Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking updates from David Miller:

 1) More iov_iter conversion work from Al Viro.

    [ The "crypto: switch af_alg_make_sg() to iov_iter" commit was
      wrong, and this pull actually adds an extra commit on top of the
      branch I'm pulling to fix that up, so that the pre-merge state is
      ok.   - Linus ]

 2) Various optimizations to the ipv4 forwarding information base trie
    lookup implementation.  From Alexander Duyck.

 3) Remove sock_iocb altogether, from CHristoph Hellwig.

 4) Allow congestion control algorithm selection via routing metrics.
    From Daniel Borkmann.

 5) Make ipv4 uncached route list per-cpu, from Eric Dumazet.

 6) Handle rfs hash collisions more gracefully, also from Eric Dumazet.

 7) Add xmit_more support to r8169, e1000, and e1000e drivers.  From
    Florian Westphal.

 8) Transparent Ethernet Bridging support for GRO, from Jesse Gross.

 9) Add BPF packet actions to packet scheduler, from Jiri Pirko.

10) Add support for uniqu flow IDs to openvswitch, from Joe Stringer.

11) New NetCP ethernet driver, from Muralidharan Karicheri and Wingman
    Kwok.

12) More sanely handle out-of-window dupacks, which can result in
    serious ACK storms.  From Neal Cardwell.

13) Various rhashtable bug fixes and enhancements, from Herbert Xu,
    Patrick McHardy, and Thomas Graf.

14) Support xmit_more in be2net, from Sathya Perla.

15) Group Policy extensions for vxlan, from Thomas Graf.

16) Remove Checksum Offload support for vxlan, from Tom Herbert.

17) Like ipv4, support lockless transmit over ipv6 UDP sockets.  From
    Vlad Yasevich.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1494+1 commits)
  crypto: fix af_alg_make_sg() conversion to iov_iter
  ipv4: Namespecify TCP PMTU mechanism
  i40e: Fix for stats init function call in Rx setup
  tcp: don't include Fast Open option in SYN-ACK on pure SYN-data
  openvswitch: Only set TUNNEL_VXLAN_OPT if VXLAN-GBP metadata is set
  ipv6: Make __ipv6_select_ident static
  ipv6: Fix fragment id assignment on LE arches.
  bridge: Fix inability to add non-vlan fdb entry
  net: Mellanox: Delete unnecessary checks before the function call "vunmap"
  cxgb4: Add support in cxgb4 to get expansion rom version via ethtool
  ethtool: rename reserved1 memeber in ethtool_drvinfo for expansion ROM version
  net: dsa: Remove redundant phy_attach()
  IB/mlx4: Reset flow support for IB kernel ULPs
  IB/mlx4: Always use the correct port for mirrored multicast attachments
  net/bonding: Fix potential bad memory access during bonding events
  tipc: remove tipc_snprintf
  tipc: nl compat add noop and remove legacy nl framework
  tipc: convert legacy nl stats show to nl compat
  tipc: convert legacy nl net id get to nl compat
  tipc: convert legacy nl net id set to nl compat
  ...

