[PATCH] D139067: [FuzzMutate] New InsertCFGStrategy

[Peter Rong via Phabricator via llvm-commits]

Peter marked an inline comment as done.
Peter added inline comments.


================
Comment at: llvm/lib/FuzzMutate/IRMutator.cpp:343-344
+                    }));
+    assert(RS && "I can't find an integer type in all allowed types, is the "
+                 "setting correct?");
+    Type *Ty = RS.getSelection();
----------------
arsenm wrote:
> assert string is a bit weird
This is to make sure that when things went wrong, we give some hint to tell the user what's wrong. Is there a recommended way of doing it?


================
Comment at: llvm/lib/FuzzMutate/IRMutator.cpp:365-374
+    for (uint64_t i = 0; i < NumCases; i++) {
+      uint64_t CaseVal = [&CaseVals, MaxCaseVal, &IB]() {
+        uint64_t tmp;
+        // Make sure we don't have two cases with same value.
+        do {
+          tmp = uniform<uint64_t>(IB.Rand, 0, MaxCaseVal);
+        } while (CaseVals.count(tmp) != 0);
----------------
arsenm wrote:
> Why is this a lambda here?
Just to enclose unnecessary variables(`tmp` required to make CaseVal unique. I can write the loop outside the lambda if you want to.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D139067/new/

https://reviews.llvm.org/D139067