brintos

brintos / llvm-project-archived public Read only

0
0
Text · 858 B · 72e5d0a Raw
22 lines · plain
1; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=fiji -mattr=-flat-for-global -relocation-model=static < %s | FileCheck %s2 3@lds = external addrspace(3) global [4 x i32]4 5; Function Attrs: nounwind6 7; Offset folding is an optimization done for global variables with relocations,8; which allows you to store the offset in the r_addend of the relocation entry.9; The offset is apllied to the variables address at link time, which eliminates10; the need to emit shader instructions to do this calculation.11; We don't use relocations for local memory, so we should never fold offsets12; for local memory globals.13 14; CHECK-LABEL: lds_no_offset:15; CHECK: ds_write_b32 v{{[0-9]+}}, v{{[0-9]+}} offset:416define amdgpu_kernel void @lds_no_offset() {17entry:18  %ptr = getelementptr [4 x i32], ptr addrspace(3) @lds, i32 0, i32 119  store i32 0, ptr addrspace(3) %ptr20  ret void21}22