[Intel-gfx] YABT - pipe_off wait timed out

[Ville Syrjälä]

On Wed, Sep 17, 2014 at 01:38:21PM +1000, Dave Airlie wrote:
> [    1.396361] ------------[ cut here ]------------
> [    1.396396] WARNING: CPU: 2 PID: 108 at
> /home/airlied/devel/kernel/drm-next/drivers/gpu/drm/i915/intel_display.c:997
> intel_disable_pipe+0x2ae/0x2c0 [i915]()
> [    1.396397] pipe_off wait timed out
> [    1.396401] Modules linked in: i915 i2c_algo_bit video
> drm_kms_helper drm i2ccore
> [    1.396403] CPU: 2 PID: 108 Comm: kworker/u16:2 Not tainted 3.17.0-rc5+ #75
> [    1.396404] Hardware name: Hewlett-Packard HP Z220 CMT
> Workstation/1790, BIOS K51 v01.65 09/03/2013
> [    1.396409] Workqueue: events_unbound async_run_entry_fn
> [    1.396412]  0000000000000009 ffff8805f847b798 ffffffff81691069
> ffff8805f847b7e0
> [    1.396413]  ffff8805f847b7d0 ffffffff81073fad ffff8805f6ab0000
> 0000000000070008
> [    1.396414]  00000000fffb7181 0000000000000000 ffff8805f8106330
> ffff8805f847b830
> [    1.396415] Call Trace:
> [    1.396421]  [<ffffffff81691069>] dump_stack+0x45/0x56
> [    1.396423]  [<ffffffff81073fad>] warn_slowpath_common+0x7d/0xa0
> [    1.396425]  [<ffffffff8107401c>] warn_slowpath_fmt+0x4c/0x50
> [    1.396447]  [<ffffffffa00e456e>] intel_disable_pipe+0x2ae/0x2c0 [i915]
> [    1.396469]  [<ffffffffa00ef8dd>] ironlake_crtc_disable+0x9d/0x7c0 [i915]
> [    1.396487]  [<ffffffffa00e1bec>] ?
> intel_dump_pipe_config.isra.51+0x3c/0x3a0 [i915]
> [    1.396489]  [<ffffffff811b3a66>] ? kmem_cache_alloc_trace+0x1d6/0x200
> [    1.396508]  [<ffffffffa00e9f31>] __intel_set_mode+0x7b1/0x1620 [i915]
> [    1.396511]  [<ffffffff8120e683>] ? ep_poll_callback+0xf3/0x160
> [    1.396529]  [<ffffffffa00f2716>] intel_set_mode+0x16/0x30 [i915]
> [    1.396547]  [<ffffffffa00f368c>] intel_crtc_set_config+0x92c/0xe50 [i915]
> [    1.396558]  [<ffffffffa0025141>]
> drm_mode_set_config_internal+0x61/0xf0 [drm]
> [    1.396563]  [<ffffffffa007e843>] restore_fbdev_mode+0xd3/0x100
> [drm_kms_helper]
> [    1.396567]  [<ffffffffa007e893>]
> drm_fb_helper_restore_fbdev_mode_unlocked+0x23/0x40 [drm_kms_helper]
> [    1.396572]  [<ffffffffa00804a8>] drm_fb_helper_set_par+0x28/0x70
> [drm_kms_helper]
> [    1.396593]  [<ffffffffa00fe3ca>] intel_fbdev_set_par+0x1a/0x60 [i915]
> [    1.396595]  [<ffffffff81383244>] fbcon_init+0x4f4/0x580
> [    1.396599]  [<ffffffff813e89a8>] visual_init+0xb8/0x120
> [    1.396602]  [<ffffffff813eb013>] do_bind_con_driver+0x163/0x330
> [    1.396604]  [<ffffffff813eb794>] do_take_over_console+0x114/0x1c0
> [    1.396606]  [<ffffffff8137e993>] do_fbcon_takeover+0x63/0xd0
> [    1.396608]  [<ffffffff81383ced>] fbcon_event_notify+0x68d/0x7e0
> [    1.396610]  [<ffffffff8109097c>] notifier_call_chain+0x4c/0x70
> [    1.396612]  [<ffffffff81090cbd>] __blocking_notifier_call_chain+0x4d/0x70
> [    1.396613]  [<ffffffff81090cf6>] blocking_notifier_call_chain+0x16/0x20
> [    1.396616]  [<ffffffff8138a23b>] fb_notifier_call_chain+0x1b/0x20
> [    1.396617]  [<ffffffff8138c386>] register_framebuffer+0x1e6/0x320
> [    1.396622]  [<ffffffffa008075f>]
> drm_fb_helper_initial_config+0x26f/0x3c0 [drm_kms_helper]
> [    1.396641]  [<ffffffffa00ff5bb>] intel_fbdev_initial_config+0x1b/0x20 [i915]
> [    1.396643]  [<ffffffff81092479>] async_run_entry_fn+0x39/0x120
> [    1.396645]  [<ffffffff8108a8a2>] process_one_work+0x142/0x3c0
> [    1.396647]  [<ffffffff8108ac3b>] worker_thread+0x11b/0x480
> [    1.396649]  [<ffffffff8108ab20>] ? process_one_work+0x3c0/0x3c0
> [    1.396651]  [<ffffffff8108fab9>] kthread+0xc9/0xe0
> [    1.396653]  [<ffffffff8108f9f0>] ? kthread_create_on_node+0x170/0x170
> [    1.396656]  [<ffffffff816984ac>] ret_from_fork+0x7c/0xb0
> [    1.396659]  [<ffffffff8108f9f0>] ? kthread_create_on_node+0x170/0x170
> [    1.396660] ---[ end trace 94ef6c67e63fd2ee ]---
> [    1.449389] [drm:ironlake_disable_pch_transcoder] *ERROR* failed to
> disable transcoder A
> [    1.451978] [drm:ivybridge_set_fifo_underrun_reporting] *ERROR*
> uncleared fifo underrun on pipe A
> [    1.451980] [drm:ivb_err_int_handler] *ERROR* Pipe A FIFO underrun
> [    1.452429] [drm:cpt_set_fifo_underrun_reporting] *ERROR* uncleared
> pch fifo underrun on pch transcoder A
> [    1.452431] [drm:cpt_serr_int_handler] *ERROR* PCH transcoder A FIFO underrun
> 
> This is on an Ivybridge desktop with one DVI and one DP monitor with drm-next.

So our modeset sequence is still crap. Was this already present or a new
problem?

Do you have this in your tree?

commit 08aff3fe26ae7a0d6f302ac2e1b7e2eb9933cd42
Author: Ville Syrjälä <ville.syrjala at linux.intel.com>
Date:   Mon Aug 18 22:16:09 2014 +0300

    drm/i915: Move DP port disable to post_disable for pch platforms

If yes, try reverting. The funy thing is that it fixes a pipe_off
timeout on my ILK and didn't cause any harm on my IVB, but perhaps
your IVB is more special than mine. It also matches the modeset
sequence in the docs better, but perhaps it's still not close enough,
or the docs are just plain wrong.

> If I don't see a fix in the next while I'll just remove the warning, deal?
> 
> Dave.
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Ville Syrjälä
Intel OTC