[ValueTracking] Teach isKnownNonZero that a vscale is never 0.

[dmgreen]

Summary

A llvm.vscale will always be at least 1, never zero. Teaching that to isKnownNonZero can help fold away some statically known compares.

[paulwalker-arm]

llvm/lib/Analysis/ValueTracking.cpp

2701

The existing coding style suggests this should be

}
// Intrinsic
else if (...

llvm/lib/Analysis/ValueTracking.cpp

else if (const auto *EEI = dyn_cast<ExtractElementInst>(V)) {
     }
   }
   // Freeze
   else if (const FreezeInst *FI = dyn_cast<FreezeInst>(V)) {
     auto *Op = FI->getOperand(0);
     if (isKnownNonZero(Op, Depth, Q) &&
         isGuaranteedNotToBePoison(Op, Q.AC, Q.CxtI, Q.DT, Depth))
       return true;
+  } else if (const auto *II = dyn_cast<IntrinsicInst>(V)) {
+    if (II->getIntrinsicID() == Intrinsic::vscale)
+      return true;
   }
 
   KnownBits Known(BitWidth);
   computeKnownBits(V, DemandedElts, Known, Depth, Q);
   return Known.One != 0;
 }
 
 bool isKnownNonZero(const Value* V, unsigned Depth, const Query& Q) {

llvm/test/Transforms/InstCombine/vscale_zero.ll

 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt < %s -passes=instcombine -S | FileCheck %s
 
 define i1 @iszero() {
 ; CHECK-LABEL: @iszero(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.vscale.i32()
-; CHECK-NEXT:    [[TMP1:%.*]] = icmp eq i32 [[TMP0]], 0
-; CHECK-NEXT:    ret i1 [[TMP1]]
+; CHECK-NEXT:    ret i1 false
 ;
   entry:
   %0 = call i32 @llvm.vscale.i32()
   %1 = icmp eq i32 %0, 0
   ret i1 %1
 }
 
 define i1 @iszero_range() vscale_range(1,16) {
 ; CHECK-LABEL: @iszero_range(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.vscale.i32()
-; CHECK-NEXT:    [[TMP1:%.*]] = icmp eq i32 [[TMP0]], 0
-; CHECK-NEXT:    ret i1 [[TMP1]]
+; CHECK-NEXT:    ret i1 false
 ;
   entry:
   %0 = call i32 @llvm.vscale.i32()
   %1 = icmp eq i32 %0, 0
   ret i1 %1
 }
 
 declare i32 @llvm.vscale.i32()