[Fwd: [Fwd: Re: Toplevel: Maybe-ize GDB_TK.]]

[Andrew Cagney <ac131313@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 110 bytes --]

(Never one to miss an oportunity to pass the buck, forwarding it to 
insight which uses the tk part).

Andrew

[-- Attachment #2: [Fwd: Re: Toplevel: Maybe-ize GDB_TK.] --]
[-- Type: message/rfc822, Size: 5600 bytes --]

From: Nathanael Nerode <neroden@twcny.rr.com>
To: gdb@sources.redhat.com
Subject: [Fwd: Re: Toplevel: Maybe-ize GDB_TK.]
Date: Sun, 24 Nov 2002 14:50:03 -0500
Message-ID: <3DE12D6B.1010007@twcny.rr.com>



-------- Original Message --------
Subject: Re: Toplevel: Maybe-ize GDB_TK.
Date: Wed, 20 Nov 2002 16:27:05 -0500
From: DJ Delorie <dj@redhat.com>
To: neroden@twcny.rr.com
References: <20021120212326.GA21509@doctormoo>


Could you send a copy of this to the gdb list?  I'm sure they have a
better understanding of the Tk dependencies than the gcc folks.

> X-Sieve: cmu-sieve 2.0
> Date: Wed, 20 Nov 2002 16:23:26 -0500
> Content-Type: text/plain; charset=us-ascii
> Content-Disposition: inline
> User-Agent: Mutt/1.4i
> From: Nathanael Nerode <neroden@twcny.rr.com>
> 
> Another maybe-izing change for b-i-b, tested on i686-pc-linux-gnu.
> 
> The long comment added here indicates what I really want to do, but it needs
> extensive testing so I'm doing the 'preliminaries' here.
> 
> 	* configure: Move GDB_TK substitution to configure.in.
> 	* configure.in: Maybe-ize GDB_TK, rearrange slightly.
> 
> --- configure.4	2002-11-20 16:12:28.000000000 -0500
> +++ configure	2002-11-20 16:18:42.000000000 -0500
> @@ -1355,8 +1355,6 @@
>        -e "s:@DEFAULT_LEX@:${DEFAULT_LEX}:" \
>        -e "s:@DEFAULT_M4@:${DEFAULT_M4}:" \
>            ./Makefile.tem > ${Makefile}
> -  sed -e "s:@GDB_TK@:${GDB_TK}:" ${Makefile} >${Makefile}.tem
> -  mv -f ${Makefile}.tem ${Makefile}
>  
>    # If this is a Canadian Cross, preset the values of many more
>    # tools.
> --- configure.in.4	2002-11-20 16:12:24.000000000 -0500
> +++ configure.in	2002-11-20 16:18:22.000000000 -0500
> @@ -848,13 +848,6 @@
>    *msdosdjgpp*)
>      enable_gdbtk=no ;;
>  esac
> -# Determine whether gdb needs tk/tcl or not.
> -case "$enable_gdbtk" in
> -  no)
> -    GDB_TK="" ;;
> -  *)
> -    GDB_TK="all-tcl all-tk all-itcl all-tix all-libgui" ;;
> -esac
>  
>  copy_dirs=
>  
> @@ -1383,6 +1376,20 @@
>    configure_target_modules="${configure_target_modules} configure-target-${module}"
>  done
>  
> +# Determine whether gdb needs tk/tcl or not.
> +# Use 'maybe' since enable_gdbtk might be true even if tk isn't available
> +# and in that case we want gdb to be built without tk.  Ugh!
> +# In fact I believe gdb is the *only* package directly dependent on tk,
> +# so we should be able to put the 'maybe's in unconditionally and
> +# leave out the maybe dependencies when enable_gdbtk is false.  I'm not
> +# 100% sure that that's safe though.
> +case "$enable_gdbtk" in
> +  no)
> +    GDB_TK="" ;;
> +  *)
> +    GDB_TK="maybe-all-tcl maybe-all-tk maybe-all-itcl maybe-all-tix maybe-all-libgui" ;;
> +esac
> +
>  # Create the 'maybe dependencies'.  This uses a temporary file.
>  rm -f maybedep.tmp
>  for item in ${all_build_modules} ${all_host_modules} ${all_target_modules} \
> @@ -1593,6 +1600,7 @@
>  
>  sedtemp=sed.$$
>  cat >$sedtemp <<EOF
> +s%@GDB_TK@%${GDB_TK}%
>  s%@all_build_modules@%${all_build_modules}%
>  s%@configure_build_modules@%${configure_build_modules}%
>  s%@all_host_modules@%${all_host_modules}%
>