1
1
Граф коммитов

660 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
176cecfb3b Fix for LT 2.1 libltdl lt_dlopen behavior of putting loaded DSO's in a
private scope.  Solves the problem of Red Hat-provided OFED not
working properly because the libibverbs plugins are not linked
against libibverbs.

This commit was SVN r13926.
2007-03-05 17:21:42 +00:00
Brian Barrett
ada2316fd3 work around issue where $CC might be multiword and would leave stuff in the
extra linker flags.  Reported by Bert Wesarg.

This commit was SVN r13774.
2007-02-24 22:18:44 +00:00
Brian Barrett
7ce72af158 fix some misuses of various environment variables when running pthread checks
reported by Ake Sandgren on the users list a couple months ago...

This commit was SVN r13773.
2007-02-24 21:22:55 +00:00
Rainer Keller
dbe6100bd2 - As there are other changes to configure-files:
Cleanup check_attributes and comments, get rid of _msg

This commit was SVN r13678.
2007-02-16 07:45:53 +00:00
Josh Hursey
5a0f268ddd fix minor typo
This commit was SVN r13677.
2007-02-16 01:36:59 +00:00
Pavel Shamis
edeab0e912 Adding Mellanox Technologies copyright to files touched by Mellanox.
This commit was SVN r13669.
2007-02-15 18:03:20 +00:00
Rainer Keller
38696e8d32 - Check for attributes, after C and C++ have been set up.
- Set ompi-specific autoconf cache-variables
 - Implement one function to check for availability of an
   attribute with the possibility for a cross-check.
 - Do cross-checks for
     __attribute__(format)
     __attribute__(nonnull)
     __attribute__(sentinel)
     __attribute__(warn_unused_result)
 - Grep the compilers warnings for keywords regarding ignored
   attributes.
 - Include also the no_instrument_function

This commit was SVN r13556.
2007-02-08 13:34:44 +00:00
Rainer Keller
80295b8f68 - Indent one space; this crept up to deliver:
icc -g -finline-functions -fno-strict-aliasing-restrict -c help.c
icc: Command line warning: ignoring option '-fno-strict-aliasing'; no argument required

This commit was SVN r13507.
2007-02-06 13:23:23 +00:00
Jeff Squyres
c91fcd7fbd Fix a bunch of minor typos submitted by Bernhard Fischer.
This commit was SVN r13505.
2007-02-06 12:00:30 +00:00
Rainer Keller
c2c8bdf8ac - Check not only C, but also C++ compiler both at one
go and have proper casting.
   Tested with linux gcc-4, icc-9.0, pathcc-2.2.1 and pgi-6.2.5

   (icc warns on several attributes, that they are ignored;
    same for pathcc and pgi)

This commit was SVN r13485.
2007-02-03 17:35:46 +00:00
Tim Prins
cc896e6292 make way for a better fix...
This commit was SVN r13484.
2007-02-03 16:49:28 +00:00
Tim Prins
0d283cb4d7 Fix for the attribute problems we have been having lately.
Check both the c and cpp compilers for support of attributes, and then check for support for each individual attribute with both compilers. Only if both compilers support a given attribute will we enable it.

This commit was SVN r13483.
2007-02-03 15:58:49 +00:00
Rolf vandeVaart
bf5113198d Update to orte-clean so it will remove files on local and
remote nodes.  It will also kill off rogue orteds and orterun
processes.  The killing of processes is ifdef'ed out for Windows
since I do not know how to do it there.  Note that this change
will requite an autogen.  

This commit was SVN r13477.
2007-02-03 00:25:42 +00:00
Rolf vandeVaart
fd1bf67119 Fix where we initialize some variables if attribute is not
supported by compiler.

This commit was SVN r13410.
2007-02-01 04:01:04 +00:00
Rainer Keller
7d799523e5 - include the attributes for always_inline and sentinel
This commit was SVN r13373.
2007-01-30 21:26:08 +00:00
Rainer Keller
2bda2f5d8c - Fix and extend the attribute checking macros for the
most common attributes. Useful attributes may be:
   - optimization (pure, const, malloc--restrictness)
   - data layout (packed)
   - interface definition, API (deprecated, visibility, weak alias)
   - compiler hint for code analysis and debugging (nonnull, format)

This commit was SVN r13369.
2007-01-30 20:52:46 +00:00
Rainer Keller
ca35881cd0 - Minor bugfixes and removed compiler warnings
This commit was SVN r13343.
2007-01-28 19:52:09 +00:00
Brian Barrett
c6ee421c78 Work around a bug involving dependent libraries when building 64 bit on Mac OS X
This commit was SVN r13246.
2007-01-22 20:31:04 +00:00
Rainer Keller
31e12cbe71 - Get --coverage to work with new gcov and cleanup the generated files.
- Use regexp to check for optimizations in flags.

This commit was SVN r13211.
2007-01-19 14:28:52 +00:00
Jeff Squyres
add3909096 Back out 13076 and 13077 in favor of a much simpler approach.
Sorry for the configure change -- hopefully it's early enough in the
morning that it won't affect people... (new approach won't have a
configure change).

Refs trac:739.

This commit was SVN r13080.

The following Trac tickets were found above:
  Ticket 739 --> https://svn.open-mpi.org/trac/ompi/ticket/739
2007-01-11 14:07:15 +00:00
Jeff Squyres
f0932a0701 A workaround for a bug in the PGI 6.2 compiler series. This bug has
been fixed in the 7.0 PGI series, but is unlikely to be fixed in the
6.2 series:

 * Add a configure test looking for the bad behavior (the PGI compiler
   chokes on C code where structs containing bool's are copied by
   value)
 * Set OMPI_BOOL_STRUCT_COPY to 1 if it's ok, 0 if it's not (i.e., PGI
   6.2 series will have this value set to 0)
 * In two places in the code base -- orte-clean and btl_openib_ini.h,
   we have a struct that contains a bool that is copied by value.  In
   these two places, check OMPI_BOOL_STRUCT_COPY and if it's 1, use
   the "int" type instead of "bool".

Fixes trac:739

This commit was SVN r13076.

The following Trac tickets were found above:
  Ticket 739 --> https://svn.open-mpi.org/trac/ompi/ticket/739
2007-01-11 02:21:26 +00:00
George Bosilca
416e5b5f6a Enable the MX extensions if and only if the mx_extensions.h header
is installed on the system.

This commit was SVN r12937.
2006-12-29 00:31:32 +00:00
George Bosilca
3eeecc3838 Add support for faster small messages. While sending a message, we check if
the data was buffered by the MX library. If it's the case then we declare
the send as completed and disable the completion event for the mx request.

This commit was SVN r12935.
2006-12-28 22:34:24 +00:00
George Bosilca
3903009b8b Add a check for the unexpected handler. If enabled, allow the zero-copy
protocol over the MX BTL. Now, we have only one matching, the one in Open
MPI.

The problem is that when the unexpected handler is triggered, not all the
message is on the host memory. In the best case we get one MX fragment (internal
MX fragment), in the worst we get NULL. The only way to fit this with the
design of the PML is to force the eager protocol at the MX internal fragment
size, and to limit the send/receive protocol at the same size. Tests show
the outcome is not far from optimal (if the pipeline depth is increased
a little bit).

Set MX_PIPELINE_LOG in order to allow MX to use internal fragments of 4K.

This commit was SVN r12930.
2006-12-28 03:35:41 +00:00
Li-Ta Lo
eac3520d00 update xcpu autoconfig checks for libxcpu and friends
This commit was SVN r12903.
2006-12-20 22:56:25 +00:00
Brian Barrett
bf38c9d8e2 Configure changes... woopie...
* Check that the C++, Fortran 77, and Objective C comilers emit code
    that can link against object files emitted by the C compiler.
    Moves some built / run time errors to configure time, which is
    nice and should help with the debugging
  * Remove unneeded -F option when building the XGrid components, 
    which started causing problems with LT 2.0.
  * Try to use the XGridFoundation library, rather than just seeing
    if we can give -framework XGridFoundation.  Should make the
    test slightly more accurate
  * Don't assume XGrid is unavailable on 64 bit platforms, as that
    won't be true on Leopard
  * Require AM 1.10 or newer if using AC 2.60 or newer, so that
    we don't have a split of AC supporting Objective C and AM
    not doing so

This commit was SVN r12701.
2006-11-30 01:59:44 +00:00
Rainer Keller
2fc3ce8efc - Get -restrict check to work
This commit was SVN r12432.
2006-11-05 20:38:44 +00:00
Brian Barrett
9534bb27d9 dont' remove .m4 files during make maintainer-clean as it prevents
the target from completing due to dependency issues 

This commit was SVN r12412.
2006-11-02 18:41:04 +00:00
Pavel Shamis
566667ac61 Adding progress thread support to OpenIB BTL.
Reviewed by Gleb.

This commit was SVN r12411.
2006-11-02 16:15:21 +00:00
George Bosilca
6745e1a7b7 Correctly detect the STDC_HEADERS, before adding flags to the CFLAGS. Now,
that we have the STDC_HEADERS we are able to detect when ptrdiff_t is defined.

This commit was SVN r12254.
2006-10-23 03:55:52 +00:00
Jeff Squyres
533b4db7b5 Fix a problem when $FC/$F77 were multiple tokens.
This commit was SVN r12246.
2006-10-21 14:24:30 +00:00
Tim Prins
45a4f2c7ed Fix a minor problem in variable naming in these configure macros.
Thanks to Martin Audet for reporting this on the users list.

This commit was SVN r12203.
2006-10-19 23:35:14 +00:00
Brian Barrett
204f5b8f52 - Clean up wrapper compiler man pages during maintainer-clean, since
they might require special tools (not sure if sed with multiple -e
    arguments is totally portable)
  - ignore the opalcc.1 man page.  Couldn't do this in the previous
    man page commit (r12192) because I was removing opalcc.1 in that
    commit.

This commit was SVN r12194.

The following SVN revision numbers were found above:
  r12192 --> open-mpi/ompi@581a4b0a4e
2006-10-19 20:14:40 +00:00
Brian Barrett
e7a7a64e4c Implement MPI::SEEK_{SET, END, POS} for the C++ bindings, working around
some issues with the C #defines SEEK_{SET, END, POS}.  The workaround
involves some hackery that should work in almost every common use case
for the C stdio constants (and all the legal issues of the MPI constants).
The one issue is that the C stdio constants are now const ints instead
of #defines, which means that #ifdef checks will fail for the constants.

Behavior can be disabled at either configure time or build time.

Refs trac:387

This commit was SVN r12121.

The following Trac tickets were found above:
  Ticket 387 --> https://svn.open-mpi.org/trac/ompi/ticket/387
2006-10-15 23:50:24 +00:00
Brian Barrett
9adde4f7b8 Allow multilib capability based on compiler flags. See:
https://svn.open-mpi.org/trac/ompi/wiki/compilerwrapper3264
for more information.

Refs trac:374

This commit was SVN r12120.

The following Trac tickets were found above:
  Ticket 374 --> https://svn.open-mpi.org/trac/ompi/ticket/374
2006-10-15 21:21:08 +00:00
Dan Lacher
ba0389723e Ticket: #346
remove requirements on .la files on wrapper scripts

Ticket: #374
  extend compilers to support 32 bit and 64 bit in one version of the wrapper

Submitted by: Dan Lacher
Reviewed by: Rolf Vandevaart

This commit was SVN r11908.
2006-09-29 23:58:58 +00:00
Brian Barrett
bc9c6d65c6 The last of the Alpha fixes. The Alpha sh shell's builtin test doesn't
like == that much...

Refs trac:380

This commit was SVN r11860.

The following Trac tickets were found above:
  Ticket 380 --> https://svn.open-mpi.org/trac/ompi/ticket/380
2006-09-28 03:45:27 +00:00
Brian Barrett
4d7f221f86 There are some compilers (with the right options) that make LOGICALS
8 bytes.  Adjust the configure tests to allow for this case.

Refs trac:427

This commit was SVN r11859.

The following Trac tickets were found above:
  Ticket 427 --> https://svn.open-mpi.org/trac/ompi/ticket/427
2006-09-28 03:25:18 +00:00
Brian Barrett
9111f74627 More escaping fixes in the build system. Anything in the Makefile needs to
have $ escaped to $$.  Also, remove one more place for the shell to
expand those $s in the format string.

This commit was SVN r11809.
2006-09-26 02:37:35 +00:00
Brian Barrett
a546b6834b Only add the ROMIO source directory into DIST_SUBDIRS and SUBDIRS if
the component is configured successfully.  Otherwise, we can end up 
trying to run make in the romio directory without any Makefiles.  This
really only happens on the targets that recurse into DIST_SUBDIRS - ie
dist, maintainer-clean, and distclean

refs trac:411

This commit was SVN r11807.

The following Trac tickets were found above:
  Ticket 411 --> https://svn.open-mpi.org/trac/ompi/ticket/411
2006-09-25 23:51:13 +00:00
Brian Barrett
ad48aa82e9 A number of changes for improved Alpha support:
* Use $31 instead of mnemonic zero for the gcc inline
    assembly test, as the GNU assembler doesn't like
    zero, but both Tru64 and GNU assembler should be fine
    with $31
  * Disable Linux timer component on Alpha.  The CPU timer
    rolls over every 10 seconds or less, so it's kinda
    worthless for our needs.
  * Fix some escaping issues when local functions are
    denoted with a $
  * Remove C++ comments from the Alpha assembly.
  * Add base assembly code for the non-inlined functions
    on Alpha

This commit was SVN r11764.
2006-09-23 03:23:57 +00:00
Brian Barrett
e3497a8bff Fix a number of places where we can cause errors in configure if CFLAGS /
CXXFLAGS are set to -Wall -Werror.  Thanks to Ralf for the patch.

refs trac:290

This commit was SVN r11762.

The following Trac tickets were found above:
  Ticket 290 --> https://svn.open-mpi.org/trac/ompi/ticket/290
2006-09-22 23:58:23 +00:00
Tim Prins
83a7f6e4de Fix for bug #369.
LoadLeveler only sets LOADL_PROCESSOR_LIST when there are 128 or less tasks allocated to a job. The POE RAS relied on this variable so I created a new RAS which uses the LoadLeveler API instead of relying on the environment variable. This still needs some testing, so for now we use the POE RAS whenever LOADL_PROCESSOR_LIST, otherwise we fall back on this component.

Unfortunately, this will require an autogen...

This commit was SVN r11732.
2006-09-21 00:08:49 +00:00
Brian Barrett
8622fe8874 * be more liberal about about which platforms we consider to be alpha,
so that we include all the ev series on platforms that report more than
  just "alpha".  Fixes one of many issues on Alpha reported by a user.

refs trac:380

This commit was SVN r11683.

The following Trac tickets were found above:
  Ticket 380 --> https://svn.open-mpi.org/trac/ompi/ticket/380
2006-09-17 23:14:17 +00:00
Jeff Squyres
8226dab86c Fixes trac:377
Add --enable-orterun-prefix-by-default (and a synonym:
--enable-mpirun-prefix-by-default) to make orterun always behave as if
"--prefix $prefix" was given on the command line (where $prefix is the
value given to the --prefix option to configure).  This prevents many
rsh/ssh users from needing to modify their shell startup files to set
the LD_LIBRARY_PATH for Open MPI (they will still need to set PATH or
otherwise find the OMPI executables to mpicc/mpirun/etc. their MPI
applications).

