.TH LIBPFM 3  "January, 2009" "" "Linux Programmer's Manual"

.SH NAME

libpfm_nehalem - support for Intel Nehalem processor family

.SH SYNOPSIS

.nf

.B #include <perfmon/pfmlib.h>

.B #include <perfmon/pfmlib_intel_nhm.h>

.sp

.SH DESCRIPTION

The libpfm library provides full support for the Intel Nehalem processor family, such as

Intel Core i7. The interface is defined in \fBpfmlib_intel_nhm.h\fR. It consists of a set

of functions and structures describing the Intel Nehalem processor specific PMU features.

The Intel Nehalem processor is a quad core, dual thread processor. It includes two types

of PMU: core and uncore. The latter measures events at the socket level and is therefore

disconnected from any of the four cores. The core PMU implements Intel architectural 

perfmon version 3 with four generic counters and three fixed counters. The uncore has

eight generic counters and one fixed counter. Each Intel Nehalem core also implement

a 16-deep branch trace buffer, called Last Branch Record (LBR), which can be used in

combination with the core PMU. Intel Nehalem implements a newer version of the 

Precise Event-Based Sampling (PEBS) mechanism which has the ability to capture

where cache misses occur.

.sp

When Intel Nehalem processor specific features are needed to support a measurement, their

descriptions must be passed as model-specific input arguments to the

\fBpfm_dispatch_events()\fR function. The Intel Nehalem processors specific input

arguments are described in the \fBpfmlib_nhm_input_param_t\fR structure. No

output parameters are currently defined. The input parameters are defined as follows:

.sp

.nf

typedef struct {

	unsigned long  cnt_mask;

	unsigned int   flags;

} pfmlib_nhm_counter_t;

typedef struct {

	unsigned int lbr_used;

	unsigned int lbr_plm;

	unsigned int lbr_filter;

} pfmlib_nhm_lbr_t;

typedef struct {

	unsigned int pebs_used;

	unsigned int ld_lat_thres;

} pfmlib_nhm_pebs_t;

typedef struct {

	pfmlib_nhm_counter_t pfp_nhm_counters[PMU_NHM_NUM_COUNTERS];

	pfmlib_nhm_pebs_t    pfp_nhm_pebs;

	pfmlib_nhm_lbr_t     pfm_nhm_lbr;

	uint64_t             reserved[4];

} pfmlib_nhm_input_param_t;

.fi

.sp

.sp

The Intel Nehalem processor provides a few additional per-event features for 

counters: thresholding, inversion, edge detection, monitoring of both

threads, occupancy. They can be set using the \fBpfp_nhm_counters\fR data

structure for each event.  The \fBflags\fR field can be initialized with

the following values, depending on the event:

.TP

.B PFMLIB_NHM_SEL_INV

Inverse the results of the \fBcnt_mask\fR comparison when set. This

flag is supported for core and uncore PMU events.

.TP

.B PFMLIB_NHM_SEL_EDGE

Enables edge detection of events. This

flag is supported for core and uncore PMU events.

.TP

.B PFMLIB_NHM_SEL_ANYTHR

Enable measuring the event in any of the two processor threads assuming hyper-threading

is enabled.  By default, only the current thread is measured. This flag is restricted

to core PMU events.

.TP

.B PFMLIB_NHM_SEL_OCC_RST

When set, the queue occupancy counter associated with the event is cleared. This flag

is only available to uncore PMU events.

.LP

The \fBcnt_mask\fR field is used to set the event threshold.

The value of the counter is incremented for each cycle in which the

number of occurrences of the event is greater or equal to the value of

the field. Thus, the event is modified to actually measure the number

of qualifying cycles.  When zero all occurrences are counted (this is the default).

This flag is supported for core and uncore PMU events.

.sp

.SH Support for Precise-Event Based Sampling (PEBS)

The library can be used to setup the PMC registers associated with PEBS. In this case,

the \fBpfp_nhm_pebs_t\fR structure must be used and the \fBpebs_used\fR field must

be set to 1.

.sp

To enable the PEBS load latency filtering capability, it is necessary to program the

\fBMEM_INST_RETIRED:LATENCY_ABOVE_THRESHOLD\fR event into one generic counter. The

latency threshold must be passed to the library in the \fBld_lat_thres\fR field.

It is expressed in core cycles and \fBmust\fR greater than 3. Note that \fBpebs_used\fR

must be set as well.

.SH Support for Last Branch Record (LBR)

The library can be used to setup LBR registers. On Intel Nehalem processors, the

LBR is 16-entry deep and it is possible to filter branches, based on privilege level

or type. To configure the LBR, the \fBpfm_nhm_lbr_t\fR structure must be used.

.sp

Like core PMU counters, LBR only distinguishes two privilege levels, 0 and the rest (1,2,3).

When running Linux natively, the kernel is at privilege level 0, applications at level 3.

It is possible to specify the privilege level of LBR using the \fBlbr_plm\fR. Any attempt

to pass \fBPFM_PLM1\fB or \fBPFM_PLM2\fR will be rejected. If \fB\lbr_plm\fR is 0, then the global

value in \fBpfmlib_input_param_t\fR and the \fBpfp_dfl_plm\fR is used.

.sp

By default, LBR captures all branches. It is possible to filter out branches by passing

a set of flags in \fBlbr_select\fR. The flags are as follows:

.TP

.B PFMLIB_NHM_LBR_JCC

When set, LBR does not capture conditional branches. Default: off.

.TP

.B PFM_NHM_LBR_NEAR_REL_CALL

When set, LBR does not capture near calls. Default: off.

.TP

.B PFM_NHM_LBR_NEAR_IND_CALL

When set, LBR does not capture indirect calls. Default: off.

.TP

.B PFM_NHM_LBR_NEAR_RET

When set, LBR does not capture return branches. Default: off.

.TP

.B PFM_NHM_LBR_NEAR_IND_JMP

When set, LBR does not capture indirect branches. Default: off.

.TP

.B PFM_NHM_LBR_NEAR_REL_JMP

When set, LBR does not capture relative branches. Default: off.

.TP

.B PFM_NHM_LBR_FAR_BRANCH

When set, LBR does not capture far branches. Default: off.

.SH Support for uncore PMU

By nature, the uncore PMU does not distinguish privilege levels, therefore

it captures events at all privilege levels. To avoid any misinterpretation,

the library enforces that uncore events be measured with both \fBPFM_PLM0\fR

and \fBPFM_PLM3\fR set.

Tools and operating system kernel interfaces may impose further restrictions

on how the uncore PMU can be accessed.

.SH SEE ALSO

pfm_dispatch_events(3) and set of examples shipped with the library

.SH AUTHOR

Stephane Eranian <eranian@gmail.com>

.PP