Re: [PATCH 2/2] vhost: disentangle vring endianness stuff from the core code

On Wed, 10 Feb 2016 14:23:33 +0100
Cornelia Huck [off-list ref] wrote:

On Wed, 10 Feb 2016 14:08:43 +0100
Greg Kurz [off-list ref] wrote:

quoted

But you are right, there is a bug: we should rollback if vhost_init_used()
fails. Something like below:

 err_used:
        vq->private_data = oldsock;
        vhost_net_enable_vq(n, vq);
+       vhost_adjust_vring_endian(vq);  

Shouldn't we switch back before we reenable? Or have I lost myself in
this maze here again?

I haven't spotted any path under vhost_net_enable_vq() that needs
the vring endianness, but indeed it looks safer to switch back
before a worker thread may be woken up.

quoted

        if (ubufs)
                vhost_net_ubuf_put_wait_and_free(ubufs);
 err_ubufs: