RankedTensorType: Use getHashValue(KeyTy) when calling getHashValue(RankedTensorTypeStorage*).
PiperOrigin-RevId: 223649958
diff --git a/lib/IR/MLIRContext.cpp b/lib/IR/MLIRContext.cpp
index 423d2cd..0726db0 100644
--- a/lib/IR/MLIRContext.cpp
+++ b/lib/IR/MLIRContext.cpp
@@ -133,9 +133,12 @@
struct RankedTensorTypeKeyInfo : DenseMapInfo<RankedTensorTypeStorage *> {
// Ranked tensors are uniqued based on their element type and shape.
using KeyTy = std::pair<Type, ArrayRef<int>>;
- using DenseMapInfo<RankedTensorTypeStorage *>::getHashValue;
using DenseMapInfo<RankedTensorTypeStorage *>::isEqual;
+ static unsigned getHashValue(RankedTensorTypeStorage *key) {
+ return getHashValue(KeyTy(key->elementType, key->getShape()));
+ }
+
static unsigned getHashValue(KeyTy key) {
return hash_combine(
DenseMapInfo<Type>::getHashValue(key.first),
diff --git a/test/IR/repro_b120295301.mlir b/test/IR/repro_b120295301.mlir
new file mode 100755
index 0000000..57438af
--- /dev/null
+++ b/test/IR/repro_b120295301.mlir
@@ -0,0 +1,110 @@
+// RUN: mlir-opt %s | FileCheck %s
+
+cfgfunc @testType(tensor<1x224x224x3xf32>) -> tensor<96xf32> {
+bb0(%arg0: tensor<1x224x224x3xf32>):
+ %1 = "constant"() {value: splat<tensor<1xf32>, 0.1>} : () -> (tensor<1xf32>)
+ %2 = "constant"() {value: splat<tensor<2xf32>, 0.1>} : () -> (tensor<2xf32>)
+ %3 = "constant"() {value: splat<tensor<3xf32>, 0.1>} : () -> (tensor<3xf32>)
+ %4 = "constant"() {value: splat<tensor<4xf32>, 0.1>} : () -> (tensor<4xf32>)
+ %5 = "constant"() {value: splat<tensor<5xf32>, 0.1>} : () -> (tensor<5xf32>)
+ %6 = "constant"() {value: splat<tensor<6xf32>, 0.1>} : () -> (tensor<6xf32>)
+ %7 = "constant"() {value: splat<tensor<7xf32>, 0.1>} : () -> (tensor<7xf32>)
+ %8 = "constant"() {value: splat<tensor<8xf32>, 0.1>} : () -> (tensor<8xf32>)
+ %9 = "constant"() {value: splat<tensor<9xf32>, 0.1>} : () -> (tensor<9xf32>)
+ %10 = "constant"() {value: splat<tensor<10xf32>, 0.1>} : () -> (tensor<10xf32>)
+ %11 = "constant"() {value: splat<tensor<11xf32>, 0.1>} : () -> (tensor<11xf32>)
+ %12 = "constant"() {value: splat<tensor<12xf32>, 0.1>} : () -> (tensor<12xf32>)
+ %13 = "constant"() {value: splat<tensor<13xf32>, 0.1>} : () -> (tensor<13xf32>)
+ %14 = "constant"() {value: splat<tensor<14xf32>, 0.1>} : () -> (tensor<14xf32>)
+ %15 = "constant"() {value: splat<tensor<15xf32>, 0.1>} : () -> (tensor<15xf32>)
+ %16 = "constant"() {value: splat<tensor<16xf32>, 0.1>} : () -> (tensor<16xf32>)
+ %17 = "constant"() {value: splat<tensor<17xf32>, 0.1>} : () -> (tensor<17xf32>)
+ %18 = "constant"() {value: splat<tensor<18xf32>, 0.1>} : () -> (tensor<18xf32>)
+ %19 = "constant"() {value: splat<tensor<19xf32>, 0.1>} : () -> (tensor<19xf32>)
+ %20 = "constant"() {value: splat<tensor<20xf32>, 0.1>} : () -> (tensor<20xf32>)
+ %21 = "constant"() {value: splat<tensor<21xf32>, 0.1>} : () -> (tensor<21xf32>)
+ %22 = "constant"() {value: splat<tensor<22xf32>, 0.1>} : () -> (tensor<22xf32>)
+ %23 = "constant"() {value: splat<tensor<23xf32>, 0.1>} : () -> (tensor<23xf32>)
+ %24 = "constant"() {value: splat<tensor<24xf32>, 0.1>} : () -> (tensor<24xf32>)
+ %25 = "constant"() {value: splat<tensor<25xf32>, 0.1>} : () -> (tensor<25xf32>)
