On 13 January 2015 at 16:07, Ben Pfaff <[email protected]> wrote: > On Fri, Dec 19, 2014 at 01:17:56PM -0800, Joe Stringer wrote: >> We previously tracked the debug enable/disable of UFID in each udpif, >> and allowed the ovs-appctl debug option to turn on UFID features even if >> the datapath doesn't support it. >> >> This commit shifts the enable_ufid debug flag to a single flag, and >> provides a helper to determine whether UFID features should be used on a >> per-udpif basis. >> >> Suggested-by: Andy Zhou <[email protected]> >> Signed-off-by: Joe Stringer <[email protected]> > > Seems OK. > > The naming would be slightly clearer if struct dpif_backer's > 'enable_ufid' were named something like 'supports_ufid' and similarly > for ofproto_dpif_get_enable_ufid(). Then the difference between > whether a dpif supports UFIDs and whether ofproto-dpif is actually > using them would be crystal clear. (That could be a different patch, > if you think it's worthwhile.) > > Acked-by: Ben Pfaff <[email protected]>
Thanks, pushed to master. I agree that differentiating "support" from "turning feature on/off" is a useful distinction for readability. I'll send a patch. _______________________________________________ dev mailing list [email protected] http://openvswitch.org/mailman/listinfo/dev
