user/sven/linux.git/tools/perf, branch v6.9.5

Revert "perf record: Reduce memory for recording PERF_RECORD_LOST_SAMPLES event"

2024-06-16T11:51:10Z

commit 5b3cde198878b2f3269d5e7efbc0d514899b1fd8 upstream. This reverts commit 7d1405c71df21f6c394b8a885aa8a133f749fa22. This causes segfaults in some cases, as reported by Milian: ``` sudo /usr/bin/perf record -z --call-graph dwarf -e cycles -e raw_syscalls:sys_enter ls ... [ perf record: Woken up 3 times to write data ] malloc(): invalid next size (unsorted) Aborted ``` Backtrace with GDB + debuginfod: ``` malloc(): invalid next size (unsorted) Thread 1 "perf" received signal SIGABRT, Aborted. __pthread_kill_implementation (threadid=, signo=signo@entry=6, no_tid=no_tid@entry=0) at pthread_kill.c:44 Downloading source file /usr/src/debug/glibc/glibc/nptl/pthread_kill.c 44 return INTERNAL_SYSCALL_ERROR_P (ret) ? INTERNAL_SYSCALL_ERRNO (ret) : 0; (gdb) bt #0 __pthread_kill_implementation (threadid=, signo=signo@entry=6, no_tid=no_tid@entry=0) at pthread_kill.c:44 #1 0x00007ffff6ea8eb3 in __pthread_kill_internal (threadid=, signo=6) at pthread_kill.c:78 #2 0x00007ffff6e50a30 in __GI_raise (sig=sig@entry=6) at ../sysdeps/posix/ raise.c:26 #3 0x00007ffff6e384c3 in __GI_abort () at abort.c:79 #4 0x00007ffff6e39354 in __libc_message_impl (fmt=fmt@entry=0x7ffff6fc22ea "%s\n") at ../sysdeps/posix/libc_fatal.c:132 #5 0x00007ffff6eb3085 in malloc_printerr (str=str@entry=0x7ffff6fc5850 "malloc(): invalid next size (unsorted)") at malloc.c:5772 #6 0x00007ffff6eb657c in _int_malloc (av=av@entry=0x7ffff6ff6ac0 , bytes=bytes@entry=368) at malloc.c:4081 #7 0x00007ffff6eb877e in __libc_calloc (n=, elem_size=) at malloc.c:3754 #8 0x000055555569bdb6 in perf_session.do_write_header () #9 0x00005555555a373a in __cmd_record.constprop.0 () #10 0x00005555555a6846 in cmd_record () #11 0x000055555564db7f in run_builtin () #12 0x000055555558ed77 in main () ``` Valgrind memcheck: ``` ==45136== Invalid write of size 8 ==45136== at 0x2B38A5: perf_event__synthesize_id_sample (in /usr/bin/perf) ==45136== by 0x157069: __cmd_record.constprop.0 (in /usr/bin/perf) ==45136== by 0x15A845: cmd_record (in /usr/bin/perf) ==45136== by 0x201B7E: run_builtin (in /usr/bin/perf) ==45136== by 0x142D76: main (in /usr/bin/perf) ==45136== Address 0x6a866a8 is 0 bytes after a block of size 40 alloc'd ==45136== at 0x4849BF3: calloc (vg_replace_malloc.c:1675) ==45136== by 0x3574AB: zalloc (in /usr/bin/perf) ==45136== by 0x1570E0: __cmd_record.constprop.0 (in /usr/bin/perf) ==45136== by 0x15A845: cmd_record (in /usr/bin/perf) ==45136== by 0x201B7E: run_builtin (in /usr/bin/perf) ==45136== by 0x142D76: main (in /usr/bin/perf) ==45136== ==45136== Syscall param write(buf) points to unaddressable byte(s) ==45136== at 0x575953D: __libc_write (write.c:26) ==45136== by 0x575953D: write (write.c:24) ==45136== by 0x35761F: ion (in /usr/bin/perf) ==45136== by 0x357778: writen (in /usr/bin/perf) ==45136== by 0x1548F7: record__write (in /usr/bin/perf) ==45136== by 0x15708A: __cmd_record.constprop.0 (in /usr/bin/perf) ==45136== by 0x15A845: cmd_record (in /usr/bin/perf) ==45136== by 0x201B7E: run_builtin (in /usr/bin/perf) ==45136== by 0x142D76: main (in /usr/bin/perf) ==45136== Address 0x6a866a8 is 0 bytes after a block of size 40 alloc'd ==45136== at 0x4849BF3: calloc (vg_replace_malloc.c:1675) ==45136== by 0x3574AB: zalloc (in /usr/bin/perf) ==45136== by 0x1570E0: __cmd_record.constprop.0 (in /usr/bin/perf) ==45136== by 0x15A845: cmd_record (in /usr/bin/perf) ==45136== by 0x201B7E: run_builtin (in /usr/bin/perf) ==45136== by 0x142D76: main (in /usr/bin/perf) ==45136== ----- Closes: https://lore.kernel.org/linux-perf-users/23879991.0LEYPuXRzz@milian-workstation/ Reported-by: Milian Wolff Tested-by: Milian Wolff Cc: Adrian Hunter Cc: Ian Rogers Cc: Jiri Olsa Cc: Kan Liang Cc: Namhyung Kim Cc: stable@kernel.org # 6.8+ Link: https://lore.kernel.org/lkml/Zl9ksOlHJHnKM70p@x1 Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Greg Kroah-Hartman

