[PATCH] mm, oom_reaper: Update rationale comment for holding oom_lock.

[PATCH] mm, oom_reaper: Update rationale comment for holding oom_lock.

[Tetsuo Handa]

Since commit 862e3073b3eed13f
("mm, oom: get rid of signal_struct::oom_victims")
changed to wait until MMF_OOM_SKIP is set rather than wait while
TIF_MEMDIE is set, rationale comment for commit e2fe14564d3316d1
("oom_reaper: close race with exiting task") needs to be updated.

Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
---
 mm/oom_kill.c | 15 +++------------
 1 file changed, 3 insertions(+), 12 deletions(-)

diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index ec9f11d..6fd076b 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -470,18 +470,9 @@ static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm)
 	bool ret = true;
 
 	/*
-	 * We have to make sure to not race with the victim exit path
-	 * and cause premature new oom victim selection:
-	 * __oom_reap_task_mm		exit_mm
-	 *   mmget_not_zero
-	 *				  mmput
-	 *				    atomic_dec_and_test
-	 *				  exit_oom_victim
-	 *				[...]
-	 *				out_of_memory
-	 *				  select_bad_process
-	 *				    # no TIF_MEMDIE task selects new victim
-	 *  unmap_page_range # frees some memory
+	 * Make sure that other threads waiting for oom_lock at
+	 * __alloc_pages_may_oom() are given a chance to call
+	 * get_page_from_freelist() after MMF_OOM_SKIP is set.
 	 */
 	mutex_lock(&oom_lock);
 
-- 
1.8.3.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH] mm, oom_reaper: Update rationale comment for holding oom_lock.

[Michal Hocko]

On Thu 22-12-16 21:57:30, Tetsuo Handa wrote:
> Since commit 862e3073b3eed13f
> ("mm, oom: get rid of signal_struct::oom_victims")
> changed to wait until MMF_OOM_SKIP is set rather than wait while
> TIF_MEMDIE is set, rationale comment for commit e2fe14564d3316d1
> ("oom_reaper: close race with exiting task") needs to be updated.
> 
> Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
> ---
>  mm/oom_kill.c | 15 +++------------
>  1 file changed, 3 insertions(+), 12 deletions(-)
> 
> diff --git a/mm/oom_kill.c b/mm/oom_kill.c
> index ec9f11d..6fd076b 100644
> --- a/mm/oom_kill.c
> +++ b/mm/oom_kill.c
> @@ -470,18 +470,9 @@ static bool __oom_reap_task_mm(struct task_struct *tsk, struct mm_struct *mm)
>  	bool ret = true;
>  
>  	/*
> -	 * We have to make sure to not race with the victim exit path
> -	 * and cause premature new oom victim selection:
> -	 * __oom_reap_task_mm		exit_mm
> -	 *   mmget_not_zero
> -	 *				  mmput
> -	 *				    atomic_dec_and_test
> -	 *				  exit_oom_victim
> -	 *				[...]
> -	 *				out_of_memory
> -	 *				  select_bad_process
> -	 *				    # no TIF_MEMDIE task selects new victim
> -	 *  unmap_page_range # frees some memory
> +	 * Make sure that other threads waiting for oom_lock at
> +	 * __alloc_pages_may_oom() are given a chance to call
> +	 * get_page_from_freelist() after MMF_OOM_SKIP is set.
>  	 */
>  	mutex_lock(&oom_lock);

I am not sure the comment clarifies things. I would either remove the
comment completely or write something like the below

	/*
	 * Exclude any oom actions while we are reaping the oom
	 * victim. This will save us from pointless searching of the
	 * new oom victim.
	 */

-- 
Michal Hocko
SUSE Labs

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>