Waker threads in the futex wake-parallel benchmark are started by a loop
using pthread_create(). However, there is no synchronization for when
the waker threads wake the waiting threads. Comparison of the waker
threads' measurement timestamps show they are not all running
concurrently because older waker threads finish their task before newer
waker threads even start.
This patch uses a barrier to better synchronize the waker threads.
Signed-off-by: James Yang <james.yang@arm.com
Cc: Kim Phillips <kim.phillips@arm.com>
Link: http://lkml.kernel.org/r/20171127042101.3659-4-dave@stgolabs.net
Signed-off-by: Davidlohr Bueso <dave@stgolabs.net>
[ Disable the wake-parallel test for systems without pthread_barrier_t ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It was reported that the whole futex bench breaks when dealing with
non-contiguously numbered cpus.
$ echo 0 | sudo tee /sys/devices/system/cpu/cpu3/online
$ ./perf bench futex all
perf: pthread_create: Operation not permitted
Run summary [PID 14934]: 7 threads, each ....
James had implemented an approach with cpumaps that use an in house
flavor. Instead of re-inventing the wheel, I've redone the patch such
that we use the perf's util/cpumap.c interface instead.
Applies to all futex benchmarks.
Suggested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Originally-from: James Yang <james.yang@arm.com>
Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Cc: Davidlohr Bueso <dbueso@suse.de>
Cc: Kim Phillips <kim.phillips@arm.com>
Link: http://lkml.kernel.org/r/20171127042101.3659-2-dave@stgolabs.net
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add support for computing 'perf stat' style metrics in 'perf script'.
When using leader sampling we can get metrics for each sampling period
by computing formulas over the values of the different group members.
This allows things like fine grained IPC tracking through sampling, much
more fine grained than with 'perf stat'.
The metric is still averaged over the sampling period, it is not just
for the sampling point.
This patch adds a new metric output field for 'perf script' that uses
the existing 'perf stat' metrics infrastructure to compute any metrics
supported by 'perf stat'.
For example to sample IPC:
$ perf record -e '{ref-cycles,cycles,instructions}:S' -a sleep 1
$ perf script -F metric,ip,sym,time,cpu,comm
...
alsa-sink-ALC32 [000] 42815.856074: 7fd65937d6cc [unknown]
alsa-sink-ALC32 [000] 42815.856074: 7fd65937d6cc [unknown]
alsa-sink-ALC32 [000] 42815.856074: 7fd65937d6cc [unknown]
alsa-sink-ALC32 [000] 42815.856074: metric: 0.13 insn per cycle
swapper [000] 42815.857961: ffffffff81655df0 __schedule
swapper [000] 42815.857961: ffffffff81655df0 __schedule
swapper [000] 42815.857961: ffffffff81655df0 __schedule
swapper [000] 42815.857961: metric: 0.23 insn per cycle
qemu-system-x86 [000] 42815.858130: ffffffff8165ad0e _raw_spin_unlock_irqrestore
qemu-system-x86 [000] 42815.858130: ffffffff8165ad0e _raw_spin_unlock_irqrestore
qemu-system-x86 [000] 42815.858130: ffffffff8165ad0e _raw_spin_unlock_irqrestore
qemu-system-x86 [000] 42815.858130: metric: 0.46 insn per cycle
:4972 [000] 42815.858312: ffffffffa080e5f2 vmx_vcpu_run
:4972 [000] 42815.858312: ffffffffa080e5f2 vmx_vcpu_run
:4972 [000] 42815.858312: ffffffffa080e5f2 vmx_vcpu_run
:4972 [000] 42815.858312: metric: 0.45 insn per cycle
TopDown:
This requires disabling SMT if you have it enabled, because SMT would
require sampling per core, which is not supported.
$ perf record -e '{ref-cycles,topdown-fetch-bubbles,\
topdown-recovery-bubbles,\
topdown-slots-retired,topdown-total-slots,\
topdown-slots-issued}:S' -a sleep 1
$ perf script --header -I -F cpu,ip,sym,event,metric,period
...
[000] 121108 ref-cycles: ffffffff8165222e copy_user_enhanced_fast_string
[000] 190350 topdown-fetch-bubbles: ffffffff8165222e copy_user_enhanced_fast_string
[000] 2055 topdown-recovery-bubbles: ffffffff8165222e copy_user_enhanced_fast_string
[000] 148729 topdown-slots-retired: ffffffff8165222e copy_user_enhanced_fast_string
[000] 144324 topdown-total-slots: ffffffff8165222e copy_user_enhanced_fast_string
[000] 160852 topdown-slots-issued: ffffffff8165222e copy_user_enhanced_fast_string
[000] metric: 33.0% frontend bound
[000] metric: 3.5% bad speculation
[000] metric: 25.8% retiring
[000] metric: 37.7% backend bound
[000] 112112 ref-cycles: ffffffff8165aec8 _raw_spin_lock_irqsave
[000] 357222 topdown-fetch-bubbles: ffffffff8165aec8 _raw_spin_lock_irqsave
[000] 3325 topdown-recovery-bubbles: ffffffff8165aec8 _raw_spin_lock_irqsave
[000] 323553 topdown-slots-retired: ffffffff8165aec8 _raw_spin_lock_irqsave
[000] 270507 topdown-total-slots: ffffffff8165aec8 _raw_spin_lock_irqsave
[000] 341226 topdown-slots-issued: ffffffff8165aec8 _raw_spin_lock_irqsave
[000] metric: 33.0% frontend bound
[000] metric: 2.9% bad speculation
[000] metric: 29.9% retiring
[000] metric: 34.2% backend bound
...
v2:
Use evsel->priv for new fields
Port to new base line, support fp output.
Handle stats in ->stats, not ->priv
Minor cleanups
Extra explanation about the use of the term 'averaging', from Andi in the
thread in the Link: tag below:
<quote Andi>
The current samples contains the sum of event counts for a sampling period.
EventA-1 EventA-2 EventA-3 EventA-4
EventB-1 EventB-2 EventC-3
gap with no events overflow
|-----------------------------------------------------------------|
period-start period-end
^ ^
| |
previous sample current sample
So EventA = 4 and EventB = 3 at the sample point
I generate a metric, let's say EventA / EventB. It applies to the whole period.
But the metric is over a longer time which does not have the same behavior. For
example the gap above doesn't have any events, while they are clustered at the
beginning and end of the sample period.
But we're summing everything together. The metric doesn't know that the gap is
different than the busy period.
That's what I'm trying to express with averaging.
</quote>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20171117214300.32746-4-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
If we're not sampling the kernel, we shouldn't care about kptr_restrict
neither synthesize anything for assisting in resolving kernel samples,
like the reference relocation symbol or kernel modules information.
Before:
$ cat /proc/sys/kernel/kptr_restrict /proc/sys/kernel/perf_event_paranoid
2
2
$ perf record sleep 1
WARNING: Kernel address maps (/proc/{kallsyms,modules}) are restricted,
check /proc/sys/kernel/kptr_restrict.
Samples in kernel functions may not be resolved if a suitable vmlinux
file is not found in the buildid cache or in the vmlinux path.
Samples in kernel modules won't be resolved at all.
If some relocation was applied (e.g. kexec) symbols may be misresolved
even with a suitable vmlinux or kallsyms file.
Couldn't record kernel reference relocation symbol
Symbol resolution may be skewed if relocation was used (e.g. kexec).
Check /proc/kallsyms permission or run as root.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.001 MB perf.data (8 samples) ]
$ perf evlist -v
cycles:uppp: size: 112, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|PERIOD, disabled: 1, inherit: 1, exclude_kernel: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1, precise_ip: 3, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1
$
After:
$ perf record sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.001 MB perf.data (10 samples) ]
$
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-t025e9zftbx2b8cq2w01g5e5@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The 'perf test' case "probe libc's inet_pton & backtrace it with ping"
fails on s390x. The reason is the 'realpath /lib64/ld*.so.* | uniq' line
which returns 2 libraries:
root@s35lp76 shell]# realpath /lib64/ld*.so.* | uniq
/usr/lib64/ld-2.26.so
/usr/lib64/ld_pre_smc.so.1.0.1
[root@s35lp76 shell]
This output makes the "perf probe" command lines invalid.
