path: root/trace_profile/perf.html

               
                                        
<!DOCTYPE HTML>
<html lang="en" class="light" dir="ltr">
    <head>
        <!-- Book generated using mdBook -->
        <meta charset="UTF-8">
        <title>perf - Notes</title>


        <!-- Custom HTML head -->
        
        <meta name="description" content="">
        <meta name="viewport" content="width=device-width, initial-scale=1">
        <meta name="theme-color" content="#ffffff">

        <link rel="icon" href="../favicon.svg">
        <link rel="shortcut icon" href="../favicon.png">
        <link rel="stylesheet" href="../css/variables.css">
        <link rel="stylesheet" href="../css/general.css">
        <link rel="stylesheet" href="../css/chrome.css">
        <link rel="stylesheet" href="../css/print.css" media="print">

        <!-- Fonts -->
        <link rel="stylesheet" href="../FontAwesome/css/font-awesome.css">
        <link rel="stylesheet" href="../fonts/fonts.css">

        <!-- Highlight.js Stylesheets -->
        <link rel="stylesheet" href="../highlight.css">
        <link rel="stylesheet" href="../tomorrow-night.css">
        <link rel="stylesheet" href="../ayu-highlight.css">

        <!-- Custom theme stylesheets -->

    </head>
    <body class="sidebar-visible no-js">
    <div id="body-container">
        <!-- Provide site root to javascript -->
        <script>
            var path_to_root = "../";
            var default_theme = window.matchMedia("(prefers-color-scheme: dark)").matches ? "navy" : "light";
        </script>

        <!-- Work around some values being stored in localStorage wrapped in quotes -->
        <script>
            try {
                var theme = localStorage.getItem('mdbook-theme');
                var sidebar = localStorage.getItem('mdbook-sidebar');

                if (theme.startsWith('"') && theme.endsWith('"')) {
                    localStorage.setItem('mdbook-theme', theme.slice(1, theme.length - 1));
                }

                if (sidebar.startsWith('"') && sidebar.endsWith('"')) {
                    localStorage.setItem('mdbook-sidebar', sidebar.slice(1, sidebar.length - 1));
                }
            } catch (e) { }
        </script>

        <!-- Set the theme before any content is loaded, prevents flash -->
        <script>
            var theme;
            try { theme = localStorage.getItem('mdbook-theme'); } catch(e) { }
            if (theme === null || theme === undefined) { theme = default_theme; }
            var html = document.querySelector('html');
            html.classList.remove('light')
            html.classList.add(theme);
            var body = document.querySelector('body');
            body.classList.remove('no-js')
            body.classList.add('js');
        </script>

        <input type="checkbox" id="sidebar-toggle-anchor" class="hidden">

        <!-- Hide / unhide sidebar before it is displayed -->
        <script>
            var body = document.querySelector('body');
            var sidebar = null;
            var sidebar_toggle = document.getElementById("sidebar-toggle-anchor");
            if (document.body.clientWidth >= 1080) {
                try { sidebar = localStorage.getItem('mdbook-sidebar'); } catch(e) { }
                sidebar = sidebar || 'visible';
            } else {
                sidebar = 'hidden';
            }
            sidebar_toggle.checked = sidebar === 'visible';
            body.classList.remove('sidebar-visible');
            body.classList.add("sidebar-" + sidebar);
        </script>

        <nav id="sidebar" class="sidebar" aria-label="Table of contents">
            <div class="sidebar-scrollbox">
                <ol class="chapter"><li class="chapter-item expanded affix "><a href="../intro.html">Introduction</a></li><li class="chapter-item expanded "><a href="../tools/index.html"><strong aria-hidden="true">1.</strong> Tools</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../tools/zsh.html"><strong aria-hidden="true">1.1.</strong> zsh</a></li><li class="chapter-item expanded "><a href="../tools/bash.html"><strong aria-hidden="true">1.2.</strong> bash</a></li><li class="chapter-item expanded "><a href="../tools/fish.html"><strong aria-hidden="true">1.3.</strong> fish</a></li><li class="chapter-item expanded "><a href="../tools/tmux.html"><strong aria-hidden="true">1.4.</strong> tmux</a></li><li class="chapter-item expanded "><a href="../tools/git.html"><strong aria-hidden="true">1.5.</strong> git</a></li><li class="chapter-item expanded "><a href="../tools/awk.html"><strong aria-hidden="true">1.6.</strong> awk</a></li><li class="chapter-item expanded "><a href="../tools/emacs.html"><strong aria-hidden="true">1.7.</strong> emacs</a></li><li class="chapter-item expanded "><a href="../tools/gpg.html"><strong aria-hidden="true">1.8.</strong> gpg</a></li><li class="chapter-item expanded "><a href="../tools/gdb.html"><strong aria-hidden="true">1.9.</strong> gdb</a></li><li class="chapter-item expanded "><a href="../tools/gdbserver.html"><strong aria-hidden="true">1.10.</strong> gdbserver</a></li><li class="chapter-item expanded "><a href="../tools/radare2.html"><strong aria-hidden="true">1.11.</strong> radare2</a></li><li class="chapter-item expanded "><a href="../tools/qemu.html"><strong aria-hidden="true">1.12.</strong> qemu</a></li><li class="chapter-item expanded "><a href="../tools/pacman.html"><strong aria-hidden="true">1.13.</strong> pacman</a></li><li class="chapter-item expanded "><a href="../tools/dot.html"><strong aria-hidden="true">1.14.</strong> dot</a></li><li class="chapter-item expanded "><a href="../tools/ffmpeg.html"><strong aria-hidden="true">1.15.</strong> ffmpeg</a></li><li class="chapter-item expanded "><a href="../tools/column.html"><strong aria-hidden="true">1.16.</strong> column</a></li><li class="chapter-item expanded "><a href="../tools/sort.html"><strong aria-hidden="true">1.17.</strong> sort</a></li><li class="chapter-item expanded "><a href="../tools/sed.html"><strong aria-hidden="true">1.18.</strong> sed</a></li></ol></li><li class="chapter-item expanded "><a href="../monitor/index.html"><strong aria-hidden="true">2.</strong> Resource analysis & monitor</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../monitor/lsof.html"><strong aria-hidden="true">2.1.</strong> lsof</a></li><li class="chapter-item expanded "><a href="../monitor/ss.html"><strong aria-hidden="true">2.2.</strong> ss</a></li><li class="chapter-item expanded "><a href="../monitor/pidstat.html"><strong aria-hidden="true">2.3.</strong> pidstat</a></li><li class="chapter-item expanded "><a href="../monitor/pgrep.html"><strong aria-hidden="true">2.4.</strong> pgrep</a></li><li class="chapter-item expanded "><a href="../monitor/ps.html"><strong aria-hidden="true">2.5.</strong> ps</a></li><li class="chapter-item expanded "><a href="../monitor/pmap.html"><strong aria-hidden="true">2.6.</strong> pmap</a></li><li class="chapter-item expanded "><a href="../monitor/pstack.html"><strong aria-hidden="true">2.7.</strong> pstack</a></li></ol></li><li class="chapter-item expanded "><a href="../trace_profile/index.html"><strong aria-hidden="true">3.</strong> Trace and Profile</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../trace_profile/time.html"><strong aria-hidden="true">3.1.</strong> time</a></li><li class="chapter-item expanded "><a href="../trace_profile/strace.html"><strong aria-hidden="true">3.2.</strong> strace</a></li><li class="chapter-item expanded "><a href="../trace_profile/ltrace.html"><strong aria-hidden="true">3.3.</strong> ltrace</a></li><li class="chapter-item expanded "><a href="../trace_profile/perf.html" class="active"><strong aria-hidden="true">3.4.</strong> perf</a></li><li class="chapter-item expanded "><a href="../trace_profile/oprofile.html"><strong aria-hidden="true">3.5.</strong> OProfile</a></li><li class="chapter-item expanded "><a href="../trace_profile/callgrind.html"><strong aria-hidden="true">3.6.</strong> callgrind</a></li></ol></li><li class="chapter-item expanded "><a href="../binary/index.html"><strong aria-hidden="true">4.</strong> Binary</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../binary/od.html"><strong aria-hidden="true">4.1.</strong> od</a></li><li class="chapter-item expanded "><a href="../binary/xxd.html"><strong aria-hidden="true">4.2.</strong> xxd</a></li><li class="chapter-item expanded "><a href="../binary/readelf.html"><strong aria-hidden="true">4.3.</strong> readelf</a></li><li class="chapter-item expanded "><a href="../binary/objdump.html"><strong aria-hidden="true">4.4.</strong> objdump</a></li><li class="chapter-item expanded "><a href="../binary/nm.html"><strong aria-hidden="true">4.5.</strong> nm</a></li></ol></li><li class="chapter-item expanded "><a href="../development/index.html"><strong aria-hidden="true">5.</strong> Development</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../development/c++filt.html"><strong aria-hidden="true">5.1.</strong> c++filt</a></li><li class="chapter-item expanded "><a href="../development/c++.html"><strong aria-hidden="true">5.2.</strong> c++</a></li><li class="chapter-item expanded "><a href="../development/glibc.html"><strong aria-hidden="true">5.3.</strong> glibc</a></li><li class="chapter-item expanded "><a href="../development/gcc.html"><strong aria-hidden="true">5.4.</strong> gcc</a></li><li class="chapter-item expanded "><a href="../development/cmake.html"><strong aria-hidden="true">5.5.</strong> cmake</a></li><li class="chapter-item expanded "><a href="../development/make.html"><strong aria-hidden="true">5.6.</strong> make</a></li><li class="chapter-item expanded "><a href="../development/ld.so.html"><strong aria-hidden="true">5.7.</strong> ld.so</a></li><li class="chapter-item expanded "><a href="../development/symbolver.html"><strong aria-hidden="true">5.8.</strong> symbol versioning</a></li><li class="chapter-item expanded "><a href="../development/python.html"><strong aria-hidden="true">5.9.</strong> python</a></li><li class="chapter-item expanded "><a href="../development/gcov.html"><strong aria-hidden="true">5.10.</strong> gcov</a></li><li class="chapter-item expanded "><a href="../development/pgo.html"><strong aria-hidden="true">5.11.</strong> pgo</a></li></ol></li><li class="chapter-item expanded "><a href="../linux/index.html"><strong aria-hidden="true">6.</strong> Linux</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../linux/systemd.html"><strong aria-hidden="true">6.1.</strong> systemd</a></li><li class="chapter-item expanded "><a href="../linux/coredump.html"><strong aria-hidden="true">6.2.</strong> coredump</a></li><li class="chapter-item expanded "><a href="../linux/ptrace_scope.html"><strong aria-hidden="true">6.3.</strong> ptrace_scope</a></li><li class="chapter-item expanded "><a href="../linux/cryptsetup.html"><strong aria-hidden="true">6.4.</strong> cryptsetup</a></li><li class="chapter-item expanded "><a href="../linux/swap.html"><strong aria-hidden="true">6.5.</strong> swap</a></li><li class="chapter-item expanded "><a href="../linux/input.html"><strong aria-hidden="true">6.6.</strong> input</a></li><li class="chapter-item expanded "><a href="../linux/acl.html"><strong aria-hidden="true">6.7.</strong> acl</a></li><li class="chapter-item expanded "><a href="../linux/zfs.html"><strong aria-hidden="true">6.8.</strong> zfs</a></li></ol></li><li class="chapter-item expanded "><a href="../network/index.html"><strong aria-hidden="true">7.</strong> Network</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../network/tcpdump.html"><strong aria-hidden="true">7.1.</strong> tcpdump</a></li><li class="chapter-item expanded "><a href="../network/tshark.html"><strong aria-hidden="true">7.2.</strong> tshark</a></li><li class="chapter-item expanded "><a href="../network/firewall-cmd.html"><strong aria-hidden="true">7.3.</strong> firewall-cmd</a></li><li class="chapter-item expanded "><a href="../network/nftables.html"><strong aria-hidden="true">7.4.</strong> nftables</a></li></ol></li><li class="chapter-item expanded "><a href="../web/index.html"><strong aria-hidden="true">8.</strong> Web</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../web/html.html"><strong aria-hidden="true">8.1.</strong> html</a></li><li class="chapter-item expanded "><a href="../web/css.html"><strong aria-hidden="true">8.2.</strong> css</a></li><li class="chapter-item expanded "><a href="../web/chartjs.html"><strong aria-hidden="true">8.3.</strong> chartjs</a></li></ol></li><li class="chapter-item expanded "><a href="../arch/index.html"><strong aria-hidden="true">9.</strong> Arch</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../arch/x86_64.html"><strong aria-hidden="true">9.1.</strong> x86_64</a></li><li class="chapter-item expanded "><a href="../arch/arm64.html"><strong aria-hidden="true">9.2.</strong> arm64</a></li><li class="chapter-item expanded "><a href="../arch/armv7.html"><strong aria-hidden="true">9.3.</strong> armv7</a></li><li class="chapter-item expanded "><a href="../arch/riscv.html"><strong aria-hidden="true">9.4.</strong> riscv</a></li></ol></li></ol>
            </div>
            <div id="sidebar-resize-handle" class="sidebar-resize-handle">
                <div class="sidebar-resize-indicator"></div>
            </div>
        </nav>

        <!-- Track and set sidebar scroll position -->
        <script>
            var sidebarScrollbox = document.querySelector('#sidebar .sidebar-scrollbox');
            sidebarScrollbox.addEventListener('click', function(e) {
                if (e.target.tagName === 'A') {
                    sessionStorage.setItem('sidebar-scroll', sidebarScrollbox.scrollTop);
                }
            }, { passive: true });
            var sidebarScrollTop = sessionStorage.getItem('sidebar-scroll');
            sessionStorage.removeItem('sidebar-scroll');
            if (sidebarScrollTop) {
                // preserve sidebar scroll position when navigating via links within sidebar
                sidebarScrollbox.scrollTop = sidebarScrollTop;
            } else {
                // scroll sidebar to current active section when navigating via "next/previous chapter" buttons
                var activeSection = document.querySelector('#sidebar .active');
                if (activeSection) {
                    activeSection.scrollIntoView({ block: 'center' });
                }
            }
        </script>

