[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240312100105.16a59086@kernel.org>
Date: Tue, 12 Mar 2024 10:01:05 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: Donald Hunter <donald.hunter@...il.com>, Jiri Pirko <jiri@...nulli.us>
Cc: Hangbin Liu <liuhangbin@...il.com>, netdev@...r.kernel.org
Subject: Re: How to display IPv4 array?
On Tue, 12 Mar 2024 16:04:56 +0000 Donald Hunter wrote:
> > "nested-array" would tell the parser to expect a nest that has attr
> > type of value of array index, "type" is the same for all array members.
> > The output will be the same as in case of "multi-attr", array index
> > ignored (I don't see what it would be good for to the user).
>
> I'd say that this construct looks more like nest-type-value
type-value is sort of a decomposed array, if we have all the entries
under one nest I reckon array extension may be more appropriate.
My gut feeling is that we should generalize the array-nest type,
when I wrote the initial spec we didn't have sub-type. How about
we replace array-nest with indexed-array (good name TBD), and instead
of assuming the value is always a nest pass the type via sub-type?
For bonding probably something like:
-
name: linkinfo-bond-attrs
name-prefix: ifla-bond-
attributes:
-
name: arp-ip-target
type: indexed-array
sub-type: u32
byte-order: big-endian
how does that sound?
exiting array-nests would change from:
-
name: bla
type: array-nest
nested-attributes: bla-attrs
to
-
name: bla
type: indexed-array
sub-type: nest
nested-attributes: bla-attrs
But that'd mean updating all existing specs and codegen.
Powered by blists - more mailing lists