1
1
Граф коммитов

329 Коммитов

Автор SHA1 Сообщение Дата
Sven Stork
0084c9469a use correct free methode for additional allocated memory
This commit was SVN r10241.
2006-06-07 10:24:28 +00:00
George Bosilca
a64a80dff4 If the user type has a size of zero let's return zero. We will have a consistent behavior
with MPICH.

This commit was SVN r10230.
2006-06-06 19:51:42 +00:00
George Bosilca
7d7e801f15 External pack/unpack fixes.
This commit was SVN r10223.
2006-06-06 03:26:32 +00:00
George Bosilca
07fb4b8012 Allow a block indexed type with a count of zero. Be user friendly and set the ub, lb,
true_lb as well as the true_ub to zero in this case.

This commit was SVN r10208.
2006-06-05 21:16:57 +00:00
George Bosilca
5c2d2fc02a Match size is supposed to return Fortran types.
This commit was SVN r10206.
2006-06-05 21:07:48 +00:00
George Bosilca
e50cdeb927 Allow the creation of strcutres with count zero. And try to have a more friendly behavior
(even if I don't agree with it) by setting the lb, ub, true_lb and true_ub to zero.

This commit was SVN r10205.
2006-06-05 21:07:16 +00:00
Jeff Squyres
299f4fdb2c Oops -- fix the comment. A victim of cut-n-paste.
This commit was SVN r9971.
2006-05-18 18:10:12 +00:00
Jeff Squyres
942f9e8f8d Fixes for ticket:14. Lengthy discussion is on that ticket and in a
comment in ompi_comm_invalid() in
source:/trunk/ompi/communicator/communicator.h.

Short version:
- ompi_comm_invalid() returns TRUE for MPI_COMM_NULL
- therefore MPI_COMM_C2F needs to explicitly check for MPI_COMM_NULL
  (because it uses ompi_comm_invalid())
- make ~20 MPI functions only call ompi_comm_invalid() instead of
  calling ompi_comm_invalid() *and* checking for MPI_COMM_NULL (~40 MPI
  functions already only called ompi_comm_invalid() -- we should be
  consistent)
- similar issue for ompi_win_invalid(), so I added a cross-referencing
  comment in win.h and fixed MPI_WIN_SET_NAME to only call
  ompi_win_invalid() (and not check for MPI_WIN_NULL)

This commit was SVN r9970.
2006-05-18 18:05:46 +00:00
Edgar Gabriel
7e99dc4301 - fix in comm_spawn_multiple: we cannot check for the errorcodes array
being null, if the constant MPI_ERRCODES_IGNORE is defined as (void *)
NULL;
- the communicator in file open has to be an intra-communicator.

This commit was SVN r9893.
2006-05-11 13:23:53 +00:00
George Bosilca
29fe545451 Remove all references to the ompi_mpi_long_long variable. The MPI_LONG_LONG is now
a synonym to MPI_LONG_LONG_INT.

This commit was SVN r9703.
2006-04-24 22:15:42 +00:00
Jeff Squyres
f710ef5b2b Commit 1.5 of 2: just like r9560, we need to make the MPI_*_C2F
functions return an invalid fortran handle (-1) if an invalid C handle
is passed in.

Just so it's logged somewhere -- it looks like commit 2 of 2 as noted
in the r9560 commit message (ok, 3 of 3 if you're really counting),
where we have to edit the C interface functions to check for NULL, is
going to be far easier than I thought -- many (but not all) of the C
interface functions already check for NULL MPI handles (either
directly or through helper functions).  

Woo hoo for foresight...

This commit was SVN r9561.

The following SVN revision numbers were found above:
  r9560 --> open-mpi/ompi@e00c6053dc
2006-04-07 04:01:22 +00:00
Jeff Squyres
e00c6053dc Commit 1 of 2: be consistent in the MPI_*_F2C functions for the
handling of invalid Fortran handles.  Per MPI-2:4.12.4, if we get an
invalid Fortran handle, we should return an invalid C handle.  Before
this commit, we checked if the fortran handle index was out of range
(i.e., <0 or >sizeof(array)).  If so, we used to return MPI_*_NULL
(i.e., a valid C handle).  But to be faithful to MPI-2:4.12.4, we
now return NULL (an invalid C handle).

If the fortran index is in bounds but is an index for an MPI object
that has already been freed, the code already returns NULL because the
entry in the array will be NULL (i.e., we already did what
MPI-2:4.12.4 said for this case).

Hence, this commit makes the handling of invalid fortran handles in
the MPI_*_F2C functions be uniform: we always return NULL.

Commit 2 of this will be to edit just about every C interface function
(!) to ensure that MPI handles are not NULL.  Otherwise, if the user
calls a fortran interface function with an invalid handle, the fortran
interface function will call MPI_*_F2C and blindly pass the result to
the back-end C function.  The C function will eventually end up trying
to dereference it -- segv.  Having a run-time check for NULL and
invoking an MPI exception is far more social (e.g., the user can get a
stack trace out of MPI_ABORT) and consistent (i.e., we're already
checking for MPI_*_NULL in the C interface functions).

Since all the C interface functions have all the machinery for
run-time parameter checking, and they all already check for
MPI_*_NULL, it's easy enough to add another check for NULL.

This commit was SVN r9560.
2006-04-07 03:49:58 +00:00
Sven Stork
e6b0828a4f - add missing lb check
This commit was SVN r9516.
2006-04-03 13:22:04 +00:00
George Bosilca
285b1af2a7 Follow the standard when the count is set to zero. Or let me say it differently,
follow the statement from the standard that make the most sense to me. In this
particular case this statement is:
If there is no active handle in the list it returns outcount = MPI_UNDEFINED.

This commit was SVN r9512.
2006-04-01 07:39:55 +00:00
George Bosilca
ae47cbfe02 Accept an array of request set to NULL if the count is set to ZERO.
This commit was SVN r9504.
2006-03-31 20:32:56 +00:00
Sven Stork
711b30fbac - introduction of user buffer checking for pt2pt operations
This commit was SVN r9449.
2006-03-29 09:26:27 +00:00
George Bosilca
85bb1a9c90 Add one more argument to the copy functions for the MPI objects. As this argument
is the last one on the list and as on C the caller "make it right" this addition
will not affect the way we handle the user defined copy functions. Only the C
version of the function has this additional parameter. As it represent the pointer
to the newly created MPI object It hold the key to allow us to modify the new
object (communicator, window or type) depending on some key stored on the initial
communicator.

This commit was SVN r9371.
2006-03-23 04:47:14 +00:00
Tim Woodall
bd870519fd - modified convertor copy_and_prepare routines to accept an addition
flag, new flags to be included when convertor is initialized
- modified pml/btl module defs and added stub functions for diagnostic
  output routines to dump state of queues / endpoints
- updates to data reliability pml

This commit was SVN r9329.
2006-03-17 18:46:48 +00:00
Mohamad Chaarawi
b8c9adc833 checking of multiply included/excluded processes in
group_range_incl/excl has now been separated from the implementation.

This commit was SVN r9311.
2006-03-16 17:51:16 +00:00
Rainer Keller
a8144cca8d - Fix the return codes -- not OMPI, but MPI..
- Use the datatype-checks of bindings.h within recv.
   However, for the time being disable the check for overlapped,
   as it is not correctly set.

This commit was SVN r9298.
2006-03-16 09:42:46 +00:00
George Bosilca
d19e924ca8 gcc print out: warning: ISO C90 forbids mixed declarations and code.
This commit was SVN r9283.
2006-03-14 22:27:19 +00:00
Mohamad Chaarawi
c5a8870f6b move the implementation of the group creation functions into according back-end
functions. 

This commit was SVN r9281.
2006-03-14 21:01:46 +00:00
George Bosilca
679ccbcac7 Correctly handle the MPI_STATUSES_IGNORE, and trigger the MPI error handler and return the
same error code.

This commit was SVN r9206.
2006-03-06 16:59:46 +00:00
George Bosilca
ca116e3cf3 Where we create a duplicate data-type for the get_content MPI function we have to attach the
args from the old data-type. As this args are only read-only, I don't want to have several
copies around, so I add a reference count to them and simply attach them to the new
datatype. The args array now get freed only when the last reference dissapear.

