Merge branch 'XDP-redirect-memory-return-API'

Jesper Dangaard Brouer says:

====================
XDP redirect memory return API

Submitted against net-next, as it contains NIC driver changes.

This patchset works towards supporting different XDP RX-ring memory
allocators.  As this will be needed by the AF_XDP zero-copy mode.

The patchset uses mlx5 as the sample driver, which gets implemented
XDP_REDIRECT RX-mode, but not ndo_xdp_xmit (as this API is subject to
change thought the patchset).

A new struct xdp_frame is introduced (modeled after cpumap xdp_pkt).
And both ndo_xdp_xmit and the new xdp_return_frame end-up using this.

Support for a driver supplied allocator is implemented, and a
refurbished version of page_pool is the first return allocator type
introduced.  This will be a integration point for AF_XDP zero-copy.

The mlx5 driver evolve into using the page_pool, and see a performance
increase (with ndo_xdp_xmit out ixgbe driver) from 6Mpps to 12Mpps.

The patchset stop at 16 patches (one over limit), but more API changes
are planned.  Specifically extending ndo_xdp_xmit and xdp_return_frame
APIs to support bulking.  As this will address some known limits.

V2: Updated according to Tariq's feedback
V3: Updated based on feedback from Jason Wang and Alex Duyck
V4: Updated based on feedback from Tariq and Jason
V5: Fix SPDX license, add Tariq's reviews, improve patch desc for perf test
V6: Updated based on feedback from Eric Dumazet and Alex Duyck
V7: Adapt to i40e that got XDP_REDIRECT support in-between
V8:
 Updated based on feedback kbuild test robot, and adjust for mlx5 changes
 page_pool only compiled into kernel when drivers Kconfig 'select' feature
V9:
 Remove some inline statements, let compiler decide what to inline
 Fix return value in virtio_net driver
 Adjust for mlx5 changes in-between submissions
V10:
 Minor adjust for mlx5 requested by Tariq
 Resubmit against net-next
V11: avoid leaking info stored in frame data on page reuse
====================

Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 613 insertions, 2 deletions

diff --git a/net/Kconfig b/net/Kconfig
index 0428f12c25c2..6fa1a4493b8c 100644
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -423,6 +423,9 @@ config MAY_USE_DEVLINK
 	  on MAY_USE_DEVLINK to ensure they do not cause link errors when
 	  devlink is a loadable module and the driver using it is built-in.
 
+config PAGE_POOL
+       bool
+
 endif   # if NET
 
 # Used by archs to tell that they support BPF JIT compiler plus which flavour.
diff --git a/net/core/Makefile b/net/core/Makefile
index 6dbbba8c57ae..7080417f8bc8 100644
--- a/net/core/Makefile
+++ b/net/core/Makefile
@@ -14,6 +14,7 @@ obj-y		     += dev.o ethtool.o dev_addr_lists.o dst.o netevent.o \
 			fib_notifier.o xdp.o
 
 obj-y += net-sysfs.o
+obj-$(CONFIG_PAGE_POOL) += page_pool.o
 obj-$(CONFIG_PROC_FS) += net-procfs.o
 obj-$(CONFIG_NET_PKTGEN) += pktgen.o
 obj-$(CONFIG_NETPOLL) += netpoll.o
diff --git a/net/core/filter.c b/net/core/filter.c
index d31aff93270d..a374b8560bc4 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -2692,6 +2692,7 @@ static unsigned long xdp_get_metalen(const struct xdp_buff *xdp)
 
 BPF_CALL_2(bpf_xdp_adjust_head, struct xdp_buff *, xdp, int, offset)
 {
+	void *xdp_frame_end = xdp->data_hard_start + sizeof(struct xdp_frame);
 	unsigned long metalen = xdp_get_metalen(xdp);
 	void *data_start = xdp->data_hard_start + metalen;
 	void *data = xdp->data + offset;
@@ -2700,6 +2701,13 @@ BPF_CALL_2(bpf_xdp_adjust_head, struct xdp_buff *, xdp, int, offset)
 		     data > xdp->data_end - ETH_HLEN))
 		return -EINVAL;
 
