On 10/23/25 1:45 AM, Jakub Kicinski wrote:
 > On Wed, 22 Oct 2025 18:26:53 +0000 Asbjørn Sloth Tønnesen wrote:
 >> This patchset adds a way to mark if an indedex array is just an
 >> array, and the index is uninteresting, as previously discussed[1].
 >>
 >> Which is the case in most of the indexed-arrays in the current specs.
 >>
 >> As the name indexed-array kinda implies that the index is interesting,
 >> then I am using `ignore-index` to mark if the index is unused.
 >>
 >> This adds some noise to YNL, and as it's only few indexed-arrays which
 >> actually use the index, then if we can come up with some good naming,
 >> it may be better to reverse it so it's the default behaviour.
 >
 > C code already does this, right? We just collect the attributes
 > completely ignoring the index.

In the userspace C code, for sub-type nest the index is preserved
in struct member idx, and elided for other sub-types.

 > So why do we need to extend the spec.

For me it's mostly the naming "indexed-array". Earlier it was
"array-nest", then we renamed it to "indexed-array" because
it doesn't always contain nests.  I think it's counter-intuitive
to elide the index by default, for something called "indexed-array".
The majority of the families using it don't care about the index.

What if we called it "ordered-array", and then always counted from 1
(for families that cares) when packing in user-space code?

 > Have you found any case where the index matters and can be
 > non-contiguous (other than the known TC kerfuffle).

IFLA_OFFLOAD_XSTATS_HW_S_INFO could be re-defined as a nest,
IFLA_OFFLOAD_XSTATS_L3_STATS is the only index atm.

IFLA_INET_CONF / IFLA_INET6_CONF is on input, but those are
also special by having different types depending on direction.

I found a bunch of other ones, using a static index, but they
can also be defined as a multi-attr wrapped in an additional
outer nest, like IFLA_VF_VLAN_LIST already is.

 > FWIW another concept is what TypeValue does.
 > "Inject" the index into the child nest as an extra member.
 > Most flexible but also prolly a PITA for user space to init those
 > for requests.

Same as is done in the userspace C code for indexed-arrays with
sub-type nest. For most families it doesn't matter if the C code
inits the index or not.