Stephane reported that we don't set properly PERIOD sample type for
events with period term defined.
Before:
$ perf record -e cpu/cpu-cycles,period=1000/u ls
$ perf evlist -v
cpu/cpu-cycles,period=1000/u: ... sample_type: IP|TID|TIME|PERIOD, ...
After:
$ perf record -e cpu/cpu-cycles,period=1000/u ls
$ perf evlist -v
cpu/cpu-cycles,period=1000/u: ... sample_type: IP|TID|TIME, ...
Setting PERIOD sample type based on period term setup.
Committer note:
When we use -c or a period=N term in the event definition, then we don't
need to ask the kernel, for this event, via perf_event_attr.sample_type
|= PERF_SAMPLE_PERIOD, to put the event period in each sample for this
event, as we know it already, it is in perf_event_attr.sample_period.
Reported-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Stephane Eranian <eranian@google.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20180201083812.11359-2-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
I've meet a strange behavior with these commands on my gentoo box:
1: perf kmem record
2: CTRL-C to stop 1
3: perf report
4: "Enter", "Enter", "Run scripts for all samples",
"event_analyzing_sample".
Then 'perf report' says:
"
No kallsyms or vmlinux with build-id xxxx was found
/lib/modules/4.10.0+/build/vmlinux with build id xxxx not found,
continuing without symbols
".
It is strange because I am sure /lib/modules/4.10.0+/build/vmlinux is
right for perf.data.
After digging, I found out the reason is that "perf report" generates
many open fds, then "script_browse" uses popen to run "perf script"
which run out of open files.
The gentoo box has a small default value for "max open files", 1024.
Yes, "ulimit -n " with a bigger number could fix it, but I think that
using O_CLOEXEC in do_open is a better way.
Signed-off-by: Wang YanQing <udknight@gmail.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20180115050448.GA20759@udknight
[ Make sure O_CLOEXEC is available in old systems by adding a patch
just before this one, to keep this bisectable in such systems ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that we can get it working for TUI, where using just pr_err() would
end up making the message emitted to stderr to be erased by the TUI exit
routine restoring the terminal to its previous state.
Now we can see that trying to use a tracepoint field as one of the
--field entries isn't working:
# perf top --stdio --no-children -e syscalls:sys_enter_write --fields pid,sym,count
Error:
Unknown --fields key: `count'
Usage: perf top [<options>]
--fields <key[,keys...]>
output field(s): overhead, period, sample plus all of sort keys
#
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: https://lkml.kernel.org/n/tip-usy9hhy7umdd4bbblkn63t8w@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When setting up DWARF callchains on specific events, without using
'record' or 'trace' --call-graph, but instead doing it like:
perf trace -e cycles/call-graph=dwarf/
The unwind__prepare_access() call in thread__insert_map() when we
process PERF_RECORD_MMAP(2) metadata events were not being performed,
precluding us from using per-event DWARF callchains, handling them just
when we asked for all events to be DWARF, using "--call-graph dwarf".
We do it in the PERF_RECORD_MMAP because we have to look at one of the
executable maps to figure out the executable type (64-bit, 32-bit) of
the DSO laid out in that mmap. Also to look at the architecture where
the perf.data file was recorded.
All this probably should be deferred to when we process a sample for
some thread that has callchains, so that we do this processing only for
the threads with samples, not for all of them.
For now, fix using DWARF on specific events.
Before:
# perf trace --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1
PING ::1(::1) 56 data bytes
64 bytes from ::1: icmp_seq=1 ttl=64 time=0.048 ms
--- ::1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.048/0.048/0.048/0.000 ms
0.000 probe_libc:inet_pton:(7fe9597bb350))
Problem processing probe_libc:inet_pton callchain, skipping...
#
After:
# perf trace --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1
PING ::1(::1) 56 data bytes
64 bytes from ::1: icmp_seq=1 ttl=64 time=0.060 ms
--- ::1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.060/0.060/0.060/0.000 ms
0.000 probe_libc:inet_pton:(7fd4aa930350))
__inet_pton (inlined)
gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so)
__GI_getaddrinfo (inlined)
[0xffffaa804e51af3f] (/usr/bin/ping)
__libc_start_main (/usr/lib64/libc-2.26.so)
[0xffffaa804e51b379] (/usr/bin/ping)
#
# perf trace --call-graph=dwarf --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1
PING ::1(::1) 56 data bytes
64 bytes from ::1: icmp_seq=1 ttl=64 time=0.057 ms
--- ::1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.057/0.057/0.057/0.000 ms
0.000 probe_libc:inet_pton:(7f9363b9e350))
__inet_pton (inlined)
gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so)
__GI_getaddrinfo (inlined)
[0xffffa9e8a14e0f3f] (/usr/bin/ping)
__libc_start_main (/usr/lib64/libc-2.26.so)
[0xffffa9e8a14e1379] (/usr/bin/ping)
#
# perf trace --call-graph=fp --no-syscalls -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1
PING ::1(::1) 56 data bytes
64 bytes from ::1: icmp_seq=1 ttl=64 time=0.077 ms
--- ::1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.077/0.077/0.077/0.000 ms
0.000 probe_libc:inet_pton:(7f4947e1c350))
__inet_pton (inlined)
gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so)
__GI_getaddrinfo (inlined)
[0xffffaa716d88ef3f] (/usr/bin/ping)
__libc_start_main (/usr/lib64/libc-2.26.so)
[0xffffaa716d88f379] (/usr/bin/ping)
#
# perf trace --no-syscalls -e probe_libc:inet_pton/call-graph=fp/ ping -6 -c 1 ::1
PING ::1(::1) 56 data bytes
64 bytes from ::1: icmp_seq=1 ttl=64 time=0.078 ms
--- ::1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.078/0.078/0.078/0.000 ms
0.000 probe_libc:inet_pton:(7fa157696350))
__GI___inet_pton (/usr/lib64/libc-2.26.so)
getaddrinfo (/usr/lib64/libc-2.26.so)
[0xffffa9ba39c74f40] (/usr/bin/ping)
#
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Hendrick Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Thomas Richter <tmricht@linux.vnet.ibm.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/r/20180116182650.GE16107@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When setting the "dwarf" unwinder for a specific event and not
specifying the max-stack, the attr.sample_max_stack ended up using an
uninitialized callchain_param.max_stack, fix it by using designated
initializers for that callchain_param variable, zeroing all non
explicitely initialized struct members.
Here is what happened:
# perf trace -vv --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1
callchain: type DWARF
callchain: stack dump size 8192
perf_event_attr:
type 2
size 112
config 0x730
{ sample_period, sample_freq } 1
sample_type IP|TID|TIME|ADDR|CALLCHAIN|CPU|PERIOD|RAW|REGS_USER|STACK_USER|DATA_SRC
exclude_callchain_user 1
{ wakeup_events, wakeup_watermark } 1
sample_regs_user 0xff0fff
sample_stack_user 8192
sample_max_stack 50656
sys_perf_event_open failed, error -75
Value too large for defined data type
# perf trace -vv --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1
callchain: type DWARF
callchain: stack dump size 8192
perf_event_attr:
type 2
size 112
config 0x730
sample_type IP|TID|TIME|ADDR|CALLCHAIN|CPU|PERIOD|RAW|REGS_USER|STACK_USER|DATA_SRC
exclude_callchain_user 1
sample_regs_user 0xff0fff
sample_stack_user 8192
sample_max_stack 30448
sys_perf_event_open failed, error -75
Value too large for defined data type
#
Now the attr.sample_max_stack is set to zero and the above works as
expected:
# perf trace --no-syscalls --max-stack 4 -e probe_libc:inet_pton/call-graph=dwarf/ ping -6 -c 1 ::1
PING ::1(::1) 56 data bytes
64 bytes from ::1: icmp_seq=1 ttl=64 time=0.072 ms
--- ::1 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.072/0.072/0.072/0.000 ms
0.000 probe_libc:inet_pton:(7feb7a998350))
__inet_pton (inlined)
gaih_inet.constprop.7 (/usr/lib64/libc-2.26.so)
__GI_getaddrinfo (inlined)
[0xffffaa39b6108f3f] (/usr/bin/ping)
#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Hendrick Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Thomas Richter <tmricht@linux.vnet.ibm.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-is9tramondqa9jlxxsgcm9iz@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
'perf record' and 'perf report --dump-raw-trace' supported in this
release.