        <div id="page-wrapper" class="page-wrapper">

            <div class="page">
                                <div id="menu-bar-hover-placeholder"></div>
                <div id="menu-bar" class="menu-bar sticky">
                    <div class="left-buttons">
                        <label id="sidebar-toggle" class="icon-button" for="sidebar-toggle-anchor" title="Toggle Table of Contents" aria-label="Toggle Table of Contents" aria-controls="sidebar">
                            <i class="fa fa-bars"></i>
                        </label>
                        <button id="theme-toggle" class="icon-button" type="button" title="Change theme" aria-label="Change theme" aria-haspopup="true" aria-expanded="false" aria-controls="theme-list">
                            <i class="fa fa-paint-brush"></i>
                        </button>
                        <ul id="theme-list" class="theme-popup" aria-label="Themes" role="menu">
                            <li role="none"><button role="menuitem" class="theme" id="light">Light</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="rust">Rust</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="coal">Coal</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="navy">Navy</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="ayu">Ayu</button></li>
                        </ul>
                        <button id="search-toggle" class="icon-button" type="button" title="Search. (Shortkey: s)" aria-label="Toggle Searchbar" aria-expanded="false" aria-keyshortcuts="S" aria-controls="searchbar">
                            <i class="fa fa-search"></i>
                        </button>
                    </div>

                    <h1 class="menu-title">Notes</h1>

                    <div class="right-buttons">
                        <a href="../print.html" title="Print this book" aria-label="Print this book">
                            <i id="print-button" class="fa fa-print"></i>
                        </a>
                        <a href="https://github.com/johannst/notes" title="Git repository" aria-label="Git repository">
                            <i id="git-repository-button" class="fa fa-github"></i>
                        </a>

                    </div>
                </div>

                <div id="search-wrapper" class="hidden">
                    <form id="searchbar-outer" class="searchbar-outer">
                        <input type="search" id="searchbar" name="searchbar" placeholder="Search this book ..." aria-controls="searchresults-outer" aria-describedby="searchresults-header">
                    </form>
                    <div id="searchresults-outer" class="searchresults-outer hidden">
                        <div id="searchresults-header" class="searchresults-header"></div>
                        <ul id="searchresults">
                        </ul>
                    </div>
                </div>

                <!-- Apply ARIA attributes after the sidebar and the sidebar toggle button are added to the DOM -->
                <script>
                    document.getElementById('sidebar-toggle').setAttribute('aria-expanded', sidebar === 'visible');
                    document.getElementById('sidebar').setAttribute('aria-hidden', sidebar !== 'visible');
                    Array.from(document.querySelectorAll('#sidebar a')).forEach(function(link) {
                        link.setAttribute('tabIndex', sidebar === 'visible' ? 0 : -1);
                    });
                </script>

                <div id="content" class="content">
                    <main>
                        <h1 id="perf1"><a class="header" href="#perf1">perf(1)</a></h1>
<pre><code>perf list     show supported hw/sw events &amp; metrics
  -v ........ print longer event descriptions
  --details . print information on the perf event names
              and expressions used internally by events

perf stat
  -p &lt;pid&gt; ..... show stats for running process
  -I &lt;ms&gt; ...... show stats periodically over interval &lt;ms&gt;
  -e &lt;ev&gt; ...... select event(s)
  -M &lt;met&gt; ..... print metric(s), this adds the metric events
  --all-user ... configure all selected events for user space
  --all-kernel . configure all selected events for kernel space

perf top
  -p &lt;pid&gt; .. show stats for running process
  -F &lt;hz&gt; ... sampling frequency
  -K ........ hide kernel threads

perf record
  -p &lt;pid&gt; ............... record stats for running process
  -F &lt;hz&gt; ................ sampling frequency
  --call-graph &lt;method&gt; .. [fp, dwarf, lbr] method how to caputre backtrace
                           fp   : use frame-pointer, need to compile with
                                  -fno-omit-frame-pointer
                           dwarf: use .cfi debug information
                           lbr  : use hardware last branch record facility
  -g ..................... short-hand for --call-graph fp
  -e &lt;ev&gt; ................ select event(s)
  --all-user ... configure all selected events for user space
  --all-kernel . configure all selected events for kernel space

perf report
  -n .................... annotate symbols with nr of samples
  --stdio ............... report to stdio, if not presen tui mode
  -g graph,0.5,callee ... show callee based call chains with value &gt;0.5
</code></pre>
<pre><code>Useful &lt;ev&gt;:
  page-faults
  minor-faults
  major-faults
  cpu-cycles`
  task-clock
</code></pre>
<h2 id="select-specific-events"><a class="header" href="#select-specific-events">Select specific events</a></h2>
<p>Events to sample are specified with the <code>-e</code> option, either pass a comma
separated list or pass <code>-e</code> multiple times.</p>
<p>Events are specified in the following form <code>name[:modifier]</code>. The list and
description of the <code>modifier</code> can be found in the
<a href="https://man7.org/linux/man-pages/man1/perf-list.1.html"><code>perf-list(1)</code></a> manpage under <code>EVENT MODIFIERS</code>.</p>
<pre><code class="language-sh"># L1 i$ misses in user space
# L2 i$ stats in user/kernel space mixed
# Sample specified events.
perf stat -e L1-icache-load-misses:u \
          -e l2_rqsts.all_code_rd:uk,l2_rqsts.code_rd_hit:k,l2_rqsts.code_rd_miss:k \
          -- stress -c 2
</code></pre>
<p>The <code>--all-user</code> and <code>--all-kernel</code> options append a <code>:u</code> and <code>:k</code> modifier to
all specified events. Therefore the following two command lines are equivalent.</p>
<pre><code class="language-sh"># 1)
perf stat -e cycles:u,instructions:u -- ls

# 2)
perf stat --all-user -e cycles,instructions -- ls
</code></pre>
<h3 id="raw-events"><a class="header" href="#raw-events">Raw events</a></h3>
<p>In case perf does not provide a <em>symbolic</em> name for an event, the event can be
specified in a <em>raw</em> form as <code>r + UMask + EventCode</code>.</p>
<p>The following is an example for the <a href="https://github.com/intel/perfmon/blob/09c155f72e1b8f14b09aea346a35467a03a7d62b/SNB/events/sandybridge_core.json#L808">L2_RQSTS.CODE_RD_HIT</a> event
with <code>EventCode=0x24</code> and <code>UMask=0x10</code> on my laptop with a <code>sandybridge</code> uarch.</p>
<pre><code class="language-sh">perf stat -e l2_rqsts.code_rd_hit -e r1024 -- ls
# Performance counter stats for 'ls':
#
#       33.942      l2_rqsts.code_rd_hit
#       33.942      r1024
</code></pre>
<h3 id="find-raw-performance-counter-events-intel"><a class="header" href="#find-raw-performance-counter-events-intel">Find raw performance counter events (intel)</a></h3>
<p>The <a href="https://github.com/intel/perfmon"><code>intel/perfmon</code></a> repository provides a performance event
databases for the different intel uarchs.</p>
<p>The table in <a href="https://github.com/intel/perfmon/blob/main/mapfile.csv"><code>mapfile.csv</code></a> can be used to lookup the
corresponding uarch, just grab the family model from the procfs.</p>
<pre><code class="language-sh"> cat /proc/cpuinfo | awk '/^vendor_id/  { V=$3 }
                          /^cpu family/ { F=$4 }
                          /^model\s*:/  { printf &quot;%s-%d-%x\n&quot;,V,F,$3 }'
</code></pre>
<blockquote>
<p>The table in <a href="https://github.com/intel/perfmon/tree/main#performance-monitoring-events">performance monitoring events</a> describes how
events are sorted into the different files.</p>
</blockquote>
<h3 id="raw-events-for-perfs-own-symbolic-names"><a class="header" href="#raw-events-for-perfs-own-symbolic-names">Raw events for perfs own symbolic names</a></h3>
<p>Perf also defines some own <em>symbolic</em> names for events. An example is the
<code>cache-references</code> event. The <a href="https://man7.org/linux/man-pages/man2/perf_event_open.2.html"><code>perf_event_open(2)</code></a> manpage
gives the following description.</p>
<pre><code class="language-man">perf_event_open(2)

PERF_COUNT_HW_CACHE_REFERENCES
    Cache accesses.  Usually this indicates Last Level Cache accesses but this
    may vary depending on your CPU.  This may include prefetches and coherency
    messages; again this depends on the design of your CPU.
</code></pre>
<p>The <code>sysfs</code> can be consulted to get the concrete performance counter on the
given system.</p>
<pre><code class="language-sh">cat /sys/devices/cpu/events/cache-misses
# event=0x2e,umask=0x41
</code></pre>
<h2 id="flamegraph"><a class="header" href="#flamegraph"><a href="https://github.com/brendangregg/FlameGraph"><code>Flamegraph</code></a></a></h2>
<h3 id="flamegraph-with-single-event-trace"><a class="header" href="#flamegraph-with-single-event-trace">Flamegraph with single event trace</a></h3>
<pre><code>perf record -g -e cpu-cycles -p &lt;pid&gt;
perf script | FlameGraph/stackcollapse-perf.pl | FlameGraph/flamegraph.pl &gt; cycles-flamegraph.svg
</code></pre>
<h3 id="flamegraph-with-multiple-event-traces"><a class="header" href="#flamegraph-with-multiple-event-traces">Flamegraph with multiple event traces</a></h3>
<pre><code class="language-sh">perf record -g -e cpu-cycles,page-faults -p &lt;pid&gt;
perf script --per-event-dump
# fold &amp; generate as above
</code></pre>
<h2 id="examples"><a class="header" href="#examples">Examples</a></h2>
<h3 id="determine-theoretical-max-instructions-per-cycle"><a class="header" href="#determine-theoretical-max-instructions-per-cycle">Determine theoretical max instructions per cycle</a></h3>
<pre><code class="language-c">#define NOP4        &quot;nop\nnop\nnop\nnop\n&quot;
#define NOP32       NOP4   NOP4   NOP4   NOP4   NOP4   NOP4   NOP4   NOP4
#define NOP256      NOP32  NOP32  NOP32  NOP32  NOP32  NOP32  NOP32  NOP32
#define NOP2048     NOP256 NOP256 NOP256 NOP256 NOP256 NOP256 NOP256 NOP256

int main() {
  for (unsigned i = 0; i &lt; 2000000; ++i) {
    asm volatile(NOP2048);
  }
}
</code></pre>
<pre><code class="language-sh">perf stat -e cycles,instructions ./noploop
# Performance counter stats for './noploop':
#
#     1.031.075.940      cycles
#     4.103.534.341      instructions       #    3,98  insn per cycle
</code></pre>
<h3 id="caller-vs-callee-callstacks"><a class="header" href="#caller-vs-callee-callstacks">Caller vs callee callstacks</a></h3>
<p>The following gives an example for a scenario where we have the following calls</p>
<ul>
<li><code>main -&gt; do_foo() -&gt; do_work()</code></li>
<li><code>main -&gt; do_bar() -&gt; do_work()</code></li>
</ul>
<pre><code class="language-sh">perf report --stdio -g graph,callee

# Children      Self  Command  Shared Object         Symbols
# ........  ........  .......  ....................  .................
#
#  49.71%    49.66%   bench    bench                 [.] do_work
#          |
#           --49.66%--_start                &lt;- callstack bottom
#                     __libc_start_main
#                     0x7ff366c62ccf
#                     main
#                     |
#                     |--25.13%--do_bar
#                     |          do_work    &lt;- callstack top
#                     |
#                      --24.53%--do_foo
#                                do_work

perf report --stdio -g graph,callee

# Children      Self  Command  Shared Object         Symbols
# ........  ........  .......  ....................  .................
#
#  49.71%    49.66%   bench    bench                 [.] do_work
#          |
#          ---do_work                       &lt;- callstack top
#             |
#             |--25.15%--do_bar
#             |          main
#             |          0x7ff366c62ccf
#             |          __libc_start_main
#             |          _start             &lt;- callstack bottom
#             |
#              --24.55%--do_foo
#                        main
#                        0x7ff366c62ccf
#                        __libc_start_main
#                        _start             &lt;- callstack bottom
</code></pre>
<h2 id="references"><a class="header" href="#references">References</a></h2>
<ul>
<li><a href="https://github.com/intel/perfmon">intel/perfmon</a> - intel PMU event database per uarch</li>
<li><a href="https://perfmon-events.intel.com/">intel/perfmon-html</a> - a html rendered version of the PMU events with search</li>
<li><a href="https://github.com/intel/perfmon/blob/main/mapfile.csv">intel/perfmon/mapfile.csv</a> - processor family to uarch mapping</li>
<li><a href="https://github.com/torvalds/linux/tree/master/tools/perf/pmu-events/arch/x86">linux/perf/events</a> - x86 PMU events known to perf tools</li>
<li><a href="https://github.com/torvalds/linux/blob/master/arch/x86/events/intel/core.c">linux/arch/events</a> - x86 PMU events linux kernel</li>
<li><a href="https://en.wikichip.org/wiki/WikiChip">wikichip</a> - computer architecture wiki</li>
<li><a href="https://man7.org/linux/man-pages/man1/perf-list.1.html">perf-list(1)</a> - manpage</li>
<li><a href="https://man7.org/linux/man-pages/man2/perf_event_open.2.html">perf_event_open(2)</a> - manpage</li>
</ul>

                    </main>

                    <nav class="nav-wrapper" aria-label="Page navigation">
                        <!-- Mobile navigation buttons -->
                            <a rel="prev" href="../trace_profile/ltrace.html" class="mobile-nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
                                <i class="fa fa-angle-left"></i>
                            </a>

                            <a rel="next prefetch" href="../trace_profile/oprofile.html" class="mobile-nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
                                <i class="fa fa-angle-right"></i>
                            </a>

                        <div style="clear: both"></div>
                    </nav>
                </div>
            </div>

            <nav class="nav-wide-wrapper" aria-label="Page navigation">
                    <a rel="prev" href="../trace_profile/ltrace.html" class="nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
                        <i class="fa fa-angle-left"></i>
                    </a>

                    <a rel="next prefetch" href="../trace_profile/oprofile.html" class="nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
                        <i class="fa fa-angle-right"></i>
                    </a>
            </nav>

        </div>


        <script>
            window.playground_copyable = true;
        </script>


        <script src="../elasticlunr.min.js"></script>
        <script src="../mark.min.js"></script>
        <script src="../searcher.js"></script>

        <script src="../clipboard.min.js"></script>
        <script src="../highlight.js"></script>
        <script src="../book.js"></script>

        <!-- Custom JS scripts -->


    </div>
    </body>
</html>
<!DOCTYPE HTML>
<html lang="en" class="light" dir="ltr">
    <head>
        <!-- Book generated using mdBook -->
        <meta charset="UTF-8">
        <title>perf - Notes</title>


