Gitiles
Code Review Sign In
gerrit-public.fairphone.software / toolchain / llvm-project / 91be65be656072a68b51a8c4e7bb751ea475d896 / . / llvm / test / CodeGen / AMDGPU / promote-alloca-globals.ll
blob: 4403d1f23066b4b531e010aefa3e7d101a2be10e [file] [log] [blame]
Yaxun Liu2a22c5d2018-02-02 16:07:16 +0000[diff] [blame]1; RUN: opt -data-layout=A5 -S -mtriple=amdgcn-unknown-unknown -amdgpu-promote-alloca < %s | FileCheck -check-prefix=IR %s
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]2; RUN: llc -march=amdgcn -mcpu=tonga < %s | FileCheck -check-prefix=ASM %s
3
4
Matt Arsenault8a028bf2016-05-16 21:19:59 +0000[diff] [blame]5@global_array0 = internal unnamed_addr addrspace(3) global [750 x [10 x i32]] undef, align 4
6@global_array1 = internal unnamed_addr addrspace(3) global [750 x [10 x i32]] undef, align 4
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]7
Matt Arsenault3dbeefa2017-03-21 21:39:51 +0000[diff] [blame]8; IR-LABEL: define amdgpu_kernel void @promote_alloca_size_256(i32 addrspace(1)* nocapture %out, i32 addrspace(1)* nocapture %in) {
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]9; IR: alloca [10 x i32]
10; ASM-LABEL: {{^}}promote_alloca_size_256:
11; ASM: ; LDSByteSize: 60000 bytes/workgroup (compile time only)
12
Matt Arsenault3dbeefa2017-03-21 21:39:51 +0000[diff] [blame]13define amdgpu_kernel void @promote_alloca_size_256(i32 addrspace(1)* nocapture %out, i32 addrspace(1)* nocapture %in) {
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]14entry:
Yaxun Liu2a22c5d2018-02-02 16:07:16 +0000[diff] [blame]15 %stack = alloca [10 x i32], align 4, addrspace(5)
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]16 %tmp = load i32, i32 addrspace(1)* %in, align 4
Yaxun Liu2a22c5d2018-02-02 16:07:16 +0000[diff] [blame]17 %arrayidx1 = getelementptr inbounds [10 x i32], [10 x i32] addrspace(5)* %stack, i32 0, i32 %tmp
18 store i32 4, i32 addrspace(5)* %arrayidx1, align 4
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]19 %arrayidx2 = getelementptr inbounds i32, i32 addrspace(1)* %in, i32 1
20 %tmp1 = load i32, i32 addrspace(1)* %arrayidx2, align 4
Yaxun Liu2a22c5d2018-02-02 16:07:16 +0000[diff] [blame]21 %arrayidx3 = getelementptr inbounds [10 x i32], [10 x i32] addrspace(5)* %stack, i32 0, i32 %tmp1
22 store i32 5, i32 addrspace(5)* %arrayidx3, align 4
23 %arrayidx10 = getelementptr inbounds [10 x i32], [10 x i32] addrspace(5)* %stack, i32 0, i32 0
24 %tmp2 = load i32, i32 addrspace(5)* %arrayidx10, align 4
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]25 store i32 %tmp2, i32 addrspace(1)* %out, align 4
Yaxun Liu2a22c5d2018-02-02 16:07:16 +0000[diff] [blame]26 %arrayidx12 = getelementptr inbounds [10 x i32], [10 x i32] addrspace(5)* %stack, i32 0, i32 1
27 %tmp3 = load i32, i32 addrspace(5)* %arrayidx12
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]28 %arrayidx13 = getelementptr inbounds i32, i32 addrspace(1)* %out, i32 1
29 store i32 %tmp3, i32 addrspace(1)* %arrayidx13
Matt Arsenault8a028bf2016-05-16 21:19:59 +0000[diff] [blame]30 %v0 = getelementptr inbounds [750 x [10 x i32]], [750 x [10 x i32]] addrspace(3)* @global_array0, i32 0, i32 0, i32 0
31 store i32 %tmp3, i32 addrspace(3)* %v0
32 %v1 = getelementptr inbounds [750 x [10 x i32]], [750 x [10 x i32]] addrspace(3)* @global_array1, i32 0, i32 0, i32 0
33 store i32 %tmp3, i32 addrspace(3)* %v1
Matt Arsenault0547b012016-04-27 21:05:08 +0000[diff] [blame]34 ret void
35}