Date: Mon, 23 Apr 2007 19:43:56 +0200 From: Roman Divacky <rdivacky@freebsd.org> To: Tim Kientzle <kientzle@freebsd.org> Cc: current@freebsd.org, "Jesper B. Rosenkilde" <jbr@humppa.dk> Subject: Re: Suggestions on Avoiding syscall Overhead Message-ID: <20070423174356.GA46913@freebsd.org> In-Reply-To: <462CD251.9060105@freebsd.org> References: <f126fae00704221639l68095de1ye7ce9ba3d921bf20@mail.gmail.com> <20070423113400.GC28587@gw.humppa.dk> <462CD251.9060105@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Apr 23, 2007 at 08:35:45AM -0700, Tim Kientzle wrote: > >>We can have 3 type of pages mapped into one process's address map. > >>1. System wide global readonly page which will help on these syscalls: > >> gethostname,getdomainname,uname > >> help on importing sysenter as syscall entry point!! > >> > >>2. Per process Readonly page. (change will still through standard syscall) > >> help on the syscalls: > >> getuid, geteuid, getpid,getgid, getegid, getpgrp, > >> > >>3. As you planed, Read+Write Page > > I don't really understand this suggestion. > > Do any real programs call these syscalls very often? > If not, this is unnecessary complexity. Making > a syscall fast that's only called once doesn't really > help performance. > > This type of approach is sometimes suggested for systime(), > which could have a big impact, as there are real programs > that call systime() thousands of times per second. > But I find it hard to believe there are programs that > call getuid() that often. on the other hand, if there is a easy to use API for this I don't see why it should not be converted. I agree that speeding up a "function" that gets called just once doesn't make much sense but if it means 5minutes + 20lines patch I think its worth it. just my 2 cents roman
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070423174356.GA46913>