^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) perf-top(1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) ===========
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) NAME
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) ----
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) perf-top - System profiling tool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) SYNOPSIS
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) --------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) [verse]
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) 'perf top' [-e <EVENT> | --event=EVENT] [<options>]
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) DESCRIPTION
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) -----------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) This command generates and displays a performance counter profile in real time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) OPTIONS
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) -------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) -a::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) --all-cpus::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) System-wide collection. (default)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) -c <count>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) --count=<count>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) Event period to sample.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) -C <cpu-list>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) --cpu=<cpu>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) Monitor only on the list of CPUs provided. Multiple CPUs can be provided as a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) Default is to monitor all CPUS.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) -d <seconds>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) --delay=<seconds>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) Number of seconds to delay between refreshes.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) -e <event>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) --event=<event>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) Select the PMU event. Selection can be a symbolic event name
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) (use 'perf list' to list all events) or a raw PMU
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) event (eventsel+umask) in the form of rNNN where NNN is a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) hexadecimal event descriptor.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) -E <entries>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) --entries=<entries>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) Display this many functions.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) -f <count>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) --count-filter=<count>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) Only display functions with more events than this.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) --group::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) Put the counters into a counter group.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) --group-sort-idx::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) Sort the output by the event at the index n in group. If n is invalid,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) sort by the first event. It can support multiple groups with different
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) amount of events. WARNING: This should be used on grouped events.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) -F <freq>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) --freq=<freq>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) Profile at this frequency. Use 'max' to use the currently maximum
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) allowed frequency, i.e. the value in the kernel.perf_event_max_sample_rate
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) sysctl.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) -i::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) --inherit::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) Child tasks do not inherit counters.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) -k <path>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) --vmlinux=<path>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) Path to vmlinux. Required for annotation functionality.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) --ignore-vmlinux::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) Ignore vmlinux files.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) --kallsyms=<file>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) kallsyms pathname
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) -m <pages>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) --mmap-pages=<pages>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) Number of mmap data pages (must be a power of two) or size
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) specification with appended unit character - B/K/M/G. The
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) size is rounded up to have nearest pages power of two value.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) -p <pid>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) --pid=<pid>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) Profile events on existing Process ID (comma separated list).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) -t <tid>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) --tid=<tid>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) Profile events on existing thread ID (comma separated list).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) -u::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) --uid=::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) Record events in threads owned by uid. Name or number.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) -r <priority>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) --realtime=<priority>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) Collect data with this RT SCHED_FIFO priority.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) --sym-annotate=<symbol>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) Annotate this symbol.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) -K::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) --hide_kernel_symbols::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) Hide kernel symbols.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) -U::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) --hide_user_symbols::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) Hide user symbols.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) --demangle-kernel::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) Demangle kernel symbols.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) -D::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) --dump-symtab::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) Dump the symbol table used for profiling.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) -v::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) --verbose::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) Be more verbose (show counter open errors, etc).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) -z::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) --zero::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) Zero history across display updates.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) -s::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) --sort::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) Sort by key(s): pid, comm, dso, symbol, parent, srcline, weight,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) local_weight, abort, in_tx, transaction, overhead, sample, period.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) Please see description of --sort in the perf-report man page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) --fields=::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) Specify output field - multiple keys can be specified in CSV format.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) Following fields are available:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) overhead, overhead_sys, overhead_us, overhead_children, sample and period.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) Also it can contain any sort key(s).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) By default, every sort keys not specified in --field will be appended
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) automatically.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) -n::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) --show-nr-samples::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) Show a column with the number of samples.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) --show-total-period::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) Show a column with the sum of periods.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) --dsos::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) Only consider symbols in these dsos. This option will affect the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) percentage of the overhead column. See --percentage for more info.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) --comms::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) Only consider symbols in these comms. This option will affect the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) percentage of the overhead column. See --percentage for more info.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) --symbols::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) Only consider these symbols. This option will affect the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) percentage of the overhead column. See --percentage for more info.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) -M::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) --disassembler-style=:: Set disassembler style for objdump.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) --prefix=PREFIX::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) --prefix-strip=N::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) Remove first N entries from source file path names in executables
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) and add PREFIX. This allows to display source code compiled on systems
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) with different file system layout.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) --source::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) Interleave source code with assembly code. Enabled by default,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) disable with --no-source.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) --asm-raw::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) Show raw instruction encoding of assembly instructions.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) -g::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) Enables call-graph (stack chain/backtrace) recording.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) --call-graph [mode,type,min[,limit],order[,key][,branch]]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) Setup and enable call-graph (stack chain/backtrace) recording,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) implies -g. See `--call-graph` section in perf-record and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) perf-report man pages for details.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) --children::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) Accumulate callchain of children to parent entry so that then can
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) show up in the output. The output will have a new "Children" column
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) and will be sorted on the data. It requires -g/--call-graph option
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) enabled. See the `overhead calculation' section for more details.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) Enabled by default, disable with --no-children.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) --max-stack::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) Set the stack depth limit when parsing the callchain, anything
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) beyond the specified depth will be ignored. This is a trade-off
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) between information loss and faster processing especially for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) workloads that can have a very long callchain stack.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) Default: /proc/sys/kernel/perf_event_max_stack when present, 127 otherwise.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) --ignore-callees=<regex>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) Ignore callees of the function(s) matching the given regex.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) This has the effect of collecting the callers of each such
