fix name for PetscOptionsFList() and get FList() and EList() working properly with SAWs legal values
Use classname in object for object name when calling PetscObjectPrintClassNamePrefixType() instead of some string with less precise name passed in
Modified interface for TS error norm calculation.Renamed TSErrorNormWRMS to TSErrorWeightedNorm.Implemented infinite weighted norm calculation.Added option -ts_adapt_wnormtype to set the type of
Modified interface for TS error norm calculation.Renamed TSErrorNormWRMS to TSErrorWeightedNorm.Implemented infinite weighted norm calculation.Added option -ts_adapt_wnormtype to set the type of norm computed.
show more ...
TSGetAdapt: restore name from TSGetTSAdaptThis function was called TSGetAdapt in petsc-3.3 and was changed toTSGetTSAdapt in: commit ad6bc4219076b084d4a49a29c04a77944f238f10 Author: Barry Smi
TSGetAdapt: restore name from TSGetTSAdaptThis function was called TSGetAdapt in petsc-3.3 and was changed toTSGetTSAdapt in: commit ad6bc4219076b084d4a49a29c04a77944f238f10 Author: Barry Smith <bsmith@mcs.anl.gov> Date: Sun Dec 9 18:14:24 2012 -0600 My first TS restart just 8 million other changes needed PETSc! TSGetAdapt() changed to TSGetTSAdapt() because it gets a TSAdapt object, not an Adapt object added TSAdapt binary view and loadThis rationale does not match the more established convention used byfunctions such as the following.MatGetInfoMatGetColoringMatGetNullSpaceMatGetLocalToGlobalMappingMatGetLocalToGlobalMappingBlock (31 characters)SNESGetConvergedReason
removed unnneeded and unused XXXRegisterDestroy() routines
removed PetscFunctionListDestroyAll() and put PetscFunctionListDestroy(&XXXList) into the package finalize
PetscFunctionListAdd: use macro to hide (PetscVoidFunction) castSame rationale as the last commit:Casting using PetscVoidFunction instead of (void(*)(void)) is neededwith C++ compilers that will
PetscFunctionListAdd: use macro to hide (PetscVoidFunction) castSame rationale as the last commit:Casting using PetscVoidFunction instead of (void(*)(void)) is neededwith C++ compilers that will not implicitly convert C++ language linkageto C language linkage. Since PetscFunctionListAdd() is meant to operateon generic function pointers, we hide this language linkage detail(which is inconsistently implemented) in a cast.
PetscFunctionListFind: use macro to hide (PetscVoidFunction*) castCasting using PetscVoidFunction* instead of (void(**)(void)) is neededwith C++ compilers that will not implicitly convert C++ lang
PetscFunctionListFind: use macro to hide (PetscVoidFunction*) castCasting using PetscVoidFunction* instead of (void(**)(void)) is neededwith C++ compilers that will not implicitly convert C++ language linkageto C language linkage. Since PetscFunctionListFind() is meant to operateon generic function pointers, we hide this language linkage detail(which is inconsistently implemented) in a cast.
removed string version of function name for XXXRegister(), PetscFunctionListAdd() and PetscObjectComposeFunction()
removed path and MPI_Comm arguments from PetscFunctionListFind/Add()
removed path argument to XXXInitializePackage() and XXXRegister()
removed XXXRegisterDynamic() but kept the APIs for everything else underneath the samephase I of the update to handling registering function pointers
Change all PETSC_EXTERN_C to PETSC_EXTERNHg-commit: 8d2ebbb193fb583bccc64015e35640c4e08c3426
updated EXTERN_C_BEGIN/END through TSHg-commit: c76fd273af402078ae0fa266ac7dcdc7b8d0306e
added PetscObjectComm() and used it to replace (((PetscObject)obj)->comm)Hg-commit: 3da37c458124ad48ae939f4e9823e4430ee0b8be
removed typeid from PetscHeaderCreate() args since it is not know at create time anywayHg-commit: 368e928141e31cf1ff6cf108913006e5886dc6cc
removed PETSC_NULL from C and Fortran (except declaration in C for backward compatibility). Kept PETSC_NULL_xxx for FortranFixed a few bugs where PETSC_NULL had been used incorrectly.Hg-commit: 05
removed PETSC_NULL from C and Fortran (except declaration in C for backward compatibility). Kept PETSC_NULL_xxx for FortranFixed a few bugs where PETSC_NULL had been used incorrectly.Hg-commit: 054705a517d7f4388a8a084415d7478cbe95dff4
Uncrustified src/ts/*Hg-commit: 688366eb7a8ecbc6948bb32d69e2de3093f22e25
if, for, while, switch now have consistently a space prior to the opening parenthesis.Hg-commit: cf69a5905ee4efb938b2bdd2762f87854175a6c5
Fixed bug in EXACT_FINAL_MATCHSTEP - If the final clippped time step is less than PETSC_SMALL then don't take it.Hg-commit: 75225b570bbaeb0a1163656ce70dfd4157a66dad
removed silly abreviationsHg-commit: 0641f9ef776cf11cc3336b494981da756bf9d14f
change PetscFListAdd() and PetscFListFind() arguments list to match PetscOpFListAdd/Find() argsHg-commit: 07f9e01e040feeb4162253a60ca63556436f4135
commit after mergeHg-commit: 76831309032f5585424cb3e5e4c7f8f754699d5d
My first TS restart just 8 million other changes needed PETSc!TSGetAdapt() changed to TSGetTSAdapt() because it gets a TSAdapt object, not an Adapt objectadded TSAdapt binary view and loadHg-comm
My first TS restart just 8 million other changes needed PETSc!TSGetAdapt() changed to TSGetTSAdapt() because it gets a TSAdapt object, not an Adapt objectadded TSAdapt binary view and loadHg-commit: c5ab7de6c480df2d7b84dc349ac7871cdc85dcbe
Changed ts_exact_final_time from boolean to enum.TSAdapt can now modify the final time step to stop exactly atthe given final time.Hg-commit: 2bbe8e4fc9d649fbb4ec4bc19259912d0788d4c2
12345678