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
94 lines
3.9 KiB
LLVM
94 lines
3.9 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --filter-out "s_setreg_imm32_b32" --version 6
|
|
; RUN: llc < %s -global-isel=1 -new-reg-bank-select -mtriple=amdgcn -mcpu=gfx1200 | FileCheck %s -check-prefixes=GCN
|
|
|
|
declare float @llvm.amdgcn.global.atomic.fmin.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
declare float @llvm.amdgcn.global.atomic.fmax.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
|
|
define amdgpu_cs void @global_atomic_fmin_num_f32_noret(ptr addrspace(1) %ptr, float %data) {
|
|
; GCN-LABEL: global_atomic_fmin_num_f32_noret:
|
|
; GCN: ; %bb.0:
|
|
; GCN: global_atomic_min_num_f32 v[0:1], v2, off
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmin.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_cs void @global_atomic_fmax_num_f32_noret(ptr addrspace(1) %ptr, float %data) {
|
|
; GCN-LABEL: global_atomic_fmax_num_f32_noret:
|
|
; GCN: ; %bb.0:
|
|
; GCN: global_atomic_max_num_f32 v[0:1], v2, off
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmax.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_cs void @global_atomic_fmax_num_f32_rtn(ptr addrspace(1) %ptr, float %data, ptr addrspace(1) %out) {
|
|
; GCN-LABEL: global_atomic_fmax_num_f32_rtn:
|
|
; GCN: ; %bb.0:
|
|
; GCN: global_atomic_max_num_f32 v0, v[0:1], v2, off th:TH_ATOMIC_RETURN
|
|
; GCN: s_wait_loadcnt 0x0
|
|
; GCN: global_store_b32 v[3:4], v0, off
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmax.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
store float %ret, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_cs void @global_atomic_fmin_num_f32_rtn(ptr addrspace(1) %ptr, float %data, ptr addrspace(1) %out) {
|
|
; GCN-LABEL: global_atomic_fmin_num_f32_rtn:
|
|
; GCN: ; %bb.0:
|
|
; GCN: global_atomic_min_num_f32 v0, v[0:1], v2, off th:TH_ATOMIC_RETURN
|
|
; GCN: s_wait_loadcnt 0x0
|
|
; GCN: global_store_b32 v[3:4], v0, off
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmin.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
store float %ret, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_ps void @global_atomic_fmin_num_f32_noret_saddr(ptr addrspace(1) inreg %ptr, float %data) {
|
|
; GCN-LABEL: global_atomic_fmin_num_f32_noret_saddr:
|
|
; GCN: ; %bb.0:
|
|
; GCN: v_mov_b32_e32 v1, 0
|
|
; GCN: global_atomic_min_num_f32 v1, v0, s[0:1]
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmin.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_ps void @global_atomic_fmax_num_f32_noret_saddr(ptr addrspace(1) inreg %ptr, float %data) {
|
|
; GCN-LABEL: global_atomic_fmax_num_f32_noret_saddr:
|
|
; GCN: ; %bb.0:
|
|
; GCN: v_mov_b32_e32 v1, 0
|
|
; GCN: global_atomic_max_num_f32 v1, v0, s[0:1]
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmax.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_ps void @global_atomic_fmin_num_f32_rtn_saddr(ptr addrspace(1) inreg %ptr, float %data, ptr addrspace(1) %out) {
|
|
; GCN-LABEL: global_atomic_fmin_num_f32_rtn_saddr:
|
|
; GCN: ; %bb.0:
|
|
; GCN: v_mov_b32_e32 v3, 0
|
|
; GCN: global_atomic_min_num_f32 v0, v3, v0, s[0:1] th:TH_ATOMIC_RETURN
|
|
; GCN: s_wait_loadcnt 0x0
|
|
; GCN: global_store_b32 v[1:2], v0, off
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmin.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
store float %ret, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_ps void @global_atomic_fmax_num_f32_rtn_saddr(ptr addrspace(1) inreg %ptr, float %data, ptr addrspace(1) %out) {
|
|
; GCN-LABEL: global_atomic_fmax_num_f32_rtn_saddr:
|
|
; GCN: ; %bb.0:
|
|
; GCN: v_mov_b32_e32 v3, 0
|
|
; GCN: global_atomic_max_num_f32 v0, v3, v0, s[0:1] th:TH_ATOMIC_RETURN
|
|
; GCN: s_wait_loadcnt 0x0
|
|
; GCN: global_store_b32 v[1:2], v0, off
|
|
; GCN: s_endpgm
|
|
%ret = call float @llvm.amdgcn.global.atomic.fmax.num.f32.p1.f32(ptr addrspace(1) %ptr, float %data)
|
|
store float %ret, ptr addrspace(1) %out
|
|
ret void
|
|
}
|