emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Philippe Waroquiers]

In an hpterm (or in an xterm), emacs core dumps if DISPLAY is not set.
When DISPLAY is set, emacs startups properly.

(gdb) bt
#0  0xc020cc68 in kill () from /usr/lib/libc.2
#1  0x0013e6e8 in fatal_error_signal ()
#2  <signal handler called>
#3  0x00276d40 in _free_internal ()
#4  0x002778b0 in free ()
#5  0x001cef50 in emacs_blocked_free ()
#6  0x0027789c in free ()
#7  0x001ce564 in xfree ()
#8  0x000e4f00 in tty_default_color_capabilities ()
#9  0x000e5cac in term_init ()
#10 0x00040a84 in init_display ()
#11 0x0014042c in main ()
(gdb)

(note: bt full gives the same results, but with only some additional
error messages telling 'No symbol table info available.'.
xbacktrace command is not found).


In GNU Emacs 22.1.1 (hppa1.1-hp-hpux11.11, Motif Version 2.1.0)
 of 2007-08-04 on bobcat
Windowing system distributor `Hewlett-Packard Company', version 11.0.600000
configured using `configure  'hppa1.1-hp-hpux11.11' '--prefix=/cm/ot/TOOL/GNU!12.43/build_G!67.IP.11/generated' '--exec-prefix=/cm/ot/TOOL/GNU!12.43/build_G!67.IP.11/generated/libexec/emacs-22.1' '--mandir=/cm/ot/TOOL/GNU!12.43/build_G!67.IP.11/generated/man' '--with-gcc=no' '--with-xpm' '--with-x11r6' '--with-x-toolkit=motif' '--x-libraries=/usr/lib/X11R6:/usr/contrib/X11R6/lib' 'build_alias=hppa1.1-hp-hpux11.11' 'host_alias=hppa1.1-hp-hpux11.11' 'target_alias=hppa1.1-hp-hpux11.11' 'CFLAGS=-g -Ae +O0 +DA1.1' 'LDFLAGS= -L/cm/ot/TOOL/GNU!12.43/build_G!67.IP.11/generated/work/emacs-22.1/xpm-3.4k/lib' 'CPPFLAGS=-I/cm/ot/TOOL/GNU!12.43/build_G!67.IP.11/generated/work/emacs-22.1/xpm-3.4k ''

Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: C
  locale-coding-system: nil
  default-enable-multibyte-characters: nil

Major mode: Lisp Interaction

Minor modes in effect:
  iswitchb-mode: t
  show-paren-mode: t
  which-function-mode: t
  tooltip-mode: t
  mouse-wheel-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  unify-8859-on-encoding-mode: t
  utf-translate-cjk-mode: t
  auto-compression-mode: t
  column-number-mode: t
  line-number-mode: t

Recent input:
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo> 
<help-echo> <menu-bar> <help-menu> <report-emacs-b
ug>

Recent messages:
Loading gpc...done
Loading gnuserv-compat...
Loading cl-macs...done
Loading gnuserv-compat...done
isearch-forwardisearch-backwardisearch-forward-regexpisearch-backward-regexpquery-replacequery-replace-regexpreplace-regexpreplace-string
Loading gnat-fix-error...done
Loading version_emacs_startup...done
Loading /cm/ot/TOOL/GNU!12.43/build_G!67.IP.11/generated/share/emacs/22.1/site-lisp/emacs_startup.el (source)...done
Loading easy-mmode...done
Loading emacsbug...done


-- 
Philippe WAROQUIERS                  Eurocontrol - Central Flow Management Unit
philippe.waroquiers@eurocontrol.int  Rue de la fusee, 96
Tel: +32 2 729 97 35                 1130 Brussels
Fax: +32 2 729 90 22                 Belgium
____

This message and any files transmitted with it are legally privileged and intended for the sole use of the individual(s) or entity to whom they are addressed. If you are not the intended recipient, please notify the sender by reply and delete the message and any attachments from your system. Any unauthorised use or disclosure of the content of this message is strictly prohibited and may be unlawful.

