ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
86f48bf0d7
openmpi/ompi/mpi/f77/constants.h
Jeff Squyres 520147f209 Clean up the Fortran MPI sentinel values per problem reported on the 
users mailing list:

  http://www.open-mpi.org/community/lists/users/2006/07/1680.php

Warning: this log message is not for the weak.  Read at your own
risk.

The problem was that we had several variables in Fortran common blocks
of various types, but their C counterparts were all of a type
equivalent to a fortran double complex.  This didn't seem to matter
for the compilers that we tested, but we never tested static builds
(which is where this problem seems to occur, at least with the Intel
compiler: the linker compilains that the variable in the common block
in the user's .o file was of one size/alignment but the one in the C
library was a different size/alignment).

So this patch fixes the sizes/types of the Fortran common block
variables and their corresponding C instantiations to be of the same
sizes/types. 

But wait, there's more.

We recently introduced a fix for the OSX linker where some C versions
of the fortran common block variables (e.g.,
_ompi_fortran_status_ignore) were not being found when linking
ompi_info (!).  Further research shows that the code path for
ompi_info to require ompi_fortran_status_ignore is, unfortunately,
necessary (a quirk of how various components pull in different
portions of the code base -- nothing in ompi_info itself requires
fortran or MPI knowledge, of course).

Hence, the real problem was that there was no code path from ompi_info
to the portion of the code base where the C globals corresponding to
the Fortran common block variables were instantiated.  This is because
the OSX linker does not automatically pull in .o files that only
contain unintialized global variables; the OSX linker typically only
pulls in a .o file from a library if it either has a function that is
used or have a global variable that is initialized (that's the short
version; lots of details and corner cases omitted).  Hence, we changed
the global C variables corresponding to the fortran common blocks to
be initialized, thereby causing the OSX linker to pull them in
automatically -- problem solved.  At the same time, we moved the
constants to another .c file with a function, just for good measure.

However, this didn't really solve the problem:

1. The function in the file with the C versions of the fortran common
   block variables (ompi/mpi/f77/test_constants_f.c) did not have a
   code path that was reachable from ompi_info, so the only reason
   that the constants were found (on OSX) was because they were
   initialized in the global scope (i.e., causing the OSX compiler to
   pull in that .o file).

2. Initializing these variable in the global scope causes problems for
   some linkers where -- once all the size/type problems mentioned
   above were fixed -- the alignments of fortran common blocks and C
   global variables do not match (even though the types of the Fortran
   and C variables match -- wow!).  Hence, initializing the C
   variables would not necessarily match the alignment of what Fortran
   expected, and the linker would issue a warning (i.e., the alignment
   warnings referenced in the original post).

The solution is two-fold:

1. Move the Fortran variables from test_constants_f.c to
   ompi/mpi/runtime/ompi_mpi_init.c where there are other global
   constants that *are* initialized (that had nothing to do with
   fortran, so the alignment issues described above are not a factor),
   and therefore all linkers (including the OSX linker) will pull in
   this .o file and find all the symbols that it needs.

2. Do not initialize the C variables corresponding to the Fortran
   common blocks in the global scope.  Indeed, never initialize them
   at all (because we never need their *values* - we only check for
   their *locations*).  Since nothing is ever written to these
   variables (particularly in the global scope), the linker does not
   see any alignment differences during initialization, but does make
   both the C and Fortran variables have the same addresses (this
   method has been working in LAM/MPI for over a decade).

There were some comments here in the OMPI code base and in the LAM
code base that stated/implied that C variables corresponding to
Fortran common blocks had to have the same alignment as the Fortran
common blocks (i.e., 16).  There were attempts in both code bases to
ensure that this was true.  However, the attempts were wrong (in both
code bases), and I have now read enough Fortran compiler documentation
to convince myself that matching alignments is not required (indeed,
it's beyond our control).  As long as C variables corresponding to
Fortran common blocks are not initialized in the global scope, the
linker will "figure it out" and adjust the alignment to whatever is
required (i.e., the greater of the alignments).  Specifically (to
counter comments that no longer exist in the OMPI code base but still
exist in the LAM code base):

- there is no need to make attempts to specially align C variables
  corresponding to Fortran common blocks
- the types and sizes of C variables corresponding to Fortran common
  blocks should match, but do not need to be on any particular
  alignment 

Finally, as a side effect of this effort, I found a bunch of
inconsistencies with the intent of status/array_of_statuses
parameters.  For all the functions that I modified they should be
"out" (not inout).

This commit was SVN r11057.
2006-07-31 15:07:09 +00:00

222 строки
9.7 KiB
C
Исходник Ответственный История

/*
* Copyright (c) 2004-2005 The Trustees of Indiana University and Indiana
* University Research and Technology
* Corporation. All rights reserved.
* Copyright (c) 2004-2005 The University of Tennessee and The University
* of Tennessee Research Foundation. All rights
* reserved.
* Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
* University of Stuttgart. All rights reserved.
* Copyright (c) 2004-2005 The Regents of the University of California.
* All rights reserved.
* Copyright (c) 2006 Cisco Systems, Inc. All rights reserved.
* $COPYRIGHT$
*
* Additional copyrights may follow
*
* $HEADER$
*/
#ifndef OMPI_F77_CONSTANTS_H
#define OMPI_F77_CONSTANTS_H
/*
* Several variables are used to link against MPI F77 constants which
* correspond to addresses, e.g. MPI_BOTTOM, and are implemented via
* common blocks.
*
* We use common blocks so that in the C wrapper functions, we can
* compare the address that comes in against known addresses (e.g., if
* the "status" argument in MPI_RECV is the address of the common
* block for the fortran equivalent of MPI_STATUS_IGNORE, then we know
* to pass the C MPI_STATUS_IGNORE to the C MPI_Recv function). As
* such, we never look at the *value* of these variables (indeed,
* they're never actually initialized), but instead only ever look at
* the *address* of these variables.
*
* As such, it is not strictly necessary that the size and type of our
* C variables matches that of the common Fortran block variables.
* However, good programming form says that we should match, so we do.
*
* Note, however, that the alignments of the Fortran common block and
* the C variable may not match (e.g., Intel 9.0 compilers on 64 bit
* platforms will put the alignment of a double on 4 bytes, but put
* the alignment of all common blocks on 16 bytes). This only matters
* (to some compilers!), however, if you initialize the C variable in
* the global scope. If the C global instantiation is not
* initialized, the compiler/linker seems to "figure it all out" and
* make the alignments match.
*
* Since we made the fundamental decision to support all 4 common
* fortran compiler symbol conventions within the same library for
* those compilers who support weak symbols, we need to have 4 symbols
* for each of the fortran address constants. As described above, we
* have to have known *pointer* values for the fortran addresses
* (e.g., MPI_STATUS_IGNORE). So when the fortran wrapper for
* MPI_RECV gets (MPI_Fint *status), it can check (status ==
* some_sentinel_value) to know that it got the Fortran equivalent of
* MPI_STATUS_IGNORE and therefore pass the C MPI_STATUS_IGNORE to the
* C MPI_Recv.
*
* We do this by having a "common" block in mpif.h:
*
* INTEGER MPI_STATUS_IGNORE(MPI_STATUS_SIZE)
* common /mpi_fortran_status_ignore/ MPI_STATUS_IGNORE
*
* This makes the fortran variable MPI_STATUS_IGNORE effectively be an
* alias for the C variable "mpi_fortran_status_ignore" -- but the C
* symbol name is according to the fortran compiler's naming symbol
* convention bais. So it could be MPI_FORTRAN_STATUS_IGNORE,
