On Tue, Mar 11, 2008 at 03:39:04PM -1000, Jeff Roberson wrote: >.... > I believe this is a cpuset/kse incompatibility. Can you try this patch: > > http://people.freebsd.org/~jeff/cpusetkse.diff I was able to verify that if I tried to run firefox on my build machine, it got a panic: Fatal trap 12: page fault while in kernel mode cpuid = 1; apic id = 00 fault virtual address = 0x4 fault code = supervisor write, page not present instruction pointer = 0x20:0xc07400f7 stack pointer = 0x28:0xe7b07b60 frame pointer = 0x28:0xe7b07b7c code segment = base 0x0, limit 0xfffff, type 0x1b = DPL 0, pres 1, def32 1, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 880 (firefox-bin) [thread pid 880 tid 100122 ] Stopped at cpuset_rel+0x17: lock xaddl %eax,0x4(%esi) db> bt Tracing pid 880 tid 100122 td 0xc5ad9d20 cpuset_rel(0,c0af3c0d,f4,c50f6d20,e7b07bb4,...) at cpuset_rel+0x17 thread_free(c50f6d20,0,c0af5fd4,131,c5ad9d20,...) at thread_free+0x15 thread_reap(c0af1e43,e7b07be4,c0762cdc,c5ad9d20,e7b07be4,...) at thread_reap+0xc0 thread_alloc(ca,c5a96000,c5a96000,2,e7b07c80,...) at thread_alloc+0xc thread_alloc_spare(c5a96090,0,c0af1e43,289,4,...) at thread_alloc_spare+0x21 kse_create(c5ad9d20,e7b07cfc,8,c0afaace,c0badfd8,...) at kse_create+0x1c2 syscall(e7b07d38) at syscall+0x2b3 Xint0x80_syscall() at Xint0x80_syscall+0x20 --- syscall (381, FreeBSD ELF32, kse_create), eip = 0x28bd64db, esp = 0xbfbfcf0c, ebp = 0xbfbfcf48 --- db> And it's obvious even to me that KSE was involved. :-} I applied the cited patch & re-built the kernel; after rebooting with the new kernel, I am unable to reproduce the panic: firefox just runs normally. I believe the patch is an improvement. :-) Peace, david -- David H. Wolfskill david_at_catwhisker.org I submit that "conspiracy" would be an appropriate collective noun for cats. See http://www.catwhisker.org/~david/publickey.gpg for my public key.
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:28 UTC