The /proc/vmstat displays counters from various sources. It is easy to
forget to add or remove a label when a new counter is added or removed.

There is a BUILD_BUG_ON() function that catches missing labels. However,
for some reason, it ignores extra labels.

Let's make the check strict. This would help to catch issues when
a counter is removed.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Konstantin Khlebnikov <koct9i@gmail.com>
---
 mm/vmstat.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/vmstat.c b/mm/vmstat.c
index ae9882063d89..0903adace423 100644
--- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -1867,7 +1867,7 @@ static void *vmstat_start(struct seq_file *m, loff_t *pos)
 	if (*pos >= NR_VMSTAT_ITEMS)
 		return NULL;
 
-	BUILD_BUG_ON(ARRAY_SIZE(vmstat_text) < NR_VMSTAT_ITEMS);
+	BUILD_BUG_ON(ARRAY_SIZE(vmstat_text) != NR_VMSTAT_ITEMS);
 	fold_vm_numa_events();
 	v = kmalloc_array(NR_VMSTAT_ITEMS, sizeof(unsigned long), GFP_KERNEL);
 	m->private = v;
-- 
2.47.2

On 29.05.25 13:05, Kirill A. Shutemov wrote:
> The /proc/vmstat displays counters from various sources. It is easy to
> forget to add or remove a label when a new counter is added or removed.
> 
> There is a BUILD_BUG_ON() function that catches missing labels. However,
> for some reason, it ignores extra labels.
> 
> Let's make the check strict. This would help to catch issues when
> a counter is removed.
> 
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> Cc: Konstantin Khlebnikov <koct9i@gmail.com>
> ---

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 
Cheers,

David / dhildenb

On Thu 29-05-25 14:05:41, Kirill A. Shutemov wrote:
> The /proc/vmstat displays counters from various sources. It is easy to
> forget to add or remove a label when a new counter is added or removed.
> 
> There is a BUILD_BUG_ON() function that catches missing labels. However,
> for some reason, it ignores extra labels.
> 
> Let's make the check strict. This would help to catch issues when
> a counter is removed.
> 
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> Cc: Konstantin Khlebnikov <koct9i@gmail.com>

LGTM
Acked-by: Michal Hocko <mhocko@suse.com>

Not sure why we have gone with the current check TBH.

> ---
>  mm/vmstat.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/vmstat.c b/mm/vmstat.c
> index ae9882063d89..0903adace423 100644
> --- a/mm/vmstat.c
> +++ b/mm/vmstat.c
> @@ -1867,7 +1867,7 @@ static void *vmstat_start(struct seq_file *m, loff_t *pos)
>  	if (*pos >= NR_VMSTAT_ITEMS)
>  		return NULL;
>  
> -	BUILD_BUG_ON(ARRAY_SIZE(vmstat_text) < NR_VMSTAT_ITEMS);
> +	BUILD_BUG_ON(ARRAY_SIZE(vmstat_text) != NR_VMSTAT_ITEMS);
>  	fold_vm_numa_events();
>  	v = kmalloc_array(NR_VMSTAT_ITEMS, sizeof(unsigned long), GFP_KERNEL);
>  	m->private = v;
> -- 
> 2.47.2

-- 
Michal Hocko
SUSE Labs

On 5/29/25 13:05, Kirill A. Shutemov wrote:
> The /proc/vmstat displays counters from various sources. It is easy to
> forget to add or remove a label when a new counter is added or removed.
> 
> There is a BUILD_BUG_ON() function that catches missing labels. However,
> for some reason, it ignores extra labels.
> 
> Let's make the check strict. This would help to catch issues when
> a counter is removed.
> 
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> Cc: Konstantin Khlebnikov <koct9i@gmail.com>

Yeah I don't see why it would be necessary to check only for missing labels.

Reviewed-by: Vlastimil Babka <vbabka@suse.cz>

> ---
>  mm/vmstat.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/vmstat.c b/mm/vmstat.c
> index ae9882063d89..0903adace423 100644
> --- a/mm/vmstat.c
> +++ b/mm/vmstat.c
> @@ -1867,7 +1867,7 @@ static void *vmstat_start(struct seq_file *m, loff_t *pos)
>  	if (*pos >= NR_VMSTAT_ITEMS)
>  		return NULL;
>  
> -	BUILD_BUG_ON(ARRAY_SIZE(vmstat_text) < NR_VMSTAT_ITEMS);
> +	BUILD_BUG_ON(ARRAY_SIZE(vmstat_text) != NR_VMSTAT_ITEMS);
>  	fold_vm_numa_events();
>  	v = kmalloc_array(NR_VMSTAT_ITEMS, sizeof(unsigned long), GFP_KERNEL);
>  	m->private = v;