Hi,

Of course, like I said in my previous emails, this is more of a scratch pad
to see what's worth it and what's working. There will obviously be one
commit per kernel per pull request.


--Albin

On Thu, May 30, 2019, 17:04 Andrej Rode <[email protected]> wrote:

> Hi Albin,
>
> just to throw in some snarky comments.
>
> On Thu, 30 May 2019 16:20:44 +0200
> Albin Stigö <[email protected]> wrote:
>
> > I've finished ARM NEON SIMD optimization for 44 kernels which means
> > only 23 kernels (not often used) are now missing a NEON
> > implementation.
>
> I've checked on your progress a couple of weeks ago and just to have
> that info beforehand: It is really beneficial to have a good
> reviewability with meaningful commits that can be merged seperately.
> Since your changeset will be quite big I think going for a commit per
> kernel might be a way to go. That will help getting reviews from other
> experts on NEON (Me not being an expert in that field).
>
> Otherwise keep up the good work!
>
> Cheers,
> Andrej
>
> _______________________________________________
> Discuss-gnuradio mailing list
> [email protected]
> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
>
_______________________________________________
Discuss-gnuradio mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

Reply via email to