[igt-dev] [PATCH i-g-t] tests/i915/kms_draw_crc : check tiling support
Juha-Pekka Heikkila
juhapekka.heikkila at gmail.com
Tue Jan 11 21:39:15 UTC 2022
Why is this patch showing up again? Let's not do this for the second
time either.
On 11.1.2022 12.54, Jeevan B wrote:
> From: Tejas Upadhyay <tejaskumarx.surendrakumar.upadhyay at intel.com>
>
> We can check tiling support before executing test to avoid
> failure and smoothly skipping expected failure.
>
> Signed-off-by: Jeevan B <jeevan.b at intel.com>
> Signed-off-by: Tejas Upadhyay <tejaskumarx.surendrakumar.upadhyay at intel.com>
> ---
> lib/intel_bufops.c | 2 +-
> tests/i915/kms_draw_crc.c | 1 +
> 2 files changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/lib/intel_bufops.c b/lib/intel_bufops.c
> index f13063fa..4ba4271a 100644
> --- a/lib/intel_bufops.c
> +++ b/lib/intel_bufops.c
> @@ -1176,7 +1176,7 @@ void intel_buf_write_aux_to_png(struct intel_buf *buf, const char *filename)
> #define DEFAULT_BUFOPS(__gen_start, __gen_end) \
> .gen_start = __gen_start, \
> .gen_end = __gen_end, \
> - .supported_hw_tiles = TILE_X | TILE_Y, \
> + .supported_hw_tiles = TILE_NONE | TILE_X | TILE_Y, \
there's no need to think about (de-)tiling linear buffers.
> .linear_to = copy_linear_to_wc, \
> .linear_to_x = copy_linear_to_gtt, \
> .linear_to_y = copy_linear_to_gtt, \
> diff --git a/tests/i915/kms_draw_crc.c b/tests/i915/kms_draw_crc.c
> index 82c368e1..ff1520d1 100644
> --- a/tests/i915/kms_draw_crc.c
> +++ b/tests/i915/kms_draw_crc.c
> @@ -182,6 +182,7 @@ static void draw_method_subtest(enum igt_draw_method method,
> igt_skip_on(method == IGT_DRAW_MMAP_GTT &&
> !gem_has_mappable_ggtt(drm_fd));
>
> + igt_require(buf_ops_has_hw_fence(bops, modifier));
This will not work on hw where there is no hw detiling. See how
__buf_ops_create(..) works in igt, if needed it just will do tiling with
sw instead.
> igt_require(format_is_supported(formats[format_index], modifier));
>
> /* Use IGT_DRAW_MMAP_GTT/WC on an untiled buffer as the parameter for
More information about the igt-dev
mailing list