update conv1d to new layout specifier gen, axis mapping, and use non-…

…singlethreaded local workgroup (pytorch#5504) Summary: Pull Request resolved: pytorch#5504 Using new load_texel_lpos for simpler updating Reviewed By: SS-JIA Differential Revision: D62990822 fbshipit-source-id: 9163b807d9095ebdb089f08aa6ea20fbbb563d02
mybigday · Sep 21, 2024 · d5fdbd4 · d5fdbd4
1 parent 0eee42a
commit d5fdbd4
Show file tree

Hide file tree

Showing 3 changed files with 39 additions and 46 deletions.
diff --git a/backends/vulkan/runtime/graph/ops/glsl/conv1d.glsl b/backends/vulkan/runtime/graph/ops/glsl/conv1d.glsl
@@ -18,32 +18,22 @@
 
 layout(std430) buffer;
 
-layout(set = 0, binding = 0, ${IMAGE_FORMAT[DTYPE]}) uniform PRECISION restrict writeonly ${IMAGE_T[NDIM][DTYPE]} image_out;
-layout(set = 0, binding = 1) uniform PRECISION sampler3D image_in;
-layout(set = 0, binding = 2) uniform PRECISION sampler3D kernel_in;
-layout(set = 0, binding = 3) uniform PRECISION sampler3D bias_in;
-
-layout(set = 0, binding = 4) uniform PRECISION restrict OutLimits {
-  ivec3 out_limits;
-};
-
-layout(set = 0, binding = 5) uniform PRECISION restrict InSizes {
-  ivec4 in_sizes;
-};
-
-layout(set = 0, binding = 6) uniform PRECISION restrict Params {
-  int kernel_size;
-  int stride;
-  int padding;
-  int dilation;
-  int in_group_size;
-  int out_group_size;
-};
-
-layout(set = 0, binding = 7) uniform PRECISION restrict OutputParams {
-  float out_min;
-  float out_max;
-};
+${layout_declare_tensor(B, "w", "t_out", DTYPE, STORAGE)}
+${layout_declare_tensor(B, "r", "t_in", DTYPE, STORAGE)}
+${layout_declare_tensor(B, "r", "kernel_in", DTYPE, STORAGE)}
+${layout_declare_tensor(B, "r", "bias_in", DTYPE, STORAGE)}
+
+${layout_declare_ubo(B, "ivec3", "out_limits")}
+${layout_declare_ubo(B, "ivec4", "in_sizes")}
+
+${layout_declare_ubo(B, "ivec4", "out_axis_map")}
+${layout_declare_ubo(B, "ivec4", "in_axis_map")}
+${layout_declare_ubo(B, "ivec4", "kernel_axis_map")}
+${layout_declare_ubo(B, "ivec4", "bias_axis_map")}
+
+${layout_declare_ubo(B,"int", "kernel_size", "int", "stride", "int", "padding", "int", "dilation", "int", "in_group_size", "int", "out_group_size")}
+
+${layout_declare_ubo(B, "float", "out_min", "float", "out_max")}
 
 layout(local_size_x_id = 0, local_size_y_id = 1, local_size_z_id = 2) in;
 
@@ -67,9 +57,9 @@ layout(local_size_x_id = 0, local_size_y_id = 1, local_size_z_id = 2) in;
 // shader invocations, where each invocation computes 1 result. But that
 // performs worse.
 void main() {
-  const ivec3 pos = ivec3(gl_GlobalInvocationID);
+  const ivec3 lpos = ivec3(gl_GlobalInvocationID);
 
-  if (any(greaterThanEqual(pos, out_limits))) {
+  if (any(greaterThanEqual(lpos, out_limits))) {
     return;
   }
 
@@ -78,8 +68,8 @@ void main() {
 
   // "out_c" is the output's channel index where we write our result.
   // Across shader invocations, this is the only value that varies.
-  int out_c = pos.y;
-  vec4 bias = texelFetch(bias_in, ivec3(out_c, 0, 0), 0);
+  int out_c = lpos.y;
+  VEC4_T bias = load_texel_lpos(bias_in, ivec3(out_c, 0, 0), bias_axis_map);
 
   // "in_c" tracks the input's channel start index.
   // We iterate over the input group that corresponds to the output group.
@@ -98,7 +88,7 @@ void main() {
     int out_l = 0;
 
     for (int in_l = l_start; in_l < l_end; in_l += stride, ++out_l) {
-      vec4 sum = vec4(0);
+      VEC4_T sum = VEC4_T(0);
 
       for (int in_c = c_start; in_c < c_end; ++in_c) {
         // "k" tracks the kernel's index for our input-kernel computation.
@@ -107,25 +97,25 @@ void main() {
         for (int k = 0; k < kernel_size; k += 4) {
           // Since the weight tensor is width-packed, which is along the length
           // dimension, we can batch-read four elements at a time.
-          const ivec3 w_pos = ivec3(k / 4, in_c % in_group_size, out_c);
-          const vec4 weight = texelFetch(kernel_in, w_pos, 0);
+          const ivec3 w_lpos = ivec3(k / 4, in_c % in_group_size, out_c);
+          const VEC4_T weight = load_texel_lpos(kernel_in, w_lpos, kernel_axis_map);
 
-          const ivec3 in_pos_0 = ivec3(in_l + k * dilation, in_c, n / 4);
-          sum = fma(weight.xxxx, texelFetch(image_in, in_pos_0, 0), sum);
+          ivec3 in_pos = lpos_to_pos(ivec3(in_l + k * dilation, in_c, n / 4), in_axis_map);
+          sum = fma(weight.xxxx, load_texel(t_in, in_pos), sum);
 
-          const ivec3 in_pos_1 = ivec3(in_l + (k+1) * dilation, in_c, n / 4);
-          sum = fma(weight.yyyy, texelFetch(image_in, in_pos_1, 0), sum);
+          in_pos[in_axis_map.x] += dilation;
+          sum = fma(weight.yyyy, load_texel(t_in, in_pos), sum);
 
-          const ivec3 in_pos_2 = ivec3(in_l + (k+2) * dilation, in_c, n / 4);
-          sum = fma(weight.zzzz, texelFetch(image_in, in_pos_2, 0), sum);
+          in_pos[in_axis_map.x] += dilation;
+          sum = fma(weight.zzzz, load_texel(t_in, in_pos), sum);
 
-          const ivec3 in_pos_3 = ivec3(in_l + (k+3) * dilation, in_c, n / 4);
-          sum = fma(weight.wwww, texelFetch(image_in, in_pos_3, 0), sum);
+          in_pos[in_axis_map.x] += dilation;
+          sum = fma(weight.wwww, load_texel(t_in, in_pos), sum);
         }
       }
 
-      ivec3 out_pos = ivec3(out_l, out_c, n / 4);
-      imageStore(image_out, out_pos, op(sum + bias.x, out_min, out_max));
+      const ivec3 out_lpos = ivec3(out_l, out_c, n / 4);
+      write_texel_lpos(t_out, out_lpos, op(sum + bias.x, out_min, out_max), out_axis_map);
     }
   }
 }
diff --git a/backends/vulkan/runtime/graph/ops/glsl/conv1d.yaml b/backends/vulkan/runtime/graph/ops/glsl/conv1d.yaml
@@ -7,9 +7,8 @@
 conv1d:
   parameter_names_with_default_values:
     OPERATOR: X
-    NDIM: 3
     DTYPE: float
-    PACKING: C_packed
+    STORAGE: texture3d
   generate_variant_forall:
     DTYPE:
       - VALUE: half

diff --git a/backends/vulkan/runtime/graph/ops/impl/Convolution.cpp b/backends/vulkan/runtime/graph/ops/impl/Convolution.cpp
@@ -444,7 +444,7 @@ void add_conv1d_node(
   int32_t out_group_size = static_cast<int64_t>(out_channels / groups_val);
 
   utils::uvec3 global_size = {1, static_cast<uint32_t>(out_channels), 1};
-  utils::uvec3 local_size = {1, 1, 1};
+  utils::uvec3 local_size = {1, 64, 1};
 
   Kernel1dParams kernel_params = {
       kernel_size,
@@ -476,6 +476,10 @@ void add_conv1d_node(
       {
           t_out->logical_limits_ubo(),
           t_in->sizes_ubo(),
+          t_out->axis_map_ubo(),
+          t_in->axis_map_ubo(),
+          t_weight->axis_map_ubo(),
+          t_bias->axis_map_ubo(),
           graph.create_params_buffer(kernel_params),
           graph.create_params_buffer(out_params),
       },