Bug 9158 - Current git radeon(4) does not work with MergedFB
Summary: Current git radeon(4) does not work with MergedFB
Status: RESOLVED NOTABUG
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/Radeon (show other bugs)
Version: 7.1 (2006.05)
Hardware: Other Linux (All)
: high normal
Assignee: Xorg Project Team
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2006-11-25 14:55 UTC by Diego Elio Pettenò
Modified: 2006-11-28 11:15 UTC (History)
0 users

See Also:
i915 platform:
i915 features:

Attachments

Description Diego Elio Pettenò 2006-11-25 14:55:46 UTC 
As noted by Donnie Berkholz in bug #9031, current GIT version of 
xf86-video-ati does not work with MergedFB as intended (in particular, the 
second screen is not displaying anything) and requires a reboot before getting 
to work (even when replacing the driver with a working copy0.

I've done a git-bisect run on the driver, and restricted the change to commit 
457178f418983041ea59df703a17513a47d17044: "Disable all ouputs on initial mode 
set.  The ones we want will get enabled by RADEONEnableDisplay()."

It seems like RADEONEnableDisplay() is not enabling the second output, I'm 
going to take a look to the code, but I don't think I'll be able to get much 
done, but at least I hope this report will be useful.

HTH,
Diego
Comment 1 Diego Elio Pettenò 2006-11-25 15:41:35 UTC 
I tried to put

    RADEONEnableDisplay(pScrn, pCRTC1, TRUE);
    RADEONEnableDisplay(pScrn, pCRTC2, TRUE);

right after the RADEONDisableDisplays call, but that didn't help, so the 
problem is either RADEONDisableDisplays which disables too much, or 
RADEONEnableDisplay that is not able to enable them as supposed to, rather 
than the logic of RADEONRestoreMode to be flawed, but I admit this is the 
opinion of a prophane.
Comment 2 Diego Elio Pettenò 2006-11-25 16:13:20 UTC 
Okay, I found what was the cause of the problem, so I'm not sure if this bug 
is still valid or if it was my configuration wrong.
When I first configured MergedFB, I set MonitorLayout to "TMDS, TMDS", with 
the setting I had with the previous card (an nvidia one) "TMDS, CRT" I had a 
blurry output on the second monitor.

With this setup, pCRTC2->pPort->MonType was set at MT_DFP, while 
pCRTC2->pPort->TMDSType was set to TMDS_UNKNOWN. This falled out of the ifs in 
RADEONEnableDisplay() and caused the second output never to be enabled again.

Now, I don't see the same blurryness as before on the second monitor, so it 
might have been some other bug fixed already, making my configuration 
obsolete, but it would be certainly nice to have a warning if the probe finds 
unknown values, I tried looking for one on the log, but I didn't.
Comment 3 Alex Deucher 2006-11-28 11:15:02 UTC 
TMDS, TMDS forces the driver to assume you have 2 DVI ports.  The fact that it
worked for you previously was more luck than anything.  Monitorlayout is not
used for probing.  It's there to force the outputs if probing doesn't work or
you want to force a particular configuration.

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.