[llvm-commits] [PATCH] ARM/NEON: Custom lower ctpop for appropriate vector types

[Pete Couperus]

Hi Anton,

Thanks for taking a look.  I had trouble with complex patterns for other
similar items, but I'll take another look at this case if that actually
sounds feasible.
Looking at LowerCTTZ, using Pat<> would be the same for cttz, although its
lowering is much simpler, correct?  Is there guidelines for when to use
Pat<> and when to lay it out in the code?
Thanks again for taking a look!

Pete



On Thu, Nov 29, 2012 at 9:53 PM, Anton Korobeynikov <anton at korobeynikov.info
> wrote:

> Pete,
>
> > The population count intrinsic (ctpop) is supported on ARM/NEON for v8i8
> and
> > v16i8 types via NEON's vcnt instruction.
> > This patch leverages vcnt and other NEON instructions to custom lower
> ctpop
> > for v2i32/v4i32 and v4i16/v8i16 types.
> > As you'll see, a fair chunk of this patch is comments describing the
> > lowering, which I am happy to adjust to people's liking.
> > Please review!
> > Thanks!
> Can't you use Pat<> magic in .td file to expand ctpop into series of
> nodes? It seems exactly what's your lowering does...
>
> --
> With best regards, Anton Korobeynikov
> Faculty of Mathematics and Mechanics, Saint Petersburg State University
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20121130/41e3d804/attachment.html>