Re: [CFT] Paravirtualized KVM clock
On 2015-Jan-04 11:56:14 -0600, Bryan Venteicher bry...@daemoninthecloset.org wrote: For the last few weeks, I've been working on adding support for KVM clock in the projects/paravirt branch. Currently, a KVM VM guest will end up selecting either the HPET or ACPI as the timecounter source. Unfortunately, this is very costly since every timecounter fetch causes a VM exit. KVM clock allows the guest to use the TSC instead; it is very similar to the existing Xen timer. A somewhat late response but have you looked at https://github.com/blitz/freebsd/commit/cdc5f872b3e48cc0dda031fc7d6bdedc65c3148f I've been running this[*] on a Google Compute Engine instance for about 6 months without problems. [*] I had to patch out the test for KVM_FEATURE_CLOCKSOURCE_STABLE_BIT but I think that's a GCE issue. -- Peter Jeremy pgpi9_M8QUFuE.pgp Description: PGP signature
panic in pmap_remove_pages()
Just got this panic. If anyone is interested I have the kenrel and core, so can do some additional poking around. troutmask.apl.washington.edu dumped core - see /var/crash/vmcore.0 Wed Jan 21 13:28:04 PST 2015 FreeBSD troutmask.apl.washington.edu 11.0-CURRENT FreeBSD 11.0-CURRENT #7 r276378M: Mon Dec 29 14:13:57 PST 2014 ka...@troutmask.apl.washington.edu:/data/obj/usr/src/sys/SPEW amd64 panic: general protection fault Unread portion of the kernel message buffer: Fatal trap 9: general protection fault while in kernel mode cpuid = 3; apic id = 13 instruction pointer = 0x20:0x8079abf9 stack pointer = 0x28:0xfe047325e360 frame pointer = 0x28:0xfe047325e440 code segment= base 0x0, limit 0xf, type 0x1b = DPL 0, pres 1, long 1, def32 0, gran 1 processor eflags= interrupt enabled, resume, IOPL = 0 current process = 41779 (z) trap number = 9 panic: general protection fault cpuid = 3 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfe047325e020 panic() at panic+0x1c1/frame 0xfe047325e0e0 trap_fatal() at trap_fatal+0x396/frame 0xfe047325e140 trap() at trap+0x6ce/frame 0xfe047325e2a0 calltrap() at calltrap+0x8/frame 0xfe047325e2a0 --- trap 0x9, rip = 0x8079abf9, rsp = 0xfe047325e360, rbp = 0xfe047325e440 --- pmap_remove_pages() at pmap_remove_pages+0x539/frame 0xfe047325e440 exec_new_vmspace() at exec_new_vmspace+0x180/frame 0xfe047325e4a0 exec_elf64_imgact() at exec_elf64_imgact+0x6c0/frame 0xfe047325e570 kern_execve() at kern_execve+0x484/frame 0xfe047325e8c0 sys_execve() at sys_execve+0x35/frame 0xfe047325e920 amd64_syscall() at amd64_syscall+0x289/frame 0xfe047325ea30 Xfast_syscall() at Xfast_syscall+0xfb/frame 0xfe047325ea30 --- syscall (59, FreeBSD ELF64, sys_execve), rip = 0x4251ba, rsp = 0x7e8ebab8, rbp = 0x7e8ec1c0 --- Uptime: 22d22h22m46s #0 doadump (textdump=1) at pcpu.h:219 219 pcpu.h: No such file or directory. in pcpu.h (kgdb) #0 doadump (textdump=1) at pcpu.h:219 #1 0x80555bd7 in kern_reboot (howto=260) at /usr/src/sys/kern/kern_shutdown.c:447 #2 0x80556040 in panic (fmt=value optimized out) at /usr/src/sys/kern/kern_shutdown.c:746 #3 0x807a2986 in trap_fatal (frame=value optimized out, eva=value optimized out) at /usr/src/sys/amd64/amd64/trap.c:867 #4 0x807a25de in trap (frame=value optimized out) at /usr/src/sys/amd64/amd64/trap.c:201 #5 0x80787ca3 in calltrap () at /usr/src/sys/amd64/amd64/exception.S:235 #6 0x8079abf9 in pmap_remove_pages (pmap=0xf801c627dec8) at /usr/src/sys/amd64/amd64/pmap.c:5389 #7 0x8051fa00 in exec_new_vmspace (imgp=0xfe047325e6e0, sv=0x80b3e8e8) at /usr/src/sys/kern/kern_exec.c:1036 #8 0x804fed20 in exec_elf64_imgact (imgp=0xfe047325e6e0) at /usr/src/sys/kern/imgact_elf.c:830 #9 0x8051e4f4 in kern_execve (td=0xf8027588f490, args=0xfe047325e8d8, mac_p=0x1da) at /usr/src/sys/kern/kern_exec.c:486 #10 0x8051de15 in sys_execve (td=value optimized out, uap=value optimized out) at /usr/src/sys/kern/kern_exec.c:210 #11 0x807a3199 in amd64_syscall (td=0xf8027588f490, traced=0) at subr_syscall.c:133 #12 0x80787f8b in Xfast_syscall () at /usr/src/sys/amd64/amd64/exception.S:395 #13 0x004251ba in ?? () -- Steve ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: [CFT] Paravirtualized KVM clock
On Wed, Jan 21, 2015 at 3:15 PM, Peter Jeremy pe...@rulingia.com wrote: On 2015-Jan-04 11:56:14 -0600, Bryan Venteicher bry...@daemoninthecloset.org wrote: For the last few weeks, I've been working on adding support for KVM clock in the projects/paravirt branch. Currently, a KVM VM guest will end up selecting either the HPET or ACPI as the timecounter source. Unfortunately, this is very costly since every timecounter fetch causes a VM exit. KVM clock allows the guest to use the TSC instead; it is very similar to the existing Xen timer. A somewhat late response but have you looked at https://github.com/blitz/freebsd/commit/cdc5f872b3e48cc0dda031fc7d6bdedc65c3148f I've been running this[*] on a Google Compute Engine instance for about 6 months without problems. A goal of my work was to put a bit of infrastructure in place so FreeBSD can support pvops across a variety of hypervisors. KVMCLOCK happens to be about the easiest to implement, and has a decent performance win for many situations. I think that commit is broken on SMP guests: CPU_FOREACH() does not switch the current CPU, so it just keeps writing to the MSR on the BSP. [*] I had to patch out the test for KVM_FEATURE_CLOCKSOURCE_STABLE_BIT but I think that's a GCE issue. -- Peter Jeremy ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
logged errors with current drm stuff
Hi! I'm seeing this with your patch now: error: [drm:pid5:i915_gem_object_unbind] *ERROR* Attempting to unbind pinned buffer Nothing seems to be problematic at the moment; but it does seem to happen quite frequently. Any ideas? -adrian ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: Head not buildin in zfs.c
On 2015-01-21 11:08, Garrett Cooper wrote: On Jan 21, 2015, at 2:05, Willem Jan Withagen w...@digiware.nl wrote: On 2015-01-21 10:51, Benjamin Perrault wrote: On Jan 21, 2015, at 1:34 AM, Willem Jan Withagen w...@digiware.nl wrote: Found this lastnight build. Just fetched the most recent HEAD. Error remains --- zfs/zfs.o --- In file included from /usr/srcs/src11/src/lib/libprocstat/zfs/../zfs.c:39: /usr/srcs/src11/src/lib/libprocstat/zfs/../../../cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h:538:9: error: 'NSEC_TO_TICK' macro redefined [-Werror,-Wmacro-redefined] #define NSEC_TO_TICK(usec) ((usec) / (NANOSEC / hz)) ^ /usr/srcs/src11/src/lib/libprocstat/zfs/../../../sys/cddl/compat/opensolaris/sys/time.h:54:9: note: previous definition is here #define NSEC_TO_TICK(nsec) ((nsec) / (NANOSEC / hz)) ^ /etc/{make,src}.conf are empty I’m seeing the exact same issue. It began for me at r277472 and is still persisting as of r277482 ( the most recent HEAD as of writing this email ). Well, the problem is reported several times during compilation. This is the first warning that also has -Werror, and hence building stops. I have not been able to find where this -Werror all of a sudden pops up. Has to be in any of those commits... It’s a mismatch in the macro definitions, made in the commit below. I’m testing out a fix for it right now. Thanks for the report! r277449 | will | 2015-01-20 14:29:27 -0800 (Tue, 20 Jan 2015) | 2 lines NSEC_TO_TICK(usec) - NSEC_TO_TICK(nsec) Right, picky nasty compilers being fussy. ;) I did in look in the svn log of both .h files, but overlooked what you just found. Need more coffee. I'll be waiting the the fix in the tree.. Thanx, --WjW ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: Head not buildin in zfs.c
On Jan 21, 2015, at 2:17, Willem Jan Withagen w...@digiware.nl wrote: … Right, picky nasty compilers being fussy. ;) I did in look in the svn log of both .h files, but overlooked what you just found. Need more coffee. I'll be waiting the the fix in the tree.. Fixed in r277484 — thanks! signature.asc Description: Message signed with OpenPGP using GPGMail
Re: Head not buildin in zfs.c
On 2015-01-21 10:51, Benjamin Perrault wrote: On Jan 21, 2015, at 1:34 AM, Willem Jan Withagen w...@digiware.nl wrote: Found this lastnight build. Just fetched the most recent HEAD. Error remains --- zfs/zfs.o --- In file included from /usr/srcs/src11/src/lib/libprocstat/zfs/../zfs.c:39: /usr/srcs/src11/src/lib/libprocstat/zfs/../../../cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h:538:9: error: 'NSEC_TO_TICK' macro redefined [-Werror,-Wmacro-redefined] #define NSEC_TO_TICK(usec) ((usec) / (NANOSEC / hz)) ^ /usr/srcs/src11/src/lib/libprocstat/zfs/../../../sys/cddl/compat/opensolaris/sys/time.h:54:9: note: previous definition is here #define NSEC_TO_TICK(nsec) ((nsec) / (NANOSEC / hz)) ^ /etc/{make,src}.conf are empty I’m seeing the exact same issue. It began for me at r277472 and is still persisting as of r277482 ( the most recent HEAD as of writing this email ). Well, the problem is reported several times during compilation. This is the first warning that also has -Werror, and hence building stops. I have not been able to find where this -Werror all of a sudden pops up. Has to be in any of those commits... --WjW ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: Head not buildin in zfs.c
On Jan 21, 2015, at 2:05, Willem Jan Withagen w...@digiware.nl wrote: On 2015-01-21 10:51, Benjamin Perrault wrote: On Jan 21, 2015, at 1:34 AM, Willem Jan Withagen w...@digiware.nl wrote: Found this lastnight build. Just fetched the most recent HEAD. Error remains --- zfs/zfs.o --- In file included from /usr/srcs/src11/src/lib/libprocstat/zfs/../zfs.c:39: /usr/srcs/src11/src/lib/libprocstat/zfs/../../../cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h:538:9: error: 'NSEC_TO_TICK' macro redefined [-Werror,-Wmacro-redefined] #define NSEC_TO_TICK(usec) ((usec) / (NANOSEC / hz)) ^ /usr/srcs/src11/src/lib/libprocstat/zfs/../../../sys/cddl/compat/opensolaris/sys/time.h:54:9: note: previous definition is here #define NSEC_TO_TICK(nsec) ((nsec) / (NANOSEC / hz)) ^ /etc/{make,src}.conf are empty I’m seeing the exact same issue. It began for me at r277472 and is still persisting as of r277482 ( the most recent HEAD as of writing this email ). Well, the problem is reported several times during compilation. This is the first warning that also has -Werror, and hence building stops. I have not been able to find where this -Werror all of a sudden pops up. Has to be in any of those commits... It’s a mismatch in the macro definitions, made in the commit below. I’m testing out a fix for it right now. Thanks for the report! r277449 | will | 2015-01-20 14:29:27 -0800 (Tue, 20 Jan 2015) | 2 lines NSEC_TO_TICK(usec) - NSEC_TO_TICK(nsec) signature.asc Description: Message signed with OpenPGP using GPGMail
Head not buildin in zfs.c
Found this lastnight build. Just fetched the most recent HEAD. Error remains --- zfs/zfs.o --- In file included from /usr/srcs/src11/src/lib/libprocstat/zfs/../zfs.c:39: /usr/srcs/src11/src/lib/libprocstat/zfs/../../../cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h:538:9: error: 'NSEC_TO_TICK' macro redefined [-Werror,-Wmacro-redefined] #define NSEC_TO_TICK(usec) ((usec) / (NANOSEC / hz)) ^ /usr/srcs/src11/src/lib/libprocstat/zfs/../../../sys/cddl/compat/opensolaris/sys/time.h:54:9: note: previous definition is here #define NSEC_TO_TICK(nsec) ((nsec) / (NANOSEC / hz)) ^ /etc/{make,src}.conf are empty --WjW ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: Head not buildin in zfs.c
On Jan 21, 2015, at 1:34 AM, Willem Jan Withagen w...@digiware.nl wrote: Found this lastnight build. Just fetched the most recent HEAD. Error remains --- zfs/zfs.o --- In file included from /usr/srcs/src11/src/lib/libprocstat/zfs/../zfs.c:39: /usr/srcs/src11/src/lib/libprocstat/zfs/../../../cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h:538:9: error: 'NSEC_TO_TICK' macro redefined [-Werror,-Wmacro-redefined] #define NSEC_TO_TICK(usec) ((usec) / (NANOSEC / hz)) ^ /usr/srcs/src11/src/lib/libprocstat/zfs/../../../sys/cddl/compat/opensolaris/sys/time.h:54:9: note: previous definition is here #define NSEC_TO_TICK(nsec) ((nsec) / (NANOSEC / hz)) ^ /etc/{make,src}.conf are empty --WjW ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org I’m seeing the exact same issue. It began for me at r277472 and is still persisting as of r277482 ( the most recent HEAD as of writing this email ). cheers and thanks, -bp ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: Head not buildin in zfs.c
On Wed, Jan 21, 2015 at 02:47:57AM -0800, Garrett Cooper wrote: On Jan 21, 2015, at 2:17, Willem Jan Withagen w...@digiware.nl wrote: … Right, picky nasty compilers being fussy. ;) I did in look in the svn log of both .h files, but overlooked what you just found. Need more coffee. I'll be waiting the the fix in the tree.. Fixed in r277484 — thanks! Thanks! Sorry for the noise. --Will. pgpdmFn4VHM6f.pgp Description: PGP signature
LOR with bufwait (bufwait) @ /usr/src/sys/ufs/ffs/ffs_vnops.c:263
Hoi, Looked at: http://sources.zabbadoz.net/freebsd/lor.html#howtoreportalor But did not find any reference to ffs_vnops, so it might be a new one? I got this from running 'make installworld' of todays HEAD IN a bhyveVM which was already running thismornings kernel. So it was not the Dom0 server, but the actual version running in the VM FreeBSD-11 11.0-CURRENT FreeBSD 11.0-CURRENT #0 r277423: Wed Jan 21 09:38:40 UTC 2015 wjw@FreeBSD-11:/usr/obj/usr/src/sys/GENERIC amd64 --WjW lock order reversal: 1st 0xf80009aa2240 ufs (ufs) @ /usr/src/sys/kern/vfs_subr.c:2164 2nd 0xfe01ea244770 bufwait (bufwait) @ /usr/src/sys/ufs/ffs/ffs_vnops.c:263 3rd 0xf801685dc418 ufs (ufs) @ /usr/src/sys/kern/vfs_subr.c:2164 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfe02332ecd60 witness_checkorder() at witness_checkorder+0xe50/frame 0xfe02332ecdf0 __lockmgr_args() at __lockmgr_args+0xacf/frame 0xfe02332ecf20 ffs_lock() at ffs_lock+0x84/frame 0xfe02332ecf70 VOP_LOCK1_APV() at VOP_LOCK1_APV+0xfc/frame 0xfe02332ecfa0 _vn_lock() at _vn_lock+0x8a/frame 0xfe02332ed010 vget() at vget+0x67/frame 0xfe02332ed050 vfs_hash_get() at vfs_hash_get+0xdc/frame 0xfe02332ed0a0 ffs_vgetf() at ffs_vgetf+0x40/frame 0xfe02332ed130 softdep_sync_buf() at softdep_sync_buf+0xa90/frame 0xfe02332ed210 ffs_syncvnode() at ffs_syncvnode+0x259/frame 0xfe02332ed290 ffs_truncate() at ffs_truncate+0x671/frame 0xfe02332ed470 ufs_direnter() at ufs_direnter+0x7d2/frame 0xfe02332ed530 ufs_makeinode() at ufs_makeinode+0x5bf/frame 0xfe02332ed6f0 ufs_symlink() at ufs_symlink+0x32/frame 0xfe02332ed740 VOP_SYMLINK_APV() at VOP_SYMLINK_APV+0xf7/frame 0xfe02332ed770 kern_symlinkat() at kern_symlinkat+0x245/frame 0xfe02332ed9a0 amd64_syscall() at amd64_syscall+0x25a/frame 0xfe02332edab0 Xfast_syscall() at Xfast_syscall+0xfb/frame 0xfe02332edab0 --- syscall (57, FreeBSD ELF64, sys_symlink), rip = 0x4101ba, rsp = 0x7ffec3d8, rbp = 0x7ffec800 --- ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org