Re: egcs-19980508 warning patches part 1/5

Re: egcs-19980508 warning patches part 1/5

[Kaveh R. Ghazi]

 > From: Jeffrey A Law <law@hurl.cygnus.com>
 > 
 >   In message < 199805121406.KAA28401@caip.rutgers.edu >you write:
 >   > 	This is patch part 1/5.  Okay to install?
 >   > 
 >   > Note: I tried to only use ATTRIBUTE_UNUSED where the function call
 >   > interface was mandated by outside constraints.  Also, the ChangeLog
 >   > is for all parts 1 to 4.
 > 
 > 1-4 basically look fine to me.  I probably wouldn't have gone
 > through the trouble of adding all those ATTRIBUTE_UNUSED, but I
 > guess we might as well go ahead use them since you've done the work.

	Installed.

	Regarding the ATTRIBUTE_UNUSED stuff, I'd like to continue
marking unused stuff when, for some reason, its not possible to remove
the unused thing outright.  This helps cut down on the number of
spurious warnings we get.  Too many spurious warnings lead to people
ignoring real warnings.




 >   > 	* cplus-dem.c (squangle_mop_up): Change return type to void.
 >   > 	(internal_cplus_demangle): Remove unused parameter `options'.
 >   > 	All callers changed.
 >   > 	(cplus_demangle_opname): Remove function wide variable `int i' and
 >   >  	replace with `size_t i' at each location where it is used.
 >   > 	(cplus_demangle_opname): change type of `i' from int to size_t.
 > Can you also apply that patch to libiberty.  It also makes one of
 > patches that I haven't checked in redundant :-)  Ahh, I see the
 > libiberty patch in a later part.
 > 
 > I assume we've always got size_t on hosts these days :-)?  I can't
 > remember if this was discussed and dealt with before or not.


	I don't recall reading a size_t discussion.  However, since
the source is already littered with size_t I assume we would have
heard by now if it was a problem. :-) If it becomes an issue for
anyone, we could always put AC_TYPE_SIZE_T in configure.in to handle
it.



 >   > 	* scan-decls.c: Include scan.h.
 > I didn't see an update of Makefile.in dependencies for this one.
 > jeff

	Correct, because for some unknown reason scan-decls.o already
depends on scan.h.  So I deemed a second dependency unnecessary. :-)

		--Kaveh
--
Kaveh R. Ghazi			Project Manager / Custom Development
ghazi@caip.rutgers.edu		Icon CMT Corp.

Re: egcs-19980508 warning patches part 1/5

[Jeffrey A Law]

  In message < 199805132005.QAA23802@caip.rutgers.edu >you write:
  > 	Regarding the ATTRIBUTE_UNUSED stuff, I'd like to continue
  > marking unused stuff when, for some reason, its not possible to remove
  > the unused thing outright.  This helps cut down on the number of
  > spurious warnings we get.  Too many spurious warnings lead to people
  > ignoring real warnings.
I consider the unused and signed comparison warnings the least
interesting and prime candidates for turning off by default.

Do we want to go scatter attributes all over the sources just to
eliminate warnings?  Especially a warning that is extremely unlikely
to actually point us at a bug?

There's certainly value in eliminating unused parameters when we
can, but when an external interface issue forces us to have
parms which may not be used, it's unclear if taking the time and
effort to install and maintain unused attributes is ever going to
pay off.

I guess it's the basic "how far do we take the warning elimination
work".  Some claim we've already gone too far, others claim we
haven't gone far enough.  I'm still undecided myself, but leaning
towards we've gone "far enough" for the unused stuff.


  > 	Correct, because for some unknown reason scan-decls.o already
  > depends on scan.h.  So I deemed a second dependency unnecessary. :-)
Ah.  Nevermind then.

jeff

Re: egcs-19980508 warning patches part 1/5

[Kaveh R. Ghazi]

 > From: Jeffrey A Law <law@hurl.cygnus.com>
 > 
 > As I said, I'm still undecided at this time.  And (of course) I'm
 > willing to keep going if that's what folks think is best.  I've
 > heard from both sides equally so far.

	Oh okay, then I'll keep sprinkling them among other warning fixes. :-)

--
Kaveh R. Ghazi			Project Manager / Custom Development
ghazi@caip.rutgers.edu		Icon CMT Corp.

