charm.git
16 years agofix of the MultiRingMulticast strategy
Filippo Gioachin [Tue, 19 Jul 2005 01:02:24 +0000 (01:02 +0000)]
fix of the MultiRingMulticast strategy

16 years agoFixed a bug handling ghosts in FEM_Node::fillNodeAdjacency
Isaac Dooley [Mon, 18 Jul 2005 21:57:09 +0000 (21:57 +0000)]
Fixed a bug handling ghosts in FEM_Node::fillNodeAdjacency

16 years agotouch up section reduction callback during initialize
Eric Bohm [Mon, 18 Jul 2005 21:53:49 +0000 (21:53 +0000)]
touch up section reduction callback during initialize

16 years agowrappers for interacting with section reduction paircalc
Eric Bohm [Mon, 18 Jul 2005 21:17:03 +0000 (21:17 +0000)]
wrappers for interacting with section reduction paircalc

16 years agoAdded ghost handling to n2n and n2e functions.
Isaac Dooley [Mon, 18 Jul 2005 21:16:03 +0000 (21:16 +0000)]
Added ghost handling to n2n and n2e functions.

16 years agoSmaller fixes missed in last checkin
Chao Huang [Mon, 18 Jul 2005 19:58:26 +0000 (19:58 +0000)]
Smaller fixes missed in last checkin

16 years agoModified the communication advance scheme to be more aggressive.
Chao Huang [Mon, 18 Jul 2005 19:57:23 +0000 (19:57 +0000)]
Modified the communication advance scheme to be more aggressive.

16 years agoModified some comments & whitespace.
Isaac Dooley [Mon, 18 Jul 2005 19:37:04 +0000 (19:37 +0000)]
Modified some comments & whitespace.

16 years agoRemoved duplicate entries in n2e creation. fillElemAdjacencyTable() now checks if...
Isaac Dooley [Mon, 18 Jul 2005 19:21:43 +0000 (19:21 +0000)]
Removed duplicate entries in n2e creation. fillElemAdjacencyTable() now checks if an entry is in the table before adding it.

16 years agoFixing some problems with initial n2e and n2n generation.
Isaac Dooley [Mon, 18 Jul 2005 18:59:24 +0000 (18:59 +0000)]
Fixing some problems with initial n2e and n2n generation.

16 years agoNow we add ghost elements to n2e tables in FEM_Node::fillElemAdjacencyTable().
Isaac Dooley [Mon, 18 Jul 2005 18:33:10 +0000 (18:33 +0000)]
Now we add ghost elements to n2e tables in FEM_Node::fillElemAdjacencyTable().

16 years agoRemoved spurious CmiMemoryChecks and got rid of the insane vi tabbing.
Terry L. Wilmarth [Mon, 18 Jul 2005 17:07:58 +0000 (17:07 +0000)]
Removed spurious CmiMemoryChecks and got rid of the insane vi tabbing.

16 years agofixed compilation error
Filippo Gioachin [Mon, 18 Jul 2005 17:07:20 +0000 (17:07 +0000)]
fixed compilation error

16 years agoAdded doxygen documentation.
Filippo Gioachin [Mon, 18 Jul 2005 05:53:55 +0000 (05:53 +0000)]
Added doxygen documentation.
Fixed bug for multicast when LastKnown is not coherent among all processors.
Now the information of the source processor is multicasted together with the
message, and is used by everyone.
MultiRing is not updated with this check-in, and thus broken. Nevertheless due
to the bug it was not working even before...

16 years agoNow removing elements and nodes doesn't crash, and they may actually be correct too.
Isaac Dooley [Mon, 18 Jul 2005 01:11:49 +0000 (01:11 +0000)]
Now removing elements and nodes doesn't crash, and they may actually be correct too.

16 years agoFEM_remove_node now works for shared nodes as well!! :)
Nilesh Choudhury [Sat, 16 Jul 2005 23:30:13 +0000 (23:30 +0000)]
FEM_remove_node now works for shared nodes as well!! :)

16 years agogspacesum is now correct
Eric Bohm [Sat, 16 Jul 2005 23:03:03 +0000 (23:03 +0000)]
gspacesum is now correct

16 years agosome bug fixes for lock unlock functionalities
Nilesh Choudhury [Sat, 16 Jul 2005 21:04:03 +0000 (21:04 +0000)]
some bug fixes for lock unlock functionalities

16 years agosome bug fixes for lock, unlock functionalities
Nilesh Choudhury [Sat, 16 Jul 2005 21:03:45 +0000 (21:03 +0000)]
some bug fixes for lock, unlock functionalities

16 years agoPair Calculator is now completely correct for S!=Grainsize
Eric Bohm [Sat, 16 Jul 2005 20:23:24 +0000 (20:23 +0000)]
Pair Calculator is now completely correct for S!=Grainsize

16 years agoAdded functions to set the coordinates for a node. Will be helpful in certain mesh...
Isaac Dooley [Sat, 16 Jul 2005 19:14:24 +0000 (19:14 +0000)]
Added functions to set the coordinates for a node. Will be helpful in certain mesh modification functions that wish to modify the locations of nodes without caring
about the underlying framework's data structures.

FEM_set_entity_coord2(int mesh, int entityType, int entityIdx, double x, double y);
FEM_set_entity_coord3(int mesh, int entityType, int entityIdx, double x, double y, double z);

16 years agosome bug fixes related to add_shared_node.
Nilesh Choudhury [Sat, 16 Jul 2005 18:13:52 +0000 (18:13 +0000)]
some bug fixes related to add_shared_node.
It should work now.

16 years agoFixing a problem if FEM_IS_VALID fields are accessed before being allocated. Now...
Isaac Dooley [Sat, 16 Jul 2005 18:02:31 +0000 (18:02 +0000)]
Fixing a problem if FEM_IS_VALID fields are accessed before being allocated. Now it won't fail during the subsequent allocate commands, it will just gracefully do
nothing.

16 years agoCreated e2n_removeAll().
Isaac Dooley [Sat, 16 Jul 2005 17:25:26 +0000 (17:25 +0000)]
Created e2n_removeAll().
Added some default parameters to calls.
Fixed a bug with the use of the tuple table in fem_mesh_modify.C

16 years agoAdding some convenient FEM_IS_VALID data accessors and modifier functions:
Isaac Dooley [Fri, 15 Jul 2005 17:06:52 +0000 (17:06 +0000)]
Adding some convenient FEM_IS_VALID data accessors and modifier functions:

  FEM_set_entity_valid(int mesh, int entityType, int entityIdx);
  FEM_set_entity_invalid(int mesh, int entityType, int entityIdx);
  FEM_is_valid(int mesh, int entityType, int entityIdx);
  FEM_count_valid(int mesh, int entityType);

The internal corresponding functions also exist in the FEM_Entity class:

   void allocateValid();
void set_valid(int idx);
void set_invalid(int idx);
int is_valid(int idx);
unsigned int count_valid();

16 years agoadded the helper functions from the .h to the .C, since it was causing multiple
Nilesh Choudhury [Fri, 15 Jul 2005 04:12:48 +0000 (04:12 +0000)]
added the helper functions from the .h to the .C, since it was causing multiple
definitions of the helper functions

16 years agoFixed the Fortran interface to FEM_REFINE2D_Coarsen to include the edges in FEM_SPARSE.
Terry L. Wilmarth [Fri, 15 Jul 2005 02:33:24 +0000 (02:33 +0000)]
Fixed the Fortran interface to FEM_REFINE2D_Coarsen to include the edges in FEM_SPARSE.
Task: TMRC2D

16 years agoChange FEM_VALID to FEM_IS_VALID
Isaac Dooley [Fri, 15 Jul 2005 01:36:19 +0000 (01:36 +0000)]
Change FEM_VALID to FEM_IS_VALID

16 years agoAdding an FEM_VALID attribute to entities. Is used internally in the mesh_modify...
Isaac Dooley [Fri, 15 Jul 2005 01:21:51 +0000 (01:21 +0000)]
Adding an FEM_VALID attribute to entities. Is used internally in the mesh_modify functions.

16 years agoThe last change didn't change all occurences of a dependency.
Isaac Dooley [Fri, 15 Jul 2005 01:16:15 +0000 (01:16 +0000)]
The last change didn't change all occurences of a dependency.

16 years agogot rid of mainmodule
Nilesh Choudhury [Thu, 14 Jul 2005 22:22:34 +0000 (22:22 +0000)]
got rid of mainmodule

16 years agofixed fem_ref_init and made FEMMeshModify mainmodule as it was not getting registered.
Nilesh Choudhury [Thu, 14 Jul 2005 21:46:47 +0000 (21:46 +0000)]
fixed fem_ref_init and made FEMMeshModify mainmodule as it was not getting registered.
Will change this later, if needed.
CV: ----------------------------------------------------------------------

16 years agochanged class definition order for doxygen
Filippo Gioachin [Thu, 14 Jul 2005 20:56:47 +0000 (20:56 +0000)]
changed class definition order for doxygen

16 years agoget rid of the hardcoded BGL_INSTALL
Gengbin Zheng [Thu, 14 Jul 2005 19:56:48 +0000 (19:56 +0000)]
get rid of the hardcoded BGL_INSTALL

16 years agoallocate one extra in pelist of ComlibSectionInfo::getRemotePelist() to avoid memory...
Gengbin Zheng [Thu, 14 Jul 2005 19:43:31 +0000 (19:43 +0000)]
allocate one extra in pelist of ComlibSectionInfo::getRemotePelist() to avoid memory corruption.

16 years agoexposed add, remove node/element, lock, unlock functions in fem.h
Nilesh Choudhury [Thu, 14 Jul 2005 17:41:16 +0000 (17:41 +0000)]
exposed add, remove node/element, lock, unlock functions in fem.h

CV:S ----------------------------------------------------------------------

16 years agoModified the various GVT phases to hopefully "expedite" the process. Initial startPh...
Terry L. Wilmarth [Thu, 14 Jul 2005 16:40:39 +0000 (16:40 +0000)]
Modified the various GVT phases to hopefully "expedite" the process.  Initial startPhases are still slow, but when the first one is triggered, it should set off expedited startPhases on all the other processors.
Task: POSE

16 years agoFixing a problem that showed up on autobuild, as well as modifying some other things.
Isaac Dooley [Thu, 14 Jul 2005 16:08:06 +0000 (16:08 +0000)]
Fixing a problem that showed up on autobuild, as well as modifying some other things.

16 years agoAdding a test program outline for testing adjacencies and mesh modification operation...
Terry L. Wilmarth [Thu, 14 Jul 2005 16:05:05 +0000 (16:05 +0000)]
Adding a test program outline for testing adjacencies and mesh modification operations.  Dima will write up the testing section.
Task: FEM

16 years agodded functions to add an entry to the IDXL lists across all shared nodes
Nilesh Choudhury [Thu, 14 Jul 2005 08:34:49 +0000 (08:34 +0000)]
dded functions to add an entry to the IDXL lists across all shared nodes
n all chunks

16 years agoCorrected a return value.
Isaac Dooley [Thu, 14 Jul 2005 02:55:56 +0000 (02:55 +0000)]
Corrected a return value.

16 years agoAdded FEM_VALID attribute for elements and nodes. Added it's corresponding create...
Isaac Dooley [Thu, 14 Jul 2005 02:54:39 +0000 (02:54 +0000)]
Added FEM_VALID attribute for elements and nodes. Added it's corresponding create functions. Also added its use to the
fem_mesh_modify.C functions

16 years agoTested and fixed isnan function.
Terry L. Wilmarth [Thu, 14 Jul 2005 02:38:08 +0000 (02:38 +0000)]
Tested and fixed isnan function.

16 years agoisnan seems not to exist on turing, replaced with my own. Not sure it works. Weird...
Terry L. Wilmarth [Thu, 14 Jul 2005 02:18:17 +0000 (02:18 +0000)]
isnan seems not to exist on turing, replaced with my own.  Not sure it works.  Weird things were happening to floating point ops near zero when building with -O. Thus the isnan usage.
Task: TMRC2D

16 years agoModified some comments.
Isaac Dooley [Thu, 14 Jul 2005 01:15:58 +0000 (01:15 +0000)]
Modified some comments.

16 years agofix hanging bug in symmetric backward path
Eric Bohm [Thu, 14 Jul 2005 01:04:00 +0000 (01:04 +0000)]
fix hanging bug in symmetric backward path

16 years agoFixing some e2e things regarding ghosts, and adding functionality to fem_mesh_modify.C
Isaac Dooley [Thu, 14 Jul 2005 00:59:42 +0000 (00:59 +0000)]
Fixing some e2e things regarding ghosts, and adding functionality to fem_mesh_modify.C

16 years agoincluded tmp/*.decl.h and tmp/*.def.h in the configuration.
Filippo Gioachin [Wed, 13 Jul 2005 23:21:45 +0000 (23:21 +0000)]
included tmp/*.decl.h and tmp/*.def.h in the configuration.
in order to work, the tmp directory must be moved so that it is not anymore a link:
i.e, if the build is net-linux-amd64 do from charm

rm tmp #remove the link
mv net-linux-amd64/tmp tmp #tmp becomes a full directory instead of a link

then run "make web"

16 years agoAdded an n2n_exists function, as well as some more code to fem_mesh_modify.C
Isaac Dooley [Wed, 13 Jul 2005 22:45:41 +0000 (22:45 +0000)]
Added an n2n_exists function, as well as some more code to fem_mesh_modify.C

16 years agoFixed many bugs to get interleaved serial refinements and coarsenings working.
Terry L. Wilmarth [Wed, 13 Jul 2005 20:55:40 +0000 (20:55 +0000)]
Fixed many bugs to get interleaved serial refinements and coarsenings working.
There is one bad hack in calculateArea.  When a bad element is sufficiently flattened, it results in NaN.  I catch this and return 0.
Task: TMRC2D

16 years agoMaking all internal functions use FEM_Mesh* instead of ints.
Isaac Dooley [Wed, 13 Jul 2005 19:23:47 +0000 (19:23 +0000)]
Making all internal functions use FEM_Mesh* instead of ints.

16 years agoforgot conv-com
Filippo Gioachin [Wed, 13 Jul 2005 16:46:51 +0000 (16:46 +0000)]
forgot conv-com

16 years agoMany Changes
Isaac Dooley [Wed, 13 Jul 2005 16:44:39 +0000 (16:44 +0000)]
Many Changes

16 years agoremove excess memory allocated in misbegotten debug attempt
Eric Bohm [Wed, 13 Jul 2005 16:24:37 +0000 (16:24 +0000)]
remove excess memory allocated in misbegotten debug attempt

16 years agoincluding commlib in doxygen
Filippo Gioachin [Wed, 13 Jul 2005 16:08:16 +0000 (16:08 +0000)]
including commlib in doxygen
exposing private methods

16 years agodded code for FMmesh
Nilesh Choudhury [Wed, 13 Jul 2005 16:04:30 +0000 (16:04 +0000)]
dded code for FMmesh

16 years agoFixed the e2n and e2e functions when they take in ghost elements as parameters. Now...
Isaac Dooley [Wed, 13 Jul 2005 02:19:35 +0000 (02:19 +0000)]
Fixed the e2n and e2e functions when they take in ghost elements as parameters. Now they correctly handle the negative ghost id's, converting
them to the needed positive indices.

16 years agoPaircalculator now produces correct results for all grainsizes
Eric Bohm [Wed, 13 Jul 2005 01:35:24 +0000 (01:35 +0000)]
Paircalculator now produces correct results for all grainsizes

16 years ago new class fem_mutil has been added
Nilesh Choudhury [Tue, 12 Jul 2005 22:41:29 +0000 (22:41 +0000)]
 new class fem_mutil has been added

16 years agoModified comments
Isaac Dooley [Tue, 12 Jul 2005 22:36:39 +0000 (22:36 +0000)]
Modified comments

16 years ago*** empty log message ***
Nilesh Choudhury [Tue, 12 Jul 2005 16:58:11 +0000 (16:58 +0000)]
*** empty log message ***

16 years agoattached a bound array to the fem object and added a set of locking and unlocking
Nilesh Choudhury [Tue, 12 Jul 2005 16:57:58 +0000 (16:57 +0000)]
attached a bound array to the fem object and added a set of locking and unlocking
functions to lock chunks.

16 years agoSlightly cleaned up code, some bugs fixed, test program enhanced.
Terry L. Wilmarth [Tue, 12 Jul 2005 16:04:36 +0000 (16:04 +0000)]
Slightly cleaned up code, some bugs fixed, test program enhanced.
Task:TMRC2D

16 years agoMade e2e functions work with ghost elements. Now the functions take an optional param...
Isaac Dooley [Tue, 12 Jul 2005 02:39:20 +0000 (02:39 +0000)]
Made e2e functions work with ghost elements. Now the functions take an optional parameter for the element type corresponding to the requested element. The do
not yet return any type information, just element ids.

16 years agoAdded vertex_remove operation to fem_adapt. Added sample code for fem_adapt_new...
Terry L. Wilmarth [Mon, 11 Jul 2005 19:33:40 +0000 (19:33 +0000)]
Added vertex_remove operation to fem_adapt.  Added sample code for fem_adapt_new that uses Isaac & Nilesh's new interface from fem_mesh_modify.
Task: FEM

16 years agoMoving all adjacency functions to fem_mesh_adjacency.C
Isaac Dooley [Mon, 11 Jul 2005 15:40:51 +0000 (15:40 +0000)]
Moving all adjacency functions to fem_mesh_adjacency.C

16 years agoAdded return statements so we don't break net-sol-cc, which expects
Isaac Dooley [Sun, 10 Jul 2005 22:08:37 +0000 (22:08 +0000)]
Added return statements so we don't break net-sol-cc, which expects
return statements. These are still just stubb functions, which will be
completed later.

16 years agoAdded blocking communication advance whenever there's outstanding
Chao Huang [Sun, 10 Jul 2005 21:01:04 +0000 (21:01 +0000)]
Added blocking communication advance whenever there's outstanding
recvs.

16 years agoRemoved ReleaseSentMsgs because the send_done callback will
Chao Huang [Sun, 10 Jul 2005 19:51:53 +0000 (19:51 +0000)]
Removed ReleaseSentMsgs because the send_done callback will
take care of the cleaning up after sending a msg.

16 years agoremoved "myr" in generated job script for turing-new.
Gengbin Zheng [Sun, 10 Jul 2005 19:42:14 +0000 (19:42 +0000)]
removed "myr" in generated job script for turing-new.

16 years agoprint num nodes.
Gengbin Zheng [Sun, 10 Jul 2005 18:23:13 +0000 (18:23 +0000)]
print num nodes.

16 years agofixed a race condition in smp version between worker and comm threads.
Gengbin Zheng [Sun, 10 Jul 2005 16:44:54 +0000 (16:44 +0000)]
fixed a race condition in smp version between worker and comm threads.
Now send down broadcast msg along spoanning tree first before push the msg into scheduler queues.

16 years agorefixed the previous change.
Gengbin Zheng [Sun, 10 Jul 2005 03:15:34 +0000 (03:15 +0000)]
refixed the previous change.

16 years agoa little cleanup in printout of the previous entry.
Gengbin Zheng [Sun, 10 Jul 2005 02:51:31 +0000 (02:51 +0000)]
a little cleanup in printout of the previous entry.

16 years agoadded more useful printout in +LBPrintSummary mode for comparison of before and after LB.
Gengbin Zheng [Sun, 10 Jul 2005 02:40:01 +0000 (02:40 +0000)]
added more useful printout in +LBPrintSummary mode for comparison of before and after LB.

16 years agofixed bug in broadcast in spanning tree/hypercube - when packetization, the packet...
Gengbin Zheng [Sun, 10 Jul 2005 02:36:51 +0000 (02:36 +0000)]
fixed bug in broadcast in spanning tree/hypercube - when packetization, the packet order can be changed along the tree when packets are sent in pipelining fashion.

16 years agodefine CMK_MOD_NAME_ALLCAPS=1 since ifc generates mod name with capitals.
Gengbin Zheng [Sat, 9 Jul 2005 19:15:00 +0000 (19:15 +0000)]
define CMK_MOD_NAME_ALLCAPS=1 since ifc generates mod name with capitals.

16 years agouse ifc instead of ifort in default icc option.
Gengbin Zheng [Sat, 9 Jul 2005 19:04:53 +0000 (19:04 +0000)]
use ifc instead of ifort in default icc option.

16 years agoa minor change in detecting fortran name mangling, if f90 is not available, try f77...
Gengbin Zheng [Sat, 9 Jul 2005 19:03:56 +0000 (19:03 +0000)]
a minor change in detecting fortran name mangling, if f90 is not available, try f77 compiler.

16 years agofixed a bug in error checking of using a wrong array index.
Gengbin Zheng [Sat, 9 Jul 2005 17:16:25 +0000 (17:16 +0000)]
fixed a bug in error checking of using a wrong array index.

16 years agoAdding one more assumption to the comments.
Isaac Dooley [Sat, 9 Jul 2005 16:29:01 +0000 (16:29 +0000)]
Adding one more assumption to the comments.

16 years agoAdding the new mesh modification functionality. The files are currently empty, but...
Isaac Dooley [Sat, 9 Jul 2005 16:26:38 +0000 (16:26 +0000)]
Adding the new mesh modification functionality. The files are currently empty, but they contain the prototypes
and general steps that need to be implemented.

16 years agoadded -module comlib to dep.
Gengbin Zheng [Sat, 9 Jul 2005 04:29:05 +0000 (04:29 +0000)]
added -module comlib to dep.

16 years agofixed in debug print
Gengbin Zheng [Sat, 9 Jul 2005 04:26:57 +0000 (04:26 +0000)]
fixed in debug print

16 years agochecked in translator changes for new section proxy ckNew syntax.
Gengbin Zheng [Sat, 9 Jul 2005 04:16:15 +0000 (04:16 +0000)]
checked in translator changes for new section proxy ckNew syntax.

16 years agoudpate to new syntax
Gengbin Zheng [Sat, 9 Jul 2005 03:45:47 +0000 (03:45 +0000)]
udpate to new syntax

16 years agominor fix for the example
Gengbin Zheng [Sat, 9 Jul 2005 03:43:23 +0000 (03:43 +0000)]
minor fix for the example

16 years agotwo new syntax for creating cksection proxy (more easily):
Gengbin Zheng [Sat, 9 Jul 2005 03:42:39 +0000 (03:42 +0000)]
two new syntax for creating cksection proxy (more easily):
1. do not have to convert CkArrayIndexXD to CkArrayIndexMax when it is built-in array index type (1D - 6D);
2. create by specifying (l:u:s) directly:
CProxySection_Hello proxy = CProxySection_Hello::ckNew(AID, 0, 10, 1);

16 years agoupdated cksection proxy creatino syntax.
Gengbin Zheng [Sat, 9 Jul 2005 03:35:28 +0000 (03:35 +0000)]
updated cksection proxy creatino syntax.

16 years agomistype fixes
Filippo Gioachin [Fri, 8 Jul 2005 21:31:05 +0000 (21:31 +0000)]
mistype fixes

16 years agosayantan:
Sayantan Chakravorty [Fri, 8 Jul 2005 21:08:32 +0000 (21:08 +0000)]
sayantan:
Fixed a problem with interpolation .. It was doing A*(1-frac) +B*frac
rather than A*frac + B*(1-frac)

16 years agoAdded elan_CmiAlloc() function prototype.
David Kunzman [Fri, 8 Jul 2005 19:45:55 +0000 (19:45 +0000)]
Added elan_CmiAlloc() function prototype.

16 years agoAdded coarsening bail-out in the case of coarsening around a vertex of degree 3.Made...
Terry L. Wilmarth [Fri, 8 Jul 2005 17:16:16 +0000 (17:16 +0000)]
Added coarsening bail-out in the case of coarsening around a vertex of degree 3.Made this and other coarsening bailouts increment the nonCoarsenCount flag on the relevant element.  This causes coarsening to avoid a retry.  These still need to be reset when new coarsening starts and when elements are altered by other means.
Also added a new test program for the purpose of beating on the code.
Task: TMRC2D

16 years agoChanged element queueing from stack to heap, added tolerance for coarsening and refin...
Dmitriy Ofman [Fri, 8 Jul 2005 16:31:46 +0000 (16:31 +0000)]
Changed element queueing from stack to heap, added tolerance for coarsening and refinement.  Modified newMesh, deriveEdges and added another addEdge function tosupport edge connectivity information.

16 years agoupdated REFINE2D_NewMesh function declaration signature with edge parameters.
Dmitriy Ofman [Fri, 8 Jul 2005 16:26:51 +0000 (16:26 +0000)]
updated REFINE2D_NewMesh function declaration signature with edge parameters.

16 years agoupdated function declaration REFINE2D_NewMesh with edge parameters.
Dmitriy Ofman [Fri, 8 Jul 2005 16:26:05 +0000 (16:26 +0000)]
updated function declaration REFINE2D_NewMesh with edge parameters.

16 years agoUpdated FEM_REFINE2D_Newmesh, added support for
Dmitriy Ofman [Fri, 8 Jul 2005 16:24:41 +0000 (16:24 +0000)]
Updated FEM_REFINE2D_Newmesh, added support for
edge connectivity  (stubs only)

16 years agoMuch more abundant header comments describing paircalculator.
Eric Bohm [Fri, 8 Jul 2005 00:17:07 +0000 (00:17 +0000)]
Much more abundant header comments describing paircalculator.

Refactored creation process to have no redundant code and a more logical
breakdown of creation process.  Trimmed down the makeleft and makeright
functions and made them ready for external calling to assist in the
planned commlib multicast refactoring.

Added support for the new section reduction.
Added initGRed message, the sumDoubleMatrix reduction client,
and the initOneGRed function for the creation of the multicast tree which
will be used by that reduction.

Added proper pup behavior for sparseReducer state.

16 years agominor change to fix the hardcoded f90 libdir.
Gengbin Zheng [Thu, 7 Jul 2005 20:08:51 +0000 (20:08 +0000)]
minor change to fix the hardcoded f90 libdir.

16 years agocalling for ifort always
Gengbin Zheng [Thu, 7 Jul 2005 20:05:03 +0000 (20:05 +0000)]
calling for ifort always