From: Qi Zheng <zhengqi.arch@bytedance.com>
The __mod_memcg_state() and __mod_memcg_lruvec_state() functions are also
used to reparent non-hierarchical stats. In this scenario, the values
passed to them are accumulated statistics that might be extremely large
and exceed the upper limit of a 32-bit integer.
Change the val parameter type from int to long in these functions and
their corresponding tracepoints (memcg_rstat_stats) to prevent potential
overflow issues.
After that, in memcg_state_val_in_pages(), if the passed val is negative,
the expression val * unit / PAGE_SIZE could be implicitly converted to a
massive positive number when compared with 1UL in the max() macro.
This leads to returning an incorrect massive positive value.
Fix this by using abs(val) to calculate the magnitude first, and then
restoring the sign of the value before returning the result. Additionally,
use mult_frac() to prevent potential overflow during the multiplication of
val and unit.
Reported-by: Harry Yoo (Oracle) <harry@kernel.org>
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
---
include/trace/events/memcg.h | 10 +++++-----
mm/memcontrol.c | 18 ++++++++++++------
2 files changed, 17 insertions(+), 11 deletions(-)
diff --git a/include/trace/events/memcg.h b/include/trace/events/memcg.h
index dfe2f51019b4c..51b62c5931fc2 100644
--- a/include/trace/events/memcg.h
+++ b/include/trace/events/memcg.h
@@ -11,14 +11,14 @@
DECLARE_EVENT_CLASS(memcg_rstat_stats,
- TP_PROTO(struct mem_cgroup *memcg, int item, int val),
+ TP_PROTO(struct mem_cgroup *memcg, int item, long val),
TP_ARGS(memcg, item, val),
TP_STRUCT__entry(
__field(u64, id)
__field(int, item)
- __field(int, val)
+ __field(long, val)
),
TP_fast_assign(
@@ -27,20 +27,20 @@ DECLARE_EVENT_CLASS(memcg_rstat_stats,
__entry->val = val;
),
- TP_printk("memcg_id=%llu item=%d val=%d",
+ TP_printk("memcg_id=%llu item=%d val=%ld",
__entry->id, __entry->item, __entry->val)
);
DEFINE_EVENT(memcg_rstat_stats, mod_memcg_state,
- TP_PROTO(struct mem_cgroup *memcg, int item, int val),
+ TP_PROTO(struct mem_cgroup *memcg, int item, long val),
TP_ARGS(memcg, item, val)
);
DEFINE_EVENT(memcg_rstat_stats, mod_memcg_lruvec_state,
- TP_PROTO(struct mem_cgroup *memcg, int item, int val),
+ TP_PROTO(struct mem_cgroup *memcg, int item, long val),
TP_ARGS(memcg, item, val)
);
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 3daab9b46429a..51d72ddf08119 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -527,7 +527,7 @@ unsigned long lruvec_page_state_local(struct lruvec *lruvec,
#ifdef CONFIG_MEMCG_V1
static void __mod_memcg_lruvec_state(struct mem_cgroup_per_node *pn,
- enum node_stat_item idx, int val);
+ enum node_stat_item idx, long val);
void reparent_memcg_lruvec_state_local(struct mem_cgroup *memcg,
struct mem_cgroup *parent, int idx)
@@ -784,14 +784,20 @@ static int memcg_page_state_unit(int item);
* Normalize the value passed into memcg_rstat_updated() to be in pages. Round
* up non-zero sub-page updates to 1 page as zero page updates are ignored.
*/
-static int memcg_state_val_in_pages(int idx, int val)
+static long memcg_state_val_in_pages(int idx, long val)
{
int unit = memcg_page_state_unit(idx);
+ long res;
if (!val || unit == PAGE_SIZE)
return val;
- else
- return max(val * unit / PAGE_SIZE, 1UL);
+
+ /* Get the absolute value of (val * unit / PAGE_SIZE). */
+ res = mult_frac(abs(val), unit, PAGE_SIZE);
+ /* Round up zero values. */
+ res = res ? : 1;
+
+ return val < 0 ? -res : res;
}
#ifdef CONFIG_MEMCG_V1
@@ -831,7 +837,7 @@ static inline void get_non_dying_memcg_end(void)
#endif
static void __mod_memcg_state(struct mem_cgroup *memcg,
- enum memcg_stat_item idx, int val)
+ enum memcg_stat_item idx, long val)
{
int i = memcg_stats_index(idx);
int cpu;
@@ -896,7 +902,7 @@ void reparent_memcg_state_local(struct mem_cgroup *memcg,
#endif
static void __mod_memcg_lruvec_state(struct mem_cgroup_per_node *pn,
- enum node_stat_item idx, int val)
+ enum node_stat_item idx, long val)
{
struct mem_cgroup *memcg = pn->memcg;
int i = memcg_stats_index(idx);
--
2.20.1On Fri, Mar 27, 2026 at 06:16:29PM +0800, Qi Zheng wrote:
> From: Qi Zheng <zhengqi.arch@bytedance.com>
>
> The __mod_memcg_state() and __mod_memcg_lruvec_state() functions are also
> used to reparent non-hierarchical stats. In this scenario, the values
> passed to them are accumulated statistics that might be extremely large
> and exceed the upper limit of a 32-bit integer.
>
> Change the val parameter type from int to long in these functions and
> their corresponding tracepoints (memcg_rstat_stats) to prevent potential
> overflow issues.