perf pmu: Count sys and cpuid JSON events separately

2024-06-12T09:39:23Z

[ Upstream commit d9c5f5f94c2d356fdf3503f7fcaf254512bc032d ] Sys events are eagerly loaded as each event has a compat option that may mean the event is or isn't associated with the PMU. These shouldn't be counted as loaded_json_events as that is used for JSON events matching the CPUID that may or may not have been loaded. The mismatch causes issues on ARM64 that uses sys events. Fixes: e6ff1eed3584362d ("perf pmu: Lazily add JSON events") Closes: https://lore.kernel.org/lkml/20240510024729.1075732-1-justin.he@arm.com/ Reported-by: Jia He Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: John Garry Cc: Kan Liang Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20240511003601.2666907-1-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Sasha Levin

perf pmu: Assume sysfs events are always the same case

2024-06-12T09:39:22Z

[ Upstream commit 7b6dd7a923281a7ccb980a0f768d6926721eb3cc ] Perf event names aren't case sensitive. For sysfs events the entire directory of events is read then iterated comparing names in a case insensitive way, most often to see if an event is present. Consider: $ perf stat -e inst_retired.any true The event inst_retired.any may be present in any PMU, so every PMU's sysfs events are loaded and then searched with strcasecmp to see if any match. This event is only present on the cpu PMU as a JSON event so a lot of events were loaded from sysfs unnecessarily just to prove an event didn't exist there. This change avoids loading all the events by assuming sysfs event names are always either lower or uppercase. It uses file exists and only loads the events when the desired event is present. For the example above, the number of openat calls measured by 'perf trace' on a tigerlake laptop goes from 325 down to 255. The reduction will be larger for machines with many PMUs, particularly replicated uncore PMUs. Ensure pmu_aliases_parse() is called before all uses of the aliases list, but remove some "pmu->sysfs_aliases_loaded" tests as they are now part of the function. Reviewed-by: Kan Liang Signed-off-by: Ian Rogers Cc: Alexander Shishkin Cc: Bjorn Helgaas Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: Jonathan Corbet Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Randy Dunlap Cc: Ravi Bangoria Cc: Thomas Richter Link: https://lore.kernel.org/r/20240502213507.2339733-7-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo Stable-dep-of: d9c5f5f94c2d ("perf pmu: Count sys and cpuid JSON events separately") Signed-off-by: Sasha Levin

perf tools: Add/use PMU reverse lookup from config to name

2024-06-12T09:39:22Z

[ Upstream commit 67ee8e71daabb8632931b7559e5c8a4b69a427f8 ] Add perf_pmu__name_from_config that does a reverse lookup from a config number to an alias name. The lookup is expensive as the config is computed for every alias by filling in a perf_event_attr, but this is only done when verbose output is enabled. The lookup also only considers config, and not config1, config2 or config3. An example of the output: $ perf stat -vv -e data_read true ... perf_event_attr: type 24 (uncore_imc_free_running_0) size 136 config 0x20ff (data_read) sample_type IDENTIFIER read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING disabled 1 inherit 1 exclude_guest 1 ... Committer notes: Fix the python binding build by adding dummies for not strictly needed perf_pmu__name_from_config() and perf_pmus__find_by_type(). Signed-off-by: Ian Rogers Tested-by: Arnaldo Carvalho de Melo Tested-by: Kan Liang Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Ravi Bangoria Cc: Yang Jihong Link: https://lore.kernel.org/r/20240308001915.4060155-7-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo Stable-dep-of: d9c5f5f94c2d ("perf pmu: Count sys and cpuid JSON events separately") Signed-off-by: Sasha Levin

perf tools: Use pmus to describe type from attribute

2024-06-12T09:39:22Z

[ Upstream commit 7093882067e2e2f88d3449c35c5f0f3f566c8a26 ] When dumping a perf_event_attr, use pmus to find the PMU and its name by the type number. This allows dynamically added PMUs to be described. Before: $ perf stat -vv -e data_read true ... perf_event_attr: type 24 size 136 config 0x20ff sample_type IDENTIFIER read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING disabled 1 inherit 1 exclude_guest 1 ... After: $ perf stat -vv -e data_read true ... perf_event_attr: type 24 (uncore_imc_free_running_0) size 136 config 0x20ff sample_type IDENTIFIER read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING disabled 1 inherit 1 exclude_guest 1 ... However, it also means that when we have a PMU name we prefer it to a hard coded name: Before: $ perf stat -vv -e faults true ... perf_event_attr: type 1 (PERF_TYPE_SOFTWARE) size 136 config 0x2 (PERF_COUNT_SW_PAGE_FAULTS) sample_type IDENTIFIER read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING disabled 1 inherit 1 enable_on_exec 1 exclude_guest 1 ... After: $ perf stat -vv -e faults true ... perf_event_attr: type 1 (software) size 136 config 0x2 (PERF_COUNT_SW_PAGE_FAULTS) sample_type IDENTIFIER read_format TOTAL_TIME_ENABLED|TOTAL_TIME_RUNNING disabled 1 inherit 1 enable_on_exec 1 exclude_guest 1 ... It feels more consistent to do this, rather than only prefer a PMU name when a hard coded name isn't available. Signed-off-by: Ian Rogers Tested-by: Arnaldo Carvalho de Melo Tested-by: Kan Liang Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Ravi Bangoria Cc: Yang Jihong Link: https://lore.kernel.org/r/20240308001915.4060155-6-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo Stable-dep-of: d9c5f5f94c2d ("perf pmu: Count sys and cpuid JSON events separately") Signed-off-by: Sasha Levin

perf stat: Don't display metric header for non-leader uncore events

2024-06-12T09:39:22Z

[ Upstream commit 193a9e30207f54777ff42d0d8be8389edc522277 ] On an Intel tigerlake laptop a metric like: { "BriefDescription": "Test", "MetricExpr": "imc_free_running@data_read@ + imc_free_running@data_write@", "MetricGroup": "Test", "MetricName": "Test", "ScaleUnit": "6.103515625e-5MiB" }, Will have 4 events: uncore_imc_free_running_0/data_read/ uncore_imc_free_running_0/data_write/ uncore_imc_free_running_1/data_read/ uncore_imc_free_running_1/data_write/ If aggregration is disabled with metric-only 2 column headers are needed: $ perf stat -M test --metric-only -A -a sleep 1 Performance counter stats for 'system wide': MiB Test MiB Test CPU0 1821.0 1820.5 But when not, the counts aggregated in the metric leader and only 1 column should be shown: $ perf stat -M test --metric-only -a sleep 1 Performance counter stats for 'system wide': MiB Test 5909.4 1.001258915 seconds time elapsed Achieve this by skipping events that aren't metric leaders when printing column headers and aggregation isn't disabled. The bug is long standing, the fixes tag is set to a refactor as that is as far back as is reasonable to backport. Fixes: 088519f318be3a41 ("perf stat: Move the display functions to stat-display.c") Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ingo Molnar Cc: Jiri Olsa Cc: Kaige Ye Cc: Kan Liang Cc: K Prateek Nayak Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Yicong Yang Link: https://lore.kernel.org/r/20240510051309.2452468-1-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Sasha Levin