This commit was SVN r9191.
2006-03-03 20:51:36 +00:00
Rainer Keller
8568bb9892 - Allow MPI_ERRCODES_IGNORE when checking is on.
This commit was SVN r9180.
2006-02-28 10:16:42 +00:00
Brian Barrett
d5e0ea3590 * Post and Start should only check their epoch types for conflicts, otherwise
you can't be in a post and a start at the same time, and that is clearly
  legal to do
* Fix interptretation of when the epochs start for MPI_Fence.  Only start
  an epoch if communication actually occurs, otherwise it isn't actually
  an epoch.  I don't know who thought that wording in the MPI standard
  was a good idea, but can't change it now...

This commit was SVN r9139.
2006-02-24 13:04:15 +00:00
Brian Barrett
27b8430e8f * update MPI_ACCUMULATE to perform it's parameter checking based on what
MPI-2:6.3.4 says about reduction operations
* Have the point-to-point one-sided component spew a warning and return
  an error if a non-predefined datatype is used with an MPI_OP other
  than MPI_REPLACE.  Yes, this violates the MPI standard, but it's the
  best we can do until George and I implement support for figuring out
  where all the locations to update are..

This commit was SVN r9134.
2006-02-23 21:07:49 +00:00
Jeff Squyres
8f7311becd Fix a few copy-n-paste errors: when an op is invalid, invoke the MPI
exception on the passed communicator, not MPI_COMM_WORLD.

This commit was SVN r9105.
2006-02-21 19:11:40 +00:00
Brian Barrett
889aeeddb2 * clean up the error checking a bit - user defined datatypes are allowed for
MPI_OP_REPLACE.  All other rules are basically the same as MPI_Reduce, so
  use that helper function.

This commit was SVN r9104.
2006-02-21 19:00:49 +00:00
Galen Shipman
0bc3cbf0db Corrections to pml_dr, now passes intel test suite (p2p_c).
Note, the checksums are not enabled currently, setting to zero as the
convertor is not ready for checksums yet. 

Also, we can't call unpack/pack on convertor with 0 bytes, otherwise it
crashes. 

This commit was SVN r9062.
2006-02-16 16:15:16 +00:00
Jeff Squyres
c674c165fd - Fix final check in TYPE_CREATE_F90_INTEGER
- Add more checks on the params to ensure that the user conformed to
  MPI-2 (i.e., can have MPI_UNDEFINED for p *or* r in the real/complex
  functions, but not both)
- Always return MPI_ERR_ARG, not MPI_ERR_OTHER -- seems to make a
  little more sense here, since any errors that are returned are
  solely because of the arguments that were passed into the function

This commit was SVN r9037.
2006-02-14 18:01:44 +00:00
Brian Barrett
566a050c23 Next step in the project split, mainly source code re-arranging
- move files out of toplevel include/ and etc/, moving it into the
    sub-projects
  - rather than including config headers with <project>/include, 
    have them as <project>
  - require all headers to be included with a project prefix, with
    the exception of the config headers ({opal,orte,ompi}_config.h
    mpi.h, and mpif.h)

This commit was SVN r8985.
2006-02-12 01:33:29 +00:00
George Bosilca
9c1547373d Match the latest changes in the datatype engine.
This commit was SVN r8967.
2006-02-10 21:06:15 +00:00
Brian Barrett
ec7b60d501 * Fix some error checking code for Lock / Unlock at the MPI layer
* Implement win_lock and win_unlock in the pt2pt component.  Not well
  tested, but appears to move bits if properly motivated...

This commit was SVN r8922.
2006-02-07 18:45:18 +00:00
George Bosilca
e20265bd2b Dont let any external to the data-type code check directly for the predefined
data-types. Instead, use the newly provided data-type function ompi_ddt_is_predefined..

This commit was SVN r8903.
2006-02-06 18:01:45 +00:00
Brian Barrett
514e58c49a * pass an ompi_win_t, not an ompi_win_t*
This commit was SVN r8866.
2006-01-31 23:08:26 +00:00
Brian Barrett
44a516d966 * Clean up error checking in the MPI interface for MPI-2 onesided
* Implement fortran handle -> c handle tracking
* Remove some unneeded locking around free lists (the free list
  macros do their own locking)
* Try to be a bit more memory friendly with the w_mode setting /
  checking

This commit was SVN r8865.
2006-01-31 21:40:12 +00:00
Brian Barrett
0d507ad970 * Fix a bunch of checks for proper access / exposure epochs
* Add implementation of Post/Wait/Start/Complete synchronization