* mpi_fortran_status_ignore, mpi_fortran_status_ignore_, or
* mpi_fortran_status_ignore__.
*
* Hence, we have to have *4* C symbols for this, and them compare for
* all of them in the fortran MPI_RECV wrapper. :-( I can't think of
* any better way to do this.
*
* I'm putting these 4 comparisons in macros (on systems where we
* don't support the 4 symbols -- e.g., OSX, where we don't have weak
* symbols -- it'll only be one comparison), so if anyone things of
* something better than this, you should only need to modify this
* file.
*/
#define DECL(type, upper_case, lower_case, single_u, double_u) \
OMPI_DECLSPEC extern type upper_case; \
OMPI_DECLSPEC extern type lower_case; \
OMPI_DECLSPEC extern type single_u; \
OMPI_DECLSPEC extern type double_u
/* Note that the rationale for the types of each of these variables is
discussed in ompi/include/mpif-common.h. Do not change the types
without also changing ompi/mpi/runtime/ompi_mpi_init.c and
ompi/include/mpif-common.h. */
DECL(int, MPI_FORTRAN_BOTTOM, mpi_fortran_bottom,
mpi_fortran_bottom_, mpi_fortran_bottom__);
DECL(int, MPI_FORTRAN_IN_PLACE, mpi_fortran_in_place,
mpi_fortran_in_place_, mpi_fortran_in_place__);
DECL(char *, MPI_FORTRAN_ARGV_NULL, mpi_fortran_argv_null,
mpi_fortran_argv_null_, mpi_fortran_argv_null__);
DECL(double, MPI_FORTRAN_ARGVS_NULL, mpi_fortran_argvs_null,
mpi_fortran_argvs_null_, mpi_fortran_argvs_null__);
DECL(int *, MPI_FORTRAN_ERRCODES_IGNORE, mpi_fortran_errcodes_ignore,
mpi_fortran_errcodes_ignore_, mpi_fortran_errcodes_ignore__);
DECL(int *, MPI_FORTRAN_STATUS_IGNORE, mpi_fortran_status_ignore,
mpi_fortran_status_ignore_, mpi_fortran_status_ignore__);
DECL(double, MPI_FORTRAN_STATUSES_IGNORE, mpi_fortran_statuses_ignore,
mpi_fortran_statuses_ignore_, mpi_fortran_statuses_ignore__);
/*
* Create macros to do the checking. Only check for all 4 if we have
* weak symbols. Otherwise, just check for the one relevant symbol.
*/
#if OMPI_HAVE_WEAK_SYMBOLS
#define OMPI_IS_FORTRAN_BOTTOM(addr) \
(addr == (void*) &MPI_FORTRAN_BOTTOM || \
addr == (void*) &mpi_fortran_bottom || \
addr == (void*) &mpi_fortran_bottom_ || \
addr == (void*) &mpi_fortran_bottom__)
#define OMPI_IS_FORTRAN_IN_PLACE(addr) \
(addr == (void*) &MPI_FORTRAN_IN_PLACE || \
addr == (void*) &mpi_fortran_in_place || \
addr == (void*) &mpi_fortran_in_place_ || \
addr == (void*) &mpi_fortran_in_place__)
#define OMPI_IS_FORTRAN_ARGV_NULL(addr) \
(addr == (void*) &MPI_FORTRAN_ARGV_NULL || \
addr == (void*) &mpi_fortran_argv_null || \
addr == (void*) &mpi_fortran_argv_null_ || \
addr == (void*) &mpi_fortran_argv_null__)
#define OMPI_IS_FORTRAN_ARGVS_NULL(addr) \
(addr == (void*) &MPI_FORTRAN_ARGVS_NULL || \
addr == (void*) &mpi_fortran_argvs_null || \
addr == (void*) &mpi_fortran_argvs_null_ || \
addr == (void*) &mpi_fortran_argvs_null__)
#define OMPI_IS_FORTRAN_ERRCODES_IGNORE(addr) \
(addr == (void*) &MPI_FORTRAN_ERRCODES_IGNORE || \
addr == (void*) &mpi_fortran_errcodes_ignore || \
addr == (void*) &mpi_fortran_errcodes_ignore_ || \
addr == (void*) &mpi_fortran_errcodes_ignore__)
#define OMPI_IS_FORTRAN_STATUS_IGNORE(addr) \
(addr == (void*) &MPI_FORTRAN_STATUS_IGNORE || \
addr == (void*) &mpi_fortran_status_ignore || \
