NAME

ppchcalls - Summarize ppc hcall counts and latencies.

SYNOPSIS

ppchcalls [-h] [-p PID] [-t TID] [-i INTERVAL] [-d DURATION] [-T TOP] [-x] [-e ERRNO] [-L] [-m] [-P] [-l] [--hcall HCALL]

DESCRIPTION

This tool traces hcall entry and exit raw tracepoints and summarizes either the number of hcalls of each type, or the number of hcalls per process. It can also collect min, max and average latency for each hcall or each process.

Since this uses BPF, only the root user can use this tool.

REQUIREMENTS

CONFIG_BPF and bcc. Linux 4.17+ is required to attach a BPF program to the raw_hcalls:hcall_{enter,exit} tracepoints, used by this tool.

OPTIONS

-h

Print usage message.

-p PID

Trace only this process.

-t TID

Trace only this thread.

-i INTERVAL

Print the summary at the specified interval (in seconds).

-d DURATION

Total duration of trace (in seconds).

-T TOP

Print only this many entries. Default: 10.

-x

Trace only failed hcalls (i.e., the return value from the hcall was < 0).

-e ERRNO

Trace only hcalls that failed with that error (e.g. -e EPERM or -e 1).

-m

Display times in milliseconds. Default: microseconds.

-P

Summarize by process and not by hcall.

-l

List the hcalls recognized by the tool (hard-coded list). Hcalls beyond this list will still be displayed, as "[unknown: nnn]" where nnn is the hcall number.

--hcall HCALL

Trace this hcall only (use option -l to get all recognized hcalls).

EXAMPLES

Summarize all hcalls by hcall:

# ppchcalls

Summarize all hcalls by process:

# ppchcalls -P

Summarize only failed hcalls:

# ppchcalls -x

Summarize only hcalls that failed with EPERM:

# ppchcalls -e EPERM

Trace PID 181 only:

# ppchcalls -p 181

Summarize hcalls counts and latencies:

# ppchcalls -L

FIELDS

PID

Process ID

COMM

Process name

HCALL

Hcall name, or "[unknown: nnn]" for hcalls that aren't recognized

COUNT

The number of events

MIN

The minimum elapsed time (in us or ms)

MAX

The maximum elapsed time (in us or ms)

AVG

The average elapsed time (in us or ms)

OVERHEAD

For most applications, the overhead should be manageable if they perform 1000's or even 10,000's of hcalls per second. For higher rates, the overhead may become considerable.

SOURCE

This is from bcc.

https://github.com/iovisor/bcc

Also look in the bcc distribution for a companion _examples.txt file containing example usage, output, and commentary for this tool.

OS

Linux

STABILITY

Unstable - in development.

AUTHOR

Harsh Prateek Bora

SEE ALSO

syscount(8)