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) function into one place in the call-graph tree.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) --percent-limit::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) Do not show entries which have an overhead under that percent.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) (Default: 0).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) --percentage::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) Determine how to display the overhead percentage of filtered entries.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) Filters can be applied by --comms, --dsos and/or --symbols options and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) Zoom operations on the TUI (thread, dso, etc).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) "relative" means it's relative to filtered entries only so that the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) sum of shown entries will be always 100%. "absolute" means it retains
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) the original value before and after the filter is applied.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) -w::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) --column-widths=<width[,width...]>::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) Force each column width to the provided list, for large terminal
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) readability. 0 means no limit (default behavior).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) --proc-map-timeout::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) When processing pre-existing threads /proc/XXX/mmap, it may take
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) a long time, because the file may be huge. A time out is needed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) in such cases.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) This option sets the time out limit. The default value is 500 ms.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) -b::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) --branch-any::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) Enable taken branch stack sampling. Any type of taken branch may be sampled.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) This is a shortcut for --branch-filter any. See --branch-filter for more infos.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) -j::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) --branch-filter::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) Enable taken branch stack sampling. Each sample captures a series of consecutive
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) taken branches. The number of branches captured with each sample depends on the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) underlying hardware, the type of branches of interest, and the executed code.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) It is possible to select the types of branches captured by enabling filters.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) For a full list of modifiers please see the perf record manpage.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) The option requires at least one branch type among any, any_call, any_ret, ind_call, cond.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) The privilege levels may be omitted, in which case, the privilege levels of the associated
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) event are applied to the branch filter. Both kernel (k) and hypervisor (hv) privilege
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) levels are subject to permissions. When sampling on multiple events, branch stack sampling
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249) is enabled for all the sampling events. The sampled branch type is the same for all events.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) The various filters must be specified as a comma separated list: --branch-filter any_ret,u,k
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) Note that this feature may not be available on all processors.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) --raw-trace::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) When displaying traceevent output, do not use print fmt or plugins.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) --hierarchy::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) Enable hierarchy output.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) --overwrite::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) Enable this to use just the most recent records, which helps in high core count
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) machines such as Knights Landing/Mill, but right now is disabled by default as
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) the pausing used in this technique is leading to loss of metadata events such
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) as PERF_RECORD_MMAP which makes 'perf top' unable to resolve samples, leading
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) to lots of unknown samples appearing on the UI. Enable this if you are in such
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) machines and profiling a workload that doesn't creates short lived threads and/or
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) doesn't uses many executable mmap operations. Work is being planed to solve
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) this situation, till then, this will remain disabled by default.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) --force::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) Don't do ownership validation.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) --num-thread-synthesize::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) The number of threads to run when synthesizing events for existing processes.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) By default, the number of threads equals to the number of online CPUs.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) --namespaces::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) Record events of type PERF_RECORD_NAMESPACES and display it with the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) 'cgroup_id' sort key.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) --all-cgroups::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) Record events of type PERF_RECORD_CGROUP and display it with the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) 'cgroup' sort key.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) --switch-on EVENT_NAME::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) Only consider events after this event is found.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) E.g.:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) Find out where broadcast packets are handled
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) perf probe -L icmp_rcv
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) Insert a probe there:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) perf probe icmp_rcv:59
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) Start perf top and ask it to only consider the cycles events when a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) broadcast packet arrives This will show a menu with two entries and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) will start counting when a broadcast packet arrives:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) perf top -e cycles,probe:icmp_rcv --switch-on=probe:icmp_rcv
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) Alternatively one can ask for --group and then two overhead columns
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) will appear, the first for cycles and the second for the switch-on event.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) perf top --group -e cycles,probe:icmp_rcv --switch-on=probe:icmp_rcv
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) This may be interesting to measure a workload only after some initialization
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) phase is over, i.e. insert a perf probe at that point and use the above
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) examples replacing probe:icmp_rcv with the just-after-init probe.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) --switch-off EVENT_NAME::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) Stop considering events after this event is found.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) --show-on-off-events::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) Show the --switch-on/off events too. This has no effect in 'perf top' now
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) but probably we'll make the default not to show the switch-on/off events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) on the --group mode and if there is only one event besides the off/on ones,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) go straight to the histogram browser, just like 'perf top' with no events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) explicitely specified does.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) --stitch-lbr::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) Show callgraph with stitched LBRs, which may have more complete
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) callgraph. The option must be used with --call-graph lbr recording.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) Disabled by default. In common cases with call stack overflows,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) it can recreate better call stacks than the default lbr call stack
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) output. But this approach is not full proof. There can be cases
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) where it creates incorrect call stacks from incorrect matches.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) The known limitations include exception handing such as
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) setjmp/longjmp will have calls/returns not match.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) ifdef::HAVE_LIBPFM[]
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) --pfm-events events::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) Select a PMU event using libpfm4 syntax (see http://perfmon2.sf.net)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) including support for event filters. For example '--pfm-events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) inst_retired:any_p:u:c=1:i'. More than one event can be passed to the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) option using the comma separator. Hardware events and generic hardware
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) events cannot be mixed together. The latter must be used with the -e
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) option. The -e option and this one can be mixed and matched. Events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) can be grouped using the {} notation.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) endif::HAVE_LIBPFM[]
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) INTERACTIVE PROMPTING KEYS
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) --------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) [d]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) Display refresh delay.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349) [e]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) Number of entries to display.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) [E]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) Event to display when multiple counters are active.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) [f]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) Profile display filter (>= hit count).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) [F]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) Annotation display filter (>= % of total).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) [s]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) Annotate symbol.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) [S]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) Stop annotation, return to full profile display.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) [K]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) Hide kernel symbols.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) [U]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) Hide user symbols.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) [z]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) Toggle event count zeroing across display updates.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) [qQ]::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) Quit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) Pressing any unmapped key displays a menu, and prompts for input.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) include::callchain-overhead-calculation.txt[]
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) SEE ALSO
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) --------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) linkperf:perf-stat[1], linkperf:perf-list[1], linkperf:perf-report[1]
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags