[Mesa-dev] [PATCH v2] freedreno/ir3: Make imageStore use num components from image format

[Eduardo Lima Mitev]

emit_intrinsic_store_image() is always using 4 components when
collecting registers for the value. When image has less than
4 components (e.g, r32f, r32i, r32ui) this results in extra mov
instructions.

This patch uses the actual number of components from the image format.

For example, in a shader like:

layout (r32f, binding=0) writeonly uniform imageBuffer u_image;
...
void main(void) {
   ...
   imageStore (u_image, some_offset, vec4(1.0));
   ...
}

instruction count is reduced in at least 3 instructions (note image
format is r32f, 1 component only).

This obviously reduces register pressure as well.

v2: - Added support for image formats from NV_image_format extension
    (Ilia Mirkin).
    - Return 4 components by default instead of asserting. (Rob Clark).
---
 src/freedreno/ir3/ir3_compiler_nir.c | 63 +++++++++++++++++++++++++++-
 1 file changed, 61 insertions(+), 2 deletions(-)

diff --git a/src/freedreno/ir3/ir3_compiler_nir.c b/src/freedreno/ir3/ir3_compiler_nir.c
index 85f14f354d2..e33d2b8fd46 100644
--- a/src/freedreno/ir3/ir3_compiler_nir.c
+++ b/src/freedreno/ir3/ir3_compiler_nir.c
@@ -1251,6 +1251,64 @@ emit_intrinsic_load_image(struct ir3_context *ctx, nir_intrinsic_instr *intr,
 	ir3_split_dest(b, dst, sam, 0, 4);
 }
 
+/* Returns the number of components for the different image formats
+ * supported by the GLES 3.1 spec, plus those added by the
+ * GL_NV_image_formats extension.
+ */
+static unsigned
+get_num_components_for_glformat(GLuint format)
+{
+	switch (format) {
+	case GL_R32F:
+	case GL_R16F:
+	case GL_R32I:
+	case GL_R32UI:
+	case GL_R16I:
+	case GL_R16UI:
+	case GL_R8I:
+	case GL_R8UI:
+	case GL_R8:
+	case GL_R8_SNORM:
+		return 1;
+
+	case GL_RG32F:
+	case GL_RG16F:
+	case GL_RG32I:
+	case GL_RG32UI:
+	case GL_RG16I:
+	case GL_RG16UI:
+	case GL_RG8I:
+	case GL_RG8UI:
+	case GL_RG8:
+	case GL_RG8_SNORM:
+		return 2;
+
+	case GL_R11F_G11F_B10F:
+		return 3;
+
+	case GL_RGBA32F:
+	case GL_RGBA16F:
+	case GL_RGBA8:
+	case GL_RGBA8_SNORM:
+	case GL_RGBA32I:
+	case GL_RGBA16I:
+	case GL_RGBA8I:
+	case GL_RGBA32UI:
+	case GL_RGBA16UI:
+	case GL_RGBA8UI:
+	case GL_RGB10_A2UI:
+	case GL_RGB10_A2:
+		return 4;
+
+	default:
+		/* Return 4 components also for all other formats we don't know
+		 * about. This is always safe. Also, the format should have been
+		 * validated already by the higher level API.
+		 */
+		return 4;
+	}
+}
+
 /* src[] = { deref, coord, sample_index, value }. const_index[] = {} */
 static void
 emit_intrinsic_store_image(struct ir3_context *ctx, nir_intrinsic_instr *intr)
@@ -1262,6 +1320,7 @@ emit_intrinsic_store_image(struct ir3_context *ctx, nir_intrinsic_instr *intr)
 	struct ir3_instruction * const *coords = ir3_get_src(ctx, &intr->src[1]);
 	unsigned ncoords = get_image_coords(var, NULL);
 	unsigned tex_idx = get_image_slot(ctx, nir_src_as_deref(intr->src[0]));
+	unsigned ncomp = get_num_components_for_glformat(var->data.image.format);
 
 	/* src0 is value
 	 * src1 is coords
@@ -1276,10 +1335,10 @@ emit_intrinsic_store_image(struct ir3_context *ctx, nir_intrinsic_instr *intr)
 	 */
 
 	stib = ir3_STIB(b, create_immed(b, tex_idx), 0,
-			ir3_create_collect(ctx, value, 4), 0,
+			ir3_create_collect(ctx, value, ncomp), 0,
 			ir3_create_collect(ctx, coords, ncoords), 0,
 			offset, 0);
-	stib->cat6.iim_val = 4;
+	stib->cat6.iim_val = ncomp;
 	stib->cat6.d = ncoords;
 	stib->cat6.type = get_image_type(var);
 	stib->cat6.typed = true;
-- 
2.19.2