libpciaccess patch for multiple videocards/int 10 problem

[Timothy S. Nelson]

On Thu, 12 Mar 2009, Francisco Jerez wrote:

> Hi,
>
> "Timothy S. Nelson" <wayland at wayland.id.au> writes:
>
>>  	Hi all.  There's a patch that's attached to bugzilla bug 18160 that
>> fixes the problem specified in that bug; now that it prevents xorg from
>> hanging the machine, other bugs have become visible, but the patch attached to
>> comment 30 seems to solve the problem.  How do we go about getting this
>> applied to libpciaccess?
>>
>> https://bugs.freedesktop.org/show_bug.cgi?id=18160
>>
>>  	:)
>>
>
> If I've understood correctly, the patch at comment 30 mainly ensures
> that the device has been pci_device_enable()'ed when the ROM is about
> to be read. That looks like something the X server should care about
> (in fact it already does, see xserver commit
> ea4ec9e9983e25d94a3edf8a77ed2ad1db193284).

 	Great!  I think the patch fixes some other things, but I didn't write 
the patch, so I'm not going to worry too much about them.

 	The patch, while it prevents my machine from locking up with two video 
cards (and I presume the other xserver commit does the same), didn't actually 
make either of them work; the only way I was able to obtain a working config 
was to comment out one of the screens.  That's why I believe there's still 
work to be done on it.  I had two screens working with a pre-libpciaccess 
version of xorg, but now they don't work.

 	Thanks,


---------------------------------------------------------------------
| Name: Tim Nelson                 | Because the Creator is,        |
| E-mail: wayland at wayland.id.au    | I am                           |
---------------------------------------------------------------------

----BEGIN GEEK CODE BLOCK----
Version 3.12
GCS d+++ s+: a- C++$ U+++$ P+++$ L+++ E- W+ N+ w--- V- 
PE(+) Y+>++ PGP->+++ R(+) !tv b++ DI++++ D G+ e++>++++ h! y-
-----END GEEK CODE BLOCK-----