Provided by: librsb-dev_1.2.0-rc7-5_amd64 
NAME
librsb - rsb.h - The reference documentation of the librsb library comes in both HTML and Unix man pages
formats. The following sections/man pages are available: The librsb library interface (rsb.h, rsb.F90) ;
The Sparse BLAS interface to librsb (blas_sparse.h, rsb_blas_sparse.F90) ; Example programs and code.
DESCRIPTION
SYNOPSIS
Macros
#define RSB_SIZEOF(TYPE) RSB_NUMERICAL_TYPE_SIZE(TYPE)
Enumerations
enum rsb_opt_t { RSB_IO_WANT_VERBOSE_INIT =0x000001, RSB_IO_WANT_VERBOSE_EXIT =0x000002,
RSB_IO_WANT_OUTPUT_STREAM =0x000003, RSB_IO_WANT_SORT_METHOD =0x000004,
RSB_IO_WANT_CACHE_BLOCKING_METHOD =0x000005, RSB_IO_WANT_SUBDIVISION_MULTIPLIER =0x000006,
RSB_IO_WANT_VERBOSE_ERRORS =0x000007, RSB_IO_WANT_BOUNDED_BOX_COMPUTATION =0x000008,
RSB_IO_WANT_EXECUTING_THREADS =0x000009, RSB_IO_WANT_EXTRA_VERBOSE_INTERFACE =0x000010,
RSB_IO_WANT_MEMORY_HIERARCHY_INFO_STRING =0x000011, RSB_IO_WANT_IS_INITIALIZED_MARKER =0x000012,
RSB_IO_WANT_MEM_ALLOC_CNT =0x000013, RSB_IO_WANT_MEM_ALLOC_TOT =0x000014,
RSB_IO_WANT_LEAF_LEVEL_MULTIVEC =0x000015, RSB_IO_WANT_MAX_MEMORY_ALLOCATIONS =0x000016,
RSB_IO_WANT_MAX_MEMORY_ALLOCATED =0x000017, RSB_IO_WANT_LIBRSB_ETIME =0x000018,
RSB_IO_WANT_VERBOSE_TUNING =0x000019 } library option values for rsb_lib_init, rsb_lib_set_opt_str,
rsb_lib_reinit, rsb_lib_exit, rsb_lib_get_opt, rsb_lib_set_opt, or (deprecated) macros
RSB_REINIT_SINGLE_VALUE_GET, RSB_REINIT_SINGLE_VALUE_SET, RSB_REINIT_SINGLE_VALUE,
RSB_REINIT_SINGLE_VALUE_C_IOP.. "
enum rsb_extff_t { RSB_EXTF_NORM_ONE =0x00001001, RSB_EXTF_NORM_TWO =0x00001002, RSB_EXTF_NORM_INF
=0x00001003, RSB_EXTF_SUMS_ROW =0x00001004, RSB_EXTF_SUMS_COL =0x00001005, RSB_EXTF_ASUMS_ROW
=0x00001006, RSB_EXTF_ASUMS_COL =0x00001007, RSB_EXTF_DIAG =0x00000004 } Extraction filter flags, to
be used with rsb_mtx_get_nrm()/rsb_mtx_get_vec(). "
enum rsb_mif_t { RSB_MIF_INDEX_STORAGE_IN_BYTES__TO__SIZE_T =0x00000001,
RSB_MIF_INDEX_STORAGE_IN_BYTES_PER_NNZ__TO__RSB_REAL_T =0x00000002,
RSB_MIF_MATRIX_ROWS__TO__RSB_COO_INDEX_T =0x00000004, RSB_MIF_MATRIX_COLS__TO__RSB_COO_INDEX_T
=0x00000008, RSB_MIF_MATRIX_NNZ__TO__RSB_NNZ_INDEX_T =0x00000010, RSB_MIF_TOTAL_SIZE__TO__SIZE_T
=0x00000020, RSB_MIF_MATRIX_FLAGS__TO__RSB_FLAGS_T =0x00000040,
RSB_MIF_MATRIX_TYPECODE__TO__RSB_TYPE_T =0x00000080, RSB_MIF_MATRIX_INFO__TO__CHAR_P =0x00000100,
RSB_MIF_LEAVES_COUNT__TO__RSB_BLK_INDEX_T =0x00000200 } Flags for getting matrix information via
rsb_mtx_get_info()/rsb_mtx_get_info_str(). "
enum rsb_elopf_t { RSB_ELOPF_MUL =0x00000001, RSB_ELOPF_DIV =0x00000002, RSB_ELOPF_POW =0x00000004,
RSB_ELOPF_NEG =0x00000008, RSB_ELOPF_SCALE_ROWS =0x00000010, RSB_ELOPF_SCALE_COLS =0x00000020,
RSB_ELOPF_SCALE_ROWS_REAL =0x00000040, RSB_ELOPF_SCALE_COLS_REAL =0x00000080 } Flags for specifying a
particular elemental/row-wise operation with rsb_mtx_upd_vals(). "
Functions
rsb_err_t rsb_lib_init (struct rsb_initopts *iop)
rsb_err_t rsb_lib_set_opt_str (const rsb_char_t *opnp, const rsb_char_t *opvp)
rsb_err_t rsb_lib_reinit (struct rsb_initopts *iop)
rsb_err_t rsb_lib_exit (struct rsb_initopts *iop)
struct rsb_mtx_t * rsb_mtx_alloc_from_coo_const (const void *VA, const rsb_coo_idx_t *IA, const
rsb_coo_idx_t *JA, rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA,
rsb_blk_idx_t brA, rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t *errvalp)
struct rsb_mtx_t * rsb_mtx_alloc_from_coo_inplace (void *VA, rsb_coo_idx_t *IA, rsb_coo_idx_t *JA,
rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA, rsb_blk_idx_t brA,
rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t *errvalp)
struct rsb_mtx_t * rsb_mtx_free (struct rsb_mtx_t *mtxAp)
rsb_err_t rsb_mtx_clone (struct rsb_mtx_t **mtxBpp, rsb_type_t typecode, rsb_trans_t transA, const void
*alphap, const struct rsb_mtx_t *mtxAp, rsb_flags_t flags)
rsb_err_t rsb_spmv (rsb_trans_t transA, const void *alphap, const struct rsb_mtx_t *mtxAp, const void
*Xp, rsb_coo_idx_t incX, const void *betap, void *Yp, rsb_coo_idx_t incY)
rsb_err_t rsb_spsv (rsb_trans_t transT, const void *alphap, const struct rsb_mtx_t *mtxTp, const void
*Xp, rsb_coo_idx_t incX, void *Yp, rsb_coo_idx_t incY)
rsb_err_t rsb_spsm (rsb_trans_t transT, const void *alphap, const struct rsb_mtx_t *mtxTp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void *betap, const void *Bp, rsb_nnz_idx_t ldB, void *Cp,
rsb_nnz_idx_t ldC)
rsb_err_t rsb_coo_sort (void *VA, rsb_coo_idx_t *IA, rsb_coo_idx_t *JA, rsb_nnz_idx_t nnzA, rsb_coo_idx_t
nrA, rsb_coo_idx_t ncA, rsb_type_t typecode, rsb_flags_t flagsA)
rsb_err_t rsb_file_mtx_get_dims (const char *filename, rsb_coo_idx_t *nrp, rsb_coo_idx_t *ncp,
rsb_coo_idx_t *nzp, rsb_flags_t *flagsp)
rsb_err_t rsb_perror (void *stream, rsb_err_t errval)
rsb_err_t rsb_strerror_r (rsb_err_t errval, rsb_char_t *buf, size_t buflen)
rsb_err_t rsb_mtx_upd_vals (struct rsb_mtx_t *mtxAp, enum rsb_elopf_t elop_flags, const void *omegap)
rsb_err_t rsb_mtx_set_vals (struct rsb_mtx_t *mtxAp, const void *VA, const rsb_coo_idx_t *IA, const
rsb_coo_idx_t *JA, rsb_nnz_idx_t nnz, rsb_flags_t flags)
rsb_err_t rsb_mtx_get_vals (const struct rsb_mtx_t *mtxAp, void *VA, const rsb_coo_idx_t *IA, const
rsb_coo_idx_t *JA, rsb_nnz_idx_t nnz, rsb_flags_t flags)
rsb_err_t rsb_file_mtx_save (const struct rsb_mtx_t *mtxAp, const rsb_char_t *filename)
rsb_err_t rsb_file_vec_save (const rsb_char_t *filename, rsb_type_t typecode, const void *Yp,
rsb_coo_idx_t yvl)
rsb_err_t rsb_file_vec_load (const rsb_char_t *filename, rsb_type_t typecode, void *Yp, rsb_coo_idx_t
*yvlp)
struct rsb_mtx_t * rsb_file_mtx_load (const rsb_char_t *filename, rsb_flags_t flagsA, rsb_type_t
typecode, rsb_err_t *errvalp)
struct rsb_mtx_t * rsb_sppsp (rsb_type_t typecode, rsb_trans_t transA, const void *alphap, const struct
rsb_mtx_t *mtxAp, rsb_trans_t transB, const void *betap, const struct rsb_mtx_t *mtxBp, rsb_err_t
*errvalp)
struct rsb_mtx_t * rsb_spmsp (rsb_type_t typecode, rsb_trans_t transA, const void *alphap, const struct
rsb_mtx_t *mtxAp, rsb_trans_t transB, const void *betap, const struct rsb_mtx_t *mtxBp, rsb_err_t
*errvalp)
rsb_err_t rsb_mtx_add_to_dense (const void *alphap, const struct rsb_mtx_t *mtxAp, rsb_nnz_idx_t ldB,
rsb_nnz_idx_t nrB, rsb_nnz_idx_t ncB, rsb_bool_t rowmajorB, void *Bp)
rsb_trans_t rsb_psblas_trans_to_rsb_trans (const char psbtrans)
struct rsb_mtx_t * rsb_mtx_alloc_from_csr_const (const void *VA, const rsb_coo_idx_t *RP, const
rsb_coo_idx_t *JA, rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA,
rsb_blk_idx_t brA, rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t *errvalp)
struct rsb_mtx_t * rsb_mtx_alloc_from_csc_const (const void *VA, const rsb_coo_idx_t *IA, const
rsb_coo_idx_t *CP, rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA,
rsb_blk_idx_t brA, rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t *errvalp)
struct rsb_mtx_t * rsb_mtx_alloc_from_csr_inplace (void *VA, rsb_nnz_idx_t *RP, rsb_coo_idx_t *JA,
rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA, rsb_blk_idx_t brA,
rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t *errvalp)
rsb_err_t rsb_mtx_switch_to_csr (struct rsb_mtx_t *mtxAp, void **VAp, rsb_coo_idx_t **IAp, rsb_coo_idx_t
**JAp, rsb_flags_t flags)
rsb_err_t rsb_mtx_get_coo (const struct rsb_mtx_t *mtxAp, void *VA, rsb_coo_idx_t *IA, rsb_coo_idx_t *JA,
rsb_flags_t flags)
rsb_err_t rsb_mtx_get_csr (rsb_type_t typecode, const struct rsb_mtx_t *mtxAp, void *VA, rsb_nnz_idx_t
*RP, rsb_coo_idx_t *JA, rsb_flags_t flags)
rsb_err_t rsb_mtx_get_rows_sparse (rsb_trans_t transA, const void *alphap, const struct rsb_mtx_t *mtxAp,
void *VA, rsb_coo_idx_t *IA, rsb_coo_idx_t *JA, rsb_coo_idx_t frA, rsb_coo_idx_t lrA, rsb_nnz_idx_t
*rnzp, rsb_flags_t flags)
rsb_err_t rsb_mtx_get_coo_block (const struct rsb_mtx_t *mtxAp, void *VA, rsb_coo_idx_t *IA,
rsb_coo_idx_t *JA, rsb_coo_idx_t frA, rsb_coo_idx_t lrA, rsb_coo_idx_t fcA, rsb_coo_idx_t lcA,
rsb_coo_idx_t *IREN, rsb_coo_idx_t *JREN, rsb_nnz_idx_t *rnzp, rsb_flags_t flags)
rsb_err_t rsb_spmm (rsb_trans_t transA, const void *alphap, const struct rsb_mtx_t *mtxAp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void *Bp, rsb_nnz_idx_t ldB, const void *betap, void *Cp,
rsb_nnz_idx_t ldC)
rsb_err_t rsb_spmsp_to_dense (rsb_type_t typecode, rsb_trans_t transA, const void *alphap, const struct
rsb_mtx_t *mtxAp, rsb_trans_t transB, const void *betap, const struct rsb_mtx_t *mtxBp, rsb_nnz_idx_t
ldC, rsb_nnz_idx_t nrC, rsb_nnz_idx_t ncC, rsb_bool_t rowmajorC, void *Cp)
rsb_err_t rsb_mtx_rndr (const char *filename, const struct rsb_mtx_t *mtxAp, rsb_coo_idx_t pmWidth,
rsb_coo_idx_t pmHeight, rsb_marf_t rflags)
rsb_err_t rsb_file_mtx_rndr (void *pmp, const char *filename, rsb_coo_idx_t pmlWidth, rsb_coo_idx_t
pmWidth, rsb_coo_idx_t pmHeight, rsb_marf_t rflags)
rsb_err_t rsb_mtx_switch_to_coo (struct rsb_mtx_t *mtxAp, void **VAp, rsb_coo_idx_t **IAp, rsb_coo_idx_t
**JAp, rsb_flags_t flags)
rsb_err_t rsb_mtx_get_prec (void *opdp, const struct rsb_mtx_t *mtxAp, rsb_precf_t prec_flags, const void
*ipdp)
rsb_err_t rsb_mtx_get_info (const struct rsb_mtx_t *mtxAp, enum rsb_mif_t miflags, void *minfop)
rsb_err_t rsb_mtx_get_info_str (const struct rsb_mtx_t *mtxAp, const rsb_char_t *mis, void *minfop,
size_t buflen)
rsb_err_t rsb_mtx_get_nrm (const struct rsb_mtx_t *mtxAp, void *Np, enum rsb_extff_t flags)
rsb_err_t rsb_mtx_get_vec (const struct rsb_mtx_t *mtxAp, void *Dp, enum rsb_extff_t flags)
rsb_time_t rsb_time (void)
rsb_err_t rsb_tune_spmm (struct rsb_mtx_t **mtxOpp, rsb_real_t *sfp, rsb_int_t *tnp, rsb_int_t maxr,
rsb_time_t maxt, rsb_trans_t transA, const void *alphap, const struct rsb_mtx_t *mtxAp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void *Bp, rsb_nnz_idx_t ldB, const void *betap, void *Cp,
rsb_nnz_idx_t ldC)
rsb_err_t rsb_tune_spsm (struct rsb_mtx_t **mtxOpp, rsb_real_t *sfp, rsb_int_t *tnp, rsb_int_t maxr,
rsb_time_t maxt, rsb_trans_t transA, const void *alphap, const struct rsb_mtx_t *mtxAp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void *Bp, rsb_nnz_idx_t ldB, const void *betap, void *Cp,
rsb_nnz_idx_t ldC)
Detailed Description
The reference documentation of the librsb library comes in both HTML and Unix man pages formats. The
following sections/man pages are available: The librsb library interface (rsb.h, rsb.F90) ; The Sparse
BLAS interface to librsb (blas_sparse.h, rsb_blas_sparse.F90) ; Example programs and code.
In general, users of this library are interested in high performance sparse matrix computations on cache
based shared memory parallel computers. For this, librsb offers a native C interface (here documented)
and a Fortran one (in rsb.F90, equivalent to the C declaration headers from rsb.h), in addition to a the
Sparse BLAS one (both C and Fortran, documented).
Configuration, build, and installation instructions are contained in the README file distributed in the
sources archive.
Typical program structure
• initialize librsb with rsb_lib_init()
• (in any order) allocate matrices (e.g.: with rsb_mtx_alloc_from_coo_inplace() or others); do any
computation with them (e.g.: rsb_spmv(), rsb_spsv() ); converting matrices (e.g.: with
rsb_mtx_switch_to_coo() ); freeing matrices (rsb_mtx_free() )
• finalize librsb with rsb_lib_exit()
Important usage notes
General program structure Before calling any librsb function, a program is required to initialize
librsb's internal status. This is done by calling rsb_lib_init() . Afterwards, any librsb function can be
safely used. When librsb functions are not intended to be called anymore, a program may call
rsb_lib_exit() to free any resource. Then, rsb_lib_init() should be called for further usage of librsb.
Manipulating matrices and vectors In order to use librsb, the user is not required to use explicitly
any of librsb's data structures: their manipulation is to be performed by librsb functions. Therefore,
knowledge of librsb's matrix type (rsb_mtx_t) is not necessary at all: this structure is intended to be
used as an opaque container.
On the contrary, arrays for numerical vectors (or more generally, dense matrices) are expected to be
managed by the user: librsb does not furnish any specific vector type. Computational functions treat
dense vectors/matrices are simple arrays of a specified type; see the Example programs and code .
Computational functions This library can be configured at build time to support a custom subset of
numerical types. To keep the programming interface compact, it has been decided to not replicate the
computational functions to each numerical type. Instead, the type is expected to be specified by the user
via a type flag. For instance, matrix assembly functions (e.g.: rsb_mtx_alloc_from_coo_const() ) accept a
type information and keep it stored in the matrix structure. Therefore, computational functions (e.g.:
rsb_spmv() ) can fetch this information from their rsb_mtx_t operand, and treat accordingly the other
parameters (e.g.: alphap, Xp, ...). Mixed type operations are currently not supported.
Memory management
Matrix structures (rsb_mtx_t) allocated by librsb shall be freed only via rsb_mtx_free() .
Benchmarking
If you want to benchmark this library, there are different possibilities:
#!/bin/sh
# systematic comparative benchmark, mostly for dense matrices
# (with Intel MKL, if linked) benchmark comparing
# produces a number of plots systematically
bench/dense.sh
# the benchmark command; assumes A.mtx is a file in Matrix Market format
./rsbench -oa -Ob -f A.mtx -qH -R -n1 -t100 --verbose -TD --compare-competitors
# rsbench is very flexible tool; see the help for it:
./rsbench -oa -Ob --help
Tuning and Customization
There are different ./configure options you may look at for tuning or customizing the library.
Macro Definition Documentation
#define RSB_SIZEOF(TYPE) RSB_NUMERICAL_TYPE_SIZE(TYPE)
Use RSB_SIZEOF macro to get the size (in bytes) of a type supported by the library (e.g.: when allocating
numerical vectors).
Enumeration Type Documentation
enum rsb_elopf_t
Flags for specifying a particular elemental/row-wise operation with rsb_mtx_upd_vals().
Enumerator
RSB_ELOPF_MUL
Elemental multiplication of the matrix by a specified scalar (usable with rsb_mtx_upd_vals(),
binary operation).
RSB_ELOPF_DIV
Elemental division by a specified scalar (usable with rsb_mtx_upd_vals(), binary operation).
RSB_ELOPF_POW
Elemental power to a specified scalar (usable with rsb_mtx_upd_vals(), binary operation).
RSB_ELOPF_NEG
Elemental negation (usable with rsb_mtx_upd_vals(), unary operation).
RSB_ELOPF_SCALE_ROWS
Row wise scaling by a specified scaling vector (usable with rsb_mtx_upd_vals(), binary operation).
RSB_ELOPF_SCALE_COLS
Column wise scaling by a specified scaling vector (usable with rsb_mtx_upd_vals(), binary
operation).
RSB_ELOPF_SCALE_ROWS_REAL
Row wise scaling by a specified scaling vector. If matrix is of a complex type, the argument is
expected to be of the corresponding real type (assumed that that type has been enabled). (usable
with rsb_mtx_upd_vals(), binary operation).
RSB_ELOPF_SCALE_COLS_REAL
Column wise scaling by a specified scaling vector. If matrix is of a complex type, the argument is
expected to be of the corresponding real type (assumed that that type has been enabled). (usable
with rsb_mtx_upd_vals(), binary operation).
enum rsb_extff_t
Extraction filter flags, to be used with rsb_mtx_get_nrm()/rsb_mtx_get_vec().
Enumerator
RSB_EXTF_NORM_ONE
rsb_mtx_get_nrm() flag value for computing the one-norm.
RSB_EXTF_NORM_TWO
rsb_mtx_get_nrm() flag value for computing the two-norm (Frobenius norm).
RSB_EXTF_NORM_INF
rsb_mtx_get_nrm() flag value for computing the infinity-norm.
RSB_EXTF_SUMS_ROW
rsb_mtx_get_vec() flag value for computing the sum along each row.
RSB_EXTF_SUMS_COL
rsb_mtx_get_vec() flag value for computing the sum along each column.
RSB_EXTF_ASUMS_ROW
rsb_mtx_get_vec() flag value for computing the absolute values sum, along each row.
RSB_EXTF_ASUMS_COL
rsb_mtx_get_vec() flag value for computing the absolute values sum, along each column.
RSB_EXTF_DIAG
rsb_mtx_get_vec() flag value for extracting the diagonal submatrix.
enum rsb_mif_t
Flags for getting matrix information via rsb_mtx_get_info()/rsb_mtx_get_info_str().
Enumerator
RSB_MIF_INDEX_STORAGE_IN_BYTES__TO__SIZE_T
Index storage occupation, in bytes. (size_t)
RSB_MIF_INDEX_STORAGE_IN_BYTES_PER_NNZ__TO__RSB_REAL_T
Index storage occupation per nnz, in bytes. (rsb_real_t)
RSB_MIF_MATRIX_ROWS__TO__RSB_COO_INDEX_T
Rows count(rsb_coo_idx_t)
RSB_MIF_MATRIX_COLS__TO__RSB_COO_INDEX_T
Columns count (rsb_coo_idx_t)
RSB_MIF_MATRIX_NNZ__TO__RSB_NNZ_INDEX_T
Nonzeroes count (rsb_nnz_idx_t)
RSB_MIF_TOTAL_SIZE__TO__SIZE_T
Total size, in bytes (size_t)
RSB_MIF_MATRIX_FLAGS__TO__RSB_FLAGS_T
Matrix flags (rsb_flags_t)
RSB_MIF_MATRIX_TYPECODE__TO__RSB_TYPE_T
Matrix type code (rsb_type_t)
RSB_MIF_MATRIX_INFO__TO__CHAR_P
Matrix info string, only for rsb_mtx_get_info_str() (rsb_char_t*)
RSB_MIF_LEAVES_COUNT__TO__RSB_BLK_INDEX_T
Leaf submatrices count (rsb_blk_idx_t)
enum rsb_opt_t
library option values for rsb_lib_init, rsb_lib_set_opt_str, rsb_lib_reinit, rsb_lib_exit,
rsb_lib_get_opt, rsb_lib_set_opt, or (deprecated) macros RSB_REINIT_SINGLE_VALUE_GET,
RSB_REINIT_SINGLE_VALUE_SET, RSB_REINIT_SINGLE_VALUE, RSB_REINIT_SINGLE_VALUE_C_IOP..
Enumerator
RSB_IO_WANT_VERBOSE_INIT
RSB_IO_WANT_VERBOSE_INIT prompts for a verbose initialization of the library: messages will be
written to the file descriptor (FILE*) pointed by the value pointer when calling rsb_lib_init.
RSB_IO_WANT_VERBOSE_EXIT
RSB_IO_WANT_VERBOSE_EXIT prompts for a verbose finalization of the library: messages will be
written to the file descriptor (FILE*) pointed by the value pointer when calling rsb_lib_exit.
RSB_IO_WANT_OUTPUT_STREAM
Specifies the default output stream. Output (debug info) info will be written to the file
descriptor (FILE*) pointed by the value pointer.
RSB_IO_WANT_SORT_METHOD
Specifies the default sorting method. Specified as a pointed integer (rsb_int_t) number, in
{[0],1}. (internal)
RSB_IO_WANT_CACHE_BLOCKING_METHOD
Specifies the default cache blocking method. Specified as a pointed integer (rsb_int_t) number, in
{-1,[0],1}. (internal)
RSB_IO_WANT_SUBDIVISION_MULTIPLIER
Specifies a multiplier for finer (if >1.0) or coarser (if <1.0) subdivisions. Specified as a
pointed (rsb_real_t) number, in {..,[1.0],..}. (internal)
RSB_IO_WANT_VERBOSE_ERRORS
Prompts for a verbose error reporting: messages will be written to the file descriptor (FILE*)
pointed by the value pointer. Only meaningful if an interface error verbosity greater than 0 was
set at configure time.
RSB_IO_WANT_BOUNDED_BOX_COMPUTATION
Prompts for bounded box computation, for a smoother submatrices locking; pointed rsb_int_t in
{0,[1]}. (internal).
RSB_IO_WANT_EXECUTING_THREADS
Specifies the number of desired executing threads; pointed rsb_int_t in {[0],1,..}.
RSB_IO_WANT_EXTRA_VERBOSE_INTERFACE
Specifies the level of interface verbosity; if setting, pointed rsb_int_t values should be in
{[0],1,..}. Support may be enabled or disabled at build time via the --enable-internals-error-
verbosity configure option. If disabled, only getting is supported and yields -1, but setting is
not supported and the RSB_ERR_NO_STREAM_OUTPUT_CONFIGURED_OUT error will be returned.
RSB_IO_WANT_MEMORY_HIERARCHY_INFO_STRING
Specifies a custom memory hierarchy info string; pointed const rsb_char_t*; (may point to a NULL
string pointer).
RSB_IO_WANT_IS_INITIALIZED_MARKER
Used for getting whether the library has been initialized (RSB_BOOL_TRUE) or not (RSB_BOOL_FALSE)
; pointed const rsb_bool_t*; (this is NOT for general users).
RSB_IO_WANT_MEM_ALLOC_CNT
Used for getting the count of memory allocations performed by librsb employing librsb's memory
allocation wrapper (if disabled, will return zero); pointed const size_t*; (this is for debugging
purposes).
RSB_IO_WANT_MEM_ALLOC_TOT
Used for getting the total amount of memory allocated by librsb employing librsb's memory
allocation wrapper (if disabled, will return zero); pointed const size_t*; (this is for debugging
purposes).
RSB_IO_WANT_LEAF_LEVEL_MULTIVEC
Specifies whether the default multi-vector ops shall act at a leaf level (default value of 0 is
yes). Specified as a pointed integer (rsb_int_t) number, in {-1,[0]}. (internal)
RSB_IO_WANT_MAX_MEMORY_ALLOCATIONS
Specifies an upper limit to the count of allocated memory areas (default value of 0 means no
limit). Specified as a pointed size_t. Only works if the memory wrapper (--enable-allocator-
wrapper) has been specified at configure time.
RSB_IO_WANT_MAX_MEMORY_ALLOCATED
Specifies an upper limit to the amount of allocated memory (default value of 0 means no limit).
Specified as a pointed size_t. Only works if the memory wrapper (--enable-allocator-wrapper) has
been specified at configure time.
RSB_IO_WANT_LIBRSB_ETIME
Represents time spent in librsb. Specified as a pointed rsb_time_t. Only works if statistics
collection (--enable-librsb-stats) was specified at configure time.
RSB_IO_WANT_VERBOSE_TUNING
Auto tuning verbosity level for rsb_tune_spmm/rsb_tune_spsm. If 0, no verbosity; if 1, verbose; if
2, verbose with trace files being dumped.
Function Documentation
rsb_err_t rsb_coo_sort (void * VA, rsb_coo_idx_t * IA, rsb_coo_idx_t * JA, rsb_nnz_idx_t nnzA, rsb_coo_idx_t
nrA, rsb_coo_idx_t ncA, rsb_type_t typecode, rsb_flags_t flagsA)
Sorts row-major the given COO input arrays representing a sparse matrix $A$.
Parameters:
VA,IA,JA Output numerical values (VA) array; output row (IA) and column (JA) indices arrays.
nnzA The number of nonzeroes in the input arrays representing matrix $A$.
nrA,ncA The number of rows and columns of the sparse matrix $A$.
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
flagsA A valid combination of matrix storage flags. If unsure, use RSB_FLAG_NOFLAGS.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_time, rsb_coo_sort
Note:
By invoking with swapped IA and JA (and swapping nrA and ncA as well) one can obtain column major
order.
rsb_err_t rsb_file_mtx_get_dims (const char * filename, rsb_coo_idx_t * nrp, rsb_coo_idx_t * ncp,
rsb_coo_idx_t * nzp, rsb_flags_t * flagsp)
Reads structural information (dimensions, structural flags) for a matrix file into user specified (and
optionally NULL) variables.
Parameters:
filename The specified matrix file name (cannot be NULL).
nrp,ncp Output pointers to rows and columns count variables (can be NULL).
nzp Output pointer to the nonzeroes count variable (can be NULL).
flagsp Output pointer to the detected structural flags variable. Will be a combination of
RSB_FLAG_LOWER, RSB_FLAG_UPPER, RSB_FLAG_SYMMETRIC, RSB_FLAG_HERMITIAN.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error. If read dimensions are illegal (see
rsb_coo_idx_t,rsb_nnz_idx_t), RSB_ERR_LIMITS will be returned.
Note:
The only sparse matrix file format currently supported is Matrix Market. E.g.:
%%MatrixMarket matrix coordinate real symmetric
%
% A Hilbert Matrix of order 3, so with 3 rows, 3 columns, and 6 nonzeroes.
%
3 3 6
1 1 1.0
2 1 0.5
2 2 0.33
3 1 0.33
3 2 0.25
3 3 0.2
In the above example header on the first line, you can specify either real or complex or pattern for
the numerical type. Either general, symmetric, hermitian can be specified for the structure. In case
of pattern matrices, only coordinate indices will be loaded (saving pattern matrices is not yet
supported); in case of real matrices, also one coefficient value will be saved/loaded; in the case of
complex matrices, both the real and imaginary parts will be saved/loaded in addition to the indices.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
struct rsb_mtx_t* rsb_file_mtx_load (const rsb_char_t * filename, rsb_flags_t flagsA, rsb_type_t typecode,
rsb_err_t * errvalp)
Loads a sparse matrix from the specified matrix file, assembling it in the format specified by flags,
using the numerical type representation as specified by the user.
Parameters:
filename The specified matrix file name (cannot be NULL).
flagsA A valid combination of matrix storage flags.
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
errvalp An optional (can be NULL) pointer to rsb_err_t where the error status will be written to.
Returns:
On success, a valid pointer (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL.
Note:
The only sparse matrix file format currently supported is Matrix Market. E.g.:
%%MatrixMarket matrix coordinate real symmetric
%
% A Hilbert Matrix of order 3, so with 3 rows, 3 columns, and 6 nonzeroes.
%
3 3 6
1 1 1.0
2 1 0.5
2 2 0.33
3 1 0.33
3 2 0.25
3 3 0.2
In the above example header on the first line, you can specify either real or complex or pattern for
the numerical type. Either general, symmetric, hermitian can be specified for the structure. In case
of pattern matrices, only coordinate indices will be loaded (saving pattern matrices is not yet
supported); in case of real matrices, also one coefficient value will be saved/loaded; in the case of
complex matrices, both the real and imaginary parts will be saved/loaded in addition to the indices.
See also:
rsb_mtx_get_info, rsb_mtx_get_info_str, rsb_file_mtx_save, rsb_file_vec_load, rsb_file_mtx_load
rsb_err_t rsb_file_mtx_rndr (void * pmp, const char * filename, rsb_coo_idx_t pmlWidth, rsb_coo_idx_t
pmWidth, rsb_coo_idx_t pmHeight, rsb_marf_t rflags)
Renders as pixel map the matrix contained in a matrix file.
Parameters:
pmp Pixel map array pointer.
filename The specified matrix file name (cannot be NULL).
pmlWidth stride between lines (in pixels; no less than pmWidth).
pmWidth Pixel map width (in pixels or points).
pmHeight Pixel map height (in pixels or points).
rflags The color mode; only RSB_MARF_RGB is supported for now (1 byte per channel, 3 channels ---
red, green, blue): this requires array pmp to be at least (3*pmlWidth*pmHeight) bytes large.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Warning:
This function has not been thoroughly tested.
Note:
At the time being, pmlWidth is required to be equal to pmWidth.
See also:
rsb_mtx_rndr, rsb_file_mtx_rndr
rsb_err_t rsb_file_mtx_save (const struct rsb_mtx_t * mtxAp, const rsb_char_t * filename)
Saves the given matrix to the specified matrix file.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
filename The specified output file name (if NULL, will write to standard output).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Warning:
Some structural info contained in the matrix structural flags may be lost in the output data.
Note:
The only sparse matrix file format currently supported is Matrix Market. E.g.:
%%MatrixMarket matrix coordinate real symmetric
%
% A Hilbert Matrix of order 3, so with 3 rows, 3 columns, and 6 nonzeroes.
%
3 3 6
1 1 1.0
2 1 0.5
2 2 0.33
3 1 0.33
3 2 0.25
3 3 0.2
In the above example header on the first line, you can specify either real or complex or pattern for
the numerical type. Either general, symmetric, hermitian can be specified for the structure. In case
of pattern matrices, only coordinate indices will be loaded (saving pattern matrices is not yet
supported); in case of real matrices, also one coefficient value will be saved/loaded; in the case of
complex matrices, both the real and imaginary parts will be saved/loaded in addition to the indices.
See also:
rsb_mtx_get_info, rsb_mtx_get_info_str, rsb_file_mtx_save, rsb_file_vec_load, rsb_file_mtx_load
rsb_err_t rsb_file_vec_load (const rsb_char_t * filename, rsb_type_t typecode, void * Yp, rsb_coo_idx_t *
yvlp)
Loads a dense vector from the specified file, using the numerical type representation as specified by the
user. This function is intended to be called in two steps: first with Yp=NULL, in order to write the
vector length to *yvlp ; then, with yvlp=NULL, to get Yp written.
Parameters:
filename The specified vector file name (cannot be NULL).
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
Yp The input array vector.
yvlp An optional pointer (can be NULL). If supplied, vector length will be written here, and no
vector will be read.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Note:
The only dense vector file format currently supported is Matrix Market. E.g.:
%%MatrixMarket matrix array complex general
6 1
11.000000000000000E+000 12.000000000000000E+000
21.000000000000000E+000 22.000000000000000E+000
31.000000000000000E+000 32.000000000000000E+000
41.000000000000000E+000 42.000000000000000E+000
51.000000000000000E+000 52.000000000000000E+000
61.000000000000000E+000 62.000000000000000E+000
In the above example header on the first line, you can specify either real or complex or pattern for
the numerical type.
