add multi-buff support for xdp running in generic mode

index 1b935ee341b4..30b6a3f601fe 100644
@@ -3319,6 +3319,7 @@ struct softnet_data {
 	int			defer_count;
 	int			defer_ipi_scheduled;
 	struct sk_buff		*defer_list;
+	struct page_pool	*page_pool;
 	call_single_data_t	defer_csd;
 };
 

index 7dc65774cde5..b0da1fdb62dc 100644
@@ -393,4 +393,9 @@ static inline void page_pool_nid_changed(struct page_pool *pool, int new_nid)
 		page_pool_update_nid(pool, new_nid);
 }
 
+static inline void page_pool_set_cpuid(struct page_pool *pool, int cpuid)
+{
+	pool->cpuid = cpuid;
+}
+
 #endif /* _NET_PAGE_POOL_HELPERS_H */

index ac286ea8ce2d..75396f107b20 100644
@@ -127,7 +127,7 @@ struct page_pool_stats {
 
 struct page_pool {
 	struct page_pool_params_fast p;
-
+	int cpuid;
 	bool has_init_callback;
 
 	long frag_users;

index 0432b04cf9b0..4c3d82fcf5b5 100644
@@ -153,6 +153,8 @@
 #include <linux/prandom.h>
 #include <linux/once_lite.h>
 #include <net/netdev_rx_queue.h>
+#include <net/page_pool/types.h>
+#include <net/page_pool/helpers.h>
 
 #include "dev.h"
 #include "net-sysfs.h"

@@ -11670,12 +11672,14 @@ static void __init net_dev_struct_check(void)
  *
  */
 
+#define SD_PAGE_POOL_RING_SIZE	256
 /*
  *       This is called single threaded during boot, so no need
  *       to take the rtnl semaphore.
  */
 static int __init net_dev_init(void)
 {
+	struct softnet_data *sd;
 	int i, rc = -ENOMEM;
 
 	BUG_ON(!dev_boot_phase);

@@ -11701,10 +11705,14 @@ static int __init net_dev_init(void)
 
 	for_each_possible_cpu(i) {
 		struct work_struct *flush = per_cpu_ptr(&flush_works, i);
-		struct softnet_data *sd = &per_cpu(softnet_data, i);
+		struct page_pool_params page_pool_params = {
+			.pool_size = SD_PAGE_POOL_RING_SIZE,
+			.nid = NUMA_NO_NODE,
+		};
 
 		INIT_WORK(flush, flush_backlog);
 
+		sd = &per_cpu(softnet_data, i);
 		skb_queue_head_init(&sd->input_pkt_queue);
 		skb_queue_head_init(&sd->process_queue);
 #ifdef CONFIG_XFRM_OFFLOAD

@@ -11722,6 +11730,13 @@ static int __init net_dev_init(void)
 		init_gro_hash(&sd->backlog);
 		sd->backlog.poll = process_backlog;
 		sd->backlog.weight = weight_p;
+
+		sd->page_pool = page_pool_create(&page_pool_params);
+		if (IS_ERR(sd->page_pool)) {
+			sd->page_pool = NULL;
+			goto out;
+		}
+		page_pool_set_cpuid(sd->page_pool, i);
 	}
 
 	dev_boot_phase = 0;

@@ -11749,6 +11764,17 @@ static int __init net_dev_init(void)
 	WARN_ON(rc < 0);
 	rc = 0;
 out:
+	if (rc < 0) {
+		for_each_possible_cpu(i) {
+			sd = &per_cpu(softnet_data, i);
+			if (!sd->page_pool)
+				continue;
+
+			page_pool_destroy(sd->page_pool);
+			sd->page_pool = NULL;
+		}
+	}
+
 	return rc;
 }
 

index b157efea5dea..4bc0a7f98241 100644
@@ -918,9 +918,10 @@ bool napi_pp_put_page(struct page *page, bool napi_safe)
 	 */
 	if (napi_safe || in_softirq()) {
 		const struct napi_struct *napi = READ_ONCE(pp->p.napi);
+		unsigned int cpuid = smp_processor_id();
 
