[lkml]   [2012]   [Jun]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    Patch in this message
    Subject[tip:perf/core] uprobes: Teach find_active_uprobe() to provide the "is_swbp" info
    Commit-ID:  d790d34653ab20c74034902f5f0889bba807949a
    Author: Oleg Nesterov <>
    AuthorDate: Tue, 29 May 2012 21:29:14 +0200
    Committer: Ingo Molnar <>
    CommitDate: Wed, 6 Jun 2012 17:15:24 +0200

    uprobes: Teach find_active_uprobe() to provide the "is_swbp" info

    A separate patch to simplify the review, and for the

    The patch adds another "int *is_swbp" argument to
    find_active_uprobe(), so far its only caller doesn't use this

    With this patch find_active_uprobe() additionally does:

    - if find_vma() + ->vm_start check fails, *is_swbp = -EFAULT

    - otherwise, if valid_vma() + find_uprobe() fails, it holds
    the result of is_swbp_at_addr(), can be negative too. The
    latter is only possible if we raced with another thread
    which did munmap/etc after we hit this bp.

    IOW. If find_active_uprobe(&is_swbp) returns NULL, the caller
    can look at is_swbp to figure out whether the current insn is bp
    or not, or detect the race with another thread if it is

    Note: I think that performance-wise this change is fine. This
    adds is_swbp_at_addr(), but only if we raced with
    uprobe_unregister() or if we hit the "normal" int3 but this mm
    has uprobes as well. And even in this case the slow
    read_opcode() path is very unlikely, this insn recently
    triggered do_int3(), __copy_from_user_inatomic() shouldn't fail
    in the likely case.

    Signed-off-by: Oleg Nesterov <>
    Acked-by: Srikar Dronamraju <>
    Cc: Ananth N Mavinakayanahalli <>
    Cc: Anton Arapov <>
    Cc: Linus Torvalds <>
    Cc: Masami Hiramatsu <>
    Cc: Peter Zijlstra <>
    Signed-off-by: Ingo Molnar <>
    kernel/events/uprobes.c | 11 ++++++++---
    1 files changed, 8 insertions(+), 3 deletions(-)

    diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
    index eaf4d55..ee3df70 100644
    --- a/kernel/events/uprobes.c
    +++ b/kernel/events/uprobes.c
    @@ -1489,7 +1489,7 @@ static bool can_skip_sstep(struct uprobe *uprobe, struct pt_regs *regs)
    return false;

    -static struct uprobe *find_active_uprobe(unsigned long bp_vaddr)
    +static struct uprobe *find_active_uprobe(unsigned long bp_vaddr, int *is_swbp)
    struct mm_struct *mm = current->mm;
    struct uprobe *uprobe = NULL;
    @@ -1497,7 +1497,6 @@ static struct uprobe *find_active_uprobe(unsigned long bp_vaddr)

    vma = find_vma(mm, bp_vaddr);
    if (vma && vma->vm_start <= bp_vaddr) {
    if (valid_vma(vma, false)) {
    struct inode *inode;
    @@ -1508,6 +1507,11 @@ static struct uprobe *find_active_uprobe(unsigned long bp_vaddr)
    offset += (vma->vm_pgoff << PAGE_SHIFT);
    uprobe = find_uprobe(inode, offset);
    + if (!uprobe)
    + *is_swbp = is_swbp_at_addr(mm, bp_vaddr);
    + } else {
    + *is_swbp = -EFAULT;

    srcu_read_unlock_raw(&uprobes_srcu, current->uprobe_srcu_id);
    @@ -1526,9 +1530,10 @@ static void handle_swbp(struct pt_regs *regs)
    struct uprobe_task *utask;
    struct uprobe *uprobe;
    unsigned long bp_vaddr;
    + int is_swbp;

    bp_vaddr = uprobe_get_swbp_addr(regs);
    - uprobe = find_active_uprobe(bp_vaddr);
    + uprobe = find_active_uprobe(bp_vaddr, &is_swbp);

    if (!uprobe) {
    /* No matching uprobe; signal SIGTRAP. */

     \ /
      Last update: 2012-06-07 03:01    [W:0.028 / U:4.908 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site