From js at alien8.de Mon Feb 2 09:16:36 2009 From: js at alien8.de (Julian Stecklina) Date: Mon Feb 2 09:16:42 2009 Subject: Time goes backwards References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> Message-ID: <87ljsokbdn.fsf@tabernacle.localnet> Mister Olli writes: > Hi, > > I've had time problems too. The clock did not go backwards, but forwards > for some time sime, and then jumped back to a certain date. Like > Sun Jan 25 00:55:12 UTC 2009 > ... > Sun Jan 25 01:08:35 UTC 2009 > Sun Jan 25 00:55:12 UTC 2009 > Sun Jan 25 00:55:13 UTC 2009 > ... > > It took me quite some time to find a solution that helped for me: > - activate the independent_wallclock in domO by 'sysctl xen.independent_wallclock=1' I booted the DomU with machdep.independent_wallclock=1 as kernel parameter instead. But that didn't work. sysctl still reported it as 0. I've tried setting machdep.xen_disable_rtc_set and machdep.independent_wallclock after booting, but it does not help either. > - edit and configure '/etc/ntpd.conf' in domU > - edit '/etc/rc.conf' and add the following lines > # NTP stuff > ntpdate_enable="YES" > ntpdate_flags="-b " > ntpd_enable="YES" > - reboot domU > > If you try it on your problem, could you please give me some feedback if > it worked? Doesn't work. :-/ MfG, -- Julian Stecklina Well, take it from an old hand: the only reason it would be easier to program in C is that you can't easily express complex problems in C, so you don't. - Erik Naggum (in comp.lang.lisp) From kmacy at freebsd.org Mon Feb 2 21:37:04 2009 From: kmacy at freebsd.org (Kip Macy) Date: Mon Feb 2 21:37:12 2009 Subject: Instable 32bit domU on 64bit dom0 (w/Xen 3.3.1) In-Reply-To: <11276352.23661233243265678.JavaMail.root@zim.freshx.de> References: <11276352.23661233243265678.JavaMail.root@zim.freshx.de> Message-ID: <3c1674c90902022136t48eef4faic3cc091082b68432@mail.gmail.com> Hi Kai, This should be fixed by r188042. Interrupts need to be disabled on entry to the page fault handler or we risk being preempted prior to reading cr2 (the faulting address). There is still another fundamental stability bug whereby timer interrupts appear to sometimes be left disabled until another interrupt (console, network, disk) comes in at which point they're re-enabled. You may or may not hit that, I only have one configuration where I'm able to reproduce that issue. Cheers, Kip On Thu, Jan 29, 2009 at 7:34 AM, Kai Mosebach wrote: > Hi, > > i recently got my first FreeBSD 8 domU working and i am quite happy to see that there is eventually progress on the important Xen topic for FreeBSD. > Since this domU as a svn -head checkout from 23.01.09 still segfaults here and there i startet to rebuild the whole world with debugging symbols in the hope that the backtraces give more info's on where to find bugs. > > Here are the first results from several core files which hang around in the system after a world build and i hope the developers here can make more sense out of it than me. > > If you need further dumps/traces or have some advises how to go on from here, please let me know! > > Best Kai > > /usr/obj/usr/src/make.core > > Program terminated with signal 11, Segmentation fault. > #0 0x0804da01 in Cmd_Exec ( > cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", > error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 > 2713 if ((ps.child_pid = vfork()) == -1) { > (gdb) bt > #0 0x0804da01 in Cmd_Exec ( > cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", > error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 > #1 0x080544ed in Parse_DoVar (line=0x28128b00 "_MAKEOBJDIRPREFIX", > ctxt=0x2810e200) at /usr/src/usr.bin/make/parse.c:1520 > #2 0x08055a1a in Parse_File (name=0xbf7fdbf8 "/usr/src/Makefile", > stream=0x80a4960) at /usr/src/usr.bin/make/parse.c:2457 > #3 0x08051e68 in ReadMakefile (p=Variable "p" is not available. > ) at /usr/src/usr.bin/make/main.c:323 > #4 0x08051f24 in TryReadMakefile (p=0x8094744 "Makefile") > at /usr/src/usr.bin/make/main.c:342 > #5 0x080527c3 in main (argc=1, argv=0xbf7feddc) > at /usr/src/usr.bin/make/main.c:1209 > (gdb) > > > > /usr/src/sbin/atacontrol/sh.core > > Core was generated by `sh'. > Program terminated with signal 11, Segmentation fault. > Reading symbols from /lib/libedit.so.6...done. > Loaded symbols for /lib/libedit.so.6 > Reading symbols from /lib/libncurses.so.7...done. > Loaded symbols for /lib/libncurses.so.7 > Reading symbols from /lib/libc.so.7...done. > Loaded symbols for /lib/libc.so.7 > Reading symbols from /libexec/ld-elf.so.1...done. > Loaded symbols for /libexec/ld-elf.so.1 > #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at /usr/src/bin/sh/jobs.c:746 > 746 pid = fork(); > (gdb) bt > #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at /usr/src/bin/sh/jobs.c:746 > #1 0x0804d759 in evalcommand (cmd=0x28304d94, flags=0, backcmd=0x0) at /usr/src/bin/sh/eval.c:747 > #2 0x0804e725 in evaltree (n=0x28304d94, flags=0) at /usr/src/bin/sh/eval.c:267 > #3 0x0804e4ff in evaltree (n=0x28304da4, flags=0) at /usr/src/bin/sh/eval.c:238 > #4 0x0804e86c in evaltree (n=0x28304210, flags=0) at /usr/src/bin/sh/eval.c:341 > #5 0x0804eba7 in evalstring ( > s=0xbf7fea84 "for entry in adjkerntz atacontrol atm badsect bsdlabel camcontrol ccdconfig clri comcontrol conscontrol ddb devd devfs dhclient dmesg dump dumpfs dumpon fdisk ffsinfo fsck fsck_"...) at /usr/src/bin/sh/eval.c:169 > #6 0x08056d98 in main (argc=3, argv=0xbf7fe9a8) at /usr/src/bin/sh/main.c:181 > > > > > /usr/obj/usr/src/libexec/ftpd/make.core > > Program terminated with signal 11, Segmentation fault. > #0 0x080898bf in memset () > (gdb) bt > #0 0x080898bf in memset () > #1 0x0000000c in ?? () > #2 0x0806f8b2 in arena_malloc () > #3 0x0807106b in malloc () > #4 0x0805a2d1 in emalloc (len=256) at /usr/src/usr.bin/make/util.c:234 > #5 0x08049d3f in Buf_Init (size=0) at /usr/src/usr.bin/make/buf.c:184 > #6 0x0805a63a in VarCreate (name=0x2810d118 ".CURDIR", value=0xbf7fe0b6 "/usr/src/libexec/ftpd", flags=0) at /usr/src/usr.bin/make/var.c:214 > #7 0x0805a68d in VarAdd (name=0x2810d118 ".CURDIR", val=0xbf7fe0b6 "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:954 > #8 0x0805d2fc in Var_Set (name=0x809409d ".CURDIR", val=0xbf7fe0b6 "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:1009 > #9 0x0805d321 in Var_SetGlobal (name=0x809409d ".CURDIR", value=0xbf7fe0b6 "/usr/src/libexec/ftpd") at /usr/src/usr.bin/make/var.c:1035 > #10 0x0805243f in main (argc=3, argv=0xbf7fed74) at /usr/src/usr.bin/make/main.c:1115 > > > > /usr/obj/usr/src/usr.bin/telnet/as.core > > Program terminated with signal 11, Segmentation fault. > #0 0x080480d6 in _start () > (gdb) bt > #0 0x080480d6 in _start () > #1 0x00000003 in ?? () > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From mister.olli at googlemail.com Tue Feb 3 05:58:31 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Tue Feb 3 05:58:37 2009 Subject: Instable 32bit domU on 64bit dom0 (w/Xen 3.3.1) In-Reply-To: <3c1674c90902022136t48eef4faic3cc091082b68432@mail.gmail.com> References: <11276352.23661233243265678.JavaMail.root@zim.freshx.de> <3c1674c90902022136t48eef4faic3cc091082b68432@mail.gmail.com> Message-ID: <1233669467.4832.6.camel@phoenix.blechhirn.net> Hi, > This should be fixed by r188042. Interrupts need to be disabled on > entry to the page fault handler or we risk being preempted prior to > reading cr2 (the faulting address). Sounds great. I will try to test it as soon as possible, since I ran in a lot of seg-faults during compiling world under Xen, too. > There is still another fundamental stability bug whereby timer > interrupts appear to sometimes be left disabled until another > interrupt (console, network, disk) comes in at which point they're > re-enabled. You may or may not hit that, I only have one configuration > where I'm able to reproduce that issue. Sounds like the problem I gave up using FreeBSD7 in Xen. My domU halted (does nothing but was running, even SSH sessions timed out) sometimes and the only way to get it back to work, was to hit a key on the console. I did a temporary fix by 'while (true); do echo "e"; sleep 10s; done | xm console ". -- Mr. Olli > > Cheers, > Kip > > > On Thu, Jan 29, 2009 at 7:34 AM, Kai Mosebach wrote: > > Hi, > > > > i recently got my first FreeBSD 8 domU working and i am quite happy to see that there is eventually progress on the important Xen topic for FreeBSD. > > Since this domU as a svn -head checkout from 23.01.09 still segfaults here and there i startet to rebuild the whole world with debugging symbols in the hope that the backtraces give more info's on where to find bugs. > > > > Here are the first results from several core files which hang around in the system after a world build and i hope the developers here can make more sense out of it than me. > > > > If you need further dumps/traces or have some advises how to go on from here, please let me know! > > > > Best Kai > > > > /usr/obj/usr/src/make.core > > > > Program terminated with signal 11, Segmentation fault. > > #0 0x0804da01 in Cmd_Exec ( > > cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", > > error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 > > 2713 if ((ps.child_pid = vfork()) == -1) { > > (gdb) bt > > #0 0x0804da01 in Cmd_Exec ( > > cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", > > error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 > > #1 0x080544ed in Parse_DoVar (line=0x28128b00 "_MAKEOBJDIRPREFIX", > > ctxt=0x2810e200) at /usr/src/usr.bin/make/parse.c:1520 > > #2 0x08055a1a in Parse_File (name=0xbf7fdbf8 "/usr/src/Makefile", > > stream=0x80a4960) at /usr/src/usr.bin/make/parse.c:2457 > > #3 0x08051e68 in ReadMakefile (p=Variable "p" is not available. > > ) at /usr/src/usr.bin/make/main.c:323 > > #4 0x08051f24 in TryReadMakefile (p=0x8094744 "Makefile") > > at /usr/src/usr.bin/make/main.c:342 > > #5 0x080527c3 in main (argc=1, argv=0xbf7feddc) > > at /usr/src/usr.bin/make/main.c:1209 > > (gdb) > > > > > > > > /usr/src/sbin/atacontrol/sh.core > > > > Core was generated by `sh'. > > Program terminated with signal 11, Segmentation fault. > > Reading symbols from /lib/libedit.so.6...done. > > Loaded symbols for /lib/libedit.so.6 > > Reading symbols from /lib/libncurses.so.7...done. > > Loaded symbols for /lib/libncurses.so.7 > > Reading symbols from /lib/libc.so.7...done. > > Loaded symbols for /lib/libc.so.7 > > Reading symbols from /libexec/ld-elf.so.1...done. > > Loaded symbols for /libexec/ld-elf.so.1 > > #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at /usr/src/bin/sh/jobs.c:746 > > 746 pid = fork(); > > (gdb) bt > > #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at /usr/src/bin/sh/jobs.c:746 > > #1 0x0804d759 in evalcommand (cmd=0x28304d94, flags=0, backcmd=0x0) at /usr/src/bin/sh/eval.c:747 > > #2 0x0804e725 in evaltree (n=0x28304d94, flags=0) at /usr/src/bin/sh/eval.c:267 > > #3 0x0804e4ff in evaltree (n=0x28304da4, flags=0) at /usr/src/bin/sh/eval.c:238 > > #4 0x0804e86c in evaltree (n=0x28304210, flags=0) at /usr/src/bin/sh/eval.c:341 > > #5 0x0804eba7 in evalstring ( > > s=0xbf7fea84 "for entry in adjkerntz atacontrol atm badsect bsdlabel camcontrol ccdconfig clri comcontrol conscontrol ddb devd devfs dhclient dmesg dump dumpfs dumpon fdisk ffsinfo fsck fsck_"...) at /usr/src/bin/sh/eval.c:169 > > #6 0x08056d98 in main (argc=3, argv=0xbf7fe9a8) at /usr/src/bin/sh/main.c:181 > > > > > > > > > > /usr/obj/usr/src/libexec/ftpd/make.core > > > > Program terminated with signal 11, Segmentation fault. > > #0 0x080898bf in memset () > > (gdb) bt > > #0 0x080898bf in memset () > > #1 0x0000000c in ?? () > > #2 0x0806f8b2 in arena_malloc () > > #3 0x0807106b in malloc () > > #4 0x0805a2d1 in emalloc (len=256) at /usr/src/usr.bin/make/util.c:234 > > #5 0x08049d3f in Buf_Init (size=0) at /usr/src/usr.bin/make/buf.c:184 > > #6 0x0805a63a in VarCreate (name=0x2810d118 ".CURDIR", value=0xbf7fe0b6 "/usr/src/libexec/ftpd", flags=0) at /usr/src/usr.bin/make/var.c:214 > > #7 0x0805a68d in VarAdd (name=0x2810d118 ".CURDIR", val=0xbf7fe0b6 "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:954 > > #8 0x0805d2fc in Var_Set (name=0x809409d ".CURDIR", val=0xbf7fe0b6 "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:1009 > > #9 0x0805d321 in Var_SetGlobal (name=0x809409d ".CURDIR", value=0xbf7fe0b6 "/usr/src/libexec/ftpd") at /usr/src/usr.bin/make/var.c:1035 > > #10 0x0805243f in main (argc=3, argv=0xbf7fed74) at /usr/src/usr.bin/make/main.c:1115 > > > > > > > > /usr/obj/usr/src/usr.bin/telnet/as.core > > > > Program terminated with signal 11, Segmentation fault. > > #0 0x080480d6 in _start () > > (gdb) bt > > #0 0x080480d6 in _start () > > #1 0x00000003 in ?? () > > _______________________________________________ > > freebsd-xen@freebsd.org mailing list > > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > > > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org From mister.olli at googlemail.com Tue Feb 3 11:00:23 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Tue Feb 3 11:00:30 2009 Subject: 8-CURRENT does not build without WITNESS activated Message-ID: <1233687579.4832.11.camel@phoenix.blechhirn.net> Hi. I'm currently unable to build the 8-CURRENT kernel in a PV domU, without "options WITNESS" in my kernel config. The output is the following: ====================================================== make -V CFILES -V SYSTEM_CFILES -V GEN_CFILES | MKDEP_CPP="cc -E" CC="cc" xargs mkdep -a -f .newdep -O -pipe -std=c99 -g -Wall -Wredundant-decls -Wnested-externs -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual -Wundef -Wno-pointer-sign -fformat-extensions -nostdinc -I. -I/usr/src/sys -I/usr/src/sys/contrib/altq -I/usr/src/sys/contrib/ipfilter -I/usr/src/sys/contrib/pf -I/usr/src/sys/dev/ath -I/usr/src/sys/dev/ath/ath_hal -I/usr/src/sys/contrib/ngatm -I/usr/src/sys/dev/twa -I/usr/src/sys/gnu/fs/xfs/FreeBSD -I/usr/src/sys/gnu/fs/xfs/FreeBSD/support -I/usr/src/sys/gnu/fs/xfs -I/usr/src/sys/contrib/opensolaris/compat -I/usr/src/sys/dev/cxgb -D_KERNEL -DHAVE_KERNEL_OPTION_HEADERS -include opt_global.h -fno-common -finline-limit=8000 --param inline-unit-growth=100 --param large-function-growth=1000 -mno-align-long-strings -mpreferred-stack-boundary=2 -mno-mmx -mno-3dnow -mno-sse -mno-sse2 -mno-sse3 -ffreestanding -fstack-protector In file included from /usr/src/sys/dev/xen/netfront/netfront.c:32: /usr/src/sys/sys/sx.h:210:2: error: #error "LOCK_DEBUG not defined, include before " mkdep: compile failed *** Error code 1 Stop in /usr/obj/usr/src/sys/XEN. *** Error code 1 Stop in /usr/src. *** Error code 1 Stop in /usr/src. template_8-CURRENT# ====================================================== -- Mr. Olli From kai at freshx.de Wed Feb 4 02:20:40 2009 From: kai at freshx.de (Kai Mosebach) Date: Wed Feb 4 02:20:51 2009 Subject: Instable 32bit domU on 64bit dom0 (w/Xen 3.3.1) In-Reply-To: <3c1674c90902022136t48eef4faic3cc091082b68432@mail.gmail.com> Message-ID: Hi Kip, After rebuilding the kernel (which improved stability by roughly 80%) and rebuilding world (which improved stability by another 19% ;) the userland seems to be much more stable now! Kernel wise I hit the bug you described below at every bootup/shutdown and the first hit of a key on the console, but the system keeps running afterwards... Booting with 2 CPUs is still ending up in Spanic: blockable sleep lock (sleep mutex) XCONS LOCK @ dev/xen/console/console.c:290 cpuid = 1 KDB: enter: panic [thread pid 11 tid 100003 ] Stopped at kdb_enter+0x3a: movl $0,kdb_why Anyway, thanks for your effort! Kai On 2/3/09 6:36 AM, "Kip Macy" wrote: > Hi Kai, > > This should be fixed by r188042. Interrupts need to be disabled on > entry to the page fault handler or we risk being preempted prior to > reading cr2 (the faulting address). > > There is still another fundamental stability bug whereby timer > interrupts appear to sometimes be left disabled until another > interrupt (console, network, disk) comes in at which point they're > re-enabled. You may or may not hit that, I only have one configuration > where I'm able to reproduce that issue. > > Cheers, > Kip > > > On Thu, Jan 29, 2009 at 7:34 AM, Kai Mosebach wrote: >> Hi, >> >> i recently got my first FreeBSD 8 domU working and i am quite happy to see >> that there is eventually progress on the important Xen topic for FreeBSD. >> Since this domU as a svn -head checkout from 23.01.09 still segfaults here >> and there i startet to rebuild the whole world with debugging symbols in the >> hope that the backtraces give more info's on where to find bugs. >> >> Here are the first results from several core files which hang around in the >> system after a world build and i hope the developers here can make more sense >> out of it than me. >> >> If you need further dumps/traces or have some advises how to go on from here, >> please let me know! >> >> Best Kai >> >> /usr/obj/usr/src/make.core >> >> Program terminated with signal 11, Segmentation fault. >> #0 0x0804da01 in Cmd_Exec ( >> cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make >> __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", >> error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 >> 2713 if ((ps.child_pid = vfork()) == -1) { >> (gdb) bt >> #0 0x0804da01 in Cmd_Exec ( >> cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make >> __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", >> error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 >> #1 0x080544ed in Parse_DoVar (line=0x28128b00 "_MAKEOBJDIRPREFIX", >> ctxt=0x2810e200) at /usr/src/usr.bin/make/parse.c:1520 >> #2 0x08055a1a in Parse_File (name=0xbf7fdbf8 "/usr/src/Makefile", >> stream=0x80a4960) at /usr/src/usr.bin/make/parse.c:2457 >> #3 0x08051e68 in ReadMakefile (p=Variable "p" is not available. >> ) at /usr/src/usr.bin/make/main.c:323 >> #4 0x08051f24 in TryReadMakefile (p=0x8094744 "Makefile") >> at /usr/src/usr.bin/make/main.c:342 >> #5 0x080527c3 in main (argc=1, argv=0xbf7feddc) >> at /usr/src/usr.bin/make/main.c:1209 >> (gdb) >> >> >> >> /usr/src/sbin/atacontrol/sh.core >> >> Core was generated by `sh'. >> Program terminated with signal 11, Segmentation fault. >> Reading symbols from /lib/libedit.so.6...done. >> Loaded symbols for /lib/libedit.so.6 >> Reading symbols from /lib/libncurses.so.7...done. >> Loaded symbols for /lib/libncurses.so.7 >> Reading symbols from /lib/libc.so.7...done. >> Loaded symbols for /lib/libc.so.7 >> Reading symbols from /libexec/ld-elf.so.1...done. >> Loaded symbols for /libexec/ld-elf.so.1 >> #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at >> /usr/src/bin/sh/jobs.c:746 >> 746 pid = fork(); >> (gdb) bt >> #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at >> /usr/src/bin/sh/jobs.c:746 >> #1 0x0804d759 in evalcommand (cmd=0x28304d94, flags=0, backcmd=0x0) at >> /usr/src/bin/sh/eval.c:747 >> #2 0x0804e725 in evaltree (n=0x28304d94, flags=0) at >> /usr/src/bin/sh/eval.c:267 >> #3 0x0804e4ff in evaltree (n=0x28304da4, flags=0) at >> /usr/src/bin/sh/eval.c:238 >> #4 0x0804e86c in evaltree (n=0x28304210, flags=0) at >> /usr/src/bin/sh/eval.c:341 >> #5 0x0804eba7 in evalstring ( >> s=0xbf7fea84 "for entry in adjkerntz atacontrol atm badsect bsdlabel >> camcontrol ccdconfig clri comcontrol conscontrol ddb devd devfs >> dhclient dmesg dump dumpfs dumpon fdisk ffsinfo fsck fsck_"...) at >> /usr/src/bin/sh/eval.c:169 >> #6 0x08056d98 in main (argc=3, argv=0xbf7fe9a8) at >> /usr/src/bin/sh/main.c:181 >> >> >> >> >> /usr/obj/usr/src/libexec/ftpd/make.core >> >> Program terminated with signal 11, Segmentation fault. >> #0 0x080898bf in memset () >> (gdb) bt >> #0 0x080898bf in memset () >> #1 0x0000000c in ?? () >> #2 0x0806f8b2 in arena_malloc () >> #3 0x0807106b in malloc () >> #4 0x0805a2d1 in emalloc (len=256) at /usr/src/usr.bin/make/util.c:234 >> #5 0x08049d3f in Buf_Init (size=0) at /usr/src/usr.bin/make/buf.c:184 >> #6 0x0805a63a in VarCreate (name=0x2810d118 ".CURDIR", value=0xbf7fe0b6 >> "/usr/src/libexec/ftpd", flags=0) at /usr/src/usr.bin/make/var.c:214 >> #7 0x0805a68d in VarAdd (name=0x2810d118 ".CURDIR", val=0xbf7fe0b6 >> "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:954 >> #8 0x0805d2fc in Var_Set (name=0x809409d ".CURDIR", val=0xbf7fe0b6 >> "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:1009 >> #9 0x0805d321 in Var_SetGlobal (name=0x809409d ".CURDIR", value=0xbf7fe0b6 >> "/usr/src/libexec/ftpd") at /usr/src/usr.bin/make/var.c:1035 >> #10 0x0805243f in main (argc=3, argv=0xbf7fed74) at >> /usr/src/usr.bin/make/main.c:1115 >> >> >> >> /usr/obj/usr/src/usr.bin/telnet/as.core >> >> Program terminated with signal 11, Segmentation fault. >> #0 0x080480d6 in _start () >> (gdb) bt >> #0 0x080480d6 in _start () >> #1 0x00000003 in ?? () >> _______________________________________________ >> freebsd-xen@freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-xen >> To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" >> From kmacy at freebsd.org Wed Feb 4 20:03:33 2009 From: kmacy at freebsd.org (Kip Macy) Date: Wed Feb 4 20:03:40 2009 Subject: Instable 32bit domU on 64bit dom0 (w/Xen 3.3.1) In-Reply-To: References: <3c1674c90902022136t48eef4faic3cc091082b68432@mail.gmail.com> Message-ID: <3c1674c90902042003jb24eef0hec2960c270b54e31@mail.gmail.com> Please run with only 1 vcpu for the next couple of weeks. Thanks, Kip On Wed, Feb 4, 2009 at 2:20 AM, Kai Mosebach wrote: > Hi Kip, > > After rebuilding the kernel (which improved stability by roughly 80%) and > rebuilding world (which improved stability by another 19% ;) the userland > seems to be much more stable now! Kernel wise I hit the bug you described > below at every bootup/shutdown and the first hit of a key on the console, > but the system keeps running afterwards... > > Booting with 2 CPUs is still ending up in > > Spanic: blockable sleep lock (sleep mutex) XCONS LOCK @ > dev/xen/console/console.c:290 > cpuid = 1 > KDB: enter: panic > [thread pid 11 tid 100003 ] > Stopped at kdb_enter+0x3a: movl $0,kdb_why > > Anyway, thanks for your effort! > > Kai > > On 2/3/09 6:36 AM, "Kip Macy" wrote: > >> Hi Kai, >> >> This should be fixed by r188042. Interrupts need to be disabled on >> entry to the page fault handler or we risk being preempted prior to >> reading cr2 (the faulting address). >> >> There is still another fundamental stability bug whereby timer >> interrupts appear to sometimes be left disabled until another >> interrupt (console, network, disk) comes in at which point they're >> re-enabled. You may or may not hit that, I only have one configuration >> where I'm able to reproduce that issue. >> >> Cheers, >> Kip >> >> >> On Thu, Jan 29, 2009 at 7:34 AM, Kai Mosebach wrote: >>> Hi, >>> >>> i recently got my first FreeBSD 8 domU working and i am quite happy to see >>> that there is eventually progress on the important Xen topic for FreeBSD. >>> Since this domU as a svn -head checkout from 23.01.09 still segfaults here >>> and there i startet to rebuild the whole world with debugging symbols in the >>> hope that the backtraces give more info's on where to find bugs. >>> >>> Here are the first results from several core files which hang around in the >>> system after a world build and i hope the developers here can make more sense >>> out of it than me. >>> >>> If you need further dumps/traces or have some advises how to go on from here, >>> please let me know! >>> >>> Best Kai >>> >>> /usr/obj/usr/src/make.core >>> >>> Program terminated with signal 11, Segmentation fault. >>> #0 0x0804da01 in Cmd_Exec ( >>> cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make >>> __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", >>> error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 >>> 2713 if ((ps.child_pid = vfork()) == -1) { >>> (gdb) bt >>> #0 0x0804da01 in Cmd_Exec ( >>> cmd=0x2812df00 "/usr/bin/env -i PATH=/sbin:/bin:/usr/sbin:/usr/bin make >>> __MAKE_CONF=/etc/make.conf -f /dev/null -V MAKEOBJDIRPREFIX dummy", >>> error=0xbf7fdb18) at /usr/src/usr.bin/make/job.c:2713 >>> #1 0x080544ed in Parse_DoVar (line=0x28128b00 "_MAKEOBJDIRPREFIX", >>> ctxt=0x2810e200) at /usr/src/usr.bin/make/parse.c:1520 >>> #2 0x08055a1a in Parse_File (name=0xbf7fdbf8 "/usr/src/Makefile", >>> stream=0x80a4960) at /usr/src/usr.bin/make/parse.c:2457 >>> #3 0x08051e68 in ReadMakefile (p=Variable "p" is not available. >>> ) at /usr/src/usr.bin/make/main.c:323 >>> #4 0x08051f24 in TryReadMakefile (p=0x8094744 "Makefile") >>> at /usr/src/usr.bin/make/main.c:342 >>> #5 0x080527c3 in main (argc=1, argv=0xbf7feddc) >>> at /usr/src/usr.bin/make/main.c:1209 >>> (gdb) >>> >>> >>> >>> /usr/src/sbin/atacontrol/sh.core >>> >>> Core was generated by `sh'. >>> Program terminated with signal 11, Segmentation fault. >>> Reading symbols from /lib/libedit.so.6...done. >>> Loaded symbols for /lib/libedit.so.6 >>> Reading symbols from /lib/libncurses.so.7...done. >>> Loaded symbols for /lib/libncurses.so.7 >>> Reading symbols from /lib/libc.so.7...done. >>> Loaded symbols for /lib/libc.so.7 >>> Reading symbols from /libexec/ld-elf.so.1...done. >>> Loaded symbols for /libexec/ld-elf.so.1 >>> #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at >>> /usr/src/bin/sh/jobs.c:746 >>> 746 pid = fork(); >>> (gdb) bt >>> #0 0x080555e6 in forkshell (jp=0x2830c0c0, n=0x28304d94, mode=0) at >>> /usr/src/bin/sh/jobs.c:746 >>> #1 0x0804d759 in evalcommand (cmd=0x28304d94, flags=0, backcmd=0x0) at >>> /usr/src/bin/sh/eval.c:747 >>> #2 0x0804e725 in evaltree (n=0x28304d94, flags=0) at >>> /usr/src/bin/sh/eval.c:267 >>> #3 0x0804e4ff in evaltree (n=0x28304da4, flags=0) at >>> /usr/src/bin/sh/eval.c:238 >>> #4 0x0804e86c in evaltree (n=0x28304210, flags=0) at >>> /usr/src/bin/sh/eval.c:341 >>> #5 0x0804eba7 in evalstring ( >>> s=0xbf7fea84 "for entry in adjkerntz atacontrol atm badsect bsdlabel >>> camcontrol ccdconfig clri comcontrol conscontrol ddb devd devfs >>> dhclient dmesg dump dumpfs dumpon fdisk ffsinfo fsck fsck_"...) at >>> /usr/src/bin/sh/eval.c:169 >>> #6 0x08056d98 in main (argc=3, argv=0xbf7fe9a8) at >>> /usr/src/bin/sh/main.c:181 >>> >>> >>> >>> >>> /usr/obj/usr/src/libexec/ftpd/make.core >>> >>> Program terminated with signal 11, Segmentation fault. >>> #0 0x080898bf in memset () >>> (gdb) bt >>> #0 0x080898bf in memset () >>> #1 0x0000000c in ?? () >>> #2 0x0806f8b2 in arena_malloc () >>> #3 0x0807106b in malloc () >>> #4 0x0805a2d1 in emalloc (len=256) at /usr/src/usr.bin/make/util.c:234 >>> #5 0x08049d3f in Buf_Init (size=0) at /usr/src/usr.bin/make/buf.c:184 >>> #6 0x0805a63a in VarCreate (name=0x2810d118 ".CURDIR", value=0xbf7fe0b6 >>> "/usr/src/libexec/ftpd", flags=0) at /usr/src/usr.bin/make/var.c:214 >>> #7 0x0805a68d in VarAdd (name=0x2810d118 ".CURDIR", val=0xbf7fe0b6 >>> "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:954 >>> #8 0x0805d2fc in Var_Set (name=0x809409d ".CURDIR", val=0xbf7fe0b6 >>> "/usr/src/libexec/ftpd", ctxt=0x28112180) at /usr/src/usr.bin/make/var.c:1009 >>> #9 0x0805d321 in Var_SetGlobal (name=0x809409d ".CURDIR", value=0xbf7fe0b6 >>> "/usr/src/libexec/ftpd") at /usr/src/usr.bin/make/var.c:1035 >>> #10 0x0805243f in main (argc=3, argv=0xbf7fed74) at >>> /usr/src/usr.bin/make/main.c:1115 >>> >>> >>> >>> /usr/obj/usr/src/usr.bin/telnet/as.core >>> >>> Program terminated with signal 11, Segmentation fault. >>> #0 0x080480d6 in _start () >>> (gdb) bt >>> #0 0x080480d6 in _start () >>> #1 0x00000003 in ?? () >>> _______________________________________________ >>> freebsd-xen@freebsd.org mailing list >>> http://lists.freebsd.org/mailman/listinfo/freebsd-xen >>> To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" >>> > > > From kmacy at freebsd.org Wed Feb 4 20:04:05 2009 From: kmacy at freebsd.org (Kip Macy) Date: Wed Feb 4 20:04:11 2009 Subject: 8-CURRENT does not build without WITNESS activated In-Reply-To: <1233687579.4832.11.camel@phoenix.blechhirn.net> References: <1233687579.4832.11.camel@phoenix.blechhirn.net> Message-ID: <3c1674c90902042004m1b7261d7o60a57a80c447582b@mail.gmail.com> adding lock.h will fix that - sorry about it Thanks, Kip On Tue, Feb 3, 2009 at 10:59 AM, Mister Olli wrote: > Hi. > > I'm currently unable to build the 8-CURRENT kernel in a PV domU, without > "options WITNESS" in my kernel config. > > The output is the following: > ====================================================== > make -V CFILES -V SYSTEM_CFILES -V GEN_CFILES | MKDEP_CPP="cc -E" > CC="cc" xargs mkdep -a -f .newdep -O -pipe -std=c99 -g -Wall > -Wredundant-decls -Wnested-externs -Wstrict-prototypes > -Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual -Wundef > -Wno-pointer-sign -fformat-extensions -nostdinc -I. -I/usr/src/sys > -I/usr/src/sys/contrib/altq -I/usr/src/sys/contrib/ipfilter > -I/usr/src/sys/contrib/pf -I/usr/src/sys/dev/ath > -I/usr/src/sys/dev/ath/ath_hal -I/usr/src/sys/contrib/ngatm > -I/usr/src/sys/dev/twa -I/usr/src/sys/gnu/fs/xfs/FreeBSD > -I/usr/src/sys/gnu/fs/xfs/FreeBSD/support -I/usr/src/sys/gnu/fs/xfs > -I/usr/src/sys/contrib/opensolaris/compat -I/usr/src/sys/dev/cxgb > -D_KERNEL -DHAVE_KERNEL_OPTION_HEADERS -include opt_global.h -fno-common > -finline-limit=8000 --param inline-unit-growth=100 --param > large-function-growth=1000 -mno-align-long-strings > -mpreferred-stack-boundary=2 -mno-mmx -mno-3dnow -mno-sse -mno-sse2 > -mno-sse3 -ffreestanding -fstack-protector > In file included from /usr/src/sys/dev/xen/netfront/netfront.c:32: > /usr/src/sys/sys/sx.h:210:2: error: #error "LOCK_DEBUG not defined, > include before " > mkdep: compile failed > *** Error code 1 > > Stop in /usr/obj/usr/src/sys/XEN. > *** Error code 1 > > Stop in /usr/src. > *** Error code 1 > > Stop in /usr/src. > template_8-CURRENT# > ====================================================== > > -- > Mr. Olli > > > > > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From kmacy at freebsd.org Wed Feb 4 20:04:37 2009 From: kmacy at freebsd.org (Kip Macy) Date: Wed Feb 4 20:04:42 2009 Subject: Time goes backwards In-Reply-To: <87ljsokbdn.fsf@tabernacle.localnet> References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> Message-ID: <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> Can you please try with the latest sources and let me know if you still see this? Thanks, Kip On Mon, Feb 2, 2009 at 9:16 AM, Julian Stecklina wrote: > Mister Olli writes: > >> Hi, >> >> I've had time problems too. The clock did not go backwards, but forwards >> for some time sime, and then jumped back to a certain date. Like >> Sun Jan 25 00:55:12 UTC 2009 >> ... >> Sun Jan 25 01:08:35 UTC 2009 >> Sun Jan 25 00:55:12 UTC 2009 >> Sun Jan 25 00:55:13 UTC 2009 >> ... >> >> It took me quite some time to find a solution that helped for me: >> - activate the independent_wallclock in domO by 'sysctl xen.independent_wallclock=1' > > I booted the DomU with machdep.independent_wallclock=1 as kernel > parameter instead. But that didn't work. sysctl still reported it as > 0. I've tried setting machdep.xen_disable_rtc_set and > machdep.independent_wallclock after booting, but it does not help either. > >> - edit and configure '/etc/ntpd.conf' in domU >> - edit '/etc/rc.conf' and add the following lines >> # NTP stuff >> ntpdate_enable="YES" >> ntpdate_flags="-b " >> ntpd_enable="YES" >> - reboot domU >> >> If you try it on your problem, could you please give me some feedback if >> it worked? > > Doesn't work. :-/ > > MfG, > -- > Julian Stecklina > > Well, take it from an old hand: the only reason it would be easier to > program in C is that you can't easily express complex problems in C, > so you don't. - Erik Naggum (in comp.lang.lisp) > > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From js at alien8.de Thu Feb 5 07:58:27 2009 From: js at alien8.de (Julian Stecklina) Date: Thu Feb 5 07:58:35 2009 Subject: Time goes backwards References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> Message-ID: <8763joeuzz.fsf@tabernacle.localnet> Kip Macy writes: > Can you please try with the latest sources and let me know if you > still see this? I've tried with r188140 and see no change. It still reverts to Jan 25 00:50 every couple of minutes. dmesg is at http://os.inf.tu-dresden.de/~jsteckli/fbsd-xen-dmsg.txt Btw, when I do 'reboot' in the virtual machine, it just stops and has to be recreated. Regards, -- Julian Stecklina Well, take it from an old hand: the only reason it would be easier to program in C is that you can't easily express complex problems in C, so you don't. - Erik Naggum (in comp.lang.lisp) From mister.olli at googlemail.com Thu Feb 5 11:58:22 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Thu Feb 5 11:58:29 2009 Subject: Time goes backwards In-Reply-To: <8763joeuzz.fsf@tabernacle.localnet> References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> <8763joeuzz.fsf@tabernacle.localnet> Message-ID: <1233863896.22204.46.camel@phoenix.blechhirn.net> hi, Am Donnerstag, den 05.02.2009, 16:58 +0100 schrieb Julian Stecklina: > Kip Macy writes: > > > Can you please try with the latest sources and let me know if you > > still see this? > > I've tried with r188140 and see no change. It still reverts to Jan 25 > 00:50 every couple of minutes. dmesg is at The same for me. Setting 'xen.independent_wallclock=0' in the domU and deactivating ntpd in the domU make the time jumps happen nearly immediately: ====================================================== virt-001 template_8-CURRENT # while (true); do echo "date"; sleep 10s; done | xm console template_8-CURRENT Thu Feb 5 20:52:22 CET 2009 template_8-CURRENT# date Thu Feb 5 20:52:32 CET 2009 template_8-CURRENT# date Thu Feb 5 20:52:42 CET 2009 template_8-CURRENT# date Thu Feb 5 20:52:52 CET 2009 template_8-CURRENT# date Sun Jan 25 00:54:16 CET 2009 template_8-CURRENT# date Sun Jan 25 00:54:26 CET 2009 template_8-CURRENT# date Sun Jan 25 00:54:36 CET 2009 template_8-CURRENT# virt-001 template_8-CURRENT # echo "uname -a" || xm console template_8-CURRENT uname -a virt-001 template_8-CURRENT # echo "uname -a" | xm console template_8-CURRENT virt-001 template_8-CURRENT # xm console template_8-CURRENT template_8-CURRENT# uname -a FreeBSD template_8-CURRENT.blechhirn.net 8.0-CURRENT FreeBSD 8.0-CURRENT #0 r188157: Thu Feb 5 14:53:10 CET 2009 root@template_8-CURRENT.blechhirn.net:/usr/obj/usr/src/sys/freebsd8_XEN i386 template_8-CURRENT# > Btw, when I do 'reboot' in the virtual machine, it just stops and has to > be recreated. The same problem here. Even with on_reboot = 'restart' in my config file. Is this a XEN problem, or is it related to the FreeBSD kernel? -- Mr. Olli From kmacy at freebsd.org Thu Feb 5 12:03:04 2009 From: kmacy at freebsd.org (Kip Macy) Date: Thu Feb 5 12:03:11 2009 Subject: Time goes backwards In-Reply-To: <1233863896.22204.46.camel@phoenix.blechhirn.net> References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> <8763joeuzz.fsf@tabernacle.localnet> <1233863896.22204.46.camel@phoenix.blechhirn.net> Message-ID: <3c1674c90902051203n338be726rc18ab1f71c513d6c@mail.gmail.com> On Thu, Feb 5, 2009 at 11:58 AM, Mister Olli wrote: > hi, > > > Am Donnerstag, den 05.02.2009, 16:58 +0100 schrieb Julian Stecklina: >> Kip Macy writes: >> >> > Can you please try with the latest sources and let me know if you >> > still see this? >> >> I've tried with r188140 and see no change. It still reverts to Jan 25 >> 00:50 every couple of minutes. dmesg is at > > The same for me. Setting 'xen.independent_wallclock=0' in the domU and > deactivating ntpd in the domU make the time jumps happen nearly > immediately: > ====================================================== > virt-001 template_8-CURRENT # while (true); do echo "date"; sleep 10s; done | xm console template_8-CURRENT > Thu Feb 5 20:52:22 CET 2009 > template_8-CURRENT# date > Thu Feb 5 20:52:32 CET 2009 > template_8-CURRENT# date > Thu Feb 5 20:52:42 CET 2009 > template_8-CURRENT# date > Thu Feb 5 20:52:52 CET 2009 > template_8-CURRENT# date > Sun Jan 25 00:54:16 CET 2009 > template_8-CURRENT# date > Sun Jan 25 00:54:26 CET 2009 > template_8-CURRENT# date > Sun Jan 25 00:54:36 CET 2009 > template_8-CURRENT# I will take a closer look. This is clearly a separate issue from the loss of timer interrupts - I had assumed that they were related. > virt-001 template_8-CURRENT # echo "uname -a" || xm console template_8-CURRENT > uname -a > virt-001 template_8-CURRENT # echo "uname -a" | xm console template_8-CURRENT > virt-001 template_8-CURRENT # xm console template_8-CURRENT > > template_8-CURRENT# uname -a > FreeBSD template_8-CURRENT.blechhirn.net 8.0-CURRENT FreeBSD 8.0-CURRENT #0 r188157: Thu Feb 5 14:53:10 CET 2009 root@template_8-CURRENT.blechhirn.net:/usr/obj/usr/src/sys/freebsd8_XEN i386 > template_8-CURRENT# > > > >> Btw, when I do 'reboot' in the virtual machine, it just stops and has to >> be recreated. > > The same problem here. Even with > on_reboot = 'restart' > in my config file. > > Is this a XEN problem, or is it related to the FreeBSD kernel? > FreeBSD. A different argument to shutdown needs to be passed for the reboot case. -Kip From mister.olli at googlemail.com Thu Feb 5 12:58:55 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Thu Feb 5 12:59:02 2009 Subject: Time goes backwards In-Reply-To: <3c1674c90902051203n338be726rc18ab1f71c513d6c@mail.gmail.com> References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> <8763joeuzz.fsf@tabernacle.localnet> <1233863896.22204.46.camel@phoenix.blechhirn.net> <3c1674c90902051203n338be726rc18ab1f71c513d6c@mail.gmail.com> Message-ID: <1233867531.22204.50.camel@phoenix.blechhirn.net> Hi, Am Donnerstag, den 05.02.2009, 12:03 -0800 schrieb Kip Macy: > > The same problem here. Even with > > on_reboot = 'restart' > > in my config file. > > > > Is this a XEN problem, or is it related to the FreeBSD kernel? > > > > FreeBSD. A different argument to shutdown needs to be passed for the > reboot case. Can this be done by a userland utility, or does it require a kernel modification? -- Mr. Olli From kmacy at freebsd.org Thu Feb 5 12:59:45 2009 From: kmacy at freebsd.org (Kip Macy) Date: Thu Feb 5 12:59:51 2009 Subject: Time goes backwards In-Reply-To: <1233867531.22204.50.camel@phoenix.blechhirn.net> References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> <8763joeuzz.fsf@tabernacle.localnet> <1233863896.22204.46.camel@phoenix.blechhirn.net> <3c1674c90902051203n338be726rc18ab1f71c513d6c@mail.gmail.com> <1233867531.22204.50.camel@phoenix.blechhirn.net> Message-ID: <3c1674c90902051259m54b09398s3e5a6d10e9eedc14@mail.gmail.com> On Thu, Feb 5, 2009 at 12:58 PM, Mister Olli wrote: > Hi, > > Am Donnerstag, den 05.02.2009, 12:03 -0800 schrieb Kip Macy: >> > The same problem here. Even with >> > on_reboot = 'restart' >> > in my config file. >> > >> > Is this a XEN problem, or is it related to the FreeBSD kernel? >> > >> >> FreeBSD. A different argument to shutdown needs to be passed for the >> reboot case. > > Can this be done by a userland utility, or does it require a kernel > modification? It is a fairly trivial kernel source change. -Kip From mister.olli at googlemail.com Thu Feb 5 13:00:29 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Thu Feb 5 13:00:36 2009 Subject: 8-CURRENT does not build without WITNESS activated In-Reply-To: <3c1674c90902042004m1b7261d7o60a57a80c447582b@mail.gmail.com> References: <1233687579.4832.11.camel@phoenix.blechhirn.net> <3c1674c90902042004m1b7261d7o60a57a80c447582b@mail.gmail.com> Message-ID: <1233867624.22204.53.camel@phoenix.blechhirn.net> Hi, thanks, worked great for me... I'm glad to see that CURRENT makes so fast progress. Great work :-)) -- Mr. Olli Am Mittwoch, den 04.02.2009, 20:04 -0800 schrieb Kip Macy: > adding lock.h will fix that - sorry about it > > Thanks, > Kip > > > On Tue, Feb 3, 2009 at 10:59 AM, Mister Olli wrote: > > Hi. > > > > I'm currently unable to build the 8-CURRENT kernel in a PV domU, without > > "options WITNESS" in my kernel config. > > > > The output is the following: > > ====================================================== > > make -V CFILES -V SYSTEM_CFILES -V GEN_CFILES | MKDEP_CPP="cc -E" > > CC="cc" xargs mkdep -a -f .newdep -O -pipe -std=c99 -g -Wall > > -Wredundant-decls -Wnested-externs -Wstrict-prototypes > > -Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual -Wundef > > -Wno-pointer-sign -fformat-extensions -nostdinc -I. -I/usr/src/sys > > -I/usr/src/sys/contrib/altq -I/usr/src/sys/contrib/ipfilter > > -I/usr/src/sys/contrib/pf -I/usr/src/sys/dev/ath > > -I/usr/src/sys/dev/ath/ath_hal -I/usr/src/sys/contrib/ngatm > > -I/usr/src/sys/dev/twa -I/usr/src/sys/gnu/fs/xfs/FreeBSD > > -I/usr/src/sys/gnu/fs/xfs/FreeBSD/support -I/usr/src/sys/gnu/fs/xfs > > -I/usr/src/sys/contrib/opensolaris/compat -I/usr/src/sys/dev/cxgb > > -D_KERNEL -DHAVE_KERNEL_OPTION_HEADERS -include opt_global.h -fno-common > > -finline-limit=8000 --param inline-unit-growth=100 --param > > large-function-growth=1000 -mno-align-long-strings > > -mpreferred-stack-boundary=2 -mno-mmx -mno-3dnow -mno-sse -mno-sse2 > > -mno-sse3 -ffreestanding -fstack-protector > > In file included from /usr/src/sys/dev/xen/netfront/netfront.c:32: > > /usr/src/sys/sys/sx.h:210:2: error: #error "LOCK_DEBUG not defined, > > include before " > > mkdep: compile failed > > *** Error code 1 > > > > Stop in /usr/obj/usr/src/sys/XEN. > > *** Error code 1 > > > > Stop in /usr/src. > > *** Error code 1 > > > > Stop in /usr/src. > > template_8-CURRENT# > > ====================================================== > > > > -- > > Mr. Olli > > > > > > > > > > _______________________________________________ > > freebsd-xen@freebsd.org mailing list > > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > > From mister.olli at googlemail.com Thu Feb 5 13:11:47 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Thu Feb 5 13:11:54 2009 Subject: Time goes backwards In-Reply-To: <3c1674c90902051259m54b09398s3e5a6d10e9eedc14@mail.gmail.com> References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> <8763joeuzz.fsf@tabernacle.localnet> <1233863896.22204.46.camel@phoenix.blechhirn.net> <3c1674c90902051203n338be726rc18ab1f71c513d6c@mail.gmail.com> <1233867531.22204.50.camel@phoenix.blechhirn.net> <3c1674c90902051259m54b09398s3e5a6d10e9eedc14@mail.gmail.com> Message-ID: <1233868302.22204.65.camel@phoenix.blechhirn.net> hi, Am Donnerstag, den 05.02.2009, 12:59 -0800 schrieb Kip Macy: > On Thu, Feb 5, 2009 at 12:58 PM, Mister Olli wrote: > > Hi, > > > > Am Donnerstag, den 05.02.2009, 12:03 -0800 schrieb Kip Macy: > >> > The same problem here. Even with > >> > on_reboot = 'restart' > >> > in my config file. > >> > > >> > Is this a XEN problem, or is it related to the FreeBSD kernel? > >> > > >> > >> FreeBSD. A different argument to shutdown needs to be passed for the > >> reboot case. > > > > Can this be done by a userland utility, or does it require a kernel > > modification? > > It is a fairly trivial kernel source change. Ok, I'll wait for it ;-))) -- Mr. Olli From mister.olli at googlemail.com Fri Feb 6 01:05:55 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Fri Feb 6 01:06:01 2009 Subject: 8-CURRENT does not build without WITNESS activated In-Reply-To: <3c1674c90902042004m1b7261d7o60a57a80c447582b@mail.gmail.com> References: <1233687579.4832.11.camel@phoenix.blechhirn.net> <3c1674c90902042004m1b7261d7o60a57a80c447582b@mail.gmail.com> Message-ID: <1233911099.22390.6.camel@phoenix.blechhirn.net> Hi, unfortunately building a kernel without options WITNESS options WITNESS_SKIPSPIN creates a unusable kernel. When starting the domain and trying to attach the console I nothing happens, and the domain doesn't seem to start up: ================================================================ virt-001 template_8-CURRENT # xm list Name ID Mem VCPUs State Time(s) Domain-0 0 116 2 r----- 8886.2 virt-001 template_8-CURRENT # xm create -c 00_template_8-CURRENT.XENconfig Using config file "./00_template_8-CURRENT.XENconfig". Started domain template_8-CURRENT xenconsole: Could not open tty `/dev/pts/7': No such file or directory virt-001 template_8-CURRENT # xm list Name ID Mem VCPUs State Time(s) Domain-0 0 116 2 r----- 8894.1 ================================================================ Sometimes I get some python errors from the 'xm create' command: ================================================================ virt-001 template_8-CURRENT # xm create -c 00_template_8-CURRENT.XENconfig Using config file "./00_template_8-CURRENT.XENconfig". Started domain template_8-CURRENT Unexpected error: Please report to xen-devel@lists.xensource.com Traceback (most recent call last): File "/usr/sbin/xm", line 10, in main.main(sys.argv) File "usr/lib/python2.5/site-packages/xen/xm/main.py", line 2883, in main File "usr/lib/python2.5/site-packages/xen/xm/main.py", line 2907, in _run_cmd File "", line 1, in File "usr/lib/python2.5/site-packages/xen/xm/main.py", line 1375, in xm_importcommand File "usr/lib/python2.5/site-packages/xen/xm/create.py", line 1321, in main File "usr/lib/python2.5/site-packages/xen/xm/create.py", line 1348, in do_console OSError: [Errno 10] No child processes virt-001 template_8-CURRENT # xm list Name ID Mem VCPUs State Time(s) Domain-0 0 116 2 r----- 8894.1 ================================================================ When having a look at '/var/log/xen/xend.log' I found the following message: [2009-02-06 09:06:15 3867] WARNING (XendDomainInfo:1617) Domain has crashed: name=template_8-CURRENT id=44. It appears right after the domain has been created. -- Mr. Olli Am Mittwoch, den 04.02.2009, 20:04 -0800 schrieb Kip Macy: > adding lock.h will fix that - sorry about it > > Thanks, > Kip > > > On Tue, Feb 3, 2009 at 10:59 AM, Mister Olli wrote: > > Hi. > > > > I'm currently unable to build the 8-CURRENT kernel in a PV domU, without > > "options WITNESS" in my kernel config. > > > > The output is the following: > > ====================================================== > > make -V CFILES -V SYSTEM_CFILES -V GEN_CFILES | MKDEP_CPP="cc -E" > > CC="cc" xargs mkdep -a -f .newdep -O -pipe -std=c99 -g -Wall > > -Wredundant-decls -Wnested-externs -Wstrict-prototypes > > -Wmissing-prototypes -Wpointer-arith -Winline -Wcast-qual -Wundef > > -Wno-pointer-sign -fformat-extensions -nostdinc -I. -I/usr/src/sys > > -I/usr/src/sys/contrib/altq -I/usr/src/sys/contrib/ipfilter > > -I/usr/src/sys/contrib/pf -I/usr/src/sys/dev/ath > > -I/usr/src/sys/dev/ath/ath_hal -I/usr/src/sys/contrib/ngatm > > -I/usr/src/sys/dev/twa -I/usr/src/sys/gnu/fs/xfs/FreeBSD > > -I/usr/src/sys/gnu/fs/xfs/FreeBSD/support -I/usr/src/sys/gnu/fs/xfs > > -I/usr/src/sys/contrib/opensolaris/compat -I/usr/src/sys/dev/cxgb > > -D_KERNEL -DHAVE_KERNEL_OPTION_HEADERS -include opt_global.h -fno-common > > -finline-limit=8000 --param inline-unit-growth=100 --param > > large-function-growth=1000 -mno-align-long-strings > > -mpreferred-stack-boundary=2 -mno-mmx -mno-3dnow -mno-sse -mno-sse2 > > -mno-sse3 -ffreestanding -fstack-protector > > In file included from /usr/src/sys/dev/xen/netfront/netfront.c:32: > > /usr/src/sys/sys/sx.h:210:2: error: #error "LOCK_DEBUG not defined, > > include before " > > mkdep: compile failed > > *** Error code 1 > > > > Stop in /usr/obj/usr/src/sys/XEN. > > *** Error code 1 > > > > Stop in /usr/src. > > *** Error code 1 > > > > Stop in /usr/src. > > template_8-CURRENT# > > ====================================================== > > > > -- > > Mr. Olli > > > > > > > > > > _______________________________________________ > > freebsd-xen@freebsd.org mailing list > > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > > From mister.olli at googlemail.com Fri Feb 6 01:06:04 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Fri Feb 6 01:06:14 2009 Subject: svn commit: r188200 - head/sys/i386/i386 In-Reply-To: <200902052141.n15LfRnB039904@svn.freebsd.org> References: <200902052141.n15LfRnB039904@svn.freebsd.org> Message-ID: <1233911138.22390.7.camel@phoenix.blechhirn.net> Hi, Just tested this. Works great :-)) -- Mr. Olli Am Donnerstag, den 05.02.2009, 21:41 +0000 schrieb Kip Macy: > Author: kmacy > Date: Thu Feb 5 21:41:27 2009 > New Revision: 188200 > URL: http://svn.freebsd.org/changeset/base/188200 > > Log: > halt APs on reboot > > Modified: > head/sys/i386/i386/vm_machdep.c > > Modified: head/sys/i386/i386/vm_machdep.c > ============================================================================== > --- head/sys/i386/i386/vm_machdep.c Thu Feb 5 21:35:40 2009 (r188199) > +++ head/sys/i386/i386/vm_machdep.c Thu Feb 5 21:41:27 2009 (r188200) > @@ -619,7 +619,7 @@ cpu_reset_real() > if (smp_processor_id() == 0) > HYPERVISOR_shutdown(SHUTDOWN_reboot); > else > - HYPERVISOR_shutdown(SHUTDOWN_reboot); > + HYPERVISOR_shutdown(SHUTDOWN_poweroff); > #endif > #ifdef CPU_ELAN > if (elan_mmcr != NULL) > _______________________________________________ > svn-src-head@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/svn-src-head > To unsubscribe, send any mail to "svn-src-head-unsubscribe@freebsd.org" From sylvain.desbureaux at gmail.com Mon Feb 9 01:18:37 2009 From: sylvain.desbureaux at gmail.com (Sylvain Desbureaux) Date: Mon Feb 9 01:18:44 2009 Subject: login/password? In-Reply-To: <652bbbf50902090105u4ecb2ab8t3234504881077308@mail.gmail.com> References: <652bbbf50902090105u4ecb2ab8t3234504881077308@mail.gmail.com> Message-ID: <652bbbf50902090112m42f9a21dt1aab36df8a7834a9@mail.gmail.com> Ok, found it! its "root/no password" :-) 2009/2/9 Sylvain Desbureaux > Hi all, > I tried FreeBSD7.1 Xen DomU and it seems to work well on a FC8 x86_64 (the > only I have). I have an easy question (I hope ;-) ): what's the default > login/password? I searched but didn't found... > > thanks in advance > > From sylvain.desbureaux at gmail.com Mon Feb 9 01:28:28 2009 From: sylvain.desbureaux at gmail.com (Sylvain Desbureaux) Date: Mon Feb 9 01:28:35 2009 Subject: login/password? Message-ID: <652bbbf50902090105u4ecb2ab8t3234504881077308@mail.gmail.com> Hi all, I tried FreeBSD7.1 Xen DomU and it seems to work well on a FC8 x86_64 (the only I have). I have an easy question (I hope ;-) ): what's the default login/password? I searched but didn't found... thanks in advance From js at alien8.de Mon Feb 9 11:39:36 2009 From: js at alien8.de (Julian Stecklina) Date: Mon Feb 9 11:39:44 2009 Subject: Xen in Virtual Machine? Message-ID: <87d4drxuvn.fsf@tabernacle.localnet> Hello, I am trying to build a kvm (or VirtualBox) virtual machine to painlessly experiment with Xen on my 32-bit Linux laptop (Core Duo L2400 with hardware virtualization foo) or my 64-bit Linux desktop box (Phenom 8450). So far neither of the two manage to run Xen with either of the above mentioned VMMs. Has someone already done this or am I out of luck and have to get a real test box? Regards, -- Julian Stecklina Well, take it from an old hand: the only reason it would be easier to program in C is that you can't easily express complex problems in C, so you don't. - Erik Naggum (in comp.lang.lisp) From mark at rootbsd.net Mon Feb 9 12:14:30 2009 From: mark at rootbsd.net (Mark Price) Date: Mon Feb 9 12:14:38 2009 Subject: Xen in Virtual Machine? In-Reply-To: <87d4drxuvn.fsf@tabernacle.localnet> References: <87d4drxuvn.fsf@tabernacle.localnet> Message-ID: <61b6fbec0902091146l1f4f7f58t852914d2b64a6a41@mail.gmail.com> Hi Julian, So, you want to run Xen hypvervisor inside a kvm or virtualbox machine? If so, I don't think its possible. I tried it also with virtualbox and was unable. I did some research and tweaks but couldn't find an answer except that "it wont work" - probably because Xen hypervisor depends on full hysical CPU access. The Xen hypervisor is pretty stable, I don't see why you couldn't get it running on a Linux workstation. Also note that with some computers you need to turn on Virtualization Technology in the BIOS. Regards, Mark http://www.rootbsd.net On Mon, Feb 9, 2009 at 2:39 PM, Julian Stecklina wrote: > Hello, > > I am trying to build a kvm (or VirtualBox) virtual machine to painlessly > experiment with Xen on my 32-bit Linux laptop (Core Duo L2400 with > hardware virtualization foo) or my 64-bit Linux desktop box (Phenom > 8450). So far neither of the two manage to run Xen with either of the > above mentioned VMMs. Has someone already done this or am I out of luck > and have to get a real test box? > > Regards, > -- > Julian Stecklina > > Well, take it from an old hand: the only reason it would be easier to > program in C is that you can't easily express complex problems in C, > so you don't. - Erik Naggum (in comp.lang.lisp) > > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From mister.olli at googlemail.com Mon Feb 9 13:09:24 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Mon Feb 9 13:09:30 2009 Subject: Xen in Virtual Machine? In-Reply-To: <87d4drxuvn.fsf@tabernacle.localnet> References: <87d4drxuvn.fsf@tabernacle.localnet> Message-ID: <1234213751.28262.3.camel@phoenix.blechhirn.net> Hi, from what I can tell this is impossible since XEN needs to have access to ring0 of the CPU which has the highest priviledge level. And there can only be one kernel at a time. I know that with VMware there's a thing like VMware ESX server in VMware workstation (combination could be slightly different ;-)) possible. I don't know that much about the XEN internas to say if it's possible to modify XEN, so that something similar is possible. I once tried building a XEN dom0 kernel under VMware workstation 6.0 but the building process itself failed with some errors on APIC. -- Mr. Olli Am Montag, den 09.02.2009, 20:39 +0100 schrieb Julian Stecklina: > Hello, > > I am trying to build a kvm (or VirtualBox) virtual machine to painlessly > experiment with Xen on my 32-bit Linux laptop (Core Duo L2400 with > hardware virtualization foo) or my 64-bit Linux desktop box (Phenom > 8450). So far neither of the two manage to run Xen with either of the > above mentioned VMMs. Has someone already done this or am I out of luck > and have to get a real test box? > > Regards, From kmacy at freebsd.org Mon Feb 9 13:28:42 2009 From: kmacy at freebsd.org (Kip Macy) Date: Mon Feb 9 13:28:48 2009 Subject: Xen in Virtual Machine? In-Reply-To: <1234213751.28262.3.camel@phoenix.blechhirn.net> References: <87d4drxuvn.fsf@tabernacle.localnet> <1234213751.28262.3.camel@phoenix.blechhirn.net> Message-ID: <3c1674c90902091328t45ce61b4q96fff69b52101a65@mail.gmail.com> Running Xen under vmware works fine (albeit slowly, and periodically exposing bugs in vmware) - I've done it many times. You're likely to run in to problems on less mature platforms that rely on hardware virtualization. However, there is no fundamental reason that paravirtualized guests should not work there. Cheers, Kip On Mon, Feb 9, 2009 at 1:09 PM, Mister Olli wrote: > Hi, > > from what I can tell this is impossible since XEN needs to have access > to ring0 of the CPU which has the highest priviledge level. And there > can only be one kernel at a time. > > I know that with VMware there's a thing like VMware ESX server in VMware > workstation (combination could be slightly different ;-)) possible. > > I don't know that much about the XEN internas to say if it's possible to > modify XEN, so that something similar is possible. > > I once tried building a XEN dom0 kernel under VMware workstation 6.0 but > the building process itself failed with some errors on APIC. > > -- > Mr. Olli > > > Am Montag, den 09.02.2009, 20:39 +0100 schrieb Julian Stecklina: >> Hello, >> >> I am trying to build a kvm (or VirtualBox) virtual machine to painlessly >> experiment with Xen on my 32-bit Linux laptop (Core Duo L2400 with >> hardware virtualization foo) or my 64-bit Linux desktop box (Phenom >> 8450). So far neither of the two manage to run Xen with either of the >> above mentioned VMMs. Has someone already done this or am I out of luck >> and have to get a real test box? >> >> Regards, > > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From mister.olli at googlemail.com Mon Feb 9 14:05:11 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Mon Feb 9 14:05:19 2009 Subject: Xen in Virtual Machine? In-Reply-To: <3c1674c90902091328t45ce61b4q96fff69b52101a65@mail.gmail.com> References: <87d4drxuvn.fsf@tabernacle.localnet> <1234213751.28262.3.camel@phoenix.blechhirn.net> <3c1674c90902091328t45ce61b4q96fff69b52101a65@mail.gmail.com> Message-ID: <1234217096.28262.8.camel@phoenix.blechhirn.net> Hi, great to hear. The machine I tested this on did not have hardware virtualization support and I think about 1 - 1 1/2 years went down since I've tested this scenario. I will definitely give it a shot the moment I own a HVM capable notebook, since I'm heavily tied to vmware from work, but love XEN in my spare time ;-)) -- Mr. Olli Am Montag, den 09.02.2009, 13:28 -0800 schrieb Kip Macy: > Running Xen under vmware works fine (albeit slowly, and periodically > exposing bugs in vmware) - I've done it many times. You're likely to > run in to problems on less mature platforms that rely on hardware > virtualization. However, there is no fundamental reason that > paravirtualized guests should not work there. > > Cheers, > Kip > > On Mon, Feb 9, 2009 at 1:09 PM, Mister Olli wrote: > > Hi, > > > > from what I can tell this is impossible since XEN needs to have access > > to ring0 of the CPU which has the highest priviledge level. And there > > can only be one kernel at a time. > > > > I know that with VMware there's a thing like VMware ESX server in VMware > > workstation (combination could be slightly different ;-)) possible. > > > > I don't know that much about the XEN internas to say if it's possible to > > modify XEN, so that something similar is possible. > > > > I once tried building a XEN dom0 kernel under VMware workstation 6.0 but > > the building process itself failed with some errors on APIC. > > > > -- > > Mr. Olli > > > > > > Am Montag, den 09.02.2009, 20:39 +0100 schrieb Julian Stecklina: > >> Hello, > >> > >> I am trying to build a kvm (or VirtualBox) virtual machine to painlessly > >> experiment with Xen on my 32-bit Linux laptop (Core Duo L2400 with > >> hardware virtualization foo) or my 64-bit Linux desktop box (Phenom > >> 8450). So far neither of the two manage to run Xen with either of the > >> above mentioned VMMs. Has someone already done this or am I out of luck > >> and have to get a real test box? > >> > >> Regards, > > > > _______________________________________________ > > freebsd-xen@freebsd.org mailing list > > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > > From js at alien8.de Mon Feb 9 16:35:24 2009 From: js at alien8.de (Julian Stecklina) Date: Mon Feb 9 16:35:30 2009 Subject: Xen in Virtual Machine? References: <87d4drxuvn.fsf@tabernacle.localnet> <1234213751.28262.3.camel@phoenix.blechhirn.net> <3c1674c90902091328t45ce61b4q96fff69b52101a65@mail.gmail.com> Message-ID: <878wofxh6p.fsf@tabernacle.localnet> Kip Macy writes: > Running Xen under vmware works fine (albeit slowly, and periodically > exposing bugs in vmware) - I've done it many times. You're likely to > run in to problems on less mature platforms that rely on hardware > virtualization. However, there is no fundamental reason that > paravirtualized guests should not work there. I'll try VMWare. Thanks for the advice. I am still puzzled why Xen presents such a problem for hardware-virtualization based VMMs. Maybe the virtualization guys at our university can clear that up... Regards, -- Julian Stecklina Well, take it from an old hand: the only reason it would be easier to program in C is that you can't easily express complex problems in C, so you don't. - Erik Naggum (in comp.lang.lisp) From kmacy at freebsd.org Mon Feb 9 17:30:19 2009 From: kmacy at freebsd.org (Kip Macy) Date: Mon Feb 9 17:30:25 2009 Subject: Xen in Virtual Machine? In-Reply-To: <878wofxh6p.fsf@tabernacle.localnet> References: <87d4drxuvn.fsf@tabernacle.localnet> <1234213751.28262.3.camel@phoenix.blechhirn.net> <3c1674c90902091328t45ce61b4q96fff69b52101a65@mail.gmail.com> <878wofxh6p.fsf@tabernacle.localnet> Message-ID: <3c1674c90902091730k71b4075bqb9a4746d684bc6e5@mail.gmail.com> On Mon, Feb 9, 2009 at 4:35 PM, Julian Stecklina wrote: > Kip Macy writes: > > I'll try VMWare. Thanks for the advice. I am still puzzled why Xen > presents such a problem for hardware-virtualization based VMMs. Maybe > the virtualization guys at our university can clear that up... > 32-bit Xen uses segments in an unusual way. They may not be implementing emulation support for this correctly. Cheers, Kip From js at alien8.de Mon Feb 9 18:23:11 2009 From: js at alien8.de (Julian Stecklina) Date: Mon Feb 9 18:23:18 2009 Subject: Xen in Virtual Machine? References: <87d4drxuvn.fsf@tabernacle.localnet> <1234213751.28262.3.camel@phoenix.blechhirn.net> <3c1674c90902091328t45ce61b4q96fff69b52101a65@mail.gmail.com> <878wofxh6p.fsf@tabernacle.localnet> <3c1674c90902091730k71b4075bqb9a4746d684bc6e5@mail.gmail.com> Message-ID: <878wofghdn.fsf@tabernacle.lan> Kip Macy writes: > On Mon, Feb 9, 2009 at 4:35 PM, Julian Stecklina wrote: >> Kip Macy writes: >> >> I'll try VMWare. Thanks for the advice. I am still puzzled why Xen >> presents such a problem for hardware-virtualization based VMMs. Maybe >> the virtualization guys at our university can clear that up... >> > > 32-bit Xen uses segments in an unusual way. They may not be > implementing emulation support for this correctly. Do you mean using additional protection rings? Yes, no one does that except Xen. Regards, -- Julian Stecklina Well, take it from an old hand: the only reason it would be easier to program in C is that you can't easily express complex problems in C, so you don't. - Erik Naggum (in comp.lang.lisp) From xen at komadev.de Tue Feb 10 06:45:04 2009 From: xen at komadev.de (xen@komadev.de) Date: Tue Feb 10 06:45:11 2009 Subject: domU ignores external xm commands In-Reply-To: <10619699.41031234276618948.JavaMail.root@zim.freshx.de> Message-ID: <27077918.41051234277101668.JavaMail.root@zim.freshx.de> Hi, my current build (r188436M) ignores external xm commands (on Xen 3.3.1), such as # xm reboot # xm shutdown # xm dump-core -L bsd8 Dumping core of domain: bsd8 ... Error: Failed to dump core: (1, 'Internal error', 'p2m_size < nr_pages -1 (0 < 3ffff') # xm save (creates a (small) snapshot file but hangs, domain gets renamed to migrating-) # xm restore (depends on xm save) # xm reset works and does a hard reset Cheers Kai From xen at komadev.de Tue Feb 10 07:59:36 2009 From: xen at komadev.de (Kai Mosebach) Date: Tue Feb 10 07:59:46 2009 Subject: domU page faults under heavy network load In-Reply-To: <17604735.41331234281418189.JavaMail.root@zim.freshx.de> Message-ID: <13732317.41351234281573360.JavaMail.root@zim.freshx.de> Hi, running r188436M on Xen3.3.1 (Centos5, x86_64). The reason is not to clear but seems related to heavy load on pipes + memory + network? It is repeatable with : bsd8# dd if=/dev/zero | nc otherhost 12345 Fatal trap 12: page fault while in kernel mode cpuid = 0; apic id = 00 fault virtual address = 0xd216dd00 fault code = supervisor read, page not present instruction pointer = 0x21:0xc02e4d40 stack pointer = 0x29:0xc332aba8 frame pointer = 0x29:0xc332abfc code segment = base 0x0, limit 0xfffff, type 0x1b = DPL 1, pres 1, def32 1, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 12 (irq137: xn) [thread pid 12 tid 100025 ] Stopped at xlvbd_add+0x2f70: movl 0x900(%edx,%eax,4),%edx db> bt On otherhost i get some stuff (between 19 and 62 MB), after nc closes down eventually otherhost# nc -l 12345 | cpipe -vt -b 65535 >/dev/null ... thru: 108930.791ms at 184.3kB/s ( 184.3kB/s avg) 19.6MB best Kai From xen at komadev.de Tue Feb 10 08:47:49 2009 From: xen at komadev.de (Kai Mosebach) Date: Tue Feb 10 08:47:59 2009 Subject: domU reboots on halt In-Reply-To: <13633077.41461234284395344.JavaMail.root@zim.freshx.de> Message-ID: <22869163.41481234284465542.JavaMail.root@zim.freshx.de> Hi, me, again ... Shuting down my domU (r188436M) simply reboots it when using # halt -p # shutdown -h now So i have to shut it down and issue an xm destroy on a certain point to get rid of it... Best Kai From xen at komadev.de Tue Feb 10 09:05:22 2009 From: xen at komadev.de (Kai Mosebach) Date: Tue Feb 10 09:05:32 2009 Subject: domU performance In-Reply-To: <20984507.41581234284966568.JavaMail.root@zim.freshx.de> Message-ID: <27423886.41631234285518798.JavaMail.root@zim.freshx.de> Hi, just as a note, since it is not important in the current state, i guess (but maybe you might have some tweaking tips) Testing the raw performance on my domU (r188436 compiled without WITNESS) seems to fall behind a Linux domU on the same machine: BSD8 Reading: # dd if=/dev/ad0s1 bs=1024k | cpipe -vt -b 65535 | cat >/dev/null ... thru: 1200.000ms at 53.3MB/s ( 57.3MB/s avg) 2.2GB ... 2401239040 bytes transferred in 40.000000 secs (60030976 bytes/sec) BSD8 Writing : # dd if=/dev/zero bs=1024k | cpipe -vt -b 65535 |dd of=/bigfile ... thru: 4400.000ms at 14.5MB/s ( 13.2MB/s avg) 2.8GB ... 3044998656 bytes transferred in 222.500000 secs (13685387 bytes/sec) On the Linux side i get Linux Reading (~15% faster): # dd if=/dev/sda bs=1024k | cat | dd of=/dev/null 4870635520 bytes (4.9 GB) copied, 73.7775 seconds, 66.0 MB/s Linux Writing: (~90% faster) # dd if=/dev/zero bs=1024k | cat | dd of=/root/test2 2473590784 bytes (2.5 GB) copied, 98.9709 seconds, 25.0 MB/s cheers Kai From kmacy at freebsd.org Tue Feb 10 11:30:02 2009 From: kmacy at freebsd.org (Kip Macy) Date: Tue Feb 10 11:30:09 2009 Subject: domU reboots on halt In-Reply-To: <22869163.41481234284465542.JavaMail.root@zim.freshx.de> References: <13633077.41461234284395344.JavaMail.root@zim.freshx.de> <22869163.41481234284465542.JavaMail.root@zim.freshx.de> Message-ID: <3c1674c90902101129g607b1df2t310f479c779e0bee@mail.gmail.com> Ok, looks like I'll have to re-visit that. Thanks, Kip On Tue, Feb 10, 2009 at 8:47 AM, Kai Mosebach wrote: > Hi, > > me, again ... > > Shuting down my domU (r188436M) simply reboots it when using > > # halt -p > # shutdown -h now > > So i have to shut it down and issue an xm destroy on a certain point to get rid of it... > > Best Kai > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From kmacy at freebsd.org Tue Feb 10 11:30:54 2009 From: kmacy at freebsd.org (Kip Macy) Date: Tue Feb 10 11:31:01 2009 Subject: domU page faults under heavy network load In-Reply-To: <13732317.41351234281573360.JavaMail.root@zim.freshx.de> References: <17604735.41331234281418189.JavaMail.root@zim.freshx.de> <13732317.41351234281573360.JavaMail.root@zim.freshx.de> Message-ID: <3c1674c90902101130r58c05f07we473c0897fe29481@mail.gmail.com> How big is the memory allocation to the VM? Thanks, Kip On Tue, Feb 10, 2009 at 7:59 AM, Kai Mosebach wrote: > Hi, > > running r188436M on Xen3.3.1 (Centos5, x86_64). The reason is not to clear but seems related to heavy load on pipes + memory + network? > It is repeatable with : > > bsd8# dd if=/dev/zero | nc otherhost 12345 > > Fatal trap 12: page fault while in kernel mode > cpuid = 0; apic id = 00 > fault virtual address = 0xd216dd00 > fault code = supervisor read, page not present > instruction pointer = 0x21:0xc02e4d40 > stack pointer = 0x29:0xc332aba8 > frame pointer = 0x29:0xc332abfc > code segment = base 0x0, limit 0xfffff, type 0x1b > = DPL 1, pres 1, def32 1, gran 1 > processor eflags = interrupt enabled, resume, IOPL = 0 > current process = 12 (irq137: xn) > [thread pid 12 tid 100025 ] > Stopped at xlvbd_add+0x2f70: movl 0x900(%edx,%eax,4),%edx > db> bt > > > > On otherhost i get some stuff (between 19 and 62 MB), after nc closes down eventually > > otherhost# nc -l 12345 | cpipe -vt -b 65535 >/dev/null > ... > thru: 108930.791ms at 184.3kB/s ( 184.3kB/s avg) 19.6MB > > best Kai > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From kai at freshx.de Tue Feb 10 11:40:13 2009 From: kai at freshx.de (Kai Mosebach) Date: Tue Feb 10 11:40:20 2009 Subject: domU page faults under heavy network load In-Reply-To: <3c1674c90902101130r58c05f07we473c0897fe29481@mail.gmail.com> Message-ID: 1024meg On 2/10/09 8:30 PM, "Kip Macy" wrote: > How big is the memory allocation to the VM? > > Thanks, > Kip > > > On Tue, Feb 10, 2009 at 7:59 AM, Kai Mosebach wrote: >> Hi, >> >> running r188436M on Xen3.3.1 (Centos5, x86_64). The reason is not to clear >> but seems related to heavy load on pipes + memory + network? >> It is repeatable with : >> >> bsd8# dd if=/dev/zero | nc otherhost 12345 >> >> Fatal trap 12: page fault while in kernel mode >> cpuid = 0; apic id = 00 >> fault virtual address = 0xd216dd00 >> fault code = supervisor read, page not present >> instruction pointer = 0x21:0xc02e4d40 >> stack pointer = 0x29:0xc332aba8 >> frame pointer = 0x29:0xc332abfc >> code segment = base 0x0, limit 0xfffff, type 0x1b >> = DPL 1, pres 1, def32 1, gran 1 >> processor eflags = interrupt enabled, resume, IOPL = 0 >> current process = 12 (irq137: xn) >> [thread pid 12 tid 100025 ] >> Stopped at xlvbd_add+0x2f70: movl 0x900(%edx,%eax,4),%edx >> db> bt >> >> >> >> On otherhost i get some stuff (between 19 and 62 MB), after nc closes down >> eventually >> >> otherhost# nc -l 12345 | cpipe -vt -b 65535 >/dev/null >> ... >> thru: 108930.791ms at 184.3kB/s ( 184.3kB/s avg) 19.6MB >> >> best Kai >> _______________________________________________ >> freebsd-xen@freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-xen >> To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" >> From bj at SerNet.DE Wed Feb 11 02:56:19 2009 From: bj at SerNet.DE (=?iso-8859-1?Q?Bj=F6rn?= JACKE) Date: Wed Feb 11 02:56:26 2009 Subject: DomU time resets Message-ID: Hi, with CURRENT FreeBSD kernel I have the problem that time periodically jumps back to Sep 16 08:46. That happens every couple of minutes. I set up a cron job setting the correct time via ntpdate every minute but that is a problem when the time jump happens during ntpdate is just running - it hangs then. There is no cpu and no io load on the machine which might correlate with the time jumps. Any idea what might cause the problem? XEN is 3.2.0 with Linux Dom0. Bj?rn From xen at komadev.de Wed Feb 11 03:03:55 2009 From: xen at komadev.de (Kai Mosebach) Date: Wed Feb 11 03:04:02 2009 Subject: DomU time resets In-Reply-To: Message-ID: <26077784.42351234350232426.JavaMail.root@zim.freshx.de> Hi Bj?rn, This is a known problem (see other mails on the list) The only workaround at the moment is setting this in the dom0: sysctl xen.independent_wallclock=1 Best Kai ----- "Bj?rn JACKE" schrieb: > Hi, > > with CURRENT FreeBSD kernel I have the problem that time periodically > jumps > back to Sep 16 08:46. That happens every couple of minutes. I set up a > cron job > setting the correct time via ntpdate every minute but that is a > problem when > the time jump happens during ntpdate is just running - it hangs then. > There is > no cpu and no io load on the machine which might correlate with the > time jumps. > Any idea what might cause the problem? XEN is 3.2.0 with Linux Dom0. > > Bj?rn > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" From mister.olli at googlemail.com Wed Feb 11 06:42:30 2009 From: mister.olli at googlemail.com (Mister Olli) Date: Wed Feb 11 06:42:37 2009 Subject: DomU time resets In-Reply-To: <26077784.42351234350232426.JavaMail.root@zim.freshx.de> References: <26077784.42351234350232426.JavaMail.root@zim.freshx.de> Message-ID: <1234361252.28262.20.camel@phoenix.blechhirn.net> Hi bj?rn, and don't forget configuring & activating 'ntpd' within the domU ;-)) -- Mr. Olli Am Mittwoch, den 11.02.2009, 12:03 +0100 schrieb Kai Mosebach: > Hi Bj?rn, > > This is a known problem (see other mails on the list) > The only workaround at the moment is setting this in the dom0: > > sysctl xen.independent_wallclock=1 > > Best Kai > > ----- "Bj?rn JACKE" schrieb: > > > Hi, > > > > with CURRENT FreeBSD kernel I have the problem that time periodically > > jumps > > back to Sep 16 08:46. That happens every couple of minutes. I set up a > > cron job > > setting the correct time via ntpdate every minute but that is a > > problem when > > the time jump happens during ntpdate is just running - it hangs then. > > There is > > no cpu and no io load on the machine which might correlate with the > > time jumps. > > Any idea what might cause the problem? XEN is 3.2.0 with Linux Dom0. > > > > Bj?rn > > _______________________________________________ > > freebsd-xen@freebsd.org mailing list > > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" From dan_johnson at doomed.us Wed Feb 11 08:31:46 2009 From: dan_johnson at doomed.us (Dan Johnson) Date: Wed Feb 11 08:31:52 2009 Subject: DomU time resets In-Reply-To: <1234361252.28262.20.camel@phoenix.blechhirn.net> References: <26077784.42351234350232426.JavaMail.root@zim.freshx.de> <1234361252.28262.20.camel@phoenix.blechhirn.net> Message-ID: I also had a problem with this, however on my setup for some reason it would still revert the time to Jan 25th 2009 every minute or so, presumably ntpd didn't know what to do with the time that far out of whack. sysctl xen.independent_wallclock=1 didn't seem to do anything, so i changed sys/i386/xen/clock.c if (shadow_tv_version != HYPERVISOR_shared_info->wc_version) { to if ((shadow_tv_version != HYPERVISOR_shared_info->wc_version) && (independent_wallclock!=1)) { Kludgy i'm sure, but it stopped the Jan25th thing. My time still skews slowly, but ntpd should be able to adjust that without changing ntpds tolerances. On Wed, Feb 11, 2009 at 9:07 AM, Mister Olli wrote: > Hi bj?rn, > > and don't forget configuring & activating 'ntpd' within the domU ;-)) > > -- > Mr. Olli > > > Am Mittwoch, den 11.02.2009, 12:03 +0100 schrieb Kai Mosebach: >> Hi Bj?rn, >> >> This is a known problem (see other mails on the list) >> The only workaround at the moment is setting this in the dom0: >> >> sysctl xen.independent_wallclock=1 >> >> Best Kai >> >> ----- "Bj?rn JACKE" schrieb: >> >> > Hi, >> > >> > with CURRENT FreeBSD kernel I have the problem that time periodically >> > jumps >> > back to Sep 16 08:46. That happens every couple of minutes. I set up a >> > cron job >> > setting the correct time via ntpdate every minute but that is a >> > problem when >> > the time jump happens during ntpdate is just running - it hangs then. >> > There is >> > no cpu and no io load on the machine which might correlate with the >> > time jumps. >> > Any idea what might cause the problem? XEN is 3.2.0 with Linux Dom0. >> > >> > Bj?rn >> > _______________________________________________ >> > freebsd-xen@freebsd.org mailing list >> > http://lists.freebsd.org/mailman/listinfo/freebsd-xen >> > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" >> _______________________________________________ >> freebsd-xen@freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-xen >> To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > > _______________________________________________ > freebsd-xen@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-xen > To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" > From kmacy at freebsd.org Sat Feb 14 22:54:06 2009 From: kmacy at freebsd.org (Kip Macy) Date: Sat Feb 14 22:54:12 2009 Subject: domU page faults under heavy network load In-Reply-To: References: <3c1674c90902101130r58c05f07we473c0897fe29481@mail.gmail.com> Message-ID: <3c1674c90902142254w62aee712j8ae9010361ec5d2a@mail.gmail.com> Is this a recent version of HEAD? I'm not able to reproduce. <...> thru: 1947.346ms at 32.9MB/s ( 27.8MB/s avg) 3.3GB thru: 1946.356ms at 32.9MB/s ( 27.9MB/s avg) 3.4GB thru: 1946.910ms at 32.9MB/s ( 27.9MB/s avg) 3.4GB <...> Maybe some variation in the options? Thanks, Kip On Tue, Feb 10, 2009 at 11:40 AM, Kai Mosebach wrote: > 1024meg > > > On 2/10/09 8:30 PM, "Kip Macy" wrote: > >> How big is the memory allocation to the VM? >> >> Thanks, >> Kip >> >> >> On Tue, Feb 10, 2009 at 7:59 AM, Kai Mosebach wrote: >>> Hi, >>> >>> running r188436M on Xen3.3.1 (Centos5, x86_64). The reason is not to clear >>> but seems related to heavy load on pipes + memory + network? >>> It is repeatable with : >>> >>> bsd8# dd if=/dev/zero | nc otherhost 12345 >>> >>> Fatal trap 12: page fault while in kernel mode >>> cpuid = 0; apic id = 00 >>> fault virtual address = 0xd216dd00 >>> fault code = supervisor read, page not present >>> instruction pointer = 0x21:0xc02e4d40 >>> stack pointer = 0x29:0xc332aba8 >>> frame pointer = 0x29:0xc332abfc >>> code segment = base 0x0, limit 0xfffff, type 0x1b >>> = DPL 1, pres 1, def32 1, gran 1 >>> processor eflags = interrupt enabled, resume, IOPL = 0 >>> current process = 12 (irq137: xn) >>> [thread pid 12 tid 100025 ] >>> Stopped at xlvbd_add+0x2f70: movl 0x900(%edx,%eax,4),%edx >>> db> bt >>> >>> >>> >>> On otherhost i get some stuff (between 19 and 62 MB), after nc closes down >>> eventually >>> >>> otherhost# nc -l 12345 | cpipe -vt -b 65535 >/dev/null >>> ... >>> thru: 108930.791ms at 184.3kB/s ( 184.3kB/s avg) 19.6MB >>> >>> best Kai >>> _______________________________________________ >>> freebsd-xen@freebsd.org mailing list >>> http://lists.freebsd.org/mailman/listinfo/freebsd-xen >>> To unsubscribe, send any mail to "freebsd-xen-unsubscribe@freebsd.org" >>> > > > From js at alien8.de Mon Feb 16 06:43:13 2009 From: js at alien8.de (Julian Stecklina) Date: Mon Feb 16 06:43:19 2009 Subject: Time goes backwards References: <877i4d1p46.fsf@tabernacle.localnet> <1233285639.24925.12.camel@phoenix.blechhirn.net> <87ljsokbdn.fsf@tabernacle.localnet> <3c1674c90902042004x299bee3ay7a231a27252e7e55@mail.gmail.com> <8763joeuzz.fsf@tabernacle.localnet> <1233863896.22204.46.camel@phoenix.blechhirn.net> <3c1674c90902051203n338be726rc18ab1f71c513d6c@mail.gmail.com> Message-ID: <87skmeqw7h.fsf@tabernacle.lan> Kip Macy writes: > I will take a closer look. This is clearly a separate issue from the > loss of timer interrupts - I had assumed that they were related. I've seen that you made some changes to the timer code. Is this related to this issue? Btw, VMWare didn't cut it either in building a virtual Xen test box. *bummer* Regards, -- Julian Stecklina Well, take it from an old hand: the only reason it would be easier to program in C is that you can't easily express complex problems in C, so you don't. - Erik Naggum (in comp.lang.lisp) From xen at komadev.de Tue Feb 17 12:16:41 2009 From: xen at komadev.de (xen@komadev.de) Date: Tue Feb 17 12:16:50 2009 Subject: domU page faults under heavy network load In-Reply-To: <13712113.49661234900587705.JavaMail.root@zim.freshx.de> Message-ID: <14830020.49681234901798034.JavaMail.root@zim.freshx.de> Hi, in the kernel i disabled: WITNESS WITNESS_SKIPSPIN in /etc/make.conf i enabled DEBUG_FLAGS=-ggdb To make sure that user/kernel are in sync i just rebuilt/installed both. I furthermore tracked down the problem, since it seems to occur only between the dom0 and its domU, but not between the dom0 and a "real" remote host. FreeBSD 8.0-CURRENT #0 r188718: Tue Feb 17 18:29:43 CET 2009 # dd if=/dev/zero bs=1024k| cpipe -vt | nc dom0 12345 thru: 5500.000ms at 23.3kB/s ( 23.3kB/s avg) 128.0kB thru: 3200.000ms at 40.0kB/s ( 29.4kB/s avg) 256.0kB ... thru: 0.000ms at infGB/s ( 398.8kB/s avg) 5.4MB thru: 0.000ms at infGB/s ( 408.1kB/s avg) 5.5MB Fatal trap 12: page fault while in kernel mode cpuid = 0; apic id = 00 fault virtual address = 0xd1d9a500 fault code = supervisor read, page not present instruction pointer = 0x21:0xc02e59b0 stack pointer = 0x29:0xc332aba8 frame pointer = 0x29:0xc332abfc code segment = base 0x0, limit 0xfffff, type 0x1b = DPL 1, pres 1, def32 1, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 12 (irq137: xn) [thread pid 12 tid 100025 ] Stopped at xlvbd_add+0x2f70: movl 0x900(%edx,%eax,4),%edx Second System (AMD64, CentOS5.2, Xen 3.3.1): .... thru: 0.000ms at infGB/s ( 598.1kB/s avg) 82.9MB thru: 0.000ms at infGB/s ( 599.0kB/s avg) 83.0MB Fatal trap 12: page fault while in kernel mode cpuid = 0; apic id = 00 fault virtual address = 0xc433a140 fault code = supervisor read, page not present instruction pointer = 0x21:0xc02db38e stack pointer = 0x29:0xc332ab7c frame pointer = 0x29:0xc332ab84 code segment = base 0x0, limit 0xfffff, type 0x1b = DPL 1, pres 1, def32 1, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 12 (irq137: xn) [thread pid 12 tid 100025 ] Stopped at gnttab_claim_grant_reference+0x2e: movl 0(%eax,%edx,4),%eax Best Kai PS : Between the dom0 and some random system via Gigabit Ethernet i get only 1/4 of the real machine speed # dd if=/dev/zero bs=1024k | cpipe -vt -b 65535 | nc other 12345 thru: 7003.500ms at 9.1MB/s ( 9.1MB/s avg) 64.0MB thru: 3301.650ms at 19.4MB/s ( 12.4MB/s avg) 128.0MB thru: 1600.800ms at 40.0MB/s ( 16.1MB/s avg) 192.0MB thru: 1800.900ms at 35.5MB/s ( 18.7MB/s avg) 256.0MB thru: 2601.300ms at 24.6MB/s ( 19.6MB/s avg) 320.0MB thru: 1400.700ms at 45.7MB/s ( 21.7MB/s avg) 384.0MB thru: 1200.600ms at 53.3MB/s ( 23.7MB/s avg) 448.0MB thru: 1600.800ms at 40.0MB/s ( 25.0MB/s avg) 512.0MB thru: 1500.750ms at 42.6MB/s ( 26.2MB/s avg) 576.0MB thru: 1900.682ms at 33.7MB/s ( 26.8MB/s avg) 640.0MB thru: 2000.035ms at 32.0MB/s ( 27.2MB/s avg) 704.0MB thru: 2300.000ms at 27.8MB/s ( 27.2MB/s avg) 768.0MB thru: 1700.000ms at 37.6MB/s ( 27.8MB/s avg) 832.0MB ----- "Kip Macy" schrieb: > Is this a recent version of HEAD? I'm not able to reproduce. > > <...> > thru: 1947.346ms at 32.9MB/s ( 27.8MB/s avg) 3.3GB > thru: 1946.356ms at 32.9MB/s ( 27.9MB/s avg) 3.4GB > thru: 1946.910ms at 32.9MB/s ( 27.9MB/s avg) 3.4GB > <...> > > Maybe some variation in the options? > > Thanks, > Kip > > On Tue, Feb 10, 2009 at 11:40 AM, Kai Mosebach wrote: > > 1024meg > > > > > > On 2/10/09 8:30 PM, "Kip Macy" wrote: > > > >> How big is the memory allocation to the VM? > >> > >> Thanks, > >> Kip > >> > >> > >> On Tue, Feb 10, 2009 at 7:59 AM, Kai Mosebach > wrote: > >>> Hi, > >>> > >>> running r188436M on Xen3.3.1 (Centos5, x86_64). The reason is not > to clear > >>> but seems related to heavy load on pipes + memory + network? > >>> It is repeatable with : > >>> > >>> bsd8# dd if=/dev/zero | nc otherhost 12345 > >>> > >>> Fatal trap 12: page fault while in kernel mode > >>> cpuid = 0; apic id = 00 > >>> fault virtual address = 0xd216dd00 > >>> fault code = supervisor read, page not present > >>> instruction pointer = 0x21:0xc02e4d40 > >>> stack pointer = 0x29:0xc332aba8 > >>> frame pointer = 0x29:0xc332abfc > >>> code segment = base 0x0, limit 0xfffff, type 0x1b > >>> = DPL 1, pres 1, def32 1, gran 1 > >>> processor eflags = interrupt enabled, resume, IOPL = 0 > >>> current process = 12 (irq137: xn) > >>> [thread pid 12 tid 100025 ] > >>> Stopped at xlvbd_add+0x2f70: movl > 0x900(%edx,%eax,4),%edx > >>> db> bt > >>> > >>> > >>> > >>> On otherhost i get some stuff (between 19 and 62 MB), after nc > closes down > >>> eventually > >>> > >>> otherhost# nc -l 12345 | cpipe -vt -b 65535 >/dev/null > >>> ... > >>> thru: 108930.791ms at 184.3kB/s ( 184.3kB/s avg) 19.6MB > >>> > >>> best Kai > >>> _______________________________________________ > >>> freebsd-xen@freebsd.org mailing list > >>> http://lists.freebsd.org/mailman/listinfo/freebsd-xen > >>> To unsubscribe, send any mail to > "freebsd-xen-unsubscribe@freebsd.org" > >>> > > > > > > From berhoeckner at googlemail.com Tue Feb 24 12:21:15 2009 From: berhoeckner at googlemail.com (=?ISO-8859-1?Q?Bernhard_H=F6ckner?=) Date: Tue Feb 24 12:21:21 2009 Subject: Building/Installing CURRENT as domU Message-ID: <59208d7f0902241157q6de1a8fbqe0ff015957c2d402@mail.gmail.com> Hi, i was looking for informations how to build and install a CURRENT as domU or how the kernel should be configured, maybe someone can give me some hints/pointers? Thanks in advance, Bernhard From js at alien8.de Tue Feb 24 12:37:33 2009 From: js at alien8.de (Julian Stecklina) Date: Tue Feb 24 12:37:40 2009 Subject: Building/Installing CURRENT as domU References: <59208d7f0902241157q6de1a8fbqe0ff015957c2d402@mail.gmail.com> Message-ID: <87mycbbmgu.fsf@tabernacle.lan> Bernhard H?ckner writes: > Hi, > > i was looking for informations how to build and install a CURRENT as > domU or how the kernel should be configured, maybe someone can give me > some hints/pointers? There is a ready-made XEN kernel configuration. Basically you just prepare a disk image, use make buildworld buildkernel installworld installkernel distribution DESTDIR=/foo KERNCONF=XEN and fiddle in /etc until you have a working system, like with every from-source install. I had no problems with that. Regards, -- Julian Stecklina Always code as if the guy who ends up maintaining your code will be a violent psychopath who knows where you live. - Martin Golding