[RFC] percpu: use nr_groups as check condition

[RFC] percpu: use nr_groups as check condition

[Peng Fan]

group_cnt array is defined with NR_CPUS entries, but normally
nr_groups will not reach up to NR_CPUS. So there is no issue
to the current code.

Checking other parts of pcpu_build_alloc_info, use nr_groups as
check condition, so make it consistent to use 'group < nr_groups'
as for loop check. In case we do have nr_groups equals with NR_CPUS,
we could also avoid memory access out of bounds.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
---
 mm/percpu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/percpu.c b/mm/percpu.c
index db86282fd024..c5c750781628 100644
--- a/mm/percpu.c
+++ b/mm/percpu.c
@@ -2384,7 +2384,7 @@ static struct pcpu_alloc_info * __init pcpu_build_alloc_info(
 	ai->atom_size = atom_size;
 	ai->alloc_size = alloc_size;
 
-	for (group = 0, unit = 0; group_cnt[group]; group++) {
+	for (group = 0, unit = 0; group < nr_groups; group++) {
 		struct pcpu_group_info *gi = &ai->groups[group];
 
 		/*
-- 
2.16.4

Re: [RFC] percpu: use nr_groups as check condition

[Dennis Zhou]

Hi Peng,

On Wed, Feb 20, 2019 at 01:32:55PM +0000, Peng Fan wrote:
> group_cnt array is defined with NR_CPUS entries, but normally
> nr_groups will not reach up to NR_CPUS. So there is no issue
> to the current code.
> 
> Checking other parts of pcpu_build_alloc_info, use nr_groups as
> check condition, so make it consistent to use 'group < nr_groups'
> as for loop check. In case we do have nr_groups equals with NR_CPUS,
> we could also avoid memory access out of bounds.
> 
> Signed-off-by: Peng Fan <peng.fan@nxp.com>
> ---
>  mm/percpu.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/percpu.c b/mm/percpu.c
> index db86282fd024..c5c750781628 100644
> --- a/mm/percpu.c
> +++ b/mm/percpu.c
> @@ -2384,7 +2384,7 @@ static struct pcpu_alloc_info * __init pcpu_build_alloc_info(
>  	ai->atom_size = atom_size;
>  	ai->alloc_size = alloc_size;
>  
> -	for (group = 0, unit = 0; group_cnt[group]; group++) {
> +	for (group = 0, unit = 0; group < nr_groups; group++) {
>  		struct pcpu_group_info *gi = &ai->groups[group];
>  
>  		/*
> -- 
> 2.16.4
> 

This seems right to me. It is quite the edge case though. I've queued
this for 5.1.

Thanks,
Dennis