Orange Pi5 kernel

Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards

3 Commits   0 Branches   0 Tags
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   1) =========================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   2) Uprobe-tracer: Uprobe-based Event Tracing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   3) =========================================
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   4) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   5) :Author: Srikar Dronamraju
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   6) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   7) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   8) Overview
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300   9) --------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  10) Uprobe based trace events are similar to kprobe based trace events.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  11) To enable this feature, build your kernel with CONFIG_UPROBE_EVENTS=y.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  12) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  13) Similar to the kprobe-event tracer, this doesn't need to be activated via
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  14) current_tracer. Instead of that, add probe points via
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  15) /sys/kernel/debug/tracing/uprobe_events, and enable it via
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  16) /sys/kernel/debug/tracing/events/uprobes/<EVENT>/enable.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  17) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  18) However unlike kprobe-event tracer, the uprobe event interface expects the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  19) user to calculate the offset of the probepoint in the object.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  20) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  21) You can also use /sys/kernel/debug/tracing/dynamic_events instead of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  22) uprobe_events. That interface will provide unified access to other
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  23) dynamic events too.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  24) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  25) Synopsis of uprobe_tracer
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  26) -------------------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  27) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  28) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  29)   p[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a uprobe
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  30)   r[:[GRP/]EVENT] PATH:OFFSET [FETCHARGS] : Set a return uprobe (uretprobe)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  31)   p[:[GRP/]EVENT] PATH:OFFSET%return [FETCHARGS] : Set a return uprobe (uretprobe)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  32)   -:[GRP/]EVENT                           : Clear uprobe or uretprobe event
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  33) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  34)   GRP           : Group name. If omitted, "uprobes" is the default value.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  35)   EVENT         : Event name. If omitted, the event name is generated based
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  36)                   on PATH+OFFSET.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  37)   PATH          : Path to an executable or a library.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  38)   OFFSET        : Offset where the probe is inserted.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  39)   OFFSET%return : Offset where the return probe is inserted.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  40) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  41)   FETCHARGS     : Arguments. Each probe can have up to 128 args.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  42)    %REG         : Fetch register REG
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  43)    @ADDR	: Fetch memory at ADDR (ADDR should be in userspace)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  44)    @+OFFSET	: Fetch memory at OFFSET (OFFSET from same file as PATH)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  45)    $stackN	: Fetch Nth entry of stack (N >= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  46)    $stack	: Fetch stack address.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  47)    $retval	: Fetch return value.(\*1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  48)    $comm	: Fetch current task comm.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  49)    +|-[u]OFFS(FETCHARG) : Fetch memory at FETCHARG +|- OFFS address.(\*2)(\*3)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  50)    \IMM		: Store an immediate value to the argument.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  51)    NAME=FETCHARG     : Set NAME as the argument name of FETCHARG.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  52)    FETCHARG:TYPE     : Set TYPE as the type of FETCHARG. Currently, basic types
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  53) 		       (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal types
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  54) 		       (x8/x16/x32/x64), "string" and bitfield are supported.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  55) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  56)   (\*1) only for return probe.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  57)   (\*2) this is useful for fetching a field of data structures.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  58)   (\*3) Unlike kprobe event, "u" prefix will just be ignored, becuse uprobe
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  59)         events can access only user-space memory.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  60) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  61) Types
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  62) -----
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  63) Several types are supported for fetch-args. Uprobe tracer will access memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  64) by given type. Prefix 's' and 'u' means those types are signed and unsigned
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  65) respectively. 'x' prefix implies it is unsigned. Traced arguments are shown
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  66) in decimal ('s' and 'u') or hexadecimal ('x'). Without type casting, 'x32'
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  67) or 'x64' is used depends on the architecture (e.g. x86-32 uses x32, and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  68) x86-64 uses x64).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  69) String type is a special type, which fetches a "null-terminated" string from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  70) user space.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  71) Bitfield is another special type, which takes 3 parameters, bit-width, bit-
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  72) offset, and container-size (usually 32). The syntax is::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  73) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  74)  b<bit-width>@<bit-offset>/<container-size>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  75) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  76) For $comm, the default type is "string"; any other type is invalid.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  77) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  78) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  79) Event Profiling
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  80) ---------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  81) You can check the total number of probe hits per event via
