1 Kprobe-based Event Tracing 2 ========================== 3 4 Documentation is written by Masami Hiramatsu 5 6 7Overview 8-------- 9These events are similar to tracepoint based events. Instead of Tracepoint, 10this is based on kprobes (kprobe and kretprobe). So it can probe wherever 11kprobes can probe (this means, all functions body except for __kprobes 12functions). Unlike the Tracepoint based event, this can be added and removed 13dynamically, on the fly. 14 15To enable this feature, build your kernel with CONFIG_KPROBE_EVENT=y. 16 17Similar to the events tracer, this doesn't need to be activated via 18current_tracer. Instead of that, add probe points via 19/sys/kernel/debug/tracing/kprobe_events, and enable it via 20/sys/kernel/debug/tracing/events/kprobes/<EVENT>/enabled. 21 22 23Synopsis of kprobe_events 24------------------------- 25 p[:[GRP/]EVENT] [MOD:]SYM[+offs]|MEMADDR [FETCHARGS] : Set a probe 26 r[:[GRP/]EVENT] [MOD:]SYM[+0] [FETCHARGS] : Set a return probe 27 -:[GRP/]EVENT : Clear a probe 28 29 GRP : Group name. If omitted, use "kprobes" for it. 30 EVENT : Event name. If omitted, the event name is generated 31 based on SYM+offs or MEMADDR. 32 MOD : Module name which has given SYM. 33 SYM[+offs] : Symbol+offset where the probe is inserted. 34 MEMADDR : Address where the probe is inserted. 35 36 FETCHARGS : Arguments. Each probe can have up to 128 args. 37 %REG : Fetch register REG 38 @ADDR : Fetch memory at ADDR (ADDR should be in kernel) 39 @SYM[+|-offs] : Fetch memory at SYM +|- offs (SYM should be a data symbol) 40 $stackN : Fetch Nth entry of stack (N >= 0) 41 $stack : Fetch stack address. 42 $retval : Fetch return value.(*) 43 $comm : Fetch current task comm. 44 +|-offs(FETCHARG) : Fetch memory at FETCHARG +|- offs address.(**) 45 NAME=FETCHARG : Set NAME as the argument name of FETCHARG. 46 FETCHARG:TYPE : Set TYPE as the type of FETCHARG. Currently, basic types 47 (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal types 48 (x8/x16/x32/x64), "string" and bitfield are supported. 49 50 (*) only for return probe. 51 (**) this is useful for fetching a field of data structures. 52 53Types 54----- 55Several types are supported for fetch-args. Kprobe tracer will access memory 56by given type. Prefix 's' and 'u' means those types are signed and unsigned 57respectively. 'x' prefix implies it is unsigned. Traced arguments are shown 58in decimal ('s' and 'u') or hexadecimal ('x'). Without type casting, 'x32' 59or 'x64' is used depends on the architecture (e.g. x86-32 uses x32, and 60x86-64 uses x64). 61String type is a special type, which fetches a "null-terminated" string from 62kernel space. This means it will fail and store NULL if the string container 63has been paged out. 64Bitfield is another special type, which takes 3 parameters, bit-width, bit- 65offset, and container-size (usually 32). The syntax is; 66 67 b<bit-width>@<bit-offset>/<container-size> 68 69For $comm, the default type is "string"; any other type is invalid. 70 71 72Per-Probe Event Filtering 73------------------------- 74 Per-probe event filtering feature allows you to set different filter on each 75probe and gives you what arguments will be shown in trace buffer. If an event 76name is specified right after 'p:' or 'r:' in kprobe_events, it adds an event 77under tracing/events/kprobes/<EVENT>, at the directory you can see 'id', 78'enabled', 'format' and 'filter'. 79 80enabled: 81 You can enable/disable the probe by writing 1 or 0 on it. 82 83format: 84 This shows the format of this probe event. 85 86filter: 87 You can write filtering rules of this event. 88 89id: 90 This shows the id of this probe event. 