[PATCH] D51411: [OpenCL] Improve diagnostic of argument in address space conversion builtins

Yaxun Liu via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Wed Sep 5 07:08:06 PDT 2018


yaxunl added a comment.

In https://reviews.llvm.org/D51411#1224615, @Anastasia wrote:

> In https://reviews.llvm.org/D51411#1217477, @yaxunl wrote:
>
> > Is this a feature requested by users?
> >
> > I can understand that this may be useful to pinpoint some conversions which are potentially harmful to performance. However such conversions can often be eliminated by optimization, which makes this warning less useful.
> >
> > On the other hand, too many warnings is a nuance to users, therefore I am wondering whether this warning should be off by default.
> >
> > Do you have any chance to have any feedback from users about how useful or intrusive this warning is.
>
>
> Hi Sam, no feedback from the user. If you think this is not that useful we won't commit or make is off by default as you suggested. Let us know what you think makes more sense.


I suggest to make it off by default. Since implicit casting to generic pointer is allowed by spec, we only want to see this warning when we want to debug performance issues.


https://reviews.llvm.org/D51411





More information about the cfe-commits mailing list