IPI-related panic on an SMP machine
From: Thierry Herbelot (thierry_at_herbelot.com)
Date: 06/30/05
- Previous message: John Baldwin: "Re: Timekeeping hosed by factor 3, high lapic[01] interrupt rates"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
To: current@freebsd.org Date: Thu, 30 Jun 2005 20:47:07 +0200
Hello,
When I installed my latest world, I have had a panic (see following log)
The machine was running single-user with a GENERIC kernel :
FreeBSD 6.0-CURRENT #702: Thu Jun 30 11:21:34 CEST 2005
XXX@YYY:/usr/obj_ini/usr/src/sys/GENERIC
What is disappointing is that I could not repeat the panic :
I have installed the world, while in single-user mode with the same "panicky"
kernel, first on a standard partition (/dev/ad9s2d), then on /usr which is a
geom_mirror partition (/dev/mirror/gm0s1f), and in both cases, the install
went to its correct end.
I could not save a memory dump as I was in single-user mode and the dump
device was not set (is it possible to set the dumpdev from the ddb prompt ?)
Cheers
TfH
PS : the machine is an ABIT BP6, running a straight GENERIC kernel
PS2 : the sources have been synced around june 29 at 20h GMT
PS2 : panic log
[make installworld in single-user mode]
....
install -s -o root -g wheel -m 555 mount_portalfs /usr/sbin
panic: APIC: Previous IPI is stuck
cpuid = 1
KDB: enter: panic
[thread pid 16777 tid 100095 ]
Stopped at kdb_enter+0x2b: nop
db> ecing pid 16777 tid 100095 td 0xc1878d80
kdb_enter(c084eb14) at kdb_enter+0x2b
panic(c0872758,1,bfbfe000,cc921a50,c07e078d) at panic+0x127
lapic_ipi_vectored(f4,fffffffd,cc921a60,c07e0485,f4) at
lapic_ipi_vectored+0x93
ipi_all_but_self(f4) at ipi_all_but_self+0xd
smp_tlb_shootdown(f4,bfbfe000,0) at smp_tlb_shootdown+0x49
smp_invlpg(bfbfe000) at smp_invlpg+0x1b
pmap_invalidate_page(c186e8f4,bfbfe000) at pmap_invalidate_page+0x6c
pmap_enter(c186e8f4,bfbfe000,c1114c80,7,0,c1c9f294,0,c086971a,37e) at
pmap_enter+0x276
vm_fault(c186e834,bfbfe000,2,8,c1878d80) at vm_fault+0x10b8
trap_pfault(cc921c18,0,bfbfe900) at trap_pfault+0xee
trap(c1510008,c1870028,cc920028,bfbfe900,cc921c74) at trap+0x33d
calltrap() at calltrap+0x5
--- trap 0xc, eip = 0xc07e608c, esp = 0xcc921c58, ebp = 0xcc921cdc ---
slow_copyout(c1878d80,cc921d04,2,4,296) at slow_copyout+0x4
syscall(3b,3b,3b,0,2) at syscall+0x22f
Xint0x80_syscall() at Xint0x80_syscall+0x1f
--- syscall (189, FreeBSD ELF32, fstat), eip = 0x804a6a7, esp = 0xbfbfdccc,
ebp = 0xbfbfea38 ---
db>
_______________________________________________
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscribe@freebsd.org"
- Previous message: John Baldwin: "Re: Timekeeping hosed by factor 3, high lapic[01] interrupt rates"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|
