View Single Post
Old 11-19-05, 09:57 PM   #5
ckitching
Registered User
 
Join Date: Nov 2005
Posts: 2
Default Re: Nvidia support for Xen

If you try to use the current drivers on a Xen0 kernel, you just get an error.

Code:
nvidia: module license 'NVIDIA' taints kernel.
PCI: Obtained IRQ 16 for device 0000:01:00.0
NVRM: loading NVIDIA Linux x86 NVIDIA Kernel Module  1.0-7676  Fri Jul 29 12:58:54 PDT 2005
general protection fault: 0000 [#1]
PREEMPT
Modules linked in: nvidia nfsd exportfs lockd sunrpc ipv6 iptable_filter ip_tables bridge 3c59x bt878 snd_bt87x e100 mii snd_intel8x0 snd_ac97_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd soundcore snd_page_alloc hw_random evdev tsdev ehci_hcd usbhid uhci_hcd usbcore bttv video_buf firmware_class v4l2_common btcx_risc tveeprom videodev tvaudio tuner ide_cd cdrom ext3 jbd mbcache reiserfs ide_disk ide_generic via82cxxx trm290 triflex slc90e66 sis5513 siimage serverworks sc1200 rz1000 piix pdc202xx_old pdc202xx_new opti621 ns87415 hpt366 hpt34x generic cy82c693 cs5530 cs5520 cmd64x atiixp amd74xx alim15x3 aec62xx unix ide_core
CPU:    0
EIP:    0061:[<cb28bd3e>]    Tainted: P      VLI
EFLAGS: 00011206   (2.6.11.12-xen0)
EIP is at __check_pat_support+0x5d/0xbb [nvidia]
eax: 3febd3f1   ebx: 00000001   ecx: 00000277   edx: 00000000
esi: cb3f2500   edi: 00000000   ebp: c92b0000   esp: c92b1f4c
ds: 007b   es: 007b   ss: 0069
Process insmod (pid: 4529, threadinfo=c92b0000 task=c89c1020)
Stack: c8a9ff00 c8a9fe00 c3b24800 00000001 cb3f2500 00000000 cb28be90 00000000
       00000001 cb3f2500 00000000 00000001 cab4e2b0 cb3f3680 cb2c4965 cb2c495e
       c92b1f98 00000000 00000000 ffffffff c92b0000 00000001 c0134b94 c03a8848
Call Trace:
 [<cb28be90>] __nv_enable_pat_support+0x15/0x93 [nvidia]
 [<cab4e2b0>] nvidia_init_module+0x2b0/0x38e [nvidia]
 [<c0134b94>] sys_init_module+0x184/0x230
 [<c0109258>] syscall_call+0x7/0xb
Code: 04 24 04 00 00 00 e8 10 43 00 00 c7 44 24 04 40 2b 2c cb c7 04 24 04 00 00 00 e8 fc 42 00 00 b8 00 00 00 00 eb 5c b9 77 02 00 00 <0f> 32 89 d7 89 c3 ba 00 00 00 00 be ff 00 00 00 8d 0c d5 00 00
Editing nv.c to set nv_disable_pat to 1 (because the install refuses to proceed if nvidia.ko refuses to load. After installation, I can load an unedited driver by passing nv_disable_pat=1 to insmod), and the module loads, but starting X gives more protection faults

