[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH net-next V3 2/3] xen-netfront: split event channels support for Xen frontend driver




On 2013-5-23 9:46, Wei Liu wrote:
On Wed, May 22, 2013 at 09:38:20PM +0100, Wei Liu wrote:
On Wed, May 22, 2013 at 9:35 PM, annie li <annie.li@xxxxxxxxxx> wrote:
On 2013-5-22 16:20, Wei Liu wrote:
On Wed, May 22, 2013 at 8:32 PM, annie li <annie.li@xxxxxxxxxx> wrote:

Originally, netfront protects access to rx shared-ring with tx_lock, you
remove this protection here. It is better to protect the ring access by a
sperate rx_lock then.

TX ring and RX ring are separate rings. I don't think that comment / code
makes sense any more. My stress test confirms that.

Yes, they are separate rings. Actually I am not sure why
RING_HAS_UNCONSUMED_RESPONSES(&np->rx) is protected by any tx_lock
originally. But for xennet_rx_interrupt, it is better to use rx_lock to
protect RING_HAS_UNCONSUMED_RESPONSES(&np->rx).

This doesn't make sense to me either. Xen ring protocol is designed to
be lock-free.
And in netfront's case there is no concurrent access to the ring.


Annie, did I answer your questions / relieve your concern?

Yes, I think you are correct.

Thanks
Annie

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.