OOM killer is a mechanism that selects and kills processes when the system
runs out of memory to reclaim resources and keep the system stable.
However, the oom victim cannot terminate on its own when it is frozen,
because __thaw_task() only thaws one thread of the victim, while
the other threads remain in the frozen state.
Since __thaw_task did not fully thaw the OOM victim for self-termination,
introduce thaw_oom_process() to properly thaw OOM victims.
Signed-off-by: zhongjinji <zhongjinji@honor.com>
---
include/linux/freezer.h | 2 ++
kernel/freezer.c | 19 +++++++++++++++++++
2 files changed, 21 insertions(+)
diff --git a/include/linux/freezer.h b/include/linux/freezer.h
index b303472255be..19a4b57950cd 100644
--- a/include/linux/freezer.h
+++ b/include/linux/freezer.h
@@ -47,6 +47,7 @@ extern int freeze_processes(void);
extern int freeze_kernel_threads(void);
extern void thaw_processes(void);
extern void thaw_kernel_threads(void);
+extern void thaw_oom_process(struct task_struct *p);
static inline bool try_to_freeze(void)
{
@@ -80,6 +81,7 @@ static inline int freeze_processes(void) { return -ENOSYS; }
static inline int freeze_kernel_threads(void) { return -ENOSYS; }
static inline void thaw_processes(void) {}
static inline void thaw_kernel_threads(void) {}
+static inline void thaw_oom_process(struct task_struct *p) {}
static inline bool try_to_freeze(void) { return false; }
diff --git a/kernel/freezer.c b/kernel/freezer.c
index 6a96149aede9..17970e0be8a7 100644
--- a/kernel/freezer.c
+++ b/kernel/freezer.c
@@ -206,6 +206,25 @@ void __thaw_task(struct task_struct *p)
wake_up_state(p, TASK_FROZEN);
}
+/*
+ * thaw_oom_process - thaw the OOM victim process
+ * @p: process to be thawed
+ *
+ * Sets TIF_MEMDIE for all threads in the process group and thaws them.
+ * Threads with TIF_MEMDIE are ignored by the freezer.
+ */
+void thaw_oom_process(struct task_struct *p)
+{
+ struct task_struct *t;
+
+ rcu_read_lock();
+ for_each_thread(p, t) {
+ set_tsk_thread_flag(t, TIF_MEMDIE);
+ __thaw_task(t);
+ }
+ rcu_read_unlock();
+}
+
/**
* set_freezable - make %current freezable
*
--
2.17.1
On Tue 09-09-25 17:06:57, zhongjinji wrote: > OOM killer is a mechanism that selects and kills processes when the system > runs out of memory to reclaim resources and keep the system stable. > However, the oom victim cannot terminate on its own when it is frozen, > because __thaw_task() only thaws one thread of the victim, while > the other threads remain in the frozen state. > > Since __thaw_task did not fully thaw the OOM victim for self-termination, > introduce thaw_oom_process() to properly thaw OOM victims. You will need s@thaw_oom_process@thaw_processes@ I would also add the caller in this patch. > Signed-off-by: zhongjinji <zhongjinji@honor.com> Other than that looks good to me. With the above fixed feel free to add Acked-by: Michal Hocko <mhocko@suse.com> > --- > include/linux/freezer.h | 2 ++ > kernel/freezer.c | 19 +++++++++++++++++++ > 2 files changed, 21 insertions(+) > > diff --git a/include/linux/freezer.h b/include/linux/freezer.h > index b303472255be..19a4b57950cd 100644 > --- a/include/linux/freezer.h > +++ b/include/linux/freezer.h > @@ -47,6 +47,7 @@ extern int freeze_processes(void); > extern int freeze_kernel_threads(void); > extern void thaw_processes(void); > extern void thaw_kernel_threads(void); > +extern void thaw_oom_process(struct task_struct *p); > > static inline bool try_to_freeze(void) > { > @@ -80,6 +81,7 @@ static inline int freeze_processes(void) { return -ENOSYS; } > static inline int freeze_kernel_threads(void) { return -ENOSYS; } > static inline void thaw_processes(void) {} > static inline void thaw_kernel_threads(void) {} > +static inline void thaw_oom_process(struct task_struct *p) {} > > static inline bool try_to_freeze(void) { return false; } > > diff --git a/kernel/freezer.c b/kernel/freezer.c > index 6a96149aede9..17970e0be8a7 100644 > --- a/kernel/freezer.c > +++ b/kernel/freezer.c > @@ -206,6 +206,25 @@ void __thaw_task(struct task_struct *p) > wake_up_state(p, TASK_FROZEN); > } > > +/* > + * thaw_oom_process - thaw the OOM victim process > + * @p: process to be thawed > + * > + * Sets TIF_MEMDIE for all threads in the process group and thaws them. > + * Threads with TIF_MEMDIE are ignored by the freezer. > + */ > +void thaw_oom_process(struct task_struct *p) > +{ > + struct task_struct *t; > + > + rcu_read_lock(); > + for_each_thread(p, t) { > + set_tsk_thread_flag(t, TIF_MEMDIE); > + __thaw_task(t); > + } > + rcu_read_unlock(); > +} > + > /** > * set_freezable - make %current freezable > * > -- > 2.17.1 -- Michal Hocko SUSE Labs
On Tue, Sep 9, 2025 at 2:15 AM Michal Hocko <mhocko@suse.com> wrote: > > On Tue 09-09-25 17:06:57, zhongjinji wrote: > > OOM killer is a mechanism that selects and kills processes when the system > > runs out of memory to reclaim resources and keep the system stable. > > However, the oom victim cannot terminate on its own when it is frozen, > > because __thaw_task() only thaws one thread of the victim, while > > the other threads remain in the frozen state. > > > > Since __thaw_task did not fully thaw the OOM victim for self-termination, > > introduce thaw_oom_process() to properly thaw OOM victims. > > You will need s@thaw_oom_process@thaw_processes@ Do you suggest renaming thaw_oom_process() into thaw_processes() (s/thaw_oom_process/thaw_processes)? If so, I don't think that's a better name considering the function sets TIF_MEMDIE flag. From that perspective less generic thaw_oom_process() seems appropriate, no? > > I would also add the caller in this patch. > > > Signed-off-by: zhongjinji <zhongjinji@honor.com> > > Other than that looks good to me. With the above fixed feel free to add > Acked-by: Michal Hocko <mhocko@suse.com> > > > --- > > include/linux/freezer.h | 2 ++ > > kernel/freezer.c | 19 +++++++++++++++++++ > > 2 files changed, 21 insertions(+) > > > > diff --git a/include/linux/freezer.h b/include/linux/freezer.h > > index b303472255be..19a4b57950cd 100644 > > --- a/include/linux/freezer.h > > +++ b/include/linux/freezer.h > > @@ -47,6 +47,7 @@ extern int freeze_processes(void); > > extern int freeze_kernel_threads(void); > > extern void thaw_processes(void); > > extern void thaw_kernel_threads(void); > > +extern void thaw_oom_process(struct task_struct *p); > > > > static inline bool try_to_freeze(void) > > { > > @@ -80,6 +81,7 @@ static inline int freeze_processes(void) { return -ENOSYS; } > > static inline int freeze_kernel_threads(void) { return -ENOSYS; } > > static inline void thaw_processes(void) {} > > static inline void thaw_kernel_threads(void) {} > > +static inline void thaw_oom_process(struct task_struct *p) {} > > > > static inline bool try_to_freeze(void) { return false; } > > > > diff --git a/kernel/freezer.c b/kernel/freezer.c > > index 6a96149aede9..17970e0be8a7 100644 > > --- a/kernel/freezer.c > > +++ b/kernel/freezer.c > > @@ -206,6 +206,25 @@ void __thaw_task(struct task_struct *p) > > wake_up_state(p, TASK_FROZEN); > > } > > > > +/* > > + * thaw_oom_process - thaw the OOM victim process > > + * @p: process to be thawed > > + * > > + * Sets TIF_MEMDIE for all threads in the process group and thaws them. > > + * Threads with TIF_MEMDIE are ignored by the freezer. > > + */ > > +void thaw_oom_process(struct task_struct *p) > > +{ > > + struct task_struct *t; > > + > > + rcu_read_lock(); > > + for_each_thread(p, t) { > > + set_tsk_thread_flag(t, TIF_MEMDIE); > > + __thaw_task(t); > > + } > > + rcu_read_unlock(); > > +} > > + > > /** > > * set_freezable - make %current freezable > > * > > -- > > 2.17.1 > > -- > Michal Hocko > SUSE Labs
On Tue 09-09-25 09:27:54, Suren Baghdasaryan wrote: > On Tue, Sep 9, 2025 at 2:15 AM Michal Hocko <mhocko@suse.com> wrote: > > > > On Tue 09-09-25 17:06:57, zhongjinji wrote: > > > OOM killer is a mechanism that selects and kills processes when the system > > > runs out of memory to reclaim resources and keep the system stable. > > > However, the oom victim cannot terminate on its own when it is frozen, > > > because __thaw_task() only thaws one thread of the victim, while > > > the other threads remain in the frozen state. > > > > > > Since __thaw_task did not fully thaw the OOM victim for self-termination, > > > introduce thaw_oom_process() to properly thaw OOM victims. > > > > You will need s@thaw_oom_process@thaw_processes@ > > Do you suggest renaming thaw_oom_process() into thaw_processes() > (s/thaw_oom_process/thaw_processes)? If so, I don't think that's a > better name considering the function sets TIF_MEMDIE flag. From that > perspective less generic thaw_oom_process() seems appropriate, no? Please see the discussion for the patch 2. TL;DR yes rename and drop TIF_MEMDIE part and update freezer to check tsk_is_oom_victim rather than TIF_MEMDIE. -- Michal Hocko SUSE Labs
On Tue, Sep 9, 2025 at 9:44 AM Michal Hocko <mhocko@suse.com> wrote: > > On Tue 09-09-25 09:27:54, Suren Baghdasaryan wrote: > > On Tue, Sep 9, 2025 at 2:15 AM Michal Hocko <mhocko@suse.com> wrote: > > > > > > On Tue 09-09-25 17:06:57, zhongjinji wrote: > > > > OOM killer is a mechanism that selects and kills processes when the system > > > > runs out of memory to reclaim resources and keep the system stable. > > > > However, the oom victim cannot terminate on its own when it is frozen, > > > > because __thaw_task() only thaws one thread of the victim, while > > > > the other threads remain in the frozen state. > > > > > > > > Since __thaw_task did not fully thaw the OOM victim for self-termination, > > > > introduce thaw_oom_process() to properly thaw OOM victims. > > > > > > You will need s@thaw_oom_process@thaw_processes@ > > > > Do you suggest renaming thaw_oom_process() into thaw_processes() > > (s/thaw_oom_process/thaw_processes)? If so, I don't think that's a > > better name considering the function sets TIF_MEMDIE flag. From that > > perspective less generic thaw_oom_process() seems appropriate, no? > > Please see the discussion for the patch 2. > TL;DR yes rename and drop TIF_MEMDIE part and update freezer to check > tsk_is_oom_victim rather than TIF_MEMDIE. Oh, sorry. For some reason that part of the email thread ended up as a separate email in my mailbox and I missed it. Your suggestion there sounds great. > > -- > Michal Hocko > SUSE Labs
© 2016 - 2025 Red Hat, Inc.