Comment 2 for bug 506444

Revision history for this message
Stefano Rivera (stefanor) wrote :

As of this morning's updates. Starting X results in a GPU hang.
Xorg spends 100% of CPU time in kernel mode, and the kernel starts complaining:

[ 360.480027] INFO: task i915/0:293 blocked for more than 120 seconds.
[ 360.480032] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 360.480034] i915/0 D 00000000ffffffff 0 293 2 0x00000000
[ 360.480039] ffff880097d6dd40 0000000000000046 0000000000015ac0 0000000000015ac0
[ 360.480044] ffff8800981a1ab0 ffff880097d6dfd8 0000000000015ac0 ffff8800981a16f0
[ 360.480048] 0000000000015ac0 ffff880097d6dfd8 0000000000015ac0 ffff8800981a1ab0
[ 360.480052] Call Trace:
[ 360.480063] [<ffffffff81575d17>] __mutex_lock_slowpath+0xe7/0x170
[ 360.480070] [<ffffffff81055200>] ? finish_task_switch+0x50/0xd0
[ 360.480073] [<ffffffff81575c0b>] mutex_lock+0x2b/0x50
[ 360.480111] [<ffffffffa00c691d>] i915_gem_retire_work_handler+0x3d/0xa0 [i915]
[ 360.480120] [<ffffffffa00c68e0>] ? i915_gem_retire_work_handler+0x0/0xa0 [i915]
[ 360.480126] [<ffffffff8107ead7>] run_workqueue+0xc7/0x1a0
[ 360.480129] [<ffffffff8107ec53>] worker_thread+0xa3/0x110
[ 360.480133] [<ffffffff81083680>] ? autoremove_wake_function+0x0/0x40
[ 360.480136] [<ffffffff8107ebb0>] ? worker_thread+0x0/0x110
[ 360.480139] [<ffffffff81083306>] kthread+0x96/0xa0
[ 360.480143] [<ffffffff810141ea>] child_rip+0xa/0x20
[ 360.480146] [<ffffffff81083270>] ? kthread+0x0/0xa0
[ 360.480149] [<ffffffff810141e0>] ? child_rip+0x0/0x20