Documentation/trace/uprobetracer.txt

   1             Uprobe-tracer: Uprobe-based Event Tracing
   2             =========================================
   3
   4            Documentation written by Srikar Dronamraju
   5
   6
   7 Overview
   8 --------
   9 Uprobe based trace events are similar to kprobe based trace events.
  10 To enable this feature, build your kernel with CONFIG_UPROBE_EVENT=y.
  11
  12 Similar to the kprobe-event tracer, this doesn't need to be activated via
  13 current_tracer. Instead of that, add probe points via
  14 /sys/kernel/debug/tracing/uprobe_events, and enable it via
  15 /sys/kernel/debug/tracing/events/uprobes/<EVENT>/enabled.
  16
  17 However unlike kprobe-event tracer, the uprobe event interface expects the
  18 user to calculate the offset of the probepoint in the object.
  19
  20 Synopsis of uprobe_tracer
  21 -------------------------
  22   p[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a uprobe
  23   r[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a return uprobe (uretprobe)
  24   -:[GRP/]EVENT                           : Clear uprobe or uretprobe event
  25
  26   GRP           : Group name. If omitted, "uprobes" is the default value.
  27   EVENT         : Event name. If omitted, the event name is generated based
  28                   on PATH+OFFSET.
  29   PATH          : Path to an executable or a library.
  30   OFFSET        : Offset where the probe is inserted.
  31
  32   FETCHARGS     : Arguments. Each probe can have up to 128 args.
  33    %REG         : Fetch register REG
  34    @ADDR        : Fetch memory at ADDR (ADDR should be in userspace)
  35    @+OFFSET     : Fetch memory at OFFSET (OFFSET from same file as PATH)
  36    $stackN      : Fetch Nth entry of stack (N >= 0)
  37    $stack       : Fetch stack address.
  38    $retval      : Fetch return value.(*)
  39    $comm        : Fetch current task comm.
  40    +|-offs(FETCHARG) : Fetch memory at FETCHARG +|- offs address.(**)
  41    NAME=FETCHARG     : Set NAME as the argument name of FETCHARG.
  42    FETCHARG:TYPE     : Set TYPE as the type of FETCHARG. Currently, basic types
  43                        (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal types
  44                        (x8/x16/x32/x64), "string" and bitfield are supported.
  45
  46   (*) only for return probe.
  47   (**) this is useful for fetching a field of data structures.
  48
  49 Types
  50 -----
  51 Several types are supported for fetch-args. Uprobe tracer will access memory
  52 by given type. Prefix 's' and 'u' means those types are signed and unsigned
  53 respectively. 'x' prefix implies it is unsigned. Traced arguments are shown
  54 in decimal ('s' and 'u') or hexadecimal ('x'). Without type casting, 'x32'
  55 or 'x64' is used depends on the architecture (e.g. x86-32 uses x32, and
  56 x86-64 uses x64).
  57 String type is a special type, which fetches a "null-terminated" string from
  58 user space.
  59 Bitfield is another special type, which takes 3 parameters, bit-width, bit-
  60 offset, and container-size (usually 32). The syntax is;
  61
  62  b<bit-width>@<bit-offset>/<container-size>
  63
  64 For $comm, the default type is "string"; any other type is invalid.
  65
  66
  67 Event Profiling
  68 ---------------
  69 You can check the total number of probe hits and probe miss-hits via
  70 /sys/kernel/debug/tracing/uprobe_profile.
  71 The first column is event name, the second is the number of probe hits,
  72 the third is the number of probe miss-hits.
