Using two dual-head Radeon cards with X on a single PC with 3 monitors

Mark Cave-Ayland mark.cave-ayland at
Mon Aug 22 09:16:05 PDT 2011

On 22/08/11 14:09, Jan Kasprzak wrote:

> 	Shouldn't the -isolateDevice be used in order to avoid
> the X servers disabling the other card? Also you may want to make
> them share a common VT (although I am not sure if this is necessary
> in dual-head setup as opposed to dual-seat). In my dual-seat desktop,
> I use the following command lines:
> X :0 -layout Primary -isolateDevice PCI:1:0:0 vt7
> X :1 -layout Secondary -isolateDevice PCI:5:0:0 -sharevts vt7
> (the -layout option is analogous to your -config, but allows you to share
> a single xorg.conf file for both X servers).
> 	However, I had problems with my multi-seat desktop with 2.6.39
> and newer kernels, so you can also try to downgrade the kernel. 2.6.38
> works for me. FWIW, the Fedora bug report I have filled is here:

Hi Yenya,

Thanks for the reference. I'd already tried with the -isolateDevice 
parameter but it didn't seem to make any difference :(

However, I found out that the PC I'm using has an internal built-in 
Intel graphics card, so I enabled it in the BIOS and tried using that 
instead. And with one server running the radeon driver and another 
running the intel driver it works just fine.

It seems there is a bug when you have multiple instances of the radeon 
driver instantiated within different X servers (maybe DRI related but I 
really don't know enough at the moment?). I'd like to try and get this 
fixed if possible - what would be the best thing to do, start a thread 
on the xorg-driver-ati mailing list or create a bug report directly in's bugzilla?



Mark Cave-Ayland - Senior Technical Architect
PostgreSQL - PostGIS
Sirius Corporation plc - control through freedom
t: +44 870 608 0063

Sirius Labs:

More information about the xorg mailing list