about autoconf >= 2.64

[pierre kestener]

Thanks to all for your answers, I found the origin of the bug.

I was testing Gaisler's RTEMS sources 
(rtems-4.10-1.1.99.10-src.tar.bz2); these sources are based on rtems cvs 
before Ralf commit on file cpukit/aclocal/check-rtems-debug.m4 (Add 
missing ','.).

As autoconf >= 2.64 is better than old versions, it do not let it pass 
this bug !!

By correcting this bug, everything works well whatever autoconf version 
on Ubuntu 9.10.

My advice to people using Gaisler's sources along with autoconf >= 2.64, 
is to temporaly fall back to autoconf 2.63, or correct the 
check-rtems-debug.m4 bug.

Pierre.


Joel Sherrill a écrit :
> On 03/02/2010 10:58 AM, pierre kestener wrote:
>> Hello,
>>
>> While upgrading from Ubuntu 9.04 to Ubuntu 9.10, autoconf was updated
>> from 2.63 to 2.64, and I noticed that some of the configure script are
>> not well generated anymore in RTEMS source tree.
>>
>>    
> The autoconf version is tied to the RTEMS version. For
> hosts we provide binaries on, automake/autoconf is included
> in the RTEMS toolset.  Updating autotools can break things
> so we tend not to update it on release branches.  You can see
> this by the latest (or terminal) autoconf RPM for each release
> branch.
>
> rtems-4.6-autoconf-rtems-2.59-0.noarch
> rtems-4.7-autoconf-2.60-4.el5.noarch
> rtems-4.8-autoconf-2.61-7.el5.noarch
> rtems-4.9-autoconf-2.62-4.fc12.noarch
> rtems-4.10-autoconf-2.65-1.fc12.noarch
>
>> For example, the generated cpukit/configure contains the following piece
>> of code (bad shell code):
>> ##############################
>> if test x"${enable_rtems_debug}" = x"yes"
>>
>> cat>>confdefs.h<<_ACEOF
>> #define RTEMS_DEBUG 1
>> _ACEOF
>> ; then :
>>
>> fi
>> ##############################
>>
>> Falling back to "old" autoconf 2.63, everything works perfectly on
>> Ubuntu 9.10
>> I also tried autoconf 2.65, the bug is still there.
>>
>> I was wondering if such a bug was Ubuntu-specific, and if someone was
>> using autoconf>=2.64 with success and on which distribution ?
>>
>> Thank you.
>>
>> Pierre Kestener.
>>
>> _______________________________________________
>> rtems-users mailing list
>> rtems-users at rtems.org
>> http://www.rtems.org/mailman/listinfo/rtems-users
>>    
>
>


-- 
+--------------------------------------------------------+
 Pierre KESTENER - Ingénieur-Chercheur
 CEA Saclay -- DSM/IRFU/SEDI/TRAPS
 91191 Gif-sur-Yvette - FRANCE
 tel : +33 (0)1 69 08 57 09
 fax : +33 (0)1 69 08 31 47
 http://pierre.kestener.org/
+--------------------------------------------------------+