[ofa-general] Re: ipoib attempting to join on junk MGID / 2.6.21-rc6 crash dump

[Or Gerlitz]

OK, I did some checks with upstream kernel, the junk mkey for child interface phenomena
does not reproduce, which probably means its either ofed or RH4 kernel issue.

However, I started on 2.6.21-rc6 under which i saw the below crash, which
does not reproduce now under 2.6.22, was there any fix that you are aware
to around this area of the code?

Or.


ib0.8007: bringing up interface
ib0.8007: IPOIB_FLAG_OPER_UP not set<7>ib0.8007: IPOIB_FLAG_OPER_UP not set<6>ADDRCONF(NETDEV_UP): ib0.8007: link is not ready
ib0.8007: IPOIB_FLAG_OPER_UP not set<7>ib0.8007: IPOIB_FLAG_OPER_UP not set<7>ib0.8007: stopping interface
ib0.8007: downing ib_dev
ib0.8007: stopping multicast thread
ib0.8007: flushing multicast list
Unable to handle kernel NULL pointer dereference at 0000000000000070 RIP:
 [<ffffffff8045d84e>] _spin_lock_irqsave+0x3/0x24
PGD 36305067 PUD 3b8fb067 PMD 0
Oops: 0002 [1] SMP
CPU 1
Modules linked in: ib_ipoib ib_cm ib_sa ipv6 ib_mthca ib_mad ib_core sg st sd_mod sr_mod scsi_mod e100 i2c_amd8111 i2c_amd756 i2c_core
Pid: 12633, comm: ifconfig Not tainted 2.6.21-rc6 #2
RIP: 0010:[<ffffffff8045d84e>]  [<ffffffff8045d84e>] _spin_lock_irqsave+0x3/0x24
RSP: 0018:ffff810026dcbc50  EFLAGS: 00010092
RAX: 0000000000000292 RBX: ffff810016425000 RCX: ffff810016425750
RDX: ffff810026dcbd48 RSI: 0000000000000000 RDI: 0000000000000070
RBP: 0000000000000000 R08: 00000000ffffffff R09: 0000000000000000
R10: ffff810000e6b2c0 R11: 0000000000000001 R12: 0000000000000070
R13: 0000000000000000 R14: ffff81003f8c6c00 R15: ffff810016425000
FS:  00002abdfeb7b740(0000) GS:ffff81003f8a7a40(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000070 CR3: 000000003897a000 CR4: 00000000000006e0
Process ifconfig (pid: 12633, threadinfo ffff810026dca000, task ffff81003d7537f0)
Stack:  ffffffff880e48f7 0000003000000010 ffff810026dcbd38 ffff810026dcbc78
 ffff81003f8c6c00 ffff810016425000 ffff810016425000 ffff810026dcbd30
 ffff810016425000 ffff810016425700 ffff810016425700 ffff810016425000
Call Trace:
 [<ffffffff880e48f7>] :ib_cm:cm_destroy_id+0x1c/0x25c
 [<ffffffff880f28a3>] :ib_ipoib:ipoib_cm_dev_stop+0x27/0xc5
 [<ffffffff880ed53c>] :ib_ipoib:ipoib_ib_dev_stop+0x25/0x2c3
 [<ffffffff8023f504>] flush_cpu_workqueue+0xb3/0xc1
 [<ffffffff802426cc>] autoremove_wake_function+0x0/0x2e
 [<ffffffff80238f8e>] lock_timer_base+0x1b/0x3c
 [<ffffffff880eea8c>] :ib_ipoib:ipoib_mcast_dev_flush+0x10e/0x159
 [<ffffffff880ece93>] :ib_ipoib:ipoib_flush_paths+0x34/0x15a
 [<ffffffff880eb927>] :ib_ipoib:ipoib_stop+0x63/0xef
 [<ffffffff8040d237>] dev_close+0x58/0x77
 [<ffffffff8040c2f6>] dev_change_flags+0x57/0x119
 [<ffffffff80443f24>] devinet_ioctl+0x265/0x5cd
 [<ffffffff80444f8e>] inet_ioctl+0x3f/0x5e
 [<ffffffff80402cae>] sock_ioctl+0x16c/0x189
 [<ffffffff80284ecd>] do_ioctl+0x29/0x6f
 [<ffffffff80285187>] vfs_ioctl+0x274/0x285
 [<ffffffff802851d4>] sys_ioctl+0x3c/0x60
 [<ffffffff802093ce>] system_call+0x7e/0x83


Code: f0 ff 0f 79 1b a9 00 02 00 00 74 0b fb f3 90 83 3f 00 7e f9
RIP  [<ffffffff8045d84e>] _spin_lock_irqsave+0x3/0x24
 RSP <ffff810026dcbc50>
CR2: 0000000000000070