Message-ID: <399BB121.2A01868A@softhome.net>
Date: Thu, 17 Aug 2000 11:32:17 +0200
From: Laurynas Biveinis <lauras AT softhome DOT net>
X-Mailer: Mozilla 4.74 [en] (Win98; U)
X-Accept-Language: lt,en
MIME-Version: 1.0
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
CC: djgpp-workers AT delorie DOT com
Subject: Re: Symlinks & fstat
References: <399B00B3 DOT 1AA2C454 AT softhome DOT net> <1858-Thu17Aug2000010725+0300-eliz AT is DOT elta DOT co DOT il> <399B954E DOT 4BC4ECC AT softhome DOT net> <2950-Thu17Aug2000115242+0300-eliz AT is DOT elta DOT co DOT il>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com

Eli Zaretskii wrote:
> What you described above should be a primitive which can be hooked.

It was something that has been overlooked. (If you have my old symlink
patches, you might check them, if they contain patches for fstat(). I
doubt that). And FSEXT just made this little flaw very clear.

> > Now if I provide
> > int __internal_readlink(const char * path, int fhandle, char * buf, int max);
> 
> Assuming that we won't get the user into trouble by requesting to have
> the same hook that is called both from __FSEXT_call_open_handlers and
> by a handle-related code, this is okay, I think.

I will document that user should call __internal_readlink() from its fstat
handler, if he/she does not support his/her own symlink format.

> > BTW, if file was opened for write only, __internal_readlink() would
> > fail here
> 
> This should be documented.  

Yes.

Laurynas