+	/* Avoid info leak, when reusing area prev used by xdp_frame */
+	if (data < xdp_frame_end) {
+		unsigned long clearlen = xdp_frame_end - data;
+
+		memset(data, 0, clearlen);
+	}
+
 	if (metalen)
 		memmove(xdp->data_meta + offset,
 			xdp->data_meta, metalen);
@@ -2749,13 +2757,18 @@ static int __bpf_tx_xdp(struct net_device *dev,
 			struct xdp_buff *xdp,
 			u32 index)
 {
+	struct xdp_frame *xdpf;
 	int err;
 
 	if (!dev->netdev_ops->ndo_xdp_xmit) {
 		return -EOPNOTSUPP;
 	}
 
-	err = dev->netdev_ops->ndo_xdp_xmit(dev, xdp);
+	xdpf = convert_to_xdp_frame(xdp);
+	if (unlikely(!xdpf))
+		return -EOVERFLOW;
+
+	err = dev->netdev_ops->ndo_xdp_xmit(dev, xdpf);
 	if (err)
 		return err;
 	dev->netdev_ops->ndo_xdp_flush(dev);
@@ -2771,11 +2784,19 @@ static int __bpf_tx_xdp_map(struct net_device *dev_rx, void *fwd,
 
 	if (map->map_type == BPF_MAP_TYPE_DEVMAP) {
 		struct net_device *dev = fwd;
+		struct xdp_frame *xdpf;
 
 		if (!dev->netdev_ops->ndo_xdp_xmit)
 			return -EOPNOTSUPP;
 
