Re: [PATCH v4] numa: make node_to_cpumask_map() NUMA_NO_NODE aware

[Michal Hocko]

On Sun 15-09-19 16:20:56, Yunsheng Lin wrote:
> When passing the return value of dev_to_node() to cpumask_of_node()
> without checking if the device's node id is NUMA_NO_NODE, there is
> global-out-of-bounds detected by KASAN.
> 
> >From the discussion [1], NUMA_NO_NODE really means no node affinity,
> which also means all cpus should be usable. So the cpumask_of_node()
> should always return all cpus online when user passes the node id as
> NUMA_NO_NODE, just like similar semantic that page allocator handles
> NUMA_NO_NODE.
> 
> But we cannot really copy the page allocator logic. Simply because the
> page allocator doesn't enforce the near node affinity. It just picks it
> up as a preferred node but then it is free to fallback to any other numa
> node. This is not the case here and node_to_cpumask_map will only restrict
> to the particular node's cpus which would have really non deterministic
> behavior depending on where the code is executed. So in fact we really
> want to return cpu_online_mask for NUMA_NO_NODE.
> 
> Some arches were already NUMA_NO_NODE aware, so only change them to return
> cpu_online_mask and use NUMA_NO_NODE instead of "-1".
> 
> Also there is a debugging version of node_to_cpumask_map() for x86 and
> arm64, which is only used when CONFIG_DEBUG_PER_CPU_MAPS is defined, this
> patch changes it to handle NUMA_NO_NODE as normal node_to_cpumask_map().
> And "fix" a sign "bug" since it is for debugging and should catch all the
> error cases.
> 
> [1] https://lore.kernel.org/patchwork/patch/1125789/
> Signed-off-by: Yunsheng Lin <linyunsheng@xxxxxxxxxx>
> Suggested-by: Michal Hocko <mhocko@xxxxxxxxxx>

The change makes sense to me. I wish this particular thing wasn't
duplicated so heavily - maybe we can unify all of them and use a common
code? In a separate patch most likely...

I would also not change cpu_all_mask -> cpu_online_mask in this patch.
That is worth a patch on its own with some explanation. I haven't
checked but I would suspect that alpha simply doesn't support cpu
hotplug so the two things are the same. But this needs some explanation.

Other than that the patch looks good to me. Feel free to add
Acked-by: Michal Hocko <mhocko@xxxxxxxx>

[...]
> diff --git a/arch/alpha/include/asm/topology.h b/arch/alpha/include/asm/topology.h
> index 5a77a40..836c9e2 100644
> --- a/arch/alpha/include/asm/topology.h
> +++ b/arch/alpha/include/asm/topology.h
> @@ -31,7 +31,7 @@ static const struct cpumask *cpumask_of_node(int node)
>  	int cpu;
>  
>  	if (node == NUMA_NO_NODE)
> -		return cpu_all_mask;
> +		return cpu_online_mask;
>  
>  	cpumask_clear(&node_to_cpumask_map[node]);
>  
[...]

> diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c
> index e6dad60..c676ffb 100644
> --- a/arch/x86/mm/numa.c
> +++ b/arch/x86/mm/numa.c
> @@ -861,9 +861,12 @@ void numa_remove_cpu(int cpu)
>   */
>  const struct cpumask *cpumask_of_node(int node)
>  {
> -	if (node >= nr_node_ids) {
> +	if (node == NUMA_NO_NODE)
> +		return cpu_online_mask;
> +
> +	if ((unsigned int)node >= nr_node_ids) {
>  		printk(KERN_WARNING
> -			"cpumask_of_node(%d): node > nr_node_ids(%u)\n",
> +			"cpumask_of_node(%d): node >= nr_node_ids(%u)\n",
>  			node, nr_node_ids);
>  		dump_stack();
>  		return cpu_none_mask;

Why do we need this?
-- 
Michal Hocko
SUSE Labs