[llvm] [SLP] Add test to demonstrate ordering changes in vectorizeStores() refactor (PR #196018)

[via llvm-commits]

llvmorg-github-actions[bot] wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-risc-v

Author: Ryan Buchner (bababuck)

<details>
<summary>Changes</summary>

Test is for #<!-- -->193616. Because this test has two overlapping chains with different VFs, when the ordering that they are attempted to vectorize in changes, the result will change since the smaller chain cannot vectorize if the larger chain hasn't already vectorized due to the extract costs.

---
Full diff: https://github.com/llvm/llvm-project/pull/196018.diff


1 Files Affected:

- (modified) llvm/test/Transforms/SLPVectorizer/RISCV/load-store.ll (+50) 


``````````diff
diff --git a/llvm/test/Transforms/SLPVectorizer/RISCV/load-store.ll b/llvm/test/Transforms/SLPVectorizer/RISCV/load-store.ll
index a6c5273dc4cd0..81bed1403cb61 100644
--- a/llvm/test/Transforms/SLPVectorizer/RISCV/load-store.ll
+++ b/llvm/test/Transforms/SLPVectorizer/RISCV/load-store.ll
@@ -265,3 +265,53 @@ entry:
   store i32 77, ptr %inc3, align 2
   ret void
 }
+
+define void @shared-chain-ordering(ptr %dest, ptr %p, i64 %offset) {
+; CHECK-LABEL: @shared-chain-ordering(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[INCS0:%.*]] = getelementptr inbounds i16, ptr [[DEST:%.*]], i64 [[OFFSET:%.*]]
+; CHECK-NEXT:    [[TMP0:%.*]] = load <4 x i16>, ptr [[P:%.*]], align 4
+; CHECK-NEXT:    store <4 x i16> [[TMP0]], ptr [[INCS0]], align 4
+; CHECK-NEXT:    [[TMP1:%.*]] = extractelement <4 x i16> [[TMP0]], i32 0
+; CHECK-NEXT:    store i16 [[TMP1]], ptr [[DEST]], align 4
+; CHECK-NEXT:    [[INCS:%.*]] = getelementptr inbounds i16, ptr [[DEST]], i64 1
+; CHECK-NEXT:    [[TMP2:%.*]] = extractelement <4 x i16> [[TMP0]], i32 1
+; CHECK-NEXT:    store i16 [[TMP2]], ptr [[INCS]], align 2
+; CHECK-NEXT:    ret void
+;
+; DEFAULT-LABEL: @shared-chain-ordering(
+; DEFAULT-NEXT:  entry:
+; DEFAULT-NEXT:    [[INCS0:%.*]] = getelementptr inbounds i16, ptr [[DEST:%.*]], i64 [[OFFSET:%.*]]
+; DEFAULT-NEXT:    [[TMP0:%.*]] = load <4 x i16>, ptr [[P:%.*]], align 4
+; DEFAULT-NEXT:    store <4 x i16> [[TMP0]], ptr [[INCS0]], align 4
+; DEFAULT-NEXT:    [[TMP1:%.*]] = extractelement <4 x i16> [[TMP0]], i32 0
+; DEFAULT-NEXT:    store i16 [[TMP1]], ptr [[DEST]], align 4
+; DEFAULT-NEXT:    [[INCS:%.*]] = getelementptr inbounds i16, ptr [[DEST]], i64 1
+; DEFAULT-NEXT:    [[TMP2:%.*]] = extractelement <4 x i16> [[TMP0]], i32 1
+; DEFAULT-NEXT:    store i16 [[TMP2]], ptr [[INCS]], align 2
+; DEFAULT-NEXT:    ret void
+;
+entry:
+  %e0 = load i16, ptr %p, align 4
+  %inc = getelementptr inbounds i16, ptr %p, i64 1
+  %e1 = load i16, ptr %inc, align 2
+  %inc1 = getelementptr inbounds i16, ptr %p, i64 2
+  %e2 = load i16, ptr %inc1, align 2
+  %inc2 = getelementptr inbounds i16, ptr %p, i64 3
+  %e3 = load i16, ptr %inc2, align 2
+
+  %incs0 = getelementptr inbounds i16, ptr %dest, i64 %offset
+  store i16 %e0, ptr %incs0, align 4
+  %incs1 = getelementptr inbounds i16, ptr %incs0, i64 1
+  store i16 %e1, ptr %incs1, align 2
+  %incs2 = getelementptr inbounds i16, ptr %incs0, i64 2
+  store i16 %e2, ptr %incs2, align 2
+  %incs3 = getelementptr inbounds i16, ptr %incs0, i64 3
+  store i16 %e3, ptr %incs3, align 2
+
+  store i16 %e0, ptr %dest, align 4
+  %incs = getelementptr inbounds i16, ptr %dest, i64 1
+  store i16 %e1, ptr %incs, align 2
+
+  ret void
+}

``````````

</details>


https://github.com/llvm/llvm-project/pull/196018