On Mon, May 10, 2004 at 01:18:26AM -0700, carbonated beverage wrote: > On Mon, May 10, 2004 at 02:12:12AM -0500, Branden Robinson wrote: > > Outstanding. You've confirmed as well that your patched X server > > doesn't get confused on 2.4.x, then? > > Yup, but it just tries the code that works for 2.4.x on my box, and when > it fails, tries the 2.6.x-based path. So it's blindly trying to open the > /proc/bus/pci/* stuff directly instead of "selecting" the right one for > the kernel. I'm not sure if there's any architectures where a PCI domain > tree and a non-domain tree are exported. If there is, and there's a video > card in both, then it probably won't do the right thing. > > As far as I can tell, though, there doesn't seem to be such hardware... > *crosses fingers* > > > If your patch works as intended and causes no breakage both on 2.4.x and > > 2.6.x, then it's probably good enough for us to apply. > > > > Please confirm (or correct) my understanding of the situation. > > I haven't tested this against any other architectures, as only my Ultra 5 > currently has a video card that's under the PCI device tree. I've added this item to the TODO list, but we do not expect to include it in 4.3.0.dfsg.1-2. It may get a closer look for the release after that. -- G. Branden Robinson | When dogma enters the brain, all Debian GNU/Linux | intellectual activity ceases. branden@debian.org | -- Robert Anton Wilson http://people.debian.org/~branden/ |
Attachment:
signature.asc
Description: Digital signature