2020-02-23 19:26:46

by Andrea Arcangeli

Subject: [PATCH 1/3] mm: use_mm: fix for arches checking mm_users to optimize TLB flushes

alpha, ia64, mips, powerpc, sh, sparc are relying on a check on
mm->mm_users to know if they can skip some remote TLB flushes for
single threaded processes.

Most callers of use_mm() tend to invoke mmget_not_zero() or
get_task_mm() before use_mm() to ensure the mm will remain alive in
between use_mm() and unuse_mm().

Some callers however don't increase mm_users and they instead rely on
serialization in __mmput() to ensure the mm will remain alive in
between use_mm() and unuse_mm(). Not increasing mm_users during
use_mm() is however unsafe for aforementioned arch TLB flushes
optimizations. So either mmget()/mmput() should be added to the
problematic callers of use_mm()/unuse_mm() or we can embed them in
use_mm()/unuse_mm() which is more robust.

Signed-off-by: Andrea Arcangeli <[email protected]>
---
mm/mmu_context.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/mm/mmu_context.c b/mm/mmu_context.c
index 3e612ae748e9..ced0e1218c0f 100644
--- a/mm/mmu_context.c
+++ b/mm/mmu_context.c
@@ -30,6 +30,7 @@ void use_mm(struct mm_struct *mm)
mmgrab(mm);
tsk->active_mm = mm;
}
+ mmget(mm);
tsk->mm = mm;
switch_mm(active_mm, mm, tsk);
task_unlock(tsk);
@@ -57,6 +58,7 @@ void unuse_mm(struct mm_struct *mm)
task_lock(tsk);
sync_mm_rss(mm);
tsk->mm = NULL;
+ mmput(mm);
/* active_mm is still 'mm' */
enter_lazy_tlb(mm, tsk);
task_unlock(tsk);

2020-03-03 04:28:58

by Rafael Aquini

Subject: Re: [PATCH 1/3] mm: use_mm: fix for arches checking mm_users to optimize TLB flushes

On Sun, Feb 23, 2020 at 02:25:18PM -0500, Andrea Arcangeli wrote:
> alpha, ia64, mips, powerpc, sh, sparc are relying on a check on
> mm->mm_users to know if they can skip some remote TLB flushes for
> single threaded processes.
>
> Most callers of use_mm() tend to invoke mmget_not_zero() or
> get_task_mm() before use_mm() to ensure the mm will remain alive in
> between use_mm() and unuse_mm().
>
> Some callers however don't increase mm_users and they instead rely on
> serialization in __mmput() to ensure the mm will remain alive in
> between use_mm() and unuse_mm(). Not increasing mm_users during
> use_mm() is however unsafe for aforementioned arch TLB flushes
> optimizations. So either mmget()/mmput() should be added to the
> problematic callers of use_mm()/unuse_mm() or we can embed them in
> use_mm()/unuse_mm() which is more robust.
>
> Signed-off-by: Andrea Arcangeli <[email protected]>
> ---
> mm/mmu_context.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/mm/mmu_context.c b/mm/mmu_context.c
> index 3e612ae748e9..ced0e1218c0f 100644
> --- a/mm/mmu_context.c
> +++ b/mm/mmu_context.c
> @@ -30,6 +30,7 @@ void use_mm(struct mm_struct *mm)
> mmgrab(mm);
> tsk->active_mm = mm;
> }
> + mmget(mm);
> tsk->mm = mm;
> switch_mm(active_mm, mm, tsk);
> task_unlock(tsk);
> @@ -57,6 +58,7 @@ void unuse_mm(struct mm_struct *mm)
> task_lock(tsk);
> sync_mm_rss(mm);
> tsk->mm = NULL;
> + mmput(mm);
> /* active_mm is still 'mm' */
> enter_lazy_tlb(mm, tsk);
> task_unlock(tsk);

Acked-by: Rafael Aquini <[email protected]>