[llvm-commits] Fix in VSQRTSSm_Int pattern in X86 AVX CodeGen - Please review

[Demikhovsky, Elena]

Thank you. I found a smaller sequence and added to avx-arith.ll.

- Elena

-----Original Message-----
From: Bruno Cardoso Lopes [mailto:bruno.cardoso at gmail.com] 
Sent: Monday, November 28, 2011 17:35
To: Demikhovsky, Elena
Cc: llvm-commits at cs.uiuc.edu
Subject: Re: [llvm-commits] Fix in VSQRTSSm_Int pattern in X86 AVX CodeGen - Please review

Hi Elena,

On Mon, Nov 28, 2011 at 8:31 AM, Demikhovsky, Elena
<elena.demikhovsky at intel.com> wrote:
> Fixed an order of operands in VSQRTSSm_Int form:
>
>   def SSm_Int : SSI<opc, MRMSrcMem, (outs VR128:$dst),
> -                (ins ssmem:$src1, VR128:$src2),
> +                (ins VR128:$src1, ssmem:$src2),
>                 !strconcat(OpcodeStr,
>                            "ss\t{$src2, $src1, $dst|$dst, $src1, $src2}"), []>;
>
> Provided a test case.

Looks great! But please provide a more reduced testcase, and place it
inside avx-arith.ll

-- 
Bruno Cardoso Lopes
http://www.brunocardoso.cc
---------------------------------------------------------------------
Intel Israel (74) Limited

This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.