thakis added a comment. Nice!
================ Comment at: include/clang/Basic/DiagnosticGroups.td:294 @@ -293,2 +293,3 @@ def SelTypeCast : DiagGroup<"cast-of-sel-type">; +def CastCallingConvention : DiagGroup<"cast-calling-convention">; def FunctionDefInObjCContainer : DiagGroup<"function-def-in-objc-container">; ---------------- nit: since this isn't referenced in this file, consider using an unnamed inline `InGroup<DiagGroup<"cast-calling-convention">` in the other file instead ================ Comment at: include/clang/Basic/DiagnosticSemaKinds.td:6515 @@ +6514,3 @@ +def warn_cast_calling_conv : Warning< + "cast between incompatible calling conventions '%0' and '%1'">, + InGroup<CastCallingConvention>, DefaultIgnore; ---------------- Maybe add ", likely to abort at runtime" to communicate the gravity of the situation better. ================ Comment at: include/clang/Basic/DiagnosticSemaKinds.td:6516 @@ +6515,3 @@ + "cast between incompatible calling conventions '%0' and '%1'">, + InGroup<CastCallingConvention>, DefaultIgnore; +def note_change_calling_conv_fixit : Note< ---------------- I'd build a large codebase of your choice with this enabled and see how it does. If it does well I'd land it default-on. (I spent some time enabling DefaultIgnore warnings that nobody turned on after they landed recently, and I'd prefer to not having to do this for this warning too :-) ) ================ Comment at: lib/Sema/SemaCast.cpp:1762 @@ +1761,3 @@ + + // The source expression is a pointer to a known function defined in this TU. + // Diagnose this cast, as it is probably bad. ---------------- If it's an inline function in a header, would we want to warn? Should this check if the body's SourceLocation is from the main file? ================ Comment at: lib/Sema/SemaCast.cpp:1775 @@ +1774,3 @@ + Self.Diag(NameLoc, diag::note_change_calling_conv_fixit) + << FD << DstCCName << FixItHint::CreateInsertion(NameLoc, CCAttrText); +} ---------------- Consider doing something like r164858 did to see if there's a define for the calling convention (WINAPI or what have you) and suggest that instead http://reviews.llvm.org/D17348 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
