Corinna Vinschen corinna-cygwin@cygwin.com
Thu Mar 4 10:23:00 GMT 2010

On Mar  3 02:45, Dave Korn wrote:
> On 02/03/2010 21:35, Corinna Vinschen wrote:
> > On Mar  2 13:32, David Rothenberger wrote:
> >> On 3/2/2010 12:52 PM, Corinna Vinschen wrote:
> >>> So, maybe we should just add an cygwin_internal hook for now which allows
> >>> to disable .exe appending?
> >> Just a quick comment here. svn uses libapr1 to do the renaming, so I'd
> >> probably use the cygwin_internal hook in libapr1 instead of just svn,
> >> unless there were some major objections.
> > 
> > That's your call, but so far the hook is an unlaid egg...
>   If it helps this discussion any, I'd be happy with any patch that teaches
> strip how to handle the .exe suffix on cygwin; there's no reason it shouldn't
> have this target-specific knowledge.

That would make sense, I guess.  The whole idea is to check the
original file for the .exe suffix and to add it to the result again:

  $ strip foo

      open("foo") already opens foo.exe
      open (tempfile)
      [strip... strip... strip]
      unlink ("foo") already removes foo.exe
      rename (tempfile, "foo") is the problem.

  Something like this would help:

      bool add_suffix = false;
      if (file_has_no_dll_or_exe_suffix)
	if stat("foo.exe",&st_exe) == 0
	  stat("foo", &st)
	  if (st.st_ino == st_exe.st_ino)
	    add_suffix = true
      open (tempfile)
      [strip... strip... strip]
      unlink ("foo")
      rename (tempfile, add_suffix ? "foo.exe" : "foo")

If that mechanism works in strip, and a similar mechanism is added
to install, I have no problems to remove what Eric calls the second
category of exe magic.

The question left is, shall I remove that exe magic part already
in 1.7.2?

If we could have patched versions of strip and install ASAP, I have
no problem with that, either.


Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

More information about the Cygwin-developers mailing list