Skip to contentSkip to conversation

net: macb: fix for fixed link, support for c45 mdio and 10G

PATCH Superseded Milind Parab 3 participants 14 messages Dec 9, 2019 – Dec 11, 2019 2 messages not in the archive

Subsystems (5)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

cover net: macb: fix for fixed link, support for c45 mdio and 10G

This patch series applies to Cadence Ethernet MAC Controller. The first patch in this series is related to the patch that converts the driver to phylink interface in net-next "net: macb: convert to phylink". Next patch is for adding support for C45 interface and the final patch adds 10G MAC support.

The patch sequences are detailed as below

1. 0001-net-macb-fix-for-fixed-link-mode
This patch fix the issue with fixed-link mode in macb phylink interface. In fixed-link we don't need to parse phandle because it's better handled by phylink_of_phy_connect()

2. 0002-net-macb-add-support-for-C45-MDIO-read-write
This patch is to support C45 interface to PHY. This upgrade is downward compatible. All versions of the MAC (old and new) using the GPL driver support both Clause 22 and Clause 45 operation. Whether the access is in Clause 22 or Clause 45 format depends on the data pattern written to the PHY management register.

3. 0003-net-macb-add-support-for-high-speed-interface This patch add support for 10G fixed mode.

Thanks,
Milind Parab

Milind Parab (3):
net: macb: fix for fixed-link mode
net: macb: add support for C45 MDIO read/write
net: macb: add support for high speed interface

 drivers/net/ethernet/cadence/macb.h      |  65 ++++++-
 drivers/net/ethernet/cadence/macb_main.c | 220 +++++++++++++++++++----
 2 files changed, 243 insertions(+), 42 deletions(-)
Signature · 1 line
2.17.1
#1 lore raw

PATCH 1/3 net: macb: fix for fixed-link mode

This patch fix the issue with fixed link. With fixed-link device opening fails due to macb_phylink_connect not handling fixed-link mode, in which case no MAC-PHY connection is needed and phylink_connect return success (0), however in current driver attempt is made to search and connect to PHY even for fixed-link.

---
 drivers/net/ethernet/cadence/macb_main.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +8 -9
index 9c767ee252ac..6b68ef34ab19 100644
@@ -615,17 +615,13 @@ static int macb_phylink_connect(struct macb *bp)
 {
 	struct net_device *dev = bp->dev;
 	struct phy_device *phydev;
+	struct device_node *dn = bp->pdev->dev.of_node;
 	int ret;
 
-	if (bp->pdev->dev.of_node &&
-	    of_parse_phandle(bp->pdev->dev.of_node, "phy-handle", 0)) {
-		ret = phylink_of_phy_connect(bp->phylink, bp->pdev->dev.of_node,
-					     0);
-		if (ret) {
-			netdev_err(dev, "Could not attach PHY (%d)\n", ret);
-			return ret;
-		}
-	} else {
+	if (dn)
+		ret = phylink_of_phy_connect(bp->phylink, dn, 0);
+
+	if (!dn || (ret && !of_parse_phandle(dn, "phy-handle", 0))) {
 		phydev = phy_find_first(bp->mii_bus);
 		if (!phydev) {
 			netdev_err(dev, "no PHY found\n");
@@ -638,6 +634,9 @@ static int macb_phylink_connect(struct macb *bp)
 			netdev_err(dev, "Could not attach to PHY (%d)\n", ret);
 			return ret;
 		}
+	} else if (ret) {
+		netdev_err(dev, "Could not attach PHY (%d)\n", ret);
+		return ret;
 	}
 
 	phylink_start(bp->phylink);
Signature · 1 line
2.17.1
#2 lore raw

Re: [PATCH 1/3] net: macb: fix for fixed-link mode

Quoted 26 lines · Milind Parab

This patch fix the issue with fixed link. With fixed-link device opening fails due to macb_phylink_connect not handling fixed-link mode, in which case no MAC-PHY connection is needed and phylink_connect return success (0), however in current driver attempt is made to search and connect to PHY even for fixed-link.

---
 drivers/net/ethernet/cadence/macb_main.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +1 -3
index 9c767ee252ac..6b68ef34ab19 100644
@@ -615,17 +615,13 @@ static int macb_phylink_connect(struct macb *bp)
 {
 	struct net_device *dev = bp->dev;
 	struct phy_device *phydev;
+	struct device_node *dn = bp->pdev->dev.of_node;
 	int ret;
 
-	if (bp->pdev->dev.of_node &&
-	    of_parse_phandle(bp->pdev->dev.of_node, "phy-handle", 0)) {
-		ret = phylink_of_phy_connect(bp->phylink, bp->pdev->dev.of_node,

drivers/net/ethernet/cadence/macb_main.c static int macb_phylink_connect(struct macb *bp) · line 624

Quoted 10 lines
drivers/net/ethernet/cadence/macb_main.c +4 -6
-					     0);
-		if (ret) {
-			netdev_err(dev, "Could not attach PHY (%d)\n", ret);
-			return ret;
-		}
-	} else {
+	if (dn)
+		ret = phylink_of_phy_connect(bp->phylink, dn, 0);
+
+	if (!dn || (ret && !of_parse_phandle(dn, "phy-handle", 0))) {

Hi,

If of_parse_phandle() returns non-null, the device_node it returns will have its reference count increased by one. That reference needs to be put.

I assume you're trying to determine whether phylink_of_phy_connect() failed because of a missing phy-handle rather than of_phy_attach() failing? Maybe those two failures ought to be distinguished by errno return value?

of_phy_attach() may fail due to of_phy_find_device() failing to find the PHY, or phy_attach_direct() failing. We could switch from using of_phy_attach(), to using of_phy_find_device() directly so we can then propagate phy_attach_direct()'s error code back, rather than losing it. That would then leave the case of of_phy_find_device() failure to be considered in terms of errno return value.

Quoted 17 lines
drivers/net/ethernet/cadence/macb_main.c +3 -0
 		phydev = phy_find_first(bp->mii_bus);
 		if (!phydev) {
 			netdev_err(dev, "no PHY found\n");
@@ -638,6 +634,9 @@ static int macb_phylink_connect(struct macb *bp)
 			netdev_err(dev, "Could not attach to PHY (%d)\n", ret);
 			return ret;
 		}
+	} else if (ret) {
+		netdev_err(dev, "Could not attach PHY (%d)\n", ret);
+		return ret;
 	}
 
 	phylink_start(bp->phylink);
Signature · 1 line
2.17.1
Signature · 3 lines
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
According to speedtest.net: 11.9Mbps down 500kbps up
#3 lore raw

RE: [PATCH 1/3] net: macb: fix for fixed-link mode

Quoted 34 lines
Quoted 34 lines

This patch fix the issue with fixed link. With fixed-link device opening fails due to macb_phylink_connect not handling fixed-link mode, in which case no MAC-PHY connection is needed and phylink_connect return success (0), however in current driver attempt is made to search and connect to PHY even for fixed-link.

---
 drivers/net/ethernet/cadence/macb_main.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +2 -9
index 9c767ee252ac..6b68ef34ab19 100644
@@ -615,17 +615,13 @@ static int macb_phylink_connect(struct macb *bp)
 {
 	struct net_device *dev = bp->dev;
 	struct phy_device *phydev;
+	struct device_node *dn = bp->pdev->dev.of_node;
 	int ret;
 
-	if (bp->pdev->dev.of_node &&
-	    of_parse_phandle(bp->pdev->dev.of_node, "phy-handle", 0)) {
-		ret = phylink_of_phy_connect(bp->phylink, bp->pdev-
 dev.of_node,
-					     0);
-		if (ret) {
-			netdev_err(dev, "Could not attach PHY (%d)\n", ret);
-			return ret;
-		}
-	} else {
+	if (dn)
Quoted 9 lines

drivers/net/ethernet/cadence/macb_main.c static int macb_phylink_connect(struct macb *bp) · line 624

Quoted 3 lines
drivers/net/ethernet/cadence/macb_main.c +3 -0
+		ret = phylink_of_phy_connect(bp->phylink, dn, 0);
+
+	if (!dn || (ret && !of_parse_phandle(dn, "phy-handle", 0))) {

Hi,
If of_parse_phandle() returns non-null, the device_node it returns will have its reference count increased by one. That reference needs to be put.

Okay, as per your suggestion below addition will be okay to store the "phy_node" and then of_node_put(phy_node) on error

phy_node = of_parse_phandle(dn, "phy-handle", 0);
        if (!dn || (ret && !phy_node)) {
                phydev = phy_find_first(bp->mii_bus);
                if (!phydev) {
                        netdev_err(dev, "no PHY found\n");
                        ret = -ENXIO;
                        goto phylink_connect_err;
                }

                /* attach the mac to the phy */
                ret = phylink_connect_phy(bp->phylink, phydev);
                if (ret) {
                        netdev_err(dev, "Could not attach to PHY (%d)\n", ret);
                        goto phylink_connect_err;
                }
        } else if (ret) {
                netdev_err(dev, "Could not attach PHY (%d)\n", ret);
                goto phylink_connect_err;
        }

        phylink_start(bp->phylink);

phylink_connect_err:

        if (phy_node)
                of_node_put(phy_node);

        return ret;
Quoted 4 lines

I assume you're trying to determine whether phylink_of_phy_connect() failed because of a missing phy-handle rather than of_phy_attach() failing? Maybe those two failures ought to be distinguished by errno return value?

Yes, PHY will be scanned only if phylink_of_phy_connect() returns error due to missing "phy-handle". Currently, phylink_of_phy_connect() returns same error for missing "phy-handle" and of_phy_attach() failure.

Quoted 35 lines

of_phy_attach() may fail due to of_phy_find_device() failing to find the PHY, or phy_attach_direct() failing. We could switch from using of_phy_attach(), to using of_phy_find_device() directly so we can then propagate phy_attach_direct()'s error code back, rather than losing it. That would then leave the case of of_phy_find_device() failure to be considered in terms of errno return value.

Quoted 14 lines
drivers/net/ethernet/cadence/macb_main.c +0 -0
 		phydev = phy_find_first(bp->mii_bus);
 		if (!phydev) {
drivers/net/ethernet/cadence/macb_main.c +3 -1
 			netdev_err(dev, "no PHY found\n");
@@ -638,6 +634,9 @@ static int macb_phylink_connect(struct macb *bp)
 			netdev_err(dev, "Could not attach to PHY (%d)\n",ret);
 			return ret;
 		}
+	} else if (ret) {
+		netdev_err(dev, "Could not attach PHY (%d)\n", ret);
+		return ret;
 	}
 
 	phylink_start(bp->phylink);
--
Quoted 3 lines

2.17.1

--RMK's Patch system: https://urldefense.proofpoint.com/v2/url?u=https-
3A__www.armlinux.org.uk_developer_patches_&d=DwIBAg&c=aUq983L2pue 2FqKFoP6PGHMJQyoJ7kl3s3GZ- _haXqY&r=BDdk1JtITE_JJ0519WwqU7IKF80Cw1i55lZOGqv2su8&m=blnuaRbic V2uF6XaoVuWN0U5yR5cFOzUSAs3ZPlxioU&s=rhp71ilc6R4_pmDsY07- kLPGbhyoyixXoHF0hMGu4Go&e=

FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up

According to speedtest.net: 11.9Mbps down 500kbps up

#4 lore raw

Re: [PATCH 1/3] net: macb: fix for fixed-link mode

Quoted 49 lines · Milind Parab
Quoted 34 lines
Quoted 34 lines

This patch fix the issue with fixed link. With fixed-link device opening fails due to macb_phylink_connect not handling fixed-link mode, in which case no MAC-PHY connection is needed and phylink_connect return success (0), however in current driver attempt is made to search and connect to PHY even for fixed-link.

---
 drivers/net/ethernet/cadence/macb_main.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +2 -9
index 9c767ee252ac..6b68ef34ab19 100644
@@ -615,17 +615,13 @@ static int macb_phylink_connect(struct macb *bp)
 {
 	struct net_device *dev = bp->dev;
 	struct phy_device *phydev;
+	struct device_node *dn = bp->pdev->dev.of_node;
 	int ret;
 
-	if (bp->pdev->dev.of_node &&
-	    of_parse_phandle(bp->pdev->dev.of_node, "phy-handle", 0)) {
-		ret = phylink_of_phy_connect(bp->phylink, bp->pdev-
 dev.of_node,
-					     0);
-		if (ret) {
-			netdev_err(dev, "Could not attach PHY (%d)\n", ret);
-			return ret;
-		}
-	} else {
+	if (dn)
Quoted 9 lines

drivers/net/ethernet/cadence/macb_main.c static int macb_phylink_connect(struct macb *bp) · line 625

Quoted 3 lines
drivers/net/ethernet/cadence/macb_main.c +3 -0
+		ret = phylink_of_phy_connect(bp->phylink, dn, 0);
+
+	if (!dn || (ret && !of_parse_phandle(dn, "phy-handle", 0))) {

Hi,
If of_parse_phandle() returns non-null, the device_node it returns will have its reference count increased by one. That reference needs to be put.

Okay, as per your suggestion below addition will be okay to store the "phy_node" and then of_node_put(phy_node) on error

phy_node = of_parse_phandle(dn, "phy-handle", 0);
        if (!dn || (ret && !phy_node)) {
                phydev = phy_find_first(bp->mii_bus);

...

Quoted 2 lines
        if (phy_node)
                of_node_put(phy_node);

As you're only interested in whether phy-handle exists or not, you could do this instead:

	phy_node = of_parse_phandle(dn, "phy-handle", 0);
	of_node_put(phy_node);
	if (!dn || (ret && !phy_node)) {
		...

Yes, it looks a bit weird, but the only thing you're interested in here is whether of_parse_phandle() returned NULL or non-NULL. You're not interested in dereferencing the pointer.

Some may raise some eye-brows at that, so it may be better to have this as a helper:

static bool macb_phy_handle_exists(struct device_node *dn)

{
	dn = of_parse_phandle(dn, "phy-handle", 0);
	of_node_put(dn);
	return dn != NULL;
}

and use it as:

	if (!dn || (ret && !macb_phy_handle_exists(dn))) {

which is more obvious what is going on.

Quoted 17 lines
        return ret;
Quoted 4 lines

I assume you're trying to determine whether phylink_of_phy_connect() failed because of a missing phy-handle rather than of_phy_attach() failing? Maybe those two failures ought to be distinguished by errno return value?

Yes, PHY will be scanned only if phylink_of_phy_connect() returns error due to missing "phy-handle". Currently, phylink_of_phy_connect() returns same error for missing "phy-handle" and of_phy_attach() failure.

Quoted 6 lines

of_phy_attach() may fail due to of_phy_find_device() failing to find the PHY, or phy_attach_direct() failing. We could switch from using of_phy_attach(), to using of_phy_find_device() directly so we can then propagate phy_attach_direct()'s error code back, rather than losing it. That would then leave the case of of_phy_find_device() failure to be considered in terms of errno return value.

Here's a patch I quickly knocked up that does this - may not apply to the kernel you're using as there's a whole bunch of work I have outstanding, but gives the outline idea. Does this help?

8<===
From: Russell King <[email protected]>
Subject: [PATCH] net: phylink: avoid of_phy_attach()

of_phy_attach() hides the return value of phy_attach_direct(), forcing us to return a "generic" ENODEV error code that is indistinguishable from the lack-of-phy-property case.

Switch to using of_phy_find_device() to find the PHY device, and then propagating any phy_attach_direct() error back to the caller.

---
 drivers/net/phy/phylink.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/net/phy/phylink.c +6 -3
index e9036b72114c..5a5109428d9e 100644
@@ -887,14 +887,17 @@ int phylink_of_phy_connect(struct phylink *pl, struct device_node *dn,
 		return 0;
 	}
 
-	phy_dev = of_phy_attach(pl->netdev, phy_node, flags,
-				pl->link_interface);
+	phy_dev = of_phy_find_device(phy_node);
 	/* We're done with the phy_node handle */
 	of_node_put(phy_node);
-
 	if (!phy_dev)
 		return -ENODEV;
 
+	ret = phy_attach_direct(pl->netdev, phy_dev, flags,
+				pl->link_interface);
+	if (ret)
+		return ret;
+
 	ret = phylink_bringup_phy(pl, phy_dev, pl->link_config.interface);
 	if (ret)
 		phy_detach(phy_dev);
Signature · 7 lines
2.20.1


-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
According to speedtest.net: 11.9Mbps down 500kbps up
#5 lore raw

RE: [PATCH 1/3] net: macb: fix for fixed-link mode

Quoted 42 lines
Quoted 11 lines
Quoted 9 lines

Review

Quoted 3 lines
(quoted diff) +3 -0
+		ret = phylink_of_phy_connect(bp->phylink, dn, 0);
+
+	if (!dn || (ret && !of_parse_phandle(dn, "phy-handle", 0))) {

Hi,
If of_parse_phandle() returns non-null, the device_node it returns will have its reference count increased by one. That reference needs to be put.

Okay, as per your suggestion below addition will be okay to store the

"phy_node" and then of_node_put(phy_node) on error

Quoted 4 lines
phy_node = of_parse_phandle(dn, "phy-handle", 0);
        if (!dn || (ret && !phy_node)) {
                phydev = phy_find_first(bp->mii_bus);

...

Quoted 2 lines
        if (phy_node)
                of_node_put(phy_node);

As you're only interested in whether phy-handle exists or not, you could do this instead:

	phy_node = of_parse_phandle(dn, "phy-handle", 0);
	of_node_put(phy_node);
	if (!dn || (ret && !phy_node)) {
		...

Yes, it looks a bit weird, but the only thing you're interested in here is whether of_parse_phandle() returned NULL or non-NULL. You're not interested in dereferencing the pointer.

Some may raise some eye-brows at that, so it may be better to have this as a helper:

static bool macb_phy_handle_exists(struct device_node *dn)

{
	dn = of_parse_phandle(dn, "phy-handle", 0);
	of_node_put(dn);
	return dn != NULL;
}
Quoted 6 lines

and use it as:

	if (!dn || (ret && !macb_phy_handle_exists(dn))) {

which is more obvious what is going on.

This is good. I will put this in the revised patch.

Quoted 19 lines
Quoted 4 lines

        return ret;
Quoted 14 lines
Quoted 4 lines

I assume you're trying to determine whether phylink_of_phy_connect() failed because of a missing phy-handle rather than of_phy_attach() failing? Maybe those two failures ought to be distinguished by errno return value?

Yes, PHY will be scanned only if phylink_of_phy_connect() returns error due to missing "phy-handle". Currently, phylink_of_phy_connect() returns same error for missing "phy-handle" and of_phy_attach() failure.

Quoted 6 lines

of_phy_attach() may fail due to of_phy_find_device() failing to find the PHY, or phy_attach_direct() failing. We could switch from using of_phy_attach(), to using of_phy_find_device() directly so we can then propagate phy_attach_direct()'s error code back, rather than losing it. That would then leave the case of of_phy_find_device() failure to be considered in terms of errno return value.

Quoted 5 lines

Here's a patch I quickly knocked up that does this - may not apply to the kernel you're using as there's a whole bunch of work I have outstanding, but gives the outline idea. Does this help?

Yes, this will help. Once available, we will adopt this change.

Quoted 14 lines

8<===
From: Russell King <[email protected]>
Subject: [PATCH] net: phylink: avoid of_phy_attach()

of_phy_attach() hides the return value of phy_attach_direct(), forcing us to return a "generic" ENODEV error code that is indistinguishable from the lack-of-phy-property case.
Switch to using of_phy_find_device() to find the PHY device, and then propagating any phy_attach_direct() error back to the caller.

---

#6 lore raw

PATCH 2/3 net: macb: add support for C45 MDIO read/write

In reply to Milind Parab

This patch modify MDIO read/write functions to support communication with C45 PHY.

---
 drivers/net/ethernet/cadence/macb.h      | 15 ++++--
 drivers/net/ethernet/cadence/macb_main.c | 61 +++++++++++++++++++-----
 2 files changed, 61 insertions(+), 15 deletions(-)
drivers/net/ethernet/cadence/macb.h +11 -4
index 19fe4f4867c7..dbf7070fcdba 100644
@@ -630,10 +630,17 @@
 #define GEM_CLK_DIV96				5
 
 /* Constants for MAN register */
-#define MACB_MAN_SOF				1
-#define MACB_MAN_WRITE				1
-#define MACB_MAN_READ				2
-#define MACB_MAN_CODE				2
+#define MACB_MAN_C22_SOF			1
+#define MACB_MAN_C22_WRITE			1
+#define MACB_MAN_C22_READ			2
+#define MACB_MAN_C22_CODE			2
+
+#define MACB_MAN_C45_SOF			0
+#define MACB_MAN_C45_ADDR			0
+#define MACB_MAN_C45_WRITE			1
+#define MACB_MAN_C45_POST_READ_INCR		2
+#define MACB_MAN_C45_READ			3
+#define MACB_MAN_C45_CODE			2
 
 /* Capability mask bits */
 #define MACB_CAPS_ISR_CLEAR_ON_WRITE		0x00000001
drivers/net/ethernet/cadence/macb_main.c +50 -11

73 lines not on this page · raw message

Signature · 1 line
2.17.1
#7 lore raw

PATCH 3/3 net: macb: add support for high speed interface

In reply to Milind Parab

This patch add support for high speed USXGMII PCS and 10G speed in Cadence ethernet controller driver.

---
 drivers/net/ethernet/cadence/macb.h      |  50 ++++++++
 drivers/net/ethernet/cadence/macb_main.c | 142 ++++++++++++++++++++---
 2 files changed, 174 insertions(+), 18 deletions(-)
drivers/net/ethernet/cadence/macb.h +50 -0
index dbf7070fcdba..b731807d1c49 100644
@@ -76,10 +76,12 @@
 #define MACB_RBQPH		0x04D4
 
 /* GEM register offsets. */
+#define GEM_NCR			0x0000 /* Network Control */
 #define GEM_NCFGR		0x0004 /* Network Config */
 #define GEM_USRIO		0x000c /* User IO */
 #define GEM_DMACFG		0x0010 /* DMA Configuration */
 #define GEM_JML			0x0048 /* Jumbo Max Length */
+#define GEM_HS_MAC_CONFIG	0x0050 /* GEM high speed config */
 #define GEM_HRB			0x0080 /* Hash Bottom */
 #define GEM_HRT			0x0084 /* Hash Top */
 #define GEM_SA1B		0x0088 /* Specific1 Bottom */
@@ -164,6 +166,9 @@
 #define GEM_DCFG7		0x0298 /* Design Config 7 */
 #define GEM_DCFG8		0x029C /* Design Config 8 */
 #define GEM_DCFG10		0x02A4 /* Design Config 10 */
+#define GEM_DCFG12		0x02AC /* Design Config 12 */
+#define GEM_USX_CONTROL		0x0A80 /* USXGMII control register */
+#define GEM_USX_STATUS		0x0A88 /* USXGMII status register */
 
 #define GEM_TXBDCTRL	0x04cc /* TX Buffer Descriptor control register */
 #define GEM_RXBDCTRL	0x04d0 /* RX Buffer Descriptor control register */
@@ -270,11 +275,19 @@
 #define MACB_IRXFCS_OFFSET	19
 #define MACB_IRXFCS_SIZE	1
 
+/* GEM specific NCR bitfields. */
+#define GEM_ENABLE_HS_MAC_OFFSET	31
+#define GEM_ENABLE_HS_MAC_SIZE		1
+
 /* GEM specific NCFGR bitfields. */
+#define GEM_FD_OFFSET		1 /* Full duplex */
+#define GEM_FD_SIZE		1
 #define GEM_GBE_OFFSET		10 /* Gigabit mode enable */
 #define GEM_GBE_SIZE		1
 #define GEM_PCSSEL_OFFSET	11
 #define GEM_PCSSEL_SIZE		1
+#define GEM_PAE_OFFSET		13 /* Pause enable */
+#define GEM_PAE_SIZE		1
 #define GEM_CLK_OFFSET		18 /* MDC clock division */
 #define GEM_CLK_SIZE		3
 #define GEM_DBW_OFFSET		21 /* Data bus width */
@@ -455,11 +468,17 @@
 #define MACB_REV_OFFSET				0
 #define MACB_REV_SIZE				16
 
+/* Bitfield in HS_MAC_CONFIG */
+#define GEM_HS_MAC_SPEED_OFFSET			0
+#define GEM_HS_MAC_SPEED_SIZE			3
+
 /* Bitfields in DCFG1. */
 #define GEM_IRQCOR_OFFSET			23
 #define GEM_IRQCOR_SIZE				1
 #define GEM_DBWDEF_OFFSET			25
 #define GEM_DBWDEF_SIZE				3
+#define GEM_NO_PCS_OFFSET			0
+#define GEM_NO_PCS_SIZE				1
 
 /* Bitfields in DCFG2. */
 #define GEM_RX_PKT_BUFF_OFFSET			20
@@ -494,6 +513,34 @@
 #define GEM_RXBD_RDBUFF_OFFSET			8
 #define GEM_RXBD_RDBUFF_SIZE			4
 
+/* Bitfields in DCFG12. */
+#define GEM_HIGH_SPEED_OFFSET			26
+#define GEM_HIGH_SPEED_SIZE			1
+
+/* Bitfields in USX_CONTROL. */
+#define GEM_USX_CTRL_SPEED_OFFSET		14
+#define GEM_USX_CTRL_SPEED_SIZE			3
+#define GEM_SERDES_RATE_OFFSET			12
+#define GEM_SERDES_RATE_SIZE			2
+#define GEM_RX_SCR_BYPASS_OFFSET		9
+#define GEM_RX_SCR_BYPASS_SIZE			1
+#define GEM_TX_SCR_BYPASS_OFFSET		8
+#define GEM_TX_SCR_BYPASS_SIZE			1
+#define GEM_RX_SYNC_RESET_OFFSET		2
+#define GEM_RX_SYNC_RESET_SIZE			1
+#define GEM_TX_EN_OFFSET			1
+#define GEM_TX_EN_SIZE				1
+#define GEM_SIGNAL_OK_OFFSET			0
+#define GEM_SIGNAL_OK_SIZE			1
+
+/* Bitfields in USX_STATUS. */
+#define GEM_USX_TX_FAULT_OFFSET			28
+#define GEM_USX_TX_FAULT_SIZE			1
+#define GEM_USX_RX_FAULT_OFFSET			27
+#define GEM_USX_RX_FAULT_SIZE			1
+#define GEM_USX_BLOCK_LOCK_OFFSET		0
+#define GEM_USX_BLOCK_LOCK_SIZE			1
+
 /* Bitfields in TISUBN */
 #define GEM_SUBNSINCR_OFFSET			0
 #define GEM_SUBNSINCRL_OFFSET			24
@@ -656,6 +703,8 @@
 #define MACB_CAPS_GIGABIT_MODE_AVAILABLE	0x20000000
 #define MACB_CAPS_SG_DISABLED			0x40000000
 #define MACB_CAPS_MACB_IS_GEM			0x80000000
+#define MACB_CAPS_PCS				0x01000000
+#define MACB_CAPS_HIGH_SPEED			0x02000000
 
 /* LSO settings */
 #define MACB_LSO_UFO_ENABLE			0x01
@@ -724,6 +773,7 @@
 	})
 
 #define MACB_READ_NSR(bp)	macb_readl(bp, NSR)
+#define GEM_READ_USX_STATUS(bp)	gem_readl(bp, USX_STATUS)
 
 /* struct macb_dma_desc - Hardware DMA descriptor
  * @addr: DMA address of data buffer
drivers/net/ethernet/cadence/macb_main.c +124 -18

196 lines not on this page · raw message

Signature · 1 line
2.17.1
#8 lore raw

Re: [PATCH 3/3] net: macb: add support for high speed interface

Quoted 2 lines · Milind Parab

This patch add support for high speed USXGMII PCS and 10G speed in Cadence ethernet controller driver.

How has this been tested?

