Re: --enable-font-backend bug (was: Re: 23.0.60; emacs segfaults when creating X display via emacsclient)

[Dan Nicolaescu <dann@ics.uci.edu>]

Kenichi Handa <handa@ni.aist.go.jp> writes:

  > In article <200802050018.m150IrgE007563@sallyv1.ics.uci.edu>, Dan Nicolaescu <dann@ics.uci.edu> writes:
  > 
  > > Run emacs in a terminal
  > > 
  > > 1) emacs -nw -Q --eval '(server-start)'
  > > 
  > > Create an X frame.
  > > 2) emacsclient -c -d :0.0
  > > 
  > > Delete the frame
  > > 3) C-x 5 0
  > > 
  > > Try and create another X frame.
  > > 4) emacsclient -c -d :0.0
  > > 
  > > leading to the following backtrace:
  > 
  > > I don't get a segfault, but emacs and emacsclient hang.
  > 
  > As I got "Fatal error (6)Aborted", I started the debugging
  > by running the first emacs under gdb.  But I found that I
  > can't suspend it by typing C-z once I did the step 2).
  > Emacs just shows this message: "Suspending an Emacs running
  > under X makes no sense".
  > 
  > This makes the debugging very difficult.  :-(

Why not start the debugger in another terminal and attach to the emacs
process?

  > Is this a multi-tty problem?

The "Suspending an Emacs running under X" message? It is, I posted a
message about that on emacs-devel a week or two ago. But that is
harmless, it should not affect anything here. Emacs just refuses to
suspend an tty frame thinking that it has an X11 frame.

Again, if emacs is not configured with --enable-font-backend everything
works fine.  So it seems that the new font backend is the culprit here.

  > ---
  > Kenichi Handa
  > handa@ni.aist.go.jp
  > 
  > PS.  I had to do this to get this "abort".
  > 
  > 1) emacs -nw -Q --eval '(server-start)'
  > 2) emacsclient -c -d :0.0
  > 3) type C-h h
  > 4) C-x 5 0
  > 5) emacsclient -c -d :0.0

After this step emacs and emacsclien hang completely for me, so I can't
do step 6.

  > 6) type C-h h