Date: Sun, 11 Nov 2001 21:53:16 +0200
From: "Eli Zaretskii" <eliz AT is DOT elta DOT co DOT il>
Sender: halo1 AT zahav DOT net DOT il
To: Richard Dawe <rich AT phekda DOT freeserve DOT co DOT uk>
Message-Id: <4634-Sun11Nov2001215315+0200-eliz@is.elta.co.il>
X-Mailer: emacs 21.1.50 (via feedmail 8 I) and Blat ver 1.8.9
CC: djgpp-workers AT delorie DOT com
In-reply-to: <3BEED119.E912DFAA@phekda.freeserve.co.uk> (message from Richard
	Dawe on Sun, 11 Nov 2001 19:27:21 +0000)
Subject: Re: RESEND: Patch to computer st_blksize in struct stat
References: <Pine DOT SUN DOT 3 DOT 91 DOT 1011111152354 DOT 6498A-100000 AT is> <3BEED119 DOT E912DFAA AT phekda DOT freeserve DOT co DOT uk>
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> Date: Sun, 11 Nov 2001 19:27:21 +0000
> From: Richard Dawe <rich AT phekda DOT freeserve DOT co DOT uk>
> 
> I'll add some text to develop.txi too.

Thanks.

> >   2) I'm a bit worried by the possible slow-down, due to the call to
> >      statfs.  Could you compare the old and the new versions, at least
> >      on hard disks, floppies, and CDs?
> 
> I'll look at that later in the week. A single stat call may be hurt badly,
> but programs that stat files on the same drive a lot shouldn't be hurt
> much, because of the caching mechanism. If it really is a lot slower, we
> could add another flag to _djstat_flags for those people really concerned
> with performance (_STAT_BLKSIZE perhaps).

If the slow-down is real, perhaps it would make sense to see if files
on a hard disk are ``punished'' much less than on floppies and CDs.
If that is indeed so, we could supply some constant values for the
latter two cases, since they are not very important, and the variation
between block sizes in those cases is small (unlike with hard disks).