* [PATCH net-next 0/2] net: systemport: Add support for SYSTEMPORT lite
From: Florian Fainelli @ 2017-01-20 19:08 UTC (permalink / raw)
To: netdev; +Cc: davem, Florian Fainelli
Hi David,
This patch series adds support for SYSTEMPORT Lite which is an evolution
of the existing SYSTEMPORT adapter.
The two generations are largely identical as far as the transmit/receive
path are concerned, and there were just a few control path changes here
and there.
Thanks!
Florian Fainelli (2):
net: systemport: Dynamically allocate number of TX rings
net: systemport: Add support for SYSTEMPORT Lite
.../devicetree/bindings/net/brcm,systemport.txt | 5 +-
drivers/net/ethernet/broadcom/bcmsysport.c | 334 +++++++++++++++++----
drivers/net/ethernet/broadcom/bcmsysport.h | 80 ++++-
3 files changed, 339 insertions(+), 80 deletions(-)
--
2.11.0
^ permalink raw reply
* [PATCH net-next 1/2] net: systemport: Dynamically allocate number of TX rings
From: Florian Fainelli @ 2017-01-20 19:08 UTC (permalink / raw)
To: netdev; +Cc: davem, Florian Fainelli
In-Reply-To: <20170120190827.26444-1-f.fainelli@gmail.com>
In preparation for adding SYSTEMPORT Lite, which has twice as less transmit
queues than SYSTEMPORT make sure we do allocate TX rings based on the
systemport,txq property to get an appropriate memory footprint.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/ethernet/broadcom/bcmsysport.c | 11 +++++++++++
drivers/net/ethernet/broadcom/bcmsysport.h | 2 +-
2 files changed, 12 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index 744ed6ddaf37..31bb2c3696ec 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -1752,6 +1752,10 @@ static int bcm_sysport_probe(struct platform_device *pdev)
if (of_property_read_u32(dn, "systemport,num-rxq", &rxq))
rxq = 1;
+ /* Sanity check the number of transmit queues */
+ if (!txq || txq > TDMA_NUM_RINGS)
+ return -EINVAL;
+
dev = alloc_etherdev_mqs(sizeof(*priv), txq, rxq);
if (!dev)
return -ENOMEM;
@@ -1759,6 +1763,13 @@ static int bcm_sysport_probe(struct platform_device *pdev)
/* Initialize private members */
priv = netdev_priv(dev);
+ /* Allocate number of TX rings */
+ priv->tx_rings = devm_kcalloc(&pdev->dev, txq,
+ sizeof(struct bcm_sysport_tx_ring),
+ GFP_KERNEL);
+ if (!priv->tx_rings)
+ return -ENOMEM;
+
priv->irq0 = platform_get_irq(pdev, 0);
priv->irq1 = platform_get_irq(pdev, 1);
priv->wol_irq = platform_get_irq(pdev, 2);
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.h b/drivers/net/ethernet/broadcom/bcmsysport.h
index 1c82e3da69a7..f051356b0274 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.h
+++ b/drivers/net/ethernet/broadcom/bcmsysport.h
@@ -659,7 +659,7 @@ struct bcm_sysport_priv {
int wol_irq;
/* Transmit rings */
- struct bcm_sysport_tx_ring tx_rings[TDMA_NUM_RINGS];
+ struct bcm_sysport_tx_ring *tx_rings;
/* Receive queue */
void __iomem *rx_bds;
--
2.11.0
^ permalink raw reply related
* [PATCH net-next 2/2] net: systemport: Add support for SYSTEMPORT Lite
From: Florian Fainelli @ 2017-01-20 19:08 UTC (permalink / raw)
To: netdev; +Cc: davem, Florian Fainelli
In-Reply-To: <20170120190827.26444-1-f.fainelli@gmail.com>
Add supporf for the SYSTEMPORT Lite Ethernet controller, this piece of hardware
is largely based on the full-blown SYSTEMPORT and differs in the following:
- no full-blown UniMAC, instead we have the MagicPacket matching from UniMAC at
same offset, and a GMII Interface Block (GIB) for the MAC-level stuff, since
we are always interfaced to an Ethernet switch which is fully Ethernet compliant
shortcuts could be made
- 16 transmit queues, whose interrupts are moved into the first Level-2 interrupt
controller bank
- slight TDMA offset change (a register was inserted after TDMA_STATUS, *sigh*)
- 256 RX descriptors (512 words) and 256 TX descriptors (not visible)
As a consequence of these two things, update the code paths accordingly to
differentiate the full-blown from the light version.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
.../devicetree/bindings/net/brcm,systemport.txt | 5 +-
drivers/net/ethernet/broadcom/bcmsysport.c | 323 +++++++++++++++++----
drivers/net/ethernet/broadcom/bcmsysport.h | 78 ++++-
3 files changed, 327 insertions(+), 79 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/brcm,systemport.txt b/Documentation/devicetree/bindings/net/brcm,systemport.txt
index 877da34145b0..83f29e0e11ba 100644
--- a/Documentation/devicetree/bindings/net/brcm,systemport.txt
+++ b/Documentation/devicetree/bindings/net/brcm,systemport.txt
@@ -1,7 +1,10 @@
* Broadcom BCM7xxx Ethernet Systemport Controller (SYSTEMPORT)
Required properties:
-- compatible: should be one of "brcm,systemport-v1.00" or "brcm,systemport"
+- compatible: should be one of:
+ "brcm,systemport-v1.00"
+ "brcm,systemportlite-v1.00" or
+ "brcm,systemport"
- reg: address and length of the register set for the device.
- interrupts: interrupts for the device, first cell must be for the rx
interrupts, and the second cell should be for the transmit queues. An
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index 31bb2c3696ec..a68d4889f5db 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -43,14 +43,43 @@ static inline void name##_writel(struct bcm_sysport_priv *priv, \
BCM_SYSPORT_IO_MACRO(intrl2_0, SYS_PORT_INTRL2_0_OFFSET);
BCM_SYSPORT_IO_MACRO(intrl2_1, SYS_PORT_INTRL2_1_OFFSET);
BCM_SYSPORT_IO_MACRO(umac, SYS_PORT_UMAC_OFFSET);
+BCM_SYSPORT_IO_MACRO(gib, SYS_PORT_GIB_OFFSET);
BCM_SYSPORT_IO_MACRO(tdma, SYS_PORT_TDMA_OFFSET);
-BCM_SYSPORT_IO_MACRO(rdma, SYS_PORT_RDMA_OFFSET);
BCM_SYSPORT_IO_MACRO(rxchk, SYS_PORT_RXCHK_OFFSET);
BCM_SYSPORT_IO_MACRO(txchk, SYS_PORT_TXCHK_OFFSET);
BCM_SYSPORT_IO_MACRO(rbuf, SYS_PORT_RBUF_OFFSET);
BCM_SYSPORT_IO_MACRO(tbuf, SYS_PORT_TBUF_OFFSET);
BCM_SYSPORT_IO_MACRO(topctrl, SYS_PORT_TOPCTRL_OFFSET);
+/* On SYSTEMPORT Lite, any register after RDMA_STATUS has the exact
+ * same layout, except it has been moved by 4 bytes up, *sigh*
+ */
+static inline u32 rdma_readl(struct bcm_sysport_priv *priv, u32 off)
+{
+ if (priv->is_lite && off >= RDMA_STATUS)
+ off += 4;
+ return __raw_readl(priv->base + SYS_PORT_RDMA_OFFSET + off);
+}
+
+static inline void rdma_writel(struct bcm_sysport_priv *priv, u32 val, u32 off)
+{
+ if (priv->is_lite && off >= RDMA_STATUS)
+ off += 4;
+ __raw_writel(val, priv->base + SYS_PORT_RDMA_OFFSET + off);
+}
+
+static inline u32 tdma_control_bit(struct bcm_sysport_priv *priv, u32 bit)
+{
+ if (!priv->is_lite) {
+ return BIT(bit);
+ } else {
+ if (bit >= ACB_ALGO)
+ return BIT(bit + 1);
+ else
+ return BIT(bit);
+ }
+}
+
/* L2-interrupt masking/unmasking helpers, does automatic saving of the applied
* mask in a software copy to avoid CPU_MASK_STATUS reads in hot-paths.
