Message-ID: <003501c2eec7$e41e2a80$0100a8c0@acp42g>
From: "Andrew Cottrell" <acottrel AT ihug DOT com DOT au>
To: <djgpp-workers AT delorie DOT com>, <rich AT phekda DOT freeserve DOT co DOT uk>
Cc: <djgpp-workers AT delorie DOT com>
References: <10303182107 DOT AA24101 AT clio DOT rice DOT edu> <3E7868E9 DOT 19949F8E AT phekda DOT freeserve DOT co DOT uk> <9003-Wed19Mar2003174940+0200-eliz AT elta DOT co DOT il> <3E78AA1D DOT 23720139 AT phekda DOT freeserve DOT co DOT uk> <3405-Wed19Mar2003223319+0200-eliz AT elta DOT co DOT il> <3E78D73C DOT 3B41ECF AT phekda DOT freeserve DOT co DOT uk> <7263-Thu20Mar2003061743+0200-eliz AT elta DOT co DOT il>
Subject: Re: DJGPP 2.04 release? [Was: Re: nmalloc revisited]
Date: Thu, 20 Mar 2003 21:03:00 +1100
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 6.00.2800.1106
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106
Reply-To: djgpp-workers AT delorie DOT com

> Are you saying that the current code does _not_ invent inodes on
> Windows/XP?  I'd be surprised, I thought we invent inodes for any
> kind of files on all the versions of Windows except 3.1.

If the invent_inode is not passed a name then under Windows 2K/XP the inode
number may change between calls and as such cause failures in directory
tranversial programs. It appears to be okay on Win 9x/DOS, but fails on
2K/XP.

In the following email you can see a patch that I did to fix the cases that
I was aware of, but the solution did not fix all cases from what I can
remember.
http://www.delorie.com/djgpp/mail-archives/browse.cgi?p=djgpp-workers/2003/0
2/08/23:50:22

The main line that was changed was:-
    From:    stat_buf->st_ino = _invent_inode("", dos_ftime,trusted_fsize);
    To:        stat_buf->st_ino = _invent_inode(fixed_filename, dos_ftime,
trusted_fsize);

As you can see _invent_inode passed NULL as the name and now uses the fixed
file name.

The original info on this can be seen in the following email:-
http://www.delorie.com/djgpp/mail-archives/browse.cgi?p=djgpp-workers/2002/0
9/15/03:27:43

Excert from the email:-
The reason for this is that in xstat.c the following piece of code is
executed if no filename is used in the  _invent_inode as is the case in
fstat, but not the case for lstat
The offending line in stat.c whcih is a few lines above the symlink check
for trusted_fsize == 510 :
                 stat_buf->st_ino = _invent_inode("", dos_ftime,
trusted_fsize);

  /* If the passed name is empty, invent a new inode unconditionally.
   * This is for those unfortunate circumstances where we couldn't
   * get a name (e.g., fstat() under Novell).  For these we want at
   * least to ensure that no two calls will get the same inode number.
   * The lossage here is that you get different inodes even if you call
   * twice with the same file.  Sigh...
   */
  if (!*name)
    {
      ino_t retval = inode_count;

      inode_count++;
      return retval;
    }


Hope this helps and does not confuse the issue more.

Regards,
Andrew