Its more intention revealing, and if we're interested in the odd cases
where this may end up truncating we can do debug checks at one
centralized place.
Motivation, of all the container builds, fedora rawhide started
complaining of:
util/machine.c: In function ‘machine__create_modules’:
util/machine.c:1419:50: error: ‘%s’ directive output may be truncated writing up to 255 bytes into a region of size between 0 and 4095 [-Werror=format-truncation=]
1419 | snprintf(path, sizeof(path), "%s/%s", dir_name, dent->d_name);
| ^~
In file included from /usr/include/stdio.h:894,
from util/branch.h:9,
from util/callchain.h:8,
from util/machine.c:7:
In function ‘snprintf’,
inlined from ‘maps__set_modules_path_dir’ at util/machine.c:1419:3,
inlined from ‘machine__set_modules_path’ at util/machine.c:1473:9,
inlined from ‘machine__create_modules’ at util/machine.c:1519:7:
/usr/include/bits/stdio2.h:71:10: note: ‘__builtin___snprintf_chk’ output between 2 and 4352 bytes into a destination of size 4096
There are other places where we should use path__join(), but lets get rid of
this one first.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Acked-by: Ian Rogers <irogers@google.com>
Link: Link: https://lore.kernel.org/r/YebZKjwgfdOz0lAs@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When the target is a pid, not started by 'perf stat' we need to disable
the events, and in that case there is no need to setup affinities as we
use a dummy CPU map, with just one entry set to -1.
So stop doing it to avoid this needless call to sched_getaffinity():
# strace -ke sched_getaffinity perf stat -e cycles -p 241957 sleep 1
<SNIP>
sched_getaffinity(0, 512, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31]) = 8
> /usr/lib64/libc-2.33.so(sched_getaffinity@@GLIBC_2.3.4+0x1a) [0xe6eea]
> /var/home/acme/bin/perf(affinity__setup+0x6a) [0x532a2a]
> /var/home/acme/bin/perf(__evlist__disable.constprop.0+0x27) [0x4b9827]
> /var/home/acme/bin/perf(cmd_stat+0x29b5) [0x431725]
> /var/home/acme/bin/perf(run_builtin+0x6a) [0x4a2cfa]
> /var/home/acme/bin/perf(main+0x612) [0x40f8c2]
> /usr/lib64/libc-2.33.so(__libc_start_main+0xd4) [0x27b74]
> /var/home/acme/bin/perf(_start+0x2d) [0x40fadd]
<SNIP>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20220117160931.1191712-5-acme@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When the target is a pid, not started by 'perf stat' we need to enable
the events, and in that case there is no need to setup affinities as we
use a dummy CPU map, with just one entry set to -1.
So stop doing it to avoid this needless call to sched_getaffinity():
# strace -ke sched_getaffinity perf stat -e cycles -p 241957 sleep 1
<SNIP>
sched_getaffinity(0, 512, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31]) = 8
> /usr/lib64/libc-2.33.so(sched_getaffinity@@GLIBC_2.3.4+0x1a) [0xe6eea]
> /var/home/acme/bin/perf(affinity__setup+0x6a) [0x5329ca]
> /var/home/acme/bin/perf(__evlist__enable.constprop.0+0x23) [0x4b9693]
> /var/home/acme/bin/perf(enable_counters+0x14d) [0x42de5d]
> /var/home/acme/bin/perf(cmd_stat+0x2358) [0x4310c8]
> /var/home/acme/bin/perf(run_builtin+0x6a) [0x4a2cfa]
> /var/home/acme/bin/perf(main+0x612) [0x40f8c2]
> /usr/lib64/libc-2.33.so(__libc_start_main+0xd4) [0x27b74]
> /var/home/acme/bin/perf(_start+0x2d) [0x40fadd]
<SNIP>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20220117160931.1191712-4-acme@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Pull perf tool updates from Arnaldo Carvalho de Melo:
"New features:
- Add 'trace' subcommand for 'perf ftrace', setting the stage for
more 'perf ftrace' subcommands. Not using a subcommand yields the
previous behaviour of 'perf ftrace'.
- Add 'latency' subcommand to 'perf ftrace', that can use the
function graph tracer or a BPF optimized one, via the -b/--use-bpf
option.
