nVidia 1.0-7185 and kernel 2.6.23-rc1-mm1

[Oleg Osovitskiy]

I want to install legacy drivers 1.0-7185 to linux kernel 2.6.23-rc1-mm1. It won't compiles. I corrected drivers around kmem_cache_t, kmem_cache_create() and unregister_chrdev(). It compiles successfully. But when I try ti run module I got such warning:

WARNING: "__you_cannot_kmalloc_that_much"
[/usr/src/NVIDIA-Linux-x86-1.0-7185-pkg1/usr/src/nv/nvidia.ko] undefined!

And loading of the module failed!

I think it try to kmalloc very much of memory, but I can't find where it is.
Can You help me?

[netllama]

Please see the forum sticky posts.

[Oleg Osovitskiy]

Originally Posted by netllama

Please see the forum sticky posts.

I'm sorry. Did You mean that I should provide additional info (nvidia-installer.log etc) or I don't observe solution for my problem?

If the 1st, I'll provide additional info tonight, when I'll back to home.

[Oleg Osovitskiy]

Ok, I've gathered information. I extracted and patched drivers, so nvidia-installer.log is not interesting here. It shows other errors.

Info about my Linux and gcc:

Linux oleg 2.6.23-rc1-mm1 #1 PREEMPT Sat Jul 28 22:06:17 EEST 2007 i686 unknown unknown GNU/Linux


gcc (GCC) 3.3.6
Copyright (C) 2003 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

I've made this patch to original 1.0-7185 drivers:

diff -rNu ./NVIDIA-Linux-x86-1.0-7185-pkg1/usr/src/nv/nv-linux.h ./NVIDIA-Linux-x86-1.0-7185-oo1/usr/src/nv/nv-linux.h
--- ./NVIDIA-Linux-x86-1.0-7185-pkg1/usr/src/nv/nv-linux.h	2007-04-03 04:58:22.000000000 +0300
+++ ./NVIDIA-Linux-x86-1.0-7185-oo1/usr/src/nv/nv-linux.h	2007-07-30 19:46:32.000000000 +0300
@@ -492,7 +492,7 @@
 #define NV_KMEM_CACHE_CREATE(kmem_cache, name, type)            \
     {                                                           \
         kmem_cache = kmem_cache_create(name, sizeof(type),      \
-                        0, 0, NULL, NULL);                      \
+                        0, 0, NULL);                            \
     } 
 
 #define NV_KMEM_CACHE_DESTROY(kmem_cache)                       \
diff -rNu ./NVIDIA-Linux-x86-1.0-7185-pkg1/usr/src/nv/nv.c ./NVIDIA-Linux-x86-1.0-7185-oo1/usr/src/nv/nv.c
--- ./NVIDIA-Linux-x86-1.0-7185-pkg1/usr/src/nv/nv.c	2007-04-03 04:58:22.000000000 +0300
+++ ./NVIDIA-Linux-x86-1.0-7185-oo1/usr/src/nv/nv.c	2007-07-30 19:43:14.000000000 +0300
@@ -90,6 +90,7 @@
 int nv_use_cpa = 1;
 #endif
 
+typedef struct kmem_cache kmem_cache_t;
 static kmem_cache_t *nv_pte_t_cache = NULL;
 
 // allow an easy way to convert all debug printfs related to events
@@ -1463,8 +1464,7 @@
         if (apm_nv_dev[i] != NULL) pm_unregister(apm_nv_dev[i]);
 #endif
 
-    if (unregister_chrdev(nv_major, "nvidia") < 0)
-        nv_printf(NV_DBG_ERRORS, "NVRM: unregister nv chrdev failed\n");
+    unregister_chrdev(nv_major, "nvidia");
 
     for (i = 0; i < num_nv_devices; i++)
     {
@@ -1488,8 +1488,7 @@
 
     nv_printf(NV_DBG_INFO, "NVRM: nvidia_exit_module\n");
 
-    if (unregister_chrdev(nv_major, "nvidia") < 0)
-        nv_printf(NV_DBG_ERRORS, "NVRM: unregister nv chrdev failed\n");
+    unregister_chrdev(nv_major, "nvidia");
 
     for (i = 0; i < num_nv_devices; i++)
     {

And when I build module I've got errors shown in the attached file

[hmrct]

Exact same problem seen here with 2.6.23-rc2 kernel and 1.0-9639 driver.

[hmrct]

Ok... Upgrading gcc from 3.X to 4.1.2 fixed the problem for me. Of course, this means rebuilding your kernel...

[Oleg Osovitskiy]

Originally Posted by hmrct

Ok... Upgrading gcc from 3.X to 4.1.2 fixed the problem for me. Of course, this means rebuilding your kernel...

Thank You! I recently upgraded to Slackware 12.0, it has gcc-4.1.2 as default compiler. And not yet try to install nvidia drivers. I'll try it tonight and post results here. Rebuilding kernel it's not a problem =) Anyway vanilla slackware kernel don't know anything about reiser4.