1
1
Граф коммитов

1174 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
26f6d9cf0c Properly protect the use of dev->transport_type and
IBV_TRANSPORT_IWARP.

This commit was SVN r18738.
2008-06-25 14:50:59 +00:00
Jeff Squyres
64034b65d0 Remove debugging message kruft
This commit was SVN r18737.
2008-06-25 11:32:33 +00:00
George Bosilca
df9e54194b Fix a deadlock in the put protocol for MX. The callback was never triggered,
as the send and the put share the same execution path and in the case of the
put the MCA_BTL_DES_SEND_ALWAYS_CALLBACK was not set.

This commit was SVN r18735.
2008-06-25 03:45:30 +00:00
George Bosilca
24316b77cf This commit is related to ticket #1362. The following devices respect the
behavior described on the ticket: elan, gm, mx, self, sm, tcp.

This commit was SVN r18734.
2008-06-25 03:35:07 +00:00
Jeff Squyres
edad3b66a2 Add warning message if you request one max_inline_data value and the
device reduces it.

This commit was SVN r18728.
2008-06-24 20:40:03 +00:00
Jon Mason
eefc8956b6 Fix a bug introduced in r18725
ompi_info will crash due to a reference to a unalloced struct.  Add a check for
this struct to prevent this crash.

This commit was SVN r18727.

The following SVN revision numbers were found above:
  r18725 --> open-mpi/ompi@c2351d39f1
2008-06-24 20:38:47 +00:00
Jon Mason
c2351d39f1 Fix Ticket #1359
If rdmacm cpc is disabled and an iwarp device is forced to be used, then
mca_btl_openib_get_iwarp_subnet_id will attempt to reference an uninitalized
list.  Modify the code to check for the uninitalized list and return zero.

Also, fix a memory leak caused by not freeing alloced addresses structs during
shutdown. 

This commit was SVN r18725.
2008-06-24 19:16:06 +00:00
George Bosilca
fbf341068a Remove the pending queue from the shared memory BTL. The PML is in charge of managing
the fragments that failed to be send, there is no need to replicate the same
mechanism in the BTL.
Force the SM BTL to empty all ack fragments in the component progress function.

This commit was SVN r18724.
2008-06-24 19:01:26 +00:00
Jeff Squyres
ea21c31f44 * MCA params btl_openib_use_eager_rdma can now override the
INI file use_eager_rdma value (fixes trac:1169)
 * fixed a typo in a MCA param help message
 * made the check for enabling short/eager RDMA more robust in the
   presence of progress threads; it now emits a show_help warning

This commit was SVN r18723.

The following Trac tickets were found above:
  Ticket 1169 --> https://svn.open-mpi.org/trac/ompi/ticket/1169
2008-06-24 18:31:46 +00:00
Jeff Squyres
e0545460ff Fixes trac:1355: allow INI file to set max_inline_data vale, and if not
specified, probe for max value supported by device.

This commit was SVN r18720.

The following Trac tickets were found above:
  Ticket 1355 --> https://svn.open-mpi.org/trac/ompi/ticket/1355
2008-06-24 17:18:07 +00:00
George Bosilca
1eb62b6c48 Remove a warning. Close ticket #1357.
This commit was SVN r18717.
2008-06-24 14:23:02 +00:00
Jeff Squyres
3f95b906c5 Fixes trac:1085: Improves SLURM configure logic to also allow OS X or any platform where srun is found in the PATH.
This commit was SVN r18714.

The following Trac tickets were found above:
  Ticket 1085 --> https://svn.open-mpi.org/trac/ompi/ticket/1085
2008-06-23 23:12:55 +00:00
Jeff Squyres
281c37afcc Ensure to ignoe the "empty" CPC components.
This commit was SVN r18702.
2008-06-21 11:39:53 +00:00
Jeff Squyres
807e2cc742 Mark a notable place where we need to return an error up to the BTL or
PML.

This commit was SVN r18701.
2008-06-20 22:11:49 +00:00
Jeff Squyres
5ded50df0e * Fix a > that should be ==
* Ensure to destroy the correct QP (local->id[num]->qp will always
   have a valid pointer in it, even if we setup a dummy qp)
 * Note two notable places where we need to figure out how to
   propagate errors up from the CPC to the main BTL / PML when errors
   occur.  Probably have the same issue in IBCM, too.

This commit was SVN r18700.
2008-06-20 22:09:30 +00:00
Jeff Squyres
0074126886 Per #1352, most iWARP adapters today cannot handle connections between
two processes on the same server (!).  So for today, we'll simply mark
all local processes that use iWARP adapters as "unreachable".

More details in #1352.

This commit was SVN r18699.
2008-06-20 22:08:00 +00:00
Jeff Squyres
f4145fce7a Ensure that we don't try to shut down a thread that is not [yet] there
(e.g., if you're excluding some devices, their destructors will be
invoked before the async event thread was setup for them).

This commit was SVN r18698.
2008-06-20 19:30:51 +00:00
Jeff Squyres
ed17b51204 Adjust the max_inline default size down so that it can be accepted on
multiple adapters (eg., Chelsio T3).

But we need to figure out how to determine a good value for the
resident adapter(s) at runtime.  It's problematic because, for
example, Mellanox ConnectX and Chelsio T3 report max_inline values
differently at run-time.  If you ibv_create_qp with a max_inline value
of 0, ConnectX reports back a value that is a formular based on a few
other values (e.g., max_send_sge and max_recv_sge).  But T3 always
reports back "64".

We're looking into this to figure out the best way -- reducing the
default right now should allow other adapters to run while we figure
it out.

This commit was SVN r18697.
2008-06-20 18:24:04 +00:00
George Bosilca
54e7e03695 One less warning.
This commit was SVN r18695.
2008-06-20 17:50:19 +00:00
Jeff Squyres
7905db57bd Slightly decrease the number of buffers for the NetXen adapter
This commit was SVN r18691.
2008-06-20 01:00:22 +00:00
Pavel Shamis
4537827973 Making the qp allocation more optimized.
- sq parameter was replaced with max_inline parameter
- inline is allocated only for relevant QPs

This commit was SVN r18675.
2008-06-19 08:40:39 +00:00
Lenny Verkhovsky
f4811d6c4d NUMA Awareness support. Gleb's patch
This commit was SVN r18658.
2008-06-15 13:43:28 +00:00
Galen Shipman
44cd373a87 I also forgot to initialize the convertor max_data, george probably copied
this dumb mistake from me. 

This commit was SVN r18653.
2008-06-13 18:33:43 +00:00
George Bosilca
170b9c344e Mea culpa. I forget to initialize the max_data before the call to the
convertor.

This commit was SVN r18651.
2008-06-12 17:24:39 +00:00
Pavel Shamis
dc3f14736d Fixing QP initialization stuff.
This commit was SVN r18650.
2008-06-11 16:31:39 +00:00
Galen Shipman
a239877b78 revert my previous boneheadedness
This commit was SVN r18634.
2008-06-10 01:19:04 +00:00
Galen Shipman
4ef4a9520f remove showhelp..
This commit was SVN r18628.
2008-06-09 20:53:01 +00:00
Galen Shipman
9efbec0383 fix normal send path
remove unneeded checks

This commit was SVN r18624.
2008-06-09 20:25:27 +00:00
Ralph Castain
9613b3176c Effectively revert the orte_output system and return to direct use of opal_output at all levels. Retain the orte_show_help subsystem to allow aggregation of show_help messages at the HNP.
After much work by Jeff and myself, and quite a lot of discussion, it has become clear that we simply cannot resolve the infinite loops caused by RML-involved subsystems calling orte_output. The original rationale for the change to orte_output has also been reduced by shifting the output of XML-formatted vs human readable messages to an alternative approach.

I have globally replaced the orte_output/ORTE_OUTPUT calls in the code base, as well as the corresponding .h file name. I have test compiled and run this on the various environments within my reach, so hopefully this will prove minimally disruptive.

This commit was SVN r18619.
2008-06-09 14:53:58 +00:00
Jeff Squyres
1a748bc7be First cut at the NetEffect NE020 NIC.
This commit was SVN r18599.
2008-06-05 20:24:24 +00:00
Jeff Squyres
f0d465c30a Slightly simplify the code and remove a compiler warning.
This commit was SVN r18596.
2008-06-05 19:08:08 +00:00
Jeff Squyres
b1999bbba3 * Use inclusive NIC/HCA language
* Add a description of receive_queues

This commit was SVN r18595.
2008-06-05 19:07:22 +00:00
Pavel Shamis
7b9024bc05 Updating Mellanox's Copyright in files touched in 2008
This commit was SVN r18592.
2008-06-05 13:40:26 +00:00
Pavel Shamis
379e00050c Fixing openib btl finalize flow. Bug fix for #1286.
This commit was SVN r18590.
2008-06-05 12:20:13 +00:00
Jeff Squyres
91a281080a Fix a compiler warning for a case that would never really happen
anyway.  Rename a variable to be a bit more descriptive.

This commit was SVN r18585.
2008-06-04 19:10:23 +00:00
Jeff Squyres
bc584dedd6 Remove a compiler warning that would never happen in practice.
This commit was SVN r18584.
2008-06-04 19:03:02 +00:00
Jeff Squyres
6e37dd0ef0 Fix some 32/64 printf errors once and for all
This commit was SVN r18582.
2008-06-04 14:39:37 +00:00
Pavel Shamis
0a8321e08d Calls to APM functions should be protected with OMPI_HAVE_THREADS.
This commit was SVN r18581.
2008-06-04 14:27:41 +00:00
Jeff Squyres
5e918ad25d Add first cut of NetXen iWARP NIC definition. May still be refined
with more experimentation.

This commit was SVN r18580.
2008-06-04 12:11:45 +00:00
Pavel Shamis
c73ed2b256 Updating cpc name from xrc to xoob.
This commit was SVN r18571.
2008-06-04 08:50:30 +00:00
Ralph Castain
c992e99035 Remove the tags from orte_output_open and the filtering operation from orte_output - this will be handled differently to improve the XML output interface
This commit was SVN r18557.
2008-06-03 14:24:01 +00:00
Jeff Squyres
69d78c6739 Fixes trac:1215: adds specific show_help messages about PP vs. SRQ/XRC RNR
retry exceeded errors.

This commit was SVN r18554.

The following Trac tickets were found above:
  Ticket 1215 --> https://svn.open-mpi.org/trac/ompi/ticket/1215
2008-06-02 11:03:48 +00:00
Jeff Squyres
8c267d50a3 Fixes trac:1121.
We already show_help when we fail to create queues, so I just made the
message a little more verbose such that it may be that OMPI is trying
to use a feature that is not supported on the hardware.

This commit was SVN r18553.

The following Trac tickets were found above:
  Ticket 1121 --> https://svn.open-mpi.org/trac/ompi/ticket/1121
2008-05-30 19:03:58 +00:00
George Bosilca
e361bcb64c Send optimizations.
1. The send path get shorter. The BTL is allowed to return > 0 to specify that the
   descriptor was pushed to the networks, and that the memory attached to it is 
   available again for the upper layer. The MCA_BTL_DES_SEND_ALWAYS_CALLBACK flag
   can be used by the PML to force the BTL to always trigger the callback.
   Unmodified BTL will continue to work as expected, as they will return OMPI_SUCCESS
   which force the PML to have exactly the same behavior as before. Some BTLs have
   been modified: self, sm, tcp, mx.
2. Add send immediate interface to BTL.
   The idea is to have a mechanism of allowing the BTL to take advantage of
   send optimizations such as the ability to deliver data "inline". Some
   network APIs such as Portals allow data to be sent using a "thin" event
   without packing data into a memory descriptor. This interface change
   allows the BTL to use such capabilities and allows for other optimizations
   in the future. All existing BTLs except for Portals and sm have this interface
   set to NULL.

This commit was SVN r18551.
2008-05-30 03:58:39 +00:00
Jeff Squyres
728ee47be4 Just check for the presents of $sysfsdir/class/infiniband and check
that it's a directory.  That's good enough to know that the
OpenFabrics kernel drivers have been loaded.  If you have no RDMA
devices and don't want to see the OMPI warning about not finding any
devices, then don't start the OpenFabrics kernel drivers.

This commit was SVN r18540.
2008-05-29 14:19:51 +00:00
Nysal Jan
25ac3629e9 eHCA does not have SRQ. Adding receive_queues value so that it works out of the box
This commit was SVN r18537.
2008-05-29 13:55:39 +00:00
Jeff Squyres
d5bf8fe005 Remove unused variables.
This commit was SVN r18532.
2008-05-29 11:58:16 +00:00
Jeff Squyres
e5ea9d08ca Fixes trac:1305: check to see if $sysfsdir/class/infiniband exists and is
non-empty.  If not, then exit the openib btl silently.  This addresses
the case where libibverbs is installed (which is getting more common)
and therefore the openib BTL was built/installed, but the kernel
drivers are not loaded (assumedly because there is no RDMA hardware
present).  In this case, "mpirun a.out" will not issue a warning.

There appears to be no good way to definitely tell if there are no
RDMA hardware devices present.  For example, if libibverbs/the openib
BTL is installed, there are no RDMA devices present, but the RDMA
hardware kernel drivers ''are'' loaded, OMPI will warn that it was
unable to find suitable devices.  This warning is easily eliminated by
unloading the kernel drivers.

This commit was SVN r18530.

The following Trac tickets were found above:
  Ticket 1305 --> https://svn.open-mpi.org/trac/ompi/ticket/1305
2008-05-28 22:05:47 +00:00
Pavel Shamis
28c763f751 Fixing the error flow when somebody tries to use XRC without XOOB.
This commit was SVN r18527.
2008-05-28 15:56:04 +00:00
Pavel Shamis
2c81b0ab9a Fixing compilation warning in btl_openib_connect_ibcm.c
This commit was SVN r18526.
2008-05-28 15:20:48 +00:00
Pavel Shamis
879a9fe45c setup_qps() may exit with error.
This commit was SVN r18523.
2008-05-28 11:36:38 +00:00
Pavel Shamis
e657a03143 Fixing broken XRC initialization flow.
This commit was SVN r18522.
2008-05-28 11:31:38 +00:00
Pavel Shamis
6596d19c90 Adding new ConnectX vendor_part_id. Fix for ticket #1310.
This commit was SVN r18495.
2008-05-26 12:25:49 +00:00
Jeff Squyres
e1f118d0e6 Remove unused variable
This commit was SVN r18491.
2008-05-24 13:05:04 +00:00
Jeff Squyres
1b50e5f6a5 Use the right variable in the output
This commit was SVN r18487.
2008-05-23 13:11:12 +00:00
Jeff Squyres
8faeeab81a Style cleanup only: s/struct foo/foo_t/g to conform to rest of code
base

This commit was SVN r18483.
2008-05-22 19:26:00 +00:00
Jeff Squyres
1f7f0e1f96 Fixes trac:1281
* s/port/tcp_port/g where relevant to disambiguate TCP port from
   device port
 * Rework ipaddrcheck to make it work in the LMC>0 case

This commit was SVN r18482.

The following Trac tickets were found above:
  Ticket 1281 --> https://svn.open-mpi.org/trac/ompi/ticket/1281
2008-05-22 19:18:15 +00:00
Jon Mason
d0e26b1cf6 Add pretty comments for *_iwarp.*
This commit was SVN r18478.
2008-05-22 18:02:20 +00:00
Jeff Squyres
62ac6533e0 * Add proper copyrights
* Ensure _iwarp.h is always included, or you'll get warnings on
   platforms that don't have the RDMACM
 * Add skeleton for function descriptions in comments in iwarp.h

This commit was SVN r18477.
2008-05-22 17:41:43 +00:00
Jeff Squyres
28b56c389a Only check if the opal_ifindex is >= 0 (opal_ifbegin() and
opal_ifnext() return -1 upon completion); don't check it against
opal_ifcount() -- the interface indexes aren't necessarily related to
how many interfaces were found.

This commit was SVN r18476.
2008-05-22 02:10:23 +00:00
Jeff Squyres
27978b29f8 Fixes trac:1302: ensure to also use the LID for identifing an incoming
IBCM request (not just the port number).

This commit was SVN r18475.

The following Trac tickets were found above:
  Ticket 1302 --> https://svn.open-mpi.org/trac/ompi/ticket/1302
2008-05-22 01:28:34 +00:00
George Bosilca
df2156568d The Elan BTL is now thread safe, and can be build in all conditions.
This commit was SVN r18471.
2008-05-21 20:44:37 +00:00
Pak Lui
1585789e8b Fix the undeclared variable.
This commit was SVN r18470.
2008-05-21 04:09:54 +00:00
Jon Mason
b9c25efbd2 Modify to comply with the "prefix rule" and remove "static inline" for
the non-rdmacm enabled case.   This should fix Ticket #1294.

This commit was SVN r18468.
2008-05-20 23:28:59 +00:00
Jeff Squyres
64f61ebd07 Fixes trac:1285. Really.
This commit has the same commit message as r18450, but without the
extra bonus memory corruption that was introduced.

This commit was SVN r18467.

The following SVN revision numbers were found above:
  r18450 --> open-mpi/ompi@5295902ebe

The following Trac tickets were found above:
  Ticket 1285 --> https://svn.open-mpi.org/trac/ompi/ticket/1285
2008-05-20 21:53:42 +00:00
Jeff Squyres
01a7f7eeb6 Switch orte_output* -> OPAL_OUTPUT* for two reasons:
1. We can't use orte_output in the CPC service thread because orte is
    not thread safe
 1. Use the macro version sso that they're compiled out of production
    builds 

This commit was SVN r18455.
2008-05-19 17:42:51 +00:00
Jeff Squyres
76fc8dd188 Revert r18450 -- there is some memory badness in there somewhere...
This commit was SVN r18451.

The following SVN revision numbers were found above:
  r18450 --> open-mpi/ompi@5295902ebe
2008-05-18 19:11:45 +00:00
Jeff Squyres
5295902ebe Fixes trac:1285:
* allow receive_queues to be specified in the INI file 
 * detect when multiple different receive_queues are specified and 
   gracefully abort 

However, accomplishing these goals ran into multiple difficulties. By 
putting receive_queues in the INI file: 

 1. we may not find the value until we've already traversed multiple HCAs 
 1. we may find multiple different receive_queues values

But since the openib btl initializes as it discovers each HCA/port/LID
(including the BSRQ data), if we find a new receive_queues value late
in the discovery process, then all the BSRQ data that was previously
initialized will likely be invalid. So I had to pull all the BSRQ
initialization out until after the rest of the discovery /
initialization process.

Additionally, note that if the user specifies the MCA parameter
btl_openib_receive_queues, it trumps whatever was in the INI file. So
in this case, there can never be a receive_queues conflict.  This
commit does the following (Jon wrote part of this, too):

 * adapt _ini.c to accept the "receive_queues" field in the file 
 * move 90% of _setup_qps() from _ini.c to _component.c 
 * move what was left of _setup_qps() into the main 
   _register_mca_params() function 
 * adapt init_one_hca() to detect conflicting receive_queues values 
   from the INI file 
 * after the _component.c loop calling init_one_hca(): 
   * call setup_qps() to parse the final receive_queues string value 
   * traverse all resulting btls and initialize their HCAs (if they
     weren't already): setup some lists and call prepare_hca_for_use()

I tested this code on a dual-HCA system where I artificially put in 
differing receive_queues values in the INI file for the two different 
types of HCAs that I have and it all seemed to work.

This commit was SVN r18450.

