[llvm] Use getSignedTargetConstant for offset (PR #141149)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-powerpc

Author: Shimin Cui (scui-ibm)

<details>
<summary>Changes</summary>

This is to fix an assertion failure with PeepholePPC64. The load/store offset can be negative. A reduced case from one of our failures is added as well.

---
Full diff: https://github.com/llvm/llvm-project/pull/141149.diff


2 Files Affected:

- (modified) llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp (+2-2) 
- (added) llvm/test/CodeGen/PowerPC/signed-offset.ll (+33) 


``````````diff
diff --git a/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp b/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
index 7147506bbd89b..f921032356d65 100644
--- a/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
+++ b/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
@@ -7858,8 +7858,8 @@ void PPCDAGToDAGISel::PeepholePPC64() {
         if (!isInt<16>(Offset))
           continue;
 
-        ImmOpnd = CurDAG->getTargetConstant(Offset, SDLoc(ImmOpnd),
-                                            ImmOpnd.getValueType());
+        ImmOpnd = CurDAG->getSignedTargetConstant(Offset, SDLoc(ImmOpnd),
+                                                  ImmOpnd.getValueType());
       } else if (Offset != 0) {
         // This optimization is performed for non-TOC-based local-[exec|dynamic]
         // accesses.
diff --git a/llvm/test/CodeGen/PowerPC/signed-offset.ll b/llvm/test/CodeGen/PowerPC/signed-offset.ll
new file mode 100644
index 0000000000000..d12f99ead644b
--- /dev/null
+++ b/llvm/test/CodeGen/PowerPC/signed-offset.ll
@@ -0,0 +1,33 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
+; RUN: llc -verify-machineinstrs < %s -mcpu=ppc -mtriple powerpc-ibm-aix-xcoff | FileCheck %s
+
+%_type = type <{ [90112 x i8] }>
+
+ at g = external global %_type
+
+define void @foo(ptr %p) {
+; CHECK-LABEL: foo:
+; CHECK:     	 stfd 31, 0(31)
+; CHECK-NEXT:    lwz 3, -32(30)
+; CHECK-NEXT:    lwz 4, -28(30)
+; CHECK-NEXT:    cmplwi 3, 0
+; CHECK-NEXT:    cmpwi 1, 3, 0
+; CHECK-NEXT:    crandc 20, 5, 2
+; CHECK-NEXT:    cmpwi 1, 4, 0
+entry:
+  %0 = load double, ptr getelementptr inbounds nuw (i8, ptr @g, i32 83272), align 8
+  %1 = call i32 @bar(ptr getelementptr inbounds nuw (i8, ptr @g, i32 83272))
+  %2 = call i32 @bar(ptr getelementptr inbounds nuw (i8, ptr @g, i32 83240))
+  store double %0, ptr %p, align 8
+  %3 = load i64, ptr getelementptr inbounds nuw (i8, ptr @g, i32 83240), align 8
+  %4 = icmp slt i64 %3, 1
+  br i1 %4, label %then, label %else
+
+then:                                                     ; preds = %entry
+  ret void
+
+else:                                                     ; preds = %entry
+  ret void
+}
+
+declare signext i32 @bar(ptr)

``````````

</details>


https://github.com/llvm/llvm-project/pull/141149