https://bugs.freedesktop.org/show_bug.cgi?id=58660
Priority: medium Bug ID: 58660 Assignee: dri-devel@lists.freedesktop.org Summary: Radeon HD 6950 completely broken. Severity: normal Classification: Unclassified OS: All Reporter: v10lator@myway.de Hardware: Other Status: NEW Version: DRI CVS Component: DRM/Radeon Product: DRI
I own a Sapphire Radeon HD 6950. This card has a dual-bios where one of them has unlocked shaders (Radeon HD 6970). This is not a modded BIOS, it's made by Sapphire and fully supported by them. Now my problem is that with the Radeon HD 6950 BIOS (locked shaders) the GPU seems to hang as soon as compiz starts. With the unlocked BIOS it starts just fine.
https://bugs.freedesktop.org/show_bug.cgi?id=58660
--- Comment #1 from Thomas Rohloff v10lator@myway.de --- Forget to say: This is on newest mesa from git with kernel 3.8-rc1 (+ this patch: http://cgit.freedesktop.org/~agd5f/linux/commit/?h=drm-fixes-3.8&id=668b... )
https://bugs.freedesktop.org/show_bug.cgi?id=58660
Alex Deucher agd5f@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|Radeon HD 6950 completely |Radeon HD 6950 completely |broken. |broken with sapphire | |unlocked vbios
--- Comment #2 from Alex Deucher agd5f@yahoo.com --- Please attach the dmesg output when booting with both video bioses. Also, please attach both video bioses. To dump the video bios: (as root) (use lspci to get the bus id) cd /sys/bus/pci/devices/<pci bus id> echo 1 > rom cat rom > /tmp/vbios.rom echo 0 > rom
https://bugs.freedesktop.org/show_bug.cgi?id=58660
--- Comment #3 from Thomas Rohloff v10lator@myway.de --- The thing is that I can't access dmesg after the crash. Not even SysMagRQ works.
See my dmesg with the other BIOS here: https://bugs.freedesktop.org/attachment.cgi?id=72006
I have the BIOSes still lying around somewhere. Let me search real quick.
https://bugs.freedesktop.org/show_bug.cgi?id=58660
--- Comment #4 from Thomas Rohloff v10lator@myway.de --- Created attachment 72008 --> https://bugs.freedesktop.org/attachment.cgi?id=72008&action=edit BIOS 0 (locked shaders)
https://bugs.freedesktop.org/show_bug.cgi?id=58660
--- Comment #5 from Thomas Rohloff v10lator@myway.de --- Created attachment 72009 --> https://bugs.freedesktop.org/attachment.cgi?id=72009&action=edit BIOS 1 (unlocked shaders)
https://bugs.freedesktop.org/show_bug.cgi?id=58660
--- Comment #6 from Thomas Rohloff v10lator@myway.de --- Whoever changed the report title: It is broken on the locked BIOS, not on the unlocked.
https://bugs.freedesktop.org/show_bug.cgi?id=58660
--- Comment #7 from Thomas Rohloff v10lator@myway.de --- Bisected mesa.
This is a mesa bug caused by http://cgit.freedesktop.org/mesa/mesa/commit/?id=6532eb17baff6e61b427f29e076...
Can anybody move this to the right place or do I have to re-create the report (and if so: Where) ?
I also think this is a duplication of https://bugs.freedesktop.org/show_bug.cgi?id=58667 - The crashes just occur way more often (right after starting compiz) with the locked BIOS.
https://bugs.freedesktop.org/show_bug.cgi?id=58660
Alex Deucher agd5f@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|Radeon HD 6950 completely |Radeon HD 6950 completely |broken with sapphire |broken with sapphire locked |unlocked vbios |vbios
https://bugs.freedesktop.org/show_bug.cgi?id=58660
Marek Olšák maraeo@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|Radeon HD 6950 completely |CAYMAN broken with HyperZ |broken with sapphire locked |on |vbios | Product|DRI |Mesa Version|DRI CVS |unspecified Component|DRM/Radeon |Drivers/Gallium/r600
https://bugs.freedesktop.org/show_bug.cgi?id=58660
Andreas Boll andreas.boll.dev@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |75112
https://bugs.freedesktop.org/show_bug.cgi?id=58660
--- Comment #8 from Marek Olšák maraeo@gmail.com --- Is this still an issue with current Mesa git?
https://bugs.freedesktop.org/show_bug.cgi?id=58660
Marek Olšák maraeo@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |WORKSFORME
--- Comment #9 from Marek Olšák maraeo@gmail.com --- I'm closing this 2-years-old bug.
The report says that a patch which adds HTILE support causes a crash when compiz starts, but compiz doesn't even use a zbuffer, so we are not really using Hyper-Z here. I believe this is just a bug caused by incorrect register programming in the initial HTILE patch and must have been fixed since then.
Feel free to reopen if you see this crash with current Mesa git.
dri-devel@lists.freedesktop.org
-
bugzilla-daemon＠freedesktop.org