Date: Sun, 15 Dec 2002 21:26:47 +0100 From: phk@FreeBSD.ORG To: Nate Lawson <nate@root.org> Cc: freebsd-hackers@FreeBSD.ORG Subject: Re: setattr() syscall as proposed by phk Message-ID: <71695.1039984007@critter.freebsd.dk> In-Reply-To: Your message of "Sun, 15 Dec 2002 12:21:16 PST." <Pine.BSF.4.21.0212151211031.44745-100000@root.org>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <Pine.BSF.4.21.0212151211031.44745-100000@root.org>, Nate Lawson wri tes: >> Because 1 syscall and 2 namei calls are faster than 4 syscalls and >> four namei calls. > >Which leaves us back at my previous point which is that something is wrong >with caching if 4 namei calls (for the SAME name) are so much slower. A >great task would be to examine/test namei(), identify why it's not cached >correctly, and fix the underlying problem. The syscall boundary crossing >is NOT significant here. When you restore a N files, it makes a difference if you need to do 4N syscalls and 4N namei's or just 1N syscalls and 2N nameis. Caching makes the difference smaller, but the fact remains that there is a difference. In particular as N climbs into the range of millions. Don't forget that restore uses (and has to!) paths relative to its root directory, each of the nameis can trivially contain a handful of component names. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?71695.1039984007>