| Message ID | 4ba9adb2-43f5-2de0-22de-f6075c1fab50@i-love.sakura.ne.jp (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] binder: Don't use mmput() from shrinker function. | expand |
On Fri 17-07-20 00:12:15, Tetsuo Handa wrote: > syzbot is reporting that mmput() from shrinker function has a risk of > deadlock [1], for delayed_uprobe_add() from update_ref_ctr() calls > kzalloc(GFP_KERNEL) with delayed_uprobe_lock held, and > uprobe_clear_state() from __mmput() also holds delayed_uprobe_lock. > > Commit a1b2289cef92ef0e ("android: binder: drop lru lock in isolate > callback") replaced mmput() with mmput_async() in order to avoid sleeping > with spinlock held. But this patch replaces mmput() with mmput_async() in > order not to start __mmput() from shrinker context. > > [1] https://syzkaller.appspot.com/bug?id=bc9e7303f537c41b2b0cc2dfcea3fc42964c2d45 > > Reported-by: syzbot <syzbot+1068f09c44d151250c33@syzkaller.appspotmail.com> > Reported-by: syzbot <syzbot+e5344baa319c9a96edec@syzkaller.appspotmail.com> > Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Reviewed-by: Michal Hocko <mhocko@suse.com> Thanks! > --- > drivers/android/binder_alloc.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/android/binder_alloc.c b/drivers/android/binder_alloc.c > index 42c672f1584e..cbe6aa77d50d 100644 > --- a/drivers/android/binder_alloc.c > +++ b/drivers/android/binder_alloc.c > @@ -947,7 +947,7 @@ enum lru_status binder_alloc_free_page(struct list_head *item, > trace_binder_unmap_user_end(alloc, index); > } > mmap_read_unlock(mm); > - mmput(mm); > + mmput_async(mm); > > trace_binder_unmap_kernel_start(alloc, index); > > -- > 2.18.4
On Thu, Jul 16, 2020 at 05:17:56PM +0200, Michal Hocko wrote: > On Fri 17-07-20 00:12:15, Tetsuo Handa wrote: > > syzbot is reporting that mmput() from shrinker function has a risk of > > deadlock [1], for delayed_uprobe_add() from update_ref_ctr() calls > > kzalloc(GFP_KERNEL) with delayed_uprobe_lock held, and > > uprobe_clear_state() from __mmput() also holds delayed_uprobe_lock. > > > > Commit a1b2289cef92ef0e ("android: binder: drop lru lock in isolate > > callback") replaced mmput() with mmput_async() in order to avoid sleeping > > with spinlock held. But this patch replaces mmput() with mmput_async() in > > order not to start __mmput() from shrinker context. > > > > [1] https://syzkaller.appspot.com/bug?id=bc9e7303f537c41b2b0cc2dfcea3fc42964c2d45 > > > > Reported-by: syzbot <syzbot+1068f09c44d151250c33@syzkaller.appspotmail.com> > > Reported-by: syzbot <syzbot+e5344baa319c9a96edec@syzkaller.appspotmail.com> > > Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> > > Reviewed-by: Michal Hocko <mhocko@suse.com> Thanks for the careful review Michal! Does this need a Cc: stable? Otherwise Acked-by: Christian Brauner <christian.brauner@ubuntu.com> Thanks! Christian
On 2020/07/17 1:29, Christian Brauner wrote:
> Does this need a Cc: stable?
Up to someone who applies this patch. I think this race is hard to hit.
On Thu, Jul 16, 2020 at 8:18 AM Michal Hocko <mhocko@kernel.org> wrote: > > On Fri 17-07-20 00:12:15, Tetsuo Handa wrote: > > syzbot is reporting that mmput() from shrinker function has a risk of > > deadlock [1], for delayed_uprobe_add() from update_ref_ctr() calls > > kzalloc(GFP_KERNEL) with delayed_uprobe_lock held, and > > uprobe_clear_state() from __mmput() also holds delayed_uprobe_lock. > > > > Commit a1b2289cef92ef0e ("android: binder: drop lru lock in isolate > > callback") replaced mmput() with mmput_async() in order to avoid sleeping > > with spinlock held. But this patch replaces mmput() with mmput_async() in > > order not to start __mmput() from shrinker context. > > > > [1] https://syzkaller.appspot.com/bug?id=bc9e7303f537c41b2b0cc2dfcea3fc42964c2d45 > > > > Reported-by: syzbot <syzbot+1068f09c44d151250c33@syzkaller.appspotmail.com> > > Reported-by: syzbot <syzbot+e5344baa319c9a96edec@syzkaller.appspotmail.com> > > Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> > > Reviewed-by: Michal Hocko <mhocko@suse.com> Acked-by: Todd Kjos <tkjos@google.com> > > Thanks! > > > --- > > drivers/android/binder_alloc.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/drivers/android/binder_alloc.c b/drivers/android/binder_alloc.c > > index 42c672f1584e..cbe6aa77d50d 100644 > > --- a/drivers/android/binder_alloc.c > > +++ b/drivers/android/binder_alloc.c > > @@ -947,7 +947,7 @@ enum lru_status binder_alloc_free_page(struct list_head *item, > > trace_binder_unmap_user_end(alloc, index); > > } > > mmap_read_unlock(mm); > > - mmput(mm); > > + mmput_async(mm); > > > > trace_binder_unmap_kernel_start(alloc, index); > > > > -- > > 2.18.4 > > -- > Michal Hocko > SUSE Labs
diff --git a/drivers/android/binder_alloc.c b/drivers/android/binder_alloc.c index 42c672f1584e..cbe6aa77d50d 100644 --- a/drivers/android/binder_alloc.c +++ b/drivers/android/binder_alloc.c @@ -947,7 +947,7 @@ enum lru_status binder_alloc_free_page(struct list_head *item, trace_binder_unmap_user_end(alloc, index); } mmap_read_unlock(mm); - mmput(mm); + mmput_async(mm); trace_binder_unmap_kernel_start(alloc, index);
syzbot is reporting that mmput() from shrinker function has a risk of deadlock [1], for delayed_uprobe_add() from update_ref_ctr() calls kzalloc(GFP_KERNEL) with delayed_uprobe_lock held, and uprobe_clear_state() from __mmput() also holds delayed_uprobe_lock. Commit a1b2289cef92ef0e ("android: binder: drop lru lock in isolate callback") replaced mmput() with mmput_async() in order to avoid sleeping with spinlock held. But this patch replaces mmput() with mmput_async() in order not to start __mmput() from shrinker context. [1] https://syzkaller.appspot.com/bug?id=bc9e7303f537c41b2b0cc2dfcea3fc42964c2d45 Reported-by: syzbot <syzbot+1068f09c44d151250c33@syzkaller.appspotmail.com> Reported-by: syzbot <syzbot+e5344baa319c9a96edec@syzkaller.appspotmail.com> Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> --- drivers/android/binder_alloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)