Date: Thu, 28 Nov 1996 14:45:28 +0100 From: Poul-Henning Kamp <phk@critter.tfs.com> To: Bruce Evans <bde@zeta.org.au> Cc: current@freebsd.org Subject: Re: users of "ft" tapes, please test! Message-ID: <275.849188728@critter.tfs.com> In-Reply-To: Your message of "Thu, 28 Nov 1996 23:48:02 %2B1100." <199611281248.XAA24059@godzilla.zeta.org.au>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <199611281248.XAA24059@godzilla.zeta.org.au>, Bruce Evans writes: >>>Unless the 512(?)-byte code bloat costs another text page? >> >>the text segment got smaller too :-) > >It must have had large code for all the array address calculations. > >I have been thinking about un-inlining spls. This saves 29K out of >1096K text. It may even save some time (because function call overhead >is small better locality more than compensates for it). I haven't found >a benchmark that shows a clear advantage either way. I tried kernel >compiles, i/o's with a small block size, `ping -f localhost's with the >usual (small) block size, and ttcp's with small and large block sizes, >on a P5 and a 486/33. ttcp is known to use a lot of spls. I generally use the rule of thumb that unless the text-size is smaller as a result, then inlining is wrong. -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@ref.tfs.com TRW Financial Systems, Inc. Future will arrive by its own means, progress not so.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?275.849188728>