DragonFlyBSD

A workaround has been provided by sjg ... by commenting out knote_remove, the
panic will not occur but small amounts of memory will be leaked.

Could you make vmcore.0 and info.0 readable please?

On 16 October 2010 20:13, Rumko <rumcic@gmail.com> wrote:
> After several detaches and reattaches, the machine paniced with "Fatal trap 12:
> page fault while in kernel mode"
>
> The core dump is available at leaf:~rumko/crash/ums/*.0
>
> #0  _get_mycpu (di=0xc04ff620) at ./machine/thread.h:83
> #1  md_dumpsys (di=0xc04ff620)
> at /usr/src/sys/platform/pc32/i386/dump_machdep.c:263
> #2  0xc01e46cd in dumpsys () at /usr/src/sys/kern/kern_shutdown.c:880
> #3  0xc01e4c8d in boot (howto=260) at /usr/src/sys/kern/kern_shutdown.c:387
> #4  0xc01e4f56 in panic (fmt=0xc0443534 "%s")
> at /usr/src/sys/kern/kern_shutdown.c:786
> #5  0xc040ffcb in trap_fatal (frame=0xea434808, eva=<value optimized out>)
> at /usr/src/sys/platform/pc32/i386/trap.c:1117
> #6  0xc04100d9 in trap_pfault (frame=0xea434808, usermode=0, eva=12)
> at /usr/src/sys/platform/pc32/i386/trap.c:1018
> #7  0xc0410b44 in trap (frame=0xea434808)
> at /usr/src/sys/platform/pc32/i386/trap.c:699
> #8  0xc03fcf17 in calltrap ()
> at /usr/src/sys/platform/pc32/i386/exception.s:785
> #9  0xc01d0854 in knote_remove (klist=0xd4292224, kn=0xea366ea0)
> at /usr/src/sys/kern/kern_event.c:1370
> #10 0xc0312c44 in devfs_detached_filter_detach (kn=0xea366ea0)
> at /usr/src/sys/vfs/devfs/devfs_core.c:2234
> #11 0xc01d0de7 in knote_detach_and_drop (kn=0xea366ea0)
> at /usr/src/sys/kern/kern_event.c:1258
> #12 0xc01d157b in kqueue_register (kq=0xea3965b4, kev=0xea4348b8)
> at /usr/src/sys/kern/kern_event.c:933
> #13 0xc020b2f3 in poll_copyout (arg=0xea434c9c, kevp=0xea4349b4, count=2,
> res=0xea434cf0) at /usr/src/sys/kern/sys_generic.c:1325
> #14 0xc01d20c5 in kern_kevent (kq=0xea3965b4, nevents=2147483647,
> res=0xea434cf0, uap=0xea434c9c, kevent_copyinfn=0xc020b4a5 <poll_copyin>,
>    kevent_copyoutfn=0xc020b290 <poll_copyout>, tsp_in=0xea434cb0)
> at /usr/src/sys/kern/kern_event.c:697
> #15 0xc020b031 in dopoll (uap=0xea434cf0)
> at /usr/src/sys/kern/sys_generic.c:1474
> #16 sys_poll (uap=0xea434cf0) at /usr/src/sys/kern/sys_generic.c:1228
> #17 0xc04113d2 in syscall2 (frame=0xea434d40)
> at /usr/src/sys/platform/pc32/i386/trap.c:1310
> #18 0xc03fcfc6 in Xint0x80_syscall ()
> at /usr/src/sys/platform/pc32/i386/exception.s:876
> #19 0x0000001f in ?? ()
> --
> Please do not CC me, since I already receive everything from these MLs.
>
> Regards,
> Rumko
>

It tries to read at address 0xC which is the offset of kn_next so
it probably crashes while dereferencing kn_next in SLIST_REMOVE().
This could happen if SLIST_REMOVE() reaches the end of the list
without finding the knode in the klist. I can't figure out how this
could happen though.

Could you chmod vmcore.0 so I can analyse the dump please?

On 16 October 2010 20:13, Rumko <rumcic@gmail.com> wrote:
> After several detaches and reattaches, the machine paniced with "Fatal trap 12:
> page fault while in kernel mode"
>
> The core dump is available at leaf:~rumko/crash/ums/*.0
>
> #0  _get_mycpu (di=0xc04ff620) at ./machine/thread.h:83
> #1  md_dumpsys (di=0xc04ff620)
> at /usr/src/sys/platform/pc32/i386/dump_machdep.c:263
> #2  0xc01e46cd in dumpsys () at /usr/src/sys/kern/kern_shutdown.c:880
> #3  0xc01e4c8d in boot (howto=260) at /usr/src/sys/kern/kern_shutdown.c:387
> #4  0xc01e4f56 in panic (fmt=0xc0443534 "%s")
> at /usr/src/sys/kern/kern_shutdown.c:786
> #5  0xc040ffcb in trap_fatal (frame=0xea434808, eva=<value optimized out>)
> at /usr/src/sys/platform/pc32/i386/trap.c:1117
> #6  0xc04100d9 in trap_pfault (frame=0xea434808, usermode=0, eva=12)
> at /usr/src/sys/platform/pc32/i386/trap.c:1018
> #7  0xc0410b44 in trap (frame=0xea434808)
> at /usr/src/sys/platform/pc32/i386/trap.c:699
> #8  0xc03fcf17 in calltrap ()
> at /usr/src/sys/platform/pc32/i386/exception.s:785
> #9  0xc01d0854 in knote_remove (klist=0xd4292224, kn=0xea366ea0)
> at /usr/src/sys/kern/kern_event.c:1370
> #10 0xc0312c44 in devfs_detached_filter_detach (kn=0xea366ea0)
> at /usr/src/sys/vfs/devfs/devfs_core.c:2234
> #11 0xc01d0de7 in knote_detach_and_drop (kn=0xea366ea0)
> at /usr/src/sys/kern/kern_event.c:1258
> #12 0xc01d157b in kqueue_register (kq=0xea3965b4, kev=0xea4348b8)
> at /usr/src/sys/kern/kern_event.c:933
> #13 0xc020b2f3 in poll_copyout (arg=0xea434c9c, kevp=0xea4349b4, count=2,
> res=0xea434cf0) at /usr/src/sys/kern/sys_generic.c:1325
> #14 0xc01d20c5 in kern_kevent (kq=0xea3965b4, nevents=2147483647,
> res=0xea434cf0, uap=0xea434c9c, kevent_copyinfn=0xc020b4a5 <poll_copyin>,
>    kevent_copyoutfn=0xc020b290 <poll_copyout>, tsp_in=0xea434cb0)
> at /usr/src/sys/kern/kern_event.c:697
> #15 0xc020b031 in dopoll (uap=0xea434cf0)
> at /usr/src/sys/kern/sys_generic.c:1474
> #16 sys_poll (uap=0xea434cf0) at /usr/src/sys/kern/sys_generic.c:1228
> #17 0xc04113d2 in syscall2 (frame=0xea434d40)
> at /usr/src/sys/platform/pc32/i386/trap.c:1310
> #18 0xc03fcfc6 in Xint0x80_syscall ()
> at /usr/src/sys/platform/pc32/i386/exception.s:876
> #19 0x0000001f in ?? ()
> --
> Please do not CC me, since I already receive everything from these MLs.
>
> Regards,
> Rumko
>

Done