Unable to correct driver from IA32 to AMD64

[Vindane]

I recently have been told that I've been running the wrong driver for my setup. The driver I have been using is IA32. I have an AMD Athlon 1GHZ processor with Nvidia GeForce 2 MX/200.

Well, I downloaded the kernel file and GLX file, unpacked them into two different directories and followed the README instructions and typed "make" in the kernel file directory and this is where it got me....



# make
cc -c -Wall -Wimplicit -Wreturn-type -Wswitch -Wformat -Wchar-subscripts -Wparentheses -Wpointer-arith -Wcast-qual -Wno-multichar -O -MD -D__KERNEL__ -DMODULE
-D_LOOSE_KERNEL_NAMES -mcmodel=kernel -DNTRM -D_GNU_SOURCE -D_LOOSE_KERNEL_NAMES -D__KERNEL__ -DMODULE -DNV_MAJOR_VERSION=1 -DNV_MINOR_VERSION=0 -DNV_PATCHLEVEL=4180 -DNV_UNIX -DNV_LINUX -DNV_INT64_OK -DNV_64_BITS -DNVCPU_X86_64 -I. -I/lib/modules/2.4.20-19.9/build/include -Wno-cast-qual nv.c
cc1: code model `kernel' not supported in the 32 bit mode
In file included from /lib/modules/2.4.20-19.9/build/include/linux/vmalloc.h:8,
from nv-linux.h:64,
from nv.c:14:
/lib/modules/2.4.20-19.9/build/include/linux/highmem.h: In function `bh_kmap':
/lib/modules/2.4.20-19.9/build/include/linux/highmem.h:20: warning: pointer of type `void *' used in arithmetic
nv.c: In function `nv_kern_mmap':
nv.c:1282: warning: passing arg 1 of `remap_page_range_Rc414bdc2' makes pointer
from integer without a cast
nv.c:1282: incompatible type for argument 4 of `remap_page_range_Rc414bdc2'
nv.c:1282: too few arguments to function `remap_page_range_Rc414bdc2'
nv.c:1301: warning: passing arg 1 of `remap_page_range_Rc414bdc2' makes pointer
from integer without a cast
nv.c:1301: incompatible type for argument 4 of `remap_page_range_Rc414bdc2'
nv.c:1301: too few arguments to function `remap_page_range_Rc414bdc2'
nv.c:1387: warning: passing arg 1 of `remap_page_range_Rc414bdc2' makes pointer
from integer without a cast
nv.c:1387: incompatible type for argument 4 of `remap_page_range_Rc414bdc2'
nv.c:1387: too few arguments to function `remap_page_range_Rc414bdc2'
nv.c: In function `nv_get_phys_address':
nv.c:2250: warning: implicit declaration of function `pte_offset'
nv.c:2250: invalid type argument of `unary *'
nv.c: In function `nv_agp_translate_address':
nv.c:2756: warning: cast from pointer to integer of different size
make: *** [nv.o] Error 1


When that failed, I went ahead into the GLX directory and tried to run "make" there but I got "permission denied". Also, I ran this whole process as root.

Thanks for any input!

Vindane

[leibold]

For an Athlon XP cpu you need the IA32 driver.

Even for the Opteron (which is an AMD 64-bit cpu) you still need the IA32 drivers as long as you run a 32-bit operating system on it.

The only time to use the AMD64 driver is if you are using an AMD 64-bit cpu (Opteron or the upcoming Athlon64) *AND* an x86-64 based operating system (e.g. SuSE Linux: ftp://ftp.suse.com/pub/suse/x86-64).

[Vindane]

See, this friend of mine told me that the "I" in IA32 stood for Intel. That's why I went for the AMD64. Also, I do believe that he mentioned that I run a 64-bit operating system (Red Hat 9). I could have gotten that mixed up though. This friend is a pretty knowledgable programmer, also pretty well-versed in Linux. But like I said, I could have gotten things mixed up that he told me.

Well, all of this curiosity about needing a different graphics driver comes from some problems I've been experiencing with Half-Life. I'm running Half-Life under Wine and it works just fine except one thing. When i'm in any part of the game that has a free-moving pointer, I can pull the pointer down to the bottom of the screen and actually make the HL screen "roll" up and I have access to most of the desktop. But the resolution is all screwed up when I do this. Basically, if you were to split up the screen in 4 square pieces, the HL screen takes up the upper-left piece. But once I'm actually IN the game and I'm using the mouse to move around and there's no free-moving pointer, it stays put.

I don't know. I've never seen anything like this before. Maybe I need to adjust something in my XF86Config file?