        <!-- Custom HTML head -->
        
        <meta name="description" content="">
        <meta name="viewport" content="width=device-width, initial-scale=1">
        <meta name="theme-color" content="#ffffff">

        <link rel="icon" href="../favicon.svg">
        <link rel="shortcut icon" href="../favicon.png">
        <link rel="stylesheet" href="../css/variables.css">
        <link rel="stylesheet" href="../css/general.css">
        <link rel="stylesheet" href="../css/chrome.css">
        <link rel="stylesheet" href="../css/print.css" media="print">

        <!-- Fonts -->
        <link rel="stylesheet" href="../FontAwesome/css/font-awesome.css">
        <link rel="stylesheet" href="../fonts/fonts.css">

        <!-- Highlight.js Stylesheets -->
        <link rel="stylesheet" href="../highlight.css">
        <link rel="stylesheet" href="../tomorrow-night.css">
        <link rel="stylesheet" href="../ayu-highlight.css">

        <!-- Custom theme stylesheets -->

    </head>
    <body class="sidebar-visible no-js">
    <div id="body-container">
        <!-- Provide site root to javascript -->
        <script>
            var path_to_root = "../";
            var default_theme = window.matchMedia("(prefers-color-scheme: dark)").matches ? "navy" : "light";
        </script>

        <!-- Work around some values being stored in localStorage wrapped in quotes -->
        <script>
            try {
                var theme = localStorage.getItem('mdbook-theme');
                var sidebar = localStorage.getItem('mdbook-sidebar');

                if (theme.startsWith('"') && theme.endsWith('"')) {
                    localStorage.setItem('mdbook-theme', theme.slice(1, theme.length - 1));
                }

                if (sidebar.startsWith('"') && sidebar.endsWith('"')) {
                    localStorage.setItem('mdbook-sidebar', sidebar.slice(1, sidebar.length - 1));
                }
            } catch (e) { }
        </script>

        <!-- Set the theme before any content is loaded, prevents flash -->
        <script>
            var theme;
            try { theme = localStorage.getItem('mdbook-theme'); } catch(e) { }
            if (theme === null || theme === undefined) { theme = default_theme; }
            var html = document.querySelector('html');
            html.classList.remove('light')
            html.classList.add(theme);
            var body = document.querySelector('body');
            body.classList.remove('no-js')
            body.classList.add('js');
        </script>

        <input type="checkbox" id="sidebar-toggle-anchor" class="hidden">

        <!-- Hide / unhide sidebar before it is displayed -->
        <script>
            var body = document.querySelector('body');
            var sidebar = null;
            var sidebar_toggle = document.getElementById("sidebar-toggle-anchor");
            if (document.body.clientWidth >= 1080) {
                try { sidebar = localStorage.getItem('mdbook-sidebar'); } catch(e) { }
                sidebar = sidebar || 'visible';
            } else {
                sidebar = 'hidden';
            }
            sidebar_toggle.checked = sidebar === 'visible';
            body.classList.remove('sidebar-visible');
            body.classList.add("sidebar-" + sidebar);
        </script>

