On Friday 03 March 2006 16:03, Kris Kennaway wrote: > On Fri, Mar 03, 2006 at 11:14:47AM -0500, John Baldwin wrote: > > On Thursday 02 March 2006 18:59, Kris Kennaway wrote: > > > Updated to -current, panicked at boot: > > > > > > Timecounters tick evpanic: blockable sleep lock (sleep mutex) Giant _at_ kern/kern_intr.c:661 > > > cpuid = 0 > > > KDB: enter: panic > > > [thread pid 30 tid 100035 ] > > > Stopped at kdb_enter+0x30: leave > > > db> wh > > > Tracing pid 30 tid 100035 td 0xcc5b7bd0 > > > kdb_enter(c072c0f6,0,c07303fc,f5893c50,cc5b7bd0) at kdb_enter+0x30 > > > panic(c07303fc,c074362d,c073e0e1,c07294ca,295) at panic+0x13f > > > witness_checkorder(c07939e8,9,c07294ca,295,c07999f8) at witness_checkorder+0xd6 > > > _mtx_lock_flags(c07939e8,0,c07294c1,295,2) at _mtx_lock_flags+0x94 > > > ithread_execute_handlers(cc5c1000,cc5f8c80,c052b648,c07939d0,cc5b7bd0) at ithread_execute_handlers+0x10c > > > ithread_loop(cc5d3440,f5893d38,c07292b7,31a,cc5d3440) at ithread_loop+0x78 > > > fork_exit(c051d959,cc5d3440,f5893d38) at fork_exit+0xc5 > > > fork_trampoline() at fork_trampoline+0x8 > > > --- trap 0x1, eip = 0, esp = 0xf5893d6c, ebp = 0 --- > > > > Something has leaked a critical section. Do you have witness_skipspin enabled? > > Can you try turning it off? It's probably a missing unlock of a spin mutex > > somewhere. > > WITNESS_SKIPSPIN was not enabled. Well, the panic means td_critnest > 1, so a critical section leaked somewhere. The box must have been really slow if you had witness checking spin mutexes. -- John Baldwin <jhb_at_FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve" = http://www.FreeBSD.orgReceived on Mon Mar 06 2006 - 14:41:39 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:53 UTC