mirror of
https://github.com/hardkernel/linux.git
synced 2026-03-27 13:00:25 +09:00
119dcafe36795a15ae53351cbbd6177aaf94ffef
15521 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Kuan-Wei Chiu
|
8da6b450a1 |
perf bench: Fix undefined behavior in cmpworker()
commit 62892e77b8a64b9dc0e1da75980aa145347b6820 upstream.
The comparison function cmpworker() violates the C standard's
requirements for qsort() comparison functions, which mandate symmetry
and transitivity:
Symmetry: If x < y, then y > x.
Transitivity: If x < y and y < z, then x < z.
In its current implementation, cmpworker() incorrectly returns 0 when
w1->tid < w2->tid, which breaks both symmetry and transitivity. This
violation causes undefined behavior, potentially leading to issues such
as memory corruption in glibc [1].
Fix the issue by returning -1 when w1->tid < w2->tid, ensuring
compliance with the C standard and preventing undefined behavior.
Link: https://www.qualys.com/2024/01/30/qsort.txt [1]
Fixes:
|
||
|
Howard Chu
|
d2cd915aea |
perf trace: Fix runtime error of index out of bounds
[ Upstream commit c7b87ce0dd10b64b68a0b22cb83bbd556e28fe81 ]
libtraceevent parses and returns an array of argument fields, sometimes
larger than RAW_SYSCALL_ARGS_NUM (6) because it includes "__syscall_nr",
idx will traverse to index 6 (7th element) whereas sc->fmt->arg holds 6
elements max, creating an out-of-bounds access. This runtime error is
found by UBsan. The error message:
$ sudo UBSAN_OPTIONS=print_stacktrace=1 ./perf trace -a --max-events=1
builtin-trace.c:1966:35: runtime error: index 6 out of bounds for type 'syscall_arg_fmt [6]'
#0 0x5c04956be5fe in syscall__alloc_arg_fmts /home/howard/hw/linux-perf/tools/perf/builtin-trace.c:1966
#1 0x5c04956c0510 in trace__read_syscall_info /home/howard/hw/linux-perf/tools/perf/builtin-trace.c:2110
#2 0x5c04956c372b in trace__syscall_info /home/howard/hw/linux-perf/tools/perf/builtin-trace.c:2436
#3 0x5c04956d2f39 in trace__init_syscalls_bpf_prog_array_maps /home/howard/hw/linux-perf/tools/perf/builtin-trace.c:3897
#4 0x5c04956d6d25 in trace__run /home/howard/hw/linux-perf/tools/perf/builtin-trace.c:4335
#5 0x5c04956e112e in cmd_trace /home/howard/hw/linux-perf/tools/perf/builtin-trace.c:5502
#6 0x5c04956eda7d in run_builtin /home/howard/hw/linux-perf/tools/perf/perf.c:351
#7 0x5c04956ee0a8 in handle_internal_command /home/howard/hw/linux-perf/tools/perf/perf.c:404
#8 0x5c04956ee37f in run_argv /home/howard/hw/linux-perf/tools/perf/perf.c:448
#9 0x5c04956ee8e9 in main /home/howard/hw/linux-perf/tools/perf/perf.c:556
#10 0x79eb3622a3b7 in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
#11 0x79eb3622a47a in __libc_start_main_impl ../csu/libc-start.c:360
#12 0x5c04955422d4 in _start (/home/howard/hw/linux-perf/tools/perf/perf+0x4e02d4) (BuildId: 5b6cab2d59e96a4341741765ad6914a4d784dbc6)
0.000 ( 0.014 ms): Chrome_ChildIO/117244 write(fd: 238, buf: !, count: 1) = 1
Fixes:
|
||
|
Chun-Tse Shao
|
55f75ce472 |
perf lock: Fix parse_lock_type which only retrieve one lock flag
[ Upstream commit 1be9264158ef4818393e5d8144887a1a5d3cc480 ]
`parse_lock_type` can only add the first lock flag in `lock_type_table`
given input `str`. For example, for `Y rwlock`, it only adds `rwlock:R`
into this perf session. Another example is for `-Y mutex`, it only adds
the mutex without `LCB_F_SPIN` flag. The patch fixes this issue, makes
sure both `rwlock:R` and `rwlock:W` will be added with `-Y rwlock`, and
so on.
Testing:
$ ./perf lock con -ab -Y mutex,rwlock -- perf bench sched pipe
# Running 'sched/pipe' benchmark:
# Executed 1000000 pipe operations between two processes
Total time: 9.313 [sec]
9.313976 usecs/op
107365 ops/sec
contended total wait max wait avg wait type caller
176 1.65 ms 19.43 us 9.38 us mutex pipe_read+0x57
34 180.14 us 10.93 us 5.30 us mutex pipe_write+0x50
7 77.48 us 16.09 us 11.07 us mutex do_epoll_wait+0x24d
7 74.70 us 13.50 us 10.67 us mutex do_epoll_wait+0x24d
3 35.97 us 14.44 us 11.99 us rwlock:W ep_done_scan+0x2d
3 35.00 us 12.23 us 11.66 us rwlock:W do_epoll_wait+0x255
2 15.88 us 11.96 us 7.94 us rwlock:W do_epoll_wait+0x47c
1 15.23 us 15.23 us 15.23 us rwlock:W do_epoll_wait+0x4d0
1 14.26 us 14.26 us 14.26 us rwlock:W ep_done_scan+0x2d
2 14.00 us 7.99 us 7.00 us mutex pipe_read+0x282
1 12.29 us 12.29 us 12.29 us rwlock:R ep_poll_callback+0x35
1 12.02 us 12.02 us 12.02 us rwlock:W do_epoll_ctl+0xb65
1 10.25 us 10.25 us 10.25 us rwlock:R ep_poll_callback+0x35
1 7.86 us 7.86 us 7.86 us mutex do_epoll_ctl+0x6c1
1 5.04 us 5.04 us 5.04 us mutex do_epoll_ctl+0x3d4
[namhyung: Add a comment and rename to 'mutex:spin' for consistency
Fixes:
|
||
|
Jiachen Zhang
|
2e09336f35 |
perf report: Fix misleading help message about --demangle
[ Upstream commit ac0ac75189a4d6a29a2765a7adbb62bc6cc650c7 ]
The wrong help message may mislead users. This commit fixes it.
Fixes:
|
||
|
Arnaldo Carvalho de Melo
|
125066c329 |
perf namespaces: Fixup the nsinfo__in_pidns() return type, its bool
[ Upstream commit 64a7617efd5ae1d57a75e464d7134eec947c3fe3 ]
When adding support for refconunt checking a cut'n'paste made this
function, that is just an accessor to a bool member of 'struct nsinfo',
return a pid_t, when that member is a boolean, fix it.
Fixes:
|
||
|
Arnaldo Carvalho de Melo
|
d43c244703 |
perf namespaces: Introduce nsinfo__set_in_pidns()
[ Upstream commit 9c6a585d257f6845731f4e36b45fe42b5c3162f5 ] When we're processing a perf.data file we will, for every thread in that file do a machine__findnew_thread(machine, pid, tid) that when that pid is seen for the first time will create a 'struct thread' representing it. That in turn will call nsinfo__new() -> nsinfo__init() and there it will assume we're running live, which is wrong and will need to be addressed in a followup patch. The nsinfo__new() assumes that if we can't access that thread it has already finished and will ignore the -1 return from nsinfo__init(), just taking notes to avoid trying to enter in that namespace, since it isn't there anymore, a race. When doing this from 'perf inject', tho, we can fill in parts of that nsinfo from what we get from the PERF_RECORD_MMAP2 (pid, tid) and in the jitdump file name, that has the form of jit-<PID>.dump. So if the pid in the jitdump file name is not the one in the PERF_RECORD_MMAP2, we can assume that its the pid of the process _inside_ the namespace, and that perf was runing outside that namespace. This will be done in the following patch. Reported-by: Francesco Nigro <fnigro@redhat.com> Reported-by: Ilan Green <igreen@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Clark Williams <williams@redhat.com> Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: James Clark <james.clark@linaro.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Yonatan Goldschmidt <yonatan.goldschmidt@granulate.io> Link: https://lore.kernel.org/r/20241206204828.507527-4-acme@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Stable-dep-of: 64a7617efd5a ("perf namespaces: Fixup the nsinfo__in_pidns() return type, its bool") Signed-off-by: Sasha Levin <sashal@kernel.org> |
||
|
Christophe Leroy
|
4318e26fe4 |
perf machine: Don't ignore _etext when not a text symbol
[ Upstream commit 7a93786c306296f15e728b1dbd949a319e4e3d19 ]
Depending on how vmlinux.lds is written, _etext might be the very first
data symbol instead of the very last text symbol.
