From: Rob Landley <landley@webofficenow.com>
To: Pavel Machek <pavel@suse.cz>, Hans Reiser <reiser@namesys.com>
Cc: Jan Harkes <jaharkes@cs.cmu.edu>,
Craig Soules <soules@happyplace.pdl.cmu.edu>,
Andi Kleen <ak@suse.de>,
linux-kernel@vger.kernel.org
Subject: Re: NFS Client patch
Date: Sun, 22 Jul 2001 11:15:07 -0400 [thread overview]
Message-ID: <01072211150706.08013@localhost.localdomain> (raw)
In-Reply-To: <Pine.LNX.3.96L.1010717180713.13980K-100000@happyplace.pdl.cmu.edu> <3B55A12C.F194DAF6@namesys.com> <20010719182405.A35@toy.ucw.cz>
In-Reply-To: <20010719182405.A35@toy.ucw.cz>
On Thursday 19 July 2001 14:24, Pavel Machek wrote:
> Hi!
>
> > > In any case, what happens if the file was renamed or removed between
> > > the 2 readdir calls. A cookie identifying a name that was returned
> > > last, or should be read next is just as volatile as a cookie that
> > > contains an offset into the directory.
> >
> > No, if the file was removed, it still tells you where to start your
> > search. A missing filename is just as good a marker as a present one.
>
> And if new file is created with same name?
> Pavel
The same thing that happens as if a new file was inserted BEFORE your cursor,
in the part of the directory you've already looked at. You ignore it.
The "filename cookie" indicates the LAST file we looked at. We've already
seen it. Therefore, whether it's the same file or not, we don't care. We
just want the next file AFTER that one.
We're doing fairly arbitrary, unlocked reads across volatile data. No
algorithm is going to behave perfectly here. We just want a behavior that is
consistent, guaranteed to complete, and doesn't violate any obvious
constraints about how filesystems should behave (like producing duplicate
entries, which returning two different sets of data with the same filename
would do).
Makes sense to me, anyway...
Rob
next prev parent reply other threads:[~2001-07-23 0:17 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <Pine.LNX.3.96L.1010709131315.16113O-200000@happyplace.pdl.cmu.edu.suse.lists.linux.kernel>
2001-07-09 18:33 ` NFS Client patch Andi Kleen
2001-07-10 13:33 ` Chris Wedgwood
2001-07-10 13:41 ` Andi Kleen
2001-07-10 16:48 ` Craig Soules
2001-07-10 17:06 ` Andi Kleen
2001-07-10 18:04 ` Chris Wedgwood
2001-07-12 20:57 ` Alan Cox
2001-07-13 11:26 ` Chris Wedgwood
2001-07-17 22:02 ` Hans Reiser
2001-07-17 22:14 ` Craig Soules
2001-07-17 22:21 ` Hans Reiser
2001-07-18 13:30 ` Daniel Phillips
2001-07-18 14:46 ` Hans Reiser
2001-07-18 14:00 ` Jan Harkes
2001-07-18 14:46 ` Hans Reiser
2001-07-19 18:24 ` Pavel Machek
2001-07-22 15:15 ` Rob Landley [this message]
2001-07-23 2:02 ` Horst von Brand
2001-07-23 9:57 ` Rob Landley
2001-07-18 13:57 ` Chris Mason
2001-07-19 11:35 ` Trond Myklebust
2001-07-19 18:02 ` Hans Reiser
2001-07-20 8:50 ` Trond Myklebust
2001-07-20 11:30 ` Hans Reiser
2001-07-20 14:07 ` Chris Mason
2001-07-09 17:28 Craig Soules
2001-07-09 18:59 ` Trond Myklebust
2001-07-09 19:45 ` Craig Soules
2001-07-09 19:53 ` Charles Cazabon
2001-07-09 20:05 ` Trond Myklebust
2001-07-09 22:09 ` Craig Soules
2001-07-10 8:22 ` Trond Myklebust
2001-07-10 13:38 ` Chris Wedgwood
2001-07-11 8:14 ` Trond Myklebust
2001-07-09 21:46 ` J. Richard Sladkey
2001-07-10 15:06 ` Craig Soules
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=01072211150706.08013@localhost.localdomain \
--to=landley@webofficenow.com \
--cc=ak@suse.de \
--cc=jaharkes@cs.cmu.edu \
--cc=linux-kernel@vger.kernel.org \
--cc=pavel@suse.cz \
--cc=reiser@namesys.com \
--cc=soules@happyplace.pdl.cmu.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox