mbox series

[0/5,SRU,OEM-5.17] Keeps rebooting with AMD W6400, W6600, and W6800 graphic cards

Message ID 20221221072647.8981-1-acelan.kao@canonical.com
Headers show
Series Keeps rebooting with AMD W6400, W6600, and W6800 graphic cards | expand

Message

AceLan Kao Dec. 21, 2022, 7:26 a.m. UTC
From: "Chia-Lin Kao (AceLan)" <acelan.kao@canonical.com>

BugLink: https://launchpad.net/bugs/2000110

[Impact]
With AMD W6400, W6600, or W6800 graphic cards the system keeps rebooting
while entering graphics mode.

[Fix]
AMD provides a list of commits to fix this issue
1. drm/amdgpu: Remove ATC L2 access for MMHUB 2.1.x
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=d2c4c1569a7d7d5c8f75963bf2d62d7aeac30e2a
2. drm/amdgpu: Don't enable LTR if not supported
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=6c20490663553cd7e07d8de8af482012329ab9d6
3. Patch series: fix PCI AER issues
drm/amdgpu: make sure to init common IP before gmc
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=a8671493d2074950553da3cf07d1be43185ef6c6
drm/amdgpu: move nbio sdma_doorbell_range() into sdma code for vega
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=e3163bc8ffdfdb405e10530b140135b2ee487f89
drm/amdgpu: move nbio ih_doorbell_range() into ih code for vega
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=dc1d85cb790f2091eea074cee24a704b2d6c4a06
4. Disable amdgpu runpm, will have a update later.
https://patchwork.freedesktop.org/patch/507366/

[Test]
Verified on the machine with AMD graphic card, and confirmed the issue is gone.

[Where problems could occur]
The main idea is to disable BACO on those cards, the fix is quite
straightforward with 2 Fixes commits, have impact on limited cards.

Alex Deucher (1):
  drm/amdgpu: make sure to init common IP before gmc

Guchun Chen (3):
  drm/amd/pm: disable BACO entry/exit completely on several sienna
    cichlid cards
  drm/amdgpu: disable BACO on special BEIGE_GOBY card
  drm/amdgpu: disable BACO support on more cards

Lijo Lazar (1):
  drm/amdgpu: Remove ATC L2 access for MMHUB 2.1.x

 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c    | 14 ++++++++--
 drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c       | 28 ++++++-------------
 .../amd/pm/swsmu/smu11/sienna_cichlid_ppt.c   | 17 +++++++++++
 3 files changed, 36 insertions(+), 23 deletions(-)

Comments

Timo Aaltonen Jan. 3, 2023, 1:55 p.m. UTC | #1
AceLan Kao kirjoitti 21.12.2022 klo 9.26:
> From: "Chia-Lin Kao (AceLan)" <acelan.kao@canonical.com>
> 
> BugLink: https://launchpad.net/bugs/2000110
> 
> [Impact]
> With AMD W6400, W6600, or W6800 graphic cards the system keeps rebooting
> while entering graphics mode.
> 
> [Fix]
> AMD provides a list of commits to fix this issue
> 1. drm/amdgpu: Remove ATC L2 access for MMHUB 2.1.x
> https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=d2c4c1569a7d7d5c8f75963bf2d62d7aeac30e2a
> 2. drm/amdgpu: Don't enable LTR if not supported
> https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=6c20490663553cd7e07d8de8af482012329ab9d6
> 3. Patch series: fix PCI AER issues
> drm/amdgpu: make sure to init common IP before gmc
> https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=a8671493d2074950553da3cf07d1be43185ef6c6
> drm/amdgpu: move nbio sdma_doorbell_range() into sdma code for vega
> https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=e3163bc8ffdfdb405e10530b140135b2ee487f89
> drm/amdgpu: move nbio ih_doorbell_range() into ih code for vega
> https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20221107&id=dc1d85cb790f2091eea074cee24a704b2d6c4a06
> 4. Disable amdgpu runpm, will have a update later.
> https://patchwork.freedesktop.org/patch/507366/
> 
> [Test]
> Verified on the machine with AMD graphic card, and confirmed the issue is gone.
> 
> [Where problems could occur]
> The main idea is to disable BACO on those cards, the fix is quite
> straightforward with 2 Fixes commits, have impact on limited cards.
> 
> Alex Deucher (1):
>    drm/amdgpu: make sure to init common IP before gmc
> 
> Guchun Chen (3):
>    drm/amd/pm: disable BACO entry/exit completely on several sienna
>      cichlid cards
>    drm/amdgpu: disable BACO on special BEIGE_GOBY card
>    drm/amdgpu: disable BACO support on more cards
> 
> Lijo Lazar (1):
>    drm/amdgpu: Remove ATC L2 access for MMHUB 2.1.x
> 
>   drivers/gpu/drm/amd/amdgpu/amdgpu_device.c    | 14 ++++++++--
>   drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c       | 28 ++++++-------------
>   .../amd/pm/swsmu/smu11/sienna_cichlid_ppt.c   | 17 +++++++++++
>   3 files changed, 36 insertions(+), 23 deletions(-)
> 

applied to oem-5.17, thanks