addr == (void*) &mpi_fortran_status_ignore_ || \
addr == (void*) &mpi_fortran_status_ignore__)
#define OMPI_IS_FORTRAN_STATUSES_IGNORE(addr) \
(addr == (void*) &MPI_FORTRAN_STATUSES_IGNORE || \
addr == (void*) &mpi_fortran_statuses_ignore || \
addr == (void*) &mpi_fortran_statuses_ignore_ || \
addr == (void*) &mpi_fortran_statuses_ignore__)
#elif OMPI_F77_CAPS
#define OMPI_IS_FORTRAN_BOTTOM(addr) \
(addr == (void*) &MPI_FORTRAN_BOTTOM)
#define OMPI_IS_FORTRAN_IN_PLACE(addr) \
(addr == (void*) &MPI_FORTRAN_IN_PLACE)
#define OMPI_IS_FORTRAN_ARGV_NULL(addr) \
(addr == (void*) &MPI_FORTRAN_ARGV_NULL)
#define OMPI_IS_FORTRAN_ARGVS_NULL(addr) \
(addr == (void*) &MPI_FORTRAN_ARGVS_NULL)
#define OMPI_IS_FORTRAN_ERRCODES_IGNORE(addr) \
(addr == (void*) &MPI_FORTRAN_ERRCODES_IGNORE)
#define OMPI_IS_FORTRAN_STATUS_IGNORE(addr) \
(addr == (void*) &MPI_FORTRAN_STATUS_IGNORE)
#define OMPI_IS_FORTRAN_STATUSES_IGNORE(addr) \
(addr == (void*) &MPI_FORTRAN_STATUSES_IGNORE)
#elif OMPI_F77_PLAIN
#define OMPI_IS_FORTRAN_BOTTOM(addr) \
(addr == (void*) &mpi_fortran_bottom)
#define OMPI_IS_FORTRAN_IN_PLACE(addr) \
(addr == (void*) &mpi_fortran_in_place)
#define OMPI_IS_FORTRAN_ARGV_NULL(addr) \
(addr == (void*) &mpi_fortran_argv_null)
#define OMPI_IS_FORTRAN_ARGVS_NULL(addr) \
(addr == (void*) &mpi_fortran_argvs_null)
#define OMPI_IS_FORTRAN_ERRCODES_IGNORE(addr) \
(addr == (void*) &mpi_fortran_errcodes_ignore)
#define OMPI_IS_FORTRAN_STATUS_IGNORE(addr) \
(addr == (void*) &mpi_fortran_status_ignore)
#define OMPI_IS_FORTRAN_STATUSES_IGNORE(addr) \
(addr == (void*) &mpi_fortran_statuses_ignore)
#elif OMPI_F77_SINGLE_UNDERSCORE
#define OMPI_IS_FORTRAN_BOTTOM(addr) \
(addr == (void*) &mpi_fortran_bottom_)
#define OMPI_IS_FORTRAN_IN_PLACE(addr) \
(addr == (void*) &mpi_fortran_in_place_)
#define OMPI_IS_FORTRAN_ARGV_NULL(addr) \
(addr == (void*) &mpi_fortran_argv_null_)
#define OMPI_IS_FORTRAN_ARGVS_NULL(addr) \
(addr == (void*) &mpi_fortran_argvs_null_)
#define OMPI_IS_FORTRAN_ERRCODES_IGNORE(addr) \
(addr == (void*) &mpi_fortran_errcodes_ignore_)
#define OMPI_IS_FORTRAN_STATUS_IGNORE(addr) \
(addr == (void*) &mpi_fortran_status_ignore_)
#define OMPI_IS_FORTRAN_STATUSES_IGNORE(addr) \
(addr == (void*) &mpi_fortran_statuses_ignore_)
#else
#define OMPI_IS_FORTRAN_BOTTOM(addr) \
(addr == (void*) &mpi_fortran_bottom__)
#define OMPI_IS_FORTRAN_IN_PLACE(addr) \
(addr == (void*) &mpi_fortran_in_place__)
#define OMPI_IS_FORTRAN_ARGV_NULL(addr) \
(addr == (void*) &mpi_fortran_argv_null__)
#define OMPI_IS_FORTRAN_ARGVS_NULL(addr) \
(addr == (void*) &mpi_fortran_argvs_null__)
#define OMPI_IS_FORTRAN_ERRCODES_IGNORE(addr) \
(addr == (void*) &mpi_fortran_errcodes_ignore__)
#define OMPI_IS_FORTRAN_STATUS_IGNORE(addr) \
(addr == (void*) &mpi_fortran_status_ignore__)
#define OMPI_IS_FORTRAN_STATUSES_IGNORE(addr) \
(addr == (void*) &mpi_fortran_statuses_ignore__)
#endif /* weak / specific symbol type */
/* Convert between Fortran and C MPI_BOTTOM */
#define OMPI_ADDR(addr) (OMPI_IS_FORTRAN_BOTTOM(addr) ? MPI_BOTTOM : (addr))
#define OMPI_IN_PLACE(addr) (OMPI_IS_FORTRAN_IN_PLACE(addr) ? MPI_IN_PLACE : (addr))
#endif /* OMPI_F77_CONSTANTS_H */
Показать git blame Копировать постоянную ссылку