LUSolveFlops(3P) Sun Performance Library LUSolveFlops(3P)NAMELUSolveFlops - A query function that returns the floating point opera‐
tion count of the solve step of the SuperLU solver.
SYNOPSIS
#include <sunperf.h>
float LUSolveFlops(SuperLUStat_t *stat);
float LUSolveFlops_64(SuperLUStat_t_64 *stat);
PURPOSELUSolveFlops returns the floating point operation count in the solve
step of the SuperLU solver. This information is collected and stored
in a structure object of type SuperLUStat_t:
typedef struct {
int *panel_histo;
double *utime;
float *ops;
int TinyPivots;
int RefineSteps;
} SuperLUStat_t;
typedef struct {
long *panel_histo;
double *utime;
float *ops;
long TinyPivots;
long RefineSteps;
} SuperLUStat_t_64;
where
panel_histo is histogram of panel size distribution
utime is running time at various phases
ops is operation count at various phases
TinyPivots is number of tiny pivots
RefineSteps is number of iterative refinement steps
LUSolveFlops returns the single precision floating point value in array ops
that represents the number of floating point operations performed in the solvestep.ARGUMENTS
SuperLUStat_t *stat (input)
On entry, stat points to a SuperLUStat_t structure.
COPYRIGHT
Copyright (c) 2003, The Regents of the University of California,
through Lawrence Berkeley National Laboratory (subject to receipt of
any required approvals from U.S. Dept. of Energy)
SEE ALSO
StatInit
StatFree
LUFactTime
LUSolveTime
LUFactFlops
SuperMatrix
http://crd.lbl.gov/~xiaoye/SuperLU/
James W. Demmel, Stanley C. Eisenstat, John R. Gilbert, Xiaoye S. Li
and Joseph W. H. Liu, "A supernodal approach to sparse partial pivot‐
ing", SIAM J. Matrix Analysis and Applications, Vol. 20, Num. 3, 1999,
pp. 720-755.
6 Mar 2009 LUSolveFlops(3P)
