New binutils release

New binutils release

[Ian Lance Taylor]

I would like to start moving toward a new binutils release.  I'm going
to be making this release, rather than Ken, the usual binutils
maintainer.

If you have any patches you want to get in, please send them in soon.
If you have sent in patches and they are waiting on copyright
assignments, please get the assignments in ASAP; it can sometimes take
a while for the FSF office to turn them around.

I do have a few patches which I haven't integrated yet; however, if
you have sent in patches and they haven't appeared in a snapshot yet,
let me know, just in case they've gotten lost.

Ian

Re: New binutils release

[manfred]

Hi there,

I found a bug in the generic output routine in linker.c, which emits symbols
even if you specify `-s' (strip all) on the linker command-line; obviously
the check for (info->strip == strip_all) is missing:

diff --context --recursive --show-c-function -x *.o -x *.info* -x *.html* -x *.elc -x *.dvi -x *.orig -x *~ -x version.el /u/b60/manfred/tmp/orig/binutils-2.6/bfd/ChangeLog binutils-2.6/bfd/ChangeLog
*** /u/b60/manfred/tmp/orig/binutils-2.6/bfd/ChangeLog	Fri Nov 17 02:42:59 1995
--- binutils-2.6/bfd/ChangeLog	Wed Jun 19 09:22:21 1996
***************
*** 1,3 ****
--- 1,8 ----
+ Wed Jun 19 07:20:56 1996  Manfred Hollstein KS/EIC5 60/3/142 #40283  <manfred@lts.sel.alcatel.de>
+ 
+ 	* linker.c (_bfd_generic_link_output_symbols): Don't output any
+ 	symbol if info->strip == strip_all.
+ 
  Thu Nov 16 17:32:55 1995  Ken Raeburn  <raeburn@cygnus.com>
  
  	* VERSION: Updated to 2.6.
diff --context --recursive --show-c-function -x *.o -x *.info* -x *.html* -x *.elc -x *.dvi -x *.orig -x *~ -x version.el /u/b60/manfred/tmp/orig/binutils-2.6/bfd/linker.c binutils-2.6/bfd/linker.c
*** /u/b60/manfred/tmp/orig/binutils-2.6/bfd/linker.c	Fri Nov  3 05:17:42 1995
--- binutils-2.6/bfd/linker.c	Wed Jun 19 09:17:22 1996
*************** _bfd_generic_link_output_symbols (output
*** 2155,2164 ****
  
        /* This switch is straight from the old code in
  	 write_file_locals in ldsym.c.  */
!       if (info->strip == strip_some
! 	  && (bfd_hash_lookup (info->keep_hash, bfd_asymbol_name (sym),
! 			       false, false)
! 	      == (struct bfd_hash_entry *) NULL))
  	output = false;
        else if ((sym->flags & (BSF_GLOBAL | BSF_WEAK)) != 0)
  	{
--- 2155,2165 ----
  
        /* This switch is straight from the old code in
  	 write_file_locals in ldsym.c.  */
!       if (info->strip == strip_all
! 	  || (info->strip == strip_some
! 	      && (bfd_hash_lookup (info->keep_hash, bfd_asymbol_name (sym),
! 				   false, false)
! 		  == (struct bfd_hash_entry *) NULL)))
  	output = false;
        else if ((sym->flags & (BSF_GLOBAL | BSF_WEAK)) != 0)
  	{

 Manfred Hollstein			Alcatel TELECOM, Dept. KS/EIC5
 e-mail: manfred@lts.sel.alcatel.de	Lorenzstrasse 10
 Phone: +49 711 821 40283		70435 Stuttgart
 Fax:	+49 711 821 40196		Germany

Re: New binutils release

[Joel Sherrill]

On Wed, 19 Jun 1996, Ian Lance Taylor wrote:

> I do have a few patches which I haven't integrated yet; however, if
> you have sent in patches and they haven't appeared in a snapshot yet,
> let me know, just in case they've gotten lost.

I submitted configuration file changes to add CPU-rtems configurations 
where CPU is one of [i386, m68k, hppa, powerpc, sparc, or i960].  I have 
not seen them in a snapshot.  Do you need the diff versus binutils 2.6 again?

Thanks. 

--joel