hung task backtraces and/or panic triggered by elxhbamgrd

Issue

• We see the following hung task backtrace and/or panic triggered

INFO: task elxhbamgrd:10352 blocked for more than 30 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
elxhbamgrd    D ffff8806787c3b78  5472 10352  10017 0x00000080
 ffff8806787c3a78 0000000000000046 ffff880678016318 0000000000000000
 ffff880678015980 ffff880600000000 ffff8806787c3a28 0000000000000046
 ffff880678015f40 ffff8806787c3fd8 000000000000fd28 ffff880678015f40
Call Trace:
 [<ffffffff8153f7b5>] schedule_timeout+0x265/0x330
 [<ffffffff8153f3db>] wait_for_common+0x12b/0x180
 [<ffffffff8153f4ed>] wait_for_completion+0x1d/0x20
 [<ffffffff81284729>] blk_execute_rq+0x99/0x100
 [<ffffffff8128aa0a>] bsg_ioctl+0x1ea/0x250
 [<ffffffff811b17f2>] vfs_ioctl+0x22/0xa0
 [<ffffffff811b1994>] do_vfs_ioctl+0x84/0x590
 [<ffffffff811b1f21>] sys_ioctl+0x81/0xa0
 [<ffffffff8100b072>] system_call_fastpath+0x16/0x1b
no locks held by elxhbamgrd/10352.
Kernel panic - not syncing: hung_task: blocked tasks
Pid: 177, comm: khungtaskd Not tainted 2.6.32-358.el6.x86_64.debug #1