[openib-general] oops removing rdma_cm

[Michael S. Tsirkin]

Sean, I see this when I remove rdma_cm:


Mar 20 23:08:01 swlab155 kernel: general protection fault: 0000 [1] SMP
Mar 20 23:08:01 swlab155 kernel: CPU 0
Mar 20 23:08:01 swlab155 kernel: Modules linked in: ib_cm ib_local_sa findex
ib_addr ib_ipoib ib_sa ib_umad ib_mthca ib_mad ib_core
Mar 20 23:08:01 swlab155 kernel: Pid: 9597, comm: rmmod Not tainted 2.6.15 #8
Mar 20 23:08:01 swlab155 kernel: RIP: 0010:[<ffffffff801508f5>]
<ffffffff801508f5>{module_unload_free+130}
Mar 20 23:08:01 swlab155 kernel: RSP: 0018:ffff8101760d7ea8  EFLAGS: 00010293
Mar 20 23:08:01 swlab155 kernel: RAX: ffff81017ec65a20 RBX: ffff810179986900
RCX: 000000000000003c
Mar 20 23:08:01 swlab155 kernel: RDX: 9b08040b8804c700 RSI: ffff81000000f000
RDI: ffff810179986900
Mar 20 23:08:01 swlab155 kernel: RBP: ffffffff8804b580 R08: 0000000000000003
R09: 000000d000000000
Mar 20 23:08:01 swlab155 kernel: R10: 0000000000000000 R11: 0000000000000000
R12: ffffffff88062200
Mar 20 23:08:01 swlab155 kernel: R13: 0000000000000000 R14: 0000000000000880
R15: 00007fffffbd5d28
Mar 20 23:08:01 swlab155 kernel: FS:  0000000000000000(0000)
GS:ffffffff805e3800(0000) knlGS:0000000000000000
Mar 20 23:08:01 swlab155 kernel: CS:  0010 DS: 0000 ES: 0000 CR0:
000000008005003b
Mar 20 23:08:01 swlab155 kernel: CR2: 00002aaaaac29cc0 CR3: 0000000177aa1000
CR4: 00000000000006e0
Mar 20 23:08:01 swlab155 kernel: Process rmmod (pid: 9597, threadinfo
ffff8101760d6000, task ffff81017ddc10c0)
Mar 20 23:08:01 swlab155 kernel: Stack: ffffffff88062250 ffffffff88062200
0000000000000002 ffffffff801520a9
Mar 20 23:08:01 swlab155 kernel:        ffffffff804834a0 ffffffff88062200
ffffffff804834a0 ffffffff801525fc
Mar 20 23:08:01 swlab155 kernel:        ffffffffffffffff 000000007d15a7b8
Mar 20 23:08:01 swlab155 kernel: Call Trace:<ffffffff801520a9>{free_module+174}
<ffffffff801525fc>{sys_delete_module+519}
Mar 20 23:08:01 swlab155 kernel:        <ffffffff80110665>{error_exit+0}
<ffffffff8010f92a>{system_call+126}

Any idea why could that be?


-- 
Michael S. Tsirkin
Staff Engineer, Mellanox Technologies