Please share the dmesg log, and what’s the chip are you using ?

 

From: Julien Isorce [mailto:julien.isorce@gmail.com]
Sent: 2017
119 17:35
To: Liu, Monk <Monk.Liu@amd.com>
Cc: amd-gfx@lists.freedesktop.org
Subject: Re: [PATCH 0/7] *** GPU recover V3 ***

 

Hi Monk.

 

I am interested on this. Currently when a "ring X stalled for more than N sec" happens it usually goes into the gpu reset routine.

Does it always cause the vram to be lost ? Could you explain what happens if the vram remains lost ?

 

I am asking this because I experienced some recurrent gpu reset that are marked succeeded from the log but fail in the "resume" step.

I would not be interested in this if it would always leave a chance to the user to cleanly reboot the machine.

 

The issue is that it can require a hard reboot without kernel panic and without keeping the keyboard responding to magic keys.

Are those patches trying to address this issue ?

 

Note that here "issue" is not referring to the root cause of a ring X stalled and it is also not referring to why "resume" step fails. 

 

Thx a lot

Julien

 

 

On 30 October 2017 at 04:15, Monk Liu <Monk.Liu@amd.com> wrote:

*** job skipping logic in scheduler part is re-implemented  ***

Monk Liu (7):
  amd/scheduler:imple job skip feature(v3)
  drm/amdgpu:implement new GPU recover(v3)
  drm/amdgpu:cleanup in_sriov_reset and lock_reset
  drm/amdgpu:cleanup ucode_init_bo
  drm/amdgpu:block kms open during gpu_reset
  drm/amdgpu/sriov:fix memory leak in psp_load_fw
  drm/amdgpu:fix random missing of FLR NOTIFY

 drivers/gpu/drm/amd/amdgpu/amdgpu.h           |   9 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c    | 311 ++++++++++++--------------
 drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c     |  10 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c       |   2 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_job.c       |  18 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c       |   3 +
 drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c       |  22 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c     |   4 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c      |   2 -
 drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h      |   2 -
 drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c         |   6 +-
 drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c         |   6 +-
 drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c         |  16 +-
 drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c         |   2 +-
 drivers/gpu/drm/amd/scheduler/gpu_scheduler.c |  39 ++--
 15 files changed, 220 insertions(+), 232 deletions(-)

--
2.7.4

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx