Skip to contentSkip to conversation

of_mdio: fix phy interrupt passing

PATCH Applied Ben Dooks 4 participants 7 messages Feb 18, 2014 – Feb 19, 2014

Subsystems (4)
lore
Share & tools
lore

Ben Dooks

author
#0 lore raw

PATCH v2 1/1 of_mdio: fix phy interrupt passing

The of_mdiobus_register_phy() is not setting phy->irq thus causing some drivers to incorrectly assume that the PHY does not have an IRQ associated with it. Not only do some drivers report no IRQ they do not install an interrupt handler for the PHY.

Simplify the code setting irq and set the phy->irq at the same time so that we cover the following issues, which should cover all the cases the code will find:

- Set phy->irq if node has irq property and mdio->irq is NULL
- Set phy->irq if node has no irq and mdio->irq is not NULL
- Leave phy->irq as PHY_POLL default if none of the above

This fixes the issue:
net eth0: attached PHY 1 (IRQ -1) to driver Micrel KSZ8041RNLI

to the correct:
net eth0: attached PHY 1 (IRQ 416) to driver Micrel KSZ8041RNLI

---
Since v1:
- Updated phy->irq setting code
- Deal with issue if mdio->irq array NULL

Notes:

It was discussed if this should be two patches, but we end up making enough changes in #2 the same area as #1 means that we basically rewrite #1 in #2. So keep it as 1 patch.

---
 drivers/of/of_mdio.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
drivers/of/of_mdio.c +9 -7
index 875b7b6..46d95fc 100644
@@ -44,7 +44,7 @@ static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi
 {
 	struct phy_device *phy;
 	bool is_c45;
-	int rc, prev_irq;
+	int rc;
 	u32 max_speed = 0;
 
 	is_c45 = of_device_is_compatible(child,
@@ -54,12 +54,14 @@ static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi
 	if (!phy || IS_ERR(phy))
 		return 1;
 
-	if (mdio->irq) {
-		prev_irq = mdio->irq[addr];
-		mdio->irq[addr] =
-			irq_of_parse_and_map(child, 0);
-		if (!mdio->irq[addr])
-			mdio->irq[addr] = prev_irq;
+	rc = irq_of_parse_and_map(child, 0);
+	if (rc > 0) {
+		phy->irq = rc;
+		if (mdio->irq)
+			mdio->irq[addr] = rc;
+	} else {
+		if (mdio->irq)
+			phy->irq = mdio->irq[addr];
 	}
 
 	/* Associate the OF node with the device structure so it
Signature · 1 line
1.8.5.3
#1 lore raw

Hello.

Quoted 39 lines · Ben Dooks

The of_mdiobus_register_phy() is not setting phy->irq thus causing some drivers to incorrectly assume that the PHY does not have an IRQ associated with it. Not only do some drivers report no IRQ they do not install an interrupt handler for the PHY.

Simplify the code setting irq and set the phy->irq at the same time so that we cover the following issues, which should cover all the cases the code will find:

- Set phy->irq if node has irq property and mdio->irq is NULL
- Set phy->irq if node has no irq and mdio->irq is not NULL
- Leave phy->irq as PHY_POLL default if none of the above

This fixes the issue:
net eth0: attached PHY 1 (IRQ -1) to driver Micrel KSZ8041RNLI

to the correct:
net eth0: attached PHY 1 (IRQ 416) to driver Micrel KSZ8041RNLI

---
Since v1:
- Updated phy->irq setting code
- Deal with issue if mdio->irq array NULL

Notes:

It was discussed if this should be two patches, but we end up making enough changes in #2 the same area as #1 means that we basically rewrite #1 in #2. So keep it as 1 patch.

---
  drivers/of/of_mdio.c | 16 +++++++++-------
  1 file changed, 9 insertions(+), 7 deletions(-)
drivers/of/of_mdio.c +0 -0
index 875b7b6..46d95fc 100644

[...]

Quoted 7 lines
drivers/of/of_mdio.c +0 -3
@@ -54,12 +54,14 @@ static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi
  	if (!phy || IS_ERR(phy))
  		return 1;
 
-	if (mdio->irq) {
-		prev_irq = mdio->irq[addr];
-		mdio->irq[addr] =

drivers/of/of_mdio.c static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi · line 63

Quoted 10 lines
drivers/of/of_mdio.c +7 -3
-			irq_of_parse_and_map(child, 0);
-		if (!mdio->irq[addr])
-			mdio->irq[addr] = prev_irq;
+	rc = irq_of_parse_and_map(child, 0);
+	if (rc > 0) {
+		phy->irq = rc;
+		if (mdio->irq)
+			mdio->irq[addr] = rc;
+	} else {
+		if (mdio->irq)

You could write that as *else if* and save a line of code...

drivers/of/of_mdio.c static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi · line 65

Quoted 2 lines
drivers/of/of_mdio.c +1 -0
+			phy->irq = mdio->irq[addr];
  	}

WBR, Sergei

Reviewed-by on v2 1/1
#2 lore raw
In reply to Ben Dooks
Quoted 13 lines · Ben Dooks

The of_mdiobus_register_phy() is not setting phy->irq thus causing some drivers to incorrectly assume that the PHY does not have an IRQ associated with it. Not only do some drivers report no IRQ they do not install an interrupt handler for the PHY.

Simplify the code setting irq and set the phy->irq at the same time so that we cover the following issues, which should cover all the cases the code will find:

- Set phy->irq if node has irq property and mdio->irq is NULL
- Set phy->irq if node has no irq and mdio->irq is not NULL
- Leave phy->irq as PHY_POLL default if none of the above

Quoted 7 lines

This fixes the issue:
net eth0: attached PHY 1 (IRQ -1) to driver Micrel KSZ8041RNLI

to the correct:
net eth0: attached PHY 1 (IRQ 416) to driver Micrel KSZ8041RNLI

Looks okay to me

Quoted 52 lines

---
Since v1:
- Updated phy->irq setting code
- Deal with issue if mdio->irq array NULL

Notes:

It was discussed if this should be two patches, but we end up making enough changes in #2 the same area as #1 means that we basically rewrite #1 in #2. So keep it as 1 patch.

---
 drivers/of/of_mdio.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
drivers/of/of_mdio.c +9 -7
index 875b7b6..46d95fc 100644
@@ -44,7 +44,7 @@ static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi
 {
 	struct phy_device *phy;
 	bool is_c45;
-	int rc, prev_irq;
+	int rc;
 	u32 max_speed = 0;
 
 	is_c45 = of_device_is_compatible(child,
@@ -54,12 +54,14 @@ static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi
 	if (!phy || IS_ERR(phy))
 		return 1;
 
-	if (mdio->irq) {
-		prev_irq = mdio->irq[addr];
-		mdio->irq[addr] =
-			irq_of_parse_and_map(child, 0);
-		if (!mdio->irq[addr])
-			mdio->irq[addr] = prev_irq;
+	rc = irq_of_parse_and_map(child, 0);
+	if (rc > 0) {
+		phy->irq = rc;
+		if (mdio->irq)
+			mdio->irq[addr] = rc;
+	} else {
+		if (mdio->irq)
+			phy->irq = mdio->irq[addr];
 	}
 
 	/* Associate the OF node with the device structure so it
Signature · 1 line
1.8.5.3
maintainer Applied
#3 lore raw

From: Grant Likely <[email protected]>
Date: Tue, 18 Feb 2014 16:15:51 +0000

Quoted 17 lines · Ben Dooks
Quoted 16 lines

The of_mdiobus_register_phy() is not setting phy->irq thus causing some drivers to incorrectly assume that the PHY does not have an IRQ associated with it. Not only do some drivers report no IRQ they do not install an interrupt handler for the PHY.

Simplify the code setting irq and set the phy->irq at the same time so that we cover the following issues, which should cover all the cases the code will find:

- Set phy->irq if node has irq property and mdio->irq is NULL
- Set phy->irq if node has no irq and mdio->irq is not NULL
- Leave phy->irq as PHY_POLL default if none of the above

This fixes the issue:
net eth0: attached PHY 1 (IRQ -1) to driver Micrel KSZ8041RNLI

Quoted 8 lines
Quoted 4 lines

to the correct:
net eth0: attached PHY 1 (IRQ 416) to driver Micrel KSZ8041RNLI

Looks okay to me

Applied, thanks.

#4 lore raw

Hello.

Quoted 19 lines · David Miller
Quoted 18 lines
Quoted 18 lines

The of_mdiobus_register_phy() is not setting phy->irq thus causing some drivers to incorrectly assume that the PHY does not have an IRQ associated with it. Not only do some drivers report no IRQ they do not install an interrupt handler for the PHY.

Simplify the code setting irq and set the phy->irq at the same time so that we cover the following issues, which should cover all the cases the code will find:

- Set phy->irq if node has irq property and mdio->irq is NULL
- Set phy->irq if node has no irq and mdio->irq is not NULL
- Leave phy->irq as PHY_POLL default if none of the above

This fixes the issue:
net eth0: attached PHY 1 (IRQ -1) to driver Micrel KSZ8041RNLI

to the correct:
net eth0: attached PHY 1 (IRQ 416) to driver Micrel KSZ8041RNLI

Quoted 7 lines
Quoted 5 lines
Quoted 1 line

Looks okay to me

Applied, thanks.

So, you decided to ignore my comment about unneeded *else* branch?

WBR, Sergei

maintainer
#5 lore raw

From: Sergei Shtylyov <[email protected]>
Date: Wed, 19 Feb 2014 17:27:39 +0400

Quoted 1 line

So, you decided to ignore my comment about unneeded *else* branch?

Not ignored, just figured that a follow-up patch could fix it.

#6 lore raw
In reply to Ben Dooks

Hello.

Quoted 20 lines · Ben Dooks

The of_mdiobus_register_phy() is not setting phy->irq thus causing some drivers to incorrectly assume that the PHY does not have an IRQ associated with it. Not only do some drivers report no IRQ they do not install an interrupt handler for the PHY.

Simplify the code setting irq and set the phy->irq at the same time so that we cover the following issues, which should cover all the cases the code will find:

- Set phy->irq if node has irq property and mdio->irq is NULL
- Set phy->irq if node has no irq and mdio->irq is not NULL
- Leave phy->irq as PHY_POLL default if none of the above

This fixes the issue:
net eth0: attached PHY 1 (IRQ -1) to driver Micrel KSZ8041RNLI

to the correct:
net eth0: attached PHY 1 (IRQ 416) to driver Micrel KSZ8041RNLI

[...]

Quoted 4 lines
drivers/of/of_mdio.c +0 -0
index 875b7b6..46d95fc 100644

[...]

Quoted 8 lines
drivers/of/of_mdio.c +0 -4
@@ -54,12 +54,14 @@ static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi
  	if (!phy || IS_ERR(phy))
  		return 1;
 
-	if (mdio->irq) {
-		prev_irq = mdio->irq[addr];
-		mdio->irq[addr] =
-			irq_of_parse_and_map(child, 0);

drivers/of/of_mdio.c static int of_mdiobus_register_phy(struct mii_bus *mdio, struct device_node *chi · line 64

Quoted 10 lines
drivers/of/of_mdio.c +8 -2
-		if (!mdio->irq[addr])
-			mdio->irq[addr] = prev_irq;
+	rc = irq_of_parse_and_map(child, 0);
+	if (rc > 0) {
+		phy->irq = rc;
+		if (mdio->irq)
+			mdio->irq[addr] = rc;
+	} else {
+		if (mdio->irq)
+			phy->irq = mdio->irq[addr];

I have now looked thru the of_mdio.c code and the code it calls.
This case seems to be cared about by phy_device_create() called earlier
via get_phy_device() call just above this hunk; in any case the value will be PHY_POLL. I don't think this branch is needed at all.

WBR, Sergei