bug-gnulib
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: coreutils-6.2: various runtime problems on Darwin-8.7.0 HFS+ (includ


From: Jim Meyering
Subject: Re: coreutils-6.2: various runtime problems on Darwin-8.7.0 HFS+ (including attachment this time)
Date: Tue, 26 Sep 2006 21:17:05 +0200

mwoehlke <address@hidden> wrote:
...
> Did you try with 5.97? I haven't tried 6.2 yet (will try to do that

No, since rm in that release is so different, and since
I don't plan to make a 5.98 release.

> soon)... Um, actually, I am probably not going to try until someone can
> tell me how the *&($ to test the exit code of 'make check' after I've
> piped it through 'tee' :-) - tips would be most welcome!
>
> Easiest test is on coreutils itself...
>
> $ uname -srvmpio
> Darwin 7.9.0 Darwin Kernel Version 7.9.0: Wed Mar 30 20:11:17 PST 2005;
> root:xnu/xnu-517.12.7.obj~1/RELEASE_PPC  Power Macintosh powerpc
> RackMac3,1 Darwin
> $ rm --version
> rm (GNU coreutils) 5.97
> Copyright (C) 2006 Free Software Foundation, Inc.
> This is free software.  You may redistribute copies of it under the terms of
> the GNU General Public License <http://www.gnu.org/licenses/gpl.html>.
> There is NO WARRANTY, to the extent permitted by law.
>
> Written by Paul Rubin, David MacKenzie, Richard Stallman, and Jim Meyering.
> $ ls coreutils-5.97
> ls: coreutils-5.97: No such file or directory
> $ tar xzf coreutils-5.97.tar.gz
> $ find coreutils-5.97/ | wc -l
> 2788
> $ rm -rf coreutils-5.97
> rm: cannot remove directory `coreutils-5.97/m4': Directory not empty
> rm: cannot remove directory `coreutils-5.97/lib': Di0rectory not empty
> $ find coreutils-5.97/ | wc -l
> 173
> $ rm -rf coreutils-5.97
> $ find coreutils-5.97/ | wc -l
> find: coreutils-5.97/: No such file or directory
> 0
> $

Thanks.  That's probably all I needed.
>From those numbers and the number of files in lib and m4, I deduce
that your system's readdir is mistakenly returning NULL when there
are around 190 entries in a directory.

Would you please try this, to confirm the theory:

#!/bin/bash
for i in $(seq 150 200); do
  echo i=$i
  mkdir b; (cd b; mkdir $(seq $i) ); Rm -r b || { echo bug at i=$i; break; }
done

That should find the minimum number of entries for which rm fails.

Once you find that number, say 190, change the enum in your
copy of coreutils' src/remove.c that looks like this:

    /* This is the maximum number of consecutive readdir/unlink calls that
       can be made (with no intervening rewinddir or closedir/opendir)
       before triggering a bug that makes readdir return NULL even though
       some directory entries have not been processed.  The bug afflicts
       SunOS's readdir when applied to ufs file systems and Darwin 6.5's
       (and OSX v.10.3.8's) HFS+.  This maximum is conservative in that
       demonstrating the problem seems to require a directory containing
       at least 254 deletable entries (which doesn't count . and ..), so
       we could conceivably increase the maximum value to 254.  */
    enum
      {
        CONSECUTIVE_READDIR_UNLINK_THRESHOLD = 200
      };

to use the number that is one smaller than the one you found:

        CONSECUTIVE_READDIR_UNLINK_THRESHOLD = 189

Then recompile, and "rm -r" should work.

Please let us know.




reply via email to

[Prev in Thread] Current Thread [Next in Thread]