Skip to content

[ET-VK] Using push constants for convolution prepack dispatches. #11306

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Jun 4, 2025
Merged

[ET-VK] Using push constants for convolution prepack dispatches. #11306

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
0346905
[ET-VK] Using push constants for convolution prepack dispatches.
trivedivivek Jun 3, 2025
d4bc9a6
Update on "[ET-VK] Using push constants for convolution prepack dispa…
trivedivivek Jun 4, 2025
0e9cf7f
Update on "[ET-VK] Using push constants for convolution prepack dispa…
trivedivivek Jun 4, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 1 addition & 4 deletions backends/vulkan/runtime/graph/ops/glsl/conv2d_dw_prepack_weights.glsl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,11 +26,8 @@ layout(set = 0, binding = 1) buffer PRECISION restrict readonly Buffer {
BUF_T buffer_in[];
};

layout(set = 0, binding = 2) uniform PRECISION restrict Sizes {
layout(push_constant) uniform PRECISION restrict Block {
ivec4 sizes;
};

layout(set = 0, binding = 3) uniform PRECISION restrict OriginalSizes {
ivec4 original_sizes;
};

Expand Down
5 changes: 1 addition & 4 deletions backends/vulkan/runtime/graph/ops/glsl/conv2d_prepack_weights.glsl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,11 +26,8 @@ layout(set = 0, binding = 1) buffer PRECISION restrict readonly Buffer {
BUF_T buffer_in[];
};

layout(set = 0, binding = 2) uniform PRECISION restrict Sizes {
layout(push_constant) uniform PRECISION restrict Block {
ivec4 sizes;
};

layout(set = 0, binding = 3) uniform PRECISION restrict OriginalSizes {
ivec4 original_sizes;
};

Expand Down
5 changes: 1 addition & 4 deletions backends/vulkan/runtime/graph/ops/glsl/conv_transpose2d_prepack_weights.glsl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,11 +26,8 @@ layout(set = 0, binding = 1) buffer PRECISION restrict readonly Buffer {
BUF_T buffer_in[];
};

layout(set = 0, binding = 2) uniform PRECISION restrict Sizes {
layout(push_constant) uniform PRECISION restrict Block {
ivec4 sizes;
};

layout(set = 0, binding = 3) uniform PRECISION restrict OriginalSizes {
ivec4 original_sizes;
};

Expand Down
9 changes: 5 additions & 4 deletions backends/vulkan/runtime/graph/ops/impl/Convolution.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -211,18 +211,19 @@ ValueRef prepack_weights(
vkapi::ShaderInfo shader =
get_conv2d_shader(graph, *t, /*prepack_weights = */ true, method, vref);

const auto sizes_pc = utils::make_ivec4(original_sizes, /*reverse = */ true);
graph.prepack_nodes().emplace_back(new PrepackNode(
graph,
shader,
graph.create_global_wg_size(v),
graph.create_local_wg_size(v),
vref,
v,
{t->sizes_ubo(),
graph.create_params_buffer(
utils::make_ivec4(original_sizes, /*reverse = */ true))},
{},
// Specialization constants
{SV(t->packed_dim())}));
{SV(t->packed_dim())},
{graph.sizes_pc_of(v),
PushConstantDataInfo(&sizes_pc, sizeof(sizes_pc))}));

return v;
}
Expand Down
Loading