Messages in this thread | | | Subject | Re: [PATCH] percpu_counter: Fix __percpu_counter_sum() | From | Peter Zijlstra <> | Date | Wed, 10 Dec 2008 08:12:42 +0100 |
| |
On Wed, 2008-12-10 at 06:09 +0100, Eric Dumazet wrote: > Now percpu_counter_sum() is 'fixed', what about "percpu_counter_add()" ? > > void __percpu_counter_add(struct percpu_counter *fbc, s64 amount, s32 batch) > { > s64 count; > s32 *pcount; > int cpu = get_cpu(); > > pcount = per_cpu_ptr(fbc->counters, cpu); > count = *pcount + amount; > if (count >= batch || count <= -batch) { > spin_lock(&fbc->lock); > fbc->count += count; > *pcount = 0; > spin_unlock(&fbc->lock); > } else { > *pcount = count; > } > put_cpu(); > } > > > If I read this well, this is not IRQ safe. > > get_cpu() only disables preemption IMHO > > For nr_files, nr_dentry, nr_inodes, it should not be a problem. > > But for network counters (only in net-next-2.6) > and lib/proportions.c, we have a problem ?
Non of percpu_counter if irqsafe, for lib/proportions I disabled irqs by hand when needed - I don't think we ought to bother with local_t, esp as it basically sucks chunks on anything !x86.
| |