Introduce the `aggr_nr` function to the metric expression parser.
`aggr_nr` allows metric formulas to dynamically utilize the number of
aggregated targets (`aggr->nr`) instead of relying on the static
`source_count` (which represents the static socket or node count).
This adds the `AGGR_NR` token to the lexer and parser, updates the
expression parsing context helpers to store `aggr_nr`, and feeds
`aggr->nr` from the aggregation structure in `prepare_metric`.
Assisted-by: Gemini:gemini-3.1-pro-preview
Signed-off-by: Chun-Tse Shao <[email protected]>
Tested-by: Zide Chen <[email protected]>
Acked-by: Namhyung Kim <[email protected]>
Cc: Adrian Hunter <[email protected]>
Cc: Alexander Shishkin <[email protected]>
Cc: Ian Rogers <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: James Clark <[email protected]>
Cc: Jiri Olsa <[email protected]>
Cc: Leo Yan <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Sandipan Das <[email protected]>
Cc: Thomas Falcon <[email protected]>
Cc: Yang Li <[email protected]>
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
Add a trailing newline for logs.
Before:
perf stat -C 5
Failure to read '#slots'Cannot find metric or group `Default'
After:
perf stat -C 5
Failure to read '#slots'
Cannot find metric or group `Default'
Signed-off-by: Leo Yan <[email protected]>
Reviewed-by: Ian Rogers <[email protected]>
Signed-off-by: Namhyung Kim <[email protected]>
expr__find_ids() propagates the parser return value directly. For syntax
errors, the parser can return a positive value, but callers treat it as
success, e.g., for below case on Arm64 platform:
metric expr 100 * (STALL_SLOT_BACKEND / (CPU_CYCLES * #slots) - BR_MIS_PRED * 3 / CPU_CYCLES) for backend_bound
parsing metric: 100 * (STALL_SLOT_BACKEND / (CPU_CYCLES * #slots) - BR_MIS_PRED * 3 / CPU_CYCLES)
Failure to read '#slots' literal: #slots = nan
syntax error
Convert positive parser returns in expr__find_ids() to -EINVAL, as a
result, the error value will be respected by callers.
Before:
perf stat -C 5
Failure to read '#slots'Failure to read '#slots'Failure to read '#slots'Failure to read '#slots'Segmentation fault
After:
perf stat -C 5
Failure to read '#slots'Cannot find metric or group `Default'
Fixes: ded80bda8b ("perf expr: Migrate expr ids table to a hashmap")
Signed-off-by: Leo Yan <[email protected]>
Reviewed-by: Ian Rogers <[email protected]>
Signed-off-by: Namhyung Kim <[email protected]>
For the sake of better documentation, add core_wide and target_cpu to
the tool.json. When the values of system_wide and
user_requested_cpu_list are unknown, use the values from the global
stat_config.
Example output showing how '-a' modifies the values in `perf stat`:
```
$ perf stat -e core_wide,target_cpu true
Performance counter stats for 'true':
0 core_wide
0 target_cpu
0.000993787 seconds time elapsed
0.001128000 seconds user
0.000000000 seconds sys
$ perf stat -e core_wide,target_cpu -a true
Performance counter stats for 'system wide':
1 core_wide
1 target_cpu
0.002271723 seconds time elapsed
$ perf list
...
tool:
core_wide
[1 if not SMT,if SMT are events being gathered on all SMT threads 1 otherwise 0. Unit: tool]
...
target_cpu
[1 if CPUs being analyzed,0 if threads/processes. Unit: tool]
...
```
Signed-off-by: Ian Rogers <[email protected]>
Signed-off-by: Namhyung Kim <[email protected]>
For CPU nanoseconds a lot of the stat-shadow metrics use either
task-clock or cpu-clock, the latter being used when
target__has_cpu. Add a #target_cpu literal so that json metrics can
perform the same test.
Signed-off-by: Ian Rogers <[email protected]>
Signed-off-by: Namhyung Kim <[email protected]>
For hybrid metrics it is useful to know the number of p-core or e-core
CPUs. If a cpumask is specified for the num_cpus or num_cpus_online
tool events, compute the value relative to the given mask rather than
for the full system.
```
$ sudo /tmp/perf/perf stat -e 'tool/num_cpus/,tool/num_cpus,cpu=cpu_core/,
tool/num_cpus,cpu=cpu_atom/,tool/num_cpus_online/,tool/num_cpus_online,
cpu=cpu_core/,tool/num_cpus_online,cpu=cpu_atom/' true
Performance counter stats for 'true':
28 tool/num_cpus/
16 tool/num_cpus,cpu=cpu_core/
12 tool/num_cpus,cpu=cpu_atom/
28 tool/num_cpus_online/
16 tool/num_cpus_online,cpu=cpu_core/
12 tool/num_cpus_online,cpu=cpu_atom/
0.000767205 seconds time elapsed
0.000938000 seconds user
0.000000000 seconds sys
```
Reviewed-by: Thomas Falcon <[email protected]>
Signed-off-by: Ian Rogers <[email protected]>
Tested-by: James Clark <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Namhyung Kim <[email protected]>
Metrics will fill in the context to have mappings from an event to a
count. When counts are added they replace existing mappings which
generally shouldn't exist with aggregation. Switch to accumulating to
better support cases where perf stat's aggregation isn't used and we
may see a counter more than once.
Signed-off-by: Ian Rogers <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Namhyung Kim <[email protected]>
Some topdown related metrics may fail on hybrid machines.
$ perf stat -M tma_frontend_bound
Cannot resolve IDs for tma_frontend_bound:
cpu_atom@TOPDOWN_FE_BOUND.ALL@ / (8 * cpu_atom@CPU_CLK_UNHALTED.CORE@)
In the find_tool_events(), the tool_pmu__event_to_str() is used to
compare the tool_events. It only checks the event name, no PMU or arch.
So the tool_events[TOOL_PMU__EVENT_SLOTS] is set to true, because the
p-core Topdown metrics has "slots" event.
The tool_events is shared. So when parsing the e-core metrics, the
"slots" is automatically added.
The "slots" event as a tool event should only be available on arm64. It
has a different meaning on X86. The tool_pmu__skip_event() intends
handle the case. Apply it for tool_pmu__event_to_str() as well.
There is a lack of sanity check in the expr__get_id(). Add the check.
Closes: https://lore.kernel.org/lkml/[email protected]/
Fixes: 069057239a ("perf tool_pmu: Move expr literals to tool_pmu")
Signed-off-by: Kan Liang <[email protected]>
Reviewed-by: Ian Rogers <[email protected]>
Cc: [email protected]
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Namhyung Kim <[email protected]>
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: 3340a08354 ("perf pmu-events: Fix testing with JEVENTS_ARCH=all")
Reviewed-by: Ian Rogers <[email protected]>
Reviewed-by: James Clark <[email protected]>
Signed-off-by: Levi Yun <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Namhyung Kim <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
Add the expr literals like "#smt_on" as tool events, this allows stat
events to give the values. On my laptop with hyperthreading enabled:
```
$ perf stat -e "has_pmem,num_cores,num_cpus,num_cpus_online,num_dies,num_packages,smt_on,system_tsc_freq" true
Performance counter stats for 'true':
0 has_pmem
8 num_cores
16 num_cpus
16 num_cpus_online
1 num_dies
1 num_packages
1 smt_on
2,496,000,000 system_tsc_freq
0.001113637 seconds time elapsed
0.001218000 seconds user
0.000000000 seconds sys
```
And with hyperthreading disabled:
```
$ perf stat -e "has_pmem,num_cores,num_cpus,num_cpus_online,num_dies,num_packages,smt_on,system_tsc_freq" true
Performance counter stats for 'true':
0 has_pmem
8 num_cores
16 num_cpus
8 num_cpus_online
1 num_dies
1 num_packages
0 smt_on
2,496,000,000 system_tsc_freq
0.000802115 seconds time elapsed
0.000000000 seconds user
0.000806000 seconds sys
```
As zero matters for these values, in stat-display
should_skip_zero_counter only skip the zero value if it is not the
first aggregation index.
The tool event implementations are used in expr but not evaluated as
events for simplicity. Also core_wide isn't made a tool event as it
requires command line parameters.
Signed-off-by: Ian Rogers <[email protected]>
Acked-by: Namhyung Kim <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Namhyung Kim <[email protected]>
Compiling perf tool with 'DEBUG_PARSER=1' leads to errors:
$> make -C tools/perf PARSER_DEBUG=1 NO_LIBTRACEEVENT=1
...
CC util/expr-flex.o
CC util/expr.o
util/parse-events.c:33:12: error: redundant redeclaration of ‘parse_events_debug’ [-Werror=redundant-decls]
33 | extern int parse_events_debug;
| ^~~~~~~~~~~~~~~~~~
In file included from util/parse-events.c:18:
util/parse-events-bison.h:43:12: note: previous declaration of ‘parse_events_debug’ with type ‘int’
43 | extern int parse_events_debug;
| ^~~~~~~~~~~~~~~~~~
util/expr.c:27:12: error: redundant redeclaration of ‘expr_debug’ [-Werror=redundant-decls]
27 | extern int expr_debug;
| ^~~~~~~~~~
In file included from util/expr.c:11:
util/expr-bison.h:43:12: note: previous declaration of ‘expr_debug’ with type ‘int’
43 | extern int expr_debug;
| ^~~~~~~~~~
cc-1: all warnings being treated as errors
Remove extern declaration from the parse-envents.c file as there is a
conflict with the ones generated using bison and yacc tools from the file
parse-events.[ly].
Signed-off-by: Clément Le Goffic <[email protected]>
Reviewed-by: Ian Rogers <[email protected]>
Cc: James Clark <[email protected]>
Cc: John Garry <[email protected]>
Signed-off-by: Namhyung Kim <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
pmu__find_core_pmu() more logically belongs in pmus.c because it
iterates over all PMUs, so move it to pmus.c
At the same time rename it to perf_pmus__find_core_pmu() to match the
naming convention in this file.
list_prepare_entry() can't be used in perf_pmus__scan_core() anymore now
that it's called from the same compilation unit. This is with -O2
(specifically -O1 -ftree-vrp -finline-functions
-finline-small-functions) which allow the bounds of the array
access to be determined at compile time. list_prepare_entry() subtracts
the offset of the 'list' member in struct perf_pmu from &core_pmus,
which isn't a struct perf_pmu. The compiler sees that pmu results in
&core_pmus - 8 and refuses to compile. At runtime this works because
list_for_each_entry_continue() always adds the offset back again before
dereferencing ->next, but it's technically undefined behavior. With
-fsanitize=undefined an additional warning is generated.
Using list_first_entry_or_null() to get the first entry here avoids
doing &core_pmus - 8 but has the same result and fixes both the compile
warning and the undefined behavior warning. There are other uses of
list_prepare_entry() in pmus.c, but the compiler doesn't seem to be
able to see that they can also be called with &core_pmus, so I won't
change any at this time.
Signed-off-by: James Clark <[email protected]>
Reviewed-by: Ian Rogers <[email protected]>
Reviewed-by: John Garry <[email protected]>
Cc: Ravi Bangoria <[email protected]>
Cc: Eduard Zingerman <[email protected]>
Cc: Will Deacon <[email protected]>
Cc: Leo Yan <[email protected]>
Cc: Mike Leach <[email protected]>
Cc: Jing Zhang <[email protected]>
Cc: Haixin Yu <[email protected]>
Cc: Kan Liang <[email protected]>
Cc: [email protected]
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Namhyung Kim <[email protected]>
Do defensive programming by using zfree() to initialize freed pointers
to NULL, so that eventual use after free result in a NULL pointer deref
instead of more subtle behaviour.
Also remove one NULL test before free(), as it accepts a NULL arg and we
get one line shaved not doing it explicitely.
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
To resolve a trivial merge conflict with c302378bc1 ("libbpf:
Hashmap interface update to allow both long and void* keys/values"),
where a function present upstream was removed in the perf tools
development tree.
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
An update for libbpf's hashmap interface from void* -> void* to a
polymorphic one, allowing both long and void* keys and values.
This simplifies many use cases in libbpf as hashmaps there are mostly
integer to integer.
Perf copies hashmap implementation from libbpf and has to be
updated as well.
Changes to libbpf, selftests/bpf and perf are packed as a single
commit to avoid compilation issues with any future bisect.
Polymorphic interface is acheived by hiding hashmap interface
functions behind auxiliary macros that take care of necessary
type casts, for example:
#define hashmap_cast_ptr(p) \
({ \
_Static_assert((p) == NULL || sizeof(*(p)) == sizeof(long),\
#p " pointee should be a long-sized integer or a pointer"); \
(long *)(p); \
})
bool hashmap_find(const struct hashmap *map, long key, long *value);
#define hashmap__find(map, key, value) \
hashmap_find((map), (long)(key), hashmap_cast_ptr(value))
- hashmap__find macro casts key and value parameters to long
and long* respectively
- hashmap_cast_ptr ensures that value pointer points to a memory
of appropriate size.
