Re: [PATCH 09/11] drm/amdgpu/vce2: Fix VCE 2 firmware size and offsets

[John Olender <john.olender@gmail.com>]

On 4/22/26 9:16 PM, Timur Kristóf wrote:
> The VCPU BO contains the actual FW at an offset, but
> it was not calculated into the VCPU BO size.
> Subtract this from the FW size to make sure there is
> no out of bounds access.
> 
> This may fix VM faults when using VCE 2.

I gave this patchset a test on Hawaii and Ellesmere.

I'm still seeing VM faults and card lockup on Hawaii.

Link: https://gitlab.freedesktop.org/drm/amd/-/work_items/4802

Adding additional padding to the VCE VCPU BO (e.g.: rounding up to 1MB)
does seem like an effective workaround, both when the the VCPU BO is
placed in VRAM or in GTT.

I don't see this problem on Ellesmere both with and without this
patchset.

Thanks,
John

> 
> Cc: John Olender <john.olender@gmail.com>
> Closes: https://gitlab.freedesktop.org/drm/amd/-/work_items/4802
> Fixes: e98226221467 ("drm/amdgpu: recalculate VCE firmware BO size")
> Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
> ---
>  drivers/gpu/drm/amd/amdgpu/vce_v2_0.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
> index 00b4037d4bc89..3b493a2e94dd0 100644
> --- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
> @@ -183,7 +183,7 @@ static void vce_v2_0_mc_resume(struct amdgpu_device *adev)
>  	WREG32(mmVCE_LMI_VCPU_CACHE_40BIT_BAR, (adev->vce.gpu_addr >> 8));
>  
>  	offset = AMDGPU_VCE_FIRMWARE_OFFSET;
> -	size = VCE_V2_0_FW_SIZE;
> +	size = VCE_V2_0_FW_SIZE - AMDGPU_VCE_FIRMWARE_OFFSET;
>  	WREG32(mmVCE_VCPU_CACHE_OFFSET0, offset & 0x7fffffff);
>  	WREG32(mmVCE_VCPU_CACHE_SIZE0, size);
>