Re: [patch -mm] cpusets: add memory_slab_hardwall flag

[Paul Menage <menage@google.com>]

Another thought - it would probably be better to call this flag
kernel_mem_hardwall or mem_hardwall_kernel, to avoid hard-coding its
name to be slab-specific.

Paul

On Sun, Mar 8, 2009 at 2:38 PM, David Rientjes <rientjes@google.com> wrote:
> On Sun, 8 Mar 2009, Paul Menage wrote:
>
>> > +/**
>> > + * current_cpuset_object_allowed - can a slab object be allocated on a node?
>> > + * @node: the node for object allocation
>> > + * @flags: allocation flags
>> > + *
>> > + * Return non-zero if object is allowed, zero otherwise.
>> > + */
>> > +int current_cpuset_object_allowed(int node, gfp_t flags)
>> > +{
>> > +       return !is_slab_hardwall(task_cs(current)) ||
>> > +              cpuset_node_allowed_hardwall(node, flags);
>> > +}
>> > +
>>
>> This should be in rcu_read_lock()/rcu_read_unlock() in order to safely
>> dereference the result of task_cs(current)
>>
>
> I've folded the following into the patch, thanks Paul.
> ---
>  kernel/cpuset.c |    8 ++++++--
>  1 files changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/cpuset.c b/kernel/cpuset.c
> --- a/kernel/cpuset.c
> +++ b/kernel/cpuset.c
> @@ -1205,8 +1205,12 @@ int current_cpuset_is_being_rebound(void)
>  */
>  int current_cpuset_object_allowed(int node, gfp_t flags)
>  {
> -       return !is_slab_hardwall(task_cs(current)) ||
> -              cpuset_node_allowed_hardwall(node, flags);
> +       int is_hardwall;
> +
> +       rcu_read_lock();
> +       is_hardwall = is_slab_hardwall(task_cs(current));
> +       rcu_read_unlock();
> +       return !is_hardwall || cpuset_node_allowed_hardwall(node, flags);
>  }
>
>  static int update_relax_domain_level(struct cpuset *cs, s64 val)