+ %26 = "constant"() {value: splat<tensor<26xf32>, 0.1>} : () -> (tensor<26xf32>)
+ %27 = "constant"() {value: splat<tensor<27xf32>, 0.1>} : () -> (tensor<27xf32>)
+ %28 = "constant"() {value: splat<tensor<28xf32>, 0.1>} : () -> (tensor<28xf32>)
+ %29 = "constant"() {value: splat<tensor<29xf32>, 0.1>} : () -> (tensor<29xf32>)
+ %30 = "constant"() {value: splat<tensor<30xf32>, 0.1>} : () -> (tensor<30xf32>)
+ %31 = "constant"() {value: splat<tensor<31xf32>, 0.1>} : () -> (tensor<31xf32>)
+ %32 = "constant"() {value: splat<tensor<32xf32>, 0.1>} : () -> (tensor<32xf32>)
+ %33 = "constant"() {value: splat<tensor<33xf32>, 0.1>} : () -> (tensor<33xf32>)
+ %34 = "constant"() {value: splat<tensor<34xf32>, 0.1>} : () -> (tensor<34xf32>)
+ %35 = "constant"() {value: splat<tensor<35xf32>, 0.1>} : () -> (tensor<35xf32>)
+ %36 = "constant"() {value: splat<tensor<36xf32>, 0.1>} : () -> (tensor<36xf32>)
+ %37 = "constant"() {value: splat<tensor<37xf32>, 0.1>} : () -> (tensor<37xf32>)
+ %38 = "constant"() {value: splat<tensor<38xf32>, 0.1>} : () -> (tensor<38xf32>)
+ %39 = "constant"() {value: splat<tensor<39xf32>, 0.1>} : () -> (tensor<39xf32>)
+ %40 = "constant"() {value: splat<tensor<40xf32>, 0.1>} : () -> (tensor<40xf32>)
+ %41 = "constant"() {value: splat<tensor<41xf32>, 0.1>} : () -> (tensor<41xf32>)
+ %42 = "constant"() {value: splat<tensor<42xf32>, 0.1>} : () -> (tensor<42xf32>)
+ %43 = "constant"() {value: splat<tensor<43xf32>, 0.1>} : () -> (tensor<43xf32>)
+ %44 = "constant"() {value: splat<tensor<44xf32>, 0.1>} : () -> (tensor<44xf32>)
+ %45 = "constant"() {value: splat<tensor<45xf32>, 0.1>} : () -> (tensor<45xf32>)
+ %46 = "constant"() {value: splat<tensor<46xf32>, 0.1>} : () -> (tensor<46xf32>)
+ %47 = "constant"() {value: splat<tensor<47xf32>, 0.1>} : () -> (tensor<47xf32>)
+ %48 = "constant"() {value: splat<tensor<48xf32>, 0.1>} : () -> (tensor<48xf32>)
+ %49 = "constant"() {value: splat<tensor<49xf32>, 0.1>} : () -> (tensor<49xf32>)
+ %50 = "constant"() {value: splat<tensor<50xf32>, 0.1>} : () -> (tensor<50xf32>)
+ %51 = "constant"() {value: splat<tensor<51xf32>, 0.1>} : () -> (tensor<51xf32>)
+ %52 = "constant"() {value: splat<tensor<52xf32>, 0.1>} : () -> (tensor<52xf32>)
+ %53 = "constant"() {value: splat<tensor<53xf32>, 0.1>} : () -> (tensor<53xf32>)
+ %54 = "constant"() {value: splat<tensor<54xf32>, 0.1>} : () -> (tensor<54xf32>)
+ %55 = "constant"() {value: splat<tensor<55xf32>, 0.1>} : () -> (tensor<55xf32>)
+ %56 = "constant"() {value: splat<tensor<56xf32>, 0.1>} : () -> (tensor<56xf32>)
+ %57 = "constant"() {value: splat<tensor<57xf32>, 0.1>} : () -> (tensor<57xf32>)
+ %58 = "constant"() {value: splat<tensor<58xf32>, 0.1>} : () -> (tensor<58xf32>)
+ %59 = "constant"() {value: splat<tensor<59xf32>, 0.1>} : () -> (tensor<59xf32>)
+ %60 = "constant"() {value: splat<tensor<60xf32>, 0.1>} : () -> (tensor<60xf32>)
+ %61 = "constant"() {value: splat<tensor<61xf32>, 0.1>} : () -> (tensor<61xf32>)
+ %62 = "constant"() {value: splat<tensor<62xf32>, 0.1>} : () -> (tensor<62xf32>)
+ %63 = "constant"() {value: splat<tensor<63xf32>, 0.1>} : () -> (tensor<63xf32>)
+ %64 = "constant"() {value: splat<tensor<64xf32>, 0.1>} : () -> (tensor<64xf32>)
