LGTM with minor tweaks.
I'm still not sure we've got the function template specialization case right,
but I think this adds most of the value as-is, and I think you're right to
blacklist an area with known false positives over not having the warning at all.
================
Comment at: lib/Lex/Pragma.cpp:934-936
@@ -929,4 +933,5 @@
+#endif
void DebugOverflowStack() {
DebugOverflowStack();
}
#ifdef _MSC_VER
----------------
Somewhat unrelated to your patch, but this looks broken (at `-O1` or above).
How about replacing this with:
static void DebugOverflowStack() {
void (*volatile Self)() = DebugOverflowStack;
self();
}
That should both suppress your warning (and presumably the MSVC warning) and
make this function work reliably.
================
Comment at: lib/Sema/AnalysisBasedWarnings.cpp:90-93
@@ +89,6 @@
+
+static void CheckForFunctionCall(Sema &S, const FunctionDecl *FD,
+ CFGBlock& Block, unsigned exitID,
+ llvm::SmallVectorImpl<RecursiveState> &states,
+ RecursiveState state) {
+ unsigned ID = Block.getBlockID();
----------------
For new code, please follow the coding style's formatting rules
(`checkForFunctionCall`, `States`, `State`, `ExitID`, and `" &"` not `"& "`).
================
Comment at: lib/Sema/AnalysisBasedWarnings.cpp:96-99
@@ +95,6 @@
+
+ if (states[ID] < state)
+ states[ID] = state;
+ else
+ return;
+
----------------
Maybe
if (States[ID] >= State)
return;
States[ID] = State;
// ...
http://llvm-reviews.chandlerc.com/D1864
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
