M1 progress: engine port compiles except pf.c; AF_DIVERT socket family; ifq framework

- compat layer extended through the full BSD API surface pf_ioctl/pf_lb/
  pf_if/pf_norm/pf_table/pf_ruleset/radix/md5/sha2 need (all compile):
  cpumem/counters model, ifaddr/ifg groups, cred view, endian+in6 names,
  NAT46/64, SipHash24, mbuf_list, frag6 helpers, ip6 forward/checksum,
  tcp options/FSM names, taskqueue & sysctl shims (documented in
  docs/porting-notes.md)
- pfsrc/divert_sock.c: real BSD divert(4) as runtime-registered PF_DIVERT=32
  socket family inside the module (sock_register; NPROTO=AF_MAX=45 -> zero
  core patches); divert_packet/divert6_packet wired for pf.c
- compat/pfk_ifq.c: OpenBSD 8.0 ifq framework with priq back end; hfsc and
  fq_codel aliases pending M5 (recorded in feature matrix as pending)
- design + plan updated for the divert decision (user-requested)
- stage-port.sh: single deterministic rename program for headers and .c

module/include/compat/sys/percpu.h

@@ -0,0 +1,106 @@
+/* PFLINUX compat: <sys/percpu.h> — OpenBSD cpumem over a
+ * per-CPU-slot array (preempt disabled during access; counters only). */
+#ifndef _PFK_SYSPERCPU_H
+#define _PFK_SYSPERCPU_H
+
+struct cpumem {
+	void	*cm_base;
+	size_t	cm_item;
+	int	cm_ncpu;
+};
+
+struct cpumem_iter {
+	unsigned int	cpumem_iter_num;
+	unsigned int	cpumem_iter_cpus;
+	unsigned int	cpumem_iter_cpu;
+	unsigned long	cpumem_iter_obj;
+	struct cpumem	*cm;
+};
+
+static inline struct cpumem *
+cpumem_malloc(size_t item, int flags)
+{
+	struct cpumem *cm;
+	gfp_t g = (flags & M_WAITOK) ? GFP_KERNEL : GFP_ATOMIC;
+
+	cm = kmalloc(sizeof(*cm), g);
+	if (!cm)
+		return NULL;
+	cm->cm_ncpu = num_possible_cpus();
+	cm->cm_base = vzalloc((size_t)cm->cm_ncpu * item);
+	if (!cm->cm_base) {
+		kfree(cm);
+		return NULL;
+	}
+	cm->cm_item = item;
+	return cm;
+}
+
+static inline void *
+cpumem_enter(struct cpumem *cm)
+{
+	preempt_disable();
+	return (char *)cm->cm_base + (size_t)smp_processor_id() * cm->cm_item;
+}
+
+static inline void
+cpumem_leave(struct cpumem *cm, void *p)
+{
+	preempt_enable();
+}
+
+static inline void
+cpumem_free(struct cpumem *cm, int size)
+{
+	if (!cm)
+		return;
+	vfree(cm->cm_base);
+	kfree(cm);
+}
+
+/* BSD CPUMEM_FOREACH: iterate the per-CPU slots of a cpumem area. */
+static inline void *
+pfk_cpumem_at(struct cpumem *cm, unsigned int cpu)
+{
+	if (!cm || !cm->cm_base || cpu >= (unsigned int)cm->cm_ncpu)
+		return NULL;
+	return (char *)cm->cm_base + (size_t)cpu * cm->cm_item;
+}
+
+static inline void *
+pfk_cpumem_first(struct cpumem_iter *it, struct cpumem *cm)
+{
+	it->cpumem_iter_num = 0;
+	it->cm = cm;
+	return pfk_cpumem_at(cm, 0);
+}
+
+static inline void *
+pfk_cpumem_next(struct cpumem_iter *it, struct cpumem *cm)
+{
+	it->cpumem_iter_num++;
+	if (it->cpumem_iter_num >= (unsigned int)nr_cpu_ids)
+		return NULL;
+	return pfk_cpumem_at(cm, it->cpumem_iter_num);
+}
+
+#define CPUMEM_FOREACH(var, it, cm)					\
+	for ((var) = pfk_cpumem_first((it), (cm));			\
+	     (var) != NULL;						\
+	     (var) = pfk_cpumem_next((it), (cm)))
+
+/* BSD cpumem_ref/cpumem_modify aliases */
+static inline void *
+cpumem_ref(struct cpumem *cm)
+{
+	preempt_disable();
+	return (char *)cm->cm_base + (size_t)smp_processor_id() * cm->cm_item;
+}
+
+static inline void
+cpumem_drop(struct cpumem *cm, void *p)
+{
+	preempt_enable();
+}
+
+#endif