*/
@@ -143,9 +172,9 @@ static int bcm_sysport_set_tx_csum(struct net_device *dev,
priv->tsb_en = !!(wanted & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM));
reg = tdma_readl(priv, TDMA_CONTROL);
if (priv->tsb_en)
- reg |= TSB_EN;
+ reg |= tdma_control_bit(priv, TSB_EN);
else
- reg &= ~TSB_EN;
+ reg &= ~tdma_control_bit(priv, TSB_EN);
tdma_writel(priv, reg, TDMA_CONTROL);
return 0;
@@ -281,11 +310,35 @@ static void bcm_sysport_set_msglvl(struct net_device *dev, u32 enable)
priv->msg_enable = enable;
}
+static inline bool bcm_sysport_lite_stat_valid(enum bcm_sysport_stat_type type)
+{
+ switch (type) {
+ case BCM_SYSPORT_STAT_NETDEV:
+ case BCM_SYSPORT_STAT_RXCHK:
+ case BCM_SYSPORT_STAT_RBUF:
+ case BCM_SYSPORT_STAT_SOFT:
+ return true;
+ default:
+ return false;
+ }
+}
+
static int bcm_sysport_get_sset_count(struct net_device *dev, int string_set)
{
+ struct bcm_sysport_priv *priv = netdev_priv(dev);
+ const struct bcm_sysport_stats *s;
+ unsigned int i, j;
+
switch (string_set) {
case ETH_SS_STATS:
- return BCM_SYSPORT_STATS_LEN;
+ for (i = 0, j = 0; i < BCM_SYSPORT_STATS_LEN; i++) {
+ s = &bcm_sysport_gstrings_stats[i];
+ if (priv->is_lite &&
+ !bcm_sysport_lite_stat_valid(s->type))
+ continue;
+ j++;
+ }
+ return j;
default:
return -EOPNOTSUPP;
}
@@ -294,14 +347,21 @@ static int bcm_sysport_get_sset_count(struct net_device *dev, int string_set)
static void bcm_sysport_get_strings(struct net_device *dev,
u32 stringset, u8 *data)
{
- int i;
+ struct bcm_sysport_priv *priv = netdev_priv(dev);
+ const struct bcm_sysport_stats *s;
+ int i, j;
switch (stringset) {
case ETH_SS_STATS:
- for (i = 0; i < BCM_SYSPORT_STATS_LEN; i++) {
- memcpy(data + i * ETH_GSTRING_LEN,
- bcm_sysport_gstrings_stats[i].stat_string,
+ for (i = 0, j = 0; i < BCM_SYSPORT_STATS_LEN; i++) {
+ s = &bcm_sysport_gstrings_stats[i];
+ if (priv->is_lite &&
+ !bcm_sysport_lite_stat_valid(s->type))
+ continue;
+
+ memcpy(data + j * ETH_GSTRING_LEN, s->stat_string,
ETH_GSTRING_LEN);
+ j++;
}
break;
default:
@@ -327,6 +387,9 @@ static void bcm_sysport_update_mib_counters(struct bcm_sysport_priv *priv)
case BCM_SYSPORT_STAT_MIB_RX:
case BCM_SYSPORT_STAT_MIB_TX:
case BCM_SYSPORT_STAT_RUNT:
+ if (priv->is_lite)
+ continue;
+
if (s->type != BCM_SYSPORT_STAT_MIB_RX)
offset = UMAC_MIB_STAT_OFFSET;
val = umac_readl(priv, UMAC_MIB_START + j + offset);
@@ -355,12 +418,12 @@ static void bcm_sysport_get_stats(struct net_device *dev,
struct ethtool_stats *stats, u64 *data)
{
struct bcm_sysport_priv *priv = netdev_priv(dev);
- int i;
+ int i, j;
if (netif_running(dev))
bcm_sysport_update_mib_counters(priv);
- for (i = 0; i < BCM_SYSPORT_STATS_LEN; i++) {
+ for (i = 0, j = 0; i < BCM_SYSPORT_STATS_LEN; i++) {
const struct bcm_sysport_stats *s;
char *p;
@@ -370,7 +433,8 @@ static void bcm_sysport_get_stats(struct net_device *dev,
else
p = (char *)priv;
p += s->stat_offset;
- data[i] = *(unsigned long *)p;
+ data[j] = *(unsigned long *)p;
+ j++;
}
}
@@ -573,8 +637,14 @@ static unsigned int bcm_sysport_desc_rx(struct bcm_sysport_priv *priv,
u16 len, status;
struct bcm_rsb *rsb;
- /* Determine how much we should process since last call */
- p_index = rdma_readl(priv, RDMA_PROD_INDEX);
+ /* Determine how much we should process since last call, SYSTEMPORT Lite
+ * groups the producer and consumer indexes into the same 32-bit
+ * which we access using RDMA_CONS_INDEX
+ */
+ if (!priv->is_lite)
+ p_index = rdma_readl(priv, RDMA_PROD_INDEX);
+ else
+ p_index = rdma_readl(priv, RDMA_CONS_INDEX);
p_index &= RDMA_PROD_INDEX_MASK;
if (p_index < priv->rx_c_index)
@@ -791,7 +861,11 @@ static int bcm_sysport_tx_poll(struct napi_struct *napi, int budget)
if (work_done == 0) {
napi_complete(napi);
/* re-enable TX interrupt */
- intrl2_1_mask_clear(ring->priv, BIT(ring->index));
+ if (!ring->priv->is_lite)
+ intrl2_1_mask_clear(ring->priv, BIT(ring->index));
+ else
+ intrl2_0_mask_clear(ring->priv, BIT(ring->index +
+ INTRL2_0_TDMA_MBDONE_SHIFT));
return 0;
}
@@ -817,7 +891,15 @@ static int bcm_sysport_poll(struct napi_struct *napi, int budget)
priv->rx_c_index += work_done;
priv->rx_c_index &= RDMA_CONS_INDEX_MASK;
- rdma_writel(priv, priv->rx_c_index, RDMA_CONS_INDEX);
+
+ /* SYSTEMPORT Lite groups the producer/consumer index, producer is
+ * maintained by HW, but writes to it will be ignore while RDMA
+ * is active
+ */
+ if (!priv->is_lite)
+ rdma_writel(priv, priv->rx_c_index, RDMA_CONS_INDEX);
+ else
+ rdma_writel(priv, priv->rx_c_index << 16, RDMA_CONS_INDEX);
if (work_done < budget) {
napi_complete_done(napi, work_done);
@@ -848,6 +930,8 @@ static irqreturn_t bcm_sysport_rx_isr(int irq, void *dev_id)
{
struct net_device *dev = dev_id;
struct bcm_sysport_priv *priv = netdev_priv(dev);
+ struct bcm_sysport_tx_ring *txr;
+ unsigned int ring, ring_bit;
priv->irq0_stat = intrl2_0_readl(priv, INTRL2_CPU_STATUS) &
~intrl2_0_readl(priv, INTRL2_CPU_MASK_STATUS);
@@ -877,6 +961,22 @@ static irqreturn_t bcm_sysport_rx_isr(int irq, void *dev_id)
bcm_sysport_resume_from_wol(priv);
}
+ if (!priv->is_lite)
+ goto out;
+
+ for (ring = 0; ring < dev->num_tx_queues; ring++) {
+ ring_bit = BIT(ring + INTRL2_0_TDMA_MBDONE_SHIFT);
+ if (!(priv->irq0_stat & ring_bit))
+ continue;
+
+ txr = &priv->tx_rings[ring];
+
+ if (likely(napi_schedule_prep(&txr->napi))) {
+ intrl2_0_mask_set(priv, ring_bit);
+ __napi_schedule(&txr->napi);
+ }
+ }
+out:
return IRQ_HANDLED;
}
@@ -930,9 +1030,11 @@ static void bcm_sysport_poll_controller(struct net_device *dev)
bcm_sysport_rx_isr(priv->irq0, priv);
enable_irq(priv->irq0);
- disable_irq(priv->irq1);
- bcm_sysport_tx_isr(priv->irq1, priv);
- enable_irq(priv->irq1);
+ if (!priv->is_lite) {
+ disable_irq(priv->irq1);
+ bcm_sysport_tx_isr(priv->irq1, priv);
+ enable_irq(priv->irq1);
+ }
}
#endif
@@ -1129,6 +1231,9 @@ static void bcm_sysport_adj_link(struct net_device *dev)
priv->old_duplex = phydev->duplex;
}
+ if (priv->is_lite)
+ goto out;
+
switch (phydev->speed) {
case SPEED_2500:
cmd_bits = CMD_SPEED_2500;
@@ -1169,8 +1274,9 @@ static void bcm_sysport_adj_link(struct net_device *dev)
reg |= cmd_bits;
umac_writel(priv, reg, UMAC_CMD);
}
-
- phy_print_status(phydev);
+out:
+ if (changed)
+ phy_print_status(phydev);
}
static int bcm_sysport_init_tx_ring(struct bcm_sysport_priv *priv,
@@ -1315,9 +1421,9 @@ static inline int tdma_enable_set(struct bcm_sysport_priv *priv,
reg = tdma_readl(priv, TDMA_CONTROL);
if (enable)
- reg |= TDMA_EN;
+ reg |= tdma_control_bit(priv, TDMA_EN);
else
- reg &= ~TDMA_EN;
+ reg &= ~tdma_control_bit(priv, TDMA_EN);
tdma_writel(priv, reg, TDMA_CONTROL);
/* Poll for TMDA disabling completion */
@@ -1342,7 +1448,7 @@ static int bcm_sysport_init_rx_ring(struct bcm_sysport_priv *priv)
int i;
/* Initialize SW view of the RX ring */
- priv->num_rx_bds = NUM_RX_DESC;
+ priv->num_rx_bds = priv->num_rx_desc_words / WORDS_PER_DESC;
priv->rx_bds = priv->base + SYS_PORT_RDMA_OFFSET;
priv->rx_c_index = 0;
priv->rx_read_ptr = 0;
@@ -1379,7 +1485,7 @@ static int bcm_sysport_init_rx_ring(struct bcm_sysport_priv *priv)
rdma_writel(priv, 0, RDMA_START_ADDR_HI);
rdma_writel(priv, 0, RDMA_START_ADDR_LO);
rdma_writel(priv, 0, RDMA_END_ADDR_HI);
- rdma_writel(priv, NUM_HW_RX_DESC_WORDS - 1, RDMA_END_ADDR_LO);
+ rdma_writel(priv, priv->num_rx_desc_words - 1, RDMA_END_ADDR_LO);
rdma_writel(priv, 1, RDMA_MBDONE_INTR);
@@ -1421,6 +1527,9 @@ static void bcm_sysport_set_rx_mode(struct net_device *dev)
struct bcm_sysport_priv *priv = netdev_priv(dev);
u32 reg;
+ if (priv->is_lite)
+ return;
+
reg = umac_readl(priv, UMAC_CMD);
if (dev->flags & IFF_PROMISC)
reg |= CMD_PROMISC;
@@ -1438,12 +1547,21 @@ static inline void umac_enable_set(struct bcm_sysport_priv *priv,
{
u32 reg;
- reg = umac_readl(priv, UMAC_CMD);
- if (enable)
- reg |= mask;
- else
- reg &= ~mask;
- umac_writel(priv, reg, UMAC_CMD);
+ if (!priv->is_lite) {
+ reg = umac_readl(priv, UMAC_CMD);
+ if (enable)
+ reg |= mask;
+ else
+ reg &= ~mask;
+ umac_writel(priv, reg, UMAC_CMD);
+ } else {
+ reg = gib_readl(priv, GIB_CONTROL);
+ if (enable)
+ reg |= mask;
+ else
+ reg &= ~mask;
+ gib_writel(priv, reg, GIB_CONTROL);
+ }
/* UniMAC stops on a packet boundary, wait for a full-sized packet
* to be processed (1 msec).
@@ -1456,6 +1574,9 @@ static inline void umac_reset(struct bcm_sysport_priv *priv)
{
u32 reg;
+ if (priv->is_lite)
+ return;
+
reg = umac_readl(priv, UMAC_CMD);
reg |= CMD_SW_RESET;
umac_writel(priv, reg, UMAC_CMD);
@@ -1468,9 +1589,17 @@ static inline void umac_reset(struct bcm_sysport_priv *priv)
static void umac_set_hw_addr(struct bcm_sysport_priv *priv,
unsigned char *addr)
{
- umac_writel(priv, (addr[0] << 24) | (addr[1] << 16) |
- (addr[2] << 8) | addr[3], UMAC_MAC0);
- umac_writel(priv, (addr[4] << 8) | addr[5], UMAC_MAC1);
+ u32 mac0 = (addr[0] << 24) | (addr[1] << 16) | (addr[2] << 8) |
+ addr[3];
+ u32 mac1 = (addr[4] << 8) | addr[5];
+
+ if (!priv->is_lite) {
+ umac_writel(priv, mac0, UMAC_MAC0);
+ umac_writel(priv, mac1, UMAC_MAC1);
+ } else {
+ gib_writel(priv, mac0, GIB_MAC0);
+ gib_writel(priv, mac1, GIB_MAC1);
+ }
}
static void topctrl_flush(struct bcm_sysport_priv *priv)
@@ -1515,8 +1644,11 @@ static void bcm_sysport_netif_start(struct net_device *dev)
phy_start(dev->phydev);
- /* Enable TX interrupts for the 32 TXQs */
- intrl2_1_mask_clear(priv, 0xffffffff);
+ /* Enable TX interrupts for the TXQs */
+ if (!priv->is_lite)
+ intrl2_1_mask_clear(priv, 0xffffffff);
+ else
+ intrl2_0_mask_clear(priv, INTRL2_0_TDMA_MBDONE_MASK);
/* Last call before we start the real business */
netif_tx_start_all_queues(dev);
@@ -1528,9 +1660,37 @@ static void rbuf_init(struct bcm_sysport_priv *priv)
reg = rbuf_readl(priv, RBUF_CONTROL);
reg |= RBUF_4B_ALGN | RBUF_RSB_EN;
+ /* Set a correct RSB format on SYSTEMPORT Lite */
+ if (priv->is_lite) {
+ reg &= ~RBUF_RSB_SWAP1;
+ reg |= RBUF_RSB_SWAP0;
+ }
rbuf_writel(priv, reg, RBUF_CONTROL);
}
+static inline void bcm_sysport_mask_all_intrs(struct bcm_sysport_priv *priv)
+{
+ intrl2_0_mask_set(priv, 0xffffffff);
+ intrl2_0_writel(priv, 0xffffffff, INTRL2_CPU_CLEAR);
+ if (!priv->is_lite) {
+ intrl2_1_mask_set(priv, 0xffffffff);
+ intrl2_1_writel(priv, 0xffffffff, INTRL2_CPU_CLEAR);
+ }
+}
+
+static inline void gib_set_pad_extension(struct bcm_sysport_priv *priv)
+{
+ u32 __maybe_unused reg;
+
+ /* Include Broadcom tag in pad extension */
+ if (netdev_uses_dsa(priv->netdev)) {
+ reg = gib_readl(priv, GIB_CONTROL);
+ reg &= ~(GIB_PAD_EXTENSION_MASK << GIB_PAD_EXTENSION_SHIFT);
+ reg |= ENET_BRCM_TAG_LEN << GIB_PAD_EXTENSION_SHIFT;
+ gib_writel(priv, reg, GIB_CONTROL);
+ }
+}
+
static int bcm_sysport_open(struct net_device *dev)
{
struct bcm_sysport_priv *priv = netdev_priv(dev);
@@ -1551,13 +1711,20 @@ static int bcm_sysport_open(struct net_device *dev)
rbuf_init(priv);
/* Set maximum frame length */
- umac_writel(priv, UMAC_MAX_MTU_SIZE, UMAC_MAX_FRAME_LEN);
+ if (!priv->is_lite)
+ umac_writel(priv, UMAC_MAX_MTU_SIZE, UMAC_MAX_FRAME_LEN);
+ else
+ gib_set_pad_extension(priv);
/* Set MAC address */
umac_set_hw_addr(priv, dev->dev_addr);
/* Read CRC forward */
- priv->crc_fwd = !!(umac_readl(priv, UMAC_CMD) & CMD_CRC_FWD);
+ if (!priv->is_lite)
+ priv->crc_fwd = !!(umac_readl(priv, UMAC_CMD) & CMD_CRC_FWD);
+ else
+ priv->crc_fwd = !!(gib_readl(priv, GIB_CONTROL) &
+ GIB_FCS_STRIP);
phydev = of_phy_connect(dev, priv->phy_dn, bcm_sysport_adj_link,
0, priv->phy_interface);
@@ -1572,12 +1739,7 @@ static int bcm_sysport_open(struct net_device *dev)
priv->old_pause = -1;
/* mask all interrupts and request them */
- intrl2_0_writel(priv, 0xffffffff, INTRL2_CPU_MASK_SET);
- intrl2_0_writel(priv, 0xffffffff, INTRL2_CPU_CLEAR);
- intrl2_0_writel(priv, 0, INTRL2_CPU_MASK_CLEAR);
- intrl2_1_writel(priv, 0xffffffff, INTRL2_CPU_MASK_SET);
- intrl2_1_writel(priv, 0xffffffff, INTRL2_CPU_CLEAR);
- intrl2_1_writel(priv, 0, INTRL2_CPU_MASK_CLEAR);
+ bcm_sysport_mask_all_intrs(priv);
ret = request_irq(priv->irq0, bcm_sysport_rx_isr, 0, dev->name, dev);
if (ret) {
@@ -1585,10 +1747,13 @@ static int bcm_sysport_open(struct net_device *dev)
goto out_phy_disconnect;
}
- ret = request_irq(priv->irq1, bcm_sysport_tx_isr, 0, dev->name, dev);
- if (ret) {
- netdev_err(dev, "failed to request TX interrupt\n");
- goto out_free_irq0;
+ if (!priv->is_lite) {
+ ret = request_irq(priv->irq1, bcm_sysport_tx_isr, 0,
+ dev->name, dev);
+ if (ret) {
+ netdev_err(dev, "failed to request TX interrupt\n");
+ goto out_free_irq0;
+ }
}
/* Initialize both hardware and software ring */
@@ -1635,7 +1800,8 @@ static int bcm_sysport_open(struct net_device *dev)
out_free_tx_ring:
for (i = 0; i < dev->num_tx_queues; i++)
bcm_sysport_fini_tx_ring(priv, i);
- free_irq(priv->irq1, dev);
+ if (!priv->is_lite)
+ free_irq(priv->irq1, dev);
out_free_irq0:
free_irq(priv->irq0, dev);
out_phy_disconnect:
@@ -1653,10 +1819,7 @@ static void bcm_sysport_netif_stop(struct net_device *dev)
phy_stop(dev->phydev);
/* mask all interrupts */
- intrl2_0_mask_set(priv, 0xffffffff);
- intrl2_0_writel(priv, 0xffffffff, INTRL2_CPU_CLEAR);
- intrl2_1_mask_set(priv, 0xffffffff);
- intrl2_1_writel(priv, 0xffffffff, INTRL2_CPU_CLEAR);
+ bcm_sysport_mask_all_intrs(priv);
}
static int bcm_sysport_stop(struct net_device *dev)
@@ -1694,7 +1857,8 @@ static int bcm_sysport_stop(struct net_device *dev)
bcm_sysport_fini_rx_ring(priv);
free_irq(priv->irq0, dev);
- free_irq(priv->irq1, dev);
+ if (!priv->is_lite)
+ free_irq(priv->irq1, dev);
/* Disconnect from PHY */
phy_disconnect(dev->phydev);
@@ -1733,8 +1897,32 @@ static const struct net_device_ops bcm_sysport_netdev_ops = {
#define REV_FMT "v%2x.%02x"
+static const struct bcm_sysport_hw_params bcm_sysport_params[] = {
+ [SYSTEMPORT] = {
+ .is_lite = false,
+ .num_rx_desc_words = SP_NUM_HW_RX_DESC_WORDS,
+ },
+ [SYSTEMPORT_LITE] = {
+ .is_lite = true,
+ .num_rx_desc_words = SP_LT_NUM_HW_RX_DESC_WORDS,
+ },
+};
+
+static const struct of_device_id bcm_sysport_of_match[] = {
+ { .compatible = "brcm,systemportlite-v1.00",
+ .data = &bcm_sysport_params[SYSTEMPORT_LITE] },
+ { .compatible = "brcm,systemport-v1.00",
+ .data = &bcm_sysport_params[SYSTEMPORT] },
+ { .compatible = "brcm,systemport",
+ .data = &bcm_sysport_params[SYSTEMPORT] },
+ { /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, bcm_sysport_of_match);
+
static int bcm_sysport_probe(struct platform_device *pdev)
{
+ const struct bcm_sysport_hw_params *params;
+ const struct of_device_id *of_id = NULL;
struct bcm_sysport_priv *priv;
struct device_node *dn;
struct net_device *dev;
@@ -1745,6 +1933,12 @@ static int bcm_sysport_probe(struct platform_device *pdev)
dn = pdev->dev.of_node;
r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ of_id = of_match_node(bcm_sysport_of_match, dn);
+ if (!of_id || !of_id->data)
+ return -EINVAL;
+
+ /* Fairly quickly we need to know the type of adapter we have */
+ params = of_id->data;
/* Read the Transmit/Receive Queue properties */
if (of_property_read_u32(dn, "systemport,num-txq", &txq))
@@ -1770,10 +1964,14 @@ static int bcm_sysport_probe(struct platform_device *pdev)
if (!priv->tx_rings)
return -ENOMEM;
+ priv->is_lite = params->is_lite;
+ priv->num_rx_desc_words = params->num_rx_desc_words;
+
priv->irq0 = platform_get_irq(pdev, 0);
- priv->irq1 = platform_get_irq(pdev, 1);
+ if (!priv->is_lite)
+ priv->irq1 = platform_get_irq(pdev, 1);
priv->wol_irq = platform_get_irq(pdev, 2);
- if (priv->irq0 <= 0 || priv->irq1 <= 0) {
+ if (priv->irq0 <= 0 || (priv->irq1 <= 0 && !priv->is_lite)) {
dev_err(&pdev->dev, "invalid interrupts\n");
ret = -EINVAL;
goto err_free_netdev;
@@ -1847,8 +2045,9 @@ static int bcm_sysport_probe(struct platform_device *pdev)
priv->rev = topctrl_readl(priv, REV_CNTL) & REV_MASK;
dev_info(&pdev->dev,
- "Broadcom SYSTEMPORT" REV_FMT
+ "Broadcom SYSTEMPORT%s" REV_FMT
" at 0x%p (irqs: %d, %d, TXQs: %d, RXQs: %d)\n",
+ priv->is_lite ? " Lite" : "",
(priv->rev >> 8) & 0xff, priv->rev & 0xff,
priv->base, priv->irq0, priv->irq1, txq, rxq);
@@ -2044,7 +2243,10 @@ static int bcm_sysport_resume(struct device *d)
rbuf_init(priv);
/* Set maximum frame length */
- umac_writel(priv, UMAC_MAX_MTU_SIZE, UMAC_MAX_FRAME_LEN);
+ if (!priv->is_lite)
+ umac_writel(priv, UMAC_MAX_MTU_SIZE, UMAC_MAX_FRAME_LEN);
+ else
+ gib_set_pad_extension(priv);
/* Set MAC address */
umac_set_hw_addr(priv, dev->dev_addr);
@@ -2080,13 +2282,6 @@ static int bcm_sysport_resume(struct device *d)
static SIMPLE_DEV_PM_OPS(bcm_sysport_pm_ops,
bcm_sysport_suspend, bcm_sysport_resume);
-static const struct of_device_id bcm_sysport_of_match[] = {
- { .compatible = "brcm,systemport-v1.00" },
- { .compatible = "brcm,systemport" },
- { /* sentinel */ }
-};
-MODULE_DEVICE_TABLE(of, bcm_sysport_of_match);
-
static struct platform_driver bcm_sysport_driver = {
.probe = bcm_sysport_probe,
.remove = bcm_sysport_remove,
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.h b/drivers/net/ethernet/broadcom/bcmsysport.h
index f051356b0274..863ddd7870b7 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.h
+++ b/drivers/net/ethernet/broadcom/bcmsysport.h
@@ -127,6 +127,10 @@ struct bcm_rsb {
#define INTRL2_0_DESC_ALLOC_ERR (1 << 10)
#define INTRL2_0_UNEXP_PKTSIZE_ACK (1 << 11)
+/* SYSTEMPORT Lite groups the TX queues interrupts on instance 0 */
+#define INTRL2_0_TDMA_MBDONE_SHIFT 12
+#define INTRL2_0_TDMA_MBDONE_MASK (0xffff << INTRL2_0_TDMA_MBDONE_SHIFT)
+
/* RXCHK offset and defines */
#define SYS_PORT_RXCHK_OFFSET 0x300
@@ -176,7 +180,9 @@ struct bcm_rsb {
#define RBUF_OK_TO_SEND_MASK 0xff
#define RBUF_CRC_REPLACE (1 << 20)
#define RBUF_OK_TO_SEND_MODE (1 << 21)
-#define RBUF_RSB_SWAP (1 << 22)
+/* SYSTEMPORT Lite uses two bits here */
+#define RBUF_RSB_SWAP0 (1 << 22)
+#define RBUF_RSB_SWAP1 (1 << 23)
#define RBUF_ACPI_EN (1 << 23)
#define RBUF_PKT_RDY_THRESH 0x04
@@ -247,6 +253,7 @@ struct bcm_rsb {
#define MIB_RUNT_CNT_RST (1 << 1)
#define MIB_TX_CNT_RST (1 << 2)
+/* These offsets are valid for SYSTEMPORT and SYSTEMPORT Lite */
#define UMAC_MPD_CTRL 0x620
#define MPD_EN (1 << 0)
#define MSEQ_LEN_SHIFT 16
@@ -258,6 +265,34 @@ struct bcm_rsb {
#define UMAC_MDF_CTRL 0x650
#define UMAC_MDF_ADDR 0x654
+/* Only valid on SYSTEMPORT Lite */
+#define SYS_PORT_GIB_OFFSET 0x1000
+
+#define GIB_CONTROL 0x00
+#define GIB_TX_EN (1 << 0)
+#define GIB_RX_EN (1 << 1)
+#define GIB_TX_FLUSH (1 << 2)
+#define GIB_RX_FLUSH (1 << 3)
+#define GIB_GTX_CLK_SEL_SHIFT 4
+#define GIB_GTX_CLK_EXT_CLK (0 << GIB_GTX_CLK_SEL_SHIFT)
+#define GIB_GTX_CLK_125MHZ (1 << GIB_GTX_CLK_SEL_SHIFT)
+#define GIB_GTX_CLK_250MHZ (2 << GIB_GTX_CLK_SEL_SHIFT)
+#define GIB_FCS_STRIP (1 << 6)
+#define GIB_LCL_LOOP_EN (1 << 7)
+#define GIB_LCL_LOOP_TXEN (1 << 8)
+#define GIB_RMT_LOOP_EN (1 << 9)
+#define GIB_RMT_LOOP_RXEN (1 << 10)
+#define GIB_RX_PAUSE_EN (1 << 11)
+#define GIB_PREAMBLE_LEN_SHIFT 12
+#define GIB_PREAMBLE_LEN_MASK 0xf
+#define GIB_IPG_LEN_SHIFT 16
+#define GIB_IPG_LEN_MASK 0x3f
+#define GIB_PAD_EXTENSION_SHIFT 22
+#define GIB_PAD_EXTENSION_MASK 0x3f
+
+#define GIB_MAC1 0x08
+#define GIB_MAC0 0x0c
+
/* Receive DMA offset and defines */
#define SYS_PORT_RDMA_OFFSET 0x2000
@@ -409,16 +444,19 @@ struct bcm_rsb {
RING_PCP_DEI_VID)
#define TDMA_CONTROL 0x600
-#define TDMA_EN (1 << 0)
-#define TSB_EN (1 << 1)
-#define TSB_SWAP (1 << 2)
-#define ACB_ALGO (1 << 3)
+#define TDMA_EN 0
+#define TSB_EN 1
+/* Uses 2 bits on SYSTEMPORT Lite and shifts everything by 1 bit, we
+ * keep the SYSTEMPORT layout here and adjust with tdma_control_bit()
+ */
+#define TSB_SWAP 2
+#define ACB_ALGO 3
#define BUF_DATA_OFFSET_SHIFT 4
#define BUF_DATA_OFFSET_MASK 0x3ff
-#define VLAN_EN (1 << 14)
-#define SW_BRCM_TAG (1 << 15)
-#define WNC_KPT_SIZE_UPDATE (1 << 16)
-#define SYNC_PKT_SIZE (1 << 17)
+#define VLAN_EN 14
+#define SW_BRCM_TAG 15
+#define WNC_KPT_SIZE_UPDATE 16
+#define SYNC_PKT_SIZE 17
#define ACH_TXDONE_DELAY_SHIFT 18
#define ACH_TXDONE_DELAY_MASK 0xff
@@ -475,12 +513,12 @@ struct dma_desc {
};
/* Number of Receive hardware descriptor words */
-#define NUM_HW_RX_DESC_WORDS 1024
-/* Real number of usable descriptors */
-#define NUM_RX_DESC (NUM_HW_RX_DESC_WORDS / WORDS_PER_DESC)
+#define SP_NUM_HW_RX_DESC_WORDS 1024
+#define SP_LT_NUM_HW_RX_DESC_WORDS 256
-/* Internal linked-list RAM has up to 1536 entries */
-#define NUM_TX_DESC 1536
+/* Internal linked-list RAM size */
+#define SP_NUM_TX_DESC 1536
+#define SP_LT_NUM_TX_DESC 256
#define WORDS_PER_DESC (sizeof(struct dma_desc) / sizeof(u32))
@@ -627,6 +665,16 @@ struct bcm_sysport_cb {
DEFINE_DMA_UNMAP_LEN(dma_len);
};
+enum bcm_sysport_type {
+ SYSTEMPORT = 0,
+ SYSTEMPORT_LITE,
+};
+
+struct bcm_sysport_hw_params {
+ bool is_lite;
+ unsigned int num_rx_desc_words;
+};
+
/* Software view of the TX ring */
struct bcm_sysport_tx_ring {
spinlock_t lock; /* Ring lock for tx reclaim/xmit */
@@ -651,6 +699,8 @@ struct bcm_sysport_priv {
u32 irq0_mask;
u32 irq1_stat;
u32 irq1_mask;
+ bool is_lite;
+ unsigned int num_rx_desc_words;
struct napi_struct napi ____cacheline_aligned;
struct net_device *netdev;
struct platform_device *pdev;
--
2.11.0
^ permalink raw reply related
* Re: [PATCH v2] xen-netfront: Fix Rx stall during network stress and OOM
From: David Miller @ 2017-01-20 19:09 UTC (permalink / raw)
To: vineethp; +Cc: netdev, linux-kernel
In-Reply-To: <66b10c64-936a-8001-6855-2ff1ed626642@amazon.com>
From: Vineeth Remanan Pillai <vineethp@amazon.com>
Date: Thu, 19 Jan 2017 08:35:39 -0800
> From: Vineeth Remanan Pillai <vineethp@amazon.com>
>
> During an OOM scenario, request slots could not be created as skb
> allocation fails. So the netback cannot pass in packets and netfront
> wrongly assumes that there is no more work to be done and it disables
> polling. This causes Rx to stall.
>
> The issue is with the retry logic which schedules the timer if the
> created slots are less than NET_RX_SLOTS_MIN. The count of new request
> slots to be pushed are calculated as a difference between new req_prod
> and rsp_cons which could be more than the actual slots, if there are
> unconsumed responses.
>
> The fix is to calculate the count of newly created slots as the
> difference between new req_prod and old req_prod.
>
> Signed-off-by: Vineeth Remanan Pillai <vineethp@amazon.com>
> Reviewed-by: Juergen Gross <jgross@suse.com>
> ---
> Changes in v2:
> - Removed the old implementation of enabling polling on
> skb allocation error.
> - Corrected the refill timer logic to schedule when newly
> created slots since last push is less than NET_RX_SLOTS_MIN.
Applied.
^ permalink raw reply
* Re: [PATCHv4 net-next 3/5] sctp: implement sender-side procedures for SSN/TSN Reset Request Parameter
From: David Miller @ 2017-01-20 19:00 UTC (permalink / raw)
To: marcelo.leitner; +Cc: lucien.xin, netdev, linux-sctp, nhorman, vyasevich
In-Reply-To: <20170120182521.GV3781@localhost.localdomain>
From: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Date: Fri, 20 Jan 2017 16:25:22 -0200
> I talked offline with Xin about this and we cannot do it this way.
> Unfortunatelly we will have to take the long road here, because then
> we may send data while sending the request, as the streams are not
> closed yet. We really need to close team, send the request, and
> re-open if the send fails.
I am expecting another spin of this series, correct?
^ permalink raw reply
* Re: [PATCH v5 0/2] Add support for the ethernet switch on the ESPRESSObin
From: David Miller @ 2017-01-20 19:17 UTC (permalink / raw)
To: gregory.clement
Cc: thomas.petazzoni, andrew, f.fainelli, jason, shjzhou,
vivien.didelot, netdev, linux-kernel, jpannell, nadavh, kostap,
dingwei, linux-arm-kernel, sebastian.hesselbarth
In-Reply-To: <20170119214934.27442-1-gregory.clement@free-electrons.com>
From: Gregory CLEMENT <gregory.clement@free-electrons.com>
Date: Thu, 19 Jan 2017 22:49:32 +0100
> I created a new family for this switch and filled the ops structure
> by selecting which seems the more appropriate functions. I rebased
> the series on net-next/master which allowed me to benefit to the
> eeprom functions introduced for the 6390.
It looks like there will be at least one more respin of this series,
specifically to remove the new family as Vivien seems to object to
this.
^ permalink raw reply
* Re: [pull request][net-next 00/15] Mellanox mlx5 updates 2017-01-19
From: David Miller @ 2017-01-20 19:23 UTC (permalink / raw)
To: saeedm; +Cc: netdev
In-Reply-To: <20170119223908.15924-1-saeedm@mellanox.com>
From: Saeed Mahameed <saeedm@mellanox.com>
Date: Fri, 20 Jan 2017 00:38:53 +0200
> This pull request includes some small mlx5 updates and two new features,
> The 1st exposes new HW counters to "ethtool -S" and the other introduces
> mlx5 ptp 1pps support. Details are down bleow.
>
> Please pull and let me know if there's any problem.
Pulled, thank you.
^ permalink raw reply
* Re: [Xen-devel] xennet_start_xmit assumptions
From: David Miller @ 2017-01-20 19:30 UTC (permalink / raw)
To: sowmini.varadhan; +Cc: Paul.Durrant, konrad.wilk, wei.liu2, netdev, xen-devel
In-Reply-To: <20170119224123.GB19618@oracle.com>
From: Sowmini Varadhan <sowmini.varadhan@oracle.com>
Date: Thu, 19 Jan 2017 17:41:23 -0500
> On (01/19/17 13:47), Sowmini Varadhan wrote:
>> > Specifically I'm talking about the dev_validate_header() check.
>> > That is supposed to protect us from these kinds of situations.
>>
>> ah, but I run my pf_packet application as root, so I have
>> capable(CAP_SYS_RAWIO), so I slip through the dev_validate_header()
>> check.
>
> and in that light, should dev_validate_header()
> always return false if len == 0?
>
> that will take care of all the send paths in af_packet.c
> but it impacts all drivers as well (even though it is the
> logically correct thing to do..)
I think dev_validate_header() almost does the correct thing in
the SYS_RAWIO case.
It clears out the not-provided hard header bytes, but it doesn't
adjust the skb->len. I think that is a real requirement in this
situation.
CAP_SYS_RAWIO or not, the contract we have with the device is that
there will be at least enough bytes to cover a link layer header.
This probably requires a little bit of an adjustment to the calling
convention. Perhaps:
int dev_validate_header(const struct net_device *dev,
char *ll_header, int len);
So then you can go:
new_len = dev_validate_header(dev, skb->data, len);
if (new_len < 0)
goto out_cleanup_err;
if (new_len > len)
__skb_put(skb, new_len - len);
Or something like that.
^ permalink raw reply
* Re: [PATCH net-next v2 0/2] net: ipv6: Improve user experience with multipath routes
From: David Ahern @ 2017-01-20 19:31 UTC (permalink / raw)
To: netdev
In-Reply-To: <1484892626-14257-1-git-send-email-dsa@cumulusnetworks.com>
On 1/19/17 11:10 PM, David Ahern wrote:
> This series closes a couple of gaps between IPv4 and IPv6 with respect
> to multipath routes:
...
> In both cases, the new behavior requires users to opt in by setting a new
> flag, RTM_F_ALL_NEXTHOPS, in the rtm_flags of struct rtmsg which is
> expected to be the ancillary header in the netlink request received from
> the user. A program must opt in to the new behavior so as to not break
> any existing applications.
>
> The opt-in behavior works for both route deletes and dumps (the two
> differences noted above), but not for notifications as notifications
> do not take user input to specify flags. The only way to have
> notifications generate RTA_MULTIPATH encodings is to have a gobal
> flag -- e.g., sysctl. I'd prefer not to add a sysctl knob for this
> backwards compatibility.
BTW, I am in favor of not requiring a user API for this but just doing it. I can't imagine anyone working with multipath routes not wanting the efficiency of the RTA_MULTIPATH attribute. These patches require an API only because of the rule not to break userspace. If we conclude to just do it without an API, the multipath_add and multipath_del need to be modified to only send a notification once at the end of the actions.
^ permalink raw reply
* Re: [PATCH] inet: don't use sk_v6_rcv_saddr directly
From: David Miller @ 2017-01-20 19:36 UTC (permalink / raw)
To: jbacik; +Cc: tom, netdev, kafai
In-Reply-To: <1484866066-25407-1-git-send-email-jbacik@fb.com>
From: Josef Bacik <jbacik@fb.com>
Date: Thu, 19 Jan 2017 17:47:46 -0500
> When comparing two sockets we need to use inet6_rcv_saddr so we get a NULL
> sk_v6_rcv_saddr if the socket isn't AF_INET6, otherwise our comparison function
> can be wrong.
>
> Fixes: 637bc8b ("inet: reset tb->fastreuseport when adding a reuseport sk")
> Signed-off-by: Josef Bacik <jbacik@fb.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH v2 0/2] net: dsa: Move temperature sensor code into PHY.
From: David Miller @ 2017-01-20 19:43 UTC (permalink / raw)
To: andrew; +Cc: netdev, vivien.didelot
In-Reply-To: <1484872670-32585-1-git-send-email-andrew@lunn.ch>
From: Andrew Lunn <andrew@lunn.ch>
Date: Fri, 20 Jan 2017 01:37:48 +0100
> Marvell Ethernet switches contain a temperature sensor. There appears
> to be one sensor, which is shared by each of the internal PHYs. Each
> PHY has independent registers to read this sensor, and to set a limit
> for when an alarm should be raised.
>
> Some Marvell discrete PHY also have the same sensor and registers.
> Moving the HWMON code from DSA into the PHY makes the sensor available
> in discrete PHYs, and removes the layering violation, the switch
> driver poking around in PHY registers.
>
> While moving the code into the PHY driver, it has been re-written to
> use the new HWMON APIs.
>
> v2:
>
> Better Cover note explaining one sensor, but multiple independent
> registers
>
> Simply error checking.
I know there was minor request for a respin, but I'm not going to hold
this up any more just for that.
Series applied, thanks Andrew.
^ permalink raw reply
* Re: [PATCH net] net: mpls: Fix multipath selection for LSR use case
From: David Miller @ 2017-01-20 19:49 UTC (permalink / raw)
To: dsa; +Cc: netdev, rshearma, roopa
In-Reply-To: <1484873463-11699-1-git-send-email-dsa@cumulusnetworks.com>
From: David Ahern <dsa@cumulusnetworks.com>
Date: Thu, 19 Jan 2017 16:51:03 -0800
> MPLS multipath for LSR is broken -- always selecting the first nexthop
> in the one label case. For example:
...
David, this doesn't apply cleanly to the net tree, please respin.
Thanks.
^ permalink raw reply
* Re: [Xen-devel] xennet_start_xmit assumptions
From: Sowmini Varadhan @ 2017-01-20 20:03 UTC (permalink / raw)
To: David Miller; +Cc: Paul.Durrant, konrad.wilk, wei.liu2, netdev, xen-devel
In-Reply-To: <20170120.143059.1390682983473502518.davem@davemloft.net>
On (01/20/17 14:30), David Miller wrote:
>
> CAP_SYS_RAWIO or not, the contract we have with the device is that
> there will be at least enough bytes to cover a link layer header.
I see. If that's the case (for all the kernel-driver interfaces),
then the xen_netfront driver is probably not required to check for
variants of sk_buffs like pure-non-linear etc.
> This probably requires a little bit of an adjustment to the calling
> convention. Perhaps:
>
> int dev_validate_header(const struct net_device *dev,
> char *ll_header, int len);
>
> So then you can go:
>
> new_len = dev_validate_header(dev, skb->data, len);
> if (new_len < 0)
> goto out_cleanup_err;
> if (new_len > len)
> __skb_put(skb, new_len - len);
>
> Or something like that.
ok let me work with that and get back (hopefully with an
RFC patch).
--Sowmini
^ permalink raw reply
* Re: [PATCH net v3] bridge: netlink: call br_changelink() during br_dev_newlink()
From: David Miller @ 2017-01-20 20:08 UTC (permalink / raw)
To: jiri; +Cc: netdev, jiri, cera, bridge
In-Reply-To: <20170120181042.GB1784@nanopsycho.orion>
From: Jiri Pirko <jiri@resnulli.us>
Date: Fri, 20 Jan 2017 19:10:42 +0100
> Fri, Jan 20, 2017 at 06:12:17PM CET, cera@cera.cz wrote:
>>Any bridge options specified during link creation (e.g. ip link add)
>>are ignored as br_dev_newlink() does not process them.
>>Use br_changelink() to do it.
>>
>>Fixes: 1332351 ("bridge: implement rtnl_link_ops->changelink")
>
> Should have 12 chars. Other than that,
>
> Reviewed-by: Jiri Pirko <jiri@mellanox.com>
I fixed up the SHA1-ID.
Applied and queued up for -stable, thanks.
^ permalink raw reply
* Re: [PATCH net 0/2] net: Fix oops on state free after lwt module unload
From: Robert Shearman @ 2017-01-20 20:21 UTC (permalink / raw)
To: David Miller; +Cc: netdev, tom, roopa
In-Reply-To: <20170120.120321.688277769193451439.davem@davemloft.net>
On 20/01/17 17:03, David Miller wrote:
> From: Robert Shearman <rshearma@brocade.com>
> Date: Wed, 18 Jan 2017 15:32:01 +0000
>
>> This patchset fixes an oops in lwtstate_free and a memory leak that
>> would otherwise be exposed by ensuring that references are taken on
>> modules that need to stay around to clean up lwt state. To faciliate
>> this all ops that implement destroy_state and that can be configured
>> to build as a module are changed specify the owner module in the
>> ops. The intersection of those two sets is just ila at the moment.
>
> Two things:
>
> 1) Under no circumstances should we allow a lwtunnel ops implementing
> module to unload while there is a rule using those ops which is
> alive.
>
> Therefore, we should not special case the destroy op. We should
> unconditionally grab the module reference.
>
> 2) Please add the new 'owner' field and add an appropriate assignment
> for ops->owner to _every_ lwtunnel implementation, and do so in
> your first patch. Please do not only do this for ILA.
>
> Thanks.
Very clear, makes sense, will do.
Thanks,
Rob
^ permalink raw reply
* [PATCH net-next 0/7] net: dsa: bcm_sf2: Add support for BCM7278
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
Hi all,
This patch series adds support for the Broadcom BCM7278 integrated switch
which is a successor of the BCM7445 switch. We have a little bit of
register shuffling going on, which is why most of the functional changes
are to deal with that.
Thanks!
Florian Fainelli (7):
net: dsa: bcm_sf2: Make SF2_IO64_MACRO() utilize 32-bit macro
net: dsa: bcm_sf2: Prepare for different register layouts
net: dsa: bcm_sf2: Add support for BCM7278 integrated switch
net: dsa: bcm_sf2: Move code enabling Broadcom tags
net: dsa: bcm_sf2: Allow non-IMP ports to have Broadcom tags enabled
net: phy: bcm7xxx: Add entry for BCM7278
net: phy: bcm7xxx: Implement EGPHY workaround for 7278
.../bindings/net/brcm,bcm7445-switch-v4.0.txt | 10 +-
drivers/net/dsa/b53/b53_common.c | 12 ++
drivers/net/dsa/b53/b53_priv.h | 4 +-
drivers/net/dsa/bcm_sf2.c | 167 ++++++++++++++++-----
drivers/net/dsa/bcm_sf2.h | 41 ++++-
drivers/net/dsa/bcm_sf2_regs.h | 47 +++---
drivers/net/phy/bcm7xxx.c | 36 +++++
include/linux/brcmphy.h | 1 +
8 files changed, 261 insertions(+), 57 deletions(-)
--
2.9.3
^ permalink raw reply
* [PATCH net-next 1/7] net: dsa: bcm_sf2: Make SF2_IO64_MACRO() utilize 32-bit macro
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
In-Reply-To: <20170120203634.2773-1-f.fainelli@gmail.com>
There is no point inlining the 32-bit direct register read/write part,
just infer it from the existing macro. This will make it easier to
centralize the address rewriting that we are going to introduce later
on.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/dsa/bcm_sf2.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/dsa/bcm_sf2.h b/drivers/net/dsa/bcm_sf2.h
index 44692673e1d5..4531c2333e86 100644
--- a/drivers/net/dsa/bcm_sf2.h
+++ b/drivers/net/dsa/bcm_sf2.h
@@ -125,7 +125,7 @@ static inline u64 name##_readq(struct bcm_sf2_priv *priv, u32 off) \
{ \
u32 indir, dir; \
spin_lock(&priv->indir_lock); \
- dir = __raw_readl(priv->name + off); \
+ dir = name##_readl(priv, off); \
indir = reg_readl(priv, REG_DIR_DATA_READ); \
spin_unlock(&priv->indir_lock); \
return (u64)indir << 32 | dir; \
@@ -135,7 +135,7 @@ static inline void name##_writeq(struct bcm_sf2_priv *priv, u64 val, \
{ \
spin_lock(&priv->indir_lock); \
reg_writel(priv, upper_32_bits(val), REG_DIR_DATA_WRITE); \
- __raw_writel(lower_32_bits(val), priv->name + off); \
+ name##_writel(priv, lower_32_bits(val), off); \
spin_unlock(&priv->indir_lock); \
}
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 2/7] net: dsa: bcm_sf2: Prepare for different register layouts
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
In-Reply-To: <20170120203634.2773-1-f.fainelli@gmail.com>
In preparation for supporting a new device with a slightly different
register layout, affecting the SWITCH_REG and SWITCH_CORE address
spaces, perform a few preparatory steps:
- allow matching the compatible string against a data description
- convert the SWITCH_REG register accesses into an indirection table
- prepare for supporting a SWITCH_CORE register alignment requirement
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/dsa/bcm_sf2.c | 57 +++++++++++++++++++++++++++++++++++++-----
drivers/net/dsa/bcm_sf2.h | 34 +++++++++++++++++++++++--
drivers/net/dsa/bcm_sf2_regs.h | 43 ++++++++++++++++++-------------
3 files changed, 109 insertions(+), 25 deletions(-)
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index 31d017086f8b..d952099afc60 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -1009,10 +1009,49 @@ static const struct dsa_switch_ops bcm_sf2_ops = {
.port_fdb_del = b53_fdb_del,
};
+struct bcm_sf2_of_data {
+ u32 type;
+ const u16 *reg_offsets;
+ unsigned int core_reg_align;
+};
+
+/* Register offsets for the SWITCH_REG_* block */
+static const u16 bcm_sf2_7445_reg_offsets[] = {
+ [REG_SWITCH_CNTRL] = 0x00,
+ [REG_SWITCH_STATUS] = 0x04,
+ [REG_DIR_DATA_WRITE] = 0x08,
+ [REG_DIR_DATA_READ] = 0x0C,
+ [REG_SWITCH_REVISION] = 0x18,
+ [REG_PHY_REVISION] = 0x1C,
+ [REG_SPHY_CNTRL] = 0x2C,
+ [REG_RGMII_0_CNTRL] = 0x34,
+ [REG_RGMII_1_CNTRL] = 0x40,
+ [REG_RGMII_2_CNTRL] = 0x4c,
+ [REG_LED_0_CNTRL] = 0x90,
+ [REG_LED_1_CNTRL] = 0x94,
+ [REG_LED_2_CNTRL] = 0x98,
+};
+
+static const struct bcm_sf2_of_data bcm_sf2_7445_data = {
+ .type = BCM7445_DEVICE_ID,
+ .core_reg_align = 0,
+ .reg_offsets = bcm_sf2_7445_reg_offsets,
+};
+
+static const struct of_device_id bcm_sf2_of_match[] = {
+ { .compatible = "brcm,bcm7445-switch-v4.0",
+ .data = &bcm_sf2_7445_data
+ },
+ { /* sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, bcm_sf2_of_match);
+
static int bcm_sf2_sw_probe(struct platform_device *pdev)
{
const char *reg_names[BCM_SF2_REGS_NUM] = BCM_SF2_REGS_NAME;
struct device_node *dn = pdev->dev.of_node;
+ const struct of_device_id *of_id = NULL;
+ const struct bcm_sf2_of_data *data;
struct b53_platform_data *pdata;
struct dsa_switch_ops *ops;
struct bcm_sf2_priv *priv;
@@ -1040,11 +1079,22 @@ static int bcm_sf2_sw_probe(struct platform_device *pdev)
if (!pdata)
return -ENOMEM;
+ of_id = of_match_node(bcm_sf2_of_match, dn);
+ if (!of_id || !of_id->data)
+ return -EINVAL;
+
+ data = of_id->data;
+
+ /* Set SWITCH_REG register offsets and SWITCH_CORE align factor */
+ priv->type = data->type;
+ priv->reg_offsets = data->reg_offsets;
+ priv->core_reg_align = data->core_reg_align;
+
/* Auto-detection using standard registers will not work, so
* provide an indication of what kind of device we are for
* b53_common to work with
*/
- pdata->chip_id = BCM7445_DEVICE_ID;
+ pdata->chip_id = priv->type;
dev->pdata = pdata;
priv->dev = dev;
@@ -1190,11 +1240,6 @@ static int bcm_sf2_resume(struct device *dev)
static SIMPLE_DEV_PM_OPS(bcm_sf2_pm_ops,
bcm_sf2_suspend, bcm_sf2_resume);
-static const struct of_device_id bcm_sf2_of_match[] = {
- { .compatible = "brcm,bcm7445-switch-v4.0" },
- { /* sentinel */ },
-};
-MODULE_DEVICE_TABLE(of, bcm_sf2_of_match);
static struct platform_driver bcm_sf2_driver = {
.probe = bcm_sf2_sw_probe,
diff --git a/drivers/net/dsa/bcm_sf2.h b/drivers/net/dsa/bcm_sf2.h
index 4531c2333e86..a1430866bd79 100644
--- a/drivers/net/dsa/bcm_sf2.h
+++ b/drivers/net/dsa/bcm_sf2.h
@@ -61,6 +61,11 @@ struct bcm_sf2_priv {
void __iomem *fcb;
void __iomem *acb;
+ /* Register offsets indirection tables */
+ u32 type;
+ const u16 *reg_offsets;
+ unsigned int core_reg_align;
+
/* spinlock protecting access to the indirect registers */
spinlock_t indir_lock;
@@ -104,6 +109,11 @@ static inline struct bcm_sf2_priv *bcm_sf2_to_priv(struct dsa_switch *ds)
return dev->priv;
}
+static inline u32 bcm_sf2_mangle_addr(struct bcm_sf2_priv *priv, u32 off)
+{
+ return off << priv->core_reg_align;
+}
+
#define SF2_IO_MACRO(name) \
static inline u32 name##_readl(struct bcm_sf2_priv *priv, u32 off) \
{ \
@@ -153,8 +163,28 @@ static inline void intrl2_##which##_mask_set(struct bcm_sf2_priv *priv, \
priv->irq##which##_mask |= (mask); \
} \
-SF2_IO_MACRO(core);
-SF2_IO_MACRO(reg);
+static inline u32 core_readl(struct bcm_sf2_priv *priv, u32 off)
+{
+ u32 tmp = bcm_sf2_mangle_addr(priv, off);
+ return __raw_readl(priv->core + tmp);
+}
+
+static inline void core_writel(struct bcm_sf2_priv *priv, u32 val, u32 off)
+{
+ u32 tmp = bcm_sf2_mangle_addr(priv, off);
+ __raw_writel(val, priv->core + tmp);
+}
+
+static inline u32 reg_readl(struct bcm_sf2_priv *priv, u16 off)
+{
+ return __raw_readl(priv->reg + priv->reg_offsets[off]);
+}
+
+static inline void reg_writel(struct bcm_sf2_priv *priv, u32 val, u16 off)
+{
+ __raw_writel(val, priv->reg + priv->reg_offsets[off]);
+}
+
SF2_IO64_MACRO(core);
SF2_IO_MACRO(intrl2_0);
SF2_IO_MACRO(intrl2_1);
diff --git a/drivers/net/dsa/bcm_sf2_regs.h b/drivers/net/dsa/bcm_sf2_regs.h
index 838fe373cd6f..f5e566304f0c 100644
--- a/drivers/net/dsa/bcm_sf2_regs.h
+++ b/drivers/net/dsa/bcm_sf2_regs.h
@@ -12,22 +12,36 @@
#define __BCM_SF2_REGS_H
/* Register set relative to 'REG' */
-#define REG_SWITCH_CNTRL 0x00
-#define MDIO_MASTER_SEL (1 << 0)
-#define REG_SWITCH_STATUS 0x04
-#define REG_DIR_DATA_WRITE 0x08
-#define REG_DIR_DATA_READ 0x0C
+enum bcm_sf2_reg_offs {
+ REG_SWITCH_CNTRL = 0,
+ REG_SWITCH_STATUS,
+ REG_DIR_DATA_WRITE,
+ REG_DIR_DATA_READ,
+ REG_SWITCH_REVISION,
+ REG_PHY_REVISION,
+ REG_SPHY_CNTRL,
+ REG_RGMII_0_CNTRL,
+ REG_RGMII_1_CNTRL,
+ REG_RGMII_2_CNTRL,
+ REG_LED_0_CNTRL,
+ REG_LED_1_CNTRL,
+ REG_LED_2_CNTRL,
+ REG_SWITCH_REG_MAX,
+};
+
+/* Relative to REG_SWITCH_CNTRL */
+#define MDIO_MASTER_SEL (1 << 0)
-#define REG_SWITCH_REVISION 0x18
+/* Relative to REG_SWITCH_REVISION */
#define SF2_REV_MASK 0xffff
#define SWITCH_TOP_REV_SHIFT 16
#define SWITCH_TOP_REV_MASK 0xffff
-#define REG_PHY_REVISION 0x1C
+/* Relative to REG_PHY_REVISION */
#define PHY_REVISION_MASK 0xffff
-#define REG_SPHY_CNTRL 0x2C
+/* Relative to REG_SPHY_CNTRL */
#define IDDQ_BIAS (1 << 0)
#define EXT_PWR_DOWN (1 << 1)
#define FORCE_DLL_EN (1 << 2)
@@ -37,13 +51,8 @@
#define PHY_PHYAD_SHIFT 8
#define PHY_PHYAD_MASK 0x1F
-#define REG_RGMII_0_BASE 0x34
-#define REG_RGMII_CNTRL 0x00
-#define REG_RGMII_IB_STATUS 0x04
-#define REG_RGMII_RX_CLOCK_DELAY_CNTRL 0x08
-#define REG_RGMII_CNTRL_SIZE 0x0C
-#define REG_RGMII_CNTRL_P(x) (REG_RGMII_0_BASE + \
- ((x) * REG_RGMII_CNTRL_SIZE))
+#define REG_RGMII_CNTRL_P(x) (REG_RGMII_0_CNTRL + (x))
+
/* Relative to REG_RGMII_CNTRL */
#define RGMII_MODE_EN (1 << 0)
#define ID_MODE_DIS (1 << 1)
@@ -61,8 +70,8 @@
#define LPI_COUNT_SHIFT 9
#define LPI_COUNT_MASK 0x3F
-#define REG_LED_CNTRL_BASE 0x90
-#define REG_LED_CNTRL(x) (REG_LED_CNTRL_BASE + (x) * 4)
+#define REG_LED_CNTRL(x) (REG_LED_0_CNTRL + (x))
+
#define SPDLNK_SRC_SEL (1 << 24)
/* Register set relative to 'INTRL2_0' and 'INTRL2_1' */
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 3/7] net: dsa: bcm_sf2: Add support for BCM7278 integrated switch
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
In-Reply-To: <20170120203634.2773-1-f.fainelli@gmail.com>
Add support for the integrated switch found on BCM7278:
- core_reg_align is set to 1, to force a translation into the target
address space which is 8 bytes aligned
- an alternate SWITCH_REG layout is provided since registers are largely
bit/masks compatible but have different offsets
- conditional for all CORE_STS_OVERRIDE_{IMP,GMII_P} since those got
moved way out of the traditional register space
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
.../bindings/net/brcm,bcm7445-switch-v4.0.txt | 2 +-
drivers/net/dsa/b53/b53_common.c | 12 +++++
drivers/net/dsa/b53/b53_priv.h | 4 +-
drivers/net/dsa/bcm_sf2.c | 56 ++++++++++++++++++----
drivers/net/dsa/bcm_sf2_regs.h | 4 ++
5 files changed, 68 insertions(+), 10 deletions(-)
diff --git a/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt b/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt
index fb40891ee606..e1b2c3e32859 100644
--- a/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt
+++ b/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt
@@ -2,7 +2,7 @@
Required properties:
-- compatible: should be "brcm,bcm7445-switch-v4.0"
+- compatible: should be "brcm,bcm7445-switch-v4.0" or "brcm,bcm7278-switch-v4.0"
- reg: addresses and length of the register sets for the device, must be 6
pairs of register addresses and lengths
- interrupts: interrupts for the devices, must be two interrupts
diff --git a/drivers/net/dsa/b53/b53_common.c b/drivers/net/dsa/b53/b53_common.c
index 5102a3701a1a..5cbb14f6a03b 100644
--- a/drivers/net/dsa/b53/b53_common.c
+++ b/drivers/net/dsa/b53/b53_common.c
@@ -1685,6 +1685,18 @@ static const struct b53_chip_data b53_switch_chips[] = {
.jumbo_pm_reg = B53_JUMBO_PORT_MASK,
.jumbo_size_reg = B53_JUMBO_MAX_SIZE,
},
+ {
+ .chip_id = BCM7278_DEVICE_ID,
+ .dev_name = "BCM7278",
+ .vlans = 4096,
+ .enabled_ports = 0x1ff,
+ .arl_entries= 4,
+ .cpu_port = B53_CPU_PORT,
+ .vta_regs = B53_VTA_REGS,
+ .duplex_reg = B53_DUPLEX_STAT_GE,
+ .jumbo_pm_reg = B53_JUMBO_PORT_MASK,
+ .jumbo_size_reg = B53_JUMBO_MAX_SIZE,
+ },
};
static int b53_switch_init(struct b53_device *dev)
diff --git a/drivers/net/dsa/b53/b53_priv.h b/drivers/net/dsa/b53/b53_priv.h
index 86f125d55aaf..a8031b382c55 100644
--- a/drivers/net/dsa/b53/b53_priv.h
+++ b/drivers/net/dsa/b53/b53_priv.h
@@ -62,6 +62,7 @@ enum {
BCM53019_DEVICE_ID = 0x53019,
BCM58XX_DEVICE_ID = 0x5800,
BCM7445_DEVICE_ID = 0x7445,
+ BCM7278_DEVICE_ID = 0x7278,
};
#define B53_N_PORTS 9
@@ -179,7 +180,8 @@ static inline int is5301x(struct b53_device *dev)
static inline int is58xx(struct b53_device *dev)
{
return dev->chip_id == BCM58XX_DEVICE_ID ||
- dev->chip_id == BCM7445_DEVICE_ID;
+ dev->chip_id == BCM7445_DEVICE_ID ||
+ dev->chip_id == BCM7278_DEVICE_ID;
}
#define B53_CPU_PORT_25 5
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index d952099afc60..02afa0598b24 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -64,7 +64,12 @@ static void bcm_sf2_imp_vlan_setup(struct dsa_switch *ds, int cpu_port)
static void bcm_sf2_imp_setup(struct dsa_switch *ds, int port)
{
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
- u32 reg, val;
+ u32 reg, val, offset;
+
+ if (priv->type == BCM7445_DEVICE_ID)
+ offset = CORE_STS_OVERRIDE_IMP;
+ else
+ offset = CORE_STS_OVERRIDE_IMP2;
/* Enable the port memories */
reg = core_readl(priv, CORE_MEM_PSM_VDD_CTRL);
@@ -121,9 +126,9 @@ static void bcm_sf2_imp_setup(struct dsa_switch *ds, int port)
core_writel(priv, reg, CORE_BRCM_HDR_TX_DIS);
/* Force link status for IMP port */
- reg = core_readl(priv, CORE_STS_OVERRIDE_IMP);
+ reg = core_readl(priv, offset);
reg |= (MII_SW_OR | LINK_STS);
- core_writel(priv, reg, CORE_STS_OVERRIDE_IMP);
+ core_writel(priv, reg, offset);
}
static void bcm_sf2_eee_enable_set(struct dsa_switch *ds, int port, bool enable)
@@ -591,7 +596,12 @@ static void bcm_sf2_sw_adjust_link(struct dsa_switch *ds, int port,
struct ethtool_eee *p = &priv->port_sts[port].eee;
u32 id_mode_dis = 0, port_mode;
const char *str = NULL;
- u32 reg;
+ u32 reg, offset;
+
+ if (priv->type == BCM7445_DEVICE_ID)
+ offset = CORE_STS_OVERRIDE_GMIIP_PORT(port);
+ else
+ offset = CORE_STS_OVERRIDE_GMIIP2_PORT(port);
switch (phydev->interface) {
case PHY_INTERFACE_MODE_RGMII:
@@ -662,7 +672,7 @@ static void bcm_sf2_sw_adjust_link(struct dsa_switch *ds, int port,
if (phydev->duplex == DUPLEX_FULL)
reg |= DUPLX_MODE;
- core_writel(priv, reg, CORE_STS_OVERRIDE_GMIIP_PORT(port));
+ core_writel(priv, reg, offset);
if (!phydev->is_pseudo_fixed_link)
p->eee_enabled = bcm_sf2_eee_init(ds, port, phydev);
@@ -672,9 +682,14 @@ static void bcm_sf2_sw_fixed_link_update(struct dsa_switch *ds, int port,
struct fixed_phy_status *status)
{
struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
- u32 duplex, pause;
+ u32 duplex, pause, offset;
u32 reg;
+ if (priv->type == BCM7445_DEVICE_ID)
+ offset = CORE_STS_OVERRIDE_GMIIP_PORT(port);
+ else
+ offset = CORE_STS_OVERRIDE_GMIIP2_PORT(port);
+
duplex = core_readl(priv, CORE_DUPSTS);
pause = core_readl(priv, CORE_PAUSESTS);
@@ -703,13 +718,13 @@ static void bcm_sf2_sw_fixed_link_update(struct dsa_switch *ds, int port,
status->duplex = !!(duplex & (1 << port));
}
- reg = core_readl(priv, CORE_STS_OVERRIDE_GMIIP_PORT(port));
+ reg = core_readl(priv, offset);
reg |= SW_OVERRIDE;
if (status->link)
reg |= LINK_STS;
else
reg &= ~LINK_STS;
- core_writel(priv, reg, CORE_STS_OVERRIDE_GMIIP_PORT(port));
+ core_writel(priv, reg, offset);
if ((pause & (1 << port)) &&
(pause & (1 << (port + PAUSESTS_TX_PAUSE_SHIFT)))) {
@@ -1038,10 +1053,35 @@ static const struct bcm_sf2_of_data bcm_sf2_7445_data = {
.reg_offsets = bcm_sf2_7445_reg_offsets,
};
+static const u16 bcm_sf2_7278_reg_offsets[] = {
+ [REG_SWITCH_CNTRL] = 0x00,
+ [REG_SWITCH_STATUS] = 0x04,
+ [REG_DIR_DATA_WRITE] = 0x08,
+ [REG_DIR_DATA_READ] = 0x0c,
+ [REG_SWITCH_REVISION] = 0x10,
+ [REG_PHY_REVISION] = 0x14,
+ [REG_SPHY_CNTRL] = 0x24,
+ [REG_RGMII_0_CNTRL] = 0xe0,
+ [REG_RGMII_1_CNTRL] = 0xec,
+ [REG_RGMII_2_CNTRL] = 0xf8,
+ [REG_LED_0_CNTRL] = 0x40,
+ [REG_LED_1_CNTRL] = 0x4c,
+ [REG_LED_2_CNTRL] = 0x58,
+};
+
+static const struct bcm_sf2_of_data bcm_sf2_7278_data = {
+ .type = BCM7278_DEVICE_ID,
+ .core_reg_align = 1,
+ .reg_offsets = bcm_sf2_7278_reg_offsets,
+};
+
static const struct of_device_id bcm_sf2_of_match[] = {
{ .compatible = "brcm,bcm7445-switch-v4.0",
.data = &bcm_sf2_7445_data
},
+ { .compatible = "brcm,bcm7278-switch-v4.0",
+ .data = &bcm_sf2_7278_data
+ },
{ /* sentinel */ },
};
MODULE_DEVICE_TABLE(of, bcm_sf2_of_match);
diff --git a/drivers/net/dsa/bcm_sf2_regs.h b/drivers/net/dsa/bcm_sf2_regs.h
index f5e566304f0c..3b33b8010cc8 100644
--- a/drivers/net/dsa/bcm_sf2_regs.h
+++ b/drivers/net/dsa/bcm_sf2_regs.h
@@ -134,6 +134,9 @@ enum bcm_sf2_reg_offs {
#define GMII_SPEED_UP_2G (1 << 6)
#define MII_SW_OR (1 << 7)
+/* Alternate layout for e.g: 7278 */
+#define CORE_STS_OVERRIDE_IMP2 0x39040
+
#define CORE_NEW_CTRL 0x00084
#define IP_MC (1 << 0)
#define OUTRANGEERR_DISCARD (1 << 1)
@@ -151,6 +154,7 @@ enum bcm_sf2_reg_offs {
#define SW_LEARN_CNTL(x) (1 << (x))
#define CORE_STS_OVERRIDE_GMIIP_PORT(x) (0x160 + (x) * 4)
+#define CORE_STS_OVERRIDE_GMIIP2_PORT(x) (0x39000 + (x) * 8)
#define LINK_STS (1 << 0)
#define DUPLX_MODE (1 << 1)
#define SPEED_SHIFT 2
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 4/7] net: dsa: bcm_sf2: Move code enabling Broadcom tags
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
In-Reply-To: <20170120203634.2773-1-f.fainelli@gmail.com>
In preparation for enabling Broadcom tags on different ports based on
configuration information, dedicate a function that is responsible for
enabling Broadcom tags for a given port and update the IMP port setup to
call it.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/dsa/bcm_sf2.c | 61 ++++++++++++++++++++++++++---------------------
1 file changed, 34 insertions(+), 27 deletions(-)
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index 02afa0598b24..571e112c8e34 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -61,34 +61,9 @@ static void bcm_sf2_imp_vlan_setup(struct dsa_switch *ds, int cpu_port)
}
}
-static void bcm_sf2_imp_setup(struct dsa_switch *ds, int port)
+static void bcm_sf2_brcm_hdr_setup(struct bcm_sf2_priv *priv, int port)
{
- struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
- u32 reg, val, offset;
-
- if (priv->type == BCM7445_DEVICE_ID)
- offset = CORE_STS_OVERRIDE_IMP;
- else
- offset = CORE_STS_OVERRIDE_IMP2;
-
- /* Enable the port memories */
- reg = core_readl(priv, CORE_MEM_PSM_VDD_CTRL);
- reg &= ~P_TXQ_PSM_VDD(port);
- core_writel(priv, reg, CORE_MEM_PSM_VDD_CTRL);
-
- /* Enable Broadcast, Multicast, Unicast forwarding to IMP port */
- reg = core_readl(priv, CORE_IMP_CTL);
- reg |= (RX_BCST_EN | RX_MCST_EN | RX_UCST_EN);
- reg &= ~(RX_DIS | TX_DIS);
- core_writel(priv, reg, CORE_IMP_CTL);
-
- /* Enable forwarding */
- core_writel(priv, SW_FWDG_EN, CORE_SWMODE);
-
- /* Enable IMP port in dumb mode */
- reg = core_readl(priv, CORE_SWITCH_CTRL);
- reg |= MII_DUMB_FWDG_EN;
- core_writel(priv, reg, CORE_SWITCH_CTRL);
+ u32 reg, val;
/* Resolve which bit controls the Broadcom tag */
switch (port) {
@@ -124,6 +99,38 @@ static void bcm_sf2_imp_setup(struct dsa_switch *ds, int port)
reg = core_readl(priv, CORE_BRCM_HDR_TX_DIS);
reg &= ~(1 << port);
core_writel(priv, reg, CORE_BRCM_HDR_TX_DIS);
+}
+
+static void bcm_sf2_imp_setup(struct dsa_switch *ds, int port)
+{
+ struct bcm_sf2_priv *priv = bcm_sf2_to_priv(ds);
+ u32 reg, offset;
+
+ if (priv->type == BCM7445_DEVICE_ID)
+ offset = CORE_STS_OVERRIDE_IMP;
+ else
+ offset = CORE_STS_OVERRIDE_IMP2;
+
+ /* Enable the port memories */
+ reg = core_readl(priv, CORE_MEM_PSM_VDD_CTRL);
+ reg &= ~P_TXQ_PSM_VDD(port);
+ core_writel(priv, reg, CORE_MEM_PSM_VDD_CTRL);
+
+ /* Enable Broadcast, Multicast, Unicast forwarding to IMP port */
+ reg = core_readl(priv, CORE_IMP_CTL);
+ reg |= (RX_BCST_EN | RX_MCST_EN | RX_UCST_EN);
+ reg &= ~(RX_DIS | TX_DIS);
+ core_writel(priv, reg, CORE_IMP_CTL);
+
+ /* Enable forwarding */
+ core_writel(priv, SW_FWDG_EN, CORE_SWMODE);
+
+ /* Enable IMP port in dumb mode */
+ reg = core_readl(priv, CORE_SWITCH_CTRL);
+ reg |= MII_DUMB_FWDG_EN;
+ core_writel(priv, reg, CORE_SWITCH_CTRL);
+
+ bcm_sf2_brcm_hdr_setup(priv, port);
/* Force link status for IMP port */
reg = core_readl(priv, offset);
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 5/7] net: dsa: bcm_sf2: Allow non-IMP ports to have Broadcom tags enabled
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
In-Reply-To: <20170120203634.2773-1-f.fainelli@gmail.com>
Parse the "brcm,use-bcm-hdr" boolean property during ports
identification to fill a bitmask of ports that should have Broadcom tags
enabled. This is needed in some configurations where per-packet metadata
can be exchanged using Broadcom tags between the switch and an on-chip
acceleration device.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
.../devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt | 8 ++++++++
drivers/net/dsa/bcm_sf2.c | 7 +++++++
drivers/net/dsa/bcm_sf2.h | 3 +++
3 files changed, 18 insertions(+)
diff --git a/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt b/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt
index e1b2c3e32859..9a734d808aa7 100644
--- a/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt
+++ b/Documentation/devicetree/bindings/net/brcm,bcm7445-switch-v4.0.txt
@@ -41,6 +41,13 @@ Optional properties:
Admission Control Block supports reporting the number of packets in-flight in a
switch queue
+Port subnodes:
+
+Optional properties:
+
+- brcm,use-bcm-hdr: boolean property, if present, indicates that the switch
+ port has Broadcom tags enabled (per-packet metadata)
+
Example:
switch_top@f0b00000 {
@@ -114,6 +121,7 @@ switch_top@f0b00000 {
port@0 {
label = "gphy";
reg = <0>;
+ brcm,use-bcm-hdr;
};
...
};
diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
index 571e112c8e34..8eecfd227e06 100644
--- a/drivers/net/dsa/bcm_sf2.c
+++ b/drivers/net/dsa/bcm_sf2.c
@@ -236,6 +236,10 @@ static int bcm_sf2_port_setup(struct dsa_switch *ds, int port,
reg &= ~P_TXQ_PSM_VDD(port);
core_writel(priv, reg, CORE_MEM_PSM_VDD_CTRL);
+ /* Enable Broadcom tags for that port if requested */
+ if (priv->brcm_tag_mask & BIT(port))
+ bcm_sf2_brcm_hdr_setup(priv, port);
+
/* Clear the Rx and Tx disable bits and set to no spanning tree */
core_writel(priv, 0, CORE_G_PCTL_PORT(port));
@@ -515,6 +519,9 @@ static void bcm_sf2_identify_ports(struct bcm_sf2_priv *priv,
if (mode == PHY_INTERFACE_MODE_MOCA)
priv->moca_port = port_num;
+
+ if (of_property_read_bool(port, "brcm,use-bcm-hdr"))
+ priv->brcm_tag_mask |= 1 << port_num;
}
}
diff --git a/drivers/net/dsa/bcm_sf2.h b/drivers/net/dsa/bcm_sf2.h
index a1430866bd79..6e1f74e4d471 100644
--- a/drivers/net/dsa/bcm_sf2.h
+++ b/drivers/net/dsa/bcm_sf2.h
@@ -100,6 +100,9 @@ struct bcm_sf2_priv {
struct device_node *master_mii_dn;
struct mii_bus *slave_mii_bus;
struct mii_bus *master_mii_bus;
+
+ /* Bitmask of ports needing BRCM tags */
+ unsigned int brcm_tag_mask;
};
static inline struct bcm_sf2_priv *bcm_sf2_to_priv(struct dsa_switch *ds)
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 6/7] net: phy: bcm7xxx: Add entry for BCM7278
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
In-Reply-To: <20170120203634.2773-1-f.fainelli@gmail.com>
Add support for the BCM7278 28nm process Gigabit Ethernet PHY.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/phy/bcm7xxx.c | 2 ++
include/linux/brcmphy.h | 1 +
2 files changed, 3 insertions(+)
diff --git a/drivers/net/phy/bcm7xxx.c b/drivers/net/phy/bcm7xxx.c
index 264b085d796b..fb11927de0ff 100644
--- a/drivers/net/phy/bcm7xxx.c
+++ b/drivers/net/phy/bcm7xxx.c
@@ -416,6 +416,7 @@ static int bcm7xxx_28nm_probe(struct phy_device *phydev)
static struct phy_driver bcm7xxx_driver[] = {
BCM7XXX_28NM_GPHY(PHY_ID_BCM7250, "Broadcom BCM7250"),
+ BCM7XXX_28NM_GPHY(PHY_ID_BCM7278, "Broadcom BCM7278"),
BCM7XXX_28NM_GPHY(PHY_ID_BCM7364, "Broadcom BCM7364"),
BCM7XXX_28NM_GPHY(PHY_ID_BCM7366, "Broadcom BCM7366"),
BCM7XXX_28NM_GPHY(PHY_ID_BCM7439, "Broadcom BCM7439"),
@@ -430,6 +431,7 @@ static struct phy_driver bcm7xxx_driver[] = {
static struct mdio_device_id __maybe_unused bcm7xxx_tbl[] = {
{ PHY_ID_BCM7250, 0xfffffff0, },
+ { PHY_ID_BCM7278, 0xfffffff0, },
{ PHY_ID_BCM7364, 0xfffffff0, },
{ PHY_ID_BCM7366, 0xfffffff0, },
{ PHY_ID_BCM7346, 0xfffffff0, },
diff --git a/include/linux/brcmphy.h b/include/linux/brcmphy.h
index 4f7d8be9ddbf..295fb3e73de5 100644
--- a/include/linux/brcmphy.h
+++ b/include/linux/brcmphy.h
@@ -24,6 +24,7 @@
#define PHY_ID_BCM57780 0x03625d90
#define PHY_ID_BCM7250 0xae025280
+#define PHY_ID_BCM7278 0xae0251a0
#define PHY_ID_BCM7364 0xae025260
#define PHY_ID_BCM7366 0x600d8490
#define PHY_ID_BCM7346 0x600d8650
--
2.9.3
^ permalink raw reply related
* [PATCH net-next 7/7] net: phy: bcm7xxx: Implement EGPHY workaround for 7278
From: Florian Fainelli @ 2017-01-20 20:36 UTC (permalink / raw)
To: netdev; +Cc: davem, vivien.didelot, andrew, Florian Fainelli
In-Reply-To: <20170120203634.2773-1-f.fainelli@gmail.com>
Implement the HW design team recommended workaround in for 7278. Since
the GPHY now returns its revision information in MII_PHYS_ID[23] we need
to check whether the revision provided in flags is 0 or not.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
drivers/net/phy/bcm7xxx.c | 34 ++++++++++++++++++++++++++++++++++
1 file changed, 34 insertions(+)
diff --git a/drivers/net/phy/bcm7xxx.c b/drivers/net/phy/bcm7xxx.c
index fb11927de0ff..aa01020ab1b9 100644
--- a/drivers/net/phy/bcm7xxx.c
+++ b/drivers/net/phy/bcm7xxx.c
@@ -167,6 +167,31 @@ static int bcm7xxx_28nm_e0_plus_afe_config_init(struct phy_device *phydev)
return 0;
}
+static int bcm7xxx_28nm_a0_patch_afe_config_init(struct phy_device *phydev)
+{
+ /* +1 RC_CAL codes for RL centering for both LT and HT conditions */
+ bcm_phy_write_misc(phydev, AFE_RXCONFIG_2, 0xd003);
+
+ /* Cut master bias current by 2% to compensate for RC_CAL offset */
+ bcm_phy_write_misc(phydev, DSP_TAP10, 0x791b);
+
+ /* Improve hybrid leakage */
+ bcm_phy_write_misc(phydev, AFE_HPF_TRIM_OTHERS, 0x10e3);
+
+ /* Change rx_on_tune 8 to 0xf */
+ bcm_phy_write_misc(phydev, 0x21, 0x2, 0x87f6);
+
+ /* Change 100Tx EEE bandwidth */
+ bcm_phy_write_misc(phydev, 0x22, 0x2, 0x017d);
+
+ /* Enable ffe zero detection for Vitesse interoperability */
+ bcm_phy_write_misc(phydev, 0x26, 0x2, 0x0015);
+
+ r_rc_cal_reset(phydev);
+
+ return 0;
+}
+
static int bcm7xxx_28nm_config_init(struct phy_device *phydev)
{
u8 rev = PHY_BRCM_7XXX_REV(phydev->dev_flags);
@@ -174,6 +199,12 @@ static int bcm7xxx_28nm_config_init(struct phy_device *phydev)
u8 count;
int ret = 0;
+ /* Newer devices have moved the revision information back into a
+ * standard location in MII_PHYS_ID[23]
+ */
+ if (rev == 0)
+ rev = phydev->phy_id & ~phydev->drv->phy_id_mask;
+
pr_info_once("%s: %s PHY revision: 0x%02x, patch: %d\n",
phydev_name(phydev), phydev->drv->name, rev, patch);
@@ -197,6 +228,9 @@ static int bcm7xxx_28nm_config_init(struct phy_device *phydev)
case 0x10:
ret = bcm7xxx_28nm_e0_plus_afe_config_init(phydev);
break;
+ case 0x01:
+ ret = bcm7xxx_28nm_a0_patch_afe_config_init(phydev);
+ break;
default:
break;
}
--
2.9.3
^ permalink raw reply related
* Re: [PATCH] net: qcom/emac: claim the irq only when the device is opened
From: Lino Sanfilippo @ 2017-01-20 20:44 UTC (permalink / raw)
To: Timur Tabi, David Miller, netdev
In-Reply-To: <1484775745-14050-1-git-send-email-timur@codeaurora.org>
Hi,
On 18.01.2017 22:42, Timur Tabi wrote:
> @@ -1029,8 +1017,6 @@ void emac_mac_down(struct emac_adapter *adpt)
> */
> writel(DIS_INT, adpt->base + EMAC_INT_STATUS);
> writel(0, adpt->base + EMAC_INT_MASK);
> - synchronize_irq(adpt->irq.irq);
There is no reason to remove the irq synchronization, is it?
Note that the desriptors are freed after that so we must be sure that
the irq handler is not running any more.
Regards,
Lino
^ permalink raw reply
* Re: [PATCHv4 net-next 3/5] sctp: implement sender-side procedures for SSN/TSN Reset Request Parameter
From: marcelo.leitner @ 2017-01-20 20:47 UTC (permalink / raw)
To: David Miller; +Cc: lucien.xin, netdev, linux-sctp, nhorman, vyasevich
In-Reply-To: <20170120.140034.1018095629148946810.davem@davemloft.net>
On Fri, Jan 20, 2017 at 02:00:34PM -0500, David Miller wrote:
> From: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
> Date: Fri, 20 Jan 2017 16:25:22 -0200
>
> > I talked offline with Xin about this and we cannot do it this way.
> > Unfortunatelly we will have to take the long road here, because then
> > we may send data while sending the request, as the streams are not
> > closed yet. We really need to close team, send the request, and
> > re-open if the send fails.
>
> I am expecting another spin of this series, correct?
Yes.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox