[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Possible spurious cycle detection with fts
From: |
James Youngman |
Subject: |
Possible spurious cycle detection with fts |
Date: |
Mon, 8 Aug 2005 22:32:05 +0100 |
User-agent: |
Mutt/1.5.9i |
Hello,
I'm still investigating this problem, but I'm seeing some odd
behaviour with fts(). I have a directory layout like this:-
$ find find1 -ls
1299693 4 drwxr-xr-x 3 james users 4096 Aug 8 20:33 find1
1299705 4 drwxr-xr-x 2 james users 4096 Aug 8 20:33 find1/dir1
1299729 0 lrwxrwxrwx 1 james users 4 Aug 8 20:33
find1/symlink_to_dir1 -> dir1
If I invoke fts_open() with the FTS_LOGICAL flag and do not set
FTS_TIGHT_CYCLE_CHECK, then fts_read() returns FTS_DC in
ent->fts_info. However, with the same directory layout FTS_DC is not
set if FTS_TIGHT_CYCLE_CHECK is set.
fts_options Result
FTS_NOSTAT|FTS_LOGICAL|FTS_TIGHT_CYCLE_CHECK OK
FTS_NOSTAT|FTS_LOGICAL Loop (wrongly) detected
FTS_NOSTAT|FTS_PHYSICAL OK
This means that I suspect that there may be a problem with
cycle_check() in this case, as called from fts-cycle.c. I have not
fully investigated the problem yet, so I am not sure if there really
is a problem.
I have noticed in particular that leave_dir() in fts-cycle.c does
nothing if FTS_TIGHT_CYCLE_CHECK is not set. Perhaps that is related
to this problem.
Regards,
James.
- Possible spurious cycle detection with fts,
James Youngman <=
- Re: Possible spurious cycle detection with fts, Jim Meyering, 2005/08/09
- Re: Possible spurious cycle detection with fts, James Youngman, 2005/08/09
- Re: Possible spurious cycle detection with fts, Jim Meyering, 2005/08/10
- Re: Possible spurious cycle detection with fts, Jim Meyering, 2005/08/14
- Re: Possible spurious cycle detection with fts, Eric Blake, 2005/08/23
- Re: Possible spurious cycle detection with fts, Jim Meyering, 2005/08/24
- Re: Possible spurious cycle detection with fts, Paul Eggert, 2005/08/24
- Re: Possible spurious cycle detection with fts, Jim Meyering, 2005/08/24