perf annotate: Fix segfault on sample histogram

2024-06-12T09:39:22Z

[ Upstream commit 9ef30265a483f0405e4f7b3f15cda251b9a2c7da ] A symbol can have no samples, then accessing the annotated_source->samples hashmap will result in a segfault. Fixes: a3f7768bcf48281d ("perf annotate: Fix memory leak in annotated_source") Reviewed-by: Ian Rogers Signed-off-by: Namhyung Kim Cc: Adrian Hunter Cc: Ingo Molnar Cc: Jiri Olsa Cc: Kan Liang Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20240510210452.2449944-1-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Sasha Levin

perf daemon: Fix file leak in daemon_session__control

2024-06-12T09:39:21Z

[ Upstream commit 09541603462c399c7408d50295db99b4b8042eaa ] The open() function returns -1 on error. The 'control' and 'ack' file descriptors are both initialized with open() and further validated with 'if' statement. 'if (!control)' would evaluate to 'true' if returned value on error were '0' but it is actually '-1'. Fixes: edcaa47958c7438b ("perf daemon: Add 'ping' command") Signed-off-by: Samasth Norway Ananda Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ian Rogers Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20240510003424.2016914-1-samasth.norway.ananda@oracle.com Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Sasha Levin

perf symbols: Fix ownership of string in dso__load_vmlinux()

2024-06-12T09:39:20Z

[ Upstream commit 25626e19ae6df34f336f235b6b3dbd1b566d2738 ] The linked commit updated dso__load_vmlinux() to call dso__set_long_name() before loading the symbols. Loading the symbols may not succeed but dso__set_long_name() takes ownership of the string. The two callers of this function free the string themselves on failure cases, resulting in the following error: $ perf record -- ls $ perf report free(): double free detected in tcache 2 Fix it by always taking ownership of the string, even on failure. This means the string is either freed at the very first early exit condition, or later when the dso is deleted or the long name is replaced. Now no special return value is needed to signify that the caller needs to free the string. Fixes: e59fea47f83e8a9a ("perf symbols: Fix DSO kernel load and symbol process to correctly map DSO to its long_name, type and adjust_symbols") Reviewed-by: Ian Rogers Signed-off-by: James Clark Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Athira Rajeev Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20240507141210.195939-5-james.clark@arm.com Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Sasha Levin

perf symbols: Update kcore map before merging in remaining symbols

2024-06-12T09:39:20Z

[ Upstream commit f30232b20fadea8c0f2f43f764bc06e51e8cfcdf ] When loading kcore, the main vmlinux map is updated in the same loop that merges the remaining maps. If a map that overlaps is merged in before kcore, the list can become unsortable when the main map addresses are updated. This will later trigger the check_invariants() assert: $ perf record $ perf report util/maps.c:96: check_invariants: Assertion `map__end(prev) <= map__start(map) || map__start(prev) == map__start(map)' failed. Aborted Fix it by moving the main map update prior to the loop so that maps__merge_in() can split it if necessary. Fixes: 659ad3492b913c90 ("perf maps: Switch from rbtree to lazily sorted array for addresses") Signed-off-by: James Clark Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Athira Rajeev Cc: Ian Rogers Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20240507141210.195939-4-james.clark@arm.com Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Sasha Levin