[PATCH] D102343: [clang][patch][FPEnv} Initialization of C++ globals not strictfp aware

[Aaron Ballman via Phabricator via cfe-commits]

aaron.ballman added inline comments.


================
Comment at: clang/include/clang/AST/Decl.h:1993
                const DeclarationNameInfo &NameInfo, QualType T,
-               TypeSourceInfo *TInfo, StorageClass S, bool isInlineSpecified,
-               ConstexprSpecKind ConstexprKind,
+               TypeSourceInfo *TInfo, StorageClass S, bool UsesFPIntrin,
+               bool isInlineSpecified, ConstexprSpecKind ConstexprKind,
----------------
mibintc wrote:
> aaron.ballman wrote:
> > I have no idea if others agree, but I have a slight preference for putting `UsesFPIntrin` towards the end of the parameter list (here and in the Create methods) -- this flag seems less important than the constexpr kind or whether inline is specified (and, it helpfully avoids putting two `bool` parameters next to one another).
> > 
> > If you agree, perhaps it could be moved to before `TrailingRequiresClause`?
> I spent a few hours recoding to rearrange the parameter list, but there are downstream types from FunctionDecl which would presumably also need to be rearranged, and those types add more parameters and some of those parameters have default values, and moving the new boolean before TrailingRequiresClause doesn't work, I decided to throw away the mess I made and ask if you could live with it this way. 
Yup, I can live with it this way if changing it causes trouble. Thanks!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D102343/new/

https://reviews.llvm.org/D102343