[PowerPC] make lwa as a valid ds candidate in ppcloopinstrformprep pass

Fix a FIXME in ppcloopinstrformprep pass.

Reviewed by: nemanjai

Differential Revision: https://reviews.llvm.org/D71346

diff --git a/llvm/lib/Target/PowerPC/PPCLoopInstrFormPrep.cpp b/llvm/lib/Target/PowerPC/PPCLoopInstrFormPrep.cpp
index 086db4e..467a119 100644 (file)
--- a/llvm/lib/Target/PowerPC/PPCLoopInstrFormPrep.cpp
+++ b/llvm/lib/Target/PowerPC/PPCLoopInstrFormPrep.cpp
@@ -846,11 +846,15 @@ bool PPCLoopInstrFormPrep::runOnLoop(Loop *L) {
   // Check if a load/store has DS form.
   auto isDSFormCandidate = [] (const Instruction *I, const Value *PtrValue) {
     assert((PtrValue && I) && "Invalid parameter!");
-    // FIXME: 32 bit instruction lwa is also DS form.
-    return !isa<IntrinsicInst>(I) &&
-           ((PtrValue->getType()->getPointerElementType()->isIntegerTy(64)) ||
-            (PtrValue->getType()->getPointerElementType()->isFloatTy()) ||
-            (PtrValue->getType()->getPointerElementType()->isDoubleTy()));
+    if (isa<IntrinsicInst>(I))
+      return false;
+    Type *PointerElementType = PtrValue->getType()->getPointerElementType();
+    return (PointerElementType->isIntegerTy(64)) ||
+           (PointerElementType->isFloatTy()) ||
+           (PointerElementType->isDoubleTy()) ||
+           (PointerElementType->isIntegerTy(32) &&
+            llvm::any_of(I->users(),
+                         [](const User *U) { return isa<SExtInst>(U); }));
   };
 
   // Check if a load/store has DQ form.

diff --git a/llvm/test/CodeGen/PowerPC/loop-instr-form-prepare.ll b/llvm/test/CodeGen/PowerPC/loop-instr-form-prepare.ll
index 297e301..42a2c78 100644 (file)
--- a/llvm/test/CodeGen/PowerPC/loop-instr-form-prepare.ll
+++ b/llvm/test/CodeGen/PowerPC/loop-instr-form-prepare.ll
@@ -774,20 +774,17 @@ define float @test_ds_combine_float_int(i8* %0, i32 signext %1) {
 
 define i64 @test_ds_lwa_prep(i8* %0, i32 signext %1) {
 ; CHECK-LABEL: test_ds_lwa_prep:
-; CHECK:         li r6, 1
-; CHECK-NEXT:    li r7, 2
-; CHECK-NEXT:    li r8, 6
-; CHECK-NEXT:    li r9, 10
+; CHECK:         addi r5, r3, 2
+; CHECK:         li r6, -1
 ; CHECK:       .LBB9_2: #
-; CHECK-NEXT:    lwax r11, r3, r6
-; CHECK-NEXT:    lwax r12, r3, r7
-; CHECK-NEXT:    lwax r0, r3, r8
-; CHECK-NEXT:    addi r10, r3, 1
-; CHECK-NEXT:    mulld r11, r12, r11
-; CHECK-NEXT:    lwax r3, r3, r9
-; CHECK-NEXT:    mulld r11, r11, r0
-; CHECK-NEXT:    maddld r5, r11, r3, r5
-; CHECK-NEXT:    mr r3, r10
+; CHECK-NEXT:    lwax r7, r5, r6
+; CHECK-NEXT:    lwa r8, 0(r5)
+; CHECK-NEXT:    lwa r9, 4(r5)
+; CHECK-NEXT:    lwa r10, 8(r5)
+; CHECK-NEXT:    addi r5, r5, 1
+; CHECK-NEXT:    mulld r7, r8, r7
+; CHECK-NEXT:    mulld r7, r7, r9
+; CHECK-NEXT:    maddld r3, r7, r10, r3
 ; CHECK-NEXT:    bdnz .LBB9_2
 
   %3 = sext i32 %1 to i64
@@ -827,3 +824,4 @@ define i64 @test_ds_lwa_prep(i8* %0, i32 signext %1) {
   %33 = add nsw i64 %32, %3
   ret i64 %33
 }
+