-		allow_direct = napi &&
-			READ_ONCE(napi->list_owner) == smp_processor_id();
+		allow_direct = napi && READ_ONCE(napi->list_owner) == cpuid;
+		allow_direct |= (pp->cpuid == cpuid);
 	}
 
 	/* Driver set this to memory recycling info. Reset it on recycle.

index afa5497f7c35..206adddff699 100644
@@ -1921,7 +1921,7 @@ static ssize_t tun_get_user(struct tun_struct *tun, struct tun_file *tfile,
 		rcu_read_lock();
 		xdp_prog = rcu_dereference(tun->xdp_prog);
 		if (xdp_prog) {
-			ret = do_xdp_generic(xdp_prog, skb);
+			ret = do_xdp_generic(xdp_prog, &skb);
 			if (ret != XDP_PASS) {
 				rcu_read_unlock();
 				local_bh_enable();

@@ -2511,7 +2511,7 @@ static int tun_xdp_one(struct tun_struct *tun,
 	skb_record_rx_queue(skb, tfile->queue_index);
 
 	if (skb_xdp) {
-		ret = do_xdp_generic(xdp_prog, skb);
+		ret = do_xdp_generic(xdp_prog, &skb);
 		if (ret != XDP_PASS) {
 			ret = 0;
 			goto out;

index 30b6a3f601fe..d850ead7f9cd 100644
@@ -3953,7 +3953,7 @@ static inline void dev_consume_skb_any(struct sk_buff *skb)
 u32 bpf_prog_run_generic_xdp(struct sk_buff *skb, struct xdp_buff *xdp,
 			     struct bpf_prog *xdp_prog);
 void generic_xdp_tx(struct sk_buff *skb, struct bpf_prog *xdp_prog);
-int do_xdp_generic(struct bpf_prog *xdp_prog, struct sk_buff *skb);
+int do_xdp_generic(struct bpf_prog *xdp_prog, struct sk_buff **pskb);
 int netif_rx(struct sk_buff *skb);
 int __netif_rx(struct sk_buff *skb);
 

index 4c3d82fcf5b5..14554698b1fa 100644
@@ -4916,10 +4916,11 @@ u32 bpf_prog_run_generic_xdp(struct sk_buff *skb, struct xdp_buff *xdp,
 	return act;
 }
 
-static u32 netif_receive_generic_xdp(struct sk_buff *skb,
+static u32 netif_receive_generic_xdp(struct sk_buff **pskb,
 				     struct xdp_buff *xdp,
 				     struct bpf_prog *xdp_prog)
 {
+	struct sk_buff *skb = *pskb;
 	u32 act = XDP_DROP;
 
 	/* Reinjected packets coming from act_mirred or similar should

@@ -5000,24 +5001,24 @@ void generic_xdp_tx(struct sk_buff *skb, struct bpf_prog *xdp_prog)
 
 static DEFINE_STATIC_KEY_FALSE(generic_xdp_needed_key);
 
-int do_xdp_generic(struct bpf_prog *xdp_prog, struct sk_buff *skb)
+int do_xdp_generic(struct bpf_prog *xdp_prog, struct sk_buff **pskb)
 {
 	if (xdp_prog) {
 		struct xdp_buff xdp;
 		u32 act;
 		int err;
 
-		act = netif_receive_generic_xdp(skb, &xdp, xdp_prog);
+		act = netif_receive_generic_xdp(pskb, &xdp, xdp_prog);
 		if (act != XDP_PASS) {
 			switch (act) {
 			case XDP_REDIRECT:
-				err = xdp_do_generic_redirect(skb->dev, skb,
+				err = xdp_do_generic_redirect((*pskb)->dev, *pskb,
 							      &xdp, xdp_prog);
 				if (err)
 					goto out_redir;
 				break;
 			case XDP_TX:
-				generic_xdp_tx(skb, xdp_prog);
+				generic_xdp_tx(*pskb, xdp_prog);
 				break;
 			}
 			return XDP_DROP;

@@ -5025,7 +5026,7 @@ int do_xdp_generic(struct bpf_prog *xdp_prog, struct sk_buff *skb)
 	}
 	return XDP_PASS;
 out_redir:
-	kfree_skb_reason(skb, SKB_DROP_REASON_XDP);
+	kfree_skb_reason(*pskb, SKB_DROP_REASON_XDP);
 	return XDP_DROP;
 }
 EXPORT_SYMBOL_GPL(do_xdp_generic);

@@ -5348,7 +5349,8 @@ static int __netif_receive_skb_core(struct sk_buff **pskb, bool pfmemalloc,
 		int ret2;
 
 		migrate_disable();
-		ret2 = do_xdp_generic(rcu_dereference(skb->dev->xdp_prog), skb);
+		ret2 = do_xdp_generic(rcu_dereference(skb->dev->xdp_prog),
+				      &skb);
 		migrate_enable();
 
 		if (ret2 != XDP_PASS) {

index 14554698b1fa..3826e0ea08e3 100644
@@ -4854,6 +4854,12 @@ u32 bpf_prog_run_generic_xdp(struct sk_buff *skb, struct xdp_buff *xdp,
 	xdp_init_buff(xdp, frame_sz, &rxqueue->xdp_rxq);
 	xdp_prepare_buff(xdp, hard_start, skb_headroom(skb) - mac_len,
 			 skb_headlen(skb) + mac_len, true);
+	if (skb_is_nonlinear(skb)) {
+		skb_shinfo(skb)->xdp_frags_size = skb->data_len;
+		xdp_buff_set_frags_flag(xdp);
+	} else {
+		xdp_buff_clear_frags_flag(xdp);
+	}
 
 	orig_data_end = xdp->data_end;
 	orig_data = xdp->data;

@@ -4883,6 +4889,14 @@ u32 bpf_prog_run_generic_xdp(struct sk_buff *skb, struct xdp_buff *xdp,
 		skb->len += off; /* positive on grow, negative on shrink */
 	}
 
+	/* XDP frag metadata (e.g. nr_frags) are updated in eBPF helpers
+	 * (e.g. bpf_xdp_adjust_tail), we need to update data_len here.
+	 */
+	if (xdp_buff_has_frags(xdp))
+		skb->data_len = skb_shinfo(skb)->xdp_frags_size;
+	else
+		skb->data_len = 0;
+
 	/* check if XDP changed eth hdr such SKB needs update */
 	eth = (struct ethhdr *)xdp->data;
 	if ((orig_eth_type != eth->h_proto) ||

@@ -4916,12 +4930,84 @@ u32 bpf_prog_run_generic_xdp(struct sk_buff *skb, struct xdp_buff *xdp,
 	return act;
 }
 
+static int netif_skb_segment_for_xdp(struct sk_buff **pskb)
+{
+	struct softnet_data *sd = this_cpu_ptr(&softnet_data);
+	u32 size, truesize, len, max_head_size, off;
+	struct sk_buff *skb = *pskb, *nskb;
+	int err, i, head_off;
+	void *data;
+
+	max_head_size = SKB_WITH_OVERHEAD(PAGE_SIZE - XDP_PACKET_HEADROOM);
+	if (skb->len > max_head_size + MAX_SKB_FRAGS * PAGE_SIZE)
+		return -ENOMEM;
+
+	size = min_t(u32, skb->len, max_head_size);
+	truesize = SKB_HEAD_ALIGN(size) + XDP_PACKET_HEADROOM;
+	data = page_pool_dev_alloc_va(sd->page_pool, &truesize);
+	if (!data)
+		return -ENOMEM;
+
+	nskb = napi_build_skb(data, truesize);
+	if (!nskb) {
+		page_pool_free_va(sd->page_pool, data, true);
+		return -ENOMEM;
+	}
+
+	skb_reserve(nskb, XDP_PACKET_HEADROOM);
+	skb_copy_header(nskb, skb);
+	skb_mark_for_recycle(nskb);
+
+	err = skb_copy_bits(skb, 0, nskb->data, size);
+	if (err) {
+		consume_skb(nskb);
+		return err;
+	}
+	skb_put(nskb, size);
+
+	head_off = skb_headroom(nskb) - skb_headroom(skb);
+	skb_headers_offset_update(nskb, head_off);
+
+	off = size;
+	len = skb->len - off;
+	for (i = 0; i < MAX_SKB_FRAGS && off < skb->len; i++) {
+		struct page *page;
+		u32 page_off;
+
+		size = min_t(u32, len, PAGE_SIZE);
+		truesize = size;
+
+		page = page_pool_dev_alloc(sd->page_pool, &page_off,
+					   &truesize);
+		if (!data) {
+			consume_skb(nskb);
+			return -ENOMEM;
+		}
+
+		skb_add_rx_frag(nskb, i, page, page_off, size, truesize);
+		err = skb_copy_bits(skb, off, page_address(page) + page_off,
+				    size);
+		if (err) {
+			consume_skb(nskb);
+			return err;
+		}
+
+		len -= size;
+		off += size;
+	}
+
+	consume_skb(skb);
+	*pskb = nskb;
+
+	return 0;
+}
+
 static u32 netif_receive_generic_xdp(struct sk_buff **pskb,
 				     struct xdp_buff *xdp,
 				     struct bpf_prog *xdp_prog)
 {
 	struct sk_buff *skb = *pskb;
-	u32 act = XDP_DROP;
+	u32 mac_len, act = XDP_DROP;
 
