think a script should be >> more (or less) verbose, please let me know. >> > > Thanks, Mike! This will help a lot. > > If I missed it, I apologize, but are there plans to MFC these changes to > RELENG_7 or RELENG6? > Yes, I plan to MFC it unless there are strong objections. Although I haven't committed it yet, there will be a knob to enable the previous verbose mode. Obviously it will default to OFF in -current, but I'm thinking of defaulting it to ON on RELENG_{6,7} for POLA reasons. Cheers. -- Mike Makonnen | GPG-KEY: http://people.freebsd.org/~mtm/mtm.asc mtm @ FreeBSD.Org | AC7B 5672 2D11 F4D0 EBF8 5279 5359 2B82 7CD4 1F55 FreeBSD | http://www.freebsd.org From jhb at FreeBSD.org Mon Jun 23 21:39:07 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Mon Jun 23 21:39:12 2008 Subject: cvs commit: src/sys/kern uipc_sem.c src/sys/security/mac mac_framework.h mac_policy.h mac_posix_sem.c src/sys/security/mac_biba mac_biba.c src/sys/security/mac_mls mac_mls.c src/sys/security/mac_stub mac_stub.c ... Message-ID: <200806232139.m5NLd6Fo030280@repoman.freebsd.org> jhb 2008-06-23 21:37:53 UTC FreeBSD src repository Modified files: sys/kern uipc_sem.c sys/security/mac mac_framework.h mac_policy.h mac_posix_sem.c sys/security/mac_biba mac_biba.c sys/security/mac_mls mac_mls.c sys/security/mac_stub mac_stub.c sys/security/mac_test mac_test.c Log: SVN rev 179963 on 2008-06-23 21:37:53Z by jhb Remove the posixsem_check_destroy() MAC check. It is semantically identical to doing a MAC check for close(), but no other types of close() (including close(2) and ksem_close(2)) have MAC checks. Discussed with: rwatson Revision Changes Path 1.33 +0 -5 src/sys/kern/uipc_sem.c 1.99 +0 -1 src/sys/security/mac/mac_framework.h 1.109 +0 -3 src/sys/security/mac/mac_policy.h 1.13 +0 -10 src/sys/security/mac/mac_posix_sem.c 1.119 +0 -1 src/sys/security/mac_biba/mac_biba.c 1.101 +0 -1 src/sys/security/mac_mls/mac_mls.c 1.82 +0 -9 src/sys/security/mac_stub/mac_stub.c 1.96 +0 -14 src/sys/security/mac_test/mac_test.c From marcel at FreeBSD.org Mon Jun 23 21:45:28 2008 From: marcel at FreeBSD.org (Marcel Moolenaar) Date: Mon Jun 23 21:46:26 2008 Subject: cvs commit: src/sys/powerpc/include atomic.h Message-ID: <200806232145.m5NLjLqa030998@repoman.freebsd.org> marcel 2008-06-23 21:45:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/powerpc/include atomic.h Log: SVN rev 179964 on 2008-06-23 21:45:04Z by marcel MFC rev. 173742-178057. Requested by: mav (twice) Revision Changes Path 1.15.2.1 +331 -264 src/sys/powerpc/include/atomic.h From mtm at FreeBSD.org Mon Jun 23 22:11:57 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 22:12:03 2008 Subject: cvs commit: src/etc/rc.d auto_linklocal power_profile sysctl src/sbin/sysctl sysctl.c Message-ID: <200806232211.m5NMBuK9034230@repoman.freebsd.org> mtm 2008-06-23 22:06:28 UTC FreeBSD src repository Modified files: etc/rc.d auto_linklocal power_profile sysctl sbin/sysctl sysctl.c Log: SVN rev 179965 on 2008-06-23 22:06:28Z by mtm The sysctl(8) program exits on some errors and only emits warnings on others. In the case where it displayed warnings it would still return succesfully. Modify it so that it returns the number of sysctls that it was not able to set. Make use of this in rc.d to display only *unsuccessfull* attempts to set sysctls. Revision Changes Path 1.6 +4 -1 src/etc/rc.d/auto_linklocal 1.13 +5 -1 src/etc/rc.d/power_profile 1.19 +3 -1 src/etc/rc.d/sysctl 1.89 +5 -2 src/sbin/sysctl/sysctl.c From mtm at FreeBSD.org Mon Jun 23 22:20:06 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 23 22:20:08 2008 Subject: cvs commit: src/etc/rc.d swap1 src/sbin/swapon swapon.8 swapon.c Message-ID: <200806232220.m5NMK5OK034866@repoman.freebsd.org> mtm 2008-06-23 22:17:08 UTC FreeBSD src repository Modified files: etc/rc.d swap1 sbin/swapon swapon.8 swapon.c Log: SVN rev 179966 on 2008-06-23 22:17:08Z by mtm Add a -q flag to swapon(8) to suppress informational messages. Use it in rc.d. Note: errors are not affected by this flag. Revision Changes Path 1.11 +1 -1 src/etc/rc.d/swap1 1.31 +11 -3 src/sbin/swapon/swapon.8 1.24 +13 -6 src/sbin/swapon/swapon.c From oberman at es.net Mon Jun 23 22:35:02 2008 From: oberman at es.net (Kevin Oberman) Date: Mon Jun 23 22:35:06 2008 Subject: svn commit: r179946 - head/etc In-Reply-To: Your message of "Tue, 24 Jun 2008 00:20:26 +0300." <4860139A.6060309@wubethiopia.com> Message-ID: <20080623223501.5BDD14500E@ptavv.es.net> > Date: Tue, 24 Jun 2008 00:20:26 +0300 > From: Mike Makonnen > > Kevin Oberman wrote: > >> This ends the bulk of the commits to reduce console-spamming by rc.d > >> during bootup. It should greatly improve the signal-to-noise ratio on > >> your console :-). I've been running these changes for a while without > >> problems, but if I've missed something or you think a script should be > >> more (or less) verbose, please let me know. > >> > > > > Thanks, Mike! This will help a lot. > > > > If I missed it, I apologize, but are there plans to MFC these changes to > > RELENG_7 or RELENG6? > > > > Yes, I plan to MFC it unless there are strong objections. Although I haven't > committed it yet, there will be a knob to enable the previous verbose mode. > Obviously it will default to OFF in -current, but I'm thinking of defaulting > it to ON on RELENG_{6,7} for POLA reasons. Excellent on all counts. I agree with the POLA issue and defaulting to the current behavior although I suspect most everyone will opt for the quiet output in short order and the current rc stuff often is just noise to most users and not really useful, even for those who understand most of it. I consider this and the recent work on the FAQ bu Gabor Pali to be real wins for OS friendliness and approachability for those without BSD experience. -- R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman@es.net Phone: +1 510 486-8634 Key fingerprint:059B 2DDF 031C 9BA3 14A4 EADA 927D EBB3 987B 3751 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 224 bytes Desc: not available Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080623/415e1cba/attachment.pgp From jkim at FreeBSD.org Mon Jun 23 23:09:59 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Mon Jun 23 23:10:04 2008 Subject: cvs commit: src/sys/amd64/amd64 bpf_jit_machdep.c bpf_jit_machdep.h Message-ID: <200806232309.m5NN9wPW039882@repoman.freebsd.org> jkim 2008-06-23 23:09:52 UTC FreeBSD src repository Modified files: sys/amd64/amd64 bpf_jit_machdep.c bpf_jit_machdep.h Log: SVN rev 179967 on 2008-06-23 23:09:52Z by jkim Rehash and clean up BPF JIT compiler macros to match AT&T notations. Revision Changes Path 1.5 +88 -88 src/sys/amd64/amd64/bpf_jit_machdep.c 1.4 +64 -64 src/sys/amd64/amd64/bpf_jit_machdep.h From jkim at FreeBSD.org Mon Jun 23 23:10:15 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Mon Jun 23 23:10:17 2008 Subject: cvs commit: src/sys/i386/i386 bpf_jit_machdep.c bpf_jit_machdep.h Message-ID: <200806232310.m5NNAELY039977@repoman.freebsd.org> jkim 2008-06-23 23:10:11 UTC FreeBSD src repository Modified files: sys/i386/i386 bpf_jit_machdep.c bpf_jit_machdep.h Log: SVN rev 179968 on 2008-06-23 23:10:11Z by jkim Rehash and clean up BPF JIT compiler macros to match AT&T notations. Revision Changes Path 1.5 +86 -86 src/sys/i386/i386/bpf_jit_machdep.c 1.4 +57 -57 src/sys/i386/i386/bpf_jit_machdep.h From emaste at FreeBSD.org Tue Jun 24 03:26:49 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Tue Jun 24 03:26:51 2008 Subject: cvs commit: src/sys/dev/aac aac.c Message-ID: <200806240326.m5O3QmWG084105@repoman.freebsd.org> emaste 2008-06-24 03:26:41 UTC FreeBSD src repository Modified files: sys/dev/aac aac.c Log: SVN rev 179969 on 2008-06-24 03:26:41Z by emaste Fix test for waiting AIFs in aac_poll(). This seems to solve the problem where Adaptec's arcconf monitoring tool hangs after producing its expected output. Submitted by: Adaptec, via driver ver 15317 MFC after: 1 week Revision Changes Path 1.138 +7 -2 src/sys/dev/aac/aac.c From davidxu at FreeBSD.org Tue Jun 24 07:37:44 2008 From: davidxu at FreeBSD.org (David Xu) Date: Tue Jun 24 07:37:53 2008 Subject: cvs commit: src/lib/libthr/thread thr_mutex.c thr_umtx.c thr_umtx.h src/sys/kern kern_umtx.c src/sys/sys umtx.h Message-ID: <200806240737.m5O7bhc5018813@repoman.freebsd.org> davidxu 2008-06-24 07:32:12 UTC FreeBSD src repository Modified files: lib/libthr/thread thr_mutex.c thr_umtx.c thr_umtx.h sys/kern kern_umtx.c sys/sys umtx.h Log: SVN rev 179970 on 2008-06-24 07:32:12Z by davidxu Add two commands to _umtx_op system call to allow a simple mutex to be locked and unlocked completely in userland. by locking and unlocking mutex in userland, it reduces the total time a mutex is locked by a thread, in some application code, a mutex only protects a small piece of code, the code's execution time is less than a simple system call, if a lock contention happens, however in current implemenation, the lock holder has to extend its locking time and enter kernel to unlock it, the change avoids this disadvantage, it first sets mutex to free state and then enters kernel and wake one waiter up. This improves performance dramatically in some sysbench mutex tests. Tested by: kris Sounds great: jeff Revision Changes Path 1.75 +33 -35 src/lib/libthr/thread/thr_mutex.c 1.17 +60 -10 src/lib/libthr/thread/thr_umtx.c 1.14 +13 -9 src/lib/libthr/thread/thr_umtx.h 1.71 +142 -35 src/sys/kern/kern_umtx.c 1.33 +3 -1 src/sys/sys/umtx.h From ronald-freebsd8 at klop.yi.org Tue Jun 24 12:14:18 2008 From: ronald-freebsd8 at klop.yi.org (Ronald Klop) Date: Tue Jun 24 12:14:28 2008 Subject: cvs commit: src/etc/rc.d savecore In-Reply-To: <200806232054.m5NKsaRS025388@repoman.freebsd.org> References: <200806232054.m5NKsaRS025388@repoman.freebsd.org> Message-ID: On Mon, 23 Jun 2008 22:54:32 +0200, Mike Makonnen wrote: > mtm 2008-06-23 20:54:32 UTC > > FreeBSD src repository > > Modified files: > etc/rc.d savecore > Log: > SVN rev 179962 on 2008-06-23 20:54:32Z by mtm > Run savecore(8) only if there is a core dump to save. If there is > no core dump hide the message to that effect behind $rc_quiet. > Revision Changes Path > 1.13 +5 -2 src/etc/rc.d/savecore So, if I enable savecore in rc.conf I don't get any feedback from the system that I enabled it correctly? I'm not a kernel developer and have nothing to say about it, but this questions comes to my mind about userfriendlyness. Cheers, Ronald. -- Ronald Klop Amsterdam, The Netherlands From gonzo at FreeBSD.org Tue Jun 24 13:58:38 2008 From: gonzo at FreeBSD.org (Oleksandr Tymoshenko) Date: Tue Jun 24 13:58:48 2008 Subject: cvs commit: src/sys/netinet in.c Message-ID: <200806241358.m5ODwc4F075629@repoman.freebsd.org> gonzo 2008-06-24 13:58:28 UTC FreeBSD src repository Modified files: sys/netinet in.c Log: SVN rev 179971 on 2008-06-24 13:58:28Z by gonzo In case of interface initialization failure remove struct in_ifaddr* from in_ifaddrhashtbl in in_ifinit because error handler in in_control removes entries only for AF_INET addresses. If in_ifinit is called for the cloned inteface that has just been created its address family is not AF_INET and therefor LIST_REMOVE is not called for respective LIST_INSERT_HEAD and freed entries remain in in_ifaddrhashtbl and lead to memory corruption. PR: kern/124384 Revision Changes Path 1.104 +8 -0 src/sys/netinet/in.c From kientzle at FreeBSD.org Tue Jun 24 15:18:55 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Tue Jun 24 15:19:02 2008 Subject: cvs commit: src/usr.bin/cpio cpio.c Message-ID: <200806241518.m5OFItri088615@repoman.freebsd.org> kientzle 2008-06-24 15:18:40 UTC FreeBSD src repository Modified files: usr.bin/cpio cpio.c Log: SVN rev 179972 on 2008-06-24 15:18:40Z by kientzle In -p mode, don't gaurd against '..' in paths. We continue to check in -i mode unless --insecure is specified. PR: bin/124924 Revision Changes Path 1.4 +1 -0 src/usr.bin/cpio/cpio.c From gnn at FreeBSD.org Tue Jun 24 18:49:58 2008 From: gnn at FreeBSD.org (George V. Neville-Neil) Date: Tue Jun 24 18:50:04 2008 Subject: cvs commit: src/sys/netgraph netgraph.h Message-ID: <200806241849.m5OInvag020252@repoman.freebsd.org> gnn 2008-06-24 18:49:49 UTC FreeBSD src repository Modified files: sys/netgraph netgraph.h Log: SVN rev 179973 on 2008-06-24 18:49:49Z by gnn Make it simpler to build netgraph modules outside of the kernel source tree. This change follows similar ones in the device tree. MFC after: 2 weeks Revision Changes Path 1.74 +2 -0 src/sys/netgraph/netgraph.h From jhb at FreeBSD.org Tue Jun 24 19:40:03 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 24 19:40:12 2008 Subject: cvs commit: src/etc/rc.d initrandom src/gnu/usr.bin/groff/tmac mdoc.local src/share/man/man4 ciss.4 src/sys/dev/ciss ciss.c src/sys/kern sched_4bsd.c src/sys/pci nfsmb.c src/usr.bin/cksum cksum.c src/usr.bin/shar shar.sh src/usr.bin/whereis whereis.c Message-ID: <200806241940.m5OJe3Uu025839@repoman.freebsd.org> jhb 2008-06-24 19:36:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) etc/rc.d initrandom gnu/usr.bin/groff/tmac mdoc.local share/man/man4 ciss.4 sys/dev/ciss ciss.c sys/kern sched_4bsd.c sys/pci nfsmb.c usr.bin/cksum cksum.c usr.bin/shar shar.sh usr.bin/whereis whereis.c Log: SVN rev 179974 on 2008-06-24 19:36:33Z by jhb MFC: Change the roundrobin implementation in the 4BSD scheduler to trigger a userland preemption directly from hardclock() via sched_clock(). Revision Changes Path 1.7.2.2 +0 -0 src/etc/rc.d/initrandom 1.59.2.3 +0 -0 src/gnu/usr.bin/groff/tmac/mdoc.local 1.14.2.5 +0 -0 src/share/man/man4/ciss.4 1.81.2.7 +0 -0 src/sys/dev/ciss/ciss.c 1.106.2.2 +8 -29 src/sys/kern/sched_4bsd.c 1.6.2.5 +0 -0 src/sys/pci/nfsmb.c 1.17.22.2 +0 -0 src/usr.bin/cksum/cksum.c 1.3.48.3 +0 -0 src/usr.bin/shar/shar.sh 1.15.10.2 +0 -0 src/usr.bin/whereis/whereis.c From jhb at FreeBSD.org Tue Jun 24 19:56:02 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 24 19:56:05 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/kern sched_4bsd.c Message-ID: <200806241956.m5OJu2kM027215@repoman.freebsd.org> jhb 2008-06-24 19:55:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/ciss ciss.c sys/kern sched_4bsd.c Log: SVN rev 179975 on 2008-06-24 19:55:22Z by jhb MFC: Change the roundrobin implementation in the 4BSD scheduler to trigger a userland preemption directly from hardclock() via sched_clock(). Revision Changes Path 1.64.2.8 +0 -0 src/sys/dev/ciss/ciss.c 1.77.2.2 +8 -29 src/sys/kern/sched_4bsd.c From jhb at FreeBSD.org Tue Jun 24 20:05:27 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 24 20:05:28 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/i386/i386 machdep.c src/sys/kern kern_mib.c src/sys/pci nfsmb.c src/sys/sys sysctl.h Message-ID: <200806242005.m5OK5Qqx029267@repoman.freebsd.org> jhb 2008-06-24 20:04:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/ciss ciss.c sys/i386/i386 machdep.c sys/kern kern_mib.c sys/pci nfsmb.c sys/sys sysctl.h Log: SVN rev 179976 on 2008-06-24 20:04:39Z by jhb MFC: Add the kern.features sysctl tree and FEATURE() macro. Add a "pae" feature for PAE kernels on i386. Revision Changes Path 1.81.2.8 +0 -0 src/sys/dev/ciss/ciss.c 1.658.2.7 +4 -0 src/sys/i386/i386/machdep.c 1.84.2.3 +2 -0 src/sys/kern/kern_mib.c 1.6.2.6 +0 -0 src/sys/pci/nfsmb.c 1.148.2.3 +8 -0 src/sys/sys/sysctl.h From jkim at FreeBSD.org Tue Jun 24 20:12:21 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Tue Jun 24 20:12:23 2008 Subject: cvs commit: src/sys/amd64/amd64 bpf_jit_machdep.c bpf_jit_machdep.h Message-ID: <200806242012.m5OKCLAZ029869@repoman.freebsd.org> jkim 2008-06-24 20:12:12 UTC FreeBSD src repository Modified files: sys/amd64/amd64 bpf_jit_machdep.c bpf_jit_machdep.h Log: SVN rev 179977 on 2008-06-24 20:12:12Z by jkim Emit opcodes closer to GNU as(1) generated codes and micro-optimize. Revision Changes Path 1.6 +11 -11 src/sys/amd64/amd64/bpf_jit_machdep.c 1.5 +38 -52 src/sys/amd64/amd64/bpf_jit_machdep.h From jkim at FreeBSD.org Tue Jun 24 20:12:52 2008 From: jkim at FreeBSD.org (Jung-uk Kim) Date: Tue Jun 24 20:12:54 2008 Subject: cvs commit: src/sys/i386/i386 bpf_jit_machdep.c bpf_jit_machdep.h Message-ID: <200806242012.m5OKCqD6029932@repoman.freebsd.org> jkim 2008-06-24 20:12:44 UTC FreeBSD src repository Modified files: sys/i386/i386 bpf_jit_machdep.c bpf_jit_machdep.h Log: SVN rev 179978 on 2008-06-24 20:12:44Z by jkim Emit opcodes closer to GNU as(1) generated codes and micro-optimize. Revision Changes Path 1.6 +11 -11 src/sys/i386/i386/bpf_jit_machdep.c 1.5 +35 -48 src/sys/i386/i386/bpf_jit_machdep.h From jhb at FreeBSD.org Tue Jun 24 20:41:52 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 24 20:41:54 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/i386/i386 machdep.c src/sys/kern kern_mib.c src/sys/sys sysctl.h Message-ID: <200806242041.m5OKfqNh035484@repoman.freebsd.org> jhb 2008-06-24 20:40:17 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/ciss ciss.c sys/i386/i386 machdep.c sys/kern kern_mib.c sys/sys sysctl.h Log: SVN rev 179979 on 2008-06-24 20:40:17Z by jhb MFC: Add the kern.features sysctl tree and FEATURE() macro. Add a "pae" feature for PAE kernels on i386. Revision Changes Path 1.64.2.9 +0 -0 src/sys/dev/ciss/ciss.c 1.616.2.14 +4 -0 src/sys/i386/i386/machdep.c 1.74.2.3 +2 -0 src/sys/kern/kern_mib.c 1.138.2.4 +8 -0 src/sys/sys/sysctl.h From wollman at FreeBSD.org Tue Jun 24 21:00:55 2008 From: wollman at FreeBSD.org (Garrett Wollman) Date: Tue Jun 24 21:01:02 2008 Subject: cvs commit: src/lib/libc/gen valloc.3 Message-ID: <200806242100.m5OL0tUG046453@repoman.freebsd.org> wollman 2008-06-24 21:00:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/gen valloc.3 Log: SVN rev 179980 on 2008-06-24 21:00:48Z by wollman MFC r173182: remove extraneous .Ef request Revision Changes Path 1.14.2.1 +1 -2 src/lib/libc/gen/valloc.3 From mtm at FreeBSD.org Tue Jun 24 21:02:11 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Tue Jun 24 21:02:13 2008 Subject: cvs commit: src/etc/rc.d syscons Message-ID: <200806242102.m5OL2A9T046575@repoman.freebsd.org> mtm 2008-06-24 21:01:56 UTC FreeBSD src repository Modified files: etc/rc.d syscons Log: SVN rev 179981 on 2008-06-24 21:01:56Z by mtm Quiet rc.d/syscons unless it has something to say. Revision Changes Path 1.22 +35 -6 src/etc/rc.d/syscons From jhb at FreeBSD.org Tue Jun 24 21:06:44 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 24 21:06:46 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/kern kern_mib.c src/sys/pci nfsmb.c Message-ID: <200806242106.m5OL6ioi048461@repoman.freebsd.org> jhb 2008-06-24 21:06:02 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/ciss ciss.c sys/kern kern_mib.c sys/pci nfsmb.c Log: SVN rev 179983 on 2008-06-24 21:06:02Z by jhb MFC: Add 'compat_freebsd[456]' features corresponding to the kernel options COMPAT_FREEBSD[456]. Revision Changes Path 1.81.2.9 +0 -0 src/sys/dev/ciss/ciss.c 1.84.2.4 +13 -0 src/sys/kern/kern_mib.c 1.6.2.7 +0 -0 src/sys/pci/nfsmb.c From jhb at FreeBSD.org Tue Jun 24 21:09:20 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 24 21:09:22 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/kern kern_mib.c Message-ID: <200806242109.m5OL9KrL048699@repoman.freebsd.org> jhb 2008-06-24 21:08:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/ciss ciss.c sys/kern kern_mib.c Log: SVN rev 179984 on 2008-06-24 21:08:30Z by jhb MFC: Add 'compat_freebsd[45]' features corresponding to the kernel options COMPAT_FREEBSD[45]. Revision Changes Path 1.64.2.10 +0 -0 src/sys/dev/ciss/ciss.c 1.74.2.4 +9 -0 src/sys/kern/kern_mib.c From njm at njm.f2s.com Tue Jun 24 22:04:20 2008 From: njm at njm.f2s.com (N.J. Mann) Date: Tue Jun 24 22:04:24 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/kern kern_mib.c src/sys/pci nfsmb.c In-Reply-To: <200806242106.m5OL6ioi048461@repoman.freebsd.org> References: <200806242106.m5OL6ioi048461@repoman.freebsd.org> Message-ID: <20080624213420.GC18932@oberon.njm.f2s.com> In message <200806242106.m5OL6ioi048461@repoman.freebsd.org>, John Baldwin (jhb@FreeBSD.org) wrote: > jhb 2008-06-24 21:06:02 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_7) > sys/dev/ciss ciss.c > sys/kern kern_mib.c > sys/pci nfsmb.c > Log: > SVN rev 179983 on 2008-06-24 21:06:02Z by jhb > > MFC: Add 'compat_freebsd[456]' features corresponding to the kernel options > COMPAT_FREEBSD[456]. > > Revision Changes Path > 1.81.2.9 +0 -0 src/sys/dev/ciss/ciss.c > 1.84.2.4 +13 -0 src/sys/kern/kern_mib.c > 1.6.2.7 +0 -0 src/sys/pci/nfsmb.c > > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/ciss/ciss.c.diff?r1=1.81.2.8&r2=1.81.2.9 > --- src/sys/dev/ciss/ciss.c 2008/06/24 20:04:39 1.81.2.8 > +++ src/sys/dev/ciss/ciss.c 2008/06/24 21:06:02 1.81.2.9 > @@ -24,7 +24,7 @@ > * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF > * SUCH DAMAGE. > * > - * $FreeBSD: /usr/local/www/cvsroot/FreeBSD/src/sys/dev/ciss/ciss.c,v 1.81.2.8 2008/06/24 20:04:39 jhb Exp $ > + * $FreeBSD: /usr/local/www/cvsroot/FreeBSD/src/sys/dev/ciss/ciss.c,v 1.81.2.9 2008/06/24 21:06:02 jhb Exp $ > */ John, Every commit I have seen from you today has include an empty "change" to ciss.c, including RELENG_7 and RELENG_6. Is there some sort of problem at you end and/or with svn? Cheers, Nick. -- From jhb at FreeBSD.org Tue Jun 24 22:09:01 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Tue Jun 24 22:09:07 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/kern kern_mutex.c src/sys/pci nfsmb.c Message-ID: <200806242209.m5OM90wN058413@repoman.freebsd.org> jhb 2008-06-24 22:07:51 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/ciss ciss.c sys/kern kern_mutex.c sys/pci nfsmb.c Log: SVN rev 179985 on 2008-06-24 22:07:51Z by jhb MFC: Add KASSERT()'s to catch attempts to recurse on spin mutexes that aren't marked recursable either via mtx_lock_spin() or thread_lock(). Revision Changes Path 1.81.2.10 +0 -0 src/sys/dev/ciss/ciss.c 1.198.2.3 +9 -1 src/sys/kern/kern_mutex.c 1.6.2.8 +0 -0 src/sys/pci/nfsmb.c From mlaier at FreeBSD.org Tue Jun 24 23:25:12 2008 From: mlaier at FreeBSD.org (Max Laier) Date: Tue Jun 24 23:25:48 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/netinet ip_carp.c src/sys/pci nfsmb.c Message-ID: <200806242325.m5ONPBAX066995@repoman.freebsd.org> mlaier 2008-06-24 23:24:38 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/ciss ciss.c sys/netinet ip_carp.c sys/pci nfsmb.c Log: SVN rev 179986 on 2008-06-24 23:24:38Z by mlaier MFC: r179490 Sort IP addresses before hashing them for the signature. Otherwise carp is sensitive to address configuration order. PR: kern/121574 Reported by: Douglas K. Rand, Wouter de Jong Obtained from: OpenBSD (rev 1.114 + fixes) Revision Changes Path 1.81.2.11 +0 -0 src/sys/dev/ciss/ciss.c 1.52.2.1 +39 -13 src/sys/netinet/ip_carp.c 1.6.2.9 +0 -0 src/sys/pci/nfsmb.c From wollman at FreeBSD.org Wed Jun 25 04:56:21 2008 From: wollman at FreeBSD.org (Garrett Wollman) Date: Wed Jun 25 04:56:28 2008 Subject: cvs commit: src/games/fortune/datfiles fortunes Message-ID: <200806250456.m5P4uLbw020494@repoman.freebsd.org> wollman 2008-06-25 04:56:08 UTC FreeBSD src repository Modified files: games/fortune/datfiles fortunes Log: SVN rev 179987 on 2008-06-25 04:56:08Z by wollman Months in English are capitalized (even when abbreviated). Revision Changes Path 1.265 +1 -1 src/games/fortune/datfiles/fortunes From kientzle at FreeBSD.org Wed Jun 25 05:01:08 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Wed Jun 25 05:01:14 2008 Subject: cvs commit: src/usr.bin/tar write.c Message-ID: <200806250501.m5P517s2020887@repoman.freebsd.org> kientzle 2008-06-25 05:01:02 UTC FreeBSD src repository Modified files: usr.bin/tar write.c Log: SVN rev 179988 on 2008-06-25 05:01:02Z by kientzle Pass the entry down into the core write loop, so we can include the filename when reporting errors. Thanks to: Dan Nelson Revision Changes Path 1.71 +6 -4 src/usr.bin/tar/write.c From ale at FreeBSD.org Wed Jun 25 06:07:09 2008 From: ale at FreeBSD.org (Alex Dupre) Date: Wed Jun 25 06:07:18 2008 Subject: cvs commit: src/share/mk bsd.cpu.mk Message-ID: <200806250607.m5P679QI028538@repoman.freebsd.org> ale 2008-06-25 06:07:03 UTC FreeBSD src repository Modified files: share/mk bsd.cpu.mk Log: SVN rev 179989 on 2008-06-25 06:07:03Z by ale Fix links to online gcc docs. Reported by: Andre Guibert de Bruet MFC after: 1 day Revision Changes Path 1.67 +3 -3 src/share/mk/bsd.cpu.mk From ed at FreeBSD.org Wed Jun 25 07:46:30 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Wed Jun 25 07:46:40 2008 Subject: cvs commit: src/sys/amd64/include iodev.h memdev.h src/sys/arm/include memdev.h src/sys/dev/io iodev.c src/sys/i386/include iodev.h memdev.h src/sys/ia64/include memdev.h src/sys/mips/include iodev.h memdev.h ... Message-ID: <200806250746.m5P7kUdH045968@repoman.freebsd.org> ed 2008-06-25 07:45:31 UTC FreeBSD src repository Modified files: sys/amd64/include iodev.h memdev.h sys/arm/include memdev.h sys/dev/io iodev.c sys/i386/include iodev.h memdev.h sys/ia64/include memdev.h sys/mips/include iodev.h memdev.h sys/powerpc/include memdev.h sys/sparc64/include memdev.h sys/sun4v/include memdev.h Log: SVN rev 179990 on 2008-06-25 07:45:31Z by ed Remove the unused major/minor numbers from iodev and memdev. Now that st_rdev is being automatically generated by the kernel, there is no need to define static major/minor numbers for the iodev and memdev. We still need the minor numbers for the memdev, however, to distinguish between /dev/mem and /dev/kmem. Approved by: philip (mentor) Revision Changes Path 1.2 +0 -3 src/sys/amd64/include/iodev.h 1.2 +0 -1 src/sys/amd64/include/memdev.h 1.2 +0 -1 src/sys/arm/include/memdev.h 1.3 +1 -1 src/sys/dev/io/iodev.c 1.2 +0 -3 src/sys/i386/include/iodev.h 1.2 +0 -1 src/sys/i386/include/memdev.h 1.3 +0 -1 src/sys/ia64/include/memdev.h 1.2 +0 -3 src/sys/mips/include/iodev.h 1.2 +0 -1 src/sys/mips/include/memdev.h 1.2 +0 -1 src/sys/powerpc/include/memdev.h 1.2 +0 -1 src/sys/sparc64/include/memdev.h 1.2 +0 -1 src/sys/sun4v/include/memdev.h From ed at FreeBSD.org Wed Jun 25 07:52:39 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Wed Jun 25 07:52:41 2008 Subject: cvs commit: src/sys/i386/include memdev.h src/sys/powerpc/include memdev.h Message-ID: <200806250752.m5P7qcR3046464@repoman.freebsd.org> ed 2008-06-25 07:52:10 UTC FreeBSD src repository Modified files: sys/i386/include memdev.h sys/powerpc/include memdev.h Log: SVN rev 179991 on 2008-06-25 07:52:10Z by ed Remove the unused M_MEMDEV from the kernel. The M_MEMDEV memory allocation pool does not seem to be used. We can live without it. Approved by: philip (mentor) Revision Changes Path 1.3 +0 -2 src/sys/i386/include/memdev.h 1.3 +0 -2 src/sys/powerpc/include/memdev.h From peter at FreeBSD.org Wed Jun 25 08:47:40 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 08:47:47 2008 Subject: cvs commit: src/usr.bin/whereis whereis.c Message-ID: <200806250847.m5P8ldgg061472@repoman.freebsd.org> peter 2008-06-25 08:47:36 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/whereis whereis.c Log: SVN rev 179992 on 2008-06-25 08:47:36Z by peter Move mergeinfo up to containing directory. Revision Changes Path 1.15.2.2 +0 -0 src/usr.bin/whereis/whereis.c From peter at FreeBSD.org Wed Jun 25 08:58:41 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 08:58:51 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c Message-ID: <200806250858.m5P8weBC062291@repoman.freebsd.org> peter 2008-06-25 08:58:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/ciss ciss.c Log: SVN rev 179993 on 2008-06-25 08:58:35Z by peter Promote mergeinfo from ciss.c to the containing directory. Revision Changes Path 1.64.2.11 +0 -0 src/sys/dev/ciss/ciss.c From peter at FreeBSD.org Wed Jun 25 09:09:26 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 09:09:28 2008 Subject: cvs commit: src/usr.bin/shar shar.sh Message-ID: <200806250909.m5P99Q9F064280@repoman.freebsd.org> peter 2008-06-25 09:09:13 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.bin/shar shar.sh Log: SVN rev 179994 on 2008-06-25 09:09:13Z by peter Move mergeinfo up to containing directory. Revision Changes Path 1.3.38.2 +0 -0 src/usr.bin/shar/shar.sh From peter at FreeBSD.org Wed Jun 25 09:24:44 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 09:24:51 2008 Subject: cvs commit: src/share/man/man4 ciss.4 Message-ID: <200806250924.m5P9Og1c065512@repoman.freebsd.org> peter 2008-06-25 09:24:35 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/man/man4 ciss.4 Log: SVN rev 179995 on 2008-06-25 09:24:35Z by peter Move mergeinfo up from file to containing directory Revision Changes Path 1.11.2.7 +0 -0 src/share/man/man4/ciss.4 From peter at FreeBSD.org Wed Jun 25 09:33:37 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 09:33:44 2008 Subject: cvs commit: src/share/sendmail Makefile src/usr.bin/cksum cksum.c Message-ID: <200806250933.m5P9XbLg066305@repoman.freebsd.org> peter 2008-06-25 09:29:46 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/sendmail Makefile usr.bin/cksum cksum.c Log: SVN rev 179996 on 2008-06-25 09:29:46Z by peter Move mergeinfo up from file to directory Revision Changes Path 1.10.2.2 +0 -0 src/share/sendmail/Makefile 1.17.12.3 +0 -0 src/usr.bin/cksum/cksum.c From peter at FreeBSD.org Wed Jun 25 09:52:07 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 09:52:09 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c Message-ID: <200806250952.m5P9q7qL067768@repoman.freebsd.org> peter 2008-06-25 09:51:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/ciss ciss.c Log: SVN rev 179998 on 2008-06-25 09:51:59Z by peter Move mergeinfo from ciss.c up to containing directory. Revision Changes Path 1.81.2.12 +0 -0 src/sys/dev/ciss/ciss.c From peter at FreeBSD.org Wed Jun 25 09:56:53 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 09:57:02 2008 Subject: cvs commit: src/sys/pci nfsmb.c Message-ID: <200806250956.m5P9ur8m068150@repoman.freebsd.org> peter 2008-06-25 09:56:44 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/pci nfsmb.c Log: SVN rev 179999 on 2008-06-25 09:56:44Z by peter Move mergeinfo from file up to parent directory Revision Changes Path 1.6.2.10 +0 -0 src/sys/pci/nfsmb.c From peter at FreeBSD.org Wed Jun 25 10:00:40 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 10:00:41 2008 Subject: cvs commit: src/share/man/man4 ciss.4 Message-ID: <200806251000.m5PA0d9g068488@repoman.freebsd.org> peter 2008-06-25 10:00:28 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 ciss.4 Log: SVN rev 180000 on 2008-06-25 10:00:28Z by peter Move mergeinfo from file up to directory; fix broken mergeinfo record. Revision Changes Path 1.14.2.6 +0 -0 src/share/man/man4/ciss.4 From peter at FreeBSD.org Wed Jun 25 10:06:10 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 10:06:17 2008 Subject: cvs commit: src/usr.bin/whereis whereis.c Message-ID: <200806251006.m5PA6AdH070068@repoman.freebsd.org> peter 2008-06-25 10:06:00 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/whereis whereis.c Log: SVN rev 180001 on 2008-06-25 10:06:00Z by peter Move mergeinfo to containing directory. Fix some very strange records that make no sense. Revision Changes Path 1.15.10.3 +0 -0 src/usr.bin/whereis/whereis.c From peter at FreeBSD.org Wed Jun 25 10:10:42 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 10:10:44 2008 Subject: cvs commit: src/usr.bin/cksum cksum.c Message-ID: <200806251010.m5PAAfdt070439@repoman.freebsd.org> peter 2008-06-25 10:10:30 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/cksum cksum.c Log: SVN rev 180002 on 2008-06-25 10:10:30Z by peter Move mergeinfo up from file to containing directory. Revision Changes Path 1.17.22.3 +0 -0 src/usr.bin/cksum/cksum.c From peter at FreeBSD.org Wed Jun 25 10:13:13 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Wed Jun 25 10:13:33 2008 Subject: cvs commit: src/usr.bin/shar shar.sh Message-ID: <200806251013.m5PADCnq070667@repoman.freebsd.org> peter 2008-06-25 10:13:03 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/shar shar.sh Log: SVN rev 180003 on 2008-06-25 10:13:03Z by peter Move mergeinfo from file to containing directory. Revision Changes Path 1.3.48.4 +0 -0 src/usr.bin/shar/shar.sh From jhb at freebsd.org Wed Jun 25 13:41:00 2008 From: jhb at freebsd.org (John Baldwin) Date: Wed Jun 25 13:41:12 2008 Subject: cvs commit: src/sys/dev/ciss ciss.c src/sys/kern kern_mib.c src/sys/pci nfsmb.c In-Reply-To: <20080624213420.GC18932@oberon.njm.f2s.com> References: <200806242106.m5OL6ioi048461@repoman.freebsd.org> <20080624213420.GC18932@oberon.njm.f2s.com> Message-ID: <200806250920.57763.jhb@freebsd.org> On Tuesday 24 June 2008 05:34:20 pm N.J. Mann wrote: > In message <200806242106.m5OL6ioi048461@repoman.freebsd.org>, > > John Baldwin (jhb@FreeBSD.org) wrote: > > jhb 2008-06-24 21:06:02 UTC > > > > FreeBSD src repository > > > > Modified files: (Branch: RELENG_7) > > sys/dev/ciss ciss.c > > sys/kern kern_mib.c > > sys/pci nfsmb.c > > Log: > > SVN rev 179983 on 2008-06-24 21:06:02Z by jhb > > > > MFC: Add 'compat_freebsd[456]' features corresponding to the kernel > > options COMPAT_FREEBSD[456]. > > > > Revision Changes Path > > 1.81.2.9 +0 -0 src/sys/dev/ciss/ciss.c > > 1.84.2.4 +13 -0 src/sys/kern/kern_mib.c > > 1.6.2.7 +0 -0 src/sys/pci/nfsmb.c > > > > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/ciss/ciss.c.diff?r1=1.8 > >1.2.8&r2=1.81.2.9 --- src/sys/dev/ciss/ciss.c 2008/06/24 20:04:39 > > 1.81.2.8 +++ src/sys/dev/ciss/ciss.c 2008/06/24 21:06:02 1.81.2.9 > > @@ -24,7 +24,7 @@ > > * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY > > OF * SUCH DAMAGE. > > * > > - * $FreeBSD: > > /usr/local/www/cvsroot/FreeBSD/src/sys/dev/ciss/ciss.c,v 1.81.2.8 > > 2008/06/24 20:04:39 jhb Exp $ + * $FreeBSD: > > /usr/local/www/cvsroot/FreeBSD/src/sys/dev/ciss/ciss.c,v 1.81.2.9 > > 2008/06/24 21:06:02 jhb Exp $ */ > > John, > > > Every commit I have seen from you today has include an empty "change" to > ciss.c, including RELENG_7 and RELENG_6. Is there some sort of problem > at you end and/or with svn? It has to do with the way SVN tracks merges. -- John Baldwin From bz at FreeBSD.org Wed Jun 25 18:11:33 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Wed Jun 25 18:11:40 2008 Subject: cvs commit: src/sbin/atacontrol atacontrol.8 Message-ID: <200806251811.m5PIBXOg037528@repoman.freebsd.org> bz 2008-06-25 18:11:22 UTC FreeBSD src repository Modified files: sbin/atacontrol atacontrol.8 Log: SVN rev 180005 on 2008-06-25 18:11:22Z by bz Document spindown constraints as given in the original commit message[1] and later clarification provided by phk. [1] http://docs.freebsd.org/cgi/mid.cgi?200803171033.m2HAXOeN055116 Reviewed by: brueffer, phk, ed Revision Changes Path 1.31 +22 -1 src/sbin/atacontrol/atacontrol.8 From ru at FreeBSD.org Wed Jun 25 20:29:36 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Wed Jun 25 20:29:45 2008 Subject: cvs commit: src/release/scripts src-install.sh Message-ID: <200806252029.m5PKTZKj050937@repoman.freebsd.org> ru 2008-06-25 20:29:22 UTC FreeBSD src repository Modified files: release/scripts src-install.sh Log: SVN rev 180008 on 2008-06-25 20:29:22Z by ru src/compat/ is gone back in March. Reported by: Mars G Miro Revision Changes Path 1.13 +2 -2 src/release/scripts/src-install.sh From ru at FreeBSD.org Wed Jun 25 20:37:21 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Wed Jun 25 20:37:23 2008 Subject: cvs commit: src/release/scripts src-install.sh Message-ID: <200806252037.m5PKbL8U051677@repoman.freebsd.org> ru 2008-06-25 20:37:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) release/scripts src-install.sh Log: SVN rev 180009 on 2008-06-25 20:37:04Z by ru MFC: src/compat is gone. Revision Changes Path 1.11.2.2 +2 -2 src/release/scripts/src-install.sh From marius at FreeBSD.org Wed Jun 25 21:03:39 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 25 21:03:41 2008 Subject: cvs commit: src/sys/sun4v/include in_cksum.h Message-ID: <200806252103.m5PL3de5064262@repoman.freebsd.org> marius 2008-06-25 21:03:26 UTC FreeBSD src repository Modified files: sys/sun4v/include in_cksum.h Log: SVN rev 180010 on 2008-06-25 21:03:26Z by marius Given that sun4u uses sparc64/sparc64/in_cksum.c, use the sparc64 here also. MFC after: 3 days Revision Changes Path 1.2 +2 -164 src/sys/sun4v/include/in_cksum.h From marius at FreeBSD.org Wed Jun 25 21:05:10 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Wed Jun 25 21:05:17 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h Message-ID: <200806252105.m5PL5AUp064418@repoman.freebsd.org> marius 2008-06-25 21:04:59 UTC FreeBSD src repository Modified files: sys/sparc64/include in_cksum.h Log: SVN rev 180011 on 2008-06-25 21:04:59Z by marius Use "__asm __volatile" rather than "__asm" for instruction sequences that modify condition codes (the carry bit, in this case). Without "__volatile", the compiler might add the inline assembler instructions between unrelated code which also uses condition codes, modifying the latter. This prevents the TCP pseudo header checksum calculation done in tcp_output() from having effects on other conditions when compiled with GCC 4.2.1 at "-O2" and "options INET6" left out. [1] Reported & tested by: Boris Kochergin [1] MFC after: 3 days Revision Changes Path 1.4 +7 -7 src/sys/sparc64/include/in_cksum.h From ru at FreeBSD.org Wed Jun 25 21:38:36 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Wed Jun 25 21:38:38 2008 Subject: cvs commit: src Makefile.inc1 src/gnu/lib Makefile src/gnu/lib/csu Makefile src/gnu/lib/libssp Makefile src/lib/csu Makefile.inc src/lib/libc Makefile src/lib/libstand Makefile src/lib/libthr Makefile src/libexec/rtld-elf Makefile src/release Makefile ... Message-ID: <200806252138.m5PLcajG067917@repoman.freebsd.org> ru 2008-06-25 21:33:28 UTC FreeBSD src repository Modified files: . Makefile.inc1 gnu/lib Makefile gnu/lib/csu Makefile gnu/lib/libssp Makefile lib/libc Makefile lib/libstand Makefile lib/libthr Makefile libexec/rtld-elf Makefile release Makefile release/picobsd/build picobsd rescue/librescue Makefile rescue/rescue Makefile share/mk bsd.sys.mk sys/boot/arm/at91 Makefile.inc sys/boot/efi Makefile.inc sys/boot/i386 Makefile.inc sys/boot/i386/loader Makefile sys/boot/ia64 Makefile.inc sys/boot/ia64/common Makefile sys/boot/ia64/efi Makefile sys/boot/ia64/ski Makefile sys/boot/pc98 Makefile.inc sys/boot/pc98/loader Makefile sys/boot/powerpc/ofw Makefile sys/boot/sparc64 Makefile.inc sys/boot/sparc64/loader Makefile sys/conf files kern.mk kern.pre.mk tools/build/options WITHOUT_SSP Added files: lib/csu Makefile.inc sys/boot Makefile.inc sys/boot/arm Makefile.inc sys/boot/ofw Makefile.inc sys/boot/powerpc Makefile.inc sys/boot/uboot Makefile.inc sys/kern stack_protector.c Log: SVN rev 180012 on 2008-06-25 21:33:28Z by ru Enable GCC stack protection (aka Propolice) for userland: - It is opt-out for now so as to give it maximum testing, but it may be turned opt-in for stable branches depending on the consensus. You can turn it off with WITHOUT_SSP. - WITHOUT_SSP was previously used to disable the build of GNU libssp. It is harmless to steal the knob as SSP symbols have been provided by libc for a long time, GNU libssp should not have been much used. - SSP is disabled in a few corners such as system bootstrap programs (sys/boot), process bootstrap code (rtld, csu) and SSP symbols themselves. - It should be safe to use -fstack-protector-all to build world, however libc will be automatically downgraded to -fstack-protector because it breaks rtld otherwise. - This option is unavailable on ia64. Enable GCC stack protection (aka Propolice) for kernel: - It is opt-out for now so as to give it maximum testing. - Do not compile your kernel with -fstack-protector-all, it won't work. Submitted by: Jeremie Le Hen Revision Changes Path 1.607 +6 -5 src/Makefile.inc1 1.43 +1 -5 src/gnu/lib/Makefile 1.29 +1 -0 src/gnu/lib/csu/Makefile 1.3 +1 -0 src/gnu/lib/libssp/Makefile 1.1 +3 -0 src/lib/csu/Makefile.inc (new) 1.74 +6 -0 src/lib/libc/Makefile 1.62 +1 -0 src/lib/libstand/Makefile 1.35 +2 -0 src/lib/libthr/Makefile 1.42 +2 -0 src/libexec/rtld-elf/Makefile 1.932 +4 -3 src/release/Makefile 1.42 +5 -4 src/release/picobsd/build/picobsd 1.10 +2 -0 src/rescue/librescue/Makefile 1.62 +1 -0 src/rescue/rescue/Makefile 1.45 +6 -0 src/share/mk/bsd.sys.mk 1.1 +3 -0 src/sys/boot/Makefile.inc (new) 1.1 +3 -0 src/sys/boot/arm/Makefile.inc (new) 1.8 +2 -0 src/sys/boot/arm/at91/Makefile.inc 1.8 +2 -0 src/sys/boot/efi/Makefile.inc 1.13 +2 -0 src/sys/boot/i386/Makefile.inc 1.86 +2 -0 src/sys/boot/i386/loader/Makefile 1.4 +2 -0 src/sys/boot/ia64/Makefile.inc 1.2 +2 -0 src/sys/boot/ia64/common/Makefile 1.29 +1 -0 src/sys/boot/ia64/efi/Makefile 1.21 +1 -0 src/sys/boot/ia64/ski/Makefile 1.1 +3 -0 src/sys/boot/ofw/Makefile.inc (new) 1.8 +2 -0 src/sys/boot/pc98/Makefile.inc 1.42 +2 -0 src/sys/boot/pc98/loader/Makefile 1.1 +3 -0 src/sys/boot/powerpc/Makefile.inc (new) 1.24 +2 -0 src/sys/boot/powerpc/ofw/Makefile 1.2 +2 -0 src/sys/boot/sparc64/Makefile.inc 1.21 +2 -0 src/sys/boot/sparc64/loader/Makefile 1.1 +3 -0 src/sys/boot/uboot/Makefile.inc (new) 1.1311 +2 -0 src/sys/conf/files 1.55 +7 -0 src/sys/conf/kern.mk 1.100 +1 -4 src/sys/conf/kern.pre.mk 1.1 +35 -0 src/sys/kern/stack_protector.c (new) 1.2 +1 -1 src/tools/build/options/WITHOUT_SSP From ru at FreeBSD.org Wed Jun 25 21:40:21 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Wed Jun 25 21:40:28 2008 Subject: cvs commit: src/share/man/man5 src.conf.5 Message-ID: <200806252140.m5PLeLlJ068157@repoman.freebsd.org> ru 2008-06-25 21:36:25 UTC FreeBSD src repository Modified files: share/man/man5 src.conf.5 Log: SVN rev 180013 on 2008-06-25 21:36:25Z by ru Regen. Revision Changes Path 1.29 +88 -88 src/share/man/man5/src.conf.5 From ru at FreeBSD.org Wed Jun 25 21:42:38 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Wed Jun 25 21:42:40 2008 Subject: cvs commit: src/share/man/man5 src.conf.5 Message-ID: <200806252142.m5PLgbHI068348@repoman.freebsd.org> ru 2008-06-25 21:42:23 UTC FreeBSD src repository Modified files: share/man/man5 src.conf.5 Log: SVN rev 180014 on 2008-06-25 21:42:23Z by ru Regen properly. Revision Changes Path 1.30 +85 -85 src/share/man/man5/src.conf.5 From ru at FreeBSD.org Wed Jun 25 22:13:39 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Wed Jun 25 22:13:45 2008 Subject: cvs commit: src/contrib/top top.X top.c Message-ID: <200806252213.m5PMDc5Q074895@repoman.freebsd.org> ru 2008-06-25 22:13:24 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) contrib/top top.X top.c Log: SVN rev 180015 on 2008-06-25 22:13:24Z by ru MFC: Removed the no-op -p; documented -P. Revision Changes Path 1.20.2.1 +4 -1 src/contrib/top/top.X 1.23.2.3 +3 -7 src/contrib/top/top.c From kientzle at FreeBSD.org Thu Jun 26 04:48:58 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Thu Jun 26 04:49:04 2008 Subject: cvs commit: src/lib/libarchive/test test_read_format_zip.c test_read_format_zip.zip.uu Message-ID: <200806260448.m5Q4mwrF033746@repoman.freebsd.org> kientzle 2008-06-26 04:48:42 UTC FreeBSD src repository Modified files: lib/libarchive/test test_read_format_zip.c Added files: lib/libarchive/test test_read_format_zip.zip.uu Log: SVN rev 180016 on 2008-06-26 04:48:42Z by kientzle Split out the reference zip file for ease of maintenance. Revision Changes Path 1.4 +3 -19 src/lib/libarchive/test/test_read_format_zip.c 1.1 +13 -0 src/lib/libarchive/test/test_read_format_zip.zip.uu (new) From ale at FreeBSD.org Thu Jun 26 05:41:44 2008 From: ale at FreeBSD.org (Alex Dupre) Date: Thu Jun 26 05:41:47 2008 Subject: cvs commit: src/share/mk bsd.cpu.mk Message-ID: <200806260541.m5Q5fiQV039012@repoman.freebsd.org> ale 2008-06-26 05:41:25 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/mk bsd.cpu.mk Log: SVN rev 180017 on 2008-06-26 05:41:25Z by ale MFC r179989: Fix links to online gcc docs. Revision Changes Path 1.62.2.3 +3 -3 src/share/mk/bsd.cpu.mk From ale at FreeBSD.org Thu Jun 26 05:46:30 2008 From: ale at FreeBSD.org (Alex Dupre) Date: Thu Jun 26 05:46:37 2008 Subject: cvs commit: src/share/mk bsd.cpu.mk Message-ID: <200806260546.m5Q5kUXj039411@repoman.freebsd.org> ale 2008-06-26 05:46:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) share/mk bsd.cpu.mk Log: SVN rev 180018 on 2008-06-26 05:46:11Z by ale MFC r179989: Fix links to online gcc docs. Revision Changes Path 1.48.2.6 +3 -3 src/share/mk/bsd.cpu.mk From sobomax at FreeBSD.org Thu Jun 26 07:03:03 2008 From: sobomax at FreeBSD.org (Maxim Sobolev) Date: Thu Jun 26 07:03:12 2008 Subject: cvs commit: src/usr.sbin/newsyslog newsyslog.conf.5 Message-ID: <200806260703.m5Q7331Q047038@repoman.freebsd.org> sobomax 2008-06-26 07:02:47 UTC FreeBSD src repository Modified files: usr.sbin/newsyslog newsyslog.conf.5 Log: SVN rev 180019 on 2008-06-26 07:02:47Z by sobomax Fix 6-year old cut&paste error. The # could be escaped with '\', not with '\\'. MFC after: 2 weeks Revision Changes Path 1.8 +2 -2 src/usr.sbin/newsyslog/newsyslog.conf.5 From mtm at FreeBSD.org Thu Jun 26 07:05:49 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Jun 26 07:05:55 2008 Subject: cvs commit: src/sbin/ggate/ggated ggated.c Message-ID: <200806260705.m5Q75mDu047893@repoman.freebsd.org> mtm 2008-06-26 07:05:35 UTC FreeBSD src repository Modified files: sbin/ggate/ggated ggated.c Log: SVN rev 180020 on 2008-06-26 07:05:35Z by mtm The signature for a pthread function requires that it return a pointer to a void. The send_thread() and disk_thread() funtions; however, do not have a return value because they run for the duration of the daemon's lifetime. This causes gcc to barf when running with -O3. Make these functions return a null pointer to quiet it. PR: bin/124342 Submitted by: Garrett Cooper (minus his comments) MFC after: 1 week Revision Changes Path 1.11 +6 -0 src/sbin/ggate/ggated/ggated.c From mtm at FreeBSD.org Thu Jun 26 07:12:49 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Thu Jun 26 07:12:51 2008 Subject: cvs commit: src/lib/libc/gen glob.c Message-ID: <200806260712.m5Q7CnXj048439@repoman.freebsd.org> mtm 2008-06-26 07:12:35 UTC FreeBSD src repository Modified files: lib/libc/gen glob.c Log: SVN rev 180021 on 2008-06-26 07:12:35Z by mtm Gcc barfs in glob.c when run with -O3. To fix this make g_strchr() work on and return (const Char *) pointers instead of just (Char *) and get rid of all the type casting. PR: kern/124334 Revision Changes Path 1.27 +5 -5 src/lib/libc/gen/glob.c From ru at FreeBSD.org Thu Jun 26 07:52:51 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Thu Jun 26 07:52:54 2008 Subject: cvs commit: src/sys/kern stack_protector.c Message-ID: <200806260752.m5Q7qphk051885@repoman.freebsd.org> ru 2008-06-26 07:52:45 UTC FreeBSD src repository Modified files: sys/kern stack_protector.c Log: SVN rev 180022 on 2008-06-26 07:52:45Z by ru Fix a chicken-and-egg problem: this files implements SSP support, so we cannot compile it with -fstack-protector[-all] flags (or it will self-recurse); this is ensured in sys/conf/files. This OTOH means that checking for defines __SSP__ and __SSP_ALL__ to determine if we should be compiling the support is impossible (which it was trying, resulting in an empty object file). Fix this by always compiling the symbols in this files. It's good because it allows us to always have SSP support, and then compile with SSP selectively. Repoted by: tinderbox Revision Changes Path 1.2 +0 -3 src/sys/kern/stack_protector.c From ru at FreeBSD.org Thu Jun 26 07:56:22 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Thu Jun 26 07:56:29 2008 Subject: cvs commit: src/sys/boot/pc98/boot2 Makefile Message-ID: <200806260756.m5Q7uMC7052275@repoman.freebsd.org> ru 2008-06-26 07:56:16 UTC FreeBSD src repository Modified files: sys/boot/pc98/boot2 Makefile Log: SVN rev 180023 on 2008-06-26 07:56:16Z by ru Fix a fallout from SSP commit, and make this compile again. Bonus: including kern.mk just to pick kernel warning flags was an extremely bad idea anyway, because it also picked up CFLAGS (it probably wasn't the case at the time of CVS rev. 1.1, I haven't checked). Remove duplicate CWARNFLAGS from CFLAGS. Revision Changes Path 1.24 +1 -2 src/sys/boot/pc98/boot2/Makefile From leafy7382 at gmail.com Thu Jun 26 08:06:44 2008 From: leafy7382 at gmail.com (Jiawei Ye) Date: Thu Jun 26 08:06:47 2008 Subject: cvs commit: src Makefile.inc1 src/gnu/lib Makefile src/gnu/lib/csu Makefile src/gnu/lib/libssp Makefile src/lib/csu Makefile.inc src/lib/libc Makefile src/lib/libstand Makefile src/lib/libthr Makefile src/libexec/rtld-elf Makefile src/release Mak Message-ID: On Thu, Jun 26, 2008 at 5:33 AM, Ruslan Ermilov wrote: > ru 2008-06-25 21:33:28 UTC > > > Enable GCC stack protection (aka Propolice) for kernel: > - It is opt-out for now so as to give it maximum testing. > - Do not compile your kernel with -fstack-protector-all, it won't work. > > Submitted by: Jeremie Le Hen > > I am getting kernel compilation error (when linking) -- error repeated -- vga_isa.o(.text+0x2ef): In function `isavga_probe': /usr/src/sys/isa/vga_isa.c:98: undefined reference to `__stack_chk_guard' vga_isa.o(.text+0x407):/usr/src/sys/isa/vga_isa.c:121: undefined reference to `__stack_chk_guard' vga_isa.o(.text+0x40e):/usr/src/sys/isa/vga_isa.c:121: undefined reference to `__stack_chk_fail' *** Error code 1 Stop in /tmp/obj/usr/src/sys/MAIL. *** Error code 1 Stop in /usr/src. *** Error code 1 Stop in /usr/src. Missing something here? Jiawei -- "If it looks like a duck, walks like a duck, and quacks like a duck, then to the end user it's a duck, and end users have made it pretty clear they want a duck; whether the duck drinks hot chocolate or coffee is irrelevant." From danger at FreeBSD.org Thu Jun 26 08:25:09 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Thu Jun 26 08:26:25 2008 Subject: cvs commit: src/lib/libc/stdlib system.3 Message-ID: <200806260825.m5Q8P8pY055996@repoman.freebsd.org> danger 2008-06-26 08:24:59 UTC FreeBSD src repository Modified files: lib/libc/stdlib system.3 Log: SVN rev 180024 on 2008-06-26 08:24:59Z by danger Mark the section describing return values with an appropriate section flag. PR: docs/122818 MFC after: 3 days Revision Changes Path 1.13 +1 -1 src/lib/libc/stdlib/system.3 From ru at freebsd.org Thu Jun 26 08:55:00 2008 From: ru at freebsd.org (Ruslan Ermilov) Date: Thu Jun 26 08:55:08 2008 Subject: cvs commit: src Makefile.inc1 src/gnu/lib Makefile src/gnu/lib/csu Makefile src/gnu/lib/libssp Makefile src/lib/csu Makefile.inc src/lib/libc Makefile src/lib/libstand Makefile src/lib/libthr Makefile src/libexec/rtld-elf Makefile src/release Mak In-Reply-To: References: Message-ID: <20080626082515.GB31554@edoofus.dev.vega.ru> On Thu, Jun 26, 2008 at 03:39:53PM +0800, Jiawei Ye wrote: > I am getting kernel compilation error (when linking) > -- error repeated -- > vga_isa.o(.text+0x2ef): In function `isavga_probe': > /usr/src/sys/isa/vga_isa.c:98: undefined reference to `__stack_chk_guard' > vga_isa.o(.text+0x407):/usr/src/sys/isa/vga_isa.c:121: undefined reference > to `__stack_chk_guard' > vga_isa.o(.text+0x40e):/usr/src/sys/isa/vga_isa.c:121: undefined reference > to `__stack_chk_fail' > *** Error code 1 > Stop in /tmp/obj/usr/src/sys/MAIL. > *** Error code 1 > Stop in /usr/src. > *** Error code 1 > Stop in /usr/src. > Missing something here? > Yes. You're missing my recent commit that fixes this. ;) Cheers, -- Ruslan Ermilov ru@FreeBSD.org FreeBSD committer From dfr at FreeBSD.org Thu Jun 26 10:27:10 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Thu Jun 26 10:27:21 2008 Subject: cvs commit: src/sys/conf files src/sys/kern kern_lockf.c src/sys/modules/nfslockd Makefile src/sys/nfsclient nfs.h nfs_node.c nfs_vfsops.c nfs_vnops.c nfsmount.h nfsnode.h src/sys/nlm nlm.h nlm_advlock.c nlm_prot.h nlm_prot_clnt.c ... Message-ID: <200806261027.m5QARA9S077372@repoman.freebsd.org> dfr 2008-06-26 10:21:54 UTC FreeBSD src repository Modified files: sys/conf files sys/kern kern_lockf.c sys/modules/nfslockd Makefile sys/nfsclient nfs.h nfs_node.c nfs_vfsops.c nfs_vnops.c nfsmount.h nfsnode.h sys/nlm nlm.h nlm_prot.h nlm_prot_clnt.c nlm_prot_impl.c nlm_prot_server.c sys/rpc auth_unix.c authunix_prot.c clnt.h clnt_dg.c clnt_rc.c clnt_vc.c svc_vc.c sys/sys fcntl.h lockf.h param.h tools/regression/file/flock flock.c usr.sbin/rpc.lockd lockd.c usr.sbin/rpc.statd file.c Added files: sys/nlm nlm_advlock.c Log: SVN rev 180025 on 2008-06-26 10:21:54Z by dfr Re-implement the client side of rpc.lockd in the kernel. This implementation provides the correct semantics for flock(2) style locks which are used by the lockf(1) command line tool and the pidfile(3) library. It also implements recovery from server restarts and ensures that dirty cache blocks are written to the server before obtaining locks (allowing multiple clients to use file locking to safely share data). Sponsored by: Isilon Systems PR: 94256 MFC after: 2 weeks Revision Changes Path 1.1312 +1 -0 src/sys/conf/files 1.65 +110 -26 src/sys/kern/kern_lockf.c 1.2 +1 -0 src/sys/modules/nfslockd/Makefile 1.100 +1 -0 src/sys/nfsclient/nfs.h 1.89 +7 -0 src/sys/nfsclient/nfs_node.c 1.206 +7 -0 src/sys/nfsclient/nfs_vfsops.c 1.286 +9 -2 src/sys/nfsclient/nfs_vnops.c 1.34 +1 -0 src/sys/nfsclient/nfsmount.h 1.63 +3 -0 src/sys/nfsclient/nfsnode.h 1.2 +120 -24 src/sys/nlm/nlm.h 1.1 +1235 -0 src/sys/nlm/nlm_advlock.c (new) 1.2 +39 -39 src/sys/nlm/nlm_prot.h 1.3 +117 -120 src/sys/nlm/nlm_prot_clnt.c 1.10 +549 -176 src/sys/nlm/nlm_prot_impl.c 1.3 +66 -112 src/sys/nlm/nlm_prot_server.c 1.3 +95 -32 src/sys/rpc/auth_unix.c 1.3 +5 -0 src/sys/rpc/authunix_prot.c 1.2 +88 -10 src/sys/rpc/clnt.h 1.3 +146 -72 src/sys/rpc/clnt_dg.c 1.4 +90 -13 src/sys/rpc/clnt_rc.c 1.3 +96 -42 src/sys/rpc/clnt_vc.c 1.3 +36 -13 src/sys/rpc/svc_vc.c 1.22 +1 -0 src/sys/sys/fcntl.h 1.23 +5 -0 src/sys/sys/lockf.h 1.358 +1 -1 src/sys/sys/param.h 1.3 +182 -50 src/tools/regression/file/flock/flock.c 1.29 +62 -45 src/usr.sbin/rpc.lockd/lockd.c 1.9 +30 -0 src/usr.sbin/rpc.statd/file.c From des at FreeBSD.org Thu Jun 26 10:53:22 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Thu Jun 26 10:53:29 2008 Subject: cvs commit: src/lib/libarchive/test Makefile Message-ID: <200806261053.m5QArMSr079539@repoman.freebsd.org> des 2008-06-26 10:53:05 UTC FreeBSD src repository Modified files: lib/libarchive/test Makefile Log: SVN rev 180026 on 2008-06-26 10:53:05Z by des Allow the tests to build without libdmalloc. Revision Changes Path 1.22 +2 -2 src/lib/libarchive/test/Makefile From des at FreeBSD.org Thu Jun 26 11:48:39 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Thu Jun 26 11:48:45 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_zip.c Message-ID: <200806261148.m5QBmcrQ085165@repoman.freebsd.org> des 2008-06-26 11:48:19 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_zip.c Log: SVN rev 180027 on 2008-06-26 11:48:19Z by des Implement CRC32 verification. Note that you have to read until EOF to trigger the check. Requested by: ache Approved by: kientzle Revision Changes Path 1.25 +12 -5 src/lib/libarchive/archive_read_support_format_zip.c From des at FreeBSD.org Thu Jun 26 11:50:24 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Thu Jun 26 11:50:34 2008 Subject: cvs commit: src/lib/libarchive/test test_read_format_zip.c test_read_format_zip.zip.uu Message-ID: <200806261150.m5QBoOR3085369@repoman.freebsd.org> des 2008-06-26 11:50:11 UTC FreeBSD src repository Modified files: lib/libarchive/test test_read_format_zip.c test_read_format_zip.zip.uu Log: SVN rev 180028 on 2008-06-26 11:50:11Z by des Add regression test for CRC32 check. The test file has been modified to include an invalid checksum for file2. Approved by: kientzle Revision Changes Path 1.5 +3 -2 src/lib/libarchive/test/test_read_format_zip.c 1.2 +3 -3 src/lib/libarchive/test/test_read_format_zip.zip.uu From des at FreeBSD.org Thu Jun 26 11:58:41 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Thu Jun 26 11:58:47 2008 Subject: cvs commit: src/lib/libarchive/test Makefile Message-ID: <200806261158.m5QBwetY085966@repoman.freebsd.org> des 2008-06-26 11:58:26 UTC FreeBSD src repository Modified files: lib/libarchive/test Makefile Log: SVN rev 180029 on 2008-06-26 11:58:26Z by des Some tests won't build at WARNS level 6 due to aliasing violations. Add missing -I. so the tests will build when ${.OBJDIR} != ${.CURDIR}. ${.OBJDIR} does not need to be spelled out. Revision Changes Path 1.23 +4 -4 src/lib/libarchive/test/Makefile From danger at FreeBSD.org Thu Jun 26 12:15:56 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Thu Jun 26 12:16:01 2008 Subject: cvs commit: src/lib/libc/sys mkdir.2 Message-ID: <200806261215.m5QCFu9f088577@repoman.freebsd.org> danger 2008-06-26 12:15:38 UTC FreeBSD src repository Modified files: lib/libc/sys mkdir.2 Log: SVN rev 180030 on 2008-06-26 12:15:38Z by danger - add description of the MLINK error PR: docs/123019 MFC after: 3 days Revision Changes Path 1.27 +3 -0 src/lib/libc/sys/mkdir.2 From jhb at FreeBSD.org Thu Jun 26 13:49:44 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 26 13:49:56 2008 Subject: cvs commit: src/sys/security/mac_test mac_test.c Message-ID: <200806261349.m5QDni1V006326@repoman.freebsd.org> jhb 2008-06-26 13:49:32 UTC FreeBSD src repository Modified files: sys/security/mac_test mac_test.c Log: SVN rev 180031 on 2008-06-26 13:49:32Z by jhb Add missing counter increments for posix shm checks. Revision Changes Path 1.97 +5 -0 src/sys/security/mac_test/mac_test.c From jhb at FreeBSD.org Thu Jun 26 13:51:45 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 26 13:51:54 2008 Subject: cvs commit: src/sys/sys semaphore.h Message-ID: <200806261351.m5QDpjda006584@repoman.freebsd.org> jhb 2008-06-26 13:51:25 UTC FreeBSD src repository Modified files: sys/sys semaphore.h Log: SVN rev 180032 on 2008-06-26 13:51:25Z by jhb Change SEM_VALUE_MAX (maximum value of a POSIX semaphore) from UINT_MAX to INT_MAX. Otherwise, a process could create a semaphore (or increase its value via ksem_post()) beyond INT_MAX and sem_getvalue() would return a negative value. sem_getvalue() is only supposed to return a negative value if that is the number of waiters for that semaphore. MFC after: 2 weeks Revision Changes Path 1.13 +3 -1 src/sys/sys/semaphore.h From kientzle at FreeBSD.org Thu Jun 26 15:46:18 2008 From: kientzle at FreeBSD.org (Tim Kientzle) Date: Thu Jun 26 15:46:25 2008 Subject: cvs commit: src/usr.bin/cpio cpio.c Message-ID: <200806261546.m5QFkHct017934@repoman.freebsd.org> kientzle 2008-06-26 15:46:01 UTC FreeBSD src repository Modified files: usr.bin/cpio cpio.c Log: SVN rev 180033 on 2008-06-26 15:46:01Z by kientzle As reported by Alexey Shuvaev, -dumpl overwrote files after linking them, with predictably bad results. Revision Changes Path 1.5 +2 -0 src/usr.bin/cpio/cpio.c From jhb at freebsd.org Thu Jun 26 17:17:09 2008 From: jhb at freebsd.org (John Baldwin) Date: Thu Jun 26 17:17:16 2008 Subject: cvs commit: src/sys/sys semaphore.h In-Reply-To: <200806261351.m5QDpjda006584@repoman.freebsd.org> References: <200806261351.m5QDpjda006584@repoman.freebsd.org> Message-ID: <200806261046.34932.jhb@freebsd.org> On Thursday 26 June 2008 09:51:25 am John Baldwin wrote: > jhb 2008-06-26 13:51:25 UTC > > FreeBSD src repository > > Modified files: > sys/sys semaphore.h > Log: > SVN rev 180032 on 2008-06-26 13:51:25Z by jhb > > Change SEM_VALUE_MAX (maximum value of a POSIX semaphore) from UINT_MAX > to INT_MAX. Otherwise, a process could create a semaphore (or increase > its value via ksem_post()) beyond INT_MAX and sem_getvalue() would return > a negative value. sem_getvalue() is only supposed to return a negative > value if that is the number of waiters for that semaphore. > > MFC after: 2 weeks Forgot to mention that other OS's I checked (glibc, Windows) use signed max constants (INT_MAX, LONG_MAX) rather than the unsigned constant. Probably for the same reason. -- John Baldwin From thompsa at FreeBSD.org Thu Jun 26 18:58:05 2008 From: thompsa at FreeBSD.org (Andrew Thompson) Date: Thu Jun 26 18:58:07 2008 Subject: cvs commit: src/sys/modules/ralfw Makefile Message-ID: <200806261858.m5QIw5jm045720@repoman.freebsd.org> thompsa 2008-06-26 18:58:01 UTC FreeBSD src repository Modified files: sys/modules/ralfw Makefile Log: SVN rev 180034 on 2008-06-26 18:58:01Z by thompsa Remove the non-existent rt2860 subdir. Note, the ralfw module is not used in the build yet. PR: kern/125015 Submitted by: Dan Cojocar Revision Changes Path 1.2 +1 -1 src/sys/modules/ralfw/Makefile From mav at FreeBSD.org Thu Jun 26 20:21:38 2008 From: mav at FreeBSD.org (Alexander Motin) Date: Thu Jun 26 20:21:41 2008 Subject: cvs commit: src/sys/netgraph ng_base.c Message-ID: <200806262021.m5QKLbvs055317@repoman.freebsd.org> mav 2008-06-26 20:21:20 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/netgraph ng_base.c Log: SVN rev 180035 on 2008-06-26 20:21:20Z by mav MFC rev. 1.155-1.156 Rewrite node's r/w/q-lock semantics using only atomics instead of mutex and atomics combination. Mutex is now used only for queue protection. Also avoid unneded extra swi scheduling calls. Add memory barriers to the node locking operations. Add some comments. Revision Changes Path 1.135.2.10 +102 -259 src/sys/netgraph/ng_base.c From jhb at FreeBSD.org Thu Jun 26 21:26:39 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 26 21:26:46 2008 Subject: cvs commit: src/tools/regression/posixshm test.c Message-ID: <200806262126.m5QLQd4Z070827@repoman.freebsd.org> jhb 2008-06-26 21:26:34 UTC FreeBSD src repository Modified files: tools/regression/posixshm test.c Log: SVN rev 180036 on 2008-06-26 21:26:34Z by jhb Fix compile on 64-bit platforms. Revision Changes Path 1.2 +1 -1 src/tools/regression/posixshm/test.c From jhb at FreeBSD.org Thu Jun 26 22:33:41 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 26 22:33:44 2008 Subject: cvs commit: src/sys/dev/mfi mfi.c Message-ID: <200806262233.m5QMXfqS077359@repoman.freebsd.org> jhb 2008-06-26 22:33:24 UTC FreeBSD src repository Modified files: sys/dev/mfi mfi.c Log: SVN rev 180037 on 2008-06-26 22:33:24Z by jhb Adjust the handling of pending log events during boot: - Fetch events from the controller in batches of 15 rather than a single event at a time. - When fetching events from the controller, honor the event class and locale settings (via hw.mfi tunables). This also allows the firmware to skip over unwanted log entries resulting in fewer requests to the controller if there many unwanted log entries since the last clean shutdown. - Don't drop the driver mutex while decoding an event. - If we get an error other than MFI_STAT_NOT_FOUND (basically EOF for hitting the end of the event log) then emit a warning and bail on processing further log entries. Reviewed by: ambrisko, scottl MFC after: 2 weeks Revision Changes Path 1.40 +75 -50 src/sys/dev/mfi/mfi.c From jhb at FreeBSD.org Thu Jun 26 22:36:42 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Thu Jun 26 22:36:44 2008 Subject: cvs commit: src/sys/dev/mfi mfi.c Message-ID: <200806262236.m5QMagU1077683@repoman.freebsd.org> jhb 2008-06-26 22:36:38 UTC FreeBSD src repository Modified files: sys/dev/mfi mfi.c Log: SVN rev 180038 on 2008-06-26 22:36:38Z by jhb Tweak the output of event log messages from the controller: - Each log entry contains a text description in the "description" field of the entry. The existing decode logic always ended up duplicating information that was already in the description string. This made the logs overly verbose. Now we just print out the description string. - Add some simple parsing of the timestamp and event classes. Reviewed by: ambrisko, scottl MFC after: 2 weeks Revision Changes Path 1.41 +48 -283 src/sys/dev/mfi/mfi.c From julian at FreeBSD.org Thu Jun 26 22:45:14 2008 From: julian at FreeBSD.org (Julian Elischer) Date: Thu Jun 26 22:45:20 2008 Subject: cvs commit: src/sys/kern kern_xxx.c Message-ID: <200806262245.m5QMjDHI078452@repoman.freebsd.org> julian 2008-06-26 22:45:04 UTC FreeBSD src repository Modified files: sys/kern kern_xxx.c Log: SVN rev 180039 on 2008-06-26 22:45:04Z by julian Someone cut and pasted a bunch of stuff here so lots of indents were spaces when they should have been tabs, screwing up diffs and patches.. Whitespace commit as my first SVN commit. (yay) MFC after: 1 week Revision Changes Path 1.50 +15 -15 src/sys/kern/kern_xxx.c From rodrigc at FreeBSD.org Thu Jun 26 22:59:43 2008 From: rodrigc at FreeBSD.org (Craig Rodrigues) Date: Thu Jun 26 22:59:46 2008 Subject: cvs commit: src/sys/kern vfs_mount.c src/sys/ufs/ffs ffs_vfsops.c Message-ID: <200806262259.m5QMxhFT079837@repoman.freebsd.org> rodrigc 2008-06-26 22:58:20 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern vfs_mount.c sys/ufs/ffs ffs_vfsops.c Log: SVN rev 180040 on 2008-06-26 22:58:20Z by rodrigc ffs_vfsops.c: MFC 179269-179270 - delete "snapshot" from options list after we convert mount option to MNT_SNAPSHOT flag - do not perform string to flag conversions for mount option which are converted further up in vfs_donmount() in vfs_mount.c vfs_mount.c: MFC 179268 - Do not convert the "snapshot" string to the MNT_SNAPSHOT flag here, since we do it further down in ffs_vfsops.c PR: 122833 Revision Changes Path 1.265.2.5 +0 -2 src/sys/kern/vfs_mount.c 1.329.2.4 +8 -22 src/sys/ufs/ffs/ffs_vfsops.c From julian at FreeBSD.org Thu Jun 26 23:00:02 2008 From: julian at FreeBSD.org (Julian Elischer) Date: Thu Jun 26 23:00:08 2008 Subject: cvs commit: src/sys/net if_gre.c Message-ID: <200806262300.m5QN00LE079899@repoman.freebsd.org> julian 2008-06-26 22:59:49 UTC FreeBSD src repository Modified files: sys/net if_gre.c Log: SVN rev 180041 on 2008-06-26 22:59:49Z by julian change a variable name ot stop it from colliding with other names in some situations. (i.e. in vimage) MFC after: 1 week Revision Changes Path 1.50 +7 -7 src/sys/net/if_gre.c From rwatson at FreeBSD.org Thu Jun 26 23:05:32 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Thu Jun 26 23:05:37 2008 Subject: cvs commit: src/sys/net if.c if_var.h Message-ID: <200806262305.m5QN5VYF081649@repoman.freebsd.org> rwatson 2008-06-26 23:05:28 UTC FreeBSD src repository Modified files: sys/net if.c if_var.h Log: SVN rev 180042 on 2008-06-26 23:05:28Z by rwatson Introduce locking around use of ifindex_table, whose use was previously unsynchronized. While races were extremely rare, we've now had a couple of reports of panics in environments involving large numbers of IPSEC tunnels being added very quickly on an active system. - Add accessor functions ifnet_byindex(), ifaddr_byindex(), ifdev_byindex() to replace existing accessor macros. These functions now acquire the ifnet lock before derefencing the table. - Add IFNET_WLOCK_ASSERT(). - Add static accessor functions ifnet_setbyindex(), ifdev_setbyindex(), which set values in the table either asserting of acquiring the ifnet lock. - Use accessor functions throughout if.c to modify and read ifindex_table. - Rework ifnet attach/detach to lock around ifindex_table modification. Note that these changes simply close races around use of ifindex_table, and make no attempt to solve the probem of disappearing ifnets. Further refinement of this work, including with respect to ifindex_table resizing, is still required. In a future change, the ifnet lock should be converted from a mutex to an rwlock in order to reduce contention. Reviewed and tested by: brooks Revision Changes Path 1.281 +69 -12 src/sys/net/if.c 1.119 +4 -4 src/sys/net/if_var.h From jhb at FreeBSD.org Fri Jun 27 00:37:56 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 00:38:03 2008 Subject: cvs commit: src/sys/kern vfs_mount.c src/sys/ufs/ffs ffs_vfsops.c Message-ID: <200806270037.m5R0buuR090129@repoman.freebsd.org> jhb 2008-06-27 00:37:08 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/kern vfs_mount.c sys/ufs/ffs ffs_vfsops.c Log: SVN rev 180043 on 2008-06-27 00:37:08Z by jhb Consolidate merge info in sys/. Revision Changes Path 1.265.2.6 +0 -0 src/sys/kern/vfs_mount.c 1.329.2.5 +0 -0 src/sys/ufs/ffs/ffs_vfsops.c From jhb at FreeBSD.org Fri Jun 27 00:48:42 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 00:48:43 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200806270048.m5R0mg4D000334@repoman.freebsd.org> jhb 2008-06-27 00:48:23 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/i386/cpufreq est.c Log: SVN rev 180044 on 2008-06-27 00:48:23Z by jhb MFC: After probing the available frequency settings, restore the CPU to run at whatever frequency it started at. Revision Changes Path 1.11.2.5 +4 -18 src/sys/i386/cpufreq/est.c From jhb at FreeBSD.org Fri Jun 27 00:56:43 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 00:56:50 2008 Subject: cvs commit: src/sys/i386/cpufreq est.c Message-ID: <200806270056.m5R0ugGl001012@repoman.freebsd.org> jhb 2008-06-27 00:56:36 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/i386/cpufreq est.c Log: SVN rev 180045 on 2008-06-27 00:56:36Z by jhb MFC: After probing the available frequency settings, restore the CPU to run at whatever frequency it started at. Revision Changes Path 1.7.2.5 +3 -3 src/sys/i386/cpufreq/est.c From jhb at FreeBSD.org Fri Jun 27 00:58:43 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 00:58:50 2008 Subject: cvs commit: src/sys/boot/i386/libi386 biossmap.c Message-ID: <200806270058.m5R0whOm001164@repoman.freebsd.org> jhb 2008-06-27 00:58:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/boot/i386/libi386 biossmap.c Log: SVN rev 180046 on 2008-06-27 00:58:39Z by jhb MFC: Workaround a bug in the BIOS of Dell R900 machines by padding the structure used to fetch SMAP entries. Revision Changes Path 1.6.2.2 +4 -1 src/sys/boot/i386/libi386/biossmap.c From jhb at FreeBSD.org Fri Jun 27 01:00:16 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 01:00:19 2008 Subject: cvs commit: src/sys/boot/i386/libi386 biossmap.c Message-ID: <200806270100.m5R10EG4001354@repoman.freebsd.org> jhb 2008-06-27 01:00:08 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/boot/i386/libi386 biossmap.c Log: SVN rev 180047 on 2008-06-27 01:00:08Z by jhb MFC: Workaround a bug in the BIOS of Dell R900 machines by padding the structure used to fetch SMAP entries. Revision Changes Path 1.3.2.4 +4 -1 src/sys/boot/i386/libi386/biossmap.c From jhb at FreeBSD.org Fri Jun 27 02:18:17 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 02:18:24 2008 Subject: cvs commit: src/libexec/rpc.rstatd Makefile rstat_proc.c Message-ID: <200806270218.m5R2IGQ7009868@repoman.freebsd.org> jhb 2008-06-27 02:18:01 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) libexec/rpc.rstatd Makefile rstat_proc.c Log: SVN rev 180048 on 2008-06-27 02:18:01Z by jhb MFC: Use sysctl to fetch stats from the kernel instead of reading variables directly via libkvm. Revision Changes Path 1.8.18.1 +2 -2 src/libexec/rpc.rstatd/Makefile 1.25.20.1 +17 -46 src/libexec/rpc.rstatd/rstat_proc.c From jhb at FreeBSD.org Fri Jun 27 02:20:48 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 02:20:50 2008 Subject: cvs commit: src/libexec/rpc.rstatd Makefile rstat_proc.c Message-ID: <200806270220.m5R2KldI010139@repoman.freebsd.org> jhb 2008-06-27 02:20:37 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) libexec/rpc.rstatd Makefile rstat_proc.c Log: SVN rev 180049 on 2008-06-27 02:20:37Z by jhb MFC: Use sysctl to fetch stats from the kernel instead of reading variables directly via libkvm. Revision Changes Path 1.8.8.1 +2 -2 src/libexec/rpc.rstatd/Makefile 1.25.10.1 +17 -46 src/libexec/rpc.rstatd/rstat_proc.c From jhb at FreeBSD.org Fri Jun 27 03:03:44 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 03:03:53 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h src/sys/dev/mii brgphy.c Message-ID: <200806270303.m5R33iJl014860@repoman.freebsd.org> jhb 2008-06-27 03:02:50 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/bge if_bge.c if_bgereg.h sys/dev/mii brgphy.c Log: SVN rev 180050 on 2008-06-27 03:02:50Z by jhb MFC: Add a flag for Ethernet@WireSpeed capability and correct chip revisions. Revision Changes Path 1.198.2.8 +9 -0 src/sys/dev/bge/if_bge.c 1.73.2.3 +2 -1 src/sys/dev/bge/if_bgereg.h 1.70.2.2 +1 -7 src/sys/dev/mii/brgphy.c From jhb at FreeBSD.org Fri Jun 27 03:11:46 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 03:11:48 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h Message-ID: <200806270311.m5R3BkUE015548@repoman.freebsd.org> jhb 2008-06-27 03:11:34 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/bge if_bge.c if_bgereg.h Log: SVN rev 180051 on 2008-06-27 03:11:34Z by jhb MFC: - Add support/quirks for the on-board BGEs found in various Sun machines. - In bge_attach() factor out calling bge_release_resources() before going to the fail label. Revision Changes Path 1.91.2.28 +92 -34 src/sys/dev/bge/if_bge.c 1.36.2.12 +6 -0 src/sys/dev/bge/if_bgereg.h From jhb at FreeBSD.org Fri Jun 27 03:19:28 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 03:19:36 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h src/sys/dev/mii brgphy.c Message-ID: <200806270319.m5R3JSak016164@repoman.freebsd.org> jhb 2008-06-27 03:19:03 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/bge if_bge.c if_bgereg.h sys/dev/mii brgphy.c Log: SVN rev 180052 on 2008-06-27 03:19:03Z by jhb MFC: Add a flag for Ethernet@WireSpeed capability and correct chip revisions. Revision Changes Path 1.91.2.29 +9 -0 src/sys/dev/bge/if_bge.c 1.36.2.13 +2 -1 src/sys/dev/bge/if_bgereg.h 1.34.2.17 +1 -7 src/sys/dev/mii/brgphy.c From jhb at FreeBSD.org Fri Jun 27 03:25:41 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 03:25:48 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h src/sys/dev/mii brgphy.c brgphyreg.h miidevs Message-ID: <200806270325.m5R3Pfdw016756@repoman.freebsd.org> jhb 2008-06-27 03:24:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/dev/bge if_bge.c if_bgereg.h sys/dev/mii brgphy.c brgphyreg.h miidevs Log: SVN rev 180053 on 2008-06-27 03:24:54Z by jhb MFC: Add support for the BCM5906[M] adapters including subsequent improvements by marius@. Revision Changes Path 1.198.2.9 +302 -95 src/sys/dev/bge/if_bge.c 1.73.2.4 +69 -1 src/sys/dev/bge/if_bgereg.h 1.70.2.3 +11 -2 src/sys/dev/mii/brgphy.c 1.10.2.1 +1 -0 src/sys/dev/mii/brgphyreg.h 1.46.2.6 +2 -0 src/sys/dev/mii/miidevs From jhb at FreeBSD.org Fri Jun 27 03:31:12 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 03:31:14 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c Message-ID: <200806270331.m5R3VCjI017188@repoman.freebsd.org> jhb 2008-06-27 03:31:05 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/bge if_bge.c Log: SVN rev 180054 on 2008-06-27 03:31:05Z by jhb MFC: Fix time out check in EEPROM read and move delays to give some settle time. Revision Changes Path 1.91.2.30 +12 -9 src/sys/dev/bge/if_bge.c From jhb at FreeBSD.org Fri Jun 27 03:36:07 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 03:36:08 2008 Subject: cvs commit: src/sys/dev/bge if_bge.c if_bgereg.h src/sys/dev/mii brgphy.c brgphyreg.h miidevs Message-ID: <200806270336.m5R3a72C017649@repoman.freebsd.org> jhb 2008-06-27 03:35:48 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/bge if_bge.c if_bgereg.h sys/dev/mii brgphy.c brgphyreg.h miidevs Log: SVN rev 180055 on 2008-06-27 03:35:48Z by jhb MFC: Add support for the BCM5906[M] adapters including subsequent improvements by marius@. Revision Changes Path 1.91.2.31 +302 -95 src/sys/dev/bge/if_bge.c 1.36.2.14 +69 -1 src/sys/dev/bge/if_bgereg.h 1.34.2.18 +11 -2 src/sys/dev/mii/brgphy.c 1.6.2.3 +1 -0 src/sys/dev/mii/brgphyreg.h 1.30.2.16 +2 -0 src/sys/dev/mii/miidevs From jhb at FreeBSD.org Fri Jun 27 03:51:00 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 03:51:02 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: <200806270351.m5R3p0iZ023807@repoman.freebsd.org> jhb 2008-06-27 03:49:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/amd64/amd64 legacy.c sys/dev/acpica acpi.c sys/dev/cpufreq ichss.c sys/i386/i386 legacy.c Log: SVN rev 180056 on 2008-06-27 03:49:22Z by jhb MFC: 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. Revision Changes Path 1.61.2.1 +23 -12 src/sys/amd64/amd64/legacy.c 1.243.2.2 +22 -6 src/sys/dev/acpica/acpi.c 1.10.2.1 +50 -61 src/sys/dev/cpufreq/ichss.c 1.63.2.1 +23 -12 src/sys/i386/i386/legacy.c From jhb at FreeBSD.org Fri Jun 27 04:46:05 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 04:46:08 2008 Subject: cvs commit: src/etc/mtree BSD.include.dist src/etc/rc.d initrandom src/gnu/usr.bin/groff/tmac mdoc.local src/include Makefile src/release/scripts src-install.sh src/sys/conf files src/sys/dev/mpt mpt.h mpt_user.c src/sys/modules/mpt Makefile ... Message-ID: <200806270446.m5R4k5MM029526@repoman.freebsd.org> jhb 2008-06-27 04:42:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) etc/mtree BSD.include.dist etc/rc.d initrandom gnu/usr.bin/groff/tmac mdoc.local include Makefile release/scripts src-install.sh sys/conf files sys/dev/mpt mpt.h sys/modules/mpt Makefile Added files: (Branch: RELENG_7) sys/dev/mpt mpt_user.c sys/sys mpt_ioctl.h Log: SVN rev 180057 on 2008-06-27 04:42:19Z by jhb MFC: Add the mpt_user personality to mpt(4). Revision Changes Path 1.120.2.1 +4 -0 src/etc/mtree/BSD.include.dist 1.7.2.3 +0 -0 src/etc/rc.d/initrandom 1.59.2.4 +0 -0 src/gnu/usr.bin/groff/tmac/mdoc.local 1.276.2.1 +3 -1 src/include/Makefile 1.11.2.3 +0 -0 src/release/scripts/src-install.sh 1.1243.2.24 +1 -0 src/sys/conf/files 1.42.2.1 +3 -0 src/sys/dev/mpt/mpt.h 1.1.2.1 +755 -0 src/sys/dev/mpt/mpt_user.c (new) 1.2.10.1 +1 -1 src/sys/modules/mpt/Makefile 1.1.2.1 +132 -0 src/sys/sys/mpt_ioctl.h (new) From jhb at FreeBSD.org Fri Jun 27 04:50:44 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 04:50:51 2008 Subject: cvs commit: src/etc/mtree BSD.include.dist src/include Makefile src/sys/conf files src/sys/dev/mpt mpt.h mpt_user.c src/sys/modules/mpt Makefile src/sys/sys mpt_ioctl.h src/usr.bin/tar COPYING Makefile bsdtar.1 bsdtar.c bsdtar.h ... Message-ID: <200806270450.m5R4oiLN039338@repoman.freebsd.org> jhb 2008-06-27 04:46:06 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) etc/mtree BSD.include.dist include Makefile sys/conf files sys/dev/mpt mpt.h sys/modules/mpt Makefile usr.bin/tar COPYING Makefile bsdtar.1 bsdtar.c bsdtar.h bsdtar_platform.h config_freebsd.h getdate.y matching.c read.c tree.c tree.h util.c write.c Added files: (Branch: RELENG_6) sys/dev/mpt mpt_user.c sys/sys mpt_ioctl.h Log: SVN rev 180058 on 2008-06-27 04:46:06Z by jhb MFC: Add the mpt_user personality to mpt(4). Revision Changes Path 1.100.2.5 +4 -0 src/etc/mtree/BSD.include.dist 1.244.2.7 +3 -1 src/include/Makefile 1.1031.2.74 +1 -0 src/sys/conf/files 1.6.2.8 +3 -0 src/sys/dev/mpt/mpt.h 1.1.4.1 +755 -0 src/sys/dev/mpt/mpt_user.c (new) 1.2.2.1 +1 -1 src/sys/modules/mpt/Makefile 1.1.4.1 +132 -0 src/sys/sys/mpt_ioctl.h (new) 1.1.8.3 +0 -0 src/usr.bin/tar/COPYING 1.24.2.6 +0 -0 src/usr.bin/tar/Makefile 1.29.2.6 +0 -0 src/usr.bin/tar/bsdtar.1 1.63.2.14 +0 -0 src/usr.bin/tar/bsdtar.c 1.23.2.6 +0 -0 src/usr.bin/tar/bsdtar.h 1.15.2.11 +0 -0 src/usr.bin/tar/bsdtar_platform.h 1.1.2.2 +0 -0 src/usr.bin/tar/config_freebsd.h 1.4.2.4 +0 -0 src/usr.bin/tar/getdate.y 1.9.2.4 +0 -0 src/usr.bin/tar/matching.c 1.23.2.9 +0 -0 src/usr.bin/tar/read.c 1.2.2.4 +0 -0 src/usr.bin/tar/tree.c 1.2.2.2 +0 -0 src/usr.bin/tar/tree.h 1.13.2.6 +0 -0 src/usr.bin/tar/util.c 1.41.2.13 +0 -0 src/usr.bin/tar/write.c From jhb at FreeBSD.org Fri Jun 27 05:45:11 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Fri Jun 27 05:45:20 2008 Subject: cvs commit: src/sys/kern kern_descrip.c uipc_sem.c uipc_shm.c src/sys/modules/sem Makefile src/sys/security/mac mac_framework.h mac_policy.h mac_posix_sem.c src/sys/security/mac_biba mac_biba.c src/sys/security/mac_mls mac_mls.c src/sys/security/mac_stub ... Message-ID: <200806270545.m5R5jBxK044999@repoman.freebsd.org> jhb 2008-06-27 05:39:04 UTC FreeBSD src repository Modified files: sys/kern kern_descrip.c uipc_sem.c uipc_shm.c sys/modules/sem Makefile sys/security/mac mac_framework.h mac_policy.h mac_posix_sem.c sys/security/mac_biba mac_biba.c sys/security/mac_mls mac_mls.c sys/security/mac_stub mac_stub.c sys/security/mac_test mac_test.c sys/sys file.h ksem.h user.h usr.bin/procstat procstat_files.c Added files: tools/regression/posixsem Makefile posixsem.c posixsem.t test.c test.h Log: SVN rev 180059 on 2008-06-27 05:39:04Z by jhb Rework the lifetime management of the kernel implementation of POSIX semaphores. Specifically, semaphores are now represented as new file descriptor type that is set to close on exec. This removes the need for all of the manual process reference counting (and fork, exec, and exit event handlers) as the normal file descriptor operations handle all of that for us nicely. It is also suggested as one possible implementation in the spec and at least one other OS (OS X) uses this approach. Some bugs that were fixed as a result include: - References to a named semaphore whose name is removed still work after the sem_unlink() operation. Prior to this patch, if a semaphore's name was removed, valid handles from sem_open() would get EINVAL errors from sem_getvalue(), sem_post(), etc. This fixes that. - Unnamed semaphores created with sem_init() were not cleaned up when a process exited or exec'd. They were only cleaned up if the process did an explicit sem_destroy(). This could result in a leak of semaphore objects that could never be cleaned up. - On the other hand, if another process guessed the id (kernel pointer to 'struct ksem' of an unnamed semaphore (created via sem_init)) and had write access to the semaphore based on UID/GID checks, then that other process could manipulate the semaphore via sem_destroy(), sem_post(), sem_wait(), etc. - As part of the permission check (UID/GID), the umask of the proces creating the semaphore was not honored. Thus if your umask denied group read/write access but the explicit mode in the sem_init() call allowed it, the semaphore would be readable/writable by other users in the same group, for example. This includes access via the previous bug. - If the module refused to unload because there were active semaphores, then it might have deregistered one or more of the semaphore system calls before it noticed that there was a problem. I'm not sure if this actually happened as the order that modules are discovered by the kernel linker depends on how the actual .ko file is linked. One can make the order deterministic by using a single module with a mod_event handler that explicitly registers syscalls (and deregisters during unload after any checks). This also fixes a race where even if the sem_module unloaded first it would have destroyed locks that the syscalls might be trying to access if they are still executing when they are unloaded. XXX: By the way, deregistering system calls doesn't do any blocking to drain any threads from the calls. - Some minor fixes to errno values on error. For example, sem_init() isn't documented to return ENFILE or EMFILE if we run out of semaphores the way that sem_open() can. Instead, it should return ENOSPC in that case. Other changes: - Kernel semaphores now use a hash table to manage the namespace of named semaphores nearly in a similar fashion to the POSIX shared memory object file descriptors. Kernel semaphores can now also have names longer than 14 chars (up to MAXPATHLEN) and can include subdirectories in their pathname. - The UID/GID permission checks for access to a named semaphore are now done via vaccess() rather than a home-rolled set of checks. - Now that kernel semaphores have an associated file object, the various MAC checks for POSIX semaphores accept both a file credential and an active credential. There is also a new posixsem_check_stat() since it is possible to fstat() a semaphore file descriptor. - A small set of regression tests (using the ksem API directly) is present in src/tools/regression/posixsem. Reported by: kris (1) Tested by: kris Reviewed by: rwatson (lightly) MFC after: 1 month Revision Changes Path 1.335 +6 -0 src/sys/kern/kern_descrip.c 1.34 +563 -625 src/sys/kern/uipc_sem.c 1.5 +4 -0 src/sys/kern/uipc_shm.c 1.3 +1 -1 src/sys/modules/sem/Makefile 1.100 +8 -3 src/sys/security/mac/mac_framework.h 1.110 +13 -6 src/sys/security/mac/mac_policy.h 1.14 +24 -6 src/sys/security/mac/mac_posix_sem.c 1.120 +25 -6 src/sys/security/mac_biba/mac_biba.c 1.102 +25 -6 src/sys/security/mac_mls/mac_mls.c 1.83 +15 -6 src/sys/security/mac_stub/mac_stub.c 1.98 +26 -9 src/sys/security/mac_test/mac_test.c 1.80 +1 -0 src/sys/sys/file.h 1.4 +24 -18 src/sys/sys/ksem.h 1.76 +1 -0 src/sys/sys/user.h 1.1 +11 -0 src/tools/regression/posixsem/Makefile (new) 1.1 +1437 -0 src/tools/regression/posixsem/posixsem.c (new) 1.1 +5 -0 src/tools/regression/posixsem/posixsem.t (new) 1.1 +128 -0 src/tools/regression/posixsem/test.c (new) 1.1 +59 -0 src/tools/regression/posixsem/test.h (new) 1.6 +4 -0 src/usr.bin/procstat/procstat_files.c From richardtector at thekeelecentre.com Fri Jun 27 07:48:31 2008 From: richardtector at thekeelecentre.com (Richard Tector) Date: Fri Jun 27 07:48:35 2008 Subject: cvs commit: src/etc/mtree BSD.include.dist src/include Makefile src/sys/conf files src/sys/dev/mpt mpt.h mpt_user.c src/sys/modules/mpt Makefile src/sys/sys mpt_ioctl.h src/usr.bin/tar COPYING Makefile bsdtar.1 bsdtar.c bsdtar.h ... In-Reply-To: <200806270450.m5R4oiLN039338@repoman.freebsd.org> References: <200806270450.m5R4oiLN039338@repoman.freebsd.org> Message-ID: <48649766.605@thekeelecentre.com> John Baldwin wrote: > jhb 2008-06-27 04:46:06 UTC > > FreeBSD src repository > > Modified files: (Branch: RELENG_6) > etc/mtree BSD.include.dist > include Makefile > sys/conf files > sys/dev/mpt mpt.h > sys/modules/mpt Makefile > usr.bin/tar COPYING Makefile bsdtar.1 bsdtar.c > bsdtar.h bsdtar_platform.h > config_freebsd.h getdate.y matching.c > read.c tree.c tree.h util.c write.c > Added files: (Branch: RELENG_6) > sys/dev/mpt mpt_user.c > sys/sys mpt_ioctl.h > Log: > SVN rev 180058 on 2008-06-27 04:46:06Z by jhb > > MFC: Add the mpt_user personality to mpt(4). > > Revision Changes Path > 1.100.2.5 +4 -0 src/etc/mtree/BSD.include.dist > 1.244.2.7 +3 -1 src/include/Makefile > 1.1031.2.74 +1 -0 src/sys/conf/files > 1.6.2.8 +3 -0 src/sys/dev/mpt/mpt.h > 1.1.4.1 +755 -0 src/sys/dev/mpt/mpt_user.c (new) > 1.2.2.1 +1 -1 src/sys/modules/mpt/Makefile > 1.1.4.1 +132 -0 src/sys/sys/mpt_ioctl.h (new) > 1.1.8.3 +0 -0 src/usr.bin/tar/COPYING > 1.24.2.6 +0 -0 src/usr.bin/tar/Makefile > 1.29.2.6 +0 -0 src/usr.bin/tar/bsdtar.1 > 1.63.2.14 +0 -0 src/usr.bin/tar/bsdtar.c > 1.23.2.6 +0 -0 src/usr.bin/tar/bsdtar.h > 1.15.2.11 +0 -0 src/usr.bin/tar/bsdtar_platform.h > 1.1.2.2 +0 -0 src/usr.bin/tar/config_freebsd.h > 1.4.2.4 +0 -0 src/usr.bin/tar/getdate.y > 1.9.2.4 +0 -0 src/usr.bin/tar/matching.c > 1.23.2.9 +0 -0 src/usr.bin/tar/read.c > 1.2.2.4 +0 -0 src/usr.bin/tar/tree.c > 1.2.2.2 +0 -0 src/usr.bin/tar/tree.h > 1.13.2.6 +0 -0 src/usr.bin/tar/util.c > 1.41.2.13 +0 -0 src/usr.bin/tar/write.c Why all the no-op commits to tar? Is this another quirk of SVN? Regards, Richard -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 2709 bytes Desc: S/MIME Cryptographic Signature Url : http://lists.freebsd.org/pipermail/cvs-src/attachments/20080627/b909c7d6/smime.bin From mtm at FreeBSD.org Fri Jun 27 09:09:59 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Fri Jun 27 09:10:00 2008 Subject: cvs commit: src/sbin/devfs devfs.8 Message-ID: <200806270909.m5R99wXw075901@repoman.freebsd.org> mtm 2008-06-27 09:09:50 UTC FreeBSD src repository Modified files: sbin/devfs devfs.8 Log: SVN rev 180060 on 2008-06-27 09:09:50Z by mtm Support for filtering on major device number was removed in rev. 1.7 or rule.c. Update man page example accordingly. Submitted by: Mateusz Guzik PR: docs/124892 Revision Changes Path 1.19 +2 -2 src/sbin/devfs/devfs.8 From luigi at FreeBSD.org Fri Jun 27 10:30:20 2008 From: luigi at FreeBSD.org (Luigi Rizzo) Date: Fri Jun 27 10:30:22 2008 Subject: cvs commit: src/sys/dev/usb usbdevs uscanner.c Message-ID: <200806271030.m5RAUFw0084885@repoman.freebsd.org> luigi 2008-06-27 10:30:08 UTC FreeBSD src repository Modified files: sys/dev/usb usbdevs uscanner.c Log: SVN rev 180061 on 2008-06-27 10:30:08Z by luigi Add USB ids for the Epson DX7400 / CX7300 multifunction scanner printer card reader. Tested operation of the scanner part with Sane, card and printer correctly recognised as /dev/da0 and /dev/ulpt0 MFC after: 3 days Revision Changes Path 1.361 +1 -0 src/sys/dev/usb/usbdevs 1.93 +1 -0 src/sys/dev/usb/uscanner.c From rpaulo at FreeBSD.org Fri Jun 27 12:08:53 2008 From: rpaulo at FreeBSD.org (Rui Paulo) Date: Fri Jun 27 12:09:00 2008 Subject: cvs commit: src/etc devd.conf src/sys/dev/acpi_support acpi_asus.c Message-ID: <200806271208.m5RC8rBa013061@repoman.freebsd.org> rpaulo 2008-06-27 12:04:36 UTC FreeBSD src repository Modified files: etc devd.conf sys/dev/acpi_support acpi_asus.c Log: SVN rev 180062 on 2008-06-27 12:04:36Z by rpaulo Add the missing support for Asus Eee PC in acpi_asus(4). This includes hotkeys support and sysctl variables to control camera and card reader. These new sysctls don't have CTFLAG_ANYBODY set. While there add entries to devd.conf related to the Eee volume keys. Reviewed by: phillip MFC after: 1 week Also tested by: lme (previous version) Revision Changes Path 1.42 +22 -0 src/etc/devd.conf 1.34 +109 -6 src/sys/dev/acpi_support/acpi_asus.c From jhb at freebsd.org Fri Jun 27 12:20:27 2008 From: jhb at freebsd.org (John Baldwin) Date: Fri Jun 27 12:20:38 2008 Subject: cvs commit: src/etc/mtree BSD.include.dist src/include Makefile src/sys/conf files src/sys/dev/mpt mpt.h mpt_user.c src/sys/modules/mpt Makefile src/sys/sys mpt_ioctl.h src/usr.bin/tar COPYING Makefile bsdtar.1 bsdtar.c bsdtar.h ... In-Reply-To: <48649766.605@thekeelecentre.com> References: <200806270450.m5R4oiLN039338@repoman.freebsd.org> <48649766.605@thekeelecentre.com> Message-ID: <200806270812.11551.jhb@freebsd.org> On Friday 27 June 2008 03:31:50 am Richard Tector wrote: > John Baldwin wrote: > > jhb 2008-06-27 04:46:06 UTC > > > > FreeBSD src repository > > > > Modified files: (Branch: RELENG_6) > > etc/mtree BSD.include.dist > > include Makefile > > sys/conf files > > sys/dev/mpt mpt.h > > sys/modules/mpt Makefile > > usr.bin/tar COPYING Makefile bsdtar.1 bsdtar.c > > bsdtar.h bsdtar_platform.h > > config_freebsd.h getdate.y matching.c > > read.c tree.c tree.h util.c write.c > > Added files: (Branch: RELENG_6) > > sys/dev/mpt mpt_user.c > > sys/sys mpt_ioctl.h > > Log: > > SVN rev 180058 on 2008-06-27 04:46:06Z by jhb > > > > MFC: Add the mpt_user personality to mpt(4). > > > > Revision Changes Path > > 1.100.2.5 +4 -0 src/etc/mtree/BSD.include.dist > > 1.244.2.7 +3 -1 src/include/Makefile > > 1.1031.2.74 +1 -0 src/sys/conf/files > > 1.6.2.8 +3 -0 src/sys/dev/mpt/mpt.h > > 1.1.4.1 +755 -0 src/sys/dev/mpt/mpt_user.c (new) > > 1.2.2.1 +1 -1 src/sys/modules/mpt/Makefile > > 1.1.4.1 +132 -0 src/sys/sys/mpt_ioctl.h (new) > > 1.1.8.3 +0 -0 src/usr.bin/tar/COPYING > > 1.24.2.6 +0 -0 src/usr.bin/tar/Makefile > > 1.29.2.6 +0 -0 src/usr.bin/tar/bsdtar.1 > > 1.63.2.14 +0 -0 src/usr.bin/tar/bsdtar.c > > 1.23.2.6 +0 -0 src/usr.bin/tar/bsdtar.h > > 1.15.2.11 +0 -0 src/usr.bin/tar/bsdtar_platform.h > > 1.1.2.2 +0 -0 src/usr.bin/tar/config_freebsd.h > > 1.4.2.4 +0 -0 src/usr.bin/tar/getdate.y > > 1.9.2.4 +0 -0 src/usr.bin/tar/matching.c > > 1.23.2.9 +0 -0 src/usr.bin/tar/read.c > > 1.2.2.4 +0 -0 src/usr.bin/tar/tree.c > > 1.2.2.2 +0 -0 src/usr.bin/tar/tree.h > > 1.13.2.6 +0 -0 src/usr.bin/tar/util.c > > 1.41.2.13 +0 -0 src/usr.bin/tar/write.c > > > Why all the no-op commits to tar? Is this another quirk of SVN? Yep. Someone merged some changes into the actual tar files rather than the tar/ directory. I think peter@ is going to fix the SVN -> CVS exporter to ignore these changes in CVS commits though. -- John Baldwin From obrien at FreeBSD.org Fri Jun 27 14:26:52 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Fri Jun 27 14:26:59 2008 Subject: cvs commit: src/usr.bin/make globals.h hash_tables.c main.c make.1 parse.c Message-ID: <200806271426.m5REQqLO035938@repoman.freebsd.org> obrien 2008-06-27 14:26:43 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.bin/make globals.h hash_tables.c main.c make.1 parse.c Log: SVN rev 180063 on 2008-06-27 14:26:43Z by obrien MFC: r177101, r177102, r177541: If the special target .MAKEFILEDEPS exists, then enable the "remaking makefiles" feature. Otherwise, follow traditional Pmake behavior. Revision Changes Path 1.11.2.1 +1 -0 src/usr.bin/make/globals.h 1.3.2.1 +26 -26 src/usr.bin/make/hash_tables.c 1.163.2.2 +2 -1 src/usr.bin/make/main.c 1.103.2.1 +23 -2 src/usr.bin/make/make.1 1.113.2.1 +5 -0 src/usr.bin/make/parse.c From dfr at FreeBSD.org Fri Jun 27 14:35:23 2008 From: dfr at FreeBSD.org (Doug Rabson) Date: Fri Jun 27 14:35:29 2008 Subject: cvs commit: src/sys/rpc auth_unix.c Message-ID: <200806271435.m5REZNH2036655@repoman.freebsd.org> dfr 2008-06-27 14:35:05 UTC FreeBSD src repository Modified files: sys/rpc auth_unix.c Log: SVN rev 180064 on 2008-06-27 14:35:05Z by dfr Include for curthread. Revision Changes Path 1.4 +1 -0 src/sys/rpc/auth_unix.c From obrien at FreeBSD.org Fri Jun 27 14:35:40 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Fri Jun 27 14:35:45 2008 Subject: cvs commit: src/usr.bin/make make.1 Message-ID: <200806271435.m5REZdcP036722@repoman.freebsd.org> obrien 2008-06-27 14:35:33 UTC FreeBSD src repository Modified files: usr.bin/make make.1 Log: SVN rev 180065 on 2008-06-27 14:35:33Z by obrien Note that the .POSIX special target disables the "Remaking Makefiles" feature. Revision Changes Path 1.111 +3 -0 src/usr.bin/make/make.1 From emaste at FreeBSD.org Fri Jun 27 14:37:40 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Fri Jun 27 14:37:47 2008 Subject: cvs commit: src/usr.sbin/config main.c Message-ID: <200806271437.m5REbe2N036889@repoman.freebsd.org> emaste 2008-06-27 14:37:32 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) usr.sbin/config main.c Log: SVN rev 180066 on 2008-06-27 14:37:32Z by emaste MFC r175163: Honour the logical current working directory ($PWD) when using config's -d destdir option. For an automounted src tree using the logical cwd in the Makefile keeps amd(8)'s mount timeout refreshed. Code to check $PWD's validity cribbed from pwd(1). Discussed on hackers@. Revision Changes Path 1.64.2.2 +18 -0 src/usr.sbin/config/main.c From obrien at FreeBSD.org Fri Jun 27 14:59:42 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Fri Jun 27 14:59:45 2008 Subject: cvs commit: src/gnu/usr.bin/cvs/contrib Makefile Message-ID: <200806271459.m5RExftq038676@repoman.freebsd.org> obrien 2008-06-27 14:59:23 UTC FreeBSD src repository Modified files: gnu/usr.bin/cvs/contrib Makefile Log: SVN rev 180067 on 2008-06-27 14:59:23Z by obrien Instead of using .POSIX:, use the cool looking "@:" rule. Submitted by: ru Revision Changes Path 1.21 +4 -2 src/gnu/usr.bin/cvs/contrib/Makefile From emaste at FreeBSD.org Fri Jun 27 15:20:28 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Fri Jun 27 15:20:30 2008 Subject: cvs commit: src/usr.sbin/config main.c Message-ID: <200806271520.m5RFKR0F041621@repoman.freebsd.org> emaste 2008-06-27 15:20:18 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) usr.sbin/config main.c Log: SVN rev 180068 on 2008-06-27 15:20:18Z by emaste MFC r175163: Honour the logical current working directory ($PWD) when using config's -d destdir option. For an automounted src tree using the logical cwd in the Makefile keeps amd(8)'s mount timeout refreshed. Code to check $PWD's validity cribbed from pwd(1). Discussed on hackers@. Revision Changes Path 1.76.2.1 +18 -0 src/usr.sbin/config/main.c From avatar at FreeBSD.org Fri Jun 27 15:29:58 2008 From: avatar at FreeBSD.org (Tai-hwa Liang) Date: Fri Jun 27 15:30:01 2008 Subject: cvs commit: src/sys/nlm nlm_prot_impl.c Message-ID: <200806271529.m5RFTwEO042320@repoman.freebsd.org> avatar 2008-06-27 15:29:48 UTC FreeBSD src repository Modified files: sys/nlm nlm_prot_impl.c Log: SVN rev 180069 on 2008-06-27 15:29:48Z by avatar Fixing NO_INET6 build. Revision Changes Path 1.11 +4 -1 src/sys/nlm/nlm_prot_impl.c From mtm at FreeBSD.org Fri Jun 27 15:45:30 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Fri Jun 27 15:45:40 2008 Subject: cvs commit: src/etc/rc.d lockd statd Message-ID: <200806271545.m5RFjTva043716@repoman.freebsd.org> mtm 2008-06-27 15:45:17 UTC FreeBSD src repository Modified files: etc/rc.d lockd statd Log: SVN rev 180070 on 2008-06-27 15:45:17Z by mtm Backout r179941. The nfsclient knob always confuses me. I should have double-checked my setup before commiting. Noticed by: Florian Smeets Pointy hat to: mtm Revision Changes Path 1.21 +2 -2 src/etc/rc.d/lockd 1.20 +2 -2 src/etc/rc.d/statd From danger at FreeBSD.org Fri Jun 27 16:32:16 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Fri Jun 27 16:32:22 2008 Subject: cvs commit: src/usr.sbin/mountd exports.5 Message-ID: <200806271632.m5RGWFUa052114@repoman.freebsd.org> danger 2008-06-27 16:32:04 UTC FreeBSD src repository Modified files: usr.sbin/mountd exports.5 Log: SVN rev 180071 on 2008-06-27 16:32:04Z by danger - markup fixes - advise to use rc script to SIGHUP mountd - add information about possiblity of using /prefix network notation [1] PR: docs/124373 Reviewed by: jhb Obtained from: NetBSD [1] MFC after: 3 days Revision Changes Path 1.29 +68 -15 src/usr.sbin/mountd/exports.5 From ru at FreeBSD.org Fri Jun 27 19:23:57 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Fri Jun 27 19:24:53 2008 Subject: cvs commit: src/etc/rc.d savecore In-Reply-To: References: <200806232054.m5NKsaRS025388@repoman.freebsd.org> Message-ID: <20080627192335.GC1976@edoofus.dev.vega.ru> On Tue, Jun 24, 2008 at 01:32:17PM +0200, Ronald Klop wrote: > On Mon, 23 Jun 2008 22:54:32 +0200, Mike Makonnen wrote: > > > mtm 2008-06-23 20:54:32 UTC > > > > FreeBSD src repository > > > > Modified files: > > etc/rc.d savecore > > Log: > > SVN rev 179962 on 2008-06-23 20:54:32Z by mtm > > Run savecore(8) only if there is a core dump to save. If there is > > no core dump hide the message to that effect behind $rc_quiet. > > Revision Changes Path > > 1.13 +5 -2 src/etc/rc.d/savecore > > > So, if I enable savecore in rc.conf I don't get any feedback from the > system that I enabled it correctly? > I'm not a kernel developer and have nothing to say about it, but this > questions comes to my mind about userfriendlyness. > If you set $rc_quiet and there's no core to save, then no info will be printed. Cheers, -- Ruslan Ermilov ru@FreeBSD.org FreeBSD committer From emaste at FreeBSD.org Fri Jun 27 19:39:48 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Fri Jun 27 19:39:55 2008 Subject: cvs commit: src/share/man/man4 aac.4 src/sys/dev/aac aac_pci.c Message-ID: <200806271939.m5RJdlnQ080645@repoman.freebsd.org> emaste 2008-06-27 19:34:47 UTC FreeBSD src repository Modified files: share/man/man4 aac.4 sys/dev/aac aac_pci.c Log: SVN rev 180072 on 2008-06-27 19:34:47Z by emaste Add explicit PCI IDs for the following Adaptec RAID Series 2 adapters: Adaptec RAID 2045 Adaptec RAID 2405 Adaptec RAID 2445 Adaptec RAID 2805 Without this change these devices are supported by the driver's family support, but they then appear as "Adaptec RAID Controller" in boot messages and the dev.aac.0.%desc sysctl. Revision Changes Path 1.40 +9 -1 src/share/man/man4/aac.4 1.68 +8 -0 src/sys/dev/aac/aac_pci.c From christoph.mallon at gmx.de Fri Jun 27 20:25:14 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Fri Jun 27 20:25:23 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <200806252105.m5PL5AUp064418@repoman.freebsd.org> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> Message-ID: <48654667.1040401@gmx.de> Marius Strobl wrote: > marius 2008-06-25 21:04:59 UTC > > FreeBSD src repository > > Modified files: > sys/sparc64/include in_cksum.h > Log: > SVN rev 180011 on 2008-06-25 21:04:59Z by marius > > Use "__asm __volatile" rather than "__asm" for instruction sequences > that modify condition codes (the carry bit, in this case). Without > "__volatile", the compiler might add the inline assembler instructions > between unrelated code which also uses condition codes, modifying the > latter. > This prevents the TCP pseudo header checksum calculation done in > tcp_output() from having effects on other conditions when compiled > with GCC 4.2.1 at "-O2" and "options INET6" left out. [1] > > Reported & tested by: Boris Kochergin [1] > MFC after: 3 days This approach seems wrong to me and I think it works only by chance. The condition codes ("cc") should be added to the clobbered list of the asm statement instead of making the statement volatile: __asm("..." : $OUT : $IN : "cc"); This very case is also mentioned in the GCC documentation: "If your assembler instruction can alter the condition code register, add `cc' to the list of clobbered registers. GCC on some machines represents the condition codes as a specific hardware register; `cc' serves to name this register. On other machines, the condition code is handled differently, and specifying `cc' has no effect. But it is valid no matter what the machine." (Section 5.35 Assembler Instructions with C Expression Operands) I wrote a letter directly to Marius about this issue two days ago, but I got no response so far. Because this change has a MFC after 3 days, I'm writing to this list. Regards Christoph From ru at FreeBSD.org Fri Jun 27 21:08:47 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Fri Jun 27 21:08:56 2008 Subject: cvs commit: src/gnu/usr.bin/cvs/contrib Makefile In-Reply-To: <200806271459.m5RExftq038676@repoman.freebsd.org> References: <200806271459.m5RExftq038676@repoman.freebsd.org> Message-ID: <20080627210828.GC2889@edoofus.dev.vega.ru> On Fri, Jun 27, 2008 at 02:59:23PM +0000, David E. O'Brien wrote: > obrien 2008-06-27 14:59:23 UTC > > FreeBSD src repository > > Modified files: > gnu/usr.bin/cvs/contrib Makefile > Log: > SVN rev 180067 on 2008-06-27 14:59:23Z by obrien > > Instead of using .POSIX:, use the cool looking "@:" rule. > > Submitted by: ru > > Revision Changes Path > 1.21 +4 -2 src/gnu/usr.bin/cvs/contrib/Makefile > Thank you, David! :-) Cheers, -- Ruslan Ermilov ru@FreeBSD.org FreeBSD committer From marius at FreeBSD.org Fri Jun 27 22:17:44 2008 From: marius at FreeBSD.org (Marius Strobl) Date: Fri Jun 27 22:17:51 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h Message-ID: <200806272217.m5RMHTd7006079@repoman.freebsd.org> marius 2008-06-27 22:17:14 UTC FreeBSD src repository Modified files: sys/sparc64/include in_cksum.h Log: SVN rev 180073 on 2008-06-27 22:17:14Z by marius Improve r180011 by explicitly adding the condition codes to the clobber list. Suggested by: Christoph Mallon Revision Changes Path 1.5 +3 -3 src/sys/sparc64/include/in_cksum.h From marius at alchemy.franken.de Fri Jun 27 22:24:11 2008 From: marius at alchemy.franken.de (Marius Strobl) Date: Fri Jun 27 22:24:14 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <48654667.1040401@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> Message-ID: <20080627222404.GJ1215@alchemy.franken.de> On Fri, Jun 27, 2008 at 09:58:31PM +0200, Christoph Mallon wrote: > Marius Strobl wrote: > >marius 2008-06-25 21:04:59 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/sparc64/include in_cksum.h > > Log: > > SVN rev 180011 on 2008-06-25 21:04:59Z by marius > > > > Use "__asm __volatile" rather than "__asm" for instruction sequences > > that modify condition codes (the carry bit, in this case). Without > > "__volatile", the compiler might add the inline assembler instructions > > between unrelated code which also uses condition codes, modifying the > > latter. > > This prevents the TCP pseudo header checksum calculation done in > > tcp_output() from having effects on other conditions when compiled > > with GCC 4.2.1 at "-O2" and "options INET6" left out. [1] > > > > Reported & tested by: Boris Kochergin [1] > > MFC after: 3 days > > This approach seems wrong to me and I think it works only by chance. The > condition codes ("cc") should be added to the clobbered list of the asm > statement instead of making the statement volatile: __asm("..." : $OUT : > $IN : "cc"); > This very case is also mentioned in the GCC documentation: > "If your assembler instruction can alter the condition code register, > add `cc' to the list of clobbered registers. GCC on some machines > represents the condition codes as a specific hardware register; `cc' > serves to name this register. On other machines, the condition code is > handled differently, and specifying `cc' has no effect. But it is > valid no matter what the machine." (Section 5.35 Assembler Instructions > with C Expression Operands) > > I wrote a letter directly to Marius about this issue two days ago, but I > got no response so far. Because this change has a MFC after 3 days, I'm > writing to this list. I wasn't aware that the clobber list allows to explicitly specify the condition codes, thanks for the hint. Though it unfortunately took me longer than two days to verify it's effect on the generated code; sparc64 could still have been one of the archs where "cc" has no effect. Besides I don't think using "__volatile" for this is that wrong, given that the sparc64 code generated by using "cc" and "__volatile" is nearly identical and given that at least i386 relies on "__volatile" telling GCC that the inline assembler uses the condition codes since quite some time. So the condition codes are probably part of what GCC treats as "important side-effects". Regarding the MFC, they don't happen automatically and the change was not wrong in general so there was no need to hurry :) Thanks again, Marius From christoph.mallon at gmx.de Fri Jun 27 22:56:11 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Fri Jun 27 22:56:16 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <200806272217.m5RMHTd7006079@repoman.freebsd.org> References: <200806272217.m5RMHTd7006079@repoman.freebsd.org> Message-ID: <48657008.4010504@gmx.de> Marius Strobl wrote: > marius 2008-06-27 22:17:14 UTC > > FreeBSD src repository > > Modified files: > sys/sparc64/include in_cksum.h > Log: > SVN rev 180073 on 2008-06-27 22:17:14Z by marius > > Improve r180011 by explicitly adding the condition codes to the > clobber list. You should remove the volatile specifier. For example volatile prevents common subexpression elimination and other types of optimisations. Regards Christoph From christoph.mallon at gmx.de Fri Jun 27 22:57:31 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Fri Jun 27 22:57:40 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080627222404.GJ1215@alchemy.franken.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> Message-ID: <48657058.6020102@gmx.de> Marius Strobl wrote: > I wasn't aware that the clobber list allows to explicitly specify > the condition codes, thanks for the hint. Though it unfortunately > took me longer than two days to verify it's effect on the generated > code; sparc64 could still have been one of the archs where "cc" has > no effect. Besides I don't think using "__volatile" for this is > that wrong, given that the sparc64 code generated by using "cc" > and "__volatile" is nearly identical and given that at least i386 > relies on "__volatile" telling GCC that the inline assembler uses > the condition codes since quite some time. So the condition codes > are probably part of what GCC treats as "important side-effects". If this is true and GCC only handles the eflags on x86 correctly, when __volatile is used, but not if "cc" is marked as clobbered, then this is clearly a bug. > Regarding the MFC, they don't happen automatically and the change > was not wrong in general so there was no need to hurry :) I still think, using __volatile only works by accident. volatile for an assembler block mostly means "this asm statement has an effect, even though the register specification looks otherwise, so do not optimise this away (i.e. no CSE, do not remove if result is unused etc.). On a related note: Is inline assembler really necessary here? For example couldn't in_addword() be written as static __inline u_short in_addword(u_short const sum, u_short const b) { u_int const t = sum + b; return t + (t >> 16); } ? This should at least produce equally good code and because the compiler has more knowledge about it than an assembler block, it potentially leads to better code. I have no SPARC compiler at hand, though. In fact the in/out specification for this asm block looks rather bad: "=&r" (__ret), "=&r" (__tmp) : "r" (sum), "r" (b) : "cc"); The "&"-modifiers (do not use the same registers as for any input operand value) force the compiler to use 4 (!) register in total for this asm block. It could be done with 2 registers if a proper in/out specification was used. At the very least the in/out specification can be improved, but I suspect using plain C is the better choice. Regards Christoph From das at FreeBSD.org Sat Jun 28 01:43:32 2008 From: das at FreeBSD.org (David Schultz) Date: Sat Jun 28 01:43:35 2008 Subject: cvs commit: src/lib/msun/man cos.3 Message-ID: <200806280143.m5S1hWj5038825@repoman.freebsd.org> das 2008-06-28 01:43:24 UTC FreeBSD src repository Modified files: lib/msun/man cos.3 Log: SVN rev 180074 on 2008-06-28 01:43:24Z by das Fix a typo in the cosl() prototype. Revision Changes Path 1.14 +2 -2 src/lib/msun/man/cos.3 From brde at optusnet.com.au Sat Jun 28 02:06:34 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sat Jun 28 02:06:43 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <48654667.1040401@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> Message-ID: <20080628114028.M89039@delplex.bde.org> On Fri, 27 Jun 2008, Christoph Mallon wrote: > Marius Strobl wrote: >> marius 2008-06-25 21:04:59 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/sparc64/include in_cksum.h Log: >> SVN rev 180011 on 2008-06-25 21:04:59Z by marius >> Use "__asm __volatile" rather than "__asm" for instruction sequences >> that modify condition codes (the carry bit, in this case). Without >> ... > > This approach seems wrong to me and I think it works only by chance. The > condition codes ("cc") should be added to the clobbered list of the asm > statement instead of making the statement volatile: __asm("..." : $OUT : $IN > : "cc"); > This very case is also mentioned in the GCC documentation: > "If your assembler instruction can alter the condition code register, > add `cc' to the list of clobbered registers. GCC on some machines > represents the condition codes as a specific hardware register; `cc' > serves to name this register. On other machines, the condition code is > handled differently, and specifying `cc' has no effect. But it is > valid no matter what the machine." (Section 5.35 Assembler Instructions with > C Expression Operands) Does sparc64 do anything good with this? Later it says that this is irrelevant for the type type of bug in in_cksum.* (expecting cc to be preserved across separate asms). From gcc.info: % Similarly, you can't expect a sequence of volatile `asm' instructions % to remain perfectly consecutive. If you want consecutive output, use a ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ % single `asm'. Also, GCC will perform some optimizations across a ^^^^^^^^^^^^^ % volatile `asm' instruction; GCC does not "forget everything" when it % encounters a volatile `asm' instruction the way some other compilers do. % % It is a natural idea to look for a way to give access to the condition % code left by the assembler instruction. However, when we attempted to % implement this, we found no way to make it work reliably. The problem ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ % is that output operands might need reloading, which would result in % additional following "store" instructions. On most machines, these % instructions would alter the condition code before there was time to % test it. This problem doesn't arise for ordinary "test" and "compare" % instructions because they don't have any output operands. % % For reasons similar to those described above, it is not possible to ^^^^^^^^^^^^^^^^^^^^^ % give an assembler instruction access to the condition code left by ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ % previous instructions. ^^^^^^^^^^^^^^^^^^^^^^ On i386, specifying cc has no effect (the compiler must always assume that cc is clobbered), and specifying cc in asms is a style bug in FreeBSD. It took about 15 years (1992-2007) for the bugs in in_cksum.* to be finally fixed on i386, by following the rule given in gcc.info. IIRC, this rule wasn't present in 1992, and wasn't really needed then because the optimizer wasn't agressive enough to move things, especially volatile asms, but it has been there for about 10 years. > I wrote a letter directly to Marius about this issue two days ago, but I got > no response so far. Because this change has a MFC after 3 days, I'm writing > to this list. I wrote a similar reply with a similar response. The i386 version, though fixed, still has a lot of historical cruft related to this bug (bogus volatile declarations, bogus comments about volatile's affect. and triplication of in_cksum.c to work around the bug being so large for INTEL_COMPILER that the volatile hack never helped). The amd64 version never had the bug since it was bogusly duplicated from the ia64 version. It is just not in asm and thus probably slow. It only has a duplicated in_cksum.c. The MI version in netinet should be used instead of n-tuplicated, but has rotted. Bruce From brde at optusnet.com.au Sat Jun 28 02:09:20 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sat Jun 28 02:09:25 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080627222404.GJ1215@alchemy.franken.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> Message-ID: <20080628120700.S89039@delplex.bde.org> On Sat, 28 Jun 2008, Marius Strobl wrote: > On Fri, Jun 27, 2008 at 09:58:31PM +0200, Christoph Mallon wrote: >> >> This approach seems wrong to me and I think it works only by chance. The > > I wasn't aware that the clobber list allows to explicitly specify > the condition codes, thanks for the hint. Though it unfortunately > took me longer than two days to verify it's effect on the generated > code; sparc64 could still have been one of the archs where "cc" has > no effect. I think it still only works by chance. > Besides I don't think using "__volatile" for this is > that wrong, given that the sparc64 code generated by using "cc" > and "__volatile" is nearly identical and given that at least i386 > relies on "__volatile" telling GCC that the inline assembler uses > the condition codes since quite some time. So the condition codes > are probably part of what GCC treats as "important side-effects". No, the comments in the i386 version are rotted bits left over from old failing attempts to fix this problem. Bruce From brde at optusnet.com.au Sat Jun 28 02:45:55 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sat Jun 28 02:46:04 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <48657058.6020102@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> Message-ID: <20080628121025.F89039@delplex.bde.org> On Sat, 28 Jun 2008, Christoph Mallon wrote: > I still think, using __volatile only works by accident. volatile for an > assembler block mostly means "this asm statement has an effect, even though > the register specification looks otherwise, so do not optimise this away > (i.e. no CSE, do not remove if result is unused etc.). Right. Though I've never seen unnecessary's __volatiles significantly affecting i386 code. This is because the code in the asms can't be removed completely, and can't be moved much either. With out of order execution, the type of moves that are permitted (not across dependencies) are precisely the type of moves that the CPU's scheduler can do or undo no matter how the compiler orders the code. > On a related note: Is inline assembler really necessary here? For example > couldn't in_addword() be written as > static __inline u_short > in_addword(u_short const sum, u_short const b) > { > u_int const t = sum + b; > return t + (t >> 16); > } ? > This should at least produce equally good code and because the compiler has > more knowledge about it than an assembler block, it potentially leads to > better code. I have no SPARC compiler at hand, though. Last time I tried on i386, I couldn't get gcc to generate operations involving carries for things like this, or the bswap instruction from C code to reorder a word. gcc4.2 -O3 on i386 now generates for the above: movzwl b, %eax # starting from b and sum in memory movzwl sum, %edx addl %eax, %edx # 32-bit add movl %edx, %eax shrl $16, %eax # it does the shift laboriously addl %edx, %eax movzwl %ax, %eax # don't really need 32-bit result # but need something to discard the high bits In non-inline asm, I would write this as: movw sum,%ax addw b,%ax adcw $0,%ax movzwl %ax,%eax Pipelining can make bloated code run better than it looks, but probably not for the generated code above, since shifts are slow on some i386's and there is an extra dependency for the extra shift operation. > In fact the in/out specification for this asm block looks rather bad: > "=&r" (__ret), "=&r" (__tmp) : "r" (sum), "r" (b) : "cc"); > The "&"-modifiers (do not use the same registers as for any input operand > value) force the compiler to use 4 (!) register in total for this asm block. > It could be done with 2 registers if a proper in/out specification was used. > At the very least the in/out specification can be improved, but I suspect > using plain C is the better choice. Hmm, the i386 version is much simpler. It just forces use of 2 registers when 1 is enough (change its constraint for (b) from "r" to "rm" to permit adcw from either register or memory, so that it can generate the above code if b happens to be in memory; on most i386's, this optimizes for space but makes no difference for time). Bruce From remko at FreeBSD.org Sat Jun 28 08:37:07 2008 From: remko at FreeBSD.org (Remko Lodder) Date: Sat Jun 28 08:37:13 2008 Subject: cvs commit: src/sys/dev/acpi_support acpi_asus.c Message-ID: <200806280837.m5S8b7nF000114@repoman.freebsd.org> remko 2008-06-28 08:36:47 UTC FreeBSD src repository Modified files: sys/dev/acpi_support acpi_asus.c Log: SVN rev 180075 on 2008-06-28 08:36:47Z by remko Style fix (space into tab). Approved by: imp (mentor, implicit) Revision Changes Path 1.35 +1 -1 src/sys/dev/acpi_support/acpi_asus.c From christoph.mallon at gmx.de Sat Jun 28 08:38:49 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sat Jun 28 08:38:57 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080628114028.M89039@delplex.bde.org> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080628114028.M89039@delplex.bde.org> Message-ID: <4865F895.8030600@gmx.de> Bruce Evans wrote: > On Fri, 27 Jun 2008, Christoph Mallon wrote: > >> Marius Strobl wrote: >>> marius 2008-06-25 21:04:59 UTC >>> >>> FreeBSD src repository >>> >>> Modified files: >>> sys/sparc64/include in_cksum.h Log: >>> SVN rev 180011 on 2008-06-25 21:04:59Z by marius >>> Use "__asm __volatile" rather than "__asm" for instruction sequences >>> that modify condition codes (the carry bit, in this case). Without >>> ... >> >> This approach seems wrong to me and I think it works only by chance. >> The condition codes ("cc") should be added to the clobbered list of >> the asm statement instead of making the statement volatile: >> __asm("..." : $OUT : $IN : "cc"); > >> This very case is also mentioned in the GCC documentation: >> "If your assembler instruction can alter the condition code register, >> add `cc' to the list of clobbered registers. GCC on some machines >> represents the condition codes as a specific hardware register; `cc' >> serves to name this register. On other machines, the condition code is >> handled differently, and specifying `cc' has no effect. But it is >> valid no matter what the machine." (Section 5.35 Assembler >> Instructions with C Expression Operands) > > Does sparc64 do anything good with this? Later it says that this is > irrelevant for the type type of bug in in_cksum.