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
74 lines
2.9 KiB
LLVM
74 lines
2.9 KiB
LLVM
; RUN: llc -mtriple=amdgcn -mattr=+max-private-element-size-4 < %s | FileCheck -check-prefix=GCN %s
|
|
; RUN: llc -mtriple=amdgcn -disable-promote-alloca-to-vector -disable-promote-alloca-to-lds -mattr=+max-private-element-size-4 < %s | FileCheck -check-prefix=GCN %s
|
|
|
|
; Pointer value is stored in a candidate for LDS usage.
|
|
|
|
; GCN-LABEL: {{^}}stored_lds_pointer_value:
|
|
; GCN: buffer_store_dword v
|
|
define amdgpu_kernel void @stored_lds_pointer_value(ptr addrspace(1) %ptr) #0 {
|
|
%tmp = alloca float, addrspace(5)
|
|
store float 0.0, ptr addrspace(5) %tmp
|
|
store ptr addrspace(5) %tmp, ptr addrspace(1) %ptr
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}stored_lds_pointer_value_offset:
|
|
; GCN: buffer_store_dword v
|
|
define amdgpu_kernel void @stored_lds_pointer_value_offset(ptr addrspace(1) %ptr) #0 {
|
|
%tmp0 = alloca float, addrspace(5)
|
|
%tmp1 = alloca float, addrspace(5)
|
|
store float 0.0, ptr addrspace(5) %tmp0
|
|
store float 0.0, ptr addrspace(5) %tmp1
|
|
store volatile ptr addrspace(5) %tmp0, ptr addrspace(1) %ptr
|
|
store volatile ptr addrspace(5) %tmp1, ptr addrspace(1) %ptr
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}stored_lds_pointer_value_gep:
|
|
; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD0
|
|
; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD1
|
|
; GCN: buffer_store_dword v
|
|
; GCN: buffer_store_dword v
|
|
define amdgpu_kernel void @stored_lds_pointer_value_gep(ptr addrspace(1) %ptr, i32 %idx) #0 {
|
|
bb:
|
|
%tmp = alloca float, i32 16, addrspace(5)
|
|
store float 0.0, ptr addrspace(5) %tmp
|
|
%tmp2 = getelementptr inbounds float, ptr addrspace(5) %tmp, i32 %idx
|
|
store ptr addrspace(5) %tmp2, ptr addrspace(1) %ptr
|
|
ret void
|
|
}
|
|
|
|
; Pointer value is stored in a candidate for vector usage
|
|
; GCN-LABEL: {{^}}stored_vector_pointer_value:
|
|
; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD0
|
|
; GCN-DAG: s_mov_b32 s{{[0-9]+}}, SCRATCH_RSRC_DWORD1
|
|
; GCN: buffer_store_dword
|
|
; GCN: buffer_store_dword
|
|
; GCN: buffer_store_dword
|
|
; GCN: buffer_store_dword
|
|
define amdgpu_kernel void @stored_vector_pointer_value(ptr addrspace(1) %out, i32 %index) {
|
|
entry:
|
|
%tmp0 = alloca [4 x i32], addrspace(5)
|
|
%y = getelementptr inbounds [4 x i32], ptr addrspace(5) %tmp0, i32 0, i32 1
|
|
%z = getelementptr inbounds [4 x i32], ptr addrspace(5) %tmp0, i32 0, i32 2
|
|
%w = getelementptr inbounds [4 x i32], ptr addrspace(5) %tmp0, i32 0, i32 3
|
|
store i32 0, ptr addrspace(5) %tmp0
|
|
store i32 1, ptr addrspace(5) %y
|
|
store i32 2, ptr addrspace(5) %z
|
|
store i32 3, ptr addrspace(5) %w
|
|
%tmp1 = getelementptr inbounds [4 x i32], ptr addrspace(5) %tmp0, i32 0, i32 %index
|
|
store ptr addrspace(5) %tmp1, ptr addrspace(1) %out
|
|
ret void
|
|
}
|
|
|
|
; GCN-LABEL: {{^}}stored_fi_to_self:
|
|
; GCN-NOT: ds_
|
|
define amdgpu_kernel void @stored_fi_to_self() #0 {
|
|
%tmp = alloca ptr addrspace(5), addrspace(5)
|
|
store volatile ptr addrspace(5) inttoptr (i32 1234 to ptr addrspace(5)), ptr addrspace(5) %tmp
|
|
store volatile ptr addrspace(5) %tmp, ptr addrspace(5) %tmp
|
|
ret void
|
|
}
|
|
|
|
attributes #0 = { nounwind }
|