See also:
rsb_mtx_get_info, rsb_mtx_get_info_str, rsb_file_mtx_save, rsb_file_vec_load, rsb_file_mtx_load
rsb_err_t rsb_file_vec_save (const rsb_char_t * filename, rsb_type_t typecode, const void * Yp, rsb_coo_idx_t
yvl)
Saves a dense vector to the specified file, using the numerical type representation as specified by the
user. This function assumes Yp!=NULL and yvl>0.
Parameters:
filename The specified vector file name (cannot be NULL).
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
Yp The output array vector.
yvl Output vector length.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Note:
The only dense vector file format currently supported is Matrix Market. E.g.:
%%MatrixMarket matrix array complex general
6 1
11.000000000000000E+000 12.000000000000000E+000
21.000000000000000E+000 22.000000000000000E+000
31.000000000000000E+000 32.000000000000000E+000
41.000000000000000E+000 42.000000000000000E+000
51.000000000000000E+000 52.000000000000000E+000
61.000000000000000E+000 62.000000000000000E+000
In the above example header on the first line, you can specify either real or complex or pattern for
the numerical type.
See also:
rsb_mtx_get_info, rsb_mtx_get_info_str, rsb_file_mtx_save, rsb_file_vec_load, rsb_file_mtx_load
rsb_err_t rsb_lib_exit (struct rsb_initopts * iop)
Finalize librsb.
rsb_lib_exit should be called after having freed all matrices.
If not all of the data structures were properly deallocated before, this function may still attempt
finalizing the library and return the RSB_ERR_MEMORY_LEAK error code (this depends on the --enable-
allocator-wrapper configure time option). Any allocated memory will be lost (librsb does not keep track
of allocated matrices).
Internal library state will be cleared. After this call, it is legal to initialize the library again, by
calling rsb_lib_init().
On an error, the library state may be inconsistent, so it is advisable to either terminate program
execution (rather than forcing a new initialization with rsb_lib_init()).
Parameter iop is reserved for future use; for now it is safe to pass RSB_NULL_EXIT_OPTIONS.
Parameters:
iop A pointer to a rsb_initopts structure with library options. It may be NULL (or better,
RSB_NULL_INIT_OPTIONS/RSB_NULL_EXIT_OPTIONS) for specifying default options.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_lib_init, rsb_lib_set_opt_str, rsb_lib_reinit, rsb_lib_exit, rsb_lib_get_opt, rsb_lib_set_opt, or
(deprecated) macros RSB_REINIT_SINGLE_VALUE_GET, RSB_REINIT_SINGLE_VALUE_SET,
RSB_REINIT_SINGLE_VALUE, RSB_REINIT_SINGLE_VALUE_C_IOP..
rsb_err_t rsb_lib_init (struct rsb_initopts * iop)
This is the library initialization function.
It must be called only once before using any other library function.
It is allowed to call it again after rsb_lib_exit().
To fine-tune the library behaviour, one may specify a number of options via the iop parameter.
Options may be specified also after rsb_lib_init() by calling rsb_lib_reinit().
One may call RSB_REINIT_SINGLE_VALUE_GET with flag RSB_IO_WANT_IS_INITIALIZED_MARKER to verify whether
the library has been initialized or not.
Parameters:
iop A pointer to a rsb_initopts structure with library options. It may be NULL (or better,
RSB_NULL_INIT_OPTIONS/RSB_NULL_EXIT_OPTIONS) for specifying default options.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_lib_init, rsb_lib_set_opt_str, rsb_lib_reinit, rsb_lib_exit, rsb_lib_get_opt, rsb_lib_set_opt, or
(deprecated) macros RSB_REINIT_SINGLE_VALUE_GET, RSB_REINIT_SINGLE_VALUE_SET,
RSB_REINIT_SINGLE_VALUE, RSB_REINIT_SINGLE_VALUE_C_IOP..
rsb_err_t rsb_lib_reinit (struct rsb_initopts * iop)
Changes the library operation options which were set at initialization time either by a user or as
defaults.
Not all options may be supported, depending on build time library settings.
If an unsupported option was specified, an appropriate error (e.g.: RSB_ERR_UNSUPPORTED_OPERATION) will
be returned.
On the first error, option processing is interrupted and the remaining options (if any) are not
processed.
Program execution may continue safely even if an error code is returned (that is, library status should
be consistent).
Parameters:
iop A pointer to a rsb_initopts structure with library options. It may be NULL (or better,
RSB_NULL_INIT_OPTIONS/RSB_NULL_EXIT_OPTIONS) for specifying default options.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_lib_init, rsb_lib_set_opt_str, rsb_lib_reinit, rsb_lib_exit, rsb_lib_get_opt, rsb_lib_set_opt, or
(deprecated) macros RSB_REINIT_SINGLE_VALUE_GET, RSB_REINIT_SINGLE_VALUE_SET,
RSB_REINIT_SINGLE_VALUE, RSB_REINIT_SINGLE_VALUE_C_IOP..
rsb_err_t rsb_lib_set_opt_str (const rsb_char_t * opnp, const rsb_char_t * opvp)
Specifies individual library options in order to fine-tune the library behaviour. Both the option name
and the value shall be expressed as strings, identical to their preprocessor identifiers (see rsb_opt_t
). The opnp string will be translated internally to the corresponding request flag values, and the passed
value will be parsed out of the opvp string.
Parameters:
opnp A pointer to a library option input name string (may not be NULL).
opvp A pointer to a library option input value string (may not be NULL).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_lib_init, rsb_lib_set_opt_str, rsb_lib_reinit, rsb_lib_exit, rsb_lib_get_opt, rsb_lib_set_opt, or
(deprecated) macros RSB_REINIT_SINGLE_VALUE_GET, RSB_REINIT_SINGLE_VALUE_SET,
RSB_REINIT_SINGLE_VALUE, RSB_REINIT_SINGLE_VALUE_C_IOP..
rsb_err_t rsb_mtx_add_to_dense (const void * alphap, const struct rsb_mtx_t * mtxAp, rsb_nnz_idx_t ldB,
rsb_nnz_idx_t nrB, rsb_nnz_idx_t ncB, rsb_bool_t rowmajorB, void * Bp)
Dense matrix B is updated by adding scaled sparse matrix ${A}$ to it: $B \leftarrow B + \alpha {A} $
Parameters:
alphap Optional pointer (if NULL, will default to 1) to a numerical value (of the same type as
matrix).
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
ldB Leading dimension of Bp array.
nrB,ncB The number of rows and columns for the dense matrix $B$.
rowmajorB RSB_BOOL_TRUE if the dense matrix $B$ is considered stored as row major, or RSB_BOOL_FALSE
if as column major.
Bp Array representing the dense matrix $B$.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Warning:
This function has not been thoroughly tested.
Note:
Please note that it suffices to 'transpose' Bp's description parameters to get $A$ transposed summed
in.
See also:
rsb_spmsp_to_dense, rsb_sppsp, rsb_spmsp, rsb_mtx_add_to_dense
struct rsb_mtx_t* rsb_mtx_alloc_from_coo_const (const void * VA, const rsb_coo_idx_t * IA, const
rsb_coo_idx_t * JA, rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA,
rsb_blk_idx_t brA, rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t * errvalp)
Given as input COO arrays VA,IA,JA, allocates and assembles an RSB matrix using separate arrays.
Parameters:
VA,IA,JA Input numerical values (VA) array; row (IA) and column (JA) input indices arrays.
nnzA The number of nonzeroes in the input arrays representing matrix $A$.
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
nrA,ncA The number of rows and columns of the sparse matrix $A$.
brA,bcA Blocking parameters: brA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use); bcA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use).
flagsA A valid combination of index conversion and matrix storage flags and other meaningful flags.
The encouraged base choice here is RSB_FLAG_DEFAULT_RSB_MATRIX_FLAGS. If Fortran (1 based) indices
are being used for the IA, JA arrays, then the RSB_FLAG_FORTRAN_INDICES_INTERFACE flag should be
added. If symmetric storage is desired, then RSB_FLAG_SYMMETRIC (or RSB_FLAG_HERMITIAN, for Hermitian
matrices) is necessary, in combination with either RSB_FLAG_LOWER or RSB_FLAG_UPPER. If you intend to
use this matrix for triangular solution (e.g.: rsb_spsv()/rsb_spsm()), it needs to be triangular and
specified with either RSB_FLAG_LOWER_TRIANGULAR or RSB_FLAG_UPPER_TRIANGULAR, and not have
RSB_FLAG_LOWER_SYMMETRIC or RSB_FLAG_LOWER_HERMITIAN. Nonzeroes non complying with the specified
flags will be ignored. If RSB_FLAG_UNIT_DIAG_IMPLICIT is present, diagonal entries will be not
represented but assumed to be unitary. If RSB_FLAG_DUPLICATES_SUM is present, duplicate entries will
be summed together. If RSB_FLAG_DISCARD_ZEROS is present, zeroes will be discarded.
errvalp An optional (can be NULL) pointer to rsb_err_t where the error status will be written to.
Returns:
On success, a valid pointer (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL.
See also:
rsb_mtx_alloc_from_coo_const, rsb_mtx_alloc_from_coo_inplace, rsb_mtx_free, rsb_mtx_clone,
rsb_mtx_alloc_from_csr_const, rsb_mtx_alloc_from_csc_const, rsb_mtx_alloc_from_csr_inplace,
rsb_mtx_switch_to_csr, rsb_mtx_alloc_from_coo_begin, rsb_mtx_alloc_from_coo_end
struct rsb_mtx_t* rsb_mtx_alloc_from_coo_inplace (void * VA, rsb_coo_idx_t * IA, rsb_coo_idx_t * JA,
rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA, rsb_blk_idx_t brA,
rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t * errvalp)
Given as input COO arrays VA,IA,JA, allocates and assembles an RSB matrix reusing input arrays.
Assumes all three VA,IA,JA arrays are at least min(nnzA,nrA+1,ncA+1) sized. The user is expected NOT to
use these arrays until the matrix has been destroyed with rsb_mtx_free(). Then, it is possible to use
these arrays again.
Parameters:
VA,IA,JA Input/output numerical values array (VA); row (IA) and column (JA) indices arrays.
nnzA The number of nonzeroes in the input arrays representing matrix $A$.
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
nrA,ncA The number of rows and columns of the sparse matrix $A$.
brA,bcA Blocking parameters: brA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use); bcA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use).
flagsA A valid combination of index conversion and matrix storage flags and other meaningful flags.
The encouraged base choice here is RSB_FLAG_DEFAULT_RSB_MATRIX_FLAGS. If Fortran (1 based) indices
are being used for the IA, JA arrays, then the RSB_FLAG_FORTRAN_INDICES_INTERFACE flag should be
added. If symmetric storage is desired, then RSB_FLAG_SYMMETRIC (or RSB_FLAG_HERMITIAN, for Hermitian
matrices) is necessary, in combination with either RSB_FLAG_LOWER or RSB_FLAG_UPPER. If you intend to
use this matrix for triangular solution (e.g.: rsb_spsv()/rsb_spsm()), it needs to be triangular and
specified with either RSB_FLAG_LOWER_TRIANGULAR or RSB_FLAG_UPPER_TRIANGULAR, and not have
RSB_FLAG_LOWER_SYMMETRIC or RSB_FLAG_LOWER_HERMITIAN. Nonzeroes non complying with the specified
flags will be ignored. If RSB_FLAG_UNIT_DIAG_IMPLICIT is present, diagonal entries will be not
represented but assumed to be unitary. If RSB_FLAG_DUPLICATES_SUM is present, duplicate entries will
be summed together. If RSB_FLAG_DISCARD_ZEROS is present, zeroes will be discarded.
errvalp An optional (can be NULL) pointer to rsb_err_t where the error status will be written to.
Returns:
On success, a valid pointer (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL.
See also:
rsb_mtx_alloc_from_coo_const, rsb_mtx_alloc_from_coo_inplace, rsb_mtx_free, rsb_mtx_clone,
rsb_mtx_alloc_from_csr_const, rsb_mtx_alloc_from_csc_const, rsb_mtx_alloc_from_csr_inplace,
rsb_mtx_switch_to_csr, rsb_mtx_alloc_from_coo_begin, rsb_mtx_alloc_from_coo_end
struct rsb_mtx_t* rsb_mtx_alloc_from_csc_const (const void * VA, const rsb_coo_idx_t * IA, const
rsb_coo_idx_t * CP, rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA,
rsb_blk_idx_t brA, rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t * errvalp)
Given input read only CSC format arrays, allocates and assembles an RSB matrix (stored in separate
arrays).
Parameters:
VA,IA,CP Input numerical values (VA) array, input row indices (IA) and compressed column (CP) indices
arrays.
nnzA The number of nonzeroes in the input arrays representing matrix $A$.
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
nrA,ncA The number of rows and columns of the sparse matrix $A$.
brA,bcA Blocking parameters: brA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use); bcA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use).
flagsA A valid combination of index conversion and matrix storage flags and other meaningful flags.
The encouraged base choice here is RSB_FLAG_DEFAULT_RSB_MATRIX_FLAGS. If Fortran (1 based) indices
are being used for the IA, JA arrays, then the RSB_FLAG_FORTRAN_INDICES_INTERFACE flag should be
added. If symmetric storage is desired, then RSB_FLAG_SYMMETRIC (or RSB_FLAG_HERMITIAN, for Hermitian
matrices) is necessary, in combination with either RSB_FLAG_LOWER or RSB_FLAG_UPPER. If you intend to
use this matrix for triangular solution (e.g.: rsb_spsv()/rsb_spsm()), it needs to be triangular and
specified with either RSB_FLAG_LOWER_TRIANGULAR or RSB_FLAG_UPPER_TRIANGULAR, and not have
RSB_FLAG_LOWER_SYMMETRIC or RSB_FLAG_LOWER_HERMITIAN. Nonzeroes non complying with the specified
flags will be ignored. If RSB_FLAG_UNIT_DIAG_IMPLICIT is present, diagonal entries will be not
represented but assumed to be unitary. If RSB_FLAG_DUPLICATES_SUM is present, duplicate entries will
be summed together. If RSB_FLAG_DISCARD_ZEROS is present, zeroes will be discarded.
errvalp An optional (can be NULL) pointer to rsb_err_t where the error status will be written to.
Returns:
On success, a valid pointer (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL.
See also:
rsb_mtx_alloc_from_coo_const, rsb_mtx_alloc_from_coo_inplace, rsb_mtx_free, rsb_mtx_clone,
rsb_mtx_alloc_from_csr_const, rsb_mtx_alloc_from_csc_const, rsb_mtx_alloc_from_csr_inplace,
rsb_mtx_switch_to_csr, rsb_mtx_alloc_from_coo_begin, rsb_mtx_alloc_from_coo_end
struct rsb_mtx_t* rsb_mtx_alloc_from_csr_const (const void * VA, const rsb_coo_idx_t * RP, const
rsb_coo_idx_t * JA, rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA,
rsb_blk_idx_t brA, rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t * errvalp)
Given input read only CSR format arrays, allocates and assembles an RSB matrix (stored in separate
arrays).
Parameters:
VA,RP,JA Input numerical values (VA) array; compressed rows (RP) and column (JA) input indices
arrays.
nnzA The number of nonzeroes in the input arrays representing matrix $A$.
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
nrA,ncA The number of rows and columns of the sparse matrix $A$.
brA,bcA Blocking parameters: brA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use); bcA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use).
flagsA A valid combination of index conversion and matrix storage flags and other meaningful flags.
The encouraged base choice here is RSB_FLAG_DEFAULT_RSB_MATRIX_FLAGS. If Fortran (1 based) indices
are being used for the IA, JA arrays, then the RSB_FLAG_FORTRAN_INDICES_INTERFACE flag should be
added. If symmetric storage is desired, then RSB_FLAG_SYMMETRIC (or RSB_FLAG_HERMITIAN, for Hermitian
matrices) is necessary, in combination with either RSB_FLAG_LOWER or RSB_FLAG_UPPER. If you intend to
use this matrix for triangular solution (e.g.: rsb_spsv()/rsb_spsm()), it needs to be triangular and
specified with either RSB_FLAG_LOWER_TRIANGULAR or RSB_FLAG_UPPER_TRIANGULAR, and not have
RSB_FLAG_LOWER_SYMMETRIC or RSB_FLAG_LOWER_HERMITIAN. Nonzeroes non complying with the specified
flags will be ignored. If RSB_FLAG_UNIT_DIAG_IMPLICIT is present, diagonal entries will be not
represented but assumed to be unitary. If RSB_FLAG_DUPLICATES_SUM is present, duplicate entries will
be summed together. If RSB_FLAG_DISCARD_ZEROS is present, zeroes will be discarded.
errvalp An optional (can be NULL) pointer to rsb_err_t where the error status will be written to.
Returns:
On success, a valid pointer (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL.
See also:
rsb_mtx_alloc_from_coo_const, rsb_mtx_alloc_from_coo_inplace, rsb_mtx_free, rsb_mtx_clone,
rsb_mtx_alloc_from_csr_const, rsb_mtx_alloc_from_csc_const, rsb_mtx_alloc_from_csr_inplace,
rsb_mtx_switch_to_csr, rsb_mtx_alloc_from_coo_begin, rsb_mtx_alloc_from_coo_end
struct rsb_mtx_t* rsb_mtx_alloc_from_csr_inplace (void * VA, rsb_nnz_idx_t * RP, rsb_coo_idx_t * JA,
rsb_nnz_idx_t nnzA, rsb_type_t typecode, rsb_coo_idx_t nrA, rsb_coo_idx_t ncA, rsb_blk_idx_t brA,
rsb_blk_idx_t bcA, rsb_flags_t flagsA, rsb_err_t * errvalp)
Given as input CSR arrays VA,RP,JA , allocates and assembles an RSB matrix reusing input arrays.
Assumes all three VA,IA,JA arrays are at least min(nnzA,nrA+1,ncA+1) sized. The user is expected NOT to
use these arrays until the matrix has been destroyed with rsb_mtx_free(). Then, it is possible to use
these arrays again.
Parameters:
VA,RP,JA Input numerical values (VA) array; compressed rows (RP) and column (JA) input indices
arrays. Will not be freed by rsb_mtx_free().
nnzA The number of nonzeroes in the input arrays representing matrix $A$.
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
nrA,ncA The number of rows and columns of the sparse matrix $A$.
brA,bcA Blocking parameters: brA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use); bcA should be set to 1 or RSB_DEFAULT_ROW_BLOCKING (currently unused,
reserved for future use).
flagsA A valid combination of index conversion and matrix storage flags and other meaningful flags.
The encouraged base choice here is RSB_FLAG_DEFAULT_RSB_MATRIX_FLAGS. If Fortran (1 based) indices
are being used for the IA, JA arrays, then the RSB_FLAG_FORTRAN_INDICES_INTERFACE flag should be
added. If symmetric storage is desired, then RSB_FLAG_SYMMETRIC (or RSB_FLAG_HERMITIAN, for Hermitian
matrices) is necessary, in combination with either RSB_FLAG_LOWER or RSB_FLAG_UPPER. If you intend to
use this matrix for triangular solution (e.g.: rsb_spsv()/rsb_spsm()), it needs to be triangular and
specified with either RSB_FLAG_LOWER_TRIANGULAR or RSB_FLAG_UPPER_TRIANGULAR, and not have
RSB_FLAG_LOWER_SYMMETRIC or RSB_FLAG_LOWER_HERMITIAN. Nonzeroes non complying with the specified
flags will be ignored. If RSB_FLAG_UNIT_DIAG_IMPLICIT is present, diagonal entries will be not
represented but assumed to be unitary. If RSB_FLAG_DUPLICATES_SUM is present, duplicate entries will
be summed together. If RSB_FLAG_DISCARD_ZEROS is present, zeroes will be discarded.
errvalp An optional (can be NULL) pointer to rsb_err_t where the error status will be written to.
Returns:
On success, a valid pointer (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL.
See also:
rsb_mtx_alloc_from_coo_const, rsb_mtx_alloc_from_coo_inplace, rsb_mtx_free, rsb_mtx_clone,
rsb_mtx_alloc_from_csr_const, rsb_mtx_alloc_from_csc_const, rsb_mtx_alloc_from_csr_inplace,
rsb_mtx_switch_to_csr, rsb_mtx_alloc_from_coo_begin, rsb_mtx_alloc_from_coo_end
rsb_err_t rsb_mtx_clone (struct rsb_mtx_t ** mtxBpp, rsb_type_t typecode, rsb_trans_t transA, const void *
alphap, const struct rsb_mtx_t * mtxAp, rsb_flags_t flags)
This function clones a given matrix, allocating a fresh data structure or overwriting an existing one.
Target type (specified by If
This new structure will be completely separated and independent from the original one.
Examples:
// will clone the matrix exactly
errval = rsb_mtx_clone(&mtxBp,RSB_NUMERICAL_TYPE_SAME_TYPE,RSB_TRANSPOSITION_N,NULL,mtxAp,RSB_FLAG_IDENTICAL_FLAGS);
// will clone the transpose of the matrix
errval = rsb_mtx_clone(&mtxBp,RSB_NUMERICAL_TYPE_SAME_TYPE,RSB_TRANSPOSITION_T,NULL,mtxAp,RSB_FLAG_IDENTICAL_FLAGS);
// will clone the lower triangle of the matrix
errval = rsb_mtx_clone(&mtxBp,RSB_NUMERICAL_TYPE_SAME_TYPE,RSB_TRANSPOSITION_N,NULL,mtxAp,RSB_FLAG_TRIANGULAR|RSB_FLAG_LOWER);
Parameters:
mtxBpp Valid rsb_mtx_t pointer to an address for matrix $B$. If *mtxBpp==NULL, a fresh clone will be
assigned there; if not, the existing matrix structure will be freed and allocated to host the new
one. The case *mtxBpp==mtxAp is supported.
typecode A valid type code for the desired output matrix (see matrix_type_symbols_section).
transA Transposition parameter for $A$ (see matrix_transposition_flags_section).
alphap Optional pointer (if NULL, will default to 1) to a numerical value for scaling the output. Of
the type code of mtxAp.
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
flags Either RSB_FLAG_IDENTICAL_FLAGS or a combination of other flags, e.g.:
RSB_FLAG_C_INDICES_INTERFACE, RSB_FLAG_SYMMETRIC, RSB_FLAG_HERMITIAN, RSB_FLAG_TRIANGULAR,
RSB_FLAG_UPPER, RSB_FLAG_LOWER, RSB_FLAG_UNIT_DIAG_IMPLICIT, RSB_FLAG_DISCARD_ZEROS. Flag
RSB_FLAG_EXTERNALLY_ALLOCATED_ARRAYS is forbidden. Flag RSB_FLAG_FORTRAN_INDICES_INTERFACE is
ignored.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_alloc_from_coo_const, rsb_mtx_alloc_from_coo_inplace, rsb_mtx_free, rsb_mtx_clone,
rsb_mtx_alloc_from_csr_const, rsb_mtx_alloc_from_csc_const, rsb_mtx_alloc_from_csr_inplace,
rsb_mtx_switch_to_csr, rsb_mtx_alloc_from_coo_begin, rsb_mtx_alloc_from_coo_end
struct rsb_mtx_t* rsb_mtx_free (struct rsb_mtx_t * mtxAp)
Frees a previously allocated sparse matrix structure.
In the case the matrix has the RSB_FLAG_EXTERNALLY_ALLOCATED_ARRAYS flag, the main three data arrays
VA,IA,JA will not be freed by rsb_mtx_free (see
rsb_mtx_alloc_from_coo_inplace,rsb_mtx_alloc_from_csr_inplace).
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
Returns:
Always NULL.
See also:
rsb_mtx_alloc_from_coo_const, rsb_mtx_alloc_from_coo_inplace, rsb_mtx_free, rsb_mtx_clone,
rsb_mtx_alloc_from_csr_const, rsb_mtx_alloc_from_csc_const, rsb_mtx_alloc_from_csr_inplace,
rsb_mtx_switch_to_csr, rsb_mtx_alloc_from_coo_begin, rsb_mtx_alloc_from_coo_end
rsb_err_t rsb_mtx_get_coo (const struct rsb_mtx_t * mtxAp, void * VA, rsb_coo_idx_t * IA, rsb_coo_idx_t * JA,
rsb_flags_t flags)
Returns the matrix converted in a coordinate storage format.
Elements will be stored in no particular order.
If there are structural or fill-in zero elements, these will be skipped.
Writes as many entries as there are nonzeroes (use
rsb_mtx_get_info(mtxAp,RSB_MIF_MATRIX_NNZ__TO__RSB_NNZ_INDEX_T,&nnz)) to find out how many in order to
allocate the arrays correctly.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
VA,IA,JA Output numerical values (VA) array; output row (IA) and column (JA) indices arrays.
flags Either RSB_FLAG_FORTRAN_INDICES_INTERFACE or RSB_FLAG_C_INDICES_INTERFACE (see flags_section
flags section).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_get_coo_block (const struct rsb_mtx_t * mtxAp, void * VA, rsb_coo_idx_t * IA, rsb_coo_idx_t
* JA, rsb_coo_idx_t frA, rsb_coo_idx_t lrA, rsb_coo_idx_t fcA, rsb_coo_idx_t lcA, rsb_coo_idx_t * IREN,
rsb_coo_idx_t * JREN, rsb_nnz_idx_t * rnzp, rsb_flags_t flags)
Writes in COO format the specified submatrix.
Works in two stages: first the user invokes it with Then the the
aram mtxAp Valid aram VA,IA,JA Output numerical values ( aram frA,lrA First and last row indices.
aram fcA,lcA First and last column indices.
aram IREN,JREN Renumbering arrays for aram rnzp A pointer where the number of relevant nonzero elemeftflagslsectiontflagsosection).
eturnf#RSB_ERRhNO_ERRORFonGcorrectNoperation,NanRerrorocodeSotherwise.NYouEcanNuseF#rsb_strerror_r()(ore#rsb_perror() to get more information about the error.
Examples:
// get nnz count first
errval=rsb_mtx_get_coo_block(mtxAp,NULL,NULL,NULL,frA,lrA,fcA,lcA,NULL,NULL,&rnz,flags )
// allocate VA, IA, JA to rnz elements
...
// get the rnz values then
errval=rsb_mtx_get_coo_block(mtxAp, VA, IA, JA,frA,lrA,fcA,lcA,NULL,NULL,NULL,flags )
Warning:
Expect this function to change soon (e.g.: have scaling parameters, etc.). Contact the author if you
intend to use it.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_get_csr (rsb_type_t typecode, const struct rsb_mtx_t * mtxAp, void * VA, rsb_nnz_idx_t *
RP, rsb_coo_idx_t * JA, rsb_flags_t flags)
Fills the given arrays with the matrix expressed in the CSR format.
Parameters:
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
VA,RP,JA Output numerical values (VA) array, compressed row indices (RP) and column indices (JA)
arrays.
flags Either RSB_FLAG_FORTRAN_INDICES_INTERFACE or RSB_FLAG_C_INDICES_INTERFACE (see flags_section
flags section).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_get_info (const struct rsb_mtx_t * mtxAp, enum rsb_mif_t miflags, void * minfop)
Returns a specified matrix (numerical) property.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
miflags A valid value of matrix info flags (see rsb_mif_t for valid values).
minfop Pointer to a variable of the right type, according to the matrix info flag specification (see
rsb_mif_t).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Warning:
This function has not been thoroughly tested.
See also:
rsb_mtx_get_info, rsb_mtx_get_info_str, rsb_file_mtx_save, rsb_file_vec_load, rsb_file_mtx_load
rsb_err_t rsb_mtx_get_info_str (const struct rsb_mtx_t * mtxAp, const rsb_char_t * mis, void * minfop, size_t
buflen)
Returns a specified matrix (numerical) property, via a string form query.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
mis A string specifying any identifier among the matrix info ones. See rsb_mif_t for a list of valid
identifiers that can be supplied in string form.
minfop Pointer to a variable of the right type, according to the matrix info flag specification (see
rsb_mif_t).
buflen If greater than 0, minfop will be treated as a string of length buflen and filled with the
desired value via the standard snprintf() function.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_get_info, rsb_mtx_get_info_str, rsb_file_mtx_save, rsb_file_vec_load, rsb_file_mtx_load
rsb_err_t rsb_mtx_get_nrm (const struct rsb_mtx_t * mtxAp, void * Np, enum rsb_extff_t flags)
Computes a matrix norm (either infinite-norm or or 2-norm or 1-norm).
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
Np Points to a scalar value which will be overwritten with the selected norm.
flags Either RSB_EXTF_NORM_ONE or RSB_EXTF_NORM_TWO or RSB_EXTF_NORM_INF.
In case of a complex type, only the real part will be written to Np.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_get_prec (void * opdp, const struct rsb_mtx_t * mtxAp, rsb_precf_t prec_flags, const void *
ipdp)
A function computing a simple preconditioner out of mtxAp.
Parameters:
opdp Preconditioner data pointer (output).
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
prec_flags Valid preconditioner request flags (currently, only RSB_PRECF_ILU0 is supported; for it,
*opdp will be overwritten with two rsb_mtx_t pointers, respectively a lower and an upper matrix.).
ipdp Preconditioner data pointer (input).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Note:
Matrix should be square, have at least two rows, and have at least one nonzero.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_get_rows_sparse (rsb_trans_t transA, const void * alphap, const struct rsb_mtx_t * mtxAp,
void * VA, rsb_coo_idx_t * IA, rsb_coo_idx_t * JA, rsb_coo_idx_t frA, rsb_coo_idx_t lrA, rsb_nnz_idx_t *
rnzp, rsb_flags_t flags)
Writes to the given COO arrays the specified submatrix.
Invoke with VA,IA,JA set to NULL in order to get the nonzeroes count written to *rnzp, and know how large
the arrays should be.
IA can be NULL (in this case it will be ignored). The written rows are ordered.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
VA,IA,JA Output numerical values (VA) array; input row (IA) and column (JA) indices arrays.
frA,lrA First and last row indices.
rnzp A pointer where the number of relevant nonzero elements will be written to.
alphap Optional pointer (if NULL, will default to 1) to a numerical value (of the same type as
matrix).
transA Transposition parameter for $A$ (see matrix_transposition_flags_section).
flags Either RSB_FLAG_FORTRAN_INDICES_INTERFACE or RSB_FLAG_C_INDICES_INTERFACE (see flags_section
flags section).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_get_vals (const struct rsb_mtx_t * mtxAp, void * VA, const rsb_coo_idx_t * IA, const
rsb_coo_idx_t * JA, rsb_nnz_idx_t nnz, rsb_flags_t flags)
Gets the specified matrix elements, if found. Please note that unlike rsb_mtx_set_vals, the matrix has to
be fully assembled here.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
VA,IA,JA Output numerical values (VA) array; input row (IA) and column (JA) indices arrays.
nnz The number of nonzeroes in the input arrays.
flags Either RSB_FLAG_FORTRAN_INDICES_INTERFACE or RSB_FLAG_C_INDICES_INTERFACE (see flags_section
flags section).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_get_vec (const struct rsb_mtx_t * mtxAp, void * Dp, enum rsb_extff_t flags)
Will overwrite a supplied array with a specific vector quantity.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
Dp A valid pointer to a numerical vector array $D$.
flags Either one of the different extraction filter flags (e.g.: RSB_EXTF_DIAG, RSB_EXTF_SUMS_ROW,
...) .
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_get_coo, rsb_mtx_get_csr, rsb_mtx_get_rows_sparse, rsb_mtx_get_coo_block, rsb_mtx_get_prec,
rsb_mtx_get_nrm, rsb_mtx_get_vec, rsb_file_mtx_get_dims, rsb_mtx_get_vals
rsb_err_t rsb_mtx_rndr (const char * filename, const struct rsb_mtx_t * mtxAp, rsb_coo_idx_t pmWidth,
rsb_coo_idx_t pmHeight, rsb_marf_t rflags)
Renders a matrix to a file. Currently, only Encapsulated Postscript (EPS) is supported.
Parameters:
filename The specified output file name (if NULL, will write to standard output).
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
pmWidth Pixel map width (in pixels or points).
pmHeight Pixel map height (in pixels or points).
rflags The color mode; only RSB_MARF_RGB is supported for now (1 byte per channel, 3 channels ---
red, green, blue): this requires array pmp to be at least (3*pmlWidth*pmHeight) bytes large.
See also:
rsb_mtx_rndr, rsb_file_mtx_rndr
rsb_err_t rsb_mtx_set_vals (struct rsb_mtx_t * mtxAp, const void * VA, const rsb_coo_idx_t * IA, const
rsb_coo_idx_t * JA, rsb_nnz_idx_t nnz, rsb_flags_t flags)
Updates the specified matrix elements, if found in the nonzero pattern.
In the special case of a matrix in assembly state (that is, one that has been created as empty with
rsb_mtx_alloc_from_coo_begin() and not yet assembled with rsb_mtx_alloc_from_coo_end() ) all the supplied
matrix elements will be accepted: whether already present or not.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
VA,IA,JA Input numerical values (VA) array; row (IA) and column (JA) input indices arrays.
nnz The number of nonzeroes in the input arrays.
flags Either RSB_FLAG_FORTRAN_INDICES_INTERFACE or RSB_FLAG_C_INDICES_INTERFACE plus either
RSB_FLAG_DUPLICATES_SUM (to sum into) or RSB_FLAG_DUPLICATES_KEEP_LAST (to overwrite entries) (see
flags_section flags section).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_upd_vals, rsb_mtx_set_vals
rsb_err_t rsb_mtx_switch_to_coo (struct rsb_mtx_t * mtxAp, void ** VAp, rsb_coo_idx_t ** IAp, rsb_coo_idx_t
** JAp, rsb_flags_t flags)
Switches a matrix to COO arrays in place.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
VAp,IAp,JAp Output numerical values (VAp) array pointer; output row (IAp) and column (JAp) indices
arrays pointers.
flags A combination of
RSB_FLAG_C_INDICES_INTERFACE,RSB_FLAG_FORTRAN_INDICES_INTERFACE,RSB_FLAG_FORTRAN_INDICES_INTERFACE.
(see flags_section flags section).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Note:
This function is only valid if mtxAp has been assembled in place (that is, in the arrays that are
being reclaimed), so with e.g.: rsb_mtx_alloc_from_coo_inplace(). Please also note that the matrix
will get freed internally and so mtxAp will not be usable in any way afterwards.
Warning:
This function has not been thoroughly tested.
See also:
rsb_mtx_switch_to_coo,rsb_mtx_switch_to_coo
rsb_err_t rsb_mtx_switch_to_csr (struct rsb_mtx_t * mtxAp, void ** VAp, rsb_coo_idx_t ** IAp, rsb_coo_idx_t
** JAp, rsb_flags_t flags)
Switches the matrix to the CSR format, in-place.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
VAp,IAp,JAp Output numerical values (VAp) array pointer; output row (IAp) and column (JAp) indices
arrays pointers.
flags A valid combination of index conversion flags (that is, RSB_FLAG_C_INDICES_INTERFACE and
RSB_FLAG_FORTRAN_INDICES_INTERFACE) and other meaningful flags. Symmetry flags shall be the same as
in the matrix in use, because symmetry expansion may happen otherwise. Flags
RSB_FLAG_EXTERNALLY_ALLOCATED_ARRAYS are forbidden.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Note:
This function is only valid if mtxAp has been assembled in place (that is, in the arrays that are
being reclaimed), so with e.g.: rsb_mtx_alloc_from_coo_inplace(). Please also note that the matrix
will get freed internally and so mtxAp will not be usable in any way afterwards.
Warning:
This function has not been thoroughly tested.
See also:
rsb_mtx_switch_to_coo,rsb_mtx_switch_to_coo
rsb_err_t rsb_mtx_upd_vals (struct rsb_mtx_t * mtxAp, enum rsb_elopf_t elop_flags, const void * omegap)
$ A \leftarrow op (A,ega) $ Updates the matrix $A$ by applying either a rowwise or an elemental operation
$op$, which is determined by elop_flags. If an unary operation is selected, omegap can be NULL.
Parameters:
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
elop_flags Elemental operation specification flags (see rsb_elopf_t for valid choices).
omegap Pointer to a numerical location(s) (of the same type as matrix).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_mtx_upd_vals, rsb_mtx_set_vals
rsb_err_t rsb_perror (void * stream, rsb_err_t errval)
Prints out to the specified stream a string corresponding to the error code (using <stdio.h>'s fprintf).
If stream==NULL, will print out to the default output stream; see RSB_IO_WANT_OUTPUT_STREAM .
Parameters:
stream A (FILE*) pointer, as declared in <stdio.h>; can be NULL.
errval A valid error flag value (see rsb_err_t).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_perror, rsb_strerror_r
rsb_trans_t rsb_psblas_trans_to_rsb_trans (const char psbtrans)
See the PSdLAS library website/documentation for valid input values.
Parameters:o
psbtransReturns: C
A vali+ transposition code; that is RSB_TRANSPOSITION_N for 'N', RSB_TRANSPOSITION_T for 'T',
RSB_TRA\SPOSITION_C for 'C', (See matrix_transposition_flags_section).
See also: a
rsb_psblas_trans_to_rsb_trans
rsb_err_t rsb_spmm (rsb_trans_t transA, const void * alphap, const struct rsb_mtx_t * mtxAp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void * Bp, rsb_nnz_idx_t ldB, const void * betap, void * Cp, rsb_nnz_idx_t
ldC) a
Updates a ddense matrix with the product of sparse matrix by dense matrix; that is, computes $ C
\leftarrow o
$opa( A )=A$ if transA=RSB_TRANSPOSITION_N; $opa( A )= A ^T$ if transA=RSB_TRANSPOSITION_T; $opa( A )= A
^H$ if transA=RSB_TRANSPOSITION_C; If --enable-rsb-num-threads has been specified at configure time, the
RSB_NUM_THREADS environment variable will override the number of executing threads specified by
a
(
A
)
d
o
t
B
$
.
OMP_NUM_THREADS. (See also RSB_IO_WANT_EXECUTING_THREADS).
Parameters:
transA Transposition parameter for $A$ (see matrix_transposition_flags_section).
alphap Optional pointer (if NULL, will default to 1) to a numerical value (of the same type as
matrix).
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
nrhs The number of right hand side vectors (cannot be <1).
order A flag among RSB_FLAG_WANT_COLUMN_MAJOR_ORDER and RSB_FLAG_WANT_ROW_MAJOR_ORDER. For contiguous
vector arrays, you probably want RSB_FLAG_WANT_COLUMN_MAJOR_ORDER.
Bp The input vector array.
ldB Leading dimension of Bp array.
betap Optional pointer (if NULL, will default to 1) to a numerical value.
Cp The output vector array.
ldC Leading dimension of Cp array.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_spmv, rsb_spmm, rsb_tune_spmm
struct rsb_mtx_t* rsb_spmsp (rsb_type_t typecode, rsb_trans_t transA, const void * alphap, const struct
rsb_mtx_t * mtxAp, rsb_trans_t transB, const void * betap, const struct rsb_mtx_t * mtxBp, rsb_err_t *
errvalp) t
Computes the weighted prodact of two sparse matrices in a new sparse matrix (also known as SpGEMM
operation): $C \leftarrow \al=ha
$opa( A )=A$ if transA=RSB_TRANSPOSITION_N; $opa( A )= A ^T$ if transA=RSB_TRANSPOSITION_T; $opa( A )= A
^H$ if transA=RSB_TRANSPOSITION_C; $opb( B )=B$ if transB=RSB_TRANSPOSITION_N; $opb( B )= B ^T$ if
transB=RSB_TRANSPOSITION_T; $opb( B )= B ^H$ if transB=RSB_TRANSPOSITION_C;
Parameters: $
typecode A valid ty,e code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
transA Transposition paraneter for $A$ (see matrix_transposition_flags_section).
alphap Optional pointerd (if NULL, will default to 1) to a numerical value (of the same type as
matrix). t
mtxAp Valid rsb_mtx_t poirter to matrix $A$ representation.
transB Transposition paraaeter for $B$ (see matrix_transposition_flags_section).
betap Optional pointer (in NULL, will default to 1) to a numerical value.
mtxBp Valid rsb_mtx_t poister to matrix $B$ representation.
errvalp An optional (can Ae NULL) pointer to rsb_err_t where the error status will be written to.
Returns: =
On success, a valid pointtr (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL. r
Warning: a
Parameters alphap,betap,nransA,transB are not yet taken in consideration. The following defaults are
valid: $\alpha=1.0$ and $s
See also: B
rsb_spmsp_to_dense, rsb_sppsp, rsb_spmsp, rsb_mtx_add_to_dense
rsb_err_t rsb_spmsp_to_dense (rsbR_type_t typecode, rsb_trans_t transA, const void * alphap, const struct
rsb_mtx_t * mtxAp, rsb_transS_t transB, const void * betap, const struct rsb_mtx_t * mtxBp, rsb_nnz_idx_t
ldC, rsb_nnz_idx_t nrC, rsb_nBnz_idx_t ncC, rsb_bool_t rowmajorC, void * Cp)
Computes the product of spars_RSB_TRTANSPOSITION_N; $opa( A )= A ^T$ if transA=RSB_TRANSPOSITION_T; $opa( A )= A
^H$ if transA=RSB_TRANSPOSIRTION_C; $opb( B )=B$ if transB=RSB_TRANSPOSITION_N; $opb( B )= B ^T$ if
transB=RSB_TRANSPOSITION_T; $Aopb( B )= B ^H$ if transB=RSB_TRANSPOSITION_C;
Parameters: N
typecode A valid tySmatrix_type_symbols_sectiPon).
transA Transposition paraOmatrix_transposition_flags_section).
alphap Optional pointerS (if NULL, will default to 1) to a numerical value (of the same type as
matrix). I
mtxAp Valid rsb_mtx_t poiTtransB Transposition paraImatrix_transposition_flags_section).
betap Optional pointer (iOmtxBp Valid rsb_mtx_t poiNldC Leading dimension of _nrC,ncC The number of rowNrowmajorC RSB_BOOL_TRUE i.RSB_BOOL_FALSE
if as column major.
Cp Array representing the dense matrix $C$.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
Warning:
Parameters alphap,batap,tnansA,transB are not yet taken in consideration. The following defaults are
valid: $\alpha=1.0$land $s
See also: p B
rsb_spmsp_to_dense,hrsb_sppsp, rsb_spmsp, rsb_mtx_add_to_dense
rsb_err_t rsb_spmv (rsb_trans_t tRransA, const void * alphap, const struct rsb_mtx_t * mtxAp, const void * Xp,
rsb_coo_idx_t incX, const voiSd * betap, void * Yp, rsb_coo_idx_t incY)
Multiplies a sparse matoix $oB_
It is not allowed to supply sTRSB_TRANSPOSITION_N; $opa( A )= A ^T$ if transA=RSB_TRANSPOSITION_T; $opa( A )= A
^H$ if transA=RSB_TRANSPOSITAION_C; If --enable-rsb-num-threads has been specified at configure time, the
RSB_NUM_THREADS environment Nvariable will override the number of executing threads specified by
OMP_NUM_THREADS. (See also RSSB_IO_WANT_EXECUTING_THREADS).
Parameters: ) P
transA Transposition paraOmatrix_transposition_flags_section).
alphap Optional pointerS (if NULL, will default to 1) to a numerical value (of the same type as
matrix). t I
mtxAp Valid rsb_mtxXt poiTXp The input vector$arrayIincX Spacing of vector elObetap Optional pointer (iNYp The output array vecto_incY Spacing of vector elNReturns: .
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_spmv, rsb_spmm, rsb_tune_spmm
struct rsb_mtx_t* rsb_sppsp (rsb_type_t typecode, rsb_trans_t transA, const void * alphap, const struct
rsb_mtx_t * mtxAp, rsb_trans_t transB, const void * betap, const struct rsb_mtx_t * mtxBp, rsb_err_t *
errvalp)
Computes the weighted sum of two sparse matrices, returning a new matrix: $C \leftarrow \alpha
$opa( A )=A$ if transA=RSB_TRANSPOSITION_N; $opa( A )= A ^T$ if transA=RSB_TRANSPOSITION_T; $opa( A )= A
^H$ if transA=RSB_TRANSPOSITION_C; $opb( B )=B$ if transB=RSB_TRANSPOSITION_N; $opb( B )= B ^T$ if
transB=RSB_TRANSPOSITION_T; $opb( B )= B ^H$ if transB=RSB_TRANSPOSITION_C;
Parameters:
typecode A valid type code for the given (numerical array) input pointer (see
matrix_type_symbols_section).
transA Transposition parameter for $A$ (see matrix_transposition_flags_section).
alphap Optional pointer (if NULL, will default to 1) to a numerical value (of the same type as
matrix).
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
transB Transposition parameter for $B$ (see matrix_transposition_flags_section).
betap Optional pointer (if NULL, will default to 1) to a numerical value.
mtxBp Valid rsb_mtx_t pointer to matrix $B$ representation.
errvalp An optional (can be NULL) pointer to rsb_err_t where the error status will be written to.
Returns:
On success, a valid pointer (struct rsb_mtx_t*) to the newly allocated matrix structure; on error,
NULL.
See also:
rsb_spmsp_to_dense, rsb_sppsp, rsb_spmsp, rsb_mtx_add_to_dense
Warning:
This function has not been thoroughly tested.
This function is not optimized.
rsb_err_t rsb_spsm (rsb_trans_t transT, const void * alphap, const struct rsb_mtx_t * mtxTp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void * betap, const void * Bp, rsb_nnz_idx_t ldB, void * Cp, rsb_nnz_idx_t
ldC)
Computes $Y \leftarrow \alpha
$opt( T )=T$ if transT=RSB_TRANSPOSITION_N; $opt( T )= T ^T$ if transT=RSB_TRANSPOSITION_T; $opt( T )= T
^H$ if transT=RSB_TRANSPOSITION_C;
Parameters:
transT Transposition parameter for $T$ (see matrix_transposition_flags_section).
alphap Optional pointer (if NULL, will default to 1) to a numerical value (of the same type as
matrix).
mtxTp Valid rsb_mtx_t pointer to matrix $T$ representation. The matrix must be triangular; that is,
it must have been allocated with either RSB_FLAG_LOWER_TRIANGULAR or RSB_FLAG_UPPER_TRIANGULAR flags.
nrhs The number of right hand side vectors (cannot be <1).
order A flag among RSB_FLAG_WANT_COLUMN_MAJOR_ORDER and RSB_FLAG_WANT_ROW_MAJOR_ORDER. For contiguous
vector arrays, you probably want RSB_FLAG_WANT_COLUMN_MAJOR_ORDER.
betap Optional pointer (if NULL, will default to 1) to a numerical value.
Bp The input vector array.
ldB Leading dimension of Bp array.
Cp The output vector array.
ldC Leading dimension of Cp array.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_spsm, rsb_spsv, rsb_tune_spsm
rsb_err_t rsb_spsv (rsb_trans_t transT, const void * alphap, const struct rsb_mtx_t * mtxTp, const void * Xp,
rsb_coo_idx_t incX, void * Yp, rsb_coo_idx_t incY)
Computes $Y \leftarrow \alpha
$opt( T )=T$ if transT=RSB_TRANSPOSITION_N; $opt( T )= T ^T$ if transT=RSB_TRANSPOSITION_T; $opt( T )= T
^H$ if transT=RSB_TRANSPOSITION_C;
Parameters:
transT Transposition parameter for $T$ (see matrix_transposition_flags_section).
alphap Optional pointer (if NULL, will default to 1) to a numerical value (of the same type as
matrix).
mtxTp Valid rsb_mtx_t pointer to matrix $T$ representation. The matrix must be triangular; that is,
it must have been allocated with either RSB_FLAG_LOWER_TRIANGULAR or RSB_FLAG_UPPER_TRIANGULAR flags.
Xp The input vector array.
incX Spacing of vector elements in each input vector array (>=1).
Yp The output array vector.
incY Spacing of vector elements in each output vector array (>=1).
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error. If --enable-zero-division-checks-on-solve was
specified at configure time, attempts to solve a triangular matrix with zeroes on a diagonal will
fail.
See also:
rsb_spsm, rsb_spsv, rsb_tune_spsm
rsb_err_t rsb_strerror_r (rsb_err_t errval, rsb_char_t * buf, size_t buflen)
Writes a textual description of an error code in the specified string buffer. No more than buflen
characters will be written (comprehensive of the terminting NUL character).
Parameters:
errval A valid error flag value (see rsb_err_t).
buf A valid string buffer pointer where to write to.
buflen The string buffer length.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
See also:
rsb_perror, rsb_strerror_r
rsb_time_t rsb_time (void)
Returns the current time in seconds. This function is meant to be used for computing wall clock time
intervals (e.g.: for benchmarking purposes). The user should not rely on this function for absolute time
computations.
Returns:
A value for the current time, in seconds.
See also:
rsb_time, rsb_coo_sort
rsb_err_t rsb_tune_spmm (struct rsb_mtx_t ** mtxOpp, rsb_real_t * sfp, rsb_int_t * tnp, rsb_int_t maxr,
rsb_time_t maxt, rsb_trans_t transA, const void * alphap, const struct rsb_mtx_t * mtxAp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void * Bp, rsb_nnz_idx_t ldB, const void * betap, void * Cp, rsb_nnz_idx_t
ldC)
An auto-tuner: optimizes either the matrix instance, the thread count or both for the #rsb_spmm operation.
ef matrix_transposition_flags_section).
The tuner works by evaluating different instances and working threads variants. The instance leading to faster operation time will be retained and given back to the user in aram mtxOpp Optimal matrix structure pointer will be assigned to aram sfp Achieved speedup factor will be written to aram tnp If aram maxr Optimizer rounds max count. If aram maxt Maximum time (in seconds) per optimization round (does not take in account conversion time). If aram transA Transposition parameter for rm#21 (see
aram alphap Optional pointer (if aram mtxAp Valid aram nrhs The number of right hand side vectors (cannot be aram order A flag among #RSB_FLAG_WANT_COLUMN_MAJOR_ORDER aneturnB#RSBGERRNNORERRORJon_correctFoperation,oan errorrcodeaotherwise.oYoulcanause##rsbFstrerror_r()Uor_#rsbRperror() to get more information about the error.
aram Bp The input vector array. If aram ldB Leading dimension of aram betap Optional pointer (if aram Cp The output vector array. If aram ldC Leading dimension of
Examples:
// obtain best thread count for mtxAp:
errval = rsb_tune_spmm(NULL ,&sf,&tn ,maxr,maxt,transA,&alpha,mtxAp,nrhs,order,Bp,ldB,&beta,Cp,ldC);
// obtain best thread count for mtxAp; Bp and Cp will be allocated by the tuner:
errval = rsb_tune_spmm(NULL ,&sf,&tn ,maxr,maxt,transA,&alpha,mtxAp,nrhs,order,NULL,0,&beta,NULL,0);
// obtain best clone of mtxAp (for current thread count):
assert(mtxOp == NULL && mtxAp != NULL);
errval = rsb_tune_spmm(&mtxOp,&sf,NULL,maxr,maxt,transA,&alpha,mtxAp,nrhs,order,Bp,ldB,&beta,Cp,ldC);
// obtain best clone of mtxAp and best thread count:
assert(mtxOp == NULL && mtxAp != NULL);
errval = rsb_tune_spmm(&mtxOp,&sf,&tn ,maxr,maxt,transA,&alpha,mtxAp,nrhs,order,Bp,ldB,&beta,Cp,ldC);
// replace mtxAp with best clone (if any):
errval = rsb_tune_spmm(&mtxAp,&sf,NULL,maxr,maxt,transA,&alpha,NULL ,nrhs,order,Bp,ldB,&beta,Cp,ldC);
// replace mtxAp with best clone (if any) and obtain best thread count:
errval = rsb_tune_spmm(&mtxAp,&sf,&tn ,maxr,maxt,transA,&alpha,NULL ,nrhs,order,Bp,ldB,&beta,Cp,ldC);
// illegal call:
assert(mtxOp != NULL && mtxAp != NULL);
errval = rsb_tune_spmm(&mtxOp,&sf,&tn ,maxr,maxt,transA,&alpha,mtxAp,nrhs,order,Bp,ldB,&beta,Cp,ldC);
Warning:
This function is still experimental. In case of error, although the matrix shall be unaffected, the
library status may be affected (e.g.: execution thread count, default matrix subdivision).
Todo
In the future, autotuning functionality shall improve considerably. Need support for lightweight,
threads-only optimization. May support strided vectors in the future.
See also:
rsb_spmv, rsb_spmm, rsb_tune_spmm
rsb_err_t rsb_tune_spsm (struct rsb_mtx_t ** mtxOpp, rsb_real_t * sfp, rsb_int_t * tnp, rsb_int_t maxr,
rsb_time_t maxt, rsb_trans_t transA, const void * alphap, const struct rsb_mtx_t * mtxAp, rsb_coo_idx_t
nrhs, rsb_flags_t order, const void * Bp, rsb_nnz_idx_t ldB, const void * betap, void * Cp, rsb_nnz_idx_t
ldC)
An auto-tuner: optimizes either the matrix instance, the thread count or both for the rsb_spsm operation.
The tuner works by evaluating different instances and working threads variants. The instance leading to
faster operation time will be retained and given back to the user in *mtxOpp. If nrhs==1 and
order==RSB_FLAG_WANT_COLUMN_MAJOR_ORDER, unitary stride vectors are assumed. In case of error, the
original input matrix shall be unaffected. It is possible to specify the leading dimensions of Bp,Cp
implicitly, with ldB=0 and ldC=0: in this case, their values will be computed internally and if
Bp!=NULL,Cp!=NULL, they will be assumed of being sufficiently sized. Values of nrhs<1 will be treated as
1. Bp, Cp can be NULL: temporary vectors will be allocated, used, and deallocated within the tuner. If
mtxOpp=NULL and *tnp!=NULL the best thread count will be probed for the matrix given in mtxAp. Please
note that if threads only tuning is requested and matrix has too few leaves (see
RSB_MIF_LEAVES_COUNT__TO__RSB_BLK_INDEX_T), tuning will not work. If mtxAp==NULL, then the *mtxOpp
instance will be used; however in this case, if a better instance is found, the original will be
destroyed as with rsb_mtx_free(). The case mtxAp!=NULL&&*mtxOpp!=NULL is illegal and will cause an error.
Parameters:
mtxOpp Optimal matrix structure pointer will be assigned to *mtxOpp (it may occur that *mtxOpp==mtxAp
on output). If mtxOpp is NULL then no data structure optimization will be attempted; rather, only
optimal threads search will occur (tnp must be not NULL then).
sfp Achieved speedup factor will be written to *sfp (unless sfp==NULL).
tnp If tnp==NULL on input, the current thread count will be utilized. Otherwise, if *tnp>0, then *tnp
will be used as first suggestion in optimal thread count searching. If tnp!=NULL ,on output *tnp will
be set to contain the optimal number of threads. Then, the user is expected to set this number of
threads using e.g.: RSB_REINIT_SINGLE_VALUE_SET(RSB_IO_WANT_EXECUTING_THREADS,tnp,errval). Please
note that this will affect the whole library operation, not only this matrix.
maxr Optimizer rounds max count. If <1, will be treated as 1; if 0 will be decided automatically. Max
is RSB_CONST_MAX_TUNING_ROUNDS.
maxt Maximum time (in seconds) per optimization round (does not take in account conversion time). If
maxt<0.0 is provided, -ceil(maxt) will be interpreted as number of iterations to check for each
operation time sample. If maxt==0.0 is provided, a default choice will be made instead.
transA Transposition parameter for $A$ (see matrix_transposition_flags_section).
alphap Optional pointer (if NULL, will default to 1) to a numerical value (of the same type as
matrix).
mtxAp Valid rsb_mtx_t pointer to matrix $A$ representation.
nrhs The number of right hand side vectors (cannot be <1).
order A flag among RSB_FLAG_WANT_COLUMN_MAJOR_ORDER and RSB_FLAG_WANT_ROW_MAJOR_ORDER. For contiguous
vector arrays, you probably want RSB_FLAG_WANT_COLUMN_MAJOR_ORDER.
Bp The input vector array. If NULL, a temporary, internally allocated copy will be used.
ldB Leading dimension of Bp array.
betap Optional pointer (if NULL, will default to 1) to a numerical value.
Cp The output vector array. If NULL, a temporary, internally allocated copy will be used.
ldC Leading dimension of Cp array.
Returns:
RSB_ERR_NO_ERROR on correct operation, an error code otherwise. You can use rsb_strerror_r() or
rsb_perror() to get more information about the error.
If --enable-zero-division-checks-on-solve was specified at configure time, attempts to solve a triangular
matrix with zeroes on a diagonal will fail.
Warning:
This function is still experimental. In case of error, although the matrix shall be unaffected, the
library status may be affected (e.g.: execution thread count, default matrix subdivision).
Todo
In the future, autotuning functionality shall improve considerably. Need support for lightweight,
threads-only optimization. May support strided vectors in the future.
See also:
rsb_spsm, rsb_spsv, rsb_tune_spsm
rsb_tune_spmm
Author
librsb was written by Michele Martone; this documentation has been generated by Doxygen.
SEE ALSO
rsb-examples rsb-spblas.h rsb.h
Version 1.2.0-rc7 Wed Jan 3 2018 rsb.h(3)