| Summary: | x11-base/xorg-server-1.3.0.0-r6 - mtrr: 0xd0000000,0x10000000 overlaps existing 0xd0000000,0x1000000 with nvidia and nv drivers | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Adrian Dannberg <neustadt> |
| Component: | Current packages | Assignee: | Gentoo X packagers <x11> |
| Status: | RESOLVED NEEDINFO | ||
| Severity: | normal | CC: | mmokrejs |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| URL: | http://www.file-upload.net/download-1215236/russland.tar.gz.html | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: |
the page to visit
requested xorg.conf requested xorg.0.log |
||
|
Description
Adrian Dannberg
2008-10-29 11:50:03 UTC
Created attachment 170218 [details]
the page to visit
Please attach also your Xorg.0.log and xorg.conf, since similar video hardware and configuration might be necessary for reproducing the bug. Created attachment 170260 [details]
requested xorg.conf
Created attachment 170264 [details]
requested xorg.0.log
Ok, I see you are using the non-free Nvidia driver. There have been bugs Nvidia driver related to firefox 3 and compiz desktop effects; are you using compiz by any chance? If so, try installing the latest ~x86 version of x11-drivers/nvidia-driver -- they may have fixed it by now. hi, it doesnt seem to be a driver issue. I just tried the "nv" driver and X crashes as well. I couldn't get X running with the "vga" properly so i couldn't test that. to provide more info I decided to recompile x11-base/xorg-server and x11-base/xorg-x11 according to http://www.gentoo.org/proj/en/qa/backtraces.xml. the problem I have now is: how do I start X with gdb properly? If I just run "gdb X", I cant run firefox to make X crash. Or can I? If I run "gdb startx" or "gdb -e startx", it tells mit startx is not in executable format. So how do I get the backtrace? You could try first to get a backtrace from firefox. dmesg output might also help figure out what is going on. ok i did a backtrace of firefox now. it looks like this: quit Starting program: /usr/lib/mozilla-firefox/firefox (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) [Thread debugging using libthread_db enabled] [New Thread 0xb7c526d0 (LWP 6075)] (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) [New Thread 0xb5f63b90 (LWP 6078)] [New Thread 0xb54beb90 (LWP 6079)] [New Thread 0xb4483b90 (LWP 6080)] [New Thread 0xb3affb90 (LWP 6081)] [Thread 0xb3affb90 (LWP 6081) exited] [New Thread 0xb32feb90 (LWP 6082)] [Thread 0xb32feb90 (LWP 6082) exited] [New Thread 0xb32feb90 (LWP 6083)] [New Thread 0xb3affb90 (LWP 6084)] [New Thread 0xb275bb90 (LWP 6085)] [New Thread 0xb1d9cb90 (LWP 6088)] [New Thread 0xb159bb90 (LWP 6089)] [New Thread 0xb0d9ab90 (LWP 6090)] [New Thread 0xb0599b90 (LWP 6091)] [New Thread 0xafd98b90 (LWP 6092)] [Thread 0xb4483b90 (LWP 6080) exited] [Thread 0xb0599b90 (LWP 6091) exited] [Thread 0xb0d9ab90 (LWP 6090) exited] [Thread 0xb159bb90 (LWP 6089) exited] [New Thread 0xb4483b90 (LWP 6093)] [Thread 0xb4483b90 (LWP 6093) exited] [New Thread 0xb4483b90 (LWP 6094)] [New Thread 0xb0599b90 (LWP 6095)] [New Thread 0xb0d9ab90 (LWP 6096)] [New Thread 0xb159bb90 (LWP 6097)] [New Thread 0xaf3ffb90 (LWP 6098)] [New Thread 0xaebfeb90 (LWP 6099)] [Thread 0xafd98b90 (LWP 6092) exited] [Thread 0xb4483b90 (LWP 6094) exited] [Thread 0xb0599b90 (LWP 6095) exited] [Thread 0xb0d9ab90 (LWP 6096) exited] [Thread 0xb159bb90 (LWP 6097) exited] [Thread 0xaf3ffb90 (LWP 6098) exited] Program received signal SIGHUP, Hangup. [Switching to Thread 0xb5f63b90 (LWP 6078)] 0xb7efc424 in __kernel_vsyscall () Dmesg output when X with nvidia driver crashes: X[5655]: segfault at ffffff08 ip b6b864a3 sp bf9711f0 error 4 in nvidia_drv.so[b69ec000+326000] mtrr: 0xd0000000,0x10000000 overlaps existing 0xd0000000,0x1000000 Dmesg output when X with nv driver crashes: uvesafb: VBE get state call failed (eax=0x13, err=0) mtrr: 0xd0000000,0x10000000 overlaps existing 0xd0000000,0x1000000 I wonder why severity was changed to normal. The bug occurred again today for me, while surfing the web randomly. luckily i didnt lose any important data, but this bug has definitely to power to do so. i am talking about stable x86 here, so a lot of users could be harmed by this. shouldn't this be _at least_ major? I am sorry firefox-3 is not stable x86. Yet I think a lot of people out there are using ~x86 firefox. Could you try using the "vesa" driver instead? The "vga" one is busted and generally useless anyway. Thanks hi exactly same behavior and dmesg output with vesa driver. I tested x11-base/xorg-x11-7.4 now as well. It does not crash, but unfortunally I cant use it, because of broken xf86-input-keyboard drivers? When a keyboard was connected to my box (tried usb and ps/2) the system froze after starting X. (box became unpingable). Without a keyboard, X seemed to run fine. Tried xf86-input-driver-keyboard-1.3.0 and 1.3.1 Will probably give x11-7.3 a shot on the weekend. Again: if s.o. tells me how to start X through gdb, I can give you a backtrace. A few things: 1) we're planning to get xorg-server 1.5.x into stable Real Soon Now (tm) so it might be definitely worth to solve your upgrade path problems rather than fix 1.3 which has been dead for years. We've been keeping it alive for much too long and we need to pull the plug. 2) If you really want to debug Xorg, you'll need to re-build xorg-server and all your drivers with debug symbols. http://www.gentoo.org/proj/en/qa/backtraces.xml will show you how. As a package maintainer, I would definitely rather spend some time on #1 rather than #2. Thanks for testing > mtrr: 0xd0000000,0x10000000 overlaps existing 0xd0000000,0x1000000
This could be cause by having PAT enabled in the kernel. Many people have
problems with PAT being enabled. It is just too new.
Xorg-server 1.5.3-r5 is now stable. Please don't hesitate to reopen this bug if you can reproduce with an up-to-date system. Thanks |
