NetBSD-Bugs archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: kern/38707: scheduler related deadlock during build.sh
The following reply was made to PR kern/38707; it has been noted by GNATS.
From: Andrew Doran <andrew%hairylemon.org@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc:
Subject: Re: kern/38707: scheduler related deadlock during build.sh
Date: Fri, 23 May 2008 16:13:35 +0100
Will close the following as a dup.
----- Forwarded message from martin%duskware.de@localhost -----
Subject: kern/38730: recursive lock in scheduler
From: martin%duskware.de@localhost
Date: Thu, 22 May 2008 19:55:00 +0000 (UTC)
To:
kern-bug-people%netbsd.org@localhost,gnats-admin%netbsd.org@localhost,netbsd-bugs%netbsd.org@localhost
Cc:
>Number: 38730
>Category: kern
>Synopsis: recursive lock in scheduler
>Confidential: no
>Severity: critical
>Priority: high
>Responsible: kern-bug-people
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Thu May 22 19:55:00 +0000 2008
>Originator: Martin Husemann
>Release: NetBSD 4.99.63
>Organization:
The NetBSD Foundation, Inc.
>Environment:
System: NetBSD sunny-weather.duskware.de 4.99.63 NetBSD 4.99.63 (SUNNY.MP)
#48: Thu May 22 17:50:52 CEST 2008
martin%sunny-weather.duskware.de@localhost:/usr/src/sys/arch/sparc64/compile/SUNNY.MP
sparc64
Architecture: sparc64
Machine: sparc64
>Description:
Mutex error: lockdebug_wantlock: locking against myself
lock address : 0x000000000d047e80 type : spin
shared holds : 0 exclusive: 1
shares wanted: 0 exclusive: 1
current cpu : 1 last held: 1
current lwp : 0x000000000f89a120 last held: 0x000000000f89a120
last locked : 0x0000000001160bf8 unlocked : 0x0000000001160610
initialized : 0x000000000114d744
owner field : 0x00ff0a0000000000 wait/spin: 0/1
panic: LOCKDEBUG
Stopped in pid 1022.1 (cc1) at netbsd:cpu_Debugger+0x4: nop
db{1}> bt
lockdebug_abort1(d803900, a, 127c598, 12c2480, 1, 20) at
netbsd:lockdebug_abort1
+0x7c
mutex_vector_enter(d047e80, 1a, d047d80, 12c1a18, de98000, 1814000) at
netbsd:mu
tex_vector_enter+0x230
sched_nextlwp(d812040, 1, 127c400, f89a120, de98000, f89a170) at
netbsd:sched_ne
xtlwp+0x288
mi_switch(f89a120, 12c3d70, 14f, 12c1c48, deb40c8, ffffffffffffffff) at
netbsd:m
i_switch+0x40c
preempt(f89a120, 17e798, 31dedb1234, 8000000000000000, 31dedb1234, 4ce800) at
ne
tbsd:preempt+0xa0
trap(f795ed0, fffffffffffffffe, 17e798, 44820092, 121a9a0, ffffffffffff9d68)
at n
etbsd:trap+0x8b4
?(1, 80, 40c12600, c, 20, 5fdf30) at 0x10090b8
db{1}> mach cpu 0
db{0}> bt
mi_switch(f9cac80, 12c3d70, 14f, 12c1c48, 1455980, 42c54a10) at
netbsd:mi_switch
+0x308
preempt(f9cac80, 3d566c, fffffffffffffff8, 0, 42c54a10, 0) at
netbsd:preempt+0xa
0
trap(f7bfed0, fffffffffffffffe, 3d566c, 820092, 121aa00, 0) at
netbsd:trap+0x8b4
?(42c00980, 42b93800, 18, 0, 30, 2) at 0x10090b8
db{0}> x/i 0x000000000114d744
netbsd:mutex_obj_alloc+0x24: call netbsd:mutex_init
db{0}> x/i 0x0000000001160bf8
netbsd:mi_switch+0x758: call netbsd:mutex_vector_enter
db{0}> x/i 0x0000000001160610
netbsd:mi_switch+0x170: call netbsd:mutex_vector_exit
0x114d744 is in mutex_obj_alloc (../../../../kern/kern_mutex.c:970).
965 mutex_obj_alloc(kmutex_type_t type, int ipl)
966 {
967 struct kmutexobj *mo;
968
969 mo = pool_cache_get(mutex_obj_cache, PR_WAITOK);
970 mutex_init(&mo->mo_lock, type, ipl);
971 mo->mo_refcnt = 1;
972
973 return (kmutex_t *)mo;
974 }
0x1160bf8 is in mi_switch (../../../../sys/lwp.h:407).
402 KASSERT(ci1 != ci2);
403 if (ci1 < ci2) {
404 mutex_spin_enter(spc1->spc_mutex);
405 mutex_spin_enter(spc2->spc_mutex);
406 } else {
407 mutex_spin_enter(spc2->spc_mutex);
408 mutex_spin_enter(spc1->spc_mutex);
409 }
410 }
0x1160610 is in mi_switch (../../../../sys/lwp.h:344).
339 * Unlock an LWP. XXXLKM
340 */
341 static inline void
342 lwp_unlock(lwp_t *l)
343 {
344 mutex_spin_exit(l->l_mutex);
345 }
346
347 static inline void
348 lwp_changepri(lwp_t *l, pri_t pri)
>How-To-Repeat:
No idea - I did a "make -j 4" in a kernel compile directory over NFS
(expecting another bug to show up)
>Fix:
n/a
----- End forwarded message -----
Home |
Main Index |
Thread Index |
Old Index