On 12/01/2014 15:29, Aaron Ballman wrote:
On Sun, Jan 12, 2014 at 10:18 AM, Alp Toker <[email protected]> wrote:
Author: alp
Date: Sun Jan 12 09:18:06 2014
New Revision: 199053
URL: http://llvm.org/viewvc/llvm-project?rev=199053&view=rev
Log:
Clarify warn_cxx98_compat_attribute diagnostic
Various attribute flavours are supported in C++98. Make it clear that this
compatibility warning relates specifically to C++11-style generalized
attributes.
Modified:
cfe/trunk/include/clang/Basic/DiagnosticParseKinds.td
cfe/trunk/test/SemaCXX/cxx98-compat.cpp
Modified: cfe/trunk/include/clang/Basic/DiagnosticParseKinds.td
URL:
http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/DiagnosticParseKinds.td?rev=199053&r1=199052&r2=199053&view=diff
==============================================================================
--- cfe/trunk/include/clang/Basic/DiagnosticParseKinds.td (original)
+++ cfe/trunk/include/clang/Basic/DiagnosticParseKinds.td Sun Jan 12 09:18:06
2014
@@ -517,7 +517,7 @@ def warn_cxx98_compat_nullptr : Warning<
def warn_cxx98_compat_alignas : Warning<"'alignas' is incompatible with
C++98">,
InGroup<CXX98Compat>, DefaultIgnore;
def warn_cxx98_compat_attribute : Warning<
- "attributes are incompatible with C++98">,
+ "generalized attributes are incompatible with C++98">,
What are "generalized" attributes? I think it would be better-worded
as "C++11 attributes are incompatible with C++98" (this is more
consistent with other parser diagnostics, as well).
Hi Aaron,
I'd say it's time to use the more natural name for C++11 attributes
because the syntax now appears in a published standard. That's the usual
lifecycle for clang language features moving from experimental to fully
supported.
As for the name itself, it's the terminology adopted by the C++
community and also the name by which we advertise the feature on our own
C++ status page at http://clang.llvm.org/cxx_status.html:
Generalized attributes N2761 Clang 3.3 (1)
There's also a compelling reason _not_ to keep the old experimental
naming scheme indefinitely -- doing so will cause terminology dissonance
as other language dialects like ISO C look to adopt the new syntax,
likewise if we decide to introduce generalized attributes as a clang
extension to C11 in the meantime.
Alp.
Otherwise, LGTM!
~Aaron
--
http://www.nuanti.com
the browser experts
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
