Re: [OMPI devel] VT configure problem?

2011-12-08 Thread Matthias Jurenz
Hello,

this issue is fixed by r25595.

It seems to be a compiler bug in GCC v4.2.? ... When using assert() within 
OpenMP-parallel regions the compiler prepends an extra '_' to the symbol 
__buildin_expect(), so the linker reports undefined references.

The solution is actually a workaround where an own implementation of assert() 
is used to prevent references to __builtin_expect(). 

Matthias

On Thursday 08 December 2011 00:17:29 Ralph Castain wrote:
> Undefined symbols for architecture x86_64:
>   "___builtin_expect", referenced from:
>   __ZN12FilterTraceC26processEventsAndStatisticsEv.omp_fn.0 in
> vtfilter-vt_filter_trc.o __Z13vtfilter_mainiPPc.omp_fn.0 in
> vtfilter-vt_tracefilter.o __Z13vtfilter_mainiPPc.omp_fn.1 in
> vtfilter-vt_tracefilter.o ld: symbol(s) not found for architecture x86_64
> 
> Fresh checkout (r25589) on an x86_64 Mac...obviously, I can --disable-vt,
> but shouldn't this be picked up in configure and corrected?
> 
> 
> 
> ___
> devel mailing list
> de...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/devel


[OMPI devel] OMPI 1.4.5rc1 posted

2011-12-08 Thread Jeff Squyres
1.4.5rc1 is now posted in the usual place:

http://www.open-mpi.org/software/ompi/v1.4/

Gearing up for a pre-Christmas release -- please test!  There have only been a 
few bug fixes since 1.4.4.  See 
http://svn.open-mpi.org/svn/ompi/branches/v1.4/NEWS for details.

(note that we'll likely have a 1.5.5rc1 shortly, too)

-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to:
http://www.cisco.com/web/about/doing_business/legal/cri/




Re: [OMPI devel] Incorrect and undefined return code/function/data type at C++ header

2011-12-08 Thread Y.MATSUMOTO

Dear Jeff and all,

Thank you for your comment.
I'm sorry for not replying sooner.

1:MPI::Fint
We checked C++ header using MPI-2.1 standard.
So, it doesn't need MPI::Fint definition.
(Please remove it!)

2:MPI::Grequest::Start
Sorry! I send you incorrect list.

Best regards.
---
Yuki MATSUMOTO
MPI development team,
Fujitsu

(2011/12/06 1:35), Jeff Squyres wrote:

Many thanks for the patch!

Two minor points:

1. I do not believe that MPI::Fint exists.  It's surprising, but I'm pretty sure we 
double checked this back in the MPI-2.2 timeframe and came to the conclusions that a) 
it does not exist, and b) it should not exist, because all C++<-->  Fortran 
interaction is supposed to go through the C translation routines.

2. Grequest::Start is a static function on the MPI namespace -- it is not marked 
"const" in MPI 2.1 or 2.2 (I don't see it in the patch, either).

On Dec 4, 2011, at 9:31 PM, Y.MATSUMOTO wrote:


Dear all,

We send next feed back.
It's about C++ header file.

In ompi/mpi/cxx/*.h,
Some definitions of return code, type and function are lacked or incorrect.
Attached patch fixes them (This Patch is for V1.4.X).

Following list is what is lacked and incorrect.

*Undefined return code
--
MPI::ERR_ACCESS
MPI::ERR_AMODE
MPI::ERR_ASSERT
MPI::ERR_BAD_FILE
MPI::ERR_CONVERSION
MPI::ERR_DISP
MPI::ERR_DUP_DATAREP
MPI::ERR_FILE_EXISTS
MPI::ERR_FILE_IN_USE
MPI::ERR_FILE
MPI::ERR_INFO
MPI::ERR_IO
MPI::ERR_LOCKTYPE
MPI::ERR_NOT_SAME
MPI::ERR_NO_SPACE
MPI::ERR_NO_SUCH_FILE
MPI::ERR_PORT
MPI::ERR_QUOTA
MPI::ERR_READ_ONLY
MPI::ERR_RMA_CONFLICT
MPI::ERR_RMA_SYNC
MPI::ERR_SIZE
MPI::ERR_UNSUPPORTED_DATAREP
MPI::ERR_UNSUPPORTED_OPERATION
--
*Undefined data type
--
MPI::LONG_LONG_INT
MPI::Fint
MPI::F_DOUBLE_COMPLEX
--

*Undefined function
--
MPI::Datatype::Create_darray
MPI::Datatype::Pack_external
MPI::Datatype::Pack_external_size
MPI::Datatype::Unpack_external
MPI::Add_error_class
MPI::Add_error_code
MPI::Add_error_string
MPI::Datatype::Create_f90_complex
MPI::Datatype::Create_f90_integer
MPI::Datatype::Create_f90_real
MPI::Datatype::Match_size
--

*Incorrect of definitions
(MPI-2.1 standard defines these as "const", but they are not "const" in code)
--
MPI::Intercomm::Merge
MPI::Cartcomm::Sub
MPI::Grequest::Start
--

*Incorrect of definitions
(MPI-2.1 standard defines these as not "const", but they are "const" in code)
--
MPI::Comm::Set_errhandler
MPI::File::Set_errhandler
MPI::Win::Set_errhandler
--

Best regards.
--
Yuki MATSUMOTO
MPI development team,
Fujitsu

___
devel mailing list
de...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/devel