Skip to contentSkip to conversation

Xilinx axienet fixes

PATCH Superseded Robert Hancock 4 participants 22 messages Jan 12, 2022 – Jan 18, 2022

Subsystems (4)

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

v2 cover Xilinx axienet fixes

Various fixes for the Xilinx AXI Ethernet driver.

Changed since v1:
-corrected a Fixes tag to point to mainline commit -split up reset changes into 3 patches -added ratelimit on netdev_warn in TX busy case

Robert Hancock (9):
net: axienet: increase reset timeout
net: axienet: Wait for PhyRstCmplt after core reset
net: axienet: reset core on initialization prior to MDIO access
net: axienet: add missing memory barriers
net: axienet: limit minimum TX ring size
net: axienet: Fix TX ring slot available check
net: axienet: fix number of TX ring slots for available check
net: axienet: fix for TX busy handling
net: axienet: increase default TX ring size to 128

 .../net/ethernet/xilinx/xilinx_axienet_main.c | 135 +++++++++++-------
 1 file changed, 84 insertions(+), 51 deletions(-)
Signature · 1 line
2.31.1
#1 lore raw

PATCH v2 1/9 net: axienet: increase reset timeout

The previous timeout of 1ms was too short to handle some cases where the core is reset just after the input clocks were started, which will be introduced in an upcoming patch. Increase the timeout to 50ms. Also simplify the reset timeout checking to use read_poll_timeout.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 .../net/ethernet/xilinx/xilinx_axienet_main.c | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +9 -10
index 90144ac7aee8..f950342f6467 100644
@@ -496,7 +496,8 @@ static void axienet_setoptions(struct net_device *ndev, u32 options)
 
 static int __axienet_device_reset(struct axienet_local *lp)
 {
-	u32 timeout;
+	u32 value;
+	int ret;
 
 	/* Reset Axi DMA. This would reset Axi Ethernet core as well. The reset
 	 * process of Axi DMA takes a while to complete as all pending
@@ -506,15 +507,13 @@ static int __axienet_device_reset(struct axienet_local *lp)
 	 * they both reset the entire DMA core, so only one needs to be used.
 	 */
 	axienet_dma_out32(lp, XAXIDMA_TX_CR_OFFSET, XAXIDMA_CR_RESET_MASK);
-	timeout = DELAY_OF_ONE_MILLISEC;
-	while (axienet_dma_in32(lp, XAXIDMA_TX_CR_OFFSET) &
-				XAXIDMA_CR_RESET_MASK) {
-		udelay(1);
-		if (--timeout == 0) {
-			netdev_err(lp->ndev, "%s: DMA reset timeout!\n",
-				   __func__);
-			return -ETIMEDOUT;
-		}
+	ret = read_poll_timeout(axienet_dma_in32, value,
+				!(value & XAXIDMA_CR_RESET_MASK),
+				DELAY_OF_ONE_MILLISEC, 50000, false, lp,
+				XAXIDMA_TX_CR_OFFSET);
+	if (ret) {
+		dev_err(lp->dev, "%s: DMA reset timeout!\n", __func__);
+		return ret;
 	}
 
 	return 0;
Signature · 1 line
2.31.1
Reviewed-by on v2 1/9
#2 lore raw

Re: [PATCH net v2 1/9] net: axienet: increase reset timeout

Quoted 7 lines · Robert Hancock

The previous timeout of 1ms was too short to handle some cases where the core is reset just after the input clocks were started, which will be introduced in an upcoming patch. Increase the timeout to 50ms. Also simplify the reset timeout checking to use read_poll_timeout.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")

Andrew

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#3 lore raw

PATCH v2 2/9 net: axienet: Wait for PhyRstCmplt after core reset

In reply to Robert Hancock

When resetting the device, wait for the PhyRstCmplt bit to be set in the interrupt status register before continuing initialization, to ensure that the core is actually ready. The MgtRdy bit could also be waited for, but unfortunately when using 7-series devices, the bit does not appear to work as documented (it seems to behave as some sort of link state indication and not just an indication the transceiver is ready) so it can't really be relied on.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +10 -0
index f950342f6467..f425a8404a9b 100644
@@ -516,6 +516,16 @@ static int __axienet_device_reset(struct axienet_local *lp)
 		return ret;
 	}
 
+	/* Wait for PhyRstCmplt bit to be set, indicating the PHY reset has finished */
+	ret = read_poll_timeout(axienet_ior, value,
+				value & XAE_INT_PHYRSTCMPLT_MASK,
+				DELAY_OF_ONE_MILLISEC, 50000, false, lp,
+				XAE_IS_OFFSET);
+	if (ret) {
+		dev_err(lp->dev, "%s: timeout waiting for PhyRstCmplt\n", __func__);
+		return ret;
+	}
+
 	return 0;
 }
 
Signature · 1 line
2.31.1
#4 lore raw

Re: [PATCH net v2 2/9] net: axienet: Wait for PhyRstCmplt after core reset

Quoted 22 lines · Robert Hancock

When resetting the device, wait for the PhyRstCmplt bit to be set in the interrupt status register before continuing initialization, to ensure that the core is actually ready. The MgtRdy bit could also be waited for, but unfortunately when using 7-series devices, the bit does not appear to work as documented (it seems to behave as some sort of link state indication and not just an indication the transceiver is ready) so it can't really be relied on.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
index f950342f6467..f425a8404a9b 100644
@@ -516,6 +516,16 @@ static int __axienet_device_reset(struct axienet_local *lp)
 		return ret;
 	}
 

drivers/net/ethernet/xilinx/xilinx_axienet_main.c static int __axienet_device_reset(struct axienet_local *lp) · line 528

Quoted 10 lines
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +10 -0
+	/* Wait for PhyRstCmplt bit to be set, indicating the PHY reset has finished */
+	ret = read_poll_timeout(axienet_ior, value,
+				value & XAE_INT_PHYRSTCMPLT_MASK,
+				DELAY_OF_ONE_MILLISEC, 50000, false, lp,
+				XAE_IS_OFFSET);
+	if (ret) {
+		dev_err(lp->dev, "%s: timeout waiting for PhyRstCmplt\n", __func__);
+		return ret;
+	}
+

Is this bit guaranteed to be clear before you start waiting for it?

Andrew

#5 lore raw

Re: [PATCH net v2 2/9] net: axienet: Wait for PhyRstCmplt after core reset

Quoted 26 lines · Andrew Lunn
Quoted 25 lines · Robert Hancock

When resetting the device, wait for the PhyRstCmplt bit to be set in the interrupt status register before continuing initialization, to ensure that the core is actually ready. The MgtRdy bit could also be waited for, but unfortunately when using 7-series devices, the bit does not appear to work as documented (it seems to behave as some sort of link state indication and not just an indication the transceiver is ready) so it can't really be relied on.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI

Ethernet driver")

---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
index f950342f6467..f425a8404a9b 100644
@@ -516,6 +516,16 @@ static int __axienet_device_reset(struct axienet_local
 *lp)
 		return ret;
 	}
 
Quoted 14 lines

drivers/net/ethernet/xilinx/xilinx_axienet_main.c static int __axienet_device_reset(struct axienet_local *lp) · line 528

Quoted 12 lines
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +1 -0
+	/* Wait for PhyRstCmplt bit to be set, indicating the PHY reset has

finished */

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +6 -0
+	ret = read_poll_timeout(axienet_ior, value,
+				value & XAE_INT_PHYRSTCMPLT_MASK,
+				DELAY_OF_ONE_MILLISEC, 50000, false, lp,
+				XAE_IS_OFFSET);
+	if (ret) {
+		dev_err(lp->dev, "%s: timeout waiting for PhyRstCmplt\n",

__func__);

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +3 -0
+		return ret;
+	}
+

Is this bit guaranteed to be clear before you start waiting for it?

The documentation for the IP core (
https://www.xilinx.com/content/dam/xilinx/support/documentation/ip_documentation/axi_ethernet/v7_2/pg138-axi-ethernet.pdf
) states for the phy_rst_n output signal: "This active-Low reset is held
active for 10 ms after power is applied and during any reset. After the reset goes inactive, the PHY cannot be accessed for an additional 5 ms." The PhyRstComplt bit definition mentions "This signal does not transition to 1 for 5 ms after PHY_RST_N transitions to 1". Given that a reset of the core has just been completed above, the PHY reset should at least have been initiated as well, so it should be sufficient to just wait for the bit to become 1 at this point.

Quoted 2 lines

Andrew

Reviewed-by
#6 lore raw

Re: [PATCH net v2 2/9] net: axienet: Wait for PhyRstCmplt after core reset

Quoted 12 lines
Quoted 1 line

Is this bit guaranteed to be clear before you start waiting for it?

The documentation for the IP core (
https://www.xilinx.com/content/dam/xilinx/support/documentation/ip_documentation/axi_ethernet/v7_2/pg138-axi-ethernet.pdf
) states for the phy_rst_n output signal: "This active-Low reset is held
active for 10 ms after power is applied and during any reset. After the reset goes inactive, the PHY cannot be accessed for an additional 5 ms." The PhyRstComplt bit definition mentions "This signal does not transition to 1 for 5 ms after PHY_RST_N transitions to 1". Given that a reset of the core has just been completed above, the PHY reset should at least have been initiated as well, so it should be sufficient to just wait for the bit to become 1 at this point.

Great, thanks for checking.

Andrew

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#7 lore raw

RE: [PATCH net v2 2/9] net: axienet: Wait for PhyRstCmplt after core reset

In reply to Robert Hancock
Quoted 10 lines

-----Original Message-----
From: Robert Hancock <[email protected]>
Sent: Wednesday, January 12, 2022 11:07 PM
To: [email protected]
Cc: Radhey Shyam Pandey <[email protected]>; [email protected]; [email protected]; [email protected]; Michal Simek <[email protected]>; [email protected]; [email protected]; Robert Hancock <[email protected]>
Subject: [PATCH net v2 2/9] net: axienet: Wait for PhyRstCmplt after core reset

Quoted 4 lines

When resetting the device, wait for the PhyRstCmplt bit to be set in the interrupt status register before continuing initialization, to ensure that the core is actually ready. The MgtRdy bit could also be waited for, but unfortunately when using 7-series devices, the bit does

Just to understand - can you share 7- series design details. Based on documentation - This MgtRdy bit indicates if the TEMAC core is out of reset and ready for use. In systems that use an serial transceiver, this bit goes to 1 when the serial transceiver is ready to use.