This commit was SVN r8848.
2006-01-31 02:44:08 +00:00
Brian Barrett
b1d2424013 Merge in present work on the MPI-2 onesided chapter. The current code is not
complete, but stable enough that it will have no impact on general development,
so into the trunk it goes.  Changes in this commit include:

 - Remove the --with option for disabling MPI-2 onesided support.  It
   complicated code, and has no real reason for existing
 - add a framework osc (OneSided Communication) for encapsulating
   all the MPI-2 onesided functionality
 - Modify the MPI interface functions for the MPI-2 onesided chapter
   to properly call the underlying framework and do the required
   error checking
 - Created an osc component pt2pt, which is layered over the BML/BTL
   for communication (although it also uses the PML for long message
   transfers).  Currently, all support functions, all communication
   functions (Put, Get, Accumulate), and the Fence synchronization
   function are implemented.  The PWSC active synchronization
   functions and Lock/Unlock passive synchronization functions are
   still not implemented

This commit was SVN r8836.
2006-01-28 15:38:37 +00:00
Rainer Keller
9b2c23bead - Do additional checks in case of MPI_PARAM_CHECK for NULL pointers.
This commit was SVN r8487.
2005-12-13 08:13:13 +00:00
Brian Barrett
a46d0d3801 Add support to Open MPI for MPI_Register_datarep, as ROMIO now has support for the
functionality.  This is a global registration, so all components have to be made
aware of the registration -- therefore, do things at the component level, not the
module level.

This commit was SVN r8314.
2005-11-29 18:25:35 +00:00
George Bosilca
55051b81c4 Activate the protection against unavailable datatypes. They get a flag DT_FLAG_UNAVAILABLE. We check now this flag in all the send/recv operations via the macros on mpi/c/bindings.h.
This flag is inherited by all datatypes create with unavailable datatypes. Basically, we let the user create the wrong datatype but we dont let him using it for any pt2pt communications or any pack/unpack.

This commit was SVN r8069.
2005-11-09 23:43:41 +00:00
Brian Barrett
28891d6de3 * Move MPI_Wtime and MPI_Wtick back out of mpi.h and into the C bindings library,
restoring the PMPI version.  A variety of reasons for this:

  - mpi.h was blinding using inline in a C header without the configrue mojo
    properly set it, as mpi.h doesn't include ompi_config.h.  This eventually
    would have caused a borked build.
  - mpi.h and mpif.h were never updated to not include PMPI_W{tick,time} as
    a proper prototype
  - The C++ and F90 bindings didn't do the right things when there was no
    PMPI version of the C call, but profiling was enabled
  - Since we only use gettimeofday, the function call overhead really doesn't
    matter

This should probably go to the 1.0 branch

This commit was SVN r8014.
2005-11-07 17:22:48 +00:00
Jeff Squyres
42ec26e640 Update the copyright notices for IU and UTK.
This commit was SVN r7999.
2005-11-05 19:57:48 +00:00
Jeff Squyres
7bdfe6557b - Update the checks in REDUCE, ALLREDUCE, SCAN, EXSCAN, and
REDUCE_SCATTER to more thoroughly check the datatype/op combination
  to see if it's valid or not.  If it's not, print a meaningful error
  message rather than "Invalid MPI_Op" indicating what specifically
  was wrong (therefore hopefully helping users track down where in the
  code the problem is, and/or telling us that there's a reduction
  operation combo that we don't support that we should)
- The check for whether a datatype is intrinsic needed to be updated
  -- it's not sufficient to check that dtype->id < DT_MAX_PREDEFINED;
  you really need to check the PREDEFINED flag on the datatype.
  Thanks to George for this fix (only intrinsics have a meaningful
  value in dtype->id).

This commit was SVN r7923.
2005-10-28 16:47:32 +00:00
Jeff Squyres
23ab9e0277 A better solution to the previous commit -- RETAIN/RELEASE the MPI_Op
at the top-level MPI API function.  This allows two kinds of
scenarios:

1. MPI_Ireduce(..., op, ...);
   MPI_Op_free(op);
   MPI_Wait(...);

For the non-blocking collectives that we're someday planning -- to
make them analogous to non-blocking point-to-point stuff.

2. Thread 1:
   MPI_Reduce(..., op, ...);
   Thread 2:
   MPI_Op_free(op);

Granted, for #2 to occur would tread a fine line between a correct and
erroneous MPI program, but it is possible (as long as the Op_free was
*after* MPI_reduce() had started to execute).  It's more realistic
with case #1, where the Op_free() could be executed in the same thread
or a different thread.

This commit was SVN r7870.
2005-10-25 19:20:42 +00:00
Jeff Squyres
e097ee635a Silence compiler warnings.
This commit was SVN r7768.
2005-10-14 22:06:25 +00:00
Edgar Gabriel
2c909383bb abstracting the group_free operation into an internal routine (required
by some other components on ompi).

This commit was SVN r7763.
2005-10-14 18:51:20 +00:00
Jeff Squyres
94bab558dd Put in a check to ensure the root is valid (all other rooted
operations have this; we somehow missed this for intracomms on reduce,
and it bit me this morning ;-) )

This commit was SVN r7612.
2005-10-04 14:38:17 +00:00
Tim Woodall
aceab46c5f use MPI_Alloc_mem/MPI_Free_mem for internally allocated buffers
This commit was SVN r7487.
2005-09-22 16:43:17 +00:00
Brian Barrett
c87babb565 * if start_rank == end_rank, there doesn't seem to be a requirement that
stride == 1, and the Intel tests explicitly test the case with
  strides != 1, expecting them to work

This commit was SVN r7411.
2005-09-16 18:44:21 +00:00
George Bosilca
789d49d613 Activate thread support ... at least at the MPI level.
This commit was SVN r7386.
2005-09-15 15:44:12 +00:00
Jeff Squyres
b8f4f88ccc Ignore sendtype/sendcount if MPI_IN_PLACE
This commit was SVN r7361.
2005-09-14 01:44:55 +00:00
Jeff Squyres
15f5294da4 Ignore another param in the IN_PLACE case
This commit was SVN r7348.
2005-09-13 19:15:01 +00:00
Jeff Squyres
c89e8c3712 Fixes for error checking -- properly ignore arguments as they should
be

This commit was SVN r7347.
2005-09-13 19:07:13 +00:00
Jeff Squyres
ad647b105d Error checking for MPI_IN_PLACE
This commit was SVN r7343.
2005-09-13 17:07:03 +00:00
George Bosilca
71b1cdb5f5 Cleanup the dependencies between the MPI layer and the rest of the ompi files.
Modify the includes to start with the root directory (i.e orte/datatype/datatype.h).

This commit was SVN r7306.
2005-09-12 09:17:44 +00:00
Brian Barrett
ed56e743b7 * update configure.ac to use the modern version of AC_INIT and
AM_INIT_AUTOMAKE, instead of the deprecated version.
* Work around dumbness in modern AC_INIT that requires the version
  number to be set at autoconf time (instead of at configure time, as
  it was before).  Set the version number, minus the subversion r number,
  at autoconf time.  Override the internal variables to include the r
  number (if needed) at configure time.  Basically, the right thing
  should always happen.  The only place it might not is the version
  reported as part of configure --help will not have an r number.
* Since AM_INIT_AUTOMAKE taks a list of options, no need to specify
  them in all the Makefile.am files.
* Addes support for subdir-objects, meaning that object files are put
  in the directory containing source files, even if the Makefile.am is
  in another directory.  This should start making it feasible to
  reduce the number of Makefile.am files we have in the tree, which
  will greatly reduce the time to run autogen and configure.

This commit was SVN r7211.
2005-09-07 05:54:53 +00:00
Rainer Keller
605e13baf4 - Just a comment to circumvent warning in check.
This commit was SVN r6933.
2005-08-19 10:38:08 +00:00
Jeff Squyres
9b7c1ca97a Forgot to include the check for 0 in the Allreduce case (already got
Reduce and Reduce_scatter)

This commit was SVN r6929.
2005-08-18 19:39:45 +00:00
Jeff Squyres
e8c103ac1f Make MPI_REDUCE and MPI_REDUCE_SCATTER return MPI_SUCCESS immediately
if the count/sum of counts is 0.  This is technically in violation of
the MPI-1 standard, but...  :-(

