FireMV 2400 not working with Xorg 7.1 anymore

[Roland Scheidegger]

Erwin Rol wrote:
> Hey all,
> 
> I am at the point of giving up on 7.1 and go back to 7.0 because I can't
> get my FireMV 2400 to work anymore. It seems that the second chip (head
> 3 & 4) is not initialized at all, and i have not found a way to do so.
> with Xorg 7.0 this did work and i had a functioning 4 head system. 
> 
> Since i didn't get very much feedback on my problem description  and i
> am not really familiar with the internals of Xorg i am stuck. 
> 
> An interesting observation was that when i ran Xorg 7.0, updated to Xorg
> 7.1 and ran Xorg 7.1 it did work. It seems xorg 7.0 did initialize the
> card in such a way that Xorg 7.1 could also use it. So i am wondering
> where i should look for the difference between Xorg 7.0 and 7.1, if
> someone could at least give me some hints or directions of things that
> changed from 7.0 to 7.1 that could have to do with this problem i would
> be very thankful. 
> 
> I assume the int10 code used to initialize the card, what is the flow on
> how this happens ? 
> 
> If someone is interested in the two rom dumps of the card i could mail
> those privately, so maybe there are some hints in there, it is
> interesting to see that the second rom is only 2k byte large. What is
> stored in those roms ? The radeontool seems to be able to extract some
> information out of them, so is the first 2k something like information
> tables and after that BIOS code ?  If so how can the second chip be
> initialized with the BIOS of the first one ? And why did it work in Xorg
> 7.0.
> 
> Ok i would really be thankful for information.
I'm not sure if this is directly related, but there were other reported 
problems with posting multiple cards with xorg (IIRC 7.1) (i.e. one 
radeon and one intel igp). Can't remember if this was fixed or not in 
cvs though. I've never looked at that code, so not sure what the code 
differences were.

Roland