nV News Forums


nV News Forums (http://www.nvnews.net/vbulletin/index.php)
-   NVIDIA Linux (http://www.nvnews.net/vbulletin/forumdisplay.php?f=14)
-   -   RenderAccel Problems (http://www.nvnews.net/vbulletin/showthread.php?t=36600)

wizdom 09-10-04 08:14 PM

RenderAccel Problems
I've already sent an email to linux-bugs@nvidia.com, no response so far.. but I thought maybe the forum would have some suggestions.

It's easy to see that a lot of people experiencing lockups with the NVIDIA driver that aren't chipset/agpgart/SBA/whatever related also note that disabling RenderAccel in their XF86Config/xorg.conf fixes the problem..

I've raised this issue about 4 times now. The first time it was on GeForce 2 Ti hardware. Now it's GeForce 4 Ti. Same difference. Adding:

Option "RenderAccel" "1"

To the Driver section of the X config causes my machine to randomly lock up, if the render extension is used heavily. Some people claim to have no problems with it and -- I'll admit -- since the 4xxx drivers I've been able to leave it enabled pretty much all the time, with only the occasional lockup.

However, like many people here (I'm sure), I downloaded the latest Xorg 6.8.0 release to try out the new (admittedly experimental) Composite extension, using the crappy xcompmgr utility. I got the utility installed and the extension registered, and all seemed to be good.

If I enable drop shadows with xcompmgr -cf, moving the window around rapidly locks up the X server. 100% CPU. I have to ssh into the machine, kill X with kill -9 `pidof X` and restart X. Disabling RenderAccel allows me to enable drop shadows, but it's all done in software then and is horrendously slow.

I did what other people here have suggested and strace'd my X binary. I did the following to test it. I started up X, with:

~$ strace -o ~/xtrace X

Then I switched back to vt7 and ran:

~$ xterm -display :0

Then I switched back to vt12 (X) and ran:

~# twm &

Then I ran xcompmgr:

~# xcompmgr -cf

Then I just randomly clicked outside of the terminal to bring up repeatedly the twm menu. Locked up in about 2 clicks. Reproduced the bug!

In ~/xtrace I see the log terminating with this, repeating indefinitely:

--- SIGALRM (Alarm clock) @ 0 (0) ---
sigreturn() = ? (mask now [])
--- SIGALRM (Alarm clock) @ 0 (0) ---
sigreturn() = ? (mask now [])

(The gettimeofday() "DoS"ing somebody was talking about elsewhere on the forum is bogus; it does this when RenderAccel isn't enabled, it seems to be the defacto way X functions. The above, however, does not happen.)

I simply cannot believe that a problem that has been so long lived, that is clearly software (I can kill X and restart; no problem), and in acceleration which will be especially relevant to users in future Xorg releases, has been ignored for so long!

Please NVIDIA, get somebody to fix it!

kmare 09-10-04 10:07 PM

Re: RenderAccel Problems
hmmm... interesting results... anyway, indeed I don't have lockups... but with xorg 6.8 I have problems when running xcompmgr with the f option (it's about fading, isn't it?). Running "xcompmgr -c" is fine. Well sometimes some transparent windows become a translucent rectangle with nothing inside, but minimizing and restoring the window does the trick. As far as I know the nvidia developers are working on drivers that will fuction properly with composite. I just hope soon... :)

PS: I got some problems resizing a window with shadows and translucency... it is really slow and most of the time the window doesn't get filled (some areas are just translucent, the minimize/restore does trick though)

PrakashP 09-11-04 02:13 AM

Re: RenderAccel Problems
COuld your try using agpgart, acpi pm timer and fast writes off? Works for me with composition and renderaccel w/o problems.

Only minor cosmetiv thing is that with render accel (even with xfree 4.3) some fonts in mozilla firefox/thunderbird which have "special" characters sometimes get screwed up. Moving the mouse over it/marking it, fixes it.

ViRgiLiO 09-11-04 04:05 AM

Re: RenderAccel Problems
AGPGART+FastWrite disable seems to be a stable combination. Almost two days without any frozen screen!

dpw2atox 09-11-04 12:38 PM

Re: RenderAccel Problems
well for me on my 5900xt renderaccel is relativly stable. however if im browsing a large text document and I scroll back up the text in the document gets blured and i have managed to crash X about 2x with renderaccel enabled. i am emailing the bug group on this issue.

wizdom 09-11-04 01:33 PM

Re: RenderAccel Problems
Please, no more suggestions for AGPGART, SBA or other chipset related settings..

The crash is in the X software driver itself, it's got nothing to do with my hardware settings.

This has happened since 2.4.8 for me, on a variety of different machines.

For what it's worth, and I really think it's worth very little, I'm running 2 AGPGART/AGP 8x/SBA off.

This is perfectly, 100% stable for all 3D functions.

Why would a 2D renderaccel lockup depend on SBA or AGP settings?

dpw2atox 09-11-04 03:00 PM

Re: RenderAccel Problems
honestly it sounds like its justa bug with the renderaccel code. before for me it used to make everything corupt on my screen before the 6x driver series so it is getting better. I think that for now just run with it disabled and wait for the next driver release and try it then.

wizdom 09-11-04 07:19 PM

Re: RenderAccel Problems
I think it's clear they're not aware of the problem, else it would have been fixed by now..

I'd rather not "wait" for a fix I've been "waiting" two years for.

Keeping this post at the top of the forum might allow an NVIDIA employee concerned with the quality of their driver, a chance to fix the problem.

wizdom 09-12-04 11:54 AM

Re: RenderAccel Problems

kmike 09-13-04 03:44 AM

Re: RenderAccel Problems
From 61.11 driver's README.txt: Option "RenderAccel" ... THIS OPTION IS EXPERIMENTAL. ENABLE IT AT YOUR OWN RISK.
That disclaimer was there since when I started to use nVidia hardware on linux. Given that, I don't think nVidia is going to fix RENDER support.

Cartman 09-13-04 05:27 AM

Re: RenderAccel Problems

Originally Posted by kmike
From 61.11 driver's README.txt: Option "RenderAccel" ... THIS OPTION IS EXPERIMENTAL. ENABLE IT AT YOUR OWN RISK.
That disclaimer was there since when I started to use nVidia hardware on linux. Given that, I don't think nVidia is going to fix RENDER support.

There is now RENDER Testsuite in X.org cvs. So NVidia can fix RenderAccel support.

wizdom 09-13-04 07:47 AM

Re: RenderAccel Problems
It's said that for ages. Their windows drivers don't depend on software acceleration, why the hell shouldn't a 99% working extension be fixed?

Wise up, the warning's there probably because they know there's some horrible hack in place. I'm just rekindling the fire.

All times are GMT -5. The time now is 03:00 AM.

Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.