This commit was SVN r6914.
2005-08-17 15:34:33 +00:00
Jeff Squyres
c465eb8567 Rename opal/threads/thread.h -> opal/threads/threads.h to avoid a
naming conflict with Solaris' <thread.h>

This commit was SVN r6879.
2005-08-15 11:02:01 +00:00
Brian Barrett
f273d84b1b * update ob1 to direct call
* don't know what I was thinking, but can't use the MCA_PML_CALL macro on
  the two data values, as they don't have things that the macro can
  expand into

This commit was SVN r6868.
2005-08-14 03:14:20 +00:00
Jeff Squyres
cf16a521c8 Ensure to get ompi/include/constants.h
This commit was SVN r6845.
2005-08-12 21:42:07 +00:00
Brian Barrett
95fd068ffa remove hard coded constants for value of MPI_TAG_UB and the max CID and add
the values to the PML structure.  This will allow PMLs that want to do
hardware matching at the cost of a smaller range of valid tags and cids.
Updated all the places that used the MPI_TAG_UB_VALUE constant to instead
look at the pml struct.

This commit was SVN r6778.
2005-08-09 14:56:04 +00:00
Rainer Keller
f5f96f424d - Use for-loop
- Minor heck for stride != 1
- indenting

This commit was SVN r6715.
2005-08-02 19:56:18 +00:00
Rainer Keller
15c4ae3391 Fix for buglet #1423.
Any non-blocking P2P function should set the request to something else
then MPI_REQUEST_NULL, when passed a source/dest of MPI_PROC_NULL.

Introduce a new ompi_request_empty, which returns the
MPI_SOURCE=MPI_PROC_NULL, MPI_TAG=MPI_ANY_TAG, count=0 as specified by
the MPI-standard.

This commit was SVN r6611.
2005-07-27 01:00:02 +00:00
Edgar Gabriel
213be28613 minor fixes to make the trunk compile, if ompi has been configured with the --enable-mpi2-one-sided option
This commit was SVN r6518.
2005-07-15 13:53:10 +00:00
George Bosilca
a7aa1a76bc Correctly use the new convertor. Now we have 2 generic predefined convertors. One is for
the external32 conversions (as specified in the MPI standard) the other one is the local
convertor that can be used for MPI_Pack and MPI_Unpack functions.

This commit was SVN r6375.
2005-07-07 23:00:55 +00:00
Brian Barrett
170ef8af1f * rename ompi_show_help to opal_show_help
* rename ompi_stacktrace to opal_stacktrace
* rename ompi_strncpy to opal_strncpy

This commit was SVN r6336.
2005-07-04 02:38:44 +00:00
Brian Barrett
39dbeeedfb * rename locking code from ompi to opal
This commit was SVN r6327.
2005-07-03 22:45:48 +00:00
Brian Barrett
ccd2624e3f * rename ompi_progress to opal_progress
This commit was SVN r6326.
2005-07-03 21:57:43 +00:00
Brian Barrett
9da0b4fe1d * rename all the atomic functions from ompi to opal
This commit was SVN r6325.
2005-07-03 21:38:51 +00:00
Brian Barrett
761402f95f * rename ompi_list to opal_list
This commit was SVN r6322.
2005-07-03 16:22:16 +00:00
Brian Barrett
499e4de1e7 * rename ompi_object and ompi_class to opal_object and opal_class
This commit was SVN r6321.
2005-07-03 16:06:07 +00:00
Jeff Squyres
35c141aef6 While we're moving directories around, move ompi/mpi/runtime ->
ompi/runtime, for consistency and parallel-ness will orte/runtime.
Also remove a few useless #includes along the way.

This commit was SVN r6317.
2005-07-03 12:07:29 +00:00
Brian Barrett
e984f48ad9 * add missing Makefile to configure output
* Fix MPI profiling layer symlink code from src -> ompi
* Change some modex includes

This commit was SVN r6309.
2005-07-02 17:32:01 +00:00
Jeff Squyres
aa056f7bfd First cut of OMPI Makefile.am's, plus a few more catchup updates in orte
This commit was SVN r6286.
2005-07-02 15:06:47 +00:00
Jeff Squyres
4ab17f019b Rename src -> ompi
This commit was SVN r6269.
2005-07-02 13:43:57 +00:00