Re: [PATCH sched-devel] ftrace: trace_entries to change trace buffersize

From: Steven Rostedt
Date: Fri Apr 18 2008 - 15:25:29 EST



On Fri, 18 Apr 2008, Andrew Morton wrote:
> >
> > Actually, that's just a way of doing spin_lock_init for raw spin locks. I
> > didn't see any other way.
>
> Maybe we need raw_spin_lock_init(). And DEFINE_RAW_SPINLOCK()?
>
> Although it isn't a great idea to be encouraging the use of raw
> spinlocks.
>

I was thinking the same thing. Raw spin_locks should not be used, unless
there's a really good reason to use them. Strange things like mcount is
one example, but I cant think of many more. I'd be hesitant to add such an
API, which might lead to other users of raw_spin_locks when they really
shouldn't.

Not to mention that we also have raw_spinlock_t in -rt that has a little
different meaning.

-- 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/