lkml.org 
[lkml]   [2009]   [Aug]   [27]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] tracing/profile: Fix profile_disable vs module_unload

On Thu, 27 Aug 2009, Peter Zijlstra wrote:

> On Thu, 2009-08-27 at 11:57 -0400, Steven Rostedt wrote:
> > On Thu, 27 Aug 2009, Peter Zijlstra wrote:
> >
> > > On Wed, 2009-08-26 at 15:48 -0400, Steven Rostedt wrote:
> > > > \
> > > > @@ -399,16 +413,20 @@ static int ftrace_profile_enable_##call(struct ftrace_event_call *event_call) \
> > > > { \
> > > > int ret = 0; \
> > > > \
> > > > - if (!atomic_inc_return(&event_call->profile_count)) \
> > > > + if (!atomic_inc_return(&event_call->profile_count)) { \
> > > > + event_trace_up_ref(); \
> > > > ret = register_trace_##call(ftrace_profile_##call); \
> > > > + } \
> > > > \
> > > > return ret; \
> > > > } \
> > > > \
> > > > static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\
> > > > { \
> > > > - if (atomic_add_negative(-1, &event_call->profile_count)) \
> > > > + if (atomic_add_negative(-1, &event_call->profile_count)) { \
> > > > unregister_trace_##call(ftrace_profile_##call); \
> > > > + event_trace_down_ref(); \
> > > > + } \
> > > > }
> > >
> > > Ha, almost, make that {un,}register_ftrace_##call() and stick the module
> > > gunk in there, and I'm ok.
> >
> > Peter, that {un,}register_tace_##call is the tracepoint code. The
> > DEFINE_TRACE creates those functions. This is the lowest level we can go
> > in the ftrace.h file.
>
> Did you miss the extra f?

I thought that was a typo since there is no register_ftrace_##call.

>
> You can wrap the regular {un,}register_trace_##call() things in
> {un,}register_/F/trace_##call() functions which also take care of the
> module stuff.

Oh, you mean you want me to create a wrapper for it? Note, this does not
affect the normal trace "enable" code, because there's a module call back
in trace_events.c that disables all the trace points in a module before
unloading. And this in turn will unregister all events.

> Your patch is virtually identical to the one that started this thread,
> everybody using the TRACE_EVENT() things will still need to worry about
> modules.

That's not true. You are not using the TRACE_EVENT, you are modifying the
TRACE_EVENT. That is different. I don't see how a user could hit this
issue. You hit it because you modified what was in ftrace.h, and that does
not count as a user. That's a developer ;-)

-- Steve



\
 
 \ /
  Last update: 2009-08-27 18:21    [W:0.042 / U:0.068 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site