1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
|
#!/bin/bash
#
# cachestat - show Linux page cache hit/miss statistics.
# Uses Linux ftrace.
#
# This is a proof of concept using Linux ftrace capabilities on older kernels,
# and works by using function profiling for in-kernel counters. Specifically,
# four kernel functions are traced:
#
# mark_page_accessed() for measuring cache accesses
# mark_buffer_dirty() for measuring cache writes
# add_to_page_cache_lru() for measuring page additions
# account_page_dirtied() for measuring page dirties
#
# It is possible that these functions have been renamed (or are different
# logically) for your kernel version, and this script will not work as-is.
# This script was written on Linux 3.13. This script is a sandcastle: the
# kernel may wash some away, and you'll need to rebuild.
#
# USAGE: cachestat [-Dht] [interval]
# eg,
# cachestat 5 # show stats every 5 seconds
#
# Run "cachestat -h" for full usage.
#
# WARNING: This uses dynamic tracing of kernel functions, and could cause
# kernel panics or freezes. Test, and know what you are doing, before use.
# It also traces cache activity, which can be frequent, and cost some overhead.
# The statistics should be treated as best-effort: there may be some error
# margin depending on unusual workload types.
#
# REQUIREMENTS: CONFIG_FUNCTION_PROFILER, awk.
#
# From perf-tools: https://github.com/brendangregg/perf-tools
#
# COPYRIGHT: Copyright (c) 2014 Brendan Gregg.
#
# This program is free software; you can redistribute it and/or
# modify it under the terms of the GNU General Public License
# as published by the Free Software Foundation; either version 2
# of the License, or (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software Foundation,
# Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
#
# (http://www.gnu.org/copyleft/gpl.html)
#
# 28-Dec-2014 Brendan Gregg Created this.
### default variables
tracing=/sys/kernel/debug/tracing
interval=1; opt_timestamp=0; opt_debug=0
trap 'quit=1' INT QUIT TERM PIPE HUP # sends execution to end tracing section
function usage {
cat <<-END >&2
USAGE: cachestat [-Dht] [interval]
-D # print debug counters
-h # this usage message
-t # include timestamp
interval # output interval in secs (default 1)
eg,
cachestat # show stats every second
cachestat 5 # show stats every 5 seconds
See the man page and example file for more info.
END
exit
}
function warn {
if ! eval "$@"; then
echo >&2 "WARNING: command failed \"$@\""
fi
}
function die {
echo >&2 "$@"
exit 1
}
### process options
while getopts Dht opt
do
case $opt in
D) opt_debug=1 ;;
t) opt_timestamp=1 ;;
h|?) usage ;;
esac
done
shift $(( $OPTIND - 1 ))
### option logic
if (( $# )); then
interval=$1
fi
echo "Counting cache functions... Output every $interval seconds."
### check permissions
cd $tracing || die "ERROR: accessing tracing. Root user? Kernel has FTRACE?
debugfs mounted? (mount -t debugfs debugfs /sys/kernel/debug)"
### enable tracing
sysctl -q kernel.ftrace_enabled=1 # doesn't set exit status
printf "mark_page_accessed\nmark_buffer_dirty\nadd_to_page_cache_lru\naccount_page_dirtied\n" > set_ftrace_filter || \
die "ERROR: tracing these four kernel functions: mark_page_accessed,"\
"mark_buffer_dirty, add_to_page_cache_lru and account_page_dirtied (unknown kernel version?). Exiting."
warn "echo nop > current_tracer"
if ! echo 1 > function_profile_enabled; then
echo > set_ftrace_filter
die "ERROR: enabling function profiling. Have CONFIG_FUNCTION_PROFILER? Exiting."
fi
(( opt_timestamp )) && printf "%-8s " TIME
printf "%8s %8s %8s %8s %12s %10s" HITS MISSES DIRTIES RATIO "BUFFERS_MB" "CACHE_MB"
(( opt_debug )) && printf " DEBUG"
echo
### summarize
quit=0; secs=0
while (( !quit && (!opt_duration || secs < duration) )); do
(( secs += interval ))
echo 0 > function_profile_enabled
echo 1 > function_profile_enabled
sleep $interval
(( opt_timestamp )) && printf "%(%H:%M:%S)T " -1
# cat both meminfo and trace stats, and let awk pick them apart
cat /proc/meminfo trace_stat/function* | awk -v debug=$opt_debug '
# match meminfo stats:
$1 == "Buffers:" && $3 == "kB" { buffers_mb = $2 / 1024 }
$1 == "Cached:" && $3 == "kB" { cached_mb = $2 / 1024 }
# identify and save trace counts:
$2 ~ /[0-9]/ && $3 != "kB" { a[$1] += $2 }
END {
mpa = a["mark_page_accessed"]
mbd = a["mark_buffer_dirty"]
apcl = a["add_to_page_cache_lru"]
apd = a["account_page_dirtied"]
total = mpa - mbd
misses = apcl - apd
if (misses < 0)
misses = 0
hits = total - misses
ratio = 100 * hits / total
printf "%8d %8d %8d %7.1f%% %12.0f %10.0f", hits, misses, mbd,
ratio, buffers_mb, cached_mb
if (debug)
printf " (%d %d %d %d)", mpa, mbd, apcl, apd
printf "\n"
}'
done
### end tracing
echo 2>/dev/null
echo "Ending tracing..." 2>/dev/null
warn "echo 0 > function_profile_enabled"
warn "echo > set_ftrace_filter"
|
