.TH LIBPFM 3  "July, 2016" "" "Linux Programmer's Manual"

.SH NAME

libpfm_intel_knl - support for Intel Kinghts Landing core PMU

.SH SYNOPSIS

.nf

.B #include <perfmon/pfmlib.h>

.sp

.B PMU name: knl

.B PMU desc: Intel Kinghts Landing

.sp

.SH DESCRIPTION

The library supports the Intel Kinghts Landing core PMU. It should be noted that

this PMU model only covers each core's PMU and not the socket level PMU.

On Knights Landing, the number of generic counters is 4. There is 4-way HyperThreading support.

The \fBpfm_get_pmu_info()\fR function returns the maximum number of generic counters

in \fBnum_cntrs\fr.

.SH MODIFIERS

The following modifiers are supported on Intel Kinghts Landing processors:

.TP

.B u

Measure at user level which includes privilege levels 1, 2, 3. This corresponds to \fBPFM_PLM3\fR.

This is a boolean modifier.

.TP

.B k

Measure at kernel level which includes privilege level 0. This corresponds to \fBPFM_PLM0\fR.

This is a boolean modifier.

.TP

.B i

Invert the meaning of the event. The counter will now count cycles in which the event is \fBnot\fR

occurring. This is a boolean modifier

.TP

.B e

Enable edge detection, i.e., count only when there is a state transition from no occurrence of the event

to at least one occurrence. This modifier must be combined with a counter mask modifier (m) with a value greater or equal to one.

This is a boolean modifier.

.TP

.B c

Set the counter mask value. The mask acts as a threshold. The counter will count the number of cycles

in which the number of occurrences of the event is greater or equal to the threshold. This is an integer

modifier with values in the range [0:255].

.TP

.B t

Measure on any of the 4 hyper-threads at the same time assuming hyper-threading is enabled. This is a boolean modifier.

This modifier is only available on fixed counters (unhalted_reference_cycles, instructions_retired, unhalted_core_cycles).

Depending on the underlying kernel interface, the event may be programmed on a fixed counter or a generic counter, except for

unhalted_reference_cycles, in which case, this modifier may be ignored or rejected.

.SH OFFCORE_RESPONSE events

Intel Knights Landing provides two offcore_response events. They are called OFFCORE_RESPONSE_0 and OFFCORE_RESPONSE_1.

Those events need special treatment in the performance monitoring infrastructure

because each event uses an extra register to store some settings. Thus, in

case multiple offcore_response events are monitored simultaneously, the kernel needs

to manage the sharing of that extra register.

The offcore_response events are exposed as normal events by the library. The extra

settings are exposed as regular umasks. The library takes care of encoding the

events according to the underlying kernel interface.

On Intel Knights Landing, the umasks are divided into 4 categories: request, supplier

and snoop and average latency. Offcore_response event has two modes of operations: normal and average latency.

In the first mode, the two offcore_respnse events operate independently of each other. The user must provide at

least one umask for each of the first 3 categories: request, supplier, snoop. In the second mode, the two

offcore_response events are combined to compute an average latency per request type.

For the normal mode, there is a special supplier (response) umask called \fBANY_RESPONSE\fR. When this umask

is used then it overrides any supplier and snoop umasks. In other words, users can

specify either \fBANY_RESPONSE\fR \fBOR\fR any combinations of supplier + snoops. In case no supplier or snoop

is specified, the library defaults to using \fBANY_RESPONSE\fR.

For instance, the following are valid event selections:

.TP

.B OFFCORE_RESPONSE_0:DMND_DATA_RD:ANY_RESPONSE

.TP

.B OFFCORE_RESPONSE_0:ANY_REQUEST

.TP

.B OFFCORE_RESPONSE_0:ANY_RFO:DDR_NEAR

.P

But the following is illegal:

.TP

.B OFFCORE_RESPONSE_0:ANY_RFO:DDR_NEAR:ANY_RESPONSE

.P

In average latency mode, \fBOFFCORE_RESPONSE_0\fR must be programmed to select the request types of interest, for instance, \fBDMND_DATA_RD\fR, and the \fBOUTSTANDING\fR umask must be set and no others. the library will enforce that restriction as soon as the \fBOUTSTANDING\fR umask is used. Then \fBOFFCORE_RESPONSE_1\fR must be set with the same request types and the \fBANY_RESPONSE\fR umask. It should be noted that the library encodes events independently of each other and therefore cannot verify that the requests are matching between the two events.

Example of average latency settings:

.TP

.B OFFCORE_RESPONSE_0:DMND_DATA_RD:OUTSTANDING+OFFCORE_RESPONSE_1:DMND_DATA_RD:ANY_RESPONSE

.TP

.B OFFCORE_RESPONSE_0:ANY_REQUEST:OUTSTANDING+OFFCORE_RESPONSE_1:ANY_REQUEST:ANY_RESPONSE

.P

The average latency for the request(s) is obtained by dividing the counts of \fBOFFCORE_RESPONSE_0\fR by the count of \fBOFFCORE_RESPONSE_1\fR. The ratio is expressed in core cycles.

.SH AUTHORS

.nf

Stephane Eranian <eranian@gmail.com>

.if

.PP