Use ldd tool to find out the libraries required by "bash" and check if
symbol "inet_pton" is part of the "libc" library. Some distros do not
have a /lib64 directory.
I have also added a check for the existence of an IPv6 network interface
before it is being used.
Committer changes:
We can't really use ldd for libc, as in some systems, such as x86_64, it
has hardlinks and then ldd sees one and the kernel the other, so grep
for libc in /proc/self/maps to get the one we'll receive from
PERF_RECORD_MMAP.
Thomas checked this change and acked it.
Signed-off-by: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Suggested-by: Hendrik Brückner <brueckner@linux.vnet.ibm.com>
Reviewed-by: Hendrik Brückner <brueckner@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/20171114133409.GN8836@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This 'perf test' case fails on s390x. The 'touch' command on s390x uses
the 'openat' system call to open the file named on the command line:
[root@s35lp76 perf]# perf probe -l
probe:vfs_getname (on getname_flags:72@fs/namei.c with pathname)
[root@s35lp76 perf]# perf trace -e open touch /tmp/abc
0.400 ( 0.015 ms): touch/27542 open(filename:
/usr/lib/locale/locale-archive, flags: CLOEXEC) = 3
[root@s35lp76 perf]#
There is no 'open' system call for file '/tmp/abc'. Instead the 'openat'
system call is used:
[root@s35lp76 perf]# strace touch /tmp/abc
execve("/usr/bin/touch", ["touch", "/tmp/abc"], 0x3ffd547ec98
/* 30 vars */) = 0
[...]
openat(AT_FDCWD, "/tmp/abc", O_WRONLY|O_CREAT|O_NOCTTY|O_NONBLOCK, 0666) = 3
[...]
On s390x the 'egrep' command does not find a matching pattern and
returns an error.
Fix this for s390x create a platform dependent command line to enable
the 'perf probe' call to listen to the 'openat' system call and get the
expected output.
Signed-off-by: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com>
LPU-Reference: 20171114071847.2381-1-tmricht@linux.vnet.ibm.com
Link: http://lkml.kernel.org/n/tip-3qf38jk0prz54rhmhyu871my@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
A recent fix for 'perf trace' introduced a bug where
machine__exit(trace->host) could be called while trace->host was still
NULL, so make this more robust by guarding against NULL, just like
free() does.
The problem happens, for instance, when !root users try to run 'perf
trace':
[acme@jouet linux]$ trace
Error: No permissions to read /sys/kernel/debug/tracing/events/raw_syscalls/sys_(enter|exit)
Hint: Try 'sudo mount -o remount,mode=755 /sys/kernel/debug/tracing'
perf: Segmentation fault
Obtained 7 stack frames.
[0x4f1b2e]
/lib64/libc.so.6(+0x3671f) [0x7f43a1dd971f]
[0x4f3fec]
[0x47468b]
[0x42a2db]
/lib64/libc.so.6(__libc_start_main+0xe9) [0x7f43a1dc3509]
[0x42a6c9]
Segmentation fault (core dumped)
[acme@jouet linux]$
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andrei Vagin <avagin@openvz.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Vasily Averin <vvs@virtuozzo.com>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: 33974a414c ("perf trace: Call machine__exit() at exit")
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When processing PERF_RECORD_AUXTRACE_INFO several perf_evsel entries
will be synthesized and inserted into session->evlist, eventually ending
in perf_script.tool.sample(), which ends up calling builtin-script.c's
process_event(), that expects evsel->priv to be a perf_evsel_script
object with a valid FILE pointer in fp.
So we need to intercept the processing of PERF_RECORD_AUXTRACE_INFO and
then setup evsel->priv for these newly created perf_evsel instances, do
it to fix the segfault in process_event() trying to use a NULL for that
FILE pointer.
Reported-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: yuzhoujian <yuzhoujian@didichuxing.com>
Fixes: a14390fde6 ("perf script: Allow creating per-event dump files")
Link: http://lkml.kernel.org/n/tip-bthnur8r8de01gxvn2qayx6e@git.kernel.org
[ Merge fix by Ravi Bangoria before pushing upstream to preserv bisectability ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The Intel PMU event aliases have a implicit period= specifier to set the
default period.
Unfortunately this breaks overriding these periods with -c or -F,
because the alias terms look like they are user specified to the
internal parser, and user specified event qualifiers override the
command line options.
Track that they are coming from aliases by adding a "weak" state to the
term. Any weak terms don't override command line options.
I only did it for -c/-F for now, I think that's the only case that's
broken currently.
Before:
$ perf record -c 1000 -vv -e uops_issued.any
...
{ sample_period, sample_freq } 2000003
After:
$ perf record -c 1000 -vv -e uops_issued.any
...
{ sample_period, sample_freq } 1000
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20171020202755.21410-2-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When we use an initial delay, e.g.: 'perf record --delay 1000', we do not
enable the events until that delay has passed after we started the workload,
including the tracking event, i.e. the one for which we have attr.mmap, etc,
enabled to ask the kernel to generate the PERF_RECORD_{MMAP,COMM,EXEC} metadata
events that will then allow us to resolve addresses in samples to the map, dso
and symbol. There will be a shadow that even synthesizing samples won't cover,
i.e. the workload that we start and other processes forking while we
wait for the initial delay to expire.
So use a dummy event to be the tracking one and make it be enabled on exec.
Before:
# perf record --delay 1000 stress --cpu 1 --timeout 5
stress: info: [9029] dispatching hogs: 1 cpu, 0 io, 0 vm, 0 hdd
stress: info: [9029] successful run completed in 5s
[ perf record: Woken up 3 times to write data ]
[ perf record: Captured and wrote 0.624 MB perf.data (15908 samples) ]
# perf script | head
:9031 9031 32001.826888: 1 cycles:ppp: ffffffff831aa30d event_function (/lib/modules/4.14.0-rc6+/build/vmlinux)
:9031 9031 32001.826893: 1 cycles:ppp: ffffffff8300d1a0 intel_bts_enable_local (/lib/modules/4.14.0-rc6+/build/vmlinux)
:9031 9031 32001.826895: 7 cycles:ppp: ffffffff83023870 sched_clock (/lib/modules/4.14.0-rc6+/build/vmlinux)
:9031 9031 32001.826897: 103 cycles:ppp: ffffffff8300c331 intel_pmu_handle_irq (/lib/modules/4.14.0-rc6+/build/vmlinux)
:9031 9031 32001.826899: 1615 cycles:ppp: ffffffff830231f8 native_sched_clock (/lib/modules/4.14.0-rc6+/build/vmlinux)
:9031 9031 32001.826902: 26724 cycles:ppp: ffffffff8384c6a7 native_irq_return_iret (/lib/modules/4.14.0-rc6+/build/vmlinux)
:9031 9031 32001.826913: 329739 cycles:ppp: 7fb2a5410932 [unknown] ([unknown])
:9031 9031 32001.827033: 1225451 cycles:ppp: 7fb2a5410930 [unknown] ([unknown])
:9031 9031 32001.827474: 1391725 cycles:ppp: 7fb2a5410930 [unknown] ([unknown])
:9031 9031 32001.827978: 1233697 cycles:ppp: 7fb2a5410928 [unknown] ([unknown])
#
After:
# perf record --delay 1000 stress --cpu 1 --timeout 5
stress: info: [9741] dispatching hogs: 1 cpu, 0 io, 0 vm, 0 hdd
stress: info: [9741] successful run completed in 5s
[ perf record: Woken up 3 times to write data ]
[ perf record: Captured and wrote 0.751 MB perf.data (15976 samples) ]
# perf script | head
stress 9742 32110.959106: 1 cycles:ppp: ffffffff831b26f6 __perf_event_task_sched_in (/lib/modules/4.14.0-rc6+/build/vmlinux)
stress 9742 32110.959110: 1 cycles:ppp: ffffffff8300c2e9 intel_pmu_handle_irq (/lib/modules/4.14.0-rc6+/build/vmlinux)
