[PATCH] #pragma vectorize

Tyler Nowicki tnowicki at apple.com
Thu Jun 5 18:43:50 PDT 2014


Ok, here is the patch without [[clang::fallthrough]].

Could you please commit the patch. I don’t have commit privileges.

Thanks,

Tyler


On Jun 5, 2014, at 6:00 PM, Richard Smith <richard at metafoo.co.uk> wrote:

> On Wed, Jun 4, 2014 at 8:18 PM, Reid Kleckner <rnk at google.com> wrote:
> On Wed, Jun 4, 2014 at 7:57 PM, Tyler Nowicki <tnowicki at apple.com> wrote:
> Hi Aaron and Richard,
> 
> Thanks for the review and putting up with me while I learn clang! Here is the patch with your changes.
> 
> 
> >> +      // Vectorization/interleaving is disabled, set width/count to 1.
> >> +      ValueInt = 1;
> >> +    // Fallthrough.
> >
> > We should add [[clang::fallthrough]] to Compiler.h and use it here. :-)
> 
> Turns out I didn’t need to make any changes to Compiler.h. It works fine for me. Will it work for all builds?
> 
> Unlikely.  MSVC's table of C++11 lists attributes as unsupported:
> http://msdn.microsoft.com/en-us/library/hh567368.aspx
> 
> GCC is also likely to warn on unrecognized attributes, which wouldn't work.
> 
> A change to add [[clang::fallthrough]] to Compiler.h should not be part of this patch. Please go ahead with this as-is, and we can deal with the fallthrough macro part separately.
> 
> IIRC, Richard Trieu sent out a patch quite a while ago adding a fallthrough macro to Compiler.h (along with patches to Clang and LLVM to use it). I don't call what happened to that.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140605/a1ce3ecb/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pragma_loop-svn.patch
Type: application/octet-stream
Size: 46517 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140605/a1ce3ecb/attachment.obj>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140605/a1ce3ecb/attachment-0001.html>


More information about the cfe-commits mailing list