Merge branch 'cxgb4'

Hariprasad Shenai says:

====================
Adds support for Chelsio T5 40G adapter and Misc. fixes

This patch series adds support for Chelsio T5 40G adapters and provides
miscelleneous fixes for cxgb4 driver.

It also adds device ids of two new T5 adapters.

We would like to request this patch series to get merged via David Miller's
'net-next' tree.

We have included all the maintainers of respective drivers. Kindly review the
change and let us know in case of any review comments.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/chelsio/cxgb4/cxgb4.h

@@ -66,6 +66,7 @@ enum {
 	SERNUM_LEN = 24,    /* Serial # length */
 	EC_LEN     = 16,    /* E/C length */
 	ID_LEN     = 16,    /* ID length */
+	PN_LEN     = 16,    /* Part Number length */
 };
 
 enum {
@@ -254,6 +255,7 @@ struct vpd_params {
 	u8 ec[EC_LEN + 1];
 	u8 sn[SERNUM_LEN + 1];
 	u8 id[ID_LEN + 1];
+	u8 pn[PN_LEN + 1];
 };
 
 struct pci_params {
@@ -306,6 +308,7 @@ struct adapter_params {
 	unsigned char bypass;
 
 	unsigned int ofldq_wr_cred;
+	bool ulptx_memwrite_dsgl;          /* use of T5 DSGL allowed */
 };
 
 #include "t4fw_api.h"
@@ -957,7 +960,7 @@ int t4_mc_read(struct adapter *adap, int idx, u32 addr, __be32 *data,
 	       u64 *parity);
 int t4_edc_read(struct adapter *adap, int idx, u32 addr, __be32 *data,
 		u64 *parity);
-
+const char *t4_get_port_type_description(enum fw_port_type port_type);
 void t4_get_port_stats(struct adapter *adap, int idx, struct port_stats *p);
 void t4_read_mtu_tbl(struct adapter *adap, u16 *mtus, u8 *mtu_log);
 void t4_tp_wr_bits_indirect(struct adapter *adap, unsigned int addr,

drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c

@@ -254,6 +254,8 @@ static DEFINE_PCI_DEVICE_TABLE(cxgb4_pci_tbl) = {
 	CH_DEVICE(0x5011, 4),
 	CH_DEVICE(0x5012, 4),
 	CH_DEVICE(0x5013, 4),
+	CH_DEVICE(0x5014, 4),
+	CH_DEVICE(0x5015, 4),
 	CH_DEVICE(0x5401, 4),
 	CH_DEVICE(0x5402, 4),
 	CH_DEVICE(0x5403, 4),
@@ -273,6 +275,8 @@ static DEFINE_PCI_DEVICE_TABLE(cxgb4_pci_tbl) = {
 	CH_DEVICE(0x5411, 4),
 	CH_DEVICE(0x5412, 4),
 	CH_DEVICE(0x5413, 4),
+	CH_DEVICE(0x5414, 4),
+	CH_DEVICE(0x5415, 4),
 	{ 0, }
 };
 
@@ -432,6 +436,9 @@ static void link_report(struct net_device *dev)
 		case SPEED_100:
 			s = "100Mbps";
 			break;
+		case 40000: /* Need a SPEED_40000 in ethtool.h */
+			s = "40Gbps";
+			break;
 		}
 
 		netdev_info(dev, "link up, %s, full-duplex, %s PAUSE\n", s,
@@ -2061,7 +2068,7 @@ static void get_regs(struct net_device *dev, struct ethtool_regs *regs,
 		0x40200, 0x40298,
 		0x402ac, 0x4033c,
 		0x403f8, 0x403fc,
-		0x41300, 0x413c4,
+		0x41304, 0x413c4,
 		0x41400, 0x4141c,
 		0x41480, 0x414d0,
 		0x44000, 0x44078,
@@ -2089,7 +2096,7 @@ static void get_regs(struct net_device *dev, struct ethtool_regs *regs,
 		0x48200, 0x48298,
 		0x482ac, 0x4833c,
 		0x483f8, 0x483fc,
-		0x49300, 0x493c4,
+		0x49304, 0x493c4,
 		0x49400, 0x4941c,
 		0x49480, 0x494d0,
 		0x4c000, 0x4c078,
@@ -2199,6 +2206,8 @@ static unsigned int from_fw_linkcaps(unsigned int type, unsigned int caps)
 	else if (type == FW_PORT_TYPE_FIBER_XFI ||
 		 type == FW_PORT_TYPE_FIBER_XAUI || type == FW_PORT_TYPE_SFP)
 		v |= SUPPORTED_FIBRE;
+	else if (type == FW_PORT_TYPE_BP40_BA)
+		v |= SUPPORTED_40000baseSR4_Full;
 
