gtk-2.2.2 and garbled graphics

segflaunt · 06-11-03, 11:34 AM

Looks like the nvidia driver for linux causes the following problem:

Apparently switching back to the "nv" driver fixes it. Anyone else see this?

**Thunderbird** · 06-11-03, 11:48 AM

Indeed looks like a drawing problem. In the past disabling Xrender could help .. (or else software Xrender). Try playing with the option NoRenderAccel.

segflaunt · 06-11-03, 11:54 AM

More details: I have an athlon XP 1800+ running Gentoo Linux and -4363.

Oh, and my XF86Config:
http://www.enodev.com/XF86Config
and XFree86.0.log:
http://www.enodev.com/XFree86.0.log

My /proc/driver/nvidia/cards/0:

Model: GeForce4 Ti 4200
IRQ: 16
Video BIOS: 04.25.00.34.00
Card Type: AGP

...and finally, the module options:

options nvidia NVreg_EnableAGPSBA=1 NVreg_EnableAGPFW=1

segflaunt · 06-11-03, 12:00 PM

Quote:

Originally posted by Thunderbird
Indeed looks like a drawing problem. In the past disabling Xrender could help .. (or else software Xrender). Try playing with the option NoRenderAccel.

Do you mean

Code:

Option "RenderAccel" "false"

in /etc/X11/XF86Config?

**Thunderbird** · 06-11-03, 12:08 PM

Yes I meant that option.

segflaunt · 06-11-03, 12:44 PM

That did it. Any chance I can help in improving this extension through offering testing?

I guess we know gtk+-2.2.2 consistently triggers the bug, so that's a start.

SnapIT · 06-11-03, 01:10 PM

Quote:

Originally posted by segflaunt
That did it. Any chance I can help in improving this extension through offering testing?

I guess we know gtk+-2.2.2 consistently triggers the bug, so that's a start.

NO, NO, NO, the Nvidia drivers is the problem, if they were not the problem would persist with other drivers...

Blame those who hold responsibility for releasing pre-alpha code as working drivers...

segflaunt · 06-11-03, 02:15 PM

I believe you misunderstand. I was asking, actually, if i could be of help testing drivers.

Simply trying to avail myself to whatever the powers that be might need testers for.

Quote:

Originally posted by SnapIT
NO, NO, NO, the Nvidia drivers is the problem, if they were not the problem would persist with other drivers...

Blame those who hold responsibility for releasing pre-alpha code as working drivers...

SnapIT · 06-11-03, 02:23 PM

Quote:

Originally posted by segflaunt
I believe you misunderstand. I was asking, actually, if i could be of help testing drivers.

Simply trying to avail myself to whatever the powers that be might need testers for.

They don't need any testers... they will just release new pre-alpha code with new problems, this IS the problem...

You think Nvidia cares enough about the Linux community to release drivers... HA!

They USE the Linux community to develop drivers...

Nvidia really couldn't care less what works on linux or not, they really, really do not care, they use the knowledge to develop their Quadro OGL drivers...

That is why all of us are unpaid programmers with suggestions, no more suggestions from me...

Nvidia either shows their support by releasing Open Source drivers, or they lure you into testing out their pre-alpha drivers some more, reporting the problems and praising them for releasing BS drivers...

Great job Nvidia...

At0mic · 06-11-03, 02:47 PM

Cool it man and stop trolling. We are just trying to get a bug fixed.

This is far from a fatal bug.

(feeding the troll)
Helping them develop their drivers? Hmm.. seeing how XFree is a _totally_ different windowing system than Windows and like Windows has it's own quirks - I _highly_ doubt this is the situation.

SnapIT · 06-11-03, 02:59 PM

Quote:

Originally posted by At0mic
Cool it man and stop trolling. We are just trying to get a bug fixed.

(feeding the troll)
Helping them develop their drivers? Hmm.. seeing how XFree is a _totally_ different windowing system than Windows and like Windows has it's own quirks - I _highly_ doubt this is the situation.

Ok, so assume you have no clue about OGL at all, obviously...

OGL stays the same, the same sets of commands and the same rendering no matter what Operating System you use, got that?

Then let's move on, as Nvidia very well knows Linux's OGL platform is one of the very best developed, Win's is just the same platform but a step behind, still following me? Then let's move ahead, if you can test your drivers for free on a community that are glad to get ANY drivers, to develop your drivers for the community that CRAVES good drivers... ehhh... why not? And that is what is happening...

When you accuse someone of trolling, as in this case, was i off topic? was i wrong?

I would expect an apology... but i doubt i will get one, you are not man enough to admit you don't know what you are talking about...

segflaunt · 06-11-03, 03:12 PM

I guess if the assumption is that NVidia owes us drivers that are production quality because they say they support linux, you *might* have a point, except that this feature is marked as experimental.

Now, regarding performance issues with some 4xxx linux drivers, that's a big issue.

The one thing they *should* do is release specs on the damn cards so the community could write drivers too. THAT's supporting Linux and open source.

In the meantime, I just want to try to make the best of the situation.

06-11-03, 11:34 AM	#1
segflaunt Registered User Join Date: Jun 2003 Posts: 16	gtk-2.2.2 and garbled graphics Looks like the nvidia driver for linux causes the following problem: Apparently switching back to the "nv" driver fixes it. Anyone else see this?

06-11-03, 11:54 AM	#3
segflaunt Registered User Join Date: Jun 2003 Posts: 16	More details: I have an athlon XP 1800+ running Gentoo Linux and -4363. Oh, and my XF86Config: http://www.enodev.com/XF86Config and XFree86.0.log: http://www.enodev.com/XFree86.0.log My /proc/driver/nvidia/cards/0: Model: GeForce4 Ti 4200 IRQ: 16 Video BIOS: 04.25.00.34.00 Card Type: AGP ...and finally, the module options: options nvidia NVreg_EnableAGPSBA=1 NVreg_EnableAGPFW=1

06-11-03, 12:44 PM	#6
segflaunt Registered User Join Date: Jun 2003 Posts: 16	Cool That did it. Any chance I can help in improving this extension through offering testing? I guess we know gtk+-2.2.2 consistently triggers the bug, so that's a start.

06-11-03, 03:12 PM	#12
segflaunt Registered User Join Date: Jun 2003 Posts: 16	Nevertheless I guess if the assumption is that NVidia owes us drivers that are production quality because they say they support linux, you might have a point, except that this feature is marked as experimental. Now, regarding performance issues with some 4xxx linux drivers, that's a big issue. The one thing they should do is release specs on the damn cards so the community could write drivers too. THAT's supporting Linux and open source. In the meantime, I just want to try to make the best of the situation.

06-11-03, 11:48 AM	#2
Thunderbird Join Date: Jul 2002 Location: Netherlands, Europe Posts: 2,105	Indeed looks like a drawing problem. In the past disabling Xrender could help .. (or else software Xrender). Try playing with the option NoRenderAccel.

06-11-03, 12:08 PM	#5
Thunderbird Join Date: Jul 2002 Location: Netherlands, Europe Posts: 2,105	Yes I meant that option.

06-11-03, 02:47 PM	#10
At0mic Registered User Join Date: Jun 2003 Posts: 2	Cool it man and stop trolling. We are just trying to get a bug fixed. This is far from a fatal bug. (feeding the troll) Helping them develop their drivers? Hmm.. seeing how XFree is a _totally_ different windowing system than Windows and like Windows has it's own quirks - I _highly_ doubt this is the situation.