[comp.sys.amiga] Track buffering. Was "A Plea..."

UH2@PSUVM.BITNET (Lee Sailer) (06/07/88)

Speaking of file systems and such.  A while back it was more or less
established (I think) that track buffering would help floppies, especially
in cases where two processes are both trying to access the disk.

Did anybody ever do anything about this?  Couldn't someone who knew about
devices make up a new trackdisk.device that mostly just buffered tracks and
called the old trackdisk.device (after renaming it, I guess)?

page@swan.ulowell.edu (Bob Page) (06/14/88)

UH2@PSUVM.BITNET (Lee Sailer) wrote:
>track buffering would help floppies ... anybody ever do anything about this?

It wasn't like nobody every thought of it.  Early Amigas were 256K
systems, remember.  Add a couple of track buffers for df0 and df1 and
you're now using about 50K ... that's a lot of space on a 256K system.

Today things are different (memory wise), but not much, as that 50K
must still come out of CHIP ram.  (Well, I could see how to change that,
but why bother)

A better solution is to have the file system keep the buffers.
Use FACC or AddBuffers to accomplish that.

..Bob
-- 
Bob Page, U of Lowell CS Dept.  page@swan.ulowell.edu  ulowell!page

michael@stb.UUCP (Michael) (06/25/88)

In article <7525@swan.ulowell.edu> page@swan.ulowell.edu (Bob Page) writes:
>[talking about track buffering of floppies]
>
>A better solution is to have the file system keep the buffers.
>Use FACC or AddBuffers to accomplish that.

Doesn't work. File system reads in 11 sectors, one gets buffered, 10 get
thrown out.

Perry, Facc III?
				Michael
: --- 
: Michael Gersten			 uunet.uu.net!denwa!stb!michael
:				sdcsvax!crash!gryphon!denwa!stb!michael
: What would have happened if we had lost World War 2. Well, the west coast
: would be owned by Japan, we would all be driving foreign cars, hmm...