Re: likely unlikely annotation (spinlock status)

[Steven Rostedt]

On Fri, 2011-01-07 at 15:57 +0530, manikanta wrote:
> Hi Rostedt,
>  
> Problem:    
> Likely and unlikely annotation branch summary is showing
> status for spinlocks(below is the result of enabling
> CONFIG_PROFILE_ANNOTATED_BRANCHES).
> Is it right(eventhough at line no 6753 i can see spin_unlock(lock))???

Your spin_unlock is probably a macro which contains an "if (likely())"
in it. If this is the case then it would definitely show a location of
the spin lock, because that spinlock is nothing more than an if
statement to gcc.

>  
> 0       19 100 __cond_resched_lock            sched.c
> 6753
> 0        0   X   dsp_event                             audio_xwb.c
> 355
>  
> I am seeing the same results for my drivers also i.e. where we have
> used spinlocks.
> It is showing the spinlocks results.At line number 355 there is a
> spinlock call in file
> audio_xwb.c.
>  
> Is it right to show spinlock status instead of likely and unlikely
> notations???

Yes, if the spinlocks are macros with likely/unlikely annotations.

>  
> Setup:Android + mach-msm

Unfortunately, I don't have this setup to look at it.

-- Steve



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/