stress 9742 32110.959112: 7 cycles:ppp: ffffffff830231e0 native_sched_clock (/lib/modules/4.14.0-rc6+/build/vmlinux)
stress 9742 32110.959115: 101 cycles:ppp: ffffffff83023870 sched_clock (/lib/modules/4.14.0-rc6+/build/vmlinux)
stress 9742 32110.959117: 1533 cycles:ppp: ffffffff830231f8 native_sched_clock (/lib/modules/4.14.0-rc6+/build/vmlinux)
stress 9742 32110.959119: 23992 cycles:ppp: ffffffff831b0900 ctx_sched_in (/lib/modules/4.14.0-rc6+/build/vmlinux)
stress 9742 32110.959129: 329406 cycles:ppp: 7f4b1b661930 __random_r (/usr/lib64/libc-2.25.so)
stress 9742 32110.959249: 1288322 cycles:ppp: 5566e1e7cbc9 hogcpu (/usr/bin/stress)
stress 9742 32110.959712: 1464046 cycles:ppp: 7f4b1b66179e __random (/usr/lib64/libc-2.25.so)
stress 9742 32110.960241: 1266918 cycles:ppp: 7f4b1b66195b __random_r (/usr/lib64/libc-2.25.so)
#
Reported-by: Bram Stolk <b.stolk@gmail.com>
Tested-by: Bram Stolk <b.stolk@gmail.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: 6619a53ef7 ("perf record: Add --initial-delay option")
Link: http://lkml.kernel.org/n/tip-nrdfchshqxf7diszhxcecqb9@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The evsel->idx field is used mainly to access the right bucket in
per-event arrays such as the annotation ones, but also to set
evsel->tracking, that in turn will decide what of the events will ask
for PERF_RECORD_{MMAP,COMM,EXEC} to be generated, i.e. which
perf_event_attr will have its mmap, etc fields set.
When we were adding the "dummy" event using perf_evlist__add_dummy() we
were not setting it correctly, which could result in multiple tracking
events.
Now that I'll try using a dummy event to be the tracking one when using
'perf record --delay', i.e. when we process the --delay
setting we may already have the evlist set up, like with:
perf record -e cycles,instructions --delay 1000 ./workload
We will need to add a "dummy" event, then reset evsel->tracking for the
first event, "cycles", and set it instead to the dummy one, and also
setting its attr.enable_on_exec, so that we get the PERF_RECORD_MMAP,
etc metadata events while waiting to enable the explicitely requested
events, so lets get this straight and set the right evsel->idx.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Bram Stolk <b.stolk@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-nrdfchshqxf7diszhxcecqb9@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
- Optimize sample parsing for ordering events, where we don't need to parse
all the PERF_SAMPLE_ bits, just the ones leading to the timestamp needed
to reorder events (Jiri Olsa)
- Use a dummy event to ask for PERF_RECORD_{MMAP,COMM,EXEC} with
'perf record --delay', when the events asked by the user will only be
enabled after the workload is started and the requested delay passes,
so we need to add the dummy event and have it .enabled_on_exec. This
then allows us to resolve symbols for the DSO executable MMAPs setup
while we wait for the delay (Arnaldo Carvalho de Melo)
- Synchronize kcmp.h and prctl.h ABI headers wrt SPDX tags (Arnaldo Carvalho de Melo)
- Generalize the annotation code to support other source information
besides objdump/DWARF obtained ones, starting with python scripts,
that will is slated to be merged soon (Jiri Olsa)
- Advance the source code lines to right after the column with the
address in asm lines (Jiri Olsa)
- Fix terminal dimensions resizing signal handling in 'perf top --stdio' (Jiri Olsa)
- Improve error messages for PMU events (Kim Phillips)
- Fix 'perf record' -c/-F options for cpu event aliases (Andi Kleen)
- Enable type checking for perf_evsel_config_term types (Andi Kleen)
- Call machine__exit() at 'perf trace' exit, so as to remove temporary
files related to VDSO (Andrei Vagin)
- Add "reject" option to parse-events.l, fixing the build with newer
flex releases. Noticed with flex 2.6.4 on Alpine Linux 3.6 and Edge (Jiri Olsa)