Also if we don't wait for phy reset - what is the issue we are seeing?

Quoted 38 lines

not appear to work as documented (it seems to behave as some sort of link state indication and not just an indication the transceiver is ready) so it can't really be relied on.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet

driver")

---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +1 -0
index f950342f6467..f425a8404a9b 100644
@@ -516,6 +516,16 @@ static int __axienet_device_reset(struct axienet_local
 *lp)
 		return ret;
 	}
 
+	/* Wait for PhyRstCmplt bit to be set, indicating the PHY reset has

finished */

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +6 -0
+	ret = read_poll_timeout(axienet_ior, value,
+				value & XAE_INT_PHYRSTCMPLT_MASK,
+				DELAY_OF_ONE_MILLISEC, 50000, false, lp,
+				XAE_IS_OFFSET);
+	if (ret) {
+		dev_err(lp->dev, "%s: timeout waiting for PhyRstCmplt\n",

__func__);

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +3 -1
+		return ret;
+	}
+
 	return 0;
 }
 
--

2.31.1

#8 lore raw

Re: [PATCH net v2 2/9] net: axienet: Wait for PhyRstCmplt after core reset

Quoted 16 lines · Radhey Shyam Pandey
Quoted 11 lines

-----Original Message-----
From: Robert Hancock <[email protected]>
Sent: Wednesday, January 12, 2022 11:07 PM
To: [email protected]
Cc: Radhey Shyam Pandey <[email protected]>; [email protected]; [email protected]; [email protected]; Michal Simek <[email protected]>; [email protected]; [email protected]; Robert Hancock <[email protected]>
Subject: [PATCH net v2 2/9] net: axienet: Wait for PhyRstCmplt after core reset

Quoted 4 lines

When resetting the device, wait for the PhyRstCmplt bit to be set in the interrupt status register before continuing initialization, to ensure that the core is actually ready. The MgtRdy bit could also be waited for, but unfortunately when using 7-series devices, the bit does

Quoted 4 lines

Just to understand - can you share 7- series design details. Based on documentation - This MgtRdy bit indicates if the TEMAC core is out of reset and ready for use. In systems that use an serial transceiver, this bit goes to 1 when the serial transceiver is ready to use.

From what I saw, the bit behaved as described on ZynqMP where it would go to 1 during initialization, but on a Kintex-7 design with this core, the bit never seemed to go to 1 until an actual link was established on the transceiver, so it wasn't really usable in this situation.

Quoted 2 lines

Also if we don't wait for phy reset - what is the issue we are seeing?

This is more for the case of using an external PHY device - we shouldn't be proceeding to MDIO initialization and potentially trying to talk to the PHY when it is potentially still in reset..

Quoted 40 lines
Quoted 40 lines

not appear to work as documented (it seems to behave as some sort of link state indication and not just an indication the transceiver is ready) so it can't really be relied on.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI

Ethernet
driver")

---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +1 -0
index f950342f6467..f425a8404a9b 100644
@@ -516,6 +516,16 @@ static int __axienet_device_reset(struct axienet_local
 *lp)
 		return ret;
 	}
 
+	/* Wait for PhyRstCmplt bit to be set, indicating the PHY reset has

finished */

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +6 -0
+	ret = read_poll_timeout(axienet_ior, value,
+				value & XAE_INT_PHYRSTCMPLT_MASK,
+				DELAY_OF_ONE_MILLISEC, 50000, false, lp,
+				XAE_IS_OFFSET);
+	if (ret) {
+		dev_err(lp->dev, "%s: timeout waiting for PhyRstCmplt\n",

__func__);

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +3 -1
+		return ret;
+	}
+
 	return 0;
 }
 
--

2.31.1

Signature · 3 lines
Robert Hancock
Senior Hardware Designer, Calian Advanced Technologies
www.calian.com
#9 lore raw

PATCH v2 3/9 net: axienet: reset core on initialization prior to MDIO access

In reply to Robert Hancock

In some cases where the Xilinx Ethernet core was used in 1000Base-X or SGMII modes, which use the internal PCS/PMA PHY, and the MGT transceiver clock source for the PCS was not running at the time the FPGA logic was loaded, the core would come up in a state where the PCS could not be found on the MDIO bus. To fix this, the Ethernet core (including the PCS) should be reset after enabling the clocks, prior to attempting to access the PCS using of_mdio_find_device.

Fixes 1a02556086fc (net: axienet: Properly handle PCS/PMA PHY for 1000BaseX mode)
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +5 -0
index f425a8404a9b..f4ae035bed35 100644
@@ -2055,6 +2055,11 @@ static int axienet_probe(struct platform_device *pdev)
 	lp->coalesce_count_rx = XAXIDMA_DFT_RX_THRESHOLD;
 	lp->coalesce_count_tx = XAXIDMA_DFT_TX_THRESHOLD;
 
+	/* Reset core now that clocks are enabled, prior to accessing MDIO */
+	ret = __axienet_device_reset(lp);
+	if (ret)
+		goto cleanup_clk;
+
 	lp->phy_node = of_parse_phandle(pdev->dev.of_node, "phy-handle", 0);
 	if (lp->phy_node) {
 		ret = axienet_mdio_setup(lp);
Signature · 1 line
2.31.1
Reviewed-by on v2 3/9
#10 lore raw

Re: [PATCH net v2 3/9] net: axienet: reset core on initialization prior to MDIO access

Quoted 10 lines · Robert Hancock

In some cases where the Xilinx Ethernet core was used in 1000Base-X or SGMII modes, which use the internal PCS/PMA PHY, and the MGT transceiver clock source for the PCS was not running at the time the FPGA logic was loaded, the core would come up in a state where the PCS could not be found on the MDIO bus. To fix this, the Ethernet core (including the PCS) should be reset after enabling the clocks, prior to attempting to access the PCS using of_mdio_find_device.

Fixes 1a02556086fc (net: axienet: Properly handle PCS/PMA PHY for 1000BaseX mode)

Andrew

#11 lore raw

PATCH v2 4/9 net: axienet: add missing memory barriers

In reply to Robert Hancock

This driver was missing some required memory barriers:

Use dma_rmb to ensure we see all updates to the descriptor after we see that an entry has been completed.

Use wmb and rmb to avoid stale descriptor status between the TX path and TX complete IRQ path.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +10 -1
index f4ae035bed35..de8f85175a6c 100644
@@ -632,6 +632,8 @@ static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd,
 		if (nr_bds == -1 && !(status & XAXIDMA_BD_STS_COMPLETE_MASK))
 			break;
 
+		/* Ensure we see complete descriptor update */
+		dma_rmb();
 		phys = desc_get_phys_addr(lp, cur_p);
 		dma_unmap_single(ndev->dev.parent, phys,
 				 (cur_p->cntrl & XAXIDMA_BD_CTRL_LENGTH_MASK),
@@ -645,8 +647,10 @@ static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd,
 		cur_p->app1 = 0;
 		cur_p->app2 = 0;
 		cur_p->app4 = 0;
-		cur_p->status = 0;
 		cur_p->skb = NULL;
+		/* ensure our transmit path and device don't prematurely see status cleared */
+		wmb();
+		cur_p->status = 0;
 
 		if (sizep)
 			*sizep += status & XAXIDMA_BD_STS_ACTUAL_LEN_MASK;
@@ -704,6 +708,9 @@ static inline int axienet_check_tx_bd_space(struct axienet_local *lp,
 					    int num_frag)
 {
 	struct axidma_bd *cur_p;
+
+	/* Ensure we see all descriptor updates from device or TX IRQ path */
+	rmb();
 	cur_p = &lp->tx_bd_v[(lp->tx_bd_tail + num_frag) % lp->tx_bd_num];
 	if (cur_p->status & XAXIDMA_BD_STS_ALL_MASK)
 		return NETDEV_TX_BUSY;
@@ -843,6 +850,8 @@ static void axienet_recv(struct net_device *ndev)
 
 		tail_p = lp->rx_bd_p + sizeof(*lp->rx_bd_v) * lp->rx_bd_ci;
 
+		/* Ensure we see complete descriptor update */
+		dma_rmb();
 		phys = desc_get_phys_addr(lp, cur_p);
 		dma_unmap_single(ndev->dev.parent, phys, lp->max_frm_size,
 				 DMA_FROM_DEVICE);
Signature · 1 line
2.31.1
#12 lore raw

RE: [PATCH net v2 4/9] net: axienet: add missing memory barriers

Quoted 50 lines

-----Original Message-----
From: Robert Hancock <[email protected]>
Sent: Wednesday, January 12, 2022 11:07 PM
To: [email protected]
Cc: Radhey Shyam Pandey <[email protected]>; [email protected]; [email protected]; [email protected]; Michal Simek <[email protected]>; [email protected]; [email protected]; Robert Hancock <[email protected]>
Subject: [PATCH net v2 4/9] net: axienet: add missing memory barriers

This driver was missing some required memory barriers:

Use dma_rmb to ensure we see all updates to the descriptor after we see that an entry has been completed.

Use wmb and rmb to avoid stale descriptor status between the TX path and TX complete IRQ path.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet

driver")

---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +2 -0
index f4ae035bed35..de8f85175a6c 100644
@@ -632,6 +632,8 @@ static int axienet_free_tx_chain(struct net_device
 *ndev, u32 first_bd,
 		if (nr_bds == -1 && !(status &
 XAXIDMA_BD_STS_COMPLETE_MASK))
 			break;
 
+		/* Ensure we see complete descriptor update */
+		dma_rmb();
 		phys = desc_get_phys_addr(lp, cur_p);
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
 		dma_unmap_single(ndev->dev.parent, phys,
 				 (cur_p->cntrl &

XAXIDMA_BD_CTRL_LENGTH_MASK), @@ -645,8 +647,10 @@ static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd,

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +1 -1
 		cur_p->app1 = 0;
 		cur_p->app2 = 0;
 		cur_p->app4 = 0;
-		cur_p->status = 0;
 		cur_p->skb = NULL;
+		/* ensure our transmit path and device don't prematurely see

status cleared */

drivers/net/ethernet/xilinx/xilinx_axienet_main.c static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd, · line 653

Quoted 2 lines
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +2 -0
+		wmb();
+		cur_p->status = 0;

Any reason for moving status initialization down?

