Re: [RFC PATCH][3.0] Tracepoint: dissociate from module mutex (v2)

[Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>]

(2011/08/11 12:23), Steven Rostedt wrote:
> On Thu, 2011-08-11 at 12:14 +0900, Masami Hiramatsu wrote:
>> (2011/08/11 4:18), Mathieu Desnoyers wrote:
>>> Copy the information needed from struct module into a local module list
>>> held within tracepoint.c from within the module coming/going notifier.
>>>
>>> This vastly simplifies locking of tracepoint registration /
>>> unregistration, because we don't have to take the module mutex to
>>> register and unregister tracepoints anymore. Steven Rostedt ran into
>>> dependency problems related to modules mutex vs kprobes mutex vs ftrace
>>> mutex vs tracepoint mutex that seems to be hard to fix without removing
>>> this dependency between tracepoint and module mutex. (note: it should be
>>> investigated whether kprobes could benefit of being dissociated from the
>>> modules mutex too.)
>>
>> Thanks, it seems that kprobes has already mostly done that.
>> It holds module_mutex only in kprobe_optimizer. However,
>> it seems meaningless, because kprobe_mutex already protects
>> kprobe_optimizer against the kprobes module notifier.
>> Thus, a module removing will stays on the notifier until
>> the optimizer runs out. So I think we can remove that mutex lock.
>>
> 
> So should I change my patch 4/5 to just remove the module_mutex?
> 
> [PATCH 4/5][RFC] kprobes: Inverse taking of module_mutex with kprobe_mutex

Right, it should be changed. :-)

Thank you,


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