-		err = dev->netdev_ops->ndo_xdp_xmit(dev, xdp);
+		xdpf = convert_to_xdp_frame(xdp);
+		if (unlikely(!xdpf))
+			return -EOVERFLOW;
+
+		/* TODO: move to inside map code instead, for bulk support
+		 * err = dev_map_enqueue(dev, xdp);
+		 */
+		err = dev->netdev_ops->ndo_xdp_xmit(dev, xdpf);
 		if (err)
 			return err;
 		__dev_map_insert_ctx(map, index);
diff --git a/net/core/page_pool.c b/net/core/page_pool.c
new file mode 100644
index 000000000000..68bf07206744
--- /dev/null
+++ b/net/core/page_pool.c
@@ -0,0 +1,317 @@
+/* SPDX-License-Identifier: GPL-2.0
+ *
+ * page_pool.c
+ *	Author:	Jesper Dangaard Brouer <netoptimizer@brouer.com>
+ *	Copyright (C) 2016 Red Hat, Inc.
+ */
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+
+#include <net/page_pool.h>
+#include <linux/dma-direction.h>
+#include <linux/dma-mapping.h>
+#include <linux/page-flags.h>
+#include <linux/mm.h> /* for __put_page() */
+
+static int page_pool_init(struct page_pool *pool,
+			  const struct page_pool_params *params)
+{
+	unsigned int ring_qsize = 1024; /* Default */
+
+	memcpy(&pool->p, params, sizeof(pool->p));
+
+	/* Validate only known flags were used */
+	if (pool->p.flags & ~(PP_FLAG_ALL))
+		return -EINVAL;
+
+	if (pool->p.pool_size)
+		ring_qsize = pool->p.pool_size;
+
+	/* Sanity limit mem that can be pinned down */
+	if (ring_qsize > 32768)
+		return -E2BIG;
+
+	/* DMA direction is either DMA_FROM_DEVICE or DMA_BIDIRECTIONAL.
+	 * DMA_BIDIRECTIONAL is for allowing page used for DMA sending,
+	 * which is the XDP_TX use-case.
+	 */
+	if ((pool->p.dma_dir != DMA_FROM_DEVICE) &&
+	    (pool->p.dma_dir != DMA_BIDIRECTIONAL))
+		return -EINVAL;
+
+	if (ptr_ring_init(&pool->ring, ring_qsize, GFP_KERNEL) < 0)
+		return -ENOMEM;
+
+	return 0;
+}
+
+struct page_pool *page_pool_create(const struct page_pool_params *params)
+{
+	struct page_pool *pool;
+	int err = 0;
+
+	pool = kzalloc_node(sizeof(*pool), GFP_KERNEL, params->nid);
+	if (!pool)
+		return ERR_PTR(-ENOMEM);
+
+	err = page_pool_init(pool, params);
+	if (err < 0) {
+		pr_warn("%s() gave up with errno %d\n", __func__, err);
+		kfree(pool);
+		return ERR_PTR(err);
+	}
+	return pool;
+}
+EXPORT_SYMBOL(page_pool_create);
+
+/* fast path */
+static struct page *__page_pool_get_cached(struct page_pool *pool)
+{
+	struct ptr_ring *r = &pool->ring;
+	struct page *page;
+
+	/* Quicker fallback, avoid locks when ring is empty */
+	if (__ptr_ring_empty(r))
+		return NULL;
+
+	/* Test for safe-context, caller should provide this guarantee */
+	if (likely(in_serving_softirq())) {
+		if (likely(pool->alloc.count)) {
+			/* Fast-path */
+			page = pool->alloc.cache[--pool->alloc.count];
+			return page;
+		}
+		/* Slower-path: Alloc array empty, time to refill
+		 *
+		 * Open-coded bulk ptr_ring consumer.
+		 *
+		 * Discussion: the ring consumer lock is not really
+		 * needed due to the softirq/NAPI protection, but
+		 * later need the ability to reclaim pages on the
+		 * ring. Thus, keeping the locks.
+		 */
+		spin_lock(&r->consumer_lock);
+		while ((page = __ptr_ring_consume(r))) {
+			if (pool->alloc.count == PP_ALLOC_CACHE_REFILL)
+				break;
+			pool->alloc.cache[pool->alloc.count++] = page;
+		}
+		spin_unlock(&r->consumer_lock);
+		return page;
+	}
+
+	/* Slow-path: Get page from locked ring queue */
+	page = ptr_ring_consume(&pool->ring);
+	return page;
+}
+
+/* slow path */
+noinline
+static struct page *__page_pool_alloc_pages_slow(struct page_pool *pool,
+						 gfp_t _gfp)
+{
+	struct page *page;
+	gfp_t gfp = _gfp;
+	dma_addr_t dma;
+
+	/* We could always set __GFP_COMP, and avoid this branch, as
+	 * prep_new_page() can handle order-0 with __GFP_COMP.
+	 */
+	if (pool->p.order)
+		gfp |= __GFP_COMP;
+
+	/* FUTURE development:
+	 *
+	 * Current slow-path essentially falls back to single page
+	 * allocations, which doesn't improve performance.  This code
+	 * need bulk allocation support from the page allocator code.
+	 */
+
+	/* Cache was empty, do real allocation */
+	page = alloc_pages_node(pool->p.nid, gfp, pool->p.order);
+	if (!page)
+		return NULL;
+
+	if (!(pool->p.flags & PP_FLAG_DMA_MAP))
+		goto skip_dma_map;
+
+	/* Setup DMA mapping: use page->private for DMA-addr
+	 * This mapping is kept for lifetime of page, until leaving pool.
+	 */
+	dma = dma_map_page(pool->p.dev, page, 0,
+			   (PAGE_SIZE << pool->p.order),
+			   pool->p.dma_dir);
+	if (dma_mapping_error(pool->p.dev, dma)) {
+		put_page(page);
+		return NULL;
+	}
+	set_page_private(page, dma); /* page->private = dma; */
+
+skip_dma_map:
+	/* When page just alloc'ed is should/must have refcnt 1. */
+	return page;
+}
+
+/* For using page_pool replace: alloc_pages() API calls, but provide
+ * synchronization guarantee for allocation side.
+ */
+struct page *page_pool_alloc_pages(struct page_pool *pool, gfp_t gfp)
+{
+	struct page *page;
+
+	/* Fast-path: Get a page from cache */
+	page = __page_pool_get_cached(pool);
+	if (page)
+		return page;
+
+	/* Slow-path: cache empty, do real allocation */
+	page = __page_pool_alloc_pages_slow(pool, gfp);
+	return page;
+}
+EXPORT_SYMBOL(page_pool_alloc_pages);
+
+/* Cleanup page_pool state from page */
+static void __page_pool_clean_page(struct page_pool *pool,
+				   struct page *page)
+{
+	if (!(pool->p.flags & PP_FLAG_DMA_MAP))
+		return;
+
+	/* DMA unmap */
+	dma_unmap_page(pool->p.dev, page_private(page),
+		       PAGE_SIZE << pool->p.order, pool->p.dma_dir);
+	set_page_private(page, 0);
+}
+
+/* Return a page to the page allocator, cleaning up our state */
+static void __page_pool_return_page(struct page_pool *pool, struct page *page)
+{
+	__page_pool_clean_page(pool, page);
+	put_page(page);
+	/* An optimization would be to call __free_pages(page, pool->p.order)
+	 * knowing page is not part of page-cache (thus avoiding a
+	 * __page_cache_release() call).
+	 */
+}
+
+static bool __page_pool_recycle_into_ring(struct page_pool *pool,
+				   struct page *page)
+{
+	int ret;
+	/* BH protection not needed if current is serving softirq */
+	if (in_serving_softirq())
+		ret = ptr_ring_produce(&pool->ring, page);
+	else
+		ret = ptr_ring_produce_bh(&pool->ring, page);
+
+	return (ret == 0) ? true : false;
+}
+
+/* Only allow direct recycling in special circumstances, into the
+ * alloc side cache.  E.g. during RX-NAPI processing for XDP_DROP use-case.
+ *
+ * Caller must provide appropriate safe context.
+ */
+static bool __page_pool_recycle_direct(struct page *page,
+				       struct page_pool *pool)
+{
+	if (unlikely(pool->alloc.count == PP_ALLOC_CACHE_SIZE))
+		return false;
+
+	/* Caller MUST have verified/know (page_ref_count(page) == 1) */
+	pool->alloc.cache[pool->alloc.count++] = page;
+	return true;
+}
+
+void __page_pool_put_page(struct page_pool *pool,
+			  struct page *page, bool allow_direct)
+{
+	/* This allocator is optimized for the XDP mode that uses
+	 * one-frame-per-page, but have fallbacks that act like the
+	 * regular page allocator APIs.
+	 *
+	 * refcnt == 1 means page_pool owns page, and can recycle it.
+	 */
+	if (likely(page_ref_count(page) == 1)) {
+		/* Read barrier done in page_ref_count / READ_ONCE */
+
+		if (allow_direct && in_serving_softirq())
+			if (__page_pool_recycle_direct(page, pool))
+				return;
+
+		if (!__page_pool_recycle_into_ring(pool, page)) {
+			/* Cache full, fallback to free pages */
+			__page_pool_return_page(pool, page);
+		}
+		return;
+	}
+	/* Fallback/non-XDP mode: API user have elevated refcnt.
+	 *
+	 * Many drivers split up the page into fragments, and some
+	 * want to keep doing this to save memory and do refcnt based
+	 * recycling. Support this use case too, to ease drivers
+	 * switching between XDP/non-XDP.
+	 *
+	 * In-case page_pool maintains the DMA mapping, API user must
+	 * call page_pool_put_page once.  In this elevated refcnt
+	 * case, the DMA is unmapped/released, as driver is likely
+	 * doing refcnt based recycle tricks, meaning another process
+	 * will be invoking put_page.
+	 */
+	__page_pool_clean_page(pool, page);
+	put_page(page);
+}
+EXPORT_SYMBOL(__page_pool_put_page);
+
+static void __page_pool_empty_ring(struct page_pool *pool)
+{
+	struct page *page;
+
+	/* Empty recycle ring */
+	while ((page = ptr_ring_consume(&pool->ring))) {
+		/* Verify the refcnt invariant of cached pages */
+		if (!(page_ref_count(page) == 1))
+			pr_crit("%s() page_pool refcnt %d violation\n",
+				__func__, page_ref_count(page));
+
+		__page_pool_return_page(pool, page);
+	}
+}
+
+static void __page_pool_destroy_rcu(struct rcu_head *rcu)
+{
+	struct page_pool *pool;
+
+	pool = container_of(rcu, struct page_pool, rcu);
+
+	WARN(pool->alloc.count, "API usage violation");
+
+	__page_pool_empty_ring(pool);
+	ptr_ring_cleanup(&pool->ring, NULL);
+	kfree(pool);
+}
+
+/* Cleanup and release resources */
+void page_pool_destroy(struct page_pool *pool)
+{
+	struct page *page;
+
+	/* Empty alloc cache, assume caller made sure this is
+	 * no-longer in use, and page_pool_alloc_pages() cannot be
+	 * call concurrently.
+	 */
+	while (pool->alloc.count) {
+		page = pool->alloc.cache[--pool->alloc.count];
+		__page_pool_return_page(pool, page);
+	}
+
+	/* No more consumers should exist, but producers could still
+	 * be in-flight.
+	 */
+	__page_pool_empty_ring(pool);
+
+	/* An xdp_mem_allocator can still ref page_pool pointer */
+	call_rcu(&pool->rcu, __page_pool_destroy_rcu);
+}
+EXPORT_SYMBOL(page_pool_destroy);
diff --git a/net/core/xdp.c b/net/core/xdp.c
index 097a0f74e004..0c86b53a3a63 100644
--- a/net/core/xdp.c
+++ b/net/core/xdp.c
@@ -5,6 +5,10 @@
  */
 #include <linux/types.h>
 #include <linux/mm.h>
+#include <linux/slab.h>
+#include <linux/idr.h>
+#include <linux/rhashtable.h>
+#include <net/page_pool.h>
 
 #include <net/xdp.h>
 
@@ -13,6 +17,104 @@
 #define REG_STATE_UNREGISTERED	0x2
 #define REG_STATE_UNUSED	0x3
 
+static DEFINE_IDA(mem_id_pool);
+static DEFINE_MUTEX(mem_id_lock);
+#define MEM_ID_MAX 0xFFFE
+#define MEM_ID_MIN 1
+static int mem_id_next = MEM_ID_MIN;
+
+static bool mem_id_init; /* false */
+static struct rhashtable *mem_id_ht;
+
+struct xdp_mem_allocator {
+	struct xdp_mem_info mem;
+	union {
+		void *allocator;
+		struct page_pool *page_pool;
+	};
+	struct rhash_head node;
+	struct rcu_head rcu;
+};
+
+static u32 xdp_mem_id_hashfn(const void *data, u32 len, u32 seed)
+{
+	const u32 *k = data;
+	const u32 key = *k;
+
+	BUILD_BUG_ON(FIELD_SIZEOF(struct xdp_mem_allocator, mem.id)
+		     != sizeof(u32));
+
+	/* Use cyclic increasing ID as direct hash key, see rht_bucket_index */
+	return key << RHT_HASH_RESERVED_SPACE;
+}
+
+static int xdp_mem_id_cmp(struct rhashtable_compare_arg *arg,
+			  const void *ptr)
+{
+	const struct xdp_mem_allocator *xa = ptr;
+	u32 mem_id = *(u32 *)arg->key;
+
+	return xa->mem.id != mem_id;
+}
+
+static const struct rhashtable_params mem_id_rht_params = {
+	.nelem_hint = 64,
+	.head_offset = offsetof(struct xdp_mem_allocator, node),
+	.key_offset  = offsetof(struct xdp_mem_allocator, mem.id),
+	.key_len = FIELD_SIZEOF(struct xdp_mem_allocator, mem.id),
+	.max_size = MEM_ID_MAX,
+	.min_size = 8,
+	.automatic_shrinking = true,
+	.hashfn    = xdp_mem_id_hashfn,
+	.obj_cmpfn = xdp_mem_id_cmp,
+};
+
+static void __xdp_mem_allocator_rcu_free(struct rcu_head *rcu)
+{
+	struct xdp_mem_allocator *xa;
+
+	xa = container_of(rcu, struct xdp_mem_allocator, rcu);
+
+	/* Allow this ID to be reused */
+	ida_simple_remove(&mem_id_pool, xa->mem.id);
+
+	/* Notice, driver is expected to free the *allocator,
+	 * e.g. page_pool, and MUST also use RCU free.
+	 */
+
+	/* Poison memory */
+	xa->mem.id = 0xFFFF;
+	xa->mem.type = 0xF0F0;
+	xa->allocator = (void *)0xDEAD9001;
+
+	kfree(xa);
+}
+
+static void __xdp_rxq_info_unreg_mem_model(struct xdp_rxq_info *xdp_rxq)
+{
+	struct xdp_mem_allocator *xa;
+	int id = xdp_rxq->mem.id;
+	int err;
+
+	if (id == 0)
+		return;
+
+	mutex_lock(&mem_id_lock);
+
+	xa = rhashtable_lookup(mem_id_ht, &id, mem_id_rht_params);
+	if (!xa) {
+		mutex_unlock(&mem_id_lock);
+		return;
+	}
+
+	err = rhashtable_remove_fast(mem_id_ht, &xa->node, mem_id_rht_params);
+	WARN_ON(err);
+
+	call_rcu(&xa->rcu, __xdp_mem_allocator_rcu_free);
+
+	mutex_unlock(&mem_id_lock);
+}
+
 void xdp_rxq_info_unreg(struct xdp_rxq_info *xdp_rxq)
 {
 	/* Simplify driver cleanup code paths, allow unreg "unused" */
@@ -21,8 +123,14 @@ void xdp_rxq_info_unreg(struct xdp_rxq_info *xdp_rxq)
 
 	WARN(!(xdp_rxq->reg_state == REG_STATE_REGISTERED), "Driver BUG");
 
+	__xdp_rxq_info_unreg_mem_model(xdp_rxq);
+
 	xdp_rxq->reg_state = REG_STATE_UNREGISTERED;
 	xdp_rxq->dev = NULL;
+
+	/* Reset mem info to defaults */
+	xdp_rxq->mem.id = 0;
+	xdp_rxq->mem.type = 0;
 }
 EXPORT_SYMBOL_GPL(xdp_rxq_info_unreg);
 
@@ -71,3 +179,164 @@ bool xdp_rxq_info_is_reg(struct xdp_rxq_info *xdp_rxq)
 	return (xdp_rxq->reg_state == REG_STATE_REGISTERED);
 }
 EXPORT_SYMBOL_GPL(xdp_rxq_info_is_reg);
+
+static int __mem_id_init_hash_table(void)
+{
+	struct rhashtable *rht;
+	int ret;
+
+	if (unlikely(mem_id_init))
+		return 0;
+
+	rht = kzalloc(sizeof(*rht), GFP_KERNEL);
+	if (!rht)
+		return -ENOMEM;
+
+	ret = rhashtable_init(rht, &mem_id_rht_params);
+	if (ret < 0) {
+		kfree(rht);
+		return ret;
+	}
+	mem_id_ht = rht;
+	smp_mb(); /* mutex lock should provide enough pairing */
+	mem_id_init = true;
+
+	return 0;
+}
+
+/* Allocate a cyclic ID that maps to allocator pointer.
+ * See: https://www.kernel.org/doc/html/latest/core-api/idr.html
+ *
+ * Caller must lock mem_id_lock.
+ */
+static int __mem_id_cyclic_get(gfp_t gfp)
+{
+	int retries = 1;
+	int id;
+
+again:
+	id = ida_simple_get(&mem_id_pool, mem_id_next, MEM_ID_MAX, gfp);
+	if (id < 0) {
+		if (id == -ENOSPC) {
+			/* Cyclic allocator, reset next id */
+			if (retries--) {
+				mem_id_next = MEM_ID_MIN;
+				goto again;
+			}
+		}
+		return id; /* errno */
+	}
+	mem_id_next = id + 1;
+
+	return id;
+}
+
+static bool __is_supported_mem_type(enum xdp_mem_type type)
+{
+	if (type == MEM_TYPE_PAGE_POOL)
+		return is_page_pool_compiled_in();
+
+	if (type >= MEM_TYPE_MAX)
+		return false;
+
+	return true;
+}
+
+int xdp_rxq_info_reg_mem_model(struct xdp_rxq_info *xdp_rxq,
+			       enum xdp_mem_type type, void *allocator)
+{
+	struct xdp_mem_allocator *xdp_alloc;
+	gfp_t gfp = GFP_KERNEL;
+	int id, errno, ret;
+	void *ptr;
+
+	if (xdp_rxq->reg_state != REG_STATE_REGISTERED) {
+		WARN(1, "Missing register, driver bug");
+		return -EFAULT;
+	}
+
+	if (!__is_supported_mem_type(type))
+		return -EOPNOTSUPP;
+
+	xdp_rxq->mem.type = type;
+
+	if (!allocator) {
+		if (type == MEM_TYPE_PAGE_POOL)
+			return -EINVAL; /* Setup time check page_pool req */
+		return 0;
+	}
+
+	/* Delay init of rhashtable to save memory if feature isn't used */
+	if (!mem_id_init) {
+		mutex_lock(&mem_id_lock);
+		ret = __mem_id_init_hash_table();
+		mutex_unlock(&mem_id_lock);
+		if (ret < 0) {
+			WARN_ON(1);
+			return ret;
+		}
+	}
+
+	xdp_alloc = kzalloc(sizeof(*xdp_alloc), gfp);
+	if (!xdp_alloc)
+		return -ENOMEM;
+
+	mutex_lock(&mem_id_lock);
+	id = __mem_id_cyclic_get(gfp);
+	if (id < 0) {
+		errno = id;
+		goto err;
+	}
+	xdp_rxq->mem.id = id;
+	xdp_alloc->mem  = xdp_rxq->mem;
+	xdp_alloc->allocator = allocator;
+
+	/* Insert allocator into ID lookup table */
+	ptr = rhashtable_insert_slow(mem_id_ht, &id, &xdp_alloc->node);
+	if (IS_ERR(ptr)) {
+		errno = PTR_ERR(ptr);
+		goto err;
+	}
+
+	mutex_unlock(&mem_id_lock);
+
+	return 0;
+err:
+	mutex_unlock(&mem_id_lock);
+	kfree(xdp_alloc);
+	return errno;
+}
+EXPORT_SYMBOL_GPL(xdp_rxq_info_reg_mem_model);
+
+void xdp_return_frame(struct xdp_frame *xdpf)
+{
+	struct xdp_mem_info *mem = &xdpf->mem;
+	struct xdp_mem_allocator *xa;
+	void *data = xdpf->data;
+	struct page *page;
+
+	switch (mem->type) {
+	case MEM_TYPE_PAGE_POOL:
+		rcu_read_lock();
+		/* mem->id is valid, checked in xdp_rxq_info_reg_mem_model() */
+		xa = rhashtable_lookup(mem_id_ht, &mem->id, mem_id_rht_params);
+		page = virt_to_head_page(data);
+		if (xa)
+			page_pool_put_page(xa->page_pool, page);
+		else
+			put_page(page);
+		rcu_read_unlock();
+		break;
+	case MEM_TYPE_PAGE_SHARED:
+		page_frag_free(data);
+		break;
+	case MEM_TYPE_PAGE_ORDER0:
+		page = virt_to_page(data); /* Assumes order0 page*/
+		put_page(page);
+		break;
+	default:
+		/* Not possible, checked in xdp_rxq_info_reg_mem_model() */
+		break;
+	}
+}
+EXPORT_SYMBOL_GPL(xdp_return_frame);