Skip to content

Pingpong minor opts #22408

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

Draft
wants to merge 7 commits into
base: main
Choose a base branch
from
Draft

Pingpong minor opts #22408

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
361e37e
Apply pingpong to epilogue f16
sebvince Oct 24, 2025
9358c0e
Same opt on bf16
sebvince Oct 24, 2025
b77a0da
Handle large expanded variants
sebvince Oct 27, 2025
f5d0bab
Handle medium pingpong ukernels
sebvince Oct 27, 2025
edef968
Add medium ukernel e2e tests
sebvince Oct 28, 2025
1f7972b
Use consistent s_setprio ordering on barriers
sebvince Oct 28, 2025
e662d51
Add missing for f8 gfx950
sebvince Oct 28, 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
73 changes: 53 additions & 20 deletions compiler/plugins/target/ROCM/builtins/mlir_ukernel/iree_uk_amdgpu_matmul_bf16.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -211,27 +211,44 @@ util.func private @pingpong_large_bf16(%lhs_base: !bf16_in_ty, %rhs_base: !bf16_
// Epilogue
%lhs_vec_0 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c0, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_0 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c0, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%lhs_vec_1 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_1 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>

gpu.barrier
rocdl.s.setprio 1 { iree_gpu.swap_mfma = 1 }
rocdl.sched.barrier 0

%dot0 = iree_codegen.inner_tiled ins(%lhs_vec_0, %rhs_vec_0) outs(%3) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
kind = #iree_gpu.mma_layout<MFMA_F32_16x16x16_BF16, col_major = true>
} : vector<8x1x1x4xbf16>, vector<4x1x1x4xbf16> into vector<8x4x1x4xf32>
%lhs_vec_1 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_1 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%dot1 = iree_codegen.inner_tiled ins(%lhs_vec_1, %rhs_vec_1) outs(%dot0) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
kind = #iree_gpu.mma_layout<MFMA_F32_16x16x16_BF16, col_major = true>
} : vector<8x1x1x4xbf16>, vector<4x1x1x4xbf16> into vector<8x4x1x4xf32>

rocdl.s.setprio 0
gpu.barrier
rocdl.sched.barrier 0

%lhs_vec_2 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_2 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%lhs_vec_3 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_3 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>

scf.if %cmp1 {
rocdl.s.setprio 1 { iree_gpu.swap_mfma = 1 }
gpu.barrier
rocdl.sched.barrier 0
}

%dot2 = iree_codegen.inner_tiled ins(%lhs_vec_2, %rhs_vec_2) outs(%dot1) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
kind = #iree_gpu.mma_layout<MFMA_F32_16x16x16_BF16, col_major = true>
} : vector<8x1x1x4xbf16>, vector<4x1x1x4xbf16> into vector<8x4x1x4xf32>
%lhs_vec_3 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_3 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%dot3 = iree_codegen.inner_tiled ins(%lhs_vec_3, %rhs_vec_3) outs(%dot2) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
Expand Down Expand Up @@ -402,27 +419,30 @@ util.func private @pingpong_medium_bf16_expanded(%lhs_base: !mexp_in_ty_bf16, %r

scf.yield %dot2 : vector<4x4x1x4xf32>
}
scf.if %cmp1 {
rocdl.s.barrier
}

// Epilogue
%lhs_vec_0 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c0, %inner_id], %cst {in_bounds = [true, true, true, true]} : !mshared_exp_bf16, vector<4x1x2x4xbf16>
%rhs_vec_0 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c0, %inner_id], %cst {in_bounds = [true, true, true, true]} : !shared_exp_bf16, vector<4x1x2x4xbf16>
%lhs_vec_0_t = vector.transpose %lhs_vec_0, [0, 2, 1, 3] : vector<4x1x2x4xbf16> to vector<4x2x1x4xbf16>
%rhs_vec_0_t = vector.transpose %rhs_vec_0, [0, 2, 1, 3] : vector<4x1x2x4xbf16> to vector<4x2x1x4xbf16>

