[PATCH] D100980: [OpenCL] Allow use of double type without extension pragma

Anastasia Stulova via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Wed Apr 28 03:33:40 PDT 2021


Anastasia added a comment.

In D100980#2722112 <https://reviews.llvm.org/D100980#2722112>, @azabaznov wrote:

> In D100980#2719322 <https://reviews.llvm.org/D100980#2719322>, @Anastasia wrote:
>
>> In D100980#2719196 <https://reviews.llvm.org/D100980#2719196>, @azabaznov wrote:
>>
>>>> When the pragma is parsed we can't know why it is in the code to be able to issue any warning.
>>>
>>> I mean diagnose once when, for example in your particular case, double type is parsed.  Does it require much effort? I think this warning might be useful for developers who already rely on pragma usage in their kernels.
>>
>> I am not sure I understand your suggestion. Could you show an example perhaps?
>
> All right. Currently, what we do have for OpenCL C < 1.2  (https://godbolt.org/z/rjYWMj7v1):
>
>   <source>:6:5: error: use of type 'double' requires cl_khr_fp64 support
>       double d;
>       ^
>   1 error generated.
>
> What I suggest is to have:
>
>   <source>:6:5: warning: pragma enable is no longer required for use of type 'double'
>       double d;
>       ^
>   1 warning generated.
>
> We can issue the warning if certain flag is provided for example. Does it make sense?

Not sure what do we want to achieve with this? Do you want to point out that the code might be somehow less portable let's say between clang revisions, etc? I think we could do it as it should not be too complicated but adds a bit extra complexity into the command line interface.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D100980/new/

https://reviews.llvm.org/D100980



More information about the cfe-commits mailing list