Folding nodes in instruction selection - please review

[Quentin Colombet]

Hi Elena,

This is looking good with some minor comments:
- Could you paste the comment on the additional argument for IsProfitableToFold in the header file too?
- Add a test case.

Thanks,

-Quentin

On Jul 10, 2013, at 12:27 AM, "Demikhovsky, Elena" <elena.demikhovsky at intel.com> wrote:

> Hi, please review this patch.
>  
> Thanks.
>  
> Elena
>  
>  
> _____________________________________________
> From: Demikhovsky, Elena 
> Sent: Monday, July 08, 2013 14:48
> To: llvm-commits at cs.uiuc.edu
> Cc: Demikhovsky, Elena; Nadav Rotem <nrotem at apple.com> (nrotem at apple.com)
> Subject: Folding nodes in instruction selection - please review
>  
>  
> Hi,
>  
> I analyzed the folding DAG nodes algorithm in LLVM, compared to code generated by Intel compiler and got into conclusion that the LLVM code is not always optimal.
> This is an example:
>  
>   %b1 = fadd <8 x float> %a1, <float AAA, float AAA, float AAA, float AAA, float AAA, float AAA, float AAA, float AAA >
>   %b2 = fadd <8 x float> %a2, <float AAA, float AAA, float AAA, float AAA, float AAA, float AAA, float AAA, float AAA >
>   %c = fmul <8 x float> %b1, %b2
>  
> The result (1) bellow is not better than (2), because loading constant is not a problem, but spilling %ymm2 that may be required  in (1) is not cheap.
>  
> (1)
>         vmovaps .LCPI1_0(%rip), %ymm2
>         vaddps  %ymm2, %ymm1, %ymm1
>         vaddps  %ymm2, %ymm0, %ymm0
>         vmulps  %ymm1, %ymm0, %ymm0
>  
> (2)
>         vaddps  .LCPI1_0(%rip), %ymm1, %ymm1
>         vaddps  .LCPI1_0(%rip),  %ymm0, %ymm0
>         vmulps  %ymm1, %ymm0, %ymm0
>  
> In the attached patch I did a minor thing that does not change the current algorithm, but allows to add some target-specific analysis.
> Please review.
>  
> Thank you.
>  
> Elena
>  
>  
>  
> ---------------------------------------------------------------------
> 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.
> 
> <folding.diff>_______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130710/72d591f9/attachment.html>