        <nav id="sidebar" class="sidebar" aria-label="Table of contents">
            <div class="sidebar-scrollbox">
                <ol class="chapter"><li class="chapter-item expanded affix "><a href="../intro.html">Introduction</a></li><li class="chapter-item expanded "><a href="../tools/index.html"><strong aria-hidden="true">1.</strong> Tools</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../tools/zsh.html"><strong aria-hidden="true">1.1.</strong> zsh</a></li><li class="chapter-item expanded "><a href="../tools/bash.html"><strong aria-hidden="true">1.2.</strong> bash</a></li><li class="chapter-item expanded "><a href="../tools/fish.html"><strong aria-hidden="true">1.3.</strong> fish</a></li><li class="chapter-item expanded "><a href="../tools/tmux.html"><strong aria-hidden="true">1.4.</strong> tmux</a></li><li class="chapter-item expanded "><a href="../tools/git.html"><strong aria-hidden="true">1.5.</strong> git</a></li><li class="chapter-item expanded "><a href="../tools/awk.html"><strong aria-hidden="true">1.6.</strong> awk</a></li><li class="chapter-item expanded "><a href="../tools/emacs.html"><strong aria-hidden="true">1.7.</strong> emacs</a></li><li class="chapter-item expanded "><a href="../tools/gpg.html"><strong aria-hidden="true">1.8.</strong> gpg</a></li><li class="chapter-item expanded "><a href="../tools/gdb.html"><strong aria-hidden="true">1.9.</strong> gdb</a></li><li class="chapter-item expanded "><a href="../tools/gdbserver.html"><strong aria-hidden="true">1.10.</strong> gdbserver</a></li><li class="chapter-item expanded "><a href="../tools/radare2.html"><strong aria-hidden="true">1.11.</strong> radare2</a></li><li class="chapter-item expanded "><a href="../tools/qemu.html"><strong aria-hidden="true">1.12.</strong> qemu</a></li><li class="chapter-item expanded "><a href="../tools/pacman.html"><strong aria-hidden="true">1.13.</strong> pacman</a></li><li class="chapter-item expanded "><a href="../tools/dot.html"><strong aria-hidden="true">1.14.</strong> dot</a></li><li class="chapter-item expanded "><a href="../tools/ffmpeg.html"><strong aria-hidden="true">1.15.</strong> ffmpeg</a></li><li class="chapter-item expanded "><a href="../tools/column.html"><strong aria-hidden="true">1.16.</strong> column</a></li><li class="chapter-item expanded "><a href="../tools/sort.html"><strong aria-hidden="true">1.17.</strong> sort</a></li><li class="chapter-item expanded "><a href="../tools/sed.html"><strong aria-hidden="true">1.18.</strong> sed</a></li></ol></li><li class="chapter-item expanded "><a href="../monitor/index.html"><strong aria-hidden="true">2.</strong> Resource analysis & monitor</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../monitor/lsof.html"><strong aria-hidden="true">2.1.</strong> lsof</a></li><li class="chapter-item expanded "><a href="../monitor/ss.html"><strong aria-hidden="true">2.2.</strong> ss</a></li><li class="chapter-item expanded "><a href="../monitor/pidstat.html"><strong aria-hidden="true">2.3.</strong> pidstat</a></li><li class="chapter-item expanded "><a href="../monitor/pgrep.html"><strong aria-hidden="true">2.4.</strong> pgrep</a></li><li class="chapter-item expanded "><a href="../monitor/ps.html"><strong aria-hidden="true">2.5.</strong> ps</a></li><li class="chapter-item expanded "><a href="../monitor/pmap.html"><strong aria-hidden="true">2.6.</strong> pmap</a></li><li class="chapter-item expanded "><a href="../monitor/pstack.html"><strong aria-hidden="true">2.7.</strong> pstack</a></li></ol></li><li class="chapter-item expanded "><a href="../trace_profile/index.html"><strong aria-hidden="true">3.</strong> Trace and Profile</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../trace_profile/time.html"><strong aria-hidden="true">3.1.</strong> time</a></li><li class="chapter-item expanded "><a href="../trace_profile/strace.html"><strong aria-hidden="true">3.2.</strong> strace</a></li><li class="chapter-item expanded "><a href="../trace_profile/ltrace.html"><strong aria-hidden="true">3.3.</strong> ltrace</a></li><li class="chapter-item expanded "><a href="../trace_profile/perf.html" class="active"><strong aria-hidden="true">3.4.</strong> perf</a></li><li class="chapter-item expanded "><a href="../trace_profile/oprofile.html"><strong aria-hidden="true">3.5.</strong> OProfile</a></li><li class="chapter-item expanded "><a href="../trace_profile/callgrind.html"><strong aria-hidden="true">3.6.</strong> callgrind</a></li></ol></li><li class="chapter-item expanded "><a href="../binary/index.html"><strong aria-hidden="true">4.</strong> Binary</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../binary/od.html"><strong aria-hidden="true">4.1.</strong> od</a></li><li class="chapter-item expanded "><a href="../binary/xxd.html"><strong aria-hidden="true">4.2.</strong> xxd</a></li><li class="chapter-item expanded "><a href="../binary/readelf.html"><strong aria-hidden="true">4.3.</strong> readelf</a></li><li class="chapter-item expanded "><a href="../binary/objdump.html"><strong aria-hidden="true">4.4.</strong> objdump</a></li><li class="chapter-item expanded "><a href="../binary/nm.html"><strong aria-hidden="true">4.5.</strong> nm</a></li></ol></li><li class="chapter-item expanded "><a href="../development/index.html"><strong aria-hidden="true">5.</strong> Development</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../development/c++filt.html"><strong aria-hidden="true">5.1.</strong> c++filt</a></li><li class="chapter-item expanded "><a href="../development/c++.html"><strong aria-hidden="true">5.2.</strong> c++</a></li><li class="chapter-item expanded "><a href="../development/glibc.html"><strong aria-hidden="true">5.3.</strong> glibc</a></li><li class="chapter-item expanded "><a href="../development/gcc.html"><strong aria-hidden="true">5.4.</strong> gcc</a></li><li class="chapter-item expanded "><a href="../development/cmake.html"><strong aria-hidden="true">5.5.</strong> cmake</a></li><li class="chapter-item expanded "><a href="../development/make.html"><strong aria-hidden="true">5.6.</strong> make</a></li><li class="chapter-item expanded "><a href="../development/ld.so.html"><strong aria-hidden="true">5.7.</strong> ld.so</a></li><li class="chapter-item expanded "><a href="../development/symbolver.html"><strong aria-hidden="true">5.8.</strong> symbol versioning</a></li><li class="chapter-item expanded "><a href="../development/python.html"><strong aria-hidden="true">5.9.</strong> python</a></li><li class="chapter-item expanded "><a href="../development/gcov.html"><strong aria-hidden="true">5.10.</strong> gcov</a></li><li class="chapter-item expanded "><a href="../development/pgo.html"><strong aria-hidden="true">5.11.</strong> pgo</a></li></ol></li><li class="chapter-item expanded "><a href="../linux/index.html"><strong aria-hidden="true">6.</strong> Linux</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../linux/systemd.html"><strong aria-hidden="true">6.1.</strong> systemd</a></li><li class="chapter-item expanded "><a href="../linux/coredump.html"><strong aria-hidden="true">6.2.</strong> coredump</a></li><li class="chapter-item expanded "><a href="../linux/ptrace_scope.html"><strong aria-hidden="true">6.3.</strong> ptrace_scope</a></li><li class="chapter-item expanded "><a href="../linux/cryptsetup.html"><strong aria-hidden="true">6.4.</strong> cryptsetup</a></li><li class="chapter-item expanded "><a href="../linux/swap.html"><strong aria-hidden="true">6.5.</strong> swap</a></li><li class="chapter-item expanded "><a href="../linux/input.html"><strong aria-hidden="true">6.6.</strong> input</a></li><li class="chapter-item expanded "><a href="../linux/acl.html"><strong aria-hidden="true">6.7.</strong> acl</a></li><li class="chapter-item expanded "><a href="../linux/zfs.html"><strong aria-hidden="true">6.8.</strong> zfs</a></li></ol></li><li class="chapter-item expanded "><a href="../network/index.html"><strong aria-hidden="true">7.</strong> Network</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../network/tcpdump.html"><strong aria-hidden="true">7.1.</strong> tcpdump</a></li><li class="chapter-item expanded "><a href="../network/tshark.html"><strong aria-hidden="true">7.2.</strong> tshark</a></li><li class="chapter-item expanded "><a href="../network/firewall-cmd.html"><strong aria-hidden="true">7.3.</strong> firewall-cmd</a></li><li class="chapter-item expanded "><a href="../network/nftables.html"><strong aria-hidden="true">7.4.</strong> nftables</a></li></ol></li><li class="chapter-item expanded "><a href="../web/index.html"><strong aria-hidden="true">8.</strong> Web</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../web/html.html"><strong aria-hidden="true">8.1.</strong> html</a></li><li class="chapter-item expanded "><a href="../web/css.html"><strong aria-hidden="true">8.2.</strong> css</a></li><li class="chapter-item expanded "><a href="../web/chartjs.html"><strong aria-hidden="true">8.3.</strong> chartjs</a></li></ol></li><li class="chapter-item expanded "><a href="../arch/index.html"><strong aria-hidden="true">9.</strong> Arch</a></li><li><ol class="section"><li class="chapter-item expanded "><a href="../arch/x86_64.html"><strong aria-hidden="true">9.1.</strong> x86_64</a></li><li class="chapter-item expanded "><a href="../arch/arm64.html"><strong aria-hidden="true">9.2.</strong> arm64</a></li><li class="chapter-item expanded "><a href="../arch/armv7.html"><strong aria-hidden="true">9.3.</strong> armv7</a></li><li class="chapter-item expanded "><a href="../arch/riscv.html"><strong aria-hidden="true">9.4.</strong> riscv</a></li></ol></li></ol>
            </div>
            <div id="sidebar-resize-handle" class="sidebar-resize-handle">
                <div class="sidebar-resize-indicator"></div>
            </div>
        </nav>

        <!-- Track and set sidebar scroll position -->
        <script>
            var sidebarScrollbox = document.querySelector('#sidebar .sidebar-scrollbox');
            sidebarScrollbox.addEventListener('click', function(e) {
                if (e.target.tagName === 'A') {
                    sessionStorage.setItem('sidebar-scroll', sidebarScrollbox.scrollTop);
                }
            }, { passive: true });
            var sidebarScrollTop = sessionStorage.getItem('sidebar-scroll');
            sessionStorage.removeItem('sidebar-scroll');
            if (sidebarScrollTop) {
                // preserve sidebar scroll position when navigating via links within sidebar
                sidebarScrollbox.scrollTop = sidebarScrollTop;
            } else {
                // scroll sidebar to current active section when navigating via "next/previous chapter" buttons
                var activeSection = document.querySelector('#sidebar .active');
                if (activeSection) {
                    activeSection.scrollIntoView({ block: 'center' });
                }
            }
        </script>

