charm.git
17 years agofixed path of METIS lib.
Gengbin Zheng [Tue, 24 Aug 2004 14:55:31 +0000 (14:55 +0000)]
fixed path of METIS lib.

17 years agofixed memory leak
Filippo Gioachin [Tue, 24 Aug 2004 01:40:56 +0000 (01:40 +0000)]
fixed memory leak

17 years agoChecking in lots of strategy fixes; opt & opt2 mystery bugs remain.
Terry L. Wilmarth [Tue, 24 Aug 2004 00:20:40 +0000 (00:20 +0000)]
Checking in lots of strategy fixes; opt & opt2 mystery bugs remain.

17 years agoMoved RBevent to evq.
Terry L. Wilmarth [Tue, 24 Aug 2004 00:17:11 +0000 (00:17 +0000)]
Moved RBevent to evq.

17 years agochanged the target METIS to point to libs/ck-libs/parmetis/MetisLib
Sayantan Chakravorty [Mon, 23 Aug 2004 21:42:33 +0000 (21:42 +0000)]
changed the target METIS to point to libs/ck-libs/parmetis/MetisLib

17 years agoRemoved the entry for metis from the makefile
Sayantan Chakravorty [Mon, 23 Aug 2004 20:47:33 +0000 (20:47 +0000)]
Removed the entry for metis from the makefile

17 years agorefix for the rand() for win32 build.
Gengbin Zheng [Mon, 23 Aug 2004 20:33:42 +0000 (20:33 +0000)]
refix for the rand() for win32 build.

17 years agoremoved inclusion of ../parmetis.h because it depends on mpi.h
Gengbin Zheng [Mon, 23 Aug 2004 20:15:28 +0000 (20:15 +0000)]
removed inclusion of ../parmetis.h because it depends on mpi.h

17 years agoadded new +LBSameCpus option to tell LB to assume all CPUs are of same speed (skip...
Gengbin Zheng [Mon, 23 Aug 2004 19:39:17 +0000 (19:39 +0000)]
added new +LBSameCpus option to tell LB to assume all CPUs are of same speed (skip speed measurement)

17 years agotested with new metis and added different level of debug prints.
Gengbin Zheng [Mon, 23 Aug 2004 19:22:12 +0000 (19:22 +0000)]
tested with new metis and added different level of debug prints.

17 years agoFixed make files so that the .so are generated and OPTS gets passed in
Sayantan Chakravorty [Mon, 23 Aug 2004 17:52:43 +0000 (17:52 +0000)]
Fixed make files so that the .so are generated and OPTS gets passed in

17 years agoFixed make files so that the .so are generated and OPTS gets passed in
Sayantan Chakravorty [Mon, 23 Aug 2004 17:52:35 +0000 (17:52 +0000)]
Fixed make files so that the .so are generated and OPTS gets passed in
Copied a fix from the old metis lib to use unsigned long instead of int
for the amount of memory allocated

17 years agoFixed make files so that the .so are generated and OPTS gets passed in
Sayantan Chakravorty [Mon, 23 Aug 2004 17:45:46 +0000 (17:45 +0000)]
Fixed make files so that the .so are generated and OPTS gets passed in

17 years agouse charmc instead ar directly to create .a's.
Gengbin Zheng [Mon, 23 Aug 2004 06:39:00 +0000 (06:39 +0000)]
use charmc instead ar directly to create .a's.

17 years ago+LBDebug can be now used both as a flag and taking extra debug level number, that...
Gengbin Zheng [Mon, 23 Aug 2004 06:34:34 +0000 (06:34 +0000)]
+LBDebug can be now used both as a flag and taking extra debug level number, that is +LBDebug 2 for example to control verbosity.

17 years agoassign proper LB name string in migration constructions.
Gengbin Zheng [Mon, 23 Aug 2004 06:28:37 +0000 (06:28 +0000)]
assign proper LB name string in migration constructions.

17 years agoadded -fpic to CMK_CC and CMK_CXX for compiling shared libs.
Gengbin Zheng [Sun, 22 Aug 2004 23:16:10 +0000 (23:16 +0000)]
added -fpic to CMK_CC and CMK_CXX for compiling shared libs.

17 years agoconfig for SGI Message Passing Toolkit(MPT), specifically for ncsa's new SGI Altix.
Gengbin Zheng [Sun, 22 Aug 2004 20:58:39 +0000 (20:58 +0000)]
config for SGI Message Passing Toolkit(MPT), specifically for ncsa's new SGI Altix.

17 years agoRedo change overwritten by Sayantan: don't build liveViz3d
Orion Lawlor [Sat, 21 Aug 2004 19:06:31 +0000 (19:06 +0000)]
Redo change overwritten by Sayantan: don't build liveViz3d
by default (yet).

17 years agoincluded parmetis and multiphasedsharedarrays in the dirs of the makefile
Sayantan Chakravorty [Sat, 21 Aug 2004 14:57:39 +0000 (14:57 +0000)]
included parmetis and multiphasedsharedarrays in the dirs of the makefile

17 years agoDon't link in parmetis, since it's not yet built by default.
Orion Lawlor [Sat, 21 Aug 2004 14:38:09 +0000 (14:38 +0000)]
Don't link in parmetis, since it's not yet built by default.
I guess we need preprocessor switches in .dep files somehow...

17 years agomemory leak fix
Filippo Gioachin [Sat, 21 Aug 2004 03:56:58 +0000 (03:56 +0000)]
memory leak fix

17 years agoadded capability of lowlovel interface with iteration
Filippo Gioachin [Sat, 21 Aug 2004 02:34:00 +0000 (02:34 +0000)]
added capability of lowlovel interface with iteration

17 years agoAdded the parallel partitioning code in files parallel_part.C and parallel_part,h.
Sayantan Chakravorty [Fri, 20 Aug 2004 19:31:00 +0000 (19:31 +0000)]
Added the parallel partitioning code in files parallel_part.C and parallel_part,h.
By default the code in parallel_part.C is not compiled. In the default case
fem.C doesnt call the parallel partitioning code. So the default compilation of fem
shouldnot change.

The parallel partitioning code is activated when FEM is compiled with FEM_PARALLEL_PART.
This activates the code in parallel_part.C and fem.C now calls the parallel partition routine.
In this case FEM needs msa and parmetis. The libmodulefem.dep will be changed to reflect that.
The parallel partition code is not yet in its final shape. The part with multiple ghost layers
needs to be debugged and the rest of the code needs to be optimized and might also need some
refactoring to speed it up.

17 years agoPARMetisLib directory of PARMETIS
Sayantan Chakravorty [Fri, 20 Aug 2004 18:50:41 +0000 (18:50 +0000)]
PARMetisLib directory of PARMETIS

17 years agoMETISLib directory of parmetis
Sayantan Chakravorty [Fri, 20 Aug 2004 18:49:53 +0000 (18:49 +0000)]
METISLib directory of parmetis

17 years agoParmetis manual directory
Sayantan Chakravorty [Fri, 20 Aug 2004 18:49:08 +0000 (18:49 +0000)]
Parmetis manual directory

17 years agoChecking in parmetis, needed for parallel partitioning of FEM.
Sayantan Chakravorty [Fri, 20 Aug 2004 18:47:23 +0000 (18:47 +0000)]
Checking in parmetis, needed for parallel partitioning of FEM.

17 years ago*** empty log message ***
Yan Shi [Fri, 20 Aug 2004 15:47:28 +0000 (15:47 +0000)]
*** empty log message ***

17 years agoPut a "cd $currentdir" into the script charmrun generates.
Orion Lawlor [Thu, 19 Aug 2004 22:11:49 +0000 (22:11 +0000)]
Put a "cd $currentdir" into the script charmrun generates.
    Without this, batch jobs (now?) run in $HOME by default.

17 years ago- implemented a new wider interface in which the python object is transferred directly
Filippo Gioachin [Thu, 19 Aug 2004 08:11:58 +0000 (08:11 +0000)]
- implemented a new wider interface in which the python object is transferred directly
  to the user code, thus allowing a greater flexibility.
- code cleanup.
- added the creation of a new interpreter every time a python code is executed, in this
  way it will be possible to have multiple threads running (for smp codes).
- TODO: the variable "pyNumber" is currently fixed, and must be updated for real
  multithreading

17 years agoSync'ed up hypercube with switch architecture
Praveen Kumar Jagadishprasad [Thu, 19 Aug 2004 04:44:41 +0000 (04:44 +0000)]
Sync'ed up hypercube with switch architecture

17 years agoFixed immediate lock interface. For non-SMP version it takes a dummy
Chao Huang [Thu, 19 Aug 2004 01:24:11 +0000 (01:24 +0000)]
Fixed immediate lock interface. For non-SMP version it takes a dummy
lock but really uses a default system lock. For SMP version, the
programmer is supposed to create the locks, use them the same way
as CmiNodeLock.

Changed the test program too.

17 years agoAdd path to default location of fortran support libraries.
Orion Lawlor [Wed, 18 Aug 2004 19:23:32 +0000 (19:23 +0000)]
Add path to default location of fortran support libraries.

17 years agoA rather hideous workaround for a bug in Intel ifc 8.0.34:
Orion Lawlor [Wed, 18 Aug 2004 19:22:43 +0000 (19:22 +0000)]
A rather hideous workaround for a bug in Intel ifc 8.0.34:
   "external" procedures that are defined in a module aren't
visible outside the module.

This means when CSAR's Rocflo include 'mpif.h' in ModMPI,
the external procedures like MPI_SUM don't show up.

The workaround is to make MPI_SUM an external *function*,
by adding a (fake) return type.  Since there's no real difference
in how functions and subroutines are passed around, this
shouldn't affect any other compilers.

17 years agoMany bug fixes:
Orion Lawlor [Wed, 18 Aug 2004 19:16:20 +0000 (19:16 +0000)]
Many bug fixes:
   - Upload images immediately after unpacking, for better cache
     performance and easier analysis.
   - Handle multiple clients in image collection (FIXME: generate
     images for multiple clients properly)
   - Include image collection in lv3d_qd.
   - Always pass "-msse", even when compiled with OPTS="-g"
     (this prevents a segfault caused by mixed use of SSE)

17 years agoPut in output buffering
Praveen Kumar Jagadishprasad [Wed, 18 Aug 2004 05:03:11 +0000 (05:03 +0000)]
Put in output buffering

17 years agoremoved unused variable
Filippo Gioachin [Tue, 17 Aug 2004 19:44:38 +0000 (19:44 +0000)]
removed unused variable

17 years agochanged comments to c style /* */
Filippo Gioachin [Tue, 17 Aug 2004 05:59:35 +0000 (05:59 +0000)]
changed comments to c style /* */

17 years agoAdd quiescence detection handler, to allow clients to detect
Orion Lawlor [Tue, 17 Aug 2004 05:18:19 +0000 (05:18 +0000)]
Add quiescence detection handler, to allow clients to detect
when we're still working.

17 years agoBug fix for socket select with timeout:
Orion Lawlor [Tue, 17 Aug 2004 05:17:27 +0000 (05:17 +0000)]
Bug fix for socket select with timeout:
   "sec" field may become negative, in which case we
   should just give up.  Otherwise, select will return -1!

   Also cut "n" field of select to the relevant fd, which
   should be slightly faster than the whole FD_SETSIZE.

17 years ago*** empty log message ***
Praveen Kumar Jagadishprasad [Tue, 17 Aug 2004 01:34:49 +0000 (01:34 +0000)]
*** empty log message ***

17 years agoAllow users to pass theUniverse in, which lets them use
Orion Lawlor [Mon, 16 Aug 2004 20:36:49 +0000 (20:36 +0000)]
Allow users to pass theUniverse in, which lets them use
custom universes.

17 years agoImmediate message lock. There are two calls:
Chao Huang [Mon, 16 Aug 2004 20:28:36 +0000 (20:28 +0000)]
Immediate message lock. There are two calls:
CmiLockImmdtMsg() and CmiUnlockImmdtMsg() protects critical code
where immediate message handling may cause incorrect result.

17 years agoCleaned and correct.
Yan Shi [Mon, 16 Aug 2004 20:19:44 +0000 (20:19 +0000)]
Cleaned and correct.

