[mpich-devel] LLONG_MAX undeclared in status_set_elements_x.c ?

[William Gropp]

Ah, that's a different problem - that's an error in the implementation file, which is incorrectly assuming that LLONG_MAX is always defined.  That needs to be fixed by the routine's author.

Bill

William Gropp
Director, Parallel Computing Institute
Deputy Director for Research
Institute for Advanced Computing Applications and Technologies
Paul and Cynthia Saylor Professor of Computer Science
University of Illinois Urbana-Champaign



On Dec 31, 2012, at 2:19 PM, Jeff Hammond wrote:

> This is my setup and what I see:
> 
> [jhammond at flogin3 build-gcc]$ uname -a
> Linux flogin3 2.6.18-274.18.1.el5 #1 SMP Fri Jan 20 15:11:18 EST 2012
> x86_64 x86_64 x86_64 GNU/Linux
> 
> [jhammond at flogin3 build-gcc]$ svn info ..
> Path: ..
> URL: https://svn.mcs.anl.gov/repos/mpi/mpich2/trunk
> Repository Root: https://svn.mcs.anl.gov/repos/mpi
> Repository UUID: a5d90c62-d51d-0410-9f91-bf5351168976
> Revision: 10808
> Node Kind: directory
> Schedule: normal
> Last Changed Author: dinan
> Last Changed Rev: 10808
> Last Changed Date: 2012-12-28 22:21:55 -0600 (Fri, 28 Dec 2012)
> 
> [jhammond at flogin3 build-gcc]$ which gcc
> /soft/gcc/4.4.2/bin/gcc
> 
> [jhammond at flogin3 build-gcc]$ gcc --version
> gcc (GCC) 4.4.2
> Copyright (C) 2009 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions.  There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> 
> [jhammond at flogin3 build-gcc]$ ../configure --enable-threads --enable-g
> --prefix=/homes/jhammond/MPI/mpich2-trunk/install-gcc && make -j8 &&
> make install
> <snip>
>  CC       src/mpi_t/mpi_t_util.lo
>  CC       src/nameserv/file/file_nameserv.lo
>  CC       src/pmi/simple/simple_pmiutil.lo
>  CC       src/pmi/simple/simple_pmi.lo
>  CC       src/mpi/datatype/status_set_elements_x.lo
>  CC       src/mpi/datatype/type_indexed.lo
>  CC       src/mpi/datatype/type_hindexed.lo
>  CC       src/mpi/datatype/type_struct.lo
>  CC       src/mpi/datatype/type_contiguous.lo
>  CC       src/mpi/datatype/type_free.lo
>  CC       src/mpi/datatype/type_hvector.lo
> ../src/mpi/datatype/status_set_elements_x.c: In function
> 'MPIR_Status_set_elements_x_impl':
> ../src/mpi/datatype/status_set_elements_x.c:41: error: 'LLONG_MAX'
> undeclared (first use in this function)
> ../src/mpi/datatype/status_set_elements_x.c:41: error: (Each
> undeclared identifier is reported only once
> ../src/mpi/datatype/status_set_elements_x.c:41: error: for each
> function it appears in.)
> make[2]: *** [src/mpi/datatype/status_set_elements_x.lo] Error 1
> make[2]: *** Waiting for unfinished jobs....
> make[2]: Leaving directory
> `/fusion/gpfs/home/jhammond/MPI/mpich2-trunk/build-gcc'
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory
> `/fusion/gpfs/home/jhammond/MPI/mpich2-trunk/build-gcc'
> make: *** [all] Error 2
> 
> Jeff
> 
> On Mon, Dec 31, 2012 at 2:07 PM, William Gropp <wgropp at illinois.edu> wrote:
>> I've committed the change that tests for LLONG_MAX - in r10802 .  What error are you seeing?
>> 
>> Bill
>> 
>> William Gropp
>> Director, Parallel Computing Institute
>> Deputy Director for Research
>> Institute for Advanced Computing Applications and Technologies
>> Paul and Cynthia Saylor Professor of Computer Science
>> University of Illinois Urbana-Champaign
>> 
>> 
>> 
>> On Dec 31, 2012, at 1:37 PM, Jeff Hammond wrote:
>> 
>>> I did a fresh checkout today and am still seeing this issue.  What's
>>> the workaround until it is fixed for real?
>>> 
>>> I am working on Fusion at Argonne using "../configure --enable-threads
>>> --enable-g" and GCC 4.4.2 compilers, btw.
>>> 
>>> Thanks,
>>> 
>>> Jeff
>>> 
>>> On Fri, Dec 21, 2012 at 8:22 AM, William Gropp <wgropp at illinois.edu> wrote:
>>>> Thanks.  This test erroneously uses long long and LLONG_MAX without ensuring that they are available (and for which the configure for the tests already checks).  I'll fix it.
>>>> 
>>>> Bill
>>>> 
>>>> William Gropp
>>>> Director, Parallel Computing Institute
>>>> Deputy Director for Research
>>>> Institute for Advanced Computing Applications and Technologies
>>>> Paul and Cynthia Saylor Professor of Computer Science
>>>> University of Illinois Urbana-Champaign
>>>> 
>>>> 
>>>> 
>>>> On Dec 21, 2012, at 2:25 AM, P.-H. Lin wrote:
>>>> 
>>>>> Hello Howard,
>>>>> I have encountered the same problem with mpich-3.0.1 today. (And
>>>>> that's why I found this thread :P)
>>>>> After raising the optimization level, it didn't complain about that
>>>>> undeclared problem anymore, and the installation goes well.
>>>>> This is my configure option:
>>>>> ./configure --prefix /opt/app/MPICH --enable-fast=all,O3
>>>>> 
>>>>> BTW, the problem still exist for "indexed-misc.c" under the test directory
>>>>> 
>>>>> Regards
>>>> 
>>> 
>>> 
>>> 
>>> --
>>> Jeff Hammond
>>> Argonne Leadership Computing Facility
>>> University of Chicago Computation Institute
>>> jhammond at alcf.anl.gov / (630) 252-5381
>>> http://www.linkedin.com/in/jeffhammond
>>> https://wiki.alcf.anl.gov/parts/index.php/User:Jhammond
>> 
> 
> 
> 
> -- 
> Jeff Hammond
> Argonne Leadership Computing Facility
> University of Chicago Computation Institute
> jhammond at alcf.anl.gov / (630) 252-5381
> http://www.linkedin.com/in/jeffhammond
> https://wiki.alcf.anl.gov/parts/index.php/User:Jhammond