Hello Sowmini, I have posted updated webrevs that include suggested changes from your review to the list. But as the updates differ from my accepted changes in the previous e-mail on two review items I have clarified them here:
Rishi Srivatsavai wrote: [..] > [email protected] wrote: >> - Lines 7975, 7980, 7983- if this encompasses every line in the fields_arr >> table, then there's no need to specify it to ofmt_open: if you specify >> a NULL fields_str, then in the human-friendly mode, the ofmt functions >> will default to printing as many fields as fit the standard 80 column >> screen. Thus you can just initialize all_str to NULL, and only >> re-assign it to some specfic string for bridgeMode (at line 8089) >> or, if parsable is set, to the optarg (at line 8020). As the output fields differ from the ones that would be displayed by default if we were to set all_str to NULL, I have left this code unchanged. >> ---------------------------------------------------------------- >> >> SFW gate changes for Quagga are at: >> http://cr.opensolaris.org/~rishi/sfw-trill/ >> csope: >> /net/zhadum.east/export/ws/rs200217/sfwtrilld/ >> [..] >> - why isn't there an isis_destroy corresponding to isis_new()? I gather >> this must be XCALLOC'ed so that we can have multiple instances of >> isisd running in the future, but if that's the case, we should also >> make sure we do the XFREE correctly. Alternatively, isis_new could >> just use a global isis structure, and not make any half-hearted >> steps toward the multiple instances support. > > The global exists throughout the life of the program. There are no > multiple instances of isisd it is XCALLOC'ed to zero the structure. > I will add a destroy call. When working on this item in the code I realized there is no need for the destroy call as isisd never reaches the exit call in main to invoke the function and the global is used for the life of the daemon. Thanks, Rishi _______________________________________________ networking-discuss mailing list [email protected]
