Date: Tue, 27 May 1997 23:01:00 -0400 (EDT) From: Christopher Sedore <cmsedore@mailbox.syr.edu> To: "Justin T. Gibbs" <gibbs@plutotech.com> Cc: "Ron G. Minnich" <rminnich@Sarnoff.COM>, FreeBSD-Hackers@FreeBSD.ORG Subject: Re: async socket stuff Message-ID: <Pine.SOL.3.95.970527225442.11962B-100000@rodan.syr.edu> In-Reply-To: <199705272205.QAA08460@pluto.plutotech.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 27 May 1997, Justin T. Gibbs wrote: > >Now, in the case of a 10MB file, you've essentially saved 20MB worth of > >memory bandwidth/time for transfer (since the data has to be copied into > >user space on read, and back again on write), plus 160*2=320-1=319 system > >calls avoided. > > If you use an async I/O facility, there is no additional copy since you've > pre-allocated the buffer and I/O from the file goes directly into and out > of your user space buffer. Since your main complaint seems to be memory > bandwidth and the system call overhead is really quite small (FreeBSD can > do thousands of system calls a second on a P90), I think that async I/O > would completely solve your problem. There is already work underway to > bring async I/O to FreeBSD. I didn't know whether the async implementation would provide this or not. I still view too many syscalls as undesirable. This overhead seems a shame. -Chris
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.SOL.3.95.970527225442.11962B-100000>