Don't require it to be a text symbol, accept any symbol type.
Comitter notes:
See the first Link for further discussion, but it all boils down to
this:
---
# grep -e _stext -e _etext -e _edata /proc/kallsyms
c0000000 T _stext
c08b8000 D _etext
So there is no _edata and _etext is not text
$ ppc-linux-objdump -x vmlinux | grep -e _stext -e _etext -e _edata
c0000000 g .head.text 00000000 _stext
c08b8000 g .rodata 00000000 _etext
c1378000 g .sbss 00000000 _edata
---
Fixes:
|
||
|
Arnaldo Carvalho de Melo
|
8487f93db8 |
perf top: Don't complain about lack of vmlinux when not resolving some kernel samples
[ Upstream commit 058b38ccd2af9e5c95590b018e8425fa148d7aca ]
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:
|
||
|
Levi Yun
|
1ee00cc23c |
perf expr: Initialize is_test value in expr__ctx_new()
[ Upstream commit 1d18ebcfd302a2005b83ae5f13df223894d19902 ]
When expr_parse_ctx is allocated by expr_ctx_new(),
expr_scanner_ctx->is_test isn't initialize, so it has garbage value.
this can affects the result of expr__parse() return when it parses
non-exist event literal according to garbage value.
Use calloc instead of malloc in expr_ctx_new() to fix this.
Fixes:
|
||
|
Zhongqiu Han
|
c7d067a47b |
perf bpf: Fix two memory leakages when calling perf_env__insert_bpf_prog_info()
[ Upstream commit 03edb7020bb920f1935c3f30acad0bb27fdb99af ]
If perf_env__insert_bpf_prog_info() returns false due to a duplicate bpf
prog info node insertion, the temporary info_node and info_linear memory
will leak. Add a check to ensure the memory is freed if the function
returns false.
Fixes:
|
||
|
Zhongqiu Han
|
bce9da3aca |
perf header: Fix one memory leakage in process_bpf_prog_info()
[ Upstream commit a7da6c7030e1aec32f0a41c7b4fa70ec96042019 ]
Function __perf_env__insert_bpf_prog_info() will return without inserting
bpf prog info node into perf env again due to a duplicate bpf prog info
node insertion, causing the temporary info_linear and info_node memory to
leak. Modify the return type of this function to bool and add a check to
ensure the memory is freed if the function returns false.
Fixes:
|
||
|
Zhongqiu Han
|
4f2582daca |
perf header: Fix one memory leakage in process_bpf_btf()
[ Upstream commit 875d22980a062521beed7b5df71fb13a1af15d83 ]
If __perf_env__insert_btf() returns false due to a duplicate btf node
insertion, the temporary node will leak. Add a check to ensure the memory
is freed if the function returns false.
Fixes:
|
||
|
Benjamin Peterson
|
2f465476ff |
perf trace: Avoid garbage when not printing a syscall's arguments
[ Upstream commit 1302e352b26f34991b619b5d0b621b76d20a3883 ]
syscall__scnprintf_args may not place anything in the output buffer
(e.g., because the arguments are all zero). If that happened in
trace__fprintf_sys_enter, its fprintf would receive an unitialized
buffer leading to garbage output.
Fix the problem by passing the (possibly zero) bounds of the argument
buffer to the output fprintf.
Fixes:
|
||
|
Benjamin Peterson
|
fad7be2395 |
perf trace: Do not lose last events in a race
[ Upstream commit 3fd7c36973a250e17a4ee305a31545a9426021f4 ]
If a perf trace event selector specifies a maximum number of events to output
(i.e., "/nr=N/" syntax), the event printing handler, trace__event_handler,
disables the event selector after the maximum number events are
printed.
Furthermore, trace__event_handler checked if the event selector was
disabled before doing any work. This avoided exceeding the maximum
number of events to print if more events were in the buffer before the
selector was disabled.
However, the event selector can be disabled for reasons other than
exceeding the maximum number of events. In particular, when the traced
subprocess exits, the main loop disables all event selectors. This meant
the last events of a traced subprocess might be lost to the printing
handler's short-circuiting logic.
This nondeterministic problem could be seen by running the following many times:
$ perf trace -e syscalls:sys_enter_exit_group true
trace__event_handler should simply check for exceeding the maximum number of
events to print rather than the state of the event selector.
Fixes:
|
||
|
Howard Chu
|
e3b4b95c2b |
perf trace: Fix tracing itself, creating feedback loops
[ Upstream commit fe4f9b4124967ffb75d66994520831231b779550 ] There exists a pids_filtered map in augmented_raw_syscalls.bpf.c that ceases to provide functionality after the BPF skeleton migration done in: |
||
|
Jean-Philippe Romain
|
f38ab496c7 |
perf list: Fix topic and pmu_name argument order
[ Upstream commit d99b3125726aade4f5ec4aae04805134ab4b0abd ]
Fix function definitions to match header file declaration. Fix two
callers to pass the arguments in the right order.
On Intel Tigerlake, before:
```
$ perf list -j|grep "\"Topic\""|sort|uniq
"Topic": "cache",
"Topic": "cpu",
"Topic": "floating point",
"Topic": "frontend",
"Topic": "memory",
"Topic": "other",
"Topic": "pfm icl",
"Topic": "pfm ix86arch",
"Topic": "pfm perf_raw",
"Topic": "pipeline",
"Topic": "tool",
"Topic": "uncore interconnect",
"Topic": "uncore memory",
"Topic": "uncore other",
"Topic": "virtual memory",
$ perf list -j|grep "\"Unit\""|sort|uniq
"Unit": "cache",
"Unit": "cpu",
"Unit": "cstate_core",
"Unit": "cstate_pkg",
"Unit": "i915",
"Unit": "icl",
"Unit": "intel_bts",
"Unit": "intel_pt",
"Unit": "ix86arch",
"Unit": "msr",
"Unit": "perf_raw",
"Unit": "power",
"Unit": "tool",
"Unit": "uncore_arb",
"Unit": "uncore_clock",
"Unit": "uncore_imc_free_running_0",
"Unit": "uncore_imc_free_running_1",
```
After:
```
$ perf list -j|grep "\"Topic\""|sort|uniq
"Topic": "cache",
"Topic": "floating point",
"Topic": "frontend",
"Topic": "memory",
"Topic": "other",
"Topic": "pfm icl",
"Topic": "pfm ix86arch",
"Topic": "pfm perf_raw",
"Topic": "pipeline",
"Topic": "tool",
"Topic": "uncore interconnect",
"Topic": "uncore memory",
"Topic": "uncore other",
"Topic": "virtual memory",
$ perf list -j|grep "\"Unit\""|sort|uniq
"Unit": "cpu",
"Unit": "cstate_core",
"Unit": "cstate_pkg",
"Unit": "i915",
"Unit": "icl",
"Unit": "intel_bts",
"Unit": "intel_pt",
"Unit": "ix86arch",
"Unit": "msr",
"Unit": "perf_raw",
"Unit": "power",
"Unit": "tool",
"Unit": "uncore_arb",
"Unit": "uncore_clock",
"Unit": "uncore_imc_free_running_0",
"Unit": "uncore_imc_free_running_1",
```
Fixes:
|
||
|
Benjamin Peterson
|
ad8cf4a18d |
perf trace: avoid garbage when not printing a trace event's arguments
[ Upstream commit 5fb8e56542a3cf469fdf25d77f50e21cbff3ae7e ]
trace__fprintf_tp_fields may not print any tracepoint arguments. E.g., if the
argument values are all zero. Previously, this would result in a totally
uninitialized buffer being passed to fprintf, which could lead to garbage on the
console. Fix the problem by passing the number of initialized bytes fprintf.
