On Thu, Jun 13, 2013 at 01:33:46PM +0900, YAMAMOTO Takashi wrote: > >> def parser(cls, datapath, version, msg_type, msg_len, xid, buf): > >> @@ -195,7 +200,7 @@ class OFPSwitchFeatures(MsgBase): > >> msg.n_buffers, > >> msg.n_tables, > >> msg.capabilities, > >> - msg.reserved) = struct.unpack_from( > >> + msg._reserved) = struct.unpack_from( > > > > OFP_SWITCH_FEATURES_PACK_STR should be fixed to drop reserved? > > i.e. '4x' instead of 'I' > > unless someone wants to inspect the reserved field. > isn't it an intention of the original code?
I don't know the intention. The spec doesn't tell how the reserved field should be handled. OF1.0 used the field for supported action type. OF1.1+ deprecated the field and renamed it to reserved. To be honest, I'm not sure if this field is useful or not. There are other reserved field used in the spec, and Ryu ignores them (except this one). So I'm inclined to ignore this for consistency. thanks, -- yamahata ------------------------------------------------------------------------------ This SF.net email is sponsored by Windows: Build for Windows Store. http://p.sf.net/sfu/windows-dev2dev _______________________________________________ Ryu-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ryu-devel