%lhs_vec_2 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !mshared_exp_bf16, vector<4x1x2x4xbf16>
%rhs_vec_2 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !shared_exp_bf16, vector<4x1x2x4xbf16>
%lhs_vec_2_t = vector.transpose %lhs_vec_2, [0, 2, 1, 3] : vector<4x1x2x4xbf16> to vector<4x2x1x4xbf16>
%rhs_vec_2_t = vector.transpose %rhs_vec_2, [0, 2, 1, 3] : vector<4x1x2x4xbf16> to vector<4x2x1x4xbf16>

rocdl.s.setprio 1 { iree_gpu.swap_mfma = 1 }
scf.if %cmp1 {
gpu.barrier
}
rocdl.sched.barrier 0

%dot0 = iree_codegen.inner_tiled ins(%lhs_vec_0_t, %rhs_vec_0_t) outs(%3) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
kind = #iree_gpu.mma_layout<MFMA_F32_16x16x16_BF16, col_major = true>
} : vector<4x2x1x4xbf16>, vector<4x2x1x4xbf16> into vector<4x4x1x4xf32>

%lhs_vec_2 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !mshared_exp_bf16, vector<4x1x2x4xbf16>
%rhs_vec_2 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !shared_exp_bf16, vector<4x1x2x4xbf16>
%lhs_vec_2_t = vector.transpose %lhs_vec_2, [0, 2, 1, 3] : vector<4x1x2x4xbf16> to vector<4x2x1x4xbf16>
%rhs_vec_2_t = vector.transpose %rhs_vec_2, [0, 2, 1, 3] : vector<4x1x2x4xbf16> to vector<4x2x1x4xbf16>

%dot2 = iree_codegen.inner_tiled ins(%lhs_vec_2_t, %rhs_vec_2_t) outs(%dot0) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
Expand Down Expand Up @@ -616,34 +636,47 @@ util.func private @pingpong_large_bf16_expanded(%lhs_base: !bf16_exp_in_ty, %rhs

scf.yield %dot3 : vector<8x4x1x4xf32>
}
scf.if %cmp1 {
rocdl.s.barrier
}

// Epilogue
%lhs_vec_0 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c0, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_0 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c0, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%lhs_vec_1 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_1 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>

rocdl.s.setprio 1 { iree_gpu.swap_mfma = 1 }
gpu.barrier
rocdl.sched.barrier 0

%dot0 = iree_codegen.inner_tiled ins(%lhs_vec_0, %rhs_vec_0) outs(%3) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
kind = #iree_gpu.mma_layout<MFMA_F32_16x16x16_BF16, col_major = true>
} : vector<8x1x1x4xbf16>, vector<4x1x1x4xbf16> into vector<8x4x1x4xf32>
%lhs_vec_1 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_1 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c1, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%dot1 = iree_codegen.inner_tiled ins(%lhs_vec_1, %rhs_vec_1) outs(%dot0) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
kind = #iree_gpu.mma_layout<MFMA_F32_16x16x16_BF16, col_major = true>
} : vector<8x1x1x4xbf16>, vector<4x1x1x4xbf16> into vector<8x4x1x4xf32>

rocdl.s.setprio 0
gpu.barrier
rocdl.sched.barrier 0

%lhs_vec_2 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_2 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c2, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%lhs_vec_3 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_3 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>

scf.if %cmp1 {
rocdl.s.setprio 1 { iree_gpu.swap_mfma = 1 }
gpu.barrier
rocdl.sched.barrier 0
}

%dot2 = iree_codegen.inner_tiled ins(%lhs_vec_2, %rhs_vec_2) outs(%dot1) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
kind = #iree_gpu.mma_layout<MFMA_F32_16x16x16_BF16, col_major = true>
} : vector<8x1x1x4xbf16>, vector<4x1x1x4xbf16> into vector<8x4x1x4xf32>
%lhs_vec_3 = vector.transfer_read %lhs_shared_expand[%m_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<8x1x1x4xbf16>
%rhs_vec_3 = vector.transfer_read %rhs_shared_expand[%n_outer_id, %ids#3, %c3, %inner_id], %cst {in_bounds = [true, true, true, true]} : !bf16_shared_exp, vector<4x1x1x4xbf16>
%dot3 = iree_codegen.inner_tiled ins(%lhs_vec_3, %rhs_vec_3) outs(%dot2) {
indexing_maps = #contraction_accesses,
iterator_types = [#linalg.iterator_type<parallel>, #linalg.iterator_type<parallel>, #linalg.iterator_type<reduction>],
Expand Down
Loading
Loading