[Intel-gfx] [PATCH] drm/i915/gt: Only trust sseu subslice fuse if it is set

Lionel Landwerlin lionel.g.landwerlin at intel.com
Sat Jan 30 12:42:12 UTC 2021


On 29/01/2021 14:46, Chris Wilson wrote:
> Since userspace cannot run without any subslices, it seems remarkable
> that any system would be configured with all fused off. Ignore the fuse
> register if it says 0.
>
> References: https://gitlab.freedesktop.org/drm/intel/-/issues/3022
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> ---
>   drivers/gpu/drm/i915/gt/intel_sseu.c | 7 ++++---
>   1 file changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/gt/intel_sseu.c b/drivers/gpu/drm/i915/gt/intel_sseu.c
> index 0d9f74aec8fe..99c992db2ce7 100644
> --- a/drivers/gpu/drm/i915/gt/intel_sseu.c
> +++ b/drivers/gpu/drm/i915/gt/intel_sseu.c
> @@ -341,9 +341,10 @@ static void gen9_sseu_info_init(struct intel_gt *gt)
>   	 * The subslice disable field is global, i.e. it applies
>   	 * to each of the enabled slices.
>   	 */
> -	subslice_mask = (1 << sseu->max_subslices) - 1;
> -	subslice_mask &= ~((fuse2 & GEN9_F2_SS_DIS_MASK) >>
> -			   GEN9_F2_SS_DIS_SHIFT);
> +	subslice_mask = (fuse2 & GEN9_F2_SS_DIS_MASK) >> GEN9_F2_SS_DIS_SHIFT;
> +	if (!subslice_mask) /* Ignore the fuse if it says there is no HW */
> +		subslice_mask = ~0u;
> +	subslice_mask &= GENMASK(sseu->max_subslices - 1, 0);
>   
>   	/*
>   	 * Iterate through enabled slices and subslices to

I'm afraid it's not enough, as far as I can tell even the slice_mask is 
wrong (user device is GT2, can't have 3 slices).

Time to reach out to HW folks.


-Lionel



More information about the Intel-gfx mailing list