Messages in this thread | | | Date | Sun, 28 Nov 2010 23:40:24 +0100 | From | Andi Kleen <> | Subject | Re: [PATCH] kthread: NUMA aware kthread_create_on_cpu() |
| |
On Sun, Nov 28, 2010 at 08:33:53PM +0100, Eric Dumazet wrote: > @@ -101,7 +103,15 @@ static int kthread(void *_create) > static void create_kthread(struct kthread_create_info *create) > { > int pid; > - > + static int last_cpu_pref = -1; > + > + if (create->cpu != last_cpu_pref) {
Is that actually thread-safe?
> +void numa_cpubind_policy(int cpu) > +{ > + nodemask_t mask; > + > + init_nodemask_of_node(&mask, cpu_to_node(cpu)); > + do_set_mempolicy(MPOL_BIND, 0, &mask);
You don't want bind, you want preferred, otherwise this will explode if the node is empty.
Also this messes up the policy of the caller process. You really need to save/restore it.
And if the slab is configured for slab interleaving in the cpuset this will be ignored I think.
Also I think the slab fast path ignores the policy anyways, the policy only acts when slab has to grab new pages. Are you sure this works at all?
It would be probably better to pass through the node to the low level allocation functions and use them there directly.
Problem is that this ends up in architecture specific code for the stack, so may be a larger patch.
-Andi -- ak@linux.intel.com -- Speaking for myself only.
| |