From: hooft@fys.ruu.nl (Rob Hooft) Subject: Re: Floppies (was Re: File system issues!) Date: 15 Jul 1992 08:39:11 GMT
In <1992Jul14.150258.29853@hubcap.clemson.edu> dawill@hubcap.clemson.edu (david williams) writes:
> I'll say one thing for the floppy driver - it's much faster than the
>MS_DOG floppy. I was exceeding pissed at my 386 floppy performance,
>and had pretty much given up on my controller as the problem. Not so!
>Running floppy I/O under Linux is a breath of fresh air, so it's not my
>hardware... It's a problem with MS_LOSS. Once more, proof that Linux
>is a really nice system.
> Anybody have an idea why MS-DOG is so slow in comparison with Linux
>if the floppy drivers for Linux are in such a sad state?
I don't really think that ms/dos is slow. I think this seems to be the
case because it is doing its best to keep the floppy in a known state
all the time. You know: first writing a few block, immediately
updating the directory and the FAT. This causes a lot of seeking, and
at the end of a floppy at 3ms step rate a simple update of the FAT
costs at least 2x(80x3ms)=480ms. Linux does this in a better way
(probably most of the credits go to the caching system...).
As long as a filesystem is used on the floppy linux is about as fast
in reading/writing a floppy as ms/dos is, according to my experience.
Once you use mtools for reading floppies, the performance drops just
enough to upset the sector offsets that I always use (formatting my
floppies under ms/dos using `fdformat x 2 y 3'). mv'ing a file of
1.4MB to a floppy with a filesystem (like `mv t.t /mnt; time sync')
costs only about 35 seconds.
Of course I can only speak for my 486/33, it is possible that linux
performs worse on 386sx platforms. This is the problem with rotational
latencies.
-- Rob Hooft, Bijvoet Center for Biomolecular Research, Chemistry department University of Utrecht, the Netherlands hooft@hutruu54.bitnet hooft@chem.ruu.nl hooft@fys.ruu.nl hooft@cc.ruu.nl