E.g.:
$ sudo perf ftrace latency -a -T mutex_lock sleep 1
# DURATION | COUNT | GRAPH |
0 - 1 us | 4596 | ######################## |
1 - 2 us | 1680 | ######### |
2 - 4 us | 1106 | ##### |
4 - 8 us | 546 | ## |
8 - 16 us | 562 | ### |
16 - 32 us | 1 | |
32 - 64 us | 0 | |
64 - 128 us | 0 | |
128 - 256 us | 0 | |
256 - 512 us | 0 | |
512 - 1024 us | 0 | |
1 - 2 ms | 0 | |
2 - 4 ms | 0 | |
4 - 8 ms | 0 | |
8 - 16 ms | 0 | |
16 - 32 ms | 0 | |
32 - 64 ms | 0 | |
64 - 128 ms | 0 | |
128 - 256 ms | 0 | |
256 - 512 ms | 0 | |
512 - 1024 ms | 0 | |
1 - ... s | 0 | |
The original implementation of this command was in the bcc tool.
- Support --cputype option for hybrid events in 'perf stat'.
Improvements:
- Call chain improvements for ARM64.
- No need to do any affinity setup when profiling pids.
- Reduce multiplexing with duration_time in 'perf stat' metrics.
- Improve error message for uncore events, stating that some event
groups are can only be used in system wide (-a) mode.
- perf stat metric group leader fixes/improvements, including arch
specific changes to better support Intel topdown events.
- Probe non-deprecated sysfs path first, i.e. try the path
/sys/devices/system/cpu/cpuN/topology/thread_siblings first, then
the old /sys/devices/system/cpu/cpuN/topology/core_cpus.
- Disable debuginfod by default in 'perf record', to avoid stalls on
distros such as Fedora 35.
- Use unbuffered output in 'perf bench' when pipe/tee'ing to a file.
- Enable ignore_missing_thread in 'perf trace'
Fixes:
- Avoid TUI crash when navigating in the annotation of recursive
functions.
- Fix hex dump character output in 'perf script'.
- Fix JSON indentation to 4 spaces standard in the ARM vendor event
files.
- Fix use after free in metric__new().
- Fix IS_ERR_OR_NULL() usage in the perf BPF loader.
- Fix up cross-arch register support, i.e. when printing register
names take into account the architecture where the perf.data file
was collected.
- Fix SMT fallback with large core counts.
- Don't lower case MetricExpr when parsing JSON files so as not to
lose info such as the ":G" event modifier in metrics.
perf test:
- Add basic stress test for sigtrap handling to 'perf test'.
- Fix 'perf test' failures on s/390
- Enable system wide for metricgroups test in 'perf test´.
- Use 3 digits for test numbering now we can have more tests.
Arch specific:
- Add events for Arm Neoverse N2 in the ARM JSON vendor event files
- Support PERF_MEM_LVLNUM encodings in powerpc, that came from a
single patch series, where I incorrectly merged the kernel bits,
that were then reverted after coordination with Michael Ellerman
and Stephen Rothwell.
- Add ARM SPE total latency as PERF_SAMPLE_WEIGHT.
- Update AMD documentation, with info on raw event encoding.
- Add support for global and local variants of the "p_stage_cyc" sort
key, applicable to perf.data files collected on powerpc.
- Remove duplicate and incorrect aux size checks in the ARM CoreSight
ETM code.
Refactorings:
- Add a perf_cpu abstraction to disambiguate CPUs and CPU map
indexes, fixing problems along the way.
- Document CPU map methods.
UAPI sync:
- Update arch/x86/lib/mem{cpy,set}_64.S copies used in 'perf bench
mem memcpy'
- Sync UAPI files with the kernel sources: drm, msr-index,
cpufeatures.
Build system
- Enable warnings through HOSTCFLAGS.
- Drop requirement for libstdc++.so for libopencsd check
libperf:
- Make libperf adopt perf_counts_values__scale() from tools/perf/util/.
- Add a stat multiplexing test to libperf"
* tag 'perf-tools-for-v5.17-2022-01-16' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: (115 commits)
perf record: Disable debuginfod by default
perf evlist: No need to do any affinity setup when profiling pids
perf cpumap: Add is_dummy() method
perf metric: Fix metric_leader
perf cputopo: Fix CPU topology reading on s/390
perf metricgroup: Fix use after free in metric__new()
libperf tests: Update a use of the new cpumap API
perf arm: Fix off-by-one directory path
tools arch x86: Sync the msr-index.h copy with the kernel sources
tools headers cpufeatures: Sync with the kernel sources
tools headers UAPI: Update tools's copy of drm.h header
tools arch: Update arch/x86/lib/mem{cpy,set}_64.S copies used in 'perf bench mem memcpy'
perf pmu-events: Don't lower case MetricExpr
perf expr: Add debug logging for literals
perf tools: Probe non-deprecated sysfs path 1st
perf tools: Fix SMT fallback with large core counts
perf cpumap: Give CPUs their own type
perf stat: Correct first_shadow_cpu to return index
perf script: Fix flipped index and cpu
perf c2c: Use more intention revealing iterator
...
Pull tracing updates from Steven Rostedt:
"New:
- The Real Time Linux Analysis (RTLA) tool is added to the tools
directory.
- Can safely filter on user space pointers with: field.ustring ~
"match-string"
- eprobes can now be filtered like any other event.
- trace_marker(_raw) now uses stream_open() to allow multiple threads
to safely write to it. Note, this could possibly break existing
user space, but we will not know until we hear about it, and then
can revert the change if need be.
- New field in events to display when bottom halfs are disabled.
- Sorting of the ftrace functions are now done at compile time
instead of at bootup.
Infrastructure changes to support future efforts:
- Added __rel_loc type for trace events. Similar to __data_loc but
the offset to the dynamic data is based off of the location of the
descriptor and not the beginning of the event. Needed for user
defined events.
- Some simplification of event trigger code.
- Make synthetic events process its callback better to not hinder
other event callbacks that are registered. Needed for user defined
events.
And other small fixes and cleanups"
* tag 'trace-v5.17' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (50 commits)
tracing: Add ustring operation to filtering string pointers
rtla: Add rtla timerlat hist documentation
rtla: Add rtla timerlat top documentation
rtla: Add rtla timerlat documentation
rtla: Add rtla osnoise hist documentation
rtla: Add rtla osnoise top documentation
rtla: Add rtla osnoise man page
rtla: Add Documentation
rtla/timerlat: Add timerlat hist mode
rtla: Add timerlat tool and timelart top mode
rtla/osnoise: Add the hist mode
rtla/osnoise: Add osnoise top mode
rtla: Add osnoise tool
rtla: Helper functions for rtla
rtla: Real-Time Linux Analysis tool
tracing/osnoise: Properly unhook events if start_per_cpu_kthreads() fails
tracing: Remove duplicate warnings when calling trace_create_file()
tracing/kprobes: 'nmissed' not showed correctly for kretprobe
tracing: Add test for user space strings when filtering on string pointers
tracing: Have syscall trace events use trace_event_buffer_lock_reserve()
...
Needed to check if a cpu_map is dummy, i.e. not a cpu map at all, for
pid monitoring scenarios.
This probably needs to move to libperf, but since perf itself is the
first and so far only user, leave it at tools/perf/util/.
Acked-by: Andi Kleen <andi@firstfloor.org>
Acked-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Commit fdf1e29b61 ("perf expr: Add metric literals for topology.")
fails on s390:
# ./perf test -Fv 7
...
# FAILED tests/expr.c:173 #num_dies >= #num_packages
---- end ----
Simple expression parser: FAILED!
#
Investigating this issue leads to these functions:
build_cpu_topology()
+--> has_die_topology(void)
{
struct utsname uts;
if (uname(&uts) < 0)
return false;
if (strncmp(uts.machine, "x86_64", 6))
return false;
....
}
which always returns false on s390. The caller build_cpu_topology()
checks has_die_topology() return value. On false the
the struct cpu_topology::die_cpu_list is not contructed and has zero
entries. This leads to the failing comparison: #num_dies >= #num_packages.
s390 of course has a positive number of packages.
Fix this by adding s390 architecture to support CPU die list.
Output after:
# ./perf test -Fv 7
7: Simple expression parser :
--- start ---
division by zero
syntax error
---- end ----
Simple expression parser: Ok
#
Fixes: fdf1e29b61 ("perf expr: Add metric literals for topology.")
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Sumanth Korikkar <sumanthk@linux.ibm.com>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Link: https://lore.kernel.org/r/20211124090343.9436-1-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
To pick up fixes and get in line with other trees, powerpc kernel
mostly this time, but BPF as well.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
John Garry
German Gomez
Lv Ruyi
Ian Rogers
Arnaldo Carvalho de Melo
Zechuan Chen
Linus Torvalds
Jiri Olsa
Thomas Richter
José Expósito