We seem to be hitting something similar, running 2.6.39.2. Did anyone make any progress on this? (I'm happy to try and gather more info but I probably won't be able to seriously debug until next week) Anyway, we have a system with two mpt2sas adapters that have 4 paths to a JBOD, and we're using dm-multipath to the drives in the JBOD. Killing one of the drives (ie yanking the drive, so all 4 paths go down at once) leads nearly instantly to: [ 768.999560] device-mapper: multipath: Failing path 8:48. [ 769.005151] device-mapper: multipath: Failing path 8:48. [ 769.010919] device-mapper: table: 252:4: multipath: error getting device [ 769.017708] device-mapper: ioctl: error adding target to table [ 769.023696] device-mapper: multipath: Failing path 8:48. [ 769.030979] BUG: unable to handle kernel paging request at 0000000200000000 [ 769.038119] IP: [<0000000200000000>] 0x1ffffffff [ 769.042859] PGD 0 [ 769.045264] Oops: 0010 [#1] SMP [ 769.048671] last sysfs file: /sys/devices/pci0000:00/0000:00:07.0/0000:09:00.0/host11/port-11:0/expander-11:0/port-11:0:1/end_device-11:0:1/target11:0:1/11:0:1:0/block/sdd/uevent [ 769.064690] CPU 6 [ 769.066835] Modules linked in: target_core_pscsi target_core_file target_core_iblock tcm_loop target_core_mod configfs ps_bdrv ipmi_devintf ipmi_si ipmi_msghandler serio_raw ioatdma i7core_edac dca edac_core ses enclosure usb_storage mpt2sas qla2xxx usbhid scsi_transport_sas ahci uas scsi_transport_fc libahci e1000e hid mlx4_core scsi_tgt raid_class [last unloaded: evbug] [ 769.102323] [ 769.104141] Pid: 30, comm: kworker/6:0 Not tainted 2.6.39.2+ #1 Xyratex Storage Server /HS-1235T-ATX [ 769.116224] RIP: 0010:[<0000000200000000>] [<0000000200000000>] 0x1ffffffff [ 769.123679] RSP: 0018:ffff880c1b1a1cc8 EFLAGS: 00010082 [ 769.129317] RAX: ffff8806165cc780 RBX: ffff880614cd06c0 RCX: 0000000000000000 [ 769.136808] RDX: 0000000000000002 RSI: 0000000000000001 RDI: ffff880614cd06c0 [ 769.144298] RBP: ffff880c1b1a1ce0 R08: 0000000000000000 R09: 0000000000000001 [ 769.151789] R10: ffff880c0ecd2af0 R11: ffff880619b40400 R12: ffff880614cd06c0 [ 769.159279] R13: 0000000000000002 R14: 0000000000000002 R15: ffff880614cd4010 [ 769.166770] FS: 0000000000000000(0000) GS:ffff880c3fc00000(0000) knlGS:0000000000000000 [ 769.175210] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 769.181284] CR2: 0000000200000000 CR3: 0000000001a03000 CR4: 00000000000006e0 [ 769.188772] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 769.196254] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 769.203746] Process kworker/6:0 (pid: 30, threadinfo ffff880c1b1a0000, task ffff880c1b1996b0) [ 769.212617] Stack: [ 769.214694] ffffffff8123290a ffff880c19d54338 ffff880c19d54338 ffff880c1b1a1d10 [ 769.222669] ffffffff81232a63 ffff880c19d54338 ffff880614cd06c0 0000000000000002 [ 769.230636] ffffc900189b7040 ffff880c1b1a1d40 ffffffff812357ad ffff880c1b1a1d60 [ 769.238600] Call Trace: [ 769.241118] [<ffffffff8123290a>] ? elv_drain_elevator+0x2a/0x80 [ 769.247452] [<ffffffff81232a63>] __elv_add_request+0x103/0x280 [ 769.253697] [<ffffffff812357ad>] add_acct_request+0x3d/0x50 [ 769.259680] [<ffffffff81235825>] blk_insert_cloned_request+0x65/0x90 [ 769.266449] [<ffffffff813b332e>] dm_dispatch_request+0x3e/0x70 [ 769.272692] [<ffffffff813b4e3b>] dm_request_fn+0x16b/0x240 [ 769.278588] [<ffffffff81237300>] ? perf_trace_block_unplug+0xe0/0xe0 [ 769.285353] [<ffffffff81237340>] blk_delay_work+0x40/0x60 [ 769.291218] [<ffffffff8106954a>] process_one_work+0x11a/0x420 [ 769.297375] [<ffffffff8106a5a3>] worker_thread+0x163/0x340 [ 769.303274] [<ffffffff8106a440>] ? manage_workers.clone.21+0x240/0x240 [ 769.310212] [<ffffffff8106f2d6>] kthread+0x96/0xa0 [ 769.315421] [<ffffffff814de1a4>] kernel_thread_helper+0x4/0x10 [ 769.321662] [<ffffffff8106f240>] ? kthread_worker_fn+0x190/0x190 [ 769.328079] [<ffffffff814de1a0>] ? gs_change+0x13/0x13 [ 769.333632] Code: Bad RIP value. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