91 92 93Event Profiling 94--------------- 95 You can check the total number of probe hits and probe miss-hits via 96/sys/kernel/debug/tracing/kprobe_profile. 97 The first column is event name, the second is the number of probe hits, 98the third is the number of probe miss-hits. 99 100 101Usage examples 102-------------- 103To add a probe as a new event, write a new definition to kprobe_events 104as below. 105 106 echo 'p:myprobe do_sys_open dfd=%ax filename=%dx flags=%cx mode=+4($stack)' > /sys/kernel/debug/tracing/kprobe_events 107 108 This sets a kprobe on the top of do_sys_open() function with recording 1091st to 4th arguments as "myprobe" event. Note, which register/stack entry is 110assigned to each function argument depends on arch-specific ABI. If you unsure 111the ABI, please try to use probe subcommand of perf-tools (you can find it 112under tools/perf/). 113As this example shows, users can choose more familiar names for each arguments. 114 115 echo 'r:myretprobe do_sys_open $retval' >> /sys/kernel/debug/tracing/kprobe_events 116 117 This sets a kretprobe on the return point of do_sys_open() function with 118recording return value as "myretprobe" event. 119 You can see the format of these events via 120/sys/kernel/debug/tracing/events/kprobes/<EVENT>/format. 121 122 cat /sys/kernel/debug/tracing/events/kprobes/myprobe/format 123name: myprobe 124ID: 780 125format: 126 field:unsigned short common_type; offset:0; size:2; signed:0; 127 field:unsigned char common_flags; offset:2; size:1; signed:0; 128 field:unsigned char common_preempt_count; offset:3; size:1;signed:0; 129 field:int common_pid; offset:4; size:4; signed:1; 130 131 field:unsigned long __probe_ip; offset:12; size:4; signed:0; 132 field:int __probe_nargs; offset:16; size:4; signed:1; 133 field:unsigned long dfd; offset:20; size:4; signed:0; 134 field:unsigned long filename; offset:24; size:4; signed:0; 135 field:unsigned long flags; offset:28; size:4; signed:0; 136 field:unsigned long mode; offset:32; size:4; signed:0; 137 138 139print fmt: "(%lx) dfd=%lx filename=%lx flags=%lx mode=%lx", REC->__probe_ip, 140REC->dfd, REC->filename, REC->flags, REC->mode 141 142 You can see that the event has 4 arguments as in the expressions you specified. 143 144 echo > /sys/kernel/debug/tracing/kprobe_events 145 146 This clears all probe points. 147 148 Or, 149 150 echo -:myprobe >> kprobe_events 151 152 This clears probe points selectively. 153 154 Right after definition, each event is disabled by default. For tracing these 155events, you need to enable it. 156 157 echo 1 > /sys/kernel/debug/tracing/events/kprobes/myprobe/enable 158 echo 1 > /sys/kernel/debug/tracing/events/kprobes/myretprobe/enable 159 160 And you can see the traced information via /sys/kernel/debug/tracing/trace. 161 162 cat /sys/kernel/debug/tracing/trace 163# tracer: nop 164# 165# TASK-PID CPU# TIMESTAMP FUNCTION 166# | | | | | 167 <...>-1447 [001] 1038282.286875: myprobe: (do_sys_open+0x0/0xd6) dfd=3 filename=7fffd1ec4440 flags=8000 mode=0 168 <...>-1447 [001] 1038282.286878: myretprobe: (sys_openat+0xc/0xe <- do_sys_open) $retval=fffffffffffffffe 169 <...>-1447 [001] 1038282.286885: myprobe: (do_sys_open+0x0/0xd6) dfd=ffffff9c filename=40413c flags=8000 mode=1b6 170 <...>-1447 [001] 1038282.286915: myretprobe: (sys_open+0x1b/0x1d <- do_sys_open) $retval=3 171 <...>-1447 [001] 1038282.286969: myprobe: (do_sys_open+0x0/0xd6) dfd=ffffff9c filename=4041c6 flags=98800 mode=10 172 <...>-1447 [001] 1038282.286976: myretprobe: (sys_open+0x1b/0x1d <- do_sys_open) $retval=3 173 174 175 Each line shows when the kernel hits an event, and <- SYMBOL means kernel 176returns from SYMBOL(e.g. "sys_open+0x1b/0x1d <- do_sys_open" means kernel 177returns from do_sys_open to sys_open+0x1b). 178 179