Fixes:
|
||
|
Arnaldo Carvalho de Melo
|
aac3361f6d |
perf ftrace latency: Fix unit on histogram first entry when using --use-nsec
[ Upstream commit 064d569e20e82c065b1dec9d20c29c7087bb1a00 ]
The use_nsec arg wasn't being taken into account when printing the first
histogram entry, fix it:
root@number:~# perf ftrace latency --use-nsec -T switch_mm_irqs_off -a sleep 2
# DURATION | COUNT | GRAPH |
0 - 1 us | 0 | |
1 - 2 ns | 0 | |
2 - 4 ns | 0 | |
4 - 8 ns | 0 | |
8 - 16 ns | 0 | |
16 - 32 ns | 0 | |
32 - 64 ns | 125 | |
64 - 128 ns | 335 | |
128 - 256 ns | 2155 | #### |
256 - 512 ns | 9996 | ################### |
512 - 1024 ns | 4958 | ######### |
1 - 2 us | 4636 | ######### |
2 - 4 us | 1053 | ## |
4 - 8 us | 15 | |
8 - 16 us | 1 | |
16 - 32 us | 0 | |
32 - 64 us | 0 | |
64 - 128 us | 0 | |
128 - 256 us | 0 | |
256 - 512 us | 0 | |
512 - 1024 us | 0 | |
1 - ... ms | 0 | |
root@number:~#
After:
root@number:~# perf ftrace latency --use-nsec -T switch_mm_irqs_off -a sleep 2
# DURATION | COUNT | GRAPH |
0 - 1 ns | 0 | |
1 - 2 ns | 0 | |
2 - 4 ns | 0 | |
4 - 8 ns | 0 | |
8 - 16 ns | 0 | |
16 - 32 ns | 0 | |
32 - 64 ns | 19 | |
64 - 128 ns | 94 | |
128 - 256 ns | 2191 | #### |
256 - 512 ns | 9719 | #################### |
512 - 1024 ns | 5330 | ########### |
1 - 2 us | 4104 | ######## |
2 - 4 us | 807 | # |
4 - 8 us | 9 | |
8 - 16 us | 0 | |
16 - 32 us | 0 | |
32 - 64 us | 0 | |
64 - 128 us | 0 | |
128 - 256 us | 0 | |
256 - 512 us | 0 | |
512 - 1024 us | 0 | |
1 - ... ms | 0 | |
root@number:~#
Fixes:
|
||
|
Leo Yan
|
cbc853c490 |
perf probe: Correct demangled symbols in C++ program
[ Upstream commit 314909f13cc12d47c468602c37dace512d225eeb ]
An issue can be observed when probe C++ demangled symbol with steps:
# nm test_cpp_mangle | grep print_data
0000000000000c94 t _GLOBAL__sub_I__Z10print_datai
0000000000000afc T _Z10print_datai
0000000000000b38 T _Z10print_dataR5Point
# perf probe -x /home/niayan01/test_cpp_mangle -F --demangle
...
print_data(Point&)
print_data(int)
...
# perf --debug verbose=3 probe -x test_cpp_mangle --add "test=print_data(int)"
probe-definition(0): test=print_data(int)
symbol:print_data(int) file:(null) line:0 offset:0 return:0 lazy:(null)
0 arguments
Open Debuginfo file: /home/niayan01/test_cpp_mangle
Try to find probe point from debuginfo.
Symbol print_data(int) address found : afc
Matched function: print_data [2ccf]
Probe point found: print_data+0
Found 1 probe_trace_events.
Opening /sys/kernel/tracing//uprobe_events write=1
Opening /sys/kernel/tracing//README write=0
Writing event: p:probe_test_cpp_mangle/test /home/niayan01/test_cpp_mangle:0xb38
...
When tried to probe symbol "print_data(int)", the log shows:
Symbol print_data(int) address found : afc
The found address is 0xafc - which is right with verifying the output
result from nm. Afterwards when write event, the command uses offset
0xb38 in the last log, which is a wrong address.
The dwarf_diename() gets a common function name, in above case, it
returns string "print_data". As a result, the tool parses the offset
based on the common name. This leads to probe at the wrong symbol
"print_data(Point&)".
To fix the issue, use the die_get_linkage_name() function to retrieve
the distinct linkage name - this is the mangled name for the C++ case.
Based on this unique name, the tool can get a correct offset for
probing. Based on DWARF doc, it is possible the linkage name is missed
in the DIE, it rolls back to use dwarf_diename().
After:
# perf --debug verbose=3 probe -x test_cpp_mangle --add "test=print_data(int)"
probe-definition(0): test=print_data(int)
symbol:print_data(int) file:(null) line:0 offset:0 return:0 lazy:(null)
0 arguments
Open Debuginfo file: /home/niayan01/test_cpp_mangle
Try to find probe point from debuginfo.
Symbol print_data(int) address found : afc
Matched function: print_data [2d06]
Probe point found: print_data+0
Found 1 probe_trace_events.
Opening /sys/kernel/tracing//uprobe_events write=1
Opening /sys/kernel/tracing//README write=0
Writing event: p:probe_test_cpp_mangle/test /home/niayan01/test_cpp_mangle:0xafc
Added new event:
probe_test_cpp_mangle:test (on print_data(int) in /home/niayan01/test_cpp_mangle)
You can now use it in all perf tools, such as:
perf record -e probe_test_cpp_mangle:test -aR sleep 1
# perf --debug verbose=3 probe -x test_cpp_mangle --add "test2=print_data(Point&)"
probe-definition(0): test2=print_data(Point&)
symbol:print_data(Point&) file:(null) line:0 offset:0 return:0 lazy:(null)
0 arguments
Open Debuginfo file: /home/niayan01/test_cpp_mangle
Try to find probe point from debuginfo.
Symbol print_data(Point&) address found : b38
Matched function: print_data [2ccf]
Probe point found: print_data+0
Found 1 probe_trace_events.
Opening /sys/kernel/tracing//uprobe_events write=1
Parsing probe_events: p:probe_test_cpp_mangle/test /home/niayan01/test_cpp_mangle:0x0000000000000afc
Group:probe_test_cpp_mangle Event:test probe:p
Opening /sys/kernel/tracing//README write=0
Writing event: p:probe_test_cpp_mangle/test2 /home/niayan01/test_cpp_mangle:0xb38
Added new event:
probe_test_cpp_mangle:test2 (on print_data(Point&) in /home/niayan01/test_cpp_mangle)
You can now use it in all perf tools, such as:
perf record -e probe_test_cpp_mangle:test2 -aR sleep 1
Fixes:
|
||
|
Ian Rogers
|
2c6f6c3843 |
perf probe: Fix libdw memory leak
[ Upstream commit 4585038b8e186252141ef86e9f0d8e97f11dce8d ]
Add missing dwarf_cfi_end to free memory associated with probe_finder
cfi_eh which is allocated and owned via a call to
dwarf_getcfi_elf. Confusingly cfi_dbg shouldn't be freed as its memory
is owned by the passed in debuginfo struct. Add comments to highlight
this.
This addresses leak sanitizer issues seen in:
tools/perf/tests/shell/test_uprobe_from_different_cu.sh
Fixes:
|
||
|
Veronika Molnarova
|
9ac8d66362 |
perf test attr: Add back missing topdown events
[ Upstream commit 6bff76af9635411214ca44ea38fc2781e78064b6 ] With the patch |
||
|
Michael Petlan
|
0c47534539 |
perf trace: Keep exited threads for summary
[ Upstream commit d29d92df410e2fb523f640478b18f70c1823e55e ]
Since 9ffa6c7512ca ("perf machine thread: Remove exited threads by
default") perf cleans exited threads up, but as said, sometimes they
are necessary to be kept. The mentioned commit does not cover all the
cases, we also need the information to construct the summary table in
perf-trace.
Before:
# perf trace -s true
Summary of events:
After:
# perf trace -s -- true
Summary of events:
true (383382), 64 events, 91.4%
syscall calls errors total min avg max stddev
(msec) (msec) (msec) (msec) (%)
--------------- -------- ------ -------- --------- --------- --------- ------
mmap 8 0 0.150 0.013 0.019 0.031 11.90%
mprotect 3 0 0.045 0.014 0.015 0.017 6.47%
openat 2 0 0.014 0.006 0.007 0.007 9.73%
munmap 1 0 0.009 0.009 0.009 0.009 0.00%
access 1 1 0.009 0.009 0.009 0.009 0.00%
pread64 4 0 0.006 0.001 0.001 0.002 4.53%
fstat 2 0 0.005 0.001 0.002 0.003 37.59%
arch_prctl 2 1 0.003 0.001 0.002 0.002 25.91%
read 1 0 0.003 0.003 0.003 0.003 0.00%
close 2 0 0.003 0.001 0.001 0.001 3.86%
brk 1 0 0.002 0.002 0.002 0.002 0.00%
rseq 1 0 0.001 0.001 0.001 0.001 0.00%
prlimit64 1 0 0.001 0.001 0.001 0.001 0.00%
set_robust_list 1 0 0.001 0.001 0.001 0.001 0.00%
set_tid_address 1 0 0.001 0.001 0.001 0.001 0.00%
execve 1 0 0.000 0.000 0.000 0.000 0.00%
[namhyung: simplified the condition]
Fixes: 9ffa6c7512ca ("perf machine thread: Remove exited threads by default")
Reported-by: Veronika Molnarova <vmolnaro@redhat.com>
Signed-off-by: Michael Petlan <mpetlan@redhat.com>
Link: https://lore.kernel.org/r/20240927151926.399474-1-mpetlan@redhat.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
||
|
Ian Rogers
|
380bc5a698 |
perf stat: Fix affinity memory leaks on error path
[ Upstream commit 7f6ccb70e465bd8c9cf8973aee1c01224e4bdb3c ]
Missed cleanup when an error occurs.
Fixes:
|
||
|
Levi Yun
|
035c6b7a13 |
perf stat: Close cork_fd when create_perf_stat_counter() failed
[ Upstream commit e880a70f8046df0dd9089fa60dcb866a2cc69194 ] When create_perf_stat_counter() failed, it doesn't close workload.cork_fd open in evlist__prepare_workload(). This could make too many open file error while __run_perf_stat() repeats. Introduce evlist__cancel_workload to close workload.cork_fd and wait workload.child_pid until exit to clear child process when create_perf_stat_counter() is failed. Signed-off-by: Levi Yun <yeoreum.yun@arm.com> Reviewed-by: James Clark <james.clark@linaro.org> Reviewed-by: Andi Kleen <ak@linux.intel.com> Cc: nd@arm.com Cc: howardchu95@gmail.com Link: https://lore.kernel.org/r/20240925132022.2650180-2-yeoreum.yun@arm.com Signed-off-by: Namhyung Kim <namhyung@kernel.org> Stable-dep-of: 7f6ccb70e465 ("perf stat: Fix affinity memory leaks on error path") Signed-off-by: Sasha Levin <sashal@kernel.org> |
||
|
James Clark
|
300b218862 |
perf cs-etm: Don't flush when packet_queue fills up
[ Upstream commit 5afd032961e8465808c4bc385c06e7676fbe1951 ]
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:
|
||
|
Aditya Gupta
|
ab656dd9fc |
libsubcmd: Don't free the usage string
[ Upstream commit 1a5efc9e13f357abc396dbf445b25d08914c8060 ]
Currently, commands which depend on 'parse_options_subcommand()' don't
show the usage string, and instead show '(null)'
$ ./perf sched
Usage: (null)
-D, --dump-raw-trace dump raw trace in ASCII
-f, --force don't complain, do it
-i, --input <file> input file name
-v, --verbose be more verbose (show symbol address, etc)
'parse_options_subcommand()' is generally expected to initialise the usage
string, with information in the passed 'subcommands[]' array
This behaviour was changed in:
230a7a71f92212e7 ("libsubcmd: Fix parse-options memory leak")
Where the generated usage string is deallocated, and usage[0] string is
reassigned as NULL.
As discussed in [1], free the allocated usage string in the main
function itself, and don't reset usage string to NULL in
parse_options_subcommand
With this change, the behaviour is restored.
$ ./perf sched
Usage: perf sched [<options>] {record|latency|map|replay|script|timehist}
-D, --dump-raw-trace dump raw trace in ASCII
-f, --force don't complain, do it
-i, --input <file> input file name
-v, --verbose be more verbose (show symbol address, etc)
[1]: https://lore.kernel.org/linux-perf-users/htq5vhx6piet4nuq2mmhk7fs2bhfykv52dbppwxmo3s7du2odf@styd27tioc6e/
Fixes: 230a7a71f92212e7 ("libsubcmd: Fix parse-options memory leak")
Suggested-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Aditya Gupta <adityag@linux.ibm.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Cc: Disha Goel <disgoel@linux.vnet.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20240904061836.55873-2-adityag@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
||
|
Yang Jihong
|
03cec19f54 |
perf sched: Move curr_pid and cpu_last_switched initialization to perf_sched__{lat|map|replay}()
[ Upstream commit bd2cdf26b9ea000339d54adc82e87fdbf22c21c3 ]
The curr_pid and cpu_last_switched are used only for the
'perf sched replay/latency/map'. Put their initialization in
perf_sched__{lat|map|replay () to reduce unnecessary actions in other
commands.
Simple functional testing:
# perf sched record perf bench sched messaging
# Running 'sched/messaging' benchmark:
# 20 sender and receiver processes per group
# 10 groups == 400 processes run
Total time: 0.209 [sec]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 16.456 MB perf.data (147907 samples) ]
# perf sched lat
-------------------------------------------------------------------------------------------------------------------------------------------
Task | Runtime ms | Switches | Avg delay ms | Max delay ms | Max delay start | Max delay end |
-------------------------------------------------------------------------------------------------------------------------------------------
sched-messaging:(401) | 2990.699 ms | 38705 | avg: 0.661 ms | max: 67.046 ms | max start: 456532.624830 s | max end: 456532.691876 s
qemu-system-x86:(7) | 179.764 ms | 2191 | avg: 0.152 ms | max: 21.857 ms | max start: 456532.576434 s | max end: 456532.598291 s
sshd:48125 | 0.522 ms | 2 | avg: 0.037 ms | max: 0.046 ms | max start: 456532.514610 s | max end: 456532.514656 s
<SNIP>
ksoftirqd/11:82 | 0.063 ms | 1 | avg: 0.005 ms | max: 0.005 ms | max start: 456532.769366 s | max end: 456532.769371 s
kworker/9:0-mm_:34624 | 0.233 ms | 20 | avg: 0.004 ms | max: 0.007 ms | max start: 456532.690804 s | max end: 456532.690812 s
migration/13:93 | 0.000 ms | 1 | avg: 0.004 ms | max: 0.004 ms | max start: 456532.512669 s | max end: 456532.512674 s
-----------------------------------------------------------------------------------------------------------------
TOTAL: | 3180.750 ms | 41368 |
---------------------------------------------------
# echo $?
0
# perf sched map
*A0 456532.510141 secs A0 => migration/0:15
*. 456532.510171 secs . => swapper:0
. *B0 456532.510261 secs B0 => migration/1:21
. *. 456532.510279 secs
<SNIP>
L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 *L7 . . . . 456532.785979 secs
L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 *L7 . . . 456532.786054 secs
L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 *L7 . . 456532.786127 secs
L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 *L7 . 456532.786197 secs
L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 L7 *L7 456532.786270 secs
# echo $?
0
# perf sched replay
run measurement overhead: 108 nsecs
sleep measurement overhead: 66473 nsecs
the run test took 1000002 nsecs
the sleep test took 1082686 nsecs
nr_run_events: 49334
nr_sleep_events: 50054
nr_wakeup_events: 34701
target-less wakeups: 165
multi-target wakeups: 766
task 0 ( swapper: 0), nr_events: 15419
task 1 ( swapper: 1), nr_events: 1
task 2 ( swapper: 2), nr_events: 1
<SNIP>
task 715 ( sched-messaging: 110248), nr_events: 1438
task 716 ( sched-messaging: 110249), nr_events: 512
task 717 ( sched-messaging: 110250), nr_events: 500
task 718 ( sched-messaging: 110251), nr_events: 537
task 719 ( sched-messaging: 110252), nr_events: 823
------------------------------------------------------------
#1 : 1325.288, ravg: 1325.29, cpu: 7823.35 / 7823.35
#2 : 1363.606, ravg: 1329.12, cpu: 7655.53 / 7806.56
#3 : 1349.494, ravg: 1331.16, cpu: 7544.80 / 7780.39
#4 : 1311.488, ravg: 1329.19, cpu: 7495.13 / 7751.86
#5 : 1309.902, ravg: 1327.26, cpu: 7266.65 / 7703.34
#6 : 1309.535, ravg: 1325.49, cpu: 7843.86 / 7717.39
#7 : 1316.482, ravg: 1324.59, cpu: 7854.41 / 7731.09
#8 : 1366.604, ravg: 1328.79, cpu: 7955.81 / 7753.57
#9 : 1326.286, ravg: 1328.54, cpu: 7466.86 / 7724.90
#10 : 1356.653, ravg: 1331.35, cpu: 7566.60 / 7709.07
# echo $?
0
Signed-off-by: Yang Jihong <yangjihong1@huawei.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20240206083228.172607-5-yangjihong1@huawei.com
Stable-dep-of: 1a5efc9e13f3 ("libsubcmd: Don't free the usage string")
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
||
|
Yang Jihong
|
4078f1d442 |
perf sched: Move curr_thread initialization to perf_sched__map()
[ Upstream commit 5e895278697c014e95ae7ae5e79a72ef68c5184e ]
The curr_thread is used only for the 'perf sched map'. Put initialization
in perf_sched__map() to reduce unnecessary actions in other commands.
Simple functional testing:
# perf sched record perf bench sched messaging
# Running 'sched/messaging' benchmark:
# 20 sender and receiver processes per group
# 10 groups == 400 processes run
Total time: 0.197 [sec]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 15.526 MB perf.data (140095 samples) ]
# perf sched map
*A0 451264.532445 secs A0 => migration/0:15
*. 451264.532468 secs . => swapper:0
. *B0 451264.532537 secs B0 => migration/1:21
. *. 451264.532560 secs
. . *C0 451264.532644 secs C0 => migration/2:27
. . *. 451264.532668 secs
. . . *D0 451264.532753 secs D0 => migration/3:33
. . . *. 451264.532778 secs
. . . . *E0 451264.532861 secs E0 => migration/4:39
. . . . *. 451264.532886 secs
. . . . . *F0 451264.532973 secs F0 => migration/5:45
<SNIP>
A7 A7 A7 A7 A7 *A7 . . . . . . . . . . 451264.790785 secs
A7 A7 A7 A7 A7 A7 *A7 . . . . . . . . . 451264.790858 secs
A7 A7 A7 A7 A7 A7 A7 *A7 . . . . . . . . 451264.790934 secs
A7 A7 A7 A7 A7 A7 A7 A7 *A7 . . . . . . . 451264.791004 secs
A7 A7 A7 A7 A7 A7 A7 A7 A7 *A7 . . . . . . 451264.791075 secs
A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 *A7 . . . . . 451264.791143 secs
A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 *A7 . . . . 451264.791232 secs
A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 *A7 . . . 451264.791336 secs
A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 *A7 . . 451264.791407 secs
A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 *A7 . 451264.791484 secs
A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 A7 *A7 451264.791553 secs
# echo $?
0
Signed-off-by: Yang Jihong <yangjihong1@huawei.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20240206083228.172607-4-yangjihong1@huawei.com
Stable-dep-of: 1a5efc9e13f3 ("libsubcmd: Don't free the usage string")
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
||
|
Yang Jihong
|
e649fb7f83 |
perf sched: Fix memory leak in perf_sched__map()
[ Upstream commit ef76a5af819743d405674f6de5d0e63320ac653e ] perf_sched__map() needs to free memory of map_cpus, color_pids and color_cpus in normal path and rollback allocated memory in error path. Signed-off-by: Yang Jihong <yangjihong1@huawei.com> Signed-off-by: Namhyung Kim <namhyung@kernel.org> Link: https://lore.kernel.org/r/20240206083228.172607-3-yangjihong1@huawei.com Stable-dep-of: 1a5efc9e13f3 ("libsubcmd: Don't free the usage string") Signed-off-by: Sasha Levin <sashal@kernel.org> |
||
|
Yang Jihong
|
85ca3d45bd |
perf sched: Move start_work_mutex and work_done_wait_mutex initialization to perf_sched__replay()
[ Upstream commit c6907863519cf97ee09653cc8ec338a2328c2b6f ]
The start_work_mutex and work_done_wait_mutex are used only for the
'perf sched replay'. Put their initialization in perf_sched__replay () to
reduce unnecessary actions in other commands.
Simple functional testing:
# perf sched record perf bench sched messaging
# Running 'sched/messaging' benchmark:
# 20 sender and receiver processes per group
# 10 groups == 400 processes run
Total time: 0.197 [sec]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 14.952 MB perf.data (134165 samples) ]
