Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 23 Dec 2004 02:48:22 -0500
From:      Lucas Madar <lucas@negaverse.org>
To:        freebsd-stable@freebsd.org
Subject:   panic in 5.3-stable related to heavy usage and libthr
Message-ID:  <41CA7846.2020604@negaverse.org>

next in thread | raw e-mail | index | archive | help
I updated a machine today to 5.3-STABLE and almost instantly it crashed 
when a heavily threaded program ran. (The machine was previously running 
5.2-release).

This crash is reproducible under heavy load with four threads using 
libthr -- I haven't tested any other variances on this scheme, and I 
will be testing 5.3-RELEASE tomorrow to see if it exhibits the same 
problems.

relevant info from kgdb, with null pointer goodness.

#23 0x00000000 in ?? ()
#24 0xc055681f in sleepq_remove_thread (sq=0x0, td=0x0) at 
/data/rootsystem/src/sys/kern/subr_sleepqueue.c:581
#25 0xc0556a16 in sleepq_signal (wchan=0xc2b91320, flags=0, pri=-1) at 
/data/rootsystem/src/sys/kern/subr_sleepqueue.c:677
#26 0xc053fa53 in wakeup_one (ident=0xc2b91320) at 
/data/rootsystem/src/sys/kern/kern_synch.c:266
#27 0xc0543052 in thr_wake (td=0xc2b91000, uap=0xc2b91000) at 
/data/rootsystem/src/sys/kern/kern_thr.c:303
#28 0xc0698c9f in syscall (frame=
      {tf_fs = 134742063, tf_es = 134742063, tf_ds = -1078001617, tf_edi 
= 134558784, tf_esi = 135229440, tf_ebp = -1079198052, tf_isp = 
-378659468, tf_ebx = 671863452, tf_edx = 134558792, tf_ecx = 0, tf_eax = 
443, tf_trapno = 22, tf_err = 2, tf_eip = 672179195, tf_cs = 31, 
tf_eflags = 662, tf_esp = -1079198096, tf_ss = 47}) at 
/data/rootsystem/src/sys/i386/i386/trap.c:1001
#29 0xc0686c1f in Xint0x80_syscall () at 
/data/rootsystem/src/sys/i386/i386/exception.s:201

Is this useful to anyone? I'll keep this around if you need to poke at 
any structures.

- lucas



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?41CA7846.2020604>