On 10/28/20 3:21 PM, Thomas Monjalon wrote: > 28/10/2020 12:51, Andrew Rybchenko: >> On 10/28/20 1:26 PM, Thomas Monjalon wrote: >>> The example used the deprecated mbuf field udata64 as input mbuf pointer. >>> It is moved to a dynamic field in order to allow removal of udata64. >>> >>> Signed-off-by: Thomas Monjalon <tho...@monjalon.net> >> >> Reviewed-by: Andrew Rybchenko <andrew.rybche...@oktetlabs.ru> >> >> I like the approach with inline function(s) more than approach >> with macros. Not sure that get/set is needed. Approach in >> David's patches looks a bit simpler to me - less duplicated >> code (i.e. just one function which returns pointer of correct >> type to a dynamic field and allow to get and set it). > > Yes I tend to agree with you. > Olivier was asking for more functions, > especially if it has a semantic meaning. > Here get/set has no real added value but I don't have strong opinion, > and I don't think it deserves a respin :) >
Agreed.