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  82) /sys/kernel/debug/tracing/uprobe_profile. The first column is the filename,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  83) the second is the event name, the third is the number of probe hits.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  84) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  85) Usage examples
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  86) --------------
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  87)  * Add a probe as a new uprobe event, write a new definition to uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  88)    as below (sets a uprobe at an offset of 0x4245c0 in the executable /bin/bash)::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  89) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  90)     echo 'p /bin/bash:0x4245c0' > /sys/kernel/debug/tracing/uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  91) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  92)  * Add a probe as a new uretprobe event::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  93) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  94)     echo 'r /bin/bash:0x4245c0' > /sys/kernel/debug/tracing/uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  95) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  96)  * Unset registered event::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  97) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  98)     echo '-:p_bash_0x4245c0' >> /sys/kernel/debug/tracing/uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  99) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100)  * Print out the events that are registered::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102)     cat /sys/kernel/debug/tracing/uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104)  * Clear all events::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106)     echo > /sys/kernel/debug/tracing/uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) Following example shows how to dump the instruction pointer and %ax register
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) at the probed text address. Probe zfree function in /bin/zsh::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111)     # cd /sys/kernel/debug/tracing/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112)     # cat /proc/`pgrep zsh`/maps | grep /bin/zsh | grep r-xp
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113)     00400000-0048a000 r-xp 00000000 08:03 130904 /bin/zsh
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114)     # objdump -T /bin/zsh | grep -w zfree
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115)     0000000000446420 g    DF .text  0000000000000012  Base        zfree
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) 0x46420 is the offset of zfree in object /bin/zsh that is loaded at
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) 0x00400000. Hence the command to uprobe would be::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120)     # echo 'p:zfree_entry /bin/zsh:0x46420 %ip %ax' > uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) And the same for the uretprobe would be::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124)     # echo 'r:zfree_exit /bin/zsh:0x46420 %ip %ax' >> uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) .. note:: User has to explicitly calculate the offset of the probe-point
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) 	in the object.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) We can see the events that are registered by looking at the uprobe_events file.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132)     # cat uprobe_events
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133)     p:uprobes/zfree_entry /bin/zsh:0x00046420 arg1=%ip arg2=%ax
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134)     r:uprobes/zfree_exit /bin/zsh:0x00046420 arg1=%ip arg2=%ax
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) Format of events can be seen by viewing the file events/uprobes/zfree_entry/format.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139)     # cat events/uprobes/zfree_entry/format
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140)     name: zfree_entry
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141)     ID: 922
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142)     format:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143)          field:unsigned short common_type;         offset:0;  size:2; signed:0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144)          field:unsigned char common_flags;         offset:2;  size:1; signed:0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145)          field:unsigned char common_preempt_count; offset:3;  size:1; signed:0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146)          field:int common_pid;                     offset:4;  size:4; signed:1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147)          field:int common_padding;                 offset:8;  size:4; signed:1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149)          field:unsigned long __probe_ip;           offset:12; size:4; signed:0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150)          field:u32 arg1;                           offset:16; size:4; signed:0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151)          field:u32 arg2;                           offset:20; size:4; signed:0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153)     print fmt: "(%lx) arg1=%lx arg2=%lx", REC->__probe_ip, REC->arg1, REC->arg2
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) Right after definition, each event is disabled by default. For tracing these
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) events, you need to enable it by::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158)     # echo 1 > events/uprobes/enable
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) Lets start tracing, sleep for some time and stop tracing.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163)     # echo 1 > tracing_on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164)     # sleep 20
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165)     # echo 0 > tracing_on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) Also, you can disable the event by::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169)     # echo 0 > events/uprobes/enable
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) And you can see the traced information via /sys/kernel/debug/tracing/trace.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) ::
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174)     # cat trace
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175)     # tracer: nop
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176)     #
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177)     #           TASK-PID    CPU#    TIMESTAMP  FUNCTION
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178)     #              | |       |          |         |
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179)                  zsh-24842 [006] 258544.995456: zfree_entry: (0x446420) arg1=446420 arg2=79
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180)                  zsh-24842 [007] 258545.000270: zfree_exit:  (0x446540 <- 0x446420) arg1=446540 arg2=0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181)                  zsh-24842 [002] 258545.043929: zfree_entry: (0x446420) arg1=446420 arg2=79
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182)                  zsh-24842 [004] 258547.046129: zfree_exit:  (0x446540 <- 0x446420) arg1=446540 arg2=0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) Output shows us uprobe was triggered for a pid 24842 with ip being 0x446420
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) and contents of ax register being 79. And uretprobe was triggered with ip at
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) 0x446540 with counterpart function entry at 0x446420.