Quoted 23 lines
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
 		if (sizep)
 			*sizep += status &

XAXIDMA_BD_STS_ACTUAL_LEN_MASK; @@ -704,6 +708,9 @@ static inline int axienet_check_tx_bd_space(struct axienet_local *lp,

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +5 -0
 					    int num_frag)
 {
 	struct axidma_bd *cur_p;
+
+	/* Ensure we see all descriptor updates from device or TX IRQ path */
+	rmb();
 	cur_p = &lp->tx_bd_v[(lp->tx_bd_tail + num_frag) % lp->tx_bd_num];
 	if (cur_p->status & XAXIDMA_BD_STS_ALL_MASK)
 		return NETDEV_TX_BUSY;
@@ -843,6 +850,8 @@ static void axienet_recv(struct net_device *ndev)
 
 		tail_p = lp->rx_bd_p + sizeof(*lp->rx_bd_v) * lp->rx_bd_ci;
 
+		/* Ensure we see complete descriptor update */
+		dma_rmb();
 		phys = desc_get_phys_addr(lp, cur_p);
 		dma_unmap_single(ndev->dev.parent, phys, lp-
 >max_frm_size,
Quoted 1 line

DMA_FROM_DEVICE);

Ideally we would also need a write barrier in xmit function just before updating tail descriptor.

Quoted 2 lines

--
2.31.1

#13 lore raw

Re: [PATCH net v2 4/9] net: axienet: add missing memory barriers

Quoted 48 lines · Radhey Shyam Pandey
Quoted 48 lines

-----Original Message-----
From: Robert Hancock <[email protected]>
Sent: Wednesday, January 12, 2022 11:07 PM
To: [email protected]
Cc: Radhey Shyam Pandey <[email protected]>; [email protected]; [email protected]; [email protected]; Michal Simek <[email protected]>; [email protected]; [email protected]; Robert Hancock <[email protected]>
Subject: [PATCH net v2 4/9] net: axienet: add missing memory barriers

This driver was missing some required memory barriers:

Use dma_rmb to ensure we see all updates to the descriptor after we see that
an entry has been completed.

Use wmb and rmb to avoid stale descriptor status between the TX path and TX complete IRQ path.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI

Ethernet
driver")

---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +2 -0
index f4ae035bed35..de8f85175a6c 100644
@@ -632,6 +632,8 @@ static int axienet_free_tx_chain(struct net_device
 *ndev, u32 first_bd,
 		if (nr_bds == -1 && !(status &
 XAXIDMA_BD_STS_COMPLETE_MASK))
 			break;
 
+		/* Ensure we see complete descriptor update */
+		dma_rmb();
 		phys = desc_get_phys_addr(lp, cur_p);
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
 		dma_unmap_single(ndev->dev.parent, phys,
 				 (cur_p->cntrl &

XAXIDMA_BD_CTRL_LENGTH_MASK), @@ -645,8 +647,10 @@ static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd,

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
 		cur_p->app1 = 0;
 		cur_p->app2 = 0;
 		cur_p->app4 = 0;
Quoted 8 lines

drivers/net/ethernet/xilinx/xilinx_axienet_main.c static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd, · line 653

Quoted 6 lines
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +1 -1
-		cur_p->status = 0;
 		cur_p->skb = NULL;
+		/* ensure our transmit path and device don't prematurely see

status cleared */

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +2 -0
+		wmb();
+		cur_p->status = 0;

Any reason for moving status initialization down?

Probably not strictly necessary, but the idea was to ensure that any of the other writes to the descriptor were visible before the device saw the status being cleared (indicating it is available to be read by the device).

Quoted 22 lines
Quoted 22 lines
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
 		if (sizep)
 			*sizep += status &

XAXIDMA_BD_STS_ACTUAL_LEN_MASK; @@ -704,6 +708,9 @@ static inline int axienet_check_tx_bd_space(struct axienet_local *lp,

drivers/net/ethernet/xilinx/xilinx_axienet_main.c +5 -0
 					    int num_frag)
 {
 	struct axidma_bd *cur_p;
+
+	/* Ensure we see all descriptor updates from device or TX IRQ path */
+	rmb();
 	cur_p = &lp->tx_bd_v[(lp->tx_bd_tail + num_frag) % lp->tx_bd_num];
 	if (cur_p->status & XAXIDMA_BD_STS_ALL_MASK)
 		return NETDEV_TX_BUSY;
@@ -843,6 +850,8 @@ static void axienet_recv(struct net_device *ndev)
 
 		tail_p = lp->rx_bd_p + sizeof(*lp->rx_bd_v) * lp->rx_bd_ci;
 
+		/* Ensure we see complete descriptor update */
+		dma_rmb();
 		phys = desc_get_phys_addr(lp, cur_p);
 		dma_unmap_single(ndev->dev.parent, phys, lp-
Quoted 5 lines
Quoted 2 lines
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -0
 > max_frm_size,

DMA_FROM_DEVICE);

Ideally we would also need a write barrier in xmit function just before updating tail descriptor.

I don't think it should be needed there because there is an implicit barrier on the MMIO write.

Quoted 3 lines
Quoted 3 lines

--
2.31.1

Signature · 3 lines
Robert Hancock
Senior Hardware Designer, Calian Advanced Technologies
www.calian.com
#14 lore raw

PATCH v2 5/9 net: axienet: limit minimum TX ring size

In reply to Robert Hancock

The driver will not work properly if the TX ring size is set to below MAX_SKB_FRAGS + 1 since it needs to hold at least one full maximally fragmented packet in the TX ring. Limit setting the ring size to below this value.

Fixes 8b09ca823ffb4 ("net: axienet: Make RX/TX ring sizes configurable")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +3 -1
index de8f85175a6c..8a60219d3bfb 100644
@@ -43,6 +43,7 @@
 /* Descriptors defines for Tx and Rx DMA */
 #define TX_BD_NUM_DEFAULT		64
 #define RX_BD_NUM_DEFAULT		1024
+#define TX_BD_NUM_MIN			(MAX_SKB_FRAGS + 1)
 #define TX_BD_NUM_MAX			4096
 #define RX_BD_NUM_MAX			4096
 
@@ -1389,7 +1390,8 @@ axienet_ethtools_set_ringparam(struct net_device *ndev,
 	if (ering->rx_pending > RX_BD_NUM_MAX ||
 	    ering->rx_mini_pending ||
 	    ering->rx_jumbo_pending ||
-	    ering->rx_pending > TX_BD_NUM_MAX)
+	    ering->tx_pending < TX_BD_NUM_MIN ||
+	    ering->tx_pending > TX_BD_NUM_MAX)
 		return -EINVAL;
 
 	if (netif_running(ndev))
Signature · 1 line
2.31.1
#15 lore raw

PATCH v2 6/9 net: axienet: Fix TX ring slot available check

In reply to Robert Hancock

The check for whether a TX ring slot was available was incorrect, since a slot which had been loaded with transmit data but the device had not started transmitting would be treated as available, potentially causing non-transmitted slots to be overwritten. The control field in the descriptor should be checked, rather than the status field (which may only be updated when the device completes the entry).

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +2 -2
index 8a60219d3bfb..ee8d656200b8 100644
@@ -643,7 +643,6 @@ static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd,
 		if (cur_p->skb && (status & XAXIDMA_BD_STS_COMPLETE_MASK))
 			dev_consume_skb_irq(cur_p->skb);
 
-		cur_p->cntrl = 0;
 		cur_p->app0 = 0;
 		cur_p->app1 = 0;
 		cur_p->app2 = 0;
@@ -651,6 +650,7 @@ static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd,
 		cur_p->skb = NULL;
 		/* ensure our transmit path and device don't prematurely see status cleared */
 		wmb();
+		cur_p->cntrl = 0;
 		cur_p->status = 0;
 
 		if (sizep)
@@ -713,7 +713,7 @@ static inline int axienet_check_tx_bd_space(struct axienet_local *lp,
 	/* Ensure we see all descriptor updates from device or TX IRQ path */
 	rmb();
 	cur_p = &lp->tx_bd_v[(lp->tx_bd_tail + num_frag) % lp->tx_bd_num];
-	if (cur_p->status & XAXIDMA_BD_STS_ALL_MASK)
+	if (cur_p->cntrl)
 		return NETDEV_TX_BUSY;
 	return 0;
 }
Signature · 1 line
2.31.1
#16 lore raw

PATCH v2 7/9 net: axienet: fix number of TX ring slots for available check

In reply to Robert Hancock

The check for the number of available TX ring slots was off by 1 since a slot is required for the skb header as well as each fragment. This could result in overwriting a TX ring slot that was still in use.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +2 -2
index ee8d656200b8..c5d214abd4d5 100644
@@ -747,7 +747,7 @@ axienet_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	num_frag = skb_shinfo(skb)->nr_frags;
 	cur_p = &lp->tx_bd_v[lp->tx_bd_tail];
 
-	if (axienet_check_tx_bd_space(lp, num_frag)) {
+	if (axienet_check_tx_bd_space(lp, num_frag + 1)) {
 		if (netif_queue_stopped(ndev))
 			return NETDEV_TX_BUSY;
 
@@ -757,7 +757,7 @@ axienet_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 		smp_mb();
 
 		/* Space might have just been freed - check again */
-		if (axienet_check_tx_bd_space(lp, num_frag))
+		if (axienet_check_tx_bd_space(lp, num_frag + 1))
 			return NETDEV_TX_BUSY;
 
 		netif_wake_queue(ndev);
Signature · 1 line
2.31.1
#17 lore raw

PATCH v2 8/9 net: axienet: fix for TX busy handling

In reply to Robert Hancock

Network driver documentation indicates we should be avoiding returning NETDEV_TX_BUSY from ndo_start_xmit in normal cases, since it requires the packets to be requeued. Instead the queue should be stopped after a packet is added to the TX ring when there may not be enough room for an additional one. Also, when TX ring entries are completed, we should only wake the queue if we know there is room for another full maximally fragmented packet.

Print a warning if there is insufficient space at the start of start_xmit, since this should no longer happen.

Combined with increasing the default TX ring size (in a subsequent patch), this appears to recover the TX performance lost by previous changes to actually manage the TX ring state properly.

