mirror of
https://github.com/RPCSX/llvm.git
synced 2026-01-31 01:05:23 +01:00
aef8c41869
Since we have no call support and late linking we can produce code only for used symbols. This saves compilation time, size of the final executable, and size of any intermediate dumps. Run Internalize pass early in the opt pipeline followed by global DCE pass. To enable it RT can pass -amdgpu-internalize-symbols option. Differential Revision: https://reviews.llvm.org/D29214 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@293549 91177308-0d34-0410-b5e6-96231b3b80d8
36 lines
953 B
LLVM
36 lines
953 B
LLVM
; RUN: opt -O1 -S -mtriple=amdgcn-unknown-amdhsa -amdgpu-internalize-symbols < %s | FileCheck %s
|
|
; CHECK-NOT: unused
|
|
; CHECK-NOT: foo_used
|
|
; CHECK: gvar_used
|
|
; CHECK: main_kernel
|
|
|
|
@gvar_unused = addrspace(1) global i32 undef, align 4
|
|
@gvar_used = addrspace(1) global i32 undef, align 4
|
|
|
|
; Function Attrs: alwaysinline nounwind
|
|
define void @foo_unused(i32 addrspace(1)* %out) local_unnamed_addr #1 {
|
|
entry:
|
|
store i32 1, i32 addrspace(1)* %out
|
|
ret void
|
|
}
|
|
|
|
; Function Attrs: alwaysinline nounwind
|
|
define void @foo_used(i32 addrspace(1)* %out, i32 %tid) local_unnamed_addr #1 {
|
|
entry:
|
|
store i32 %tid, i32 addrspace(1)* %out
|
|
ret void
|
|
}
|
|
|
|
define amdgpu_kernel void @main_kernel() {
|
|
entry:
|
|
%tid = call i32 @llvm.amdgcn.workitem.id.x()
|
|
tail call void @foo_used(i32 addrspace(1)* @gvar_used, i32 %tid) nounwind
|
|
ret void
|
|
}
|
|
|
|
declare i32 @llvm.amdgcn.workitem.id.x() #0
|
|
|
|
attributes #0 = { nounwind readnone }
|
|
|
|
attributes #1 = { alwaysinline nounwind }
|