[PATCH] D11146: [ValueTracking] Add support for floating-point minnum and maxnum

[David Majnemer]

majnemer added inline comments.

================
Comment at: lib/Analysis/ValueTracking.cpp:3323
@@ +3322,3 @@
+
+  if (ConstantFP *C = dyn_cast<ConstantFP>(V))
+    return !C->isNaN();
----------------
Use `auto *`, we know the type from the LHS.

================
Comment at: lib/Analysis/ValueTracking.cpp:3329
@@ +3328,3 @@
+static bool isKnownNonZero(Value *V) {
+  if (ConstantFP *C = dyn_cast<ConstantFP>(V))
+    return !C->isZero();
----------------
Ditto.


http://reviews.llvm.org/D11146