On 11 Jan 2003, Michel Dänzer wrote: > On Sam, 2003-01-11 at 03:19, [EMAIL PROTECTED] wrote: > > On Sat, Jan 11, 2003 at 03:09:49AM +0100, Michel D?nzer wrote: > > > On Don, 2003-01-09 at 18:14, [EMAIL PROTECTED] wrote: > > > > > > > > I have been in contact with Hui Yu regarding this matter. He gave me > > > > some other options to try in place of 0x000e0020. On his advice, i > > > > tried 0x000e0000 and the AGP lockups I was experiencing with a 64 MB > > > > aperture were fixed (on a Mobile Radeon 7500). I think it is best to > > > > change the code accordingly because if you do nothing as in the case > > > > of the above-mentioned fix, my Mobility locks up as described. > > > > > > You mean the attached patch? If so, I can commit that. > > > > Yes. That would fix my AGP lockups. > > Okay, committed it to the trunk and mesa-4-0-4-branch.
I'm just curious why AGP_CNTL isn't always programmed with the same aperture size as info->agpSize as it is with Rage 128 (and mach64)? Is it possible for AGP_CNTL to be set (at server startup) for an aperture size smaller than info->agpSize -- which is configurable by the user? Also, the meaning of 0x000e0000 doesn't seem to be documented in radeon_reg.h, so I don't know if it's similar to mach64, but I found that on mach64 enabling high priority reads (bit 16 of AGP_CNTL) causes instability at 2x but not at 1x. -- Leif Delgass http://www.retinalburn.net ------------------------------------------------------- This SF.NET email is sponsored by: SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! http://www.vasoftware.com _______________________________________________ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
