[rfc] gdb_stdint.h.

[rfc] gdb_stdint.h.

[Daniel Jacobowitz]

You want uintptr_t?  I got your uintptr_t right here.

This patch creates a header in the build directory named gdb_stdint.h.
It defines all of the things which a stdint.h ought to define.  It uses
the system's stdint.h if there is one, and provides anything missing,
up to and including the whole file.

There is, of course, a huge hairy chunk of m4 and portability knowledge
that you don't see in this patch.  That's because it's already in
config/ and was written for GCC.  I'm OK with that :-)  It means it's
quite well tested at this point.

This would be enough to allow us to start using uint64_t, uintptr_t,
et cetera in the common code of GDB.  Which, in my opinion, would be a
good idea.  They're nice to have around.

Any comments?

-- 
Daniel Jacobowitz
CodeSourcery

2006-05-05  Daniel Jacobowitz  <dan@codesourcery.com>

	* configure.ac: Use GCC_HEADER_STDINT.
	* acinclude.m4: Include stdint.m4.
	* Makefile.in (gdb_stdint_h): Define.
	(distclean): Remove gdb_stdint.h.
	(Makefile, stamp-h): Update rules to generate only the correct
	files.
	(gdb_stdint.h, stamp-int): New rules.
	* config.in, configure: Regenerated.

Index: configure.ac
===================================================================
RCS file: /cvs/src/src/gdb/configure.ac,v
retrieving revision 1.31
diff -u -p -r1.31 configure.ac
--- configure.ac	29 Mar 2006 22:58:54 -0000	1.31
+++ configure.ac	5 May 2006 17:23:04 -0000
@@ -422,6 +422,9 @@ AC_CHECK_HEADERS(term.h, [], [],
 # unconditionally, so what's the point in checking these?
 AC_CHECK_HEADERS(ctype.h time.h)
 
+# Create a header we can use portably to get the standard integer types.
+GCC_HEADER_STDINT(gdb_stdint.h)
+
 # ------------------------- #
 # Checks for declarations.  #
 # ------------------------- #
Index: acinclude.m4
===================================================================
RCS file: /cvs/src/src/gdb/acinclude.m4,v
retrieving revision 1.13
diff -u -p -r1.13 acinclude.m4
--- acinclude.m4	17 Dec 2005 22:33:59 -0000	1.13
+++ acinclude.m4	5 May 2006 17:23:05 -0000
@@ -7,6 +7,9 @@ sinclude(../bfd/bfd.m4)
 dnl This gets the standard macros, like the TCL, TK, etc ones.
 sinclude(../config/acinclude.m4)
 
+dnl This gets GCC_HEADER_STDINT.
+sinclude(../config/stdint.m4)
+
 sinclude(../gettext.m4)
 dnl The lines below arrange for aclocal not to bring gettext.m4's
 dnl CY_GNU_GETTEXT into aclocal.m4.
Index: Makefile.in
===================================================================
RCS file: /cvs/src/src/gdb/Makefile.in,v
retrieving revision 1.811
diff -u -p -r1.811 Makefile.in
--- Makefile.in	23 Apr 2006 14:15:01 -0000	1.811
+++ Makefile.in	5 May 2006 17:23:05 -0000
@@ -613,6 +613,7 @@ hashtab_h =	$(INCLUDE_DIR)/hashtab.h
 config_h = config.h
 exc_request_U_h = exc_request_U.h
 exc_request_S_h = exc_request_S.h
+gdb_stdint_h = gdb_stdint.h
 msg_reply_S_h = msg_reply_S.h
 msg_U_h = msg_U.h
 notify_S_h = notify_S.h
@@ -1280,6 +1281,7 @@ distclean: clean
 	rm -f gdbserver/tm.h gdbserver/xm.h gdbserver/nm.h
 	rm -f gdbserver/Makefile gdbserver/config.cache
 	rm -f nm.h tm.h xm.h config.status config.h stamp-h .gdbinit
+	rm -f gdb_stdint.h
 	rm -f y.output yacc.acts yacc.tmp y.tab.h
 	rm -f config.log config.cache
 	rm -f Makefile
@@ -1317,11 +1319,28 @@ subdir_do: force
 	done
 
 Makefile: Makefile.in config.status @frags@
-	$(SHELL) config.status
+	# Regenerate the Makefile and the tm.h / nm.h links.
+	CONFIG_FILES=Makefile \
+	  CONFIG_COMMANDS= \
+	  CONFIG_HEADERS= \
+	  $(SHELL) config.status
 
 config.h: stamp-h ; @true
 stamp-h: config.in config.status
-	CONFIG_HEADERS=config.h:config.in $(SHELL) config.status
+	CONFIG_HEADERS=config.h:config.in \
+	  CONFIG_COMMANDS=default \
+	  CONFIG_FILES= \
+	  CONFIG_LINKS= \
+	  $(SHELL) config.status
+
+gdb_stdint.h: stamp-int ; @true
+stamp-int: config.status
+	CONFIG_COMMANDS=gdb_stdint.h \
+	  CONFIG_FILES= \
+	  CONFIG_HEADERS= \
+	  CONFIG_LINKS= \
+	  $(SHELL) config.status
+	echo stamp > stamp-int
 
 config.status: configure configure.tgt configure.host
 	$(SHELL) config.status --recheck

Re: [rfc] gdb_stdint.h.

[Jim Blandy]

Daniel Jacobowitz <drow@false.org> writes:
> You want uintptr_t?  I got your uintptr_t right here.
>
> This patch creates a header in the build directory named gdb_stdint.h.
> It defines all of the things which a stdint.h ought to define.  It uses
> the system's stdint.h if there is one, and provides anything missing,
> up to and including the whole file.
>
> There is, of course, a huge hairy chunk of m4 and portability knowledge
> that you don't see in this patch.  That's because it's already in
> config/ and was written for GCC.  I'm OK with that :-)  It means it's
> quite well tested at this point.
>
> This would be enough to allow us to start using uint64_t, uintptr_t,
> et cetera in the common code of GDB.  Which, in my opinion, would be a
> good idea.  They're nice to have around.

Oh, I double-plus-good-feeling <stdint.h>.  Criminal not to have it in
the language from the beginning.

Re: [rfc] gdb_stdint.h.

[Eli Zaretskii]

> From: Jim Blandy <jimb@codesourcery.com>
> Date: Fri, 05 May 2006 12:32:17 -0700
> 
> Oh, I double-plus-good-feeling <stdint.h>.  Criminal not to have it in
> the language from the beginning.

In the beginning, as you well know, there was only one int.

Re: [rfc] gdb_stdint.h.

[Michael Snyder]

Eli Zaretskii wrote:
>>From: Jim Blandy <jimb@codesourcery.com>
>>Date: Fri, 05 May 2006 12:32:17 -0700
>>
>>Oh, I double-plus-good-feeling <stdint.h>.  Criminal not to have it in
>>the language from the beginning.
> 
> 
> In the beginning, as you well know, there was only one int.
> 

And the int said, let there be *.
And there was *.

Re: [rfc] gdb_stdint.h.

[Daniel Jacobowitz]

On Fri, May 05, 2006 at 01:27:13PM -0400, Daniel Jacobowitz wrote:
> You want uintptr_t?  I got your uintptr_t right here.
> 
> This patch creates a header in the build directory named gdb_stdint.h.
> It defines all of the things which a stdint.h ought to define.  It uses
> the system's stdint.h if there is one, and provides anything missing,
> up to and including the whole file.
> 
> There is, of course, a huge hairy chunk of m4 and portability knowledge
> that you don't see in this patch.  That's because it's already in
> config/ and was written for GCC.  I'm OK with that :-)  It means it's
> quite well tested at this point.
> 
> This would be enough to allow us to start using uint64_t, uintptr_t,
> et cetera in the common code of GDB.  Which, in my opinion, would be a
> good idea.  They're nice to have around.
> 
> Any comments?

> 2006-05-05  Daniel Jacobowitz  <dan@codesourcery.com>
> 
> 	* configure.ac: Use GCC_HEADER_STDINT.
> 	* acinclude.m4: Include stdint.m4.
> 	* Makefile.in (gdb_stdint_h): Define.
> 	(distclean): Remove gdb_stdint.h.
> 	(Makefile, stamp-h): Update rules to generate only the correct
> 	files.
> 	(gdb_stdint.h, stamp-int): New rules.
> 	* config.in, configure: Regenerated.

Comments sounded supportive to me!  I have committed it.

Debian unstable appears to have gone up to autoconf 2.59d, so configure
has changed a bit with this commit; I think there won't be any
problems that affect GDB, but please do let me know if I'm wrong.



-- 
Daniel Jacobowitz
CodeSourcery