home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.os2.advocacy
- Path: sparky!uunet!caen!destroyer!news.iastate.edu!jws
- From: jws@iastate.edu (James William Shoemaker)
- Subject: Re: PC Week, November 9
- Message-ID: <Bxz7B9.B7H@news.iastate.edu>
- Sender: news@news.iastate.edu (USENET News System)
- Organization: Iowa State University, Ames, IA
- References: <1992Nov13.161920.24065@panix.com> <pathak-171192171854@virtual.mitre.org>
- Date: Thu, 19 Nov 1992 18:28:19 GMT
- Lines: 16
-
-
- > In some cases, even this may be true. If my I/O device buffers input, I
- > have the choice of processing data as a single byte (interrupt-driven I/O)
- > or in chunks (polled I/O). In many cases, it is more efficient to process
- > data in chunks (in some cases one might be required to process data in
- > chucks). Then the choice is to either let the I/O hardware buffer and to
- > process the data in chunks or to grab each byte and buffer the data
- > manually until enough data is received so processing can be done. I don't
- > know about you, but I would rather let the hardware do the buffering.
-
- If the device can buffer input, why not tell it to only interrupt you
- if the buffer is X% full? The 16550 works that way. Just because a device is
- interrupt driven doesn't mean it interrupts you every byte, just when it wants
- servicing.
-
- JWS
-