Example usage:
# perf record -e arm_spe/ts_enable=1,pa_enable=1/ dd if=/dev/zero of=/dev/null count=10000
# perf report --dump-raw-trace
Note that the perf.data file is portable, so the report can be run on
another architecture host if necessary.
Output will contain raw SPE data and its textual representation, such
as:
0x5c8 [0x30]: PERF_RECORD_AUXTRACE size: 0x200000 offset: 0 ref: 0x1891ad0e idx: 1 tid: 2227 cpu: 1
.
. ... ARM SPE data: size 2097152 bytes
. 00000000: 49 00 LD
. 00000002: b2 c0 3b 29 0f 00 00 ff ff VA 0xffff00000f293bc0
. 0000000b: b3 c0 eb 24 fb 00 00 00 80 PA 0xfb24ebc0 ns=1
. 00000014: 9a 00 00 LAT 0 XLAT
. 00000017: 42 16 EV RETIRED L1D-ACCESS TLB-ACCESS
. 00000019: b0 00 c4 15 08 00 00 ff ff PC 0xff00000815c400 el3 ns=1
. 00000022: 98 00 00 LAT 0 TOT
. 00000025: 71 36 6c 21 2c 09 00 00 00 TS 39395093558
. 0000002e: 49 00 LD
. 00000030: b2 80 3c 29 0f 00 00 ff ff VA 0xffff00000f293c80
. 00000039: b3 80 ec 24 fb 00 00 00 80 PA 0xfb24ec80 ns=1
. 00000042: 9a 00 00 LAT 0 XLAT
. 00000045: 42 16 EV RETIRED L1D-ACCESS TLB-ACCESS
. 00000047: b0 f4 11 16 08 00 00 ff ff PC 0xff0000081611f4 el3 ns=1
. 00000050: 98 00 00 LAT 0 TOT
. 00000053: 71 36 6c 21 2c 09 00 00 00 TS 39395093558
. 0000005c: 48 00 INSN-OTHER
. 0000005e: 42 02 EV RETIRED
. 00000060: b0 2c ef 7f 08 00 00 ff ff PC 0xff0000087fef2c el3 ns=1
. 00000069: 98 00 00 LAT 0 TOT
. 0000006c: 71 d1 6f 21 2c 09 00 00 00 TS 39395094481
...
Other release notes:
- applies to acme's perf/{core,urgent} branches, likely elsewhere
- Report is self-contained within the tool.
Record requires enabling the kernel SPE driver by
setting CONFIG_ARM_SPE_PMU.
- The intel-bts implementation was used as a starting point; its
min/default/max buffer sizes and power of 2 pages granularity need to be
revisited for ARM SPE
- Recording across multiple SPE clusters/domains not supported
- Snapshot support (record -S), and conversion to native perf events
(e.g., via 'perf inject --itrace'), are also not supported
- Technically both cs-etm and spe can be used simultaneously, however
disabled for simplicity in this release
Signed-off-by: Kim Phillips <kim.phillips@arm.com>
Reviewed-by: Dongjiu Geng <gengdongjiu@huawei.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rob Herring <robh@kernel.org>
Cc: Suzuki Poulouse <suzuki.poulose@arm.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Will Deacon <will.deacon@arm.com>
Link: http://lkml.kernel.org/r/20180114132850.0b127434b704a26bad13268f@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
There could be different types of memory in the system. E.g normal
System Memory, Persistent Memory. To understand how the workload maps to
those memories, it's important to know the I/O statistics of them. Perf
can collect physical addresses, but those are raw data. It still needs
extra work to resolve the physical addresses. Provide a script to
facilitate the physical addresses resolving and I/O statistics.
Profile with MEM_INST_RETIRED.ALL_LOADS or MEM_UOPS_RETIRED.ALL_LOADS
event if any of them is available.
Look up the /proc/iomem and resolve the physical address. Provide
memory type summary.
Here is an example output:
# perf script report mem-phys-addr
Event: mem_inst_retired.all_loads:P
Memory type count percentage
---------------------------------------- ----------- -----------
System RAM 74 53.2%
Persistent Memory 55 39.6%
N/A
---
Changes since V2:
- Apply the new license rules.
- Add comments for globals
Changes since V1:
- Do not mix DLA and Load Latency. Do not compare the loads and stores.
Only profile the loads.
- Use event name to replace the RAW event
Signed-off-by: Kan Liang <Kan.liang@intel.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Philippe Ombredanne <pombredanne@nexb.com>
Cc: Stephane Eranian <eranian@google.com>
Link: https://lkml.kernel.org/r/1515099595-34770-1-git-send-email-kan.liang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Adding support to display sample misc field in form
of letter for each bit:
# perf script -F +misc ...
sched-messaging 1414 K 28690.636582: 4590 cycles ...
sched-messaging 1407 U 28690.636600: 325620 cycles ...
sched-messaging 1414 K 28690.636608: 19473 cycles ...
misc field __________/
The misc bits are assigned to following letters:
PERF_RECORD_MISC_KERNEL K
PERF_RECORD_MISC_USER U
PERF_RECORD_MISC_HYPERVISOR H
PERF_RECORD_MISC_GUEST_KERNEL G
PERF_RECORD_MISC_GUEST_USER g
PERF_RECORD_MISC_MMAP_DATA* M
PERF_RECORD_MISC_COMM_EXEC E
PERF_RECORD_MISC_SWITCH_OUT S
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20180107160356.28203-9-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Previous patch supports the multiple time range.
For example, select the first and second 10% time slices.
perf report --time 10%/1,10%/2
We need a function to check if a timestamp is in the ranges of
[0, 10%) and [10%, 20%].
Note that it includes the last element in [10%, 20%] but it doesn't
include the last element in [0, 10%). It's to avoid the overlap.
This patch implments a new function perf_time__ranges_skip_sample
for this checking.
Change log:
v4: Let perf_time__ranges_skip_sample be compatible with
perf_time__skip_sample when only one time range.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1512738826-2628-5-git-send-email-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Current perf report/script/... have a --time option to limit the time
range of output. But right now it only supports absolute time, add
support for time percentage.
For example:
1. Select the second 10% time slice
perf report --time 10%/2
2. Select from 0% to 10% time slice
perf report --time 0%-10%
It also support the multiple time ranges.
3. Select the first and second 10% time slices
perf report --time 10%/1,10%/2
4. Select from 0% to 10% and 30% to 40% slices
perf report --time 0%-10%,30%-40%
Changelog:
v4: An issue is found. Following passes.
perf script --time 10%/10x12321xsdfdasfdsafdsafdsa
Now it uses strtol to replace atoi.
Committer notes:
This just puts in place the infrastructure, so the examples in this cset
comment will only work later, after more patches in this series are
applied.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1512738826-2628-4-git-send-email-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf report/script/... have a --time option to limit the time range of
output. That's very useful to slice large traces, e.g. when processing
the output of perf script for some analysis.
