.\"
.\" Copyright (c) 2007, Sun Microsystems, Inc. All Rights Reserved.
.\" Copyright 2021 Oxide Computer Company
.\"
.\" The contents of this file are subject to the terms of the
.\" Common Development and Distribution License (the "License").
.\" You may not use this file except in compliance with the License.
.\"
.\" You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
.\" or http://www.opensolaris.org/os/licensing.
.\" See the License for the specific language governing permissions
.\" and limitations under the License.
.\"
.\" When distributing Covered Code, include this CDDL HEADER in each
.\" file and include the License file at usr/src/OPENSOLARIS.LICENSE.
.\" If applicable, add the following below this CDDL HEADER, with the
.\" fields enclosed by brackets "[]" replaced with your own identifying
.\" information: Portions Copyright [yyyy] [name of copyright owner]
.\"
.Dd November 29, 2021
.Dt KSTAT 3KSTAT
.Os
.Sh NAME
.Nm kstat
.Nd kernel statistics facility
.Sh LIBRARY
.Lb libkstat
.Sh DESCRIPTION
The
.Nm
facility is a general-purpose mechanism for providing kernel statistics to
users.
.Ss The kstat model
The kernel maintains a linked list of statistics structures, or kstats.
Each kstat has a common header section and a type-specific data section.
The header section is defined by the
.Vt kstat_t
structure:
.Ss "kstat header"
.Bd -literal -offset indent
typedef int kid_t; /* unique kstat id */
typedef struct kstat {
/*
* Fields relevant to both kernel and user
*/
hrtime_t ks_crtime; /* creation time */
struct kstat *ks_next; /* kstat chain linkage */
kid_t ks_kid; /* unique kstat ID */
char ks_module[KSTAT_STRLEN]; /* module name */
uchar_t ks_resv; /* reserved */
int ks_instance; /* module's instance */
char ks_name[KSTAT_STRLEN]; /* kstat name */
uchar_t ks_type; /* kstat data type */
char ks_class[KSTAT_STRLEN]; /* kstat class */
uchar_t ks_flags; /* kstat flags */
void *ks_data; /* kstat type-specific
data */
uint_t ks_ndata; /* # of data records */
size_t ks_data_size; /* size of kstat data
section */
hrtime_t ks_snaptime; /* time of last data
snapshot */
/*
* Fields relevant to kernel only
*/
int(*ks_update)(struct kstat *, int);
void *ks_private;
int(*ks_snapshot)(struct kstat *, void *, int);
void *ks_lock;
} kstat_t;
.Ed
.Pp
The fields that are of significance to the user are:
.Bl -tag -width Ds
.It Fa ks_crtime
The time the kstat was created.
This allows you to compute the rates of various counters since the kstat was
created;
.Dq rate since boot
is replaced by the more general concept of
.Dq rate since kstat creation .
All times associated with kstats
.Po
such as creation time, last snapshot time,
.Vt kstat_timer_t
and
.Vt kstat_io_t
timestamps, and the like
.Pc
are 64-bit nanosecond values.
The accuracy of kstat timestamps is machine dependent, but the precision
.Pq units
is the same across all platforms.
See
.Xr gethrtime 3C
for general information about high-resolution timestamps.
.It Fa ks_next
kstats are stored as a linked list, or chain.
.Fa ks_next
points to the next kstat in the chain.
.It Fa ks_kid
A unique identifier for the kstat.
.It Fa ks_module , Fa ks_instance
contain the name and instance of the module that created the kstat.
In cases where there can only be one instance,
.Fa ks_instance
is 0.
.It Fa ks_name
gives a meaningful name to a kstat.
The full kstat namespace is
.Ao
.Fa ks_module ,
.Fa ks_instance ,
.Fa ks_name
.Ac ,
so the name only need be unique within a module.
.It Fa ks_type
The type of data in this kstat.
kstat data types are discussed below.
.It Fa ks_class
Each kstat can be characterized as belonging to some broad class of statistics,
such as disk, tape, net, vm, and streams.
This field can be used as a filter to extract related kstats.
The following values are currently in use: disk, tape, controller, net, rpc, vm,
kvm, hat, streams, kmem, kmem_cache, kstat, and misc.
.Po
The kstat class encompasses things like
.Fa kstat_types .
.Pc
.It Fa ks_data , Fa ks_ndata , Fa ks_data_size
.Fa ks_data
is a pointer to the kstat's data section.
The type of data stored there depends on
.Fa ks_type .
.Fa ks_ndata
indicates the number of data records.
Only some kstat types support multiple data records.
Currently,
.Dv KSTAT_TYPE_RAW ,
.Dv KSTAT_TYPE_NAMED ,
and
.Dv KSTAT_TYPE_TIMER
kstats support multiple data records.
.Dv KSTAT_TYPE_INTR
and
.Dv KSTAT_TYPE_IO
kstats support only one data record.
.Fa ks_data_size
is the total size of the data section, in bytes.
.It Fa ks_snaptime
The timestamp for the last data snapshot.
This allows you to compute activity rates:
.Bd -literal
rate = (new_count - old_count) / (new_snaptime - old_snaptime);
.Ed
.El
.Ss kstat data types
The following types of kstats are currently available:
.Bd -literal -offset indent
#define KSTAT_TYPE_RAW 0 /* can be anything */
#define KSTAT_TYPE_NAMED 1 /* name/value pairs */
#define KSTAT_TYPE_INTR 2 /* interrupt statistics */
#define KSTAT_TYPE_IO 3 /* I/O statistics */
#define KSTAT_TYPE_TIMER 4 /* event timers */
.Ed
.Pp
To get a list of all kstat types currently supported in the system, tools can
read out the standard system kstat
.Fa kstat_types
.Po
full name spec is
.Aq unix, 0, kstat_types
.Pc .
This is a
.Dv KSTAT_TYPE_NAMED
kstat in which the
.Fa name
field describes the type of kstat, and the
.Fa value
field is the kstat type number
.Po
for example,
.Dv KSTAT_TYPE_IO
is type 3 \(em see above
.Pc .
.Ss "Raw kstat"
.Dv KSTAT_TYPE_RAW
\(em raw data
.Pp
The
.Dq raw
kstat type is just treated as an array of bytes.
This is generally used to export well-known structures, like
.Vt sysinfo .
.Ss "Name=value kstat"
.Dv KSTAT_TYPE_NAMED
\(em A list of arbitrary
.Fa name=value
statistics.
.Bd -literal -offset indent
typedef struct kstat_named {
char name[KSTAT_STRLEN]; /* name of counter */
uchar_t data_type; /* data type */
union {
charc[16]; /* enough for 128-bit ints */
struct {
union {
char *ptr; /* NULL-terminated string */
} addr;
uint32_t len; /* length of string */
} str;
int32_t i32;
uint32_t ui32;
int64_t i64;
uint64_t ui64;
/* These structure members are obsolete */
int32_t l;
uint32_t ul;
int64_t ll;
uint64_t ull;
} value; /* value of counter */
} kstat_named_t;
/* The following types are Stable
KSTAT_DATA_CHAR
KSTAT_DATA_INT32
KSTAT_DATA_LONG
KSTAT_DATA_UINT32
KSTAT_DATA_ULONG
KSTAT_DATA_INT64
KSTAT_DATA_UINT64
/* The following type is Evolving */
KSTAT_DATA_STRING
/* The following types are Obsolete */
KSTAT_DATA_LONGLONG
KSTAT_DATA_ULONGLONG
KSTAT_DATA_FLOAT
KSTAT_DATA_DOUBLE
.Ed
.Pp
Some devices need to publish strings that exceed the maximum value for
.Dv KSTAT_DATA_CHAR
in length;
.Dv KSTAT_DATA_STRING
is a data type that allows arbitrary-length strings to be associated with a
named kstat.
The macros below are the supported means to read the pointer to the string and
its length.
.Bd -literal -offset indent
#define KSTAT_NAMED_STR_PTR(knptr) ((knptr)->value.str.addr.ptr)
#define KSTAT_NAMED_STR_BUFLEN(knptr) ((knptr)->value.str.len)
.Ed
.Pp
.Fn KSTAT_NAMED_STR_BUFLEN
returns the number of bytes required to store the string pointed to by
.Fn KSTAT_NAMED_STR_PTR ;
that is,
.Fo strlen
.Fa "KSTAT_NAMED_STR_PTR() + 1"
.Fc .
.Ss "Interrupt kstat"
.Dv KSTAT_TYPE_INTR
\(em Interrupt statistics.
.Pp
An interrupt is a hard interrupt
.Pq sourced from the hardware device itself ,
a soft interrupt
.Pq induced by the system via the use of some system interrupt source ,
a watchdog interrupt
.Pq induced by a periodic timer call ,
spurious
.Pq an interrupt entry point was entered but there was no interrupt to service ,
or multiple service
.Po
an interrupt was detected and serviced just prior to returning from any of the
other types
.Pc .
.Bd -literal -offset indent
#define KSTAT_INTR_HARD 0
#define KSTAT_INTR_SOFT 1
#define KSTAT_INTR_WATCHDOG 2
#define KSTAT_INTR_SPURIOUS 3
#define KSTAT_INTR_MULTSVC 4
#define KSTAT_NUM_INTRS 5
typedef struct kstat_intr {
uint_t intrs[KSTAT_NUM_INTRS]; /* interrupt counters */
} kstat_intr_t;
.Ed
.Ss Event timer kstat
.Dv KSTAT_TYPE_TIMER
\(em Event timer statistics.
.Pp
These provide basic counting and timing information for any type of event.
.Bd -literal -offset indent
typedef struct kstat_timer {
char name[KSTAT_STRLEN]; /* event name */
uchar_t resv; /* reserved */
u_longlong_t num_events; /* number of events */
hrtime_t elapsed_time; /* cumulative elapsed time */
hrtime_t min_time; /* shortest event duration */
hrtime_t max_time; /* longest event duration */
hrtime_t start_time; /* previous event start time */
hrtime_t stop_time; /* previous event stop time */
} kstat_timer_t;
.Ed
.Ss I/O kstat
.Dv KSTAT_TYPE_IO
\(em I/O statistics.
.Bd -literal -offset indent
typedef struct kstat_io {
/*
* Basic counters.
*/
u_longlong_t nread; /* number of bytes read */
u_longlong_t nwritten; /* number of bytes written */
uint_t reads; /* number of read operations */
uint_t writes; /* number of write operations */
/*
* Accumulated time and queue length statistics.
*
* Time statistics are kept as a running sum of "active" time.
* Queue length statistics are kept as a running sum of the
* product of queue length and elapsed time at that length --
* that is, a Riemann sum for queue length integrated against time.
*
* ^
* | _________
* 8 | i4 |
* | | |
* Queue 6 | |
* Length | _________ | |
* 4 | i2 |_______| |
* | | i3 |
* 2_______| |
* | i1 |
* |_______________________________|
* Time-> t1 t2 t3 t4
*
* At each change of state (entry or exit from the queue),
* we add the elapsed time (since the previous state change)
* to the active time if the queue length was non-zero during
* that interval; and we add the product of the elapsed time
* times the queue length to the running length*time sum.
*
* This method is generalizable to measuring residency
* in any defined system: instead of queue lengths, think
* of "outstanding RPC calls to server X".
*
* A large number of I/O subsystems have at least two basic
* "lists" of transactions they manage: one for transactions
* that have been accepted for processing but for which processing
* has yet to begin, and one for transactions which are actively
* being processed (but not done). For this reason, two cumulative
* time statistics are defined here: pre-service (wait) time,
* and service (run) time.
*
* The units of cumulative busy time are accumulated nanoseconds.
* The units of cumulative length*time products are elapsed time
* times queue length.
*/
hrtime_t wtime; /* cumulative wait (pre-service) time */
hrtime_t wlentime; /* cumulative wait length*time product*/
hrtime_t wlastupdate; /* last time wait queue changed */
hrtime_t rtime; /* cumulative run (service) time */
hrtime_t rlentime; /* cumulative run length*time product */
hrtime_t rlastupdate; /* last time run queue changed */
uint_t wcnt; /* count of elements in wait state */
uint_t rcnt; /* count of elements in run state */
} kstat_io_t;
.Ed
.Ss Using libkstat
The kstat library,
.Sy libkstat ,
defines the user interface
.Pq API
to the system's kstat facility.
.Pp
You begin by opening libkstat with
.Xr kstat_open 3KSTAT ,
which returns a pointer to a fully initialized kstat control structure.
This is your ticket to subsequent libkstat operations:
.Bd -literal -offset indent
typedef struct kstat_ctl {
kid_t kc_chain_id; /* current kstat chain ID */
kstat_t *kc_chain; /* pointer to kstat chain */
int kc_kd; /* /dev/kstat descriptor */
} kstat_ctl_t;
.Ed
.Pp
Only the first two fields,
.Fa kc_chain_id
and
.Fa kc_chain ,
are of
interest to
libkstat clients.
.Po
.Fa kc_kd
is the descriptor for
.Pa /dev/kstat ,
the kernel statistics driver.
libkstat functions are built on top of
.Pa /dev/kstat
.Xr ioctl 2
primitives.
Direct interaction with
.Pa /dev/kstat
is strongly discouraged, since it is
.Em not
a public interface.
.Pc
.Pp
.Fa kc_chain
points to your copy of the kstat chain.
You typically walk the chain to find and process a certain kind of kstat.
For example, to display all
I/O kstats:
.Bd -literal -offset indent
kstat_ctl_t *kc;
kstat_t *ksp;
kstat_io_t kio;
kc = kstat_open();
for (ksp = kc->kc_chain; ksp != NULL; ksp = ksp->ks_next) {
if (ksp->ks_type == KSTAT_TYPE_IO) {
kstat_read(kc, ksp, &kio);
my_io_display(kio);
}
}
.Ed
.Pp
.Fa kc_chain_id
is the kstat chain ID, or
.Sy KCID ,
of your copy of the kstat chain.
See
.Xr kstat_chain_update 3KSTAT
for an explanation of KCIDs.
.Sh FILES
.Bl -tag -width Pa
.It Pa /dev/kstat
kernel statistics driver character device
.It Pa /usr/include/kstat.h
kstat library header file
.It Pa /usr/include/sys/kstat.h
system kstat header
.El
.Sh SEE ALSO
.Xr ioctl 2 ,
.Xr gethrtime 3C ,
.Xr kstat_chain_update 3KSTAT ,
.Xr kstat_close 3KSTAT ,
.Xr kstat_data_lookup 3KSTAT ,
.Xr kstat_lookup 3KSTAT ,
.Xr kstat_open 3KSTAT ,
.Xr kstat_read 3KSTAT ,
.Xr kstat_write 3KSTAT ,
.Xr attributes 7