From: Matthew Wilcox <matthew@wil.cx>
To: Steve Lord <lord@sgi.com>
Cc: Matthew Wilcox <matthew@wil.cx>,
"Halfmann, Klaus" <khalfmann@libra.de>,
Roman Zippel <roman@augan.com>,
linuxppc-dev@lists.linuxppc.org, linux-fsdevel@vger.kernel.org
Subject: Re: Btree directories (Re: Status of HFS+ support)
Date: Tue, 29 Aug 2000 22:25:13 +0100 [thread overview]
Message-ID: <20000829222513.A830@parcelfarce.linux.theplanet.co.uk> (raw)
In-Reply-To: <200008291845.NAA17251@jen.americas.sgi.com>; from lord@sgi.com on Tue, Aug 29, 2000 at 01:45:51PM -0500
On Tue, Aug 29, 2000 at 01:45:51PM -0500, Steve Lord wrote:
> f_pos being 64 bit is one thing, go look at the source code for glibc
> getdents, it is not 64 bit friendly, actually it is only 31 bit friendly,
> if you return anything bigger than a signed integer it gets confused and
> can skip entries - or go into an infinite loop.
hmm.. glibc 2.2 looks better from this PoV.. but i haven't tested it,
is the bug gone?
> > There's an additional complication of rewinddir/seekdir/telldir, but
> > let's get onto that later.
>
> Which glibc getdents uses because it's dirent is different in size from
> the kernel's dirent, it uses a heuristic to decide how much data to ask
> the kernel for. If it guesses wrong and gets more data back from the
> kernel than will fit in the user's buffer it seeks backwards again to
> reposition for the next call.
Exactly. So, as a minimum, we have to be able to support seeking to one
of the keys in the B-tree. It'd be nice if the search algorithm coped
with searching for an entry which is between two values in the B-tree
and returned the lower one.
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
next prev parent reply other threads:[~2000-08-29 21:25 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2000-08-29 16:40 Status of HFS+ support (was hfs support for blocksize != 512) Halfmann, Klaus
2000-08-29 17:18 ` Btree directories (Re: Status of HFS+ support) Matthew Wilcox
2000-08-29 18:45 ` Steve Lord
2000-08-29 21:25 ` Matthew Wilcox [this message]
2000-08-30 2:22 ` flar
2000-08-30 8:35 ` Andi Kleen
2000-08-30 14:25 ` Chris Mason
2000-08-29 18:16 ` Status of HFS+ support (was hfs support for blocksize != 512) Roman Zippel
-- strict thread matches above, loose matches on Subject: below --
2000-08-30 12:06 Btree directories (Re: Status of HFS+ support) Steve Lord
2000-08-30 15:07 ` Andi Kleen
2000-08-30 16:07 ` flar
2000-08-30 20:49 ` David A. Gatwood
2000-08-30 20:45 ` Steve Lord
2000-08-30 21:05 ` Alexander Viro
2000-08-30 21:51 ` David A. Gatwood
2000-08-30 21:49 ` Alexander Viro
2000-08-30 22:13 ` David A. Gatwood
2000-08-30 22:17 ` Alexander Viro
2000-08-30 23:11 ` David A. Gatwood
2000-08-31 0:10 ` Alexander Viro
2000-08-31 5:49 ` flar
2000-08-31 10:33 ` Alexander Viro
2000-08-31 17:48 ` flar
2000-08-31 19:54 ` Alexander Viro
2000-08-31 22:09 ` flar
2000-09-01 2:40 ` Alexander Viro
2000-09-01 3:52 ` flar
2000-09-02 4:04 ` Tony Mantler
2000-09-01 17:05 Halfmann, Klaus
2000-09-01 17:40 ` flar
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=20000829222513.A830@parcelfarce.linux.theplanet.co.uk \
--to=matthew@wil.cx \
--cc=khalfmann@libra.de \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linuxppc-dev@lists.linuxppc.org \
--cc=lord@sgi.com \
--cc=roman@augan.com \
/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;
as well as URLs for NNTP newsgroup(s).