|
|
#1 | |
|
Registered User
Join Date: Jun 2010
Location: Australia
Posts: 28
|
Hi, when attempting to play Nexuiz 2.5.2 (or other games using the Darkplaces engine, for example Xonotic compiled from git) I regularly experience huge slowdowns (less than 1fps) or complete X hangs. Usually this happens when I try to task switch, such as by pressing alt+tab, but not always. Once X hangs, it is not possible to switch to a VT or otherwise remedy the situation except by killing Xorg over SSH.
Kernel 2.6.31, Xserver 1.7.7 GPU temp ~45C The first bug report was run whilst the hung Xorg is still running, the second is from after it was killed. |
|
|
|
|
|
|
#2 | |
|
Registered User
Join Date: Nov 2004
Location: Between the keyboard and the chair.
Posts: 490
|
can you read http://www.nvnews.net/vbulletin/showthread.php?t=46678 and generate a good report with the debug ( -logverbose 6 ) info as instructed after the issue appears ?
some info for the [mi] overflowing which may or may not be nVidia related |
|
|
|
|
|
|
#3 |
|
Registered User
Join Date: Jun 2010
Location: Australia
Posts: 28
|
Ok sure, next time it happens... usually X has to have been running for a while before it will happen (but not always).
EDIT: here's one with verbosity 6, after a hang (this one was recoverable without killing X) |
|
|
|
|
|
#4 | |
|
Registered User
Join Date: Jun 2010
Location: Australia
Posts: 28
|
Here we go, a nice hang this time, running sudo chvt over ssh doesn't work when it's like this, the command just hangs until ^C
|
|
|
|
|
|
|
#5 |
|
Registered User
Join Date: Jun 2010
Location: Australia
Posts: 28
|
Fixed an MTRR issue that CONFIG_MTRR_SANITIZER was not able to correct (ftp://ftp.cs.utoronto.ca/pub/hugh/), upgraded to Xserver 1.9.4 (from 1.7.7) and 270.26 but still getting X hangs here, this one caused some messages about recovering a GPU error, which is also displayed in nvidia-settings (count: 1, so far).
EDIT: upgraded from 2.6.31.13 to 2.6.37 also. Code:
[209779.760] (II) NVIDIA(0): The NVIDIA X driver has encountered an error; attempting to [209779.760] (II) NVIDIA(0): recover... [209779.770] (II) NVIDIA(0): Error recovery was successful. [209805.710] (II) Open ACPI successful (/var/run/acpid.socket) [209805.764] (II) NVIDIA(0): Setting mode "nvidia-auto-select" |
|
|
|
|
|
#6 | |
|
Registered User
Join Date: Nov 2004
Location: Between the keyboard and the chair.
Posts: 490
|
|
|
|
|
|
|
|
#7 |
|
Registered User
Join Date: Jun 2010
Location: Australia
Posts: 28
|
Maybe, but no "GPU errors" were ever reported under X server 1.7 or 1.6.
I don't really have any way of finding whether this is a hardware fault other than to ask nvidia (which doesn't seem to be helping) - the alternative is buy another card and hope it's more stable than this one. However, it took several years of driver updates for most of the major bugs and instability with this one to get fixed - so I'm not keen to get a more recent GPU and wait through all that again. Also, these cards are quite expensive in Australia, and I don't like spending money on graphics cards as they always seem so unreliable. My current GTX275 is actually a warranty return - when I first received it, it was DOA. If it did have a hardware problem during the warranty period, there is no way I could have known about it - the drivers during that period were so unstable there was no way to distinguish driver failure from hardware failure. |
|
|
|
|
|
#8 |
|
Registered User
Join Date: Jun 2010
Location: Australia
Posts: 28
|
Evening...
|
|
|
|
|
|
#9 |
|
Registered User
Join Date: Sep 2008
Posts: 25
|
Have you had any progress on this? I've just upgraded to a GTX580 and now I'm on the latest drivers and I'm getting what sounds like exactly the same issue, with two exceptions
- my hangs aren't caused by anything (except playing a 3d game - both native and wine) but I'm not trying to alt-tab out or anything like that - and they usually last for about 30 seconds and then go away. Sometimes they last until I ssh in and nuke the game. Sometimes it's enough just to load a game and X will freeze. |
|
|
|
|
|
#10 |
|
Registered User
Join Date: Nov 2004
Location: Between the keyboard and the chair.
Posts: 490
|
@Oliw: can you read http://www.nvnews.net/vbulletin/showthread.php?t=46678 and generate a good report with the debug ( -logverbose 6 ) info as instructed after the issue appears ?
|
|
|
|
|
|
#11 |
|
Registered User
Join Date: Jun 2010
Location: Australia
Posts: 28
|
OliW: No progress... It's been this way for years now, and NVIDIA have made no comment (I have also emailed them a bug report, and I tried one of these threads last year). When my card was a lot newer / had buggier drivers, it behaved more like yours - lots of temporary hangs (even when just web browsing), and it would not recover from some of the hangs.
This also seems quite similar http://www.nvnews.net/vbulletin/showthread.php?t=156372 I'm glad I haven't gone and bought a new card yet - it sounds like I could end up with exactly the same problem again. If NVIDIA can't make any response soon, I'll have little choice but to switch to ATI. I'd prefer to get a GTX560, they look like a great GPU, but I need to know it's going to work so I'm not stuck with this freezing crap again. |
|
|
|
|
|
#12 |
|
Registered User
Join Date: Nov 2004
Location: Between the keyboard and the chair.
Posts: 490
|
@bones_was_here: he had a different (solved) problem: http://www.nvnews.net/vbulletin/showthread.php?t=159771
|
|
|
|
![]() |
| Thread Tools | |
|
|