Description of problem: on certain Nvidia cards (mainly XFX cards), the screen will not be displayed properly, therefor anaconda is inusable (in my particular case, i get green stripes, and no image. the problem lies in the fact that Anaconda forces itself to use the 'nv' driver instead of 'vesa' (it would work with vesa) Version-Release number of selected component (if applicable): How reproducible: Boot the FC5 install, with a XFX graphics card) and just run Anaconda Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: you should make an option where the user can choose what driver Anaconda will use
*** Bug 186034 has been marked as a duplicate of this bug. ***
The driver package exports what PCI ids it supports. Note that you can force anaconda to use vesa instead by booting with 'linux vesa'
I can confirm this with a Geforce 6800 GT. The "nv" driver just corrupts the screen. The "vesa" driver works OK, but is very limited... Note this happens both during install, and afterwards. As long as I set the driver to "vesa", I can use the system. If I try to use the "nv" driver (which is what it should be), I get nothing but corrupt graphics, to the point of having to actually reboot (just killing the X-server doesn't restore the video). This is with FC5 x86_64, on an ASUS A8V Deluxe, AMD64 3800+, 1 GB RAM. Works (worked) fine in Fedora Core 3 and 4. I'm normally using a dual-monitor setup (one LCD, one CRT), but I've tried it with just the CRT attached as well.
Using the 'nv' driver, add the following option to the xorg.conf: Option "noaccel" Restart the X server and report back the results. Attach the X server log file and config file from the new configuration as individual uncompressed file attachments using the link below.
Noaccel option makes no difference. Attaching config file and X server log from the attempt.
Created attachment 126621 [details] Config file
Created attachment 126622 [details] X server log
I've been looking on the xorg and freedesktop sites - this bug may be related to https://bugs.freedesktop.org/show_bug.cgi?id=6037 (in fact, it most likely is...) which links to https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=182396 . I don't know why I didn't find that one on my first look through bugzilla... might be a duplicate.
Our bugzilla database had a catastrophic failure on June 13, 2006 which lost all bug changes from Thurs June 9th through 13th. I have bugzilla email records of the time period however, and will be restoring as many changes as I can manually.
Ok if "noaccel" does not work, try booting into runlevel 3, logging in as root, and running "startx", or alternatively disable or uninstall rhgb and reboot into X again. Does the problem still occur in runlevel 3, or with rhgb out of the way? rhgb is known to break otherwise working video setups, in the name of eye candy, and we know it breaks various things on nvidia setups thus far. That may help narrow things down...
------- Additional Comments From peterd 2006-06-10 12:10 EST ------- Nope, runlevel 3 "startx" (as root) doesn't help.
Alright, it appears that the "nv" driver does not properly support this chip adequately enough. We will track the issue in the upstream bug report 6037 as referenced in comment #8. When Nvidia has submitted patches to X.org which fix the problem, and they've been committed to CVS, we'll review them for consideration in a future 'nv' driver update. Thanks for your testing. Setting status to "UPSTREAM"