* (expecting cc to > be preserved across separate asms). From gcc.info: I think the bug was the following: subcc %foo, %bar, %g0 /* SPARC compare */ #APP /* inline assembler of in_addword() here, which modifies the condition codes */ #NO_APP bpl $somewhere /* condition branch depending on condition code */ The bpl is supposed to jump depending on the condition of the subcc, but if the compiler schedules the inline assembler block between the subcc and the bpl, it jumps depending on garbage. It was not expected to preserve the condition codes across separate inline assembler blocks, but it was not specified that the single inline assembler block did not modify the condition codes. > % Similarly, you can't expect a sequence of volatile `asm' instructions > % to remain perfectly consecutive. If you want consecutive output, use a > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > % single `asm'. Also, GCC will perform some optimizations across a > ^^^^^^^^^^^^^ > % volatile `asm' instruction; GCC does not "forget everything" when it > % encounters a volatile `asm' instruction the way some other compilers do. > % % It is a natural idea to look for a way to give access to the condition > % code left by the assembler instruction. However, when we attempted to > % implement this, we found no way to make it work reliably. The problem > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > % is that output operands might need reloading, which would result in > % additional following "store" instructions. On most machines, these > % instructions would alter the condition code before there was time to > % test it. This problem doesn't arise for ordinary "test" and "compare" > % instructions because they don't have any output operands. > % % For reasons similar to those described above, it is not possible to > ^^^^^^^^^^^^^^^^^^^^^ > % give an assembler instruction access to the condition code left by > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > % previous instructions. > ^^^^^^^^^^^^^^^^^^^^^^ I think the excerpt does not apply here, because the problem is the other way round (inline assembler interrupting C, not C interrupting multiple inline assembler blocks), see above. > On i386, specifying cc has no effect (the compiler must always assume that > cc is clobbered), and specifying cc in asms is a style bug in FreeBSD. I have to disagree. Where does the GCC documentation state, that "cc" has no effect on x86? It is the other way round: The compiler assumes, the condition codes are *not* modified, if it is not explicitely stated. Exactly this caused the bug (though here on SPARC), which was tried to be solved by volatile. I still am convinced that specifying "cc" in the clobber list and not using volatile is the correct solution. If it is a style bug to specify "cc", the style should be changed, otherwise you cannot use inline assembler correctly. > It took about 15 years (1992-2007) for the bugs in in_cksum.* to be > finally fixed on i386, by following the rule given in gcc.info. IIRC, > this rule wasn't present in 1992, and wasn't really needed then because > the optimizer wasn't agressive enough to move things, especially volatile > asms, but it has been there for about 10 years. As far as I can see the rule in the gcc documentation was not followed in the x86 version: The assembler block is volatile and "cc" is not specified as clobbered. This, imo, is a bug. Regards Christoph From christoph.mallon at gmx.de Sat Jun 28 08:38:56 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sat Jun 28 08:39:07 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080628121025.F89039@delplex.bde.org> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628121025.F89039@delplex.bde.org> Message-ID: <4865F89D.4090207@gmx.de> Bruce Evans wrote: > On Sat, 28 Jun 2008, Christoph Mallon wrote: > >> I still think, using __volatile only works by accident. volatile for >> an assembler block mostly means "this asm statement has an effect, >> even though the register specification looks otherwise, so do not >> optimise this away (i.e. no CSE, do not remove if result is unused etc.). > > Right. Though I've never seen unnecessary's __volatiles significantly > affecting i386 code. This is because the code in the asms can't be > removed completely, and can't be moved much either. With out of order > execution, the type of moves that are permitted (not across dependencies) > are precisely the type of moves that the CPU's scheduler can do or undo > no matter how the compiler orders the code. I disagree. For example look at the use of in_addword() in dev/sk/if_sk.cv in line 2819: csum1 = htons(csum & 0xffff); csum2 = htons((csum >> 16) & 0xffff); ipcsum = in_addword(csum1, ~csum2 & 0xffff); /* checksum fixup for IP options */ len = hlen - sizeof(struct ip); if (len > 0) { return; } The calculation will be executed even if the following if (len > 0) leaves the function and the value of ipcsum is unused. If in_addword() is not marked volatile it can be moved after the if and not be executed in all cases. csum1 and csum2 can be moved after the if, too. >> On a related note: Is inline assembler really necessary here? For >> example couldn't in_addword() be written as >> static __inline u_short >> in_addword(u_short const sum, u_short const b) >> { >> u_int const t = sum + b; >> return t + (t >> 16); >> } ? >> This should at least produce equally good code and because the >> compiler has more knowledge about it than an assembler block, it >> potentially leads to better code. I have no SPARC compiler at hand, >> though. > > Last time I tried on i386, I couldn't get gcc to generate operations > involving carries for things like this, or the bswap instruction from > C code to reorder a word. gcc4.2 -O3 on i386 now generates for the above: > > movzwl b, %eax # starting from b and sum in memory > movzwl sum, %edx > addl %eax, %edx # 32-bit add > movl %edx, %eax > shrl $16, %eax # it does the shift laboriously > addl %edx, %eax > movzwl %ax, %eax # don't really need 32-bit result > # but need something to discard the high bits If the upper 16 bits are not "looked at" then the final movzwl can be optimised away. Many instructions, like add, shl and mul, can live with "garbage" in the upper 16 bits. Only if a "bad" instruction, like shr or div, is encountered, the upper 16 bits have to be cleared. The current x86 implementation of in_addword() using inline assembler causes the compiler to add a movzwl, too, before the return. > In non-inline asm, I would write this as: > > movw sum,%ax > addw b,%ax > adcw $0,%ax > movzwl %ax,%eax You do not want to use 16bit instructions on modern x86 processors. These instructions are slow. Intel states that decoding a 16bit operation takes 6 cycles instead of the usual 1. (Intel? 64 and IA-32 Architectures Optimization Reference Manual, section 2.1.2.2 Instruction PreDecode) > Pipelining can make bloated code run better than it looks, but probably > not for the generated code above, since shifts are slow on some i386's > and there is an extra dependency for the extra shift operation. Shifts were slow on early generations of the Pentium 4. Intel corrected this "glitch" in later generations. >> In fact the in/out specification for this asm block looks rather bad: >> "=&r" (__ret), "=&r" (__tmp) : "r" (sum), "r" (b) : "cc"); >> The "&"-modifiers (do not use the same registers as for any input >> operand value) force the compiler to use 4 (!) register in total for >> this asm block. It could be done with 2 registers if a proper in/out >> specification was used. At the very least the in/out specification can >> be improved, but I suspect using plain C is the better choice. > > Hmm, the i386 version is much simpler. It just forces use of 2 registers > when 1 is enough (change its constraint for (b) from "r" to "rm" to permit > adcw from either register or memory, so that it can generate the above code > if b happens to be in memory; on most i386's, this optimizes for space but > makes no difference for time). Sometimes a few less bytes in an inner loop can have dramatic effects. Saving a register is always (ok, most of the time) a good idea on x86. On the other hand, the function is inlined so proably its operands do not originate from memory locations. But I still think the better solution is to use simple C instead of inline assembler. Regards Christoph From marius at alchemy.franken.de Sat Jun 28 11:24:03 2008 From: marius at alchemy.franken.de (Marius Strobl) Date: Sat Jun 28 11:24:12 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080628120700.S89039@delplex.bde.org> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <20080628120700.S89039@delplex.bde.org> Message-ID: <20080628112358.GK1215@alchemy.franken.de> On Sat, Jun 28, 2008 at 12:09:13PM +1000, Bruce Evans wrote: > On Sat, 28 Jun 2008, Marius Strobl wrote: > > >On Fri, Jun 27, 2008 at 09:58:31PM +0200, Christoph Mallon wrote: > >> > >>This approach seems wrong to me and I think it works only by chance. The > > > >I wasn't aware that the clobber list allows to explicitly specify > >the condition codes, thanks for the hint. Though it unfortunately > >took me longer than two days to verify it's effect on the generated > >code; sparc64 could still have been one of the archs where "cc" has > >no effect. > > I think it still only works by chance. > > >Besides I don't think using "__volatile" for this is > >that wrong, given that the sparc64 code generated by using "cc" > >and "__volatile" is nearly identical and given that at least i386 > >relies on "__volatile" telling GCC that the inline assembler uses > >the condition codes since quite some time. So the condition codes > >are probably part of what GCC treats as "important side-effects". > > No, the comments in the i386 version are rotted bits left over from > old failing attempts to fix this problem. > So by now you are saying that using "__volatile" in this case is the wrong solution and that using "cc" is a style bug. How am I supposed to tell the compiler that the inline assembler alters the condition codes then, which it apparently needs to know as it at shown to otherwise generate broken code even when using a single __asm() for the istructions. Marius From marius at alchemy.franken.de Sat Jun 28 11:44:22 2008 From: marius at alchemy.franken.de (Marius Strobl) Date: Sat Jun 28 11:44:31 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <48657058.6020102@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> Message-ID: <20080628114417.GL1215@alchemy.franken.de> On Sat, Jun 28, 2008 at 12:57:28AM +0200, Christoph Mallon wrote: > Marius Strobl wrote: > >I wasn't aware that the clobber list allows to explicitly specify > >the condition codes, thanks for the hint. Though it unfortunately > >took me longer than two days to verify it's effect on the generated > >code; sparc64 could still have been one of the archs where "cc" has > >no effect. Besides I don't think using "__volatile" for this is > >that wrong, given that the sparc64 code generated by using "cc" > >and "__volatile" is nearly identical and given that at least i386 > >relies on "__volatile" telling GCC that the inline assembler uses > >the condition codes since quite some time. So the condition codes > >are probably part of what GCC treats as "important side-effects". > > If this is true and GCC only handles the eflags on x86 correctly, when > __volatile is used, but not if "cc" is marked as clobbered, then this is > clearly a bug. > > >Regarding the MFC, they don't happen automatically and the change > >was not wrong in general so there was no need to hurry :) > > I still think, using __volatile only works by accident. volatile for an > assembler block mostly means "this asm statement has an effect, even > though the register specification looks otherwise, so do not optimise > this away (i.e. no CSE, do not remove if result is unused etc.). > > > On a related note: Is inline assembler really necessary here? For > example couldn't in_addword() be written as > static __inline u_short > in_addword(u_short const sum, u_short const b) > { > u_int const t = sum + b; > return t + (t >> 16); > } ? > This should at least produce equally good code and because the compiler > has more knowledge about it than an assembler block, it potentially > leads to better code. I have no SPARC compiler at hand, though. With GCC 4.2.1 at -O2 the code generated for the above C version takes on more instruction than the inline assembler so if one wants to go for micro-optimizing one should certainly prefer the inline assembler version. > > In fact the in/out specification for this asm block looks rather bad: > "=&r" (__ret), "=&r" (__tmp) : "r" (sum), "r" (b) : "cc"); > The "&"-modifiers (do not use the same registers as for any input > operand value) force the compiler to use 4 (!) register in total for > this asm block. It could be done with 2 registers if a proper in/out > specification was used. At the very least the in/out specification can > be improved, but I suspect using plain C is the better choice. > The "&"-modifiers are necessary as the inline assembler in question consumes output operands before all input operands are consumed. Omitting them caused GCC to generate broken code in the past. Marius From christoph.mallon at gmx.de Sat Jun 28 12:08:13 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sat Jun 28 12:08:18 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080628114417.GL1215@alchemy.franken.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628114417.GL1215@alchemy.franken.de> Message-ID: <486629AA.1050409@gmx.de> Marius Strobl wrote: >> On a related note: Is inline assembler really necessary here? For >> example couldn't in_addword() be written as >> static __inline u_short >> in_addword(u_short const sum, u_short const b) >> { >> u_int const t = sum + b; >> return t + (t >> 16); >> } ? >> This should at least produce equally good code and because the compiler >> has more knowledge about it than an assembler block, it potentially >> leads to better code. I have no SPARC compiler at hand, though. > > With GCC 4.2.1 at -O2 the code generated for the above C version > takes on more instruction than the inline assembler so if one On SPARC? What code does it produce? I have not SPARC compiler at hand. Even if it is one more instruction, I think the reduced register pressure makes more than up for it. > wants to go for micro-optimizing one should certainly prefer the > inline assembler version. As a compiler construction I can tell you, that regarding optimisation there is no such thing as "certainty". The worst part about inline assembler is, that the compiler knows nothing about the instructions in there and has to copy them verbatim. For example it can not do any clever things with the two shifts at the beginning of the inline assembler block of in_addword(). >> In fact the in/out specification for this asm block looks rather bad: >> "=&r" (__ret), "=&r" (__tmp) : "r" (sum), "r" (b) : "cc"); >> The "&"-modifiers (do not use the same registers as for any input >> operand value) force the compiler to use 4 (!) register in total for >> this asm block. It could be done with 2 registers if a proper in/out >> specification was used. At the very least the in/out specification can >> be improved, but I suspect using plain C is the better choice. >> > > The "&"-modifiers are necessary as the inline assembler in > question consumes output operands before all input operands > are consumed. Omitting them caused GCC to generate broken > code in the past. This should work fine and only use two registers (though the compiler can choose to use three, if it deems it beneficial): static __inline u_short in_addword(u_short const sum, u_short const b) { u_long const sum16 = sum << 16; u_long const b16 = b << 16; u_long ret; __asm( "addcc %1, %2, %0\n\t" "srl %0, 16, %0\n\t" "addc %0, 0, %0\n" : "=r" (ret) : "r" (sum16), "r" (b16) : "cc"); return (ret); } But I still prefer the C version. Regards Christoph From christoph.mallon at gmx.de Sat Jun 28 12:10:59 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sat Jun 28 12:11:09 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <486629AA.1050409@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628114417.GL1215@alchemy.franken.de> <486629AA.1050409@gmx.de> Message-ID: <48662A50.5070804@gmx.de> Christoph Mallon wrote: > As a compiler construction I can tell you, that regarding optimisation > there is no such thing as "certainty". Uh...right... s/construction/constructor/ From mtm at FreeBSD.org Sat Jun 28 12:31:40 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Sat Jun 28 12:31:43 2008 Subject: cvs commit: src/usr.sbin/ngctl main.c Message-ID: <200806281231.m5SCVejc032444@repoman.freebsd.org> mtm 2008-06-28 12:31:30 UTC FreeBSD src repository Modified files: usr.sbin/ngctl main.c Log: SVN rev 180076 on 2008-06-28 12:31:30Z by mtm Modify the DoParseCommand() to work on (const char *) instead of just (char *). This is a slightly simplified version of the patch in the PR. It fixes compilitation issues with -O3. PR: misc/124385 Revision Changes Path 1.24 +4 -4 src/usr.sbin/ngctl/main.c From marius at alchemy.franken.de Sat Jun 28 12:52:45 2008 From: marius at alchemy.franken.de (Marius Strobl) Date: Sat Jun 28 12:52:49 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <48657008.4010504@gmx.de> References: <200806272217.m5RMHTd7006079@repoman.freebsd.org> <48657008.4010504@gmx.de> Message-ID: <20080628125241.GN1215@alchemy.franken.de> On Sat, Jun 28, 2008 at 12:56:08AM +0200, Christoph Mallon wrote: > Marius Strobl wrote: > >marius 2008-06-27 22:17:14 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/sparc64/include in_cksum.h > > Log: > > SVN rev 180073 on 2008-06-27 22:17:14Z by marius > > > > Improve r180011 by explicitly adding the condition codes to the > > clobber list. > > You should remove the volatile specifier. For example volatile prevents > common subexpression elimination and other types of optimisations. > I had to adjust the constraint strings in this source file twice now in order to keep GCC from generating broken code, thus I prefer to be conservative by using a slightly bigger hammer and leave the "__volatile" in in order to keep these kind of problems from coming back to haunt us over and over again. Especially when it comes to something as vaguely ("important side-effects", "access memory in an unpredictable fashion", etc) documented as the GCC assembler constraints and thus hard to get right without studying the GCC source and maybe requiring "__volatile" in the future anyway. Marius From marius at alchemy.franken.de Sat Jun 28 13:22:46 2008 From: marius at alchemy.franken.de (Marius Strobl) Date: Sat Jun 28 13:22:55 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <486629AA.1050409@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628114417.GL1215@alchemy.franken.de> <486629AA.1050409@gmx.de> Message-ID: <20080628132241.GO1215@alchemy.franken.de> On Sat, Jun 28, 2008 at 02:08:10PM +0200, Christoph Mallon wrote: > Marius Strobl wrote: > >>On a related note: Is inline assembler really necessary here? For > >>example couldn't in_addword() be written as > >>static __inline u_short > >>in_addword(u_short const sum, u_short const b) > >>{ > >> u_int const t = sum + b; > >> return t + (t >> 16); > >>} ? > >>This should at least produce equally good code and because the compiler > >>has more knowledge about it than an assembler block, it potentially > >>leads to better code. I have no SPARC compiler at hand, though. > > > >With GCC 4.2.1 at -O2 the code generated for the above C version > >takes on more instruction than the inline assembler so if one > > On SPARC? What code does it produce? I have not SPARC compiler at hand. > Even if it is one more instruction, I think the reduced register > pressure makes more than up for it. Correct, it only uses two registers: 0000000000000000 : 0: 92 02 00 09 add %o0, %o1, %o1 4: 91 32 60 10 srl %o1, 0x10, %o0 8: 90 02 00 09 add %o0, %o1, %o0 c: 91 2a 20 10 sll %o0, 0x10, %o0 10: 91 32 20 10 srl %o0, 0x10, %o0 14: 81 c3 e0 08 retl 18: 91 3a 20 00 sra %o0, 0, %o0 1c: 01 00 00 00 nop > > >wants to go for micro-optimizing one should certainly prefer the > >inline assembler version. > > As a compiler construction I can tell you, that regarding optimisation > there is no such thing as "certainty". > The worst part about inline assembler is, that the compiler knows > nothing about the instructions in there and has to copy them verbatim. > For example it can not do any clever things with the two shifts at the > beginning of the inline assembler block of in_addword(). That's why my statement regarding micro-optimizing actually was meant with sarcasm. In order to decide whether it's still worth to do certain code as inline assembler rather than C one would have to re-check, i.e. re-benchmark, every time the compiler or the consumers change. Obviously that's not doable. So the bottom line is that the best we can do is to investigate once and if we come to the conclusion that doing something as inline assembler generally is worth it most of the times, stick with this (though not necessarily forever). > > >>In fact the in/out specification for this asm block looks rather bad: > >>"=&r" (__ret), "=&r" (__tmp) : "r" (sum), "r" (b) : "cc"); > >>The "&"-modifiers (do not use the same registers as for any input > >>operand value) force the compiler to use 4 (!) register in total for > >>this asm block. It could be done with 2 registers if a proper in/out > >>specification was used. At the very least the in/out specification can > >>be improved, but I suspect using plain C is the better choice. > >> > > > >The "&"-modifiers are necessary as the inline assembler in > >question consumes output operands before all input operands > >are consumed. Omitting them caused GCC to generate broken > >code in the past. > > This should work fine and only use two registers (though the compiler > can choose to use three, if it deems it beneficial): > > static __inline u_short > in_addword(u_short const sum, u_short const b) > { > u_long const sum16 = sum << 16; > u_long const b16 = b << 16; > u_long ret; > > __asm( > "addcc %1, %2, %0\n\t" > "srl %0, 16, %0\n\t" > "addc %0, 0, %0\n" > : "=r" (ret) : "r" (sum16), "r" (b16) : "cc"); > > return (ret); > } This is ten instructions with two registers. Where is the break even regarding instructions vs. registers for sparc64? :) > > But I still prefer the C version. > And I prefer to not re-write otherwise working code for micro-optimizations, there are enough unfixed real bugs to deal with. Similarly we should not waste time discussing how to possibly optimize MD versions even more but rather spend the time improving the MI version so it's good enough that using MD versions isn't worth the effort. Marius From christoph.mallon at gmx.de Sat Jun 28 13:35:36 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sat Jun 28 13:35:45 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080628125241.GN1215@alchemy.franken.de> References: <200806272217.m5RMHTd7006079@repoman.freebsd.org> <48657008.4010504@gmx.de> <20080628125241.GN1215@alchemy.franken.de> Message-ID: <48663E25.9080703@gmx.de> Marius Strobl wrote: > On Sat, Jun 28, 2008 at 12:56:08AM +0200, Christoph Mallon wrote: >> Marius Strobl wrote: >>> marius 2008-06-27 22:17:14 UTC >>> >>> FreeBSD src repository >>> >>> Modified files: >>> sys/sparc64/include in_cksum.h >>> Log: >>> SVN rev 180073 on 2008-06-27 22:17:14Z by marius >>> >>> Improve r180011 by explicitly adding the condition codes to the >>> clobber list. >> You should remove the volatile specifier. For example volatile prevents >> common subexpression elimination and other types of optimisations. >> > > I had to adjust the constraint strings in this source file > twice now in order to keep GCC from generating broken code, > thus I prefer to be conservative by using a slightly bigger > hammer and leave the "__volatile" in in order to keep these > kind of problems from coming back to haunt us over and over > again. Especially when it comes to something as vaguely > ("important side-effects", "access memory in an unpredictable > fashion", etc) documented as the GCC assembler constraints > and thus hard to get right without studying the GCC source > and maybe requiring "__volatile" in the future anyway. volatile is for stuff, which cannot be expressed as data dependencies in the constraints, like writing to a machine status register of the CPU, accessing memory mapped hardware registers, which triggers something, stuff like that. The code in question only does some adds and shifts on registers, which is harmless and covered by the data dependecies of the input/output/clobber constraints. volatile is simply the wrong hammer. Regards Christoph From philip at FreeBSD.org Sat Jun 28 13:39:12 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sat Jun 28 13:39:19 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200806281339.m5SDdCJq048312@repoman.freebsd.org> philip 2008-06-28 13:38:53 UTC FreeBSD src repository Modified files: sys/conf NOTES Log: SVN rev 180077 on 2008-06-28 13:38:53Z by philip Remove stray "miibus0" reference from ancient kernel config file times. MFC after: 1 day Revision Changes Path 1.1496 +1 -1 src/sys/conf/NOTES From christoph.mallon at gmx.de Sat Jun 28 13:52:19 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sat Jun 28 13:52:27 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080628132241.GO1215@alchemy.franken.