        <div id="page-wrapper" class="page-wrapper">

            <div class="page">
                                <div id="menu-bar-hover-placeholder"></div>
                <div id="menu-bar" class="menu-bar sticky">
                    <div class="left-buttons">
                        <label id="sidebar-toggle" class="icon-button" for="sidebar-toggle-anchor" title="Toggle Table of Contents" aria-label="Toggle Table of Contents" aria-controls="sidebar">
                            <i class="fa fa-bars"></i>
                        </label>
                        <button id="theme-toggle" class="icon-button" type="button" title="Change theme" aria-label="Change theme" aria-haspopup="true" aria-expanded="false" aria-controls="theme-list">
                            <i class="fa fa-paint-brush"></i>
                        </button>
                        <ul id="theme-list" class="theme-popup" aria-label="Themes" role="menu">
                            <li role="none"><button role="menuitem" class="theme" id="light">Light</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="rust">Rust</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="coal">Coal</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="navy">Navy</button></li>
                            <li role="none"><button role="menuitem" class="theme" id="ayu">Ayu</button></li>
                        </ul>
                        <button id="search-toggle" class="icon-button" type="button" title="Search. (Shortkey: s)" aria-label="Toggle Searchbar" aria-expanded="false" aria-keyshortcuts="S" aria-controls="searchbar">
                            <i class="fa fa-search"></i>
                        </button>
                    </div>

                    <h1 class="menu-title">Notes</h1>

                    <div class="right-buttons">
                        <a href="../print.html" title="Print this book" aria-label="Print this book">
                            <i id="print-button" class="fa fa-print"></i>
                        </a>
                        <a href="https://github.com/johannst/notes" title="Git repository" aria-label="Git repository">
                            <i id="git-repository-button" class="fa fa-github"></i>
                        </a>

                    </div>
                </div>

                <div id="search-wrapper" class="hidden">
                    <form id="searchbar-outer" class="searchbar-outer">
                        <input type="search" id="searchbar" name="searchbar" placeholder="Search this book ..." aria-controls="searchresults-outer" aria-describedby="searchresults-header">
                    </form>
                    <div id="searchresults-outer" class="searchresults-outer hidden">
                        <div id="searchresults-header" class="searchresults-header"></div>
                        <ul id="searchresults">
                        </ul>
                    </div>
                </div>

                <!-- Apply ARIA attributes after the sidebar and the sidebar toggle button are added to the DOM -->
                <script>
                    document.getElementById('sidebar-toggle').setAttribute('aria-expanded', sidebar === 'visible');
                    document.getElementById('sidebar').setAttribute('aria-hidden', sidebar !== 'visible');
                    Array.from(document.querySelectorAll('#sidebar a')).forEach(function(link) {
                        link.setAttribute('tabIndex', sidebar === 'visible' ? 0 : -1);
                    });
                </script>

                <div id="content" class="content">
                    <main>
                        <h1 id="perf1"><a class="header" href="#perf1">perf(1)</a></h1>
<pre><code>perf list     show supported hw/sw events &amp; metrics
  -v ........ print longer event descriptions
  --details . print information on the perf event names
              and expressions used internally by events

perf stat
  -p &lt;pid&gt; ..... show stats for running process
  -I &lt;ms&gt; ...... show stats periodically over interval &lt;ms&gt;
  -e &lt;ev&gt; ...... select event(s)
  -M &lt;met&gt; ..... print metric(s), this adds the metric events
  --all-user ... configure all selected events for user space
  --all-kernel . configure all selected events for kernel space

perf top
  -p &lt;pid&gt; .. show stats for running process
  -F &lt;hz&gt; ... sampling frequency
  -K ........ hide kernel threads

perf record
  -p &lt;pid&gt; ............... record stats for running process
  -F &lt;hz&gt; ................ sampling frequency
  --call-graph &lt;method&gt; .. [fp, dwarf, lbr] method how to caputre backtrace
                           fp   : use frame-pointer, need to compile with
                                  -fno-omit-frame-pointer
                           dwarf: use .cfi debug information
                           lbr  : use hardware last branch record facility
  -g ..................... short-hand for --call-graph fp
  -e &lt;ev&gt; ................ select event(s)
  --all-user ... configure all selected events for user space
  --all-kernel . configure all selected events for kernel space

perf report
  -n .................... annotate symbols with nr of samples
  --stdio ............... report to stdio, if not presen tui mode
  -g graph,0.5,callee ... show callee based call chains with value &gt;0.5
</code></pre>
<pre><code>Useful &lt;ev&gt;:
  page-faults
  minor-faults
  major-faults
  cpu-cycles`
  task-clock
</code></pre>
<h2 id="select-specific-events"><a class="header" href="#select-specific-events">Select specific events</a></h2>
<p>Events to sample are specified with the <code>-e</code> option, either pass a comma
separated list or pass <code>-e</code> multiple times.</p>
<p>Events are specified in the following form <code>name[:modifier]</code>. The list and
description of the <code>modifier</code> can be found in the
<a href="https://man7.org/linux/man-pages/man1/perf-list.1.html"><code>perf-list(1)</code></a> manpage under <code>EVENT MODIFIERS</code>.</p>
<pre><code class="language-sh"># L1 i$ misses in user space
# L2 i$ stats in user/kernel space mixed
# Sample specified events.
perf stat -e L1-icache-load-misses:u \
          -e l2_rqsts.all_code_rd:uk,l2_rqsts.code_rd_hit:k,l2_rqsts.code_rd_miss:k \
          -- stress -c 2
</code></pre>
<p>The <code>--all-user</code> and <code>--all-kernel</code> options append a <code>:u</code> and <code>:k</code> modifier to
all specified events. Therefore the following two command lines are equivalent.</p>
<pre><code class="language-sh"># 1)
perf stat -e cycles:u,instructions:u -- ls

# 2)
perf stat --all-user -e cycles,instructions -- ls
</code></pre>
<h3 id="raw-events"><a class="header" href="#raw-events">Raw events</a></h3>
<p>In case perf does not provide a <em>symbolic</em> name for an event, the event can be
specified in a <em>raw</em> form as <code>r + UMask + EventCode</code>.</p>
<p>The following is an example for the <a href="https://github.com/intel/perfmon/blob/09c155f72e1b8f14b09aea346a35467a03a7d62b/SNB/events/sandybridge_core.json#L808">L2_RQSTS.CODE_RD_HIT</a> event
with <code>EventCode=0x24</code> and <code>UMask=0x10</code> on my laptop with a <code>sandybridge</code> uarch.</p>
<pre><code class="language-sh">perf stat -e l2_rqsts.code_rd_hit -e r1024 -- ls
# Performance counter stats for 'ls':
#
#       33.942      l2_rqsts.code_rd_hit
#       33.942      r1024
</code></pre>
<h3 id="find-raw-performance-counter-events-intel"><a class="header" href="#find-raw-performance-counter-events-intel">Find raw performance counter events (intel)</a></h3>
<p>The <a href="https://github.com/intel/perfmon"><code>intel/perfmon</code></a> repository provides a performance event
databases for the different intel uarchs.</p>
<p>The table in <a href="https://github.com/intel/perfmon/blob/main/mapfile.csv"><code>mapfile.csv</code></a> can be used to lookup the
corresponding uarch, just grab the family model from the procfs.</p>
<pre><code class="language-sh"> cat /proc/cpuinfo | awk '/^vendor_id/  { V=$3 }
                          /^cpu family/ { F=$4 }
                          /^model\s*:/  { printf &quot;%s-%d-%x\n&quot;,V,F,$3 }'
</code></pre>
<blockquote>
<p>The table in <a href="https://github.com/intel/perfmon/tree/main#performance-monitoring-events">performance monitoring events</a> describes how
events are sorted into the different files.</p>
</blockquote>
<h3 id="raw-events-for-perfs-own-symbolic-names"><a class="header" href="#raw-events-for-perfs-own-symbolic-names">Raw events for perfs own symbolic names</a></h3>
<p>Perf also defines some own <em>symbolic</em> names for events. An example is the
<code>cache-references</code> event. The <a href="https://man7.org/linux/man-pages/man2/perf_event_open.2.html"><code>perf_event_open(2)</code></a> manpage
gives the following description.</p>
<pre><code class="language-man">perf_event_open(2)

