[PATCH] drm/amdgpu: extend xnack limit page fault timeout

[Felix Kuehling]

Am 2021-04-19 um 1:22 p.m. schrieb Alex Sierra:
> Extending this timeout will prevent IH from storm interrupts coming
> from SDMA while a page fault is active. Currently, on Aldebaran,
> handling that many interrupts can take a lot of CPU time
> (up to 4 seconds).
> This eventually causes timeouts in other tasks.
>
> Signed-off-by: Alex Sierra <alex.sierra at amd.com>

Reviewed-by: Felix Kuehling <Felix.Kuehling at amd.com>


> ---
>  drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
> index 5715be6770ec..823a367990bf 100644
> --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
> @@ -1109,6 +1109,8 @@ static void sdma_v4_0_ctx_switch_enable(struct amdgpu_device *adev, bool enable)
>  		if (adev->asic_type == CHIP_ARCTURUS &&
>  		    adev->sdma.instance[i].fw_version >= 14)
>  			WREG32_SDMA(i, mmSDMA0_PUB_DUMMY_REG2, enable);
> +		/* Extend page fault timeout to avoid interrupt storm */
> +		WREG32_SDMA(i, mmSDMA0_UTCL1_TIMEOUT, 0x00800080);
>  	}
>  
>  }