Using 4GB with amd64 - problems with NVidia graphics kernel driver
Hi,
I've recently installed the debian amd64 distribution on an MSI K8N
Platinum board with 4GB memory. Installation was no problem, once I found
the correct installer. I've also compiled the nvidia-kernel module
(1.0.7174-3) NVidia module for a 6600GT graphics card, and it works well.
I'm about to test video playback, so I should be able to say how
fast it is soon.
I'm using my own compiled 2.6.11 kernel (based on the kernel sources
from the unofficial apt servers, with patches for sk98lin
(downloaded from their homepage, non-open source - the internal
sk98lin driver does not support the Marvell Yukon3 which is
on-board))
During install I've not set memory hole remapping in the BIOS, and therefore
less than 4GB are recognized by the system.
This is of course becoming an issue now, so I've tested both
settings which are available: Continuous and Discrete. Both lead to
a Kernel oops in XFree86, which may be depending on the NVidia
model. See attached log excerpt. Is this a known problem? Is there a
workaround, or might the newer NVidia drivers / a newer kernel solve
this problem?
Some other comments / issues:
* forcedeth (up to 0.35): setting the interface to promiscuous mode does
not seem to work properly. I've noticed that I need to set this
flag to be able to get packets over my gigabit switch. Packets
from the forcedeth interface always got "host not reachable".
I've deactivated the interface and switched to a 100MBit card with
NatSemi. The internal NVidia chipset is clearly more suited for
gigabit ethernet and is connected outside the PCI bus, but
currently I do not need gigabit ethernet.
* Somebody should really go and compile Xvnc (vncserver) for the apt
servers. Minimal patches are needed to make it work with XFree86
4.3.0 (i.e. you have to add -fPIC everywhere :->), but it took me
almost a day to get it right. I used VNC 4.1.1, but earlier
versions are probably ok as well (>=4.0) from what I saw on the
web.
Other than that, everything works (sound has not been tested yet,
wlan is an external PCI card and supported by rt2500 -
rt2400.sourceforge.net ), so the K8N Platinum (+54G) is definitely
supported up to 3GB. ;-)
Best,
Alex
--
Dr.techn. Alexander K. Seewald
Austrian Research Institute alexsee@oefai.at
for Artificial Intelligence www.oefai.at
------------------------------------------------
Information wants to be free;
Information also wants to be expensive (S.Brant)
--------------- alex.seewald.at ----------------
Jun 28 16:12:37 localhost kernel: NVRM: loading NVIDIA Linux x86_64 NVIDIA Kernel Module 1.0-7174 Tue Mar 22 06:45:40 PST 2005
Jun 28 16:12:37 localhost kernel: Unable to handle kernel NULL pointer dereference at 0000000000000000 RIP:
Jun 28 16:12:37 localhost kernel: <ffffffff881cba7d>{:nvidia:_nv001656rm+9}
Jun 28 16:12:37 localhost kernel: PGD 13b302067 PUD 13b4ab067 PMD 0
Jun 28 16:12:37 localhost kernel: Oops: 0000 [1]
Jun 28 16:12:37 localhost kernel: CPU 0
Jun 28 16:12:37 localhost kernel: Modules linked in: nvidia ipv6 af_packet parport_pc parport floppy pcspkr shpchp pci_hotplug amd74xx snd_intel8x0 snd_ac97_codec snd_pcm snd_timer snd soundcore snd_page_alloc ehci_hcd ohci_hcd ide_scsi reiserfs tsdev mousedev evdev psmouse ide_generic ide_disk ide_cd ide_core cdrom forcedeth genrtc unix
Jun 28 16:12:37 localhost kernel: Pid: 3468, comm: XFree86 Tainted: P 2.6.11
Jun 28 16:12:37 localhost kernel: RIP: 0010:[__nosave_end+131816061/2132713472] <ffffffff881cba7d>{:nvidia:_nv001656rm+9}
Jun 28 16:12:37 localhost kernel: RSP: 0018:ffff81013b4a5cc0 EFLAGS: 00010292
Jun 28 16:12:37 localhost kernel: RAX: 0000000000000000 RBX: ffff81013e109800 RCX: 0000000000000000
Jun 28 16:12:37 localhost kernel: RDX: 0000000000000000 RSI: ffff81013e109800 RDI: ffff81013f198000
Jun 28 16:12:37 localhost kernel: RBP: ffff81013e109800 R08: ffff81013f198000 R09: ffff81013b4a5d0c
Jun 28 16:12:37 localhost kernel: R10: 00000000000f04a1 R11: 00000010bffed3c0 R12: ffff81013d015e00
Jun 28 16:12:37 localhost kernel: R13: ffff81013d015e00 R14: 0000000000000000 R15: ffff81013f198000
Jun 28 16:12:37 localhost kernel: FS: 00002aaaab19e700(0000) GS:ffffffff803ebf80(0000) knlGS:0000000000000000
Jun 28 16:12:37 localhost kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
Jun 28 16:12:37 localhost kernel: CR2: 0000000000000000 CR3: 000000013b4a8000 CR4: 00000000000006e0
Jun 28 16:12:37 localhost kernel: Process XFree86 (pid: 3468, threadinfo ffff81013b4a4000, task ffff81013f4aedd0)
Jun 28 16:12:37 localhost kernel: Stack: ffffffff8834e6d3 ffffffff88464f90 ffffffff8834e2e6 ffff81013e109800
Jun 28 16:12:37 localhost kernel: ffff81013d580000 ffff81013d01a000 ffffffff882833aa ffff81013f198000
Jun 28 16:12:37 localhost kernel: 00000000881cc583 0001000000000000
Jun 28 16:12:37 localhost kernel: Call Trace:<ffffffff8834e6d3>{:nvidia:_nv003393rm+29} <ffffffff8834e2e6>{:nvidia:_nv003395rm+36}
Jun 28 16:12:37 localhost kernel: <ffffffff882833aa>{:nvidia:_nv001871rm+260} <ffffffff881cc7c3>{:nvidia:_nv001208rm+285}
Jun 28 16:12:37 localhost kernel: <ffffffff881cd103>{:nvidia:_nv001214rm+297} <ffffffff881d04f1>{:nvidia:rm_init_adapter+107}
Jun 28 16:12:37 localhost kernel: <ffffffff883d1413>{:nvidia:nv_kern_open+623} <ffffffff801658a0>{chrdev_open+285}
Jun 28 16:12:37 localhost kernel: <ffffffff8015dc49>{dentry_open+205} <ffffffff8015db70>{filp_open+62}
Jun 28 16:12:37 localhost kernel: <ffffffff8015dd8b>{get_unused_fd+89} <ffffffff8015de9a>{sys_open+57}
Jun 28 16:12:37 localhost kernel: <ffffffff8010ce26>{system_call+126}
Jun 28 16:12:37 localhost kernel:
Jun 28 16:12:37 localhost kernel: Code: 0f b6 04 02 0f b6 c0 c3 90 c1 ea 02 89 d2 48 8b 86 c0 01 00
Jun 28 16:12:37 localhost kernel: RIP <ffffffff881cba7d>{:nvidia:_nv001656rm+9} RSP <ffff81013b4a5cc0>
Jun 28 16:12:37 localhost kernel: CR2: 0000000000000000
Jun 28 16:12:45 localhost kernel: <7>eth0: no IPv6 routers present
Jun 28 16:12:46 localhost shutdown[3518]: shutting down for system reboot
Reply to: