summaryrefslogtreecommitdiff
path: root/src/comm
Commit message (Collapse)AuthorAge
* Implement CCTK_Error and friendseschnett2013-02-26
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@4963 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Move MPI support from flesh to a thorneschnett2012-07-21
| | | | | | | | | | | | | | | | Move MPI support from flesh to thorn ExternalLibraries/MPI. This also requires thorns that call MPI directly to declare this in their configuration.ccl. Existing configurations using MPI need to include ExternalLibraries/MPI into their thorn list. * Declare that the flesh optionally uses MPI (because it needs to call MPI_Init) * Update test case mechanism to find out whether MPI is used * Handle configuration bindings for flesh * Don't special-case Crays any more when configuring Linux git-svn-id: http://svn.cactuscode.org/flesh/trunk@4847 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Introduce cctk_ash, retire cctk_lssheschnett2012-07-04
| | | | | | | | | | | | Introduce cctk_ash, describing the process-local array shape that has been allocated. This may be larger than cctk_lsh, the process-local shape that should be used. Retire cctk_lssh and related infrastructure to handle staggered grid functions. git-svn-id: http://svn.cactuscode.org/flesh/trunk@4841 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Add final newlineeschnett2011-09-25
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@4731 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Provide Fortran wrapper for CCTK_SyncGroupsI().rideout2010-11-03
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@4646 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Add a new function CCTK_MaxGFDim to the flesh which returns theschnetter2008-04-07
| | | | | | | | | | | | | | | | | (maximum) dimension of all grid functions. Cactus can currently not handle the case where one uses grid arrays of dimension D while grid functions have a dimension less than D. During initialisation, Cactus sets cctk_dim to the largest dimension of all grid variables in all active thorns. This is arguably wrong; cctk_dim should be set to the (maximum) dimension of the grid functions. This patch adds a new function CCTK_MaxGFDim to the flesh which returns the (maximum) dimension of all grid functions. This function is used to set cctk_dim. git-svn-id: http://svn.cactuscode.org/flesh/trunk@4476 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Fix a bug in CactusDefaultSyncGroupsByDirI() where the flesh ran into antradke2006-07-24
| | | | | | | | | infinite recursion if a driver didn't overload CCTK_SyncGroupsByDirI(). Thanks to Daniel Herbst and David Fiske for reporting this problem. git-svn-id: http://svn.cactuscode.org/flesh/trunk@4352 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Declare and make use of new overloadable CCTK_SyncGroupsByDirI() whichtradke2006-07-06
| | | | | | | | | | | synchronises multiple groups in a single call. This function deprecates the old overloadable function CCTK_SyncGroup(). This patch concludes the discussion thread starting on http://www.cactuscode.org/old/pipermail/developers/2006-June/004933.html. git-svn-id: http://svn.cactuscode.org/flesh/trunk@4348 17b73243-c579-4c4c-a9d2-2d5706c11dac
* bumped the warning level from 3 to 1 on missing reduction operators (now ↵yye002005-12-12
| | | | | | that they are going mainstream). Fixed fortran interface to reduce local arrays, and a few minor cosmetic changes git-svn-id: http://svn.cactuscode.org/flesh/trunk@4204 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Correct typos in comments.schnetter2005-11-19
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@4197 17b73243-c579-4c4c-a9d2-2d5706c11dac
* migration from misnomer pointwise-reduction to reduce arrays globallyyye002005-11-14
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@4195 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Pointwise reduction flesh supportyye002005-11-08
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@4194 17b73243-c579-4c4c-a9d2-2d5706c11dac
* A fix to routine documentation, more of these will be made as time allows ↵yye002005-10-21
| | | | | | (stumbled on it during reduction code review). git-svn-id: http://svn.cactuscode.org/flesh/trunk@4186 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Remove unused function declarations.schnetter2005-09-26
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@4151 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Clean up the definitions of CCTK_ReduceLocalArrays andschnetter2005-08-04
| | | | | | | | | | | | | | | | | | | | | CCTK_ReduceGridArrays: Correct indentation of declaration. Use one line per argument in declaration. Rename the *output_number* arguments to *output_value*, as we did for CCTK_ReduceGridArrays. Rename *array_dim* arguments to *array_size*; otherwise the name is too similar to N_dims, which means something different. Correct grdoc headers, which were just copied from somewhere else. Invent description of return type. git-svn-id: http://svn.cactuscode.org/flesh/trunk@4105 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Do not cast t_reduce_operator** to void** -- this is illegal in C,schnetter2005-08-04
| | | | | | | | because it may lead to writing a t_reduce_operator* into a void*, and these may have different sizes. git-svn-id: http://svn.cactuscode.org/flesh/trunk@4103 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Remove self field from cGH structureschnetter2005-01-29
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3971 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Protect call to sleep, as it is not available everywhere.goodale2004-11-05
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3899 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Also change the prototype for the Fortran ReduceGridArrays routine to match.hawke2004-11-03
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3893 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Add Fortran prototype for CCTK_NumLocalArrayReduceOperators.hawke2004-11-03
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3891 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Correct the argument list for the Fortran interface for the new Reduction API.hawke2004-11-03
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3890 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Patch from Yaakoub to add a processor-number to reduce to to thegoodale2004-10-19
| | | | | | | Global reduction api. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3869 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Tidying patch from Yaakoub - removing unused vars and fixing some prototypes.goodale2004-09-09
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3861 17b73243-c579-4c4c-a9d2-2d5706c11dac
* patch from Yaakoub to add new GA reduction API.goodale2004-08-30
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3858 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Comment out debug message about registration of new reduction ops.tradke2004-08-18
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3844 17b73243-c579-4c4c-a9d2-2d5706c11dac
* New reduction APIs - as done by Yaakoub.goodale2004-08-12
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3829 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Removed old interpolator API code.tradke2004-06-20
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3771 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Added fortran wrapper for CCTK_InterpLocalUniform().tradke2004-05-17
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3733 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Fix previous patch so we only warn at most onece per processor per API.jthorn2004-05-14
| | | | | | | | This avoids problems with millions of warnings being printed if user code calls the interpolator separately at each grid point. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3729 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Add level 1 warnings to CCTK_InterpGV() and CCTK_InterpLocal()jthorn2004-05-11
| | | | | | | | | | | | (both C and Fortran versions) warning that these APIs are obsolescent and people should convert their code to the replacement APIs CCTK_InterpGridArrays() and CCTK_InterpLocalUniform() respectively. Still to be done: a Fortran wrapper for CCTK_InterpLocalUniform() ... I am working on this next... git-svn-id: http://svn.cactuscode.org/flesh/trunk@3716 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Reformat commentschnetter2004-04-04
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3641 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Remove empty filesschnetter2004-04-01
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3636 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Add CVS headerschnetter2004-04-01
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3635 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Put macros on one line so that Cactus CPP can deal with themallen2004-03-22
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3593 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Fix the return values of CactusDefaultGroupStorage{In,De}crease. Theyschnetter2004-01-16
| | | | | | | | | cannot rely on CCTK_{En,Dis}ableGroupStorage, becuase those just return a flag. Explicltly call CCTK_QueryGroupStorageI and CCTK_NumTimeLevelsI. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3516 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Fix memory leak in CactusDefaultGroupStorageDecreaseschnetter2004-01-16
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3511 17b73243-c579-4c4c-a9d2-2d5706c11dac
* CCTK_{En,Dis}ableGroupStorage return not the number of time levels,schnetter2004-01-16
| | | | | | | | | | | | but only a flag. The default GroupStorage{In,De}crease routines have to explicitly convert the flag. Also only enable group storage if more than 0 time levels were requested. This makes the pattern "enable 0 time levels in order to find the number of active time levels" work. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3510 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Fix typoschnetter2004-01-16
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3508 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Change the way in which pointers are passed to and from Fortran.schnetter2003-10-28
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3446 17b73243-c579-4c4c-a9d2-2d5706c11dac
* /* flush stdout/stderr and then wait a few seconds before callingtradke2003-10-09
| | | | | | | | | | | MPI_Abort() This cures a problem where processor 0 is slightly behind, eg. with activating thorns, and would print an error message (to stdout!!) about a thorn missing only a few milliseconds later. If other processors call CCTK_Abort() before that those messages wouldn't be seen. */ git-svn-id: http://svn.cactuscode.org/flesh/trunk@3428 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Replace CCTK_NumTimeLevels with CCTK_MaxTimeLevelsallen2003-07-16
| | | | git-svn-id: http://svn.cactuscode.org/flesh/trunk@3282 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Added a const qualifier to the cGH argument of all CCTK_SyncGroup*() fleshtradke2003-07-15
| | | | | | | routines. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3275 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Added a const qualifier to the 'cGH *' argument of the following functionstradke2003-07-14
| | | | | | | | | | | | | (and their default overloadables): - CCTK_SyncGroup - CCTK_EnableGroupCommI / CCTK_DisableGroupCommI - CCTK_EnableGroupStorageI / CCTK_DisableGroupStorageI - CCTK_GroupStorageIncrease / CCTK_GroupStorageDecrease - CCTK_CoordRegisterRange / CCTK_CoordRegisterRangePhysIndex You should also update all driver thorns (eg. PUGH) now. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3274 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Free the string returned by CCTK_GroupName();tradke2003-07-08
| | | | | | | This closes PR Cactus/1560. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3267 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Changed CactusDefaultAbort() to return with the given return code to abort withtradke2003-06-19
| | | | | | | | (and not with 0 as it was before). This fixes PR CactusPUGH/1546. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3242 17b73243-c579-4c4c-a9d2-2d5706c11dac
* cGH.h: Add three new fields cctk_timefac, cctk_levoff, andschnetter2003-06-17
| | | | | | | | | | | | | cctk_levoffdenom to the cGH structure that simplify handling coordinates in mesh refinement settings. cctk.h: Pass the new fields to Fortran routines. Add a new macro CCTK_ORIGIN_SPACE to simplify using these fields. CactusDefaultComm.c: Initialise the new fields. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3240 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Store function pointers in a struct before adding them to the array reductiontradke2003-03-26
| | | | | | | | | operator registry database. This prevents potential problems on platforms where function pointers have a different memory representation than data pointers. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3191 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Added const qualifiers to all the pointer function arguments in the flesh'stradke2003-03-10
| | | | | | | | reduction API which refer to input-only arrays and structures. This fixes former problems with passing a 'const cGH *' pointer to this API. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3164 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Added CactusDefaultInterpGridArrays() as a default for the overloadabletradke2003-01-24
| | | | | | | | | routine CCTK_InterpGridArrays(). If no driver thorn was activated to overloaded the latter then the default routine will be called and print a level-0 warning saying so. git-svn-id: http://svn.cactuscode.org/flesh/trunk@3099 17b73243-c579-4c4c-a9d2-2d5706c11dac
* Changed the order of function parameters in CCTK_InterpGridArrays().tradke2003-01-22
| | | | | | | Now it is consistent with the order used in CCTK_InterpLocalUniform(). git-svn-id: http://svn.cactuscode.org/flesh/trunk@3094 17b73243-c579-4c4c-a9d2-2d5706c11dac