Summary: | sys-kernel/gentoo-sources-2.6.34-r3 causes DRI initialization with ati-drivers to fail | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | rainer <rbachl> |
Component: | [OLD] Core system | Assignee: | Gentoo Kernel Bug Wranglers and Kernel Maintainers <kernel> |
Status: | RESOLVED WORKSFORME | ||
Severity: | normal | CC: | dhp_gentoo |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
Whiteboard: | linux-2.6.34.1-regression | ||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | 317231 | ||
Bug Blocks: |
Description
rainer
2010-08-04 19:55:43 UTC
I just saw the typos in my bug report. [...] > This is from Xorg.log.0 in case I am booting with kernel compiled from > gentoo-sources-2.6.24-r3: [...] > This is the corresponding section from Xorg.log.0 in case I am booting with > kernel compiled from gentoo-sources-2.6.24-r2: As stated in the introduction, the correct versions to mention here are 2.6.34-r3 and 2.6.34-r2. Which was the latest working kernel? Could you please try gentoo-sources-2.6.35? Sorry, according to your report gentoo-sources-2.6.34-r2 was the latest working kernel. Please try 2.6.35 (In reply to comment #3) > Sorry, according to your report gentoo-sources-2.6.34-r2 was the latest working > kernel. Please try 2.6.35 > Sorry, currently I cannot help with confirming the bug in 2.6.35 or 2.6.35-r1. Booting with these kernels changes something with the devices so that booting proceeds with a wrong device. The last kernel working for me is 2.6.34-r2. FGLRX driver is always designed for one specific kernel. Using a kernel that is not exactly the one recommended by AMD is doomed to fail. Please wait for a driver update. => INVALID. This bug is (should) blocking bug 317231 . Please update the status. I have got ati-drivers-10.8 with DRI working with gentoo-sources-2.6.35-r5 and xorg-server-1.7.7-r1, although there are some artifacts. In particular, some window areas are not always updated and remain black until the window is scrolled or moved. How should one know which kernel version ati-drivers is working with? If any ati-drivers version is only working for a specific kernel version, then all other kernel versions should be prohibited/excluded in the ebuild. It's bit more complex. Check the version you have installed, and, either read the provided documentation, or go on AMD website to grab the doc, and compatibility list. A driver is always written for a specific (verbatim) kernel. Older kernel usually don't work. Recent kernel may work, or not. A driver often work with the two or three next kernels; it depends how much mess have been done. Kernels can not be excluded for security reasons: if a kernel has a security issue, everyone need to update fast; blocking a potential security update can not be accepted. Anyway, it USUALLY work fine for several next kernels. 2.6.34 is a broken kernel anyway; but your bug does not seem to be dup of bug 328889 ... because you say you have the problem as you start X. I also have black lines and garbage, but, only after few minutes or hours; and when this starts, KP usually occurs very fast, within minutes. Your report let me think that your box remains usable for hours with garbage. Anyway, 2.6.34 is broken. Try other combination: play with BIOS features (do you have SurroundView in BIOS ?). Also, this kind of problem is very frequent with close source. Try the free driver instead: remove the fglrx flag, and put radeon (then remerge world: emerge -DaNuv world); emerge the ucode, and update your xorg.conf if any. DRI should keep working; your system should be usable, and should not have garbage ... or it can get worse :) Note: some say that 2.6.35 is NOT better :) Oh, the bug is already closed by maint; so, you have nothing to loose :) |