derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
53 lines
2.4 KiB
MLIR
53 lines
2.4 KiB
MLIR
// RUN: mlir-opt %s --convert-amdgpu-to-rocdl=chipset=gfx1100 | FileCheck %s
|
|
|
|
// CHECK-LABEL: @dot_fdot2_f16_f16
|
|
func.func @dot_fdot2_f16_f16(%a: vector<2xf16>, %b: vector<2xf16>, %c: f16) -> f16 {
|
|
// CHECK: rocdl.fdot2.f16.f16 %{{.+}}, %{{.+}}, %{{.+}} : (vector<2xf16>, vector<2xf16>, f16) -> f16
|
|
%r = amdgpu.dot %a * %b + %c : vector<2xf16>, vector<2xf16>, f16
|
|
func.return %r : f16
|
|
}
|
|
|
|
// CHECK-LABEL: @dot_fdot2_bf16_bf16
|
|
func.func @dot_fdot2_bf16_bf16(%a: vector<2xbf16>, %b: vector<2xbf16>, %c: bf16) -> bf16 {
|
|
// CHECK: rocdl.fdot2.bf16.bf16 %{{.+}}, %{{.+}}, %{{.+}} : (vector<2xbf16>, vector<2xbf16>, bf16) -> bf16
|
|
%r = amdgpu.dot %a * %b + %c : vector<2xbf16>, vector<2xbf16>, bf16
|
|
func.return %r : bf16
|
|
}
|
|
|
|
// CHECK-LABEL: @dot_fdot2_f32_bf16
|
|
func.func @dot_fdot2_f32_bf16(%a: vector<2xbf16>, %b: vector<2xbf16>, %c: f32) -> f32 {
|
|
// CHECK: rocdl.fdot2.f32.bf16 %{{.+}}, %{{.+}}, %{{.+}} : (vector<2xbf16>, vector<2xbf16>, f32) -> f32
|
|
%r = amdgpu.dot %a * %b + %c : vector<2xbf16>, vector<2xbf16>, f32
|
|
func.return %r : f32
|
|
}
|
|
|
|
// Uniform-sign sdot4 still dispatches to the dedicated rocdl.sdot4 (not
|
|
// sudot4) on gfx11+. The backend aliases v_dot4_i32_i8 to v_dot4_i32_iu8.
|
|
// CHECK-LABEL: @dot_sdot4_gfx11_uniform_sign
|
|
func.func @dot_sdot4_gfx11_uniform_sign(%a: vector<4xi8>, %b: vector<4xi8>, %c: i32) -> i32 {
|
|
// CHECK: rocdl.sdot4 %{{.+}}, %{{.+}}, %{{.+}} : (i32, i32, i32) -> i32
|
|
%r = amdgpu.dot %a * %b + %c : vector<4xi8>, vector<4xi8>, i32
|
|
func.return %r : i32
|
|
}
|
|
|
|
// CHECK-LABEL: @dot_sudot4_signA_unsignedB
|
|
func.func @dot_sudot4_signA_unsignedB(%a: vector<4xi8>, %b: vector<4xi8>, %c: i32) -> i32 {
|
|
// CHECK: rocdl.sudot4 %{{.+}}, %{{.+}}, %{{.+}} {signA = true} : (i32, i32, i32) -> i32
|
|
%r = amdgpu.dot %a * %b + %c {unsignedB} : vector<4xi8>, vector<4xi8>, i32
|
|
func.return %r : i32
|
|
}
|
|
|
|
// CHECK-LABEL: @dot_sudot4_unsignedA_signB_clamp
|
|
func.func @dot_sudot4_unsignedA_signB_clamp(%a: vector<4xi8>, %b: vector<4xi8>, %c: i32) -> i32 {
|
|
// CHECK: rocdl.sudot4 %{{.+}}, %{{.+}}, %{{.+}} {clamp = true, signB = true} : (i32, i32, i32) -> i32
|
|
%r = amdgpu.dot %a * %b + %c {unsignedA, clamp} : vector<4xi8>, vector<4xi8>, i32
|
|
func.return %r : i32
|
|
}
|
|
|
|
// CHECK-LABEL: @dot_sudot8
|
|
func.func @dot_sudot8(%a: vector<8xi4>, %b: vector<8xi4>, %c: i32) -> i32 {
|
|
// CHECK: rocdl.sudot8 %{{.+}}, %{{.+}}, %{{.+}} {signA = true} : (i32, i32, i32) -> i32
|
|
%r = amdgpu.dot %a * %b + %c {unsignedB} : vector<8xi4>, vector<8xi4>, i32
|
|
func.return %r : i32
|
|
}
|