Nothing in this e-mail message amounts to a contractual or legal commitment on the part of EUROCONTROL, unless it is confirmed by appropriately signed hard copy.

Any views expressed in this message are those of the sender.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

Please try to debug it and find the cause.
First I suggest recompiling using -g.

    (note: bt full gives the same results, but with only some additional
    error messages telling 'No symbol table info available.'.
    xbacktrace command is not found).

That means the .gdbinit file wasn't found or didn't work.
Please try to figure out why.

RE: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[WAROQUIERS Philippe]

>Please try to debug it and find the cause.
>First I suggest recompiling using -g.
>That means the .gdbinit file wasn't found or didn't work.
>Please try to figure out why.

.gdbinit was not found because I was not in the good directory.

After recompiling with debug and using gdb, I was able to see
what is happening, and tested a fix.

The problem is that when emacs starts up without DISPLAY, it calls
the function tty_default_color_capabilities.
For a reason not clear to me (my knowledge of C is *very* rusty),
the static local variables in tty_default_color_capabilities are
not properly initialized to NULL or 0 at startup.

I first tried to just put a  " = NULL;" and a "= 0;"
but that did not solve the problem.

At the end, the problem was solved by moving the static variables
just before the function tty_default_color_capabilities 
(and also initializing them to NULL/0).

In other words, the below fixes the problem:

/* Save or restore the default color-related capabilities of this
   terminal.  */
  static char
    *default_orig_pair, *default_set_foreground, *default_set_background
= NULL; /* need init ??? */
  static int default_max_colors, default_max_pairs,
default_no_color_video = 0; /* need init ??? */
static void
tty_default_color_capabilities (save)
     int save;
{

  if (save)
    {
      if (default_orig_pair)


As I do not understand too much the problem, if you need any other
investigation, I can
always see what I can do.

Thanks for your help ...



____

This message and any files transmitted with it are legally privileged and intended for the sole use of the individual(s) or entity to whom they are addressed. If you are not the intended recipient, please notify the sender by reply and delete the message and any attachments from your system. Any unauthorised use or disclosure of the content of this message is strictly prohibited and may be unlawful.

Nothing in this e-mail message amounts to a contractual or legal commitment on the part of EUROCONTROL, unless it is confirmed by appropriately signed hard copy.

Any views expressed in this message are those of the sender.

RE: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[WAROQUIERS Philippe]

I have done a test to see if emacs compiles and runs properly
when commenting the
#define static
line in s/hpux.h

I tested with the below versions of hp-ux and c compiler.
The emacs was properly configured/compiled with the below change
in hpux.h.
I started the resulting emacs on an X display and on a tty.
With this basic test, everything seems ok.
Conclusion: hpux 11.11 and the below version of the compiler
does not have a problem with the "static".

Hoping this helps ...


/* #define static ??? test if still needed for hp-ux 11.11 */
/* tested with the below
wao@steer: cc -V bidule.c
cpp.ansi: HP92453-01 B.11.X.35175-35176.GP HP C Preprocessor (ANSI)
ccom: HP92453-01 B.11.X.35098-35101.GP HP C Compiler
cc: "bidule.c", line 1: warning 501: Empty source file.
/usr/ccs/bin/ld: 92453-07 linker linker ld B.11.58 060929
/usr/ccs/bin/ld: Unsatisfied symbols:
   main (Not referenced yet! Probably due to -u option)
wao@steer: uname -a
HP-UX steer B.11.11 U 9000/800 2316717407 unlimited-user license
wao@steer:
*/
____

This message and any files transmitted with it are legally privileged and intended for the sole use of the individual(s) or entity to whom they are addressed. If you are not the intended recipient, please notify the sender by reply and delete the message and any attachments from your system. Any unauthorised use or disclosure of the content of this message is strictly prohibited and may be unlawful.

