Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
docs: perf: convert to ReST
Rename the perf documentation files to ReST, add an
index for them and adjust in order to produce a nice html
output via the Sphinx build system.
At its new index.rst, let's add a :orphan: while this is not linked to
the main index.rst file, in order to avoid build warnings.
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
+72
-42
+10
-8
Documentation/perf/arm-ccn.txt
Documentation/perf/arm-ccn.rst
+10
-8
Documentation/perf/arm-ccn.txt
Documentation/perf/arm-ccn.rst
···
1
+
==========================
1
2
ARM Cache Coherent Network
2
3
==========================
3
4
···
30
29
require "xp" and "vc" as as above plus "port" (device port index),
31
30
"dir" (transmit/receive direction), comparator values ("cmp_l"
32
31
and "cmp_h") and "mask", being index of the comparator mask.
32
+
33
33
Masks are defined separately from the event description
34
34
(due to limited number of the config values) in the "cmp_mask"
35
35
directory, with first 8 configurable by user and additional
···
46
44
will be overwritten anyway). In case of this processor being offlined,
47
45
the events are migrated to another one and the attribute is updated.
48
46
49
-
Example of perf tool use:
47
+
Example of perf tool use::
50
48
51
-
/ # perf list | grep ccn
52
-
ccn/cycles/ [Kernel PMU event]
53
-
<...>
54
-
ccn/xp_valid_flit,xp=?,port=?,vc=?,dir=?/ [Kernel PMU event]
55
-
<...>
49
+
/ # perf list | grep ccn
50
+
ccn/cycles/ [Kernel PMU event]
51
+
<...>
52
+
ccn/xp_valid_flit,xp=?,port=?,vc=?,dir=?/ [Kernel PMU event]
53
+
<...>
56
54
57
-
/ # perf stat -a -e ccn/cycles/,ccn/xp_valid_flit,xp=1,port=0,vc=1,dir=1/ \
58
-
sleep 1
55
+
/ # perf stat -a -e ccn/cycles/,ccn/xp_valid_flit,xp=1,port=0,vc=1,dir=1/ \
56
+
sleep 1
59
57
60
58
The driver does not support sampling, therefore "perf record" will
61
59
not work. Per-task (without "-a") perf sessions are not supported.
+3
-2
Documentation/perf/arm_dsu_pmu.txt
Documentation/perf/arm_dsu_pmu.rst
+3
-2
Documentation/perf/arm_dsu_pmu.txt
Documentation/perf/arm_dsu_pmu.rst
···
1
+
==================================
1
2
ARM DynamIQ Shared Unit (DSU) PMU
2
3
==================================
3
4
···
14
13
The DSU provides a bitmap for a subset of implemented events via hardware
15
14
registers. There is no way for the driver to determine if the other events
16
15
are available or not. Hence the driver exposes only those events advertised
17
-
by the DSU, in "events" directory under :
16
+
by the DSU, in "events" directory under::
18
17
19
18
/sys/bus/event_sources/devices/arm_dsu_<N>/
20
19
···
24
23
The driver also exposes the CPUs connected to the DSU instance in "associated_cpus".
25
24
26
25
27
-
e.g usage :
26
+
e.g usage::
28
27
29
28
perf stat -a -e arm_dsu_0/cycles/
+21
-14
Documentation/perf/hisi-pmu.txt
Documentation/perf/hisi-pmu.rst
+21
-14
Documentation/perf/hisi-pmu.txt
Documentation/perf/hisi-pmu.rst
···
1
+
======================================================
1
2
HiSilicon SoC uncore Performance Monitoring Unit (PMU)
2
3
======================================================
4
+
3
5
The HiSilicon SoC chip includes various independent system device PMUs
4
6
such as L3 cache (L3C), Hydra Home Agent (HHA) and DDRC. These PMUs are
5
7
independent and have hardware logic to gather statistics and performance
···
13
11
two HHAs (0 - 1) and four DDRCs (0 - 3), respectively.
14
12
15
13
HiSilicon SoC uncore PMU driver
16
-
---------------------------------------
14
+
-------------------------------
15
+
17
16
Each device PMU has separate registers for event counting, control and
18
17
interrupt, and the PMU driver shall register perf PMU drivers like L3C,
19
18
HHA and DDRC etc. The available events and configuration options shall
20
-
be described in the sysfs, see :
19
+
be described in the sysfs, see:
20
+
21
21
/sys/devices/hisi_sccl{X}_<l3c{Y}/hha{Y}/ddrc{Y}>/, or
22
22
/sys/bus/event_source/devices/hisi_sccl{X}_<l3c{Y}/hha{Y}/ddrc{Y}>.
23
23
The "perf list" command shall list the available events from sysfs.
···
28
24
name will appear in event listing as hisi_sccl<sccl-id>_module<index-id>.
29
25
where "sccl-id" is the identifier of the SCCL and "index-id" is the index of
30
26
module.
27
+
31
28
e.g. hisi_sccl3_l3c0/rd_hit_cpipe is READ_HIT_CPIPE event of L3C index #0 in
32
29
SCCL ID #3.
30
+
33
31
e.g. hisi_sccl1_hha0/rx_operations is RX_OPERATIONS event of HHA index #0 in
34
32
SCCL ID #1.
35
33
36
34
The driver also provides a "cpumask" sysfs attribute, which shows the CPU core
37
35
ID used to count the uncore PMU event.
38
36
39
-
Example usage of perf:
40
-
$# perf list
41
-
hisi_sccl3_l3c0/rd_hit_cpipe/ [kernel PMU event]
42
-
------------------------------------------
43
-
hisi_sccl3_l3c0/wr_hit_cpipe/ [kernel PMU event]
44
-
------------------------------------------
45
-
hisi_sccl1_l3c0/rd_hit_cpipe/ [kernel PMU event]
46
-
------------------------------------------
47
-
hisi_sccl1_l3c0/wr_hit_cpipe/ [kernel PMU event]
48
-
------------------------------------------
37
+
Example usage of perf::
49
38
50
-
$# perf stat -a -e hisi_sccl3_l3c0/rd_hit_cpipe/ sleep 5
51
-
$# perf stat -a -e hisi_sccl3_l3c0/config=0x02/ sleep 5
39
+
$# perf list
40
+
hisi_sccl3_l3c0/rd_hit_cpipe/ [kernel PMU event]
41
+
------------------------------------------
42
+
hisi_sccl3_l3c0/wr_hit_cpipe/ [kernel PMU event]
43
+
------------------------------------------
44
+
hisi_sccl1_l3c0/rd_hit_cpipe/ [kernel PMU event]
45
+
------------------------------------------
46
+
hisi_sccl1_l3c0/wr_hit_cpipe/ [kernel PMU event]
47
+
------------------------------------------
48
+
49
+
$# perf stat -a -e hisi_sccl3_l3c0/rd_hit_cpipe/ sleep 5
50
+
$# perf stat -a -e hisi_sccl3_l3c0/config=0x02/ sleep 5
52
51
53
52
The current driver does not support sampling. So "perf record" is unsupported.
54
53
Also attach to a task is unsupported as the events are all uncore.
+16
Documentation/perf/index.rst
+16
Documentation/perf/index.rst
+2
-1
Documentation/perf/qcom_l2_pmu.txt
Documentation/perf/qcom_l2_pmu.rst
+2
-1
Documentation/perf/qcom_l2_pmu.txt
Documentation/perf/qcom_l2_pmu.rst
···
1
+
=====================================================================
1
2
Qualcomm Technologies Level-2 Cache Performance Monitoring Unit (PMU)
2
3
=====================================================================
3
4
···
29
28
consisting of one CPU per cluster which will be used to handle all the PMU
30
29
events on that cluster.
31
30
32
-
Examples for use with perf:
31
+
Examples for use with perf::
33
32
34
33
perf stat -e l2cache_0/config=0x001/,l2cache_0/config=0x042/ -a sleep 1
35
34
+2
-1
Documentation/perf/qcom_l3_pmu.txt
Documentation/perf/qcom_l3_pmu.rst
+2
-1
Documentation/perf/qcom_l3_pmu.txt
Documentation/perf/qcom_l3_pmu.rst
···
1
+
===========================================================================
1
2
Qualcomm Datacenter Technologies L3 Cache Performance Monitoring Unit (PMU)
2
3
===========================================================================
3
4
···
18
17
exposed via the "event" format attribute. In addition to the 32bit physical
19
18
counters the driver supports virtual 64bit hardware counters by using hardware
20
19
counter chaining. This feature is exposed via the "lc" (long counter) format
21
-
flag. E.g.:
20
+
flag. E.g.::
22
21
23
22
perf stat -e l3cache_0_0/read-miss,lc/
24
23
+13
-12
Documentation/perf/thunderx2-pmu.txt
Documentation/perf/thunderx2-pmu.rst
+13
-12
Documentation/perf/thunderx2-pmu.txt
Documentation/perf/thunderx2-pmu.rst
···
1
+
=============================================================
1
2
Cavium ThunderX2 SoC Performance Monitoring Unit (PMU UNCORE)
2
3
=============================================================
3
4
···
25
24
The driver does not support sampling, therefore "perf record" will not
26
25
work. Per-task perf sessions are also not supported.
27
26
28
-
Examples:
27
+
Examples::
29
28
30
-
# perf stat -a -e uncore_dmc_0/cnt_cycles/ sleep 1
29
+
# perf stat -a -e uncore_dmc_0/cnt_cycles/ sleep 1
31
30
32
-
# perf stat -a -e \
33
-
uncore_dmc_0/cnt_cycles/,\
34
-
uncore_dmc_0/data_transfers/,\
35
-
uncore_dmc_0/read_txns/,\
36
-
uncore_dmc_0/write_txns/ sleep 1
31
+
# perf stat -a -e \
32
+
uncore_dmc_0/cnt_cycles/,\
33
+
uncore_dmc_0/data_transfers/,\
34
+
uncore_dmc_0/read_txns/,\
35
+
uncore_dmc_0/write_txns/ sleep 1
37
36
38
-
# perf stat -a -e \
39
-
uncore_l3c_0/read_request/,\
40
-
uncore_l3c_0/read_hit/,\
41
-
uncore_l3c_0/inv_request/,\
42
-
uncore_l3c_0/inv_hit/ sleep 1
37
+
# perf stat -a -e \
38
+
uncore_l3c_0/read_request/,\
39
+
uncore_l3c_0/read_hit/,\
40
+
uncore_l3c_0/inv_request/,\
41
+
uncore_l3c_0/inv_hit/ sleep 1
+2
-1
Documentation/perf/xgene-pmu.txt
Documentation/perf/xgene-pmu.rst
+2
-1
Documentation/perf/xgene-pmu.txt
Documentation/perf/xgene-pmu.rst
···
1
+
================================================
1
2
APM X-Gene SoC Performance Monitoring Unit (PMU)
2
3
================================================
3
4
···
34
33
Each perf driver also provides a "cpumask" sysfs attribute, which contains a
35
34
single CPU ID of the processor which will be used to handle all the PMU events.
36
35
37
-
Example for perf tool use:
36
+
Example for perf tool use::
38
37
39
38
/ # perf list | grep -e l3c -e iob -e mcb -e mc
40
39
l3c0/ackq-full/ [Kernel PMU event]
+2
-2
MAINTAINERS
+2
-2
MAINTAINERS
···
1155
1155
M: Khuong Dinh <khuong@os.amperecomputing.com>
1156
1156
S: Supported
1157
1157
F: drivers/perf/xgene_pmu.c
1158
-
F: Documentation/perf/xgene-pmu.txt
1158
+
F: Documentation/perf/xgene-pmu.rst
1159
1159
F: Documentation/devicetree/bindings/perf/apm-xgene-pmu.txt
1160
1160
1161
1161
APTINA CAMERA SENSOR PLL
···
7262
7262
W: http://www.hisilicon.com
7263
7263
S: Supported
7264
7264
F: drivers/perf/hisilicon
7265
-
F: Documentation/perf/hisi-pmu.txt
7265
+
F: Documentation/perf/hisi-pmu.rst
7266
7266
7267
7267
HISILICON ROCE DRIVER
7268
7268
M: Lijun Ou <oulijun@huawei.com>
+1
-1
drivers/perf/qcom_l3_pmu.c
+1
-1
drivers/perf/qcom_l3_pmu.c
···
8
8
* the slices. User space needs to aggregate to individual counts to provide
9
9
* a global picture.
10
10
*
11
-
* See Documentation/perf/qcom_l3_pmu.txt for more details.
11
+
* See Documentation/perf/qcom_l3_pmu.rst for more details.
12
12
*
13
13
* Copyright (c) 2015-2017, The Linux Foundation. All rights reserved.
14
14
*/