On Wed, Jul 23, 2025 at 02:20:45PM +0200, Florian Westphal wrote: > Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx> wrote: > > > The fourth patch is the main change, it removes the control-plane > > > only C implementation of the pipapo lookup algorithm. > > > > > > The last patch allows the scratch maps to be backed by vmalloc. > > > > This clashes with: > > > > netfilter: nft_set_pipapo: Use nested-BH locking for nft_pipapo_scratch > > > > https://patchwork.ozlabs.org/project/netfilter-devel/list/?series=463248 > > > > It seems you and Sebastian have been working on the same area at the > > same time. > > > > Do you have a preference on how to proceed with this clash? > > If the clash is only in last patch, then how about this: > You apply the first few patches plus Sebastians work, push this out > (main or testing branch is fine too) and I will rebase + resend? Testing branch: https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next.git/log/?h=testing I still have to make another pass on patchwork, regarding your work, there is one series of you to avoid the GFP_KERNEL allocation I would like to review. > Or, we can defer to 6.18, your call.
