[PATCH] Implement aarch64 neon instruction class SIMD misc - LLVM

Kevin Qin kevinqindev at gmail.com
Wed Nov 13 18:56:27 PST 2013


I gathered all (Neon_movi (i32 255), (i32 14)) into Neon_AllOne and
commited it as http://llvm.org/viewvc/llvm-project?rev=194656&view=rev<http://llvm.org/viewvc/llvm-project?rev=194657&view=rev>
.


2013/11/14 Tim Northover <t.p.northover at gmail.com>

>
>   Hi Kevin,
>
>   Just spotted one small thing. It looks like there's a Neon_AllOne you
> could create along the lines of the Neon_AllZero. Everything else looks
> fine, so as far as I'm concerned you can make that change and commit
> immediately. I'll take a look at the svn log when it happens.
>
>   Cheers.
>
>   Tim.
>
>
> ================
> Comment at: lib/Target/AArch64/AArch64InstrNEON.td:6647
> @@ +6646,3 @@
> +            (v8i16 VPR128:$Rn),
> +            (v8i16 (bitconvert (v16i8 (Neon_movi (i32 255), (i32
> 14))))))),
> +          (NOT16b VPR128:$Rn)>;
> ----------------
> And is this AllOnes?
>
>
> http://llvm-reviews.chandlerc.com/D2118
>



-- 
Best Regards,

Kevin Qin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20131114/e51fd21d/attachment.html>


More information about the cfe-commits mailing list