Robert Norris wrote: > > > Hi all, > > Until recently, I've had XFree86 from potato installed on my Toshiba > Satellite Pro 490XCDT, without issue. This laptop has a S3 ViRGE/MX > video chip, so I was using xserver-svga 3.3.6-11. > > Today I upgraded the system to woody. xserver-xfree86 4.0.2-1 was > installed. > > After configuring and starting the server, the screen went blank and the > system hung. The machine did not respond to pings from the network, and > the Alt-SysRQ-<x> key combo did nothing. > > I found bug #78406, which seems to describe this exact same problem, but > it also states that it has been fixed. > > I have some logfiles and other information available at > http://nauseum.org/x4-virge/ . This is everything that I would submit in > a bug report, though I'm reluctant to do so initially in case I missed > something simple. If people prefer it, I'll post them all to the list, > but they are large. > > I've generated these logs by remounting the filesystems with the "sync" > option and then holding Alt-SysRQ-S until the disks and output from ping > on another machine stopped. > > I've tried five different virge drivers - the one included with 4.0.2-1 > and all the alternate drivers available from > http://www.user1.netcarrier.com/~kbrosius/virge.html . The log output > from all of these is included, however, none have made any real > difference. > > Does anyone have any other ideas I can try? Should I submit a proper bug > report? > > TIA, > Rob. >
As far as XFree86 is concerned, you don't need to submit any more than this. I've forwarded your message to the private ViRGE mailing list, we'll see if anyone with a ViRGE MX can take a look. Will you be taking the log files down soon? Have you tried 'XFree86 -configure' yet? That will generate a generic XF86Config file. Let me know if you see any difference. It can't be any worse... There's also a small chance that trying: Option "noaccel" Option "usefb" "no" one at a time in the device section of XF86Config might make a difference. (These are work-arounds. If they show improvement it helps us narrow down the bug.) -- Kevin
