util::MProbe< T, SubMa, Ma > Class Template Reference

M-way partitioning algorithm feasibility check. The check is done on a set of arrays as described in section 3.2.2 of "E. Saule, E. O. Bas, U. V. Catalyurek, Load-Balancing Spatially Located Computations using Rectangular Partitions Paper, Tech. Report, ArXiv, no. arXiv:1104.2566v1, 2011" More...

#include <util/m_probe.hpp>

---

Detailed Description

template<typename T, typename SubMa, typename Ma>
class util::MProbe< T, SubMa, Ma >

M-way partitioning algorithm feasibility check. The check is done on a set of arrays as described in section 3.2.2 of "E. Saule, E. O. Bas, U. V. Catalyurek, Load-Balancing Spatially Located Computations using Rectangular Partitions Paper, Tech. Report, ArXiv, no. arXiv:1104.2566v1, 2011"

List of all members.

Static Public Member Functions
static bool	probe (T b, const Ma &arrs, int m, int **lb, int **ub, int *procs, int **cuts)
Static Private Member Functions
static int	rprobe (const SubMa &wpre, int length, const T &bottleneck, int numproc, int *s, int *sl, int *sh)

---

Member Function Documentation

template<typename T, typename SubMa, typename Ma>

static bool util::MProbe< T, SubMa, Ma >::probe	(	T	b,
		const Ma &	arrs,
		int	m,
		int **	lb,
		int **	ub,
		int *	procs,
		int **	cuts
	)			[inline, static]

Given a bottleneck, processor count and multiple arrays, returns feasiblity (yes/no) and cut points if feasible

Parameters:

	b	Bottleneck value to be tested
	arrs	A data structure that holds multiple prefix sum arrays
	m	the number of processors to be tested
	lb	lower bound array
	ub	upper bound array
	procs	the resulting processor count per array
	cuts	the resulting cut points per array. Cut points are described nicol+

Returns:

*procs and **cuts are assigned. True or false returned depending on feasibility

template<typename T, typename SubMa, typename Ma>

static int util::MProbe< T, SubMa, Ma >::rprobe	(	const SubMa &	wpre,
		int	length,
		const T &	bottleneck,
		int	numproc,
		int *	s,
		int *	sl,
		int *	sh
	)			[inline, static, private]

This funtion is very similar to rprobe of the nicol+. The only difference is that this one returns processor number if bottleneck b is feasible

Parameters:

	wpre	a 1d prefixsum array
	length	the length of the 1d prefix sum array
	bottleneck	bottleneck value to be tesed
	numproc	the number of processors to be tried
	s	cut points
	sl	cut point lower bounds per processor
	sh	cut point upper bounds per processor

Returns:

p if it is possible to partition an array with a given bottleneck B. retuns -1 otherwise. *s includes cut points if feasible

---

The documentation for this class was generated from the following file:

• include/util/m_probe.hpp