Quoted 128 lines
---
 drivers/net/ethernet/cadence/macb.h      |  50 ++++++++
 drivers/net/ethernet/cadence/macb_main.c | 142 ++++++++++++++++++++---
 2 files changed, 174 insertions(+), 18 deletions(-)
drivers/net/ethernet/cadence/macb.h +50 -0
index dbf7070fcdba..b731807d1c49 100644
@@ -76,10 +76,12 @@
 #define MACB_RBQPH		0x04D4
 
 /* GEM register offsets. */
+#define GEM_NCR			0x0000 /* Network Control */
 #define GEM_NCFGR		0x0004 /* Network Config */
 #define GEM_USRIO		0x000c /* User IO */
 #define GEM_DMACFG		0x0010 /* DMA Configuration */
 #define GEM_JML			0x0048 /* Jumbo Max Length */
+#define GEM_HS_MAC_CONFIG	0x0050 /* GEM high speed config */
 #define GEM_HRB			0x0080 /* Hash Bottom */
 #define GEM_HRT			0x0084 /* Hash Top */
 #define GEM_SA1B		0x0088 /* Specific1 Bottom */
@@ -164,6 +166,9 @@
 #define GEM_DCFG7		0x0298 /* Design Config 7 */
 #define GEM_DCFG8		0x029C /* Design Config 8 */
 #define GEM_DCFG10		0x02A4 /* Design Config 10 */
+#define GEM_DCFG12		0x02AC /* Design Config 12 */
+#define GEM_USX_CONTROL		0x0A80 /* USXGMII control register */
+#define GEM_USX_STATUS		0x0A88 /* USXGMII status register */
 
 #define GEM_TXBDCTRL	0x04cc /* TX Buffer Descriptor control register */
 #define GEM_RXBDCTRL	0x04d0 /* RX Buffer Descriptor control register */
@@ -270,11 +275,19 @@
 #define MACB_IRXFCS_OFFSET	19
 #define MACB_IRXFCS_SIZE	1
 
+/* GEM specific NCR bitfields. */
+#define GEM_ENABLE_HS_MAC_OFFSET	31
+#define GEM_ENABLE_HS_MAC_SIZE		1
+
 /* GEM specific NCFGR bitfields. */
+#define GEM_FD_OFFSET		1 /* Full duplex */
+#define GEM_FD_SIZE		1
 #define GEM_GBE_OFFSET		10 /* Gigabit mode enable */
 #define GEM_GBE_SIZE		1
 #define GEM_PCSSEL_OFFSET	11
 #define GEM_PCSSEL_SIZE		1
+#define GEM_PAE_OFFSET		13 /* Pause enable */
+#define GEM_PAE_SIZE		1
 #define GEM_CLK_OFFSET		18 /* MDC clock division */
 #define GEM_CLK_SIZE		3
 #define GEM_DBW_OFFSET		21 /* Data bus width */
@@ -455,11 +468,17 @@
 #define MACB_REV_OFFSET				0
 #define MACB_REV_SIZE				16
 
+/* Bitfield in HS_MAC_CONFIG */
+#define GEM_HS_MAC_SPEED_OFFSET			0
+#define GEM_HS_MAC_SPEED_SIZE			3
+
 /* Bitfields in DCFG1. */
 #define GEM_IRQCOR_OFFSET			23
 #define GEM_IRQCOR_SIZE				1
 #define GEM_DBWDEF_OFFSET			25
 #define GEM_DBWDEF_SIZE				3
+#define GEM_NO_PCS_OFFSET			0
+#define GEM_NO_PCS_SIZE				1
 
 /* Bitfields in DCFG2. */
 #define GEM_RX_PKT_BUFF_OFFSET			20
@@ -494,6 +513,34 @@
 #define GEM_RXBD_RDBUFF_OFFSET			8
 #define GEM_RXBD_RDBUFF_SIZE			4
 
+/* Bitfields in DCFG12. */
+#define GEM_HIGH_SPEED_OFFSET			26
+#define GEM_HIGH_SPEED_SIZE			1
+
+/* Bitfields in USX_CONTROL. */
+#define GEM_USX_CTRL_SPEED_OFFSET		14
+#define GEM_USX_CTRL_SPEED_SIZE			3
+#define GEM_SERDES_RATE_OFFSET			12
+#define GEM_SERDES_RATE_SIZE			2
+#define GEM_RX_SCR_BYPASS_OFFSET		9
+#define GEM_RX_SCR_BYPASS_SIZE			1
+#define GEM_TX_SCR_BYPASS_OFFSET		8
+#define GEM_TX_SCR_BYPASS_SIZE			1
+#define GEM_RX_SYNC_RESET_OFFSET		2
+#define GEM_RX_SYNC_RESET_SIZE			1
+#define GEM_TX_EN_OFFSET			1
+#define GEM_TX_EN_SIZE				1
+#define GEM_SIGNAL_OK_OFFSET			0
+#define GEM_SIGNAL_OK_SIZE			1
+
+/* Bitfields in USX_STATUS. */
+#define GEM_USX_TX_FAULT_OFFSET			28
+#define GEM_USX_TX_FAULT_SIZE			1
+#define GEM_USX_RX_FAULT_OFFSET			27
+#define GEM_USX_RX_FAULT_SIZE			1
+#define GEM_USX_BLOCK_LOCK_OFFSET		0
+#define GEM_USX_BLOCK_LOCK_SIZE			1
+
 /* Bitfields in TISUBN */
 #define GEM_SUBNSINCR_OFFSET			0
 #define GEM_SUBNSINCRL_OFFSET			24
@@ -656,6 +703,8 @@
 #define MACB_CAPS_GIGABIT_MODE_AVAILABLE	0x20000000
 #define MACB_CAPS_SG_DISABLED			0x40000000
 #define MACB_CAPS_MACB_IS_GEM			0x80000000
+#define MACB_CAPS_PCS				0x01000000
+#define MACB_CAPS_HIGH_SPEED			0x02000000
 
 /* LSO settings */
 #define MACB_LSO_UFO_ENABLE			0x01
@@ -724,6 +773,7 @@
 	})
 
 #define MACB_READ_NSR(bp)	macb_readl(bp, NSR)
+#define GEM_READ_USX_STATUS(bp)	gem_readl(bp, USX_STATUS)
 
 /* struct macb_dma_desc - Hardware DMA descriptor
  * @addr: DMA address of data buffer
drivers/net/ethernet/cadence/macb_main.c +0 -0
index 17297b01e85e..710ee18a0ef0 100644

drivers/net/ethernet/cadence/macb_main.c struct sifive_fu540_macb_mgmt { · line 89

Quoted 9 lines
drivers/net/ethernet/cadence/macb_main.c +6 -0
@@ -81,6 +81,18 @@ struct sifive_fu540_macb_mgmt {
 #define MACB_WOL_HAS_MAGIC_PACKET	(0x1 << 0)
 #define MACB_WOL_ENABLED		(0x1 << 1)
 
+enum {
+	HS_MAC_SPEED_100M,
+	HS_MAC_SPEED_1000M,
+	HS_MAC_SPEED_2500M,
+	HS_MAC_SPEED_5000M,
+	HS_MAC_SPEED_10000M,

Are these chip register definitions? Shouldn't you be relying on fixed values for these, rather than their position in an enumerated list?

