Re: [RFA] Patch to imported readline to allow DJGPP build

Re: [RFA] Patch to imported readline to allow DJGPP build

[Michael Elizabeth Chastain]

Eli Zaretskii writes:
eli> Btw, I don't understand what is going on with Readline: I looked at
eli> its official home page (to figure out how to submit the patches below
eli> to the Readline maintainer) and found there an old version 4.3 of
eli> Readline where the DJGPP-specific code is ripped out.  The version
eli> that we have in the GDB repository seems a newer one, but it also has
eli> the DJGPP-specific code in it.  Can someone please tell me what is the
eli> story behind this?  TIA

Our readline has some changes since the last import, which was
2002-12-06 by Elena Z.  Check ChangeLog.gdb in the readline directory.
But I don't see any mention of "add djgpp code" in ChangeLog.gdb so
I am baffled too.

Michael C

Re: [RFA] Patch to imported readline to allow DJGPP build

[Eli Zaretskii]

> Date: Mon, 29 Dec 2003 14:56:49 -0500 (EST)
> From: mec.gnu@mindspring.com (Michael Elizabeth Chastain)
> 
> Our readline has some changes since the last import, which was
> 2002-12-06 by Elena Z.  Check ChangeLog.gdb in the readline directory.
> But I don't see any mention of "add djgpp code" in ChangeLog.gdb so
> I am baffled too.

Thanks for looking into this.

Elena, any insights?

[RFA] Patch to imported readline to allow DJGPP build

[Eli Zaretskii]

The current imported readline is broken for DJGPP.  The following
patch is needed to allow GDB to build the DJGPP port.  Okay to commit?

Btw, I don't understand what is going on with Readline: I looked at
its official home page (to figure out how to submit the patches below
to the Readline maintainer) and found there an old version 4.3 of
Readline where the DJGPP-specific code is ripped out.  The version
that we have in the GDB repository seems a newer one, but it also has
the DJGPP-specific code in it.  Can someone please tell me what is the
story behind this?  TIA

2003-12-28  Eli Zaretskii  <eliz@elta.co.il>

	* readline.c (rl_save_state, rl_restore_state): Support systems
	that don't define SIGWINCH.

	* terminal.c (_rl_get_screen_size) [__DJGPP__]: Compute the
	screen width and height using console I/O.
	(_rl_init_terminal_io) [__MSDOS__]: Zero out all the _rl_term_*
	variables.  Convert to _rl_* naming scheme.
	(_rl_set_cursor) [__MSDOS__]: Ifdef away this function.

	* display.c (_rl_move_vert) [__MSDOS__]: Don't use undeclared
	variable `l'.  Use `delta' instead recomputing its value anew.
	Assign -delta to i instead of the other way around.

--- readline/readline.c~0	2002-12-08 22:31:36.000000000 +0000
+++ readline/readline.c	2003-12-28 10:33:08.000000000 +0000
@@ -933,7 +933,9 @@ rl_save_state (sp)
   sp->macro = rl_executing_macro;
 
   sp->catchsigs = rl_catch_signals;
+#if defined (SIGWINCH)
   sp->catchsigwinch = rl_catch_sigwinch;
+#endif
 
   return (0);
 }
@@ -967,7 +969,9 @@ rl_restore_state (sp)
   rl_executing_macro = sp->macro;
 
   rl_catch_signals = sp->catchsigs;
+#if defined (SIGWINCH)
   rl_catch_sigwinch = sp->catchsigwinch;
+#endif
 
   return (0);
 }
--- readline/terminal.c~0	2002-12-08 22:31:36.000000000 +0000
+++ readline/terminal.c	2003-12-28 10:34:46.000000000 +0000
@@ -220,7 +220,10 @@ _rl_get_screen_size (tty, ignore_env)
       if (ignore_env == 0 && (ss = sh_get_env_value ("COLUMNS")))
 	_rl_screenwidth = atoi (ss);
 
-#if !defined (__DJGPP__)
+#if defined (__DJGPP__)
+      if (_rl_screenwidth <= 0)
+	_rl_screenwidth = ScreenCols ();
+#else
       if (_rl_screenwidth <= 0 && term_string_buffer)
 	_rl_screenwidth = tgetnum ("co");
 #endif
@@ -233,7 +236,10 @@ _rl_get_screen_size (tty, ignore_env)
       if (ignore_env == 0 && (ss = sh_get_env_value ("LINES")))
 	_rl_screenheight = atoi (ss);
 