The following Trac tickets were found above:
  Ticket 1285 --> https://svn.open-mpi.org/trac/ompi/ticket/1285
2008-05-18 18:50:56 +00:00
Jeff Squyres
caacaadb0a Minor shuffling of code: no need to query the GID in the iWARP case.
This commit was SVN r18446.
2008-05-16 03:36:48 +00:00
Jeff Squyres
9f1b5237fe Ensure to return an error rather than continue
This commit was SVN r18445.
2008-05-16 03:36:11 +00:00
Jeff Squyres
6546898f09 Minor style cleanups; nothing very important in this commit.
This commit was SVN r18444.
2008-05-16 03:28:20 +00:00
Jeff Squyres
5c91f53848 Fix a minor memory leak
This commit was SVN r18443.
2008-05-16 03:27:42 +00:00
Jeff Squyres
671f0c379d Remove a whole pile of orte/util/show_help.h's that I missed. :-(
This commit was SVN r18437.
2008-05-14 11:32:33 +00:00
Jeff Squyres
e7ecd56bd2 This commit represents a bunch of work on a Mercurial side branch. As
such, the commit message back to the master SVN repository is fairly
long.

= ORTE Job-Level Output Messages =

Add two new interfaces that should be used for all new code throughout
the ORTE and OMPI layers (we already make the search-and-replace on
the existing ORTE / OMPI layers):

 * orte_output(): (and corresponding friends ORTE_OUTPUT,
   orte_output_verbose, etc.)  This function sends the output directly
   to the HNP for processing as part of a job-specific output
   channel.  It supports all the same outputs as opal_output()
   (syslog, file, stdout, stderr), but for stdout/stderr, the output
   is sent to the HNP for processing and output.  More on this below.
 * orte_show_help(): This function is a drop-in-replacement for
   opal_show_help(), with two differences in functionality:
   1. the rendered text help message output is sent to the HNP for
      display (rather than outputting directly into the process' stderr
      stream)
   1. the HNP detects duplicate help messages and does not display them
      (so that you don't see the same error message N times, once from
      each of your N MPI processes); instead, it counts "new" instances
      of the help message and displays a message every ~5 seconds when
      there are new ones ("I got X new copies of the help message...")

opal_show_help and opal_output still exist, but they only output in
the current process.  The intent for the new orte_* functions is that
they can apply job-level intelligence to the output.  As such, we
recommend that all new ORTE and OMPI code use the new orte_*
functions, not thei opal_* functions.

=== New code ===

For ORTE and OMPI programmers, here's what you need to do differently
in new code:

 * Do not include opal/util/show_help.h or opal/util/output.h.
   Instead, include orte/util/output.h (this one header file has
   declarations for both the orte_output() series of functions and
   orte_show_help()).
 * Effectively s/opal_output/orte_output/gi throughout your code.
   Note that orte_output_open() takes a slightly different argument
   list (as a way to pass data to the filtering stream -- see below),
   so you if explicitly call opal_output_open(), you'll need to
   slightly adapt to the new signature of orte_output_open().
 * Literally s/opal_show_help/orte_show_help/.  The function signature
   is identical.

=== Notes ===

 * orte_output'ing to stream 0 will do similar to what
   opal_output'ing did, so leaving a hard-coded "0" as the first
   argument is safe.
 * For systems that do not use ORTE's RML or the HNP, the effect of
   orte_output_* and orte_show_help will be identical to their opal
   counterparts (the additional information passed to
   orte_output_open() will be lost!).  Indeed, the orte_* functions
   simply become trivial wrappers to their opal_* counterparts.  Note
   that we have not tested this; the code is simple but it is quite
   possible that we mucked something up.

= Filter Framework =

Messages sent view the new orte_* functions described above and
messages output via the IOF on the HNP will now optionally be passed
through a new "filter" framework before being output to
stdout/stderr.  The "filter" OPAL MCA framework is intended to allow
preprocessing to messages before they are sent to their final
destinations.  The first component that was written in the filter
framework was to create an XML stream, segregating all the messages
into different XML tags, etc.  This will allow 3rd party tools to read
the stdout/stderr from the HNP and be able to know exactly what each
text message is (e.g., a help message, another OMPI infrastructure
message, stdout from the user process, stderr from the user process,
etc.).

Filtering is not active by default.  Filter components must be
specifically requested, such as:

{{{
$ mpirun --mca filter xml ...
}}}

There can only be one filter component active.

= New MCA Parameters =

The new functionality described above introduces two new MCA
parameters:

 * '''orte_base_help_aggregate''': Defaults to 1 (true), meaning that
   help messages will be aggregated, as described above.  If set to 0,
   all help messages will be displayed, even if they are duplicates
   (i.e., the original behavior).
 * '''orte_base_show_output_recursions''': An MCA parameter to help
   debug one of the known issues, described below.  It is likely that
   this MCA parameter will disappear before v1.3 final.

= Known Issues =

 * The XML filter component is not complete.  The current output from
   this component is preliminary and not real XML.  A bit more work
   needs to be done to configure.m4 search for an appropriate XML
   library/link it in/use it at run time.
 * There are possible recursion loops in the orte_output() and
   orte_show_help() functions -- e.g., if RML send calls orte_output()
   or orte_show_help().  We have some ideas how to fix these, but
   figured that it was ok to commit before feature freeze with known
   issues.  The code currently contains sub-optimal workarounds so
   that this will not be a problem, but it would be good to actually
   solve the problem rather than have hackish workarounds before v1.3 final.

This commit was SVN r18434.
2008-05-13 20:00:55 +00:00
Jon Mason
125eb5a2ed Convert from the Linux ifaddrs to the OMPI ifaddrs, which should unbreak Solaris.
This commit was SVN r18433.
2008-05-13 18:34:22 +00:00
Jeff Squyres
d8e5608053 Remove all retransmission code; the IBCM kernel module handles all of
that for us.

This commit was SVN r18432.
2008-05-13 16:10:34 +00:00
Jon Mason
74bf1ae25f Fix compiler warnings
This commit was SVN r18431.
2008-05-13 16:01:58 +00:00
Jon Mason
4ead9442b5 Add in IDs for all Chelsio iWARP capable adapters
This commit was SVN r18428.
2008-05-12 21:59:03 +00:00
Jeff Squyres
6b26895ad4 A little style update -- constants on the left...
This commit was SVN r18426.
2008-05-12 12:05:16 +00:00
Jeff Squyres
16cde0e5fa Fix compile error on older OFED systems
This commit was SVN r18425.
2008-05-12 11:56:14 +00:00
Gleb Natapov
6844ff32ba Return OMPI_ERR_RESOURCE_BUSY from sm->btl_send() function if there is no place in cb. This will prevent OB1 from doing early completion of small sends.
This commit was SVN r18424.
2008-05-12 07:15:29 +00:00
Gleb Natapov
0827e537fa Don't include rdma/rdma_cma.h if !OMPI_HAVE_RDMACM.
This commit was SVN r18422.
2008-05-11 11:58:02 +00:00
Jon Mason
99ab66e131 RDMACM code cleanup
This patch adds some much needed comments, reduces the amount of code
wrapping, and rearrges and removes redundant code.

This commit was SVN r18417.
2008-05-08 21:20:12 +00:00
Jon Mason
88e5f2a339 Abstract iWARP subnet ID functions (sans build break)
The iWARP subnet ID determination should not be in the RDMACM cpc, as
it was in the preversion, as this violates the cpc abstract that is
present throughout the code.  Also, this patch uses the opal_list_t
data struct instead of using its own linked lists.

This attempt includes *iwarp.c and *iwarp.h

This commit was SVN r18414.
2008-05-08 14:38:14 +00:00
Jeff Squyres
60f39a30f6 Revert r18409; that commit broke the build because it forgot to add
the btl_openib_iwarp.c and btl_openib_iwarp.h files.

This commit was SVN r18410.

The following SVN revision numbers were found above:
  r18409 --> open-mpi/ompi@056bbb68c8
2008-05-08 00:22:21 +00:00
Jon Mason
056bbb68c8 Abstract iWARP subnet ID functions
The iWARP subnet ID determination should not be in the RDMACM cpc, as
it was in the preversion, as this violates the cpc abstract that is
present throughout the code.  Also, this patch uses the opal_list_t
data struct instead of using its own linked lists.

This commit was SVN r18409.
2008-05-07 23:59:43 +00:00
Ralph Castain
7c7b9b0486 Do a little cleanup on the opal graph class and opal carto framework to conform to OMPI naming conventions and avoid potential conflict with user applications - no change in functionality, passes carto test program
This commit was SVN r18407.
2008-05-07 19:33:49 +00:00
Jeff Squyres
157cea378f * A few fixes to make IP address and port number comparisons properly
* A few indenting and style fixes

This commit was SVN r18405.
2008-05-07 16:56:07 +00:00
Jeff Squyres
bfae8ea828 The comment wasn't long enough; I felt the need to make it longer (and
explain a little more ;-) ).

This commit was SVN r18404.
2008-05-07 16:53:05 +00:00
Jeff Squyres
63abb3eb9b Clarify a comment / fix typos.
This commit was SVN r18402.
2008-05-07 14:51:36 +00:00
Jon Mason
502d164908 Create subnet ID's for iWARP.
This enables subnet differientation for iWARP devices, and rearrange
initilization so that the services are available when they are needed.

This commit was SVN r18393.
2008-05-06 22:43:52 +00:00
Jon Mason
9c724128f8 Handle no IP Address in rdmacm more resiliently
If there is no IP Address, have rdmacm log the correct error and let
another cpc have a go at it.  This is being done by splitting off the
IP address checking logic for the modex message creation, and having
it log the correct error in the error case.

This commit was SVN r18392.
2008-05-06 22:31:29 +00:00
Jon Mason
46bfd42c09 Fix compile warnings in rdmacm
Fix some reported compiler warnings and make the code a little prettier.

This commit was SVN r18391.
2008-05-06 22:19:28 +00:00
Jon Mason
9066168cd1 Prevent iWARP qp flush errors.
For iWARP, the TCP connection is tied to the QP once the QP is in RTS.  
And destroying the QP is thus tied to connection teardown for iWARP.  
This is a key distinction from IB, I think.   Anyway, to destroy the 
connection in iWARP you must move the QP out of RTS, either into CLOSING 
for a nice graceful close, or to ERROR if you want to be rude.  In both 
cases, all pending non-completed SQ and RQ WRs must be flushed.

This patch ignores all flush errors reaped by the cq and removes an
earlier attempt to work around this in the rdmacm cpc.

This commit was SVN r18388.
2008-05-06 21:57:40 +00:00
Jeff Squyres
a06d4023b8 Oops -- missed one sys_errlist -> strerror().
This commit was SVN r18378.
2008-05-06 13:22:36 +00:00
Jeff Squyres
4154e587de strerror() is much better.
This commit was SVN r18376.
2008-05-05 21:06:07 +00:00
Jon Mason
a3bf503e01 Remove error on rdma cm
If there are multiple QP's, RDMACM will not send a message if the
qpnum != 0.  In doing so, it will log an error unecessarily.  This
removes that.

This commit was SVN r18363.
2008-05-02 20:12:01 +00:00
Jon Mason
3989981578 Enable support of num_proc > num_nodes
Add the logic to support using port numbers, instead of simply using
the IP address of the sending node to determine which endpoint to
connect.  Since each process calls the cpc query function, it will
generate its own port to listen on thus enablign this to work.

This commit was SVN r18362.
2008-05-02 16:20:28 +00:00
Jeff Squyres
ba5615a18f Merge in /tmp-public/cpc3 branch to trunk. oob/xoob still remains the
default CPC.

This commit was SVN r18356.
2008-05-02 11:52:33 +00:00
Donald Kerr
843a35094f adding local work queue accounting
This commit was SVN r18352.
2008-05-01 21:01:51 +00:00