Make num_blocks a uniform.

[narabu] / decoder.shader

diff --git a/decoder.shader b/decoder.shader
index 8b772c70fbefdd228ce6be25c6faff7ae1cea045..e7b83e0ccbb52402dcb0a827119fe272eff892fc 100644 (file)
--- a/decoder.shader
+++ b/decoder.shader
@@ -10,6 +10,7 @@ layout(r8ui) uniform restrict readonly uimage2D cum2sym_tex;
 layout(rg16ui) uniform restrict readonly uimage2D dsyms_tex;
 layout(r8) uniform restrict writeonly image2D out_tex;
 layout(r16i) uniform restrict writeonly iimage2D coeff_tex;
+uniform int num_blocks;
 
 const uint prob_bits = 12;
 const uint prob_scale = 1 << prob_bits;
@@ -40,6 +41,16 @@ const uint ff_zigzag_direct[64] = {
     58, 59, 52, 45, 38, 31, 39, 46,
     53, 60, 61, 54, 47, 55, 62, 63
 };
+const uint stream_mapping[64] = {
+       0, 0, 1, 1, 2, 2, 3, 3,
+       0, 0, 1, 2, 2, 2, 3, 3,
+       1, 1, 2, 2, 2, 3, 3, 3,
+       1, 1, 2, 2, 2, 3, 3, 3,
+       1, 2, 2, 2, 2, 3, 3, 3,
+       2, 2, 2, 2, 3, 3, 3, 3,
+       2, 2, 3, 3, 3, 3, 3, 3,
+       3, 3, 3, 3, 3, 3, 3, 3,
+};
 
 layout(std430, binding = 9) buffer layoutName
 {
@@ -196,7 +207,6 @@ void main()
        const uint local_y = (gl_LocalInvocationID.x / 8) % 8;
        const uint local_z = gl_LocalInvocationID.x / 64;
 
-       const uint num_blocks = 720 / 16;  // FIXME: make a uniform
        const uint slice_num = local_z;
        const uint thread_num = local_y * 8 + local_x;
 
@@ -204,7 +214,7 @@ void main()
        //const uint coeff_num = ff_zigzag_direct[thread_num];
        const uint coeff_num = thread_num;
        const uint stream_num = coeff_num * num_blocks + block_row;
-       const uint model_num = min((coeff_num % 8) + (coeff_num / 8), 7);
+       const uint model_num = stream_mapping[coeff_num];
        const uint sign_bias = sign_bias_per_model[model_num];
 
        // Initialize rANS decoder.