Re: egcs-19980508 warning patches part 1/5

[Jeffrey A Law]

  In message < 199805141603.MAA25681@caip.rutgers.edu >you write:
  > 	I agree they are unlikely to point to an actual bug.  That's
  > why I try to eliminate them, so they don't distract you and others
  > from more serious warnings.
Yup.  Understood.  I've made the same argument myself :-)

  > 	Well if I'm the one investing the effort... :-)
Yup.  Thus the patches went in :-)

  > 	Seriously, I think I explained above why I see this as useful.
  > But I can understand your point of view and I'm not going to oppose if
  > you don't want these fixed.
As I said, I'm still undecided at this time.  And (of course) I'm
willing to keep going if that's what folks think is best.  I've
heard from both sides equally so far.

  > So maybe we should add:
  > 
  > 	-Wno-sign-compare
  > 	-Wno-unused
This kinda what I've been thinking about.  We can turn them on (particularly
-Wunused) at times to search for dead functions/parameters.

  > 	-Wno-uninitialized  (This one appears to be too broken to be useful.)
Yea.  I looked into improving this, but we need another fairly
significant concept to be added to the compiler.  I suspect we'll
have it one day, but not soon.

I actually did go fix a few of these.  But they're time consuming
since I actually did verify that the variable would be initialized
in each case before I put in a default initialization.

Even though these are the hardest warnings to fix properly, they
are probably the most important ones left to go after since they're
the most likely to expose a real (and bloody obscure) problem.

  > 	On the flip side, I think there are some other warnings (not
  > enabled by -W -Wall) that we should _add_ to WARN_CFLAGS.
I've considered this too.

Particularly aggregate-return (which isn't supported in some older
compilers, or has semantics which are rather unfriendly).

I've been nailed by shadowing problems in the past, so it's a
warning I like. :-)

  > 	BTW, what's a good value for LEN in -Wid-clash-LEN?
No clue.

  > 	-Wstrict-prototypes
  > 	-Wmissing-prototypes
  > 	-Wmissing-declarations
  > 	-Wredundant-decls
  > 
  > These might be useful to turn on once, fix the problems and turn it
  > off again since I think many spurious warnings would result.
There's certainly lots of warnings when you turn them on.  But
the prototype stuff is critical for fixing problems when HOST_WIDE_INT
is a different size than "int" or "long".

jeff

Re: egcs-19980508 warning patches part 1/5

[Joern Rennecke]

> 	On the flip side, I think there are some other warnings (not
> enabled by -W -Wall) that we should _add_ to WARN_CFLAGS.
> 
> Here are some suggestions:
> 	
> 	-Wtraditional

This doesn't make sense for code that is conditionalized on __STDC__

Re: egcs-19980508 warning patches part 1/5

[Kaveh R. Ghazi]

 > From: Jeffrey A Law <law@hurl.cygnus.com>
 > 
 >   In message < 199805132005.QAA23802@caip.rutgers.edu >you write:
 >   > 	Regarding the ATTRIBUTE_UNUSED stuff, I'd like to continue
 >   > marking unused stuff when, for some reason, its not possible to remove
 >   > the unused thing outright.  This helps cut down on the number of
 >   > spurious warnings we get.  Too many spurious warnings lead to people
 >   > ignoring real warnings.
 > I consider the unused and signed comparison warnings the least
 > interesting and prime candidates for turning off by default.
 > 
 > Do we want to go scatter attributes all over the sources just to
 > eliminate warnings?  Especially a warning that is extremely unlikely
 > to actually point us at a bug?

	I agree they are unlikely to point to an actual bug.  That's
why I try to eliminate them, so they don't distract you and others
from more serious warnings.


 > 
 > There's certainly value in eliminating unused parameters when we
 > can, but when an external interface issue forces us to have
 > parms which may not be used, it's unclear if taking the time and
 > effort to install and maintain unused attributes is ever going to
 > pay off.
 > 

	Well if I'm the one investing the effort... :-)

	Seriously, I think I explained above why I see this as useful.
But I can understand your point of view and I'm not going to oppose if
you don't want these fixed.


 > I guess it's the basic "how far do we take the warning elimination
 > work".  Some claim we've already gone too far, others claim we
 > haven't gone far enough.  I'm still undecided myself, but leaning
 > towards we've gone "far enough" for the unused stuff.
 > jeff

	Okay, I'll work on other stuff.