+ %65 = "constant"() {value: splat<tensor<65xf32>, 0.1>} : () -> (tensor<65xf32>)
+ %66 = "constant"() {value: splat<tensor<66xf32>, 0.1>} : () -> (tensor<66xf32>)
+ %67 = "constant"() {value: splat<tensor<67xf32>, 0.1>} : () -> (tensor<67xf32>)
+ %68 = "constant"() {value: splat<tensor<68xf32>, 0.1>} : () -> (tensor<68xf32>)
+ %69 = "constant"() {value: splat<tensor<69xf32>, 0.1>} : () -> (tensor<69xf32>)
+ %70 = "constant"() {value: splat<tensor<70xf32>, 0.1>} : () -> (tensor<70xf32>)
+ %71 = "constant"() {value: splat<tensor<71xf32>, 0.1>} : () -> (tensor<71xf32>)
+ %72 = "constant"() {value: splat<tensor<72xf32>, 0.1>} : () -> (tensor<72xf32>)
+ %73 = "constant"() {value: splat<tensor<73xf32>, 0.1>} : () -> (tensor<73xf32>)
+ %74 = "constant"() {value: splat<tensor<74xf32>, 0.1>} : () -> (tensor<74xf32>)
+ %75 = "constant"() {value: splat<tensor<75xf32>, 0.1>} : () -> (tensor<75xf32>)
+ %76 = "constant"() {value: splat<tensor<76xf32>, 0.1>} : () -> (tensor<76xf32>)
+ %77 = "constant"() {value: splat<tensor<77xf32>, 0.1>} : () -> (tensor<77xf32>)
+ %78 = "constant"() {value: splat<tensor<78xf32>, 0.1>} : () -> (tensor<78xf32>)
+ %79 = "constant"() {value: splat<tensor<79xf32>, 0.1>} : () -> (tensor<79xf32>)
+ %80 = "constant"() {value: splat<tensor<80xf32>, 0.1>} : () -> (tensor<80xf32>)
+ %81 = "constant"() {value: splat<tensor<81xf32>, 0.1>} : () -> (tensor<81xf32>)
+ %82 = "constant"() {value: splat<tensor<82xf32>, 0.1>} : () -> (tensor<82xf32>)
+ %83 = "constant"() {value: splat<tensor<83xf32>, 0.1>} : () -> (tensor<83xf32>)
+ %84 = "constant"() {value: splat<tensor<84xf32>, 0.1>} : () -> (tensor<84xf32>)
+ %85 = "constant"() {value: splat<tensor<85xf32>, 0.1>} : () -> (tensor<85xf32>)
+ %86 = "constant"() {value: splat<tensor<86xf32>, 0.1>} : () -> (tensor<86xf32>)
+ %87 = "constant"() {value: splat<tensor<87xf32>, 0.1>} : () -> (tensor<87xf32>)
+ %88 = "constant"() {value: splat<tensor<88xf32>, 0.1>} : () -> (tensor<88xf32>)
+ %89 = "constant"() {value: splat<tensor<89xf32>, 0.1>} : () -> (tensor<89xf32>)
+ %90 = "constant"() {value: splat<tensor<90xf32>, 0.1>} : () -> (tensor<90xf32>)
+ %91 = "constant"() {value: splat<tensor<91xf32>, 0.1>} : () -> (tensor<91xf32>)
+ %92 = "constant"() {value: splat<tensor<92xf32>, 0.1>} : () -> (tensor<92xf32>)
+ %93 = "constant"() {value: splat<tensor<93xf32>, 0.1>} : () -> (tensor<93xf32>)
+ %94 = "constant"() {value: splat<tensor<94xf32>, 0.1>} : () -> (tensor<94xf32>)
+ %95 = "constant"() {value: splat<tensor<95xf32>, 0.1>} : () -> (tensor<95xf32>)
+ %96 = "constant"() {value: splat<tensor<96xf32>, 0.1>} : () -> (tensor<96xf32>)
+ %97 = "constant"() {value: splat<tensor<97xf32>, 0.1>} : () -> (tensor<97xf32>)
+ %98 = "constant"() {value: splat<tensor<98xf32>, 0.1>} : () -> (tensor<98xf32>)
+ %99 = "constant"() {value: splat<tensor<99xf32>, 0.1>} : () -> (tensor<99xf32>)
+ %100 = "constant"() {value: splat<tensor<100xf32>, 0.1>} : () -> (tensor<100xf32>)
+ %101 = "constant"() {value: splat<tensor<101xf32>, 0.1>} : () -> (tensor<101xf32>)
+ %102 = "constant"() {value: splat<tensor<102xf32>, 0.1>} : () -> (tensor<102xf32>)
+ return %96 : tensor<96xf32>
+}
+// CHECK: testType
+// CHECK: return %cst_94