Nothing in this e-mail message amounts to a contractual or legal commitment on the part of EUROCONTROL, unless it is confirmed by appropriately signed hard copy.

Any views expressed in this message are those of the sender.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

That is useful information -- thanks.  To see what conclusions we can
draw from it, I need some more info:

Is HPUX 11.11 the current version?

How many years ago was HPUX 11.11
released?

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Richard Stallman wrote:

> Is HPUX 11.11 the current version?
>
> How many years ago was HPUX 11.11 released?

According to the internet, it was released in 2000 and is 3 or 4
versions old.

I looked for non-file scope statics in src/*.c. I think there are only
a handful that aren't definitely in Windows, Mac, or VMS specific code:

grep -E '^[ \t]+static ' *.c | grep -Ev '^(w32|msdos|mac)'

alloca.c      find_stack_direction
fringe.c      init_fringe_bitmap
gtkutil.c     menu_grab_callback, xg_separator_p
image.c       our_memory_fill_input_buffer, gif_load
xselect.c     x_reply_selection_request


Not relevant:

emacs.c (commment), getloadavg.c (vms), regex.c (not emacs),
strftime.c (indented file scope), sunfns.c, sysdep.c (vms), unexsunos4.c

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

    According to the internet, it was released in 2000 and is 3 or 4
    versions old.

That's good.  I think it is ok to desupport versions older than that.

So what about USG systems.  There is no problem with USG 5.4.
Are the USG 5.3 derivatives totally gone now?

However, given this information

    I looked for non-file scope statics in src/*.c. I think there are only
    a handful that aren't definitely in Windows, Mac, or VMS specific code:

    alloca.c      find_stack_direction
    fringe.c      init_fringe_bitmap
    gtkutil.c     menu_grab_callback, xg_separator_p
    image.c       our_memory_fill_input_buffer, gif_load
    xselect.c     x_reply_selection_request

I think we may as well fix the last four, unless that is hard.  Can
you fix them?

alloca.c is not an issue.  It checks for #define static and gives an
error.  It is also surely obsolete since 15 years ago; let's delete it
and its whole support mechanism in Emacs 23.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Richard Stallman wrote:

> So what about USG systems.  There is no problem with USG 5.4.
> Are the USG 5.3 derivatives totally gone now?

No idea, sorry.

>     fringe.c      init_fringe_bitmap
>     gtkutil.c     menu_grab_callback, xg_separator_p
>     image.c       our_memory_fill_input_buffer, gif_load
>     xselect.c     x_reply_selection_request
>
> I think we may as well fix the last four, unless that is hard.  Can
> you fix them?

I guess so. Do I just move all static variables to file scope,
irrespective of type? Does it matter whether they are explicitly
initialized, or if the values never change?

> alloca.c is not an issue.  It checks for #define static and gives an
> error.  It is also surely obsolete since 15 years ago; let's delete it
> and its whole support mechanism in Emacs 23.

Remove src/alloca.c from the trunk, you mean?

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

    I guess so. Do I just move all static variables to file scope,
    irrespective of type? Does it matter whether they are explicitly
    initialized, or if the values never change?

If the values never change, it is ok to have an initializer.
Variables with initializer may be put in pure space when Emacs is
dumped.

If the value can change, there must be no initializer.  So initialize
the variable in init_FILENAME or some such function that is called at
a suitable time.

It is a good idea to rename the variable
to a name that relates to the function that uses it.
Don't keep a short name like `cnt'.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Richard Stallman wrote:

> If the values never change, it is ok to have an initializer.
> Variables with initializer may be put in pure space when Emacs is
> dumped.

I really meant: if they are explicitly initialized and never change,
is there any need to move them to file scope? In that case, they don't
actually have to _be_ static, AFAICS, so it's no problem if the HPUX
stuff defines "static" as a no-op.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

    I really meant: if they are explicitly initialized and never change,
    is there any need to move them to file scope? In that case, they don't
    actually have to _be_ static, AFAICS, so it's no problem if the HPUX
    stuff defines "static" as a no-op.