So maybe we should add:

	-Wno-sign-compare
	-Wno-unused
	-Wno-uninitialized  (This one appears to be too broken to be useful.)

to WARN_CFLAGS.  (I think this would be a loss, but not a big deal.)
This would eliminate 75-80% of the remaining warnings.



	On the flip side, I think there are some other warnings (not
enabled by -W -Wall) that we should _add_ to WARN_CFLAGS.

Here are some suggestions:
	
	-Wtraditional
	-Wshadow
	-Wid-clash-LEN
	-Wcast-align
	-Wconversion
	-Waggregate-return
	-Wnested-externs

	Some of the above can be especially useful in tracking down
problems when bootstrapping with CC != gcc.  Eg, that problem
bootstrapping on alphas that I think Jim tracked down to argument
casting not occuring because of missing prototype args when !__STDC__.

	BTW, what's a good value for LEN in -Wid-clash-LEN?


Here are some more:

	-Wstrict-prototypes
	-Wmissing-prototypes
	-Wmissing-declarations
	-Wredundant-decls

These might be useful to turn on once, fix the problems and turn it
off again since I think many spurious warnings would result.

	Anyway, I'll try bootstrapping with some of these and see what
happens.  If you approve having any of the above being added, I'll
submit a patch to change WARN_CFLAGS. 

		--Kaveh
--
Kaveh R. Ghazi			Project Manager / Custom Development
ghazi@caip.rutgers.edu		Icon CMT Corp.

Re: egcs-19980508 warning patches part 1/5

[Jeffrey A Law]

  In message < 199805121406.KAA28401@caip.rutgers.edu >you write:
  > 	This is patch part 1/5.  Okay to install?
  > 
  > Note: I tried to only use ATTRIBUTE_UNUSED where the function call
  > interface was mandated by outside constraints.  Also, the ChangeLog
  > is for all parts 1 to 4.
1-4 basically look fine to me.  I probably wouldn't have gone
through the trouble of adding all those ATTRIBUTE_UNUSED, but I
guess we might as well go ahead use them since you've done the work.





  > 	* cplus-dem.c (squangle_mop_up): Change return type to void.
  > 	(internal_cplus_demangle): Remove unused parameter `options'.
  > 	All callers changed.
  > 	(cplus_demangle_opname): Remove function wide variable `int i' and
  >  	replace with `size_t i' at each location where it is used.
  > 	(cplus_demangle_opname): change type of `i' from int to size_t.
Can you also apply that patch to libiberty.  It also makes one of
patches that I haven't checked in redundant :-)  Ahh, I see the
libiberty patch in a later part.

I assume we've always got size_t on hosts these days :-)?  I can't
remember if this was discussed and dealt with before or not.

  > 	* scan-decls.c: Include scan.h.
I didn't see an update of Makefile.in dependencies for this one.


jeff

egcs-19980508 warning patches part 1/5

[Kaveh R. Ghazi]

	This is patch part 1/5.  Okay to install?

Note: I tried to only use ATTRIBUTE_UNUSED where the function call
interface was mandated by outside constraints.  Also, the ChangeLog
is for all parts 1 to 4.




Mon May 11 21:28:22 1998  Kaveh R. Ghazi  <ghazi@caip.rutgers.edu>

	* Makefile.in (c-lang.o): Depend on c-tree.h, c-lex.h and toplev.h.
	(c-lex.o): Depend on output.h.
	(c-common.o): Likewise.
	(stmt.o): Likewise.
	(calls.o): Likewise.
	(integrate.o): Depend on toplev.h.
	(regclass.o): Depend on output.h.
	(final.o): Depend on reload.h.

	* c-common.c: Include output.h.
	(check_format_info): Remove unused variable `integral_format'.

	* c-decl.c (print_lang_decl): Mark parameters `file', `node' and
	`indent' with ATTRIBUTE_UNUSED.
	(print_lang_type): Likewise.
	(maybe_build_cleanup): Likewise for parameter `decl'.
	(copy_lang_decl): Likewise for parameter `node'.

	* c-lang.c: Include c-tree.h, c-lex.h and toplev.h.
	(lang_print_xnode): Mark parameters `file', `node' and `indent'
 	with ATTRIBUTE_UNUSED.
	(lookup_interface): Likewise for parameter `arg'.
	(is_class_name): Likewise.
	(maybe_objc_check_decl): Likewise for parameter `decl'.
	(maybe_objc_comptypes): Likewise for parameters `lhs', `rhs' and
	`reflexive'.
	(maybe_objc_method_name): Likewise for parameter `decl'.
	(build_objc_string): Likewise for parameters `len' and `str'.

	* c-lex.c: Include output.h.

	* c-lex.h (position_after_white_space): Correct typo in prototype.

	* c-tree.h (finish_file, c_expand_start_cond, c_expand_start_else,
 	c_expand_end_cond, init_iterators): Add prototypes.

	* caller-save.c (set_reg_live): Mark parameters `reg' and `setter'
	with ATTRIBUTE_UNUSED.

	* calls.c: Include output.h.

	* cccp.c (pipe_closed): Mark parameter `signo' with
 	ATTRIBUTE_UNUSED.

	* combine.c: Move inclusion of expr.h to after insn-config.h.

	* iris6.h (ASM_IDENTIFY_GCC, ASM_IDENTIFY_LANGUAGE): Don't define
	as empty, rather define as ((void)0).

	* sparc.c (sparc_check_64): Add braces around ambiguous `else'.
	Add parentheses around assignment used as truth value.

	* cplus-dem.c (squangle_mop_up): Change return type to void.
	(internal_cplus_demangle): Remove unused parameter `options'.
	All callers changed.
	(cplus_demangle_opname): Remove function wide variable `int i' and
 	replace with `size_t i' at each location where it is used.
	(cplus_demangle_opname): change type of `i' from int to size_t.

	* cppexp.c (right_shift): Mark parameter `pfile' with
	ATTRIBUTE_UNUSED.

	* cpphash.c (cpp_lookup): Likewise.
	(cpp_hash_cleanup): Likewise.

	* cpplib.c (parse_name): Add a prototype and make it static.
	(null_underflow): Mark parameter `pfile' with ATTRIBUTE_UNUSED.
	(null_cleanup): Likewise for parameters `pbuf' and `pfile'.
	(macro_cleanup): Likewise for parameter `pfile'.
	(file_cleanup): Likewise.

	* cpplib.h (cpp_reader_init, cpp_options_init, cpp_start_read,
 	cpp_read_check_assertion, skip_rest_of_line): Add prototypes.

	* crtstuff.c (force_to_data, __CTOR_LIST__, force_to_data,
	__DTOR_END__, __FRAME_END__): Mark with ATTRIBUTE_UNUSED.

	* cse.c (cse_check_loop_start): Mark parameter `set' with
 	ATTRIBUTE_UNUSED.

	* dbxout.c (flag_minimal_debug, have_used_extensions,
	source_label_number): Move inside macro wrapper check against
	defined (DBX_DEBUGGING_INFO) || defined (XCOFF_DEBUGGING_INFO).

	* dwarf2out.c (gen_entry_point_die): Hide prototype and definition.

	* except.h (doing_eh): Provide prototype.

	* expr.c: Move inclusion of expr.h to after insn-config.h.

	* final.c: Include reload.h.
	(shorten_branches): Cast the first argument of bzero to char *.

	* fix-header.c (cpp_print_containing_files): Mark parameter
 	`pfile' with ATTRIBUTE_UNUSED.
	(cpp_fatal): Likewise.

	* flow.c (find_basic_blocks_1): Cast the first argument of bzero
	to char *.

	* genattrtab.c (make_length_attrs): Change the type of variable
	`i' from int to size_t.
	(zero_fn): Mark parameter `exp' with ATTRIBUTE_UNUSED.
	(one_fn): Likewise.
	
	* genextract.c (main): When generating insn-extract.c, mark
	variable `junk' with ATTRIBUTE_UNUSED.

	* gengenrtl.c (gencode): When generating genrtl.c, cast the first
	argument of bzero to char*.

	* integrate.c: Include toplev.h.

	* libgcc2.c: Wrap `struct exception_table' and
	`find_exception_handler' in macro DWARF2_UNWIND_INFO.

	* objc/Make-lang.in (objc-act.o): Depend on toplev.h.
	
	* objc/objc-act.c: Include toplev.h.
	(lang_print_xnode): Mark parameters `file', `node' and `indent'
	with ATTRIBUTE_UNUSED.
	(finish_protocol): Likewise for parameter `protocol'.

	* output.h (declare_weak): Add prototype.
	(decode_reg_name): Don't wrap with TREE_CODE macro.
	(assemble_alias): Add prototype.

	* regclass.c: Include output.h.

	* reload.h (reloads_conflict): Add prototype.

	* rtl.h (print_rtl_single, mark_elimiation, reg_class_subset_p,
	output_func_start_profiler): Add prototypes.

	* rtlanal.c (reg_set_p_1): Mark parameters `x' and `pat' with
 	ATTRIBUTE_UNUSED.

	* scan-decls.c: Include scan.h.

	* scan.h (recognized_function, recognized_extern): Add prototypes.

	* stmt.c: Include output.h.

	* toplev.c (error_for_asm, warning_for_asm): Remove prototypes.
	(output_lang_identify): Hide prototype and definition.
	(float_signal): Mark parameter `signo' with ATTRIBUTE_UNUSED.
	(pipe_closed): Likewise.

	* toplev.h (count_error, strip_off_ending, error_for_asm,
 	warning_for_asm): Add prototypes.



diff -rup orig/egcs-19980508/gcc/Makefile.in egcs-19980508/gcc/Makefile.in
--- orig/egcs-19980508/gcc/Makefile.in	Fri May  8 02:51:03 1998
+++ egcs-19980508/gcc/Makefile.in	Sun May 10 19:40:43 1998
@@ -1254,9 +1254,10 @@ c-decl.o : c-decl.c $(CONFIG_H) system.h
     output.h toplev.h
 c-typeck.o : c-typeck.c $(CONFIG_H) system.h $(TREE_H) c-tree.h flags.h \
     output.h expr.h insn-codes.h $(RTL_H) toplev.h
-c-lang.o : c-lang.c $(CONFIG_H) system.h $(TREE_H)
+c-lang.o : c-lang.c $(CONFIG_H) system.h $(TREE_H) c-tree.h c-lex.h toplev.h
 c-lex.o : c-lex.c $(CONFIG_H) system.h $(TREE_H) $(RTL_H) c-lex.h c-tree.h \
-    $(srcdir)/c-parse.h input.h flags.h $(srcdir)/c-gperf.h c-pragma.h toplev.h
+    $(srcdir)/c-parse.h input.h flags.h $(srcdir)/c-gperf.h c-pragma.h \
+    toplev.h output.h
 c-aux-info.o : c-aux-info.c  $(CONFIG_H) system.h $(TREE_H) c-tree.h flags.h
 c-convert.o : c-convert.c $(CONFIG_H) system.h $(TREE_H) flags.h toplev.h
 c-pragma.o: c-pragma.c $(CONFIG_H) system.h $(TREE_H) except.h function.h \
@@ -1299,7 +1300,7 @@ s-under: $(GCC_PASSES)
 # A file used by all variants of C.
 
 c-common.o : c-common.c $(CONFIG_H) system.h $(TREE_H) c-tree.h c-lex.h \
-	flags.h toplev.h
+	flags.h toplev.h output.h
 
 # Language-independent files.
 
@@ -1354,7 +1355,7 @@ function.o : function.c $(CONFIG_H) syst
    insn-config.h $(RECOG_H) output.h toplev.h
 stmt.o : stmt.c $(CONFIG_H) system.h $(RTL_H) $(TREE_H) flags.h function.h  \
    insn-flags.h insn-config.h insn-codes.h hard-reg-set.h expr.h except.h \
-   loop.h $(RECOG_H) toplev.h
+   loop.h $(RECOG_H) toplev.h output.h
 except.o : except.c $(CONFIG_H) system.h $(RTL_H) $(TREE_H) flags.h \
    function.h insn-flags.h insn-codes.h expr.h regs.h hard-reg-set.h \
    insn-config.h $(RECOG_H) output.h except.h toplev.h
@@ -1362,7 +1363,7 @@ expr.o : expr.c $(CONFIG_H) system.h $(R
    regs.h insn-flags.h insn-codes.h expr.h insn-config.h $(RECOG_H) output.h \
    typeclass.h hard-reg-set.h toplev.h
 calls.o : calls.c $(CONFIG_H) system.h $(RTL_H) $(TREE_H) flags.h expr.h \
-   insn-codes.h insn-flags.h regs.h toplev.h
+   insn-codes.h insn-flags.h regs.h toplev.h output.h
 expmed.o : expmed.c $(CONFIG_H) system.h $(RTL_H) $(TREE_H) flags.h  \
    insn-flags.h insn-config.h insn-codes.h expr.h $(RECOG_H) real.h
 explow.o : explow.c $(CONFIG_H) system.h $(RTL_H) $(TREE_H) flags.h \
@@ -1389,7 +1390,7 @@ getpwd.o : getpwd.c $(CONFIG_H) system.h
 
 integrate.o : integrate.c $(CONFIG_H) system.h $(RTL_H) $(TREE_H) flags.h \
    integrate.h insn-flags.h insn-config.h insn-codes.h expr.h real.h regs.h \
-   function.h output.h $(RECOG_H) except.h
+   function.h output.h $(RECOG_H) except.h toplev.h
 
 jump.o : jump.c $(CONFIG_H) system.h $(RTL_H) flags.h hard-reg-set.h regs.h \
    insn-config.h insn-codes.h insn-flags.h $(RECOG_H) expr.h real.h except.h
@@ -1409,7 +1410,8 @@ combine.o : combine.c $(CONFIG_H) system
    insn-config.h insn-flags.h insn-codes.h insn-attr.h regs.h expr.h \
    $(BASIC_BLOCK_H) $(RECOG_H) real.h hard-reg-set.h
 regclass.o : regclass.c $(CONFIG_H) system.h $(RTL_H) hard-reg-set.h flags.h \
-   $(BASIC_BLOCK_H) regs.h insn-config.h $(RECOG_H) reload.h real.h toplev.h
+   $(BASIC_BLOCK_H) regs.h insn-config.h $(RECOG_H) reload.h real.h toplev.h \
+   output.h
 local-alloc.o : local-alloc.c $(CONFIG_H) system.h $(RTL_H) flags.h \
    $(BASIC_BLOCK_H) regs.h hard-reg-set.h insn-config.h $(RECOG_H) output.h
 bitmap.o : bitmap.c $(CONFIG_H) system.h $(RTL_H) flags.h $(BASIC_BLOCK_H) \
@@ -1439,7 +1441,7 @@ $(SCHED_PREFIX)sched.o : $(SCHED_PREFIX)
 final.o : final.c $(CONFIG_H) system.h $(RTL_H) $(TREE_H) flags.h regs.h \
    $(RECOG_H) conditions.h insn-config.h insn-attr.h except.h real.h output.h \
    hard-reg-set.h insn-flags.h insn-codes.h gstab.h xcoffout.h defaults.h \
-   toplev.h
+   toplev.h reload.h
 recog.o : recog.c $(CONFIG_H) system.h $(RTL_H)  \
    regs.h $(RECOG_H) hard-reg-set.h flags.h insn-config.h insn-attr.h \
    insn-flags.h insn-codes.h real.h
diff -rup orig/egcs-19980508/gcc/c-common.c egcs-19980508/gcc/c-common.c
--- orig/egcs-19980508/gcc/c-common.c	Wed May  6 17:06:41 1998
+++ egcs-19980508/gcc/c-common.c	Sun May 10 18:12:01 1998
@@ -26,6 +26,7 @@ Boston, MA 02111-1307, USA.  */
 #include "flags.h"
 #include "obstack.h"
 #include "toplev.h"
+#include "output.h"
 
 #ifndef WCHAR_TYPE_SIZE
 #ifdef INT_TYPE_SIZE
@@ -1235,7 +1236,6 @@ check_format_info (info, params)
   int length_char;
   int format_char;
   int format_length;
-  int integral_format;
   tree format_tree;
   tree cur_param;
   tree cur_type;
diff -rup orig/egcs-19980508/gcc/c-decl.c egcs-19980508/gcc/c-decl.c
--- orig/egcs-19980508/gcc/c-decl.c	Wed May  6 17:06:42 1998
+++ egcs-19980508/gcc/c-decl.c	Sun May 10 18:02:39 1998
@@ -808,17 +808,17 @@ c_decode_option (p)
 
 void
 print_lang_decl (file, node, indent)
-     FILE *file;
-     tree node;
-     int indent;
+     FILE *file ATTRIBUTE_UNUSED;
+     tree node ATTRIBUTE_UNUSED;
+     int indent ATTRIBUTE_UNUSED;
 {
 }
 
 void
 print_lang_type (file, node, indent)
-     FILE *file;
-     tree node;
-     int indent;
+     FILE *file ATTRIBUTE_UNUSED;
+     tree node ATTRIBUTE_UNUSED;
+     int indent ATTRIBUTE_UNUSED;
 {
 }
 
@@ -4083,7 +4083,7 @@ finish_decl (decl, init, asmspec_tree)
 
 tree
 maybe_build_cleanup (decl)
-     tree decl;
+     tree decl ATTRIBUTE_UNUSED;
 {
   /* There are no cleanups in C.  */
   return NULL_TREE;
@@ -7296,6 +7296,6 @@ pop_c_function_context ()
 
 void
 copy_lang_decl (node)
-     tree node;
+     tree node ATTRIBUTE_UNUSED;
 {
 }
diff -rup orig/egcs-19980508/gcc/c-lang.c egcs-19980508/gcc/c-lang.c
--- orig/egcs-19980508/gcc/c-lang.c	Fri Apr  3 11:33:20 1998
+++ egcs-19980508/gcc/c-lang.c	Sun May 10 17:47:41 1998
@@ -23,6 +23,9 @@ Boston, MA 02111-1307, USA.  */
 #include "system.h"
 #include "tree.h"
 #include "input.h"
+#include "c-tree.h"
+#include "c-lex.h"
+#include "toplev.h"
 
 /* Each of the functions defined here
    is an alternative to a function in objc-actions.c.  */
@@ -65,9 +68,9 @@ print_lang_statistics ()
 
 void
 lang_print_xnode (file, node, indent)
-     FILE *file;
-     tree node;
-     int indent;
+     FILE *file ATTRIBUTE_UNUSED;
+     tree node ATTRIBUTE_UNUSED;
+     int indent ATTRIBUTE_UNUSED;
 {
 }
 
@@ -75,35 +78,36 @@ lang_print_xnode (file, node, indent)
 
 tree
 lookup_interface (arg)
-     tree arg;
+     tree arg ATTRIBUTE_UNUSED;
 {
   return 0;
 }
 
 tree
 is_class_name (arg)
-    tree arg;
+    tree arg ATTRIBUTE_UNUSED;
 {
   return 0;
 }
 
 void
 maybe_objc_check_decl (decl)
-     tree decl;
+     tree decl ATTRIBUTE_UNUSED;
 {
 }
 
 int
 maybe_objc_comptypes (lhs, rhs, reflexive)
-     tree lhs, rhs;
-     int reflexive;
+     tree lhs ATTRIBUTE_UNUSED;
+     tree rhs ATTRIBUTE_UNUSED;
+     int reflexive ATTRIBUTE_UNUSED;
 {
   return -1;
 }
 
 tree
 maybe_objc_method_name (decl)
-    tree decl;
+    tree decl ATTRIBUTE_UNUSED;
 {
   return 0;
 }
@@ -122,8 +126,8 @@ recognize_objc_keyword ()
 
 tree
 build_objc_string (len, str)
-    int len;
-    char *str;
+    int len ATTRIBUTE_UNUSED;
+    char *str ATTRIBUTE_UNUSED;
 {
   abort ();
   return NULL_TREE;
diff -rup orig/egcs-19980508/gcc/c-lex.c egcs-19980508/gcc/c-lex.c
--- orig/egcs-19980508/gcc/c-lex.c	Wed May  6 17:06:43 1998
+++ egcs-19980508/gcc/c-lex.c	Sun May 10 17:55:40 1998
@@ -25,6 +25,7 @@ Boston, MA 02111-1307, USA.  */
 #include "rtl.h"
 #include "tree.h"
 #include "input.h"
+#include "output.h"
 #include "c-lex.h"
 #include "c-tree.h"
 #include "flags.h"
diff -rup orig/egcs-19980508/gcc/c-lex.h egcs-19980508/gcc/c-lex.h
--- orig/egcs-19980508/gcc/c-lex.h	Wed Apr  8 05:57:52 1998
+++ egcs-19980508/gcc/c-lex.h	Sun May 10 17:33:57 1998
@@ -75,7 +75,7 @@ extern char *token_buffer;	/* Pointer to
 
 extern tree make_pointer_declarator PROTO((tree, tree));
 extern void reinit_parse_for_function PROTO((void));
-extern void position_after_whitespace PROTO((void));
+extern void position_after_white_space PROTO((void));
 extern int check_newline PROTO((void));
 
 extern int yylex PROTO((void));
diff -rup orig/egcs-19980508/gcc/c-tree.h egcs-19980508/gcc/c-tree.h
--- orig/egcs-19980508/gcc/c-tree.h	Tue Apr 21 09:25:56 1998
+++ egcs-19980508/gcc/c-tree.h	Sun May 10 18:00:47 1998
@@ -149,6 +149,7 @@ struct lang_type
 extern tree lookup_interface			PROTO((tree));
 extern tree is_class_name			PROTO((tree));
 extern void maybe_objc_check_decl		PROTO((tree));
+extern void finish_file				PROTO((void));
 extern int maybe_objc_comptypes                 PROTO((tree, tree, int));
 extern tree maybe_building_objc_message_expr    PROTO((void));
 extern tree maybe_objc_method_name		PROTO((tree));
@@ -167,6 +168,9 @@ extern void check_function_format		PROTO
    NOP_EXPR is used as a special case (see truthvalue_conversion).  */
 extern void binary_op_error                     PROTO((enum tree_code));
 extern void c_expand_expr_stmt                  PROTO((tree));
+extern void c_expand_start_cond                 PROTO((tree, int, int));
+extern void c_expand_start_else                 PROTO((void));
+extern void c_expand_end_cond                   PROTO((void));
 /* Validate the expression after `case' and apply default promotions.  */
 extern tree check_case_value                    PROTO((tree));
 /* Concatenate a list of STRING_CST nodes into one STRING_CST.  */
@@ -366,6 +370,7 @@ extern void c_expand_return			PROTO((tre
 extern tree c_expand_start_case                 PROTO((tree));
 
 /* in c-iterate.c */
+extern void init_iterators			PROTO((void));
 extern void iterator_expand			PROTO((tree));
 extern void iterator_for_loop_start		PROTO((tree));
 extern void iterator_for_loop_end		PROTO((tree));
diff -rup orig/egcs-19980508/gcc/caller-save.c egcs-19980508/gcc/caller-save.c
--- orig/egcs-19980508/gcc/caller-save.c	Fri Apr  3 11:33:33 1998
+++ egcs-19980508/gcc/caller-save.c	Sun May 10 19:30:55 1998
@@ -505,7 +505,8 @@ save_call_clobbered_regs (insn_mode)
 
 static void
 set_reg_live (reg, setter)
-     rtx reg, setter;
+     rtx reg;
+     rtx setter ATTRIBUTE_UNUSED;
 {
   register int regno, endregno, i;
   enum machine_mode mode = GET_MODE (reg);
diff -rup orig/egcs-19980508/gcc/calls.c egcs-19980508/gcc/calls.c
--- orig/egcs-19980508/gcc/calls.c	Wed May  6 00:45:47 1998
+++ egcs-19980508/gcc/calls.c	Sun May 10 18:29:08 1998
@@ -32,6 +32,7 @@ Boston, MA 02111-1307, USA.  */
 #include "regs.h"
 #include "insn-flags.h"
 #include "toplev.h"
+#include "output.h"
 
 /* Decide whether a function's arguments should be processed
    from first to last or from last to first.
diff -rup orig/egcs-19980508/gcc/cccp.c egcs-19980508/gcc/cccp.c
--- orig/egcs-19980508/gcc/cccp.c	Thu May  7 07:44:01 1998
+++ egcs-19980508/gcc/cccp.c	Sun May 10 17:25:49 1998
@@ -10224,7 +10224,7 @@ pfatal_with_name (name)
 static void
 pipe_closed (signo)
      /* If this is missing, some compilers complain.  */
-     int signo;
+     int signo ATTRIBUTE_UNUSED;
 {
   fatal ("output pipe has been closed");
 }
diff -rup orig/egcs-19980508/gcc/combine.c egcs-19980508/gcc/combine.c
--- orig/egcs-19980508/gcc/combine.c	Thu Apr 16 19:56:01 1998
+++ egcs-19980508/gcc/combine.c	Sun May 10 18:48:29 1998
@@ -88,9 +88,10 @@ Boston, MA 02111-1307, USA.  */
 #include "flags.h"
 #include "regs.h"
 #include "hard-reg-set.h"
-#include "expr.h"
 #include "basic-block.h"
 #include "insn-config.h"
+/* Include expr.h after insn-config.h so we get HAVE_conditional_move. */
+#include "expr.h"
 #include "insn-flags.h"
 #include "insn-codes.h"
 #include "insn-attr.h"