055bfc0271
GEPs are often in the form `gep [N x %T], ptr %p, i64 0, i64 %idx`. Canonicalize these to `gep %T, ptr %p, i64 %idx`. This enables transforms that only support one GEP index to work and improves CSE. Various transforms were recently hardened to make sure they still work without the leading index.
27 lines
1.2 KiB
LLVM
27 lines
1.2 KiB
LLVM
; RUN: opt -S -mtriple=amdgcn-- -data-layout=A5 -passes='amdgpu-promote-alloca,sroa,instcombine' < %s | FileCheck -check-prefix=OPT %s
|
|
|
|
; Should give up promoting alloca to vector with an addrspacecast.
|
|
|
|
; OPT-LABEL: @vector_addrspacecast(
|
|
; OPT: alloca [3 x i32]
|
|
; OPT: store i32 0, ptr addrspace(5) %alloca, align 4
|
|
; OPT: store i32 1, ptr addrspace(5) %a1, align 4
|
|
; OPT: store i32 2, ptr addrspace(5) %a2, align 4
|
|
; OPT: %tmp = getelementptr i32, ptr addrspace(5) %alloca, i64 %index
|
|
; OPT: %ac = addrspacecast ptr addrspace(5) %tmp to ptr
|
|
; OPT: %data = load i32, ptr %ac, align 4
|
|
define amdgpu_kernel void @vector_addrspacecast(ptr addrspace(1) %out, i64 %index) {
|
|
entry:
|
|
%alloca = alloca [3 x i32], addrspace(5)
|
|
%a1 = getelementptr [3 x i32], ptr addrspace(5) %alloca, i32 0, i32 1
|
|
%a2 = getelementptr [3 x i32], ptr addrspace(5) %alloca, i32 0, i32 2
|
|
store i32 0, ptr addrspace(5) %alloca
|
|
store i32 1, ptr addrspace(5) %a1
|
|
store i32 2, ptr addrspace(5) %a2
|
|
%tmp = getelementptr [3 x i32], ptr addrspace(5) %alloca, i64 0, i64 %index
|
|
%ac = addrspacecast ptr addrspace(5) %tmp to ptr
|
|
%data = load i32, ptr %ac
|
|
store i32 %data, ptr addrspace(1) %out
|
|
ret void
|
|
}
|