This is the mail archive of the newlib@sourceware.org mailing list for the newlib project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: readdir and d_ino set to 0

From: Chris Johns <chrisj at rtems dot org>
To: Eric Blake <ebb9 at byu dot net>
Cc: Eric Norum <norume at aps dot anl dot gov>, newlib at sourceware dot org
Date: Fri, 04 Sep 2009 15:38:02 +1000
Subject: Re: readdir and d_ino set to 0
References: <4AA05AC5.4040603@rtems.org> <F62D0A9C-0837-4349-A2B8-7D57C42D6D00@aps.anl.gov> <4AA06CC3.5090307@byu.net>

Eric Blake wrote:

According to Eric Norum on 9/3/2009 7:00 PM:

If I recall my Unix V7 code correctly, I think that inode 0 was reserved
for holding a list of bad blocks.  Of course if your disk had a bad
block that was *in* the inode table you were pretty much out of luck.....


Also, modern-day BSD union mounts read all the entries from several layers
of directories, sort them, then discard duplicate names by changing d_ino
to 0.  In other words, there are still live systems where readdir returns
such dead entries.

Many thanks. I will keep clear of 0 as an ino number.

Regards
Chris

References:
- readdir and d_ino set to 0
  - From: Chris Johns
- Re: readdir and d_ino set to 0
  - From: Eric Norum
- Re: readdir and d_ino set to 0
  - From: Eric Blake

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]