[Mesa-dev] [PATCH 4/6] st/mesa: implement ARB_shader_clock
Nicolai Hähnle
nhaehnle at gmail.com
Thu Mar 30 11:41:22 UTC 2017
On 30.03.2017 11:41, Samuel Pitoiset wrote:
>
>
> On 03/30/2017 09:38 AM, Nicolai Hähnle wrote:
>> From: Nicolai Hähnle <nicolai.haehnle at amd.com>
>>
>> ---
>> src/mesa/state_tracker/st_extensions.c | 1 +
>> src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 11 ++++++++++-
>> 2 files changed, 11 insertions(+), 1 deletion(-)
>>
>> diff --git a/src/mesa/state_tracker/st_extensions.c
>> b/src/mesa/state_tracker/st_extensions.c
>> index 16f8685..a48c22e 100644
>> --- a/src/mesa/state_tracker/st_extensions.c
>> +++ b/src/mesa/state_tracker/st_extensions.c
>> @@ -592,20 +592,21 @@ void st_init_extensions(struct pipe_screen *screen,
>> { o(ARB_indirect_parameters),
>> PIPE_CAP_MULTI_DRAW_INDIRECT_PARAMS },
>> { o(ARB_instanced_arrays),
>> PIPE_CAP_VERTEX_ELEMENT_INSTANCE_DIVISOR },
>> { o(ARB_occlusion_query),
>> PIPE_CAP_OCCLUSION_QUERY },
>> { o(ARB_occlusion_query2),
>> PIPE_CAP_OCCLUSION_QUERY },
>> { o(ARB_pipeline_statistics_query),
>> PIPE_CAP_QUERY_PIPELINE_STATISTICS },
>> { o(ARB_point_sprite),
>> PIPE_CAP_POINT_SPRITE },
>> { o(ARB_query_buffer_object),
>> PIPE_CAP_QUERY_BUFFER_OBJECT },
>> { o(ARB_robust_buffer_access_behavior),
>> PIPE_CAP_ROBUST_BUFFER_ACCESS_BEHAVIOR },
>> { o(ARB_sample_shading),
>> PIPE_CAP_SAMPLE_SHADING },
>> { o(ARB_seamless_cube_map),
>> PIPE_CAP_SEAMLESS_CUBE_MAP },
>> + { o(ARB_shader_clock),
>> PIPE_CAP_TGSI_CLOCK },
>> { o(ARB_shader_draw_parameters),
>> PIPE_CAP_DRAW_PARAMETERS },
>> { o(ARB_shader_group_vote),
>> PIPE_CAP_TGSI_VOTE },
>> { o(ARB_shader_stencil_export),
>> PIPE_CAP_SHADER_STENCIL_EXPORT },
>> { o(ARB_shader_texture_image_samples),
>> PIPE_CAP_TGSI_TXQS },
>> { o(ARB_shader_texture_lod),
>> PIPE_CAP_SM3 },
>> { o(ARB_shadow),
>> PIPE_CAP_TEXTURE_SHADOW_MAP },
>> { o(ARB_texture_buffer_object),
>> PIPE_CAP_TEXTURE_BUFFER_OBJECTS },
>> { o(ARB_texture_cube_map_array),
>> PIPE_CAP_CUBE_MAP_ARRAY },
>> { o(ARB_texture_gather),
>> PIPE_CAP_MAX_TEXTURE_GATHER_COMPONENTS },
>> { o(ARB_texture_mirror_clamp_to_edge),
>> PIPE_CAP_TEXTURE_MIRROR_CLAMP },
>> diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
>> b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
>> index 63de74a..6e3bccd 100644
>> --- a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
>> +++ b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
>> @@ -3988,32 +3988,41 @@ glsl_to_tgsi_visitor::visit(ir_call *ir)
>> case ir_intrinsic_image_atomic_and:
>> case ir_intrinsic_image_atomic_or:
>> case ir_intrinsic_image_atomic_xor:
>> case ir_intrinsic_image_atomic_exchange:
>> case ir_intrinsic_image_atomic_comp_swap:
>> case ir_intrinsic_image_size:
>> case ir_intrinsic_image_samples:
>> visit_image_intrinsic(ir);
>> return;
>>
>> + case ir_intrinsic_shader_clock: {
>> + ir->return_deref->accept(this);
>> +
>> + st_dst_reg dst = st_dst_reg(this->result);
>> + dst.writemask = TGSI_WRITEMASK_XY;
>> +
>> + emit_asm(ir, TGSI_OPCODE_CLOCK, dst);
>> + return;
>> + }
>
> Maybe add visit_clock_intrinsic() for consistency? Although, there is
> only one intrinsic for that. Your call.
I didn't do that on purpose, because it's just a few lines.
Thanks for taking a look!
Nicolai
>
>> +
>> case ir_intrinsic_invalid:
>> case ir_intrinsic_generic_load:
>> case ir_intrinsic_generic_store:
>> case ir_intrinsic_generic_atomic_add:
>> case ir_intrinsic_generic_atomic_and:
>> case ir_intrinsic_generic_atomic_or:
>> case ir_intrinsic_generic_atomic_xor:
>> case ir_intrinsic_generic_atomic_min:
>> case ir_intrinsic_generic_atomic_max:
>> case ir_intrinsic_generic_atomic_exchange:
>> case ir_intrinsic_generic_atomic_comp_swap:
>> - case ir_intrinsic_shader_clock:
>> unreachable("Invalid intrinsic");
>> }
>> }
>>
>> void
>> glsl_to_tgsi_visitor::calc_deref_offsets(ir_dereference *tail,
>> unsigned *array_elements,
>> uint16_t *index,
>> st_src_reg *indirect,
>> unsigned *location)
>>
--
Lerne, wie die Welt wirklich ist,
Aber vergiss niemals, wie sie sein sollte.
More information about the mesa-dev
mailing list