Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 03-17-05, 12:22 AM   #145
kcrudup
Registered User
 
Join Date: Apr 2004
Location: Los Angeles, CA 90034
Posts: 96
Default Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by kcrudup
I was able to get [ swsusp2 ] to work, by enabling the option "UseDummyXServer" in hibernate.conf . Using 2.6.9, with ACPI on, NvAGP=0 in /etc/X11/XF86Config, APM off, and AGP turned off in the kernel config, I've been able to resume perfectly with "swsusp2" reliably with 7167
FWIW, I've been able to use "NvAGP=1" (but AGPGART off in the kernel, of course) with swsusp as well.
kcrudup is offline   Reply With Quote
Old 03-17-05, 01:27 AM   #146
warlock_ba
Registered User
 
Join Date: Mar 2004
Location: Townsville
Posts: 25
Send a message via ICQ to warlock_ba
Lightbulb Re: 1.0-7167 for Linux x86 released

I managed to fix the hangup of the X due to the new nVidia drivers. I installed the SMP version of the kernel, not the usual one, allthoug I have a single processor, and this fixed this issue it seems ...
Hope this helps ...
If any of the linux kernel experts is online, could someone tell me what impact has a SMP kernel on a PC? I mean, is it bad to put a SMP kernel on a linux ? What's the drawback, if any ? Since i don't believe that if there were no difference, we wouldn't use all the SMP kernel?

Regards,
Me
warlock_ba is offline   Reply With Quote
Old 03-17-05, 02:06 AM   #147
Thetargos
Registered User
 
Join Date: Dec 2004
Posts: 52
Send a message via ICQ to Thetargos Send a message via AIM to Thetargos
Default Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by warlock_ba
I managed to fix the hangup of the X due to the new nVidia drivers. I installed the SMP version of the kernel, not the usual one, allthoug I have a single processor, and this fixed this issue it seems ...
Hope this helps ...
If any of the linux kernel experts is online, could someone tell me what impact has a SMP kernel on a PC? I mean, is it bad to put a SMP kernel on a linux ? What's the drawback, if any ? Since i don't believe that if there were no difference, we wouldn't use all the SMP kernel?

Regards,
Me
SMP stands for Symetric MultiProccesing. In other words support for multiple processors... I can't say why installing an SMP kernel onto an uniprocessor system had such a posstive impact... However if you happen to have a Pentium IV with hyperthreading, an SMP kernel will actually make use of the HT feature on such processors...
Thetargos is offline   Reply With Quote
Old 03-17-05, 05:42 AM   #148
warlock_ba
Registered User
 
Join Date: Mar 2004
Location: Townsville
Posts: 25
Send a message via ICQ to warlock_ba
Default Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by Thetargos
SMP stands for Symetric MultiProccesing. In other words support for multiple processors... I can't say why installing an SMP kernel onto an uniprocessor system had such a posstive impact... However if you happen to have a Pentium IV with hyperthreading, an SMP kernel will actually make use of the HT feature on such processors...
I know what SMP stands for. But what I wanted to know was what impact does it have installing such a kernel on a single processor machine, like working slower, using more memory, crashing all the time ? Since i saw that the new driver works now ok with it. And on the single chip kernel it didn't. So it improved something, but i didn't have time to check for sideeffects yet ...
warlock_ba is offline   Reply With Quote
Old 03-17-05, 08:42 AM   #149
mackey
Registered User
 
Join Date: Mar 2005
Posts: 2
Default Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by Vic-20
With 7167 I can get to KDM login, but after I login X freezes before the KDE splash screen appears. The mouse still moves, but keyboard or mouse clicks are ignored. Have to reboot to recover.
I see the exact same problem on my laptop. It's a Dell Inspiron 8000 with GeForce 2 Go. I am running Debian unstable with Debian packaged kernel 2.6.10. Nothing shows up on logs, ksplash hangs and keyboard doesn't work, but mouse cursor still moves. Luckily, 6629 works fine.

Update: After disabling RenderAccel, 7167 seems to work. This is with XFree86 4.3.0.1.

Last edited by mackey; 03-17-05 at 09:01 AM.
mackey is offline   Reply With Quote
Old 03-17-05, 09:39 AM   #150
warlock_ba
Registered User
 
Join Date: Mar 2004
Location: Townsville
Posts: 25
Send a message via ICQ to warlock_ba
Default Re: 1.0-7167 for Linux x86 released

Ok: So , for me single thing that worked was putting up a smp kernel. It didn't help to patch it. It didn't help to Disable/Enable any god forsaken option Nothing, only SMP saved my thingie ... so it's not related to render accel either. There is something deeper within.
warlock_ba is offline   Reply With Quote
Old 03-17-05, 10:18 AM   #151
dave_maxwell
Registered User
 
Join Date: Mar 2005
Posts: 4
Default Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by warlock_ba
Ok: So , for me single thing that worked was putting up a smp kernel. It didn't help to patch it. It didn't help to Disable/Enable any god forsaken option Nothing, only SMP saved my thingie ... so it's not related to render accel either. There is something deeper within.