This hack was suggested by Andrii Nakryiko in [1].
This is a follow up for [2].
[1] https://lore.kernel.org/bpf/CAEf4BzZ8KFneEJxFAaNCCFPGqp20hSpS2aCj76uRk3-qZUH5xg@mail.gmail.com/
[2] https://lore.kernel.org/bpf/[email protected]/T/#m65b28f1d6d969fcd318b556db6a3ad499a42607d
Signed-off-by: Eduard Zingerman <[email protected]>
Signed-off-by: Andrii Nakryiko <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]
The TSC frequency information is required for the event metrics with the
literal, system_tsc_freq. For the newer Intel platform, the TSC
frequency information can be retrieved from the CPUID leaf 0x15. If the
TSC frequency information isn't present the /proc/cpuinfo approach is
used.
Refactor cpuid() for this use. Note, the previous stack pushing/popping
approach was broken on x86-64 that has stack red zones that would be
clobbered.
Committer testing:
Before:
$ perf record sleep 0.0001
[ perf record: Woken up 1 times to write data ]
$ perf report --header-only |& grep cpuid
# cpuid : AuthenticAMD,25,33,0
$
After the patch:
$ perf record sleep 0.0001
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.002 MB perf.data (8 samples) ]
$ perf report --header-only |& grep cpuid
# cpuid : AuthenticAMD,25,33,0
$
Signed-off-by: Kan Liang <[email protected]>
Tested-by: Arnaldo Carvalho de Melo <[email protected]>
Cc: Alexander Shishkin <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Andi Kleen <[email protected]>
Cc: Caleb Biggers <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: James Clark <[email protected]>
Cc: Jiri Olsa <[email protected]>
Cc: John Garry <[email protected]>
Cc: Kshipra Bopardikar <[email protected]>
Cc: Mark Rutland <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Perry Taylor <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Stephane Eranian <[email protected]>
Cc: Xing Zhengjun <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Ian Rogers <[email protected]>
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
This patch resolves some undefined behavior where variables in
expr_id_data were accessed (for debugging) without being defined. To
better enforce the tagged union behavior, the struct is moved into
expr.c and accessors provided. Tag values (kinds) are explicitly
identified.
Signed-off-by: Ian Rogers <[email protected]>
Reviewed-By: Kajol Jain<[email protected]>
Acked-by: Jiri Olsa <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
Keeping the stack of nested metrics via 'struct expr_id' objects
and checking if we are in recursion via already processed metric.
The stack is implemented as static array within the struct egroup
with 100 entries, which should be enough nesting depth for any
metric we have or plan to have at the moment.
Adding test that simulates the recursion and checks we can
detect it.
Committer notes:
Bumped RECURSION_ID_MAX to 1000 as per Jiri's reply to Paul Clark on the
patch series e-mail discussion.
Fixed these:
tests/parse-metric.c:308:7: error: missing field 'val' initializer [-Werror,-Wmissing-field-initializers]
{ 0 },
^
util/metricgroup.c:924:28: error: missing field 'parent' initializer [-Werror,-Wmissing-field-initializers]
struct expr_ids ids = { 0 };
^
util/metricgroup.c:924:26: error: suggest braces around initialization of subobject [-Werror,-Wmissing-braces]
struct expr_ids ids = { 0 };
^
{}
util/metricgroup.c:924:26: error: suggest braces around initialization of subobject [-Werror,-Wmissing-braces]
struct expr_ids ids = { 0 };
^
{}
util/metricgroup.c:924:28: error: missing field 'cnt' initializer [-Werror,-Wmissing-field-initializers]
struct expr_ids ids = { 0 };
^
Signed-off-by: Jiri Olsa <[email protected]>
Reviewed-by: Kajol Jain <[email protected]>
Acked-by: Ian Rogers <[email protected]>
Cc: Alexander Shishkin <[email protected]>
Cc: Andi Kleen <[email protected]>
Cc: John Garry <[email protected]>
Cc: Michael Petlan <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Paul Clarke <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Stephane Eranian <[email protected]>
Link: http://lore.kernel.org/lkml/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>
Chun-Tse Shao