Bug 200611 - BISECTED 3cd091a77393: System may get unresponsive after resume from suspend-to-RAM
Summary: BISECTED 3cd091a77393: System may get unresponsive after resume from suspend-...
Status: CLOSED UNREPRODUCIBLE
Alias: None
Product: ACPI
Classification: Unclassified
Component: EC (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Zhang Rui
URL:
Keywords:
: 201091 201153 202675 (view as bug list)
Depends on:
Blocks:
 
Reported: 2018-07-20 13:41 UTC by Timo Kalliomäki
Modified: 2020-11-18 14:08 UTC (History)
6 users (show)

See Also:
Kernel Version: 4.18-rc4
Subsystem:
Regression: Yes
Bisected commit-id:


Attachments
Dmesg on 4.16-rc1+3cd09 (freeze after resume) (68.87 KB, text/plain)
2018-07-20 13:41 UTC, Timo Kalliomäki
Details
Dmesg on 4.16-rc1+3cd09 (successful resume) (69.11 KB, text/plain)
2018-07-20 13:43 UTC, Timo Kalliomäki
Details
Dmesg on 4.16-rc1 (80.98 KB, text/plain)
2018-07-20 13:43 UTC, Timo Kalliomäki
Details
Dmesg on 4.18-rc4 (freeze after resume) (74.52 KB, text/plain)
2018-07-20 13:44 UTC, Timo Kalliomäki
Details

Description Timo Kalliomäki 2018-07-20 13:41:51 UTC
Created attachment 277437 [details]
Dmesg on 4.16-rc1+3cd09 (freeze after resume)

When resuming from RAM, the laptop screen is black, with only the mouse cursor visible. Switching to a virtual console and back brings up the unlocking screen. However, on maybe one third of resumes, the system eventually (after 1–10 seconds) becomes unresponsive, close to the point of being frozen (only the mouse cursor moves).

Dmesg logs featuring successful and unsuccessful resumes are attached.

Bisecting shows that this behavior was introduced in 3cd091a773936c54344a519f7ee1379ccb620bee (“ACPI / EC: Restore polling during noirq suspend/resume phases”, which was in response to bug 197863). Thus 4.16-rc1 was the last non-freezing mainline version, and the behavior is still present in 4.18-rc4.

Hardware: Dell Latitude E7450 laptop (Intel i7-5600U processor with HD Graphics 5500 IGP), attached via DP to an external screen.
Comment 1 Timo Kalliomäki 2018-07-20 13:43:11 UTC
Created attachment 277439 [details]
Dmesg on 4.16-rc1+3cd09 (successful resume)
Comment 2 Timo Kalliomäki 2018-07-20 13:43:41 UTC
Created attachment 277441 [details]
Dmesg on 4.16-rc1
Comment 3 Timo Kalliomäki 2018-07-20 13:44:11 UTC
Created attachment 277443 [details]
Dmesg on 4.18-rc4 (freeze after resume)
Comment 4 Zhang Rui 2018-07-26 01:12:31 UTC
(In reply to Timo Kalliomäki from comment #0)
> Created attachment 277437 [details]
> Dmesg on 4.16-rc1+3cd09 (freeze after resume)
> 
> When resuming from RAM, the laptop screen is black, with only the mouse
> cursor visible. Switching to a virtual console and back brings up the
> unlocking screen.

is this also caused by the EC commit?
Comment 5 Timo Kalliomäki 2018-07-26 06:27:57 UTC
(In reply to Zhang Rui from comment #4)
> (In reply to Timo Kalliomäki from comment #0)
> > Created attachment 277437 [details]
> > Dmesg on 4.16-rc1+3cd09 (freeze after resume)
> > 
> > When resuming from RAM, the laptop screen is black, with only the mouse
> > cursor visible. Switching to a virtual console and back brings up the
> > unlocking screen.
> 
> is this also caused by the EC commit?

Yes, both badnesses A: “immediate black screen that can be recovered from” and B: “eventual freeze” were introduced by the same commit.

Another thing I should have been more clear on: the external monitor is necessary to reproduce A and so far B hasn’t happened with it unplugged either. (Could it be possible there was some sort of latent bug in the i915 driver that this commit just made surface?)
Comment 6 Zhang Rui 2018-12-27 15:34:26 UTC
*** Bug 201091 has been marked as a duplicate of this bug. ***
Comment 7 Zhang Rui 2018-12-27 15:43:26 UTC
*** Bug 201153 has been marked as a duplicate of this bug. ***
Comment 8 Zhang Rui 2019-09-03 07:09:27 UTC
*** Bug 202675 has been marked as a duplicate of this bug. ***
Comment 9 Zhang Rui 2020-06-29 03:14:44 UTC
We have quite some EC changes recently, can you please check if the problem still exists in the latest upstream kernel?
Comment 10 Zhang Rui 2020-11-18 14:08:33 UTC
I think this should be fixed in the latest upstream kernel.
please feel free to reopen it if the problem still exists in the latest upstream kernel.

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