Quoted 80 lines
drivers/net/ethernet/cadence/macb_main.c +45 -0
+};
+
+enum {
+	MACB_SERDES_RATE_10G = 1,
+};
+
 /* Graceful stop timeouts in us. We should allow up to
  * 1 frame time (10 Mbits/s, full-duplex, ignoring collisions)
  */
@@ -90,6 +102,8 @@ struct sifive_fu540_macb_mgmt {
 
 #define MACB_MDIO_TIMEOUT	1000000 /* in usecs */
 
+#define MACB_USX_BLOCK_LOCK_TIMEOUT	1000000 /* in usecs */
+
 /* DMA buffer descriptor might be different size
  * depends on hardware configuration:
  *
@@ -506,6 +520,7 @@ static void macb_validate(struct phylink_config *config,
 	    state->interface != PHY_INTERFACE_MODE_RMII &&
 	    state->interface != PHY_INTERFACE_MODE_GMII &&
 	    state->interface != PHY_INTERFACE_MODE_SGMII &&
+	    state->interface != PHY_INTERFACE_MODE_USXGMII &&
 	    !phy_interface_mode_is_rgmii(state->interface)) {
 		bitmap_zero(supported, __ETHTOOL_LINK_MODE_MASK_NBITS);
 		return;
@@ -518,6 +533,13 @@ static void macb_validate(struct phylink_config *config,
 		return;
 	}
 
+	if (state->interface == PHY_INTERFACE_MODE_USXGMII &&
+	    !(bp->caps & MACB_CAPS_HIGH_SPEED &&
+	      bp->caps & MACB_CAPS_PCS)) {
+		bitmap_zero(supported, __ETHTOOL_LINK_MODE_MASK_NBITS);
+		return;
+	}
+
 	phylink_set_port_modes(mask);
 	phylink_set(mask, Autoneg);
 	phylink_set(mask, Asym_Pause);
@@ -527,6 +549,22 @@ static void macb_validate(struct phylink_config *config,
 	phylink_set(mask, 100baseT_Half);
 	phylink_set(mask, 100baseT_Full);
 
+	if (bp->caps & MACB_CAPS_GIGABIT_MODE_AVAILABLE &&
+	    (state->interface == PHY_INTERFACE_MODE_NA ||
+	     state->interface == PHY_INTERFACE_MODE_USXGMII)) {
+		phylink_set(mask, 10000baseCR_Full);
+		phylink_set(mask, 10000baseER_Full);
+		phylink_set(mask, 10000baseKR_Full);
+		phylink_set(mask, 10000baseLR_Full);
+		phylink_set(mask, 10000baseLRM_Full);
+		phylink_set(mask, 10000baseSR_Full);
+		phylink_set(mask, 10000baseT_Full);
+		phylink_set(mask, 5000baseT_Full);
+		phylink_set(mask, 2500baseX_Full);
+		phylink_set(mask, 1000baseX_Full);
+		phylink_set(mask, 1000baseT_Full);
+	}
+
 	if (bp->caps & MACB_CAPS_GIGABIT_MODE_AVAILABLE &&
 	    (state->interface == PHY_INTERFACE_MODE_NA ||
 	     state->interface == PHY_INTERFACE_MODE_GMII ||
@@ -544,6 +582,60 @@ static void macb_validate(struct phylink_config *config,
 		   __ETHTOOL_LINK_MODE_MASK_NBITS);
 }
 
+static int gem_mac_usx_configure(struct macb *bp,
+				 const struct phylink_link_state *state)
+{
+	u32 speed, config, val;
+	int ret;
+
+	val = gem_readl(bp, NCFGR);
+	val = GEM_BIT(PCSSEL) | (~GEM_BIT(SGMIIEN) & val);
+	if (state->pause & MLO_PAUSE_TX)
+		val |= GEM_BIT(PAE);
+	gem_writel(bp, NCFGR, val);
+	gem_writel(bp, NCR, gem_readl(bp, NCR) | GEM_BIT(ENABLE_HS_MAC));
+	gem_writel(bp, NCFGR, gem_readl(bp, NCFGR) | GEM_BIT(FD));

drivers/net/ethernet/cadence/macb_main.c static void macb_validate(struct phylink_config *config, · line 607

Quoted 10 lines
drivers/net/ethernet/cadence/macb_main.c +10 -0
+	config = gem_readl(bp, USX_CONTROL);
+	config = GEM_BFINS(SERDES_RATE, MACB_SERDES_RATE_10G, config);
+	config &= ~GEM_BIT(TX_SCR_BYPASS);
+	config &= ~GEM_BIT(RX_SCR_BYPASS);
+	gem_writel(bp, USX_CONTROL, config | GEM_BIT(TX_EN));
+	config = gem_readl(bp, USX_CONTROL);
+	gem_writel(bp, USX_CONTROL, config | GEM_BIT(SIGNAL_OK));
+	ret = readx_poll_timeout(GEM_READ_USX_STATUS, bp, val,
+				 val & GEM_BIT(USX_BLOCK_LOCK),
+				 1, MACB_USX_BLOCK_LOCK_TIMEOUT);

What if there's no signal to lock on to? That's treated as link down and is not a failure.

Quoted 13 lines
drivers/net/ethernet/cadence/macb_main.c +13 -0
+	if (ret < 0) {
+		netdev_warn(bp->dev, "USXGMII block lock failed");
+		return -ETIMEDOUT;
+	}
+
+	switch (state->speed) {
+	case SPEED_10000:
+		speed = HS_MAC_SPEED_10000M;
+		break;
+	case SPEED_5000:
+		speed = HS_MAC_SPEED_5000M;
+		break;
+	case SPEED_2500:

drivers/net/ethernet/cadence/macb_main.c static void macb_validate(struct phylink_config *config, · line 630

Quoted 10 lines
drivers/net/ethernet/cadence/macb_main.c +10 -0
+		speed = HS_MAC_SPEED_2500M;
+		break;
+	case SPEED_1000:
+		speed = HS_MAC_SPEED_1000M;
+		break;
+	default:
+	case SPEED_100:
+		speed = HS_MAC_SPEED_100M;
+		break;
+	}

So you only support fixed-mode (phy and fixed links) and not in-band links here.

Quoted 85 lines
drivers/net/ethernet/cadence/macb_main.c +40 -18
+
+	gem_writel(bp, HS_MAC_CONFIG, GEM_BFINS(HS_MAC_SPEED, speed,
+						gem_readl(bp, HS_MAC_CONFIG)));
+	gem_writel(bp, USX_CONTROL, GEM_BFINS(USX_CTRL_SPEED, speed,
+					      gem_readl(bp, USX_CONTROL)));
+	return 0;
+}
+
 static void macb_mac_pcs_get_state(struct phylink_config *config,
 				   struct phylink_link_state *state)
 {
@@ -565,30 +657,39 @@ static void macb_mac_config(struct phylink_config *config, unsigned int mode,
 
 	spin_lock_irqsave(&bp->lock, flags);
 
-	old_ctrl = ctrl = macb_or_gem_readl(bp, NCFGR);
+	if (bp->phy_interface == PHY_INTERFACE_MODE_USXGMII) {
+		if (gem_mac_usx_configure(bp, state) < 0) {
+			spin_unlock_irqrestore(&bp->lock, flags);
+			phylink_mac_change(bp->phylink, false);
+			return;
+		}
+	} else {
+		old_ctrl = ctrl = macb_or_gem_readl(bp, NCFGR);
 
-	/* Clear all the bits we might set later */
-	ctrl &= ~(GEM_BIT(GBE) | MACB_BIT(SPD) | MACB_BIT(FD) | MACB_BIT(PAE) |
-		  GEM_BIT(SGMIIEN) | GEM_BIT(PCSSEL));
+		/* Clear all the bits we might set later */
+		ctrl &= ~(GEM_BIT(GBE) | MACB_BIT(SPD) |
+			  MACB_BIT(FD) | MACB_BIT(PAE) |
+			  GEM_BIT(SGMIIEN) | GEM_BIT(PCSSEL));
 
-	if (state->speed == SPEED_1000)
-		ctrl |= GEM_BIT(GBE);
-	else if (state->speed == SPEED_100)
-		ctrl |= MACB_BIT(SPD);
+		if (state->speed == SPEED_1000)
+			ctrl |= GEM_BIT(GBE);
+		else if (state->speed == SPEED_100)
+			ctrl |= MACB_BIT(SPD);
 
-	if (state->duplex)
-		ctrl |= MACB_BIT(FD);
+		if (state->duplex)
+			ctrl |= MACB_BIT(FD);
 
-	/* We do not support MLO_PAUSE_RX yet */
-	if (state->pause & MLO_PAUSE_TX)
-		ctrl |= MACB_BIT(PAE);
+		/* We do not support MLO_PAUSE_RX yet */
+		if (state->pause & MLO_PAUSE_TX)
+			ctrl |= MACB_BIT(PAE);
 
-	if (state->interface == PHY_INTERFACE_MODE_SGMII)
-		ctrl |= GEM_BIT(SGMIIEN) | GEM_BIT(PCSSEL);
+		if (state->interface == PHY_INTERFACE_MODE_SGMII)
+			ctrl |= GEM_BIT(SGMIIEN) | GEM_BIT(PCSSEL);
 
-	/* Apply the new configuration, if any */
-	if (old_ctrl ^ ctrl)
-		macb_or_gem_writel(bp, NCFGR, ctrl);
+		/* Apply the new configuration, if any */
+		if (old_ctrl ^ ctrl)
+			macb_or_gem_writel(bp, NCFGR, ctrl);
+	}
 
 	bp->speed = state->speed;
 
@@ -3399,6 +3500,11 @@ static void macb_configure_caps(struct macb *bp,
 		dcfg = gem_readl(bp, DCFG1);
 		if (GEM_BFEXT(IRQCOR, dcfg) == 0)
 			bp->caps |= MACB_CAPS_ISR_CLEAR_ON_WRITE;
+		if (GEM_BFEXT(NO_PCS, dcfg) == 0)
+			bp->caps |= MACB_CAPS_PCS;
+		dcfg = gem_readl(bp, DCFG12);
+		if (GEM_BFEXT(HIGH_SPEED, dcfg) == 1)
+			bp->caps |= MACB_CAPS_HIGH_SPEED;
 		dcfg = gem_readl(bp, DCFG2);
 		if ((dcfg & (GEM_BIT(RX_PKT_BUFF) | GEM_BIT(TX_PKT_BUFF))) == 0)
 			bp->caps |= MACB_CAPS_FIFO_MODE;
Signature · 1 line
2.17.1
Signature · 3 lines
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
According to speedtest.net: 11.9Mbps down 500kbps up
#9 lore raw

RE: [PATCH 3/3] net: macb: add support for high speed interface

Quoted 5 lines
Quoted 2 lines

This patch add support for high speed USXGMII PCS and 10G speed in Cadence ethernet controller driver.

How has this been tested?

This patch is tested in 10G fixed mode on SFP+ module.

In our own lab, we have various hardware test platforms supporting SGMII (through a TI PHY and another build that connects to a Marvell 1G PHY), GMII (through a Marvell PHY), 10GBASE-R (direct connection to SFP+), USXGMII (currently we can emulate this using an SFP+ connection from/to our own hardware)

Quoted 4 lines
Quoted 4 lines
drivers/net/ethernet/cadence/macb_main.c +0 -0
@@ -81,6 +81,18 @@ struct sifive_fu540_macb_mgmt {
 #define MACB_WOL_HAS_MAGIC_PACKET	(0x1 << 0)
 #define MACB_WOL_ENABLED		(0x1 << 1)
 
Quoted 10 lines

drivers/net/ethernet/cadence/macb_main.c struct sifive_fu540_macb_mgmt { · line 89

Quoted 6 lines
drivers/net/ethernet/cadence/macb_main.c +6 -0
+enum {
+	HS_MAC_SPEED_100M,
+	HS_MAC_SPEED_1000M,
+	HS_MAC_SPEED_2500M,
+	HS_MAC_SPEED_5000M,
+	HS_MAC_SPEED_10000M,

Are these chip register definitions? Shouldn't you be relying on fixed values for these, rather than their position in an enumerated list?

Okay, yes it is safe to use fixed values instead of enum. I will change this.

Quoted 11 lines

drivers/net/ethernet/cadence/macb_main.c static void macb_validate(struct phylink_config *config, · line 607

Quoted 7 lines
drivers/net/ethernet/cadence/macb_main.c +6 -0
+	gem_writel(bp, USX_CONTROL, config | GEM_BIT(TX_EN));
+	config = gem_readl(bp, USX_CONTROL);
+	gem_writel(bp, USX_CONTROL, config | GEM_BIT(SIGNAL_OK));
+	ret = readx_poll_timeout(GEM_READ_USX_STATUS, bp, val,
+				 val & GEM_BIT(USX_BLOCK_LOCK),
+				 1, MACB_USX_BLOCK_LOCK_TIMEOUT);

What if there's no signal to lock on to? That's treated as link down and is not a failure.

Yes, if there is no signal to lock on, that is treated as link down. Is this okay or should there be some additional error handling needed?

Quoted 14 lines
Quoted 14 lines
drivers/net/ethernet/cadence/macb_main.c +14 -0
+
+	switch (state->speed) {
+	case SPEED_10000:
+		speed = HS_MAC_SPEED_10000M;
+		break;
+	case SPEED_5000:
+		speed = HS_MAC_SPEED_5000M;
+		break;
+	case SPEED_2500:
+		speed = HS_MAC_SPEED_2500M;
+		break;
+	case SPEED_1000:
+		speed = HS_MAC_SPEED_1000M;
+		break;
Quoted 9 lines

drivers/net/ethernet/cadence/macb_main.c static void macb_validate(struct phylink_config *config, · line 630

Quoted 5 lines
drivers/net/ethernet/cadence/macb_main.c +5 -0
+	default:
+	case SPEED_100:
+		speed = HS_MAC_SPEED_100M;
+		break;
+	}

So you only support fixed-mode (phy and fixed links) and not in-band links here.

Yes, this is right.

maintainer
#10 lore raw

Re: [PATCH 3/3] net: macb: add support for high speed interface

Quoted 6 lines
Quoted 2 lines

How has this been tested?

This patch is tested in 10G fixed mode on SFP+ module.

In our own lab, we have various hardware test platforms supporting SGMII (through a TI PHY and another build that connects to a Marvell 1G PHY), GMII (through a Marvell PHY), 10GBASE-R (direct connection to SFP+), USXGMII (currently we can emulate this using an SFP+ connection from/to our own hardware)

Are any of these PHY using C45?

Thanks
Andrew

#11 lore raw

RE: [PATCH 3/3] net: macb: add support for high speed interface

Quoted 5 lines
Quoted 5 lines
Quoted 2 lines

How has this been tested?

This patch is tested in 10G fixed mode on SFP+ module.

Quoted 7 lines
Quoted 1 line

In our own lab, we have various hardware test platforms supporting SGMII

(through a TI PHY and another build that connects to a Marvell 1G PHY), GMII (through a Marvell PHY), 10GBASE-R (direct connection to SFP+), USXGMII (currently we can emulate this using an SFP+ connection from/to our own hardware)

Are any of these PHY using C45?

No, none of these PHYs use C45.
For C45 testing we had a simulated PHY. The simulated PHY implemented a Clause 45 slave interface.

Quoted 3 lines

Thanks
Andrew

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): BY5PR07MB65147AEF32345E351E920188D35A0@BY5PR07MB6514.namprd07.prod.outlook.com

#14 lore raw

Re: [PATCH 3/3] net: macb: add support for high speed interface

[private email content deleted, added Cc list back since this is important.]

I'm still not getting a good enough view of what you are doing and how my understanding of your hardware fits with what you're doing with the software.

My understanding is it's something like:

	----+
        SOC |             PCS
	MAC --(USXGMII)-- PHY ----- PHY or SFP
	    |
	----+

And you are just modelling the MAC part in phylink, where as phylink has so far been used on systems which have this model - where phylink knows about both the MAC and the PCS PHY:

	---------------+
	         PCS   |
	MAC ---- PHY ----- PHY or SFP
	     SOC       |
      	---------------+

This is why I recently renamed mac_link_state() to mac_pcs_get_state() to make it clearer that it reads from the PCS not from the current settings of the MAC. So far, all such setups do not implement the PCS PHY as an 802.3 register set; they implement it as part of the MAC register set.

In the former case, if phylink is used to manage the connection between the MAC and the PCS PHY, phylink has nothing to do with the SFP at all.

In the latter case, phylink is used to manage the connection between the PCS PHY and external device, controlling the MAC as appropriate.

My problem is I believe your hardware is the former case, but you are trying to implement the latter case by ignoring in-band mode. As SFPs rely on in-band mode, that isn't going to work.

The options for the former case are:

1) implement phylink covering both the MAC and the external PCS PHY
2) implement phylink just for the MAC to PCS PHY connection but not SFPs, and implement SFP support separately in the PCS PHY driver.

Maybe phylink needs to split mac_pcs_get_state() so it can be supplied by a separate driver, or by the MAC driver as appropriate - but that brings with it other problems; phylink with a directly attached SFP considers the state of the link between the PCS PHY and the external device - not only speed but also interface mode for that part of the link. What you'd see in the mac_config() callback are interface modes for that part of the link, not between the MAC and the PCS PHY.

To change that would require reworking almost every driver that has already converted over to somehow remodel the built-in PCS and COMPHY as a separate PCS PHY for phylink. I'm not entirely clear whether that would work though.

Signature · 3 lines
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
According to speedtest.net: 11.9Mbps down 500kbps up
#15 lore raw

RE: [PATCH 3/3] net: macb: add support for high speed interface

Quoted 16 lines

I'm still not getting a good enough view of what you are doing and how my understanding of your hardware fits with what you're doing with the software.

My understanding is it's something like:

	----+
       SOC |             PCS
	MAC --(USXGMII)-- PHY ----- PHY or SFP
	    |
	----+

And you are just modelling the MAC part in phylink, where as phylink has so far been used on systems which have this model - where phylink knows about both the MAC and the PCS PHY:

Quoted 6 lines
	---------------+
	         PCS   |
	MAC ---- PHY ----- PHY or SFP
	     SOC       |
     	---------------+

Here, I am describing the GEM component followed by the test setup. Please see, if the below explanation is depicting the correct picture

The Cadence MAC, referred to as GEM is a hardware module which implements 10/100/1000Mbps Ethernet MAC with the following interface types:
MII, RMII, GMII, RGMII which can operate in either half or full duplex mode;
and also as a separate instance a full duplex 10G MAC with an XGMII interface.
GEM comprises the following constituent components:
- MAC controlling transmit, receive, address checking and loopback
- Configuration registers (REG_TOP) providing control and status registers, statistics registers and synchronization logic
- Two Physical Coding Sublayer (PCS) components; one comprising 8B/10B encode/decode, PCS transmit, PCS receive, and PCS auto-negotiation and another implementing USXGMII functionality.

As our ethernet controller (GEM) have MAC as well as USXGMII PCS, we program both appropriately based on the values passed from Phylink. And for fixed-link, Phylink correctly read out these values from device tree node and relay it to mac_config. Also note that we are not setting "sfp" node in device tree.

We are modelling MAC + USXGMII PCS.
The test configuration is as below

                                    +-------------------------+
                                    |              GEM        |
Host PC1   < ------------------->   | MAC ---- USXGMII PCS    |----- SerDes ------ SFP+ <------ Direct attach cable ------->  Chelsio 10G Card   <--->   Host PC2
                                    |        PCI based NIC    | 
                                    +-------------------------+

The setup has a 10G fixed link between PCS and SFP+.
Our test setup is emulated on Xilinx Virtex VCU118 FPGA base board placed in a PCIe slot of a PC running Ubuntu. The FPGA base board contained an image implementing a NIC using Cadence PCIe and Ethernet IP cores.
We had a separate PC also running Ubuntu containing a Chelsio 10G NIC.
We connected these with an SFP+ direct attach passive twinx-ax copper cable. Testing of the link was done with ping and iperf3 Also note that there is no PHY in the SFP+ cage.
As I understand it everything regarding SFP+ in our setup is passive so there is nothing for PHYLINK to talk to. We do not have a module in the SFP+ cage, just a direct connection to a copper cable

Quoted 47 lines

This is why I recently renamed mac_link_state() to mac_pcs_get_state() to make it clearer that it reads from the PCS not from the current settings of the MAC. So far, all such setups do not implement the PCS PHY as an 802.3 register set; they implement it as part of the MAC register set.

In the former case, if phylink is used to manage the connection between the MAC and the PCS PHY, phylink has nothing to do with the SFP at all.

In the latter case, phylink is used to manage the connection between the PCS PHY and external device, controlling the MAC as appropriate.

My problem is I believe your hardware is the former case, but you are trying to implement the latter case by ignoring in-band mode. As SFPs rely on in-band mode, that isn't going to work.

The options for the former case are:

1) implement phylink covering both the MAC and the external PCS PHY
2) implement phylink just for the MAC to PCS PHY connection but not SFPs, and implement SFP support separately in the PCS PHY driver.

Maybe phylink needs to split mac_pcs_get_state() so it can be supplied by a separate driver, or by the MAC driver as appropriate - but that brings with it other problems; phylink with a directly attached SFP considers the state of the link between the PCS PHY and the external device - not only speed but also interface mode for that part of the link. What you'd see in the mac_config() callback are interface modes for that part of the link, not between the MAC and the PCS PHY.

To change that would require reworking almost every driver that has already converted over to somehow remodel the built-in PCS and COMPHY as a separate PCS PHY for phylink. I'm not entirely clear whether that would work though.

--
RMK's Patch system: https://urldefense.proofpoint.com/v2/url?u=https-
3A__www.armlinux.org.uk_developer_patches_&d=DwIBAg&c=aUq983L2pue 2FqKFoP6PGHMJQyoJ7kl3s3GZ- _haXqY&r=BDdk1JtITE_JJ0519WwqU7IKF80Cw1i55lZOGqv2su8&m=ei26OYsu0 JYGaBZjJMg7WhXT8l_kdzu_QwlOu3RTUhY&s=YHxF2EUKwhleTZ- fO9lorELZWnn9kArzxliO1KM0uMc&e=

FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up

According to speedtest.net: 11.9Mbps down 500kbps up