View Full Version : yet another poor soul with nvnet woes (running RH9)
Jasonofabitch
05-04-03, 01:20 AM
I'm running Red Hat 9 with the 2.4.20-8 kernel. I've tried using both the SRPM and the tar file from nVidia's site, and I can't seem to get either working for the life of me.
First I tried:
rpmbuild --rebuild <package name>
Then I tried (after unpacking the tar of course):
cd nvnet
make clean
make
make install
In either case, I'm getting errors. When using the SRPM I get this:
*** [nvnet.o] Error 1
Leaving directory '/usr/src/redhat/BUILD/nforce/nvnet'
*** [nvnet_make] Error 2
Bad exit status from /var/tmp/rpm-tmp.3756 (%build)
When going through the tar install I get this:
mkdir -p //lib/modules/2.4.20-8/kernel/drivers/net
install -b -m 644 nvnet.o //lib/modules/2.4.20-8/kernel/drivers/net
install: cannot stat 'nvnet.o': No such file or directory
make: *** Error 1
I'm at my wit's end. I've read countless posts on these forums and tried anything and everything I can think of. I'm quite the newbie when it comes to linux so this is starting to get rather frustrating. Can anyone give me any idea as to what I need to do to get this working?!
*Note* I forgot to mention before that I've added the line [i]alias eth0 nvnet to modules.conf.
pixelxeno
05-04-03, 08:22 AM
In all cases your error reporting is not enough. When it says
*** [nvnet.o] Error 1
it means the true error message is one or more lines above, and this error message is really necessary to help you out on this one ;-)
Brad Smith
05-04-03, 03:45 PM
When I try to rebuild the nforce source rpm I get the following:
nvcrush.c: In function `Nvaudio_mmapwrite':
nvcrush.c:426: warning: passing arg 1 of `remap_page_range_R2baf18f2' makes pointer from integer without a cast
nvcrush.c:426: incompatible type for argument 4 of `remap_page_range_R2baf18f2'
nvcrush.c:426: too few arguments to function `remap_page_range_R2baf18f2'
nvcrush.c: In function `Nvaudio_mmap':
nvcrush.c:465: warning: passing arg 1 of `remap_page_range_R2baf18f2' makes pointer from integer without a cast
nvcrush.c:465: incompatible type for argument 4 of `remap_page_range_R2baf18f2'
nvcrush.c:465: too few arguments to function `remap_page_range_R2baf18f2'
make[2]: *** [nvcrush.o] Error 1
make[2]: Leaving directory `/usr/src/redhat/BUILD/nforce/nvaudio'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/usr/src/redhat/BUILD/nforce/nvaudio'
make: *** [nvaudio_make] Error 2
error: Bad exit status from /var/tmp/rpm-tmp.89202 (%build)
The rpmbuild command is:
rpmbuild --rebuild NVIDIA_nforce-1.0-0256.src.rpm
I have not tried the tar file yet.
Thanks,
Brad
Jasonofabitch
05-04-03, 04:43 PM
Give me a few minute pixelxeno and I'll get back to you with the full error message. :)
Jasonofabitch
05-04-03, 09:53 PM
Ok, so that was far more than a few minutes. At any rate, there are a ton of warning messages with both the SRPM and the tar file. I couldn't begin to type them all here. The bottom line is I'm getting exactly the same type of errors that everyone else has been posting about and none of the solutions that have been posted have worked for me.
:(
pixelxeno
05-05-03, 05:21 AM
I suspect that the audio part of the nforce driver from nvidia is taking you down on this one. Try going into the "nvnet" subdirectory and doing a single "make" there. It should compile fine. Then, do make install.
The nforce driver from nvidia has yet a poor installation method. Maybe they'll come with a better installation system, like they do with the NVidia GLX driver.
In all cases, I think you can safely take the tarball and not the srpm because you only want the nvnet part of the driver.
Jasonofabitch
05-05-03, 06:13 PM
I tried going into just the nvnet directory with the tarball. I did:
make clean
make
make install
This didn't work. =/
pixelxeno
05-05-03, 06:32 PM
Still with a huge error message too big to be pasted here?
Jasonofabitch
05-08-03, 12:28 AM
No, it's not that it's too big to be pasted here... it's just that I can't get any networking set up, therefore I can't connect to the internet in order to access this forum. I would just copy/paste the error into a .txt file and throw it on a floppy, but the box doesn't have a floppy drive. I'm in quite a pickle. :/
I'd same problems with Mandrake, but after I installed kernel sources from install cd, I managed to install nforce drivers from SRPM.
Eci
I have just installed Mdk 9.1 on A7V8X
When I compile nvnet it gives a mass of text and error "cc -c -DLINUX -DMODULE -DEXPORT_SYMTAB -D__KERNEL__ -O -Wstrict-prototypes -DCONFIG_PM -fno-strict-aliasing -mpreferred-stack-boundary=2 -march=i686 -falign-functions=4 -DMODULE -I/usr/src/linux-2.4.21/include nvnet.c"
Still the quastion how to compile that module?
