[LLVMdev] Please benchmark new x86 vector shuffle lowering, planning to make it the default very soon!
Chandler Carruth
chandlerc at google.com
Fri Sep 19 13:39:12 PDT 2014
FWIW, I agree with Quentin. I'm actually confident I have test cases
covering all of these.
On Fri, Sep 19, 2014 at 1:22 PM, Andrea Di Biagio <andrea.dibiagio at gmail.com
> wrote:
> 3. When zero extending 2 packed 32-bit integers, we should try to
> emit a vpmovzxdq
> Example:
> vmovq 20(%rbx), %xmm0
> vpshufd $80, %xmm0, %xmm0 # %xmm0 = %xmm0[0,0,1,1]
>
> Before:
> vpmovzxdq 20(%rbx), %xmm0
>
This one is already fixed.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140919/880c7012/attachment.html>
More information about the llvm-dev
mailing list