# perf sched replay
run measurement overhead: 108 nsecs
sleep measurement overhead: 65658 nsecs
the run test took 999991 nsecs
the sleep test took 1079324 nsecs
nr_run_events: 42378
nr_sleep_events: 43102
nr_wakeup_events: 31852
target-less wakeups: 17
multi-target wakeups: 712
task 0 ( swapper: 0), nr_events: 10451
task 1 ( swapper: 1), nr_events: 3
task 2 ( swapper: 2), nr_events: 1
<SNIP>
task 717 ( sched-messaging: 74483), nr_events: 152
task 718 ( sched-messaging: 74484), nr_events: 1944
task 719 ( sched-messaging: 74485), nr_events: 73
task 720 ( sched-messaging: 74486), nr_events: 163
task 721 ( sched-messaging: 74487), nr_events: 942
task 722 ( sched-messaging: 74488), nr_events: 78
task 723 ( sched-messaging: 74489), nr_events: 1090
------------------------------------------------------------
#1 : 1366.507, ravg: 1366.51, cpu: 7682.70 / 7682.70
#2 : 1410.072, ravg: 1370.86, cpu: 7723.88 / 7686.82
#3 : 1396.296, ravg: 1373.41, cpu: 7568.20 / 7674.96
#4 : 1381.019, ravg: 1374.17, cpu: 7531.81 / 7660.64
#5 : 1393.826, ravg: 1376.13, cpu: 7725.25 / 7667.11
#6 : 1401.581, ravg: 1378.68, cpu: 7594.82 / 7659.88
#7 : 1381.337, ravg: 1378.94, cpu: 7371.22 / 7631.01
#8 : 1373.842, ravg: 1378.43, cpu: 7894.92 / 7657.40
#9 : 1364.697, ravg: 1377.06, cpu: 7324.91 / 7624.15
#10 : 1363.613, ravg: 1375.72, cpu: 7209.55 / 7582.69
# echo $?
0
Signed-off-by: Yang Jihong <yangjihong1@huawei.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20240206083228.172607-2-yangjihong1@huawei.com
Stable-dep-of: 1a5efc9e13f3 ("libsubcmd: Don't free the usage string")
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
||
|
Greg Kroah-Hartman
|
1e0f696469 |
Revert "perf callchain: Fix stitch LBR memory leaks"
This reverts commit
|
||
|
Arnaldo Carvalho de Melo
|
e334ae4a0c |
perf python: Allow checking for the existence of warning options in clang
commit b81162302001f41157f6e93654aaccc30e817e2a upstream. We'll need to check if an warning option introduced in clang 19 is available on the clang version being used, so cover the error message emitted when testing for a -W option. Tested-by: Sedat Dilek <sedat.dilek@gmail.com> Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Nathan Chancellor <nathan@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/lkml/CA+icZUVtHn8X1Tb_Y__c-WswsO0K8U9uy3r2MzKXwTA5THtL7w@mail.gmail.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> |
||
|
Namhyung Kim
|
776ebdeee6 |
perf report: Fix segfault when 'sym' sort key is not used
commit 9af2efee41b27a0f386fb5aa95d8d0b4b5d9fede upstream. The fields in the hist_entry are filled on-demand which means they only have meaningful values when relevant sort keys are used. So if neither of 'dso' nor 'sym' sort keys are used, the map/symbols in the hist entry can be garbage. So it shouldn't access it unconditionally. I got a segfault, when I wanted to see cgroup profiles. $ sudo perf record -a --all-cgroups --synth=cgroup true $ sudo perf report -s cgroup Program received signal SIGSEGV, Segmentation fault. 0x00005555557a8d90 in map__dso (map=0x0) at util/map.h:48 48 return RC_CHK_ACCESS(map)->dso; (gdb) bt #0 0x00005555557a8d90 in map__dso (map=0x0) at util/map.h:48 #1 0x00005555557aa39b in map__load (map=0x0) at util/map.c:344 #2 0x00005555557aa592 in map__find_symbol (map=0x0, addr=140736115941088) at util/map.c:385 #3 0x00005555557ef000 in hists__findnew_entry (hists=0x555556039d60, entry=0x7fffffffa4c0, al=0x7fffffffa8c0, sample_self=true) at util/hist.c:644 #4 0x00005555557ef61c in __hists__add_entry (hists=0x555556039d60, al=0x7fffffffa8c0, sym_parent=0x0, bi=0x0, mi=0x0, ki=0x0, block_info=0x0, sample=0x7fffffffaa90, sample_self=true, ops=0x0) at util/hist.c:761 #5 0x00005555557ef71f in hists__add_entry (hists=0x555556039d60, al=0x7fffffffa8c0, sym_parent=0x0, bi=0x0, mi=0x0, ki=0x0, sample=0x7fffffffaa90, sample_self=true) at util/hist.c:779 #6 0x00005555557f00fb in iter_add_single_normal_entry (iter=0x7fffffffa900, al=0x7fffffffa8c0) at util/hist.c:1015 #7 0x00005555557f09a7 in hist_entry_iter__add (iter=0x7fffffffa900, al=0x7fffffffa8c0, max_stack_depth=127, arg=0x7fffffffbce0) at util/hist.c:1260 #8 0x00005555555ba7ce in process_sample_event (tool=0x7fffffffbce0, event=0x7ffff7c14128, sample=0x7fffffffaa90, evsel=0x555556039ad0, machine=0x5555560388e8) at builtin-report.c:334 #9 0x00005555557b30c8 in evlist__deliver_sample (evlist=0x555556039010, tool=0x7fffffffbce0, event=0x7ffff7c14128, sample=0x7fffffffaa90, evsel=0x555556039ad0, machine=0x5555560388e8) at util/session.c:1232 #10 0x00005555557b32bc in machines__deliver_event (machines=0x5555560388e8, evlist=0x555556039010, event=0x7ffff7c14128, sample=0x7fffffffaa90, tool=0x7fffffffbce0, file_offset=110888, file_path=0x555556038ff0 "perf.data") at util/session.c:1271 #11 0x00005555557b3848 in perf_session__deliver_event (session=0x5555560386d0, event=0x7ffff7c14128, tool=0x7fffffffbce0, file_offset=110888, file_path=0x555556038ff0 "perf.data") at util/session.c:1354 #12 0x00005555557affaf in ordered_events__deliver_event (oe=0x555556038e60, event=0x555556135aa0) at util/session.c:132 #13 0x00005555557bb605 in do_flush (oe=0x555556038e60, show_progress=false) at util/ordered-events.c:245 #14 0x00005555557bb95c in __ordered_events__flush (oe=0x555556038e60, how=OE_FLUSH__ROUND, timestamp=0) at util/ordered-events.c:324 #15 0x00005555557bba46 in ordered_events__flush (oe=0x555556038e60, how=OE_FLUSH__ROUND) at util/ordered-events.c:342 #16 0x00005555557b1b3b in perf_event__process_finished_round (tool=0x7fffffffbce0, event=0x7ffff7c15bb8, oe=0x555556038e60) at util/session.c:780 #17 0x00005555557b3b27 in perf_session__process_user_event (session=0x5555560386d0, event=0x7ffff7c15bb8, file_offset=117688, file_path=0x555556038ff0 "perf.data") at util/session.c:1406 As you can see the entry->ms.map was NULL even if he->ms.map has a value. This is because 'sym' sort key is not given, so it cannot assume whether he->ms.sym and entry->ms.sym is the same. I only checked the 'sym' sort key here as it implies 'dso' behavior (so maps are the same). Fixes: ac01c8c4246546fd ("perf hist: Update hist symbol when updating maps") Signed-off-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: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Matt Fleming <matt@readmodwrite.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: https://lore.kernel.org/r/20240826221045.1202305-2-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> |
||
|
Matt Fleming
|
be8d32ebfa |
perf hist: Update hist symbol when updating maps
commit ac01c8c4246546fd8340a232f3ada1921dc0ee48 upstream.
AddressSanitizer found a use-after-free bug in the symbol code which
manifested as 'perf top' segfaulting.
==1238389==ERROR: AddressSanitizer: heap-use-after-free on address 0x60b00c48844b at pc 0x5650d8035961 bp 0x7f751aaecc90 sp 0x7f751aaecc80
READ of size 1 at 0x60b00c48844b thread T193
#0 0x5650d8035960 in _sort__sym_cmp util/sort.c:310
#1 0x5650d8043744 in hist_entry__cmp util/hist.c:1286
#2 0x5650d8043951 in hists__findnew_entry util/hist.c:614
#3 0x5650d804568f in __hists__add_entry util/hist.c:754
#4 0x5650d8045bf9 in hists__add_entry util/hist.c:772
#5 0x5650d8045df1 in iter_add_single_normal_entry util/hist.c:997
#6 0x5650d8043326 in hist_entry_iter__add util/hist.c:1242
#7 0x5650d7ceeefe in perf_event__process_sample /home/matt/src/linux/tools/perf/builtin-top.c:845
#8 0x5650d7ceeefe in deliver_event /home/matt/src/linux/tools/perf/builtin-top.c:1208
#9 0x5650d7fdb51b in do_flush util/ordered-events.c:245
#10 0x5650d7fdb51b in __ordered_events__flush util/ordered-events.c:324
#11 0x5650d7ced743 in process_thread /home/matt/src/linux/tools/perf/builtin-top.c:1120
#12 0x7f757ef1f133 in start_thread nptl/pthread_create.c:442
#13 0x7f757ef9f7db in clone3 ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81
When updating hist maps it's also necessary to update the hist symbol
reference because the old one gets freed in map__put().
While this bug was probably introduced with
|
||
|
Arnaldo Carvalho de Melo
|
bebb4c2405 |
perf python: Disable -Wno-cast-function-type-mismatch if present on clang
commit 00dc514612fe98cfa117193b9df28f15e7c9db9c upstream. The -Wcast-function-type-mismatch option was introduced in clang 19 and its enabled by default, since we use -Werror, and python bindings do casts that are valid but trips this warning, disable it if present. Closes: https://lore.kernel.org/all/CA+icZUXoJ6BS3GMhJHV3aZWyb5Cz2haFneX0C5pUMUUhG-UVKQ@mail.gmail.com Reported-by: Sedat Dilek <sedat.dilek@gmail.com> Tested-by: Sedat Dilek <sedat.dilek@gmail.com> Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Nathan Chancellor <nathan@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: stable@vger.kernel.org # To allow building with the upcoming clang 19 Link: https://lore.kernel.org/lkml/CA+icZUVtHn8X1Tb_Y__c-WswsO0K8U9uy3r2MzKXwTA5THtL7w@mail.gmail.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> |
||
|
Ian Rogers
|
42cd165b4c |
perf callchain: Fix stitch LBR memory leaks
[ Upstream commit 599c19397b17d197fc1184bbc950f163a292efc9 ]
The 'struct callchain_cursor_node' has a 'struct map_symbol' whose maps
and map members are reference counted. Ensure these values use a _get
routine to increment the reference counts and use map_symbol__exit() to
release the reference counts.
