1
1
Граф коммитов

428 Коммитов

Автор SHA1 Сообщение Дата
George Bosilca
0cf537a826 Automatic text replacement work only when is done correctly.
This commit was SVN r11266.
2006-08-20 13:18:26 +00:00
Brian Barrett
c93288a96e * __USE_GNU must be set in order to get the paffinity declarations.
So define the constant if it isn't already defined.  Something else
  includes stdio.h, which has a bunch of declarations that really
  confuse non-GNU compilers, so be sure to include that one before
  setting __USE_GNU.

refs trac:280

This commit was SVN r11265.

The following Trac tickets were found above:
  Ticket 280 --> https://svn.open-mpi.org/trac/ompi/ticket/280
2006-08-20 01:45:23 +00:00
Brian Barrett
a1b0f2d5bd * opal already sets __builtin_expect, so don't try to redefine it if it's
already set.  This can annoy compilers that aren't GNUish
* __align is technically a reserved token and IBM XL appears to be doing
  something with it that causes compile badness.  So use a different
  variable name.

refs trac:279

This commit was SVN r11264.

The following Trac tickets were found above:
  Ticket 279 --> https://svn.open-mpi.org/trac/ompi/ticket/279
2006-08-20 01:42:36 +00:00
Brian Barrett
25273f57e3 * don't close debugging stream if we're just closing the unused components in
a "select only one" framework

This commit was SVN r11226.
2006-08-16 16:23:35 +00:00
Brian Barrett
7d8d5d0c75 * silence compiler warning
This commit was SVN r11194.
2006-08-15 16:11:27 +00:00
Brian Barrett
d9e06ad499 * -pedantic causes warnings in a bunch of files that Apple provided
This commit was SVN r11192.
2006-08-15 15:49:06 +00:00
Ralph Castain
d2912f03e0 Cleanup a historical naming convention problem. Move the socket_errno definitions to the OPAL layer and change the name accordingly. This cleans up some interrelationship issues as well as removing a name confusion.
This commit was SVN r11186.
2006-08-14 20:14:44 +00:00
Brian Barrett
c1040c97c6 silence some compiler warnings
This commit was SVN r11184.
2006-08-14 19:39:49 +00:00
Brian Barrett
3a7edebc71 * Use the internal (non-locking) functions when in code that will already
be holding the lock.  Fixes ticket #251.

This commit was SVN r11172.
2006-08-14 16:09:17 +00:00
Brian Barrett
dbce03c470 * don't make OPAL_EVLOOP_ONELOOP spin hard, which was causing some latency
issues Gleb saw

This commit was SVN r11169.
2006-08-14 15:02:03 +00:00
Brian Barrett
1bac566976 * update event library to libevent-1.1b
This commit was SVN r11168.
2006-08-14 14:52:39 +00:00
Brian Barrett
bc16f462b9 * print framework and component name during load errors
* return a failure from mtl select code if we don't have a
  component that can run

This commit was SVN r11092.
2006-08-02 20:59:58 +00:00
Brian Barrett
9d5f91997f * updates to build on OS X 10.3
This commit was SVN r11082.
2006-08-01 22:41:51 +00:00
Brian Barrett
a84e557815 Add new loop mode OPAL_EVLOOP_ONELOOP that behaved like OPAL_EVLOOP_ONCE
did pre-libevent update.  The problem is that the behavior of 
OPAL_EVLOOP_ONCE was changed by the OMPI team, which them broke things
during the update, so it had to be reverted to the old meaning of
loop until one event occurs.  OPAL_EVLOOP_ONELOOP will go through the
event loop once (like EVLOOP_NONBLOCK) but will pause in the event
library for a bit (like EVLOOP_ONCE).

fixes trac:234

This commit was SVN r11081.

The following Trac tickets were found above:
  Ticket 234 --> https://svn.open-mpi.org/trac/ompi/ticket/234
2006-08-01 22:23:57 +00:00
Jeff Squyres
7aac77a37c When on the IA64 platform, if we're using the Intel 9.0 v20051201
compiler, automatically disable the ptmalloc component.  It seems that
optimization level -O2 or higher will cause the generated code to do
Bad Things (e.g., opalcc will segv).  Upgrading to the Intel 9.1
compiler seems to fix the problem.

This closes ticket #227.

This commit was SVN r11076.
2006-08-01 18:48:34 +00:00
Brian Barrett
3b17a9d5ee * Older installations of XCode (even on 10.4) don't have the defines needed
for compiling the x86 stack trace unwinding code.  So don't build it on
  those platforms

This commit was SVN r11073.
2006-08-01 02:06:11 +00:00
Brian Barrett
1cc3acd6d8 * fix some warnings from the new event code
This commit was SVN r11060.
2006-07-31 17:03:30 +00:00
Brian Barrett
63f0a5f52d * Don reports that /dev/poll is borked with his version of Solaris. For
now, don't use it and go back to poll/select like everyone else.

This commit was SVN r11034.
2006-07-27 16:56:09 +00:00
Terry Dontje
9c070dafef Added missing include files and fixed some compilation errors
in the original code.

This commit was SVN r11031.
2006-07-27 14:44:54 +00:00
Sven Stork
456c872f52 - one more fix for the dist target
This commit was SVN r11029.
2006-07-27 14:20:37 +00:00
Sven Stork
8c10ea1ceb - fix dist target
This commit was SVN r11026.
2006-07-27 13:25:22 +00:00
Brian Barrett
aaf31c6ade * Make the backtrace printing functionality a framework
* Copy Linux and Solaris backtrace support from util/stacktrace.c
* Added backtrace support for Mac OS X.

This commit was SVN r11023.
2006-07-27 02:56:02 +00:00
Brian Barrett
7ea33eac02 Merge in rest of event library update branch, updating the event library to
libevent-1.1a.

svn merge -r10917:11006 https://svn.open-mpi.org/svn/ompi/tmp/libevent-update

This commit was SVN r11022.

The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
  r10917
  r11006
2006-07-27 01:51:18 +00:00
Brian Barrett
7c2ca73a1b * merge in tmp/libevent-update branch, r10914 - 10917. Rest coming soon.
This commit was SVN r11021.

The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
  r10914
2006-07-27 01:48:09 +00:00
Terry Dontje
215542ad38 Added code to pull in ucontext.h to provide the prototype for
the printstack function for Solaris systems.

This commit was SVN r10967.
2006-07-24 12:14:24 +00:00
Sven Stork
f9fd98449c - add missing prefetch header to fix dist target
This commit was SVN r10934.
2006-07-21 08:20:42 +00:00
Jeff Squyres
32c1f38976 Be a little smarter when checking the MCA parameters that specify what
components to load:

- only allow the ^ to be the first character of the value
- if we find ^ elsewhere in the value, print an error and fail  

This commit was SVN r10880.
2006-07-19 14:19:44 +00:00
Brian Barrett
2185c059e8 * use opal_free_list_item_t as the type of items stored in an opal_free_list_t,
rather than assuing it's an opal_list_item_t.

This commit was SVN r10860.
2006-07-17 21:51:50 +00:00
George Bosilca
33a7634009 Silence the compiler.
This commit was SVN r10851.
2006-07-17 17:13:28 +00:00
Terry Dontje
00694bc65c Added stacktrace support when running on Solaris.
This commit was SVN r10845.
2006-07-17 11:16:58 +00:00
Brian Barrett
dfa1221c3b * AC_CONFIG_LINKS has a minor problem in that it always uses ln -s, rather
than $(LN_S).  This causes problems with with Windows and probably
  elsewhere (re: #200).  So use a slightly different trick to get the
  right header selected for the MEMCPY and TIMER components.

* Using the same trick used to solve the AC_CONFIG_LINKS problem, 
  stop using a separate header file for direct calling in the
  PML and MTL.  This lets me remove some icky code in ompi_mca.m4
  that was more fragile than I really liked.

This commit was SVN r10841.
2006-07-16 04:23:52 +00:00
George Bosilca
a43eb4b43e It's not about how much memory we use, but about how we use it.
Keeping the cache misses as low as possible is always a good approach.
The opal_list_t is widely used, it should be a highly optimized class.
The same functionality can be reached with one one sentinel instead
of 2 currently used.
I don't have anything against the STL version, but so far nothing can
compare with the Knuth algorithm. I replace the current implementation
with a modified version of the Knuth algorithm (the one described in
The Art of Computer Programming). As expected, the latency went down.

This commit was SVN r10776.
2006-07-13 04:56:15 +00:00
Brian Barrett
6ce2b2f989 * need to check for exact equality, not just that some bits are still set.
This commit was SVN r10761.
2006-07-12 16:58:52 +00:00
Brian Barrett
4b70bb92db * Per ticket #112, localhost checks should check against 127.0.0.1/8, rather
than just 127.0.0.1.

This commit was SVN r10750.
2006-07-11 20:54:49 +00:00
Josh Hursey
682a6a123e - os_dirpath.c : reset the is_dir var each time through the loop.
- orte-clean.c : check to see if the base session directory is empty 
                 and delete it if it is.

- orte_universe_exists.c : Fix a down stread problem resulting from 
      George's r10718 commit. Don't use the 'fulldirpath' since
      that is no longer guarenteed to be the absolute path
      to the session directory. Construct this value outside of that
      function from the prefix and frontend vars.

This commit was SVN r10741.

The following SVN revision numbers were found above:
  r10718 --> open-mpi/ompi@47eef2e002
2006-07-11 17:31:05 +00:00
George Bosilca
3075225ea3 Remove the executable extension if any.
This commit was SVN r10714.
2006-07-11 05:19:54 +00:00
Josh Hursey
13f1f4d86e fix a typo when checking the return code
This commit was SVN r10686.
2006-07-06 20:49:09 +00:00
Brian Barrett
27d9e26721 Fix for ticket #179. Print a reasonable error message if we fail to parse
the compiler data file.  Also, actually fix the bug by expanding out
datarootdir before letting it get in install_dirs.h.

This commit was SVN r10643.
2006-07-04 03:00:01 +00:00
Josh Hursey
d082a63734 Add some new OPAL functionality.
After seeing the uglyness that is removing directories in the
codebase I decided to push down this to the OPAL by extending the
opal/os_create_dirpath.(c|h) to contain some more functionality.

In this process I renamed 'os_create_dirpath' to 'os_dirpath' since it
is a bit more general now.

Added a few functions to:
 - check if an directory is empty
 - check to see if the access permissions are set correctly
 - destroy the directory at the end of the dirpath
   - By using a caller callback function (a la Perl, I believe)
     for every file, the caller can have fine grained control over
     whether a specific file is deleted or not.

This simplifies things a bit for orte_session_dir_(finalize|cleanup)
as it should no longer contain any of this functionality, but uses
these functions to do the work.

From the external perspective nothing has changed, from the 
developer point of view we have some cleaner, more generic code.

This commit was SVN r10640.
2006-07-03 22:23:07 +00:00
Josh Hursey
f8541ea644 Actually define the "mca_base_param_env_var" function that was declared.
This commit was SVN r10551.
2006-06-28 21:54:31 +00:00
Brian Barrett
01913be4e6 * IA64 can have a weaker consistency model for memory than x86, so we
need memory barriers to actually do something other than hint
  to the compiler not to reorder memory-related instructions.  The
  IA64 instruction for memory barriers is "mf".

  Fixes bug #137.

This commit was SVN r10401.
2006-06-17 05:01:54 +00:00
Brian Barrett
b9e5acfbe3 * fix issue with PGI compilers at -O3 not liking the structure assignment -
they only appeared to copy the top half of the structure.  Just do
  pointer casting instead and avoid the whole issue.

This commit was SVN r10338.
2006-06-13 19:43:44 +00:00
Brian Barrett
f0a690933f Clean up the opal_free_list code. We don't use the user_data flag in the
opal_free_list_item_t and everyone was just treating them as
opal_list_itme_t structures, which meant that they were overwriting the
user_data memory anyway.  So make an opal_free_lsit_item_t be just a
opal_list_item_t.

Also, opal_free_list_item_t wasn't a proper class, which made it slightly
less than nice to use, so fix that part up.  Also update the
opal_free_list_t to use the class macros, so that it is more obvious we
actually do the right thing.

This commit was SVN r10304.
2006-06-12 14:49:03 +00:00
George Bosilca
85b60cf2bd Having access to the full power of the prefetch is way more interesting. There
are 3 arguments: the pointer to the memory location to prefetch, the type of
operation that will be done on the memory (read or write) and the expected
locality.

This commit was SVN r10294.
2006-06-11 20:10:36 +00:00
Brian Barrett
4c80624fb3 * As promised to George and per #93, add macros for hinting to the compiler
whether an if statement is likely to be taken and for prefetching memory.
  Current macros:

    OPAL_LIKELY(expression)
    OPAL_UNLIKELY(expression)
    OPAL_PREFETC(address)

This commit was SVN r10278.
2006-06-09 19:50:51 +00:00
Jeff Squyres
18f095db16 Update comments to clarify how this file is used.
This commit was SVN r10190.
2006-06-05 13:59:43 +00:00
George Bosilca
372ae03535 There should be one gap between the constructors and the destructors, otherwise
the last constructor will be set to NULL overwriting the first destructor. This
prevent us from calling the destructors on some classes.

This commit was SVN r9969.
2006-05-18 16:21:29 +00:00
Brian Barrett
e76c0ceadb mutex_trylock returns 0 (success) if locked and non-zero (error) if not
locked.  Make the non-threaded case always return "locked", similar to
the non-threaded case for mutex_lock.

This commit was SVN r9956.
2006-05-17 15:50:21 +00:00
Brian Barrett
dc47dd39aa * mvapi's deregister call calls free, and the fast bins in ptmalloc2 aren't
reentrant for free(), so we can't call free() from inside an sbrk() handler.
  The solution is never call sbrk() with a negative number.  The mmap() allocator
  used for large allocations does not have this problem and continues to give
  memory back to the OS as soon as possible.

This should go to both the v1.1 and v1.0 branches.

This commit was SVN r9943.
2006-05-16 21:16:55 +00:00
Jeff Squyres
19a51e2492 Clarify the documentation.
This commit was SVN r9904.
2006-05-12 02:14:57 +00:00