From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.1) Gecko/20020827 Description of problem: redhat-config-xfree86 chooses 'savage' driver for the above card. This driver does not work for the card. A working driver is 'VESA driver (generic)'. Xconfiguratior (in RedHat 8.0) used to set the driver correctly. The workaround for the problem is to install the system, then manually change the driver in /etc/X11/XF86Config configuration file to 'VESA driver (generic)'. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Install RedHat 9.0 Actual Results: X windows doesn't work. Expected Results: X windows should work. Additional info:
The savage driver supports this card. If it doesn't work, then it needs to be fixed. I will not change this chip to use the vesa driver except as a last ditch resort after receiving many bug reports from various users (this is the first bug report), and after exhausting all options to try to fix the problem occuring in the savage driver. Please file a bug report against XFree86 indicating in as much detail as possible what exactly is wrong. "doesn't work" isn't enough detail to be able to do anything. State exactly what doesn't work, what you tried, and then attach your config file and log file as bugzilla file attachments to the report and I will try to troubleshoot with you to narrow the problem down. Closing request to change to "vesa" driver by default as WONTFIX. We want the drivers that don't work to get fixed.
There's a bug which is already open for this chip. What happens is that X won't run at all in any resolution higher that 640x480. Changing the driver to the one found at Tim Robert's page won't work at all. As Buturovic said, the step to reproduce this bug is just try to install redhat 9. It won't work unless you change display's resolution to 640x480, and the system get VERY unstable with some applications. I really would like to see a correction for this, there's a new machine here which is stopped just because of that.
In order for us to provide an update, would require someone outside of Red Hat who has access to both the Savage hardware that is problematic, as well as the specifications to that chip from VIA/S3, and a knowledge of the chip's operation and of the XFree86 driver to investigate the issue and hopefully fix it. Once that occurs, and fixes are made available from XFree86.org or from the Savage driver maintainer or VIA/S3, then I can investigate incorporating them into future versions of Red Hat Linux, and possibly even update the driver for future XFree86 security erratum for Red Hat Linux 9. Unfortunately, there is nothing that Red Hat can do directly about this problem as we do not have any knowledge of how Savage video hardware works, nor do we have the Savage video hardware specifications as VIA/S3 do not make them publically available. I also have no Savage hardware to work with either, so all bugfixes we apply to the Savage driver come either directly from the upstream maintainer, or come from XFree86 CVS. As such, until it is fixed upstream, the official Red Hat workaround to this problem is for users to reconfigure X to use the VESA driver. This is not the most optimal solution, but there is nothing else we can do. Your best bet is to file a bug report at http://bugs.xfree86.org and hope it gets fixed in 4.4.0. If you do this, please paste the upstream bug report URL here in this bug report, and it does get fixed upstream, feel free to indicate that here as well in order for me to investigate the possiblity of updating the driver to the new fixed version, or of backporting the specific fix. Thanks.