Messages in this thread Patch in this message | | | Date | Fri, 6 Apr 2012 00:21:06 +0200 | From | Oleg Nesterov <> | Subject | [PATCH 2/6] uprobes: introduce is_swbp_at_addr_fast() |
| |
Add the new helper, is_swbp_at_addr_fast(), will be used by find_active_uprobe().
It is almost the same as is_swbp_at_addr(), but since it plays with current->mm it can avoid the slow get_user_pages() in the likely case. --- kernel/events/uprobes.c | 23 +++++++++++++++++++++++ 1 files changed, 23 insertions(+), 0 deletions(-)
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c index 3d0a4d6..2050b1a 100644 --- a/kernel/events/uprobes.c +++ b/kernel/events/uprobes.c @@ -1474,6 +1474,29 @@ static bool can_skip_sstep(struct uprobe *uprobe, struct pt_regs *regs) return false; } +int __weak is_swbp_at_addr_fast(unsigned long vaddr) +{ + uprobe_opcode_t opcode; + int fault; + + pagefault_disable(); + fault = __copy_from_user_inatomic(&opcode, (void __user*)vaddr, + sizeof(opcode)); + pagefault_enable(); + + if (unlikely(fault)) { + /* + * XXX: read_opcode() lacks FOLL_FORCE, it can fail if + * we race with another thread which does mprotect(NONE) + * after we hit bp. + */ + if (read_opcode(current->mm, vaddr, &opcode)) + return -EFAULT; + } + + return is_swbp_insn(&opcode); +} + static struct uprobe *find_active_uprobe(unsigned long bp_vaddr) { struct mm_struct *mm = current->mm; -- 1.5.5.1
| |