Skip to content

L
linux
Commit d1c9f7d1 authored by Ingo Molnar's avatar Ingo Molnar
Browse files
Options

Merge tag 'perf-core-for-mingo-5.7-20200317' of... 

Merge tag 'perf-core-for-mingo-5.7-20200317' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:

perf record:

  Alexey Budankov:

  - Fix binding of AIO user space buffers to nodes

maps:

  Dominik b. Czarnota:

  - Fix off by one in strncpy() size argument.

  Arnaldo Carvalho de Melo:

  - Use strstarts() to look for Android libraries.

  Ian Rogers:

  - Give synthetic mmap events an inode generation.

man pages:

  Ian Rogers:

  - Set man page date to last git commit.

perf test:

  Ian Rogers:

  - Print if shell directory isn't present.

perf report:

  Jin Yao:

  - Fix no branch type statistics report issue.

perf expr:

  Jiri Olsa:

  - Fix copy/paste mistake

vendor events:

  Kan Liang:

  - Support metric constraints.

vendor events intel:

  Kan Liang:

  - Add NO_NMI_WATCHDOG metric constraint.

vendor events s390:

  Thomas Richter:

 - Add new deflate counters for IBM z15.

ARM cs-etm:

  Leo Yan:

  - Last branch improvements.

intel-pt:

  Adrian Hunter:

  - Update intel-pt.txt file with new location of the documentation.

  - Add Intel PT man page references.

  - Rename intel-pt.txt and put it in man page format.

perl scripting:

  Michael Petlan:

 - Add common_callchain to fix argument order.
Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>

Conflicts:
	tools/perf/util/map.c
parents 409e1a31 59a08b4b
Hide whitespace changes
Inline Side-by-side
Showing with 1340 additions and 1123 deletions
tools/perf/Documentation/Makefile
View file @ d1c9f7d1
......@@ -295,7 +295,10 @@ $(OUTPUT)%.1 $(OUTPUT)%.5 $(OUTPUT)%.7 : $(OUTPUT)%.xml
$(OUTPUT)%.xml : %.txt
$(QUIET_ASCIIDOC)$(RM) $@+ $@ && \
$(ASCIIDOC) -b docbook -d manpage \
$(ASCIIDOC_EXTRA) -aperf_version=$(PERF_VERSION) -o $@+ $< && \
$(ASCIIDOC_EXTRA) -aperf_version=$(PERF_VERSION) \
-aperf_date=$(shell git log -1 --pretty="format:%cd" \
--date=short $<) \
-o $@+ $< && \
mv $@+ $@
XSLT = docbook.xsl
......
tools/perf/Documentation/intel-pt.txt
View file @ d1c9f7d1
Intel Processor Trace
=====================
Overview
========
Intel Processor Trace (Intel PT) is an extension of Intel Architecture that
collects information about software execution such as control flow, execution
modes and timings and formats it into highly compressed binary packets.
Technical details are documented in the Intel 64 and IA-32 Architectures
Software Developer Manuals, Chapter 36 Intel Processor Trace.
Intel PT is first supported in Intel Core M and 5th generation Intel Core
processors that are based on the Intel micro-architecture code name Broadwell.
Trace data is collected by 'perf record' and stored within the perf.data file.
See below for options to 'perf record'.
Trace data must be 'decoded' which involves walking the object code and matching
the trace data packets. For example a TNT packet only tells whether a
conditional branch was taken or not taken, so to make use of that packet the
decoder must know precisely which instruction was being executed.
Decoding is done on-the-fly. The decoder outputs samples in the same format as
samples output by perf hardware events, for example as though the "instructions"
or "branches" events had been recorded. Presently 3 tools support this:
'perf script', 'perf report' and 'perf inject'. See below for more information
on using those tools.
The main distinguishing feature of Intel PT is that the decoder can determine
the exact flow of software execution. Intel PT can be used to understand why
and how did software get to a certain point, or behave a certain way. The
software does not have to be recompiled, so Intel PT works with debug or release
builds, however the executed images are needed - which makes use in JIT-compiled
environments, or with self-modified code, a challenge. Also symbols need to be
provided to make sense of addresses.
A limitation of Intel PT is that it produces huge amounts of trace data
(hundreds of megabytes per second per core) which takes a long time to decode,
for example two or three orders of magnitude longer than it took to collect.
Another limitation is the performance impact of tracing, something that will
vary depending on the use-case and architecture.
Quickstart
==========
It is important to start small. That is because it is easy to capture vastly
more data than can possibly be processed.
The simplest thing to do with Intel PT is userspace profiling of small programs.
Data is captured with 'perf record' e.g. to trace 'ls' userspace-only:
perf record -e intel_pt//u ls
And profiled with 'perf report' e.g.
perf report
To also trace kernel space presents a problem, namely kernel self-modifying
code. A fairly good kernel image is available in /proc/kcore but to get an
accurate image a copy of /proc/kcore needs to be made under the same conditions
as the data capture. A script perf-with-kcore can do that, but beware that the
script makes use of 'sudo' to copy /proc/kcore. If you have perf installed
locally from the source tree you can do:
~/libexec/perf-core/perf-with-kcore record pt_ls -e intel_pt// -- ls
which will create a directory named 'pt_ls' and put the perf.data file and
copies of /proc/kcore, /proc/kallsyms and /proc/modules into it. Then to use
'perf report' becomes:
~/libexec/perf-core/perf-with-kcore report pt_ls
Because samples are synthesized after-the-fact, the sampling period can be
selected for reporting. e.g. sample every microsecond
~/libexec/perf-core/perf-with-kcore report pt_ls --itrace=i1usge
See the sections below for more information about the --itrace option.
Beware the smaller the period, the more samples that are produced, and the
longer it takes to process them.
Also note that the coarseness of Intel PT timing information will start to
distort the statistical value of the sampling as the sampling period becomes
smaller.
To represent software control flow, "branches" samples are produced. By default
a branch sample is synthesized for every single branch. To get an idea what
data is available you can use the 'perf script' tool with all itrace sampling
options, which will list all the samples.
perf record -e intel_pt//u ls
perf script --itrace=ibxwpe
An interesting field that is not printed by default is 'flags' which can be
displayed as follows:
perf script --itrace=ibxwpe -F+flags
The flags are "bcrosyiABEx" which stand for branch, call, return, conditional,
system, asynchronous, interrupt, transaction abort, trace begin, trace end, and
in transaction, respectively.
Another interesting field that is not printed by default is 'ipc' which can be
displayed as follows:
perf script --itrace=be -F+ipc
There are two ways that instructions-per-cycle (IPC) can be calculated depending
on the recording.
If the 'cyc' config term (see config terms section below) was used, then IPC is
calculated using the cycle count from CYC packets, otherwise MTC packets are
used - refer to the 'mtc' config term. When MTC is used, however, the values
are less accurate because the timing is less accurate.
Because Intel PT does not update the cycle count on every branch or instruction,
the values will often be zero. When there are values, they will be the number
of instructions and number of cycles since the last update, and thus represent
the average IPC since the last IPC for that event type. Note IPC for "branches"
events is calculated separately from IPC for "instructions" events.
Also note that the IPC instruction count may or may not include the current
instruction. If the cycle count is associated with an asynchronous branch
(e.g. page fault or interrupt), then the instruction count does not include the
current instruction, otherwise it does. That is consistent with whether or not
that instruction has retired when the cycle count is updated.
Another note, in the case of "branches" events, non-taken branches are not
presently sampled, so IPC values for them do not appear e.g. a CYC packet with a
TNT packet that starts with a non-taken branch. To see every possible IPC
value, "instructions" events can be used e.g. --itrace=i0ns
While it is possible to create scripts to analyze the data, an alternative
approach is available to export the data to a sqlite or postgresql database.
Refer to script export-to-sqlite.py or export-to-postgresql.py for more details,
and to script exported-sql-viewer.py for an example of using the database.
There is also script intel-pt-events.py which provides an example of how to
unpack the raw data for power events and PTWRITE.
As mentioned above, it is easy to capture too much data. One way to limit the
data captured is to use 'snapshot' mode which is explained further below.
Refer to 'new snapshot option' and 'Intel PT modes of operation' further below.
Another problem that will be experienced is decoder errors. They can be caused
by inability to access the executed image, self-modified or JIT-ed code, or the
inability to match side-band information (such as context switches and mmaps)
which results in the decoder not knowing what code was executed.
There is also the problem of perf not being able to copy the data fast enough,
resulting in data lost because the buffer was full. See 'Buffer handling' below
for more details.
perf record
===========
new event
---------
The Intel PT kernel driver creates a new PMU for Intel PT. PMU events are
selected by providing the PMU name followed by the "config" separated by slashes.
An enhancement has been made to allow default "config" e.g. the option
-e intel_pt//
will use a default config value. Currently that is the same as
-e intel_pt/tsc,noretcomp=0/
which is the same as
-e intel_pt/tsc=1,noretcomp=0/
Note there are now new config terms - see section 'config terms' further below.
The config terms are listed in /sys/devices/intel_pt/format. They are bit
fields within the config member of the struct perf_event_attr which is
passed to the kernel by the perf_event_open system call. They correspond to bit
fields in the IA32_RTIT_CTL MSR. Here is a list of them and their definitions:
$ grep -H . /sys/bus/event_source/devices/intel_pt/format/*
/sys/bus/event_source/devices/intel_pt/format/cyc:config:1
/sys/bus/event_source/devices/intel_pt/format/cyc_thresh:config:19-22
/sys/bus/event_source/devices/intel_pt/format/mtc:config:9
/sys/bus/event_source/devices/intel_pt/format/mtc_period:config:14-17
/sys/bus/event_source/devices/intel_pt/format/noretcomp:config:11
/sys/bus/event_source/devices/intel_pt/format/psb_period:config:24-27
/sys/bus/event_source/devices/intel_pt/format/tsc:config:10
Note that the default config must be overridden for each term i.e.
-e intel_pt/noretcomp=0/
is the same as:
-e intel_pt/tsc=1,noretcomp=0/
So, to disable TSC packets use:
-e intel_pt/tsc=0/
It is also possible to specify the config value explicitly:
-e intel_pt/config=0x400/
Note that, as with all events, the event is suffixed with event modifiers:
u userspace
k kernel
h hypervisor
G guest
H host
p precise ip
'h', 'G' and 'H' are for virtualization which is not supported by Intel PT.
'p' is also not relevant to Intel PT. So only options 'u' and 'k' are
meaningful for Intel PT.
perf_event_attr is displayed if the -vv option is used e.g.
------------------------------------------------------------
perf_event_attr:
type 6
size 112
config 0x400
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|CPU|IDENTIFIER
read_format ID
disabled 1
inherit 1
exclude_kernel 1
exclude_hv 1
enable_on_exec 1
sample_id_all 1
------------------------------------------------------------
sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
------------------------------------------------------------
config terms
------------
The June 2015 version of Intel 64 and IA-32 Architectures Software Developer
Manuals, Chapter 36 Intel Processor Trace, defined new Intel PT features.
Some of the features are reflect in new config terms. All the config terms are
described below.
tsc Always supported. Produces TSC timestamp packets to provide
timing information. In some cases it is possible to decode
without timing information, for example a per-thread context
that does not overlap executable memory maps.
The default config selects tsc (i.e. tsc=1).
noretcomp Always supported. Disables "return compression" so a TIP packet
is produced when a function returns. Causes more packets to be
produced but might make decoding more reliable.
The default config does not select noretcomp (i.e. noretcomp=0).
psb_period Allows the frequency of PSB packets to be specified.
The PSB packet is a synchronization packet that provides a
starting point for decoding or recovery from errors.
Support for psb_period is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/psb_cyc
which contains "1" if the feature is supported and "0"
otherwise.
Valid values are given by:
/sys/bus/event_source/devices/intel_pt/caps/psb_periods
which contains a hexadecimal value, the bits of which represent
valid values e.g. bit 2 set means value 2 is valid.
The psb_period value is converted to the approximate number of
trace bytes between PSB packets as:
2 ^ (value + 11)
e.g. value 3 means 16KiB bytes between PSBs
If an invalid value is entered, the error message
will give a list of valid values e.g.
$ perf record -e intel_pt/psb_period=15/u uname
Invalid psb_period for intel_pt. Valid values are: 0-5
If MTC packets are selected, the default config selects a value
of 3 (i.e. psb_period=3) or the nearest lower value that is
supported (0 is always supported). Otherwise the default is 0.
If decoding is expected to be reliable and the buffer is large
then a large PSB period can be used.
Because a TSC packet is produced with PSB, the PSB period can
also affect the granularity to timing information in the absence
of MTC or CYC.
mtc Produces MTC timing packets.
MTC packets provide finer grain timestamp information than TSC
packets. MTC packets record time using the hardware crystal
clock (CTC) which is related to TSC packets using a TMA packet.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/mtc
which contains "1" if the feature is supported and
"0" otherwise.
The frequency of MTC packets can also be specified - see
mtc_period below.
mtc_period Specifies how frequently MTC packets are produced - see mtc
above for how to determine if MTC packets are supported.
Valid values are given by:
/sys/bus/event_source/devices/intel_pt/caps/mtc_periods
which contains a hexadecimal value, the bits of which represent
valid values e.g. bit 2 set means value 2 is valid.
The mtc_period value is converted to the MTC frequency as:
CTC-frequency / (2 ^ value)
e.g. value 3 means one eighth of CTC-frequency
Where CTC is the hardware crystal clock, the frequency of which
can be related to TSC via values provided in cpuid leaf 0x15.
If an invalid value is entered, the error message
will give a list of valid values e.g.
$ perf record -e intel_pt/mtc_period=15/u uname
Invalid mtc_period for intel_pt. Valid values are: 0,3,6,9
The default value is 3 or the nearest lower value
that is supported (0 is always supported).
cyc Produces CYC timing packets.
CYC packets provide even finer grain timestamp information than
MTC and TSC packets. A CYC packet contains the number of CPU
cycles since the last CYC packet. Unlike MTC and TSC packets,
CYC packets are only sent when another packet is also sent.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/psb_cyc
which contains "1" if the feature is supported and
"0" otherwise.
The number of CYC packets produced can be reduced by specifying
a threshold - see cyc_thresh below.
cyc_thresh Specifies how frequently CYC packets are produced - see cyc
above for how to determine if CYC packets are supported.
Valid cyc_thresh values are given by:
/sys/bus/event_source/devices/intel_pt/caps/cycle_thresholds
which contains a hexadecimal value, the bits of which represent
valid values e.g. bit 2 set means value 2 is valid.
The cyc_thresh value represents the minimum number of CPU cycles
that must have passed before a CYC packet can be sent. The
number of CPU cycles is:
2 ^ (value - 1)
e.g. value 4 means 8 CPU cycles must pass before a CYC packet
can be sent. Note a CYC packet is still only sent when another
packet is sent, not at, e.g. every 8 CPU cycles.
If an invalid value is entered, the error message
will give a list of valid values e.g.
$ perf record -e intel_pt/cyc,cyc_thresh=15/u uname
Invalid cyc_thresh for intel_pt. Valid values are: 0-12
CYC packets are not requested by default.
pt Specifies pass-through which enables the 'branch' config term.
The default config selects 'pt' if it is available, so a user will
never need to specify this term.
branch Enable branch tracing. Branch tracing is enabled by default so to
disable branch tracing use 'branch=0'.
The default config selects 'branch' if it is available.
ptw Enable PTWRITE packets which are produced when a ptwrite instruction
is executed.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/ptwrite
which contains "1" if the feature is supported and
"0" otherwise.
fup_on_ptw Enable a FUP packet to follow the PTWRITE packet. The FUP packet
provides the address of the ptwrite instruction. In the absence of
fup_on_ptw, the decoder will use the address of the previous branch
if branch tracing is enabled, otherwise the address will be zero.
Note that fup_on_ptw will work even when branch tracing is disabled.
pwr_evt Enable power events. The power events provide information about
changes to the CPU C-state.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/power_event_trace
which contains "1" if the feature is supported and
"0" otherwise.
AUX area sampling option
------------------------
To select Intel PT "sampling" the AUX area sampling option can be used:
--aux-sample
Optionally it can be followed by the sample size in bytes e.g.
--aux-sample=8192
In addition, the Intel PT event to sample must be defined e.g.
-e intel_pt//u
Samples on other events will be created containing Intel PT data e.g. the
following will create Intel PT samples on the branch-misses event, note the
events must be grouped using {}:
perf record --aux-sample -e '{intel_pt//u,branch-misses:u}'
An alternative to '--aux-sample' is to add the config term 'aux-sample-size' to
events. In this case, the grouping is implied e.g.
perf record -e intel_pt//u -e branch-misses/aux-sample-size=8192/u
is the same as:
perf record -e '{intel_pt//u,branch-misses/aux-sample-size=8192/u}'
but allows for also using an address filter e.g.:
perf record -e intel_pt//u --filter 'filter * @/bin/ls' -e branch-misses/aux-sample-size=8192/u -- ls
It is important to select a sample size that is big enough to contain at least
one PSB packet. If not a warning will be displayed:
Intel PT sample size (%zu) may be too small for PSB period (%zu)
The calculation used for that is: if sample_size <= psb_period + 256 display the
warning. When sampling is used, psb_period defaults to 0 (2KiB).
The default sample size is 4KiB.
The sample size is passed in aux_sample_size in struct perf_event_attr. The
sample size is limited by the maximum event size which is 64KiB. It is
difficult to know how big the event might be without the trace sample attached,
but the tool validates that the sample size is not greater than 60KiB.
new snapshot option
-------------------
The difference between full trace and snapshot from the kernel's perspective is
that in full trace we don't overwrite trace data that the user hasn't collected
yet (and indicated that by advancing aux_tail), whereas in snapshot mode we let
the trace run and overwrite older data in the buffer so that whenever something
interesting happens, we can stop it and grab a snapshot of what was going on
around that interesting moment.
To select snapshot mode a new option has been added:
-S
Optionally it can be followed by the snapshot size e.g.
-S0x100000
The default snapshot size is the auxtrace mmap size. If neither auxtrace mmap size
nor snapshot size is specified, then the default is 4MiB for privileged users
(or if /proc/sys/kernel/perf_event_paranoid < 0), 128KiB for unprivileged users.
If an unprivileged user does not specify mmap pages, the mmap pages will be
reduced as described in the 'new auxtrace mmap size option' section below.
The snapshot size is displayed if the option -vv is used e.g.
Intel PT snapshot size: %zu
new auxtrace mmap size option
---------------------------
Intel PT buffer size is specified by an addition to the -m option e.g.
-m,16
selects a buffer size of 16 pages i.e. 64KiB.
Note that the existing functionality of -m is unchanged. The auxtrace mmap size
is specified by the optional addition of a comma and the value.
The default auxtrace mmap size for Intel PT is 4MiB/page_size for privileged users
(or if /proc/sys/kernel/perf_event_paranoid < 0), 128KiB for unprivileged users.
If an unprivileged user does not specify mmap pages, the mmap pages will be
reduced from the default 512KiB/page_size to 256KiB/page_size, otherwise the
user is likely to get an error as they exceed their mlock limit (Max locked
memory as shown in /proc/self/limits). Note that perf does not count the first
512KiB (actually /proc/sys/kernel/perf_event_mlock_kb minus 1 page) per cpu
against the mlock limit so an unprivileged user is allowed 512KiB per cpu plus
their mlock limit (which defaults to 64KiB but is not multiplied by the number
of cpus).
In full-trace mode, powers of two are allowed for buffer size, with a minimum
size of 2 pages. In snapshot mode or sampling mode, it is the same but the
minimum size is 1 page.
The mmap size and auxtrace mmap size are displayed if the -vv option is used e.g.
mmap length 528384
auxtrace mmap length 4198400
Intel PT modes of operation
---------------------------
Intel PT can be used in 2 modes:
full-trace mode
sample mode
snapshot mode
Full-trace mode traces continuously e.g.
perf record -e intel_pt//u uname
Sample mode attaches a Intel PT sample to other events e.g.
perf record --aux-sample -e intel_pt//u -e branch-misses:u
Snapshot mode captures the available data when a signal is sent e.g.
perf record -v -e intel_pt//u -S ./loopy 1000000000 &
[1] 11435
kill -USR2 11435
Recording AUX area tracing snapshot
Note that the signal sent is SIGUSR2.
Note that "Recording AUX area tracing snapshot" is displayed because the -v
option is used.
The 2 modes cannot be used together.
Buffer handling
---------------
There may be buffer limitations (i.e. single ToPa entry) which means that actual
buffer sizes are limited to powers of 2 up to 4MiB (MAX_ORDER). In order to
provide other sizes, and in particular an arbitrarily large size, multiple
buffers are logically concatenated. However an interrupt must be used to switch
between buffers. That has two potential problems:
a) the interrupt may not be handled in time so that the current buffer
becomes full and some trace data is lost.
b) the interrupts may slow the system and affect the performance
results.
If trace data is lost, the driver sets 'truncated' in the PERF_RECORD_AUX event
which the tools report as an error.
In full-trace mode, the driver waits for data to be copied out before allowing
the (logical) buffer to wrap-around. If data is not copied out quickly enough,
again 'truncated' is set in the PERF_RECORD_AUX event. If the driver has to
wait, the intel_pt event gets disabled. Because it is difficult to know when
that happens, perf tools always re-enable the intel_pt event after copying out
data.
Intel PT and build ids
----------------------
By default "perf record" post-processes the event stream to find all build ids
for executables for all addresses sampled. Deliberately, Intel PT is not
decoded for that purpose (it would take too long). Instead the build ids for
all executables encountered (due to mmap, comm or task events) are included
in the perf.data file.
To see buildids included in the perf.data file use the command:
perf buildid-list
If the perf.data file contains Intel PT data, that is the same as:
perf buildid-list --with-hits
Snapshot mode and event disabling
---------------------------------
In order to make a snapshot, the intel_pt event is disabled using an IOCTL,
namely PERF_EVENT_IOC_DISABLE. However doing that can also disable the
collection of side-band information. In order to prevent that, a dummy
software event has been introduced that permits tracking events (like mmaps) to
continue to be recorded while intel_pt is disabled. That is important to ensure
there is complete side-band information to allow the decoding of subsequent
snapshots.
A test has been created for that. To find the test:
perf test list
...
23: Test using a dummy software event to keep tracking
To run the test:
perf test 23
23: Test using a dummy software event to keep tracking : Ok
perf record modes (nothing new here)
------------------------------------
perf record essentially operates in one of three modes:
per thread
per cpu
workload only
"per thread" mode is selected by -t or by --per-thread (with -p or -u or just a
workload).
"per cpu" is selected by -C or -a.
"workload only" mode is selected by not using the other options but providing a
command to run (i.e. the workload).
In per-thread mode an exact list of threads is traced. There is no inheritance.
Each thread has its own event buffer.
In per-cpu mode all processes (or processes from the selected cgroup i.e. -G
option, or processes selected with -p or -u) are traced. Each cpu has its own
buffer. Inheritance is allowed.
In workload-only mode, the workload is traced but with per-cpu buffers.
Inheritance is allowed. Note that you can now trace a workload in per-thread
mode by using the --per-thread option.
Privileged vs non-privileged users
----------------------------------
Unless /proc/sys/kernel/perf_event_paranoid is set to -1, unprivileged users
have memory limits imposed upon them. That affects what buffer sizes they can
have as outlined above.
The v4.2 kernel introduced support for a context switch metadata event,
PERF_RECORD_SWITCH, which allows unprivileged users to see when their processes
are scheduled out and in, just not by whom, which is left for the
PERF_RECORD_SWITCH_CPU_WIDE, that is only accessible in system wide context,
which in turn requires CAP_SYS_ADMIN.
Please see the 45ac1403f564 ("perf: Add PERF_RECORD_SWITCH to indicate context
switches") commit, that introduces these metadata events for further info.
When working with kernels < v4.2, the following considerations must be taken,
as the sched:sched_switch tracepoints will be used to receive such information:
Unless /proc/sys/kernel/perf_event_paranoid is set to -1, unprivileged users are
not permitted to use tracepoints which means there is insufficient side-band
information to decode Intel PT in per-cpu mode, and potentially workload-only
mode too if the workload creates new processes.
Note also, that to use tracepoints, read-access to debugfs is required. So if
debugfs is not mounted or the user does not have read-access, it will again not
be possible to decode Intel PT in per-cpu mode.
sched_switch tracepoint
-----------------------
The sched_switch tracepoint is used to provide side-band data for Intel PT
decoding in kernels where the PERF_RECORD_SWITCH metadata event isn't
available.
The sched_switch events are automatically added. e.g. the second event shown
below:
$ perf record -vv -e intel_pt//u uname
------------------------------------------------------------
perf_event_attr:
type 6
size 112
config 0x400
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|CPU|IDENTIFIER
read_format ID
disabled 1
inherit 1
exclude_kernel 1
exclude_hv 1
enable_on_exec 1
sample_id_all 1
------------------------------------------------------------
sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
------------------------------------------------------------
perf_event_attr:
type 2
size 112
config 0x108
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|CPU|PERIOD|RAW|IDENTIFIER
read_format ID
inherit 1
sample_id_all 1
exclude_guest 1
------------------------------------------------------------
sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 3 group_fd -1 flags 0x8
------------------------------------------------------------
perf_event_attr:
type 1
size 112
config 0x9
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|IDENTIFIER
read_format ID
disabled 1
inherit 1
exclude_kernel 1
exclude_hv 1
mmap 1
comm 1
enable_on_exec 1
task 1
sample_id_all 1
mmap2 1
comm_exec 1
------------------------------------------------------------
sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
mmap size 528384B
AUX area mmap length 4194304
perf event ring buffer mmapped per cpu
Synthesizing auxtrace information
Linux
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.042 MB perf.data ]
Note, the sched_switch event is only added if the user is permitted to use it
and only in per-cpu mode.
Note also, the sched_switch event is only added if TSC packets are requested.
That is because, in the absence of timing information, the sched_switch events
cannot be matched against the Intel PT trace.
perf script
===========
By default, perf script will decode trace data found in the perf.data file.
This can be further controlled by new option --itrace.
New --itrace option
-------------------
Having no option is the same as
--itrace
which, in turn, is the same as
--itrace=cepwx
The letters are:
i synthesize "instructions" events
b synthesize "branches" events
x synthesize "transactions" events
w synthesize "ptwrite" events
p synthesize "power" events
c synthesize branches events (calls only)
r synthesize branches events (returns only)
e synthesize tracing error events
d create a debug log
g synthesize a call chain (use with i or x)
l synthesize last branch entries (use with i or x)
s skip initial number of events
"Instructions" events look like they were recorded by "perf record -e
instructions".
"Branches" events look like they were recorded by "perf record -e branches". "c"
and "r" can be combined to get calls and returns.
"Transactions" events correspond to the start or end of transactions. The
'flags' field can be used in perf script to determine whether the event is a
tranasaction start, commit or abort.
Note that "instructions", "branches" and "transactions" events depend on code
flow packets which can be disabled by using the config term "branch=0". Refer
to the config terms section above.
"ptwrite" events record the payload of the ptwrite instruction and whether
"fup_on_ptw" was used. "ptwrite" events depend on PTWRITE packets which are
recorded only if the "ptw" config term was used. Refer to the config terms
section above. perf script "synth" field displays "ptwrite" information like
this: "ip: 0 payload: 0x123456789abcdef0" where "ip" is 1 if "fup_on_ptw" was
used.
"Power" events correspond to power event packets and CBR (core-to-bus ratio)
packets. While CBR packets are always recorded when tracing is enabled, power
event packets are recorded only if the "pwr_evt" config term was used. Refer to
the config terms section above. The power events record information about
C-state changes, whereas CBR is indicative of CPU frequency. perf script
"event,synth" fields display information like this:
cbr: cbr: 22 freq: 2189 MHz (200%)
mwait: hints: 0x60 extensions: 0x1
pwre: hw: 0 cstate: 2 sub-cstate: 0
exstop: ip: 1
pwrx: deepest cstate: 2 last cstate: 2 wake reason: 0x4
Where:
"cbr" includes the frequency and the percentage of maximum non-turbo
"mwait" shows mwait hints and extensions
"pwre" shows C-state transitions (to a C-state deeper than C0) and
whether initiated by hardware
"exstop" indicates execution stopped and whether the IP was recorded
exactly,
"pwrx" indicates return to C0
For more details refer to the Intel 64 and IA-32 Architectures Software
Developer Manuals.
Error events show where the decoder lost the trace. Error events
are quite important. Users must know if what they are seeing is a complete
picture or not.
The "d" option will cause the creation of a file "intel_pt.log" containing all
decoded packets and instructions. Note that this option slows down the decoder
and that the resulting file may be very large.
In addition, the period of the "instructions" event can be specified. e.g.
--itrace=i10us
sets the period to 10us i.e. one instruction sample is synthesized for each 10
microseconds of trace. Alternatives to "us" are "ms" (milliseconds),
"ns" (nanoseconds), "t" (TSC ticks) or "i" (instructions).
"ms", "us" and "ns" are converted to TSC ticks.
The timing information included with Intel PT does not give the time of every
instruction. Consequently, for the purpose of sampling, the decoder estimates
the time since the last timing packet based on 1 tick per instruction. The time
on the sample is *not* adjusted and reflects the last known value of TSC.
For Intel PT, the default period is 100us.
Setting it to a zero period means "as often as possible".
In the case of Intel PT that is the same as a period of 1 and a unit of
'instructions' (i.e. --itrace=i1i).
Also the call chain size (default 16, max. 1024) for instructions or
transactions events can be specified. e.g.
--itrace=ig32
--itrace=xg32
Also the number of last branch entries (default 64, max. 1024) for instructions or
transactions events can be specified. e.g.
--itrace=il10
--itrace=xl10
Note that last branch entries are cleared for each sample, so there is no overlap
from one sample to the next.
To disable trace decoding entirely, use the option --no-itrace.
It is also possible to skip events generated (instructions, branches, transactions)
at the beginning. This is useful to ignore initialization code.
--itrace=i0nss1000000
skips the first million instructions.
dump option
-----------
perf script has an option (-D) to "dump" the events i.e. display the binary
data.
When -D is used, Intel PT packets are displayed. The packet decoder does not
pay attention to PSB packets, but just decodes the bytes - so the packets seen
by the actual decoder may not be identical in places where the data is corrupt.
One example of that would be when the buffer-switching interrupt has been too
slow, and the buffer has been filled completely. In that case, the last packet
in the buffer might be truncated and immediately followed by a PSB as the trace
continues in the next buffer.
To disable the display of Intel PT packets, combine the -D option with
--no-itrace.
perf report
===========
By default, perf report will decode trace data found in the perf.data file.
This can be further controlled by new option --itrace exactly the same as
perf script, with the exception that the default is --itrace=igxe.
perf inject
===========
perf inject also accepts the --itrace option in which case tracing data is
removed and replaced with the synthesized events. e.g.
perf inject --itrace -i perf.data -o perf.data.new
Below is an example of using Intel PT with autofdo. It requires autofdo
(https://github.com/google/autofdo) and gcc version 5. The bubble
sort example is from the AutoFDO tutorial (https://gcc.gnu.org/wiki/AutoFDO/Tutorial)
amended to take the number of elements as a parameter.
$ gcc-5 -O3 sort.c -o sort_optimized
$ ./sort_optimized 30000
Bubble sorting array of 30000 elements
2254 ms
$ cat ~/.perfconfig
[intel-pt]
mispred-all = on
$ perf record -e intel_pt//u ./sort 3000
Bubble sorting array of 3000 elements
58 ms
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 3.939 MB perf.data ]
$ perf inject -i perf.data -o inj --itrace=i100usle --strip
$ ./create_gcov --binary=./sort --profile=inj --gcov=sort.gcov -gcov_version=1
$ gcc-5 -O3 -fauto-profile=sort.gcov sort.c -o sort_autofdo
$ ./sort_autofdo 30000
Bubble sorting array of 30000 elements
2155 ms
Note there is currently no advantage to using Intel PT instead of LBR, but
that may change in the future if greater use is made of the data.
PEBS via Intel PT
=================
Some hardware has the feature to redirect PEBS records to the Intel PT trace.
Recording is selected by using the aux-output config term e.g.
perf record -c 10000 -e '{intel_pt/branch=0/,cycles/aux-output/ppp}' uname
Note that currently, software only supports redirecting at most one PEBS event.
To display PEBS events from the Intel PT trace, use the itrace 'o' option e.g.
perf script --itrace=oe
Documentation for support for Intel Processor Trace within perf tools' has moved to file perf-intel-pt.txt
tools/perf/Documentation/perf-inject.txt
View file @ d1c9f7d1
......@@ -66,4 +66,5 @@ include::itrace.txt[]
SEE ALSO
--------
linkperf:perf-record[1], linkperf:perf-report[1], linkperf:perf-archive[1]
linkperf:perf-record[1], linkperf:perf-report[1], linkperf:perf-archive[1],
linkperf:perf-intel-pt[1]
tools/perf/Documentation/perf-intel-pt.txt 0 → 100644
View file @ d1c9f7d1
perf-intel-pt(1)
================
NAME
----
perf-intel-pt - Support for Intel Processor Trace within perf tools
SYNOPSIS
--------
[verse]
'perf record' -e intel_pt//
DESCRIPTION
-----------
Intel Processor Trace (Intel PT) is an extension of Intel Architecture that
collects information about software execution such as control flow, execution
modes and timings and formats it into highly compressed binary packets.
Technical details are documented in the Intel 64 and IA-32 Architectures
Software Developer Manuals, Chapter 36 Intel Processor Trace.
Intel PT is first supported in Intel Core M and 5th generation Intel Core
processors that are based on the Intel micro-architecture code name Broadwell.
Trace data is collected by 'perf record' and stored within the perf.data file.
See below for options to 'perf record'.
Trace data must be 'decoded' which involves walking the object code and matching
the trace data packets. For example a TNT packet only tells whether a
conditional branch was taken or not taken, so to make use of that packet the
decoder must know precisely which instruction was being executed.
Decoding is done on-the-fly. The decoder outputs samples in the same format as
samples output by perf hardware events, for example as though the "instructions"
or "branches" events had been recorded. Presently 3 tools support this:
'perf script', 'perf report' and 'perf inject'. See below for more information
on using those tools.
The main distinguishing feature of Intel PT is that the decoder can determine
the exact flow of software execution. Intel PT can be used to understand why
and how did software get to a certain point, or behave a certain way. The
software does not have to be recompiled, so Intel PT works with debug or release
builds, however the executed images are needed - which makes use in JIT-compiled
environments, or with self-modified code, a challenge. Also symbols need to be
provided to make sense of addresses.
A limitation of Intel PT is that it produces huge amounts of trace data
(hundreds of megabytes per second per core) which takes a long time to decode,
for example two or three orders of magnitude longer than it took to collect.
Another limitation is the performance impact of tracing, something that will
vary depending on the use-case and architecture.
Quickstart
----------
It is important to start small. That is because it is easy to capture vastly
more data than can possibly be processed.
The simplest thing to do with Intel PT is userspace profiling of small programs.
Data is captured with 'perf record' e.g. to trace 'ls' userspace-only:
perf record -e intel_pt//u ls
And profiled with 'perf report' e.g.
perf report
To also trace kernel space presents a problem, namely kernel self-modifying
code. A fairly good kernel image is available in /proc/kcore but to get an
accurate image a copy of /proc/kcore needs to be made under the same conditions
as the data capture. A script perf-with-kcore can do that, but beware that the
script makes use of 'sudo' to copy /proc/kcore. If you have perf installed
locally from the source tree you can do:
~/libexec/perf-core/perf-with-kcore record pt_ls -e intel_pt// -- ls
which will create a directory named 'pt_ls' and put the perf.data file and
copies of /proc/kcore, /proc/kallsyms and /proc/modules into it. Then to use
'perf report' becomes:
~/libexec/perf-core/perf-with-kcore report pt_ls
Because samples are synthesized after-the-fact, the sampling period can be
selected for reporting. e.g. sample every microsecond
~/libexec/perf-core/perf-with-kcore report pt_ls --itrace=i1usge
See the sections below for more information about the --itrace option.
Beware the smaller the period, the more samples that are produced, and the
longer it takes to process them.
Also note that the coarseness of Intel PT timing information will start to
distort the statistical value of the sampling as the sampling period becomes
smaller.
To represent software control flow, "branches" samples are produced. By default
a branch sample is synthesized for every single branch. To get an idea what
data is available you can use the 'perf script' tool with all itrace sampling
options, which will list all the samples.
perf record -e intel_pt//u ls
perf script --itrace=ibxwpe
An interesting field that is not printed by default is 'flags' which can be
displayed as follows:
perf script --itrace=ibxwpe -F+flags
The flags are "bcrosyiABEx" which stand for branch, call, return, conditional,
system, asynchronous, interrupt, transaction abort, trace begin, trace end, and
in transaction, respectively.
Another interesting field that is not printed by default is 'ipc' which can be
displayed as follows:
perf script --itrace=be -F+ipc
There are two ways that instructions-per-cycle (IPC) can be calculated depending
on the recording.
If the 'cyc' config term (see config terms section below) was used, then IPC is
calculated using the cycle count from CYC packets, otherwise MTC packets are
used - refer to the 'mtc' config term. When MTC is used, however, the values
are less accurate because the timing is less accurate.
Because Intel PT does not update the cycle count on every branch or instruction,
the values will often be zero. When there are values, they will be the number
of instructions and number of cycles since the last update, and thus represent
the average IPC since the last IPC for that event type. Note IPC for "branches"
events is calculated separately from IPC for "instructions" events.
Also note that the IPC instruction count may or may not include the current
instruction. If the cycle count is associated with an asynchronous branch
(e.g. page fault or interrupt), then the instruction count does not include the
current instruction, otherwise it does. That is consistent with whether or not
that instruction has retired when the cycle count is updated.
Another note, in the case of "branches" events, non-taken branches are not
presently sampled, so IPC values for them do not appear e.g. a CYC packet with a
TNT packet that starts with a non-taken branch. To see every possible IPC
value, "instructions" events can be used e.g. --itrace=i0ns
While it is possible to create scripts to analyze the data, an alternative
approach is available to export the data to a sqlite or postgresql database.
Refer to script export-to-sqlite.py or export-to-postgresql.py for more details,
and to script exported-sql-viewer.py for an example of using the database.
There is also script intel-pt-events.py which provides an example of how to
unpack the raw data for power events and PTWRITE.
As mentioned above, it is easy to capture too much data. One way to limit the
data captured is to use 'snapshot' mode which is explained further below.
Refer to 'new snapshot option' and 'Intel PT modes of operation' further below.
Another problem that will be experienced is decoder errors. They can be caused
by inability to access the executed image, self-modified or JIT-ed code, or the
inability to match side-band information (such as context switches and mmaps)
which results in the decoder not knowing what code was executed.
There is also the problem of perf not being able to copy the data fast enough,
resulting in data lost because the buffer was full. See 'Buffer handling' below
for more details.
perf record
-----------
new event
~~~~~~~~~
The Intel PT kernel driver creates a new PMU for Intel PT. PMU events are
selected by providing the PMU name followed by the "config" separated by slashes.
An enhancement has been made to allow default "config" e.g. the option
-e intel_pt//
will use a default config value. Currently that is the same as
-e intel_pt/tsc,noretcomp=0/
which is the same as
-e intel_pt/tsc=1,noretcomp=0/
Note there are now new config terms - see section 'config terms' further below.
