Previously arch_support_sort_key and arch_perf_header_entry used a
weak symbol to compile as appropriate for x86 and powerpc. A
limitation to this is that the handling of a data file could vary in
cross-platform development. Change to using the perf_env of the
current session to determine the architecture kind and set the sort
key and header entries as appropriate.
Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250724163302.596743-23-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
The session holds a perf_env pointer env. In UI code container_of is
used to turn the env to a session, but this assumes the session
header's env is in use. Rather than a dubious container_of, hold the
session in the evlist and derive the env from the session with
evsel__env, perf_session__env, etc.
Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250724163302.596743-11-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Calling perf top with branch filters enabled on Intel CPU's
with branch counters logging (A.K.A LBR event logging [1]) support
results in a segfault.
$ perf top -e '{cpu_core/cpu-cycles/,cpu_core/event=0xc6,umask=0x3,frontend=0x11,name=frontend_retired_dsb_miss/}' -j any,counter
...
Thread 27 "perf" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fffafff76c0 (LWP 949003)]
perf_env__find_br_cntr_info (env=0xf66dc0 <perf_env>, nr=0x0, width=0x7fffafff62c0) at util/env.c:653
653 *width = env->cpu_pmu_caps ? env->br_cntr_width :
(gdb) bt
#0 perf_env__find_br_cntr_info (env=0xf66dc0 <perf_env>, nr=0x0, width=0x7fffafff62c0) at util/env.c:653
#1 0x00000000005b1599 in symbol__account_br_cntr (branch=0x7fffcc3db580, evsel=0xfea2d0, offset=12, br_cntr=8) at util/annotate.c:345
#2 0x00000000005b17fb in symbol__account_cycles (addr=5658172, start=5658160, sym=0x7fffcc0ee420, cycles=539, evsel=0xfea2d0, br_cntr=8) at util/annotate.c:389
#3 0x00000000005b1976 in addr_map_symbol__account_cycles (ams=0x7fffcd7b01d0, start=0x7fffcd7b02b0, cycles=539, evsel=0xfea2d0, br_cntr=8) at util/annotate.c:422
#4 0x000000000068d57f in hist__account_cycles (bs=0x110d288, al=0x7fffafff6540, sample=0x7fffafff6760, nonany_branch_mode=false, total_cycles=0x0, evsel=0xfea2d0) at util/hist.c:2850
#5 0x0000000000446216 in hist_iter__top_callback (iter=0x7fffafff6590, al=0x7fffafff6540, single=true, arg=0x7fffffff9e00) at builtin-top.c:737
#6 0x0000000000689787 in hist_entry_iter__add (iter=0x7fffafff6590, al=0x7fffafff6540, max_stack_depth=127, arg=0x7fffffff9e00) at util/hist.c:1359
#7 0x0000000000446710 in perf_event__process_sample (tool=0x7fffffff9e00, event=0x110d250, evsel=0xfea2d0, sample=0x7fffafff6760, machine=0x108c968) at builtin-top.c:845
#8 0x0000000000447735 in deliver_event (qe=0x7fffffffa120, qevent=0x10fc200) at builtin-top.c:1211
#9 0x000000000064ccae in do_flush (oe=0x7fffffffa120, show_progress=false) at util/ordered-events.c:245
#10 0x000000000064d005 in __ordered_events__flush (oe=0x7fffffffa120, how=OE_FLUSH__TOP, timestamp=0) at util/ordered-events.c:324
#11 0x000000000064d0ef in ordered_events__flush (oe=0x7fffffffa120, how=OE_FLUSH__TOP) at util/ordered-events.c:342
#12 0x00000000004472a9 in process_thread (arg=0x7fffffff9e00) at builtin-top.c:1120
#13 0x00007ffff6e7dba8 in start_thread (arg=<optimized out>) at pthread_create.c:448
#14 0x00007ffff6f01b8c in __GI___clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:78
The cause is that perf_env__find_br_cntr_info tries to access a
null pointer pmu_caps in the perf_env struct. A similar issue exists
for homogeneous core systems which use the cpu_pmu_caps structure.
Fix this by populating cpu_pmu_caps and pmu_caps structures with
values from sysfs when calling perf top with branch stack sampling
enabled.
[1], LBR event logging introduced here:
https://lore.kernel.org/all/20231025201626.3000228-5-kan.liang@linux.intel.com/
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
Link: https://lore.kernel.org/r/20250612163659.1357950-2-thomas.falcon@intel.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
It's a prepartion to support code annotation and data type
annotation at the same time. Data type annotation needs more
information in the hist_entry so it needs to be passed deeper.
Also rename a function with the same name in the builtin-annotate.c
to hist_entry__stdio_annotate since it matches better to the command
line option. And change the condition inside to be simpler.
Reviewed-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250310224925.799005-5-namhyung@kernel.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
The BPF sideband information is tracked using a separate thread and
evlist. But it's only useful for profiling kernel and we can skip it
when users profile their application only.
It seems it already fails to open the sideband event in that case.
Let's remove the noise in the verbose output anyway.
Reviewed-by: Ian Rogers <irogers@google.com>
Acked-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20250226203039.1099131-1-namhyung@kernel.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
The struct dump_regs contains 512 bytes of cache_regs, meaning the two
values in perf_sample contribute 1088 bytes of its total 1384 bytes
size. Initializing this much memory has a cost reported by Tavian
Barnes <tavianator@tavianator.com> as about 2.5% when running `perf
script --itrace=i0`:
https://lore.kernel.org/lkml/d841b97b3ad2ca8bcab07e4293375fb7c32dfce7.1736618095.git.tavianator@tavianator.com/
Adrian Hunter <adrian.hunter@intel.com> replied that the zero
initialization was necessary and couldn't simply be removed.
This patch aims to strike a middle ground of still zeroing the
perf_sample, but removing 79% of its size by make user_regs and
intr_regs optional pointers to zalloc-ed memory. To support the
allocation accessors are created for user_regs and intr_regs. To
support correct cleanup perf_sample__init and perf_sample__exit
functions are created and added throughout the code base.
Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250113194345.1537821-1-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Recently we got a case where a kernel sample wasn't being resolved due
to a bug that was not setting the end address on kernel functions
implemented in assembly (see Link: tag), and then those were not being
found by machine__resolve() -> map__find_symbol().
So we ended up with:
# perf top --stdio
PerfTop: 0 irqs/s kernel: 0% exact: 0% lost: 0/0 drop: 0/0 [cycles/P]
-----------------------------------------------------------------------
Warning:
A vmlinux file was not found.
Kernel samples will not be resolved.
^Z
[1]+ Stopped perf top --stdio
#
But then resolving all other kernel symbols.
So just fixup the logic to only print that warning when there are no
symbols in the kernel map.
Fixes: d88205db9c ("perf dso: Add dso__has_symbols() method")
Reviewed-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Link: https://lore.kernel.org/lkml/Z3buKhcCsZi3_aGb@x1
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When annotating a basic block, it's useful to display the occurrences
of other events in the block.
The branch counter feature is only available for newer Intel platforms.
So a dedicated option to display the branch counters is not introduced.
Reuse the existing --total-cycles option, which triggers the annotation
of a basic block and displays the cycle-related annotation.
When the branch counters information is available, the branch counters
are automatically appended after all the cycle-related annotation.
Accounting the branch counters as well when accounting the cycles in
hist__account_cycles().
In 'struct annotated_branch', introduce a br_cntr array to save the
accumulation of each branch counter.
In a sample, all the branch counters for a branch are saved in a u64
space.
Because the saturation of a branch counter is small, e.g., for Intel
Sierra Forest, the saturation is only 3.
Add ANNOTATION__BR_CNTR_SATURATED_FLAG to indicate if a branch counter
once saturated. That can be used to indicate a potential event lost
because of the saturation.
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: https://lore.kernel.org/r/20240813160208.2493643-5-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Allow filters to be added to perf top events. One use is to workaround
issues with:
```
$ perf top --uid="$(id -u)"
```
which tries to scan /proc find processes belonging to the uid and can
fail in such a pid terminates between the scan and the
perf_event_open reporting:
```
Error:
The sys_perf_event_open() syscall returned with 3 (No such process) for event (cycles:P).
