sbuf-related panic

Ian Lepore ian at freebsd.org
Mon Mar 16 18:52:53 UTC 2015


On Mon, 2015-03-16 at 14:03 -0400, Shawn Webb wrote:
> On amd64, doing a Poudriere run. On r280133:
> 
> (kgdb) bt
> #0  doadump (textdump=Unhandled dwarf expression opcode 0x93
> ) at pcpu.h:219
> #1  0xffffffff809726a5 in kern_reboot (howto=Unhandled dwarf expression opcode 0x93
> ) at /usr/src/sys/kern/kern_shutdown.c:448
> #2  0xffffffff80972c98 in vpanic (fmt=<value optimized out>, ap=<value optimized out>) at /usr/src/sys/kern/kern_shutdown.c:747
> #3  0xffffffff80972ac2 in kassert_panic (fmt=<value optimized out>) at /usr/src/sys/kern/kern_shutdown.c:635
> #4  0xffffffff809bd5bd in sbuf_delete (s=0xfffffe085d468750) at /usr/src/sys/kern/subr_sbuf.c:103
> #5  0xffffffff809669dd in sysctl_kern_proc_args (oidp=<value optimized out>, arg1=<value optimized out>, arg2=<value optimized out>, req=0xfffffe085d468868) at /usr/src/sys/kern/kern_proc.c:1858
> #6  0xffffffff8097e7b4 in sysctl_root_handler_locked (oid=0xffffffff81533738, arg1=0xfffffe085d46893c, arg2=1, req=0xfffffe085d468868) at /usr/src/sys/kern/kern_sysctl.c:183
> #7  0xffffffff8097e038 in sysctl_root (arg1=<value optimized out>, arg2=<value optimized out>) at /usr/src/sys/kern/kern_sysctl.c:1660
> #8  0xffffffff8097e5b2 in userland_sysctl (td=<value optimized out>, name=0xfffffe085d468930, namelen=<value optimized out>, old=<value optimized out>, oldlenp=<value optimized out>, inkernel=<value optimized out>, 
>     new=<value optimized out>, newlen=<value optimized out>, retval=<value optimized out>, flags=0) at /usr/src/sys/kern/kern_sysctl.c:1764
> #9  0xffffffff8097e3e4 in sys___sysctl (td=0xfffff802521324b0, uap=0xfffffe085d468a40) at /usr/src/sys/kern/kern_sysctl.c:1690
> #10 0xffffffff80da1bca in amd64_syscall (td=0xfffff802521324b0, traced=0) at subr_syscall.c:133
> #11 0xffffffff80d7f3cb in Xfast_syscall () at /usr/src/sys/amd64/amd64/exception.S:395
> #12 0x000002759e502b7a in ?? ()

That one should be fixed in r280149, sorry about that.  For some reason
I never hit the triggering condition in testing, but it seems that just
about everyone else in the world did. :/

There are a couple locking-related fixes still needed which might
trigger a witness warning, they should be fixed soon as well.

-- Ian




More information about the freebsd-current mailing list