Re: [PATCH] kprobes: x86: cleanup __recover_probed_insn().

2015-03-03 Thread Wang Nan
On 2015/3/3 16:21, Masami Hiramatsu wrote: > (2015/03/03 15:39), Wang Nan wrote: >> Since kernel kconfig forbids turning off KPROBES_ON_FTRACE for x86, we >> don't need to consider the situation that a kprobe probing on a ftrace >> location. The only exception should be early kprobe with >>

Re: [PATCH] kprobes: x86: cleanup __recover_probed_insn().

2015-03-03 Thread Masami Hiramatsu
(2015/03/03 15:39), Wang Nan wrote: > Since kernel kconfig forbids turning off KPROBES_ON_FTRACE for x86, we > don't need to consider the situation that a kprobe probing on a ftrace > location. The only exception should be early kprobe with > KPROBES_ON_FTRACE enabled. However, it is still

Re: [PATCH] kprobes: x86: cleanup __recover_probed_insn().

2015-03-03 Thread Masami Hiramatsu
(2015/03/03 15:39), Wang Nan wrote: Since kernel kconfig forbids turning off KPROBES_ON_FTRACE for x86, we don't need to consider the situation that a kprobe probing on a ftrace location. The only exception should be early kprobe with KPROBES_ON_FTRACE enabled. However, it is still impossible

Re: [PATCH] kprobes: x86: cleanup __recover_probed_insn().

2015-03-03 Thread Wang Nan
On 2015/3/3 16:21, Masami Hiramatsu wrote: (2015/03/03 15:39), Wang Nan wrote: Since kernel kconfig forbids turning off KPROBES_ON_FTRACE for x86, we don't need to consider the situation that a kprobe probing on a ftrace location. The only exception should be early kprobe with

[PATCH] kprobes: x86: cleanup __recover_probed_insn().

2015-03-02 Thread Wang Nan
Since kernel kconfig forbids turning off KPROBES_ON_FTRACE for x86, we don't need to consider the situation that a kprobe probing on a ftrace location. The only exception should be early kprobe with KPROBES_ON_FTRACE enabled. However, it is still impossible for it to get a tainted by ftrace if it

[PATCH] kprobes: x86: cleanup __recover_probed_insn().

2015-03-02 Thread Wang Nan
Since kernel kconfig forbids turning off KPROBES_ON_FTRACE for x86, we don't need to consider the situation that a kprobe probing on a ftrace location. The only exception should be early kprobe with KPROBES_ON_FTRACE enabled. However, it is still impossible for it to get a tainted by ftrace if it