Sender: rich AT phekda DOT freeserve DOT co DOT uk
Message-ID: <3C05577B.590A4890@phekda.freeserve.co.uk>
Date: Wed, 28 Nov 2001 21:30:35 +0000
From: Richard Dawe <rich AT phekda DOT freeserve DOT co DOT uk>
X-Mailer: Mozilla 4.77 [en] (X11; U; Linux 2.2.19 i586)
X-Accept-Language: de,fr
MIME-Version: 1.0
To: djgpp-workers AT delorie DOT com
Subject: Re: RESEND: Patch to computer st_blksize in struct stat
References: <Pine DOT SUN DOT 3 DOT 91 DOT 1011128103005 DOT 659H-100000 AT is>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com

Hello.

Eli Zaretskii wrote:
> 
> On Tue, 27 Nov 2001, Richard Dawe wrote:
>
> > So should I call statfs() by default for all drives?
> 
> I'd prefer not to do that for floppies and networked drives, even
> though the above data suggests the slow-down is negligible in the
> normal case.  The reason is that floppies could have damaged FATs that
> could fail stat where it previously would succeed, and networked
> drives could incur delays due to network problems.  Since the block
> size on a floppy is almost always known in advance, and for networked
> drives the size is of no practical use, it doesn't seem like a risk
> worth taking.

Righty-ho, I'll put the code back in for 512B blocks for floppies and for
4KB blocks for remote drives.
 
> Oh, btw, statfs is non-Posix, while stat and lstat are.  So we need to
> make statfs a stub.  Did we do that?

Nope, but I'll fix that. 8)
 
> > However, it seems to me that we need someone with access to a
> > machine with plain ol' DOS to test it, to see what the overhead is
> > without caching.
> 
> If you send me a source of a test program, I can run it on DOS 5.0.

Since we've decided to hardcode floppies and remote drives, there's no
need to do this.

> > Eli Zaretskii wrote:
> > 
> > > I think a constant default for networked drives is okay, since it
> > > isn't meaningful in many cases, anyway.  However, 32K seems a bit
> > > large; how about 4KB instead?
> >
> > I chose 32K, because that's what I found on my system.
> 
> How did you find that?  What methods of reporting the block size of a
> remote drive did you use?

statfs() returned 32K on my network drives. Is this not reliable on
network drives?

It may not be relevant, but I mapped a share from my Linux box, which is
running kernel 2.2.19 with Samba 2.0.10. The block size on the ext2
partition containing the share is 4K.

> > Oops, actually xstat.h includes sys/stat.h. But: _STAT_* are defined
> > in sys/stat.h and xstat.c; _STFAIL_* are defined in sys/stat.h and
> > xstat.h. Since the definitions are the same, there are no warnings.
> 
> I think we should remove the duplicate definitions.  They are probably
> a left-over from the initial coding of these functions.

OK, I'll clean those up.

Bye, Rich =]

-- 
Richard Dawe
http://www.phekda.freeserve.co.uk/richdawe/