Message-Id: <s1fdcf77.098@MAIL.TAIT.CO.NZ>
Date: Tue, 30 Jul 1996 09:03:22 +1200
From: Bill Currie <bill_currie AT MAIL DOT TAIT DOT CO DOT NZ>
To: djgpp-workers AT delorie DOT com
Subject:  Re: Opinion

On 29/7/96  6:36 pm, Eli Zaretskii <eliz AT is DOT elta DOT co DOT il> did thus say >>>
> 
> On Mon, 29 Jul 1996, Bill Currie wrote:
> 
> > What about a bios swap file (like windows) or partition? Sure, it limits swap
> > space to a fixed amout, but it should be faster than dos.
> 
> Are you sure it will be significantly faster?  AFAIK, the BIOS is one of 
> the causes of the slow real-mode disk I/O, because it ties the processor 
> until the operation completes.

No, I'm not sure, it's more the known HD consumption factor I think.  The bios
DOES (or at least in the ones I've looked at) make an int 15 (?) call while it's
waiting for IO so that the OS can keep going while the bios is waiting.
 
> Going to the BIOS level also means CWSDPMI will have to implement a disk
> cache (because third-party caches operate at the DOS/IOCTL level); 
> without a cache I think you will see a slow-down rather than a speed-up. 

I thought smartdrv grabbed the int 13 vector?