I am using the latest version of Xen 4 from the mercurial repository,
and the Xen/stable kernel from git, with the latest Nvidia drivers
(195.36.15) and a GeForce 8600GT card (but same problem with inbuilt
8200GT)
I am aware that it is difficult to solve these problems because of the
limited information from Nvidia.
The problem is that although the module compiles and loads fine, it
crashes as soon as X is started, taking down all displays and the VTs,
leaving only ssh'ing into the machine possible.
The output in dmesg is:
[ 632.715475] nvidia 0000:02:00.0: setting latency timer to 64
[ 632.715480] vgaarb: device changed decodes:
PCI:0000:02:00.0,olddecodes=none,decodes=none:owns=none
[ 632.715645] nvidia 0000:03:00.0: setting latency timer to 64
[ 632.715648] vgaarb: device changed decodes:
PCI:0000:03:00.0,olddecodes=none,decodes=none:owns=none
[ 632.715762] NVRM: loading NVIDIA UNIX x86_64 Kernel Module
195.36.15 Fri Mar 12 00:29:13 PST 2010
[ 632.715822] NVRM: CPU does not support the PAT, falling back to MTRRs.
[ 667.196544] X: Corrupted page table at address 7f581d79d000
[ 667.196549] PGD b3123067 PUD 9c670067 PMD 9c760067 PTE fffffffffffff237
[ 667.196555] Bad pagetable: 000f [#1] SMP
[ 667.196558] last sysfs file:
/sys/devices/pci0000:00/0000:00:10.0/0000:03:00.0/resource
[ 667.196561] CPU 0
[ 667.196563] Modules linked in: nvidia(P) gspca_spca561 gspca_main
videodev v4l1_compat v4l2_compat_ioctl32 [last unloaded: nvidia]
[ 667.196573] Pid: 8105, comm: X Tainted: P W 2.6.32.10 #11 MS-7578
[ 667.196575] RIP: e033:[<00007f58175cc6be>] [<00007f58175cc6be>]
0x7f58175cc6be
[ 667.196579] RSP: e02b:00007fff981d96c0 EFLAGS: 00010202
[ 667.196581] RAX: 00007f581d79d000 RBX: 000000000156ad00 RCX: 0000000000000020
[ 667.196583] RDX: 0000000000000001 RSI: 0000000000000020 RDI: 000000000156ad00
[ 667.196585] RBP: 00007fff981d9790 R08: 0000000000000010 R09: 0101010101010101
[ 667.196587] R10: 00007fff981d9450 R11: 00007f581ae5d1e0 R12: 000000000156ad00
[ 667.196588] R13: 000000000156ad00 R14: 0000000000000001 R15: 0000000001568290
[ 667.196592] FS: 00007f581d7676f0(0000) GS:ffff8800023fb000(0000)
knlGS:0000000000000000
[ 667.196594] CS: e033 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 667.196596] CR2: 00007f581d79d000 CR3: 000000009d32c000 CR4: 0000000000000660
[ 667.196598] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 667.196602] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 667.196604] Process X (pid: 8105, threadinfo ffff88009d31e000, task
ffff8800b2d27260)
[ 667.196606]
[ 667.196607] RIP [<00007f58175cc6be>] 0x7f58175cc6be
[ 667.196609] RSP <00007fff981d96c0>
[ 667.196612] ---[ end trace 4eaa2a86a8e2da24 ]---
Xorg.log reports just a lack of a fully setup system:
(**) NVIDIA(0): Depth 24, (--) framebuffer bpp 32
(==) NVIDIA(0): RGB weight 888
(==) NVIDIA(0): Default visual is TrueColor
(==) NVIDIA(0): Using gamma correction (1.0, 1.0, 1.0)
(**) NVIDIA(0): Option "UseEDID" "False"
(**) NVIDIA(0): Option "ModeValidation" "DFP-0: NoMaxPClkCheck,
NoEdidMaxPClkCheck, AllowNon60HzDFPModes"
(**) Mar 24 21:51:50 NVIDIA(0): Enabling RENDER acceleration
(**) Mar 24 21:51:50 NVIDIA(0): Ignoring EDIDs
(II) Mar 24 21:51:50 NVIDIA(0): Support for GLX with the Damage and
Composite X extensions is
(II) Mar 24 21:51:50 NVIDIA(0): enabled.
(II) Mar 24 21:51:56 NVIDIA(GPU-0): Not probing EDID on CRT-0.
(II) Mar 24 21:51:56 NVIDIA(0): NVIDIA GPU GeForce 8600 GT (G84) at
PCI:3:0:0 (GPU-0)
(--) Mar 24 21:51:56 NVIDIA(0): Memory: 524288 kBytes
(--) Mar 24 21:51:56 NVIDIA(0): VideoBIOS: 60.84.5e.00.00
(II) Mar 24 21:51:56 NVIDIA(0): Detected PCI Express Link width: 16X
(--) Mar 24 21:51:56 NVIDIA(0): Interlaced video modes are supported on this GPU
(--) Mar 24 21:51:56 NVIDIA(0): Connected display device(s) on GeForce
8600 GT at PCI:3:0:0:
(--) Mar 24 21:51:56 NVIDIA(0): CRT-0
(--) Mar 24 21:51:56 NVIDIA(0): CRT-0: 400.0 MHz maximum pixel clock
The log then stops there as the NVIDIA module crashes, the next lines
that should be there are:
(II) Mar 24 22:31:44 NVIDIA(0): Assigned Display Device: CRT-0
(II) Mar 24 22:31:44 NVIDIA(0): Validated modes:
(II) Mar 24 22:31:44 NVIDIA(0): "1920x1200"
This suggests that the driver is calling some hard-coded location Xen
has taken over to attempt to try the chosen display mode (EDID is
disabled, so the driver has been told the specific resolution to use).
Using the NV drivers works, but is not feasible because this is a
workstation, so 3D acceleration is needed.
Is more information needed, or is there enough here to attempt to find
a solution to this?
Thanks.
Ali.
_______________________________________________
Xci-devel mailing list
Xci-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/mailman/listinfo/xci-devel
|
Home