Date: Fri, 23 Jan 1998 08:43:07 -0500 From: "Alfred Perlstein" <perlsta@sunyit.edu> To: "Mike Smith" <mike@smith.net.au> Cc: <daniel_sobral@voga.com.br>, <mike@smith.net.au>, <hackers@FreeBSD.ORG> Subject: Re: uiomove() Message-ID: <199801230951.JAA12460@fang.cs.sunyit.edu>
next in thread | raw e-mail | index | archive | help
> > I'm hoping i don't get yelled at here, but why not queue all subsequent > > open operations and block those processes operations on the device? > > at least until a read is done? > This doesn't necessarily help; if a process holding an open descriptor > on your device forks, there are now two processes holding open > descriptors but there has been no second open() call. This has been > discussed to death. ooops :) > > what's the point of an encryption card if you can't have multiple processes > > acesses it, at least "transparently" in parrallel? > I dunno, but the card in question performs stream, not block > encryption, and there is no mechanism (that Daniel seems to know of > anyway) to recover context from the card to allow switching streams. is this a "Fred's(tm) encryption card"? :) -Alfred
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199801230951.JAA12460>