.TH "hwlocality_helper_distribute" 3 "Version 2.14.0" "Hardware Locality (hwloc)" \" -*- nroff -*-
.ad l
.nh
.SH NAME
hwlocality_helper_distribute \- Distributing items over a topology
.SH SYNOPSIS
.br
.PP
.SS "Enumerations"

.in +1c
.ti -1c
.RI "enum \fBhwloc_distrib_flags_e\fP { \fBHWLOC_DISTRIB_FLAG_REVERSE\fP }"
.br
.in -1c
.SS "Functions"

.in +1c
.ti -1c
.RI "int \fBhwloc_distrib\fP (\fBhwloc_topology_t\fP topology, \fBhwloc_obj_t\fP *roots, unsigned n_roots, \fBhwloc_cpuset_t\fP *set, unsigned n, int until, unsigned long flags)"
.br
.in -1c
.SH "Detailed Description"
.PP 

.SH "Enumeration Type Documentation"
.PP 
.SS "enum \fBhwloc_distrib_flags_e\fP"

.PP
Flags to be given to \fBhwloc_distrib()\fP\&. 
.PP
\fBEnumerator\fP
.in +1c
.TP
\f(BIHWLOC_DISTRIB_FLAG_REVERSE \fP
Distrib in reverse order, starting from the last objects\&. 
.SH "Function Documentation"
.PP 
.SS "int hwloc_distrib (\fBhwloc_topology_t\fP topology, \fBhwloc_obj_t\fP * roots, unsigned n_roots, \fBhwloc_cpuset_t\fP * set, unsigned n, int until, unsigned long flags)\fR [inline]\fP"

.PP
Distribute \fRn\fP items over the topology under \fRroots\fP\&. Array \fRset\fP will be filled with \fRn\fP cpusets recursively distributed linearly over the topology under objects \fRroots\fP, down to depth \fRuntil\fP (which can be INT_MAX to distribute down to the finest level)\&.

.PP
\fRn_roots\fP is usually 1 and \fRroots\fP only contains the topology root object so as to distribute over the entire topology\&.

.PP
This is typically useful when an application wants to distribute \fRn\fP threads over a machine, giving each of them as much private cache as possible and keeping them locally in number order\&.

.PP
The caller may typically want to also call \fBhwloc_bitmap_singlify()\fP before binding a thread so that it does not move at all\&.

.PP
\fRflags\fP should be 0 or a OR'ed set of \fBhwloc_distrib_flags_e\fP\&.

.PP
\fBReturns\fP
.RS 4
0 on success, -1 on error\&.
.RE
.PP
\fBNote\fP
.RS 4
On hybrid CPUs (or asymmetric platforms), distribution may be suboptimal since the number of cores or PUs inside packages or below caches may vary (the top-down recursive partitioning ignores these numbers until reaching their levels)\&. Hence it is recommended to distribute only inside a single homogeneous domain\&. For instance on a CPU with energy-efficient E-cores and high-performance P-cores, one should distribute separately N tasks on E-cores and M tasks on P-cores instead of trying to distribute directly M+N tasks on the entire CPUs\&.

.PP
This function requires the \fRroots\fP objects to have a CPU set\&. 
.RE
.PP

.SH "Author"
.PP 
Generated automatically by Doxygen for Hardware Locality (hwloc) from the source code\&.