Bug 92865 - MSI R7 370 Armor 2X dpm issues
Summary: MSI R7 370 Armor 2X dpm issues
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Radeon (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: high major
Assignee: Default DRI bug account
QA Contact:
Depends on:
Reported: 2015-11-09 07:23 UTC by Maxim Sheviakov
Modified: 2015-11-27 08:12 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Description Maxim Sheviakov 2015-11-09 07:23:15 UTC
Long story short: sent a patch to Alex Deucher, the patch got submitted and it's in 4.3 kernel now. However, I get a black screen after system startup, or it can happen randomly, especially after bug screen updates. I'm on freshly installed Arch Linux 64-bit UEFI. I think the issue is connected with the recent commit to the si_dpm.c, maybe there are ID conflicts. Here's a link to that new patch: https://github.com/torvalds/linux/commit/2b02ec79004388a8c65e227bc289ed891b5ac8c6

And link to my patch: https://github.com/torvalds/linux/commit/e78654799135a788a941bacad3452fbd7083e518

Maybe we should make some identification code in that file? Oh, I'll also try to compile the 4.3 kernel with that Asus line removed. Any ideas other than those above?
Comment 1 Maxim Sheviakov 2015-11-09 07:26:59 UTC
Oh, by the way, the GPU worked flawlessly with Linux 4.2.X + my patch.
Comment 2 Maxim Sheviakov 2015-11-09 10:36:31 UTC
Strange. Looks like changes between rc5 and rc6 (or rc4 and rc5) in radeon_pm.c may be causing issues, so I'll check that. Hey, has anyone noticed this bug yet? I hope someone will help me.
Comment 3 Maxim Sheviakov 2015-11-09 10:39:32 UTC
Here's the link. Check the pm... line, it's likely to cause problems.
Comment 4 Maxim Sheviakov 2015-11-09 10:39:54 UTC
(In reply to Maxim Sheviakov from comment #3)
> Here's the link. Check the pm... line, it's likely to cause problems.

Comment 5 Maxim Sheviakov 2015-11-09 18:04:10 UTC
(In reply to Maxim Sheviakov from comment #4)
> (In reply to Maxim Sheviakov from comment #3)
> > Here's the link. Check the pm... line, it's likely to cause problems.
> http://fossies.org/diffs/linux/4.3-rc5_vs_4.3-rc6/drivers/gpu/drm/radeon/
> radeon_pm.c-diff.html

Nope, not those pm lines. Will try using rc4 file, will report.
Comment 6 Maxim Sheviakov 2015-11-09 19:34:34 UTC
Whaaaaaaaaaaaat? Where the hell is the line "{ PCI_VENDOR_ID_ATI, 0x6811, 0x1462, 0x2015, 0, 120000 }"? It was proposed to 4.3-rc4, where the hell is it? Please consider adding the line to si_dpm.c, quirk list is its place.
Comment 7 Maxim Sheviakov 2015-11-09 19:39:26 UTC
Oh crap, the line in my patch somehow got changed. Have no idea how could it happen. Anyway, will send another patch to 4.4, if the merge window is still open.
Comment 8 Maxim Sheviakov 2015-11-10 15:01:57 UTC
The thing is, commiter (isn't it Alex?) changed the 1462 to 1762, so that the ID was there for nothing and made no sense and the GPU wouldn't run with dpm. That's it. Will probably send another patch.

Dear Alex, could you change that line from 1762 to 1462?
Comment 9 Maxim Sheviakov 2015-11-27 08:12:58 UTC
So, the patch got upstreamed, now this should work. With 4.4 kernel at least.

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.