Bug 75965

Summary: weird 'pixel static' when monitor is woken from standby
Product: xorg Reporter: Brian Paterni <bpaterni>
Component: Driver/RadeonAssignee: xf86-video-ati maintainers <xorg-driver-ati>
Status: RESOLVED INVALID QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: medium    
Version: unspecified   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
xorg.log
none
static on top of desktop background
none
static on top of terminal
none
static on top of terminal with cursor
none
static on top of terminal, full screen
none
static on top of terminal with mouse cursor
none
dmesg none

Description Brian Paterni 2014-03-10 04:46:05 UTC
Created attachment 95484 [details]
xorg.log

Hi

Sometimes when wake my monitor from standby weird 'pixel static'(?) appears on the screen. By this I mean some pixels are not colored the way they are supposed to. The 'static' also isn't stable and it seems to jitter about similar to analog tv static. I could try continuing to describe it, but that would probably only lead to more confusion. To compensate, I've attached a few photos of what I'm seeing.

Here are a few things that I gather to be known of this problem:

* The static is triggered when the monitor comes out of standby, but is not guaranteed to happen everytime the monitor is woken up
* The static does NOT appear in screenshots
* turning the monitor off/on via power switch does NOT fix
* turning the monitor off/on via xrandr DOES fix

Hardware specific details:

* graphics card is radeon hd 4850 (rv770)
* running dual monitors, both Asus VS248H-P LED-backlit
* static is only ever present on DVI-0
Comment 1 Brian Paterni 2014-03-10 04:48:54 UTC
Created attachment 95485 [details]
static on top of desktop background
Comment 2 Brian Paterni 2014-03-10 04:55:54 UTC
Created attachment 95486 [details]
static on top of terminal

note the regular pattern
Comment 3 Brian Paterni 2014-03-10 04:57:28 UTC
Created attachment 95487 [details]
static on top of terminal with cursor
Comment 4 Brian Paterni 2014-03-10 04:58:56 UTC
Created attachment 95488 [details]
static on top of terminal, full screen
Comment 5 Brian Paterni 2014-03-10 05:02:31 UTC
Created attachment 95489 [details]
static on top of terminal with mouse cursor
Comment 6 Brian Paterni 2014-03-10 05:13:00 UTC
(In reply to comment #0)

Given that the static doesn't appear in screenshots and the fact that it's only present on DVI-0, I am a bit skeptical that this might indeed be a hardware problem, but I wanted to hear an expert's thoughts on this problem before I discount software entirely.
Comment 7 Alex Deucher 2014-03-10 16:09:58 UTC
Please attach your dmesg output.  It's possible that it's a software issue, possibly a timing issue with respect to initializing part of the display pipe.  E.g., you need a 10 us delay between turning on blocks A and B for block a to synchronize, but in certain causes you only get 9 us or in certain causes you really need 11 us, etc.
Comment 8 Brian Paterni 2014-03-10 16:29:44 UTC
Created attachment 95517 [details]
dmesg

Here is my dmesg as requested. From what I can tell, it is all boot messages though. Nothing in particular is output when the static occurs.
Comment 9 Alex Deucher 2014-03-10 16:35:43 UTC
Is it still an issue if you disable dpm?  E.g., boot with radeon.dpm=0 on the kernel command line in grub?
Comment 10 Brian Paterni 2014-03-10 17:23:21 UTC
I haven't tried disabling dpm on 3.13 yet, but I do know this has occurred from time-to-time before dpm was enabled by default in 3.13 (The first occurance would probably be a year or more ago) so I don't know if that changes anything...

I will try disabling dpm though, but the problem with reproducing this bug is that it doesn't happen with full probability when coming out of standby. Is there some way to force standby state without waiting for the system to timeout?
Comment 11 Alex Deucher 2014-03-10 18:13:07 UTC
(In reply to comment #10)
> I haven't tried disabling dpm on 3.13 yet, but I do know this has occurred
> from time-to-time before dpm was enabled by default in 3.13 (The first
> occurance would probably be a year or more ago) so I don't know if that
> changes anything...

Ok.  If you saw it without dpm, then it's probably not related.

> 
> I will try disabling dpm though, but the problem with reproducing this bug
> is that it doesn't happen with full probability when coming out of standby.
> Is there some way to force standby state without waiting for the system to
> timeout?

Try running:

sleep 5; xset dpms force off

from a terminal
Comment 12 Brian Paterni 2016-01-11 21:09:23 UTC
Some time ago, I replaced the old 4850 with a newer Tonga card and haven't seen this problem since. As a result, I'm going to claim the source of this issue had been problems that were the result of aging hardware.

Marking RESOLVED - INVALID

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.