Red Hat Bugzilla – Bug 454279
Display flickering issues on RHEL 5.2
Last modified: 2011-06-29 14:50:44 EDT
When using the i810 driver in RHEL 5.2 the customer experiences a similar
flickering as described when using the vesa driver. It's different though in
that once it is flickering cycling between Ctrl+Alt+F1|F7 will not reconcile the
problem, the only solution is to reboot. The vesa driver does similar
flickering, but can be fixed by cycling between Ctl+Alt+F1|F7. Customer has
tried this with various types of Monitors so we've ruled out a sync issue.
1. Provide time and date of the problem
2. Indicate the platform(s) (architectures) the problem is being reported against.
Dell Optiplex 745 (or 755) with intel 82Q963/Q965 cards.
3. Provide clear and concise problem description as it is understood at the
time of escalation
Screen starts flickering/pulsating with the last color on the screen until
either reboot or cycle through virtual terminals depending on the driver being used.
It shouldn't flicker/pulsate like this.
4. State specific action requested of SEG
Help in setting up a bugzilla or linking it to the existing:
5. State whether or not a defect in the product is suspected
* Provide Bugzilla if one already exists
https://bugzilla.redhat.com/show_bug.cgi?id=220337 might be related.
6. If there is a proposed patch, make sure it is in unified diff format (diff
No patch that I'm aware of, but the previous bugzilla resulted in the following
Customer provided the following:
The kernel no longer panics on shutting down X.
When using the i810 driver now I experience a similar flickering as described
when using the vesa driver. It's different though in that once it is flickering
cycling between Ctrl+Alt+F1/F7 will not reconcile the problem, the only solution
I've found so far is to reboot.
Vesa driver seems to be the same as before (flickering can occur some of the
time, but cycling between Ctrl+Alt+F1/F7 will usually fix it).
All tests done with:
Older sosreport from customer before update to RHEL 5.2
I'll try to go ahead and ask for an updated sosreport.
I suggested a change to the intel driver and got the following response:
Customer updated with the following. It also looks like kdump is having issues
but hopefully we can get a vmcore soon:
If I turn on the intel driver, the system crashes when it attempts to utilize it
(both when you start X or when you try to boot with bootsplash). The screen
goes blank (monitor light stays on), keyboard/numlock is unresponive, all
network traffic stops, I assume this is a kernel panic, but no "kernel panic"
info is displayed. I can turn on kdump and grab a dump if you want.
I've attached the logs that show the output leading up to the crash, they don't
show anything interesting, just substantiate that loading the intel driver
caused it. See the 'Xorg.0.log' and 'messages' attachments.
If I run 'startx' to cause it to lock up, the screen goes blank and nothing
happens. I let it sit there for 30 minutes before power cycling it. Nothing
was written to the crash location. I'm pretty sure it's a kernel panic though
since the host stops responding to pings and my ssh session to the machine hangs.
Created attachment 311150 [details]
X-org.log after trying interl driver
Created attachment 311153 [details]
Created attachment 311155 [details]
Created attachment 311892 [details]
Xorg.0.log from the sosreport
Created attachment 311893 [details]
xorg.conf from the sosreport
1.6.509.15.el5 has been built against the new randr sources, which should at least make things fail differently.
Test this on Q963/Q965 with snap4 x86_64. I do not see any flickering when I change to text consoles and back to X console. Nor do I see any flickering with changing modes with xrandr. I am using the intel driver.
Moving to verified. Change it you retest and it doesn't work.
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.