Also added --noprefix option to orterun to disable this behavior.
Finally, note that even if --enable-orterun-prefix-by-default is
specified, if the user specifies --prefix or /path/to/mpirun, these
options will override the default value of the prefix ($prefix).

This commit was SVN r11669.

The following Trac tickets were found above:
  Ticket 377 --> https://svn.open-mpi.org/trac/ompi/ticket/377
2006-09-15 02:52:08 +00:00
Galen Shipman
877b819ddb Initial commit of QLogic PSM MTL.
This provides support for the Infinipath interconnect using the PSM API. 

Of note: 
This version has a "hackaround" we always return 1 or greater from
the MTL PSM progress function, this should be examined further.

This commit was SVN r11655.
2006-09-14 16:44:02 +00:00
Jeff Squyres
c9d244a298 Rename some OMPI_* macros to be OPAL_* macros.
This commit was SVN r11598.
2006-09-08 23:42:32 +00:00
Jeff Squyres
c068bc155a First steps towards IPv6 support. Mainly to support the guys working
on it, even though there's no other IPv6 code in the tree yet.

This commit was SVN r11561.
2006-09-08 00:10:40 +00:00
Sven Stork
78413e62b3 - fix lost compiler flag
This commit was SVN r11451.
2006-08-27 16:28:05 +00:00
George Bosilca
c697a25040 If we want to compile any application on Windows we need at least
these 2 libraries. Save and resotre (to the default values) the LDFLAGS.

This commit was SVN r11439.
2006-08-27 04:52:46 +00:00
Brian Barrett
53925dfd88 * Use AC_TRY_LINK instead of AC_TRY_cOMPILE to see if the compiler hints
for prefetch and branch prediction work.  A non-happy compiler could
  just think these were functions and we wouldn't get the error, because
  we didn't try to link.

  Refs trac:287

This commit was SVN r11333.

The following Trac tickets were found above:
  Ticket 287 --> https://svn.open-mpi.org/trac/ompi/ticket/287
2006-08-22 22:20:12 +00:00
George Bosilca
8db4a55a57 Reorder the C++ compiler discovery stages. Check first the compiler vendor
before checking if we are able to compile the test program. This is required
for windows as the C++ conftest.c file generated by configure cannot be
compiled with the Microsoft cl.exe compiler (because of the exit function
prototype). So if we detect a vendor equal to microsoft we will assume
that the compiler is correctly installed (which is true on Windows most
of the time anyway).

This commit was SVN r11268.
2006-08-20 13:53:27 +00:00
George Bosilca
6b01358cc1 Generate static-component.h file C++ friendly.
This commit was SVN r11267.
2006-08-20 13:48:24 +00:00
Brian Barrett
1daa21e1e3 It appears that most versions of the IBM XL compiler (including the latest
releases on Linux and OS X) don't handle const_cast<> of 2-dimensional 
arrays properly.  If we're using one of the compilers that isn't friendly
to such casts, fall back to a standard C-style cast.

refs: #271

This commit was SVN r11263.
2006-08-19 22:55:59 +00:00
Brian Barrett
43116cd13c * Don't define _GNU_SOURCE if we are using the IBM XL compilers, as it seems
to cause them to miscompile TCP-related code.

  refs trac:276

This commit was SVN r11258.

The following Trac tickets were found above:
  Ticket 276 --> https://svn.open-mpi.org/trac/ompi/ticket/276
2006-08-18 04:12:24 +00:00
Brian Barrett
4176e61049 * Add support for building the F90 bindings library as a shared library
on almost all platforms (except OS X... sigh...).  This is the merge 
  of r10846 - 10894 from the tmp/f90-shared branch to the trunk.

This commit was SVN r11103.

The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
  r10846
2006-08-03 00:17:31 +00:00
Rainer Keller
12166eb0d7 - The intel-based assembler on ia64 (such as NEC's ecc) needs
.proc/.endp-declarations for functions in order to be able to
   link successfully.
   Currently used in configure, only.
   
   There has not been found another arch, where this is necessary.
   So asm-data.txt and base/default.conf has not been changed.

This commit was SVN r11068.
2006-07-31 22:30:07 +00:00
Rainer Keller
bfadfa9eb6 - Get code-coverage to work with gcc-4.x (needs --coverage flag also
when linking)
   This allows ompi-branch optimization. Works now with mpicc wrappers.

This commit was SVN r11067.
2006-07-31 21:55:01 +00:00
Galen Shipman
c9e0eda190 Initialize the completion queue to a reasonable size based on maximum number
of send/receives outstanding.

Use ibv_cq_resize if available after initial creation of completion queue if
cq_size is too small (based on number of peers). 

This commit was SVN r11053.
2006-07-30 00:58:40 +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
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
Jeff Squyres
4eb3ee7835 Another project that has been brewing for a week or so...
We have repeatedly seen users inadvertantly try to use a C compiler
for $CXX (e.g., using icc instead of icpc in recent versions of the
Intel compiler).  Unfortunately, this would "sorta work", meaning that
configure would complete successfully and the build would fail much
later in the process (when $CXX was used to try to link a C++
compiler).  This was further compounded by the fact that many C
compilers will switch into "C++ mode" when they compile files that end
in .cc -- meaning that they'll *compile* C++ codes properly, but they
won't *link* properly.  Hence, users would get all the way down to
compiling the C++ MPI bindings or ompi_info (i.e., very late in the
build process) before the problem became evident.

We already have a test in configure that tries to compile, link, and
run a sample C++ program.  This helped ensure that $CXX was a valid
compiler, but it did not catch if the user accidentally supplied a C
compiler instead of a C++ compiler because the test program was simply
"return 0".  This commit updates the test program to use some
C++-specific constructs (std::string) so that if the user supplies a C
compiler in $CXX, the program may *compile*, but it will definitely
fail to *link*.

Hence, the process will fail early in configure (with a descriptive
message about how the compiler failed to work properly) rather than
late in the build.

This commit was SVN r10829.
2006-07-15 10:27:09 +00:00
Dan Lacher
94d5c9eed5 Defect 166: configure fails to crops libtool extras
Reviewed by: jsquyres, brbarret

This fix allows for support of LT2.0

This commit was SVN r10752.
2006-07-11 23:39:17 +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
Brian Barrett
47725c9b02 * Add new PML (CM) and network drivers (MTL) for high speed
interconnects that provide matching logic in the library.
  Currently includes support for MX and some support for
  Portals
* Fix overuse of proc_pml pointer on the ompi_proc structuer, 
  splitting into proc_pml for pml data and proc_bml for
  the BML endpoint data
* bug fixes in bsend init code, which wasn't being used by
  the OB1 or DR PMLs...

This commit was SVN r10642.
2006-07-04 01:20:20 +00:00
Brian Barrett
56a86c89b1 * Fix for bug #152. Make --with-FOO-libdir=DIR only check DIR, and not the
DIR/lib and DIR/lib64, which really never made sense.

This commit was SVN r10535.
2006-06-28 03:33:30 +00:00
Brian Barrett
03cc22842c fix error message typo and correct the flag that should be provided to gcc
to force v8+ ABI

This commit was SVN r10438.
2006-06-20 16:25:31 +00:00
Brian Barrett
d96554a482 * Fix for issue #144. With the opal_atomic_lifo code, we can no longer build
on platforms without a 32-bit compare and swap implementation.  There isn't
  such a thing on Sparc v8 chips, so we can't support those platforms right
  now.  Should be possible to make opal_atomic_lifo less atomic on these
  platforms, but not in the time allotted.

