secondary PCI card causes SIGFPE in X

Mike Russo miker at readq.com
Thu Jun 2 13:14:32 PDT 2005


Greetings,

I have a rather strange situation. I'm attempting to set up a
dual-headed workstation, so I acquired two old NVIDIA TNT2 PCI cards and
installed them (might as well show you lspci):

[miker at ozark ~]$ /sbin/lspci
00:00.0 Host bridge: ServerWorks CNB20LE Host Bridge (rev 06)
00:00.1 Host bridge: ServerWorks CNB20LE Host Bridge (rev 06)
00:01.0 Multimedia audio controller: Fortemedia, Inc Xwave QS3000A
[FM801] (rev b2)
00:01.1 Input device controller: Fortemedia, Inc Xwave QS3000A [FM801
game port] (rev b2)
00:03.0 RAID bus controller: Silicon Image, Inc. (formerly CMD
Technology Inc) SiI 0649 Ultra ATA/100 PCI to ATA Host Controller (rev 02)
00:04.0 Ethernet controller: Intel Corp. 82557/8/9 [Ethernet Pro 100]
(rev 08)
00:06.0 System peripheral: Compaq Computer Corporation Advanced System
Management Controller
00:0f.0 ISA bridge: ServerWorks CSB5 South Bridge (rev 92)
00:0f.1 IDE interface: ServerWorks CSB5 IDE Controller (rev 92)
00:0f.2 USB Controller: ServerWorks OSB4/CSB5 OHCI USB Controller (rev 05)
00:0f.3 Host bridge: ServerWorks CSB5 LPC bridge
02:01.0 VGA compatible controller: nVidia Corporation NV5M64 [RIVA TNT2
Model 64/Model 64 Pro] (rev 15)
02:04.0 VGA compatible controller: nVidia Corporation NV4 [RIVA TNT]
(rev 04)

The built-in video was irrevocably disabled when I added another video
card, so I had to add two video cards (how annoying is that?). I did "X
-configure" to detect the two cards and write out a sample config, using
the standard "nv" driver, but whenever I attempt to use it, I lose
control of my console (although my machine has not crashed). In the
Xorg.0.log file, I see that the server exited with signal 8 -- a
floating point exception (probably a divide-by-zero somewhere?). This
even happens when I disable the second screen and configure the primary
screen to use the second video card and ignore the first. The question
is why should this be happening, and how can I make it go away?

The chances are that I am NOT going to make it go away short of getting
a whole new motherboard, or possibly trying another video card as the
secondary. The proprietary NVIDIA driver crashes the kernel with a
variety of different errors when I try to use it, and if I do "modprobe
rivafb" and then set UseFBDev to true in my xorg.conf, I get a "divide
error 0000" in the kernel as well. I've tried various things like
turning acpi=off and pci=noacpi/nobios on the kernel command line. It's
so frustrating that the second card is seen and probed but can never be
used. I was hoping someone would at least know why or have a good
idea... anyway, I'm attaching one of the log files that I get when I try
to use the second video card as my primary display (which is why there's
only references to NV(0) as opposed to both NV(0) and NV(1)).  If anyone
has any ideas let me know. :)

-- 
Mike Russo
ReadQ Systems, Inc.
(212) 425 3680 x105

Random quote of the last-time-I-ran-bash:
This is a test of the emergency broadcast system.  Had there been an
actual emergency, then you would no longer be here.


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: Xorg.bad
URL: <http://lists.x.org/archives/xorg/attachments/20050602/da8605e8/attachment.ksh>


More information about the xorg mailing list