Code:
general protection fault: 0000 [#1]
PREEMPT
Modules linked in: nvidia ipv6 nfsd exportfs lockd sunrpc iptable_filter ip_tables bridge 3c59x bt878 snd_bt87x e100 mii snd_intel8x0 snd_ac97_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_timer snd soundcore snd_page_alloc hw_random ehci_hcd tsdev evdev usbhid uhci_hcd usbcore bttv video_buf firmware_class v4l2_common btcx_risc tveeprom videodev tvaudio tuner ide_cd cdrom ext3 jbd mbcache reiserfs ide_disk ide_generic via82cxxx trm290 triflex slc90e66 sis5513 siimage serverworks sc1200 rz1000 piix pdc202xx_old pdc202xx_new opti621 ns87415 hpt366 hpt34x generic cy82c693 cs5530 cs5520 cmd64x atiixp amd74xx alim15x3 aec62xx unix ide_core
CPU:    0
EIP:    0061:[<cb0c0b80>]    Tainted: P      VLI
EFLAGS: 00013297   (2.6.11.12-xen0)
EIP is at _nv002495rm+0x0/0x10 [nvidia]
eax: cb0c0b80   ebx: c4ad5cd8   ecx: 00000000   edx: 00000009
esi: cb420700   edi: c8d5d800   ebp: c4ad5cf4   esp: c4ad5c98
ds: 007b   es: 007b   ss: 0069
Process Xorg (pid: 6364, threadinfo=c4ad4000 task=c3c47520)
Stack: cb0a6e3d c8d5d800 c4ad5ccc c4ad5cd0 cb0b9d6d c0e21800 c1236c00 cb421780
       cb0b9dc4 c4ad5d40 cb2dcbd0 c4ad5cf4 00010809 00000000 3febfbff 00000f24
       756e6547 49656e69 6c65746e 3febfbff 00000000 0004020f 00000009 c4ad5d34
Call Trace:
 [<cb0a6e3d>] _nv001458rm+0x139/0x31c [nvidia]
 [<cb0b9d6d>] _nv001615rm+0x3d/0x6c [nvidia]
 [<cb0b9dc4>] _nv002038rm+0x28/0x70 [nvidia]
 [<cb0bac0d>] _nv003614rm+0x6d/0x138 [nvidia]
 [<cb23a310>] _nv003293rm+0x68/0x9c [nvidia]
 [<cb0abe31>] _nv001126rm+0x15/0x24 [nvidia]
 [<cb23a0d7>] _nv003251rm+0x73/0x94 [nvidia]
 [<cb23a0ca>] _nv003251rm+0x66/0x94 [nvidia]
 [<cb0b7fae>] _nv002098rm+0x16/0x1c [nvidia]
 [<cb1e75fa>] _nv004547rm+0x1d6/0x1e4 [nvidia]
 [<cb1e7454>] _nv004547rm+0x30/0x1e4 [nvidia]
 [<cb1e6159>] _nv004398rm+0xa9/0xe8 [nvidia]
 [<cb1e7201>] _nv004376rm+0x79/0xe0 [nvidia]
 [<cb1e71f7>] _nv004376rm+0x6f/0xe0 [nvidia]
 [<cb0ab14a>] _nv002191rm+0x36/0xe0 [nvidia]
 [<cb1919bb>] _nv000398rm+0x1b/0x7c [nvidia]
 [<cb0baa0f>] _nv001455rm+0x2f/0x150 [nvidia]
 [<cb0baa4b>] _nv001455rm+0x6b/0x150 [nvidia]
 [<cb0bd811>] _nv001428rm+0x91/0xac [nvidia]
 [<cb0bd7eb>] _nv001428rm+0x6b/0xac [nvidia]
 [<cb0bb418>] _nv001460rm+0x270/0x3a4 [nvidia]
 [<cb0bb408>] _nv001460rm+0x260/0x3a4 [nvidia]
 [<cb0be81e>] rm_init_adapter+0x5e/0x90 [nvidia]
 [<cb0be812>] rm_init_adapter+0x52/0x90 [nvidia]
 [<cb2ba498>] nv_kern_open+0x1ae/0x23a [nvidia]
 [<cb2baf74>] nv_kern_isr+0x0/0x6b [nvidia]
 [<cb2ba2ea>] nv_kern_open+0x0/0x23a [nvidia]
 [<c0162d6a>] chrdev_open+0xea/0x1d0
 [<c015864a>] dentry_open+0x15a/0x240
 [<c01584e8>] filp_open+0x68/0x70
 [<c01583b9>] sys_chown+0x59/0x60
 [<c015878e>] get_unused_fd+0x5e/0xd0
 [<c01588c9>] sys_open+0x49/0x90
 [<c0109258>] syscall_call+0x7/0xb
Code: 00 dc 90 0b cb 89 ec 5d c3 89 f6 55 89 e5 89 ec 5d c3 90 90 90 90 90 90 90 90 90 0f 31 c3 8d b6 00 00 00 00 8d bc 27 00 00 00 00 <0f> 20 e0 c3 8d b6 00 00 00 00 8d bf 00 00 00 00 0f 20 d8 31 d2
The system doesn't completely hang at this point, but the video is gone until reboot.

This is driver 7676 on Ubuntu Breezy 5.10 on a custom Xen0 kernel (2.6.11.12-xen0). The nv driver works fine with X (although with considerably fewer features). The card is a Geforce 2 MX.
ckitching is offline   Reply With Quote