Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!usc!elroy.jpl.nasa.gov!ames!pasteur!ucbvax!hplabs!nsc!amdahl!rbw00
From: rbw00@uts.amdahl.com (Richard Wilmont)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Need input for future DOS release
Keywords: future DOS release,I/O Overlap
Message-ID: <29Nq02FZ954W01@amdahl.uts.amdahl.com>
Date: 23 Mar 90 06:05:15 GMT
References: <53686@microsoft.UUCP> <2017@clyde.concordia.ca> <1990Mar22.202023.25752@seri.gov>
Reply-To: rbw00@amdahl.uts.amdahl.com (Richard Wilmot)
Organization: Amdahl Corporation, Sunnyvale CA
Lines: 22


And wouldn't it be nice if we could do asynchronous I/O. I understand that
the restriction of synchronous I/O is usually dictated by BIOS coding (?)
but if the DOS *standard* included interrupts for doing it then we could
put more effective pressure on BIOS vendors. I could use this facility in
ways: My search program could read the first block, issue an asynchronous
read for the next block and begin searching the one just read.

We used this technique on early computers to obtain a benefit we called
*I/O OVERLAP*.

Oh, this facility will also need a way to check on the status of a
previously issued I/O command (also called WAIT or CHECK).

I run 2 hard drives in my home machine and would like to be able to start
an I/O to physical drive 1 before the previous command to physical drive
2 has finished.

Not asking for multiprogramming or multitasking, just peripheral/compute
overlap.

     - Dick Wilmot