Maintenance patch for protoize

[Jeffrey A Law]

  In message < 199908160546.PAA00745@gluttony.geoffk.wattle.id.au >you write:
  > 
  > Attached.  OK to commit?
  > 
  > As far as I can see, the patch only neatens things up, it doesn't
  > provide any new features (oh, except that 'test-protoize-simple'
  > passes now).
  > 
  > -- 
  > Geoffrey Keating <geoffk@cygnus.com>
  > 
  > ===File ~/patches/cygnus/protoize-arg-2.patch===============
  > Index: egcs/gcc/ChangeLog
  > 0a
  > Tue May 18 12:06:52 1999  Fred Fish  <fnf@cygnus.com>
  > 			  Geoffrey Keating <geoffk@cygnus.com>
  > 
  > 	* Makefile.in (PREPROCESSOR_DEFINES): New macro.
  > 	(protoize.o):  Use PREPROCESSOR_DEFINES and DRIVER_DEFINES.
  > 	(unprotoize.o): Ditto.
  > 	(test-protoize-simple): Don't define STD_PROTO_DIR.
  > 	* protoize.c: Use PARAMS rather than PROTO.  Minor whitespace
  > 	changes to make 'test-protoize-simple' pass.
  >  	(STD_PROTO_DIR): Remove define.
  > 	(STANDARD_EXEC_PREFIX): Supply default define.
  > 	(standard_exec_prefix): New variable, init to STANDARD_EXEC_PREFIX.
  > 	(target_machine): New variable, init to DEFAULT_TARGET_MACHINE.
  > 	(target_version): New variable, init to DEFAULT_TARGET_VERSION.
  > 	(GET_ENV_PATH_LIST): New macro.
  > 	(default_syscalls_dir): No longer initialized to STD_PROTO_DIR.
  > 	(do_processing): Initialize default_syscalls_dir using new
  > 	macros.  Use it to initialize syscalls_absolute_filename.
This is fine.

Are people still using protoize/unprotoize?  Do they really belong in gcc, or
should they be separate?  I guess I'm wondering about what we gain by keeping
this stuff around in the first place.

Thanks,
jeff