0day-ci
diff --git a/‎Documentation/networking/segmentation-offloads.txt‎
Lines changed: 34 additions & 4 deletions b/‎Documentation/networking/segmentation-offloads.txt‎
Lines changed: 34 additions & 4 deletions
diff --git a/‎MAINTAINERS‎
Lines changed: 0 additions & 1 deletion b/‎MAINTAINERS‎
Lines changed: 0 additions & 1 deletion
diff --git a/‎drivers/net/ethernet/broadcom/tg3.c‎
Lines changed: 24 additions & 11 deletions b/‎drivers/net/ethernet/broadcom/tg3.c‎
Lines changed: 24 additions & 11 deletions
diff --git a/‎drivers/net/ethernet/broadcom/tg3.h‎
Lines changed: 5 additions & 0 deletions b/‎drivers/net/ethernet/broadcom/tg3.h‎
Lines changed: 5 additions & 0 deletions
diff --git a/‎drivers/net/ethernet/cavium/common/cavium_ptp.c‎
Lines changed: 2 additions & 0 deletions b/‎drivers/net/ethernet/cavium/common/cavium_ptp.c‎
Lines changed: 2 additions & 0 deletions
diff --git a/‎drivers/net/ethernet/cavium/thunder/nicvf_main.c‎
Lines changed: 26 additions & 84 deletions b/‎drivers/net/ethernet/cavium/thunder/nicvf_main.c‎
Lines changed: 26 additions & 84 deletions
@@ -13,6 +13,7 @@ The following technologies are described:
  * Generic Segmentation Offload - GSO
  * Generic Receive Offload - GRO
  * Partial Generic Segmentation Offload - GSO_PARTIAL
+ * SCTP accelleration with GSO - GSO_BY_FRAGS
 
 TCP Segmentation Offload
 ========================
@@ -49,6 +50,10 @@ datagram into multiple IPv4 fragments.  Many of the requirements for UDP
 fragmentation offload are the same as TSO.  However the IPv4 ID for
 fragments should not increment as a single IPv4 datagram is fragmented.
 
+UFO is deprecated: modern kernels will no longer generate UFO skbs, but can
+still receive them from tuntap and similar devices. Offload of UDP-based
+tunnel protocols is still supported.
+
 IPIP, SIT, GRE, UDP Tunnel, and Remote Checksum Offloads
 ========================================================
 
@@ -83,10 +88,10 @@ SKB_GSO_UDP_TUNNEL_CSUM.  These two additional tunnel types reflect the
 fact that the outer header also requests to have a non-zero checksum
 included in the outer header.
 
-Finally there is SKB_GSO_REMCSUM which indicates that a given tunnel header
-has requested a remote checksum offload.  In this case the inner headers
-will be left with a partial checksum and only the outer header checksum
-will be computed.
+Finally there is SKB_GSO_TUNNEL_REMCSUM which indicates that a given tunnel
+header has requested a remote checksum offload.  In this case the inner
+headers will be left with a partial checksum and only the outer header
+checksum will be computed.
 
 Generic Segmentation Offload
 ============================
@@ -128,3 +133,28 @@ values for if the header was simply duplicated.  The one exception to this
 is the outer IPv4 ID field.  It is up to the device drivers to guarantee
 that the IPv4 ID field is incremented in the case that a given header does
 not have the DF bit set.
+
+SCTP accelleration with GSO
+===========================
+
+SCTP - despite the lack of hardware support - can still take advantage of
+GSO to pass one large packet through the network stack, rather than
+multiple small packets.
+
+This requires a different approach to other offloads, as SCTP packets
+cannot be just segmented to (P)MTU. Rather, the chunks must be contained in
+IP segments, padding respected. So unlike regular GSO, SCTP can't just
+generate a big skb, set gso_size to the fragmentation point and deliver it
+to IP layer.
+
+Instead, the SCTP protocol layer builds an skb with the segments correctly
+padded and stored as chained skbs, and skb_segment() splits based on those.
+To signal this, gso_size is set to the special value GSO_BY_FRAGS.
+
+Therefore, any code in the core networking stack must be aware of the
+possibility that gso_size will be GSO_BY_FRAGS and handle that case
+appropriately. (For size checks, the skb_gso_validate_*_len family of
+helpers do this automatically.)
+
+This also affects drivers with the NETIF_F_FRAGLIST & NETIF_F_GSO_SCTP bits
+set. Note also that NETIF_F_GSO_SCTP is included in NETIF_F_GSO_SOFTWARE.
@@ -7909,7 +7909,6 @@ S:	Maintained
 F:	scripts/leaking_addresses.pl
 
 LED SUBSYSTEM
-M:	Richard Purdie <[email protected]>
 M:	Jacek Anaszewski <[email protected]>
 M:	Pavel Machek <[email protected]>
 L:	[email protected]
 
@@ -820,7 +820,7 @@ static int tg3_ape_event_lock(struct tg3 *tp, u32 timeout_us)
 
 		tg3_ape_unlock(tp, TG3_APE_LOCK_MEM);
 
-		udelay(10);
+		usleep_range(10, 20);
 		timeout_us -= (timeout_us > 10) ? 10 : timeout_us;
 	}
 
@@ -922,8 +922,8 @@ static int tg3_ape_send_event(struct tg3 *tp, u32 event)
 	if (!(apedata & APE_FW_STATUS_READY))
 		return -EAGAIN;
 
-	/* Wait for up to 1 millisecond for APE to service previous event. */
-	err = tg3_ape_event_lock(tp, 1000);
+	/* Wait for up to 20 millisecond for APE to service previous event. */
+	err = tg3_ape_event_lock(tp, 20000);
 	if (err)
 		return err;
 
@@ -946,6 +946,7 @@ static void tg3_ape_driver_state_change(struct tg3 *tp, int kind)
 
 	switch (kind) {
 	case RESET_KIND_INIT:
+		tg3_ape_write32(tp, TG3_APE_HOST_HEARTBEAT_COUNT, tp->ape_hb++);
 		tg3_ape_write32(tp, TG3_APE_HOST_SEG_SIG,
 				APE_HOST_SEG_SIG_MAGIC);
 		tg3_ape_write32(tp, TG3_APE_HOST_SEG_LEN,
@@ -962,13 +963,6 @@ static void tg3_ape_driver_state_change(struct tg3 *tp, int kind)
 		event = APE_EVENT_STATUS_STATE_START;
 		break;
 	case RESET_KIND_SHUTDOWN:
-		/* With the interface we are currently using,
-		 * APE does not track driver state.  Wiping
-		 * out the HOST SEGMENT SIGNATURE forces
-		 * the APE to assume OS absent status.
-		 */
-		tg3_ape_write32(tp, TG3_APE_HOST_SEG_SIG, 0x0);
-
 		if (device_may_wakeup(&tp->pdev->dev) &&
 		    tg3_flag(tp, WOL_ENABLE)) {
 			tg3_ape_write32(tp, TG3_APE_HOST_WOL_SPEED,
@@ -990,6 +984,18 @@ static void tg3_ape_driver_state_change(struct tg3 *tp, int kind)
 	tg3_ape_send_event(tp, event);
 }
 
+static void tg3_send_ape_heartbeat(struct tg3 *tp,
+				   unsigned long interval)
+{
+	/* Check if hb interval has exceeded */
+	if (!tg3_flag(tp, ENABLE_APE) ||
+	    time_before(jiffies, tp->ape_hb_jiffies + interval))
+		return;
+
+	tg3_ape_write32(tp, TG3_APE_HOST_HEARTBEAT_COUNT, tp->ape_hb++);
+	tp->ape_hb_jiffies = jiffies;
+}
+
 static void tg3_disable_ints(struct tg3 *tp)
 {
 	int i;
@@ -7262,6 +7268,7 @@ static int tg3_poll_msix(struct napi_struct *napi, int budget)
 		}
 	}
 
+	tg3_send_ape_heartbeat(tp, TG3_APE_HB_INTERVAL << 1);
 	return work_done;
 
 tx_recovery:
@@ -7344,6 +7351,7 @@ static int tg3_poll(struct napi_struct *napi, int budget)
 		}
 	}
 
+	tg3_send_ape_heartbeat(tp, TG3_APE_HB_INTERVAL << 1);
 	return work_done;
 
 tx_recovery:
@@ -10732,7 +10740,7 @@ static int tg3_reset_hw(struct tg3 *tp, bool reset_phy)
 	if (tg3_flag(tp, ENABLE_APE))
 		/* Write our heartbeat update interval to APE. */
 		tg3_ape_write32(tp, TG3_APE_HOST_HEARTBEAT_INT_MS,
-				APE_HOST_HEARTBEAT_INT_DISABLE);
+				APE_HOST_HEARTBEAT_INT_5SEC);
 
 	tg3_write_sig_post_reset(tp, RESET_KIND_INIT);
 
@@ -11077,6 +11085,9 @@ static void tg3_timer(struct timer_list *t)
 		tp->asf_counter = tp->asf_multiplier;
 	}
 
+	/* Update the APE heartbeat every 5 seconds.*/
+	tg3_send_ape_heartbeat(tp, TG3_APE_HB_INTERVAL);
+
 	spin_unlock(&tp->lock);
 
 restart_timer:
@@ -16653,6 +16664,8 @@ static int tg3_get_invariants(struct tg3 *tp, const struct pci_device_id *ent)
 				       pci_state_reg);
 
 		tg3_ape_lock_init(tp);
+		tp->ape_hb_interval =
+			msecs_to_jiffies(APE_HOST_HEARTBEAT_INT_5SEC);
 	}
 
 	/* Set up tp->grc_local_ctrl before calling
 
@@ -2508,6 +2508,7 @@
 #define TG3_APE_LOCK_PHY3		5
 #define TG3_APE_LOCK_GPIO		7
 
+#define TG3_APE_HB_INTERVAL             (tp->ape_hb_interval)
 #define TG3_EEPROM_SB_F1R2_MBA_OFF	0x10
 
 
@@ -3423,6 +3424,10 @@ struct tg3 {
 	struct device			*hwmon_dev;
 	bool				link_up;
 	bool				pcierr_recovery;
+
+	u32                             ape_hb;
+	unsigned long                   ape_hb_interval;
+	unsigned long                   ape_hb_jiffies;
 };
 
 /* Accessor macros for chip and asic attributes
 
@@ -75,6 +75,8 @@ EXPORT_SYMBOL(cavium_ptp_get);
 
 void cavium_ptp_put(struct cavium_ptp *ptp)
 {
+	if (!ptp)
+		return;
 	pci_dev_put(ptp->pdev);
 }
 EXPORT_SYMBOL(cavium_ptp_put);
 
@@ -67,11 +67,6 @@ module_param(cpi_alg, int, S_IRUGO);
 MODULE_PARM_DESC(cpi_alg,
 		 "PFC algorithm (0=none, 1=VLAN, 2=VLAN16, 3=IP Diffserv)");
 
-struct nicvf_xdp_tx {
-	u64 dma_addr;
-	u8  qidx;
-};
-
 static inline u8 nicvf_netdev_qidx(struct nicvf *nic, u8 qidx)
 {
 	if (nic->sqs_mode)
@@ -507,29 +502,14 @@ static int nicvf_init_resources(struct nicvf *nic)
 	return 0;
 }
 
-static void nicvf_unmap_page(struct nicvf *nic, struct page *page, u64 dma_addr)
-{
-	/* Check if it's a recycled page, if not unmap the DMA mapping.
-	 * Recycled page holds an extra reference.
-	 */
-	if (page_ref_count(page) == 1) {
-		dma_addr &= PAGE_MASK;
-		dma_unmap_page_attrs(&nic->pdev->dev, dma_addr,
-				     RCV_FRAG_LEN + XDP_HEADROOM,
-				     DMA_FROM_DEVICE,
-				     DMA_ATTR_SKIP_CPU_SYNC);
-	}
-}
-
 static inline bool nicvf_xdp_rx(struct nicvf *nic, struct bpf_prog *prog,
 				struct cqe_rx_t *cqe_rx, struct snd_queue *sq,
 				struct rcv_queue *rq, struct sk_buff **skb)
 {
 	struct xdp_buff xdp;
 	struct page *page;
-	struct nicvf_xdp_tx *xdp_tx = NULL;
 	u32 action;
-	u16 len, err, offset = 0;
+	u16 len, offset = 0;
 	u64 dma_addr, cpu_addr;
 	void *orig_data;
 
@@ -543,7 +523,7 @@ static inline bool nicvf_xdp_rx(struct nicvf *nic, struct bpf_prog *prog,
 	cpu_addr = (u64)phys_to_virt(cpu_addr);
 	page = virt_to_page((void *)cpu_addr);
 
-	xdp.data_hard_start = page_address(page) + RCV_BUF_HEADROOM;
+	xdp.data_hard_start = page_address(page);
 	xdp.data = (void *)cpu_addr;
 	xdp_set_data_meta_invalid(&xdp);
 	xdp.data_end = xdp.data + len;
@@ -563,7 +543,18 @@ static inline bool nicvf_xdp_rx(struct nicvf *nic, struct bpf_prog *prog,
 
 	switch (action) {
 	case XDP_PASS:
-		nicvf_unmap_page(nic, page, dma_addr);
+		/* Check if it's a recycled page, if not
+		 * unmap the DMA mapping.
+		 *
+		 * Recycled page holds an extra reference.
+		 */
+		if (page_ref_count(page) == 1) {
+			dma_addr &= PAGE_MASK;
+			dma_unmap_page_attrs(&nic->pdev->dev, dma_addr,
+					     RCV_FRAG_LEN + XDP_PACKET_HEADROOM,
+					     DMA_FROM_DEVICE,
+					     DMA_ATTR_SKIP_CPU_SYNC);
+		}
 
 		/* Build SKB and pass on packet to network stack */
 		*skb = build_skb(xdp.data,
@@ -576,28 +567,25 @@ static inline bool nicvf_xdp_rx(struct nicvf *nic, struct bpf_prog *prog,
 	case XDP_TX:
 		nicvf_xdp_sq_append_pkt(nic, sq, (u64)xdp.data, dma_addr, len);
 		return true;
-	case XDP_REDIRECT:
-		/* Save DMA address for use while transmitting */
-		xdp_tx = (struct nicvf_xdp_tx *)page_address(page);
-		xdp_tx->dma_addr = dma_addr;
-		xdp_tx->qidx = nicvf_netdev_qidx(nic, cqe_rx->rq_idx);
-
-		err = xdp_do_redirect(nic->pnicvf->netdev, &xdp, prog);
-		if (!err)
-			return true;
-
-		/* Free the page on error */
-		nicvf_unmap_page(nic, page, dma_addr);
-		put_page(page);
-		break;
 	default:
 		bpf_warn_invalid_xdp_action(action);
 		/* fall through */
 	case XDP_ABORTED:
 		trace_xdp_exception(nic->netdev, prog, action);
 		/* fall through */
 	case XDP_DROP:
-		nicvf_unmap_page(nic, page, dma_addr);
+		/* Check if it's a recycled page, if not
+		 * unmap the DMA mapping.
+		 *
+		 * Recycled page holds an extra reference.
+		 */
+		if (page_ref_count(page) == 1) {
+			dma_addr &= PAGE_MASK;
+			dma_unmap_page_attrs(&nic->pdev->dev, dma_addr,
+					     RCV_FRAG_LEN + XDP_PACKET_HEADROOM,
+					     DMA_FROM_DEVICE,
+					     DMA_ATTR_SKIP_CPU_SYNC);
+		}
 		put_page(page);
 		return true;
 	}
@@ -1864,50 +1852,6 @@ static int nicvf_xdp(struct net_device *netdev, struct netdev_bpf *xdp)
 	}
 }
 
-static int nicvf_xdp_xmit(struct net_device *netdev, struct xdp_buff *xdp)
-{
-	struct nicvf *nic = netdev_priv(netdev);
-	struct nicvf *snic = nic;
-	struct nicvf_xdp_tx *xdp_tx;
-	struct snd_queue *sq;
-	struct page *page;
-	int err, qidx;
-
-	if (!netif_running(netdev) || !nic->xdp_prog)
-		return -EINVAL;
-
-	page = virt_to_page(xdp->data);
-	xdp_tx = (struct nicvf_xdp_tx *)page_address(page);
-	qidx = xdp_tx->qidx;
-
-	if (xdp_tx->qidx >= nic->xdp_tx_queues)
-		return -EINVAL;
-
-	/* Get secondary Qset's info */
-	if (xdp_tx->qidx >= MAX_SND_QUEUES_PER_QS) {
-		qidx = xdp_tx->qidx / MAX_SND_QUEUES_PER_QS;
-		snic = (struct nicvf *)nic->snicvf[qidx - 1];
-		if (!snic)
-			return -EINVAL;
-		qidx = xdp_tx->qidx % MAX_SND_QUEUES_PER_QS;
-	}
-
-	sq = &snic->qs->sq[qidx];
-	err = nicvf_xdp_sq_append_pkt(snic, sq, (u64)xdp->data,
-				      xdp_tx->dma_addr,
-				      xdp->data_end - xdp->data);
-	if (err)
-		return -ENOMEM;
-
-	nicvf_xdp_sq_doorbell(snic, sq, qidx);
-	return 0;
-}
-
-static void nicvf_xdp_flush(struct net_device *dev)
-{
-	return;
-}
-
 static int nicvf_config_hwtstamp(struct net_device *netdev, struct ifreq *ifr)
 {
 	struct hwtstamp_config config;
@@ -1986,8 +1930,6 @@ static const struct net_device_ops nicvf_netdev_ops = {
 	.ndo_fix_features       = nicvf_fix_features,
 	.ndo_set_features       = nicvf_set_features,
 	.ndo_bpf		= nicvf_xdp,
-	.ndo_xdp_xmit		= nicvf_xdp_xmit,
-	.ndo_xdp_flush          = nicvf_xdp_flush,
 	.ndo_do_ioctl           = nicvf_ioctl,
 };
Original file line number	Diff line number	Diff line change
`@@ -75,6 +75,8 @@ EXPORT_SYMBOL(cavium_ptp_get);`
`75`	`75`
`76`	`76`	`void cavium_ptp_put(struct cavium_ptp *ptp)`
`77`	`77`	`{`
	`78`	`+ if (!ptp)`
	`79`	`+ return;`
`78`	`80`	`pci_dev_put(ptp->pdev);`
`79`	`81`	`}`
`80`	`82`	`EXPORT_SYMBOL(cavium_ptp_put);`