If it works, don't fix it. :-)
Perhaps there are other assumptions in the code such that the -f is
necessary. It's hard to say without seeing the whole code, and since
it's working, it doesn't seem to be worth it to chase down.
On Mar 30, 2007, at 4:09 PM, Yu Chen wrote:
> Thanks Jeff, and Davide,
>
> I actually sucessfully compiled everthing with passing the FFLAGS
> to LAM
> configure/build, although I have not run the application yet.
>
> Do you mean even if it compiles, don't do the -f thing?
> I think after I took out the FFLAGS from both, the undefined errors
> are
> MPI unrelated, but I don't see any other libriaries being used. I
> am not
> actually a programmer, it's other people's software. So I don't
> know too
> much about the program. Could it be the '-f' is in his code, and
> has to be
> there?
>
> Regards,
> Chen
>
>>>>> FFLAGS = -fno-underscoring -O
>>>>> FFLAGS2 = -fno-underscoring -O
>>>>
>>>> What are FFLAGS and FFLAGS2 used for? Are those flags passed to
>>>> LAM's configure/build process, or just to the configure/build
>>>> process
>>>> of your application?
>>>>
>>>> If they're just passed to your application, I suspect that this is
>>>> your problem (i.e., remove them and things should be fine).
>>>
>>> Yeah, it just passed to the application, I removed them, but
>>> still not
>>> working, although it gave lot less undefined errors, I will try to
>>> pass
>>> that to LAM's configure/build process, see it that works.
>>
>> No, don't do that.
>>
>> The issue here is that all Fortran compilers have a specific name-
>> mangling scheme. There are 4 typical schemes:
>>
>> - all lower case
>> - all upper case
>> - add one underscore to the end of the symbol
>> - add one underscore to the end of the symbol unless there's already
>> an underscore in the symbol -- then add 2 underscores
>>
>> You just need LAM and your application to match. If you use the -f
>> switch, you're telling the fortran compiler to change its name
>> mangling scheme. Using the "native" scheme is usually the easiest.
>> So just don't pass the -f flag to either LAM's or your application's
>> build process.
>
>>
>> If you're getting other missing symbols, are they LAM or MPI
>> related? If not, you might want to check the -f settings for your
>> other libraries, ensure that you're linking in all the right
>> libraries, etc.
>>
>>> Regards,
>>> Chen
>>>
>>>
>>>>
>>>>> Regards,
>>>>> Chen
>>>>>
>>>>>
>>>>>> It sounds like you have a case of compiling LAM/MPI with one
>>>>>> fortran
>>>>>> compiler and compiling your application with a different Fortran
>>>>>> compiler.
>>>>>>
>>>>>> Can you verify that you're using the same Fortran compiler for
>>>>>> both?
>>>>>>
>>>>>>
>>>>>> On Mar 29, 2007, at 1:39 PM, Yu Chen wrote:
>>>>>>
>>>>>>> Hi, wondering if someone can help me out;
>>>>>>>
>>>>>>> I am using RHEL 4, with g95 version of G95 (GCC 4.0.3 (g95
>>>>>>> 0.90!)
>>>>>>> Jul 27
>>>>>>> 2006), when installing LAM-MPI 7.1.3, everything is fine, the
>>>>>>> "laminfo"
>>>>>>> and "mpif77 --showme" outputs follows. Then when I
>>>>>>> tried to compile our program, it gave out following error.
>>>>>>>
>>>>>>> ......
>>>>>>> In function `MAIN_':
>>>>>>> : undefined reference to `mpi_init'
>>>>>>> In function `MAIN_':
>>>>>>> : undefined reference to `mpi_comm_rank'
>>>>>>> In function `MAIN_':
>>>>>>> : undefined reference to `mpi_comm_size'
>>>>>>> In function `MAIN_':
>>>>>>> : undefined reference to `mpi_barrier'
>>>>>>> In function `MAIN_':
>>>>>>> : undefined reference to `mpi_bcast'
>>>>>>> In function `MAIN_':
>>>>>>> : undefined reference to `mpi_bcast'
>>>>>>> undefined reference to `mpi_comm_dup_fn'
>>>>>>> undefined reference to `mpi_comm_null_delete_fn'
>>>>>>> undefined reference to `mpi_comm_null_copy_fn'
>>>>>>> undefined reference to `mpi_dup_fn'
>>>>>>> undefined reference to `mpi_null_copy_fn'
>>>>>>> undefined reference to `mpi_null_delete_fn'
>>>>>>> undefined reference to `mpi_type_null_copy_fn'
>>>>>>> undefined reference to `mpi_type_null_delete_fn'
>>>>>>> undefined reference to `mpi_type_dup_fn'
>>>>>>> undefined reference to `mpi_win_null_copy_fn'
>>>>>>> undefined reference to `mpi_win_dup_fn'
>>>>>>> undefined reference to `mpi_win_null_delete_fn'
>>>>>>> ......