If they don't need to be static, it is ok to delete `static'.
But please don't leave them file-scope static.
That is asking for confusion.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Richard Stallman wrote:

>     I really meant: if they are explicitly initialized and never change,
>     is there any need to move them to file scope? In that case, they don't
>     actually have to _be_ static, AFAICS, so it's no problem if the HPUX
>     stuff defines "static" as a no-op.
>
> If they don't need to be static, it is ok to delete `static'.
> But please don't leave them file-scope static.

I'm confused. Did you mean to say "function-scope" there?

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

    > If they don't need to be static, it is ok to delete `static'.
    > But please don't leave them file-scope static.

    I'm confused. Did you mean to say "function-scope" there?

Yes, I did.  Sorry.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Of the static uses I identified, five remain. I don't understand why
any of them need to be static, and I'm reluctant to change things I
don't understand. Which is better: leave them in function scope and
delete the `static', or move them to file scope?


fringe.c (init_fringe_bitmap) <swap_nibble>
gtkutil.c (xg_separator_p) <separator_names>
image.c  (our_memory_fill_input_buffer) <buffer>
         (gif_load) <interlace_start, interlace_increment>

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Jan Djärv]

Glenn Morris skrev:
> Of the static uses I identified, five remain. I don't understand why
> any of them need to be static, and I'm reluctant to change things I
> don't understand. Which is better: leave them in function scope and
> delete the `static', or move them to file scope?
> 

Move them to file scope.

On bug-gnu-emacs it was suggested in "emacs-22.1 with GTK problems (with 
patches)" that this in s/hpux.h:

/* USG systems tend to put everything declared static
    into the initialized data area, which becomes pure after dumping Emacs.
    Foil this.  Emacs carefully avoids static vars inside functions.  */

#define static

be removed, as it seems to interfere with threading.  How does that relate to 
the work you are doing?

	Jan D.

> 
> fringe.c (init_fringe_bitmap) <swap_nibble>
> gtkutil.c (xg_separator_p) <separator_names>
> image.c  (our_memory_fill_input_buffer) <buffer>
>          (gif_load) <interlace_start, interlace_increment>
> 

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Jan Djärv wrote:

> Move them to file scope.

Will do, thanks.

> On bug-gnu-emacs it was suggested in "emacs-22.1 with GTK problems (with 
> patches)" that this in s/hpux.h:
[...]
> #define static
>
> be removed, as it seems to interfere with threading.

I already removed this in the trunk, as this thread seems to show it
is not needed in recent versions of hpux.h. I left it alone in the
release branch, but if it is actively causing problems, it should
probably be removed there too.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

    I already removed this in the trunk, as this thread seems to show it
    is not needed in recent versions of hpux.h. I left it alone in the
    release branch, but if it is actively causing problems, it should
    probably be removed there too.

Please do remove that #define from hpux.h in Emacs 22.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Richard Stallman wrote:

> Please do remove that #define from hpux.h in Emacs 22.

Done. I believe this issue is now dealt with.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

    fringe.c (init_fringe_bitmap) <swap_nibble>

That should be file-scope.

    image.c  (our_memory_fill_input_buffer) <buffer>

That should be file-scope.

	     (gif_load) <interlace_start, interlace_increment>

They should be file-scope.

I did not check the gtk file.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Richard Stallman]

    > alloca.c is not an issue.  It checks for #define static and gives an
    > error.  It is also surely obsolete since 15 years ago; let's delete it
    > and its whole support mechanism in Emacs 23.

    Remove src/alloca.c from the trunk, you mean?

Delete that file, and everything in Makefile.in and configure and m/
and s/ files that supports using it.

Re: emacs 22.1 on hp-ux 11.11 core dumps when DISPLAY not set

[Glenn Morris]

Richard Stallman wrote:

>     Remove src/alloca.c from the trunk, you mean?
>
> Delete that file, and everything in Makefile.in and configure and m/
> and s/ files that supports using it.

Done.