>
> After that, in memcg_state_val_in_pages(), if the passed val is negative,
> the expression val * unit / PAGE_SIZE could be implicitly converted to a
> massive positive number when compared with 1UL in the max() macro.
> This leads to returning an incorrect massive positive value.
>
> Fix this by using abs(val) to calculate the magnitude first, and then
> restoring the sign of the value before returning the result. Additionally,
> use mult_frac() to prevent potential overflow during the multiplication of
> val and unit.
>
> Reported-by: Harry Yoo (Oracle) <harry@kernel.org>
> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
> Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
> ---
Looks good to me,
Reviewed-by: Harry Yoo (Oracle) <harry@kernel.org>
> @@ -831,7 +837,7 @@ static inline void get_non_dying_memcg_end(void)
> #endif
>
> static void __mod_memcg_state(struct mem_cgroup *memcg,
> - enum memcg_stat_item idx, int val)
> + enum memcg_stat_item idx, long val)
> {
> int i = memcg_stats_index(idx);
> int cpu;
> @@ -896,7 +902,7 @@ void reparent_memcg_state_local(struct mem_cgroup *memcg,
> #endif
>
> static void __mod_memcg_lruvec_state(struct mem_cgroup_per_node *pn,
> - enum node_stat_item idx, int val)
> + enum node_stat_item idx, long val)
> {
> struct mem_cgroup *memcg = pn->memcg;
> int i = memcg_stats_index(idx);
Some of code paths that calls mod_memcg{,_lruvec}_state still passes
int values (which is quite subtle to notice), but it should be fine
as reparenting is not involved in the path and could be cleaned up later.
--
Cheers,
Harry / Hyeonggon
On 3/30/26 9:25 AM, Harry Yoo (Oracle) wrote:
> On Fri, Mar 27, 2026 at 06:16:29PM +0800, Qi Zheng wrote:
>> From: Qi Zheng <zhengqi.arch@bytedance.com>
>>
>> The __mod_memcg_state() and __mod_memcg_lruvec_state() functions are also
>> used to reparent non-hierarchical stats. In this scenario, the values
>> passed to them are accumulated statistics that might be extremely large
>> and exceed the upper limit of a 32-bit integer.
>>
>> Change the val parameter type from int to long in these functions and
>> their corresponding tracepoints (memcg_rstat_stats) to prevent potential
>> overflow issues.
>>
>> After that, in memcg_state_val_in_pages(), if the passed val is negative,
>> the expression val * unit / PAGE_SIZE could be implicitly converted to a
>> massive positive number when compared with 1UL in the max() macro.
>> This leads to returning an incorrect massive positive value.
>>
>> Fix this by using abs(val) to calculate the magnitude first, and then
>> restoring the sign of the value before returning the result. Additionally,
>> use mult_frac() to prevent potential overflow during the multiplication of
>> val and unit.
>>
>> Reported-by: Harry Yoo (Oracle) <harry@kernel.org>
>> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
>> Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
>> ---
>
> Looks good to me,
> Reviewed-by: Harry Yoo (Oracle) <harry@kernel.org>
Thanks!
>
>> @@ -831,7 +837,7 @@ static inline void get_non_dying_memcg_end(void)
>> #endif
>>
>> static void __mod_memcg_state(struct mem_cgroup *memcg,
>> - enum memcg_stat_item idx, int val)
>> + enum memcg_stat_item idx, long val)
>> {
>> int i = memcg_stats_index(idx);
>> int cpu;
>> @@ -896,7 +902,7 @@ void reparent_memcg_state_local(struct mem_cgroup *memcg,
>> #endif
>>
>> static void __mod_memcg_lruvec_state(struct mem_cgroup_per_node *pn,
>> - enum node_stat_item idx, int val)
>> + enum node_stat_item idx, long val)
>> {
>> struct mem_cgroup *memcg = pn->memcg;
>> int i = memcg_stats_index(idx);
>
> Some of code paths that calls mod_memcg{,_lruvec}_state still passes
> int values (which is quite subtle to notice), but it should be fine
Right, it happens in too many places, for example, the callers of
mod_lruvec_state().
> as reparenting is not involved in the path and could be cleaned up later.
Agree.
>
On 27 Mar 2026, at 6:16, Qi Zheng wrote: > From: Qi Zheng <zhengqi.arch@bytedance.com> > > The __mod_memcg_state() and __mod_memcg_lruvec_state() functions are also > used to reparent non-hierarchical stats. In this scenario, the values > passed to them are accumulated statistics that might be extremely large > and exceed the upper limit of a 32-bit integer. > > Change the val parameter type from int to long in these functions and > their corresponding tracepoints (memcg_rstat_stats) to prevent potential > overflow issues. > > After that, in memcg_state_val_in_pages(), if the passed val is negative, > the expression val * unit / PAGE_SIZE could be implicitly converted to a > massive positive number when compared with 1UL in the max() macro. > This leads to returning an incorrect massive positive value. > > Fix this by using abs(val) to calculate the magnitude first, and then > restoring the sign of the value before returning the result. Additionally, > use mult_frac() to prevent potential overflow during the multiplication of > val and unit. > > Reported-by: Harry Yoo (Oracle) <harry@kernel.org> > Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com> > Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org> > --- > include/trace/events/memcg.h | 10 +++++----- > mm/memcontrol.c | 18 ++++++++++++------ > 2 files changed, 17 insertions(+), 11 deletions(-) > Acked-by: Zi Yan <ziy@nvidia.com> Best Regards, Yan, Zi
© 2016 - 2026 Red Hat, Inc.