>>>>>>>
>>>>>>> #### Laminfo output ####
>>>>>>> LAM/MPI: 7.1.3
>>>>>>> Prefix: /usr/local/lam-g95
>>>>>>> Architecture: x86_64-unknown-linux-gnu
>>>>>>> Configured by: cheny
>>>>>>> Configured on: Thu Mar 29 10:26:01 EDT 2007
>>>>>>> Memory manager: ptmalloc2
>>>>>>> C bindings: yes
>>>>>>> C++ bindings: yes
>>>>>>> Fortran bindings: yes
>>>>>>> C compiler: gcc
>>>>>>> C++ compiler: g++
>>>>>>> Fortran compiler: /usr/local/g95-32/bin/g95
>>>>>>> Fortran symbols: double_underscore
>>>>>>> C profiling: yes
>>>>>>> C++ profiling: yes
>>>>>>> Fortran profiling: yes
>>>>>>> C++ exceptions: no
>>>>>>> Thread support: yes
>>>>>>> ROMIO support: yes
>>>>>>> IMPI support: no
>>>>>>> Debug support: no
>>>>>>> Purify clean: no
>>>>>>> SSI boot: globus (API v1.1, Module v0.6)
>>>>>>> SSI boot: rsh (API v1.1, Module v1.1)
>>>>>>> SSI boot: slurm (API v1.1, Module v1.0)
>>>>>>> SSI coll: lam_basic (API v1.1, Module v7.1)
>>>>>>> SSI coll: shmem (API v1.1, Module v1.0)
>>>>>>> SSI coll: smp (API v1.1, Module v1.2)
>>>>>>> SSI rpi: crtcp (API v1.1, Module v1.1)
>>>>>>> SSI rpi: lamd (API v1.0, Module v7.1)
>>>>>>> SSI rpi: sysv (API v1.0, Module v7.1)
>>>>>>> SSI rpi: tcp (API v1.0, Module v7.1)
>>>>>>> SSI rpi: usysv (API v1.0, Module v7.1)
>>>>>>> SSI cr: self (API v1.0, Module v1.0)
>>>>>>>
>>>>>>> #### mpif77 --showme outputs ####
>>>>>>> /usr/local/g95-32/bin/g95 -I/usr/local/lam-g95/include -pthread
>>>>>>> -L/usr/local/lam-g95/lib -llammpio -llamf77mpi -lmpi -llam -
>>>>>>> laio -
>>>>>>> laio
>>>>>>> -lutil -ldl
>>>>>>>
>>>>>>> Thank you
>>>>>>>
>>>>>>> Chen
>>>>>>>
>>>>>>> ===========================================
>>>>>>> Yu Chen
>>>>>>> Howard Hughes Medical Institute
>>>>>>> Chemistry Building, Rm 182
>>>>>>> University of Maryland at Baltimore County
>>>>>>> 1000 Hilltop Circle
>>>>>>> Baltimore, MD 21250
>>>>>>>
>>>>>>> phone: (410)455-6347 (primary)
>>>>>>> (410)455-2718 (secondary)
>>>>>>> fax: (410)455-1174
>>>>>>> email: chen_at_[hidden]
>>>>>>> ===========================================
>>>>>>> _______________________________________________
>>>>>>> This list is archived at http://www.lam-mpi.org/MailArchives/
>>>>>>> lam/
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> ===========================================
>>>>> Yu Chen
>>>>> Howard Hughes Medical Institute
>>>>> Chemistry Building, Rm 182
>>>>> University of Maryland at Baltimore County
>>>>> 1000 Hilltop Circle
>>>>> Baltimore, MD 21250
>>>>>
>>>>> phone: (410)455-6347 (primary)
>>>>> (410)455-2718 (secondary)
>>>>> fax: (410)455-1174
>>>>> email: chen_at_[hidden]
>>>>> ===========================================
>>>>> _______________________________________________
>>>>> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>>>>
>>>>
>>>>
>>>
>>>
>>> ===========================================
>>> Yu Chen
>>> Howard Hughes Medical Institute
>>> Chemistry Building, Rm 182
>>> University of Maryland at Baltimore County
>>> 1000 Hilltop Circle
>>> Baltimore, MD 21250
>>>
>>> phone: (410)455-6347 (primary)
>>> (410)455-2718 (secondary)
>>> fax: (410)455-1174
>>> email: chen_at_[hidden]
>>> ===========================================
>>> _______________________________________________
>>> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>>
>>
>>
>
>
> ===========================================
> Yu Chen
> Howard Hughes Medical Institute
> Chemistry Building, Rm 182
> University of Maryland at Baltimore County
> 1000 Hilltop Circle
> Baltimore, MD 21250
>
> phone: (410)455-6347 (primary)
> (410)455-2718 (secondary)
> fax: (410)455-1174
> email: chen_at_[hidden]
> ===========================================
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
--
Jeff Squyres
Cisco Systems
|
