[llvm] [InstCombine] Fold umax(nuw_mul(x, C0), x + 1) into (x == 0 ? 1 : nuw_mul(x, C0)) (PR #123468)

[Nikita Popov via llvm-commits]

================
@@ -1847,6 +1847,29 @@ Instruction *InstCombinerImpl::visitCallInst(CallInst &CI) {
         return CastInst::Create(Instruction::ZExt, NarrowMaxMin, II->getType());
       }
     }
+    // If C0 is not 0:
+    //   umax(nuw_shl(x, C0), x + 1) -> x == 0 ? 1 : nuw_shl(x, C0)
+    // If C0 is not 0 or 1:
+    //   umax(nuw_mul(x, C0), x + 1) -> x == 0 ? 1 : nuw_mul(x, C0)
+    const APInt *C0;
+    auto matchShiftOrMul = [&](Value *I) {
+      return ((match(I, m_OneUse(m_NUWShl(m_Value(X), m_APInt(C0))))) ||
+              (match(I, m_OneUse(m_NUWMul(m_Value(X), m_APInt(C0)))) &&
+               !C0->isOne())) &&
+             !C0->isZero();
+    };
+    bool matchI0 = false;
+    if (IID == Intrinsic::umax &&
+        (((matchI0 = matchShiftOrMul(I0)) &&
+          match(I1, m_OneUse(m_Add(m_Specific(X), m_One())))) ||
+         (matchShiftOrMul(I1) &&
+          match(I0, m_OneUse(m_Add(m_Specific(X), m_One())))))) {
----------------
nikic wrote:

I think it would be cleaner if you structured the fold like moveNotAfterMinMax below (i.e. implement it in a lambda, call with commuted args).

https://github.com/llvm/llvm-project/pull/123468