- Document some missing perf.data headers (Andi Kleen)
- Allow printing period for non freq mod groups (Andi Kleen)
- Do not warn the user about kernel.kptr_restrict when not sampling the
kernel (Arnaldo Carvalho de Melo)
- Fix bug in 'perf help' introduced during conversion to strstart() (Namhyung Kim)
- Do not truncate ASM instruction mnemonics at 6 characters in the annotation
output, PowerPC has long ones (Ravi Bangoria)
- Document some missing command line options (Sihyeon Jang)
- Update POWER9 vendor event tables (Sukadev Bhattiprolu)
- Fix 'perf test' shell entries on s390x, where the 'openat' syscall
is used instead of 'open' in one of the tests and
- No need to use overwrite mmap mode in 'perf test', those tests
do not generate massive amount of events to fill the ring buffer (Wang Nan)
- Add missing command line options (mostly --force/-f) to the man pages (Sihyeon Jang)
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull second round of s390 updates from Martin Schwidefsky:
- rework of the vdso code to avoid the use of the access register mode
- use perf AUX buffers for the transport of diagnostic sample data
- add perf_regs and user stack dump support
- enable perf call graphs for user space programs
- add perf register support for floating-point registers
- all remaining s390 related timer_setup conversions
- bug fixes and cleanups
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (30 commits)
s390: remove unused parameter from Makefile
zfcp: purely mechanical update using timer API, plus blank lines
s390/scsi: Convert timers to use timer_setup()
s390/cpum_sf: correctly set the PID and TID in perf samples
s390/cpum_sf: load program parameter at sampler enablement
s390/perf: add perf register support for floating-point registers
s390/perf: extend perf_regs support to include floating-point registers
s390/perf: define common DWARF register string table
s390/perf: add support for perf_regs and libdw
s390/perf: add perf_regs support and user stack dump
s390/cpum_sf: do not register PMU if no sampling mode is authorized
s390/cpumf: remove raw event support in basic-only sampling mode
s390/perf: add callback to perf to enable using AUX buffer
s390/cpumf: enable using AUX buffer
s390/cpumf: introduce AUX buffer for dump diagnostic sample data
s390/disassembler: increase show_code buffer size
s390: Remove CONFIG_HARDENED_USERCOPY
s390: enable CPU alternatives unconditionally
s390/nmi: remove unused code
s390/mm: remove unused code
...
Currently when using ordered events we parse the sample twice (the
perf_evlist__parse_sample function). Once before we queue the sample for
sorting:
perf_session__process_event
perf_evlist__parse_sample(sample)
perf_session__queue_event(sample.time)
And then when we deliver the sorted sample:
ordered_events__deliver_event
perf_evlist__parse_sample
perf_session__deliver_event
We can skip the initial full sample parsing by using
perf_evlist__parse_sample_timestamp function, which got introduced
earlier. The new path looks like:
perf_session__process_event
perf_evlist__parse_sample_timestamp
perf_session__queue_event
ordered_events__deliver_event
perf_session__deliver_event
perf_evlist__parse_sample
It saves some instructions and is slightly faster:
Before:
Performance counter stats for './perf.old report --stdio' (5 runs):
64,396,007,225 cycles:u ( +- 0.97% )
105,882,112,735 instructions:u # 1.64 insn per cycle ( +- 0.00% )
21.618103465 seconds time elapsed ( +- 1.12% )
After:
Performance counter stats for './perf report --stdio' (5 runs):
60,567,807,182 cycles:u ( +- 0.40% )
104,853,333,514 instructions:u # 1.73 insn per cycle ( +- 0.00% )
20.168895243 seconds time elapsed ( +- 0.32% )
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-cjp2tuk0qkjs9dxzlpmm34ua@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
