Bug 107300

Summary: [REGRESSION] NULL pointer dereference when waking up monitor
Product: DRI Reporter: mr.nuke.me
Component: DRM/AMDgpuAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED MOVED QA Contact:
Severity: normal    
Priority: medium CC: harry.wentland, nicholas.kazlauskas, sunpeng.li
Version: XOrg git   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
dmesg none

Description mr.nuke.me 2018-07-19 21:41:19 UTC
Created attachment 140719 [details]
dmesg

I'm noticing this with linux-17.6 on an RX 480, but did not experience this problem with linux-4.16 and earlier.

I let the machine  without any input for a while, and it automatically puts the monitor to sleep. To wake up the screen, I move the mouse or hit the keyboard.

About a third of the time, the login screen comes up, and there's no response. SSH then dmesg points to an amazingly well crafted NULL pointer dereference in [amdgpu].

Surprisingly, this isn't some 4billionk triple-gazillion monitor setup, which I've always had driver issues with. This issue happens with a single display.

For fast access, this is the backtrace: 
[59233.951991] Call Trace:
[59233.952068]  dm_update_crtcs_state+0x26c/0x4d0 [amdgpu]
[59233.952140]  amdgpu_dm_atomic_check+0x1b1/0x3b0 [amdgpu]
[59233.952172]  drm_atomic_check_only+0x360/0x4f0 [drm]
[59233.952192]  drm_atomic_commit+0x13/0x50 [drm]
[59233.952211]  drm_atomic_connector_commit_dpms+0xdb/0x100 [drm]
[59233.952230]  drm_mode_obj_set_property_ioctl+0x178/0x280 [drm]
[59233.952249]  ? drm_mode_connector_set_obj_prop+0x80/0x80 [drm]
[59233.952266]  drm_mode_connector_property_set_ioctl+0x39/0x60 [drm]
[59233.952283]  drm_ioctl_kernel+0x5b/0xb0 [drm]
[59233.952300]  drm_ioctl+0x1b3/0x370 [drm]
[59233.952318]  ? drm_mode_connector_set_obj_prop+0x80/0x80 [drm]
[59233.952375]  amdgpu_drm_ioctl+0x49/0x80 [amdgpu]
[59233.952382]  do_vfs_ioctl+0xa4/0x610
[59233.952386]  ksys_ioctl+0x60/0x90
[59233.952389]  __x64_sys_ioctl+0x16/0x20
[59233.952394]  do_syscall_64+0x5b/0x160
[59233.952399]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
Comment 1 Michel Dänzer 2018-07-20 07:42:57 UTC
Does the patch attached to bug 107153 help?
Comment 2 mr.nuke.me 2018-08-05 17:11:15 UTC
I was not able to observe any improvement with the patch in bug 107153

I should have mentioned that the monitor is connected via DP.
Comment 3 Martin Peres 2019-11-19 08:44:53 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/457.

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.