[cfe-commits] r161650 - in /cfe/trunk: include/clang/Frontend/VerifyDiagnosticConsumer.h lib/ARCMigrate/ARCMT.cpp lib/Frontend/VerifyDiagnosticConsumer.cpp test/ARCMT/verify.m test/ASTMerge/function.c test/Frontend/verify.c test/Frontend/verify2.c test/Frontend/veri

Andy Gibbs andyg1001 at hotmail.co.uk
Fri Aug 17 08:15:53 PDT 2012


On Thursday, August 16, 2012 8:12 PM, Jordan Rose wrote:
> On Aug 16, 2012, at 11:00 , Andy Gibbs wrote:
>> ---When we get a diagnostic---
>> if (is a macro expansion)
>> ignore [[big culprit this one!]]
>
> Why not look at where the macro is instantiated?

Is it necessary?  A macro expansion itself will not contain comments,
and hence no directive?... (that's right isn't it?)

>> What do you think to this approach?
>
> I think it works pretty well, but I'd like to see it in code. :-)
> Apparently my reservations always appear on second reading.

So how about your reservations with the patch attached? :o)

It is roughly the implementation as discussed, but I've also tidied
a few other loose scraggly bits too (e.g. findDirectives only needs
to look up to the first directive).  I also did a bit more on the
comments too.

Preprocessor now only needed to do the CommentHandler and for the
modules header file detection.  And, of course, the major change
compared with trunk is that we now better support memory buffers
as well as files (cling's use-case).

I'll look forward to hearing if it passes your critical second
reading!  Thanks for your persistence!

Cheers
Andy


-------------- next part --------------
A non-text attachment was scrubbed...
Name: verify.diff
Type: application/octet-stream
Size: 14475 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20120817/c9867ffb/attachment.obj>


More information about the cfe-commits mailing list