[llvm-branch-commits] [llvm] bcbdeaf - Revert "[SLP]Need shrink the load vector after reordering."

[Alexander Belyaev via llvm-branch-commits]

Author: Alexander Belyaev
Date: 2021-01-08T14:42:26+01:00
New Revision: bcbdeafa9cb3469a7bf367c17f32d913076a4d55

URL: https://github.com/llvm/llvm-project/commit/bcbdeafa9cb3469a7bf367c17f32d913076a4d55
DIFF: https://github.com/llvm/llvm-project/commit/bcbdeafa9cb3469a7bf367c17f32d913076a4d55.diff

LOG: Revert "[SLP]Need shrink the load vector after reordering."

This reverts commit 4284afdf9432f7d756f56b0ab21d69191adefa8d.

This changes computed values in fused_batchnorm_test_cpu.

Not equal to tolerance rtol=1e-06, atol=0.001
Mismatched value: a is different from b.
not close where = (array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1]), array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
       1, 1]), array([0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1,
       1, 1]), array([0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3,
       4, 5]))
not close lhs = [-0.6636615  -0.9804948  -1.148275   -0.68193716 -0.8572368  -0.65046215
 -0.6993756  -1.2244141  -1.0938729  -0.50369143 -0.51830524 -0.738452
 -0.7214286  -0.48115745 -0.9380924  -0.9341769  -0.5916775  -1.2896856
 -0.7264182  -0.9746917  -0.783249   -0.7659018  -0.86214024 -0.47784212]
not close rhs = [ 0.44102234  0.12418899 -0.04359123  0.42274666  0.24744703  0.45422167
  0.40530816 -0.11973029  0.01081094  0.6009924   0.5863786   0.3662318
  0.38325527  0.62352633  0.1665914   0.1705069   0.5130063  -0.18500176
  0.37826565  0.12999213  0.3214348   0.338782    0.24254355  0.62684166]
not close dif = [1.1046839 1.1046838 1.1046838 1.1046839 1.1046839 1.1046839 1.1046838
 1.1046839 1.1046839 1.1046839 1.1046839 1.1046839 1.1046839 1.1046838
 1.1046839 1.1046839 1.1046839 1.1046839 1.1046839 1.1046839 1.1046839
 1.1046839 1.1046838 1.1046838]
not close tol = [0.00100044 0.00100012 0.00100004 0.00100042 0.00100025 0.00100045
 0.00100041 0.00100012 0.00100001 0.0010006  0.00100059 0.00100037
 0.00100038 0.00100062 0.00100017 0.00100017 0.00100051 0.00100019
 0.00100038 0.00100013 0.00100032 0.00100034 0.00100024 0.00100063]

Added: 
    

Modified: 
    llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
    llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
index ef0dea0f11d3..f124dd8ef374 100644
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -4260,13 +4260,18 @@ Value *BoUpSLP::vectorizeTree(ArrayRef<Value *> VL) {
       if (E->isSame(VL)) {
         Value *V = vectorizeTree(E);
         if (VL.size() == E->Scalars.size() && !E->ReuseShuffleIndices.empty()) {
-          // Reshuffle to get only unique values.
-          SmallVector<int, 4> UniqueIdxs;
-          SmallSet<int, 4> UsedIdxs;
-          for (int Idx : E->ReuseShuffleIndices)
-            if (UsedIdxs.insert(Idx).second)
-              UniqueIdxs.emplace_back(Idx);
-          V = Builder.CreateShuffleVector(V, UniqueIdxs, "shrink.shuffle");
+          // We need to get the vectorized value but without shuffle.
+          if (auto *SV = dyn_cast<ShuffleVectorInst>(V)) {
+            V = SV->getOperand(0);
+          } else {
+            // Reshuffle to get only unique values.
+            SmallVector<int, 4> UniqueIdxs;
+            SmallSet<int, 4> UsedIdxs;
+            for (int Idx : E->ReuseShuffleIndices)
+              if (UsedIdxs.insert(Idx).second)
+                UniqueIdxs.emplace_back(Idx);
+            V = Builder.CreateShuffleVector(V, UniqueIdxs);
+          }
         }
         return V;
       }

diff  --git a/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll b/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll
index 533b48eec295..110e1177774b 100644
--- a/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll
+++ b/llvm/test/Transforms/SLPVectorizer/X86/shrink_after_reorder.ll
@@ -10,8 +10,7 @@ define void @wombat(i32* %ptr, i32* %ptr1) {
 ; CHECK-NEXT:    [[TMP1:%.*]] = load <2 x i32>, <2 x i32>* [[TMP0]], align 8
 ; CHECK-NEXT:    [[SHUFFLE:%.*]] = shufflevector <2 x i32> [[TMP1]], <2 x i32> poison, <4 x i32> <i32 1, i32 0, i32 1, i32 0>
 ; CHECK-NEXT:    [[TMP27:%.*]] = getelementptr inbounds i32, i32* [[PTR1:%.*]], i32 3
-; CHECK-NEXT:    [[SHRINK_SHUFFLE:%.*]] = shufflevector <4 x i32> [[SHUFFLE]], <4 x i32> poison, <2 x i32> <i32 0, i32 1>
-; CHECK-NEXT:    [[TMP2:%.*]] = add nsw <2 x i32> [[SHRINK_SHUFFLE]], <i32 -1, i32 -1>
+; CHECK-NEXT:    [[TMP2:%.*]] = add nsw <2 x i32> [[TMP1]], <i32 -1, i32 -1>
 ; CHECK-NEXT:    [[SHUFFLE1:%.*]] = shufflevector <2 x i32> [[TMP2]], <2 x i32> poison, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
 ; CHECK-NEXT:    [[TMP34:%.*]] = getelementptr inbounds i32, i32* [[PTR1]], i32 4
 ; CHECK-NEXT:    [[TMP40:%.*]] = getelementptr inbounds i32, i32* [[PTR1]], i32 5