[llvm-bugs] [Bug 35295] Potential improvement for vectorization of sub

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=35295

Sanjay Patel <spatel+llvm at rotateright.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|NEW                         |RESOLVED

--- Comment #6 from Sanjay Patel <spatel+llvm at rotateright.com> ---
(In reply to Serguei Katkov from comment #3)
> Also the code for mul generated by icc is better than code generated by
> clang as well... I guess the reason should be close to this sub but I'm not
> sure.

Yes - this is very similar to bug 35299.

This should be fixed in IR with:
https://reviews.llvm.org/rL318404

We might still want to add ISD::SUB to x86's list of narrowable binops in
combineTruncatedArithmetic(), but I think we should open another bug if that's
needed.

Also, someone looking at non-x86 optimizations might want to open a bug for the
problem mentioned in comment 5.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20171116/bce117b8/attachment.html>