Summary: | no DRI with sis760 and agp card... | ||
---|---|---|---|
Product: | DRI | Reporter: | enrico <enricoss> |
Component: | General | Assignee: | Default DRI bug account <dri-devel> |
Status: | RESOLVED WONTFIX | QA Contact: | |
Severity: | normal | ||
Priority: | high | CC: | bugzi11.fdo.tormod, sybillel |
Version: | XOrg git | ||
Hardware: | x86 (IA32) | ||
OS: | Linux (All) | ||
Whiteboard: | |||
i915 platform: | i915 features: |
Description
enrico
2006-10-31 08:16:45 UTC
Just a note to say that this issue is still present in kernel 2.6.20 in Gentoo. I'm a Gentoo user, but I opened the bug in Ubuntu's launchpad, https://launchpad.net/distros/ubuntu/+source/linux-source-2.6.15/+bug/34144 . I encountered the problem when using Ubuntu to install Gentoo. On Gentoo, I initially did not have the problem of agpgart misidentifying the agp aperture as 4M instead of 128M (set in BIOS) and the consequence that the radeon driver reports that agp could not bind so no DRI. I believe things changed when the device ids for SIS 760 were added to the vanilla kernel and then the Gentoo sources from the Ubuntu kernel where they were first. When my SIS chipset was identified generically, it would be recognized by agpgart during boot as having a 128M aperture, which is correct. AND DRI worked. When the chip is identified as SIS 760, the result is incorrect - 4M aperture. In order to test this theory, I attempted to patch my current kernel ( 2.6.20-gentoo-r8) by commenting out the lines added to amd64-agp.c that mention SIS 760, in the hope that it would then fall back to the generic detection. But unfortunately this did not work. So I don't know. It's quite possible that I did something wrong. Anyway, I'm just hoping to draw some attention to this still-unresolved issue in the hopes of being able to use the DRI capabilities of my radeon 9250 card once again. Thanks. As an interim solution you can always force the card to run in PCI mode: Option "BusType" "PCI" Thank you, Alex Deucher, that was a good idea for a workaround. I wasn't aware that PCI mode was a possibility. :) Just wanted to add that the problem still exists as of kernel 2.6.21.5 (I believe the problem lies in the sis-agp kernel module, not the x.org agp driver). A million thanks to Alex Deucher for the workaround. It was very discouraging to get a brand new motherboard this week and not be able to use my Radeon 7000 to its full potential, but the PCI workaround is exactly what I needed to get back that missing functionality! The sis driver was deleted over seven years ago. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.