 	if (caps & FW_PORT_CAP_ANEG)
 		v |= SUPPORTED_Autoneg;
@@ -2215,6 +2224,8 @@ static unsigned int to_fw_linkcaps(unsigned int caps)
 		v |= FW_PORT_CAP_SPEED_1G;
 	if (caps & ADVERTISED_10000baseT_Full)
 		v |= FW_PORT_CAP_SPEED_10G;
+	if (caps & ADVERTISED_40000baseSR4_Full)
+		v |= FW_PORT_CAP_SPEED_40G;
 	return v;
 }
 
@@ -2269,6 +2280,8 @@ static unsigned int speed_to_caps(int speed)
 		return FW_PORT_CAP_SPEED_1G;
 	if (speed == SPEED_10000)
 		return FW_PORT_CAP_SPEED_10G;
+	if (speed == 40000) /* Need SPEED_40000 in ethtool.h */
+		return FW_PORT_CAP_SPEED_40G;
 	return 0;
 }
 
@@ -2296,8 +2309,10 @@ static int set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
 	if (cmd->autoneg == AUTONEG_DISABLE) {
 		cap = speed_to_caps(speed);
 
-		if (!(lc->supported & cap) || (speed == SPEED_1000) ||
-		    (speed == SPEED_10000))
+		if (!(lc->supported & cap) ||
+		    (speed == SPEED_1000) ||
+		    (speed == SPEED_10000) ||
+		    (speed == 40000))
 			return -EINVAL;
 		lc->requested_speed = cap;
 		lc->advertising = 0;
@@ -3765,6 +3780,7 @@ static void uld_attach(struct adapter *adap, unsigned int uld)
 	lli.dbfifo_int_thresh = dbfifo_int_thresh;
 	lli.sge_pktshift = adap->sge.pktshift;
 	lli.enable_fw_ofld_conn = adap->flags & FW_OFLD_CONN;
+	lli.ulptx_memwrite_dsgl = adap->params.ulptx_memwrite_dsgl;
 
 	handle = ulds[uld].add(&lli);
 	if (IS_ERR(handle)) {
@@ -5369,6 +5385,21 @@ static int adap_init0(struct adapter *adap)
 	val[0] = 1;
 	(void) t4_set_params(adap, adap->mbox, adap->fn, 0, 1, params, val);
 
+	/*
+	 * Find out whether we're allowed to use the T5+ ULPTX MEMWRITE DSGL
+	 * capability.  Earlier versions of the firmware didn't have the
+	 * ULPTX_MEMWRITE_DSGL so we'll interpret a query failure as no
+	 * permission to use ULPTX MEMWRITE DSGL.
+	 */
+	if (is_t4(adap->params.chip)) {
+		adap->params.ulptx_memwrite_dsgl = false;
+	} else {
+		params[0] = FW_PARAM_DEV(ULPTX_MEMWRITE_DSGL);
+		ret = t4_query_params(adap, adap->mbox, adap->fn, 0,
+				      1, params, val);
+		adap->params.ulptx_memwrite_dsgl = (ret == 0 && val[0] != 0);
+	}
+
 	/*
 	 * Get device capabilities so we can determine what resources we need
 	 * to manage.
@@ -5603,9 +5634,10 @@ static const struct pci_error_handlers cxgb4_eeh = {
 	.resume         = eeh_resume,
 };
 
-static inline bool is_10g_port(const struct link_config *lc)
+static inline bool is_x_10g_port(const struct link_config *lc)
 {
-	return (lc->supported & FW_PORT_CAP_SPEED_10G) != 0;
+	return (lc->supported & FW_PORT_CAP_SPEED_10G) != 0 ||
+	       (lc->supported & FW_PORT_CAP_SPEED_40G) != 0;
 }
 
 static inline void init_rspq(struct sge_rspq *q, u8 timer_idx, u8 pkt_cnt_idx,
@@ -5629,7 +5661,7 @@ static void cfg_queues(struct adapter *adap)
 	int i, q10g = 0, n10g = 0, qidx = 0;
 
 	for_each_port(adap, i)
-		n10g += is_10g_port(&adap2pinfo(adap, i)->link_cfg);
+		n10g += is_x_10g_port(&adap2pinfo(adap, i)->link_cfg);
 
