[PATCH] D20732: Don't use static variables in LambdaCapture

John Brawn via cfe-commits cfe-commits at lists.llvm.org
Wed Jun 8 09:52:10 PDT 2016


john.brawn added a comment.

In http://reviews.llvm.org/D20732#452055, @faisalv wrote:

> I'm assuming you've given some thought to any existing pathological code that might break if the actual compile-time alignment of Decl itself is changed (since currently it seems to be that the alignment requirements are established at run-time and only through allocation via new (not that anyone would be creating Decls on the stack)) - and whether there are any consequences of relevance there...


I can't think of anything that could go wrong because of it. Reducing the alignment could cause problems, if there were things that check that the alignment is at least some value, but here I'm increasing it. Decl and all of its subclasses only have protected or private constructors and the only way to create them is through the various XYZDecl::Create methods which go through new so there should be no problem there either.


Repository:
  rL LLVM

http://reviews.llvm.org/D20732





More information about the cfe-commits mailing list