This commit was SVN r10407.
2006-06-17 18:45:29 +00:00
Sushant Sharma
428d81bc78 No need to look for xcpu libraries now. If user specifies --with-xcpu=yes, xcpu component will be compiled. Default is do not compile xcpu component.
This commit was SVN r10344.
2006-06-13 23:28:26 +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
0e6cf49341 * fix tests for __builtin_expect to not use NULL
This commit was SVN r10282.
2006-06-10 00:32:39 +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
d4b5c98247 Add check for <sysfs/libsysfs.h> since it's needed by OpenIB.
This commit was SVN r10213.
2006-06-05 22:02:40 +00:00
Brian Barrett
5163f2b296 Fix for bug #36. The MX, MVAPI, and OpenIB components don't have
support for progress threads, so we shouldn't build them or try to use
them when support for progress threads has been requested.  The TCP, GM,
SELF, and SM BTLs should have progress thread support, so they aren't
disabled.  The Portals BTL isn't compiled on platforms with threads,
so it doens't need to be updated.

This commit was SVN r10156.
2006-06-01 01:30:16 +00:00
Brian Barrett
0125bdb94d * add missing " that somehow slipped into the commit last night.
This commit was SVN r10140.
2006-05-31 13:57:38 +00:00
George Bosilca
2f4f49c7ba Check for the mx_register_unexp_callback availability (starting from MX 1.1.1).
This commit was SVN r10134.
2006-05-31 06:29:16 +00:00
Brian Barrett
7bbdad2821 * The (undocumented!) Autoconf variable $ac_configure_args needs
to be evaluated, to account for the quoting done.


Submitted by: Ralf Wildenhues
Reviewed by: Brian Barrett

This commit was SVN r10132.
2006-05-31 03:41:57 +00:00
Brian Barrett
f4a7e9be78 * Fix shell quoting to be more consistent with what AC does
Submitted by: Ralf Wildenhues
Reviewed by: Brian Barrett

This commit was SVN r10130.
2006-05-31 03:40:26 +00:00
Brian Barrett
7e1befaab8 * Fix M4 quoting so that AC 2.60 won't complain
Submitted by: Ralf Wildenhues
Reviewed by: Brian Barrett

This commit was SVN r10128.
2006-05-31 03:37:31 +00:00
Brian Barrett
0b09ba928c Fix for bug #17. Solaris provides stubs (who knows why, but they do) for
the pthreads library that don't do anything but are there when no special
options are given.  Both the GNU compiler and the Sun compiler could
sometimes ignore the -K* options, causing badness when building with
posix threads.  Don't try those options ;).

We still try -pthread and -pthreads because the compilers *do* error
when they see those options and some versions of the GNU compiler do
understand those flags (and do all the right things in that case).

This commit was SVN r10126.
2006-05-31 00:23:49 +00:00
Jeff Squyres
5f356edb64 Bring over changes from the /tmp/fortran-stuff series:
- Make the F90 bindings compile and link properly with gfortran 4.0,
  4.1, Intel 9.0, PGI 6.1, Sun (don't know version offhand -- the most
  current as of this writing, I think), and NAG 5.2, although some
  have limitations (e.g., NAG can't seem to handle the medium and
  large sizes)
- Building the F90 "small" module size is now the default, even for
  developers
- Split up mpif.h into multiple files because parts of it were toxic
  to the F90 bindings
- Properly specify unsized/unshaped arrays to make the bindings work
  on all known compilers
- Make ompi_info show Fortran 90 bindings size
- XML somewhat lags the generated scripts as of this commit, but
  functionality was my main goal -- the XML can be updated later (if
  at all).

This commit was SVN r10118.
2006-05-30 14:37:41 +00:00
Rainer Keller
772bba620d - Allow --enable-mca-direct for VPATH builds.
This commit was SVN r10007.
2006-05-22 14:24:30 +00:00
George Bosilca
1fbccda986 -g3 is definitively not a standard gcc option, at least not on anything
else than a quite recent version. Using this option prevent gdb from accessing
the contents of some of the structures. The error message is:
Unexpected type (0) encountered for integer constant.

This commit was SVN r9994.
2006-05-19 22:09:29 +00:00
Jeff Squyres
8e26fd653d Torque changed the name of their library from libpbs to libtorque; this
commit updates our configure test to check for both names.

This commit was SVN r9976.
2006-05-18 21:03:38 +00:00
Jeff Squyres
4080646606 Fix some variable names that are generated via m4 macros.
Sorry for committing configure changes during the day, but this is
necessary for some testing via multiple people at different sites.

This commit was SVN r9926.
2006-05-15 19:33:06 +00:00
Jeff Squyres
0d092abb81 Clarify help string, per change to 1.0 branch.
This commit was SVN r9906.
2006-05-12 03:06:54 +00:00
Andrew Friedley
5ba264d854 -ldapl isn't needed on all uDAPL platforms, so don't use it unless it is needed.
This commit was SVN r9866.
2006-05-09 20:02:38 +00:00
Brian Barrett
52369307f8 Add a feature to the build system that Terry from Sun and I talked about
in San Jose.  Allow the configure option --disable-binaries to build OMPI,
but not build or install the support binaries (so basically, just build
the libraries).

This commit was SVN r9777.
2006-04-29 02:16:41 +00:00
Jeff Squyres
2a2b27bf57 Forgot to remove f90_get_alignment.m4 from Makefile.am.
This commit was SVN r9725.
2006-04-26 12:07:26 +00:00
Jeff Squyres
79a3678924 Fix typo
This commit was SVN r9723.
2006-04-26 11:52:25 +00:00
Brian Barrett
5bd1be7ac4 * clean up some configure --help output as reported on OMPI mailing list.
This should probably go to 1.1...

This commit was SVN r9722.
2006-04-26 02:07:19 +00:00
Jeff Squyres
c6a753f41b Finish the rest of r9698:
- Add a lengthy comment explaining why we don't get F90 data type
  alignments
- Remove config/f90_check.m4 and its entry in acinclude.m4

This commit was SVN r9707.

The following SVN revision numbers were found above:
  r9698 --> open-mpi/ompi@e57300da4c
2006-04-25 10:45:06 +00:00
Jeff Squyres
e57300da4c Don't bother to get the alignmnet of an f90 type -- the DDT engine
only uses the F77 alignments anyway.  Plus, it is currently unknown
how to reliably get the F90 alignments because compilers *may* reorder
things and/or give different alignments than the f77 tests.

This commit was SVN r9698.
2006-04-24 14:28:51 +00:00
Brian Barrett
2ad29df0a1 * Add option to disable the adding of -g to CFLAGS/CXXFLAGS when --enable-debug
is given.  Generally not useful, unless you're on a platform without a
  debugger...

This commit was SVN r9684.
2006-04-22 19:23:26 +00:00
Rainer Keller
990496bc80 - Small bugfix to get lib-directory checking correct
(prevented orte/mca/pls/tm from detecting libpbs).

This commit was SVN r9657.
2006-04-19 15:22:55 +00:00
Brian Barrett
6617d48e00 * fix small issue with cross-compiling - we don't need
to run the working test when cross-compiling, so just
  ignore it for now..

This commit was SVN r9653.
2006-04-18 18:22:00 +00:00
Brian Barrett
d5012bf485 Two patches from Ralf Wildenhues to improve configure:
- include stdlib.h for tests that call exit(), as AC no longer
    prototypes exit in 2.60
  - Fix a compile issue in our cross-compiling test for C/C++
    alignment checks, matching what AC did for 2.60.

This commit was SVN r9634.
2006-04-14 15:33:51 +00:00
Jeff Squyres
f8e634d6ca Bring over /tmp/f90-stuff branch to the trunk.
svn merge -r 9453:9609 https://svn.open-mpi.org/svn/ompi/tmp/f90-stuff .

