[AArch64LoadStoreOptimizer] Check aliasing correctly when creating paired loads/stores.

[efriedma]

Summary

The existing code accidentally skipped the aliasing check in edge cases.

[mcrosier]

Ouch. Thanks for catching this, Eli. LGTM.

lib/Target/AArch64/AArch64LoadStoreOptimizer.cpp

if (areCandidatesToMergeOrPair(FirstMI, MI, Flags, TII) &&
       if (IsUnscaled != MIIsUnscaled) {
         // We're trying to pair instructions that differ in how they are scaled.
         // If FirstMI is scaled then scale the offset of MI accordingly.
         // Otherwise, do the opposite (i.e., make MI's offset unscaled).
         int MemSize = getMemScale(MI);
         if (MIIsUnscaled) {
           // If the unscaled offset isn't a multiple of the MemSize, we can't
           // pair the operations together: bail and keep looking.
-          if (MIOffset % MemSize)
+          if (MIOffset % MemSize) {
+            trackRegDefsUses(MI, ModifiedRegs, UsedRegs, TRI);
+            MemInsns.push_back(&MI);
             continue;
+          }
           MIOffset /= MemSize;
         } else {
           MIOffset *= MemSize;
         }
       }
 
       if (BaseReg == MIBaseReg && ((Offset == MIOffset + OffsetStride) ||
                                    (Offset + OffsetStride == MIOffset))) {

test/CodeGen/AArch64/ldst-paired-aliasing.ll

+; RUN: llc -mcpu cortex-a53 < %s | FileCheck %s
+target datalayout = "e-m:e-i64:64-i128:128-n8:16:32:64-S128"
+target triple = "aarch64--linux-gnu"
+
+declare void @f(i8*, i8*)
+declare void @f2(i8*, i8*)
+declare void @_Z5setupv()
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #3
+
+define i32 @main() local_unnamed_addr #1 {
+; Make sure the stores happen in the correct order (the exact instructions could change).
+; CHECK-LABEL: main:
+; CHECK: str q0, [sp, #48]
+; CHECK: ldr w8, [sp, #48]
+; CHECK: stur q1, [sp, #72]
+; CHECK: str q0, [sp, #64]
+; CHECK: str w9, [sp, #80]
+
+for.body.lr.ph.i.i.i.i.i.i63:
+  %b1 = alloca [10 x i32], align 16
+  %x0 = bitcast [10 x i32]* %b1 to i8*
+  %b2 = alloca [10 x i32], align 16
+  %x1 = bitcast [10 x i32]* %b2 to i8*
+  tail call void @_Z5setupv()
+  %x2 = getelementptr inbounds [10 x i32], [10 x i32]* %b1, i64 0, i64 6
+  %x3 = bitcast i32* %x2 to i8*
+  call void @llvm.memset.p0i8.i64(i8* %x3, i8 0, i64 16, i32 8, i1 false)
+  %arraydecay2 = getelementptr inbounds [10 x i32], [10 x i32]* %b1, i64 0, i64 0
+  %x4 = bitcast [10 x i32]* %b1 to <4 x i32>*
+  store <4 x i32> <i32 1, i32 1, i32 1, i32 1>, <4 x i32>* %x4, align 16
+  %incdec.ptr.i7.i.i.i.i.i.i64.3 = getelementptr inbounds [10 x i32], [10 x i32]* %b1, i64 0, i64 4
+  %x5 = bitcast i32* %incdec.ptr.i7.i.i.i.i.i.i64.3 to <4 x i32>*
+  store <4 x i32> <i32 1, i32 1, i32 1, i32 1>, <4 x i32>* %x5, align 16
+  %incdec.ptr.i7.i.i.i.i.i.i64.7 = getelementptr inbounds [10 x i32], [10 x i32]* %b1, i64 0, i64 8
+  store i32 1, i32* %incdec.ptr.i7.i.i.i.i.i.i64.7, align 16
+  %x6 = load i32, i32* %arraydecay2, align 16
+  %cmp6 = icmp eq i32 %x6, 1
+  br i1 %cmp6, label %for.inc, label %if.then
+
+for.inc:
+  call void @f(i8* %x0, i8* %x1)
+  ret i32 0
+
+if.then:
+  call void @f2(i8* %x0, i8* %x1)
+  ret i32 0
+}