mm> On Thu, 25 Feb 1999, Romano Giannetti wrote:
>> Where is WCHAN of the stalled process? I've seen some stall too, but
>> one of these was totally network-unrelated, just a pipe, and the stall
>> was on pipe_read (or pipe_write). I can reproduce one with jed, but I
>> do not know if it's a jed bug or some other thing.
mm> My self written server and client stall in do_select. A subproc of
mm> INN ("-localhost XOVER") usually stalls in wait_for_tcp_memory.
I had an Ooops in wait_for_tcp_memory (kernel 2.2.1)
It was during a amanda backup run on the client machine. I didn't pay
much attention to it, because ksymoops reported weird opcodes and i
thougt that perhaps the CPU or the memory is flaky. If i look at the
assembler listing of wait_for_tcp_memory i see that thee opcodes c404
don't appear there only 84c404
608 04d6 83C404 addl $4,%esp
^^^^ here starts the Code: line
609 .L3035:
610 04d9 8B466C movl 108(%esi),%eax
611 04dc 394650 cmpl %eax,80(%esi)
612 04df 0F8C9D00 jl .L3038
612 0000
613 04e5 BB00E0FF movl $-8192,%ebx
613 FF
614 #APP
615 04ea 21E3 andl %esp,%ebx;
616 #NO_APP
Feb 23 14:54:30 loki kernel: Unable to handle kernel paging request at virtual address 041be840
Feb 23 14:54:30 loki kernel: current->tss.cr3 = 0120e000, `r3 = 0120e000
Feb 23 14:54:30 loki kernel: *pde = 00000000
Feb 23 14:54:30 loki kernel: Oops: 0000
Feb 23 14:54:30 loki kernel: CPU: 0
Feb 23 14:54:30 loki kernel: EIP: 0010:[<c016e017>]
Feb 23 14:54:30 loki kernel: EFLAGS: 00010246
Feb 23 14:54:30 loki kernel: eax: 00000000 ebx: 00000000 ecx: c106fa10 edx: c197afa0
Feb 23 14:54:30 loki kernel: esi: c106f980 edi: c106f980 ebp: c4e3fed4 esp: c4e3fe88
Feb 23 14:54:30 loki kernel: ds: 0018 es: 0018 ss: 0018
Feb 23 14:54:30 loki kernel: Process gzip (pid: 9706, process nr: 66, stackpage=c4e3f000)
Feb 23 14:54:30 loki kernel: Stack: 000005b4 c106f980 c4e3fed4 c106f980 00000684 c016e456 c106f980 c106f980
Feb 23 14:54:30 loki kernel: c4e3ff7c 00004000 c008da00 c512f684 00000000 08092090 00001770 00002890
Feb 23 14:54:30 loki kernel: 000005b4 c106fa8c 00000000 ffffffea c0174ea3 c106f980 00000000 c4e3ff7c
Feb 23 14:54:30 loki kernel: Call Trace: [<c016e456>] [<c0174ea3>] [<c017c1ff>] [<c017c170>] [<c015311a>] [<c017c170>] [<c0153331>]
Feb 23 14:54:30 loki kernel: [<c0153298>] [<c012305e>] [<c0108918>]
Feb 23 14:54:30 loki kernel: Code: c4 04 8b 46 6c 39 46 50 0f 8c 9d 00 00 00 bb 00 e0 ff ff 21
>>EIP: c016e017 <wait_for_tcp_memory+1f/d8>
Trace: c016e456 <tcp_do_sendmsg+386/5a4>
Trace: c0174ea3 <tcp_v4_sendmsg+87/138>
Trace: c017c1ff <inet_sendmsg+8f/a4>
Trace: c017c170 <inet_sendmsg+0/a4>
Trace: c015311a <sock_sendmsg+8a/b0>
Trace: c017c170 <inet_sendmsg+0/a4>
Trace: c0153331 <sock_write+99/a4>
Trace: c0153298 <sock_write+0/a4>
Code: c016e017 <wait_for_tcp_memory+1f/d8> 00000000 <_EIP>:
Code: c016e017 <wait_for_tcp_memory+1f/d8> 0: c4 04 8b lesl (%ebx,%ecx,4),%eax
Code: c016e01a <wait_for_tcp_memory+22/d8> 3: 46 incl %esi
Code: c016e01b <wait_for_tcp_memory+23/d8> 4: 6c insb (%dx),%es:(%edi)
Code: c016e01c <wait_for_tcp_memory+24/d8> 5: 39 46 50 cmpl %eax,0x50(%esi)
Code: c016e01f <wait_for_tcp_memory+27/d8> 8: 0f 8c 9d 00 00 jl ab <_EIP+0xab> c016e0c2 <wait_for_tcp_memory+ca/d8>
Code: c016e024 <wait_for_tcp_memory+2c/d8> d: 00
Code: c016e025 <wait_for_tcp_memory+2d/d8> e: bb 00 e0 ff ff movl $0xffffe000,%ebx
Code: c016e02a <wait_for_tcp_memory+32/d8> 13: 21 00 andl %eax,(%eax)
mm> Matthias
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/