On 8/26/05, maxim wexler <[EMAIL PROTECTED]> wrote: > > Well, at least X works. What video card will you be > using?
Contrary to many folks preferences I got an inexpensive PCI-E 1x Radeon to try out. I've had pretty good luck with ATI before. Anyway, I'm going right down the same path you're having trouble with so we're still in this together. It would be thread hijacking to start another conversation about why NVidia is considered 'more open source'. It seems to me that when I emerge nvidia-kernel it doesn't build the driver from source so what's the difference? I do agree from the one NVidia card I got last week (AGP 4X for about $40) that NVidia installed easily and works well, but since I got the ATI fglrx driver working on my laptop I'm getting about 2X the glxgears results (meaningly) vs. what I'm getting from the NVidia. Anyway, I'm not worried about that as my use is audio recording and 3D gaming is but a 30 minute break time occurance here. > > > > Here's my guess. If you look at 2.6.12-gentoo-r9 the > > NVidia AGP > > support clearly says Nforce/Nforce2. My suspicion is > > that the AGP > > chipset support for this chipset (NForce4) isn't in > > Whoa! NForce4?! What about 2, 3? I'm looking at the > Asus User Guide, p xi, in the box, Chipset: NVIDIA > nForce3 250Gb You're right. Sorry. > > This is all new territory for me. for us... ;-) > > UPDATE: did a sync and was able to emerge the new > ati-drivers. This time fglrx.ko *was* generated and is > loaded at boot. But now I get > > $startx > (WW) fglrx: No matching Device section for instance > (BusID PCI:1:0:1) found # faq says ignore OK, since the device probably has two outputs for dual screens the system found them both. Look at lspci and you'll probably see devices at PCI:1:0:0 PCI:1:0:1 If the xorgconfig program put a line in somewhere in /etc/X11/xorg.conf telling the system that a screen or device was BusID 1:0:0 then (I guess...) that it's complaining because there isn't some similar version for the second device. Try commenting out the one that is there and then I think the system will accept the definition for both devices. > (EE) fglrx(0): [agp] unable to acquire AGP, error > "xf86_ENODEV" OK, so does this kernel have /dev/agpgart and AMD Opteron/Athlon64 on-CPU GART support built as modules? Have you loaded both of the modules? (modprobe agpgart and [[POSSIBLY - I'M GUESSING]] agp-amd64) > (EE) fglrx(0): cannot init AGP Again because it's still not happy about the way the kernel part of the AGP stuff is built or installed apparently. Here's how I load the ATI stuff on my laptop: [EMAIL PROTECTED] ~ $ cat /etc/modules.autoload.d/kernel-2.6 # /etc/modules.autoload.d/kernel-2.6: kernel modules to load when system boots.# $Header: /var/cvsroot/gentoo-src/rc-scripts/etc/modules.autoload.d/kernel-2.6,v 1.1 <SNIP> snd_atiixp realtime gid=408 any=1 sbp2 agpgart ati-agp fglrx [EMAIL PROTECTED] ~ $ > Could not init font path element > /usr/share/fonts/local/, removing from list! > > The fix according to the faq: I need to clean up the > sources make oldconfig etc, re-emerge the drivers. > > WRONG :( > You'll get there. Search out the real name of the AMD agp-gart driver if you haven't already. good luck, Mark -- gentoo-user@gentoo.org mailing list
