1perf-trace(1) 2============= 3 4NAME 5---- 6perf-trace - strace inspired tool 7 8SYNOPSIS 9-------- 10[verse] 11'perf trace' 12'perf trace record' 13 14DESCRIPTION 15----------- 16This command will show the events associated with the target, initially 17syscalls, but other system events like pagefaults, task lifetime events, 18scheduling events, etc. 19 20This is a live mode tool in addition to working with perf.data files like 21the other perf tools. Files can be generated using the 'perf record' command 22but the session needs to include the raw_syscalls events (-e 'raw_syscalls:*'). 23Alternatively, 'perf trace record' can be used as a shortcut to 24automatically include the raw_syscalls events when writing events to a file. 25 26The following options apply to perf trace; options to perf trace record are 27found in the perf record man page. 28 29OPTIONS 30------- 31 32-a:: 33--all-cpus:: 34 System-wide collection from all CPUs. 35 36-e:: 37--expr:: 38 List of syscalls to show, currently only syscall names. 39 Prefixing with ! shows all syscalls but the ones specified. You may 40 need to escape it. 41 42-o:: 43--output=:: 44 Output file name. 45 46-p:: 47--pid=:: 48 Record events on existing process ID (comma separated list). 49 50-t:: 51--tid=:: 52 Record events on existing thread ID (comma separated list). 53 54-u:: 55--uid=:: 56 Record events in threads owned by uid. Name or number. 57 58--filter-pids=:: 59 Filter out events for these pids and for 'trace' itself (comma separated list). 60 61-v:: 62--verbose=:: 63 Verbosity level. 64 65--no-inherit:: 66 Child tasks do not inherit counters. 67 68-m:: 69--mmap-pages=:: 70 Number of mmap data pages (must be a power of two) or size 71 specification with appended unit character - B/K/M/G. The 72 size is rounded up to have nearest pages power of two value. 73 74-C:: 75--cpu:: 76Collect samples only on the list of CPUs provided. Multiple CPUs can be provided as a 77comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-2. 78In per-thread mode with inheritance mode on (default), Events are captured only when 79the thread executes on the designated CPUs. Default is to monitor all CPUs. 80 81--duration: 82 Show only events that had a duration greater than N.M ms. 83 84--sched: 85 Accrue thread runtime and provide a summary at the end of the session. 86 87-i 88--input 89 Process events from a given perf data file. 90 91-T 92--time 93 Print full timestamp rather time relative to first sample. 94 95--comm:: 96 Show process COMM right beside its ID, on by default, disable with --no-comm. 97 98-s:: 99--summary:: 100 Show only a summary of syscalls by thread with min, max, and average times 101 (in msec) and relative stddev. 102 103-S:: 104--with-summary:: 105 Show all syscalls followed by a summary by thread with min, max, and 106 average times (in msec) and relative stddev. 107 108--tool_stats:: 109 Show tool stats such as number of times fd->pathname was discovered thru 110 hooking the open syscall return + vfs_getname or via reading /proc/pid/fd, etc. 111 112-F=[all|min|maj]:: 113--pf=[all|min|maj]:: 114 Trace pagefaults. Optionally, you can specify whether you want minor, 115 major or all pagefaults. Default value is maj. 116 117--syscalls:: 118 Trace system calls. This options is enabled by default. 119 120--call-graph [mode,type,min[,limit],order[,key][,branch]]:: 121 Setup and enable call-graph (stack chain/backtrace) recording. 122 See `--call-graph` section in perf-record and perf-report 123 man pages for details. The ones that are most useful in 'perf trace' 124 are 'dwarf' and 'lbr', where available, try: 'perf trace --call-graph dwarf'. 125 126--kernel-syscall-graph:: 127 Show the kernel callchains on the syscall exit path. 128 129--event:: 130 Trace other events, see 'perf list' for a complete list. 131 132--max-stack:: 133 Set the stack depth limit when parsing the callchain, anything 134 beyond the specified depth will be ignored. Note that at this point 135 this is just about the presentation part, i.e. the kernel is still 136 not limiting, the overhead of callchains needs to be set via the 137 knobs in --call-graph dwarf. 138 139 Default: 127 140 141--min-stack:: 142 Set the stack depth limit when parsing the callchain, anything 143 below the specified depth will be ignored. Disabled by default. 144 145--proc-map-timeout:: 146 When processing pre-existing threads /proc/XXX/mmap, it may take a long time, 147 because the file may be huge. A time out is needed in such cases. 148 This option sets the time out limit. The default value is 500 ms. 149 150PAGEFAULTS 151---------- 152 153When tracing pagefaults, the format of the trace is as follows: 154 155<min|maj>fault [<ip.symbol>+<ip.offset>] => <addr.dso@addr.offset> (<map type><addr level>). 156 157- min/maj indicates whether fault event is minor or major; 158- ip.symbol shows symbol for instruction pointer (the code that generated the 159 fault); if no debug symbols available, perf trace will print raw IP; 160- addr.dso shows DSO for the faulted address; 161- map type is either 'd' for non-executable maps or 'x' for executable maps; 162- addr level is either 'k' for kernel dso or '.' for user dso. 163 164For symbols resolution you may need to install debugging symbols. 165 166Please be aware that duration is currently always 0 and doesn't reflect actual 167time it took for fault to be handled! 168 169When --verbose specified, perf trace tries to print all available information 170for both IP and fault address in the form of dso@symbol+offset. 171 172EXAMPLES 173-------- 174 175Trace only major pagefaults: 176 177 $ perf trace --no-syscalls -F 178 179Trace syscalls, major and minor pagefaults: 180 181 $ perf trace -F all 182 183 1416.547 ( 0.000 ms): python/20235 majfault [CRYPTO_push_info_+0x0] => /lib/x86_64-linux-gnu/libcrypto.so.1.0.0@0x61be0 (x.) 184 185 As you can see, there was major pagefault in python process, from 186 CRYPTO_push_info_ routine which faulted somewhere in libcrypto.so. 187 188SEE ALSO 189-------- 190linkperf:perf-record[1], linkperf:perf-script[1] 191