[PATCH] D53438: [llvm-exegesis] Reject x86 instructions that use non uniform memory accesses

[Clement Courbet via Phabricator via llvm-commits]

courbet added inline comments.


================
Comment at: tools/llvm-exegesis/lib/X86/Target.cpp:28
+
+static std::vector<llvm::ArrayRef<Operand>>
+getMemoryOperandRanges(llvm::ArrayRef<Operand> Operands) {
----------------
doc ?


================
Comment at: tools/llvm-exegesis/lib/X86/Target.cpp:28
+
+static std::vector<llvm::ArrayRef<Operand>>
+getMemoryOperandRanges(llvm::ArrayRef<Operand> Operands) {
----------------
courbet wrote:
> doc ?

```
struct MemoryOperandRange {
  bool isValid() const { return Range.size() == 5; }
  llvm::ArrayRef<Operand>Ops;
};

std::vector<MemoryOperandRange>
```


================
Comment at: tools/llvm-exegesis/lib/X86/Target.cpp:317
+    for (auto Ops : getMemoryOperandRanges(IT.Instr.Operands)) {
+      switch (Ops.size()) {
+      case 5:
----------------
is (Ops.isValid())


Repository:
  rL LLVM

https://reviews.llvm.org/D53438