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
70 lines
2.5 KiB
LLVM
70 lines
2.5 KiB
LLVM
; RUN: llc -mtriple=amdgcn < %s | FileCheck -check-prefix=GCN %s
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_f64_to_f16:
|
|
; GCN: s_endpgm
|
|
define amdgpu_kernel void @global_truncstore_f64_to_f16(ptr addrspace(1) %out, ptr addrspace(1) %in) #0 {
|
|
%val = load double, ptr addrspace(1) %in
|
|
%cvt = fptrunc double %val to half
|
|
store half %cvt, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_v2f64_to_v2f16:
|
|
; GCN: s_endpgm
|
|
define amdgpu_kernel void @global_truncstore_v2f64_to_v2f16(ptr addrspace(1) %out, ptr addrspace(1) %in) #0 {
|
|
%val = load <2 x double>, ptr addrspace(1) %in
|
|
%cvt = fptrunc <2 x double> %val to <2 x half>
|
|
store <2 x half> %cvt, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_v3f64_to_v3f16:
|
|
; GCN: s_endpgm
|
|
define amdgpu_kernel void @global_truncstore_v3f64_to_v3f16(ptr addrspace(1) %out, ptr addrspace(1) %in) #0 {
|
|
%val = load <3 x double>, ptr addrspace(1) %in
|
|
%cvt = fptrunc <3 x double> %val to <3 x half>
|
|
store <3 x half> %cvt, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_v4f64_to_v4f16:
|
|
; GCN: s_endpgm
|
|
define amdgpu_kernel void @global_truncstore_v4f64_to_v4f16(ptr addrspace(1) %out, ptr addrspace(1) %in) #0 {
|
|
%val = load <4 x double>, ptr addrspace(1) %in
|
|
%cvt = fptrunc <4 x double> %val to <4 x half>
|
|
store <4 x half> %cvt, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_v8f64_to_v8f16:
|
|
; GCN: s_endpgm
|
|
define amdgpu_kernel void @global_truncstore_v8f64_to_v8f16(ptr addrspace(1) %out, ptr addrspace(1) %in) #0 {
|
|
%val = load <8 x double>, ptr addrspace(1) %in
|
|
%cvt = fptrunc <8 x double> %val to <8 x half>
|
|
store <8 x half> %cvt, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_v16f64_to_v16f16:
|
|
; GCN: s_endpgm
|
|
define amdgpu_kernel void @global_truncstore_v16f64_to_v16f16(ptr addrspace(1) %out, ptr addrspace(1) %in) #0 {
|
|
%val = load <16 x double>, ptr addrspace(1) %in
|
|
%cvt = fptrunc <16 x double> %val to <16 x half>
|
|
store <16 x half> %cvt, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_v6f64_to_v6f16:
|
|
define void @global_truncstore_v6f64_to_v6f16(ptr addrspace(1) %ptr, <6 x double> %src) {
|
|
%trunc = fptrunc <6 x double> %src to <6 x half>
|
|
store <6 x half> %trunc, ptr addrspace(1) %ptr
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}global_truncstore_v6f64_to_v6bf16:
|
|
define void @global_truncstore_v6f64_to_v6bf16(ptr addrspace(1) %ptr, <6 x double> %src) {
|
|
%trunc = fptrunc <6 x double> %src to <6 x bfloat>
|
|
store <6 x bfloat> %trunc, ptr addrspace(1) %ptr
|
|
ret void
|
|
}
|