Merge branch 'bulk-cpumap-redirect'

Jesper Dangaard Brouer says:

====================
This patchset utilize a number of different kernel bulk APIs for optimizing
the performance for the XDP cpumap redirect feature.

Benchmark details are available here:
https://github.com/xdp-project/xdp-project/blob/master/areas/cpumap/cpumap03-optimizations.org

Performance measurements can be considered micro benchmarks, as they measure
dropping packets at different stages in the network stack.
Summary based on above:

Baseline benchmarks
- baseline-redirect: UdpNoPorts: 3,180,074
- baseline-redirect: iptables-raw drop: 6,193,534

Patch1: bpf: cpumap use ptr_ring_consume_batched
- redirect: UdpNoPorts: 3,327,729
- redirect: iptables-raw drop: 6,321,540

Patch2: net: core: introduce build_skb_around
- redirect: UdpNoPorts: 3,221,303
- redirect: iptables-raw drop: 6,320,066

Patch3: bpf: cpumap do bulk allocation of SKBs
- redirect: UdpNoPorts: 3,290,563
- redirect: iptables-raw drop: 6,650,112

Patch4: bpf: cpumap memory prefetchw optimizations for struct page
- redirect: UdpNoPorts: 3,520,250
- redirect: iptables-raw drop: 7,649,604

In this V2 submission I have chosen drop the SKB-list patch using
netif_receive_skb_list() as it was not showing a performance improvement for
these micro benchmarks.
====================

Signed-off-by: Alexei Starovoitov <[e-mail]>

include/linux/skbuff.h

···
 			    int node);
 struct sk_buff *__build_skb(void *data, unsigned int frag_size);
 struct sk_buff *build_skb(void *data, unsigned int frag_size);
+struct sk_buff *build_skb_around(struct sk_buff *skb,
+				 void *data, unsigned int frag_size);
 
 /**
  * alloc_skb - allocate a network buffer

kernel/bpf/cpumap.c

···
 }
 
 static struct sk_buff *cpu_map_build_skb(struct bpf_cpu_map_entry *rcpu,
-					 struct xdp_frame *xdpf)
+					 struct xdp_frame *xdpf,
+					 struct sk_buff *skb)
 {
 	unsigned int hard_start_headroom;
 	unsigned int frame_size;
 	void *pkt_data_start;
-	struct sk_buff *skb;
 
 	/* Part of headroom was reserved to xdpf */
 	hard_start_headroom = sizeof(struct xdp_frame) +  xdpf->headroom;
···
 		SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
 
 	pkt_data_start = xdpf->data - hard_start_headroom;
-	skb = build_skb(pkt_data_start, frame_size);
-	if (!skb)
+	skb = build_skb_around(skb, pkt_data_start, frame_size);
+	if (unlikely(!skb))
 		return NULL;
 
 	skb_reserve(skb, hard_start_headroom);
···
 	}
 }
 
+#define CPUMAP_BATCH 8
+
 static int cpu_map_kthread_run(void *data)
 {
 	struct bpf_cpu_map_entry *rcpu = data;
···
 	 * kthread_stop signal until queue is empty.
 	 */
 	while (!kthread_should_stop() || !__ptr_ring_empty(rcpu->queue)) {
-		unsigned int processed = 0, drops = 0, sched = 0;
-		struct xdp_frame *xdpf;
+		unsigned int drops = 0, sched = 0;
+		void *frames[CPUMAP_BATCH];
+		void *skbs[CPUMAP_BATCH];
+		gfp_t gfp = __GFP_ZERO | GFP_ATOMIC;
+		int i, n, m;
 
 		/* Release CPU reschedule checks */
 		if (__ptr_ring_empty(rcpu->queue)) {
···
 			sched = cond_resched();
 		}
 
-		/* Process packets in rcpu->queue */
-		local_bh_disable();
 		/*
 		 * The bpf_cpu_map_entry is single consumer, with this
 		 * kthread CPU pinned. Lockless access to ptr_ring
 		 * consume side valid as no-resize allowed of queue.
 		 */
-		while ((xdpf = __ptr_ring_consume(rcpu->queue))) {
-			struct sk_buff *skb;
+		n = ptr_ring_consume_batched(rcpu->queue, frames, CPUMAP_BATCH);
+
+		for (i = 0; i < n; i++) {
+			void *f = frames[i];
+			struct page *page = virt_to_page(f);
+
+			/* Bring struct page memory area to curr CPU. Read by
+			 * build_skb_around via page_is_pfmemalloc(), and when
+			 * freed written by page_frag_free call.
+			 */
+			prefetchw(page);
+		}
+
+		m = kmem_cache_alloc_bulk(skbuff_head_cache, gfp, n, skbs);
+		if (unlikely(m == 0)) {
+			for (i = 0; i < n; i++)
+				skbs[i] = NULL; /* effect: xdp_return_frame */
+			drops = n;
+		}
+
+		local_bh_disable();
+		for (i = 0; i < n; i++) {
+			struct xdp_frame *xdpf = frames[i];
+			struct sk_buff *skb = skbs[i];
 			int ret;
 
-			skb = cpu_map_build_skb(rcpu, xdpf);
+			skb = cpu_map_build_skb(rcpu, xdpf, skb);
 			if (!skb) {
 				xdp_return_frame(xdpf);
 				continue;
···
 			ret = netif_receive_skb_core(skb);
 			if (ret == NET_RX_DROP)
 				drops++;
-
-			/* Limit BH-disable period */
-			if (++processed == 8)
-				break;
 		}
 		/* Feedback loop via tracepoint */
-		trace_xdp_cpumap_kthread(rcpu->map_id, processed, drops, sched);
+		trace_xdp_cpumap_kthread(rcpu->map_id, n, drops, sched);
 
 		local_bh_enable(); /* resched point, may call do_softirq() */
 	}

net/core/skbuff.c

···
 }
 EXPORT_SYMBOL(__alloc_skb);
 
+/* Caller must provide SKB that is memset cleared */
+static struct sk_buff *__build_skb_around(struct sk_buff *skb,
+					  void *data, unsigned int frag_size)
+{
+	struct skb_shared_info *shinfo;
+	unsigned int size = frag_size ? : ksize(data);
+
+	size -= SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
+
+	/* Assumes caller memset cleared SKB */
+	skb->truesize = SKB_TRUESIZE(size);
+	refcount_set(&skb->users, 1);
+	skb->head = data;
+	skb->data = data;
+	skb_reset_tail_pointer(skb);
+	skb->end = skb->tail + size;
+	skb->mac_header = (typeof(skb->mac_header))~0U;
+	skb->transport_header = (typeof(skb->transport_header))~0U;
+
+	/* make sure we initialize shinfo sequentially */
+	shinfo = skb_shinfo(skb);
+	memset(shinfo, 0, offsetof(struct skb_shared_info, dataref));
+	atomic_set(&shinfo->dataref, 1);
+
+	return skb;
+}
+
 /**
  * __build_skb - build a network buffer
  * @data: data buffer provided by caller
···
  */
 struct sk_buff *__build_skb(void *data, unsigned int frag_size)
 {
-	struct skb_shared_info *shinfo;
 	struct sk_buff *skb;
-	unsigned int size = frag_size ? : ksize(data);
 
 	skb = kmem_cache_alloc(skbuff_head_cache, GFP_ATOMIC);
-	if (!skb)
+	if (unlikely(!skb))
 		return NULL;
 
-	size -= SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
-
 	memset(skb, 0, offsetof(struct sk_buff, tail));
-	skb->truesize = SKB_TRUESIZE(size);
-	refcount_set(&skb->users, 1);
-	skb->head = data;
-	skb->data = data;
-	skb_reset_tail_pointer(skb);
-	skb->end = skb->tail + size;
-	skb->mac_header = (typeof(skb->mac_header))~0U;
-	skb->transport_header = (typeof(skb->transport_header))~0U;
 
-	/* make sure we initialize shinfo sequentially */
-	shinfo = skb_shinfo(skb);
-	memset(shinfo, 0, offsetof(struct skb_shared_info, dataref));
-	atomic_set(&shinfo->dataref, 1);
-
-	return skb;
+	return __build_skb_around(skb, data, frag_size);
 }
 
 /* build_skb() is wrapper over __build_skb(), that specifically
···
 	return skb;
 }
 EXPORT_SYMBOL(build_skb);
+
+/**
+ * build_skb_around - build a network buffer around provided skb
+ * @skb: sk_buff provide by caller, must be memset cleared
+ * @data: data buffer provided by caller
+ * @frag_size: size of data, or 0 if head was kmalloced
+ */
+struct sk_buff *build_skb_around(struct sk_buff *skb,
+				 void *data, unsigned int frag_size)
+{
+	if (unlikely(!skb))
+		return NULL;
+
+	skb = __build_skb_around(skb, data, frag_size);
+
+	if (skb && frag_size) {
+		skb->head_frag = 1;
+		if (page_is_pfmemalloc(virt_to_head_page(data)))
+			skb->pfmemalloc = 1;
+	}
+	return skb;
+}
+EXPORT_SYMBOL(build_skb_around);
 
 #define NAPI_SKB_CACHE_SIZE	64