From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: References: <5d1347dfd611729cd82ac5bc0ca79c92@coraid.com> Date: Tue, 8 Dec 2009 14:52:14 -0500 Message-ID: <7d3530220912081152u519e803dy2029af21451ba943@mail.gmail.com> From: John Floren To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [9fans] etherigbe.c using _xinc? Topicbox-Message-UUID: aca73728-ead5-11e9-9d60-3106f5b1d025 On Tue, Dec 8, 2009 at 2:35 PM, Russ Cox wrote: >> because it's not a Ref. =C2=A0unfortunately, if it were >> a Ref, it would be much faster. =C2=A0_xinc is deadly >> slow even if there is no contention on x86. > > do you have numbers to back up this claim? > I don't have the code or the numbers in front of me, but I recall seeing quite a bit of speed improvement when I experimentally replaced incref/decref with direct calls to _xinc/_xdec. I don't remember what the test was, but I do remember that I got something like 35% improvement on it. I ran that kernel on my terminal for the rest of the summer without trouble; while I didn't notice a blazing speed increase, it didn't slow me down either. John --=20 "Object-oriented design is the roman numerals of computing" -- Rob Pike