aaron.ballman added inline comments.
================
Comment at: clang/include/clang/Sema/ParsedAttr.h:1105
void clearListOnly() {
ParsedAttributesView::clearListOnly();
Range = SourceRange();
----------------
erichkeane wrote:
> This is... oh boy. I'm hopeful you can remove this type as well.
+1, it'd be fantastic if we could, otherwise we're storing the range twice for
this type (and it's named `Range` both times).
================
Comment at: clang/lib/Parse/ParseDeclCXX.cpp:3207
+ if (Tok.is(tok::kw___attribute)) {
+ ParsedAttributes Attrs(AttrFactory);
MaybeParseGNUAttributes(Attrs);
----------------
erichkeane wrote:
> This seems like a particularly strange change, What is the reasoning for
> this? Is it just that the Attrs are unused?
The comment above is that we parse and discard any trailing attributes, so this
change is scoping the `ParsedAttributes` object more tightly to the only scope
it's needed, but otherwise not changing behavior.
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D121201/new/
https://reviews.llvm.org/D121201
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
