Poster of Linux kernelThe best gift for a Linux geek
LIBPFM

LIBPFM

Section: Linux Programmer's Manual (3) Updated: November, 2003
Local index Up
 

NAME

libpfm - a helper library to program the hardware performance counters of IA-64 CPUs  

SYNOPSIS

#include <perfmon/pfmlib.h>
 

DESCRIPTION

The libpfm library is a helper library which can be used by applications to program the IA-64 Performance Monitoring Unit (PMU). While it is being developed on Linux in conjunction with the Linux/ia64 subsystem. It is generic enough to be used on other IA-64 operating systems. It is important to realize that the library does not make the actual kernel calls to program the PMU, it simply helps you figure out which PMU registers to use to measure certains events. On Linux/ia64, for instance, it does not make any perfmonctl call.

There are two categories of PMU registers. The performance monitoring data registers (PMD) are used to collects counts or serve as hardware buffers. The performance monitoring control registers (PMCS) are used to indicate what events need to be monitored. Programming the PMU consists in setting up the PMC registers to monitor certain events. The PMDS are commonly set to zero unless sampling is used.

The number of PMC and PMD registers varies from one implementation of the IA-64 architecture to another. The association of PMC to PMD can also change. Moreover the number and encoding of events can also widely change. Finally, the structure of a PMC register can also change. All these factors make it quite difficult to write monitoring tools.

This library is designed to simplify the programming of the PMC registers by hiding the complexity behind simple interfaces. The library does this without limiting accessibility to model specific features by using a layered design.

The library is structured in two layers. The common layer provides an interface that is shared by all implementations. For instance, getting an event descriptor using the event name is the same operation on all IA-64 implementation, even though the returned descriptor and the event name can vary. This layer is good enough to setup simple monitoring sessions which count occurrences of simple events. Then there is a model specific layer which gives access to the model-specific features. For instance, on Itanium, you can use the library to figure out how to initialize PMC12 which controls the Branch Trace Buffer. Model specific interfaces have the abbreviated PMU model name in their names. For instance, pfm_ita2_get_event_umask() is an Itanium2 (ita2) specific function.

When the library is initialized, it automatically probes the host CPU and enables the right set of interfaces. The layered design uses a modular implementation, where support for each CPU is well separated and can be compiled in or out of the library.

The common interface is defined in the pfmlib.h header file. The Itanium specific interface is in pfmlib_itanium.h. The pfmlib_itanium2.h provides the Itanium2 interface.

The common interface is described in individual man pages. There is one man page for each model specific set of interfaces.

 

AUTHOR

Stephane Eranian <eranian@hpl.hp.com>  

SEE ALSO

pfmlib_itanium(3), pfmlib_itanium2(3) and the set of examples shipped with the library


 

Index

NAME
SYNOPSIS
DESCRIPTION
AUTHOR
SEE ALSO

This document was created by man2html, using the manual pages.
Time: 21:49:20 GMT, April 16, 2011