Skip to contentSkip to conversation

net: stmmac: Fixes for -net

PATCH In review Jose Abreu 2 participants 11 messages Dec 18, 2019

Subsystems (4)
lore
Share & tools
lore

Jose Abreu

author
#0 lore raw

v3 cover net: stmmac: Fixes for -net

Fixes for stmmac.

1) Fixes the filtering selftests (again) for cases when the number of multicast filters are not enough.

2) Fixes SPH feature for MTU > default.

3) Fixes the behavior of accepting invalid MTU values.

4) Fixes FCS stripping for multi-descriptor packets.

5) Fixes the change of RX buffer size in XGMAC.

6) Fixes RX buffer size alignment.

7) Fixes the 16KB buffer alignment.

8) Fixes the enabling of 16KB buffer size feature.

9) Always arm the TX coalesce timer so that missed interrupts do not cause a TX queue timeout.

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
---

Jose Abreu (9):
net: stmmac: selftests: Needs to check the number of Multicast regs
net: stmmac: Determine earlier the size of RX buffer
net: stmmac: Do not accept invalid MTU values
net: stmmac: Only the last buffer has the FCS field
net: stmmac: xgmac: Clear previous RX buffer size
net: stmmac: RX buffer size must be 16 byte aligned
net: stmmac: 16KB buffer must be 16 byte aligned
net: stmmac: Enable 16KB buffer size
net: stmmac: Always arm TX Timer at end of transmission start

 drivers/net/ethernet/stmicro/stmmac/common.h       |  5 +-
 drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h     |  2 +
 drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c |  3 +-
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c  | 53 +++++++++++++---------
 .../net/ethernet/stmicro/stmmac/stmmac_selftests.c |  4 ++
 5 files changed, 42 insertions(+), 25 deletions(-)
Signature · 1 line
2.7.4
author
#1 lore raw

PATCH v3 1/9 net: stmmac: selftests: Needs to check the number of Multicast regs

When running the MC and UC filter tests we setup a multicast address that its expected to be blocked. If the number of available multicast registers is zero, driver will always pass the multicast packets which will fail the test.

Check if available multicast addresses is enough before running the tests.

Fixes 091810dbded9 ("net: stmmac: Introduce selftests support")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +4 -0
index f3d8b9336b8e..13227909287c 100644
@@ -624,6 +624,8 @@ static int stmmac_test_mcfilt(struct stmmac_priv *priv)
 		return -EOPNOTSUPP;
 	if (netdev_uc_count(priv->dev) >= priv->hw->unicast_filter_entries)
 		return -EOPNOTSUPP;
+	if (netdev_mc_count(priv->dev) >= priv->hw->multicast_filter_bins)
+		return -EOPNOTSUPP;
 
 	while (--tries) {
 		/* We only need to check the mc_addr for collisions */
@@ -666,6 +668,8 @@ static int stmmac_test_ucfilt(struct stmmac_priv *priv)
 
 	if (stmmac_filter_check(priv))
 		return -EOPNOTSUPP;
+	if (netdev_uc_count(priv->dev) >= priv->hw->unicast_filter_entries)
+		return -EOPNOTSUPP;
 	if (netdev_mc_count(priv->dev) >= priv->hw->multicast_filter_bins)
 		return -EOPNOTSUPP;
 
Signature · 1 line
2.7.4
author
#2 lore raw

PATCH v3 2/9 net: stmmac: Determine earlier the size of RX buffer

In reply to Jose Abreu

Split Header feature needs to know the size of RX buffer but current code is determining it too late. Fix this by moving the RX buffer computation to earlier stage.

Changes from v2:
- Do not try to align already aligned buffer size

Fixes 67afd6d1cfdf ("net: stmmac: Add Split Header support and enable it in XGMAC cores")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 24 +++++++++++------------
 1 file changed, 11 insertions(+), 13 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +11 -13
index bbc65bd332a8..eccbf5daf9ed 100644
@@ -1293,19 +1293,9 @@ static int init_dma_rx_desc_rings(struct net_device *dev, gfp_t flags)
 	struct stmmac_priv *priv = netdev_priv(dev);
 	u32 rx_count = priv->plat->rx_queues_to_use;
 	int ret = -ENOMEM;
-	int bfsize = 0;
 	int queue;
 	int i;
 
-	bfsize = stmmac_set_16kib_bfsize(priv, dev->mtu);
-	if (bfsize < 0)
-		bfsize = 0;
-
-	if (bfsize < BUF_SIZE_16KiB)
-		bfsize = stmmac_set_bfsize(dev->mtu, priv->dma_buf_sz);
-
-	priv->dma_buf_sz = bfsize;
-
 	/* RX INITIALIZATION */
 	netif_dbg(priv, probe, priv->dev,
 		  "SKB addresses:\nskb\t\tskb data\tdma data\n");
@@ -1347,8 +1337,6 @@ static int init_dma_rx_desc_rings(struct net_device *dev, gfp_t flags)
 		}
 	}
 
-	buf_sz = bfsize;
-
 	return 0;
 
 err_init_rx_buffers:
@@ -2658,6 +2646,7 @@ static void stmmac_hw_teardown(struct net_device *dev)
 static int stmmac_open(struct net_device *dev)
 {
 	struct stmmac_priv *priv = netdev_priv(dev);
+	int bfsize = 0;
 	u32 chan;
 	int ret;
 
@@ -2677,7 +2666,16 @@ static int stmmac_open(struct net_device *dev)
 	memset(&priv->xstats, 0, sizeof(struct stmmac_extra_stats));
 	priv->xstats.threshold = tc;
 
-	priv->dma_buf_sz = STMMAC_ALIGN(buf_sz);
+	bfsize = stmmac_set_16kib_bfsize(priv, dev->mtu);
+	if (bfsize < 0)
+		bfsize = 0;
+
+	if (bfsize < BUF_SIZE_16KiB)
+		bfsize = stmmac_set_bfsize(dev->mtu, priv->dma_buf_sz);
+
+	priv->dma_buf_sz = bfsize;
+	buf_sz = bfsize;
+
 	priv->rx_copybreak = STMMAC_RX_COPYBREAK;
 
 	ret = alloc_dma_desc_resources(priv);
Signature · 1 line
2.7.4
author
#3 lore raw

PATCH v3 3/9 net: stmmac: Do not accept invalid MTU values

In reply to Jose Abreu

The maximum MTU value is determined by the maximum size of TX FIFO so that a full packet can fit in the FIFO. Add a check for this in the MTU change callback.

Also check if provided and rounded MTU does not passes the maximum limit of 16K.

Changes from v2:
- Align MTU before checking if its valid

Fixes 7ac6653a085b ("stmmac: Move the STMicroelectronics driver")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +12 -0
index eccbf5daf9ed..8b7861909ef9 100644
@@ -3827,12 +3827,24 @@ static void stmmac_set_rx_mode(struct net_device *dev)
 static int stmmac_change_mtu(struct net_device *dev, int new_mtu)
 {
 	struct stmmac_priv *priv = netdev_priv(dev);
+	int txfifosz = priv->plat->tx_fifo_size;
+
+	if (txfifosz == 0)
+		txfifosz = priv->dma_cap.tx_fifo_size;
+
+	txfifosz /= priv->plat->tx_queues_to_use;
 
 	if (netif_running(dev)) {
 		netdev_err(priv->dev, "must be stopped to change its MTU\n");
 		return -EBUSY;
 	}
 
+	new_mtu = STMMAC_ALIGN(new_mtu);
+
+	/* If condition true, FIFO is too small or MTU too large */
+	if ((txfifosz < new_mtu) || (new_mtu > BUF_SIZE_16KiB))
+		return -EINVAL;
+
 	dev->mtu = new_mtu;
 
 	netdev_update_features(dev);
Signature · 1 line
2.7.4
author
#4 lore raw

PATCH v3 4/9 net: stmmac: Only the last buffer has the FCS field

In reply to Jose Abreu

Only the last received buffer contains the FCS field. Check for end of packet before trying to strip the FCS field.