Several improvements over the current F90 MPI bindings:

- The capability to make 4 sizes of the F90 bindings:
  - trivial: only the F90-specific MPI functions (sizeof and a few
    others)
  - small: (this is the default) all MPI functions that do not take
    choice buffers
  - medium: small + all MPI functions that take one choice buffer
    (e.g., MPI_SEND)
  - large: all MPI functions, but those that take 2 choice buffers
    (e.g., MPI_GATHER) only allow both buffers to be of the same type
- Remove all non-standard MPI types (LOGICAL*x, CHARACTER*x)
- Remove use of selected_*_kind() and only use MPI-defined types
  (INTEGER*x, etc.)
- Decrease complexity of the F90 configure and build system

This commit was SVN r9610.
2006-04-11 03:33:38 +00:00
Sushant Sharma
8d5289b2b8 Corrected Makefile.am files for pls and soh xcpu-components as per Brian's suggestion.
This commit was SVN r9519.
2006-04-03 17:14:47 +00:00
Sushant Sharma
46f84b1e8e Added xcpu component in pls and soh.
This commit was SVN r9491.
2006-03-31 02:19:52 +00:00
George Bosilca
99ba9bea10 Correct the CRAY define (here is a interesting webpage with all the compilers defines
http://predef.sourceforge.net/precomp.html).

This commit was SVN r9472.
2006-03-30 00:20:43 +00:00
Brian Barrett
ae85aae6e5 * add some debugging if conftestval the file doesn't exist
* Try fclose()ing the file to see if that helps with the errors reported
  with Absoft 8.2

This commit was SVN r9439.
2006-03-28 13:55:39 +00:00
Brian Barrett
06e61ea03b * have to use same case for creating a function name as using it in the
Fortran tests or the mixed case (*caugh* Absoft *cough*) compilers
  don't compile the tests too well...

This commit was SVN r9406.
2006-03-24 02:34:13 +00:00
George Bosilca
aef1358808 First import or peruse. As it look like SVN doesn't like to import simultaneously
2 directories having the same name I have to split the import in 2. I start with
the test and the configure.

This commit was SVN r9372.
2006-03-23 04:54:10 +00:00
Brian Barrett
9bd236e1f0 * need to expand out $prefix if it is NONE as well
This commit was SVN r9346.
2006-03-21 00:21:12 +00:00
Brian Barrett
db5bc19785 * work around bumm assembly result on linux ppc32, which was claiming
support for 64 bit opcodes

This commit was SVN r9284.
2006-03-15 03:33:46 +00:00
Brian Barrett
1f6e85af4c Let's try this again, this time with less suck.
* Don't do the .in -> .tmp -> header thing for the prefixes and versions.
  It causes some severe cleanup issues all to save 4 files from rebuilding
  when configure is run.
* Clean up some makefiles so it's clear what is being installed/disted

This commit was SVN r9260.
2006-03-12 17:56:58 +00:00
Brian Barrett
ea7b9cfc81 * Only enable SLURM support in ORTE if on a platform currently supported by
SLURM.  Currently, this includes AIX and Linux.  If the user wants to build
  SLURM on another platform, they can specify --with-slurm.
* Enable/disable the SLURM sds component using the same logic as the PLS and
  RDS components.

This commit was SVN r9259.
2006-03-12 05:32:35 +00:00
Brian Barrett
c42da09796 * Fix a small bug George noticed - if you change the prefix (or any of the
installation directories) in configure, the files that depend on this
  information are not properly rebuilt.  If you need this information,
  don't setup a -D in the Makefile.am - instead, include 
  opal/install_dirs.h.
* Use the : option in AC_CONFIG_FILES to avoid needing to expose that
  we are playing around with temporary files with our headers to avoid
  rebuilding
* Clean up the version file information a bit, and like the install 
  directory stuff, make sure that there is a dependency so that 
  ompi_info gets rebuilt properly when a version number changes.

This commit was SVN r9256.
2006-03-12 04:35:01 +00:00
Brian Barrett
710bca2df4 * properly enable threading support inside ptmalloc2=. Thanks to Tim for
figuring out that ptmalloc2 was borked.

This commit was SVN r9251.
2006-03-11 02:37:55 +00:00
Brian Barrett
b57a6b4a62 * PGI 6.1 kind of (but not completely) supports GCC inline assembly. Previous
versions didn't even make an attempt.  Disable the not-quite-right issues
  inline assembly support with PGI 6.1.

This commit was SVN r9240.
2006-03-10 02:11:41 +00:00
George Bosilca
c604c5da2f -g3 is a GNU gcc specific flag. We are using the gnu compiler only if
the vendor is gnu. Otherwise, if GCC is set for all the compilers who
impersonate gcc like icc.

This commit was SVN r9217.
2006-03-08 17:17:17 +00:00
Jeff Squyres
fc2d95d263 Also recognize -g1, -g2, and -g3 as debugging flags so that the
now-default GNU debug flags (-g3) don't accidentally cause the
configure system to add -O3 to the build flags (because it didn't
recognize that -g3 was a debugging flag).

This commit was SVN r9207.
2006-03-06 17:32:51 +00:00
Rainer Keller
42f4a6fe03 - use AC_C_BIGENDIAN instead of deprecate AC_WORDS_BIGENDIAN
- set the highest debugging levels for macro debug support on gcc/gdb

This commit was SVN r9197.
2006-03-04 13:59:14 +00:00
Brian Barrett
c2fded586a * Andrew F. pointed out I can't spell...
This commit was SVN r9186.
2006-03-01 03:13:22 +00:00
Brian Barrett
7fa1c0eeb2 * if the default compile mode is 64bit, config.guess will call a MIPS
machine mips64

This commit was SVN r9185.
2006-03-01 02:25:09 +00:00
Brian Barrett
ee73cbb25c * remove unused F90 range and precision tests
This commit was SVN r9162.
2006-02-27 05:19:13 +00:00
George Bosilca
426c587dcc Dont use exit(0) when we try to run a simple C++ program. Use return instead
to avoid having problems on Windows with the C++ mangling scheme.

This commit was SVN r9161.
2006-02-27 05:14:35 +00:00
George Bosilca
f4b853d038 Allow Windows absolute prefix which start with a letter followed by :
This commit was SVN r9143.
2006-02-25 21:23:14 +00:00
George Bosilca
454e133ae2 Move all the Microsoft related checks outside the configure.ac
This commit was SVN r9100.
2006-02-21 07:42:00 +00:00
George Bosilca
890e2ec06e Accept Windows friendly paths for --prefix.
This commit was SVN r9099.
2006-02-21 07:41:07 +00:00
Jeff Squyres
e7528cad6f Remove AC_FC_SRCEXT(f95) because a) we don't have any files in the
source tree that are named .f95, and b) it sets the future file
extension for Fortran AC tests to be .f95.  This is not a problem for
gfortran, but other compilers (e.g., ifort) don't like that.

This commit was SVN r9040.
2006-02-14 20:28:18 +00:00
Jeff Squyres
65161c1dfe Fix the fortran 90 checks in two ways:
- Make the F90 checks use the identical types to the F77 checks (e.g.,
  INTEGER*4, as opposed to potential selected_int_kind() equivalents)
- Comment out the F77 and F90 alignment check failure -- there is some
  confusion here; some vendors apparently told Craig R. yesterday that
  they change the alignments of types based on whether the instance is
  a standalone variable, in an array, or a member of a struct.  

More investigation is needed (and potentially clarification for MPI
users -- this could be quite problematic!), but this alignment test is
otherwise hosing the nightlies, so it needs to be taken out for the
moment.

This commit was SVN r9031.
2006-02-14 13:05:37 +00:00
Jeff Squyres
681d46ac56 Fix problem with the AC_DEFINE's for the F90 check. Must use m4
mutilations in AC_DEFINE or autoheader will not pick up the symbols
and add them to opal_config.h.in.

