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
87 lines
2.7 KiB
LLVM
87 lines
2.7 KiB
LLVM
; RUN: llc < %s -mtriple=nvptx64 -mcpu=sm_89 -mattr=+ptx81 | FileCheck %s
|
|
; RUN: %if ptxas-sm_89 && ptxas-isa-8.1 %{ llc < %s -mtriple=nvptx64 -mcpu=sm_89 -mattr=+ptx81 | %ptxas-verify -arch=sm_89 %}
|
|
|
|
; CHECK-LABEL: cvt_rn_e4m3x2_f32
|
|
define i16 @cvt_rn_e4m3x2_f32(float %f1, float %f2) {
|
|
; CHECK: cvt.rn.satfinite.e4m3x2.f32
|
|
%val = call i16 @llvm.nvvm.ff.to.e4m3x2.rn(float %f1, float %f2);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_relu_e4m3x2_f32
|
|
define i16 @cvt_rn_relu_e4m3x2_f32(float %f1, float %f2) {
|
|
; CHECK: cvt.rn.satfinite.relu.e4m3x2.f32
|
|
%val = call i16 @llvm.nvvm.ff.to.e4m3x2.rn.relu(float %f1, float %f2);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_e5m2x2_f32
|
|
define i16 @cvt_rn_e5m2x2_f32(float %f1, float %f2) {
|
|
; CHECK: cvt.rn.satfinite.e5m2x2.f32
|
|
%val = call i16 @llvm.nvvm.ff.to.e5m2x2.rn(float %f1, float %f2);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_relu_e5m2x2_f32
|
|
define i16 @cvt_rn_relu_e5m2x2_f32(float %f1, float %f2) {
|
|
; CHECK: cvt.rn.satfinite.relu.e5m2x2.f32
|
|
%val = call i16 @llvm.nvvm.ff.to.e5m2x2.rn.relu(float %f1, float %f2);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_e4m3x2_f16x2
|
|
define i16 @cvt_rn_e4m3x2_f16x2(<2 x half> %in) {
|
|
; CHECK: cvt.rn.satfinite.e4m3x2.f16x2
|
|
%val = call i16 @llvm.nvvm.f16x2.to.e4m3x2.rn(<2 x half> %in);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_relu_e4m3x2_f16x2
|
|
define i16 @cvt_rn_relu_e4m3x2_f16x2(<2 x half> %in) {
|
|
; CHECK: cvt.rn.satfinite.relu.e4m3x2.f16x2
|
|
%val = call i16 @llvm.nvvm.f16x2.to.e4m3x2.rn.relu(<2 x half> %in);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_e5m2x2_f16x2
|
|
define i16 @cvt_rn_e5m2x2_f16x2(<2 x half> %in) {
|
|
; CHECK: cvt.rn.satfinite.e5m2x2.f16x2
|
|
%val = call i16 @llvm.nvvm.f16x2.to.e5m2x2.rn(<2 x half> %in);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_relu_e5m2x2_f16x2
|
|
define i16 @cvt_rn_relu_e5m2x2_f16x2(<2 x half> %in) {
|
|
; CHECK: cvt.rn.satfinite.relu.e5m2x2.f16x2
|
|
%val = call i16 @llvm.nvvm.f16x2.to.e5m2x2.rn.relu(<2 x half> %in);
|
|
ret i16 %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_f16x2_e4m3x2
|
|
define <2 x half> @cvt_rn_f16x2_e4m3x2(i16 %in) {
|
|
; CHECK: cvt.rn.f16x2.e4m3x2
|
|
%val = call <2 x half> @llvm.nvvm.e4m3x2.to.f16x2.rn(i16 %in);
|
|
ret <2 x half> %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_relu_f16x2_e4m3x2
|
|
define <2 x half> @cvt_rn_relu_f16x2_e4m3x2(i16 %in) {
|
|
; CHECK: cvt.rn.relu.f16x2.e4m3x2
|
|
%val = call <2 x half> @llvm.nvvm.e4m3x2.to.f16x2.rn.relu(i16 %in);
|
|
ret <2 x half> %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_f16x2_e5m2x2
|
|
define <2 x half> @cvt_rn_f16x2_e5m2x2(i16 %in) {
|
|
; CHECK: cvt.rn.f16x2.e5m2x2
|
|
%val = call <2 x half> @llvm.nvvm.e5m2x2.to.f16x2.rn(i16 %in);
|
|
ret <2 x half> %val
|
|
}
|
|
|
|
; CHECK-LABEL: cvt_rn_relu_f16x2_e5m2x2
|
|
define <2 x half> @cvt_rn_relu_f16x2_e5m2x2(i16 %in) {
|
|
; CHECK: cvt.rn.relu.f16x2.e5m2x2
|
|
%val = call <2 x half> @llvm.nvvm.e5m2x2.to.f16x2.rn.relu(i16 %in);
|
|
ret <2 x half> %val
|
|
}
|