Description of problem: Installing in VPC works ok but after first reboot there is a segmentation fault that makes logging in impossible. Version-Release number of selected component (if applicable): Fedora Core 2 Test 2+3 How reproducible: Always happens when booting Steps to Reproduce: 1. Install FC2 Test-build in VPC 2. After completed installation reboot 3. Enjoy Actual results: /etc/X11/prefdm: line 80: 291 segmentation fault unicode_start $SYSFONT $SYSFONTACM INIT: lx "x" respawning too fast: disabled for 5 minutes /etc/X11/prefdm: line 80: 291 segmentation fault unicode_start $SYSFONT $SYSFONTACM Incorrect Linehigth 0. Expected results: gdm should be started, loginscreen should be displayed, logging in should be possible Additional info: - this is not a dupe to issue 119838. - Platform: VPC for OS/2 (==VPC for Windows Version 5.1) - fc1 works w/o problems,
Created attachment 99733 [details] Errormessage on display of FC2 in VPC (OS/2)
- Same thing happens on Virtual PC 2004 with Windows XP as the host OS. - Fedora Core 1 runs fine in VPC. - You can actually log in as root from the command line when the SEGV occurs. When you do, you get random SEGVs on basic commands like "cat", "file", etc.
Okay, I can't make any sense out of the prefdm error message - e.g. there is no line 80 in /etc/X11/prefdm You've also logged this against the gdm package, even though I don't see anything immediately obvious to suggest that its a gdm problem. You'll need to do more analysis of the problem before anyone who can't test this on VPC can figure out a fix for whatever the problem is. Also, you may not be seeing a single problem, but several different ones. If you can figure out what the individual problems are and log bugs against the appropriate packages you'll have a much better change of getting the bugs fixed.
Hi Joerg, Chances are very likely that this is a VPC bug and not a Fedora bug. You may try upgrading VPC, or upgrading to FC3. If neither of those steps work, you may want to report this issue to the makers of VPC.
I agree that the issue is a problem in VPC. Apparently there is something wrong with the memory management - nothing i would be able to debug. Probably SELinux does something VPC cannot handle. Strange thing is that all other distros run flawlessly, no matter what kernel is used. And since VPC has been bought by M$ there will never ever be a fix for that. I gues i'll better take a look at SVista instead.