On 12.12.2016 21:25, Ben Pfaff wrote:
On Mon, Dec 12, 2016 at 04:56:56PM +0500, Valentine Sinitsyn wrote:
Suppose you are implementing a custom OpenFlow action, and you need some
per-bridge configuration to translate it into a datapath action.
Which would be the architecturally correct way to promote this bit of
information from OVSDB to somewhere inside struct xlate_ctx?
Usually you have to push it down from bridge.c to ofproto.c to
ofproto-dpif.c to ofproto-dpif-xlate.c through the various levels of
abstraction. It's awkward, unfortunately.
I keep thinking lately about getting rid of the abstraction between
ofproto and ofproto-dpif, because no one has ever contributed a second
implementation of ofproto.
Thanks for the quick answer. I traced the path from sources, and looks
like many ofproto_set_something() functions ultimately use mutexes to
protect data they change. What's the contract here; I mean, should
ofproto methods always be thread safe? The only place I call my function
is from bridge_reconfigure(). Besides, shall I introduce a separate
mutex for the data structure I have embedded in struct ofproto?
Thanks,
Valentine
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