DeVioSiTY
05-22-03, 10:28 PM
the nvnet driver also fails to compile on my system. Here's the error:
/usr/src/nforce/nvnet# make
cc -c -Wall -DLINUX -DMODULE -DEXPORT_SYMTAB -D__KERNEL__ -O -Wstrict-prototypes -DCONFIG_PM -fno-strict-aliasing -mpreferred-stack-boundary=2 -march=i686 -falign-functions=4 -DMODULE -I/usr/src/linux-2.4.21/include nvnet.c
In file included from nvnet.h:20,
from nvnet.c:21:
/usr/src/linux-2.4.21/include/linux/module.h:21:34: linux/modversions.h: No such file or directory
make: *** [nvnet.o] Error 1
Is this only suppose to work as a module? I like to run a tight ship, so if you have module support disabled it makes a tighter kernel. Plus it can help keep the script kiddies at bay.
I have been able to get the object file created at one point, but now I'm not certain how to get it into the kernel. :(
Hi there.
I have overcamed that problem by downloading new nforce driver from http://download.nvidia.com/XFree86/nforce/1.0-0256/NVIDIA_nforce-1.0-0256.tar.gz
Then make; make install; insmod nvnet; /etc/init.d/netwrok start
It should work. At least it worked for my configuration.
Good luck!
DeVioSiTY
05-23-03, 01:49 PM
That is the only version I have tried, since I cannot find any version older or newer than that one. Still doesn't work with the same error. Thanks anyway. I'm glad to hear that it at least works for you.
I think my real problem is that I need some sort of kernel patch for their driver.
cc -c -Wall -DLINUX -DMODULE -DEXPORT_SYMTAB -D__KERNEL__ -O -Wstrict-prototypes -DCONFIG_PM -fno-strict-aliasing -mpreferred-stack-boundary=2 -march=i686 -falign-functions=4 -DMODULE -I/usr/src/linux-2.4.21/include nvnet.c
In file included from nvnet.h:20,
Looks like this is being compiled as a module (the -DMODULE directive). Since this isn't part of the standard kernel source tree, I expect a fair amount of work would be required to compile this into the kernel if it is even possible. In the past with binary only releases and other such things, I found it was just simpler to compile as many things into my kernel that I wanted and turn on module support for any third party things.
I have the remap_page_range_R2baf18f2 error fixed, see my thread:
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=11630
titled: error: remap_page_range_R2baf18f2
:baaa::
tkloppel
06-05-03, 01:52 AM
In redhat 9, getting the same issue as everyone else... tried make, make install on both 0248, and 0256 tars, got the same error each time.
Here is the ENTIRE error...
# make
make -C nvnet
make[1]: Entering directory `/root/Downloads/Setup/nforce/nforce/nvnet'
cc -c -Wall -DLINUX -DMODULE -DEXPORT_SYMTAB -D__KERNEL__ -O -Wstrict-prototypes -DCONFIG_PM -fno-strict-aliasing -mpreferred-stack-boundary=2 -march=i686 -malign-functions=4 -DMODULE -I/lib/modules/2.4.20-8/build/include nvnet.c
make[1]: cc: Command not found
make[1]: *** [nvnet.o] Error 127
make[1]: Leaving directory `/root/Downloads/Setup/nforce/nforce/nvnet'
make: *** [nvnet_make] Error 2
any ideas at all are appreciated :-)
tkloppel
06-05-03, 03:17 AM
If you didnt have gcc installed this is one of the problems.... After i installed gcc, i got a whole new set of errors.. very long list of them infact... anyways...
im trying to figure out how to get libc6-dev installed... on another forum this appears to have fixed the problem for another person... think that may fix it, but i cant find it heh...
yes, definitly looks like gcc is not installed in that post.
Also if you are doing rebuilds, you will need to install the dev rpms to get it to build, and everything that it depends upon.
Get the dev rpm that is on you CDROM, and then do something like this...
rpm -ivh glibc-devel-2.3.2-11.9.i386.rpm
and/or/whatever
rpm -ivh libstdc++-devel-3.2.2-5.i386.rpm
Tell me a bit more on what you are looking for....
I did these steps...
root @ sun 14 > locate libc6
/usr/lib/libstdc++-2-libc6.1-1-2.9.0.so
/usr/lib/libstdc++-libc6.1-1.so.2
/usr/lib/libstdc++-3-libc6.2-2-2.10.0.so
/usr/lib/libstdc++-libc6.2-2.so.3
root @ sun 16 > rpm -qf /usr/lib/libstdc++-2-libc6.1-1-2.9.0.so
compat-libstdc++-7.3-2.96.118
root @ sun 17 > locate libstd | grep rpm
/rh9/RedHat/RPMS/libstdc++-3.2.2-5.i386.rpm
/rh9/RedHat/RPMS/libstdc++-devel-3.2.2-5.i386.rpm
/rh9/RedHat/RPMS/compat-libstdc++-7.3-2.96.118.i386.rpm
/rh9/RedHat/RPMS/compat-libstdc++-devel-7.3-2.96.118.i386.rpm
where rh9 is my local directory where I manually copied the RPMs.
vBulletin® v3.7.1, Copyright ©2000-2012, Jelsoft Enterprises Ltd.