[PATCH i-g-t 2/3] lib/intel_batchbuffer: Select xe2 rendercopy for LunarLake

[Matthew Auld]

Hi,

On Mon, 8 Jan 2024 at 11:36, Zbigniew Kempczyński
<zbigniew.kempczynski at intel.com> wrote:
>
> Along with rendercopy xe2 pipeline / shader selection for LunarLake
> reorganize if/else conditionals to handle specific selection first.
>
> Signed-off-by: Zbigniew Kempczyński <zbigniew.kempczynski at intel.com>
> Cc: Juha-Pekka Heikkila <juhapekka.heikkila at gmail.com>
> ---
>  lib/intel_batchbuffer.c | 38 ++++++++++++++++++++------------------
>  1 file changed, 20 insertions(+), 18 deletions(-)
>
> diff --git a/lib/intel_batchbuffer.c b/lib/intel_batchbuffer.c
> index ccab55cec7..d374645d99 100644
> --- a/lib/intel_batchbuffer.c
> +++ b/lib/intel_batchbuffer.c
> @@ -656,28 +656,30 @@ igt_render_copyfunc_t igt_get_render_copyfunc(int devid)
>  {
>         igt_render_copyfunc_t copy = NULL;
>
> -       if (IS_GEN2(devid))
> -               copy = gen2_render_copyfunc;
> -       else if (IS_GEN3(devid))
> -               copy = gen3_render_copyfunc;
> -       else if (IS_GEN4(devid) || IS_GEN5(devid))
> -               copy = gen4_render_copyfunc;
> -       else if (IS_GEN6(devid))
> -               copy = gen6_render_copyfunc;
> -       else if (IS_GEN7(devid))
> -               copy = gen7_render_copyfunc;
> -       else if (IS_GEN8(devid))
> -               copy = gen8_render_copyfunc;
> -       else if (IS_GEN9(devid) || IS_GEN10(devid))
> -               copy = gen9_render_copyfunc;
> -       else if (IS_GEN11(devid))
> -               copy = gen11_render_copyfunc;
> +       if (IS_METEORLAKE(devid))
> +               copy = mtl_render_copyfunc;
> +       else if (IS_LUNARLAKE(devid))
> +               copy = genxe2_render_copyfunc;

This doesn't work with any xe2? Can we not make this conditional on
graphics version == 20?

>         else if (HAS_FLATCCS(devid))
>                 copy = gen12p71_render_copyfunc;

Should we not not make this a normal platform or graphics version
check? This will also catch all xe2+ platforms that are not LNL, but
have flat_ccs support. I think it is better to just return NULL for
new platforms, that way we get a clear skip in tests using this
instead of seeing job timeouts due to using the wrong shader/kernel.
See VLK-54191 for example. What do you think?

> -       else if (IS_METEORLAKE(devid))
> -               copy = mtl_render_copyfunc;
>         else if (IS_GEN12(devid))
>                 copy = gen12_render_copyfunc;
> +       else if (IS_GEN11(devid))
> +               copy = gen11_render_copyfunc;
> +       else if (IS_GEN9(devid) || IS_GEN10(devid))
> +               copy = gen9_render_copyfunc;
> +       else if (IS_GEN8(devid))
> +               copy = gen8_render_copyfunc;
> +       else if (IS_GEN7(devid))
> +               copy = gen7_render_copyfunc;
> +       else if (IS_GEN6(devid))
> +               copy = gen6_render_copyfunc;
> +       else if (IS_GEN4(devid) || IS_GEN5(devid))
> +               copy = gen4_render_copyfunc;
> +       else if (IS_GEN3(devid))
> +               copy = gen3_render_copyfunc;
> +       else if (IS_GEN2(devid))
> +               copy = gen2_render_copyfunc;
>
>         return copy;
>  }
> --
> 2.34.1
>