 	/* Reinjected packets coming from act_mirred or similar should
 	 * not get XDP generic processing.

@@ -4929,12 +5015,10 @@ static u32 netif_receive_generic_xdp(struct sk_buff **pskb,
 	if (skb_is_redirected(skb))
 		return XDP_PASS;
 
-	/* XDP packets must be linear and must have sufficient headroom
-	 * of XDP_PACKET_HEADROOM bytes. This is the guarantee that also
-	 * native XDP provides, thus we need to do it here as well.
+	/* XDP does not support fraglist so we need to linearize
+	 * the skb.
 	 */
-	if (skb_cloned(skb) || skb_is_nonlinear(skb) ||
-	    skb_headroom(skb) < XDP_PACKET_HEADROOM) {
+	if (skb_has_frag_list(skb) || !xdp_prog->aux->xdp_has_frags) {
 		int hroom = XDP_PACKET_HEADROOM - skb_headroom(skb);
 		int troom = skb->tail + skb->data_len - skb->end;
 

@@ -4947,23 +5031,40 @@ static u32 netif_receive_generic_xdp(struct sk_buff **pskb,
 			goto do_drop;
 		if (skb_linearize(skb))
 			goto do_drop;
+
+		goto do_xdp;
 	}
 
-	act = bpf_prog_run_generic_xdp(skb, xdp, xdp_prog);
+	/* XDP packets must have sufficient headroom of XDP_PACKET_HEADROOM
+	 * bytes. This is the guarantee that also native XDP provides,
+	 * thus we need to do it here as well.
+	 */
+	mac_len = skb->data - skb_mac_header(skb);
+	__skb_push(skb, mac_len);
+
+	if (skb_cloned(skb) || skb_shinfo(skb)->nr_frags ||
+	    skb_headroom(skb) < XDP_PACKET_HEADROOM) {
+		if (netif_skb_segment_for_xdp(pskb))
+			goto do_drop;
+	}
+
+	__skb_pull(*pskb, mac_len);
+do_xdp:
+	act = bpf_prog_run_generic_xdp(*pskb, xdp, xdp_prog);
 	switch (act) {
 	case XDP_REDIRECT:
 	case XDP_TX:
 	case XDP_PASS:
 		break;
 	default:
-		bpf_warn_invalid_xdp_action(skb->dev, xdp_prog, act);
+		bpf_warn_invalid_xdp_action((*pskb)->dev, xdp_prog, act);
 		fallthrough;
 	case XDP_ABORTED:
-		trace_xdp_exception(skb->dev, xdp_prog, act);
+		trace_xdp_exception((*pskb)->dev, xdp_prog, act);
 		fallthrough;
 	case XDP_DROP:
 	do_drop:
-		kfree_skb(skb);
+		kfree_skb(*pskb);
 		break;
 	}