lkml.org 
[lkml]   [2011]   [Jul]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH -tip 02/13] [CLEANUP]tracing/kprobes: merge trace probe enable/disable functions
(2011/07/09 1:37), Steven Rostedt wrote:
> On Mon, 2011-06-27 at 16:26 +0900, Masami Hiramatsu wrote:
>> Merge redundant enable/disable functions into enable_trace_probe()
>> and disable_trace_probe().
>>
>> Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
>> Cc: Steven Rostedt <rostedt@goodmis.org>
>> Cc: Frederic Weisbecker <fweisbec@gmail.com>
>> Cc: Ingo Molnar <mingo@redhat.com>
>> ---
>>
>> kernel/trace/trace_kprobe.c | 88 +++++++++++++++++--------------------------
>> 1 files changed, 34 insertions(+), 54 deletions(-)
>>
>> diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c
>> index bad87e9..ce5e6aa 100644
>> --- a/kernel/trace/trace_kprobe.c
>> +++ b/kernel/trace/trace_kprobe.c
>> @@ -683,6 +683,34 @@ static struct trace_probe *find_trace_probe(const char
*event,
>> return NULL;
>> }
>>
>> +/* Enable trace_probe - @flag must be TP_FLAG_TRACE or TP_FLAG_PROFILE */
>> +static int enable_trace_probe(struct trace_probe *tp, int flag)
>> +{
>> + int ret = 0;
>> +
>> + tp->flags |= flag;
>> + if (tp->flags & (TP_FLAG_TRACE | TP_FLAG_PROFILE)) {
>> + if (trace_probe_is_return(tp))
>> + ret = enable_kretprobe(&tp->rp);
>> + else
>> + ret = enable_kprobe(&tp->rp.kp);
>> + }
>
> Hmm, this seems weird. Should we have any protection here? I mean, is it
> ok to call the enable_kprobe() twice? Or should we have something like:
>
> {
> int old_flags = tp->flags;
> int ret = 0;
>
> tp->flags |= flag;
>
> if (!(old_flags & (TP_FLAG_TRACE | TP_FLAG_PROFILE)) &&
> flag & (TP_FLAG_TRACE | TP_FLAG_PROFILE)) {
> [...]
> }
>
> return ret;
> }


Ah, no problem, enable_kprobe() enables given kprobe only if it is
disabled.

>
>
>> +
>> + return ret;
>> +}
>> +
>> +/* Disable trace_probe - @flag must be TP_FLAG_TRACE or TP_FLAG_PROFILE */
>> +static void disable_trace_probe(struct trace_probe *tp, int flag)
>> +{
>> + tp->flags &= ~flag;
>> + if (!(tp->flags & (TP_FLAG_TRACE | TP_FLAG_PROFILE))) {
>> + if (trace_probe_is_return(tp))
>> + disable_kretprobe(&tp->rp);
>> + else
>> + disable_kprobe(&tp->rp.kp);
>> + }
>> +}
>
> Same here.

Same as above.

>
> Or do we want to reenable or re disable the probe?

Yeah, we do it for enabling/disabling each event.

Thank you,


>
> -- Steve

--
Masami HIRAMATSU
Software Platform Research Dept. Linux Technology Center
Hitachi, Ltd., Yokohama Research Laboratory
E-mail: masami.hiramatsu.pt@hitachi.com


\
 
 \ /
  Last update: 2011-07-09 06:45    [from the cache]
©2003-2011 Jasper Spaans