Fixes 88ebe2cf7f3f ("net: stmmac: Rework stmmac_rx()")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +3 -2
index 8b7861909ef9..acb14a96243e 100644
@@ -3644,8 +3644,9 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue)
 		 * feature is always disabled and packets need to be
 		 * stripped manually.
 		 */
-		if (unlikely(priv->synopsys_id >= DWMAC_CORE_4_00) ||
-		    unlikely(status != llc_snap)) {
+		if (likely(!(status & rx_not_ls)) &&
+		    (likely(priv->synopsys_id >= DWMAC_CORE_4_00) ||
+		     unlikely(status != llc_snap))) {
 			if (buf2_len)
 				buf2_len -= ETH_FCS_LEN;
 			else
Signature · 1 line
2.7.4
author
#5 lore raw

PATCH v3 5/9 net: stmmac: xgmac: Clear previous RX buffer size

In reply to Jose Abreu

When switching between buffer sizes we need to clear the previous value.

Fixes d6ddfacd95c7 ("net: stmmac: Add DMA related callbacks for XGMAC2")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h     | 2 ++
 drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c | 3 ++-
 2 files changed, 4 insertions(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h +2 -0
index 3b6e559aa0b9..ef8a07c68ca7 100644
@@ -343,6 +343,8 @@
 #define XGMAC_DMA_CH_RX_CONTROL(x)	(0x00003108 + (0x80 * (x)))
 #define XGMAC_RxPBL			GENMASK(21, 16)
 #define XGMAC_RxPBL_SHIFT		16
+#define XGMAC_RBSZ			GENMASK(14, 1)
+#define XGMAC_RBSZ_SHIFT		1
 #define XGMAC_RXST			BIT(0)
 #define XGMAC_DMA_CH_TxDESC_HADDR(x)	(0x00003110 + (0x80 * (x)))
 #define XGMAC_DMA_CH_TxDESC_LADDR(x)	(0x00003114 + (0x80 * (x)))
drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c +2 -1
index 22a7f0cc1b90..f3f08ccc379b 100644
@@ -482,7 +482,8 @@ static void dwxgmac2_set_bfsize(void __iomem *ioaddr, int bfsize, u32 chan)
 	u32 value;
 
 	value = readl(ioaddr + XGMAC_DMA_CH_RX_CONTROL(chan));
-	value |= bfsize << 1;
+	value &= ~XGMAC_RBSZ;
+	value |= bfsize << XGMAC_RBSZ_SHIFT;
 	writel(value, ioaddr + XGMAC_DMA_CH_RX_CONTROL(chan));
 }
 
Signature · 1 line
2.7.4
author
#6 lore raw

PATCH v3 6/9 net: stmmac: RX buffer size must be 16 byte aligned

In reply to Jose Abreu

We need to align the RX buffer size to at least 16 byte so that IP doesn't mis-behave. This is required by HW.

Changes from v2:
- Align UP and not DOWN (David)

Fixes 7ac6653a085b ("stmmac: Move the STMicroelectronics driver")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +1 -1
index acb14a96243e..f507a9bee15b 100644
@@ -46,7 +46,7 @@
 #include "dwxgmac2.h"
 #include "hwif.h"
 
-#define	STMMAC_ALIGN(x)		__ALIGN_KERNEL(x, SMP_CACHE_BYTES)
+#define	STMMAC_ALIGN(x)		ALIGN(ALIGN(x, SMP_CACHE_BYTES), 16)
 #define	TSO_MAX_BUFF_SIZE	(SZ_16K - 1)
 
 /* Module parameters */
Signature · 1 line
2.7.4
author
#7 lore raw

PATCH v3 7/9 net: stmmac: 16KB buffer must be 16 byte aligned

In reply to Jose Abreu

The 16KB RX Buffer must also be 16 byte aligned. Fix it.

Fixes 7ac6653a085b ("stmmac: Move the STMicroelectronics driver")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/common.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/net/ethernet/stmicro/stmmac/common.h +2 -3
index b210e987a1db..94f94686cf7d 100644
@@ -365,9 +365,8 @@ struct dma_features {
 	unsigned int arpoffsel;
 };
 
-/* GMAC TX FIFO is 8K, Rx FIFO is 16K */
-#define BUF_SIZE_16KiB 16384
-/* RX Buffer size must be < 8191 and multiple of 4/8/16 bytes */
+/* RX Buffer size must be multiple of 4/8/16 bytes */
+#define BUF_SIZE_16KiB 16368
 #define BUF_SIZE_8KiB 8188
 #define BUF_SIZE_4KiB 4096
 #define BUF_SIZE_2KiB 2048
Signature · 1 line
2.7.4
author
#8 lore raw

PATCH v3 8/9 net: stmmac: Enable 16KB buffer size

In reply to Jose Abreu

XGMAC supports maximum MTU that can go to 16KB. Lets add this check in the calculation of RX buffer size.

Fixes 7ac6653a085b ("stmmac: Move the STMicroelectronics driver")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +3 -1
index f507a9bee15b..1cb466dd5b3f 100644
@@ -1109,7 +1109,9 @@ static int stmmac_set_bfsize(int mtu, int bufsize)
 {
 	int ret = bufsize;
 
-	if (mtu >= BUF_SIZE_4KiB)
+	if (mtu >= BUF_SIZE_8KiB)
+		ret = BUF_SIZE_16KiB;
+	else if (mtu >= BUF_SIZE_4KiB)
 		ret = BUF_SIZE_8KiB;
 	else if (mtu >= BUF_SIZE_2KiB)
 		ret = BUF_SIZE_4KiB;
Signature · 1 line
2.7.4
author
#9 lore raw

PATCH v3 9/9 net: stmmac: Always arm TX Timer at end of transmission start

In reply to Jose Abreu

If TX Coalesce timer is enabled we should always arm it, otherwise we may hit the case where an interrupt is missed and the TX Queue will timeout.

Arming the timer does not necessarly mean it will run the tx_clean() because this function is wrapped around NAPI launcher.

Fixes 9125cdd1be11 ("stmmac: add the initial tx coalesce schema")

---
Cc: Giuseppe Cavallaro <[email protected]>
Cc: Alexandre Torgue <[email protected]>
Cc: Jose Abreu <[email protected]>
Cc: "David S. Miller" <[email protected]>
Cc: Maxime Coquelin <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +2 -4
index 1cb466dd5b3f..6f51a265459d 100644
@@ -3053,8 +3053,6 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev)
 		tx_q->tx_count_frames = 0;
 		stmmac_set_tx_ic(priv, desc);
 		priv->xstats.tx_set_ic_bit++;
-	} else {
-		stmmac_tx_timer_arm(priv, queue);
 	}
 
 	/* We've used all descriptors we need for this skb, however,
@@ -3125,6 +3123,7 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * sizeof(*desc));
 	stmmac_set_tx_tail_ptr(priv, priv->ioaddr, tx_q->tx_tail_addr, queue);
+	stmmac_tx_timer_arm(priv, queue);
 
 	return NETDEV_TX_OK;
 
@@ -3276,8 +3275,6 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
 		tx_q->tx_count_frames = 0;
 		stmmac_set_tx_ic(priv, desc);
 		priv->xstats.tx_set_ic_bit++;
-	} else {
-		stmmac_tx_timer_arm(priv, queue);
 	}
 
 	/* We've used all descriptors we need for this skb, however,
@@ -3366,6 +3363,7 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev)
 
 	tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * sizeof(*desc));
 	stmmac_set_tx_tail_ptr(priv, priv->ioaddr, tx_q->tx_tail_addr, queue);
+	stmmac_tx_timer_arm(priv, queue);
 
 	return NETDEV_TX_OK;
 
Signature · 1 line
2.7.4
maintainer
#10 lore raw
In reply to Jose Abreu

From: Jose Abreu <[email protected]>
Date: Wed, 18 Dec 2019 11:17:34 +0100

Quoted 1 line

Fixes for stmmac.

Series applied.

But realistically I doubt you'll ever find a configuration where SMP_CACHE_BYTES is less than 16 (seriously, it is so non-sensible to build a cpu like that). So you could have done something like:

#if SMP_CACHE_BYTES < 16
#error SMP_CACHE_BYTES too small
#endif

and then add the funky double alignment code if that ever triggered.

But again it never will.