Bug 106571 - Stoney [Radeon R5 Graphics] hangs on hibernate
Summary: Stoney [Radeon R5 Graphics] hangs on hibernate
Status: RESOLVED MOVED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/amdkfd (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium major
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-05-19 01:47 UTC by Hagar
Modified: 2019-11-19 07:53 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Hardware Listing (15.13 KB, text/plain)
2018-05-19 01:47 UTC, Hagar
no flags Details
Hardware list (21.57 KB, text/plain)
2019-03-20 16:49 UTC, ukbeast89
no flags Details

Description Hagar 2018-05-19 01:47:19 UTC
Created attachment 139632 [details]
Hardware Listing

I have a laptop with "AMD A9-9420e RADEON R5, 5 COMPUTE CORES 2C+3G"

Without the amdgpu module it always hibernates.
With just the amdgou module (without amdkfd and amd_iommu_v2) it reboots instead.
With all 3 loaded it hangs at "PM: Syncing filesystems ... "
All I get in the logs is -
May 19 08:32:53 diamond systemd[1]: Reached target Sleep.
May 19 08:32:53 diamond systemd[1]: Starting Hibernate...
May 19 08:32:53 diamond systemd-sleep[10963]: Suspending system...
May 19 08:32:53 diamond kernel: PM: hibernation entry
May 19 08:32:53 diamond kernel: PM: Syncing filesystems ... 
-- Reboot --

The device suspends just fine.

I currently use the following module options - 
options amdgpu dc=0 si=0 cik=0 exp_hw_support=0 aspm=1 dpm=1 bapm=1 hw_i2c=1
radeon is not loaded (doesnt work on the card anyway)

Kernels tried -
arch 4.16.8
arch lts 4.14.41
arch mainline 4.17rc5
and a couple of others

All showed the same symptoms and log entries.

Thanks
Comment 1 lonelywoolf 2019-03-18 22:46:34 UTC
Still not working on 5.1-rc1. Also tested 4.18, 4.19. 5.0.

I tried some debug, and as I can see, system freezes after
error = dpm_suspend(PMSG_FREEZE);
in hibernate.c

System is A10-9600p + R7 M440. With nomodeset or without amdgpu system hibernates, but completely unusable. Same bug searched with AMD FX-9830P + RX560, and AMD A10-8700P.
Comment 2 lonelywoolf 2019-03-18 22:49:40 UTC
suspend works ok. Only hibernate hanging.
Comment 3 ukbeast89 2019-03-20 16:49:11 UTC
Created attachment 143742 [details]
Hardware list

Tested 5.1-rc1, Lenovo laptop will hang after waking from suspend.
Comment 4 ukbeast89 2019-03-27 02:30:24 UTC
Linux 5.0-rc2 had changes to amdgpu that mentioned resume changes.

"drm/amdgpu: Don't ignore rc from drm_dp_mst_topology_mgr_resume()
drm/amdgpu: Don't fail resume process if resuming atomic state fails
drm/dp_mst: Add __must_check to drm_dp_mst_topology_mgr_resume()"


https://lkml.org/lkml/2019/1/13/146

could someone please test 5.0-rc1?
Comment 5 lonelywoolf 2019-03-27 08:34:49 UTC
(In reply to ukbeast89 from comment #4)
> Linux 5.0-rc2 had changes to amdgpu that mentioned resume changes.
> 
> "drm/amdgpu: Don't ignore rc from drm_dp_mst_topology_mgr_resume()
> drm/amdgpu: Don't fail resume process if resuming atomic state fails
> drm/dp_mst: Add __must_check to drm_dp_mst_topology_mgr_resume()"
> 
> 
> https://lkml.org/lkml/2019/1/13/146
> 
> could someone please test 5.0-rc1?

hibernate hang. System freezes as usual, nothing changes. Original bug is'nt regression. I'm waiting for 5.1-rc3 for some tests, because it's some work with amdgpu/radeon, which can be related.
Comment 6 lonelywoolf 2019-03-27 09:27:52 UTC
(In reply to ukbeast89 from comment #3)
> Created attachment 143742 [details]
> Hardware list
> 
> Tested 5.1-rc1, Lenovo laptop will hang after waking from suspend.

You must create other bug. You have different problem.
Comment 7 ukbeast89 2019-03-27 11:09:52 UTC
(In reply to lonelywoolf from comment #6)
> (In reply to ukbeast89 from comment #3)
> > Created attachment 143742 [details]
> > Hardware list
> > 
> > Tested 5.1-rc1, Lenovo laptop will hang after waking from suspend.
> 
> You must create other bug. You have different problem.

Different bug, but similar.

4.20 gave me no issues, but 4.19, 5.0 and 5.1rc2 are regressions for me.
I will create a separate bug report.
Comment 8 Martin Peres 2019-11-19 07:53:38 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/3.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.