bcc-softirqs(8)

1softirqs(8)                 System Manager's Manual                softirqs(8)
2
3
4

NAME

6       softirqs  -  Measure  soft  IRQ (soft interrupt) event time. Uses Linux
7       eBPF/bcc.
8

SYNOPSIS

10       softirqs [-h] [-T] [-N] [-d] [interval] [count]
11

DESCRIPTION

13       This summarizes the time spent servicing soft IRQs  (soft  interrupts),
14       and  can  show this time as either totals or histogram distributions. A
15       system-wide summary of this time  is  shown  by  the  %soft  column  of
16       mpstat(1),  and  soft IRQ event counts (but not times) are available in
17       /proc/softirqs.
18
19       This tool uses the irq:softirq_enter and irq:softirq_exit kernel trace‐
20       points, which is a stable tracing mechanism. BPF programs can attach to
21       tracepoints from Linux 4.7 only. An  older  version  of  this  tool  is
22       available in tools/old, and uses kprobes instead of tracepoints.
23
24       Since this uses BPF, only the root user can use this tool.
25

REQUIREMENTS

27       CONFIG_BPF and bcc.
28

OPTIONS

30       -h     Print usage message.
31
32       -T     Include timestamps on output.
33
34       -N     Output in nanoseconds
35
36       -d     Show IRQ time distribution as histograms
37

EXAMPLES

39       Sum soft IRQ event time until Ctrl-C:
40              # softirqs
41
42       Show soft IRQ event time as histograms:
43              # softirqs -d
44
45       Print 1 second summaries, 10 times:
46              # softirqs 1 10
47
48       1 second summaries, printed in nanoseconds, with timestamps:
49              # softirqs -NT 1
50

FIELDS

52       SOFTIRQ
53              The kernel function name that performs the soft IRQ action.
54
55       TOTAL_usecs
56              Total time spent in this soft IRQ function in microseconds.
57
58       TOTAL_nsecs
59              Total time spent in this soft IRQ function in nanoseconds.
60
61       usecs  Range of microseconds for this bucket.
62
63       nsecs  Range of nanoseconds for this bucket.
64
65       count  Number of soft IRQs in this time range.
66
67       distribution
68              ASCII representation of the distribution (the count column).
69

OVERHEAD

71       This  traces kernel functions and maintains in-kernel counts, which are
72       asynchronously copied to user-space. While the rate  of  interrupts  be
73       very  high (>1M/sec), this is a relatively efficient way to trace these
74       events, and so the overhead is expected to be small  for  normal  work‐
75       loads,  but  could  become noticeable for heavy workloads. Measure in a
76       test environment before use.
77

SOURCE

79       This is from bcc.
80
81              https://github.com/iovisor/bcc
82
83       Also look in the bcc distribution for a  companion  _examples.txt  file
84       containing example usage, output, and commentary for this tool.
85

OS

87       Linux
88

STABILITY

90       Unstable - in development.
91

AUTHORS

93       Brendan Gregg, Sasha Goldshtein
94