Re: [PATCH] percpu_counter: Fix __percpu_counter_sum()

[Peter Zijlstra <a.p.zijlstra@chello.nl>]

On Mon, 2008-12-08 at 17:12 -0500, Theodore Tso wrote:
> On Sun, Dec 07, 2008 at 08:52:50PM -0800, Andrew Morton wrote:
> > 
> > The first patch which was added (pre-2.6.27) was "percpu_counter: new
> > function percpu_counter_sum_and_set".  This added the broken-by-design
> > percpu_counter_sum_and_set() function, **and used it in ext4**.
> > 
> 
> Mea culpa, I was the one who reviewed Mingming's patch, and missed
> this.  Part of the problem was that percpu_counter.c isn't well
> documented, and I so saw the spinlock, but didn't realize it only
> protected reference counter, and not the per-cpu array.  I should have
> read through code more thoroughly before approving the patch.
> 
> I suppose if we wanted we could add a rw spinlock which mediates
> access to a "foreign" cpu counter (i.e., percpu_counter_add gets a
> shared lock, and percpu_counter_set needs an exclusive lock) but it's
> probably not worth it.

rwlocks are utter suck and should be banished from the kernel - adding
one would destroy the whole purpose of the code.

> Actually, if all popular architectures had a hardware-implemented
> atomic_t, I wonder how much ext4 really needs the percpu counter,
> especially given ext4's multiblock allocator; with ext3, given that
> each block allocation required taking a per-filesystem spin lock,
> optimizing away that spinlock was far more important for improving
> ext3's scalability.  But with the multiblock allocator, it may that
> we're going through a lot more effort than what is truly necessary.

atomic_t is pretty good on all archs, but you get to keep the cacheline
ping-pong.