Do similar for 'struct thread's prev_lbr_cursor, but save the size of
the prev_lbr_cursor array so that it may be iterated.
Ensure that when stitch_nodes are placed on the free list the
map_symbols are exited.
Fix resolve_lbr_callchain_sample() by replacing list_replace_init() to
list_splice_init(), so the whole list is moved and nodes aren't leaked.
A reproduction of the memory leaks is possible with a leak sanitizer
build in the perf report command of:
```
$ perf record -e cycles --call-graph lbr perf test -w thloop
$ perf report --stitch-lbr
```
Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Fixes:
|
||
|
Ian Rogers
|
3ba5a2e91c |
perf time-utils: Fix 32-bit nsec parsing
[ Upstream commit 38e2648a81204c9fc5b4c87a8ffce93a6ed91b65 ]
The "time utils" test fails in 32-bit builds:
...
parse_nsec_time("18446744073.709551615")
Failed. ptime 4294967295709551615 expected 18446744073709551615
...
Switch strtoul to strtoull as an unsigned long in 32-bit build isn't
64-bits.
Fixes:
|
||
|
Yang Jihong
|
022f9328ef |
perf sched timehist: Fixed timestamp error when unable to confirm event sched_in time
[ Upstream commit 39c243411bdb8fb35777adf49ee32549633c4e12 ]
If sched_in event for current task is not recorded, sched_in timestamp
will be set to end_time of time window interest, causing an error in
timestamp show. In this case, we choose to ignore this event.
Test scenario:
perf[1229608] does not record the first sched_in event, run time and sch delay are both 0
# perf sched timehist
Samples of sched_switch event do not have callchains.
time cpu task name wait time sch delay run time
[tid/pid] (msec) (msec) (msec)
--------------- ------ ------------------------------ --------- --------- ---------
2090450.763231 [0000] perf[1229608] 0.000 0.000 0.000
2090450.763235 [0000] migration/0[15] 0.000 0.001 0.003
2090450.763263 [0001] perf[1229608] 0.000 0.000 0.000
2090450.763268 [0001] migration/1[21] 0.000 0.001 0.004
2090450.763302 [0002] perf[1229608] 0.000 0.000 0.000
2090450.763309 [0002] migration/2[27] 0.000 0.001 0.007
2090450.763338 [0003] perf[1229608] 0.000 0.000 0.000
2090450.763343 [0003] migration/3[33] 0.000 0.001 0.004
Before:
arbitrarily specify a time window of interest, timestamp will be set to an incorrect value
# perf sched timehist --time 100,200
Samples of sched_switch event do not have callchains.
time cpu task name wait time sch delay run time
[tid/pid] (msec) (msec) (msec)
--------------- ------ ------------------------------ --------- --------- ---------
200.000000 [0000] perf[1229608] 0.000 0.000 0.000
200.000000 [0001] perf[1229608] 0.000 0.000 0.000
200.000000 [0002] perf[1229608] 0.000 0.000 0.000
200.000000 [0003] perf[1229608] 0.000 0.000 0.000
200.000000 [0004] perf[1229608] 0.000 0.000 0.000
200.000000 [0005] perf[1229608] 0.000 0.000 0.000
200.000000 [0006] perf[1229608] 0.000 0.000 0.000
200.000000 [0007] perf[1229608] 0.000 0.000 0.000
After:
# perf sched timehist --time 100,200
Samples of sched_switch event do not have callchains.
time cpu task name wait time sch delay run time
[tid/pid] (msec) (msec) (msec)
--------------- ------ ------------------------------ --------- --------- ---------
Fixes:
|
||
|
Yicong Yang
|
fa0720b32a |
perf stat: Display iostat headers correctly
[ Upstream commit 2615639352420e6e3115952c5b8f46846e1c6d0e ]
Currently we'll only print metric headers for metric leader in
aggregration mode. This will make `perf iostat` header not shown
since it'll aggregrated globally but don't have metric events:
root@ubuntu204:/home/yang/linux/tools/perf# ./perf stat --iostat --timeout 1000
Performance counter stats for 'system wide':
port
0000:00 0 0 0 0
0000:80 0 0 0 0
[...]
Fix this by excluding the iostat in the check of printing metric
headers. Then we can see the headers:
root@ubuntu204:/home/yang/linux/tools/perf# ./perf stat --iostat --timeout 1000
Performance counter stats for 'system wide':
port Inbound Read(MB) Inbound Write(MB) Outbound Read(MB) Outbound Write(MB)
0000:00 0 0 0 0
0000:80 0 0 0 0
[...]
Fixes: 193a9e30207f5477 ("perf stat: Don't display metric header for non-leader uncore events")
Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Cameron <jonathan.cameron@huawei.com>
Cc: Junhao He <hejunhao3@huawei.com>
Cc: linuxarm@huawei.com
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Shameerali Kolothum Thodi <shameerali.kolothum.thodi@huawei.com>
Cc: Zeng Tao <prime.zeng@hisilicon.com>
Link: https://lore.kernel.org/r/20240802065800.48774-1-yangyicong@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
|
||
|
Yang Jihong
|
505ec05002 |
perf sched timehist: Fix missing free of session in perf_sched__timehist()
[ Upstream commit 6bdf5168b6fb19541b0c1862bdaa596d116c7bfb ]
When perf_time__parse_str() fails in perf_sched__timehist(),
need to free session that was previously created, fix it.
Fixes:
|
||
|
Kan Liang
|
88c4b5dd21 |
perf report: Fix --total-cycles --stdio output error
[ Upstream commit 3ef44458071a19e5b5832cdfe6f75273aa521b6e ]
The --total-cycles may output wrong information with the --stdio.
For example:
# perf record -e "{cycles,instructions}",cache-misses -b sleep 1
# perf report --total-cycles --stdio
The total cycles output of {cycles,instructions} and cache-misses are
almost the same.
# Samples: 938 of events 'anon group { cycles, instructions }'
# Event count (approx.): 938
#
# Sampled Cycles% Sampled Cycles Avg Cycles% Avg Cycles [Program Block Range]
# ............... .............. ........... .......... ..................................................>
#
11.19% 2.6K 0.10% 21 [perf_iterate_ctx+48 -> >
5.79% 1.4K 0.45% 97 [__intel_pmu_enable_all.constprop.0+80 -> __intel_>
5.11% 1.2K 0.33% 71 [native_write_msr+0 ->>
# Samples: 293 of event 'cache-misses'
# Event count (approx.): 293
#
# Sampled Cycles% Sampled Cycles Avg Cycles% Avg Cycles [Program Block Range]
# ............... .............. ........... .......... ..................................................>
#
11.19% 2.6K 0.13% 21 [perf_iterate_ctx+48 -> >
5.79% 1.4K 0.59% 97 [__intel_pmu_enable_all.constprop.0+80 -> __intel_>
5.11% 1.2K 0.43% 71 [native_write_msr+0 ->>
With the symbol_conf.event_group, the 'perf report' should only report the
block information of the leader event in a group.
However, the current implementation retrieves the next event's block
information, rather than the next group leader's block information.
Make sure the index is updated even if the event is skipped.
With the patch,
# Samples: 293 of event 'cache-misses'
# Event count (approx.): 293
#
# Sampled Cycles% Sampled Cycles Avg Cycles% Avg Cycles [Program Block Range]
# ............... .............. ........... .......... ..................................................>
#
37.98% 9.0K 4.05% 299 [perf_event_addr_filters_exec+0 -> perf_event_a>
11.19% 2.6K 0.28% 21 [perf_iterate_ctx+48 -> >
5.79% 1.4K 1.32% 97 [__intel_pmu_enable_all.constprop.0+80 -> __intel_>
Fixes:
|
||
|
Namhyung Kim
|
297871cb51 |
perf ui/browser/annotate: Use global annotation_options
[ Upstream commit 22197fb296913f83c7182befd2a8b23bf042f279 ] Now it can use the global options and no need save local browser options separately. Reviewed-by: Ian Rogers <irogers@google.com> Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/r/20231128175441.721579-6-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Stable-dep-of: 3ef44458071a ("perf report: Fix --total-cycles --stdio output error") Signed-off-by: Sasha Levin <sashal@kernel.org> |
||
|
Namhyung Kim
|
4c857dcf34 |
perf annotate: Move some source code related fields from 'struct annotation' to 'struct annotated_source'
[ Upstream commit 0aae4c99c5f8f748c6cb5ca03bb3b3ae8cfb10df ] Some fields in the 'struct annotation' are only used with 'struct annotated_source' so better to be moved there in order to reduce memory consumption for other symbols. Reviewed-by: Ian Rogers <irogers@google.com> Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Christophe JAILLET <christophe.jaillet@wanadoo.fr> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/r/20231103191907.54531-5-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Stable-dep-of: 3ef44458071a ("perf report: Fix --total-cycles --stdio output error") Signed-off-by: Sasha Levin <sashal@kernel.org> |
||
|
Namhyung Kim
|
4ef032d899 |
perf annotate: Split branch stack cycles info from 'struct annotation'
[ Upstream commit b7f87e32590bf48eca84f729d3422be7b8dc22d3 ] The cycles info is only meaningful when sample has branch stacks. To save the memory for normal cases, move those fields to a new 'struct annotated_branch' and dynamically allocate it when needed. Also move cycles_hist from annotated_source as it's related here. Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Christophe JAILLET <christophe.jaillet@wanadoo.fr> Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/r/20231103191907.54531-3-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Stable-dep-of: 3ef44458071a ("perf report: Fix --total-cycles --stdio output error") Signed-off-by: Sasha Levin <sashal@kernel.org> |
||
|
Ian Rogers
|
ba18185bea |
perf inject: Fix leader sampling inserting additional samples
[ Upstream commit 79bcd34e0f3da39fda841406ccc957405e724852 ]
The processing of leader samples would turn an individual sample with
a group of read values into multiple samples. 'perf inject' would pass
through the additional samples increasing the output data file size:
$ perf record -g -e "{instructions,cycles}:S" -o perf.orig.data true
$ perf script -D -i perf.orig.data | sed -e 's/perf.orig.data/perf.data/g' > orig.txt
$ perf inject -i perf.orig.data -o perf.new.data
$ perf script -D -i perf.new.data | sed -e 's/perf.new.data/perf.data/g' > new.txt
$ diff -u orig.txt new.txt
--- orig.txt 2024-07-29 14:29:40.606576769 -0700
+++ new.txt 2024-07-29 14:30:04.142737434 -0700
...
-0xc550@perf.data [0x30]: event: 3
+0xc550@perf.data [0xd0]: event: 9
+.
+. ... raw event: size 208 bytes
+. 0000: 09 00 00 00 01 00 d0 00 fc 72 01 86 ff ff ff ff .........r......
+. 0010: 74 7d 2c 00 74 7d 2c 00 fb c3 79 f9 ba d5 05 00 t},.t},...y.....
+. 0020: e6 cb 1a 00 00 00 00 00 01 00 00 00 00 00 00 00 ................
+. 0030: 02 00 00 00 00 00 00 00 76 01 00 00 00 00 00 00 ........v.......
+. 0040: e6 cb 1a 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
+. 0050: 62 18 00 00 00 00 00 00 f6 cb 1a 00 00 00 00 00 b...............
+. 0060: 00 00 00 00 00 00 00 00 0c 00 00 00 00 00 00 00 ................
+. 0070: 80 ff ff ff ff ff ff ff fc 72 01 86 ff ff ff ff .........r......
+. 0080: f3 0e 6e 85 ff ff ff ff 0c cb 7f 85 ff ff ff ff ..n.............
+. 0090: bc f2 87 85 ff ff ff ff 44 af 7f 85 ff ff ff ff ........D.......
+. 00a0: bd be 7f 85 ff ff ff ff 26 d0 7f 85 ff ff ff ff ........&.......
+. 00b0: 6d a4 ff 85 ff ff ff ff ea 00 20 86 ff ff ff ff m......... .....
+. 00c0: 00 fe ff ff ff ff ff ff 57 14 4f 43 fc 7e 00 00 ........W.OC.~..
+
+1642373909693435 0xc550 [0xd0]: PERF_RECORD_SAMPLE(IP, 0x1): 2915700/2915700: 0xffffffff860172fc period: 1 addr: 0
+... FP chain: nr:12
+..... 0: ffffffffffffff80
+..... 1: ffffffff860172fc
+..... 2: ffffffff856e0ef3
+..... 3: ffffffff857fcb0c
+..... 4: ffffffff8587f2bc
+..... 5: ffffffff857faf44
+..... 6: ffffffff857fbebd
+..... 7: ffffffff857fd026
+..... 8: ffffffff85ffa46d
+..... 9: ffffffff862000ea
+..... 10: fffffffffffffe00
+..... 11: 00007efc434f1457
+... sample_read:
+.... group nr 2
+..... id 00000000001acbe6, value 0000000000000176, lost 0
+..... id 00000000001acbf6, value 0000000000001862, lost 0
+
+0xc620@perf.data [0x30]: event: 3
...
This behavior is incorrect as in the case above 'perf inject' should
have done nothing. Fix this behavior by disabling separating samples
for a tool that requests it. Only request this for `perf inject` so as
to not affect other perf tools. With the patch and the test above
there are no differences between the orig.txt and new.txt.
Fixes:
|
||
|
Namhyung Kim
|
1490a5dbd5 |
perf mem: Free the allocated sort string, fixing a leak
[ Upstream commit 3da209bb1177462b6fe8e3021a5527a5a49a9336 ]
The get_sort_order() returns either a new string (from strdup) or NULL
but it never gets freed.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Fixes:
|
||
|
Casey Chen
|
1b46b23561 |
perf tool: fix dereferencing NULL al->maps
[ Upstream commit 4c17736689ccfc44ec7dcc472577f25c34cf8724 ] With |
||
|
Kan Liang
|
7d049cd134 |
perf stat: Fix the hard-coded metrics calculation on the hybrid
commit 3612ca8e2935c4c142d99e33b8effa7045ce32b5 upstream.
The hard-coded metrics is wrongly calculated on the hybrid machine.
$ perf stat -e cycles,instructions -a sleep 1
Performance counter stats for 'system wide':
18,205,487 cpu_atom/cycles/
9,733,603 cpu_core/cycles/
9,423,111 cpu_atom/instructions/ # 0.52 insn per cycle
4,268,965 cpu_core/instructions/ # 0.23 insn per cycle
The insn per cycle for cpu_core should be 4,268,965 / 9,733,603 = 0.44.
When finding the metric events, the find_stat() doesn't take the PMU
type into account. The cpu_atom/cycles/ is wrongly used to calculate
the IPC of the cpu_core.
In the hard-coded metrics, the events from a different PMU are only
SW_CPU_CLOCK and SW_TASK_CLOCK. They both have the stat type,
STAT_NSECS. Except the SW CLOCK events, check the PMU type as well.
Fixes:
|
||
|
Adrian Hunter
|
3fc173fb4f |
perf intel-pt: Fix exclude_guest setting
[ Upstream commit b40934ae32232140e85dc7dc1c3ea0e296986723 ]
In the past, the exclude_guest setting has had no effect on Intel PT
tracing, but that may not be the case in the future.
Set the flag correctly based upon whether KVM is using Intel PT
"Host/Guest" mode, which is determined by the kvm_intel module
parameter pt_mode:
pt_mode=0 System-wide mode : host and guest output to host buffer
pt_mode=1 Host/Guest mode : host/guest output to host/guest
buffers respectively
Fixes:
|
||
|
Adrian Hunter
|
f24f95be74 |
perf intel-pt: Fix aux_watermark calculation for 64-bit size
[ Upstream commit 36b4cd990a8fd3f5b748883050e9d8c69fe6398d ]
aux_watermark is a u32. For a 64-bit size, cap the aux_watermark
calculation at UINT_MAX instead of truncating it to 32-bits.
Fixes:
|