This commit was SVN r9026.
2006-02-14 05:22:49 +00:00
Brian Barrett
74860649bc * skip some F90 tests if we aren't compilng F90
* Make sure to invalidate the cache on a couple of tests when we don't
  want F77/F90

This commit was SVN r8994.
2006-02-12 16:59:56 +00:00
Jeff Squyres
bc47a24bee Bring the f90 configuration up to speed with the f77 configuration
changes.  The two Big Changes are elegance (much more re-use of code
rather than cut-n-pasting the same code over and over and over and...)
and enabling cross-compilation for F77 and F90 (because we actually
have to *run* some compiled F77 and F90 programs for some of the
tests, which obviously won't work in a cross-compilation environment
-- so enable the use of config.cache to load such values in
cross-compiling environments).

This commit was SVN r8991.
2006-02-12 15:11:36 +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
Brian Barrett
d751d3bacf * Try to execute a simple application with a compiler (in non-cross-compile
situations) before going on to tests that run executables with the
  compiler.  Print a friendly error message if it fails.  Hopefully, this
  will help people with borked compilers.

This commit was SVN r8898.
2006-02-05 01:28:05 +00:00
Brian Barrett
c1390278ad * Look for the static version of the mpi F90 library, as it is always built
statically without libtool (so there is no .la file)

This commit was SVN r8896.
2006-02-04 23:30:20 +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
Brian Barrett
dc088d0023 * Generate ompi_get_version.sh from ompi_get_version.m4, removing a bunch of
duplicated logic that wasn't being duplicated all that well.  Thanks goes
  to Ralf Wildenhues for the bulk of the patch.

This commit was SVN r8779.
2006-01-21 22:53:16 +00:00
Brian Barrett
22c50c2cca * allow loading of configure options from an external file, for use in
cross-compile environment

This commit was SVN r8706.
2006-01-16 23:38:42 +00:00
Brian Barrett
17d2c907fc Some changes to better support cross-compiling OMPI:
- fall back to compile test for windows paffinity component
    when cross compiling
  - fall back to platform guess when checking for threads having
    different pids with pthreads (yes on linux, no elsewhere)
  - pass the proper host, target, and build flags to the
    ROMIO configure script

With these changes, cross-compiling should be possible with the exception
of the Fortran 77 and Fortran 90 bindings.  Fortran 77 can be cross-
compiled if cache values are provided for type sizes and alignment.

This commit was SVN r8702.
2006-01-16 04:00:44 +00:00
Andrew Friedley
a4abe3bdbe Checkpoint:
- Borrow configure.m4 from the mvapi btl.  One of the uDAPL headers emits a
   warning when -pedantic is enabled, so strip it out.
 - Change function check in ompi_check_dapl.m4 from dat_ia_open to
   dat_registry_list_providers.. dat_ia_open wasn't working right
 - Make the references to prepare_dst, put, and get NULL for now
 - Add opal_output() calls in all the udapl interface functions for debugging
 - Add evd_qlen component parameter to control event dispatcher queue length
 - First stab at component_init and module_init
 - Misc cleanups - whitespace, dead code removal
 - Update copyrights to 2006

This commit was SVN r8701.
2006-01-16 03:01:12 +00:00
Brian Barrett
c96f870674 * Merge of wrapper compiler updates from the bwb-wrapper-fix branch (r8690 -
r8698), with changes below:

  - Split wrapper flags into those required for each of the three projects,
    and cleaned up some cruft (including the LIBMPI_EXTRA_*FLAGS) through-
    out the build system
  - Added opal_init_util and opal_finalize_util to allow init / cleanup
    of all the opal code that doesn't require the MCA system
  - Create standalone key=value file parser, based on the one that used
    to be in the mca param parser, so that it can be shared in multiple
    places
  - Add wrapper datafiles for opal, orte, and ompi wrappers, and add
    wrapper compiler with support for all the old features

This commit was SVN r8699.

The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
  r8690
  r8698
2006-01-16 01:48:03 +00:00
George Bosilca
5f37949b10 Correct few typos and make the "no fortran" a valid choice.
This commit was SVN r8689.
2006-01-14 00:20:13 +00:00
Jeff Squyres
74d759dc47 Patch submitted by Roland Dreier to look for a change in the OpenIB
API.

This commit was SVN r8687.
2006-01-13 23:20:49 +00:00
Brian Barrett
92aaaad611 * Fix issue when a type existed but we didn't have a corresponding C type -
the failure wasn't being properly propogated up from ompi_find_type and
  resulting in bad #define values in ompi_config.h
* Fix issue where we could emit illegal sh code if we were checking for a
  type with no corresponding C types listed.  Thanks to Ralf for tracking
  this one down.
* Fix a couple more messages to match all the others.

This commit was SVN r8685.
2006-01-13 14:51:20 +00:00
Brian Barrett
48f82db838 Convert all Fortran 77 tests to use config cache so that we have a way to
determine values like Fortran alignment (which can only be determined by
running a program) when cross-compiling.  By providing cache values, the
programs will not be run at all, and life will be good.  Also clean up
some macro interfaces so that they are a bit easier to use, at the cost
of horrid internals ;).

This commit was SVN r8684.
2006-01-13 04:08:40 +00:00
Andrew Friedley
c0bad339af - Use the GM BTL as a template instead, per Tim's suggestion
- Begin adding uDAPL-specific stuff
- Added config/ompi_check_udapl.m4 - hopefully I did this right

This commit was SVN r8681.
2006-01-12 04:05:02 +00:00
Brian Barrett
80c5e5c8ba * fix assembly issues for linux, which wants a notation to make the stack
non-executable.

  This should go to the v1.0 branch

This commit was SVN r8673.
2006-01-11 04:49:40 +00:00
Brian Barrett
9310fd6f85 When debugging code is turned on with --enable-debug, try to use error checking
mutexes instead of fast mutexes.  If an error occurs, a message will be
printed, and abort() will be called.

This commit was SVN r8671.
2006-01-11 04:34:29 +00:00
George Bosilca
670716c489 There is no need to do the checking for socklen_t twice.
This commit was SVN r8612.
2005-12-31 12:19:27 +00:00
Brian Barrett
f74e24df69 * don't declare functions being checked as extern char * - that's not a
function and causes problems with false negatives on some linux distros.
  Instead, use the test code used by AC_CHECK_FUNC.

This commit was SVN r8592.
2005-12-22 14:26:24 +00:00
Jeff Squyres
7497babb1f - Add an explicit example to the --enable-mca-no-build option because
I never remember the right format.  :-)
- Some capitolization consistency fixes.

This commit was SVN r8576.
2005-12-21 12:47:02 +00:00
Brian Barrett
f9018a148c per conversation with George, don't try to figure out all the assembly
information on Windows with the CL compiler.  We don't use it, and we
were running into trouble with the compiler on some tests

This commit was SVN r8558.
2005-12-20 01:34:27 +00:00
Brian Barrett
79396ea5a5 * Fix small issues when using an assembler directly, rather than through
a compiler
* Work around fact that ias (Intel Assembler) emits a banner at all times...

This commit was SVN r8555.
2005-12-20 00:49:59 +00:00
Brian Barrett
08799e4a48 * Make the format of the --with-BTL options the same for all the BTLs
* Make sure --without-BTL works for all BTLs
* Fix copy-n-paste error in aix timer configrue help string

This should go to the v1.0 branch

This commit was SVN r8554.
2005-12-20 00:47:36 +00:00
Brian Barrett
a5af07cd6b fixes suggested by Ralf for supporting both Libtool 1 and 2 in Open MPI...
This commit was SVN r8538.
2005-12-19 03:10:23 +00:00
George Bosilca
bb26d44166 Comment out the second check for socklen_t. We already check it once in the
main configure. Should be removed in the near future.

This commit was SVN r8536.
2005-12-17 22:12:45 +00:00
Brian Barrett
32e01a8e80 * one more thing that we need to ignore (given by libtool 2.0 with CL)
This commit was SVN r8503.
2005-12-15 05:20:15 +00:00
Brian Barrett
e574fd7fbe - update some svn ignores
- don't fail to configure if threads aren't found, since
  we don't have thread support on windows
- rather than setting the asm file name to none in asm-data,
  special case windows w/ CL in the ompi_config_asm macros.
  Otherwise, there were some warnings during make dist and
  configure that didn't need to be there.

This commit was SVN r8502.
2005-12-15 04:19:16 +00:00
Brian Barrett
e8cca1710f couple of changes to make Open MPI compile under Cygwin with gcc:
- add the right asm format
  - add checks for some constants / fields that cygwin
    doesn't have in the stacktrace code
  - fix for slightly more verbose libtool 2 betas that
    have multiple lines for link output

This commit was SVN r8501.
2005-12-15 00:51:28 +00:00
George Bosilca
29c75a4690 libtool 2.0 get confused if the source file is followed by -c (or any other
value that is not internally translated). In order to make all the version
of libtool happy, the simplest way is to have the source file at the end of
the command line.

This commit was SVN r8451.
2005-12-12 06:41:48 +00:00
Brian Barrett
73f4f407c1 Extend test for compiler vendor to include many, many more compilers, and
unify the C / C++ tests, as they are often the same defines to check...

This commit was SVN r8450.
2005-12-12 00:41:01 +00:00
George Bosilca
1602dd12ba Typo correction. GNU gcc define __GNUC__ not __GNU__.
This commit was SVN r8423.
2005-12-09 05:09:37 +00:00
George Bosilca
93cff86017 Include the vendor and nickname detection directly in the setup macros.
This commit was SVN r8422.
2005-12-09 04:19:53 +00:00
George Bosilca
66629c0974 Two scripts to detect the flavor and vendor of the used compiler. The flavor
is a short name usually the one that the compiler have. If we are using
wrappers around a compiler the flavor will be the real name of the compiler.
The vendor is a short name for the maker of the compiler. The GNU compilers
get gnu and the microsoft one get microsoft.

These 2 macros provide a simple way to check for specific brand of compilers
and to generate the correct code to test with the autoconf macros. They are
not yet completed, I add just some of the compilers/vendors but feel free to
add more if you need/want.

This commit was SVN r8417.
2005-12-08 17:40:32 +00:00
Brian Barrett
996fc4ca8e * allow both mx and gm to be disabled with --without-{gm,mx}
This commit was SVN r8411.
2005-12-08 04:54:42 +00:00
Brian Barrett
372f9b0f29 * it appears Solaris on Opteron identifies itself as i386-pc-solaris2.10, not
x86_64 as most opterons do.  So add test to see how many bytes a long is,
  and if it's 8 bytes, assume we're on an AMD64 instead of an IA32 for the
  purposes of assembly.

This commit was SVN r8384.
2005-12-06 13:31:05 +00:00
Brian Barrett
3a07de29f3 - add -lgen if it's needed to get dirname()
- since we now have three places where we need to add a -l<foo> if
  and only if it isn't already in something in LIBS that provides
  symbol bar...

This commit was SVN r8370.
2005-12-02 01:02:19 +00:00
Jeff Squyres
0fc36a016f Patch from David Daniel to add $LDFLAGS and $LIBS into the f90 tests
so that they're just like the f77 tests.

This commit was SVN r8348.
2005-12-01 13:19:32 +00:00
George Bosilca
906d19a73e A simpler version for the C function who is supposed to print the value of TRUE from Fortran. We support
only the fortran compilers who define the LOGICAL as being at most of the same size as an int.

This commit was SVN r8306.
2005-11-29 06:35:46 +00:00
Galen Shipman
ff1a820384 set the lib path correctly, build system doesn't like white space..
This commit was SVN r8291.
2005-11-28 21:08:39 +00:00
Galen Shipman
209bc9ca93 fix..
This commit was SVN r8288.
2005-11-28 20:44:53 +00:00
George Bosilca
c9c205302d Correct the TRUE detection code to be more Fortran friendly and not based on some compiler
features. Now it works with gfortran as well.

The problem was that the cast between LOGICAL and INTEGER is not a Fortran language feature. The easiest solution around this problem is to have a C function that print the value. We set the value in fortran and then call the C function to print it out.

This commit was SVN r8266.
2005-11-28 16:26:00 +00:00
Rainer Keller
bf0892bb32 - Implement correct Fortran Logical-handling in f77/f90 interface in
case of:
    sizeof(MPI_Flogical) != sizeof (int)
  and
    Fortran value of .TRUE. != 1
  as is often the case.
- Check in configure the value of .TRUE., the C-type coresponding to
  logical and check, that fortran compiler does not do something strange
  with arrays of logicals
- Convert all occurrences of logicals in the fortran wrappers, only
  in case it is needed.
  *Please note* Implementation of MPI_Cart_sub needed special treatment.
- Output these value in ompi_info -a
- Clean up the prototypes_mpi.h to just have a single definition and
  thereby deleting the necessity for prototypes_pmpi.h

- configured, compiled and tested with F90-program, which uses
  MPI_Cart_create and MPI_Cart_get:
  linux ia32, gcc (no testing, as no f90)
  linux ia32, gcc --disable-mpi-f77 --disable-mpi-f90 (had a bug there)
  linux ia32, icc-8.1
  linux opteron, gcc-3.3.5, pgcc, pathccx/pathf90 (tested just
pgi-compiler)
  linux em64t, gcc, icc-8.1 (tested just icc)

This commit was SVN r8254.
2005-11-24 16:52:35 +00:00
Brian Barrett
129451277e * add missing constant when using the MPI-2 onesided shell functions.
This should probably go to the branch

This commit was SVN r8222.
2005-11-21 23:43:48 +00:00
Brian Barrett
8faa1884f0 * The last of the build system optimizations. Combine the component and
component/base Makefile.am files, reducing the time configure spends
  stamping out Makefiles at the end
* Install base_impl.h file when devel-headers are being installed

This commit was SVN r8200.
2005-11-20 01:03:01 +00:00
Brian Barrett
e5525ad5f8 fix for issue found by Dries Kimpe -- if a user specifies two identical
options as part of the --with-wrapper-{*flags,libs} options to configure,
we'll eat one of them, which can have unintended issues.  For example,

  --with-wrapper-ldflags="-Wl,-rpath -Wl,pathA -Wl,-rpath -Wl,pathB"

will result in us removing the second -Wl,-rpath, which will result
in wrapper ldflags of -Wl,-rpath -Wl,pathA -Wl,pathB which doesn't
really do what the user intended.  We now always pass the user-
specified flags completely unmodified, bypassing the OMPI_UNIQ
routine

This commit was SVN r8193.
2005-11-18 20:55:38 +00:00
Brian Barrett
1066518f3b Fix output of configure --help for the --with-threads option to be posix
instead of pthread.  The code expects posix.

This commit was SVN r8130.
2005-11-12 03:10:52 +00:00
Brian Barrett
660d2f61b6 Don't add external declarations for the PMPI_W{TICK,TIME} functions
if profiling isn't enabled.  It appers that some compilers (g95)
will try to resolve the symbols if they are prototyped.

This commit was SVN r8110.
2005-11-11 00:12:40 +00:00
Jeff Squyres
bcd037315f Some Fortran compilers actually will return that a type exists even if
it doesn't support it -- the compiler will automatically convert the
unsupported type to a type that it *does* support.  For example, if
you try to use INTEGER*16 and the compiler doesn't support it, it may
well automatically convert it to INTEGER*8 for you (!).  So we have to
check the actual size of the type once we determine that the compiler
doesn't error if we try to use it (i.e,. the compiler *might* support
that type).  If the size doesn't match the expected size, then the
compiler doesn't really support it.

The F77 configure code actually handled this properly.  The F90 code
did not quite do it right.  This patch brings the F90 code up to the
same structure as the F77 code, albiet not m4-ized properly.  I also
added a comment to config/f77_check.m4 that explains *why* we do this
extra size check (because no explanation was given).

The impetus for this was that xlf* on OS X 10.3 was not recognizing
that INTEGER*16 was not supported, and mpi-f90-interfaces.h was being
assembled incorrectly.  This patch fixes this problem.

There is still one more problem, but waiting for some help from Craig
R on that (function pointers in F90 declarations).

This commit was SVN r8107.
2005-11-10 23:35:36 +00:00
Brian Barrett
5bf0a7bc62 * allow for the fact that svnversion might fail
This commit was SVN r8077.
2005-11-10 02:00:38 +00:00
Brian Barrett
bc8e6300fd Clean up the TM configure macro. Original bug was that it didn't add -lpbs to
the component's LIBS if the library was found in the default search paths.
However, it would also add -Is for default search paths and other nasty things,
so it seemed best to convert it to the OMPI_CHECK_PACKAGE macro so that it
behaved like all the other packages out there.

This commit was SVN r8007.
2005-11-07 13:21:46 +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
Brian Barrett
1c47b56163 * add ability to add compiler / linker flags to the MPI wrapper compilers
from configure (in addition to all the ones that the configure script
  finds)

This commit was SVN r7906.
2005-10-27 23:23:08 +00:00
Brian Barrett
65bcc283c0 * Change the --enable-{cxx,f77,f90} options to
--enable-mpi-{cxx,f77,f90} so that people aren't confused about what
  they are actually disabling.

  This should go to the 1.0 branch

This commit was SVN r7851.
2005-10-25 02:53:54 +00:00
Brian Barrett
76d1a808e9 disable GCC-style inline assembly when on AIX. The IBM compiler does a really,
really crappy job of trying to emulate the inline assembly mode of GCC (and will
completely rewite the assembly, which seems to be bad in my opinion).  GCC and
the AIX assembler don't see eye-to-eye on what GCC emits when doing inline
assembly.  That's two compilers and no actual working support.  So just punt and
fall back to XLC inline assembly or non-inlined assembly.

This commit was SVN r7800.
2005-10-18 19:04:35 +00:00
Brian Barrett
1302cb4072 The next in a long line of crazed build system changes from Brian. This was
originally suggested by Ralf Wildenhues, to try to speed autogen, configure,
and make (and possibly even make install).  Use automake's include directive
to drastically reduce the number of Makefile files (although the number of
Makefile.am files is the same - most are just included in a top-level
Makefile.am).  Also use an Automake SUBDIRs feature to eliminate the
dynamic-mca tree, which was no longer really needed.  This makes adding
a framework easier (since you don't have to remember the dynamic-mca
tree) and makes building faster (as make doesn't have to recurse through
the dynamic-mca tree)

This commit was SVN r7777.
2005-10-17 00:21:10 +00:00
Jeff Squyres
c760d51cc2 Fixes for INTEGER and int max values; original problem report from
Brooks Davis.  Now handle everything as hex, and we don't have to
worry about signed/unsigned implementations of integers in shell and
expr. 

This commit was SVN r7729.
2005-10-12 16:08:17 +00:00
Brian Barrett
6ee4116b22 * properly restore CPPFLAGS so that they don't accidently disappear down
the line

This commit was SVN r7718.
2005-10-12 03:59:48 +00:00
Brian Barrett
4df7d936bf make dist would create the VERSION file with incorrect timestamps if running in
a VPATH tree.  Fix that problem, as well as some potential shell escaping badness
if a version number or path ever had a space in it.

This commit was SVN r7586.
2005-10-02 21:14:50 +00:00
Jeff Squyres
e72e1f0050 Fix some incorrect fortran parameter values
This commit was SVN r7584.
2005-10-02 14:59:27 +00:00
Brian Barrett
57781484ef * check for shared receive queue support, and set OMPI_MCA_BTL_OPENIB_HAVE_SRQ
if we have said support

This commit was SVN r7575.
2005-10-01 18:44:31 +00:00
Brian Barrett
997644af31 * There are now two forms of ibv_create_cq, one with 3 params and one with 5.
Try to detect which form this version of Open IB uses, defaulting to the 5
  version if we can't figure it out (the new version has 5 params)
* Only add -lcm if it exists on the system - some versions of Open IB
  apparently don't need it.

This commit was SVN r7542.
2005-09-29 13:35:57 +00:00
Brian Barrett
80ac5c2efd * there are now two upcoming points where we want to release a version with
a random string of characters as part of the version number (the really
  soon to happen 1.0lanl release and the 1.1sc2005 release that we've
  talked about).  So rather than having alpha and beta fields that must
  be numeric values, have a general field that can be any alphanumeric
  value.

This commit was SVN r7511.
2005-09-27 02:06:05 +00:00
Jeff Squyres
fbd6db142b Remove some old references to --with-btl-mvapi (replaced with
--with-mvapi).

This commit was SVN r7495.
2005-09-23 13:18:41 +00:00
Jeff Squyres
4637ca6027 Use AC_FD_CC uniformly (one & was missing).
This commit was SVN r7407.
2005-09-16 12:28:53 +00:00
Brian Barrett
2787d993a9 * Add checks for fork/execve/setpgid for slurm components so that they
automagically don't build on platforms without such things
* Fix for mistaken use of cache variable in assembly setup
* one more cached test hits the books

This commit was SVN r7404.
2005-09-16 04:51:09 +00:00
Brian Barrett
cf04b2da3c * bunch of changes to make better use of config cache. Almost everything in the
C / C++ tests now use the caching scheme.  Only exception is C++ exception
  handling code, which is disabled by default, so I'm not too worried about making
  that configure faster :)

This commit was SVN r7403.
2005-09-16 02:49:31 +00:00
Brian Barrett
0bdbdb0cf8 * It seems that some versions of the Intel compiler don't properly handle
the way we use weak symbols (you have to use the actual symbol in the
  .o file or the weak symbol isn't generated).  It worked for our test
  case, but not for libmpi.  Make our test case look more like what we
  do in libmpi by generating a .o file with a weak symbol and attempting
  to use it from another object file.

This commit was SVN r7360.
2005-09-14 01:40:25 +00:00
George Bosilca
6a4f4acebf Typo.
This commit was SVN r7359.
2005-09-13 22:28:51 +00:00
Jeff Squyres
809a31518e Some changes:
- fix the first test
- make the #define's always have a value (0 or 1)
- prefix them with OMPI
- remove the extra "__"'s in the names, since they're now
  OMPI-specific #define names

This commit was SVN r7339.
2005-09-13 12:26:45 +00:00
George Bosilca
7a9a427c30 Check for __attribute__ to support advanced flags on the functions and structures.
They can be used to give additional knowledge to the compiler about the behaviour of a
function so it can optimize further. In the begining we just need it for the alias weak
a better way to create function aliases.

The problem was that the intel compiler version 8 to 9 support pragma weak but do not include
the function in the .o file if it's not used at least once in the same .c file. With the
alias weak this problem is solved as even intel support it correctly.

This commit was SVN r7330.
2005-09-12 22:49:09 +00:00
Jeff Squyres
67cde6c212 - Minor cleanups
- Add --enable-trace which turns on some internal tracing and dumps a
  file per process in the session directory tree.  Meant for internal
  developer tracing, *not* for tracing MPI applications in the
  traditional sense.

This commit was SVN r7229.
2005-09-08 09:44:50 +00:00
Brian Barrett
aee3316243 * just for George, don't try to get the SVN r number if we don't need it. This
should speed up autogen.sh on platforms with really slow svnversion searches,
  as we only need the base version for autogen, which means we don't need the
  r number.
* Since I can't type, add an error message if an invalid argument is given

This commit was SVN r7224.
2005-09-07 19:59:28 +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
Brian Barrett
17f60894c5 * per conversation with Tim, registration macro should be on by default
This commit was SVN r7210.
2005-09-07 05:01:51 +00:00