Bug 16361
| Summary: | KMS not working on Radeon Mobility 3430 | ||
|---|---|---|---|
| Product: | Drivers | Reporter: | Ruslan (ruslan.st) |
| Component: | Video(DRI - non Intel) | Assignee: | drivers_video-dri |
| Status: | CLOSED INVALID | ||
| Severity: | normal | CC: | alexdeucher, glisse, maciej.rutecki, rjw, zajec5 |
| Priority: | P1 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 2.6.34.1 | Subsystem: | |
| Regression: | Yes | Bisected commit-id: | |
| Bug Depends on: | |||
| Bug Blocks: | 15310 | ||
| Attachments: |
log of non-working kernel 2.6.34-rc1
log of working kernel 2.6.33.6 |
||
|
Description
Ruslan
2010-07-10 13:32:26 UTC
Can you bisect what commit caused the problem? Also does 2.6.35 work ok? Currently I have checked this: 2.6.33.6 - works 2.6.34-rc1 - does not work 2.6.35-rc4 - does not work (In reply to comment #1) > Can you bisect what commit caused the problem? Also does 2.6.35 work ok? Currently I have checked this: 2.6.33.6 - works 2.6.34-rc1 - does not work 2.6.35-rc4 - does not work I did bisection. Bad commit is: 9f93ed39804a9cfe10577cfae66059fe6bc6e3a5 commit 9f93ed39804a9cfe10577cfae66059fe6bc6e3a5 Author: Jerome Glisse <jglisse@redhat.com> Date: Thu Jan 28 18:22:31 2010 +0100 drm/radeon/kms: bogus cs recorder utilities This add an utilities function radeon_ib_bogus_add which will save an ib into a list of ib which can then be dumped using debugfs. Once dumped the ib is removed from the list. This should allow to save & capute ib for further debugging. Signed-off-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Dave Airlie <airlied@redhat.com> First-Bad-Commit : 9f93ed39804a9cfe10577cfae66059fe6bc6e3a5 I would find very hard to believe that this patch is the root of your issue, did you try to reverse it ? (it should reverse cleanly i think). I am sorry, I posted wrong commit number. I re-checked my bisect process and correct first bad commit is: d594e46ace22afa1621254f6f669e65430048153 Can you attach your dmesg? Working & non working dmesg would be helpfull thanks First-Bad-Commit : d594e46ace22afa1621254f6f669e65430048153 Created attachment 27125 [details]
log of non-working kernel 2.6.34-rc1
this is dmesg log of non-working kernel 2.6.34-rc1
Created attachment 27126 [details]
log of working kernel 2.6.33.6
dmesg log of working kernel 2.6.33.6
It seems your issue is a missing firmware, you likely forgot to add firmware to your initramfs or to build them in with the kernel. Try fixing that and also boot with drm.debug=4 option so we can have more debug informations on modesetting issue. You might also want to try radeon.new_pll=0 boot option. I attached two logs: for good and bad kernels. I just noticed that firmware fails to load in both cases. I never checked this and thought that with 2.6.33.6 it loads successfully. This means that KMS is not working with my card at all. But in bad case kernel 2.6.34-rc1 is unable to switch into good video mode after firmware fail. (In reply to comment #14) > It seems your issue is a missing firmware, you likely forgot to add firmware > to > your initramfs or to build them in with the kernel. Try fixing that and also > boot with drm.debug=4 option so we can have more debug informations on > modesetting issue. > > You might also want to try radeon.new_pll=0 boot option. ok. I will work on this I compiled in firmware into kernel and got KMS working on 2.6.34.1. It was incorrect config. So the problem is that after failed firmware load newer kernels cannot switch into good framebuffer or just plain VGA mode. But I am not sure this can be treated as a bug. |