 	/*
 	 * We default to 1 queue per non-10G port and up to # of cores queues
@@ -5644,7 +5676,7 @@ static void cfg_queues(struct adapter *adap)
 		struct port_info *pi = adap2pinfo(adap, i);
 
 		pi->first_qset = qidx;
-		pi->nqsets = is_10g_port(&pi->link_cfg) ? q10g : 1;
+		pi->nqsets = is_x_10g_port(&pi->link_cfg) ? q10g : 1;
 		qidx += pi->nqsets;
 	}
 
@@ -5799,11 +5831,6 @@ static int init_rss(struct adapter *adap)
 
 static void print_port_info(const struct net_device *dev)
 {
-	static const char *base[] = {
-		"R XFI", "R XAUI", "T SGMII", "T XFI", "T XAUI", "KX4", "CX4",
-		"KX", "KR", "R SFP+", "KR/KX", "KR/KX/KX4"
-	};
-
 	char buf[80];
 	char *bufp = buf;
 	const char *spd = "";
@@ -5821,18 +5848,20 @@ static void print_port_info(const struct net_device *dev)
 		bufp += sprintf(bufp, "1000/");
 	if (pi->link_cfg.supported & FW_PORT_CAP_SPEED_10G)
 		bufp += sprintf(bufp, "10G/");
+	if (pi->link_cfg.supported & FW_PORT_CAP_SPEED_40G)
+		bufp += sprintf(bufp, "40G/");
 	if (bufp != buf)
 		--bufp;
-	sprintf(bufp, "BASE-%s", base[pi->port_type]);
+	sprintf(bufp, "BASE-%s", t4_get_port_type_description(pi->port_type));
 
 	netdev_info(dev, "Chelsio %s rev %d %s %sNIC PCIe x%d%s%s\n",
 		    adap->params.vpd.id,
 		    CHELSIO_CHIP_RELEASE(adap->params.chip), buf,
 		    is_offload(adap) ? "R" : "", adap->params.pci.width, spd,
 		    (adap->flags & USING_MSIX) ? " MSI-X" :
 		    (adap->flags & USING_MSI) ? " MSI" : "");
-	netdev_info(dev, "S/N: %s, E/C: %s\n",
-		    adap->params.vpd.sn, adap->params.vpd.ec);
+	netdev_info(dev, "S/N: %s, P/N: %s\n",
+		    adap->params.vpd.sn, adap->params.vpd.pn);
 }
 
 static void enable_pcie_relaxed_ordering(struct pci_dev *dev)

drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h

@@ -253,6 +253,7 @@ struct cxgb4_lld_info {
 					     /*	packet data */
 	bool enable_fw_ofld_conn;            /* Enable connection through fw */
 					     /* WR */
+	bool ulptx_memwrite_dsgl;            /* use of T5 DSGL allowed */
 };
 
 struct cxgb4_uld_info {

drivers/net/ethernet/chelsio/cxgb4/sge.c

@@ -706,11 +706,17 @@ static inline unsigned int flits_to_desc(unsigned int n)
  *	@skb: the packet
  *
  *	Returns whether an Ethernet packet is small enough to fit as
- *	immediate data.
+ *	immediate data. Return value corresponds to headroom required.
  */
 static inline int is_eth_imm(const struct sk_buff *skb)
 {
-	return skb->len <= MAX_IMM_TX_PKT_LEN - sizeof(struct cpl_tx_pkt);
+	int hdrlen = skb_shinfo(skb)->gso_size ?
+			sizeof(struct cpl_tx_pkt_lso_core) : 0;
+
+	hdrlen += sizeof(struct cpl_tx_pkt);
+	if (skb->len <= MAX_IMM_TX_PKT_LEN - hdrlen)
+		return hdrlen;
+	return 0;
 }
 
 /**
@@ -723,9 +729,10 @@ static inline int is_eth_imm(const struct sk_buff *skb)
 static inline unsigned int calc_tx_flits(const struct sk_buff *skb)
 {
 	unsigned int flits;
+	int hdrlen = is_eth_imm(skb);
 
-	if (is_eth_imm(skb))
-		return DIV_ROUND_UP(skb->len + sizeof(struct cpl_tx_pkt), 8);
+	if (hdrlen)
+		return DIV_ROUND_UP(skb->len + hdrlen, sizeof(__be64));
 
 	flits = sgl_len(skb_shinfo(skb)->nr_frags + 1) + 4;
 	if (skb_shinfo(skb)->gso_size)
@@ -971,6 +978,7 @@ static inline void txq_advance(struct sge_txq *q, unsigned int n)
  */
 netdev_tx_t t4_eth_xmit(struct sk_buff *skb, struct net_device *dev)
 {
+	int len;
 	u32 wr_mid;
 	u64 cntrl, *end;
 	int qidx, credits;
@@ -982,6 +990,7 @@ netdev_tx_t t4_eth_xmit(struct sk_buff *skb, struct net_device *dev)
 	struct cpl_tx_pkt_core *cpl;
 	const struct skb_shared_info *ssi;
 	dma_addr_t addr[MAX_SKB_FRAGS + 1];
+	bool immediate = false;
 
 	/*
 	 * The chip min packet length is 10 octets but play safe and reject
@@ -1011,7 +1020,10 @@ out_free:	dev_kfree_skb(skb);
 		return NETDEV_TX_BUSY;
 	}
 
-	if (!is_eth_imm(skb) &&
+	if (is_eth_imm(skb))
+		immediate = true;
+
+	if (!immediate &&
 	    unlikely(map_skb(adap->pdev_dev, skb, addr) < 0)) {
 		q->mapping_err++;
 		goto out_free;
@@ -1028,15 +1040,18 @@ out_free:	dev_kfree_skb(skb);
 	wr->r3 = cpu_to_be64(0);
 	end = (u64 *)wr + flits;
 
+	len = immediate ? skb->len : 0;
+	len += sizeof(*cpl);
 	ssi = skb_shinfo(skb);
 	if (ssi->gso_size) {
 		struct cpl_tx_pkt_lso *lso = (void *)wr;
 		bool v6 = (ssi->gso_type & SKB_GSO_TCPV6) != 0;
 		int l3hdr_len = skb_network_header_len(skb);
 		int eth_xtra_len = skb_network_offset(skb) - ETH_HLEN;
 
+		len += sizeof(*lso);
 		wr->op_immdlen = htonl(FW_WR_OP(FW_ETH_TX_PKT_WR) |
-				       FW_WR_IMMDLEN(sizeof(*lso)));
+				       FW_WR_IMMDLEN(len));
 		lso->c.lso_ctrl = htonl(LSO_OPCODE(CPL_TX_PKT_LSO) |
 					LSO_FIRST_SLICE | LSO_LAST_SLICE |
 					LSO_IPV6(v6) |
@@ -1054,9 +1069,6 @@ out_free:	dev_kfree_skb(skb);
 		q->tso++;
 		q->tx_cso += ssi->gso_segs;
 	} else {
-		int len;
-
-		len = is_eth_imm(skb) ? skb->len + sizeof(*cpl) : sizeof(*cpl);
 		wr->op_immdlen = htonl(FW_WR_OP(FW_ETH_TX_PKT_WR) |
 				       FW_WR_IMMDLEN(len));
 		cpl = (void *)(wr + 1);
@@ -1078,7 +1090,7 @@ out_free:	dev_kfree_skb(skb);
 	cpl->len = htons(skb->len);
 	cpl->ctrl1 = cpu_to_be64(cntrl);
 
-	if (is_eth_imm(skb)) {
+	if (immediate) {
 		inline_tx_skb(skb, &q->q, cpl + 1);
 		dev_kfree_skb(skb);
 	} else {
@@ -1467,8 +1479,12 @@ static inline int ofld_send(struct adapter *adap, struct sk_buff *skb)
 {
 	unsigned int idx = skb_txq(skb);
 
-	if (unlikely(is_ctrl_pkt(skb)))
+	if (unlikely(is_ctrl_pkt(skb))) {
+		/* Single ctrl queue is a requirement for LE workaround path */
+		if (adap->tids.nsftids)
+			idx = 0;
 		return ctrl_xmit(&adap->sge.ctrlq[idx], skb);
+	}
 	return ofld_xmit(&adap->sge.ofldtxq[idx], skb);
 }