replica
/
hanchenye-llvm-project
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[SROA] Fix APInt size when alloca address space is not 0 

SROA assumes alloca address space is 0, which causes assertion. This patch fixes that.

Differential Revision: https://reviews.llvm.org/D34104

llvm-svn: 306440
This commit is contained in:
Yaxun Liu 2017-06-27 18:26:06 +00:00
parent 6851ddf942
commit 7c44f340de
2 changed files with 33 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
llvm/lib/Transforms/Scalar/SROA.cpp
View File

@ -3591,10 +3591,11 @@ bool SROA::presplitLoadsAndStores(AllocaInst &AI, AllocaSlices &AS) {
int Idx = 0, Size = Offsets.Splits.size();
for (;;) {
auto *PartTy = Type::getIntNTy(Ty->getContext(), PartSize * 8);
auto *PartPtrTy = PartTy->getPointerTo(LI->getPointerAddressSpace());
auto AS = LI->getPointerAddressSpace();
auto *PartPtrTy = PartTy->getPointerTo(AS);
LoadInst *PLoad = IRB.CreateAlignedLoad(
getAdjustedPtr(IRB, DL, BasePtr,
APInt(DL.getPointerSizeInBits(), PartOffset),
APInt(DL.getPointerSizeInBits(AS), PartOffset),
PartPtrTy, BasePtr->getName() + "."),
getAdjustedAlignment(LI, PartOffset, DL), /*IsVolatile*/ false,
LI->getName());

31
llvm/test/Transforms/SROA/alloca-address-space.ll
View File

@ -1,5 +1,5 @@
; RUN: opt < %s -sroa -S | FileCheck %s
target datalayout = "e-p:64:64:64-p1:16:16:16-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-n8:16:32:64-A2"
target datalayout = "e-p:64:64:64-p1:16:16:16-p2:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-n8:16:32:64-A2"
declare void @llvm.memcpy.p2i8.p2i8.i32(i8 addrspace(2)* nocapture, i8 addrspace(2)* nocapture readonly, i32, i32, i1)
declare void @llvm.memcpy.p1i8.p2i8.i32(i8 addrspace(1)* nocapture, i8 addrspace(2)* nocapture readonly, i32, i32, i1)
@ -82,3 +82,32 @@ define void @pr27557() {
store i32 addrspace(3)* @l, i32 addrspace(3)* addrspace(2)* %3, align 8
ret void
}
; Test load from and store to non-zero address space.
define void @test_load_store_diff_addr_space([2 x float] addrspace(1)* %complex1, [2 x float] addrspace(1)* %complex2) {
; CHECK-LABEL: @test_load_store_diff_addr_space
; CHECK-NOT: alloca
; CHECK: load i32, i32 addrspace(1)*
; CHECK: load i32, i32 addrspace(1)*
; CHECK: store i32 %{{.*}}, i32 addrspace(1)*
; CHECK: store i32 %{{.*}}, i32 addrspace(1)*
%a0 = alloca [2 x i64], align 8, addrspace(2)
%a = getelementptr [2 x i64], [2 x i64] addrspace(2)* %a0, i32 0, i32 0
%a.cast = bitcast i64 addrspace(2)* %a to [2 x float] addrspace(2)*
%a.gep1 = getelementptr [2 x float], [2 x float] addrspace(2)* %a.cast, i32 0, i32 0
%a.gep2 = getelementptr [2 x float], [2 x float] addrspace(2)* %a.cast, i32 0, i32 1
%complex1.gep = getelementptr [2 x float], [2 x float] addrspace(1)* %complex1, i32 0, i32 0
%p1 = bitcast float addrspace(1)* %complex1.gep to i64 addrspace(1)*
%v1 = load i64, i64 addrspace(1)* %p1
store i64 %v1, i64 addrspace(2)* %a
%f1 = load float, float addrspace(2)* %a.gep1
%f2 = load float, float addrspace(2)* %a.gep2
%sum = fadd float %f1, %f2
store float %sum, float addrspace(2)* %a.gep1
store float %sum, float addrspace(2)* %a.gep2
%v2 = load i64, i64 addrspace(2)* %a
%complex2.gep = getelementptr [2 x float], [2 x float] addrspace(1)* %complex2, i32 0, i32 0
%p2 = bitcast float addrspace(1)* %complex2.gep to i64 addrspace(1)*
store i64 %v2, i64 addrspace(1)* %p2
ret void
}