From: Charlie Somerville Date: 2014-03-12T11:09:11+11:00 Subject: [ruby-core:61425] Re: [REJECT?] xmalloc/xfree: reduce atomic ops w/ thread-locals --531fa5a7_1190cde7_ab4f Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline On Wednesday, 12 March 2014 at 11:06 am, Eric Wong wrote: > This gives a =7E12% speedup to the bm=5Fso=5Fcount=5Fwords.rb benchmark= > which does many small mallocs. This pattern is common in some Ruby > scripts doing text processing, so maybe it is worth doing. This is pretty common in large applications (eg. Rails apps), so this may= be beneficial there. I=E2=80=99ll perform some experiments and report back. --531fa5a7_1190cde7_ab4f Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline
On We= dnesday, 12 March 2014 at 11:06 am, Eric Wong wrote:
This gives a =7E12% speed= up to the bm=5Fso=5Fcount=5Fwords.rb benchmark
which does many small mallocs. This pattern is common in some Ruby
=
scripts doing text processing, so maybe it is worth= doing.
=20 =20 =20 =20
=20

This is pretty common in large applications (e= g. Rails apps), so this may be beneficial there.

I=E2=80=99ll perform some experiments and report back.

--531fa5a7_1190cde7_ab4f--