build failure in TUI

build failure in TUI

[Tom Tromey]

I just updated and rebuilt the gdb trunk on x86 Red Hat Linux 7.3.

I get this:

gcc -c -g -O2    -I. -I../../src/gdb -I../../src/gdb/config -DLOCALEDIR="\"/home/tromey/gnu/baseline-gdb/install/share/locale\"" -DHAVE_CONFIG_H -I../../src/gdb/../include/opcode -I../../src/gdb/../readline/.. -I../bfd -I../../src/gdb/../bfd  -I../../src/gdb/../include -I../intl -I../../src/gdb/../intl  -DMI_OUT=1 -DTUI=1 -I../../src/gdb/tui -DGDBTK -DUI_OUT=1 -Wimplicit -Wreturn-type -Wcomment -Wtrigraphs -Wformat -Wparentheses -Wpointer-arith -Wuninitialized -Werror ../../src/gdb/tui/tuiIO.c
cc1: warnings being treated as errors
../../src/gdb/tui/tuiIO.c: In function `tui_redisplay_readline':
../../src/gdb/tui/tuiIO.c:190: warning: comparison is always true due to limited range of data type
../../src/gdb/tui/tuiIO.c:193: warning: comparison is always true due to limited range of data type


This causes a failure because I configure with both --enable-tui and
--enable-gdb-build-warnings=,-Werror.

I looked at the code a bit.  I think the simplest fix may be to change
the definition of CTRL_CHAR (in readline/chardefs.h) to cast its
argument to `unsigned char' (and then remove the `>= 0' check).
However, I don't know the policy on changing readline, so I decided
not to send a patch.

Tom

Re: build failure in TUI

[Daniel Jacobowitz]

On Mon, Aug 26, 2002 at 03:15:14PM -0600, Tom Tromey wrote:
> I just updated and rebuilt the gdb trunk on x86 Red Hat Linux 7.3.
> 
> I get this:
> 
> gcc -c -g -O2    -I. -I../../src/gdb -I../../src/gdb/config -DLOCALEDIR="\"/home/tromey/gnu/baseline-gdb/install/share/locale\"" -DHAVE_CONFIG_H -I../../src/gdb/../include/opcode -I../../src/gdb/../readline/.. -I../bfd -I../../src/gdb/../bfd  -I../../src/gdb/../include -I../intl -I../../src/gdb/../intl  -DMI_OUT=1 -DTUI=1 -I../../src/gdb/tui -DGDBTK -DUI_OUT=1 -Wimplicit -Wreturn-type -Wcomment -Wtrigraphs -Wformat -Wparentheses -Wpointer-arith -Wuninitialized -Werror ../../src/gdb/tui/tuiIO.c
> cc1: warnings being treated as errors
> ../../src/gdb/tui/tuiIO.c: In function `tui_redisplay_readline':
> ../../src/gdb/tui/tuiIO.c:190: warning: comparison is always true due to limited range of data type
> ../../src/gdb/tui/tuiIO.c:193: warning: comparison is always true due to limited range of data type
> 
> 
> This causes a failure because I configure with both --enable-tui and
> --enable-gdb-build-warnings=,-Werror.
> 
> I looked at the code a bit.  I think the simplest fix may be to change
> the definition of CTRL_CHAR (in readline/chardefs.h) to cast its
> argument to `unsigned char' (and then remove the `>= 0' check).
> However, I don't know the policy on changing readline, so I decided
> not to send a patch.

Debian has this:
#define CTRL_CHAR(c) ((c) < control_character_threshold && (((c) & 0x80) == 0))

I'm not sure offhand if this is fixed in stock readline 4.3, or if it
is Debian local.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: build failure in TUI

[Elena Zannoni]

Daniel Jacobowitz writes:
 > On Mon, Aug 26, 2002 at 03:15:14PM -0600, Tom Tromey wrote:
 > > I just updated and rebuilt the gdb trunk on x86 Red Hat Linux 7.3.
 > > 
 > > I get this:
 > > 
 > > gcc -c -g -O2    -I. -I../../src/gdb -I../../src/gdb/config -DLOCALEDIR="\"/home/tromey/gnu/baseline-gdb/install/share/locale\"" -DHAVE_CONFIG_H -I../../src/gdb/../include/opcode -I../../src/gdb/../readline/.. -I../bfd -I../../src/gdb/../bfd  -I../../src/gdb/../include -I../intl -I../../src/gdb/../intl  -DMI_OUT=1 -DTUI=1 -I../../src/gdb/tui -DGDBTK -DUI_OUT=1 -Wimplicit -Wreturn-type -Wcomment -Wtrigraphs -Wformat -Wparentheses -Wpointer-arith -Wuninitialized -Werror ../../src/gdb/tui/tuiIO.c
 > > cc1: warnings being treated as errors
 > > ../../src/gdb/tui/tuiIO.c: In function `tui_redisplay_readline':
 > > ../../src/gdb/tui/tuiIO.c:190: warning: comparison is always true due to limited range of data type
 > > ../../src/gdb/tui/tuiIO.c:193: warning: comparison is always true due to limited range of data type
 > > 
 > > 
 > > This causes a failure because I configure with both --enable-tui and
 > > --enable-gdb-build-warnings=,-Werror.
 > > 
 > > I looked at the code a bit.  I think the simplest fix may be to change
 > > the definition of CTRL_CHAR (in readline/chardefs.h) to cast its
 > > argument to `unsigned char' (and then remove the `>= 0' check).
 > > However, I don't know the policy on changing readline, so I decided
 > > not to send a patch.
 > 
 > Debian has this:
 > #define CTRL_CHAR(c) ((c) < control_character_threshold && (((c) & 0x80) == 0))
 > 
 > I'm not sure offhand if this is fixed in stock readline 4.3, or if it
 > is Debian local.
 > 

It's in 4.3.

Elena


 > -- 
 > Daniel Jacobowitz
 > MontaVista Software                         Debian GNU/Linux Developer

Re: build failure in TUI

[Stephane Carrez]

Hi!

Elena Zannoni wrote:

> [...]
>  > Debian has this:
>  > #define CTRL_CHAR(c) ((c) < control_character_threshold && (((c) & 0x80) == 0))
>  > 
>  > I'm not sure offhand if this is fixed in stock readline 4.3, or if it
>  > is Debian local.
>  > 
> 
> It's in 4.3.
> 
> Elena
> 


We can temporarily override it in tuiIO.c to use 4.3 definition and until we have readline 4.3.


The TUI does not build with readline 4.3 (on the branch) because it uses readline internal

variables that have disappeared.  I filed a gdb PR for that.  But this must not prevent
a possible use of readline 4.3 for gdb 5.3.

	Stephane

-----------------------------------------------------------------------
         Home                               Office
E-mail: stcarrez@nerim.fr                  Stephane.Carrez@solsoft.fr
WWW:    http://stephane.carrez.free.fr     http://www.solsoft.com
         Free the Software!                 Visual Security Policy Management

Re: build failure in TUI

[Elena Zannoni]

Stephane Carrez writes:
 > Hi!
 > 
 > Elena Zannoni wrote:
 > 
 > > [...]
 > >  > Debian has this:
 > >  > #define CTRL_CHAR(c) ((c) < control_character_threshold && (((c) & 0x80) == 0))
 > >  > 
 > >  > I'm not sure offhand if this is fixed in stock readline 4.3, or if it
 > >  > is Debian local.
 > >  > 
 > > 
 > > It's in 4.3.
 > > 
 > > Elena
 > > 
 > 
 > 
 > We can temporarily override it in tuiIO.c to use 4.3 definition and until we have readline 4.3.
 > 

Yes.

 > 
 > The TUI does not build with readline 4.3 (on the branch) because it uses readline internal
 > 
 > variables that have disappeared.  I filed a gdb PR for that.  But this must not prevent
 > a possible use of readline 4.3 for gdb 5.3.
 > 

Is there a workaround? Maybe the variables have disappeared from the
interface, but some accessor functions have been added to readline?

Elena


 > 	Stephane
 > 
 > -----------------------------------------------------------------------
 >          Home                               Office
 > E-mail: stcarrez@nerim.fr                  Stephane.Carrez@solsoft.fr
 > WWW:    http://stephane.carrez.free.fr     http://www.solsoft.com
 >          Free the Software!                 Visual Security Policy Management
 > 

Re: build failure in TUI

[Daniel Jacobowitz]

On Tue, Aug 27, 2002 at 01:51:45AM +0200, Stephane Carrez wrote:
> Hi!
> 
> Elena Zannoni wrote:
> 
> >[...]
> > > Debian has this:
> > > #define CTRL_CHAR(c) ((c) < control_character_threshold && (((c) & 
> > 0x80) == 0))
> > > 
> > > I'm not sure offhand if this is fixed in stock readline 4.3, or if it
> > > is Debian local.
> > > 
> >
> >It's in 4.3.
> >
> >Elena
> >
> 
> 
> We can temporarily override it in tuiIO.c to use 4.3 definition and until 
> we have readline 4.3.
> 
> 
> The TUI does not build with readline 4.3 (on the branch) because it uses 
> readline internal
> 
> variables that have disappeared.  I filed a gdb PR for that.  But this must 
> not prevent
> a possible use of readline 4.3 for gdb 5.3.

They've been renamed, into _rl_.  They're visible in at least the
Debian readline packages, since I've built TUI with readline 4.3.  I
suspect there are accessors you're supposed to be using, though... The
patch I'm using:

Debian local - some variables were renamed in readline 4.2.

diff -urNp gdb-5.2.cvs20010401/gdb/tui.orig/tuiWin.c gdb-5.2.cvs20010401/gdb/tui/tuiWin.c
--- gdb-5.2.cvs20010401/gdb/tui.orig/tuiWin.c	Sun Mar 31 22:09:42 2002
+++ gdb-5.2.cvs20010401/gdb/tui/tuiWin.c	Sun Mar 31 23:01:27 2002
@@ -639,10 +639,10 @@ void
 tuiResizeAll (void)
 {
   int heightDiff, widthDiff;
-  extern int screenheight, screenwidth;		/* in readline */
+  extern int _rl_screenheight, _rl_screenwidth;		/* in readline */
 
-  widthDiff = screenwidth - termWidth ();
-  heightDiff = screenheight - termHeight ();
+  widthDiff = _rl_screenwidth - termWidth ();
+  heightDiff = _rl_screenheight - termHeight ();
   if (heightDiff || widthDiff)
     {
       TuiLayoutType curLayout = currentLayout ();
@@ -656,8 +656,8 @@ tuiResizeAll (void)
       if (winWithFocus != cmdWin)
 	keypad (cmdWin->generic.handle, FALSE);
       init_page_info ();
-      setTermHeightTo (screenheight);
-      setTermWidthTo (screenwidth);
+      setTermHeightTo (_rl_screenheight);
+      setTermWidthTo (_rl_screenwidth);
       if (curLayout == SRC_DISASSEM_COMMAND ||
 	curLayout == SRC_DATA_COMMAND || curLayout == DISASSEM_DATA_COMMAND)
 	numWinsDisplayed++;
@@ -684,8 +684,8 @@ tuiResizeAll (void)
 	  if (heightDiff == 0)
 	    newHeight = firstWin->generic.height;
 	  else if ((firstWin->generic.height + splitDiff) >=
-		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
-	    newHeight = screenheight - MIN_CMD_WIN_HEIGHT - 1;
+		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
+	    newHeight = _rl_screenheight - MIN_CMD_WIN_HEIGHT - 1;
 	  else if ((firstWin->generic.height + splitDiff) <= 0)
 	    newHeight = MIN_WIN_HEIGHT;
 	  else
@@ -694,7 +694,7 @@ tuiResizeAll (void)
 	  _makeInvisibleAndSetNewHeight (firstWin, newHeight);
 	  cmdWin->generic.origin.y = locator->origin.y + 1;
 	  cmdWin->generic.width += widthDiff;
-	  newHeight = screenheight - cmdWin->generic.origin.y;
+	  newHeight = _rl_screenheight - cmdWin->generic.origin.y;
 	  _makeInvisibleAndSetNewHeight (cmdWin, newHeight);
 	  _makeVisibleWithNewHeight (firstWin);
 	  _makeVisibleWithNewHeight (cmdWin);
@@ -722,8 +722,8 @@ tuiResizeAll (void)
 	    newHeight = firstWin->generic.height;
 	  else if ((firstWin->generic.height +
 		    secondWin->generic.height + (splitDiff * 2)) >=
-		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
-	    newHeight = (screenheight - MIN_CMD_WIN_HEIGHT - 1) / 2;
+		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
+	    newHeight = (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1) / 2;
 	  else if ((firstWin->generic.height + splitDiff) <= 0)
 	    newHeight = MIN_WIN_HEIGHT;
 	  else
@@ -742,9 +742,9 @@ tuiResizeAll (void)
 	    newHeight = secondWin->generic.height;
 	  else if ((firstWin->generic.height +
 		    secondWin->generic.height + (splitDiff * 2)) >=
-		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
+		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
 	    {
-	      newHeight = screenheight - MIN_CMD_WIN_HEIGHT - 1;
+	      newHeight = _rl_screenheight - MIN_CMD_WIN_HEIGHT - 1;
 	      if (newHeight % 2)
 		newHeight = (newHeight / 2) + 1;
 	      else

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: build failure in TUI

[Elena Zannoni]

Daniel Jacobowitz writes:
 > On Tue, Aug 27, 2002 at 01:51:45AM +0200, Stephane Carrez wrote:
 > > Hi!
 > > 
 > > Elena Zannoni wrote:
 > > 
 > > >[...]
 > > > > Debian has this:
 > > > > #define CTRL_CHAR(c) ((c) < control_character_threshold && (((c) & 
 > > > 0x80) == 0))
 > > > > 
 > > > > I'm not sure offhand if this is fixed in stock readline 4.3, or if it
 > > > > is Debian local.
 > > > > 
 > > >
 > > >It's in 4.3.
 > > >
 > > >Elena
 > > >
 > > 
 > > 
 > > We can temporarily override it in tuiIO.c to use 4.3 definition and until 
 > > we have readline 4.3.
 > > 
 > > 
 > > The TUI does not build with readline 4.3 (on the branch) because it uses 
 > > readline internal
 > > 
 > > variables that have disappeared.  I filed a gdb PR for that.  But this must 
 > > not prevent
 > > a possible use of readline 4.3 for gdb 5.3.
 > 
 > They've been renamed, into _rl_.  They're visible in at least the
 > Debian readline packages, since I've built TUI with readline 4.3.  I
 > suspect there are accessors you're supposed to be using, though... The

I see these documented in readline/CHANGES as having been made
internal only. I don't see any accessor functions. You can still use
it, but they shouldn't be exported.

Elena


 > patch I'm using:
 > 
 > Debian local - some variables were renamed in readline 4.2.
 > 
 > diff -urNp gdb-5.2.cvs20010401/gdb/tui.orig/tuiWin.c gdb-5.2.cvs20010401/gdb/tui/tuiWin.c
 > --- gdb-5.2.cvs20010401/gdb/tui.orig/tuiWin.c	Sun Mar 31 22:09:42 2002
 > +++ gdb-5.2.cvs20010401/gdb/tui/tuiWin.c	Sun Mar 31 23:01:27 2002
 > @@ -639,10 +639,10 @@ void
 >  tuiResizeAll (void)
 >  {
 >    int heightDiff, widthDiff;
 > -  extern int screenheight, screenwidth;		/* in readline */
 > +  extern int _rl_screenheight, _rl_screenwidth;		/* in readline */
 >  
 > -  widthDiff = screenwidth - termWidth ();
 > -  heightDiff = screenheight - termHeight ();
 > +  widthDiff = _rl_screenwidth - termWidth ();
 > +  heightDiff = _rl_screenheight - termHeight ();
 >    if (heightDiff || widthDiff)
 >      {
 >        TuiLayoutType curLayout = currentLayout ();
 > @@ -656,8 +656,8 @@ tuiResizeAll (void)
 >        if (winWithFocus != cmdWin)
 >  	keypad (cmdWin->generic.handle, FALSE);
 >        init_page_info ();
 > -      setTermHeightTo (screenheight);
 > -      setTermWidthTo (screenwidth);
 > +      setTermHeightTo (_rl_screenheight);
 > +      setTermWidthTo (_rl_screenwidth);
 >        if (curLayout == SRC_DISASSEM_COMMAND ||
 >  	curLayout == SRC_DATA_COMMAND || curLayout == DISASSEM_DATA_COMMAND)
 >  	numWinsDisplayed++;
 > @@ -684,8 +684,8 @@ tuiResizeAll (void)
 >  	  if (heightDiff == 0)
 >  	    newHeight = firstWin->generic.height;
 >  	  else if ((firstWin->generic.height + splitDiff) >=
 > -		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
 > -	    newHeight = screenheight - MIN_CMD_WIN_HEIGHT - 1;
 > +		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
 > +	    newHeight = _rl_screenheight - MIN_CMD_WIN_HEIGHT - 1;
 >  	  else if ((firstWin->generic.height + splitDiff) <= 0)
 >  	    newHeight = MIN_WIN_HEIGHT;
 >  	  else
 > @@ -694,7 +694,7 @@ tuiResizeAll (void)
 >  	  _makeInvisibleAndSetNewHeight (firstWin, newHeight);
 >  	  cmdWin->generic.origin.y = locator->origin.y + 1;
 >  	  cmdWin->generic.width += widthDiff;
 > -	  newHeight = screenheight - cmdWin->generic.origin.y;
 > +	  newHeight = _rl_screenheight - cmdWin->generic.origin.y;
 >  	  _makeInvisibleAndSetNewHeight (cmdWin, newHeight);
 >  	  _makeVisibleWithNewHeight (firstWin);
 >  	  _makeVisibleWithNewHeight (cmdWin);
 > @@ -722,8 +722,8 @@ tuiResizeAll (void)
 >  	    newHeight = firstWin->generic.height;
 >  	  else if ((firstWin->generic.height +
 >  		    secondWin->generic.height + (splitDiff * 2)) >=
 > -		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
 > -	    newHeight = (screenheight - MIN_CMD_WIN_HEIGHT - 1) / 2;
 > +		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
 > +	    newHeight = (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1) / 2;
 >  	  else if ((firstWin->generic.height + splitDiff) <= 0)
 >  	    newHeight = MIN_WIN_HEIGHT;
 >  	  else
 > @@ -742,9 +742,9 @@ tuiResizeAll (void)
 >  	    newHeight = secondWin->generic.height;
 >  	  else if ((firstWin->generic.height +
 >  		    secondWin->generic.height + (splitDiff * 2)) >=
 > -		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
 > +		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
 >  	    {
 > -	      newHeight = screenheight - MIN_CMD_WIN_HEIGHT - 1;
 > +	      newHeight = _rl_screenheight - MIN_CMD_WIN_HEIGHT - 1;
 >  	      if (newHeight % 2)
 >  		newHeight = (newHeight / 2) + 1;
 >  	      else
 > 
 > -- 
 > Daniel Jacobowitz
 > MontaVista Software                         Debian GNU/Linux Developer

Re: build failure in TUI

[Elena Zannoni]

Elena Zannoni writes:
 > Daniel Jacobowitz writes:
 >  > On Tue, Aug 27, 2002 at 01:51:45AM +0200, Stephane Carrez wrote:
 >  > > Hi!
 >  > > 
 >  > > Elena Zannoni wrote:
 >  > > 
 >  > > >[...]
 >  > > > > Debian has this:
 >  > > > > #define CTRL_CHAR(c) ((c) < control_character_threshold && (((c) & 
 >  > > > 0x80) == 0))
 >  > > > > 
 >  > > > > I'm not sure offhand if this is fixed in stock readline 4.3, or if it
 >  > > > > is Debian local.
 >  > > > > 
 >  > > >
 >  > > >It's in 4.3.
 >  > > >
 >  > > >Elena
 >  > > >
 >  > > 
 >  > > 
 >  > > We can temporarily override it in tuiIO.c to use 4.3 definition and until 
 >  > > we have readline 4.3.
 >  > > 
 >  > > 
 >  > > The TUI does not build with readline 4.3 (on the branch) because it uses 
 >  > > readline internal
 >  > > 
 >  > > variables that have disappeared.  I filed a gdb PR for that.  But this must 
 >  > > not prevent
 >  > > a possible use of readline 4.3 for gdb 5.3.
 >  > 
 >  > They've been renamed, into _rl_.  They're visible in at least the
 >  > Debian readline packages, since I've built TUI with readline 4.3.  I
 >  > suspect there are accessors you're supposed to be using, though... The
 > 
 > I see these documented in readline/CHANGES as having been made
 > internal only. I don't see any accessor functions. You can still use
 > it, but they shouldn't be exported.
 > 
 > Elena
 > 
 >

OK, answering my own question:

the TUI should be using this (untested, uncompiled):
Index: tuiWin.c
===================================================================
RCS file: /cvs/src/src/gdb/tui/tuiWin.c,v
retrieving revision 1.19
diff -u -r1.19 tuiWin.c
--- tuiWin.c    1 Mar 2002 06:19:28 -0000       1.19
+++ tuiWin.c    27 Aug 2002 00:35:42 -0000
@@ -639,8 +639,9 @@
 tuiResizeAll (void)
 {
   int heightDiff, widthDiff;
-  extern int screenheight, screenwidth;                /* in readline */
+  int screenheight, screenwidth;               /* in readline */
 
+  rl_get_screen_size (&screenheight, &screenwidth);
   widthDiff = screenwidth - termWidth ();
   heightDiff = screenheight - termHeight ();
   if (heightDiff || widthDiff)


 


 >  > patch I'm using:
 >  > 
 >  > Debian local - some variables were renamed in readline 4.2.
 >  > 
 >  > diff -urNp gdb-5.2.cvs20010401/gdb/tui.orig/tuiWin.c gdb-5.2.cvs20010401/gdb/tui/tuiWin.c
 >  > --- gdb-5.2.cvs20010401/gdb/tui.orig/tuiWin.c	Sun Mar 31 22:09:42 2002
 >  > +++ gdb-5.2.cvs20010401/gdb/tui/tuiWin.c	Sun Mar 31 23:01:27 2002
 >  > @@ -639,10 +639,10 @@ void
 >  >  tuiResizeAll (void)
 >  >  {
 >  >    int heightDiff, widthDiff;
 >  > -  extern int screenheight, screenwidth;		/* in readline */
 >  > +  extern int _rl_screenheight, _rl_screenwidth;		/* in readline */
 >  >  
 >  > -  widthDiff = screenwidth - termWidth ();
 >  > -  heightDiff = screenheight - termHeight ();
 >  > +  widthDiff = _rl_screenwidth - termWidth ();
 >  > +  heightDiff = _rl_screenheight - termHeight ();
 >  >    if (heightDiff || widthDiff)
 >  >      {
 >  >        TuiLayoutType curLayout = currentLayout ();
 >  > @@ -656,8 +656,8 @@ tuiResizeAll (void)
 >  >        if (winWithFocus != cmdWin)
 >  >  	keypad (cmdWin->generic.handle, FALSE);
 >  >        init_page_info ();
 >  > -      setTermHeightTo (screenheight);
 >  > -      setTermWidthTo (screenwidth);
 >  > +      setTermHeightTo (_rl_screenheight);
 >  > +      setTermWidthTo (_rl_screenwidth);
 >  >        if (curLayout == SRC_DISASSEM_COMMAND ||
 >  >  	curLayout == SRC_DATA_COMMAND || curLayout == DISASSEM_DATA_COMMAND)
 >  >  	numWinsDisplayed++;
 >  > @@ -684,8 +684,8 @@ tuiResizeAll (void)
 >  >  	  if (heightDiff == 0)
 >  >  	    newHeight = firstWin->generic.height;
 >  >  	  else if ((firstWin->generic.height + splitDiff) >=
 >  > -		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
 >  > -	    newHeight = screenheight - MIN_CMD_WIN_HEIGHT - 1;
 >  > +		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
 >  > +	    newHeight = _rl_screenheight - MIN_CMD_WIN_HEIGHT - 1;
 >  >  	  else if ((firstWin->generic.height + splitDiff) <= 0)
 >  >  	    newHeight = MIN_WIN_HEIGHT;
 >  >  	  else
 >  > @@ -694,7 +694,7 @@ tuiResizeAll (void)
 >  >  	  _makeInvisibleAndSetNewHeight (firstWin, newHeight);
 >  >  	  cmdWin->generic.origin.y = locator->origin.y + 1;
 >  >  	  cmdWin->generic.width += widthDiff;
 >  > -	  newHeight = screenheight - cmdWin->generic.origin.y;
 >  > +	  newHeight = _rl_screenheight - cmdWin->generic.origin.y;
 >  >  	  _makeInvisibleAndSetNewHeight (cmdWin, newHeight);
 >  >  	  _makeVisibleWithNewHeight (firstWin);
 >  >  	  _makeVisibleWithNewHeight (cmdWin);
 >  > @@ -722,8 +722,8 @@ tuiResizeAll (void)
 >  >  	    newHeight = firstWin->generic.height;
 >  >  	  else if ((firstWin->generic.height +
 >  >  		    secondWin->generic.height + (splitDiff * 2)) >=
 >  > -		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
 >  > -	    newHeight = (screenheight - MIN_CMD_WIN_HEIGHT - 1) / 2;
 >  > +		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
 >  > +	    newHeight = (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1) / 2;
 >  >  	  else if ((firstWin->generic.height + splitDiff) <= 0)
 >  >  	    newHeight = MIN_WIN_HEIGHT;
 >  >  	  else
 >  > @@ -742,9 +742,9 @@ tuiResizeAll (void)
 >  >  	    newHeight = secondWin->generic.height;
 >  >  	  else if ((firstWin->generic.height +
 >  >  		    secondWin->generic.height + (splitDiff * 2)) >=
 >  > -		   (screenheight - MIN_CMD_WIN_HEIGHT - 1))
 >  > +		   (_rl_screenheight - MIN_CMD_WIN_HEIGHT - 1))
 >  >  	    {
 >  > -	      newHeight = screenheight - MIN_CMD_WIN_HEIGHT - 1;
 >  > +	      newHeight = _rl_screenheight - MIN_CMD_WIN_HEIGHT - 1;
 >  >  	      if (newHeight % 2)
 >  >  		newHeight = (newHeight / 2) + 1;
 >  >  	      else
 >  > 
 >  > -- 
 >  > Daniel Jacobowitz
 >  > MontaVista Software                         Debian GNU/Linux Developer