17 years ago*** empty log message ***
Praveen Kumar Jagadishprasad [Mon, 16 Aug 2004 06:09:22 +0000 (06:09 +0000)]
*** empty log message ***

17 years agoAdded choice between adaptive and deterministic routing for Bluegene
Praveen Kumar Jagadishprasad [Mon, 16 Aug 2004 06:07:44 +0000 (06:07 +0000)]
Added choice between adaptive and deterministic routing for Bluegene
Added choice for receiving multiple messages simulataneously (Bluegene has 6 reception fifos operating simulataneously)
Added input speedup variable which constrains the number of input vcs which can be simulataneously transfer data
put switch/channel/nic code into seperate files
Added util.* files in machine directories so that they can be extended as needed
For now Bluegene and IB are working. Simple changes to be made to HypercubeArch and Redstorm too.

17 years agoadded ifort fortran library "-lifport" for system().
Gengbin Zheng [Mon, 16 Aug 2004 05:03:25 +0000 (05:03 +0000)]
added ifort fortran library "-lifport" for system().

17 years agofat-tree works
Praveen Kumar Jagadishprasad [Sun, 15 Aug 2004 02:36:50 +0000 (02:36 +0000)]
fat-tree works

17 years agoNeed liveViz3d.ci to link to level 0 and level 1 .ci files.
Orion Lawlor [Fri, 13 Aug 2004 22:58:14 +0000 (22:58 +0000)]
Need liveViz3d.ci to link to level 0 and level 1 .ci files.

17 years agoRemoved old files.
Orion Lawlor [Fri, 13 Aug 2004 22:23:26 +0000 (22:23 +0000)]
Removed old files.

17 years agoLatest version of liveViz3d. Renamed lots of files, but
Orion Lawlor [Fri, 13 Aug 2004 22:22:32 +0000 (22:22 +0000)]
Latest version of liveViz3d.  Renamed lots of files, but
still basically the same interface.

17 years agoRemove liveViz3d from default build, until the new version
Orion Lawlor [Fri, 13 Aug 2004 22:20:52 +0000 (22:20 +0000)]
Remove liveViz3d from default build, until the new version
is fully stablized.

17 years agoReplace "pixel_t" with "channel_t". This is for the latest liveViz3d.
Orion Lawlor [Fri, 13 Aug 2004 22:19:55 +0000 (22:19 +0000)]
Replace "pixel_t" with "channel_t".  This is for the latest liveViz3d.

17 years agoSome bug fixes to opt.
Terry L. Wilmarth [Fri, 13 Aug 2004 20:47:49 +0000 (20:47 +0000)]
Some bug fixes to opt.

17 years agoRemove annoying "passing unrecognized option ... to all compilers and linkers"
Orion Lawlor [Fri, 13 Aug 2004 19:16:52 +0000 (19:16 +0000)]
Remove annoying "passing unrecognized option ... to all compilers and linkers"
warning.  This warning is quite annoying anytime you have to pass
extra arguments to the compiler, e.g.:
   - Machine or optimization flags, like -msse for g++.
   - Bounds checks to Fortran compiler (-B111), like the rocket code.

I suspect this warning annoys many more people than it educates,
so it doesn't belong there.

17 years ago*** empty log message ***
Praveen Kumar Jagadishprasad [Fri, 13 Aug 2004 14:44:42 +0000 (14:44 +0000)]
*** empty log message ***

17 years agoAdd fake return statements after abort, to satisfy (stupid)
Orion Lawlor [Fri, 13 Aug 2004 14:32:23 +0000 (14:32 +0000)]
Add fake return statements after abort, to satisfy (stupid)
Solaris CC compiler.

17 years agoSemi-final version of adapt3 strategy.
Terry L. Wilmarth [Thu, 12 Aug 2004 16:44:35 +0000 (16:44 +0000)]
Semi-final version of adapt3 strategy.

17 years ago*** empty log message ***
Praveen Kumar Jagadishprasad [Thu, 12 Aug 2004 15:01:43 +0000 (15:01 +0000)]
*** empty log message ***

17 years ago*** empty log message ***
Praveen Kumar Jagadishprasad [Wed, 11 Aug 2004 20:59:42 +0000 (20:59 +0000)]
*** empty log message ***

17 years agoTo make MPI_ANY_SOURCE work, it must match the value -1 of CmmWildCard in msgmgr.c,
Chao Huang [Wed, 11 Aug 2004 16:36:52 +0000 (16:36 +0000)]
To make MPI_ANY_SOURCE work, it must match the value -1 of CmmWildCard in msgmgr.c,
so we have to move MPI_PROC_NULL to -2.

17 years agoin MPI_Irecv we should set *request to be MPI_REQEUST_NULL
Chao Huang [Wed, 11 Aug 2004 04:25:59 +0000 (04:25 +0000)]
in MPI_Irecv we should set *request to be MPI_REQEUST_NULL
rather than setting it to NULL pointer.

17 years agoMPI_ANY_SOURCE should be different than MPI_PROC_NULL to eliminate confusion
Chao Huang [Tue, 10 Aug 2004 20:49:38 +0000 (20:49 +0000)]
MPI_ANY_SOURCE should be different than MPI_PROC_NULL to eliminate confusion

17 years agoNull process MPI_PROC_NULL in point-to-point as dest/source should be
Chao Huang [Tue, 10 Aug 2004 20:40:02 +0000 (20:40 +0000)]
Null process MPI_PROC_NULL in point-to-point as dest/source should be
treated according to the standard: the call returns with success
immediately, for receive the status is set.

This is true for non-blocking communications.

17 years agobugfix
Filippo Gioachin [Tue, 10 Aug 2004 19:52:58 +0000 (19:52 +0000)]
bugfix

17 years agoSeries of bug fixes:
Orion Lawlor [Tue, 10 Aug 2004 19:23:42 +0000 (19:23 +0000)]
Series of bug fixes:
  1.) MPI_Isend with MPI_COMM_SELF didn't work, but MPI_Send did.
   Why?  Because MPI_Send has a special case for MPI_COMM_SELF
   missing from MPI_Isend (and MPI_Scan, and possibly others).
   Fixed by making ampi::getRank()/getSize() take a communicator,
   so the special case goes in one place instead of scattered through
   the source.

  2.) Make all the arithmetic reduction operations work with all the MPI datatypes.
   Removed horrible copy-and-paste code (at least from arithmetic ops).

  3.) Added (disabled) tracing statements, used to track down the above
   problems.

17 years agoUseless changes to support inheritance.
Eric Bohm [Tue, 10 Aug 2004 18:57:24 +0000 (18:57 +0000)]
Useless changes to support inheritance.
Checking them in because they'll come in handy later on when we figure out
the deeper changes required for inheritance.

These should make no difference to any existing code.  And might make multiple inheritance from non-posers work, but that capacity is untested.

17 years agoAdded sequential endtime approximation.
Terry L. Wilmarth [Tue, 10 Aug 2004 18:43:04 +0000 (18:43 +0000)]
Added sequential endtime approximation.

17 years agoFixed a bug in srtable:Restructure. Removed debug messages.
Terry L. Wilmarth [Tue, 10 Aug 2004 18:31:00 +0000 (18:31 +0000)]
Fixed a bug in srtable:Restructure.  Removed debug messages.

17 years agoAdding out of bounds check for CkVec. This is only enabled when CMK_PARANOID is tuned on.
Sameer Kumar [Tue, 10 Aug 2004 17:47:20 +0000 (17:47 +0000)]
Adding out of bounds check for CkVec. This is only enabled when CMK_PARANOID is tuned on.

17 years agoTwo bug fixes:
Orion Lawlor [Tue, 10 Aug 2004 17:20:25 +0000 (17:20 +0000)]
Two bug fixes:
  1.) Use CMK_LD to test shared library creation.  This is what
      charmc does, and with icc on the SuSE machines, CMK_CXX
      doesn't work (missing link options) but CMK_LD does.

  2.) Remove garbage (chunk of Makefile!) at end of configure.in.
     It's not clear if Chao checked this in by accident, or if
     CVS or the filesystem somehow screwed it up.

17 years ago- Moved 4 files (ComlibStrategy.*, ComlibArrayListener.*) from ck-core to the
Filippo Gioachin [Tue, 10 Aug 2004 07:20:14 +0000 (07:20 +0000)]
- Moved 4 files (ComlibStrategy.*, ComlibArrayListener.*) from ck-core to the
better place ck-com.
- Updated the benchmark for pipeBroadcast to test more strategies simulataneously
- Fixed a bug in comlib for which the variable "myInstanceID" did not get set
correctly when a Charm strategy embedded a Converse strategy in it (the converse
strategy variable was not set)
- There is still a TODO: usually Comlib remember the last strategy used (and set
through cinst.beginIteration()), but not at the very beginning of the program,
because when the StrategyTable is unPupped, it overwrite the currStrategy index
to the last inserted strategy.

17 years agodefined the missing CMK_MOD_EXT definition.
Gengbin Zheng [Mon, 9 Aug 2004 21:40:13 +0000 (21:40 +0000)]
defined the missing CMK_MOD_EXT definition.

17 years agoFix silly IDXL registration bug.
Orion Lawlor [Mon, 9 Aug 2004 17:02:18 +0000 (17:02 +0000)]
Fix silly IDXL registration bug.
Previously, would re-register after migration
*always*, even if no previous value existed.

IDXL registration is needed so that users can keep
IDXL list indices even across migration, like:
int list=FEM_Comm_ghost(mesh,FEM_ELEM);
        FEM_Migate();
        IDXL_...(list,...);
To do this, FEM needs to re-register its lists
in the same order on the new processor.

17 years agodeleted an unused (and misleading) constructor
Filippo Gioachin [Sat, 7 Aug 2004 04:49:58 +0000 (04:49 +0000)]
deleted an unused (and misleading) constructor

17 years agofixed a wrong invocation of CmiRegisterHandler
Filippo Gioachin [Sat, 7 Aug 2004 04:48:25 +0000 (04:48 +0000)]
fixed a wrong invocation of CmiRegisterHandler

17 years agoChecking in some garbage because I need to debug elsewhere...
Terry L. Wilmarth [Fri, 6 Aug 2004 22:31:45 +0000 (22:31 +0000)]
Checking in some garbage because I need to debug elsewhere...

17 years agoSeveral optimizations to speed up srTable.
Terry L. Wilmarth [Fri, 6 Aug 2004 16:07:13 +0000 (16:07 +0000)]
Several optimizations to speed up srTable.

17 years agoCompleted adapt3 strategy improves on adapt2.
Terry L. Wilmarth [Thu, 5 Aug 2004 22:25:02 +0000 (22:25 +0000)]
Completed adapt3 strategy improves on adapt2.

17 years agoFixed make test line
Eric Bohm [Thu, 5 Aug 2004 17:10:41 +0000 (17:10 +0000)]
Fixed make test line

17 years agosnippet about -s flag
Eric Bohm [Thu, 5 Aug 2004 17:05:28 +0000 (17:05 +0000)]
snippet about -s flag

17 years agoUse -s option to etrans.pl to get sequential support.
Eric Bohm [Thu, 5 Aug 2004 17:03:09 +0000 (17:03 +0000)]
Use -s option to etrans.pl to get sequential support.
Note, you still have to define SEQUENTIAL_POSE in your pose_config

17 years agoReorganized code.
Yan Shi [Thu, 5 Aug 2004 14:40:10 +0000 (14:40 +0000)]
Reorganized code.
Fixed optimization of zgemm routine, and streaming at first phase.

17 years agoRearrange IBM SP charm versions:
Orion Lawlor [Thu, 5 Aug 2004 00:22:22 +0000 (00:22 +0000)]
Rearrange IBM SP charm versions:
   mpi-sp mpcc32 is always 32 bit
   mpi-sp mpcc64 is always 64 bit
   mpi-sp (no options) uses OBJECT_MODE to decide on the number of bits.

I think this is what most people will want: those who don't care
about the bit setting will use the default OBJECT_MODE in their dot files;
while those who do care have a sensible override.

17 years agoAdded integrated sequential simulation to POSE.
Terry L. Wilmarth [Wed, 4 Aug 2004 21:32:25 +0000 (21:32 +0000)]
Added integrated sequential simulation to POSE.

17 years agoAdapt3 updates, backwards commits.
Terry L. Wilmarth [Wed, 4 Aug 2004 17:17:12 +0000 (17:17 +0000)]
Adapt3 updates, backwards commits.

17 years agoforgot the change in translator - remove unused setImmediate().
Gengbin Zheng [Wed, 4 Aug 2004 14:37:22 +0000 (14:37 +0000)]
forgot the change in translator - remove unused setImmediate().

17 years agobugfixes for pipelined broadcast in converse
Filippo Gioachin [Tue, 3 Aug 2004 20:41:31 +0000 (20:41 +0000)]
bugfixes for pipelined broadcast in converse

17 years agovirtualizing a function
Filippo Gioachin [Tue, 3 Aug 2004 20:38:16 +0000 (20:38 +0000)]
virtualizing a function

17 years agouse special bg trace macro for mpi_barrier so that the simulator can identify the...
Gengbin Zheng [Tue, 3 Aug 2004 17:11:42 +0000 (17:11 +0000)]
use special bg trace macro for mpi_barrier so that the simulator can identify the barrier and swap it with its own machine implementation.

17 years agodefine a special trace for MPI_Barrier so that it potentially allows a simulator...
Gengbin Zheng [Tue, 3 Aug 2004 17:08:00 +0000 (17:08 +0000)]
define a special trace for MPI_Barrier so that it potentially allows a simulator to swap the implementation of barrier instead of using Charm++'s barrier implementation.

17 years agoremoved unused (obsolete) code about setting immediate bit in charm envelop.
Gengbin Zheng [Tue, 3 Aug 2004 16:43:05 +0000 (16:43 +0000)]
removed unused (obsolete) code about setting immediate bit in charm envelop.

17 years agoan optimized version to substantially reduced bigsim overhead of unnecessarily access...
Gengbin Zheng [Tue, 3 Aug 2004 16:41:10 +0000 (16:41 +0000)]
an optimized version to substantially reduced bigsim overhead of unnecessarily accessing thread private varaibles.

17 years agoupdated loadlog utility program.
Gengbin Zheng [Tue, 3 Aug 2004 16:34:59 +0000 (16:34 +0000)]
updated loadlog utility program.

17 years agoSimplified and accellerated Mgc volume computation routines,
Orion Lawlor [Tue, 3 Aug 2004 02:20:00 +0000 (02:20 +0000)]
Simplified and accellerated Mgc volume computation routines,
and added check against volume_planes.h routines.

17 years agoAdd volume computation routines for arbitrary sets of
Orion Lawlor [Tue, 3 Aug 2004 02:18:59 +0000 (02:18 +0000)]
Add volume computation routines for arbitrary sets of
halfspaces.  This is GPL code from the package "VINCI",
   http://www.lix.polytechnique.fr/Labo/Andreas.Enge/Vinci.html
although I've heavily modified it for speed (in 3D) and
portability.

To avoid running afoul of the GPL, this source file is not
built or linked in by default in Charm++.  If these methods
prove useful (and for now they're still to slow to compete
with the Mgc routines), we'll have to write a non-GPL version.

17 years agohardcoded -q32 to force 32 bit compilation.
Gengbin Zheng [Mon, 2 Aug 2004 18:18:08 +0000 (18:18 +0000)]
hardcoded -q32 to force 32 bit compilation.

17 years agoBug fix:
Orion Lawlor [Mon, 2 Aug 2004 17:10:02 +0000 (17:10 +0000)]
Bug fix:
   Really blow away charm/bin, lib, include directories, even
if they're not just softlinks.  This lets ./build work properly
even if somebody does "cp -r" on the charm source tree--without
this change, the build would succeed, but charm/bin wouldn't change,
leading to horribly confusing errors.

17 years agoSome optimization breaks running on IBM SP and BG/L - Wrong answers are given and...
Yan Shi [Sat, 31 Jul 2004 19:07:04 +0000 (19:07 +0000)]
Some optimization breaks running on IBM SP and BG/L  -  Wrong answers are given and causes NANs. So have to convert back to a non-opt version just to run.