On Wednesday 04 February 2004 03:26 am, Poul-Henning Kamp wrote: > It seems somewhat bogus to me that I can create and run a kthread > without Giant but I need to grab Giant before it exits... > > panic: mutex Giant not owned at ../../../kern/kern_exit.c:123 > at line 719 in file ../../../kern/kern_mutex.c > cpuid = 0; > Debugger("panic") > Stopped at 0xc06a7cf5 = Debugger+0x55: xchgl %ebx,0xc07bab24 = > in_Debugger.0 db> trace > Debugger(c0712b15,0,2cf,c0711f91,100) at 0xc06a7cf5 = Debugger+0x55 > __panic(c0711f91,2cf,c07120f9,c0712234,c0710532) at 0xc0532782 = > __panic+0x172 _mtx_assert(c076b460,1,c0710532,7b,c05398e3) at 0xc0528dec = > _mtx_assert+0x11c exit1(c18ebd20,0,88,87,0) at 0xc0519e43 = exit1+0x33 > kthread_exit(0,0,c070d581,29e,64) at 0xc0520126 = kthread_exit+0xb6 > g_bde_worker(c1a0fc80,c54b3d48,c0710609,317,c194adc0) at 0xc04f6db5 = > g_bde_worker+0x4a5 fork_exit(c04f6910,c1a0fc80,c54b3d48) at 0xc051d065 = > fork_exit+0xc5 fork_trampoline() at 0xc06a94bc = fork_trampoline+0x8 > --- trap 0x1, eip = 0, esp = 0xc54b3d7c, ebp = 0 --- Because kthread_exit() isn't safe. :) However, it can probably just grab Giant itself. Peter wants me to look at pushing Giant down into fork/wait/ exit as much as I can, so I'll fix kthread_exit() to get Giant if needed as part of that. -- John Baldwin <jhb_at_FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve" = http://www.FreeBSD.orgReceived on Wed Feb 04 2004 - 09:45:45 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:41 UTC