[ValueTracking] Add logic for `isKnownNonZero(umax X, Y)`

[goldstein.w.n]

Summary

(umax X, Y) != 0 -> X != 0 || Y != 0

Alive2 Link:

https://alive2.llvm.org/ce/z/_Z9AUT

[nikic]

LGTM

[goldstein.w.n]

Rebase

[goldstein.w.n]

Rebase

[goldstein.w.n]

Rebase

llvm/lib/Analysis/ValueTracking.cpp

if (auto *II = dyn_cast<IntrinsicInst>(I)) {
         std::optional<bool> ugt = KnownBits::ugt(XKnown, YKnown);
         return ugt && *ugt;
       }
       case Intrinsic::sadd_sat:
         if (isNonZeroAdd(DemandedElts, Depth, Q, BitWidth, II->getArgOperand(0),
                          II->getArgOperand(1)))
           return true;
         break;
+      case Intrinsic::umax:
       case Intrinsic::uadd_sat:
         if (isKnownNonZero(II->getArgOperand(0), DemandedElts, Depth, Q) ||
             isKnownNonZero(II->getArgOperand(1), DemandedElts, Depth, Q))
           return true;
         break;
       case Intrinsic::cttz:
         return computeKnownBits(II->getArgOperand(0), DemandedElts, Depth, Q)
             .Zero[0];

llvm/test/Analysis/ValueTracking/known-non-zero.ll

;
   %r = icmp eq i8 %z, 0
   ret i1 %r
 }
 
 define i1 @umax_nonzero(i8 %xx, i8 %yy, i8 %ind) {
 ; CHECK-LABEL: @umax_nonzero(
 ; CHECK-NEXT:    [[X_NZ:%.*]] = icmp ne i8 [[XX:%.*]], 0
 ; CHECK-NEXT:    call void @llvm.assume(i1 [[X_NZ]])
-; CHECK-NEXT:    [[X:%.*]] = call i8 @llvm.umax.i8(i8 [[XX]], i8 [[YY:%.*]])
-; CHECK-NEXT:    [[Z:%.*]] = or i8 [[X]], [[IND:%.*]]
-; CHECK-NEXT:    [[R:%.*]] = icmp eq i8 [[Z]], 0
-; CHECK-NEXT:    ret i1 [[R]]
+; CHECK-NEXT:    ret i1 false
 ;
   %x_nz = icmp ne i8 %xx, 0
   call void @llvm.assume(i1 %x_nz)
   %x = call i8 @llvm.umax.i8(i8 %xx, i8 %yy)
   %z = or i8 %x, %ind
   %r = icmp eq i8 %z, 0
   ret i1 %r
 }

llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll

 ;     B[i/2] = A[i] * 2;
 ; }
 
 
 define void @even_load_dynamic_tc(ptr noalias nocapture readonly %A, ptr noalias nocapture %B, i64 %N) {
 ; CHECK-LABEL: @even_load_dynamic_tc(
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    [[UMAX:%.*]] = call i64 @llvm.umax.i64(i64 [[N:%.*]], i64 2)
-; CHECK-NEXT:    [[TMP0:%.*]] = add i64 [[UMAX]], -1
-; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 8
+; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[UMAX]], 9
 ; CHECK-NEXT:    br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
 ; CHECK:       vector.ph:
+; CHECK-NEXT:    [[TMP0:%.*]] = add i64 [[UMAX]], -1
 ; CHECK-NEXT:    [[TMP1:%.*]] = lshr i64 [[TMP0]], 1
 ; CHECK-NEXT:    [[TMP2:%.*]] = add nuw i64 [[TMP1]], 1
 ; CHECK-NEXT:    [[N_MOD_VF:%.*]] = and i64 [[TMP2]], 3
 ; CHECK-NEXT:    [[TMP3:%.*]] = icmp eq i64 [[N_MOD_VF]], 0
 ; CHECK-NEXT:    [[TMP4:%.*]] = select i1 [[TMP3]], i64 4, i64 [[N_MOD_VF]]
 ; CHECK-NEXT:    [[N_VEC:%.*]] = sub i64 [[TMP2]], [[TMP4]]
 ; CHECK-NEXT:    [[IND_END:%.*]] = shl i64 [[N_VEC]], 1
 ; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]