You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@apr.apache.org by Jeff Trawick <tr...@gmail.com> on 2010/02/04 21:46:48 UTC
q about libtool used in apr 1.3.9
I have libtool 1.5.26 on Linux (seems to be same version as used for
1.3.9) but I see this difference in my test roll of 1.3.9:
diff -ru opensource/apr-1.3.9/build/libtool.m4 apr-1.3.9/build/libtool.m4
--- opensource/apr-1.3.9/build/libtool.m4 2009-09-21
17:59:31.000000000 -0400
+++ apr-1.3.9/build/libtool.m4 2010-02-04 15:36:25.000000000 -0500
@@ -6042,7 +6042,25 @@
;;
darwin* | rhapsody*)
- _LT_AC_TAGVAR(allow_undefined_flag, $1)="$_lt_dar_allow_undefined"
+ case $host_os in
+ rhapsody* | darwin1.[[012]])
+ _LT_AC_TAGVAR(allow_undefined_flag, $1)='${wl}-undefined
${wl}suppress'
+ ;;
+ *) # Darwin 1.3 on
+ if test -z ${MACOSX_DEPLOYMENT_TARGET} ; then
+ _LT_AC_TAGVAR(allow_undefined_flag,
$1)='${wl}-flat_namespace ${wl}-undefined ${wl}suppress'
+ else
+ case ${MACOSX_DEPLOYMENT_TARGET} in
+ 10.[[012]])
+ _LT_AC_TAGVAR(allow_undefined_flag,
$1)='${wl}-flat_namespace ${wl}-undefined ${wl}suppress'
+ ;;
+ 10.*)
+ _LT_AC_TAGVAR(allow_undefined_flag,
$1)='${wl}-undefined ${wl}dynamic_lookup'
+ ;;
+ esac
+ fi
+ ;;
+ esac
_LT_AC_TAGVAR(archive_cmds_need_lc, $1)=no
_LT_AC_TAGVAR(hardcode_direct, $1)=no
_LT_AC_TAGVAR(hardcode_automatic, $1)=yes
Does anyone recognize that diff and know where it comes from? Or, can
anybody confirm that the diff isn't important?
(similar diff in the generated configure)
Thanks!
Re: q about libtool used in apr 1.3.9
Posted by Ruediger Pluem <rp...@apache.org>.
On 05.02.2010 00:36, Jeff Trawick wrote:
> On Thu, Feb 4, 2010 at 5:32 PM, Sander Temme <sa...@temme.net> wrote:
>> On Feb 4, 2010, at 12:46 PM, Jeff Trawick wrote:
>>
>>> I have libtool 1.5.26 on Linux (seems to be same version as used for
>>> 1.3.9) but I see this difference in my test roll of 1.3.9:
>> libtool.m4 is copied in when you run libtoolize as called by buildconf. A release tarball has whatever version the RM pulled in.
>
> Sure... In trying to duplicate the libtool version from the last
> tarball (1.3.9) I don't get the same libtool.m4. I guess that libtool
> was 1.5.26 + some patch that isn't reflected in the libtool version
Not sure how it was done when 1.3.8 was rolled but the general recommendation is
to download the whole auto* libtool chain, compile it yourself and use it for
rolling. This should avoid any vendor specific patches that are sometimes part
of OS distributed versions of the above.
Regards
Rüdiger
Re: q about libtool used in apr 1.3.9
Posted by Jeff Trawick <tr...@gmail.com>.
On Thu, Feb 4, 2010 at 5:32 PM, Sander Temme <sa...@temme.net> wrote:
>
> On Feb 4, 2010, at 12:46 PM, Jeff Trawick wrote:
>
>> I have libtool 1.5.26 on Linux (seems to be same version as used for
>> 1.3.9) but I see this difference in my test roll of 1.3.9:
>
> libtool.m4 is copied in when you run libtoolize as called by buildconf. A release tarball has whatever version the RM pulled in.
Sure... In trying to duplicate the libtool version from the last
tarball (1.3.9) I don't get the same libtool.m4. I guess that libtool
was 1.5.26 + some patch that isn't reflected in the libtool version
Re: q about libtool used in apr 1.3.9
Posted by Sander Temme <sa...@temme.net>.
On Feb 4, 2010, at 12:46 PM, Jeff Trawick wrote:
> I have libtool 1.5.26 on Linux (seems to be same version as used for
> 1.3.9) but I see this difference in my test roll of 1.3.9:
libtool.m4 is copied in when you run libtoolize as called by buildconf. A release tarball has whatever version the RM pulled in.
S.
> diff -ru opensource/apr-1.3.9/build/libtool.m4 apr-1.3.9/build/libtool.m4
> --- opensource/apr-1.3.9/build/libtool.m4 2009-09-21
> 17:59:31.000000000 -0400
> +++ apr-1.3.9/build/libtool.m4 2010-02-04 15:36:25.000000000 -0500
> @@ -6042,7 +6042,25 @@
> ;;
>
> darwin* | rhapsody*)
> - _LT_AC_TAGVAR(allow_undefined_flag, $1)="$_lt_dar_allow_undefined"
> + case $host_os in
> + rhapsody* | darwin1.[[012]])
> + _LT_AC_TAGVAR(allow_undefined_flag, $1)='${wl}-undefined
> ${wl}suppress'
> + ;;
> + *) # Darwin 1.3 on
> + if test -z ${MACOSX_DEPLOYMENT_TARGET} ; then
> + _LT_AC_TAGVAR(allow_undefined_flag,
> $1)='${wl}-flat_namespace ${wl}-undefined ${wl}suppress'
> + else
> + case ${MACOSX_DEPLOYMENT_TARGET} in
> + 10.[[012]])
> + _LT_AC_TAGVAR(allow_undefined_flag,
> $1)='${wl}-flat_namespace ${wl}-undefined ${wl}suppress'
> + ;;
> + 10.*)
> + _LT_AC_TAGVAR(allow_undefined_flag,
> $1)='${wl}-undefined ${wl}dynamic_lookup'
> + ;;
> + esac
> + fi
> + ;;
> + esac
> _LT_AC_TAGVAR(archive_cmds_need_lc, $1)=no
> _LT_AC_TAGVAR(hardcode_direct, $1)=no
> _LT_AC_TAGVAR(hardcode_automatic, $1)=yes
>
> Does anyone recognize that diff and know where it comes from? Or, can
> anybody confirm that the diff isn't important?
>
> (similar diff in the generated configure)
>
> Thanks!
>
>
--
sander@temme.net http://www.temme.net/sander/
PGP FP: 51B4 8727 466A 0BC3 69F4 B7B8 B2BE BC40 1529 24AF