[Mesa-dev] [PATCH 2/2] nir: add lowering for gl_HelperInvocation

[Rob Clark]

On Mon, Jun 11, 2018 at 6:34 PM, Jason Ekstrand <jason at jlekstrand.net> wrote:
> On Mon, Jun 11, 2018 at 3:32 PM, Jason Ekstrand <jason at jlekstrand.net>
> wrote:
>>
>> On Wed, Jun 6, 2018 at 7:43 AM, Rob Clark <robdclark at gmail.com> wrote:
>>>
>>> Signed-off-by: Rob Clark <robdclark at gmail.com>
>>> ---
>>> I can't say for sure that this will work on all drivers, but it is
>>> what the blob driver does, and it seems to make deqp happy.  I could
>>> move this to it's own pass inside ir3, but that seemed like overkill
>>>
>>>  src/compiler/nir/nir.h                      | 10 ++++++++++
>>>  src/compiler/nir/nir_lower_system_values.c  | 17 +++++++++++++++++
>>>  src/gallium/drivers/freedreno/ir3/ir3_nir.c |  1 +
>>>  3 files changed, 28 insertions(+)
>>>
>>> diff --git a/src/compiler/nir/nir.h b/src/compiler/nir/nir.h
>>> index 073ab4e82ea..de3d55d83af 100644
>>> --- a/src/compiler/nir/nir.h
>>> +++ b/src/compiler/nir/nir.h
>>> @@ -1963,6 +1963,16 @@ typedef struct nir_shader_compiler_options {
>>>      */
>>>     bool lower_base_vertex;
>>>
>>> +   /**
>>> +    * If enabled, gl_HelperInvocation will be lowered as:
>>> +    *
>>> +    *   !((1 << gl_SampleID) & gl_SampleMaskIN[0]))
>>
>>
>> This only works for multi-sampling.  What about the single-sampled case?
>
>
> Actually, I'm not even sure that it would work for multisampling for us.
> What about 2x MSAA?  There you are probably going to have two pixels
> involved in order to get derivatives.
>

so it definitely works on single-sampling case, at least on adreno.
That is really the only case I've tested yet, but afaict blob does
same thing in various MSAA cases, based on what I see in cmdstream
traces..

Maybe it is relying on something arguably hw specific, ie. the hw
isn't going to schedule a thread unless it is (a) covered, or (b)
helper.. which maybe isn't true on other hw, but then I guess in those
cases the driver wouldn't need to lower gl_HelperInvocation since
there would have to be some other way to implement
gl_HelperInvocation.

If this seems like something too driver specific, I'll just roll it
into my own ir3 private pass.. but that seemed overkill.

BR,
-R


> --Jason
>
>
>>>
>>> +    *
>>> +    * TODO any hw w/ more than 32 samples?  For them (if they
>>> +    * used this option), a bit more math would be involved.
>>> +    */
>>> +   bool lower_helper_invocation;
>>> +
>>>     bool lower_cs_local_index_from_id;
>>>
>>>     bool lower_device_index_to_zero;
>>> diff --git a/src/compiler/nir/nir_lower_system_values.c
>>> b/src/compiler/nir/nir_lower_system_values.c
>>> index 487da042620..6668cbb5dcd 100644
>>> --- a/src/compiler/nir/nir_lower_system_values.c
>>> +++ b/src/compiler/nir/nir_lower_system_values.c
>>> @@ -136,6 +136,23 @@ convert_block(nir_block *block, nir_builder *b)
>>>                                nir_load_first_vertex(b));
>>>           break;
>>>
>>> +      case SYSTEM_VALUE_HELPER_INVOCATION:
>>> +         if (b->shader->options->lower_helper_invocation) {
>>> +            nir_ssa_def *tmp;
>>> +
>>> +            tmp = nir_ushr(b,
>>> +                           nir_imm_int(b, 1),
>>> +                           nir_load_sample_id(b));
>>> +
>>> +            tmp = nir_iand(b,
>>> +                           nir_load_sample_mask_in(b),
>>> +                           tmp);
>>> +
>>> +            sysval = nir_inot(b, nir_i2b(b, tmp));
>>> +         }
>>> +
>>> +         break;
>>> +
>>>        case SYSTEM_VALUE_INSTANCE_INDEX:
>>>           sysval = nir_iadd(b,
>>>                             nir_load_instance_id(b),
>>> diff --git a/src/gallium/drivers/freedreno/ir3/ir3_nir.c
>>> b/src/gallium/drivers/freedreno/ir3/ir3_nir.c
>>> index cd1f9c526f2..341d990b269 100644
>>> --- a/src/gallium/drivers/freedreno/ir3/ir3_nir.c
>>> +++ b/src/gallium/drivers/freedreno/ir3/ir3_nir.c
>>> @@ -51,6 +51,7 @@ static const nir_shader_compiler_options options = {
>>>                 .lower_extract_byte = true,
>>>                 .lower_extract_word = true,
>>>                 .lower_all_io_to_temps = true,
>>> +               .lower_helper_invocation = true,
>>>  };
>>>
>>>  struct nir_shader *
>>> --
>>> 2.17.0
>>>
>>> _______________________________________________
>>> mesa-dev mailing list
>>> mesa-dev at lists.freedesktop.org
>>> https://lists.freedesktop.org/mailman/listinfo/mesa-dev
>>
>>
>