[lkml]   [2009]   [Jul]   [16]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    Subject[PATCH -tip -v12 00/11] tracing: kprobe-based event tracer and x86 instruction decoder

    Here are the v12 patches. I updated it for the latest -tip and
    add fix some bugs.

    Here are the patches of kprobe-based event tracer for x86, version 12,
    which allows you to probe various kernel events through ftrace interface.
    The tracer supports per-probe filtering which allows you to set filters
    on each probe and shows formats of each probe. I think this is more
    generic integration with ftrace, especially event-tracer.

    This version includes below small fixes.
    - Fix a buffer overflow bug. (PATCH 8/11)
    - Fix indirect memory access string bug. (PATCH 8/11)
    - Remove subsystem event directory if it is empty. (PATCH 6/11)
    - Cleanup code an remove redundant checks. (PATCH 8/11, 11/11)

    This patchset also includes x86(-64) instruction decoder which
    supports non-SSE/FP opcodes and includes x86 opcode map. The decoder
    is used for finding the instruction boundaries when inserting new
    kprobes. I think it will be possible to share this opcode map
    with KVM's decoder.
    The decoder is tested when building kernel, the test compares the
    results of objdump and the decoder right after building vmlinux.
    You can enable that test by CONFIG_X86_DECODER_SELFTEST=y.

    This series can be applied on the latest linux-2.6.31-rc3-tip.

    This supports only x86(-32/-64) (but porting it on other arch
    just needs kprobes/kretprobes and register and stack access APIs).

    Enhancement ideas will be added after merging:
    - Make a stress test of kprobes on this tracer.
    - Easy probe setting wrapper which analyzes dwarf..
    - .init function tracing support.
    - Support primitive types(long, ulong, int, uint, etc) for args.

    Kprobe-based Event Tracer

    This tracer is similar to the events tracer which is based on Tracepoint
    infrastructure. Instead of Tracepoint, this tracer is based on kprobes(kprobe
    and kretprobe). It probes anywhere where kprobes can probe(this means, all
    functions body except for __kprobes functions).

    Unlike the function tracer, this tracer can probe instructions inside of
    kernel functions. It allows you to check which instruction has been executed.

    Unlike the Tracepoint based events tracer, this tracer can add new probe points
    on the fly.

    Similar to the events tracer, this tracer doesn't need to be activated via
    current_tracer, instead of that, just set probe points via
    /sys/kernel/debug/tracing/kprobe_events. And you can set filters on each
    probe events via /sys/kernel/debug/tracing/events/kprobes/<EVENT>/filter.

    Synopsis of kprobe_events
    p[:EVENT] SYMBOL[+offs|-offs]|MEMADDR [FETCHARGS] : Set a probe
    r[:EVENT] SYMBOL[+0] [FETCHARGS] : Set a return probe

    EVENT : Event name. If omitted, the event name is generated
    based on SYMBOL+offs or MEMADDR.
    SYMBOL[+offs|-offs] : Symbol+offset where the probe is inserted.
    MEMADDR : Address where the probe is inserted.

    FETCHARGS : Arguments. Each probe can have up to 128 args.
    %REG : Fetch register REG
    sN : Fetch Nth entry of stack (N >= 0)
    @ADDR : Fetch memory at ADDR (ADDR should be in kernel)
    @SYM[+|-offs] : Fetch memory at SYM +|- offs (SYM should be a data symbol)
    aN : Fetch function argument. (N >= 0)(*)
    rv : Fetch return value.(**)
    ra : Fetch return address.(**)
    +|-offs(FETCHARG) : fetch memory at FETCHARG +|- offs address.(***)

    (*) aN may not correct on asmlinkaged functions and at the middle of
    function body.
    (**) only for return probe.
    (***) this is useful for fetching a field of data structures.

    Per-Probe Event Filtering
    Per-probe event filtering feature allows you to set different filter on each
    probe and gives you what arguments will be shown in trace buffer. If an event
    name is specified right after 'p:' or 'r:' in kprobe_events, the tracer adds
    an event under tracing/events/kprobes/<EVENT>, at the directory you can see
    'id', 'enabled', 'format' and 'filter'.

    You can enable/disable the probe by writing 1 or 0 on it.

    It shows the format of this probe event. It also shows aliases of arguments
    which you specified to kprobe_events.

    You can write filtering rules of this event. And you can use both of aliase
    names and field names for describing filters.

    Event Profiling
    You can check the total number of probe hits and probe miss-hits via
    The first column is event name, the second is the number of probe hits,
    the third is the number of probe miss-hits.

    Usage examples
    To add a probe as a new event, write a new definition to kprobe_events
    as below.

    echo p:myprobe do_sys_open a0 a1 a2 a3 > /sys/kernel/debug/tracing/kprobe_events

    This sets a kprobe on the top of do_sys_open() function with recording
    1st to 4th arguments as "myprobe" event.

    echo r:myretprobe do_sys_open rv ra >> /sys/kernel/debug/tracing/kprobe_events

    This sets a kretprobe on the return point of do_sys_open() function with
    recording return value and return address as "myretprobe" event.
    You can see the format of these events via

    cat /sys/kernel/debug/tracing/events/kprobes/myprobe/format
    name: myprobe
    ID: 23
    field:unsigned short common_type; offset:0; size:2;
    field:unsigned char common_flags; offset:2; size:1;
    field:unsigned char common_preempt_count; offset:3; size:1;
    field:int common_pid; offset:4; size:4;
    field:int common_tgid; offset:8; size:4;

    field: unsigned long ip; offset:16;tsize:8;
    field: int nargs; offset:24;tsize:4;
    field: unsigned long arg0; offset:32;tsize:8;
    field: unsigned long arg1; offset:40;tsize:8;
    field: unsigned long arg2; offset:48;tsize:8;
    field: unsigned long arg3; offset:56;tsize:8;

    alias: a0; original: arg0;
    alias: a1; original: arg1;
    alias: a2; original: arg2;
    alias: a3; original: arg3;

    print fmt: "%lx: 0x%lx 0x%lx 0x%lx 0x%lx", ip, arg0, arg1, arg2, arg3

    You can see that the event has 4 arguments and alias expressions
    corresponding to it.

    echo > /sys/kernel/debug/tracing/kprobe_events

    This clears all probe points. and you can see the traced information via

    cat /sys/kernel/debug/tracing/trace
    # tracer: nop
    # | | | | |
    <...>-1447 [001] 1038282.286875: do_sys_open+0x0/0xd6: 0x3 0x7fffd1ec4440 0x8000 0x0
    <...>-1447 [001] 1038282.286878: sys_openat+0xc/0xe <- do_sys_open: 0xfffffffffffffffe 0xffffffff81367a3a
    <...>-1447 [001] 1038282.286885: do_sys_open+0x0/0xd6: 0xffffff9c 0x40413c 0x8000 0x1b6
    <...>-1447 [001] 1038282.286915: sys_open+0x1b/0x1d <- do_sys_open: 0x3 0xffffffff81367a3a
    <...>-1447 [001] 1038282.286969: do_sys_open+0x0/0xd6: 0xffffff9c 0x4041c6 0x98800 0x10
    <...>-1447 [001] 1038282.286976: sys_open+0x1b/0x1d <- do_sys_open: 0x3 0xffffffff81367a3a

    Each line shows when the kernel hits a probe, and <- SYMBOL means kernel
    returns from SYMBOL(e.g. "sys_open+0x1b/0x1d <- do_sys_open" means kernel
    returns from do_sys_open to sys_open+0x1b).

    Thank you,


    Masami Hiramatsu (11):
    tracing: Add kprobes event profiling interface
    tracing: Generate names for each kprobe event automatically
    tracing: Kprobe-tracer supports more than 6 arguments
    tracing: add kprobe-based event tracer
    tracing: Introduce TRACE_FIELD_ZERO() macro
    tracing: ftrace dynamic ftrace_event_call support
    x86: add pt_regs register and stack access APIs
    kprobes: cleanup fix_riprel() using insn decoder on x86
    kprobes: checks probe address is instruction boudary on x86
    x86: x86 instruction decoder build-time selftest
    x86: instruction decoder API

    Documentation/trace/kprobetrace.txt | 147 ++++
    arch/x86/Kconfig.debug | 9
    arch/x86/Makefile | 3
    arch/x86/include/asm/inat.h | 127 +++
    arch/x86/include/asm/insn.h | 136 +++
    arch/x86/include/asm/ptrace.h | 62 ++
    arch/x86/kernel/kprobes.c | 197 ++---
    arch/x86/kernel/ptrace.c | 112 +++
    arch/x86/lib/Makefile | 13
    arch/x86/lib/inat.c | 82 ++
    arch/x86/lib/insn.c | 473 ++++++++++++
    arch/x86/lib/x86-opcode-map.txt | 711 ++++++++++++++++++
    arch/x86/scripts/Makefile | 19
    arch/x86/scripts/distill.awk | 42 +
    arch/x86/scripts/gen-insn-attr-x86.awk | 314 ++++++++
    arch/x86/scripts/test_get_len.c | 99 +++
    arch/x86/scripts/user_include.h | 49 +
    include/linux/ftrace_event.h | 13
    include/trace/ftrace.h | 22 -
    kernel/trace/Kconfig | 12
    kernel/trace/Makefile | 1
    kernel/trace/trace.h | 29 +
    kernel/trace/trace_event_types.h | 4
    kernel/trace/trace_events.c | 72 +-
    kernel/trace/trace_export.c | 43 +
    kernel/trace/trace_kprobe.c | 1245 ++++++++++++++++++++++++++++++++
    26 files changed, 3873 insertions(+), 163 deletions(-)
    create mode 100644 Documentation/trace/kprobetrace.txt
    create mode 100644 arch/x86/include/asm/inat.h
    create mode 100644 arch/x86/include/asm/insn.h
    create mode 100644 arch/x86/lib/inat.c
    create mode 100644 arch/x86/lib/insn.c
    create mode 100644 arch/x86/lib/x86-opcode-map.txt
    create mode 100644 arch/x86/scripts/Makefile
    create mode 100644 arch/x86/scripts/distill.awk
    create mode 100644 arch/x86/scripts/gen-insn-attr-x86.awk
    create mode 100644 arch/x86/scripts/test_get_len.c
    create mode 100644 arch/x86/scripts/user_include.h
    create mode 100644 kernel/trace/trace_kprobe.c

    Masami Hiramatsu

    Software Engineer
    Hitachi Computer Products (America), Inc.
    Software Solutions Division


     \ /
      Last update: 2009-07-16 18:01    [W:0.045 / U:8.900 seconds]
    ©2003-2017 Jasper Spaans. hosted at Digital OceanAdvertise on this site