Bug 117931 - Black screen after resume from hibernate on HP Pavilion dv7 with 2 AMD GPUs
Summary: Black screen after resume from hibernate on HP Pavilion dv7 with 2 AMD GPUs
Status: NEW
Alias: None
Product: Drivers
Classification: Unclassified
Component: Video(DRI - non Intel) (show other bugs)
Hardware: x86-64 Linux
: P1 normal
Assignee: drivers_video-dri
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-05-10 09:04 UTC by Eugene A. Shatokhin
Modified: 2016-09-06 02:33 UTC (History)
1 user (show)

See Also:
Kernel Version: 4.6-rc5 (mainline), 3.14.53, 4.4.7, 4.4.8 (Fedora)
Subsystem:
Regression: No
Bisected commit-id:


Attachments
The system log from Fedora 23 after resume. drm.debug=0x1e was set, just in case (594.22 KB, text/plain)
2016-05-10 09:04 UTC, Eugene A. Shatokhin
Details
lspci -vnn (14.20 KB, text/plain)
2016-05-10 09:06 UTC, Eugene A. Shatokhin
Details
Xorg.0.log after resume (49.97 KB, text/plain)
2016-05-10 09:07 UTC, Eugene A. Shatokhin
Details
System log after resume (kernel 4.7) (1.93 MB, text/plain)
2016-07-25 09:09 UTC, Eugene A. Shatokhin
Details
Xorg.0.log after resume (kernel 4.7) (57.03 KB, text/plain)
2016-07-25 09:10 UTC, Eugene A. Shatokhin
Details
lspci -vnn (kernel 4.7) (13.64 KB, text/plain)
2016-07-25 09:10 UTC, Eugene A. Shatokhin
Details

Description Eugene A. Shatokhin 2016-05-10 09:04:28 UTC
Created attachment 215701 [details]
The system log from Fedora 23 after resume. drm.debug=0x1e was set, just in case

When the laptop resumes from hibernate, there is no output on the screen. In some cases, I can connect to the machine via SSH after that, so I can see that the system does not crash. Looks like it is just graphics that is not turned on.

The system uses 'radeon' driver.

Laptop: HP Pavilion dv7 with A8-3530MX APU
Two graphics cards:
* Radeon HD 7690M (pci:1002-6741-103c-1860)
* Radeon HD 6620G (pci:1002-9641-103c-1860)
More hardware info for that system: http://linux-hardware.org/index.php?probe=4c12fda921

OS, kernel:
I tried both ROSA R7 x86_64 with kernels 4.6-rc5 (mainline), 4.4.7, 4.1.22 and Fedora 23 with kernel 4.4.8-300.fc23.x86_64. The problem showed up with each of these.

The log below is from Fedora where I was able to connect to the machine via SSH after the problem happened.

Resuming from suspend-to-RAM works fine though.

I also tried kernel 4.1.x with the proprietary fglrx driver (v.15.201.1151) instead of 'radeon' - the graphics worked OK after resume from hibernate this time.
Comment 1 Eugene A. Shatokhin 2016-05-10 09:06:53 UTC
Created attachment 215711 [details]
lspci -vnn
Comment 2 Eugene A. Shatokhin 2016-05-10 09:07:18 UTC
Created attachment 215721 [details]
Xorg.0.log after resume
Comment 3 Eugene A. Shatokhin 2016-07-25 09:07:07 UTC
Re-checked on Fedora 23 x64 with the mainline kernel 4.7-rc7, git rev 68093c43f352, with no additional patches.

Resume from hibernation is not fully working yet but the problem looks different this time.

When I put the system to suspend-to-disk then try to resume it, the desktop environment appears this time but does not respond to keyboard, mouse, whatever.

The system log contains lots of meesages "radeon: The kernel rejected CS, see dmesg for more information.". 

Xorg.0.log ends with the following:
[  1608.239] (WW) RADEON(0): radeon_dri2_flip_event_handler: Pageflip completion event has impossible msc 78924 < target_msc 78925

I will attach the logs below.

Besides, after I have done suspend-to-ram and resumed the system (without prior reboot after the problem happened), the desktop seemed to be  operational again after a couple minutes.

Interestingly enough, lspci shows incomplete info for one of the GPUs this time (was OK with the earlier kernels):
-------------------
01:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Whistler [Radeon HD 6630M/6650M/6750M/7670M/7690M] [1002:6741] (rev ff) (prog-if ff)
	!!! Unknown header type 7f
	Kernel driver in use: radeon
	Kernel modules: radeon
-------------------
May be it is not related to the problem, may be it is.
Comment 4 Eugene A. Shatokhin 2016-07-25 09:09:35 UTC
Created attachment 225761 [details]
System log after resume (kernel 4.7)
Comment 5 Eugene A. Shatokhin 2016-07-25 09:10:09 UTC
Created attachment 225771 [details]
Xorg.0.log after resume (kernel 4.7)
Comment 6 Eugene A. Shatokhin 2016-07-25 09:10:45 UTC
Created attachment 225781 [details]
lspci -vnn (kernel 4.7)

Note You need to log in before you can comment on or make changes to this bug.