Date: Sun, 22 Jul 2012 17:53:58 +0100 (BST) From: Robert Watson <rwatson@FreeBSD.org> To: Matthew Jacob <mj@feral.com> Cc: freebsd-arch@freebsd.org Subject: Re: aio in GENERIC? Message-ID: <alpine.BSF.2.00.1207221752310.83786@fledge.watson.org> In-Reply-To: <500752CD.9030107@feral.com> References: <3CE55F29-A5B2-44A7-8854-1ED38BAE6F16@FreeBSD.org> <50075072.5050906@gmail.com> <500752CD.9030107@feral.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 18 Jul 2012, Matthew Jacob wrote: > What practically does aio offer that is not achieved via pthreads other than > slightly simpler code? Although the VFS side of the AIO code blocks kernel threads during in-progress I/O, the socket side is able to do fully asynchronous I/O down the stack without committing a kernel thread to it. As such, it probably is actually significantly more scalable, allowing larger numbers of simultaneous outstanding I/Os in the network layer, and more efficiently than simply using threads. Robert
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.1207221752310.83786>