[Gnash-commit] [bug #27981] fscommand:quit ineffective if in first frame

[Nutchanon Wetchasit]

Follow-up Comment #2, bug #27981 (project gnash):

This is still reproducible under current Gnash git version.
(`set ignoreFSCommand` statement in `~/.gnashrc` must be set to `false`)

I have created few variants of dummy Flash files (Flash 7) that tries to run
`fscommand("quit")` on the first frame, using libming's makeswf and MTASC:

* Single-frame SWF with `fscommand("quit")` code on its timeline.
* Two-frame SWF with `fscommand("quit")` code on its first frame of timeline.
* Twenty-four frame SWF with `fscommand("quit")` code on its first frame of
timeline.
* Single-frame SWF with `fscommand("quit")` code in MTASC's `main()`.

and played them under multiple variants of Gnash (and also Flash Player):

* `gtk-gnash`
** Single frame case: *assertion errors appeared*, ActionScript executed to
the end of frame, *then freezes*.
** Two frame case: *assertion errors appeared*, ActionScript executed to the
end of first frame, *then freezes*.
** Twenty-four frame case: *assertion errors appeared*, ActionScript executed
to the end of first frame, *then freezes*.
** MTASC case: *assertion errors appeared*, ActionScript executed to the end
of `main()`, *then freezes*.
* `qt4-gnash`
** Single frame case: ActionScript executed to the end of frame, and
terminated.
** Two frame case: ActionScript executed to the end of first frame, and
terminated.
** Twenty-four frame case: ActionScript executed to the end of first frame,
and terminated.
** MTASC case: ActionScript executed to the end of `main()`, and terminated.
* `sdl-gnash`
** Single frame case: ActionScript terminated at `fscommand()` call.
** Two frame case: ActionScript terminated at `fscommand()` call.
** Twenty-four frame case: ActionScript terminated at `fscommand()` call.
** MTASC case: ActionScript terminated at `fscommand()` call.
* `fb-gnash`
** Single frame case: ActionScript terminated at `fscommand()` call.
** Two frame case: ActionScript terminated at `fscommand()` call.
** Twenty-four frame case: ActionScript terminated at `fscommand()` call.
** MTASC case: ActionScript terminated at `fscommand()` call.
* `dump-gnash`
** Single frame case: ActionScript executed to the end of first frame, and
terminated
** Two frame case: ActionScript executed to the end of first frame, and
terminated.
** Twenty-four frame case: ActionScript executed to the end of first frame,
and terminated.
** MTASC case: ActionScript executed to the end of `main()`, and terminated.
* Flash Player:
** Single frame case: ActionScript executed to the end of frame, and
terminated.
** Two frame case: ActionScript executed to the end of _second_ frame, and
terminated.
** Twenty-four frame case: ActionScript executed to the end of _second_ frame,
and terminated.
** MTASC case: ActionScript executed to the end of `main()`, and terminated.

>From the result and GDB backtrace, this issue seems to be *specific to*
*`gtk-gnash`* and is very similar to bug #46308 (`gtk-gnash -1` freezes on
single-frame SWF) even the exit is triggered by a different reason.

Debug log (verbosity 2, with action error log) of all SWF-Gnash variant
combination, trace log from Flash player, GDB backtrace from `gtk-gnash`
(collected from SIGQUIT core dump) are attached as `firstframexit.zip`
together with SWFs and their source files.

Gnash: 0.8.11dev (git 95763ce 8-Dec-2015)
System: Debian GNU/Linux 7.0 Wheezy i386

Flash Player: 10.1.82.76 Standalone
System: Microsoft Windows XP Professional SP3

Note #1: `fltk-gnash` is not tested as Gnash's configure script doesn't
correctly detect FLTK 1.1/1.3 header in my configuration.

Note #2: Adobe Flash Player doesn't seem to be strict on exit timing,
as it let another frame played after `fscommand()` call before terminating.

(file #35727)
    _______________________________________________________

Additional Item Attachment:

File name: firstframeexit.zip             Size:58 KB


    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?27981>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/