PERF_COUNT_HW_CACHE_REFERENCES
    Cache accesses.  Usually this indicates Last Level Cache accesses but this
    may vary depending on your CPU.  This may include prefetches and coherency
    messages; again this depends on the design of your CPU.
</code></pre>
<p>The <code>sysfs</code> can be consulted to get the concrete performance counter on the
given system.</p>
<pre><code class="language-sh">cat /sys/devices/cpu/events/cache-misses
# event=0x2e,umask=0x41
</code></pre>
<h2 id="flamegraph"><a class="header" href="#flamegraph"><a href="https://github.com/brendangregg/FlameGraph"><code>Flamegraph</code></a></a></h2>
<h3 id="flamegraph-with-single-event-trace"><a class="header" href="#flamegraph-with-single-event-trace">Flamegraph with single event trace</a></h3>
<pre><code>perf record -g -e cpu-cycles -p &lt;pid&gt;
perf script | FlameGraph/stackcollapse-perf.pl | FlameGraph/flamegraph.pl &gt; cycles-flamegraph.svg
</code></pre>
<h3 id="flamegraph-with-multiple-event-traces"><a class="header" href="#flamegraph-with-multiple-event-traces">Flamegraph with multiple event traces</a></h3>
<pre><code class="language-sh">perf record -g -e cpu-cycles,page-faults -p &lt;pid&gt;
perf script --per-event-dump
# fold &amp; generate as above
</code></pre>
<h2 id="examples"><a class="header" href="#examples">Examples</a></h2>
<h3 id="determine-theoretical-max-instructions-per-cycle"><a class="header" href="#determine-theoretical-max-instructions-per-cycle">Determine theoretical max instructions per cycle</a></h3>
<pre><code class="language-c">#define NOP4        &quot;nop\nnop\nnop\nnop\n&quot;
#define NOP32       NOP4   NOP4   NOP4   NOP4   NOP4   NOP4   NOP4   NOP4
#define NOP256      NOP32  NOP32  NOP32  NOP32  NOP32  NOP32  NOP32  NOP32
#define NOP2048     NOP256 NOP256 NOP256 NOP256 NOP256 NOP256 NOP256 NOP256

int main() {
  for (unsigned i = 0; i &lt; 2000000; ++i) {
    asm volatile(NOP2048);
  }
}
</code></pre>
<pre><code class="language-sh">perf stat -e cycles,instructions ./noploop
# Performance counter stats for './noploop':
#
#     1.031.075.940      cycles
#     4.103.534.341      instructions       #    3,98  insn per cycle
</code></pre>
<h3 id="caller-vs-callee-callstacks"><a class="header" href="#caller-vs-callee-callstacks">Caller vs callee callstacks</a></h3>
<p>The following gives an example for a scenario where we have the following calls</p>
<ul>
<li><code>main -&gt; do_foo() -&gt; do_work()</code></li>
<li><code>main -&gt; do_bar() -&gt; do_work()</code></li>
</ul>
<pre><code class="language-sh">perf report --stdio -g graph,callee

# Children      Self  Command  Shared Object         Symbols
# ........  ........  .......  ....................  .................
#
#  49.71%    49.66%   bench    bench                 [.] do_work
#          |
#           --49.66%--_start                &lt;- callstack bottom
#                     __libc_start_main
#                     0x7ff366c62ccf
#                     main
#                     |
#                     |--25.13%--do_bar
#                     |          do_work    &lt;- callstack top
#                     |
#                      --24.53%--do_foo
#                                do_work

perf report --stdio -g graph,callee

# Children      Self  Command  Shared Object         Symbols
# ........  ........  .......  ....................  .................
#
#  49.71%    49.66%   bench    bench                 [.] do_work
#          |
#          ---do_work                       &lt;- callstack top
#             |
#             |--25.15%--do_bar
#             |          main
#             |          0x7ff366c62ccf
#             |          __libc_start_main
#             |          _start             &lt;- callstack bottom
#             |
#              --24.55%--do_foo
#                        main
#                        0x7ff366c62ccf
#                        __libc_start_main
#                        _start             &lt;- callstack bottom
</code></pre>
<h2 id="references"><a class="header" href="#references">References</a></h2>
<ul>
<li><a href="https://github.com/intel/perfmon">intel/perfmon</a> - intel PMU event database per uarch</li>
<li><a href="https://perfmon-events.intel.com/">intel/perfmon-html</a> - a html rendered version of the PMU events with search</li>
<li><a href="https://github.com/intel/perfmon/blob/main/mapfile.csv">intel/perfmon/mapfile.csv</a> - processor family to uarch mapping</li>
<li><a href="https://github.com/torvalds/linux/tree/master/tools/perf/pmu-events/arch/x86">linux/perf/events</a> - x86 PMU events known to perf tools</li>
<li><a href="https://github.com/torvalds/linux/blob/master/arch/x86/events/intel/core.c">linux/arch/events</a> - x86 PMU events linux kernel</li>
<li><a href="https://en.wikichip.org/wiki/WikiChip">wikichip</a> - computer architecture wiki</li>
<li><a href="https://man7.org/linux/man-pages/man1/perf-list.1.html">perf-list(1)</a> - manpage</li>
<li><a href="https://man7.org/linux/man-pages/man2/perf_event_open.2.html">perf_event_open(2)</a> - manpage</li>
</ul>

                    </main>

                    <nav class="nav-wrapper" aria-label="Page navigation">
                        <!-- Mobile navigation buttons -->
                            <a rel="prev" href="../trace_profile/ltrace.html" class="mobile-nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
                                <i class="fa fa-angle-left"></i>
                            </a>

                            <a rel="next prefetch" href="../trace_profile/oprofile.html" class="mobile-nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
                                <i class="fa fa-angle-right"></i>
                            </a>

                        <div style="clear: both"></div>
                    </nav>
                </div>
            </div>

            <nav class="nav-wide-wrapper" aria-label="Page navigation">
                    <a rel="prev" href="../trace_profile/ltrace.html" class="nav-chapters previous" title="Previous chapter" aria-label="Previous chapter" aria-keyshortcuts="Left">
                        <i class="fa fa-angle-left"></i>
                    </a>

                    <a rel="next prefetch" href="../trace_profile/oprofile.html" class="nav-chapters next" title="Next chapter" aria-label="Next chapter" aria-keyshortcuts="Right">
                        <i class="fa fa-angle-right"></i>
                    </a>
            </nav>

        </div>


        <script>
            window.playground_copyable = true;
        </script>


        <script src="../elasticlunr.min.js"></script>
        <script src="../mark.min.js"></script>
        <script src="../searcher.js"></script>

        <script src="../clipboard.min.js"></script>
        <script src="../highlight.js"></script>
        <script src="../book.js"></script>

        <!-- Custom JS scripts -->


    </div>
    </body>
</html>