Re: [nft PATCH 2/5] cache: Respect family in all list commands

[Phil Sutter <phil@nwl.cc>]

On Wed, Mar 11, 2026 at 01:11:56PM +0100, Pablo Neira Ayuso wrote:
> On Wed, Mar 11, 2026 at 11:19:58AM +0100, Phil Sutter wrote:
> > On Wed, Mar 11, 2026 at 10:34:59AM +0100, Pablo Neira Ayuso wrote:
> > > On Wed, Mar 11, 2026 at 12:11:12AM +0100, Phil Sutter wrote:
> > > > Some list commands did not set filter->list.family even if one was given
> > > > on command line, fix this.
> > > > 
> > > 
> > > Fixes: a1a6b0a5c3c4 ("cache: finer grain cache population for list commands")
> > 
> > Hmm. At that point, we didn't have 'filter' parameter in
> > evaluate_cache_list(). Struct nft_cache_filter was introduced later, in
> > commit 3f1d3912c3a6b ("cache: filter out tables that are not
> > requested").
> > 
> > Assuming that Fixes: tags are used for semi-automated backporting (at
> > least I do ;), pointing at that commit will cause trouble.
> 
> Good point.
> 
> Helping identify backporting in a semi-automated way is good,
> specially for small fixes like this.
> 
> At least for me, it helps me identify if it is an
> update/enhancement/feature or fix, it is just a bit more context
> information.
> 
> > Do you still think we should add that tag?
> 
> I get your point that tags need to be right if we use them.

How about:

Fixes: b3ed8fd8c9f33 ("cache: missing family in cache filtering")