The "perf all PMU test" fails on a Coffee Lake machine.
The failure is caused by the below change in the commit e2641db83f
("perf vendor events: Add/update skylake events/metrics").
+ {
+ "BriefDescription": "This 48-bit fixed counter counts the UCLK cycles",
+ "Counter": "FIXED",
+ "EventCode": "0xff",
+ "EventName": "UNC_CLOCK.SOCKET",
+ "PerPkg": "1",
+ "PublicDescription": "This 48-bit fixed counter counts the UCLK cycles.",
+ "Unit": "cbox_0"
}
The other cbox events have the unit name "CBOX", while the fixed counter
has a unit name "cbox_0". So the events_table will maintain separate
entries for cbox and cbox_0.
The perf_pmus__print_pmu_events() calculates the total number of events,
allocate an aliases buffer, store all the events into the buffer, sort,
and print all the aliases one by one.
The problem is that the calculated total number of events doesn't match
the stored events in the aliases buffer.
The perf_pmu__num_events() is used to calculate the number of events. It
invokes the pmu_events_table__num_events() to go through the entire
events_table to find all events. Because of the
pmu_uncore_alias_match(), the suffix of uncore PMU will be ignored. So
the events for cbox and cbox_0 are all counted.
When storing events into the aliases buffer, the
perf_pmu__for_each_event() only process the events for cbox.
Since a bigger buffer was allocated, the last entry are all 0.
When printing all the aliases, null will be outputted, and trigger the
failure.
The mismatch was introduced from the commit e3edd6cf63 ("perf
pmu-events: Reduce processed events by passing PMU"). The
pmu_events_table__for_each_event() stops immediately once a pmu is set.
But for uncore, especially this case, the method is wrong and mismatch
what perf does in the perf_pmu__num_events().
With the patch,
$ perf list pmu | grep -A 1 clock.socket
unc_clock.socket
[This 48-bit fixed counter counts the UCLK cycles. Unit: uncore_cbox_0
$ perf test "perf all PMU test"
107: perf all PMU test : Ok
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/all/202407101021.2c8baddb-oliver.sang@intel.com/
Fixes: e3edd6cf63 ("perf pmu-events: Reduce processed events by passing PMU")
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Sandipan Das <sandipan.das@amd.com>
Cc: Benjamin Gray <bgray@linux.ibm.com>
Cc: Xu Yang <xu.yang_2@nxp.com>
Cc: John Garry <john.g.garry@oracle.com>
Link: https://lore.kernel.org/r/20241001021431.814811-1-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
when running below perf command, we say error is reported.
perf record -e "{slots,instructions,topdown-retiring}:S" -vv -C0 sleep 1
------------------------------------------------------------
perf_event_attr:
type 4 (cpu)
size 168
config 0x400 (slots)
sample_type IP|TID|TIME|READ|CPU|PERIOD|IDENTIFIER
read_format ID|GROUP|LOST
disabled 1
sample_id_all 1
exclude_guest 1
------------------------------------------------------------
sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8 = 5
------------------------------------------------------------
perf_event_attr:
type 4 (cpu)
size 168
config 0x8000 (topdown-retiring)
{ sample_period, sample_freq } 4000
sample_type IP|TID|TIME|READ|CPU|PERIOD|IDENTIFIER
read_format ID|GROUP|LOST
freq 1
sample_id_all 1
exclude_guest 1
------------------------------------------------------------
sys_perf_event_open: pid -1 cpu 0 group_fd 5 flags 0x8
sys_perf_event_open failed, error -22
Error:
The sys_perf_event_open() syscall returned with 22 (Invalid argument) for
event (topdown-retiring).
The reason of error is that the events are regrouped and
topdown-retiring event is moved to closely after the slots event and
topdown-retiring event needs to do the sampling, but Intel PMU driver
doesn't support to sample topdown metrics events.
For topdown metrics events, it just requires to be in a group which has
slots event as leader. It doesn't require topdown metrics event must be
closely after slots event. Thus it's a overkill to move topdown metrics
event closely after slots event in events regrouping and furtherly cause
the above issue.
Thus don't move topdown metrics events forward if they are already in a
group.
Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Cc: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lore.kernel.org/r/20240913084712.13861-4-dapeng1.mi@linux.intel.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Addresses an issue where, in the absence of a topdown metrics event
within a sampling group, the slots event was incorrectly bypassed as
the sampling leader when sample_read was enabled.
perf record -e '{slots,branches}:S' -c 10000 -vv sleep 1
In this case, the slots event should be sampled as leader but the
branches event is sampled in fact like the verbose output shows.
perf_event_attr:
type 4 (cpu)
size 168
config 0x400 (slots)
sample_type IP|TID|TIME|READ|CPU|IDENTIFIER
read_format ID|GROUP|LOST
disabled 1
sample_id_all 1
exclude_guest 1
------------------------------------------------------------
sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8 = 5
------------------------------------------------------------
perf_event_attr:
type 0 (PERF_TYPE_HARDWARE)
size 168
config 0x4 (PERF_COUNT_HW_BRANCH_INSTRUCTIONS)
{ sample_period, sample_freq } 10000
sample_type IP|TID|TIME|READ|CPU|IDENTIFIER
read_format ID|GROUP|LOST
sample_id_all 1
exclude_guest 1
The sample period of slots event instead of branches event is reset to
0.
This fix ensures the slots event remains the leader under these
conditions.
Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Cc: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lore.kernel.org/r/20240913084712.13861-3-dapeng1.mi@linux.intel.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
It's not complete to check whether an event is a topdown slots or
topdown metrics event by only comparing the event name since user
may assign the event by RAW format, e.g.
perf stat -e '{instructions,cpu/r400/,cpu/r8300/}' sleep 1
Performance counter stats for 'sleep 1':
<not counted> instructions
<not counted> cpu/r400/
<not supported> cpu/r8300/
1.002917796 seconds time elapsed
0.002955000 seconds user
0.000000000 seconds sys
The RAW format slots and topdown-be-bound events are not recognized and
not regroup the events, and eventually cause error.
Thus add two helpers arch_is_topdown_slots()/arch_is_topdown_metrics()
to detect whether an event is topdown slots/metrics event by comparing
the event config directly, and use these two helpers to replace the
original event name comparisons.
Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Cc: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lore.kernel.org/r/20240913084712.13861-2-dapeng1.mi@linux.intel.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
perf test 70 takes a long time. One culprit is the output of command
perf annotate. Per default enabled are
- demangle symbol names
- interleave source code with assembly code.
Disable demangle of symbols and abort the annotation
after the first 250 lines.
This speeds up the test case considerable, for example
on s390:
Output before:
# time perf test 70
70: perf annotate basic tests : Ok
.....
real 2m7.467s
user 1m26.869s
sys 0m34.086s
#
Output after:
# time perf test 70
70: perf annotate basic tests : Ok
real 0m3.341s
user 0m1.606s
sys 0m0.362s
#
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com>
Reviewed-by: James Clark <james.clark@linaro.org>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: sumanthk@linux.ibm.com
Link: https://lore.kernel.org/r/20240917085706.249691-1-tmricht@linux.ibm.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
With commit 8ec9497d3e ("tools/include: Sync uapi/linux/perf.h
with the kernel sources"), 'perf mem report' gives an incorrect memory
access string.
...
0.02% 1 3644 L5 hit [.] 0x0000000000009b0e mlc [.] 0x00007fce43f59480
...
This occurs because, if no entry exists in mem_lvlnum, perf_mem__lvl_scnprintf
will default to 'L%d, lvl', which in this case for PERF_MEM_LVLNUM_L2_MHB is 0x05.
Add entries for PERF_MEM_LVLNUM_L2_MHB and PERF_MEM_LVLNUM_MSC to mem_lvlnum,
so that the correct strings are printed.
...
0.02% 1 3644 L2 MHB hit [.] 0x0000000000009b0e mlc [.] 0x00007fce43f59480
...
Fixes: 8ec9497d3e ("tools/include: Sync uapi/linux/perf.h with the kernel sources")
Suggested-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
Reviewed-by: Leo Yan <leo.yan@arm.com>
Link: https://lore.kernel.org/r/20240926144040.77897-1-thomas.falcon@intel.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
The sleep_sem semaphore and the specific_wait field (member of sched_atom)
are initialized but not used anywhere in the code, so this patch removes
them.
The SCHED_EVENT_MIGRATION case in perf_sched__process_event() is currently
not used and is also removed.
Additionally, prev_state in add_sched_event_sleep() is marked with
__maybe_unused and is not utilized anywhere in the function. This patch
removes the parameter.
If the task_state parameter was intended for future use, it can be
reintroduced when needed.
No functionality change intended.
Signed-off-by: Madadi Vineeth Reddy <vineethr@linux.ibm.com>
Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Link: https://lore.kernel.org/r/20240917090100.42783-1-vineethr@linux.ibm.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
install_doc of tools/lib/perf/Makefile invokes install-man,
install-html, and install-examples of
tools/lib/perf/Documentation/Makefile at once. This invocation succeeds
when make runs in serial but can fail when make runs in parallel because
while install-man of tools/lib/perf/Documentation/Makefile depends on
all, install-html depends on nothing and can run ahead of all.
Explicitly specify the dependencies of install-html to ensure that
they are resolved before install-html.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Link: https://lore.kernel.org/r/20240915-perf-v1-1-cbfd9cd1d482@daynix.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
When it loads symbols from an ELF file, it loads label symbols which is
0 size. Sometimes it has the same address with other symbols and might
shadow the original symbols because it fixes up the size of the symbol.
For example, in my system __do_softirq is shadowed and only accepts the
__softirqentry_text_start instead. But it should accept __do_softirq.
$ readelf -sW vmlinux | grep -e __do_softirq -e __softirqentry_text_start
105089: ffffffff82000000 814 FUNC GLOBAL DEFAULT 1 __do_softirq
111954: ffffffff82000000 0 NOTYPE GLOBAL DEFAULT 1 __softirqentry_text_start
$ perf annotate --stdio __do_softirq
Error:
The perf.data data has no samples!
$ perf annotate --stdio __softirqentry_text_start | head
Percent | Source code & Disassembly of vmlinux for cycles (26 samples, percent: local period)
---------------------------------------------------------------------------------------------------
: 0 0xffffffff82000000 <__softirqentry_text_start>:
0.00 : ffffffff82000000: nopl (%rax,%rax)
30.77 : ffffffff82000005: pushq %rbp
3.85 : ffffffff82000006: movq %rsp, %rbp
0.00 : ffffffff82000009: pushq %r15
3.85 : ffffffff8200000b: pushq %r14
3.85 : ffffffff8200000d: pushq %r13
0.00 : ffffffff8200000f: pushq %r12
We can ignore NOTYPE symbols in the symbols__fixup_end() so that it can
pick the __do_softirq() in choose_best_symbol(). This should be fine
since most symbols have either STT_FUNC or STT_OBJECT.
Link: https://lore.kernel.org/r/20240912224208.3360116-1-namhyung@kernel.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Exit when run_perf_stat() returns an error to avoid continuously
repeating the same error message. It's not expected that COUNTER_FATAL
or internal errors are recoverable so there's no point in retrying.
This fixes the following flood of error messages for permission issues,
for example when perf_event_paranoid==3:
perf stat -r 1044 -- false
Error:
Access to performance monitoring and observability operations is limited.
...
Error:
Access to performance monitoring and observability operations is limited.
...
(repeating for 1044 times).
Signed-off-by: Levi Yun <yeoreum.yun@arm.com>
Reviewed-by: James Clark <james.clark@linaro.org>
Cc: nd@arm.com
Cc: howardchu95@gmail.com
Link: https://lore.kernel.org/r/20240925132022.2650180-3-yeoreum.yun@arm.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
cs_etm__flush(), like cs_etm__sample() is an operation that generates a
sample and then swaps the current with the previous packet. Calling
flush after processing the queues results in two swaps which corrupts
the next sample. Therefore it wasn't appropriate to call flush here so
remove it.
Flushing is still done on a discontinuity to explicitly clear the last
branch buffer, but when the packet_queue fills up before reaching a
timestamp, that's not a discontinuity and the call to
cs_etm__process_traceid_queue() already generated samples and drained
the buffers correctly.
This is visible by looking for a branch that has the same target as the
previous branch and the following source is before the address of the
last target, which is impossible as execution would have had to have
gone backwards:
ffff800080849d40 _find_next_and_bit+0x78 => ffff80008011cadc update_sg_lb_stats+0x94
(packet_queue fills here before a timestamp, resulting in a flush and
branch target ffff80008011cadc is duplicated.)
ffff80008011cb1c update_sg_lb_stats+0xd4 => ffff80008011cadc update_sg_lb_stats+0x94
ffff8000801117c4 cpu_util+0x24 => ffff8000801117d4 cpu_util+0x34
After removing the flush the correct branch target is used for the
second sample, and ffff8000801117c4 is no longer before the previous
address:
ffff800080849d40 _find_next_and_bit+0x78 => ffff80008011cadc update_sg_lb_stats+0x94
ffff80008011cb1c update_sg_lb_stats+0xd4 => ffff8000801117a0 cpu_util+0x0
ffff8000801117c4 cpu_util+0x24 => ffff8000801117d4 cpu_util+0x34
Make sure that a final branch stack is output at the end of the trace
by calling cs_etm__end_block(). This is already done for both the
timeless decode paths.
Fixes: 21fe8dc119 ("perf cs-etm: Add support for CPU-wide trace scenarios")
Reported-by: Ganapatrao Kulkarni <gankulkarni@os.amperecomputing.com>
Closes: https://lore.kernel.org/all/20240719092619.274730-1-gankulkarni@os.amperecomputing.com/
Reviewed-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: James Clark <james.clark@linaro.org>
Tested-by: Ganapatrao Kulkarni <gankulkarni@os.amperecomputing.com>
Cc: Ben Gainey <ben.gainey@arm.com>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Ruidong Tian <tianruidong@linux.alibaba.com>
Cc: Benjamin Gray <bgray@linux.ibm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: coresight@lists.linaro.org
Cc: John Garry <john.g.garry@oracle.com>
Cc: scclevenger@os.amperecomputing.com
Link: https://lore.kernel.org/r/20240916135743.1490403-2-james.clark@linaro.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
With that it uses the generic BTF based pretty printer:
This one we need to think about, not being acquainted with this syscall,
should we _traverse_ that list somehow? Would that be useful?
root@number:~# perf trace -e set_robust_list sleep 1
0.000 ( 0.004 ms): sleep/1206493 set_robust_list(head: (struct robust_list_head){.list = (struct robust_list){.next = (struct robust_list *)0x7f48a9a02a20,},.futex_offset = (long int)-32,}, len: 24) =
root@number:~#
strace prints the default integer args:
root@number:~# strace -e set_robust_list sleep 1
set_robust_list(0x7efd99559a20, 24) = 0
+++ exited with 0 +++
root@number:~#
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alan Maguire <alan.maguire@oracle.com>
Cc: Howard Chu <howardchu95@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org
Link: https://lore.kernel.org/lkml/ZuH6MquMraBvODRp@x1
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
No event is printed in the "Branch Counter" column on hybrid machines.
For example,
$ perf record -e "{cpu_core/branch-instructions/pp,cpu_core/branches/}:S" -j any,counter
$ perf report --total-cycles
# Branch counter abbr list:
# cpu_core/branch-instructions/pp = A
# cpu_core/branches/ = B
# '-' No event occurs
# '+' Event occurrences may be lost due to branch counter saturated
#
# Sampled Cycles% Sampled Cycles Avg Cycles% Avg Cycles Branch Counter
# ............... .............. ........... .......... ..............
44.54% 727.1K 0.00% 1 |+ |+ |
36.31% 592.7K 0.00% 2 |+ |+ |
17.83% 291.1K 0.00% 1 |+ |+ |
The branch counter information (br_cntr_width and br_cntr_nr) in the
perf_env is retrieved from the CPU_PMU_CAPS. However, the CPU_PMU_CAPS
is not available on hybrid machines. Without the width information, the
number of occurrences of an event cannot be calculated.
