https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104124

--- Comment #8 from Surya Kumari Jangala <jskumari at gcc dot gnu.org> ---
(In reply to Avinash Jayakar from comment #7)
> I had a question regarding the patch that fixed this PR. 
> In the commit message it is mentioned that for 0 and -1 the splat and unpack
> will not work.
> 1. But in the code, the check is for +1
>   value = INTVAL (elt);
>   if (value == 0 || value == 1
>       || !EASY_VECTOR_15 (value))
>     return false;
> 
> should this be corrected?
> 
> 2. Any reason why not for 0 and -1?

Yes, it should be -1. I have a patch for this at
https://gcc.gnu.org/pipermail/gcc-patches/2025-August/692871.html

Reply via email to