Update of /cvsroot/monetdb/buildtools/conf
In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv6458/conf

Modified Files:
        MonetDB.m4 
Log Message:
propagated changes of Tuesday Nov 20 2007 - Friday Nov 23 2007
from the MonetDB_1-20 branch to the development trunk


Index: MonetDB.m4
===================================================================
RCS file: /cvsroot/monetdb/buildtools/conf/MonetDB.m4,v
retrieving revision 1.48
retrieving revision 1.49
diff -u -d -r1.48 -r1.49
--- MonetDB.m4  9 Nov 2007 22:20:13 -0000       1.48
+++ MonetDB.m4  23 Nov 2007 11:38:32 -0000      1.49
@@ -465,59 +465,51 @@
                [obsolete: use --enable-bits instead]),
        AC_MSG_ERROR([argument --with-bits is obsolete: use --enable-bits 
instead]))
 
-bits="no"
+AC_CHECK_SIZEOF(long)
+native_bits=`expr $ac_cv_sizeof_long \* 8`
+bits=$native_bits
 AC_ARG_ENABLE(bits,
        AC_HELP_STRING([--enable-bits=BITS],
                [specify number of bits (32 or 64)]), [
-case $enableval in
-32)    case "$host" in
-       ia64*)  AC_MSG_ERROR([we do not support 32 bits on $host, yet]);;
-       esac
+case "$GCC-$CC-$enableval" in
+*-*-$native_bits)
        ;;
-64)    case "$host-$GCC-$CC" in
-       i?86*-*-*)  AC_MSG_ERROR([$host does not support 64 bits]);;
-       esac
+-*icc*-*)
+       AC_MSG_ERROR([icc does not support a command line option to switch 
between 32- & 64-bit compilation; please use CC & PATH (instead of 
--enable-bits=$enableval) to point to the $enableval-bit version of icc])
+       ;;
+*-*-32|*-*-64)
+       ;;
+*)     
+       AC_MSG_ERROR(--enable-bits argument must be either 32 or 64)
        ;;
-*)     AC_MSG_ERROR(--enable-bits argument must be either 32 or 64);;
 esac
 bits=$enableval
 ])
-if test "$bits" = "64"; then
-       dnl  Keep in mind how to call the 32-bit compiler.
-       case "$GCC-$CC-$host_os-$host" in
-       yes-*-linux*-x86_64*)
-               dnl  On our x86_64 machine, "gcc" defaults to "gcc -m64" ...
-               CC32="$CC -m32";;
-       -pgcc*-linux*-x86_64*)
-               dnl  On our x86_64 machine, "pgcc" defaults to "pgcc -tp=k8-64" 
...
-               CC32="$CC -tp=k8-32";;
-       *)      CC32="$CC";;
-       esac
-fi
-if test x"$bits" != xno; then
 case "$GCC-$CC-$host_os-$host-$bits" in
-yes-*-solaris*-64)
-       case `$CC -v 2>&1` in
-       *'gcc version 3.'*)     ;;
+*-*-*-*-$native_bits)  
+       ;;
+yes-*-solaris*-*-*)
+       case `$bits-$CC -v 2>&1` in
+       32-*|*-*'gcc version 3.'*)      ;;
        *)      AC_MSG_ERROR([need GCC version 3.X for 64 bits]);;
        esac
        CC="$CC -m$bits"
        ;;
--*-solaris*-64)
+-*-solaris*-*-64)
        CC="$CC -xarch=v9"
        ;;
-yes-*-irix*-64)
+yes-*-irix*-*-*)
        CC="$CC -mabi=$bits"
        ;;
--*-irix*-64)
+-*-irix*-*-*)
        CC="$CC -$bits"
        ;;
-yes-*-aix*-64)
+yes-*-aix*-*-64)
        CC="$CC -maix$bits"
        AR="ar -X64"
        NM="nm -X64 -B"
        ;;
--*-aix*-64)
+-*-aix*-*-64)
        CC="$CC -q$bits"
        AR="ar -X64"
        NM="nm -X64 -B"
@@ -525,16 +517,23 @@
 yes-*-linux*-x86_64*-*)
        CC="$CC -m$bits"
        ;;
--pgcc*-linux*-x86_64*-*)
+-*pgcc*-linux*-x86_64*-*)
        CC="$CC -tp=k8-$bits"
        ;;
 yes-*-darwin8*-powerpc*-*)
        CC="$CC -m$bits"
        ;;
+*)
+       AC_MSG_ERROR([case "$GCC-$CC-$host_os-$host-$bits" not handled with 
--enable-bits, yet; please use CC, CFLAGS and/or LDFLAGS (instead of 
--enable-bits=$bits) to make "$CC" produce $bits-bit code])
+       ;;
 esac
-else
+if test x"$bits" != x"$native_bits"; then
+       unset ac_cv_sizeof_long
        AC_CHECK_SIZEOF(long)
-       bits=`expr $ac_cv_sizeof_long \* 8`
+       alien_bits=`expr $ac_cv_sizeof_long \* 8`
+       if test x"$alien_bits" != x"$bits"; then
+               AC_MSG_ERROR([CC="$CC" seems to produce $alien_bits-bit code, 
not matching the "--enable-bits=$bits" request])
+       fi
 fi
 AC_SUBST(bits)
 
@@ -618,7 +617,7 @@
                ;;
        esac
        ;;
--icc*-linux*|-ecc*-linux*|-pgcc*-linux*)
+-*icc*-linux*|-*ecc*-linux*|-*pgcc*-linux*)
        dnl Define the same settings as for gcc, as we use the same
        dnl header files
        AC_DEFINE(_POSIX_C_SOURCE, 200112L, [Compiler flag])
@@ -760,7 +759,7 @@
 
        NO_INLINE_CFLAGS="-fno-inline -fno-inline-functions"
        ;;
--icc*-linux*|-ecc*-linux*)
+-*icc*-linux*|-*ecc*-linux*)
        dnl  Intel ([ie]cc/[ie]cpc on Linux)
        LDFLAGS="$LDFLAGS -i_dynamic"
        dnl  Let warning #140 "too many arguments in function call"
@@ -840,7 +839,7 @@
 
        NO_INLINE_CFLAGS="-fno-inline -fno-inline-functions"
        ;;
--pgcc*-linux*)
+-*pgcc*-linux*)
        dnl  Portland Group (PGI) (pgcc/pgCC on Linux)
        dnl  required for "scale" in module "decimal"
        CFLAGS="$CFLAGS -Msignextend"
@@ -892,13 +891,13 @@
                ;;
        esac
        ;;
--icc*-linux*|-ecc*-linux*)
+-*icc*-linux*|-*ecc*-linux*)
        case "$host-$icc_ver" in
         *-*-*-10.*)    CFLAGS="$CFLAGS -std=c99"       ;;
        *-*-*-*)        CFLAGS="$CFLAGS -c99"           ;;
        esac   
        ;;
--pgcc*-linux*)
+-*pgcc*-linux*)
        CFLAGS="$CFLAGS -c9x"
        ;;
 esac


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Monetdb-checkins mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-checkins

Reply via email to