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628114417.GL1215@alchemy.franken.de> <486629AA.1050409@gmx.de> <20080628132241.GO1215@alchemy.franken.de> Message-ID: <48664210.80204@gmx.de> Marius Strobl wrote: > On Sat, Jun 28, 2008 at 02:08:10PM +0200, Christoph Mallon wrote: >> Marius Strobl wrote: >>>> On a related note: Is inline assembler really necessary here? For >>>> example couldn't in_addword() be written as >>>> static __inline u_short >>>> in_addword(u_short const sum, u_short const b) >>>> { >>>> u_int const t = sum + b; >>>> return t + (t >> 16); >>>> } ? >>>> This should at least produce equally good code and because the compiler >>>> has more knowledge about it than an assembler block, it potentially >>>> leads to better code. I have no SPARC compiler at hand, though. >>> With GCC 4.2.1 at -O2 the code generated for the above C version >>> takes on more instruction than the inline assembler so if one >> On SPARC? What code does it produce? I have not SPARC compiler at hand. >> Even if it is one more instruction, I think the reduced register >> pressure makes more than up for it. > > Correct, it only uses two registers: > > 0000000000000000 : > 0: 92 02 00 09 add %o0, %o1, %o1 > 4: 91 32 60 10 srl %o1, 0x10, %o0 > 8: 90 02 00 09 add %o0, %o1, %o0 > c: 91 2a 20 10 sll %o0, 0x10, %o0 > 10: 91 32 20 10 srl %o0, 0x10, %o0 > 14: 81 c3 e0 08 retl > 18: 91 3a 20 00 sra %o0, 0, %o0 > 1c: 01 00 00 00 nop One more instruction? That's five instructions for the actual calculation afaict, just like the inline assembler version. The sra in the delay slot should be present in the inline assembler version, too. >> This should work fine and only use two registers (though the compiler >> can choose to use three, if it deems it beneficial): >> >> static __inline u_short >> in_addword(u_short const sum, u_short const b) >> { >> u_long const sum16 = sum << 16; >> u_long const b16 = b << 16; >> u_long ret; >> >> __asm( >> "addcc %1, %2, %0\n\t" >> "srl %0, 16, %0\n\t" >> "addc %0, 0, %0\n" >> : "=r" (ret) : "r" (sum16), "r" (b16) : "cc"); >> >> return (ret); >> } > > This is ten instructions with two registers. Where is the > break even regarding instructions vs. registers for sparc64? :) I still have no SPARC compiler. Ten instructions? All I did was write the two shifts in C and adjust the register constraints. It should produce identical code. >> But I still prefer the C version. >> > > And I prefer to not re-write otherwise working code for > micro-optimizations, there are enough unfixed real bugs Obviously the inline assembler magic did not work and is/was a real bug. > to deal with. Similarly we should not waste time discussing > how to possibly optimize MD versions even more but rather > spend the time improving the MI version so it's good enough > that using MD versions isn't worth the effort. The C alternative is MI and in length on par with the inline assembler version, isn't it? Regards Christoph From obrien at FreeBSD.org Sat Jun 28 15:17:15 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sat Jun 28 15:17:22 2008 Subject: cvs commit: src/contrib/gcc/config/i386 freebsd.h Message-ID: <200806281517.m5SFHEUi059185@repoman.freebsd.org> obrien 2008-06-28 15:17:02 UTC FreeBSD src repository Modified files: contrib/gcc/config/i386 freebsd.h Log: SVN rev 180078 on 2008-06-28 15:17:02Z by obrien Reduce diff to vendor. Revision Changes Path 1.74 +2 -2 src/contrib/gcc/config/i386/freebsd.h From obrien at FreeBSD.org Sat Jun 28 15:28:31 2008 From: obrien at FreeBSD.org (David E. O'Brien) Date: Sat Jun 28 15:28:33 2008 Subject: cvs commit: src/contrib/gcc/config/i386 freebsd.h Message-ID: <200806281528.m5SFSUB9060125@repoman.freebsd.org> obrien 2008-06-28 15:28:17 UTC FreeBSD src repository Modified files: contrib/gcc/config/i386 freebsd.h Log: SVN rev 180079 on 2008-06-28 15:28:17Z by obrien White space fixes. Revision Changes Path 1.75 +8 -8 src/contrib/gcc/config/i386/freebsd.h From das at FreeBSD.org Sat Jun 28 17:55:48 2008 From: das at FreeBSD.org (David Schultz) Date: Sat Jun 28 17:55:56 2008 Subject: cvs commit: src/lib/libc/amd64/gen _setjmp.S setjmp.S Message-ID: <200806281755.m5SHtl8h083641@repoman.freebsd.org> das 2008-06-28 17:55:43 UTC FreeBSD src repository Modified files: lib/libc/amd64/gen _setjmp.S setjmp.S Log: SVN rev 180080 on 2008-06-28 17:55:43Z by das Two FP-related setjmp/longjmp changes: 1. Save and restore the control part of the MXCSR in addition to the i387 control word to ensure that the two are consistent. Note that standards don't require longjmp to restore either control word, and none of Linux, MacOS X 10.3 and earlier, NetBSD, OpenBSD, or Solaris do it. However, it is historical FreeBSD behavior, and bde points out that it is needed to make longjmping out of a signal handler work properly, given the way FreeBSD clobbers the FPU state on signal handler entry. 2. Don't clobber the FPU exception flags in longjmp. C99 requires them to remain unchanged. Revision Changes Path 1.20 +10 -1 src/lib/libc/amd64/gen/_setjmp.S 1.28 +10 -1 src/lib/libc/amd64/gen/setjmp.S From das at FreeBSD.org Sat Jun 28 17:58:18 2008 From: das at FreeBSD.org (David Schultz) Date: Sat Jun 28 17:58:27 2008 Subject: cvs commit: src/lib/libc/i386/gen _setjmp.S setjmp.S Message-ID: <200806281758.m5SHwIl2083857@repoman.freebsd.org> das 2008-06-28 17:58:06 UTC FreeBSD src repository Modified files: lib/libc/i386/gen _setjmp.S setjmp.S Log: SVN rev 180081 on 2008-06-28 17:58:06Z by das We should also save and restore the MXCSR as on amd64, but detecting whether the CPU supports SSE or not here is rather odious. Revision Changes Path 1.18 +0 -1 src/lib/libc/i386/gen/_setjmp.S 1.24 +0 -1 src/lib/libc/i386/gen/setjmp.S From das at FreeBSD.ORG Sat Jun 28 18:01:50 2008 From: das at FreeBSD.ORG (David Schultz) Date: Sat Jun 28 18:01:54 2008 Subject: cvs commit: src/lib/libc/i386/gen _setjmp.S setjmp.S In-Reply-To: <200806281758.m5SHwIl2083857@repoman.freebsd.org> References: <200806281758.m5SHwIl2083857@repoman.freebsd.org> Message-ID: <20080628180230.GA37313@zim.MIT.EDU> On Sat, Jun 28, 2008, David Schultz wrote: > das 2008-06-28 17:58:06 UTC > > FreeBSD src repository > > Modified files: > lib/libc/i386/gen _setjmp.S setjmp.S > Log: > SVN rev 180081 on 2008-06-28 17:58:06Z by das > > We should also save and restore the MXCSR as on amd64, but detecting > whether the CPU supports SSE or not here is rather odious. Err, the first line of the commit message read: Don't clobber the FPU exception flags in longjmp. C99 requires them to remain unchanged. ...but got cut off somehow. From jhb at FreeBSD.org Sat Jun 28 23:27:34 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Sat Jun 28 23:27:40 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: <200806282327.m5SNRXM5030027@repoman.freebsd.org> jhb 2008-06-28 23:26:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) 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/include intr_machdep.h sys/sparc64/sparc64 intr_machdep.c sys/sun4v/sun4v intr_machdep.c sys/sys interrupt.h Log: SVN rev 180082 on 2008-06-28 23:26:57Z by jhb MFC: Add preliminary support for binding interrupts to CPUs: - Add a new method to interrupt events (ie_assign_cpu) for binding events to CPUs. - Bind ithreads to CPUs in their main loop (this is the original implementation, it may be changed if the cpuset stuff is MFC'd). - Add intr_event_bind(). - Implement ie_assign_cpu() and a 'intr_bind()' wrapper on amd64, i386, and sparc64. Revision Changes Path 1.34.2.3 +52 -9 src/sys/amd64/amd64/intr_machdep.c 1.18.2.1 +3 -0 src/sys/amd64/include/intr_machdep.h 1.17.2.1 +2 -2 src/sys/arm/arm/intr.c 1.29.2.3 +52 -9 src/sys/i386/i386/intr_machdep.c 1.20.2.1 +3 -0 src/sys/i386/include/intr_machdep.h 1.61.2.2 +1 -1 src/sys/ia64/ia64/interrupt.c 1.147.2.1 +92 -4 src/sys/kern/kern_intr.c 1.13.2.2 +1 -1 src/sys/powerpc/powerpc/intr_machdep.c 1.17.2.3 +1 -0 src/sys/sparc64/include/intr_machdep.h 1.27.2.4 +58 -4 src/sys/sparc64/sparc64/intr_machdep.c 1.7.2.1 +1 -1 src/sys/sun4v/sun4v/intr_machdep.c 1.37.2.1 +9 -4 src/sys/sys/interrupt.h From jhb at FreeBSD.org Sat Jun 28 23:55:59 2008 From: jhb at FreeBSD.org (John Baldwin) Date: Sat Jun 28 23:56:07 2008 Subject: cvs commit: src/sys/amd64/amd64 nexus.c src/sys/i386/i386 nexus.c src/sys/kern bus_if.m subr_bus.c src/sys/sparc64/sparc64 nexus.c src/sys/sys bus.h Message-ID: <200806282355.m5SNtxvu032400@repoman.freebsd.org> jhb 2008-06-28 23:54:51 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/amd64/amd64 nexus.c sys/i386/i386 nexus.c sys/kern bus_if.m subr_bus.c sys/sparc64/sparc64 nexus.c sys/sys bus.h Log: SVN rev 180083 on 2008-06-28 23:54:51Z by jhb MFC: Add a BUS_BIND_INTR() method to the bus interface to bind an IRQ resource to a CPU. A bus_bind_intr(9) wrapper routine similar to bus_setup/teardown_intr() is added for device drivers to use. Currently it is supported on amd64, i386, and sparc64 via nexus(4) methods that simply call the intr_bind() routine. Revision Changes Path 1.77.2.2 +14 -0 src/sys/amd64/amd64/nexus.c 1.73.2.1 +14 -0 src/sys/i386/i386/nexus.c 1.34.2.1 +17 -0 src/sys/kern/bus_if.m 1.201.2.3 +31 -0 src/sys/kern/subr_bus.c 1.20.2.1 +15 -0 src/sys/sparc64/sparc64/nexus.c 1.78.2.2 +3 -0 src/sys/sys/bus.h From julian at FreeBSD.org Sun Jun 29 00:17:58 2008 From: julian at FreeBSD.org (Julian Elischer) Date: Sun Jun 29 00:18:04 2008 Subject: cvs commit: src/sys/netinet6 in6_rmx.c Message-ID: <200806290017.m5T0HvOi035374@repoman.freebsd.org> julian 2008-06-29 00:17:45 UTC FreeBSD src repository Modified files: sys/netinet6 in6_rmx.c Log: SVN rev 180084 on 2008-06-29 00:17:45Z by julian Rename two vars so that they are different from the same vars in ipv4. They are static so it was not a problem 'per se' but it was confusing to the reader. Obtained from: vimage tree Revision Changes Path 1.22 +7 -7 src/sys/netinet6/in6_rmx.c From julian at FreeBSD.org Sun Jun 29 00:25:29 2008 From: julian at FreeBSD.org (Julian Elischer) Date: Sun Jun 29 00:25:31 2008 Subject: cvs commit: src/sys/netinet6 ip6_ipsec.c Message-ID: <200806290025.m5T0PSn3036319@repoman.freebsd.org> julian 2008-06-29 00:25:16 UTC FreeBSD src repository Modified files: sys/netinet6 ip6_ipsec.c Log: SVN rev 180085 on 2008-06-29 00:25:16Z by julian Oops, we've been incrementing the wrong cantforward variable. Obtained from: vimage tree Revision Changes Path 1.10 +2 -1 src/sys/netinet6/ip6_ipsec.c From julian at FreeBSD.org Sun Jun 29 00:50:00 2008 From: julian at FreeBSD.org (Julian Elischer) Date: Sun Jun 29 00:50:07 2008 Subject: cvs commit: src/sys/netipsec key.c Message-ID: <200806290050.m5T0o01W047468@repoman.freebsd.org> julian 2008-06-29 00:49:50 UTC FreeBSD src repository Modified files: sys/netipsec key.c Log: SVN rev 180086 on 2008-06-29 00:49:50Z by julian Enter the 1990s. Use real function declaration. Revision Changes Path 1.33 +1 -1 src/sys/netipsec/key.c From julian at FreeBSD.org Sun Jun 29 01:05:01 2008 From: julian at FreeBSD.org (Julian Elischer) Date: Sun Jun 29 01:05:08 2008 Subject: cvs commit: src/sys/rpc rpcb_clnt.c Message-ID: <200806290105.m5T151hx049743@repoman.freebsd.org> julian 2008-06-29 01:04:48 UTC FreeBSD src repository Modified files: sys/rpc rpcb_clnt.c Log: SVN rev 180087 on 2008-06-29 01:04:48Z by julian It may be #if 0'd out code, but change a varname to not shadow a global. Revision Changes Path 1.3 +4 -4 src/sys/rpc/rpcb_clnt.c From brde at optusnet.com.au Sun Jun 29 01:42:16 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sun Jun 29 01:42:25 2008 Subject: cvs commit: src/lib/libc/i386/gen _setjmp.S setjmp.S In-Reply-To: <20080628180230.GA37313@zim.MIT.EDU> References: <200806281758.m5SHwIl2083857@repoman.freebsd.org> <20080628180230.GA37313@zim.MIT.EDU> Message-ID: <20080629110524.W92369@delplex.bde.org> On Sat, 28 Jun 2008, David Schultz wrote: > On Sat, Jun 28, 2008, David Schultz wrote: >> das 2008-06-28 17:58:06 UTC >> >> FreeBSD src repository >> >> Modified files: >> lib/libc/i386/gen _setjmp.S setjmp.S >> Log: >> SVN rev 180081 on 2008-06-28 17:58:06Z by das >> >> We should also save and restore the MXCSR as on amd64, but detecting >> whether the CPU supports SSE or not here is rather odious. > > Err, the first line of the commit message read: > > Don't clobber the FPU exception flags in longjmp. C99 requires them > to remain unchanged. > > ...but got cut off somehow. This is wrong. It breaks longjmp() from all COMPAT_[3-4] signal handlers (not just ones for SIGFPE). I don't like the corresponding change for amd64 either, and only approved it since amd64 doesn't support COMPAT_[3-4] signal handlers. With COMPAT_[3-4] signal handlers on i386, the entire FP environment of the interrupted context (except for the exception flags in the case of a SIGFPE for the FPU) are passed raw to the signal handler. We still restore the control word in longjmp(). Now we leave garbage in the tag and status words, (except for the exception flags in the case of a SIGFPE for the FPU -- these are clobbered (reset to 0) earlier and we have no way to recover them (except in RELENG_[1-4], they may be recovered from the saved exception status word in the pcb (probably need kmem to read this)). With !COMPAT_[3-4] signal handlers, the exception flags are (now unnecessarily) clobbered by signal handling before longjmp() from a signal handler can clobber them. longjmp() on i386 must clobber (reset to 0 or another safe state) the tag and status words (and perhaps other state like the error pointers) on i386. This is less controversial than restoring the control word and exception status words, since they are an implementation-defined part of the fenv API and ABI and resetting them has effects invisible to the API and ABI (only setting them to an unsafe state would have visible effects). Resetting these while preserving the status word would be slow; the unbroken version just used fninit to reset them together with the exception flags. Bruce From das at FreeBSD.ORG Sun Jun 29 02:22:39 2008 From: das at FreeBSD.ORG (David Schultz) Date: Sun Jun 29 02:22:43 2008 Subject: cvs commit: src/lib/libc/i386/gen _setjmp.S setjmp.S In-Reply-To: <20080629110524.W92369@delplex.bde.org> References: <200806281758.m5SHwIl2083857@repoman.freebsd.org> <20080628180230.GA37313@zim.MIT.EDU> <20080629110524.W92369@delplex.bde.org> Message-ID: <20080629022323.GA39584@zim.MIT.EDU> On Sun, Jun 29, 2008, Bruce Evans wrote: > On Sat, 28 Jun 2008, David Schultz wrote: > > Don't clobber the FPU exception flags in longjmp. C99 requires them > > to remain unchanged. > > > >...but got cut off somehow. > > This is wrong. It breaks longjmp() from all COMPAT_[3-4] signal > handlers (not just ones for SIGFPE). I don't like the corresponding > change for amd64 either, and only approved it since amd64 doesn't > support COMPAT_[3-4] signal handlers. How is it possible for an application linked against an 8.X libc, and which is doing anything even remotely close to reasonable, to get COMPAT_3 or COMPAT_4 signal behavior? From brde at optusnet.com.au Sun Jun 29 03:25:42 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sun Jun 29 03:25:52 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <4865F89D.4090207@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628121025.F89039@delplex.bde.org> <4865F89D.4090207@gmx.de> Message-ID: <20080629121025.K92490@delplex.bde.org> On Sat, 28 Jun 2008, Christoph Mallon wrote: > Bruce Evans wrote: >> On Sat, 28 Jun 2008, Christoph Mallon wrote: >> >>> I still think, using __volatile only works by accident. volatile for an >>> assembler block mostly means "this asm statement has an effect, even >>> though the register specification looks otherwise, so do not optimise this >>> away (i.e. no CSE, do not remove if result is unused etc.). >> >> Right. Though I've never seen unnecessary's __volatiles significantly >> affecting i386 code. This is because the code in the asms can't be >> removed completely, and can't be moved much either. With out of order >> execution, the type of moves that are permitted (not across dependencies) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> are precisely the type of moves that the CPU's scheduler can do or undo >> no matter how the compiler orders the code. > > I disagree. For example look at the use of in_addword() in dev/sk/if_sk.cv in > line 2819: > csum1 = htons(csum & 0xffff); > csum2 = htons((csum >> 16) & 0xffff); > ipcsum = in_addword(csum1, ~csum2 & 0xffff); > /* checksum fixup for IP options */ > len = hlen - sizeof(struct ip); > if (len > 0) { > return; > } > > The calculation will be executed even if the following if (len > 0) leaves > the function and the value of ipcsum is unused. > If in_addword() is not marked volatile it can be moved after the if and not > be executed in all cases. csum1 and csum2 can be moved after the if, too. No, volatile has no effect on whether the above calculation will be executed, since the early return has no dependencies on the caclulation. Old versions of gcc used to handle volatile like that, but this changed in gcc-3 or earlier. gcc.info now says: % The `volatile' keyword indicates that the instruction has important % side-effects. GCC will not delete a volatile `asm' if it is reachable. ^^^^^^^^^^^^^^^^^^^ % (The instruction can still be deleted if GCC can prove that % control-flow will never reach the location of the instruction.) Note % that even a volatile `asm' instruction can be moved relative to other ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ % code, including across jump instructions. For example, on many targets ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Even if gcc didn't move the caclulation, then CPUs with out of order execution might schedule it so that it is effectively never executed (most likely by executing it in otherwise-unused pipelines while the main pipeline returns). This is valid for the same reasons that gcc can move the volatile asms -- the return doesn't depend on the result of the caclulation. The above C code is fairly bad, but generates not so bad code on i386: % % movl %esi, %eax % #APP % xchgb %ah, %al # byte operations can be slow; this one not # too bad, but I wonder if rorw $8 is better # (rorl $16 is already used for corresponding # 32-bit operations) where there is no xchg # alternative % #NO_APP % shrl $16, %esi % movl %esi, %edx % #APP % xchgb %dh, %dl # as above % #NO_APP % notl %edx # poor asm code -- the top 16 bits are unused # except here to stall for merging them with # the previous byte operation % movzwl %ax, %eax % #APP % addw %dx, %ax % adcw $0, %ax % #NO_APP % movl %eax, %edx >>> On a related note: Is inline assembler really necessary here? For example >>> couldn't in_addword() be written as >>> static __inline u_short >>> in_addword(u_short const sum, u_short const b) >>> { >>> u_int const t = sum + b; >>> return t + (t >> 16); >>> } ? >>> This should at least produce equally good code and because the compiler >>> has more knowledge about it than an assembler block, it potentially leads >>> to better code. I have no SPARC compiler at hand, though. >> >> Last time I tried on i386, I couldn't get gcc to generate operations >> involving carries for things like this, or the bswap instruction from >> C code to reorder a word. gcc4.2 -O3 on i386 now generates for the above: >> >> movzwl b, %eax # starting from b and sum in memory >> movzwl sum, %edx >> addl %eax, %edx # 32-bit add >> movl %edx, %eax >> shrl $16, %eax # it does the shift laboriously >> addl %edx, %eax >> movzwl %ax, %eax # don't really need 32-bit result >> # but need something to discard the high bits > > If the upper 16 bits are not "looked at" then the final movzwl can be > optimised away. Many instructions, like add, shl and mul, can live with > "garbage" in the upper 16 bits. This depends on whether the bits are "looked at". In general on i386 and amd64, operating on garbage in the top bits is a a pessimization. It causes "partial register stalls" on some CPUs starting with about PPro. This slowness is smaller on newer CPUs starting with about Athlon from Amd and PentiumM (?) from Intel. But Athlons have similar stalls for mismatched sizes in loads after stores. OTOH, movzwl is fast starting with about the same genration of CPUs that has partial register stalls. > Only if a "bad" instruction, like shr or div, > is encountered, the upper 16 bits have to be cleared. > The current x86 implementation of in_addword() using inline assembler causes > the compiler to add a movzwl, too, before the return. It is the compiler doing this, presumably because something needs a full 32-bit word. Everything in the asm and C code in the inline function deals with 16-bit words. >> In non-inline asm, I would write this as: >> >> movw sum,%ax >> addw b,%ax >> adcw $0,%ax >> movzwl %ax,%eax > > You do not want to use 16bit instructions on modern x86 processors. These > instructions are slow. Intel states that decoding a 16bit operation takes 6 > cycles instead of the usual 1. (Intel® 64 and IA-32 Architectures > Optimization Reference Manual, section 2.1.2.2 Instruction PreDecode) Yes I do, especially on modern x86's. Back in 1988, prefix bytes always took longer to fetch and decode. Back in 1997, partial register stalls made code optimized for space to use subregister operations run slow, but I think that was more for code completely unaware of the problem (operating in bytes and then using the whole word would give the P.R. stall, and I think operating on the 2 low bytes gives a similar stall on some CPUs). The above is not so bad. I would only expect it to run slowly on very old machines where the prefixes cost and the movzwl is slow. But on the old machines, to avoid the prefixes you would have to use something like movzwl to expand everything to 32 bits before operating, and the code would be slow for other reasons. It takes API/ABI changes (use 32 bits for all interfaces) to avoid the prefixes. On newer x86's with better pipelines, prefix bytes are almost free. E.g., the Athlon64 optimization manual says to use operand size prefixes for fill bytes in many cases, because this takes less resources than nop's. gcc and/or gas implements this. I don't believe 6 cycles extra just for decoding on any CPU. >> Pipelining can make bloated code run better than it looks, but probably >> not for the generated code above, since shifts are slow on some i386's >> and there is an extra dependency for the extra shift operation. > > Shifts were slow on early generations of the Pentium 4. Intel corrected this > "glitch" in later generations. Also in original i386 and some Athlons (speed of shifts is bad and/or went backwards relative to add). Fixed in i486 and Phenom(?). Athlons made sh[lr]d especially slow, and IIRC there is a case where add is better than shl $1. Bruce From brde at optusnet.com.au Sun Jun 29 04:00:22 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sun Jun 29 04:00:27 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <4865F895.8030600@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080628114028.M89039@delplex.bde.org> <4865F895.8030600@gmx.de> Message-ID: <20080629132901.Q92490@delplex.bde.org> On Sat, 28 Jun 2008, Christoph Mallon wrote: > Bruce Evans wrote: >> Does sparc64 do anything good with this? Later it says that this is [cc clobber in asm statements] >> irrelevant for the type type of bug in in_cksum.* (expecting cc to >> be preserved across separate asms). From gcc.info: > > I think the bug was the following: > subcc %foo, %bar, %g0 /* SPARC compare */ > #APP > /* inline assembler of in_addword() here, which modifies the condition codes > */ > #NO_APP > bpl $somewhere /* condition branch depending on condition code */ Is the non-#APP part all from C code? > The bpl is supposed to jump depending on the condition of the subcc, but if > the compiler schedules the inline assembler block between the subcc and the > bpl, it jumps depending on garbage. > It was not expected to preserve the condition codes across separate inline > assembler blocks, but it was not specified that the single inline assembler > block did not modify the condition codes. Certainly if gcc wants to put asm block in the middle of generated code, then it needs to know all the clobbers. >> % code left by the assembler instruction. However, when we attempted to >> % implement this, we found no way to make it work reliably. The problem >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> % is that output operands might need reloading, which would result in >> % additional following "store" instructions. On most machines, these >> % instructions would alter the condition code before there was time to >> % test it. This problem doesn't arise for ordinary "test" and "compare" >> % instructions because they don't have any output operands. >> % % For reasons similar to those described above, it is not possible to >> ^^^^^^^^^^^^^^^^^^^^^ >> % give an assembler instruction access to the condition code left by >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> % previous instructions. >> ^^^^^^^^^^^^^^^^^^^^^^ > > I think the excerpt does not apply here, because the problem is the other way > round (inline assembler interrupting C, not C interrupting multiple inline > assembler blocks), see above. Yes, it doesn't apply to any asm code wrapped in inline functions (except to the internals of the functions). No one would expect an inline function to return results in the condition codes. >> On i386, specifying cc has no effect (the compiler must always assume that >> cc is clobbered), and specifying cc in asms is a style bug in FreeBSD. > > I have to disagree. Where does the GCC documentation state, that "cc" has no > effect on x86? It is the other way round: The compiler assumes, the condition > codes are *not* modified, if it is not explicitely stated. Exactly this > caused the bug (though here on SPARC), which was tried to be solved by > volatile. I still am convinced that specifying "cc" in the clobber list and > not using volatile is the correct solution. If it is a style bug to specify > "cc", the style should be changed, otherwise you cannot use inline assembler > correctly. Well, in gcc-1.40, gcc.info says nothing about "cc", apparently because "cc" didn't exist in gcc-1.40. gcc-1.40 had to assume that cc was always clobbered. I think there are enough old i386 asm statements around (e.g., almost all in FreeBSD and many in Linux) for gcc to preserve compatiblity by not changing its assumption (for insignificant benefits since gcc doesn't really understand cc in asms). (Linux-2.6.10/include/ asm-i386/*.h has 177 __asm__ statements (133 with volatile) with only 11 "cc"'s. Examples of ones without "cc"'s are everything in string.h, where things like scasb for strcmp() normally clobber "cc".) Newer versions of gcc.info have to say to use it in clobber lists, to support CPUs with a shorter history than i386 and to prepare for changing the assumptions on i386. Bruce From brde at optusnet.com.au Sun Jun 29 04:07:25 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sun Jun 29 04:07:28 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080628112358.GK1215@alchemy.franken.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <20080628120700.S89039@delplex.bde.org> <20080628112358.GK1215@alchemy.franken.de> Message-ID: <20080629140037.H92490@delplex.bde.org> On Sat, 28 Jun 2008, Marius Strobl wrote: > On Sat, Jun 28, 2008 at 12:09:13PM +1000, Bruce Evans wrote: >> On Sat, 28 Jun 2008, Marius Strobl wrote: >> I think it still only works by chance. [volatile] >> >>> Besides I don't think using "__volatile" for this is >>> that wrong, given that the sparc64 code generated by using "cc" >>> and "__volatile" is nearly identical and given that at least i386 >>> relies on "__volatile" telling GCC that the inline assembler uses >>> the condition codes since quite some time. So the condition codes >>> are probably part of what GCC treats as "important side-effects". >> >> No, the comments in the i386 version are rotted bits left over from >> old failing attempts to fix this problem. > > So by now you are saying that using "__volatile" in this case > is the wrong solution and that using "cc" is a style bug. How > am I supposed to tell the compiler that the inline assembler > alters the condition codes then, which it apparently needs to > know as it at shown to otherwise generate broken code even > when using a single __asm() for the istructions. Using volatile is wrong in this and most cases. Use "cc" when it isn't a style bug :-). I guess this is on some CPUs including sparc64 now, and in all new asms on i386. Bruce From brde at optusnet.com.au Sun Jun 29 04:13:39 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sun Jun 29 04:13:48 2008 Subject: cvs commit: src/lib/libc/i386/gen _setjmp.S setjmp.S In-Reply-To: <20080629022323.GA39584@zim.MIT.EDU> References: <200806281758.m5SHwIl2083857@repoman.freebsd.org> <20080628180230.GA37313@zim.MIT.EDU> <20080629110524.W92369@delplex.bde.org> <20080629022323.GA39584@zim.MIT.EDU> Message-ID: <20080629140956.G92769@delplex.bde.org> On Sat, 28 Jun 2008, David Schultz wrote: > On Sun, Jun 29, 2008, Bruce Evans wrote: >> On Sat, 28 Jun 2008, David Schultz wrote: >>> Don't clobber the FPU exception flags in longjmp. C99 requires them >>> to remain unchanged. >>> >>> ...but got cut off somehow. >> >> This is wrong. It breaks longjmp() from all COMPAT_[3-4] signal >> handlers (not just ones for SIGFPE). I don't like the corresponding >> change for amd64 either, and only approved it since amd64 doesn't >> support COMPAT_[3-4] signal handlers. > > How is it possible for an application linked against an 8.X libc, > and which is doing anything even remotely close to reasonable, > to get COMPAT_3 or COMPAT_4 signal behavior? Nevermind. I forgot the layering. Bruce From kan at FreeBSD.org Sun Jun 29 04:33:52 2008 From: kan at FreeBSD.org (Alexander Kabaev) Date: Sun Jun 29 04:33:58 2008 Subject: cvs commit: src/sys/netinet6 in6_rmx.c Message-ID: <200806290433.m5T4Xp9H069741@repoman.freebsd.org> kan 2008-06-29 04:33:45 UTC FreeBSD src repository Modified files: sys/netinet6 in6_rmx.c Log: SVN rev 180088 on 2008-06-29 04:33:45Z by kan Repair botched variable rename. Pointy hat to: julian Revision Changes Path 1.23 +4 -4 src/sys/netinet6/in6_rmx.c From christoph.mallon at gmx.de Sun Jun 29 05:04:10 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sun Jun 29 05:04:15 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080629140037.H92490@delplex.bde.org> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <20080628120700.S89039@delplex.bde.org> <20080628112358.GK1215@alchemy.franken.de> <20080629140037.H92490@delplex.bde.org> Message-ID: <486717C7.4050504@gmx.de> Bruce Evans wrote: >> So by now you are saying that using "__volatile" in this case >> is the wrong solution and that using "cc" is a style bug. How >> am I supposed to tell the compiler that the inline assembler >> alters the condition codes then, which it apparently needs to >> know as it at shown to otherwise generate broken code even >> when using a single __asm() for the istructions. > > Using volatile is wrong in this and most cases. > > Use "cc" when it isn't a style bug :-). I guess this is on some CPUs > including sparc64 now, and in all new asms on i386. What is the rule for "cc" being a style bug? In style(9) nothing is mentioned. The only sensible rule I can think of is "if the assembler code destroys the condition codes, add "cc" to the clobber list, so GCC is informed about this and can get the data depedencies right". Regards Christoph From christoph.mallon at gmx.de Sun Jun 29 06:55:51 