That's kinda funny. If you look at one of my earlier postings, you'll see that my experience on a machine with an SMP kernel was good and I had trouble with a plain jane uniprocessor kernel; normally I'd expect trouble with SMT/SMP first. It's just an HT uniprocessor P-IV box but the SMP kernel enables the HT goodness.

The really odd thing is that this problem doesn't truly lock the box, just X. I could always ssh in and restart X when it happens. Usually when the kernel is involved, the sucker is going to lock...hard.

For those having trouble let's summarize:

Uniprocessor kernel + Render Accell == bad
Uniprocessor kernel - Render Accell == good
SMP kernel + Render Accell == good (maybe)

I can also tell you that the performance impact of an SMP kernel on a uniprocessor box is minimal. The kernel is larger and will take up a tad more memory though. I'll definitely give your SMP on Uniprocessor trick a try and let the list know how it goes.

Hopefully, the nvidia gods have been reading all this. I would hope there is enough information to chase down the hoseup.
dave_maxwell is offline   Reply With Quote
Old 03-17-05, 10:36 AM   #152
Iskendar
Registered User
 
Join Date: Jul 2003
Posts: 29
Default Re: 1.0-7167 for Linux x86 released

Does it fix the Ti4200 TV-out bug?

I.
Iskendar is offline   Reply With Quote

Old 03-17-05, 10:53 AM   #153
IronMan
Registered User
 
Join Date: Dec 2004
Posts: 10
Default Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by dave_maxwell
Uniprocessor kernel + Render Accell == bad
Uniprocessor kernel - Render Accell == good
SMP kernel + Render Accell == good (maybe)
My computer is a dual Athlon MP system with Linux 2.6.11.4, Xorg 6.8.2, KDE 3.4.0 (final), and RenderAccel makes no difference with the 7167 driver (hard lockup). Everything works fine with the 6629 driver, so that's what I'm working with right now. BTW, video is BFG 5900XT OC.
IronMan is offline   Reply With Quote
Old 03-17-05, 01:35 PM   #154
korgull
Registered User
 
Join Date: Jan 2005
Posts: 20
Default Re: 1.0-7167 for Linux x86 released

I get the following using the 7167 driver for my FX5700 :

agpgart: Found an AGP 3.5 compliant device at 0000:00:00.0.
agpgart: X passes broken AGP3 flags (c). Fixed.
agpgart: Putting AGP V3 device at 0000:00:00.0 into 4x mode
agpgart: Putting AGP V3 device at 0000:01:00.0 into 4x mode
agpgart: Found an AGP 3.5 compliant device at 0000:00:00.0.
agpgart: X passes broken AGP3 flags (c). Fixed.
agpgart: Putting AGP V3 device at 0000:00:00.0 into 4x mode
agpgart: Putting AGP V3 device at 0000:01:00.0 into 4x mode

it is a 8x AGP card and the BIOS is also set to 8x.

Any way I can make this work 8x ?
korgull is offline   Reply With Quote
Old 03-17-05, 06:02 PM   #155
axobeauvi
axobeauvi
 
Join Date: Dec 2003
Location: home
Posts: 14
Default Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by korgull
I get the following using the 7167 driver for my FX5700 :

agpgart: Found an AGP 3.5 compliant device at 0000:00:00.0.
agpgart: X passes broken AGP3 flags (c). Fixed.
agpgart: Putting AGP V3 device at 0000:00:00.0 into 4x mode
agpgart: Putting AGP V3 device at 0000:01:00.0 into 4x mode
agpgart: Found an AGP 3.5 compliant device at 0000:00:00.0.
agpgart: X passes broken AGP3 flags (c). Fixed.
agpgart: Putting AGP V3 device at 0000:00:00.0 into 4x mode
agpgart: Putting AGP V3 device at 0000:01:00.0 into 4x mode

it is a 8x AGP card and the BIOS is also set to 8x.

Any way I can make this work 8x ?
just make sure you don't have a mobo with a VIA chipset
there are issues in putting AGP to full
axobeauvi is offline   Reply With Quote
Old 03-18-05, 02:00 AM   #156
bani
Registered User
 
Join Date: Aug 2004
Posts: 24
Exclamation Re: 1.0-7167 for Linux x86 released

Quote:
Originally Posted by Alphohsu Pugh
Zander,

--- Previous (Mar. 11):
<< Linux-GNU 2.6.10 on Intel P4 3.06GHz, TNT2 M64 32M, XFree86 4.4

7167 compiles without any problems BUT when I try to bring up XFree86:

FATAL: Module nvidia already in kernel.
(EE) NVIDIA(0): Failed to load the NVIDIA kernel module!
(EE) NVIDIA(0): *** Aborting *** >>

--- Zander:


You were right on the mark!
I changed my mit from 0.9.14 (Sep. 11, 2003) to 3.1 (Nov. 14, 2004)
(“What a Difference A Year Makes”, and “24 little hours” of upgrade).

Everything works perfectly now INCLUDING the SMP and HyperThreading.
I still cannot believe!

Many thanks to you and NVIDIA for an outstanding job.
nice. this fixed it for me too.

so the change from 6629 to 7167 requires a new module-init-tools, i didn't see that in any of the nvidia readmes though...
bani is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 08:36 AM.


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