But right now --time only supports absolute time. Also there is no fast
way to get the start/end times of a given trace except for looking at
it. This makes it hard to e.g. only decode the first half of the trace,
which is useful for parallelization of scripts
Another problem is that perf records are variable size and there is no
synchronization mechanism. So the only way to find the last sample
reliably would be to walk all samples. But we want to avoid that in perf
report/... because it is already quite expensive. That is why storing
the first sample time and last sample time in perf record is better.
This patch creates a new header feature type HEADER_SAMPLE_TIME and
related ops. Save the first sample time and the last sample time to the
feature section in perf file header. That will be done when, for
instance, processing build-ids, where we already have to process all
samples to create the build-id table, take advantage of that to further
amortize that processing by storing HEADER_SAMPLE_TIME to make 'perf
report/script' faster when using --time.
Committer testing:
After this patch is applied the header is written with zeroes, we need
the next patch, for "perf record" to actually write the timestamps:
# perf report -D | grep PERF_RECORD_SAMPLE\(
22501155244406 0x44f0 [0x28]: PERF_RECORD_SAMPLE(IP, 0x4001): 25016/25016: 0xffffffffa21be8c5 period: 1 addr: 0
<SNIP>
22501155793625 0x4a30 [0x28]: PERF_RECORD_SAMPLE(IP, 0x4001): 25016/25016: 0xffffffffa21ffd50 period: 2828043 addr: 0
# perf report --header | grep "time of "
# time of first sample : 0.000000
# time of last sample : 0.000000
#
Changelog:
v7: 1. Rebase to latest perf/core branch.
2. Add following clarification in patch description according to
Arnaldo's suggestion.
"That will be done when, for instance, processing build-ids,
where we already have to process all samples to create the
build-id table, take advantage of that to further amortize
that processing by storing HEADER_SAMPLE_TIME to make
'perf report/script' faster when using --time."
v4: Use perf script time style for timestamp printing. Also add with
the printing of sample duration.
v3: Remove the definitions of first_sample_time/last_sample_time from
perf_session. Just define them in perf_evlist
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1512738826-2628-2-git-send-email-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When a valid vmlinux is not found, 'perf report' falls back to look at
/proc/kcore. In this case, it will report the impossible large offset.
For example:
# perf record -b -e cycles:k find /etc/ > /dev/null
# perf report --stdio --branch-history
22.77% _vm_normal_page+18446603336221188162
|
---page_remove_rmap +18446603336221188324
page_remove_rmap +18446603336221188487 (cycles:5)
unlock_page_memcg +18446603336221188096
page_remove_rmap +18446603336221188327 (cycles:1)
The issue is the value which is passed to parameter 'addr' in
__get_srcline() is the objdump address. It's not correct if we calculate
the offset by using 'addr - sym->start'.
This patch creates a new parameter 'ip' in __get_srcline(). It is not
converted to objdump address.
With this patch, the perf report output is:
22.77% _vm_normal_page+66
|
---page_remove_rmap +228
page_remove_rmap +391 (cycles:5)
unlock_page_memcg +0
page_remove_rmap +231 (cycles:1)
page_remove_rmap +236
Committer testing:
Make sure you get any valid vmlinux out of the way, using '-v' on the
'perf report' case and deleting it from places where perf searches them,
like your kernel build dir and the build-id cache, in ~/.debug/.
Reported-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1514564812-17344-1-git-send-email-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
While monitoring a multithread process with pid option, perf sometimes
may return sys_perf_event_open failure with 3(No such process) if any of
the process's threads die before we open the event. However, we want
perf continue monitoring the remaining threads and do not exit with
error.
Here, the patch enables perf_evsel::ignore_missing_thread for -p option
to ignore complete failure if any of threads die before we open the event.
But it may still return sys_perf_event_open failure with 22(Invalid) if we
monitors several event groups.
sys_perf_event_open: pid 28960 cpu 40 group_fd 118202 flags 0x8
sys_perf_event_open: pid 28961 cpu 40 group_fd 118203 flags 0x8
WARNING: Ignored open failure for pid 28962
sys_perf_event_open: pid 28962 cpu 40 group_fd [118203] flags 0x8
sys_perf_event_open failed, error -22
That is because when we ignore a missing thread, we change the thread_idx
without dealing with its fds, FD(evsel, cpu, thread). Then get_group_fd()
may return a wrong group_fd for the next thread and sys_perf_event_open()
return with 22.
sys_perf_event_open(){
...
if (group_fd != -1)
perf_fget_light()//to get corresponding group_leader by group_fd
...
if (group_leader)
if (group_leader->ctx->task != ctx->task)//should on the same task
goto err_context
...
}
This patch also fixes this bug by introducing perf_evsel__remove_fd() and
update_fds to allow removing fds for the missing thread.
Changes since v1:
- Change group_fd__remove() into a more genetic way without changing code logic
- Remove redundant condition
Changes since v2:
- Use a proper function name and add some comment.
- Multiline comment style fixes.
Committer testing:
Before this patch the recently added 'perf stat --per-thread' for system
wide counting would race while enumerating all threads using /proc:
[root@jouet ~]# perf stat --per-thread
failed to parse CPUs map: No such file or directory
Usage: perf stat [<options>] [<command>]
-C, --cpu <cpu> list of cpus to monitor in system-wide
-a, --all-cpus system-wide collection from all CPUs
[root@jouet ~]# perf stat --per-thread
failed to parse CPUs map: No such file or directory
Usage: perf stat [<options>] [<command>]
-C, --cpu <cpu> list of cpus to monitor in system-wide
-a, --all-cpus system-wide collection from all CPUs
[root@jouet ~]#
When, say, the kernel was being built, so lots of shortlived threads,
after this patch this doesn't happen.
Signed-off-by: Mengting Zhang <zhangmengting@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Cheng Jian <cj.chengjian@huawei.com>
Cc: Li Bin <huawei.libin@huawei.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1513148513-6974-1-git-send-email-zhangmengting@huawei.com
[ Remove one use 'evlist' alias variable ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When we detect a different endianity we swap event before processing.
It's tricky for samples because we have no idea what's inside. We treat
it as an array of u64s, swap them and later on we swap back parts which
are different.
We mangle this way also the tracepoint raw data, which ends up in report
showing wrong data:
1.95% comm=Q^B pid=29285 prio=16777216 target_cpu=000
1.67% comm=l^B pid=0 prio=16777216 target_cpu=000
Luckily the traceevent library handles the endianity by itself (thank
you Steven!), so we can pass the RAW data directly in the other
endianity.
2.51% comm=beah-rhts-task pid=1175 prio=120 target_cpu=002
2.23% comm=kworker/0:0 pid=11566 prio=120 target_cpu=000
The fix is basically to swap back the raw data if different endianity is
detected.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/20171129184346.3656-1-jolsa@kernel.org
[ Add util/memswap.c to python-ext-sources to link missing mem_bswap_64() ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Support the special characters escaped by '\' in parser. This allows
user to specify versions directly like below.
=====
# ./perf probe -x /lib64/libc-2.25.so malloc_get_state\\@GLIBC_2.2.5
Added new event:
probe_libc:malloc_get_state (on malloc_get_state@GLIBC_2.2.5 in /usr/lib64/libc-2.25.so)
You can now use it in all perf tools, such as:
perf record -e probe_libc:malloc_get_state -aR sleep 1
=====
Or, you can use separators in source filename, e.g.
=====
# ./perf probe -x /opt/test/a.out foo+bar.c:3
Semantic error :There is non-digit character in offset.
Error: Command Parse Error.