1 files changed, 109 insertions, 42 deletions

diff --git a/drivers/net/wireless/ath/wil6210/txrx.c b/drivers/net/wireless/ath/wil6210/txrx.c
index e3f8bdce5abc..8439f65db259 100644
--- a/drivers/net/wireless/ath/wil6210/txrx.c
+++ b/drivers/net/wireless/ath/wil6210/txrx.c
@@ -463,7 +463,7 @@ static struct sk_buff *wil_vring_reap_rx(struct wil6210_priv *wil,
 	 * and in case of error drop the packet
 	 * higher stack layers will handle retransmission (if required)
 	 */
-	if (d->dma.status & RX_DMA_STATUS_L4_IDENT) {
+	if (d->dma.status & RX_DMA_STATUS_L4I) {
 		/* L4 protocol identified, csum calculated */
 		if ((d->dma.error & RX_DMA_ERROR_L4_ERR) == 0)
 			skb->ip_summed = CHECKSUM_UNNECESSARY;
@@ -581,14 +581,8 @@ void wil_rx_handle(struct wil6210_priv *wil, int *quota)
 			skb->protocol = htons(ETH_P_802_2);
 			wil_netif_rx_any(skb, ndev);
 		} else {
-			struct ethhdr *eth = (void *)skb->data;
-
 			skb->protocol = eth_type_trans(skb, ndev);
-
-			if (is_unicast_ether_addr(eth->h_dest))
-				wil_rx_reorder(wil, skb);
-			else
-				wil_netif_rx_any(skb, ndev);
+			wil_rx_reorder(wil, skb);
 		}
 	}
 	wil_rx_refill(wil, v->size);
@@ -645,7 +639,7 @@ int wil_vring_init_tx(struct wil6210_priv *wil, int id, int size,
 		.vring_cfg = {
 			.tx_sw_ring = {
 				.max_mpdu_size =
-					cpu_to_le16(mtu_max + ETH_HLEN),
+					cpu_to_le16(wil_mtu2macbuf(mtu_max)),
 				.ring_size = cpu_to_le16(size),
 			},
 			.ringid = id,
@@ -653,7 +647,7 @@ int wil_vring_init_tx(struct wil6210_priv *wil, int id, int size,
 			.encap_trans_type = WMI_VRING_ENC_TYPE_802_3,
 			.mac_ctrl = 0,
 			.to_resolution = 0,
-			.agg_max_wsize = 16,
+			.agg_max_wsize = 0,
 			.schd_params = {
 				.priority = cpu_to_le16(0),
 				.timeslot_us = cpu_to_le16(0xfff),
@@ -677,6 +671,7 @@ int wil_vring_init_tx(struct wil6210_priv *wil, int id, int size,
 	}
 
 	memset(txdata, 0, sizeof(*txdata));
+	spin_lock_init(&txdata->lock);
 	vring->size = size;
 	rc = wil_vring_alloc(wil, vring);
 	if (rc)
@@ -701,6 +696,8 @@ int wil_vring_init_tx(struct wil6210_priv *wil, int id, int size,
 	vring->hwtail = le32_to_cpu(reply.cmd.tx_vring_tail_ptr);
 
 	txdata->enabled = 1;
+	if (wil->sta[cid].data_port_open && (agg_wsize >= 0))
+		wil_addba_tx_request(wil, id, agg_wsize);
 
 	return 0;
  out_free:
@@ -713,6 +710,7 @@ int wil_vring_init_tx(struct wil6210_priv *wil, int id, int size,
 void wil_vring_fini_tx(struct wil6210_priv *wil, int id)
 {
 	struct vring *vring = &wil->vring_tx[id];
+	struct vring_tx_data *txdata = &wil->vring_tx_data[id];
 
 	WARN_ON(!mutex_is_locked(&wil->mutex));
 
@@ -721,12 +719,15 @@ void wil_vring_fini_tx(struct wil6210_priv *wil, int id)
 
 	wil_dbg_misc(wil, "%s() id=%d\n", __func__, id);
 
+	spin_lock_bh(&txdata->lock);
+	txdata->enabled = 0; /* no Tx can be in progress or start anew */
+	spin_unlock_bh(&txdata->lock);
 	/* make sure NAPI won't touch this vring */
-	wil->vring_tx_data[id].enabled = 0;
-	if (test_bit(wil_status_napi_en, &wil->status))
+	if (test_bit(wil_status_napi_en, wil->status))
 		napi_synchronize(&wil->napi_tx);
 
 	wil_vring_free(wil, vring, 1);
+	memset(txdata, 0, sizeof(*txdata));
 }
 
 static struct vring *wil_find_tx_vring(struct wil6210_priv *wil,
@@ -773,6 +774,38 @@ static void wil_set_da_for_vring(struct wil6210_priv *wil,
 
 static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
 			struct sk_buff *skb);
+
+static struct vring *wil_find_tx_vring_sta(struct wil6210_priv *wil,
+					   struct sk_buff *skb)
+{
+	struct vring *v;
+	int i;
+	u8 cid;
+
+	/* In the STA mode, it is expected to have only 1 VRING
+	 * for the AP we connected to.
+	 * find 1-st vring and see whether it is eligible for data
+	 */
+	for (i = 0; i < WIL6210_MAX_TX_RINGS; i++) {
+		v = &wil->vring_tx[i];
+		if (!v->va)
+			continue;
+
+		cid = wil->vring2cid_tid[i][0];
+		if (!wil->sta[cid].data_port_open &&
+		    (skb->protocol != cpu_to_be16(ETH_P_PAE)))
+			break;
+
+		wil_dbg_txrx(wil, "Tx -> ring %d\n", i);
+
+		return v;
+	}
+
+	wil_dbg_txrx(wil, "Tx while no vrings active?\n");
+
+	return NULL;
+}
+
 /*
  * Find 1-st vring and return it; set dest address for this vring in skb
  * duplicate skb and send it to other active vrings
@@ -843,9 +876,6 @@ static int wil_tx_desc_map(struct vring_tx_desc *d, dma_addr_t pa, u32 len,
 	d->mac.d[1] = 0;
 	d->mac.d[2] = 0;
 	d->mac.ucode_cmd = 0;
-	/* use dst index 0 */
-	d->mac.d[1] |= BIT(MAC_CFG_DESC_TX_1_DST_INDEX_EN_POS) |
-		       (0 << MAC_CFG_DESC_TX_1_DST_INDEX_POS);
 	/* translation type:  0 - bypass; 1 - 802.3; 2 - native wifi */
 	d->mac.d[2] = BIT(MAC_CFG_DESC_TX_2_SNAP_HDR_INSERTION_EN_POS) |
 		      (1 << MAC_CFG_DESC_TX_2_L2_TRANSLATION_TYPE_POS);
@@ -908,8 +938,8 @@ static int wil_tx_desc_offload_cksum_set(struct wil6210_priv *wil,
 	return 0;
 }
 
-static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
-			struct sk_buff *skb)
+static int __wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
+			  struct sk_buff *skb)
 {
 	struct device *dev = wil_to_dev(wil);
 	struct vring_tx_desc dd, *d = &dd;
@@ -925,18 +955,21 @@ static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
 
 	wil_dbg_txrx(wil, "%s()\n", __func__);
 
+	if (unlikely(!txdata->enabled))
+		return -EINVAL;
+
 	if (avail < 1 + nr_frags) {
 		wil_err_ratelimited(wil,
-				    "Tx ring full. No space for %d fragments\n",
-				    1 + nr_frags);
+				    "Tx ring[%2d] full. No space for %d fragments\n",
+				    vring_index, 1 + nr_frags);
 		return -ENOMEM;
 	}
 	_d = &vring->va[i].tx;
 
 	pa = dma_map_single(dev, skb->data, skb_headlen(skb), DMA_TO_DEVICE);
 
-	wil_dbg_txrx(wil, "Tx skb %d bytes 0x%p -> %pad\n", skb_headlen(skb),
-		     skb->data, &pa);
+	wil_dbg_txrx(wil, "Tx[%2d] skb %d bytes 0x%p -> %pad\n", vring_index,
+		     skb_headlen(skb), skb->data, &pa);
 	wil_hex_dump_txrx("Tx ", DUMP_PREFIX_OFFSET, 16, 1,
 			  skb->data, skb_headlen(skb), false);
 
@@ -947,15 +980,13 @@ static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
 	wil_tx_desc_map(d, pa, skb_headlen(skb), vring_index);
 	/* Process TCP/UDP checksum offloading */
 	if (wil_tx_desc_offload_cksum_set(wil, d, skb)) {
-		wil_err(wil, "VRING #%d Failed to set cksum, drop packet\n",
+		wil_err(wil, "Tx[%2d] Failed to set cksum, drop packet\n",
 			vring_index);
 		goto dma_error;
 	}
 
 	vring->ctx[i].nr_frags = nr_frags;
 	wil_tx_desc_set_nr_frags(d, nr_frags);
-	if (nr_frags)
-		*_d = *d;
 
 	/* middle segments */
 	for (; f < nr_frags; f++) {
@@ -963,6 +994,10 @@ static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
 				&skb_shinfo(skb)->frags[f];
 		int len = skb_frag_size(frag);
 
+		*_d = *d;
+		wil_dbg_txrx(wil, "Tx[%2d] desc[%4d]\n", vring_index, i);
+		wil_hex_dump_txrx("TxD ", DUMP_PREFIX_NONE, 32, 4,
+				  (const void *)d, sizeof(*d), false);
 		i = (swhead + f + 1) % vring->size;
 		_d = &vring->va[i].tx;
 		pa = skb_frag_dma_map(dev, frag, 0, skb_frag_size(frag),
@@ -976,13 +1011,15 @@ static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
 		 * it will succeed here too
 		 */
 		wil_tx_desc_offload_cksum_set(wil, d, skb);
-		*_d = *d;
 	}
 	/* for the last seg only */
 	d->dma.d0 |= BIT(DMA_CFG_DESC_TX_0_CMD_EOP_POS);
 	d->dma.d0 |= BIT(DMA_CFG_DESC_TX_0_CMD_MARK_WB_POS);
 	d->dma.d0 |= BIT(DMA_CFG_DESC_TX_0_CMD_DMA_IT_POS);
 	*_d = *d;
+	wil_dbg_txrx(wil, "Tx[%2d] desc[%4d]\n", vring_index, i);
+	wil_hex_dump_txrx("TxD ", DUMP_PREFIX_NONE, 32, 4,
+			  (const void *)d, sizeof(*d), false);
 
 	/* hold reference to skb
 	 * to prevent skb release before accounting
@@ -990,15 +1027,13 @@ static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
 	 */
 	vring->ctx[i].skb = skb_get(skb);
 
-	wil_hex_dump_txrx("Tx ", DUMP_PREFIX_NONE, 32, 4,
-			  (const void *)d, sizeof(*d), false);
-
 	if (wil_vring_is_empty(vring)) /* performance monitoring */
 		txdata->idle += get_cycles() - txdata->last_idle;
 
 	/* advance swhead */
 	wil_vring_advance_head(vring, nr_frags + 1);
-	wil_dbg_txrx(wil, "Tx swhead %d -> %d\n", swhead, vring->swhead);
+	wil_dbg_txrx(wil, "Tx[%2d] swhead %d -> %d\n", vring_index, swhead,
+		     vring->swhead);
 	trace_wil6210_tx(vring_index, swhead, skb->len, nr_frags);
 	iowrite32(vring->swhead, wil->csr + HOSTADDR(vring->hwtail));
 
@@ -1025,6 +1060,19 @@ static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
 	return -EINVAL;
 }
 
+static int wil_tx_vring(struct wil6210_priv *wil, struct vring *vring,
+			struct sk_buff *skb)
+{
+	int vring_index = vring - wil->vring_tx;
+	struct vring_tx_data *txdata = &wil->vring_tx_data[vring_index];
+	int rc;
+
+	spin_lock(&txdata->lock);
+	rc = __wil_tx_vring(wil, vring, skb);
+	spin_unlock(&txdata->lock);
+	return rc;
+}
+
 netdev_tx_t wil_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 {
 	struct wil6210_priv *wil = ndev_to_wil(ndev);
@@ -1034,14 +1082,14 @@ netdev_tx_t wil_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	int rc;
 
 	wil_dbg_txrx(wil, "%s()\n", __func__);
-	if (!test_bit(wil_status_fwready, &wil->status)) {
+	if (!test_bit(wil_status_fwready, wil->status)) {
 		if (!pr_once_fw) {
 			wil_err(wil, "FW not ready\n");
 			pr_once_fw = true;
 		}
 		goto drop;
 	}
-	if (!test_bit(wil_status_fwconnected, &wil->status)) {
+	if (!test_bit(wil_status_fwconnected, wil->status)) {
 		wil_err(wil, "FW not connected\n");
 		goto drop;
 	}
@@ -1052,15 +1100,19 @@ netdev_tx_t wil_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	pr_once_fw = false;
 
 	/* find vring */
-	if (is_unicast_ether_addr(eth->h_dest))
-		vring = wil_find_tx_vring(wil, skb);
-	else
-		vring = wil_tx_bcast(wil, skb);
+	if (wil->wdev->iftype == NL80211_IFTYPE_STATION) {
+		/* in STA mode (ESS), all to same VRING */
+		vring = wil_find_tx_vring_sta(wil, skb);
+	} else { /* direct communication, find matching VRING */
+		if (is_unicast_ether_addr(eth->h_dest))
+			vring = wil_find_tx_vring(wil, skb);
+		else
+			vring = wil_tx_bcast(wil, skb);
+	}
 	if (!vring) {
 		wil_dbg_txrx(wil, "No Tx VRING found for %pM\n", eth->h_dest);
 		goto drop;
 	}
-
 	/* set up vring entry */
 	rc = wil_tx_vring(wil, vring, skb);
 
@@ -1087,6 +1139,22 @@ netdev_tx_t wil_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	return NET_XMIT_DROP;
 }
 
+static inline bool wil_need_txstat(struct sk_buff *skb)
+{
+	struct ethhdr *eth = (void *)skb->data;
+
+	return is_unicast_ether_addr(eth->h_dest) && skb->sk &&
+	       (skb_shinfo(skb)->tx_flags & SKBTX_WIFI_STATUS);
+}
+
+static inline void wil_consume_skb(struct sk_buff *skb, bool acked)
+{
+	if (unlikely(wil_need_txstat(skb)))
+		skb_complete_wifi_ack(skb, acked);
+	else
+		acked ? dev_consume_skb_any(skb) : dev_kfree_skb_any(skb);
+}
+
 /**
  * Clean up transmitted skb's from the Tx VRING
  *
@@ -1147,10 +1215,10 @@ int wil_tx_complete(struct wil6210_priv *wil, int ringid)
 			trace_wil6210_tx_done(ringid, vring->swtail, dmalen,
 					      d->dma.error);
 			wil_dbg_txrx(wil,
-				     "Tx[%3d] : %d bytes, status 0x%02x err 0x%02x\n",
-				     vring->swtail, dmalen, d->dma.status,
-				     d->dma.error);
-			wil_hex_dump_txrx("TxC ", DUMP_PREFIX_NONE, 32, 4,
+				     "TxC[%2d][%3d] : %d bytes, status 0x%02x err 0x%02x\n",
+				     ringid, vring->swtail, dmalen,
+				     d->dma.status, d->dma.error);
+			wil_hex_dump_txrx("TxCD ", DUMP_PREFIX_NONE, 32, 4,
 					  (const void *)d, sizeof(*d), false);
 
 			wil_txdesc_unmap(dev, d, ctx);
@@ -1165,8 +1233,7 @@ int wil_tx_complete(struct wil6210_priv *wil, int ringid)
 					ndev->stats.tx_errors++;
 					stats->tx_errors++;
 				}
-
-				dev_kfree_skb_any(skb);
+				wil_consume_skb(skb, d->dma.error == 0);
 			}
 			memset(ctx, 0, sizeof(*ctx));
 			/* There is no need to touch HW descriptor: