6.1: Quote macro names being defined by AC_DEFUN

6.1: Quote macro names being defined by AC_DEFUN

[Maciej W. Rozycki]

Hello,

 As of automake 1.8, aclocal requires all macro names being defined by
AC_DEFUN to be quoted.  Here's an obvious fix.

2004-04-19  Maciej W. Rozycki  <macro@ds2.pg.gda.pl>

	* acinclude.m4: Quote macro names being defined by AC_DEFUN 
	througout.

 Please apply.

  Maciej

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@ds2.pg.gda.pl, PGP key available        +

gdb-6.1-am18.patch
diff -up --recursive --new-file gdb-6.1.macro/gdb/acinclude.m4 gdb-6.1/gdb/acinclude.m4
--- gdb-6.1.macro/gdb/acinclude.m4	2003-06-08 18:27:12.000000000 +0000
+++ gdb-6.1/gdb/acinclude.m4	2004-04-17 09:12:42.000000000 +0000
@@ -8,7 +8,7 @@ dnl This gets the standard macros, like 
 sinclude(../config/acinclude.m4)
 
 dnl CYGNUS LOCAL: This gets the right posix flag for gcc
-AC_DEFUN(CY_AC_TCL_LYNX_POSIX,
+AC_DEFUN([CY_AC_TCL_LYNX_POSIX],
 [AC_REQUIRE([AC_PROG_CC])AC_REQUIRE([AC_PROG_CPP])
 AC_MSG_CHECKING([if running LynxOS])
 AC_CACHE_VAL(ac_cv_os_lynx,
@@ -48,7 +48,7 @@ fi
 # makes configure think it's cross compiling. If --target wasn't used, then
 # we can't configure, so something is wrong. We don't use the cache
 # here cause if somebody fixes their compiler install, we want this to work.
-AC_DEFUN(CY_AC_C_WORKS,
+AC_DEFUN([CY_AC_C_WORKS],
 [# If we cannot compile and link a trivial program, we can't expect anything to work
 AC_MSG_CHECKING(whether the compiler ($CC) actually works)
 AC_TRY_COMPILE(, [/* don't need anything here */],
@@ -67,7 +67,7 @@ fi
 AC_MSG_RESULT(yes)
 ])
 
-AC_DEFUN(CY_AC_PATH_TCLH, [
+AC_DEFUN([CY_AC_PATH_TCLH], [
 #
 # Ok, lets find the tcl source trees so we can use the headers
 # Warning: transition of version 9 to 10 will break this algorithm
@@ -156,7 +156,7 @@ AC_SUBST(TCLHDIR)
 ])
 
 
-AC_DEFUN(CY_AC_PATH_TCLCONFIG, [
+AC_DEFUN([CY_AC_PATH_TCLCONFIG], [
 #
 # Ok, lets find the tcl configuration
 # First, look for one uninstalled.  
@@ -229,7 +229,7 @@ fi
 
 # Defined as a separate macro so we don't have to cache the values
 # from PATH_TCLCONFIG (because this can also be cached).
-AC_DEFUN(CY_AC_LOAD_TCLCONFIG, [
+AC_DEFUN([CY_AC_LOAD_TCLCONFIG], [
     . $TCLCONFIG
 
     AC_SUBST(TCL_VERSION)
@@ -285,7 +285,7 @@ dnl    AC_SUBST(TCL_UNSHARED_LIB_SUFFIX)
 #  - Symbols in tkConfig.sh are different than tclConfig.sh
 #  - Acceptable for Tk to be missing but not Tcl.
 
-AC_DEFUN(CY_AC_PATH_TKH, [
+AC_DEFUN([CY_AC_PATH_TKH], [
 #
 # Ok, lets find the tk source trees so we can use the headers
 # If the directory (presumably symlink) named "tk" exists, use that one
@@ -379,7 +379,7 @@ AC_SUBST(TKHDIR)
 ])
 
 
-AC_DEFUN(CY_AC_PATH_TKCONFIG, [
+AC_DEFUN([CY_AC_PATH_TKCONFIG], [
 #
 # Ok, lets find the tk configuration
 # First, look for one uninstalled.  
@@ -453,7 +453,7 @@ fi
 
 # Defined as a separate macro so we don't have to cache the values
 # from PATH_TKCONFIG (because this can also be cached).
-AC_DEFUN(CY_AC_LOAD_TKCONFIG, [
+AC_DEFUN([CY_AC_LOAD_TKCONFIG], [
     if test -f "$TKCONFIG" ; then
       . $TKCONFIG
     fi
@@ -484,7 +484,7 @@ dnl    AC_SUBST(TK_EXEC_PREFIX)
 
 # check for Itcl headers. 
 
-AC_DEFUN(CY_AC_PATH_ITCLCONFIG, [
+AC_DEFUN([CY_AC_PATH_ITCLCONFIG], [
 #
 # Ok, lets find the itcl configuration
 # First, look for one uninstalled.  
@@ -557,7 +557,7 @@ fi
 
 # Defined as a separate macro so we don't have to cache the values
 # from PATH_ITCLCONFIG (because this can also be cached).
-AC_DEFUN(CY_AC_LOAD_ITCLCONFIG, [
+AC_DEFUN([CY_AC_LOAD_ITCLCONFIG], [
     if test -f "$ITCLCONFIG" ; then
       . $ITCLCONFIG
     fi
@@ -586,7 +586,7 @@ dnl    AC_SUBST(ITCL_EXEC_PREFIX)
 
 # check for Itcl headers. 
 
-AC_DEFUN(CY_AC_PATH_ITCLH, [
+AC_DEFUN([CY_AC_PATH_ITCLH], [
 AC_MSG_CHECKING(for Itcl private headers. srcdir=${srcdir})
 if test x"${ac_cv_c_itclh}" = x ; then
   for i in ${srcdir}/../itcl ${srcdir}/../../itcl ${srcdir}/../../../itcl ${srcdir}/../itcl/itcl; do
@@ -610,7 +610,7 @@ AC_SUBST(ITCLHDIR)
 ])
 
 
-AC_DEFUN(CY_AC_PATH_ITKCONFIG, [
+AC_DEFUN([CY_AC_PATH_ITKCONFIG], [
 #
 # Ok, lets find the itk configuration
 # First, look for one uninstalled.  
@@ -684,7 +684,7 @@ fi
 
 # Defined as a separate macro so we don't have to cache the values
 # from PATH_ITKCONFIG (because this can also be cached).
-AC_DEFUN(CY_AC_LOAD_ITKCONFIG, [
+AC_DEFUN([CY_AC_LOAD_ITKCONFIG], [
     if test -f "$ITKCONFIG" ; then
       . $ITKCONFIG
     fi
@@ -711,7 +711,7 @@ dnl    AC_SUBST(ITK_EXEC_PREFIX)
     AC_SUBST(ITK_LIB_SPEC)
 ])
 
-AC_DEFUN(CY_AC_PATH_ITKH, [
+AC_DEFUN([CY_AC_PATH_ITKH], [
 AC_MSG_CHECKING(for Itk private headers. srcdir=${srcdir})
 if test x"${ac_cv_c_itkh}" = x ; then
   for i in ${srcdir}/../itcl ${srcdir}/../../itcl ${srcdir}/../../../itcl ${srcdir}/../itcl/itk; do
@@ -949,7 +949,7 @@ dnl to use a different cache variable na
 dnl in a different context somewhere else.
 dnl gcc_AC_CHECK_DECL(SYMBOL,
 dnl 	[ACTION-IF-FOUND [, ACTION-IF-NOT-FOUND [, INCLUDES]]])
-AC_DEFUN(gcc_AC_CHECK_DECL,
+AC_DEFUN([gcc_AC_CHECK_DECL],
 [AC_MSG_CHECKING([whether $1 is declared])
 AC_CACHE_VAL(gcc_cv_have_decl_$1,
 [AC_TRY_COMPILE([$4],
@@ -967,7 +967,7 @@ dnl Check multiple functions to see whet
 dnl Arrange to define HAVE_DECL_<FUNCTION> to 0 or 1 as appropriate.
 dnl gcc_AC_CHECK_DECLS(SYMBOLS,
 dnl 	[ACTION-IF-NEEDED [, ACTION-IF-NOT-NEEDED [, INCLUDES]]])
-AC_DEFUN(gcc_AC_CHECK_DECLS,
+AC_DEFUN([gcc_AC_CHECK_DECLS],
 [for ac_func in $1
 do
 changequote(, )dnl

Re: 6.1: Quote macro names being defined by AC_DEFUN

[Andrew Cagney]

Daniel, were you looking at upgrading GDB to autoconf 2.5.x?

Andrew

Re: 6.1: Quote macro names being defined by AC_DEFUN

[Daniel Jacobowitz]

On Thu, Apr 22, 2004 at 03:54:27PM -0400, Andrew Cagney wrote:
> Daniel, were you looking at upgrading GDB to autoconf 2.5.x?

I've been looking at it on and off, but not working on it - not until I
get done with binutils.  I'll need to touch GDB for a change to
top-level ylwrap but I have been putting that off until I have more
time to test it.

Maciej's patch looks both right and safe to me.  If no one objects,
I can check it in tomorrow - or, since Maciej has an account for
binutils CVS, I imagine we could add him as write-after-approval
assuming he has an assignment on file that covers GDB.

[If not, I think this patch is sufficiently obvious.  Does anyone
disagree?]

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: 6.1: Quote macro names being defined by AC_DEFUN

[Jim Blandy]

Daniel Jacobowitz <drow@mvista.com> writes:
> Maciej's patch looks both right and safe to me.  If no one objects,
> I can check it in tomorrow - or, since Maciej has an account for
> binutils CVS, I imagine we could add him as write-after-approval
> assuming he has an assignment on file that covers GDB.

He does have an assignment on file for GDB.

Re: 6.1: Quote macro names being defined by AC_DEFUN

[Maciej W. Rozycki]

On Thu, 22 Apr 2004, Daniel Jacobowitz wrote:

> > Daniel, were you looking at upgrading GDB to autoconf 2.5.x?
> 
> I've been looking at it on and off, but not working on it - not until I
> get done with binutils.  I'll need to touch GDB for a change to
> top-level ylwrap but I have been putting that off until I have more
> time to test it.

 With the other patch (config-subdirs) I've proposed her recently gdb/
appears to be safe for use with autoconf 2.59, i.e. there are no errors
from autoconf and the configuration output seems correct.  But I did
limited testing only (mipsel-linux and i386-linux targets only) and with
the top-level converted to autoconf 2.59, too.  Additional testing would
be definitely desirable.

 The sim/ directory requires additional patches, which I've implemented
and made available.  Unfortunately it also hits a nasty bug in autoconf
2.58/2.59 (and likely older versions), which I'm told "would require
considerable poking around in m4sugar.m4 to fix."  I have a workaround for 
the bug for sim/, but given the current setup works, perhaps it's better 
to wait for autoconf to be fixed.

 This (am18) patch is orthogonal to autoconf -- it's required to shut up
aclocal (which promises to bail out in this case in a future version), and
it's just a sanitization of macro definitions.  The quoting was always
valid here and perhaps recommended for autoconf 2.13 vintage tools, too.  
It wasn't forced back then, but the tools were much less picky about
unsafe constructs in general.

  Maciej

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@ds2.pg.gda.pl, PGP key available        +