The config terms are listed in /sys/devices/intel_pt/format. They are bit
fields within the config member of the struct perf_event_attr which is
passed to the kernel by the perf_event_open system call. They correspond to bit
fields in the IA32_RTIT_CTL MSR. Here is a list of them and their definitions:
$ grep -H . /sys/bus/event_source/devices/intel_pt/format/*
/sys/bus/event_source/devices/intel_pt/format/cyc:config:1
/sys/bus/event_source/devices/intel_pt/format/cyc_thresh:config:19-22
/sys/bus/event_source/devices/intel_pt/format/mtc:config:9
/sys/bus/event_source/devices/intel_pt/format/mtc_period:config:14-17
/sys/bus/event_source/devices/intel_pt/format/noretcomp:config:11
/sys/bus/event_source/devices/intel_pt/format/psb_period:config:24-27
/sys/bus/event_source/devices/intel_pt/format/tsc:config:10
Note that the default config must be overridden for each term i.e.
-e intel_pt/noretcomp=0/
is the same as:
-e intel_pt/tsc=1,noretcomp=0/
So, to disable TSC packets use:
-e intel_pt/tsc=0/
It is also possible to specify the config value explicitly:
-e intel_pt/config=0x400/
Note that, as with all events, the event is suffixed with event modifiers:
u userspace
k kernel
h hypervisor
G guest
H host
p precise ip
'h', 'G' and 'H' are for virtualization which is not supported by Intel PT.
'p' is also not relevant to Intel PT. So only options 'u' and 'k' are
meaningful for Intel PT.
perf_event_attr is displayed if the -vv option is used e.g.
------------------------------------------------------------
perf_event_attr:
type 6
size 112
config 0x400
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|CPU|IDENTIFIER
read_format ID
disabled 1
inherit 1
exclude_kernel 1
exclude_hv 1
enable_on_exec 1
sample_id_all 1
------------------------------------------------------------
sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
------------------------------------------------------------
config terms
~~~~~~~~~~~~
The June 2015 version of Intel 64 and IA-32 Architectures Software Developer
Manuals, Chapter 36 Intel Processor Trace, defined new Intel PT features.
Some of the features are reflect in new config terms. All the config terms are
described below.
tsc Always supported. Produces TSC timestamp packets to provide
timing information. In some cases it is possible to decode
without timing information, for example a per-thread context
that does not overlap executable memory maps.
The default config selects tsc (i.e. tsc=1).
noretcomp Always supported. Disables "return compression" so a TIP packet
is produced when a function returns. Causes more packets to be
produced but might make decoding more reliable.
The default config does not select noretcomp (i.e. noretcomp=0).
psb_period Allows the frequency of PSB packets to be specified.
The PSB packet is a synchronization packet that provides a
starting point for decoding or recovery from errors.
Support for psb_period is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/psb_cyc
which contains "1" if the feature is supported and "0"
otherwise.
Valid values are given by:
/sys/bus/event_source/devices/intel_pt/caps/psb_periods
which contains a hexadecimal value, the bits of which represent
valid values e.g. bit 2 set means value 2 is valid.
The psb_period value is converted to the approximate number of
trace bytes between PSB packets as:
2 ^ (value + 11)
e.g. value 3 means 16KiB bytes between PSBs
If an invalid value is entered, the error message
will give a list of valid values e.g.
$ perf record -e intel_pt/psb_period=15/u uname
Invalid psb_period for intel_pt. Valid values are: 0-5
If MTC packets are selected, the default config selects a value
of 3 (i.e. psb_period=3) or the nearest lower value that is
supported (0 is always supported). Otherwise the default is 0.
If decoding is expected to be reliable and the buffer is large
then a large PSB period can be used.
Because a TSC packet is produced with PSB, the PSB period can
also affect the granularity to timing information in the absence
of MTC or CYC.
mtc Produces MTC timing packets.
MTC packets provide finer grain timestamp information than TSC
packets. MTC packets record time using the hardware crystal
clock (CTC) which is related to TSC packets using a TMA packet.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/mtc
which contains "1" if the feature is supported and
"0" otherwise.
The frequency of MTC packets can also be specified - see
mtc_period below.
mtc_period Specifies how frequently MTC packets are produced - see mtc
above for how to determine if MTC packets are supported.
Valid values are given by:
/sys/bus/event_source/devices/intel_pt/caps/mtc_periods
which contains a hexadecimal value, the bits of which represent
valid values e.g. bit 2 set means value 2 is valid.
The mtc_period value is converted to the MTC frequency as:
CTC-frequency / (2 ^ value)
e.g. value 3 means one eighth of CTC-frequency
Where CTC is the hardware crystal clock, the frequency of which
can be related to TSC via values provided in cpuid leaf 0x15.
If an invalid value is entered, the error message
will give a list of valid values e.g.
$ perf record -e intel_pt/mtc_period=15/u uname
Invalid mtc_period for intel_pt. Valid values are: 0,3,6,9
The default value is 3 or the nearest lower value
that is supported (0 is always supported).
cyc Produces CYC timing packets.
CYC packets provide even finer grain timestamp information than
MTC and TSC packets. A CYC packet contains the number of CPU
cycles since the last CYC packet. Unlike MTC and TSC packets,
CYC packets are only sent when another packet is also sent.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/psb_cyc
which contains "1" if the feature is supported and
"0" otherwise.
The number of CYC packets produced can be reduced by specifying
a threshold - see cyc_thresh below.
cyc_thresh Specifies how frequently CYC packets are produced - see cyc
above for how to determine if CYC packets are supported.
Valid cyc_thresh values are given by:
/sys/bus/event_source/devices/intel_pt/caps/cycle_thresholds
which contains a hexadecimal value, the bits of which represent
valid values e.g. bit 2 set means value 2 is valid.
The cyc_thresh value represents the minimum number of CPU cycles
that must have passed before a CYC packet can be sent. The
number of CPU cycles is:
2 ^ (value - 1)
e.g. value 4 means 8 CPU cycles must pass before a CYC packet
can be sent. Note a CYC packet is still only sent when another
packet is sent, not at, e.g. every 8 CPU cycles.
If an invalid value is entered, the error message
will give a list of valid values e.g.
$ perf record -e intel_pt/cyc,cyc_thresh=15/u uname
Invalid cyc_thresh for intel_pt. Valid values are: 0-12
CYC packets are not requested by default.
pt Specifies pass-through which enables the 'branch' config term.
The default config selects 'pt' if it is available, so a user will
never need to specify this term.
branch Enable branch tracing. Branch tracing is enabled by default so to
disable branch tracing use 'branch=0'.
The default config selects 'branch' if it is available.
ptw Enable PTWRITE packets which are produced when a ptwrite instruction
is executed.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/ptwrite
which contains "1" if the feature is supported and
"0" otherwise.
fup_on_ptw Enable a FUP packet to follow the PTWRITE packet. The FUP packet
provides the address of the ptwrite instruction. In the absence of
fup_on_ptw, the decoder will use the address of the previous branch
if branch tracing is enabled, otherwise the address will be zero.
Note that fup_on_ptw will work even when branch tracing is disabled.
pwr_evt Enable power events. The power events provide information about
changes to the CPU C-state.
Support for this feature is indicated by:
/sys/bus/event_source/devices/intel_pt/caps/power_event_trace
which contains "1" if the feature is supported and
"0" otherwise.
AUX area sampling option
~~~~~~~~~~~~~~~~~~~~~~~~
To select Intel PT "sampling" the AUX area sampling option can be used:
--aux-sample
Optionally it can be followed by the sample size in bytes e.g.
--aux-sample=8192
In addition, the Intel PT event to sample must be defined e.g.
-e intel_pt//u
Samples on other events will be created containing Intel PT data e.g. the
following will create Intel PT samples on the branch-misses event, note the
events must be grouped using {}:
perf record --aux-sample -e '{intel_pt//u,branch-misses:u}'
An alternative to '--aux-sample' is to add the config term 'aux-sample-size' to
events. In this case, the grouping is implied e.g.
perf record -e intel_pt//u -e branch-misses/aux-sample-size=8192/u
is the same as:
perf record -e '{intel_pt//u,branch-misses/aux-sample-size=8192/u}'
but allows for also using an address filter e.g.:
perf record -e intel_pt//u --filter 'filter * @/bin/ls' -e branch-misses/aux-sample-size=8192/u -- ls
It is important to select a sample size that is big enough to contain at least
one PSB packet. If not a warning will be displayed:
Intel PT sample size (%zu) may be too small for PSB period (%zu)
The calculation used for that is: if sample_size <= psb_period + 256 display the
warning. When sampling is used, psb_period defaults to 0 (2KiB).
The default sample size is 4KiB.
The sample size is passed in aux_sample_size in struct perf_event_attr. The
sample size is limited by the maximum event size which is 64KiB. It is
difficult to know how big the event might be without the trace sample attached,
but the tool validates that the sample size is not greater than 60KiB.
new snapshot option
~~~~~~~~~~~~~~~~~~~
The difference between full trace and snapshot from the kernel's perspective is
that in full trace we don't overwrite trace data that the user hasn't collected
yet (and indicated that by advancing aux_tail), whereas in snapshot mode we let
the trace run and overwrite older data in the buffer so that whenever something
interesting happens, we can stop it and grab a snapshot of what was going on
around that interesting moment.
To select snapshot mode a new option has been added:
-S
Optionally it can be followed by the snapshot size e.g.
-S0x100000
The default snapshot size is the auxtrace mmap size. If neither auxtrace mmap size
nor snapshot size is specified, then the default is 4MiB for privileged users
(or if /proc/sys/kernel/perf_event_paranoid < 0), 128KiB for unprivileged users.
If an unprivileged user does not specify mmap pages, the mmap pages will be
reduced as described in the 'new auxtrace mmap size option' section below.
The snapshot size is displayed if the option -vv is used e.g.
Intel PT snapshot size: %zu
new auxtrace mmap size option
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Intel PT buffer size is specified by an addition to the -m option e.g.
-m,16
selects a buffer size of 16 pages i.e. 64KiB.
Note that the existing functionality of -m is unchanged. The auxtrace mmap size
is specified by the optional addition of a comma and the value.
The default auxtrace mmap size for Intel PT is 4MiB/page_size for privileged users
(or if /proc/sys/kernel/perf_event_paranoid < 0), 128KiB for unprivileged users.
If an unprivileged user does not specify mmap pages, the mmap pages will be
reduced from the default 512KiB/page_size to 256KiB/page_size, otherwise the
user is likely to get an error as they exceed their mlock limit (Max locked
memory as shown in /proc/self/limits). Note that perf does not count the first
512KiB (actually /proc/sys/kernel/perf_event_mlock_kb minus 1 page) per cpu
against the mlock limit so an unprivileged user is allowed 512KiB per cpu plus
their mlock limit (which defaults to 64KiB but is not multiplied by the number
of cpus).
In full-trace mode, powers of two are allowed for buffer size, with a minimum
size of 2 pages. In snapshot mode or sampling mode, it is the same but the
minimum size is 1 page.
The mmap size and auxtrace mmap size are displayed if the -vv option is used e.g.
mmap length 528384
auxtrace mmap length 4198400
Intel PT modes of operation
~~~~~~~~~~~~~~~~~~~~~~~~~~~
Intel PT can be used in 2 modes:
full-trace mode
sample mode
snapshot mode
Full-trace mode traces continuously e.g.
perf record -e intel_pt//u uname
Sample mode attaches a Intel PT sample to other events e.g.
perf record --aux-sample -e intel_pt//u -e branch-misses:u
Snapshot mode captures the available data when a signal is sent e.g.
perf record -v -e intel_pt//u -S ./loopy 1000000000 &
[1] 11435
kill -USR2 11435
Recording AUX area tracing snapshot
Note that the signal sent is SIGUSR2.
Note that "Recording AUX area tracing snapshot" is displayed because the -v
option is used.
The 2 modes cannot be used together.
Buffer handling
~~~~~~~~~~~~~~~
There may be buffer limitations (i.e. single ToPa entry) which means that actual
buffer sizes are limited to powers of 2 up to 4MiB (MAX_ORDER). In order to
provide other sizes, and in particular an arbitrarily large size, multiple
buffers are logically concatenated. However an interrupt must be used to switch
between buffers. That has two potential problems:
a) the interrupt may not be handled in time so that the current buffer
becomes full and some trace data is lost.
b) the interrupts may slow the system and affect the performance
results.
If trace data is lost, the driver sets 'truncated' in the PERF_RECORD_AUX event
which the tools report as an error.
In full-trace mode, the driver waits for data to be copied out before allowing
the (logical) buffer to wrap-around. If data is not copied out quickly enough,
again 'truncated' is set in the PERF_RECORD_AUX event. If the driver has to
wait, the intel_pt event gets disabled. Because it is difficult to know when
that happens, perf tools always re-enable the intel_pt event after copying out
data.
Intel PT and build ids
~~~~~~~~~~~~~~~~~~~~~~
By default "perf record" post-processes the event stream to find all build ids
for executables for all addresses sampled. Deliberately, Intel PT is not
decoded for that purpose (it would take too long). Instead the build ids for
all executables encountered (due to mmap, comm or task events) are included
in the perf.data file.
To see buildids included in the perf.data file use the command:
perf buildid-list
If the perf.data file contains Intel PT data, that is the same as:
perf buildid-list --with-hits
Snapshot mode and event disabling
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In order to make a snapshot, the intel_pt event is disabled using an IOCTL,
namely PERF_EVENT_IOC_DISABLE. However doing that can also disable the
collection of side-band information. In order to prevent that, a dummy
software event has been introduced that permits tracking events (like mmaps) to
continue to be recorded while intel_pt is disabled. That is important to ensure
there is complete side-band information to allow the decoding of subsequent
snapshots.
A test has been created for that. To find the test:
perf test list
...
23: Test using a dummy software event to keep tracking
To run the test:
perf test 23
23: Test using a dummy software event to keep tracking : Ok
perf record modes (nothing new here)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
perf record essentially operates in one of three modes:
per thread
per cpu
workload only
"per thread" mode is selected by -t or by --per-thread (with -p or -u or just a
workload).
"per cpu" is selected by -C or -a.
"workload only" mode is selected by not using the other options but providing a
command to run (i.e. the workload).
In per-thread mode an exact list of threads is traced. There is no inheritance.
Each thread has its own event buffer.
In per-cpu mode all processes (or processes from the selected cgroup i.e. -G
option, or processes selected with -p or -u) are traced. Each cpu has its own
buffer. Inheritance is allowed.
In workload-only mode, the workload is traced but with per-cpu buffers.
Inheritance is allowed. Note that you can now trace a workload in per-thread
mode by using the --per-thread option.
Privileged vs non-privileged users
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Unless /proc/sys/kernel/perf_event_paranoid is set to -1, unprivileged users
have memory limits imposed upon them. That affects what buffer sizes they can
have as outlined above.
The v4.2 kernel introduced support for a context switch metadata event,
PERF_RECORD_SWITCH, which allows unprivileged users to see when their processes
are scheduled out and in, just not by whom, which is left for the
PERF_RECORD_SWITCH_CPU_WIDE, that is only accessible in system wide context,
which in turn requires CAP_SYS_ADMIN.
Please see the 45ac1403f564 ("perf: Add PERF_RECORD_SWITCH to indicate context
switches") commit, that introduces these metadata events for further info.
When working with kernels < v4.2, the following considerations must be taken,
as the sched:sched_switch tracepoints will be used to receive such information:
Unless /proc/sys/kernel/perf_event_paranoid is set to -1, unprivileged users are
not permitted to use tracepoints which means there is insufficient side-band
information to decode Intel PT in per-cpu mode, and potentially workload-only
mode too if the workload creates new processes.
Note also, that to use tracepoints, read-access to debugfs is required. So if
debugfs is not mounted or the user does not have read-access, it will again not
be possible to decode Intel PT in per-cpu mode.
sched_switch tracepoint
~~~~~~~~~~~~~~~~~~~~~~~
The sched_switch tracepoint is used to provide side-band data for Intel PT
decoding in kernels where the PERF_RECORD_SWITCH metadata event isn't
available.
The sched_switch events are automatically added. e.g. the second event shown
below:
$ perf record -vv -e intel_pt//u uname
------------------------------------------------------------
perf_event_attr:
type 6
size 112
config 0x400
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|CPU|IDENTIFIER
read_format ID
disabled 1
inherit 1
exclude_kernel 1
exclude_hv 1
enable_on_exec 1
sample_id_all 1
------------------------------------------------------------
sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
------------------------------------------------------------
perf_event_attr:
type 2
size 112
config 0x108
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|CPU|PERIOD|RAW|IDENTIFIER
read_format ID
inherit 1
sample_id_all 1
exclude_guest 1
------------------------------------------------------------
sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid -1 cpu 3 group_fd -1 flags 0x8
------------------------------------------------------------
perf_event_attr:
type 1
size 112
config 0x9
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|IDENTIFIER
read_format ID
disabled 1
inherit 1
exclude_kernel 1
exclude_hv 1
mmap 1
comm 1
enable_on_exec 1
task 1
sample_id_all 1
mmap2 1
comm_exec 1
------------------------------------------------------------
sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
mmap size 528384B
AUX area mmap length 4194304
perf event ring buffer mmapped per cpu
Synthesizing auxtrace information
Linux
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.042 MB perf.data ]
Note, the sched_switch event is only added if the user is permitted to use it
and only in per-cpu mode.
Note also, the sched_switch event is only added if TSC packets are requested.
That is because, in the absence of timing information, the sched_switch events
cannot be matched against the Intel PT trace.
perf script
-----------
By default, perf script will decode trace data found in the perf.data file.
This can be further controlled by new option --itrace.
New --itrace option
~~~~~~~~~~~~~~~~~~~
Having no option is the same as
--itrace
which, in turn, is the same as
--itrace=cepwx
The letters are:
i synthesize "instructions" events
b synthesize "branches" events
x synthesize "transactions" events
w synthesize "ptwrite" events
p synthesize "power" events
c synthesize branches events (calls only)
r synthesize branches events (returns only)
e synthesize tracing error events
d create a debug log
g synthesize a call chain (use with i or x)
l synthesize last branch entries (use with i or x)
s skip initial number of events
"Instructions" events look like they were recorded by "perf record -e
instructions".
"Branches" events look like they were recorded by "perf record -e branches". "c"
and "r" can be combined to get calls and returns.
"Transactions" events correspond to the start or end of transactions. The
'flags' field can be used in perf script to determine whether the event is a
tranasaction start, commit or abort.
Note that "instructions", "branches" and "transactions" events depend on code
flow packets which can be disabled by using the config term "branch=0". Refer
to the config terms section above.
"ptwrite" events record the payload of the ptwrite instruction and whether
"fup_on_ptw" was used. "ptwrite" events depend on PTWRITE packets which are
recorded only if the "ptw" config term was used. Refer to the config terms
section above. perf script "synth" field displays "ptwrite" information like
this: "ip: 0 payload: 0x123456789abcdef0" where "ip" is 1 if "fup_on_ptw" was
used.
"Power" events correspond to power event packets and CBR (core-to-bus ratio)
packets. While CBR packets are always recorded when tracing is enabled, power
event packets are recorded only if the "pwr_evt" config term was used. Refer to
the config terms section above. The power events record information about
C-state changes, whereas CBR is indicative of CPU frequency. perf script
"event,synth" fields display information like this:
cbr: cbr: 22 freq: 2189 MHz (200%)
mwait: hints: 0x60 extensions: 0x1
pwre: hw: 0 cstate: 2 sub-cstate: 0
exstop: ip: 1
pwrx: deepest cstate: 2 last cstate: 2 wake reason: 0x4
Where:
"cbr" includes the frequency and the percentage of maximum non-turbo
"mwait" shows mwait hints and extensions
"pwre" shows C-state transitions (to a C-state deeper than C0) and
whether initiated by hardware
"exstop" indicates execution stopped and whether the IP was recorded
exactly,
"pwrx" indicates return to C0
For more details refer to the Intel 64 and IA-32 Architectures Software
Developer Manuals.
Error events show where the decoder lost the trace. Error events
are quite important. Users must know if what they are seeing is a complete
picture or not.
The "d" option will cause the creation of a file "intel_pt.log" containing all
decoded packets and instructions. Note that this option slows down the decoder
and that the resulting file may be very large.
In addition, the period of the "instructions" event can be specified. e.g.
--itrace=i10us
sets the period to 10us i.e. one instruction sample is synthesized for each 10
microseconds of trace. Alternatives to "us" are "ms" (milliseconds),
"ns" (nanoseconds), "t" (TSC ticks) or "i" (instructions).
"ms", "us" and "ns" are converted to TSC ticks.
The timing information included with Intel PT does not give the time of every
instruction. Consequently, for the purpose of sampling, the decoder estimates
the time since the last timing packet based on 1 tick per instruction. The time
on the sample is *not* adjusted and reflects the last known value of TSC.
For Intel PT, the default period is 100us.
Setting it to a zero period means "as often as possible".
In the case of Intel PT that is the same as a period of 1 and a unit of
'instructions' (i.e. --itrace=i1i).
Also the call chain size (default 16, max. 1024) for instructions or
transactions events can be specified. e.g.
--itrace=ig32
--itrace=xg32
Also the number of last branch entries (default 64, max. 1024) for instructions or
transactions events can be specified. e.g.
--itrace=il10
--itrace=xl10
Note that last branch entries are cleared for each sample, so there is no overlap
from one sample to the next.
To disable trace decoding entirely, use the option --no-itrace.
It is also possible to skip events generated (instructions, branches, transactions)
at the beginning. This is useful to ignore initialization code.
--itrace=i0nss1000000
skips the first million instructions.
dump option
~~~~~~~~~~~
perf script has an option (-D) to "dump" the events i.e. display the binary
data.
When -D is used, Intel PT packets are displayed. The packet decoder does not
pay attention to PSB packets, but just decodes the bytes - so the packets seen
by the actual decoder may not be identical in places where the data is corrupt.
One example of that would be when the buffer-switching interrupt has been too
slow, and the buffer has been filled completely. In that case, the last packet
in the buffer might be truncated and immediately followed by a PSB as the trace
continues in the next buffer.
To disable the display of Intel PT packets, combine the -D option with
--no-itrace.
perf report
-----------
By default, perf report will decode trace data found in the perf.data file.
This can be further controlled by new option --itrace exactly the same as
perf script, with the exception that the default is --itrace=igxe.
perf inject
-----------
perf inject also accepts the --itrace option in which case tracing data is
removed and replaced with the synthesized events. e.g.
perf inject --itrace -i perf.data -o perf.data.new
Below is an example of using Intel PT with autofdo. It requires autofdo
(https://github.com/google/autofdo) and gcc version 5. The bubble
sort example is from the AutoFDO tutorial (https://gcc.gnu.org/wiki/AutoFDO/Tutorial)
amended to take the number of elements as a parameter.
$ gcc-5 -O3 sort.c -o sort_optimized
$ ./sort_optimized 30000
Bubble sorting array of 30000 elements
2254 ms
$ cat ~/.perfconfig
[intel-pt]
mispred-all = on
$ perf record -e intel_pt//u ./sort 3000
Bubble sorting array of 3000 elements
58 ms
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 3.939 MB perf.data ]
$ perf inject -i perf.data -o inj --itrace=i100usle --strip
$ ./create_gcov --binary=./sort --profile=inj --gcov=sort.gcov -gcov_version=1
$ gcc-5 -O3 -fauto-profile=sort.gcov sort.c -o sort_autofdo
$ ./sort_autofdo 30000
Bubble sorting array of 30000 elements
2155 ms
Note there is currently no advantage to using Intel PT instead of LBR, but
that may change in the future if greater use is made of the data.
PEBS via Intel PT
-----------------
Some hardware has the feature to redirect PEBS records to the Intel PT trace.
Recording is selected by using the aux-output config term e.g.
perf record -c 10000 -e '{intel_pt/branch=0/,cycles/aux-output/ppp}' uname
Note that currently, software only supports redirecting at most one PEBS event.
To display PEBS events from the Intel PT trace, use the itrace 'o' option e.g.
perf script --itrace=oe
SEE ALSO
--------
linkperf:perf-record[1], linkperf:perf-script[1], linkperf:perf-report[1],
linkperf:perf-inject[1]
tools/perf/Documentation/perf-record.txt
View file @ d1c9f7d1
......@@ -589,4 +589,4 @@ appended unit character - B/K/M/G
SEE ALSO
--------
linkperf:perf-stat[1], linkperf:perf-list[1]
linkperf:perf-stat[1], linkperf:perf-list[1], linkperf:perf-intel-pt[1]
tools/perf/Documentation/perf-report.txt
View file @ d1c9f7d1
......@@ -546,4 +546,5 @@ include::callchain-overhead-calculation.txt[]
SEE ALSO
--------
linkperf:perf-stat[1], linkperf:perf-annotate[1], linkperf:perf-record[1]
linkperf:perf-stat[1], linkperf:perf-annotate[1], linkperf:perf-record[1],
linkperf:perf-intel-pt[1]
tools/perf/Documentation/perf-script.txt
View file @ d1c9f7d1
......@@ -429,4 +429,4 @@ include::itrace.txt[]
SEE ALSO
--------
linkperf:perf-record[1], linkperf:perf-script-perl[1],
linkperf:perf-script-python[1]
linkperf:perf-script-python[1], linkperf:perf-intel-pt[1]
tools/perf/builtin-report.c
View file @ d1c9f7d1
......@@ -186,24 +186,23 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,
{
struct hist_entry *he = iter->he;
struct report *rep = arg;
struct branch_info *bi;
struct branch_info *bi = he->branch_info;
struct perf_sample *sample = iter->sample;
struct evsel *evsel = iter->evsel;
int err;
branch_type_count(&rep->brtype_stat, &bi->flags,
bi->from.addr, bi->to.addr);
if (!ui__has_annotation() && !rep->symbol_ipc)
return 0;
bi = he->branch_info;
err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
if (err)
goto out;
err = addr_map_symbol__inc_samples(&bi->to, sample, evsel);
branch_type_count(&rep->brtype_stat, &bi->flags,
bi->from.addr, bi->to.addr);
out:
return err;
}
......
tools/perf/pmu-events/arch/s390/cf_z15/crypto6.json
View file @ d1c9f7d1
......@@ -4,27 +4,27 @@
"EventCode": "80",
"EventName": "ECC_FUNCTION_COUNT",
"BriefDescription": "ECC Function Count",
"PublicDescription": "Long ECC function Count"
"PublicDescription": "This counter counts the total number of the elliptic-curve cryptography (ECC) functions issued by the CPU."
},
{
"Unit": "CPU-M-CF",
"EventCode": "81",
"EventName": "ECC_CYCLES_COUNT",
"BriefDescription": "ECC Cycles Count",
"PublicDescription": "Long ECC Function cycles count"
"PublicDescription": "This counter counts the total number of CPU cycles when the ECC coprocessor is busy performing the elliptic-curve cryptography (ECC) functions issued by the CPU."
},
{
"Unit": "CPU-M-CF",
"EventCode": "82",
"EventName": "ECC_BLOCKED_FUNCTION_COUNT",
"BriefDescription": "Ecc Blocked Function Count",
"PublicDescription": "Long ECC blocked function count"
"PublicDescription": "This counter counts the total number of the elliptic-curve cryptography (ECC) functions that are issued by the CPU and are blocked because the ECC coprocessor is busy performing a function issued by another CPU."
},
{
"Unit": "CPU-M-CF",
"EventCode": "83",
"EventName": "ECC_BLOCKED_CYCLES_COUNT",
"BriefDescription": "ECC Blocked Cycles Count",
"PublicDescription": "Long ECC blocked cycles count"
"PublicDescription": "This counter counts the total number of CPU cycles blocked for the elliptic-curve cryptography (ECC) functions issued by the CPU because the ECC coprocessor is busy performing a function issued by another CPU."
},
]
tools/perf/pmu-events/arch/s390/cf_z15/extended.json
View file @ d1c9f7d1
......@@ -25,7 +25,7 @@
"EventCode": "131",
"EventName": "DTLB2_HPAGE_WRITES",
"BriefDescription": "DTLB2 One-Megabyte Page Writes",
"PublicDescription": "A translation entry was written into the Combined Region and Segment Table Entry array in the Level-2 TLB for a one-megabyte page or a Last Host Translation was done"
"PublicDescription": "A translation entry was written into the Combined Region and Segment Table Entry array in the Level-2 TLB for a one-megabyte page"
},
{
"Unit": "CPU-M-CF",
......@@ -356,6 +356,34 @@
"BriefDescription": "Aborted transactions in constrained TX mode using special completion logic",
"PublicDescription": "A transaction abort has occurred in a constrained transactional-execution mode and the CPU is using special logic to allow the transaction to complete"
},
{
"Unit": "CPU-M-CF",
"EventCode": "247",
"EventName": "DFLT_ACCESS",
"BriefDescription": "Cycles CPU spent obtaining access to Deflate unit",
"PublicDescription": "Cycles CPU spent obtaining access to Deflate unit"
},
{
"Unit": "CPU-M-CF",
"EventCode": "252",
"EventName": "DFLT_CYCLES",
"BriefDescription": "Cycles CPU is using Deflate unit",
"PublicDescription": "Cycles CPU is using Deflate unit"
},
{
"Unit": "CPU-M-CF",
"EventCode": "264",
"EventName": "DFLT_CC",
"BriefDescription": "Increments by one for every DEFLATE CONVERSION CALL instruction executed",
"PublicDescription": "Increments by one for every DEFLATE CONVERSION CALL instruction executed"
},
{
"Unit": "CPU-M-CF",
"EventCode": "265",
"EventName": "DFLT_CCERROR",
"BriefDescription": "Increments by one for every DEFLATE CONVERSION CALL instruction executed that ended in Condition Codes 0, 1 or 2",
"PublicDescription": "Increments by one for every DEFLATE CONVERSION CALL instruction executed that ended in Condition Codes 0, 1 or 2"
},
{
"Unit": "CPU-M-CF",
"EventCode": "448",
......
tools/perf/pmu-events/arch/x86/cascadelakex/clx-metrics.json
View file @ d1c9f7d1
......@@ -215,7 +215,8 @@
"BriefDescription": "Utilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accesses",
"MetricExpr": "( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING + EPT.WALK_PENDING ) / ( 2 * cycles )",
"MetricGroup": "TLB",
"MetricName": "Page_Walks_Utilization"
"MetricName": "Page_Walks_Utilization",
"MetricConstraint": "NO_NMI_WATCHDOG"
},
{
"BriefDescription": "Utilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accesses",
......
tools/perf/pmu-events/arch/x86/skylake/skl-metrics.json
View file @ d1c9f7d1
......@@ -215,7 +215,8 @@
"BriefDescription": "Utilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accesses",
"MetricExpr": "( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING + EPT.WALK_PENDING ) / ( 2 * cycles )",
"MetricGroup": "TLB",
"MetricName": "Page_Walks_Utilization"
"MetricName": "Page_Walks_Utilization",
"MetricConstraint": "NO_NMI_WATCHDOG"
},
{
"BriefDescription": "Utilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accesses",
......
tools/perf/pmu-events/arch/x86/skylakex/skx-metrics.json
View file @ d1c9f7d1
......@@ -215,7 +215,8 @@
"BriefDescription": "Utilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accesses",
"MetricExpr": "( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING + EPT.WALK_PENDING ) / ( 2 * cycles )",
"MetricGroup": "TLB",
"MetricName": "Page_Walks_Utilization"
"MetricName": "Page_Walks_Utilization",
"MetricConstraint": "NO_NMI_WATCHDOG"
},
{
"BriefDescription": "Utilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accesses",
......
tools/perf/pmu-events/jevents.c
View file @ d1c9f7d1
......@@ -323,7 +323,7 @@ static int print_events_table_entry(void *data, char *name, char *event,
char *pmu, char *unit, char *perpkg,
char *metric_expr,
char *metric_name, char *metric_group,
char *deprecated)
char *deprecated, char *metric_constraint)
{
struct perf_entry_data *pd = data;
FILE *outfp = pd->outfp;
......@@ -357,6 +357,8 @@ static int print_events_table_entry(void *data, char *name, char *event,
fprintf(outfp, "\t.metric_group = \"%s\",\n", metric_group);
if (deprecated)
fprintf(outfp, "\t.deprecated = \"%s\",\n", deprecated);
if (metric_constraint)
fprintf(outfp, "\t.metric_constraint = \"%s\",\n", metric_constraint);
fprintf(outfp, "},\n");
return 0;
......@@ -375,6 +377,7 @@ struct event_struct {
char *metric_name;
char *metric_group;
char *deprecated;
char *metric_constraint;
};
#define ADD_EVENT_FIELD(field) do { if (field) { \
......@@ -422,7 +425,7 @@ static int save_arch_std_events(void *data, char *name, char *event,
char *desc, char *long_desc, char *pmu,
char *unit, char *perpkg, char *metric_expr,
char *metric_name, char *metric_group,
char *deprecated)
char *deprecated, char *metric_constraint)
{
struct event_struct *es;
......@@ -486,7 +489,7 @@ try_fixup(const char *fn, char *arch_std, char **event, char **desc,
char **name, char **long_desc, char **pmu, char **filter,
char **perpkg, char **unit, char **metric_expr, char **metric_name,
char **metric_group, unsigned long long eventcode,
char **deprecated)
char **deprecated, char **metric_constraint)
{
/* try to find matching event from arch standard values */
struct event_struct *es;
......@@ -515,7 +518,7 @@ int json_events(const char *fn,
char *pmu, char *unit, char *perpkg,
char *metric_expr,
char *metric_name, char *metric_group,
char *deprecated),
char *deprecated, char *metric_constraint),
void *data)
{
int err;
......@@ -545,6 +548,7 @@ int json_events(const char *fn,
char *metric_name = NULL;
char *metric_group = NULL;
char *deprecated = NULL;
char *metric_constraint = NULL;
char *arch_std = NULL;
unsigned long long eventcode = 0;
struct msrmap *msr = NULL;
......@@ -629,6 +633,8 @@ int json_events(const char *fn,
addfield(map, &metric_name, "", "", val);
} else if (json_streq(map, field, "MetricGroup")) {
addfield(map, &metric_group, "", "", val);
} else if (json_streq(map, field, "MetricConstraint")) {
addfield(map, &metric_constraint, "", "", val);
} else if (json_streq(map, field, "MetricExpr")) {
addfield(map, &metric_expr, "", "", val);
for (s = metric_expr; *s; s++)
......@@ -670,13 +676,13 @@ int json_events(const char *fn,
&long_desc, &pmu, &filter, &perpkg,
&unit, &metric_expr, &metric_name,
&metric_group, eventcode,
&deprecated);
&deprecated, &metric_constraint);
if (err)
goto free_strings;
}
err = func(data, name, real_event(name, event), desc, long_desc,
pmu, unit, perpkg, metric_expr, metric_name,
metric_group, deprecated);
metric_group, deprecated, metric_constraint);
free_strings:
free(event);
free(desc);
......@@ -691,6 +697,7 @@ int json_events(const char *fn,
free(metric_expr);
free(metric_name);
free(metric_group);
free(metric_constraint);
free(arch_std);
if (err)
......
tools/perf/pmu-events/jevents.h
View file @ d1c9f7d1
......@@ -8,7 +8,7 @@ int json_events(const char *fn,
char *pmu,
char *unit, char *perpkg, char *metric_expr,
char *metric_name, char *metric_group,
char *deprecated),
char *deprecated, char *metric_constraint),
void *data);
char *get_cpu_str(void);
......
tools/perf/pmu-events/pmu-events.h
View file @ d1c9f7d1
......@@ -18,6 +18,7 @@ struct pmu_event {
const char *metric_name;
const char *metric_group;
const char *deprecated;
const char *metric_constraint;
};
/*
......
tools/perf/scripts/perl/check-perf-trace.pl
View file @ d1c9f7d1
......@@ -28,7 +28,7 @@ sub trace_end
sub irq::softirq_entry
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$vec) = @_;
print_header($event_name, $common_cpu, $common_secs, $common_nsecs,
......@@ -43,7 +43,7 @@ sub irq::softirq_entry
sub kmem::kmalloc
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$call_site, $ptr, $bytes_req, $bytes_alloc,
$gfp_flags) = @_;
......@@ -92,7 +92,7 @@ sub print_unhandled
sub trace_unhandled
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm) = @_;
$common_pid, $common_comm, $common_callchain) = @_;
$unhandled{$event_name}++;
}
......
tools/perf/scripts/perl/failed-syscalls.pl
View file @ d1c9f7d1
......@@ -18,7 +18,7 @@ my %failed_syscalls;
sub raw_syscalls::sys_exit
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$id, $ret) = @_;
if ($ret < 0) {
......
tools/perf/scripts/perl/rw-by-file.pl
View file @ d1c9f7d1
......@@ -28,7 +28,7 @@ my %writes;
sub syscalls::sys_enter_read
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm, $nr, $fd, $buf, $count) = @_;
$common_pid, $common_comm, $common_callchain, $nr, $fd, $buf, $count) = @_;
if ($common_comm eq $for_comm) {
$reads{$fd}{bytes_requested} += $count;
......@@ -39,7 +39,7 @@ sub syscalls::sys_enter_read
sub syscalls::sys_enter_write
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm, $nr, $fd, $buf, $count) = @_;
$common_pid, $common_comm, $common_callchain, $nr, $fd, $buf, $count) = @_;
if ($common_comm eq $for_comm) {
$writes{$fd}{bytes_written} += $count;
......@@ -98,7 +98,7 @@ sub print_unhandled
sub trace_unhandled
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm) = @_;
$common_pid, $common_comm, $common_callchain) = @_;
$unhandled{$event_name}++;
}
......
tools/perf/scripts/perl/rw-by-pid.pl
View file @ d1c9f7d1
......@@ -24,7 +24,7 @@ my %writes;
sub syscalls::sys_exit_read
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $ret) = @_;
if ($ret > 0) {
......@@ -40,7 +40,7 @@ sub syscalls::sys_exit_read
sub syscalls::sys_enter_read
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $fd, $buf, $count) = @_;
$reads{$common_pid}{bytes_requested} += $count;
......@@ -51,7 +51,7 @@ sub syscalls::sys_enter_read
sub syscalls::sys_exit_write
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $ret) = @_;
if ($ret <= 0) {
......@@ -62,7 +62,7 @@ sub syscalls::sys_exit_write
sub syscalls::sys_enter_write
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $fd, $buf, $count) = @_;
$writes{$common_pid}{bytes_written} += $count;
......@@ -178,7 +178,7 @@ sub print_unhandled
sub trace_unhandled
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm) = @_;
$common_pid, $common_comm, $common_callchain) = @_;
$unhandled{$event_name}++;
}
tools/perf/scripts/perl/rwtop.pl
View file @ d1c9f7d1
......@@ -35,7 +35,7 @@ if (!$interval) {
sub syscalls::sys_exit_read
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $ret) = @_;
print_check();
......@@ -53,7 +53,7 @@ sub syscalls::sys_exit_read
sub syscalls::sys_enter_read
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $fd, $buf, $count) = @_;
print_check();
......@@ -66,7 +66,7 @@ sub syscalls::sys_enter_read
sub syscalls::sys_exit_write
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $ret) = @_;
print_check();
......@@ -79,7 +79,7 @@ sub syscalls::sys_exit_write
sub syscalls::sys_enter_write
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$nr, $fd, $buf, $count) = @_;
print_check();
......@@ -197,7 +197,7 @@ sub print_unhandled
sub trace_unhandled
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm) = @_;
$common_pid, $common_comm, $common_callchain) = @_;
$unhandled{$event_name}++;
}
tools/perf/scripts/perl/wakeup-latency.pl
View file @ d1c9f7d1
......@@ -28,7 +28,7 @@ my $total_wakeups = 0;
sub sched::sched_switch
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$prev_comm, $prev_pid, $prev_prio, $prev_state, $next_comm, $next_pid,
$next_prio) = @_;
......@@ -51,7 +51,7 @@ sub sched::sched_switch
sub sched::sched_wakeup
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm,
$common_pid, $common_comm, $common_callchain,
$comm, $pid, $prio, $success, $target_cpu) = @_;
$last_wakeup{$target_cpu}{ts} = nsecs($common_secs, $common_nsecs);
......@@ -101,7 +101,7 @@ sub print_unhandled
sub trace_unhandled
{
my ($event_name, $context, $common_cpu, $common_secs, $common_nsecs,
$common_pid, $common_comm) = @_;
$common_pid, $common_comm, $common_callchain) = @_;
$unhandled{$event_name}++;
}
tools/perf/tests/builtin-test.c
View file @ d1c9f7d1
......@@ -543,8 +543,11 @@ static int run_shell_tests(int argc, const char *argv[], int i, int width)
return -1;
dir = opendir(st.dir);
if (!dir)
if (!dir) {
pr_err("failed to open shell test directory: %s\n",
st.dir);
return -1;
}
for_each_shell_test(dir, st.dir, ent) {
int curr = i++;
......
tools/perf/util/cs-etm.c
View file @ d1c9f7d1
......@@ -363,6 +363,23 @@ struct cs_etm_packet_queue
return NULL;
}
static void cs_etm__packet_swap(struct cs_etm_auxtrace *etm,
struct cs_etm_traceid_queue *tidq)
{
struct cs_etm_packet *tmp;
if (etm->sample_branches || etm->synth_opts.last_branch ||
etm->sample_instructions) {
/*
* Swap PACKET with PREV_PACKET: PACKET becomes PREV_PACKET for
* the next incoming packet.
*/
tmp = tidq->packet;
tidq->packet = tidq->prev_packet;
tidq->prev_packet = tmp;
}
}
static void cs_etm__packet_dump(const char *pkt_string)
{
const char *color = PERF_COLOR_BLUE;
......@@ -945,7 +962,7 @@ static inline u64 cs_etm__instr_addr(struct cs_etm_queue *etmq,
if (packet->isa == CS_ETM_ISA_T32) {
u64 addr = packet->start_addr;
while (offset > 0) {
while (offset) {
addr += cs_etm__t32_instr_size(etmq,
trace_chan_id, addr);
offset--;
......@@ -1134,10 +1151,8 @@ static int cs_etm__synth_instruction_sample(struct cs_etm_queue *etmq,
cs_etm__copy_insn(etmq, tidq->trace_chan_id, tidq->packet, &sample);
if (etm->synth_opts.last_branch) {
cs_etm__copy_last_branch_rb(etmq, tidq);
if (etm->synth_opts.last_branch)
sample.branch_stack = tidq->last_branch;
}
if (etm->synth_opts.inject) {
ret = cs_etm__inject_event(event, &sample,
......@@ -1153,9 +1168,6 @@ static int cs_etm__synth_instruction_sample(struct cs_etm_queue *etmq,
"CS ETM Trace: failed to deliver instruction event, error %d\n",
ret);
if (etm->synth_opts.last_branch)
cs_etm__reset_last_branch_rb(tidq);
return ret;
}
......@@ -1342,12 +1354,14 @@ static int cs_etm__sample(struct cs_etm_queue *etmq,
struct cs_etm_traceid_queue *tidq)
{
struct cs_etm_auxtrace *etm = etmq->etm;
struct cs_etm_packet *tmp;
int ret;
u8 trace_chan_id = tidq->trace_chan_id;
u64 instrs_executed = tidq->packet->instr_count;
u64 instrs_prev;
/* Get instructions remainder from previous packet */
instrs_prev = tidq->period_instructions;
tidq->period_instructions += instrs_executed;
tidq->period_instructions += tidq->packet->instr_count;
/*
* Record a branch when the last instruction in
......@@ -1365,26 +1379,80 @@ static int cs_etm__sample(struct cs_etm_queue *etmq,
* TODO: allow period to be defined in cycles and clock time
*/
/* Get number of instructions executed after the sample point */
u64 instrs_over = tidq->period_instructions -
etm->instructions_sample_period;
/*
* Below diagram demonstrates the instruction samples
* generation flows:
*
* Instrs Instrs Instrs Instrs
* Sample(n) Sample(n+1) Sample(n+2) Sample(n+3)
* | | | |
* V V V V
* --------------------------------------------------
* ^ ^
* | |
* Period Period
* instructions(Pi) instructions(Pi')
*
* | |
* \---------------- -----------------/
* V
* tidq->packet->instr_count
*
* Instrs Sample(n...) are the synthesised samples occurring
* every etm->instructions_sample_period instructions - as
* defined on the perf command line. Sample(n) is being the
* last sample before the current etm packet, n+1 to n+3
* samples are generated from the current etm packet.
*
* tidq->packet->instr_count represents the number of
* instructions in the current etm packet.
*
* Period instructions (Pi) contains the the number of
* instructions executed after the sample point(n) from the
* previous etm packet. This will always be less than
* etm->instructions_sample_period.
*
* When generate new samples, it combines with two parts
* instructions, one is the tail of the old packet and another
* is the head of the new coming packet, to generate
* sample(n+1); sample(n+2) and sample(n+3) consume the
* instructions with sample period. After sample(n+3), the rest
* instructions will be used by later packet and it is assigned
* to tidq->period_instructions for next round calculation.
*/
/*
* Calculate the address of the sampled instruction (-1 as
* sample is reported as though instruction has just been
* executed, but PC has not advanced to next instruction)
* Get the initial offset into the current packet instructions;
* entry conditions ensure that instrs_prev is less than
* etm->instructions_sample_period.
*/
u64 offset = (instrs_executed - instrs_over - 1);
u64 addr = cs_etm__instr_addr(etmq, trace_chan_id,
tidq->packet, offset);
u64 offset = etm->instructions_sample_period - instrs_prev;
u64 addr;
ret = cs_etm__synth_instruction_sample(
etmq, tidq, addr, etm->instructions_sample_period);
if (ret)
return ret;
/* Prepare last branches for instruction sample */
if (etm->synth_opts.last_branch)
cs_etm__copy_last_branch_rb(etmq, tidq);
/* Carry remaining instructions into next sample period */
tidq->period_instructions = instrs_over;
while (tidq->period_instructions >=
etm->instructions_sample_period) {
/*
* Calculate the address of the sampled instruction (-1
* as sample is reported as though instruction has just
* been executed, but PC has not advanced to next
* instruction)
*/
addr = cs_etm__instr_addr(etmq, trace_chan_id,
tidq->packet, offset - 1);
ret = cs_etm__synth_instruction_sample(
etmq, tidq, addr,
etm->instructions_sample_period);
if (ret)
return ret;
offset += etm->instructions_sample_period;
tidq->period_instructions -=
etm->instructions_sample_period;
}
}
if (etm->sample_branches) {
......@@ -1406,15 +1474,7 @@ static int cs_etm__sample(struct cs_etm_queue *etmq,
}
}
if (etm->sample_branches || etm->synth_opts.last_branch) {
/*
* Swap PACKET with PREV_PACKET: PACKET becomes PREV_PACKET for
* the next incoming packet.
*/
tmp = tidq->packet;
tidq->packet = tidq->prev_packet;
tidq->prev_packet = tmp;
}
cs_etm__packet_swap(etm, tidq);
return 0;
}
......@@ -1443,7 +1503,6 @@ static int cs_etm__flush(struct cs_etm_queue *etmq,
{
int err = 0;
struct cs_etm_auxtrace *etm = etmq->etm;
struct cs_etm_packet *tmp;
/* Handle start tracing packet */
if (tidq->prev_packet->sample_type == CS_ETM_EMPTY)
......@@ -1451,6 +1510,11 @@ static int cs_etm__flush(struct cs_etm_queue *etmq,
if (etmq->etm->synth_opts.last_branch &&
tidq->prev_packet->sample_type == CS_ETM_RANGE) {
u64 addr;
/* Prepare last branches for instruction sample */
cs_etm__copy_last_branch_rb(etmq, tidq);
/*
* Generate a last branch event for the branches left in the
* circular buffer at the end of the trace.
......@@ -1458,7 +1522,7 @@ static int cs_etm__flush(struct cs_etm_queue *etmq,
* Use the address of the end of the last reported execution
* range
*/
u64 addr = cs_etm__last_executed_instr(tidq->prev_packet);
addr = cs_etm__last_executed_instr(tidq->prev_packet);
err = cs_etm__synth_instruction_sample(
etmq, tidq, addr,
......@@ -1478,15 +1542,11 @@ static int cs_etm__flush(struct cs_etm_queue *etmq,
}
swap_packet:
if (etm->sample_branches || etm->synth_opts.last_branch) {
/*
* Swap PACKET with PREV_PACKET: PACKET becomes PREV_PACKET for
* the next incoming packet.
*/
tmp = tidq->packet;
tidq->packet = tidq->prev_packet;
tidq->prev_packet = tmp;
}
cs_etm__packet_swap(etm, tidq);
/* Reset last branches after flush the trace */
if (etm->synth_opts.last_branch)
cs_etm__reset_last_branch_rb(tidq);
return err;
}
......@@ -1507,11 +1567,16 @@ static int cs_etm__end_block(struct cs_etm_queue *etmq,
*/
if (etmq->etm->synth_opts.last_branch &&
tidq->prev_packet->sample_type == CS_ETM_RANGE) {
u64 addr;
/* Prepare last branches for instruction sample */
cs_etm__copy_last_branch_rb(etmq, tidq);
/*
* Use the address of the end of the last reported execution
* range.
*/
u64 addr = cs_etm__last_executed_instr(tidq->prev_packet);
addr = cs_etm__last_executed_instr(tidq->prev_packet);
err = cs_etm__synth_instruction_sample(
etmq, tidq, addr,
......
tools/perf/util/expr.l
View file @ d1c9f7d1
......@@ -79,10 +79,10 @@ symbol {spec}*{sym}*{spec}*{sym}*
{
int start_token;
start_token = parse_events_get_extra(yyscanner);
start_token = expr_get_extra(yyscanner);
if (start_token) {
parse_events_set_extra(NULL, yyscanner);
expr_set_extra(NULL, yyscanner);
return start_token;
}
}
......
tools/perf/util/map.c
View file @ d1c9f7d1
......@@ -44,8 +44,8 @@ static inline int is_no_dso_memory(const char *filename)
static inline int is_android_lib(const char *filename)
{
return !strncmp(filename, "/data/app-lib", 13) ||
!strncmp(filename, "/system/lib", 11);
return strstarts(filename, "/data/app-lib/") ||
strstarts(filename, "/system/lib/");
}
static inline bool replace_android_lib(const char *filename, char *newfilename)
......@@ -65,7 +65,7 @@ static inline bool replace_android_lib(const char *filename, char *newfilename)
app_abi_length = strlen(app_abi);
if (!strncmp(filename, "/data/app-lib", 13)) {
if (strstarts(filename, "/data/app-lib/")) {
char *apk_path;
if (!app_abi_length)
......@@ -89,7 +89,7 @@ static inline bool replace_android_lib(const char *filename, char *newfilename)
return true;
}
if (!strncmp(filename, "/system/lib/", 12)) {
if (strstarts(filename, "/system/lib/")) {
char *ndk, *app;
const char *arch;
size_t ndk_length;
......
tools/perf/util/metricgroup.c
View file @ d1c9f7d1
......@@ -22,6 +22,8 @@
#include <linux/string.h>
#include <linux/zalloc.h>
#include <subcmd/parse-options.h>
#include <api/fs/fs.h>
#include "util.h"
struct metric_event *metricgroup__lookup(struct rblist *metric_events,
struct evsel *evsel,
......@@ -399,13 +401,85 @@ void metricgroup__print(bool metrics, bool metricgroups, char *filter,
strlist__delete(metriclist);
}
static void metricgroup__add_metric_weak_group(struct strbuf *events,
const char **ids,
int idnum)
{
bool no_group = false;
int i;
for (i = 0; i < idnum; i++) {
pr_debug("found event %s\n", ids[i]);
/*
* Duration time maps to a software event and can make
* groups not count. Always use it outside a
* group.
*/
if (!strcmp(ids[i], "duration_time")) {
if (i > 0)
strbuf_addf(events, "}:W,");
strbuf_addf(events, "duration_time");
no_group = true;
continue;
}
strbuf_addf(events, "%s%s",
i == 0 || no_group ? "{" : ",",
ids[i]);
no_group = false;
}
if (!no_group)
strbuf_addf(events, "}:W");
}
static void metricgroup__add_metric_non_group(struct strbuf *events,
const char **ids,
int idnum)
{
int i;
for (i = 0; i < idnum; i++)
strbuf_addf(events, ",%s", ids[i]);
}
static void metricgroup___watchdog_constraint_hint(const char *name, bool foot)
{
static bool violate_nmi_constraint;
if (!foot) {
pr_warning("Splitting metric group %s into standalone metrics.\n", name);
violate_nmi_constraint = true;
return;
}
if (!violate_nmi_constraint)
return;
pr_warning("Try disabling the NMI watchdog to comply NO_NMI_WATCHDOG metric constraint:\n"
" echo 0 > /proc/sys/kernel/nmi_watchdog\n"
" perf stat ...\n"
" echo 1 > /proc/sys/kernel/nmi_watchdog\n");
}
static bool metricgroup__has_constraint(struct pmu_event *pe)
{
if (!pe->metric_constraint)
return false;
if (!strcmp(pe->metric_constraint, "NO_NMI_WATCHDOG") &&
sysctl__nmi_watchdog_enabled()) {
metricgroup___watchdog_constraint_hint(pe->metric_name, false);
return true;
}
return false;
}
static int metricgroup__add_metric(const char *metric, struct strbuf *events,
struct list_head *group_list)
{
struct pmu_events_map *map = perf_pmu__find_map(NULL);
struct pmu_event *pe;
int ret = -EINVAL;
int i, j;
int i, ret = -EINVAL;
if (!map)
return 0;
......@@ -422,7 +496,6 @@ static int metricgroup__add_metric(const char *metric, struct strbuf *events,
const char **ids;
int idnum;
struct egroup *eg;
bool no_group = false;
pr_debug("metric expr %s for %s\n", pe->metric_expr, pe->metric_name);
......@@ -431,27 +504,11 @@ static int metricgroup__add_metric(const char *metric, struct strbuf *events,
continue;
if (events->len > 0)
strbuf_addf(events, ",");
for (j = 0; j < idnum; j++) {
pr_debug("found event %s\n", ids[j]);
/*
* Duration time maps to a software event and can make
* groups not count. Always use it outside a
* group.
*/
if (!strcmp(ids[j], "duration_time")) {
if (j > 0)
strbuf_addf(events, "}:W,");
strbuf_addf(events, "duration_time");
no_group = true;
continue;
}
strbuf_addf(events, "%s%s",
j == 0 || no_group ? "{" : ",",
ids[j]);
no_group = false;
}
if (!no_group)
strbuf_addf(events, "}:W");
if (metricgroup__has_constraint(pe))
metricgroup__add_metric_non_group(events, ids, idnum);
else
metricgroup__add_metric_weak_group(events, ids, idnum);
eg = malloc(sizeof(struct egroup));
if (!eg) {
......@@ -493,6 +550,10 @@ static int metricgroup__add_metric_list(const char *list, struct strbuf *events,
}
}
free(nlist);
if (!ret)
metricgroup___watchdog_constraint_hint(NULL, true);
return ret;
}
......
tools/perf/util/mmap.c
View file @ d1c9f7d1
......@@ -98,20 +98,29 @@ static int perf_mmap__aio_bind(struct mmap *map, int idx, int cpu, int affinity)
{
void *data;
size_t mmap_len;
unsigned long node_mask;
unsigned long *node_mask;
unsigned long node_index;
int err = 0;
if (affinity != PERF_AFFINITY_SYS && cpu__max_node() > 1) {
data = map->aio.data[idx];
mmap_len = mmap__mmap_len(map);
node_mask = 1UL << cpu__get_node(cpu);
if (mbind(data, mmap_len, MPOL_BIND, &node_mask, 1, 0)) {
pr_err("Failed to bind [%p-%p] AIO buffer to node %d: error %m\n",
data, data + mmap_len, cpu__get_node(cpu));
node_index = cpu__get_node(cpu);
node_mask = bitmap_alloc(node_index + 1);
if (!node_mask) {
pr_err("Failed to allocate node mask for mbind: error %m\n");
return -1;
}
set_bit(node_index, node_mask);
if (mbind(data, mmap_len, MPOL_BIND, node_mask, node_index + 1 + 1, 0)) {
pr_err("Failed to bind [%p-%p] AIO buffer to node %lu: error %m\n",
data, data + mmap_len, node_index);
err = -1;
}
bitmap_free(node_mask);
}
return 0;
return err;
}
#else /* !HAVE_LIBNUMA_SUPPORT */
static int perf_mmap__aio_alloc(struct mmap *map, int idx)
......
tools/perf/util/stat-display.c
View file @ d1c9f7d1
......@@ -16,6 +16,7 @@
#include <linux/ctype.h>
#include "cgroup.h"
#include <api/fs/fs.h>
#include "util.h"
#define CNTR_NOT_SUPPORTED "<not supported>"
#define CNTR_NOT_COUNTED "<not counted>"
......@@ -1097,7 +1098,6 @@ static void print_footer(struct perf_stat_config *config)
{
double avg = avg_stats(config->walltime_nsecs_stats) / NSEC_PER_SEC;
FILE *output = config->output;
int n;
if (!config->null_run)
fprintf(output, "\n");
......@@ -1131,9 +1131,7 @@ static void print_footer(struct perf_stat_config *config)
}
fprintf(output, "\n\n");
if (config->print_free_counters_hint &&
sysctl__read_int("kernel/nmi_watchdog", &n) >= 0 &&
n > 0)
if (config->print_free_counters_hint && sysctl__nmi_watchdog_enabled())
fprintf(output,
"Some events weren't counted. Try disabling the NMI watchdog:\n"
" echo 0 > /proc/sys/kernel/nmi_watchdog\n"
......
tools/perf/util/synthetic-events.c
View file @ d1c9f7d1
......@@ -345,6 +345,7 @@ int perf_event__synthesize_mmap_events(struct perf_tool *tool,
continue;
event->mmap2.ino = (u64)ino;
event->mmap2.ino_generation = 0;
/*
* Just like the kernel, see __perf_event_mmap in kernel/perf_event.c
......
tools/perf/util/util.c
View file @ d1c9f7d1
......@@ -55,6 +55,24 @@ int sysctl__max_stack(void)
return sysctl_perf_event_max_stack;
}
bool sysctl__nmi_watchdog_enabled(void)
{
static bool cached;
static bool nmi_watchdog;
int value;
if (cached)
return nmi_watchdog;
if (sysctl__read_int("kernel/nmi_watchdog", &value) < 0)
return false;
nmi_watchdog = (value > 0) ? true : false;
cached = true;
return nmi_watchdog;
}
bool test_attr__enabled;
bool perf_host = true;
......
tools/perf/util/util.h
View file @ d1c9f7d1
......@@ -29,6 +29,8 @@ size_t hex_width(u64 v);
int sysctl__max_stack(void);
bool sysctl__nmi_watchdog_enabled(void);
int fetch_kernel_version(unsigned int *puint,
char *str, size_t str_sz);
#define KVER_VERSION(x) (((x) >> 16) & 0xff)
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7