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 .../net/ethernet/xilinx/xilinx_axienet_main.c | 86 ++++++++++---------
 1 file changed, 47 insertions(+), 39 deletions(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +47 -39
index c5d214abd4d5..8ac277ef1f99 100644
@@ -660,6 +660,32 @@ static int axienet_free_tx_chain(struct net_device *ndev, u32 first_bd,
 	return i;
 }
 
+/**
+ * axienet_check_tx_bd_space - Checks if a BD/group of BDs are currently busy
+ * @lp:		Pointer to the axienet_local structure
+ * @num_frag:	The number of BDs to check for
+ *
+ * Return: 0, on success
+ *	    NETDEV_TX_BUSY, if any of the descriptors are not free
+ *
+ * This function is invoked before BDs are allocated and transmission starts.
+ * This function returns 0 if a BD or group of BDs can be allocated for
+ * transmission. If the BD or any of the BDs are not free the function
+ * returns a busy status. This is invoked from axienet_start_xmit.
+ */
+static inline int axienet_check_tx_bd_space(struct axienet_local *lp,
+					    int num_frag)
+{
+	struct axidma_bd *cur_p;
+
+	/* Ensure we see all descriptor updates from device or TX IRQ path */
+	rmb();
+	cur_p = &lp->tx_bd_v[(lp->tx_bd_tail + num_frag) % lp->tx_bd_num];
+	if (cur_p->cntrl)
+		return NETDEV_TX_BUSY;
+	return 0;
+}
+
 /**
  * axienet_start_xmit_done - Invoked once a transmit is completed by the
  * Axi DMA Tx channel.
@@ -689,33 +715,8 @@ static void axienet_start_xmit_done(struct net_device *ndev)
 	/* Matches barrier in axienet_start_xmit */
 	smp_mb();
 
-	netif_wake_queue(ndev);
-}
-
-/**
- * axienet_check_tx_bd_space - Checks if a BD/group of BDs are currently busy
- * @lp:		Pointer to the axienet_local structure
- * @num_frag:	The number of BDs to check for
- *
- * Return: 0, on success
- *	    NETDEV_TX_BUSY, if any of the descriptors are not free
- *
- * This function is invoked before BDs are allocated and transmission starts.
- * This function returns 0 if a BD or group of BDs can be allocated for
- * transmission. If the BD or any of the BDs are not free the function
- * returns a busy status. This is invoked from axienet_start_xmit.
- */
-static inline int axienet_check_tx_bd_space(struct axienet_local *lp,
-					    int num_frag)
-{
-	struct axidma_bd *cur_p;
-
-	/* Ensure we see all descriptor updates from device or TX IRQ path */
-	rmb();
-	cur_p = &lp->tx_bd_v[(lp->tx_bd_tail + num_frag) % lp->tx_bd_num];
-	if (cur_p->cntrl)
-		return NETDEV_TX_BUSY;
-	return 0;
+	if (!axienet_check_tx_bd_space(lp, MAX_SKB_FRAGS + 1))
+		netif_wake_queue(ndev);
 }
 
 /**
@@ -748,19 +749,14 @@ axienet_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	cur_p = &lp->tx_bd_v[lp->tx_bd_tail];
 
 	if (axienet_check_tx_bd_space(lp, num_frag + 1)) {
-		if (netif_queue_stopped(ndev))
-			return NETDEV_TX_BUSY;
-
+		/* Should not happen as last start_xmit call should have
+		 * checked for sufficient space and queue should only be
+		 * woken when sufficient space is available.
+		 */
 		netif_stop_queue(ndev);
-
-		/* Matches barrier in axienet_start_xmit_done */
-		smp_mb();
-
-		/* Space might have just been freed - check again */
-		if (axienet_check_tx_bd_space(lp, num_frag + 1))
-			return NETDEV_TX_BUSY;
-
-		netif_wake_queue(ndev);
+		if (net_ratelimit())
+			netdev_warn(ndev, "TX ring unexpectedly full\n");
+		return NETDEV_TX_BUSY;
 	}
 
 	if (skb->ip_summed == CHECKSUM_PARTIAL) {
@@ -821,6 +817,18 @@ axienet_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	if (++lp->tx_bd_tail >= lp->tx_bd_num)
 		lp->tx_bd_tail = 0;
 
+	/* Stop queue if next transmit may not have space */
+	if (axienet_check_tx_bd_space(lp, MAX_SKB_FRAGS + 1)) {
+		netif_stop_queue(ndev);
+
+		/* Matches barrier in axienet_start_xmit_done */
+		smp_mb();
+
+		/* Space might have just been freed - check again */
+		if (!axienet_check_tx_bd_space(lp, MAX_SKB_FRAGS + 1))
+			netif_wake_queue(ndev);
+	}
+
 	return NETDEV_TX_OK;
 }
 
Signature · 1 line
2.31.1
#18 lore raw

PATCH v2 9/9 net: axienet: increase default TX ring size to 128

In reply to Robert Hancock

With previous changes to make the driver handle the TX ring size more correctly, the default TX ring size of 64 appears to significantly bottleneck TX performance to around 600 Mbps on a 1 Gbps link on ZynqMP. Increasing this to 128 seems to bring performance up to near line rate and shouldn't cause excess bufferbloat (this driver doesn't yet support modern byte-based queue management).