For a hybrid machine, the caps information should be retrieved from the
PMU_CAPS, and stored in the perf_env->pmu_caps.
Add a perf_env__find_br_cntr_info() to return the correct branch counter
information from the corresponding fields.
Committer notes:
While testing I couldn't s ee those "Branch counter" columns enabled by
pressing 'B' on the TUI, after reporting it to the list Kan explained
the situation:
<quote Kan Liang>
For a hybrid client, the "Branch Counter" feature is only supported
starting from the just released Lunar Lake. Perf falls back to only
"ANY" on your Raptor Lake.
The "The branch counter is not available" message is expected.
Here is the 'perf evlist' result from my Lunar Lake machine,
# perf evlist -v
cpu_core/branch-instructions/pp: type: 4 (cpu_core), size: 136, config: 0xc4 (branch-instructions), { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|READ|PERIOD|BRANCH_STACK|IDENTIFIER, read_format: ID|GROUP|LOST, disabled: 1, freq: 1, enable_on_exec: 1, precise_ip: 2, sample_id_all: 1, exclude_guest: 1, branch_sample_type: ANY|COUNTERS
#
</quote>
Fixes: 6f9d8d1de2 ("perf script: Add branch counters")
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20240909184201.553519-1-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
An event group is a critical relationship. There is a -g option that can
display the relationship. But it's hard for a user to know when should
this option be applied.
If there is an event group in the perf record, print a hint to suggest
the user apply the -g to display the group information.
With the patch,
$ perf record -e "{cycles,instructions},instructions" sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.024 MB perf.data (4 samples) ]
$
$ perf evlist
cycles
instructions
instructions
# Tip: use 'perf evlist -g' to show group information
$ perf evlist -g
{cycles,instructions}
instructions
$
Committer testing:
So for a perf.data file _with_ a group:
root@number:~# perf evlist -g
{cpu_core/branch-instructions/pp,cpu_core/branches/}
dummy:u
root@number:~# perf evlist
cpu_core/branch-instructions/pp
cpu_core/branches/
dummy:u
# Tip: use 'perf evlist -g' to show group information
root@number:~#
Then for something _without_ a group, no hint:
root@number:~# perf record ls
<SNIP>
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.035 MB perf.data (7 samples) ]
root@number:~# perf evlist
cpu_atom/cycles/P
cpu_core/cycles/P
dummy:u
root@number:~#
No suggestion, good.
Suggested-by: Arnaldo Carvalho de Melo <acme@kernel.org>
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Closes: https://lore.kernel.org/lkml/ZttgvduaKsVn1r4p@x1/
Link: https://lore.kernel.org/r/20240908202847.176280-1-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
To make error messages more accurate, this change detects whether ftrace is
enabled on system by checking trace file "set_ftrace_pid".
Before:
# perf ftrace
failed to reset ftrace
#
After:
# perf ftrace
ftrace is not supported on this system
#
Committer testing:
Doing it in an unprivileged toolbox container on Fedora 40:
Before:
acme@number:~/git/perf-tools-next$ toolbox enter perf
⬢[acme@toolbox perf-tools-next]$ sudo su -
⬢[root@toolbox ~]# ~acme/bin/perf ftrace
failed to reset ftrace
⬢[root@toolbox ~]#
After this patch:
⬢[root@toolbox ~]# ~acme/bin/perf ftrace
ftrace is not supported on this system
⬢[root@toolbox ~]#
Maybe we could check if we are in such as situation, inside an
unprivileged container, and provide a HINT line?
Reviewed-by: James Clark <james.clark@linaro.org>
Signed-off-by: Changbin Du <changbin.du@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Changbin Du <changbin.du@huawei.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20240911100126.900779-1-changbin.du@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
