Re: bypassing unified buffer cache?
site_archiver@lists.apple.com Delivered-To: darwin-kernel@lists.apple.com User-agent: Mutt/1.2.5i On Thu, Jul 27, 2006 at 04:59:41PM +0100, Quinn wrote:
At 11:50 -0400 27/7/06, Travis Rose wrote:
My other question: is this code going to be platform-specific, or is it making calls into BSD-kernel which potentially can map to linux?
I believe that Linux has similar functionality, but I don't know them off the top of my head.
Most modern Unix alike systems support direct I/O these days (incl. Linux), and many applications are coded to use that interface, e.g. http://www.dba-oracle.com/oracle_tips_direct_io.htm ... the second half of that page gives advice on some OSes allowing direct I/O. Its typically implemented as a flag to open(2) and an fcntl(F_SETFL) also enables it. On the surface, it'd seem to be a straightforward exercise to map O_DIRECT to F_NOCACHE or vice-versa but probably the devil's in the details. cheers. -- Nathan _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.a... This email sent to site_archiver@lists.apple.com
participants (1)
-
Nathan Scott