linux-kernel - Re: [PATCH V4 7/9] vhost: do not use RCU to synchronize MMU notifier with worker

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20190807120738.GB1557@ziepe.ca>
Date:   Wed, 7 Aug 2019 09:07:38 -0300
From:   Jason Gunthorpe <jgg@...pe.ca>
To:     Jason Wang <jasowang@...hat.com>
Cc:     mst@...hat.com, kvm@...r.kernel.org,
        virtualization@...ts.linux-foundation.org, netdev@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-mm@...ck.org
Subject: Re: [PATCH V4 7/9] vhost: do not use RCU to synchronize MMU notifier
 with worker

On Wed, Aug 07, 2019 at 03:06:15AM -0400, Jason Wang wrote:
> We used to use RCU to synchronize MMU notifier with worker. This leads
> calling synchronize_rcu() in invalidate_range_start(). But on a busy
> system, there would be many factors that may slow down the
> synchronize_rcu() which makes it unsuitable to be called in MMU
> notifier.
> 
> So this patch switches use seqlock counter to track whether or not the
> map was used. The counter was increased when vq try to start or finish
> uses the map. This means, when it was even, we're sure there's no
> readers and MMU notifier is synchronized. When it was odd, it means
> there's a reader we need to wait it to be even again then we are
> synchronized. Consider the read critical section is pretty small the
> synchronization should be done very fast.
> 
> Reported-by: Michael S. Tsirkin <mst@...hat.com>
> Fixes: 7f466032dc9e ("vhost: access vq metadata through kernel virtual address")
> Signed-off-by: Jason Wang <jasowang@...hat.com>
>  drivers/vhost/vhost.c | 141 ++++++++++++++++++++++++++----------------
>  drivers/vhost/vhost.h |   7 ++-
>  2 files changed, 90 insertions(+), 58 deletions(-)
> 
> diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
> index cfc11f9ed9c9..57bfbb60d960 100644
> +++ b/drivers/vhost/vhost.c
> @@ -324,17 +324,16 @@ static void vhost_uninit_vq_maps(struct vhost_virtqueue *vq)
>  
>  	spin_lock(&vq->mmu_lock);
>  	for (i = 0; i < VHOST_NUM_ADDRS; i++) {
> -		map[i] = rcu_dereference_protected(vq->maps[i],
> -				  lockdep_is_held(&vq->mmu_lock));
> +		map[i] = vq->maps[i];
>  		if (map[i]) {
>  			vhost_set_map_dirty(vq, map[i], i);
> -			rcu_assign_pointer(vq->maps[i], NULL);
> +			vq->maps[i] = NULL;
>  		}
>  	}
>  	spin_unlock(&vq->mmu_lock);
>  
> -	/* No need for synchronize_rcu() or kfree_rcu() since we are
> -	 * serialized with memory accessors (e.g vq mutex held).
> +	/* No need for synchronization since we are serialized with
> +	 * memory accessors (e.g vq mutex held).
>  	 */
>  
>  	for (i = 0; i < VHOST_NUM_ADDRS; i++)
> @@ -362,6 +361,40 @@ static bool vhost_map_range_overlap(struct vhost_uaddr *uaddr,
>  	return !(end < uaddr->uaddr || start > uaddr->uaddr - 1 + uaddr->size);
>  }
>  
> +static void inline vhost_vq_access_map_begin(struct vhost_virtqueue *vq)
> +{
> +	write_seqcount_begin(&vq->seq);
> +}
> +
> +static void inline vhost_vq_access_map_end(struct vhost_virtqueue *vq)
> +{
> +	write_seqcount_end(&vq->seq);
> +}

The write side of a seqlock only provides write barriers. Access to

	map = vq->maps[VHOST_ADDR_USED];

Still needs a read side barrier, and then I think this will be no
better than a normal spinlock.

It also doesn't seem like this algorithm even needs a seqlock, as this
is just a one bit flag

atomic_set_bit(using map)
smp_mb__after_atomic()
.. maps [...]
atomic_clear_bit(using map)


map = NULL;
smp_mb__before_atomic();
while (atomic_read_bit(using map))
   relax()

Again, not clear this could be faster than a spinlock when the
barriers are correct...

Jason