=====
Usually "+" in source file cause parser error, but
=====
# ./perf probe -x /opt/test/a.out foo\\+bar.c:4
Added new event:
probe_a:main (on @foo+bar.c:4 in /opt/test/a.out)
You can now use it in all perf tools, such as:
perf record -e probe_a:main -aR sleep 1
=====
escaped "\+" allows you to specify that.
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: bhargavb <bhargavaramudu@gmail.com>
Cc: linux-rt-users@vger.kernel.org
Link: http://lkml.kernel.org/r/151309111236.18107.5634753157435343410.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Commit d80406453a ("perf symbols: Allow user probes on versioned
symbols") allows user to find default versioned symbols (with "@@") in
map. However, it did not enable normal versioned symbol (with "@") for
perf-probe. E.g.
=====
# ./perf probe -x /lib64/libc-2.25.so malloc_get_state
Failed to find symbol malloc_get_state in /usr/lib64/libc-2.25.so
Error: Failed to add events.
=====
This solves above issue by improving perf-probe symbol search function,
as below.
=====
# ./perf probe -x /lib64/libc-2.25.so malloc_get_state
Added new event:
probe_libc:malloc_get_state (on malloc_get_state in /usr/lib64/libc-2.25.so)
You can now use it in all perf tools, such as:
perf record -e probe_libc:malloc_get_state -aR sleep 1
# ./perf probe -l
probe_libc:malloc_get_state (on malloc_get_state@GLIBC_2.2.5 in /usr/lib64/libc-2.25.so)
=====
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: bhargavb <bhargavaramudu@gmail.com>
Cc: linux-rt-users@vger.kernel.org
Link: http://lkml.kernel.org/r/151275049269.24652.1639103455496216255.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add __return suffix for function return events automatically. Without
this, user have to give --force option and will see the number suffix
for each event like "function_1", which is not easy to recognize.
Instead, this adds __return suffix to it automatically. E.g.
=====
# ./perf probe -x /lib64/libc-2.25.so 'malloc*%return'
Added new events:
probe_libc:malloc_printerr__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_consolidate__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_check__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_hook_ini__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_trim__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_usable_size__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_stats__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_info__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:mallochook__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_get_state__return (on malloc*%return in /usr/lib64/libc-2.25.so)
probe_libc:malloc_set_state__return (on malloc*%return in /usr/lib64/libc-2.25.so)
You can now use it in all perf tools, such as:
perf record -e probe_libc:malloc_set_state__return -aR sleep 1
=====
Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org>
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: bhargavb <bhargavaramudu@gmail.com>
Cc: linux-rt-users@vger.kernel.org
Link: http://lkml.kernel.org/r/151275046418.24652.6696011972866498489.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cut off the version suffix (e.g. @GLIBC_2.2.5 etc.) from automatic
generated event name. This fixes wildcard event adding like below case;
=====
# perf probe -x /lib64/libc-2.25.so malloc*
Internal error: "malloc_get_state@GLIBC_2" is wrong event name.
Error: Failed to add events.
=====
This failure was caused by a versioned suffix symbol.
With this fix, perf probe automatically cuts the suffix after @ as
below.
=====
# ./perf probe -x /lib64/libc-2.25.so malloc*
Added new events:
probe_libc:malloc_printerr (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_consolidate (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_check (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_hook_ini (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_trim (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_usable_size (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_stats (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_info (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:mallochook (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_get_state (on malloc* in /usr/lib64/libc-2.25.so)
probe_libc:malloc_set_state (on malloc* in /usr/lib64/libc-2.25.so)
You can now use it in all perf tools, such as:
perf record -e probe_libc:malloc_set_state -aR sleep 1
=====
Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org>
Reported-by: bhargavb <bhargavaramudu@gmail.com>
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com>
Reviewed-by: Thomas Richter <tmricht@linux.vnet.ibm.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Clarke <pc@us.ibm.com>
Cc: linux-rt-users@vger.kernel.org
Link: http://lkml.kernel.org/r/None
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
