4.1.0-0pre1v1.3 confused by 2 mga cards
For reasons I have yet to figure out, my xinerama setup broke with my
last upgrade. It's hard to debug, since the usual symptom is a hard
lockup, but it looks like despite my BusID entries, the mga driver is
talking to my G200tv pci marvel when it thinks it's talking to my
secondary Millennium I pci. I managed to get it to not lock my system
running the Mil I as an fbdev, and it used the g200 for both screens,
and complained that the two devices shared i/o ranges and whatnot.
Interestingly, even with the g200 commented out in the XF86Config, it
still clears the screen on the g200 rather than the Mil1 before
crashing the kernel... I suspect this is either because it is noticing
I'm running from a console on /dev/fb0 (the g200) or because it is
defaulting to the lowest pci-numbered matrox card despite instructions
otherwise. X -scanpci and scanpci both report correct values, and this
setup ran with an earlier 4.1.0 release (although I had a crash&reboot
at about the same time as the problem started, because nis+pam was
broken so I couldn't log in).
Anyway, just curious if anyone else is seeing this sort of thing, or
if anyone has hints as to recent changes that might impact this. I'll
report if I figure anything meaningful out, but since it's .5hr of
fscking for each failure, it's eating into my work schedule.
thanks/ fyi
- M
--
Mark "Monty" Montague | monty@gg.caltech.edu | I don't do Windows(tm)
main(){printf("I am self-aware\\n");}
<URL:http://www.gg.caltech.edu/~monty/monty.shtml>
X-PGP-Fingerprint: E4 EA 6D B1 82 46 DB A1 B0 FF 60 B9 F9 5D 5C F7
Reply to: