ACK: [SRU][P][PATCH 0/6] Poweroff not working consistently after upgrading kernel 6.14.0-17.17 or later

Sarah Emery sarah.emery at canonical.com
Wed Nov 12 16:25:59 UTC 2025


On 03/11/2025 16:15, AceLan Kao wrote:
> From: "Chia-Lin Kao (AceLan)" <acelan.kao at canonical.com>
> 
> BugLink: https://bugs.launchpad.net/bugs/2115860
> 
> [Impact]
> The system with amdgpu can't poweroff.
> 
> [Fix]
> This series of patches fix the issue
> https://lore.kernel.org/amd-gfx/20251002174245.1691343-1-mario.limonciello@amd.com/
> 
> Included in 6.18
> 531df041f2a5 drm/amd: Avoid evicting resources at S5
> 
> Included in linux-next, should be included in 6.19
> 173360fe49c2 drm/amd: Pass IP suspend errors up to callers
> b7ff2e79240a drm/amd: Don't always set IP block HW status to false
> f35f2541780b drm/amd: Remove comment about handling errors in amdgpu_device_ip_suspend_phase1()
> 6062ede68097 drm/amd: Stop exporting amdgpu_device_ip_suspend() outside amdgpu_device
> e6af507de859 drm/amd: Unify shutdown() callback behavior
> 
> [Test]
> Try powering off AMD platforms which runs with amdgpu igpu or dgpu couple times.
> It should always be powered off.
> 
> [Where problems could occur]
> These patches address about the amdgpu shutdown or suspend process, I don't
> see there is any potential issue.
> 
> Mario Limonciello (5):
>    drm/amd: Unify shutdown() callback behavior
>    drm/amd: Stop exporting amdgpu_device_ip_suspend() outside
>      amdgpu_device
>    drm/amd: Remove comment about handling errors in
>      amdgpu_device_ip_suspend_phase1()
>    drm/amd: Don't always set IP block HW status to false
>    drm/amd: Pass IP suspend errors up to callers
> 
> Mario Limonciello (AMD) (1):
>    drm/amd: Avoid evicting resources at S5
> 
>   drivers/gpu/drm/amd/amdgpu/amdgpu.h        |  1 -
>   drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 19 +++++++++++++------
>   drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c    |  3 ++-
>   3 files changed, 15 insertions(+), 8 deletions(-)
> 

Acked-by: Sarah Emery <sarah.emery at canonical.com>



More information about the kernel-team mailing list