[PATCH v6 0/3] MR188: vkd3d-shader/tpf: Add support for emitting sample_l instructions

[Ethan Lee (＠flibitijibibo)]

From what I can tell, the recent work on SampleBias/SampleLevel did all of the work for us, we just need to take the same framework and include the case for the `sample_l` instruction.

Tested with some shaders from the native Linux version of Little Racers STREET.

--
  v6: vkd3d-shader/tpf: For sample_l/sample_b, set lod swizzle to SCALAR.
      vkd3d-shader/tpf: Add support for emitting sample_l instructions
      tests: Add a test for SampleLevel() function.

https://gitlab.winehq.org/wine/vkd3d/-/merge_requests/188

[PATCH v6 1/3] tests: Add a test for SampleLevel() function.

[Ethan Lee]

From: Ethan Lee

Currently marked TODO, since the SM4 emitter does not support sample_l yet.

Signed-off-by: Ethan Lee 
---
 tests/sampler.shader_test | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/tests/sampler.shader_test b/tests/sampler.shader_test
index 23741a35..81b0d2d0 100644
--- a/tests/sampler.shader_test
+++ b/tests/sampler.shader_test
@@ -33,6 +33,19 @@ float4 main() : sv_target
 draw quad
 probe all rgba (0.25, 0, 0.25, 0)

+[pixel shader todo]
+SamplerState s;
+Texture2D t;
+
+float4 main() : sv_target
+{
+    return t.SampleLevel(s, float2(0.5, 0.5), 0.0);
+}
+
+[test]
+todo draw quad
+todo probe all rgba (0.25, 0, 0.25, 0)
+
 [pixel shader]
 SamplerState s;
 Texture2D t;

[PATCH v6 2/3] vkd3d-shader/tpf: Add support for emitting sample_l instructions

[Ethan Lee]

From: Ethan Lee

This gets the SampleLevel() HLSL test to pass.

Signed-off-by: Ethan Lee 
---
 libs/vkd3d-shader/tpf.c   | 12 +++++++-----
 tests/sampler.shader_test |  6 +++---
 2 files changed, 10 insertions(+), 8 deletions(-)

diff --git a/libs/vkd3d-shader/tpf.c b/libs/vkd3d-shader/tpf.c
index 38d5c180..f39491e2 100644
--- a/libs/vkd3d-shader/tpf.c
+++ b/libs/vkd3d-shader/tpf.c
@@ -3708,6 +3708,10 @@ static void write_sm4_sample(struct hlsl_ctx *ctx, struct vkd3d_bytecode_buffer
             instr.opcode = VKD3D_SM4_OP_SAMPLE;
             break;

+        case HLSL_RESOURCE_SAMPLE_LOD:
+            instr.opcode = VKD3D_SM4_OP_SAMPLE_LOD;
+            break;
+
         case HLSL_RESOURCE_SAMPLE_LOD_BIAS:
             instr.opcode = VKD3D_SM4_OP_SAMPLE_B;
             break;
@@ -3734,7 +3738,8 @@ static void write_sm4_sample(struct hlsl_ctx *ctx, struct vkd3d_bytecode_buffer
     sm4_src_from_deref(ctx, &instr.srcs[2], sampler, sampler->var->data_type, VKD3DSP_WRITEMASK_ALL);
     instr.src_count = 3;

-    if (load->load_type == HLSL_RESOURCE_SAMPLE_LOD_BIAS)
+    if (load->load_type == HLSL_RESOURCE_SAMPLE_LOD
+           || load->load_type == HLSL_RESOURCE_SAMPLE_LOD_BIAS)
     {
         sm4_src_from_node(&instr.srcs[3], load->lod.node, VKD3DSP_WRITEMASK_ALL);
         ++instr.src_count;
@@ -4540,6 +4545,7 @@ static void write_sm4_resource_load(struct hlsl_ctx *ctx,
             break;

case HLSL_RESOURCE_SAMPLE:
+        case HLSL_RESOURCE_SAMPLE_LOD:
         case HLSL_RESOURCE_SAMPLE_LOD_BIAS:
             if (!load->sampler.var)
             {
@@ -4568,10 +4574,6 @@ static void write_sm4_resource_load(struct hlsl_ctx *ctx,
             write_sm4_gather(ctx, buffer, resource_type, &load->node, &load->resource,
                     &load->sampler, coords, HLSL_SWIZZLE(W, W, W, W), texel_offset);
             break;
-
-        case HLSL_RESOURCE_SAMPLE_LOD:
-            hlsl_fixme(ctx, &load->node.loc, "SM4 sample-LOD expression.");
-            break;
     }
 }

diff --git a/tests/sampler.shader_test b/tests/sampler.shader_test
index 81b0d2d0..f03d050d 100644
--- a/tests/sampler.shader_test
+++ b/tests/sampler.shader_test
@@ -33,7 +33,7 @@ float4 main() : sv_target
 draw quad
 probe all rgba (0.25, 0, 0.25, 0)

-[pixel shader todo]
+[pixel shader]
 SamplerState s;
 Texture2D t;

@@ -43,8 +43,8 @@ float4 main() : sv_target
 }

[test]
-todo draw quad
-todo probe all rgba (0.25, 0, 0.25, 0)
+draw quad
+probe all rgba (0.25, 0, 0.25, 0)

[pixel shader]
 SamplerState s;

[PATCH v6 3/3] vkd3d-shader/tpf: For sample_l/sample_b, set lod swizzle to SCALAR.

[Ethan Lee]

From: Ethan Lee

Per Francisco Casas :

I just realized one more thing. I think we are not properly following the specification for both sample_l and sample_b.

In the Restrictions section of the documentation for sample_b it is mentioned that the last src register must use a single component selector if it is not a scalar immediate. This "select_component" is also ilustrated in the sample_l format:

sample_l[_aoffimmi(u,v,w)] dest[.mask], srcAddress[.swizzle], srcResource[.swizzle], srcSampler, srcLOD.select_component

Currently, we are not respecting that since sm4_src_from_node() sets

src->swizzle_type = VKD3D_SM4_SWIZZLE_VEC4;

by default.

It should be VKD3D_SM4_SWIZZLE_SCALAR in this case... The disassembly should show a single component in the swizzle of the last src register.

Signed-off-by: Ethan Lee 
---
 libs/vkd3d-shader/tpf.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/libs/vkd3d-shader/tpf.c b/libs/vkd3d-shader/tpf.c
index f39491e2..a82b79bb 100644
--- a/libs/vkd3d-shader/tpf.c
+++ b/libs/vkd3d-shader/tpf.c
@@ -3742,6 +3742,7 @@ static void write_sm4_sample(struct hlsl_ctx *ctx, struct vkd3d_bytecode_buffer
            || load->load_type == HLSL_RESOURCE_SAMPLE_LOD_BIAS)
     {
         sm4_src_from_node(&instr.srcs[3], load->lod.node, VKD3DSP_WRITEMASK_ALL);
+        instr.srcs[3].swizzle_type = VKD3D_SM4_SWIZZLE_SCALAR;
         ++instr.src_count;
     }

[Ethan Lee (＠flibitijibibo)]

The test commit in this MR could be obsoleted by #191, depending on when each MR is merged.