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sun Jun 29 06:56:00 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080629132901.Q92490@delplex.bde.org> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080628114028.M89039@delplex.bde.org> <4865F895.8030600@gmx.de> <20080629132901.Q92490@delplex.bde.org> Message-ID: <486731F4.9050400@gmx.de> Bruce Evans wrote: >> I think the bug was the following: >> subcc %foo, %bar, %g0 /* SPARC compare */ >> #APP >> /* inline assembler of in_addword() here, which modifies the condition >> codes */ >> #NO_APP >> bpl $somewhere /* condition branch depending on condition code */ > > Is the non-#APP part all from C code? Yes, this is an example what happend here: The compiler moved an assembler block between code generated from C. The C code uses the condition codes, but the assembler code destroys the contents. This happens, when "cc" is not in the clobber list, because the C compiler (correctly) assumes that the assembler code does not modify the condition codes. > Certainly if gcc wants to put asm block in the middle of generated code, > then > it needs to know all the clobbers. Exactly. It's always a good idea not to lie to your compiler. (: >>> % % For reasons similar to those described above, it is not possible to >>> ^^^^^^^^^^^^^^^^^^^^^ >>> % give an assembler instruction access to the condition code left by >>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >>> % previous instructions. >>> ^^^^^^^^^^^^^^^^^^^^^^ >> >> I think the excerpt does not apply here, because the problem is the >> other way round (inline assembler interrupting C, not C interrupting >> multiple inline assembler blocks), see above. > > Yes, it doesn't apply to any asm code wrapped in inline functions (except > to the internals of the functions). No one would expect an inline function > to return results in the condition codes. This is not about returing values in condition codes or inline functions. It is about an assembler block, which simply destroys the contents of the condition codes and the compiler not expecting this (because nobody told it). >>> On i386, specifying cc has no effect (the compiler must always assume >>> that >>> cc is clobbered), and specifying cc in asms is a style bug in FreeBSD. >> >> I have to disagree. Where does the GCC documentation state, that "cc" >> has no effect on x86? It is the other way round: The compiler assumes, >> the condition codes are *not* modified, if it is not explicitely >> stated. Exactly this caused the bug (though here on SPARC), which was >> tried to be solved by volatile. I still am convinced that specifying >> "cc" in the clobber list and not using volatile is the correct >> solution. If it is a style bug to specify "cc", the style should be >> changed, otherwise you cannot use inline assembler correctly. > > Well, in gcc-1.40, gcc.info says nothing about "cc", apparently because > "cc" didn't exist in gcc-1.40. gcc-1.40 had to assume that cc was > always clobbered. I think there are enough old i386 asm statements GCC 1.4? Would you please bury this corpse again? It smells. (; > around (e.g., almost all in FreeBSD and many in Linux) for gcc to preserve > compatiblity by not changing its assumption (for insignificant benefits I've seen quite a bit assembler code break when going from 3.x to 4.x, because GCC moves assembler blocks more while scheduling and takes the given in/out constraints much more seriously. > since gcc doesn't really understand cc in asms). (Linux-2.6.10/include/ It's not about "understanding cc in asms", it's just about telling GCC that the "cc" does not survive the assembler block undamaged. > asm-i386/*.h has 177 __asm__ statements (133 with volatile) with only > 11 "cc"'s. Examples of ones without "cc"'s are everything in string.h, > where things like scasb for strcmp() normally clobber "cc".) > > Newer versions of gcc.info have to say to use it in clobber lists, to > support CPUs with a shorter history than i386 and to prepare for changing > the assumptions on i386. Three fourth of the assembler statements needing volatile seems way to high to me. I did not find the strcmp() implementation, which uses scasb. I only found strcmp.S, which is not inline assembler and does not use scas either, which would have suprised me (nor cmpsb, which would be more logical). Regards Christoph From jb at FreeBSD.org Sun Jun 29 07:16:09 2008 From: jb at FreeBSD.org (John Birrell) Date: Sun Jun 29 07:16:16 2008 Subject: cvs commit: src/usr.sbin/config mkmakefile.c Message-ID: <200806290716.m5T7G81D095005@repoman.freebsd.org> jb 2008-06-29 07:15:57 UTC FreeBSD src repository Modified files: usr.sbin/config mkmakefile.c Log: SVN rev 180089 on 2008-06-29 07:15:57Z by jb Revice the way the CTF conversion is done per object. Avoid creating a second shell (which was the problem with the original implementation) and avoid letting make see an empty definition (which was the problem with the current implementation). Revision Changes Path 1.94 +1 -1 src/usr.sbin/config/mkmakefile.c From bz at FreeBSD.org Sun Jun 29 07:34:40 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Jun 29 07:34:42 2008 Subject: cvs commit: src/sys/netinet6 ip6_ipsec.c Message-ID: <200806290734.m5T7Ydcp096413@repoman.freebsd.org> bz 2008-06-29 07:34:21 UTC FreeBSD src repository Modified files: sys/netinet6 ip6_ipsec.c Log: SVN rev 180090 on 2008-06-29 07:34:21Z by bz Try to fix errors introduced in svn180085/cvs rev. 1.10: * Include ip6_var.h for ip6stat. * Use the correct name under ip6stat: `ip6s_cantforward' instead of its IPv4 counterpart. MFC after: 10 days Revision Changes Path 1.11 +2 -1 src/sys/netinet6/ip6_ipsec.c From christoph.mallon at gmx.de Sun Jun 29 08:36:33 2008 From: christoph.mallon at gmx.de (Christoph Mallon) Date: Sun Jun 29 08:36:36 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <20080629121025.K92490@delplex.bde.org> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628121025.F89039@delplex.bde.org> <4865F89D.4090207@gmx.de> <20080629121025.K92490@delplex.bde.org> Message-ID: <4867498D.5050409@gmx.de> Bruce Evans wrote: >>> Right. Though I've never seen unnecessary's __volatiles significantly >>> affecting i386 code. This is because the code in the asms can't be >>> removed completely, and can't be moved much either. With out of order >>> execution, the type of moves that are permitted (not across >>> dependencies) > > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >>> are precisely the type of moves that the CPU's scheduler can do or undo >>> no matter how the compiler orders the code. >> >> I disagree. For example look at the use of in_addword() in >> dev/sk/if_sk.cv in line 2819: >> csum1 = htons(csum & 0xffff); >> csum2 = htons((csum >> 16) & 0xffff); >> ipcsum = in_addword(csum1, ~csum2 & 0xffff); >> /* checksum fixup for IP options */ >> len = hlen - sizeof(struct ip); >> if (len > 0) { >> return; >> } >> >> The calculation will be executed even if the following if (len > 0) >> leaves the function and the value of ipcsum is unused. >> If in_addword() is not marked volatile it can be moved after the if >> and not be executed in all cases. csum1 and csum2 can be moved after >> the if, too. > > No, volatile has no effect on whether the above calculation will be > executed, since the early return has no dependencies on the caclulation. The volatile induces a dependency. > Old versions of gcc used to handle volatile like that, but this changed > in gcc-3 or earlier. gcc.info now says: > > % The `volatile' keyword indicates that the instruction has important > % side-effects. GCC will not delete a volatile `asm' if it is reachable. > ^^^^^^^^^^^^^^^^^^^ This is not about whether the code is reachable or not (it is reachable), it is about whether the result is used (i.e. whether the code is dead). > % (The instruction can still be deleted if GCC can prove that > % control-flow will never reach the location of the instruction.) Note > % that even a volatile `asm' instruction can be moved relative to other > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > % code, including across jump instructions. For example, on many targets > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ jump != conditional jump. If it moves the *volatile* asm statement across the if, it would only appear on *some* execution paths, which is wrong. It is perfectly fine to move it, when the statement is not volatile, though. > Even if gcc didn't move the caclulation, then CPUs with out of order > execution might schedule it so that it is effectively never executed > (most likely by executing it in otherwise-unused pipelines while the > main pipeline returns). This is valid for the same reasons that gcc > can move the volatile asms -- the return doesn't depend on the result > of the caclulation. This is for the CPU to decide. If the assembler block really contains "important" stuff like memory barriers, writes to machine control registers etc., the CPU will not "effectively never execute" the code. The compiler does not know this, all it sees is the word "volatile". > The above C code is fairly bad, but generates not so bad code on i386: > > % % movl %esi, %eax > % #APP > % xchgb %ah, %al # byte operations can be slow; this one not > # too bad, but I wonder if rorw $8 is better > # (rorl $16 is already used for corresponding > # 32-bit operations) where there is no xchg > # alternative > % #NO_APP And this again is an example why not to use inline assembler, but let the compiler decide this: unsigned short swap16(unsigned short x) { return x >> 8 | x << 8; } is compiled to swap16: movl 4(%esp), %eax rolw $8, %ax ret The compiler is even able to do optimisations, which it absolutely cannot do, if inline assembler is used, example: unsigned short id(unsigned short x) { return swap16(swap16(x)); } results in id: movl 4(%esp), %eax ret Maybe the MD htons() macros should be replaced by MI code. > % shrl $16, %esi > % movl %esi, %edx > % #APP > % xchgb %dh, %dl # as above > % #NO_APP > % notl %edx # poor asm code -- the top 16 bits are unused > # except here to stall for merging them with > # the previous byte operation The compiler simply does not know, that the inline assembler only operates on parts of the register. Another reason not to use inline assembler: u_int g(u_short x) { return ~swap16(x); } g: movl 4(%esp), %eax rolw $8, %ax movzwl %ax, %eax # avoid stall notl %eax ret Regards Christoph From brde at optusnet.com.au Sun Jun 29 09:47:45 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sun Jun 29 09:47:55 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <486731F4.9050400@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080628114028.M89039@delplex.bde.org> <4865F895.8030600@gmx.de> <20080629132901.Q92490@delplex.bde.org> <486731F4.9050400@gmx.de> Message-ID: <20080629174222.V93221@delplex.bde.org> On Sun, 29 Jun 2008, Christoph Mallon wrote: > Bruce Evans wrote: >>> I have to disagree. Where does the GCC documentation state, that "cc" has >>> no effect on x86? It is the other way round: The compiler assumes, the >> >> Well, in gcc-1.40, gcc.info says nothing about "cc", apparently because >> "cc" didn't exist in gcc-1.40. gcc-1.40 had to assume that cc was >> always clobbered. I think there are enough old i386 asm statements > > GCC 1.4? Would you please bury this corpse again? It smells. (; I don't remember using gcc.1.4, but many of the basic asm statements in FreeBSD and Linux have the same age as gcc-1.40. >> since gcc doesn't really understand cc in asms). (Linux-2.6.10/include/ >> asm-i386/*.h has 177 __asm__ statements (133 with volatile) with only >> 11 "cc"'s. Examples of ones without "cc"'s are everything in string.h, >> where things like scasb for strcmp() normally clobber "cc".) >> >> Newer versions of gcc.info have to say to use it in clobber lists, to >> support CPUs with a shorter history than i386 and to prepare for changing >> the assumptions on i386. > > Three fourth of the assembler statements needing volatile seems way to high > to me. Three fourths of asm statements would need cc clobbers with your rule, and only 1 hundredth of them currently have them :-(. I exaggerate slightly -- the ratios might be more like 2/4 and 1/10, depending on how common arithmetic operations are in asm statements (they aren't in most asm statements because C handles plain arithmetic operations well). BTW, what happens for floating point condition codes? Nothing is documented, so we have to guess that the cc clobber covers FP condition codes or that all condition codes are assumed to be clobbered by all asms. > I did not find the strcmp() implementation, which uses scasb. I only > found strcmp.S, which is not inline assembler and does not use scas either, > which would have suprised me (nor cmpsb, which would be more logical). Try linux-x.y.z/include/asm-i386/string.h. Some of this has been imported into FreeBSD, in i386-bitops.h in ext2fs. File systems don't benefit from this much micro-optimization (for i486's), but checksumming does. Bruce From jkoshy at FreeBSD.org Sun Jun 29 10:30:13 2008 From: jkoshy at FreeBSD.org (Joseph Koshy) Date: Sun Jun 29 10:30:19 2008 Subject: cvs commit: src/usr.sbin/pmcstat pmcstat.8 Message-ID: <200806291030.m5TAUCn2022927@repoman.freebsd.org> jkoshy 2008-06-29 10:30:06 UTC FreeBSD src repository Modified files: usr.sbin/pmcstat pmcstat.8 Log: SVN rev 180091 on 2008-06-29 10:30:06Z by jkoshy Document messages printed in verbose mode. Requested by: Fabien Thomas Revision Changes Path 1.14 +36 -2 src/usr.sbin/pmcstat/pmcstat.8 From brde at optusnet.com.au Sun Jun 29 11:28:45 2008 From: brde at optusnet.com.au (Bruce Evans) Date: Sun Jun 29 11:28:55 2008 Subject: cvs commit: src/sys/sparc64/include in_cksum.h In-Reply-To: <4867498D.5050409@gmx.de> References: <200806252105.m5PL5AUp064418@repoman.freebsd.org> <48654667.1040401@gmx.de> <20080627222404.GJ1215@alchemy.franken.de> <48657058.6020102@gmx.de> <20080628121025.F89039@delplex.bde.org> <4865F89D.4090207@gmx.de> <20080629121025.K92490@delplex.bde.org> <4867498D.5050409@gmx.de> Message-ID: <20080629201900.V93605@delplex.bde.org> On Sun, 29 Jun 2008, Christoph Mallon wrote: > Bruce Evans wrote: >>>> Right. Though I've never seen unnecessary's __volatiles significantly >>>> affecting i386 code. This is because the code in the asms can't be >>>> removed completely, and can't be moved much either. With out of order >>>> execution, the type of moves that are permitted (not across dependencies) >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >>>> are precisely the type of moves that the CPU's scheduler can do or undo >>>> no matter how the compiler orders the code. >>> >>> I disagree. For example look at the use of in_addword() in dev/sk/if_sk.cv >>> in line 2819: >>> csum1 = htons(csum & 0xffff); >>> csum2 = htons((csum >> 16) & 0xffff); >>> ipcsum = in_addword(csum1, ~csum2 & 0xffff); >>> /* checksum fixup for IP options */ >>> len = hlen - sizeof(struct ip); >>> if (len > 0) { >>> return; >>> } >>> >>> The calculation will be executed even if the following if (len > 0) leaves >>> the function and the value of ipcsum is unused. >>> If in_addword() is not marked volatile it can be moved after the if and >>> not be executed in all cases. csum1 and csum2 can be moved after the if, >>> too. >> >> No, volatile has no effect on whether the above calculation will be >> executed, since the early return has no dependencies on the caclulation. > > The volatile induces a dependency. Dependency on what? Perhaps the only guarantee is that a volatile asm with no other dependencies is executed at some point before the next function call or return. gcc.info doesn't guarantee anything except not removing a volatile asm, but moving it far enough away to not execute it before return would be equivalent to removing it. >> Old versions of gcc used to handle volatile like that, but this changed >> in gcc-3 or earlier. gcc.info now says: >> >> % The `volatile' keyword indicates that the instruction has important >> % side-effects. GCC will not delete a volatile `asm' if it is reachable. >> ^^^^^^^^^^^^^^^^^^^ > > This is not about whether the code is reachable or not (it is reachable), it > is about whether the result is used (i.e. whether the code is dead). > >> % (The instruction can still be deleted if GCC can prove that >> % control-flow will never reach the location of the instruction.) Note >> % that even a volatile `asm' instruction can be moved relative to other >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> % code, including across jump instructions. For example, on many targets >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > jump != conditional jump. > If it moves the *volatile* asm statement across the if, it would only appear > on *some* execution paths, which is wrong. It is perfectly fine to move it, > when the statement is not volatile, though. I don't see why a plain conditional jump would be restricted. Only returns and possibly function calls must be restricted, since the compiler can't do any dependency analysis across returns so it must execute undead code before return. The above paragraph from gcc.info continues as follows: % there is a system register which can be set to control the rounding % mode of floating point operations. You might try setting it with a % volatile `asm', like this PowerPC example: % % asm volatile("mtfsf 255,%0" : : "f" (fpenv)); % sum = x + y; % % This will not work reliably, as the compiler may move the addition back % before the volatile `asm'. To make it work you need to add an % artificial dependency to the `asm' referencing a variable in the code % you don't want moved, for example: Suppose that this example has an if statement: asm volatile(...); if (independent_var) sum = x + y; else sum = y + z; Here there is no reason for this simple if statement to prevent moving the asm any more than the simple assignment, since nothing in any clause of the if statement has any explicit dependencies on the asm, just like the simple assignment has no explicit dependencies. >> Even if gcc didn't move the caclulation, then CPUs with out of order >> execution might schedule it so that it is effectively never executed >> (most likely by executing it in otherwise-unused pipelines while the >> main pipeline returns). This is valid for the same reasons that gcc >> can move the volatile asms -- the return doesn't depend on the result >> of the caclulation. > > This is for the CPU to decide. If the assembler block really contains > "important" stuff like memory barriers, writes to machine control registers > etc., the CPU will not "effectively never execute" the code. The compiler > does not know this, all it sees is the word "volatile". The above example shows that volatile is not enough for a memory barrier. A memory clobber is probably needed for barriers, but interferes with optimization. On ia64, barrier instructions seem to just use volatile, while on i386 the only fairly explicit barrier function is bus_space_barrier(), which uses a dummy locked instruction plus a memory clobber for read and a null instruction plus a mmory clobber for write. The hardware is barely involved, but the memory clobbers are needed to tell the compiler not to move loads and stores across the asm, or at least to force them to memory. >> The above C code is fairly bad, but generates not so bad code on i386: >> >> % % movl %esi, %eax >> % #APP >> % xchgb %ah, %al # byte operations can be slow; this one not >> # too bad, but I wonder if rorw $8 is better >> # (rorl $16 is already used for corresponding >> # 32-bit operations) where there is no xchg >> # alternative >> % #NO_APP > > And this again is an example why not to use inline assembler, but let the > compiler decide this: > > unsigned short swap16(unsigned short x) > { > return x >> 8 | x << 8; > } > > is compiled to > > swap16: > movl 4(%esp), %eax > rolw $8, %ax > ret Ah, that's an example of the compiler turning naive C code into shifts that I couldn't find last time I looked. gcc-4.2.1 doesn't generate quite such good code here: % swap16: % movzwl 4(%esp), %eax % rolw $8, %ax % movzwl %ax, %eax % ret This takes -fomit-frame-pointer and -O (arch=native). For some arches it generates movl instead of the first movzwl. For plain i386 it generates movl but should generate movw. It seems to generate the final movzwl. This is bogus, but is probably required by the i386 ABI (to be backwards compatible with unprototyped functions). On amd64, it does a direct rolw on a register arg (I think the movzwl still gets done, but in the caller), and it still does the movzwl on the return value. gcc-3.3.3 generates horrible code (movl; movl; shrw; movzwl; sall; orl; movzwl with -O and movzwl; movl; shrw; sall; orl; movzwl with -O2). gcc-4.2.1 produces worse code than the above for the __byte_swap_word_const() macro in i386/include/endian.h, and good code like yours for __word_swap_int_const() there, and disgustingly bad code for __bswap64() there. i386.md doesn't mention bswap, so the asm version is out of reach, but surely the 64-bit ints can be handled better. >> % shrl $16, %esi >> % movl %esi, %edx >> % #APP >> % xchgb %dh, %dl # as above >> % #NO_APP >> % notl %edx # poor asm code -- the top 16 bits are unused >> # except here to stall for merging them with >> # the previous byte operation > > The compiler simply does not know, that the inline assembler only operates on > parts of the register. Another reason not to use inline assembler: Actually, this is poor compiler code and the compiler does know here -- the asm returned a u_short, and the compiler doesn't use the top 16 bits except to let garbage accumulate in them. This might not actually stall, so doing this would be good, for the target CPU. I forget what the target CPU was. > u_int g(u_short x) > { > return ~swap16(x); > } > > g: > movl 4(%esp), %eax > rolw $8, %ax > movzwl %ax, %eax # avoid stall > notl %eax > ret Inlining is good for avoiding a movzwl's in both the callee and the caller. Here you have changed the interface a little -- g() returns 32 bits -- so the movzwl and the notl are good for implementing C's promotion rules and it isn't clear that they are to avoid the stall. However, when I changed g() to return u_short, they are still generated, even on plain i386 where there is a slow movzwl and no stalls. Bruce From philip at FreeBSD.org Sun Jun 29 12:52:00 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sun Jun 29 12:52:07 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200806291252.m5TCq0Ar046004@repoman.freebsd.org> philip 2008-06-29 12:51:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/conf NOTES Log: SVN rev 180092 on 2008-06-29 12:51:49Z by philip MFC r180077: Remove stray "miibus0" reference from ancient kernel config file times. Revision Changes Path 1.1454.2.10 +1 -1 src/sys/conf/NOTES From philip at FreeBSD.org Sun Jun 29 12:54:46 2008 From: philip at FreeBSD.org (Philip Paeps) Date: Sun Jun 29 12:54:53 2008 Subject: cvs commit: src/sys/conf NOTES Message-ID: <200806291254.m5TCsjcH046199@repoman.freebsd.org> philip 2008-06-29 12:54:39 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/conf NOTES Log: SVN rev 180093 on 2008-06-29 12:54:39Z by philip MFC r180077: Remove stray "miibus0" reference from ancient kernel config file times. Revision Changes Path 1.1325.2.41 +1 -1 src/sys/conf/NOTES From ed at FreeBSD.org Sun Jun 29 13:17:17 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Sun Jun 29 13:17:18 2008 Subject: cvs commit: src/sys/net if_loop.c Message-ID: <200806291317.m5TDHGGP049151@repoman.freebsd.org> ed 2008-06-29 13:17:01 UTC FreeBSD src repository Modified files: sys/net if_loop.c Log: SVN rev 180094 on 2008-06-29 13:17:01Z by ed Remove the unused softc from the lo(4) driver. Now that the pseudo-interface cloner has an internal list of instances, there is no need to create a softc. The softc only contains a pointer to the ifp, which means there is no valid reason to keep it. While there, remove the corresponding malloc-pool. Approved by: philip (mentor) Revision Changes Path 1.117 +2 -20 src/sys/net/if_loop.c From das at FreeBSD.org Sun Jun 29 16:35:49 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 16:35:52 2008 Subject: cvs commit: src/contrib/gdtoa gethex.c Message-ID: <200806291635.m5TGZnk6068582@repoman.freebsd.org> das 2008-06-29 16:35:33 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) contrib/gdtoa gethex.c Log: SVN rev 180095 on 2008-06-29 16:35:33Z by das MFC r179918: Bring in the vendor's fix for a bug in strtod() whereby strtod("0xyz", &endp) resulted in endp pointing to "0xyz" instead of "xyz". Revision Changes Path 1.1.1.4.2.1 +5 -2 src/contrib/gdtoa/gethex.c From marck at FreeBSD.org Sun Jun 29 16:56:36 2008 From: marck at FreeBSD.org (Dmitry Morozovsky) Date: Sun Jun 29 16:56:38 2008 Subject: cvs commit: src/usr.sbin/cron/cron cron.8 cron.c cron.h do_command.c Message-ID: <200806291656.m5TGuZso079529@repoman.freebsd.org> marck 2008-06-29 16:56:18 UTC FreeBSD src repository Modified files: usr.sbin/cron/cron cron.8 cron.c cron.h do_command.c Log: SVN rev 180096 on 2008-06-29 16:56:18Z by marck Add -m option to cron(8), overriding default mail recipient for cron mails, unless explicitly provided by MAILTO= line in crontab. This feature can be useful in massive hosting environment, where most users do not care about autogenerated mails. Setting recipient to null string disables default mails at all. Approved by: yar MFC after: 4 weeks Revision Changes Path 1.26 +19 -1 src/usr.sbin/cron/cron/cron.8 1.20 +5 -2 src/usr.sbin/cron/cron/cron.c 1.19 +4 -2 src/usr.sbin/cron/cron/cron.h 1.29 +8 -12 src/usr.sbin/cron/cron/do_command.c From das at FreeBSD.org Sun Jun 29 17:17:22 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 17:17:24 2008 Subject: cvs commit: src/lib/libc/sys socket.2 Message-ID: <200806291717.m5THHLel082390@repoman.freebsd.org> das 2008-06-29 17:17:14 UTC FreeBSD src repository Modified files: lib/libc/sys socket.2 Log: SVN rev 180097 on 2008-06-29 17:17:14Z by das Make it clearer that it is possible to disable the generation of SIGPIPE for individual sockets (PR: kern/118626). While here, s/insure/ensure/. Revision Changes Path 1.32 +4 -4 src/lib/libc/sys/socket.2 From bz at FreeBSD.org Sun Jun 29 17:59:19 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Jun 29 17:59:27 2008 Subject: cvs commit: src/sys/kern kern_cpuset.c src/sys/sys priv.h Message-ID: <200806291759.m5THxIv5085828@repoman.freebsd.org> bz 2008-06-29 17:58:16 UTC FreeBSD src repository Modified files: sys/kern kern_cpuset.c sys/sys priv.h Log: SVN rev 180098 on 2008-06-29 17:58:16Z by bz Add a new priv 'PRIV_SCHED_CPUSET' to check if manipulating cpusets is allowed and replace the suser() call. Do not allow it in jails. Reviewed by: rwatson Revision Changes Path 1.11 +1 -1 src/sys/kern/kern_cpuset.c 1.20 +1 -0 src/sys/sys/priv.h From bz at FreeBSD.org Sun Jun 29 18:26:20 2008 From: bz at FreeBSD.org (Bjoern A. Zeeb) Date: Sun Jun 29 18:26:27 2008 Subject: cvs commit: src/sys/kern kern_intr.c Message-ID: <200806291826.m5TIQKs4089167@repoman.freebsd.org> bz 2008-06-29 18:26:07 UTC FreeBSD src repository Modified files: sys/kern kern_intr.c Log: SVN rev 180099 on 2008-06-29 18:26:07Z by bz Remove an unneeded error variable to make clear that if reaching the end of the function we never return an error. Revision Changes Path 1.164 +1 -3 src/sys/kern/kern_intr.c From alc at FreeBSD.org Sun Jun 29 18:35:23 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sun Jun 29 18:35:26 2008 Subject: cvs commit: src/sys/amd64/include pmap.h Message-ID: <200806291835.m5TIZNLx089927@repoman.freebsd.org> alc 2008-06-29 18:35:00 UTC FreeBSD src repository Modified files: sys/amd64/include pmap.h Log: SVN rev 180100 on 2008-06-29 18:35:00Z by alc Increase the size of the kernel virtual address space to 6GB. Until the maximum size of the kmem map can be greater than 4GB, there is little point in making the kernel virtual address space larger than 6GB. Tested by: kris@ Revision Changes Path 1.141 +4 -5 src/sys/amd64/include/pmap.h From alc at FreeBSD.org Sun Jun 29 19:13:42 2008 From: alc at FreeBSD.org (Alan Cox) Date: Sun Jun 29 19:13:49 2008 Subject: cvs commit: src/sys/amd64/include pmap.h vmparam.h Message-ID: <200806291913.m5TJDg0q096570@repoman.freebsd.org> alc 2008-06-29 19:13:27 UTC FreeBSD src repository Modified files: sys/amd64/include pmap.h vmparam.h Log: SVN rev 180101 on 2008-06-29 19:13:27Z by alc Strictly speaking, the definition of VM_MAX_KERNEL_ADDRESS is wrong. However, in practice, the error (currently) makes no difference because the computation performed by KVADDR() hides the error. This revision fixes the error. Also, eliminate a (now) unused definition. Revision Changes Path 1.142 +0 -1 src/sys/amd64/include/pmap.h 1.52 +1 -1 src/sys/amd64/include/vmparam.h From das at FreeBSD.org Sun Jun 29 21:01:44 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 21:01:47 2008 Subject: cvs commit: src/lib/libc/stdio vfprintf.c vfwprintf.c Message-ID: <200806292101.m5TL1ieH017339@repoman.freebsd.org> das 2008-06-29 21:01:27 UTC FreeBSD src repository Modified files: lib/libc/stdio vfprintf.c vfwprintf.c Log: SVN rev 180102 on 2008-06-29 21:01:27Z by das Begin de-spaghettifying the code that handles positional arguments. In particular, encapsulate the state of the type table in a struct, and add inline functions to initialize, free, and manipulate that state. This replaces some ugly macros that made proper error handling impossible. While here, remove an unneeded test for NULL and a variable that is initialized (many times!) but never used. The compiler didn't catch these because of rampant use of the same variable to mean different things in different places. This commit should not cause any changes in functionality. Revision Changes Path 1.79 +153 -94 src/lib/libc/stdio/vfprintf.c 1.30 +153 -94 src/lib/libc/stdio/vfwprintf.c From das at FreeBSD.org Sun Jun 29 21:03:36 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 21:03:56 2008 Subject: cvs commit: src/tools/regression/lib/libc/stdio Makefile test-print-positional.c Message-ID: <200806292103.m5TL3Vn1018598@repoman.freebsd.org> das 2008-06-29 21:03:14 UTC FreeBSD src repository Modified files: tools/regression/lib/libc/stdio Makefile Added files: tools/regression/lib/libc/stdio test-print-positional.c Log: SVN rev 180103 on 2008-06-29 21:03:14Z by das Add some regression tests for printf() with positional arguments. The first test comes from OpenBSD, and the others are additions or adaptations. This is based on OpenBSD's src/regress/lib/libc/sprintf/sprintf_test.c, v1.3. I deliberately did not use v1.4 because it's bogus. Revision Changes Path 1.5 +1 -1 src/tools/regression/lib/libc/stdio/Makefile 1.1 +132 -0 src/tools/regression/lib/libc/stdio/test-print-positional.c (new) From ru at FreeBSD.org Sun Jun 29 21:13:38 2008 From: ru at FreeBSD.org (Ruslan Ermilov) Date: Sun Jun 29 21:13:48 2008 Subject: cvs commit: src/sys/amd64/include pmap.h In-Reply-To: <200806291835.m5TIZNLx089927@repoman.freebsd.org> References: <200806291835.m5TIZNLx089927@repoman.freebsd.org> Message-ID: <20080629211316.GA20319@edoofus.dev.vega.ru> On Sun, Jun 29, 2008 at 06:35:00PM +0000, Alan Cox wrote: > alc 2008-06-29 18:35:00 UTC > > FreeBSD src repository > > Modified files: > sys/amd64/include pmap.h > Log: > SVN rev 180100 on 2008-06-29 18:35:00Z by alc > > Increase the size of the kernel virtual address space to 6GB. Until the > maximum size of the kmem map can be greater than 4GB, there is little point > in making the kernel virtual address space larger than 6GB. > > Tested by: kris@ > > Revision Changes Path > 1.141 +4 -5 src/sys/amd64/include/pmap.h > Can you please draw a new layout, similar to the one made by Peter? http://lists.freebsd.org/pipermail/freebsd-amd64/2005-July/005578.html Cheers, -- Ruslan Ermilov ru@FreeBSD.org FreeBSD committer From alc at cs.rice.edu Sun Jun 29 21:44:27 2008 From: alc at cs.rice.edu (Alan Cox) Date: Sun Jun 29 21:44:36 2008 Subject: cvs commit: src/sys/amd64/include pmap.h In-Reply-To: <20080629211316.GA20319@edoofus.dev.vega.ru> References: <200806291835.m5TIZNLx089927@repoman.freebsd.org> <20080629211316.GA20319@edoofus.dev.vega.ru> Message-ID: <48680232.4010200@cs.rice.edu> Ruslan Ermilov wrote: >On Sun, Jun 29, 2008 at 06:35:00PM +0000, Alan Cox wrote: > > >>alc 2008-06-29 18:35:00 UTC >> >> FreeBSD src repository >> >> Modified files: >> sys/amd64/include pmap.h >> Log: >> SVN rev 180100 on 2008-06-29 18:35:00Z by alc >> >> Increase the size of the kernel virtual address space to 6GB. Until the >> maximum size of the kmem map can be greater than 4GB, there is little point >> in making the kernel virtual address space larger than 6GB. >> >> Tested by: kris@ >> >> Revision Changes Path >> 1.141 +4 -5 src/sys/amd64/include/pmap.h >> >> >> >Can you please draw a new layout, similar to the one made by Peter? > >http://lists.freebsd.org/pipermail/freebsd-amd64/2005-July/005578.html > > > Change the following line ffffffff80000000 - ffffffffffffffff 2GB kva to fffffffe80000000 - ffffffffffffffff 6GB kva Regards, Alan From das at FreeBSD.org Sun Jun 29 21:52:48 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 21:52:55 2008 Subject: cvs commit: src/lib/libc/stdio Makefile.inc printf-pos.c printflocal.h vfprintf.c vfwprintf.c Message-ID: <200806292152.m5TLqmdB023044@repoman.freebsd.org> das 2008-06-29 21:52:40 UTC FreeBSD src repository Modified files: lib/libc/stdio Makefile.inc vfprintf.c vfwprintf.c Added files: lib/libc/stdio printf-pos.c printflocal.h Log: SVN rev 180104 on 2008-06-29 21:52:40Z by das Reduce the level of duplication between vfprintf() and vfwprintf() by moving the positional argument handling code to a new file, printf-pos.c, and moving common definitions to printflocal.h. No functional change intended. Revision Changes Path 1.39 +2 -1 src/lib/libc/stdio/Makefile.inc 1.1 +788 -0 src/lib/libc/stdio/printf-pos.c (new) 1.1 +94 -0 src/lib/libc/stdio/printflocal.h (new) 1.80 +1 -486 src/lib/libc/stdio/vfprintf.c 1.31 +3 -488 src/lib/libc/stdio/vfwprintf.c From ed at 80386.nl Sun Jun 29 21:56:58 2008 From: ed at 80386.nl (Ed Schouten) Date: Sun Jun 29 21:57:05 2008 Subject: cvs commit: src/sys/amd64/include pmap.h In-Reply-To: <20080629211316.GA20319@edoofus.dev.vega.ru> References: <200806291835.m5TIZNLx089927@repoman.freebsd.org> <20080629211316.GA20319@edoofus.dev.vega.ru> Message-ID: <20080629215657.GI14567@hoeg.nl> Alan, * Ruslan Ermilov wrote: > Can you please draw a new layout, similar to the one made by Peter? > > http://lists.freebsd.org/pipermail/freebsd-amd64/2005-July/005578.html I'm not familiar with the design of amd64 or its page tables, but is there some kind of overhead when all 512 GB's would be allocated as kva? Yours, -- Ed Schouten WWW: http://80386.nl/ -------------- 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/20080629/bcb1ca7f/attachment.pgp From das at FreeBSD.org Sun Jun 29 22:04:34 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 22:04:41 2008 Subject: cvs commit: src/lib/libc/stdio printf-pos.c Message-ID: <200806292204.m5TM4XUF025474@repoman.freebsd.org> das 2008-06-29 22:04:25 UTC FreeBSD src repository Modified files: lib/libc/stdio printf-pos.c Log: SVN rev 180105 on 2008-06-29 22:04:25Z by das Factor out the code that builds the argument table. We don't need separate normal and wide character versions of it. No functional change. Revision Changes Path 1.2 +47 -130 src/lib/libc/stdio/printf-pos.c From das at FreeBSD.org Sun Jun 29 22:54:35 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 22:54:44 2008 Subject: cvs commit: src/lib/libc/stdio printf-pos.c printflocal.h vfprintf.c vfwprintf.c Message-ID: <200806292254.m5TMsZ5d030173@repoman.freebsd.org> das 2008-06-29 22:54:26 UTC FreeBSD src repository Modified files: lib/libc/stdio printf-pos.c printflocal.h vfprintf.c vfwprintf.c Log: SVN rev 180106 on 2008-06-29 22:54:26Z by das Correctly handle malloc() failure. While here, reduce the code size a bit by removing some calls to the inline function addtype(). Revision Changes Path 1.3 +142 -95 src/lib/libc/stdio/printf-pos.c 1.2 +2 -2 src/lib/libc/stdio/printflocal.h 1.81 +9 -3 src/lib/libc/stdio/vfprintf.c 1.32 +9 -3 src/lib/libc/stdio/vfwprintf.c From das at FreeBSD.org Sun Jun 29 23:46:22 2008 From: das at FreeBSD.org (David Schultz) Date: Sun Jun 29 23:46:29 2008 Subject: cvs commit: src/lib/libc/stdio printf-pos.c Message-ID: <200806292346.m5TNkLbK035862@repoman.freebsd.org> das 2008-06-29 23:46:06 UTC FreeBSD src repository Modified files: lib/libc/stdio printf-pos.c Log: SVN rev 180107 on 2008-06-29 23:46:06Z by das Fix a bogon in the previous commit and add some missing error checks. Revision Changes Path 1.4 +12 -9 src/lib/libc/stdio/printf-pos.c From alc at FreeBSD.org Mon Jun 30 02:36:09 2008 From: alc at FreeBSD.org (Alan Cox) Date: Mon Jun 30 02:36:11 2008 Subject: cvs commit: src/sys/amd64/include pmap.h Message-ID: <200806300236.m5U2a8GO062003@repoman.freebsd.org> alc 2008-06-30 02:35:55 UTC FreeBSD src repository Modified files: sys/amd64/include pmap.h Log: SVN rev 180108 on 2008-06-30 02:35:55Z by alc Compute NKPDPE from NKPT. This reduces the number of knobs that must be turned in order to change the size of the kernel virtual address space. Revision Changes Path 1.143 +1 -1 src/sys/amd64/include/pmap.h From alc at FreeBSD.org Mon Jun 30 03:14:55 2008 From: alc at FreeBSD.org (Alan Cox) Date: Mon Jun 30 03:14:57 2008 Subject: cvs commit: src/sys/amd64/include vmparam.h Message-ID: <200806300314.m5U3EtJS066537@repoman.freebsd.org> alc 2008-06-30 03:14:39 UTC FreeBSD src repository Modified files: sys/amd64/include vmparam.h Log: SVN rev 180109 on 2008-06-30 03:14:39Z by alc Document the layout of the address space, borrowing heavily from http://lists.freebsd.org/pipermail/freebsd-amd64/2005-July/005578.html Revision Changes Path 1.53 +12 -2 src/sys/amd64/include/vmparam.h From delphij at FreeBSD.org Mon Jun 30 05:14:57 2008 From: delphij at FreeBSD.org (Xin LI) Date: Mon Jun 30 05:14:59 2008 Subject: cvs commit: src/sys/dev/sound/pci atiixp.h Message-ID: <200806300514.m5U5Euqm088936@repoman.freebsd.org> delphij 2008-06-30 05:14:44 UTC FreeBSD src repository Modified files: sys/dev/sound/pci atiixp.h Log: SVN rev 180110 on 2008-06-30 05:14:44Z by delphij Correct SB600 AC97 Audio PCI ID, otherwise we will attach to the SBx00 PCI to PCI Bridge which causes problems. Revision Changes Path 1.5 +1 -1 src/sys/dev/sound/pci/atiixp.h From mtm at FreeBSD.org Mon Jun 30 08:01:59 2008 From: mtm at FreeBSD.org (Mike Makonnen) Date: Mon Jun 30 08:02:07 2008 Subject: cvs commit: src/etc/periodic/security 800.loginfail Message-ID: <200806300801.m5U81xx9007605@repoman.freebsd.org> mtm 2008-06-30 08:01:47 UTC FreeBSD src repository Modified files: etc/periodic/security 800.loginfail Log: SVN rev 180111 on 2008-06-30 08:01:47Z by mtm Rev. 1.8 broke matching on lines where the failure mode is at the head of the message, such as: Jun 30 10:49:21 rogue sshd[17553]: Invalid user iceman from 127.0.0.1 PR: conf/124569 Submitted by: Taku Revision Changes Path 1.9 +1 -1 src/etc/periodic/security/800.loginfail From danger at FreeBSD.org Mon Jun 30 08:29:16 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 08:29:27 2008 Subject: cvs commit: src/usr.sbin/mountd exports.5 Message-ID: <200806300829.m5U8TGW1010793@repoman.freebsd.org> danger 2008-06-30 08:29:02 UTC FreeBSD src repository Modified files: usr.sbin/mountd exports.5 Log: SVN rev 180112 on 2008-06-30 08:29:02Z by danger - add some missing words - we don't have ne(4), replace it with re(4) - fix markup - bump date Submitted by: ru MFC after: 3 days Revision Changes Path 1.30 +44 -25 src/usr.sbin/mountd/exports.5 From danger at FreeBSD.org Mon Jun 30 08:46:17 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 08:46:20 2008 Subject: cvs commit: src/lib/libc/sys mkdir.2 Message-ID: <200806300846.m5U8kHUQ012360@repoman.freebsd.org> danger 2008-06-30 08:46:09 UTC FreeBSD src repository Modified files: lib/libc/sys mkdir.2 Log: SVN rev 180113 on 2008-06-30 08:46:09Z by danger - Forgot to bump a date after last change Revision Changes Path 1.28 +1 -1 src/lib/libc/sys/mkdir.2 From danger at FreeBSD.org Mon Jun 30 09:06:34 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 09:06:36 2008 Subject: cvs commit: src/lib/libc/sys mkdir.2 Message-ID: <200806300906.m5U96XCR024339@repoman.freebsd.org> danger 2008-06-30 09:06:19 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/sys mkdir.2 Log: SVN rev 180114 on 2008-06-30 09:06:19Z by danger - MFC rev. 180030 and 180113: - add description of the MLINK error - bump a date PR: docs/123019 Revision Changes Path 1.24.2.2 +4 -1 src/lib/libc/sys/mkdir.2 From danger at FreeBSD.org Mon Jun 30 09:08:34 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 09:08:40 2008 Subject: cvs commit: src/lib/libc/sys mkdir.2 Message-ID: <200806300908.m5U98Yp5024496@repoman.freebsd.org> danger 2008-06-30 09:08:29 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libc/sys mkdir.2 Log: SVN rev 180115 on 2008-06-30 09:08:29Z by danger - MFC rev. 180030 and 180113: - add description of the MLINK error - bump a date PR: docs/123019 Revision Changes Path 1.19.12.2 +4 -1 src/lib/libc/sys/mkdir.2 From danger at FreeBSD.org Mon Jun 30 09:10:22 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 09:10:25 2008 Subject: cvs commit: src/lib/libc/sys mkdir.2 Message-ID: <200806300910.m5U9AMAg024711@repoman.freebsd.org> danger 2008-06-30 09:10:04 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) lib/libc/sys mkdir.2 Log: SVN rev 180116 on 2008-06-30 09:10:04Z by danger - MFC rev. 180030 and 180113: - add description of the MLINK error - bump a date PR: docs/123019 Revision Changes Path 1.19.6.1 +4 -1 src/lib/libc/sys/mkdir.2 From danger at FreeBSD.org Mon Jun 30 09:16:23 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 09:16:25 2008 Subject: cvs commit: src/lib/libc/stdlib system.3 Message-ID: <200806300916.m5U9GMlD025235@repoman.freebsd.org> danger 2008-06-30 09:16:22 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) lib/libc/stdlib system.3 Log: SVN rev 180117 on 2008-06-30 09:16:22Z by danger MFC rev. 180024: - Mark the section describing return values with an appropriate section flag. PR: docs/122818 Revision Changes Path 1.12.2.1 +1 -1 src/lib/libc/stdlib/system.3 From danger at FreeBSD.org Mon Jun 30 09:18:09 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 09:18:10 2008 Subject: cvs commit: src/lib/libc/stdlib system.3 Message-ID: <200806300918.m5U9I8no025372@repoman.freebsd.org> danger 2008-06-30 09:17:54 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) lib/libc/stdlib system.3 Log: SVN rev 180118 on 2008-06-30 09:17:54Z by danger MFC rev. 180024: - Mark the section describing return values with an appropriate section flag. PR: docs/122818 Revision Changes Path 1.11.14.1 +1 -1 src/lib/libc/stdlib/system.3 From danger at FreeBSD.org Mon Jun 30 09:18:54 2008 From: danger at FreeBSD.org (Daniel Gerzo) Date: Mon Jun 30 09:18:56 2008 Subject: cvs commit: src/lib/libc/stdlib system.3 Message-ID: <200806300918.m5U9IskL025443@repoman.freebsd.org> danger 2008-06-30 09:18:49 UTC FreeBSD src repository Modified files: (Branch: RELENG_5) lib/libc/stdlib system.3 Log: SVN rev 180119 on 2008-06-30 09:18:49Z by danger MFC rev. 180024: - Mark the section describing return values with an appropriate section flag. PR: 180024 Revision Changes Path 1.11.8.1 +1 -1 src/lib/libc/stdlib/system.3 From delphij at FreeBSD.org Mon Jun 30 15:21:58 2008 From: delphij at FreeBSD.org (Xin LI) Date: Mon Jun 30 15:22:00 2008 Subject: cvs commit: src/sys/geom/virstor g_virstor.c Message-ID: <200806301521.m5UFLwQ5079409@repoman.freebsd.org> delphij 2008-06-30 15:21:42 UTC FreeBSD src repository Modified files: sys/geom/virstor g_virstor.c Log: SVN rev 180120 on 2008-06-30 15:21:42Z by delphij Avoid NULL deference. Reviewed by: ivoras Revision Changes Path 1.4 +5 -0 src/sys/geom/virstor/g_virstor.c From lulf at FreeBSD.org Mon Jun 30 15:42:00 2008 From: lulf at FreeBSD.org (Ulf Lilleengen) Date: Mon Jun 30 15:42:07 2008 Subject: cvs commit: src/sys/fs/fdescfs fdesc.h fdesc_vfsops.c fdesc_vnops.c Message-ID: <200806301542.m5UFg0MT081778@repoman.freebsd.org> lulf 2008-06-30 15:41:57 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/fs/fdescfs fdesc.h fdesc_vfsops.c fdesc_vnops.c Log: SVN rev 180121 on 2008-06-30 15:41:57Z by lulf MFC rev r179288: - Add locking to all filesystem operations in fdescfs and flag it as MPSAFE. - Use proper synhronization primitives to protect the internal fdesc node cache used in fdescfs. - Properly initialize and uninitalize hash. - Remove unused functions. Since fdescfs might recurse on itself, adding proper locking to it needed some tricky workarounds in some parts to make it work. For instance, a descriptor in fdescfs could refer to an open descriptor to itself, thus forcing the thread to recurse on vnode locks. Because of this, other race conditions also had to be fixed. Approved by: kib (mentor) Reviewed by: kib (mentor) Tested by: pho Revision Changes Path 1.20.10.1 +7 -2 src/sys/fs/fdescfs/fdesc.h 1.56.2.1 +34 -10 src/sys/fs/fdescfs/fdesc_vfsops.c 1.104.2.1 +158 -70 src/sys/fs/fdescfs/fdesc_vnops.c From des at FreeBSD.org Mon Jun 30 15:49:31 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Mon Jun 30 15:49:33 2008 Subject: cvs commit: src/lib/libarchive/test test_compat_zip_1.zip.uu Message-ID: <200806301549.m5UFnUFd082513@repoman.freebsd.org> des 2008-06-30 15:49:12 UTC FreeBSD src repository Modified files: lib/libarchive/test test_compat_zip_1.zip.uu Log: SVN rev 180122 on 2008-06-30 15:49:12Z by des Add missing $FreeBSD$ Revision Changes Path 1.2 +1 -0 src/lib/libarchive/test/test_compat_zip_1.zip.uu From des at FreeBSD.org Mon Jun 30 16:19:32 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Mon Jun 30 16:19:34 2008 Subject: cvs commit: src/lib/libarchive archive_read_support_format_zip.c Message-ID: <200806301619.m5UGJWi6087884@repoman.freebsd.org> des 2008-06-30 16:19:26 UTC FreeBSD src repository Modified files: lib/libarchive archive_read_support_format_zip.c Log: SVN rev 180123 on 2008-06-30 16:19:26Z by des More explicit error message. Revision Changes Path 1.26 +2 -1 src/lib/libarchive/archive_read_support_format_zip.c From alc at cs.rice.edu Mon Jun 30 16:56:44 2008 From: alc at cs.rice.edu (Alan Cox) Date: Mon Jun 30 16:56:54 2008 Subject: cvs commit: src/sys/amd64/include pmap.h In-Reply-To: <20080629215657.GI14567@hoeg.nl> References: <200806291835.m5TIZNLx089927@repoman.freebsd.org> <20080629211316.GA20319@edoofus.dev.vega.ru> <20080629215657.GI14567@hoeg.nl> Message-ID: <48691043.3080403@cs.rice.edu> Ed Schouten wrote: >Alan, > >* Ruslan Ermilov wrote: > > >>Can you please draw a new layout, similar to the one made by Peter? >> >>http://lists.freebsd.org/pipermail/freebsd-amd64/2005-July/005578.html >> >> > >I'm not familiar with the design of amd64 or its page tables, but is >there some kind of overhead when all 512 GB's would be allocated as kva? > > Yes, it would consume a rather large amount of (physical) memory. Roughly speaking, we preallocate memory for the page table that implements the kernel's address space. In other words, mapping a page into the kernel's address space never requires the allocation of (page table) memory, and unmapping a page never frees memory. This has some benefits, e.g., we will never block while mapping physical memory into the kernel's address space, and we can use a very simple, very fast method for mapping part of a file into the buffer cache. In general, that is, on most machine architectures, we can use a very simple technique to avoid preallocating page table memory that can't possibly be used. However, on amd64, the need to have kernel code and global/static variables in the highest 2GB of the kernel's address space undermines the effectiveness of that technique. So, it is still impractical to make the amd64 kernel address space arbitrarily large, as in 512GB. Alan From des at FreeBSD.org Mon Jun 30 17:11:35 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Mon Jun 30 17:11:37 2008 Subject: cvs commit: src/usr.bin/unzip unzip.c Message-ID: <200806301711.m5UHBXRP005159@repoman.freebsd.org> des 2008-06-30 17:11:27 UTC FreeBSD src repository Modified files: usr.bin/unzip unzip.c Log: SVN rev 180124 on 2008-06-30 17:11:27Z by des Quick shot at implementing -t (test). Requested by: ache MFC after: 2 weeks Revision Changes Path 1.2 +42 -4 src/usr.bin/unzip/unzip.c From des at FreeBSD.org Mon Jun 30 17:16:19 2008 From: des at FreeBSD.org (Dag-Erling Smorgrav) Date: Mon Jun 30 17:16:27 2008 Subject: cvs commit: src/usr.bin/unzip unzip.1 Message-ID: <200806301716.m5UHGJ4a005557@repoman.freebsd.org> des 2008-06-30 17:16:05 UTC FreeBSD src repository Modified files: usr.bin/unzip unzip.1 Log: SVN rev 180125 on 2008-06-30 17:16:05Z by des Update man page for -t. Revision Changes Path 1.2 +5 -2 src/usr.bin/unzip/unzip.1 From delphij at FreeBSD.org Mon Jun 30 17:45:51 2008 From: delphij at FreeBSD.org (Xin LI) Date: Mon Jun 30 17:46:02 2008 Subject: cvs commit: src/usr.bin/gzip gzip.1 Message-ID: <200806301745.m5UHjoAA008332@repoman.freebsd.org> delphij 2008-06-30 17:45:47 UTC FreeBSD src repository Modified files: usr.bin/gzip gzip.1 Log: SVN rev 180126 on 2008-06-30 17:45:47Z by delphij Make it clear that gzcat expects each argument to be separate files. PR: docs/123010 Submitted by: Andrew Wright MFC after: 2 weeks Revision Changes Path 1.5 +13 -1 src/usr.bin/gzip/gzip.1 From rwatson at FreeBSD.org Mon Jun 30 18:26:52 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Jun 30 18:26:59 2008 Subject: cvs commit: src/sys/netinet udp_usrreq.c Message-ID: <200806301826.m5UIQpsl014625@repoman.freebsd.org> rwatson 2008-06-30 18:26:43 UTC FreeBSD src repository Modified files: sys/netinet udp_usrreq.c Log: SVN rev 180127 on 2008-06-30 18:26:43Z by rwatson In udp_append() and udp_input(), make use of read locking on incpbs rather than write locking: while we need to maintain a valid reference to the inpcb and fix its state, no protocol layer state is modified during an IPv4 UDP receive -- there are only changes at the socket layer, which is separately protected by socket locking. While parallel concurrent receive on a single UDP socket is currently relatively unusual, introducing read locking in the transmit path, allowing concurrent receive and transmit, will significantly improve performance for loads such as BIND, memcached, etc. MFC after: 2 months Tested by: gnn, kris, ps Revision Changes Path 1.223 +8 -8 src/sys/netinet/udp_usrreq.c From rwatson at FreeBSD.org Mon Jun 30 18:35:35 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Jun 30 18:35:43 2008 Subject: cvs commit: src/share/man/man4 auditpipe.4 Message-ID: <200806301835.m5UIZZ0S015463@repoman.freebsd.org> rwatson 2008-06-30 18:35:11 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) share/man/man4 auditpipe.4 Log: SVN rev 180128 on 2008-06-30 18:35:11Z by rwatson Merge auditpipe.4:1.6 (r178726) from head to stable/7: Fix type name: struct auditpipe_preselect is actually struct auditpipe_ioctl_preselect. Submitted by: Stacey D. Son Revision Changes Path 1.5.2.1 +1 -1 src/share/man/man4/auditpipe.4 From ed at FreeBSD.org Mon Jun 30 20:24:02 2008 From: ed at FreeBSD.org (Ed Schouten) Date: Mon Jun 30 20:24:04 2008 Subject: cvs commit: src/sbin/dhclient dhclient.c Message-ID: <200806302024.m5UKO1cs033380@repoman.freebsd.org> ed 2008-06-30 20:23:49 UTC FreeBSD src repository Modified files: sbin/dhclient dhclient.c Log: SVN rev 180130 on 2008-06-30 20:23:49Z by ed Run the privileged dhclient process in its own session. In the MPSAFE TTY branch, I noticed PTY's to be leaked, because dhclient's privileged process was run inside the session of, say, the login shell. Make sure we call setsid() here. Approved by: philip (mentor), brooks Revision Changes Path 1.24 +1 -0 src/sbin/dhclient/dhclient.c From rwatson at FreeBSD.org Mon Jun 30 20:43:19 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Jun 30 20:43:21 2008 Subject: cvs commit: src/sys/nfsserver nfs_syscalls.c Message-ID: <200806302043.m5UKhI13037012@repoman.freebsd.org> rwatson 2008-06-30 20:43:06 UTC FreeBSD src repository Modified files: sys/nfsserver nfs_syscalls.c Log: SVN rev 180131 on 2008-06-30 20:43:06Z by rwatson Remove spls from NFS server setup call; expand receive socket buffer locking to cover full setup of socket upcalls; remove XXX about locking. MFC after: 3 weeks Revision Changes Path 1.119 +3 -8 src/sys/nfsserver/nfs_syscalls.c From peter at FreeBSD.org Mon Jun 30 22:01:44 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Mon Jun 30 22:01:45 2008 Subject: cvs commit: CVSROOT cfg_local.pm Message-ID: <200806302201.m5UM1hbO054754@repoman.freebsd.org> peter 2008-06-30 22:01:43 UTC FreeBSD src repository Modified files: . cfg_local.pm Log: Turn off MFC after syntax checking in cvs. Revision Changes Path 1.33 +2 -1 CVSROOT/cfg_local.pm From rik at FreeBSD.org Mon Jun 30 22:08:23 2008 From: rik at FreeBSD.org (Roman Kurakin) Date: Mon Jun 30 22:08:31 2008 Subject: cvs commit: src/sbin/sconfig sconfig.c src/sys/dev/ce ceddk.h if_ce.c src/sys/dev/cp cpddk.h if_cp.c src/sys/dev/ctau ctddk.h if_ct.c src/sys/dev/cx cxddk.h if_cx.c Message-ID: <200806302208.m5UM8MWI056518@repoman.freebsd.org> rik 2008-06-30 21:18:27 UTC FreeBSD src repository Modified files: sbin/sconfig sconfig.c sys/dev/ce ceddk.h if_ce.c sys/dev/cp cpddk.h if_cp.c sys/dev/ctau ctddk.h if_ct.c sys/dev/cx cxddk.h if_cx.c Log: SVN rev 180132 on 2008-06-30 21:18:27Z by rik Do not set IFF_DEBUG directly from the driver. MFC after: 1 month. Revision Changes Path 1.6 +21 -1 src/sbin/sconfig/sconfig.c 1.2 +1 -0 src/sys/dev/ce/ceddk.h 1.10 +12 -7 src/sys/dev/ce/if_ce.c 1.5 +1 -0 src/sys/dev/cp/cpddk.h 1.35 +12 -7 src/sys/dev/cp/if_cp.c 1.3 +1 -0 src/sys/dev/ctau/ctddk.h 1.36 +12 -7 src/sys/dev/ctau/if_ct.c 1.3 +1 -0 src/sys/dev/cx/cxddk.h 1.59 +23 -8 src/sys/dev/cx/if_cx.c From emaste at FreeBSD.org Mon Jun 30 22:13:53 2008 From: emaste at FreeBSD.org (Ed Maste) Date: Mon Jun 30 22:13:55 2008 Subject: cvs commit: src/sys/dev/aac aac.c aac_cam.c aac_debug.c aac_disk.c aac_pci.c aacreg.h aacvar.h src/sys/sys aac_ioctl.h Message-ID: <200806302213.m5UMDrmo057097@repoman.freebsd.org> emaste 2008-06-30 21:22:59 UTC FreeBSD src repository Modified files: (Branch: RELENG_6) sys/dev/aac aac.c aac_cam.c aac_debug.c aac_disk.c aac_pci.c aacreg.h aacvar.h sys/sys aac_ioctl.h Log: SVN rev 180133 on 2008-06-30 21:22:59Z by emaste MFC additional driver enhancements. r174385 Allow simultaneous opens of the device for issuing commands to the controller. This is merged from Adaptec driver build 11669. r177462 Add ioctls FSACTL_SEND_LARGE_FIB, FSACTL_SEND_RAW_SRB, FSACTL_LNX_SEND_LARGE_FIB, and FSACTL_LNX_SEND_RAW_SRB, and correct size checks on FIBs passed in from userspace. Both changes were obtained from Adaptec's driver build 15317. Adaptec's commandline RAID tool arcconf uses these ioctls when creating a RAID-10 array (and probably other operations too). r177567 Diff reduction to Adaptec's driver (around build 15317): catch up with a change in debugging routines. The fwprintf macro in the AAC_DEBUG case (mapping to printf) isn't from the Adaptec driver. r177619 Add 64-bit array support for RAIDs > 2TB. This corresponds to ~ Adaptec driver build 15317. r177635 Add \n to the end of a printf string and remove it from panic strings. r177695 Implement FSACTL_LNX_GET_FEATURES and FSACTL_GET_FEATURES ioctls. RAID tools (e.g. arcconf) need this to be able to create arrays larger than 2TB. r179969 Fix test for waiting AIFs in aac_poll(). This seems to solve the problem where Adaptec's arcconf monitoring tool hangs after producing its expected output. Revision Changes Path 1.109.2.18 +325 -192 src/sys/dev/aac/aac.c 1.20.2.7 +5 -7 src/sys/dev/aac/aac_cam.c 1.20.2.4 +1 -0 src/sys/dev/aac/aac_debug.c 1.43.2.1 +13 -14 src/sys/dev/aac/aac_disk.c 1.54.2.12 +6 -6 src/sys/dev/aac/aac_pci.c 1.21.2.4 +1 -0 src/sys/dev/aac/aacreg.h 1.46.8.2 +27 -24 src/sys/dev/aac/aacvar.h 1.11.4.1 +26 -0 src/sys/sys/aac_ioctl.h From rwatson at FreeBSD.org Mon Jun 30 22:19:30 2008 From: rwatson at FreeBSD.org (Robert Watson) Date: Mon Jun 30 22:19:38 2008 Subject: cvs commit: src/sys/ufs/ufs extattr.h ufs_extattr.c Message-ID: <200806302219.m5UMJUnN058292@repoman.freebsd.org> rwatson 2008-06-30 22:19:18 UTC FreeBSD src repository Modified files: (Branch: RELENG_7) sys/ufs/ufs extattr.h ufs_extattr.c Log: SVN rev 180134 on 2008-06-30 22:19:18Z by rwatson Merge extattr.h:1.22, ufs_extattr.c:1.91 (r176797) from head to stable/7: Continue on-going campaign to replace lockmgr locks with sx locks where the specific semantics of lockmgr aren't required: update UFS1 extended attributes to protect its data structures using an sx lock. While here, update comments on lock granularity. Revision Changes Path 1.21.2.1 +3 -2 src/sys/ufs/ufs/extattr.h 1.86.2.2 +11 -10 src/sys/ufs/ufs/ufs_extattr.c From peter at FreeBSD.org Mon Jun 30 22:22:21 2008 From: peter at FreeBSD.org (Peter Wemm) Date: Mon Jun 30 22:22:28 2008 Subject: cvs commit: src/sys/compat/pecoff imgact_pecoff.c imgact_pecoff.h src/sys/modules/pecoff Makefile Message-ID: <200806302222.m5UMMKGN058637@repoman.freebsd.org> peter 2008-06-30 22:22:20 UTC FreeBSD src repository Removed files: sys/compat/pecoff imgact_pecoff.c imgact_pecoff.h sys/modules/pecoff Makefile Log: Manually commit removed directory part of svn change 179785. r179785 | wkoszek | 2008-06-14 05:51:44 -0700 (Sat, 14 Jun 2008) ------------------------------------------------------------------------ Remove obselete PECOFF image activator support. PRs assigned at the time of removal: kern/80742 Discussed on: freebsd-current (silence), IRC Tested by: make universe Approved by: cognet (mentor) Revision Changes Path 1.45 +0 -606 src/sys/compat/pecoff/imgact_pecoff.c (dead) 1.3 +0 -142 src/sys/compat/pecoff/imgact_pecoff.h (dead) 1.8 +0 -10 src/sys/modules/pecoff/Makefile (dead) From delphij at FreeBSD.org Mon Jun 30 23:23:49 2008 From: delphij at FreeBSD.org (Xin LI) Date: Mon Jun 30 23:23:52 2008 Subject: cvs commit: src/usr.bin/gzip unbzip2.c Message-ID: <200806302323.m5UNNn5b065330@repoman.freebsd.org> delphij 2008-06-30 23:23:31 UTC FreeBSD src repository Modified files: usr.bin/gzip unbzip2.c Log: SVN rev 180135 on 2008-06-30 23:23:31Z by delphij Remove clause 3 and 4 of NetBSD license. Obtained from: NetBSD Revision Changes Path 1.2 +1 -8 src/usr.bin/gzip/unbzip2.c From delphij at FreeBSD.org Mon Jun 30 23:53:36 2008 From: delphij at FreeBSD.org (Xin LI) Date: Mon Jun 30 23:53:43 2008 Subject: cvs commit: src/usr.bin/gzip znew Message-ID: <200806302353.m5UNrZPd067788@repoman.freebsd.org> delphij 2008-06-30 23:53:15 UTC FreeBSD src repository Modified files: usr.bin/gzip znew Log: SVN rev 180136 on 2008-06-30 23:53:15Z by delphij Reflect the fact that we actually have NetBSD revision 1.3. Revision Changes Path 1.3 +1 -1 src/usr.bin/gzip/znew