On 2022/4/8 11:54 AM, Andrew Morton wrote:
On Fri, 8 Apr 2022 11:50:05 +0800 Qi Zheng <zhengqi.arch@xxxxxxxxxxxxx> wrote:
On 2022/4/8 10:54 AM, Muchun Song wrote:
On Thu, Apr 07, 2022 at 06:33:35PM +0800, Qi Zheng wrote:
In the percpu_ref_call_confirm_rcu(), we call the wake_up_all()
before calling percpu_ref_put(), which will cause the value of
percpu_ref to be unstable when percpu_ref_switch_to_atomic_sync()
returns.
CPU0 CPU1
percpu_ref_switch_to_atomic_sync(&ref)
--> percpu_ref_switch_to_atomic(&ref)
--> percpu_ref_get(ref); /* put after confirmation */
call_rcu(&ref->data->rcu, percpu_ref_switch_to_atomic_rcu);
percpu_ref_switch_to_atomic_rcu
--> percpu_ref_call_confirm_rcu
--> data->confirm_switch = NULL;
wake_up_all(&percpu_ref_switch_waitq);
/* here waiting to wake up */
wait_event(percpu_ref_switch_waitq, !ref->data->confirm_switch);
(A)percpu_ref_put(ref);
/* The value of &ref is unstable! */
percpu_ref_is_zero(&ref)
(B)percpu_ref_put(ref);
As shown above, assuming that the counts on each cpu add up to 0 before
calling percpu_ref_switch_to_atomic_sync(), we expect that after switching
to atomic mode, percpu_ref_is_zero() can return true. But actually it will
return different values in the two cases of A and B, which is not what
we expected.
Maybe the original purpose of percpu_ref_switch_to_atomic_sync() is
just to ensure that the conversion to atomic mode is completed, but it
should not return with an extra reference count.
Calling wake_up_all() after percpu_ref_put() ensures that the value of
percpu_ref is stable after percpu_ref_switch_to_atomic_sync() returns.
So just do it.
Signed-off-by: Qi Zheng <zhengqi.arch@xxxxxxxxxxxxx>
Are any users affected by this? If so, I think a Fixes tag
is necessary.
Looks all current users(blk_pre_runtime_suspend() and set_in_sync()) are
affected by this.
I see that this patch has been merged into the mm tree, can Andrew help
me add the following Fixes tag?
Andrew is helpful ;)
Do you see reasons why we should backport this into -stable trees?
It's 8 years old, so my uninformed guess is "no"?
Hmm, although the commit 490c79a65708 add wake_up_all(), it is no
problem for the usage at that time, maybe the correct Fixes tag is the
following:
Fixes: 210f7cdcf088 ("percpu-refcount: support synchronous switch to
atomic mode.")
But in fact, there is no problem with it, but all current users expect
the refcount is stable after percpu_ref_switch_to_atomic_sync() returns.
I have no idea as which Fixes tag to add.
--
Thanks,
Qi