Fixes 8a3b7a252dca9 ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +1 -1
index 8ac277ef1f99..d5ea093fbd32 100644
@@ -41,7 +41,7 @@
 #include "xilinx_axienet.h"
 
 /* Descriptors defines for Tx and Rx DMA */
-#define TX_BD_NUM_DEFAULT		64
+#define TX_BD_NUM_DEFAULT		128
 #define RX_BD_NUM_DEFAULT		1024
 #define TX_BD_NUM_MIN			(MAX_SKB_FRAGS + 1)
 #define TX_BD_NUM_MAX			4096
Signature · 1 line
2.31.1
#19 lore raw
In reply to Robert Hancock
Quoted 18 lines · Robert Hancock

Various fixes for the Xilinx AXI Ethernet driver.

Changed since v1:

(quoted diff) +0 -3
-corrected a Fixes tag to point to mainline commit
-split up reset changes into 3 patches
-added ratelimit on netdev_warn in TX busy case

Robert Hancock (9):
net: axienet: increase reset timeout
net: axienet: Wait for PhyRstCmplt after core reset
net: axienet: reset core on initialization prior to MDIO access
net: axienet: add missing memory barriers
net: axienet: limit minimum TX ring size
net: axienet: Fix TX ring slot available check
net: axienet: fix number of TX ring slots for available check
net: axienet: fix for TX busy handling
net: axienet: increase default TX ring size to 128

Quoted 3 lines
 .../net/ethernet/xilinx/xilinx_axienet_main.c | 135 +++++++++++-------
 1 file changed, 84 insertions(+), 51 deletions(-)

Hi all,

Any other comments/reviews on this patch set? It's marked as Changes Requested in Patchwork, but I don't think I saw any discussions that ended up with any changes being asked for?

Signature · 3 lines
Robert Hancock
Senior Hardware Designer, Calian Advanced Technologies
www.calian.com
maintainer
#20 lore raw
Quoted 8 lines · On Tue, 18 Jan 2022 20:45:31 +0000 Robert Hancock wrote:
Quoted 7 lines · Robert Hancock

Various fixes for the Xilinx AXI Ethernet driver.

Changed since v1:

(quoted diff) +0 -3
-corrected a Fixes tag to point to mainline commit
-split up reset changes into 3 patches
-added ratelimit on netdev_warn in TX busy case
Quoted 14 lines
Quoted 10 lines

Robert Hancock (9):
net: axienet: increase reset timeout
net: axienet: Wait for PhyRstCmplt after core reset
net: axienet: reset core on initialization prior to MDIO access
net: axienet: add missing memory barriers
net: axienet: limit minimum TX ring size
net: axienet: Fix TX ring slot available check
net: axienet: fix number of TX ring slots for available check
net: axienet: fix for TX busy handling
net: axienet: increase default TX ring size to 128

Any other comments/reviews on this patch set? It's marked as Changes Requested in Patchwork, but I don't think I saw any discussions that ended up with any changes being asked for?

Perhaps it was done in anticipation to follow up to Radhey's or Andrew's question but seems like you answered those. Or maybe because of the missing CC on of [email protected] on patch 5? Not sure.

Could you fold some of the explanations into commit messages, add Andrew's Acks and post a v3?

We could probably apply as is but since it was marked as Changes Requested I can't be sure someone hasn't stopped reviewing in anticipation of v3.

Thanks!

#21 lore raw
Quoted 26 lines · Jakub Kicinski
Quoted 8 lines · On Tue, 18 Jan 2022 20:45:31 +0000 Robert Hancock wrote:
Quoted 7 lines · Robert Hancock

Various fixes for the Xilinx AXI Ethernet driver.

Changed since v1:

(quoted diff) +0 -3
-corrected a Fixes tag to point to mainline commit
-split up reset changes into 3 patches
-added ratelimit on netdev_warn in TX busy case
Quoted 16 lines
Quoted 10 lines

Robert Hancock (9):
net: axienet: increase reset timeout
net: axienet: Wait for PhyRstCmplt after core reset
net: axienet: reset core on initialization prior to MDIO access
net: axienet: add missing memory barriers
net: axienet: limit minimum TX ring size
net: axienet: Fix TX ring slot available check
net: axienet: fix number of TX ring slots for available check
net: axienet: fix for TX busy handling
net: axienet: increase default TX ring size to 128

Any other comments/reviews on this patch set? It's marked as Changes Requested
in Patchwork, but I don't think I saw any discussions that ended up with any
changes being asked for?

Quoted 3 lines

Perhaps it was done in anticipation to follow up to Radhey's or Andrew's question but seems like you answered those. Or maybe because of the missing CC on of [email protected] on patch 5?

That's actually my old email address (before the renaming from SED Systems to Calian AT). I think get_maintainer.pl picks that up as I had some past commits to the driver with that email.

Quoted 8 lines

Not sure.

Could you fold some of the explanations into commit messages, add Andrew's Acks and post a v3?

We could probably apply as is but since it was marked as Changes Requested I can't be sure someone hasn't stopped reviewing in anticipation of v3.

Can do.

Quoted 2 lines

Thanks!

Signature · 3 lines
Robert Hancock
Senior Hardware Designer, Calian Advanced Technologies
www.calian.com