-#if !defined (__DJGPP__)
+#if defined (__DJGPP__)
+      if (_rl_screenheight <= 0)
+	_rl_screenheight = ScreenRows ();
+#else
       if (_rl_screenheight <= 0 && term_string_buffer)
 	_rl_screenheight = tgetnum ("li");
 #endif
@@ -382,12 +388,19 @@ _rl_init_terminal_io (terminal_name)
     term = "dumb";
 
 #ifdef __MSDOS__
-  term_im = term_ei = term_ic = term_IC = (char *)NULL;
-  term_up = term_dc = term_DC = visible_bell = (char *)NULL;
-  term_ku = term_kd = term_kl = term_kr = (char *)NULL;
-  term_mm = term_mo = (char *)NULL;
-  terminal_can_insert = term_has_meta = _rl_term_autowrap = 0;
-  term_cr = "\r";
+  _rl_term_im = _rl_term_ei = _rl_term_ic = _rl_term_IC = (char *)NULL;
+  _rl_term_up = _rl_term_dc = _rl_term_DC = _rl_visible_bell = (char *)NULL;
+  _rl_term_ku = _rl_term_kd = _rl_term_kl = _rl_term_kr = (char *)NULL;
+  _rl_term_mm = _rl_term_mo = (char *)NULL;
+  _rl_terminal_can_insert = term_has_meta = _rl_term_autowrap = 0;
+  _rl_term_cr = "\r";
+  _rl_term_clreol = _rl_term_clrpag = _rl_term_backspace = (char *)NULL;
+  _rl_term_goto = _rl_term_pc = _rl_term_ip = (char *)NULL;
+  _rl_term_ks = _rl_term_ke =_rl_term_vs = _rl_term_ve = (char *)NULL;
+  _rl_term_kh = _rl_term_kH = _rl_term_at7 = _rl_term_kI = (char *)NULL;
+#if defined(HACK_TERMCAP_MOTION)
+  _rl_term_forward_char = (char *)NULL;
+#endif
 
   _rl_get_screen_size (tty, 0);
 #else  /* !__MSDOS__ */
@@ -677,6 +690,7 @@ void
 _rl_set_cursor (im, force)
      int im, force;
 {
+#ifndef __MSDOS__
   if (_rl_term_ve && _rl_term_vs)
     {
       if (force || im != rl_insert_mode)
@@ -687,4 +701,5 @@ _rl_set_cursor (im, force)
 	    tputs (_rl_term_ve, 1, _rl_output_character_function);
 	}
     }
+#endif
 }
--- readline/display.c~0	2003-01-11 00:43:52.000000000 +0000
+++ readline/display.c	2003-12-23 18:47:36.000000000 +0000
@@ -1642,10 +1642,10 @@ _rl_move_vert (to)
 #ifdef __MSDOS__
       int row, col;
 
-      l = fflush (rl_outstream); /* make sure the cursor pos is current! */
+      fflush (rl_outstream); /* make sure the cursor pos is current! */
       ScreenGetCursor (&row, &col);
-      ScreenSetCursor ((row + to - _rl_last_v_pos), col);
-      delta = i;
+      ScreenSetCursor (row + delta, col);
+      i = -delta;    /* in case someone wants to use it after the loop */
 #else /* !__MSDOS__ */
       if (_rl_term_up && *_rl_term_up)
 	for (i = 0; i < -delta; i++)

Re: [RFA] Patch to imported readline to allow DJGPP build

[Elena Zannoni]

Eli Zaretskii writes:
 > The current imported readline is broken for DJGPP.  The following
 > patch is needed to allow GDB to build the DJGPP port.  Okay to commit?
 > 
 > Btw, I don't understand what is going on with Readline: I looked at
 > its official home page (to figure out how to submit the patches below
 > to the Readline maintainer) and found there an old version 4.3 of
 > Readline where the DJGPP-specific code is ripped out.  The version
 > that we have in the GDB repository seems a newer one, but it also has
 > the DJGPP-specific code in it.  Can someone please tell me what is the
 > story behind this?  TIA
 > 

I remember having exchanged some mail with Chet and you about this,
but I don't remember the outcome. I can't find anything in my old mbox
either.  The readline in our tree includes some patches on top of pure
4.3.  One of the patches is this one, which I kept from the previous
version:

2000-07-10  Eli Zaretskii  <eliz@is.elta.co.il>
                                                                                
        * terminal.c (_rl_get_screen_size) [__DJGPP__]: Determine screen
        size via DJGPP-specific calls.
        (_rl_init_terminal_io) [__MSDOS__]: DJGPP-specific terminal
        initialization.
        (_rl_backspace) [__MSDOS__]: Don't call tputs.
        (ding) [__MSDOS__]: Use DJGPP-specific calls to support visible
        bell.
        * display.c (_rl_move_vert) [__MSDOS__]: Support cursor movement
        upwards with DJGPP-specific calls.
        (_rl_clear_to_eol) [__MSDOS__]: Don't call tputs.
        (_rl_clear_screen) [__MSDOS__]: Support clear-screen with
        DJGPP-specific calls.
        (insert_some_chars) [__MSDOS__]: Don't call tputs.
        (delete_chars) [__MSDOS__]: Don't call tputs.

Other changes are those listed in ChangeLog.gdb. You can do a diff
with a pristine 4.3 readline. There are some other local changes that
I keep integrating at each import.  I should probably have specified
that in the changelog entry at the time.

the files with local changes are:
 ./aclocal.m4
 ./config.h.in
 ./configure
 ./configure.in
 ./display.c
 ./Makefile.in
 ./rlmbutil.h
 ./rltty.c
 ./shell.c
 ./signals.c
 ./terminal.c


You changes are OK. I think you should audit the current diffs,
though, to see if they are all still needed for DJGPP.

On top of that there are some 'official' patches that are advertised
on the readline home page, which I have not merged. 

elena

Re: [RFA] Patch to imported readline to allow DJGPP build

[Eli Zaretskii]

> From: Elena Zannoni <ezannoni@redhat.com>
> Date: Mon, 29 Dec 2003 17:32:44 -0500
> 
> I remember having exchanged some mail with Chet and you about this,
> but I don't remember the outcome. I can't find anything in my old mbox
> either.  The readline in our tree includes some patches on top of pure
> 4.3.  One of the patches is this one, which I kept from the previous
> version:
> 
> 2000-07-10  Eli Zaretskii  <eliz@is.elta.co.il>

Oh, I see now: you've patched readline 4.3 with my patches from an
older version.

> Other changes are those listed in ChangeLog.gdb. You can do a diff
> with a pristine 4.3 readline.

I did both, but still couldn't figure out the issue, since there were
clearly changes that didn't come from any of these two sources.

> There are some other local changes that I keep integrating at each
> import.  I should probably have specified that in the changelog
> entry at the time.

Yes, a good idea, IMHO.

> You changes are OK.

Thanks, I will commit them shortly.

> I think you should audit the current diffs, though, to see if they
> are all still needed for DJGPP.

I did that, and they are.

(There was a plan to have a termcap library for DJGPP that would
eliminate the need for those ugly hacks, but I don't have time to
carry it over, not just yet.)

> On top of that there are some 'official' patches that are advertised
> on the readline home page, which I have not merged. 

Looked at those as well as part of my attempts to understand what
happened.  AFAICS, some of those patches (or their close equivalents)
are already in GDB's repository.

Thanks again.

Re: [RFA] Patch to imported readline to allow DJGPP build

[Elena Zannoni]

Eli Zaretskii writes:
 > 
 > > There are some other local changes that I keep integrating at each
 > > import.  I should probably have specified that in the changelog
 > > entry at the time.
 > 
 > Yes, a good idea, IMHO.
 > 

I modified that Changelog entry. 

Index: ChangeLog.gdb
===================================================================
RCS file: /cvs/src/src/readline/ChangeLog.gdb,v
retrieving revision 1.15
diff -u -p -r1.15 ChangeLog.gdb
--- ChangeLog.gdb       30 Dec 2003 07:25:18 -0000      1.15
+++ ChangeLog.gdb       15 Jan 2004 19:02:59 -0000
@@ -76,7 +76,8 @@
                                                                                    
 2002-12-06  Elena Zannoni  <ezannoni@redhat.com>
  
-        Import of readline 4.3:
+        Import of readline 4.3. NB: This import includes those gdb
+        local changes that aren't in the official readline sources.
  

Re: [RFA] Patch to imported readline to allow DJGPP build

[Eli Zaretskii]

> From: Elena Zannoni <ezannoni@redhat.com>
> Date: Thu, 15 Jan 2004 14:01:42 -0500
> 
> Eli Zaretskii writes:
>  > 
>  > > There are some other local changes that I keep integrating at each
>  > > import.  I should probably have specified that in the changelog
>  > > entry at the time.
>  > 
>  > Yes, a good idea, IMHO.
>  > 
> 
> I modified that Changelog entry. 

Thank you!

Re: [RFA] Patch to imported readline to allow DJGPP build

[Eli Zaretskii]

> From: Elena Zannoni <ezannoni@redhat.com>
> Date: Mon, 29 Dec 2003 17:32:44 -0500
> 
> You changes are OK.

Committed.

Thanks for the prompt response.