Date: Fri, 22 Jun 2001 21:22:23 -0700 From: Terry Lambert <tlambert2@mindspring.com> To: Josh Osborne <stripes@mac.com> Cc: Richard Hodges <rh@matriplex.com>, freebsd-hackers@FreeBSD.ORG Subject: Re: question: aio / nbio / kqueue Message-ID: <3B34197F.55BC93F@mindspring.com> References: <200106230044.RAA18157@smtpout.mac.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Josh Osborne wrote: > BSD/OS had select working for FFS files (returns ready to read > if the block the file pointer is at is in the buffer cache, and > sends a read ahead request). Or at least they (Paul?) calmed > they did, I never tested it. This would be good to see in FreeBSD. > I try to avoid anything that makes me write signal handlers (AIO > is done with signals, right?), Not traditionally. aiowait/aiocancel act on previously instituted aioread/aiowrite. I don't know what damage POSIC has done to them. I personally never enabled SIGPOLL, even when using SystemV, since signals are persistant conditions, not events, so you can not trust that you will get the signals you need to get when an event happens that would trigger the signal. This is just like SIGCHLD, which will deliver two signals for 40 processes dying (the one you first get, the one that gets set while the first handler is blocking, and then the other 38 that set the same flag which is already 1 back to 1... don't get an event at all). -- Terry 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?3B34197F.55BC93F>