Created attachment 354506 [details] Screenshot.jpg Description of problem: Using i586 live image posted for the Fit'n'Finish test day, the colors appear almost in greyscale on a system using the nouveau driver. Version-Release number of selected component (if applicable): * xorg-x11-drv-nouveau-0.0.14-3.20090717gitb1b2330.fc12 * kernel-2.6.31-0.81.rc3.git4.fc12 How reproducible: * Everytime on this system ... does not reproduce with F-11 Steps to Reproduce: 1. Download live image posted to https://fedoraproject.org/wiki/Test_Day:2009-07-21_Fit_and_Finish:Batteries_and_Suspend 2. Boot on system 3. Login to desktop Actual results: * See attached screenshot Expected results: * Normal looking colors Additional info: * No xorg.conf file is used * See attached Xorg information (as recommended by https://fedoraproject.org/wiki/Xorg/Debugging).
Created attachment 354507 [details] smoltprofile.txt
Created attachment 354508 [details] Xorg.0.log
Created attachment 354509 [details] dmesg
For the sake of the 'f12beta' blocking, I will note that I did not reproduce this, loading the same kernel and nouveau versions yesterday. Colors looked fine. My hardware is a 9400 GT, which Ben knows all about :) -- Fedora Bugzappers volunteer triage team https://fedoraproject.org/wiki/BugZappers
If I'm reading this correctly, there seem to be a significant amount of smolt profiles with this adapter. http://smolts.org/reports/view_device/NV44%20%5BQuadro%20NVS%20285%5D I'm trying to see if there's any way to reach out to them.
yeah, I think you are reading it right. looks like a common workstation card. -- Fedora Bugzappers volunteer triage team https://fedoraproject.org/wiki/BugZappers
No real ideas here at the moment, especially if it didn't happen in F11, not a lot changed that could be responsible in those builds. How're things with latest kernel, xorg-x11-drv-nouveau and libdrm from rawhide? Nouveau will default to kernel modesetting for the card now, it'd be interesting to know results with/without nouveau.modeset=0.
Well, good news. I'm no longer seeing this issue with the F-12-Alpha test compose on the hardware previously exhibiting this issue. I'm comfortable with closing this out for now and reopening should the problem resurface.
Ok, glad to hear it! Please reopen if it reoccurs!