/bin/dmesg | grep -i perf may provide additional information.
```
A similar filter:
```
$ perf top -e cycles:P --filter "uid == $(id -u)"
```
doesn't fail this way.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Changbin Du <changbin.du@huawei.com>
Cc: Yang Jihong <yangjihong1@huawei.com>
Cc: Andrii Nakryiko <andrii@kernel.org>
Cc: bpf@vger.kernel.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20240524205227.244375-4-irogers@google.com
It's hard to distinguish the default cycles events among hybrid PMUs.
For example,
$ perf top
Available samples
385 cycles:P
903 cycles:P
The other tool, e.g., perf record, uniforms the event name and adds the
hybrid PMU name before opening the event. So the events can be easily
distinguished. Apply the same methodology for the perf top as well.
The evlist__uniquify_name() will be invoked by both record and top.
Move it to util/evlist.c
With the patch:
$ perf top
Available samples
148 cpu_atom/cycles:P/
1K cpu_core/cycles:P/
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: Hector Martin <marcan@marcan.st>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20231214144612.1092028-2-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf top errors out on a hybrid machine
$perf top
Error:
The cycles:P event is not supported.
The perf top expects that the "cycles" is collected on all CPUs in the
system. But for hybrid there is no single "cycles" event which can cover
all CPUs. Perf has to split it into two cycles events, e.g.,
cpu_core/cycles/ and cpu_atom/cycles/. Each event has its own CPU mask.
If a event is opened on the unsupported CPU. The open fails. That's the
reason of the above error out.
Perf should only open the cycles event on the corresponding CPU. The
commit ef91871c96 ("perf evlist: Propagate user CPU maps intersecting
core PMU maps") intersect the requested CPU map with the CPU map of the
PMU. Use the evsel's cpus to replace user_requested_cpus.
The evlist's threads are also propagated to the evsel's threads in
__perf_evlist__propagate_maps(). For a system-wide event, perf appends
a dummy event and assign it to the evsel's threads. For a per-thread
event, the evlist's thread_map is assigned to the evsel's threads. The
same as the other tools, e.g., perf record, using the evsel's threads
when opening an event.
Reported-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: Hector Martin <marcan@marcan.st>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Closes: https://lore.kernel.org/linux-perf-users/ZXNnDrGKXbEELMXV@kernel.org/
Link: https://lore.kernel.org/r/20231214144612.1092028-1-kan.liang@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
While debugging a segfault on 'perf lock contention' without an
available perf.data file I noticed that it was basically calling:
perf_session__delete(ERR_PTR(-1))
Resulting in:
(gdb) run lock contention
Starting program: /root/bin/perf lock contention
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib64/libthread_db.so.1".
failed to open perf.data: No such file or directory (try 'perf record' first)
Initializing perf session failed
Program received signal SIGSEGV, Segmentation fault.
0x00000000005e7515 in auxtrace__free (session=0xffffffffffffffff) at util/auxtrace.c:2858
2858 if (!session->auxtrace)
(gdb) p session
$1 = (struct perf_session *) 0xffffffffffffffff
(gdb) bt
#0 0x00000000005e7515 in auxtrace__free (session=0xffffffffffffffff) at util/auxtrace.c:2858
#1 0x000000000057bb4d in perf_session__delete (session=0xffffffffffffffff) at util/session.c:300
#2 0x000000000047c421 in __cmd_contention (argc=0, argv=0x7fffffffe200) at builtin-lock.c:2161
#3 0x000000000047dc95 in cmd_lock (argc=0, argv=0x7fffffffe200) at builtin-lock.c:2604
#4 0x0000000000501466 in run_builtin (p=0xe597a8 <commands+552>, argc=2, argv=0x7fffffffe200) at perf.c:322
#5 0x00000000005016d5 in handle_internal_command (argc=2, argv=0x7fffffffe200) at perf.c:375
#6 0x0000000000501824 in run_argv (argcp=0x7fffffffe02c, argv=0x7fffffffe020) at perf.c:419
#7 0x0000000000501b11 in main (argc=2, argv=0x7fffffffe200) at perf.c:535
(gdb)
So just set it to NULL after using PTR_ERR(session) to decode the error
as perf_session__delete(NULL) is supported.
The same problem was found in 'perf top' after an audit of all
perf_session__new() failure handling.
Fixes: 6ef81c55a2 ("perf session: Return error code for perf_session__new() function on failure")
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexey Budankov <alexey.budankov@linux.intel.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jeremie Galarneau <jeremie.galarneau@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kate Stewart <kstewart@linuxfoundation.org>
Cc: Mamatha Inamdar <mamatha4@linux.vnet.ibm.com>
Cc: Mukesh Ojha <mojha@codeaurora.org>
Cc: Nageswara R Sastry <rnsastry@linux.vnet.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ravi Bangoria <ravi.bangoria@linux.ibm.com>
Cc: Shawn Landden <shawn@git.icu>
Cc: Song Liu <songliubraving@fb.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tzvetomir Stoyanov <tstoyanov@vmware.com>
Link: https://lore.kernel.org/lkml/ZN4Q2rxxsL08A8rd@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In __cmd_top(), perf_set_multithreaded() is used to enable
pthread_rwlock, thus down_read() and down_write () are not nops,
handling concurrency problems
Then 'perf top' uses perf_set_singlethreaded(), switching to the single
threaded phase, assuming that no thread concurrency will happen later.
However, a use after free problem could occur in the single threaded
phase, the concurrent procedure is this:
display_thread process_thread
-------------- --------------
thread__comm_len
-> thread__comm_str
-> __thread__comm_str(thread)
thread__delete
-> comm__free
-> comm_str__put
-> zfree(&cs->str)
-> thread->comm_len = strlen(comm);
Since in single thread phase, perf_singlethreaded is true, down_read()
and down_write() do nothing to avoid concurrency problems.
This patch moves the perf_set_singlethreaded() call to the function tail
to expand the multithreaded phase range, making display_thread() and
process_thread() concurrency safe.
Reviewed-by: Yunfeng Ye <yeyunfeng@huawei.com>
Signed-off-by: Hangliang Lai <laihangliang1@huawei.com>
Co-developed-by: Wenyu Liu <liuwenyu7@huawei.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Feilong Lin <linfeilong@huawei.com>
Cc: Hewenliang <hewenliang4@huawei.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Link: https://lore.kernel.org/r/20230411013224.2079-1-laihangliang1@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Rename the fallthrough attribute to better align with the kernel
version. Copy the definition from include/linux/compiler_attributes.h
including the #else clause. Adding the #else clause allows the tools
compiler.h header to drop the check for a definition entirely and keeps
both definitions together.
Change any __fallthrough statements to fallthrough anywhere it was used
within perf.
This allows other tools to use the same key word as the kernel.
Committer notes:
Did some missing conversions to:
builtin-list.c
Also included gtk.h before the 'fallthrough' definition in:
tools/perf/ui/gtk/hists.c
tools/perf/ui/gtk/helpline.c
tools/perf/ui/gtk/browser.c
As it is the arg name for a macro in glib.h:
/var/home/acme/git/perf-tools-next/tools/include/linux/compiler-gcc.h:16:55: error: missing binary operator before token "("
16 | # define fallthrough __attribute__((__fallthrough__))
| ^
/usr/include/glib-2.0/glib/gmacros.h:637:28: note: in expansion of macro ‘fallthrough’
637 | #if g_macro__has_attribute(fallthrough)
Reviewed-by: Miguel Ojeda <ojeda@kernel.org>
Signed-off-by: Liam Howlett <Liam.Howlett@oracle.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Miguel Ojeda <ojeda@kernel.org>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: Tom Rix <trix@redhat.com>
Cc: linux-sparse@vger.kernel.org <linux-sparse@vger.kernel.org>
Cc: llvm@lists.linux.dev <llvm@lists.linux.dev>
Link: https://lore.kernel.org/r/20221125154947.2163498-1-Liam.Howlett@oracle.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
