Yesterday our SMP machine crashed. It's running 2.1.103 on
a Compaq Proliant 850R with two PPro 200 CPUs and 128Mb RAM.
Disc controller is a NCR 875 and it's using the internal
Thunderlan ethernet controller. All drivers are compiled
into the kernel.
Any ideas what happened? Is this a IRQ deadlock?
Below is the relevant information that we could get.
This was on the console, printed periodically:
wait_on_bh, CPU1:
IRQ: 0 [0 0]
BH: 1 [1 0]
<[c016b8d6]> <[c01660d7]> <[c0166378]> <[c0172cb2]> <[c0155dff]>
<[c01561ff]> <[c01287c0]> <[c012883b]>
Sysrq-P:
EIP: 0010:[<c01147c8>] EFLAGS: 00000283
EAX: c35cfef8 EBX: c4707540 ECX: c01d95c4 EDX: c4707790
ESI: c0160748 EDI: 00000009 EBP: c0093f90 DS: 0018 ES: 0018
Sysrq-M:
Mem-info:
Free Pages: 4356kB
( 353*4kB 112*8kB 86*16kB 15*32kB 1*64kB 1*128kB = 4356kB)
Swap cache: add 161856/161856, delete 161571/161571, find 0/0
Free swap: 118996kB
32512 pages of RAM
743 reserved pages
3873 pages shared
285 pages swap cached
Buffer memory: 7356kB
Buffer heads: 7392
Buffer blocks: 7356
CLEAN: 2433 buffers, 35 used (last=36), 0 locked, 0 protected, 0 dirty
LOCKED: 4917 buffers, 86 used (last=4904), 0 locked, 0 protected, 0
dirty
DIRTY: 3 buffers, 0 used (last=0), 0 locked, 0 protected, 3 dirty
Networking buffers in use : 407
Total network buffer allocations : 48587401
Total failed network buffer allocs : 0
IP Fragment buffer size : 0
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu