librsb/librsb.changes

132 lines
5.5 KiB
Plaintext

-------------------------------------------------------------------
Sat Jul 16 20:13:42 UTC 2022 - Jan Engelhardt <jengelh@inai.de>
- Update to release 1.3.0.1
* rsb.h: additional input checks in rsb_mtx_rndr().
* rsb.hpp: additional input checks in the C++ wrappers.
* No more unaligned memory access with `rsbench
--read-performance-record` (fix for armhf,sparc64 architectures).
-------------------------------------------------------------------
Thu Apr 21 10:20:35 UTC 2022 - Jan Engelhardt <jengelh@inai.de>
- Update to release 1.3.0.0
* Considerably improved performance of rsb_spmm()/usmm() via
new kernels (can be turned off at runtime by setting
'RSB_WANT_RSBPP=0' in the environment).
* Added a C++ API (classes RsbMatrix and RsbLib) in new header
<rsb.hpp>.
* Environment variable 'RSB_NUM_THREADS' now used by default
(implicit configure '--enable-rsb-num-threads').
* Fixed integer overflow situations with limit-large matrices.
* Improved parallelism of rsb_spmv()/rsb_spmm() a bit (no
overly strict locking).
* Improved performance of rsb_spmv()/rsb_spmm() beta-scaling.
-------------------------------------------------------------------
Tue Nov 2 20:24:59 UTC 2021 - Jan Engelhardt <jengelh@inai.de>
- Update to release 1.2.0.10
* rsb_spmm(...,RSB_FLAG_WANT_ROW_MAJOR_ORDER,...) internals had
swapped ldB and ldC so having ldB > ldC could lead to a
crash; with ldC < ldB, to wrong results.
* rsb_spmm(...,RSB_FLAG_WANT_ROW_MAJOR_ORDER,...) on matrix
with RSB_FLAG_UNIT_DIAG_IMPLICIT could have computed wrong
results.
* rsb_tune_spmm()/rsb_tune_spsm() could have crashed if called
with order==RSB_FLAG_WANT_ROW_MAJOR_ORDER and auto leading
dimensions and implicit operands (ldB==0 and Bp==NULL, ldC==0
and Cp==NULL).
* rsb_mtx_get_rows_sparse() with: was ignoring
RSB_TRANSPOSITION_C.
-------------------------------------------------------------------
Thu Dec 24 12:33:02 UTC 2020 - Jan Engelhardt <jengelh@inai.de>
- Update to release 1.2.0.9
* Fixed: rsb_mtx_rndr(..RSB_MARF_EPS) and
rsb_file_mtx_rndr(..RSB_MARF_EPS): could crash on
uniform-values matrix input.
* Fixed: rsb_spmsp was crashing if result matrix has nnz<=rows
* Fixed: BLAS_zusaxpy and BLAS_cusaxpy in C were crashing
* Fixed: rsb_sppsp() was ignoring transB and could crash on
transA
-------------------------------------------------------------------
Thu Mar 5 16:24:51 UTC 2020 - Jan Engelhardt <jengelh@inai.de>
- Update to release 1.2.0.8
* bugfix: rsb_spmv/rsb_spmm/BLAS_cusmv/BLAS_zusmv/BLAS_cusmm/
BLAS_zusmm could compute wrong values in transpose or
conjugated transpose on very sparse complex hermitian
matrices.
-------------------------------------------------------------------
Thu Jul 12 05:02:16 UTC 2018 - bwiedemann@suse.com
- Add reproducible.patch and set an explicit memory hierarchy
to make build reproducible (boo#1100677)
-------------------------------------------------------------------
Tue Dec 26 14:10:41 UTC 2017 - jengelh@inai.de
- Update to new upstream release 1.2.0~rc7
* bugfix: rsb_spmv/rsb_spmm/BLAS_cusmv/BLAS_zusmv/BLAS_cusmm/
BLAS_zusmm could compute wrong values on complex hermitian
matrices if rhs imaginary part non null.
* bugfix: complex conjugated transpose
rsb_spsv/rsb_spsm/BLAS_cussv/BLAS_zussv/BLAS_cussm/BLAS_zussm
could compute wrong values if rhs imaginary part non null.
* bugfix: rsb_sppsp/rsb_mtx_clone would compute scaled
conjugate of complex matrices wrong if alpha imaginary part
non null.
* might detect a forgotten rsb_lib_init() at first matrix
allocation and return an error.
-------------------------------------------------------------------
Sat Apr 29 23:53:40 UTC 2017 - jengelh@inai.de
- Update to new upstream release 1.2.0~rc6
* BLAS_zusget_element & co will behave one-based in Fortran.
* rsb_sppsp was summing incorrectly certain non-overlapping
sparse matrices.
* rsb_file_mtx_save and rsb_file_vec_save's output use
full-precision.
- Add pun.diff
-------------------------------------------------------------------
Thu Oct 20 23:10:35 UTC 2016 - jengelh@inai.de
- Update to new upstream release 1.2.0~rc5
* Fixed EPS rendering of matrices, e.g.
`./rsbench --plot-matrix -aRzd -f matrix.mtx >matrix.eps`
- Drop shlib.diff (applied upstream)
-------------------------------------------------------------------
Wed Jun 8 21:11:26 UTC 2016 - jengelh@inai.de
- Update to new upstream release 1.2.0~rc3
* Extension: if parameter flagsA of mtx_set_vals() has
RSB_FLAG_DUPLICATES_SUM, then values will be summed up into the
matrix.
* Bugfix: rsb_mtx_get_nrm on symmetric matrices was buggy.
* Bugfix: rsb_spsm potentially wrong in --enable-openmp and
(nrhs>1). (ussm affected)
* Bugfix: rsb_spsm wrong in --disable-openmp version and
(nrhs>1). (ussm affected)
* Bugfix: rsb_spsm used to scale only first rhs when (*alphap!=1
and nrhs>1). (ussm affected)
* Bugfix: rsb_spsm used to solve only first rhs when (y != x).
(ussm not affected)
* Bugfix: rsb_spmm used to scale only first rhs when (*betap!=1
and nrhs>1). (usmm not affected)
* Bugfix: rsb_tune_spmm/rsb_tune_spsm returned (false positive)
error on ( mtxAp != NULL && mtxOpp != NULL ) rather than on (
mtxAp != NULL && mtxOpp != NULL && *mtxOpp != NULL ).
* Will use memset() on systems with no bzero() (e.g. mingw).
-------------------------------------------------------------------
Sun Jan 17 14:15:30 UTC 2016 - jengelh@inai.de
- Initial package (versoin 1.2.0~rc2) for build.opensuse.org