| #
87843682
|
| 30-Nov-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 88dece74cfbfd056afd52d9c6c93ea90a2af64a9
|
| #
2fcbe264
|
| 30-Nov-2009 |
Jed Brown <jed@59A2.org> |
Move PetscLikely/PetscUnlikely up because they cannot appear in executable code before being defined.
Hg-commit: f967915c1bfbd14ea0ceb0839e396ab6a896a54d
|
| #
9800092a
|
| 30-Nov-2009 |
Jed Brown <jed@59A2.org> |
PetscLikely/PetscUnlikely directives using __builtin_expect
* Added configure test to see if __builtin_expect is available
* Define PetscLikely and PetscUnlikely macros
* Use PetscUnlikely in CHKE
PetscLikely/PetscUnlikely directives using __builtin_expect
* Added configure test to see if __builtin_expect is available
* Define PetscLikely and PetscUnlikely macros
* Use PetscUnlikely in CHKERRQ to produce cleaner object code at higher
optimization levels.
Hg-commit: 51e933be84ca2b4bb231d3ae339861d74921da64
show more ...
|
| #
37dd5ef8
|
| 17-Nov-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 0dd26807c4bc910027bf57bdca0afb5fd6272442
|
| #
753ebd1d
|
| 17-Nov-2009 |
Jed Brown <jed@59A2.org> |
Align all pointers returned by PetscMallocN (in optimized mode)
Configurable via --with-memalign=<4,8,16,32,64> (default is 16).
New macro PetscAddrAlign returns an aligned address.
Hg-commit: a369
Align all pointers returned by PetscMallocN (in optimized mode)
Configurable via --with-memalign=<4,8,16,32,64> (default is 16).
New macro PetscAddrAlign returns an aligned address.
Hg-commit: a36906938641ecef3f9b33ca4a3273b89dcdd566
show more ...
|
| #
65460251
|
| 06-Nov-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
changed PETSC_USE_SINGLE etc to PETSC_USE_SCALAR_SINGLE to make clearer that this variable is related to defining the scalar type
Hg-commit: 091ca03e9ad7a91d59048671631a058ed1b9044d
|
| #
25bba741
|
| 20-Oct-2009 |
Matthew Knepley <knepley@mcs.anl.gov> |
Merge
- Builder was almost certainly screwed up
Hg-commit: 2571e8ea748f6220b54dbcda3f10127411ca11ff
|
| #
f25b1546
|
| 19-Oct-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: c05b0044ccc0039b87cfe343b689e3dcca014cf4
|
| #
a6d0e24f
|
| 18-Oct-2009 |
Jed Brown <jed@59A2.org> |
PetscSleep now takes a PetscReal
PetscDrawGetPause and PetscDrawSetPause also take PetscReal
-draw_pause takes real values
Hg-commit: 5704078f50576596756b730ece1fd2c58071fd38
|
| #
458c5f23
|
| 17-Sep-2009 |
Matthew Knepley <knepley@mcs.anl.gov> |
Merge
Hg-commit: 5cfc927ddf10a2bff062bd23b3f8b5a57e4767bf
|
| #
502ef8fd
|
| 16-Sep-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: a71d78359d21f54c7745faf8b2356d209ffabb11
|
| #
4716f0a6
|
| 16-Sep-2009 |
Satish Balay <balay@mcs.anl.gov> |
cannot have '__FUNCT__' defined for a macro
Hg-commit: 3462f76b568482be59b4268da24fb02ea9cde76f
|
| #
7e48e62b
|
| 08-Sep-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 87438658935f07f4d480aa09aeb41a071a074b16
|
| #
cf89d91a
|
| 02-Sep-2009 |
Matthew Knepley <knepley@mcs.anl.gov> |
Turned off wrapper for PetscMemcpy()
Hg-commit: f0aa8508c8cc64d8f9c1af315bbd32ba18738192
|
| #
47980e60
|
| 11-Aug-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
proper include file is now included if _mm_prefetch() is detected.
Hg-commit: c3abeb1c41fcf4977ee5a91c347c5fdc4a267a7c
|
| #
0230a03c
|
| 31-Jul-2009 |
Matthew Knepley <knepley@mcs.anl.gov> |
Merge
Hg-commit: 349518ceeca84623e101fa73fa2e7aa9b404d154
|
| #
8397fe1a
|
| 27-Jul-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
inclined PetscMemcpy() and PetscMemzero(), stupid maybe but let's see how it goes
Hg-commit: a0c866eb70319ddc2a5d86cf7b851aed3de2afd0
|
| #
2f119e2b
|
| 23-Mar-2009 |
Matthew Knepley <knepley@mcs.anl.gov> |
Merge
Hg-commit: c7e2f947f4ca87be153389e8832c1a8fda8c8255
|
| #
f59f7829
|
| 19-Mar-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
added PetscHDF5IntCast() and support for using the HDF5 viewers with 64 bit indices on 32 bit systems.
changed hdf5.py to allow building with 64 bit indices.
Note: eventually we need to adding casti
added PetscHDF5IntCast() and support for using the HDF5 viewers with 64 bit indices on 32 bit systems.
changed hdf5.py to allow building with 64 bit indices.
Note: eventually we need to adding casting in the other direction. When HDF5 is using 64 bit indexing but PETSc is using 32 bit we should check that the value fits before setting into PETSc variables.
Hg-commit: bfff2b1b93762223f2e236137ac3bc8a8ed833e5
show more ...
|
| #
0a0e8b50
|
| 11-Feb-2009 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after bad merge
Hg-commit: 635f3af8457122a4c4145532c5a56b653766f1eb
|
| #
4e6ed268
|
| 08-Feb-2009 |
Matthew Knepley <knepley@mcs.anl.gov> |
Fixed compile problem
Hg-commit: 99803aaf5223b0e4b1cf795cfe94614ddf49316e
|
| #
f98b64f3
|
| 19-Dec-2008 |
Matthew Knepley <knepley@mcs.anl.gov> |
Merge
Hg-commit: e092dc0de297d68c11c2bc7cb80d34539597c562
|
| #
017a47e3
|
| 19-Dec-2008 |
Satish Balay <balay@mcs.anl.gov> |
add in PETSC_HAVE_POPEN check for python.c [required on windows]
Hg-commit: 9daadf508e1f47291aa782d5626d9e29438d1c66
|
| #
7a03c64e
|
| 19-Dec-2008 |
Barry Smith <bsmith@mcs.anl.gov> |
C:\cygwin\home\sbalay\PETSC-~1\src\mat\impls\dense\mpi\mmdense.c(413) : warning C4022: 'PetscTrFree' : pointer mismatch for actual parameter 1
Added cast to void* in PetscFree() to
C:\cygwin\home\sbalay\PETSC-~1\src\mat\impls\dense\mpi\mmdense.c(413) : warning C4022: 'PetscTrFree' : pointer mismatch for actual parameter 1
Added cast to void* in PetscFree() to get correct argument type to PetscTrFree(). Not a perfect solution but
the definition of PetscFree() where a = 0
Hg-commit: 85b1c27ad2ac80756bf01fd2766b0359c8400d2e
show more ...
|
| #
bd52ade1
|
| 11-Dec-2008 |
Barry Smith <bsmith@mcs.anl.gov> |
commit after merge
Hg-commit: 7bf0eb275ccc1d46ddeb44e0da98632bcfcbe9b2
|