  73
  74 Usage examples
  75 --------------
  76  * Add a probe as a new uprobe event, write a new definition to uprobe_events
  77 as below: (sets a uprobe at an offset of 0x4245c0 in the executable /bin/bash)
  78
  79     echo 'p: /bin/bash:0x4245c0' > /sys/kernel/debug/tracing/uprobe_events
  80
  81  * Add a probe as a new uretprobe event:
  82
  83     echo 'r: /bin/bash:0x4245c0' > /sys/kernel/debug/tracing/uprobe_events
  84
  85  * Unset registered event:
  86
  87     echo '-:bash_0x4245c0' >> /sys/kernel/debug/tracing/uprobe_events
  88
  89  * Print out the events that are registered:
  90
  91     cat /sys/kernel/debug/tracing/uprobe_events
  92
  93  * Clear all events:
  94
  95     echo > /sys/kernel/debug/tracing/uprobe_events
  96
  97 Following example shows how to dump the instruction pointer and %ax register
  98 at the probed text address. Probe zfree function in /bin/zsh:
  99
 100     # cd /sys/kernel/debug/tracing/
 101     # cat /proc/`pgrep zsh`/maps | grep /bin/zsh | grep r-xp
 102     00400000-0048a000 r-xp 00000000 08:03 130904 /bin/zsh
 103     # objdump -T /bin/zsh | grep -w zfree
 104     0000000000446420 g    DF .text  0000000000000012  Base        zfree
 105
 106   0x46420 is the offset of zfree in object /bin/zsh that is loaded at
 107   0x00400000. Hence the command to uprobe would be:
 108
 109     # echo 'p:zfree_entry /bin/zsh:0x46420 %ip %ax' > uprobe_events
 110
 111   And the same for the uretprobe would be:
 112
 113     # echo 'r:zfree_exit /bin/zsh:0x46420 %ip %ax' >> uprobe_events
 114
 115 Please note: User has to explicitly calculate the offset of the probe-point
 116 in the object. We can see the events that are registered by looking at the
 117 uprobe_events file.
 118
 119     # cat uprobe_events
 120     p:uprobes/zfree_entry /bin/zsh:0x00046420 arg1=%ip arg2=%ax
 121     r:uprobes/zfree_exit /bin/zsh:0x00046420 arg1=%ip arg2=%ax
 122
 123 Format of events can be seen by viewing the file events/uprobes/zfree_entry/format
 124
 125     # cat events/uprobes/zfree_entry/format
 126     name: zfree_entry
 127     ID: 922
 128     format:
 129          field:unsigned short common_type;         offset:0;  size:2; signed:0;
 130          field:unsigned char common_flags;         offset:2;  size:1; signed:0;
 131          field:unsigned char common_preempt_count; offset:3;  size:1; signed:0;
 132          field:int common_pid;                     offset:4;  size:4; signed:1;
 133          field:int common_padding;                 offset:8;  size:4; signed:1;
 134
 135          field:unsigned long __probe_ip;           offset:12; size:4; signed:0;
 136          field:u32 arg1;                           offset:16; size:4; signed:0;
 137          field:u32 arg2;                           offset:20; size:4; signed:0;
 138
 139     print fmt: "(%lx) arg1=%lx arg2=%lx", REC->__probe_ip, REC->arg1, REC->arg2
 140
 141 Right after definition, each event is disabled by default. For tracing these
 142 events, you need to enable it by:
 143
 144     # echo 1 > events/uprobes/enable
 145
 146 Lets disable the event after sleeping for some time.
 147
 148     # sleep 20
 149     # echo 0 > events/uprobes/enable
 150
 151 And you can see the traced information via /sys/kernel/debug/tracing/trace.
 152
 153     # cat trace
 154     # tracer: nop
 155     #
 156     #           TASK-PID    CPU#    TIMESTAMP  FUNCTION
 157     #              | |       |          |         |
 158                  zsh-24842 [006] 258544.995456: zfree_entry: (0x446420) arg1=446420 arg2=79
 159                  zsh-24842 [007] 258545.000270: zfree_exit:  (0x446540 <- 0x446420) arg1=446540 arg2=0
 160                  zsh-24842 [002] 258545.043929: zfree_entry: (0x446420) arg1=446420 arg2=79
 161                  zsh-24842 [004] 258547.046129: zfree_exit:  (0x446540 <- 0x446420) arg1=446540 arg2=0
 162
 163 Output shows us uprobe was triggered for a pid 24842 with ip being 0x446420
 164 and contents of ax register being 79. And uretprobe was triggered with ip at
 165 0x446540 with counterpart function entry at 0x446420.