i2c: mux: Propagate firmware nodes to channel adapters

PATCH Applied Ahmad Byagowi 4 participants 10 messages Sep 8 – Sep 16

Subsystems (3)

I2C subsystem host drivers I2C I2C muxes

lore

Share & tools

lore

Ahmad Byagowi

author

2026-09-08 06:26 UTC #0 lore raw

v10 cover i2c: mux: Propagate firmware nodes to channel adapters

Device Tree channel nodes are associated with the adapters created by i2c-mux, but equivalent firmware-node descriptions are not. Convert the existing lookup to generic firmware-node operations so software-node and ACPI descriptions can use the same property traversal.

This is the remaining I2C prerequisite split out of the Time Card series. The channel-node leak fix from v7 was applied to i2c-next as commit 385c7af4e3b9 ("i2c: mux: Fix channel node leak on adapter add failure"), so it is no longer included here. The first patch factors the Device Tree lookup into a helper without changing behavior. The second converts the helper and node lifetime handling to generic firmware-node operations.

The existing acpi_preset_companion() call remains in place after the generic lookup. A separate reference to the node returned by that lookup is retained because ACPI companion setup may replace the adapter's primary firmware node. The dependent ptp_ocp board profiles will be resent to net-next after the LED and I2C prerequisites land.

Changes since v9:
- Declare dev_node, mux_node, and child on separate lines in both patches, as requested by Peter Rosin and agreed by Andy Shevchenko.

The broader __free(fwnode_handle) cleanup remains a separate follow-up.

Validation:
- git diff --check
- strict checkpatch with an 80-column limit
- x86-64 GCC W=1 object build of drivers/i2c/i2c-mux.o
- mailbox replay onto the declared base with an identical resulting tree

v9: https://lore.kernel.org/r/[email protected]/

Ahmad Byagowi (2):
i2c: mux: Factor out channel node lookup
i2c: mux: Propagate firmware nodes to channel adapters

 drivers/i2c/i2c-mux.c | 98 +++++++++++++++++++++++--------------------
 1 file changed, 52 insertions(+), 46 deletions(-)

base-commit: 1f3e66348d2527c31b9bedb0b4d9e28bdad9bc83

Signature · 1 line

2.50.1 (Apple Git-155)

Ahmad Byagowi

author

2026-09-08 06:26 UTC #1 lore raw

PATCH v10 1/2 i2c: mux: Factor out channel node lookup

Move the existing Device Tree channel-node lookup into a helper in preparation for using generic firmware-node operations.

This is a pure refactoring with no functional change.

Signed-off-by Ahmad Byagowi <[email protected]>

---
 drivers/i2c/i2c-mux.c | 84 +++++++++++++++++++++++--------------------
 1 file changed, 46 insertions(+), 38 deletions(-)

drivers/i2c/i2c-mux.c +46 ~~-38~~

index 68a4c34b5987..d207f0e81807 100644
@@ -264,6 +264,51 @@ static const struct i2c_lock_operations i2c_parent_lock_ops = {
 	.unlock_bus =  i2c_parent_unlock_bus,
 };
 
+static struct device_node *
+i2c_mux_get_channel_node(struct i2c_mux_core *muxc, u32 chan_id)
+{
+	struct device_node *dev_node;
+	struct device_node *mux_node;
+	struct device_node *child;
+	u32 reg;
+	int ret;
+
+	dev_node = dev_of_node(muxc->dev);
+	if (!dev_node)
+		return NULL;
+
+	if (muxc->arbitrator)
+		mux_node = of_get_child_by_name(dev_node, "i2c-arb");
+	else if (muxc->gate)
+		mux_node = of_get_child_by_name(dev_node, "i2c-gate");
+	else
+		mux_node = of_get_child_by_name(dev_node, "i2c-mux");
+
+	if (mux_node) {
+		/* A "reg" property indicates an old-style DT entry */
+		if (!of_property_read_u32(mux_node, "reg", &reg)) {
+			of_node_put(mux_node);
+			mux_node = NULL;
+		}
+	}
+
+	if (!mux_node)
+		mux_node = of_node_get(dev_node);
+	else if (muxc->arbitrator || muxc->gate)
+		return mux_node;
+
+	for_each_child_of_node(mux_node, child) {
+		ret = of_property_read_u32(child, "reg", &reg);
+		if (ret)
+			continue;
+		if (chan_id == reg)
+			break;
+	}
+
+	of_node_put(mux_node);
+	return child;
+}
+
 int i2c_mux_add_adapter(struct i2c_mux_core *muxc,
 			u32 force_nr, u32 chan_id)
 {

@@ -327,44 +372,7 @@ int i2c_mux_add_adapter(struct i2c_mux_core *muxc,
 	 * Try to populate the mux adapter's of_node, expands to
 	 * nothing if !CONFIG_OF.
 	 */
-	if (muxc->dev->of_node) {
-		struct device_node *dev_node = muxc->dev->of_node;
-		struct device_node *mux_node, *child = NULL;
-		u32 reg;
-
-		if (muxc->arbitrator)
-			mux_node = of_get_child_by_name(dev_node, "i2c-arb");
-		else if (muxc->gate)
-			mux_node = of_get_child_by_name(dev_node, "i2c-gate");
-		else
-			mux_node = of_get_child_by_name(dev_node, "i2c-mux");
-
-		if (mux_node) {
-			/* A "reg" property indicates an old-style DT entry */
-			if (!of_property_read_u32(mux_node, "reg", &reg)) {
-				of_node_put(mux_node);
-				mux_node = NULL;
-			}
-		}
-
-		if (!mux_node)
-			mux_node = of_node_get(dev_node);
-		else if (muxc->arbitrator || muxc->gate)
-			child = of_node_get(mux_node);
-
-		if (!child) {
-			for_each_child_of_node(mux_node, child) {
-				ret = of_property_read_u32(child, "reg", &reg);
-				if (ret)
-					continue;
-				if (chan_id == reg)
-					break;
-			}
-		}
-
-		priv->adap.dev.of_node = child;
-		of_node_put(mux_node);
-	}
+	priv->adap.dev.of_node = i2c_mux_get_channel_node(muxc, chan_id);
 
 	/*
 	 * Associate the mux channel with an ACPI node.

Signature · 1 line

2.50.1 (Apple Git-155)

Andy Shevchenko

Reviewed-by on v10 1/2

2026-09-08 10:39 UTC #2 lore raw

Re: [PATCH v10 1/2] i2c: mux: Factor out channel node lookup

Quoted 4 lines · Ahmad Byagowi

Move the existing Device Tree channel-node lookup into a helper in preparation for using generic firmware-node operations.
This is a pure refactoring with no functional change.

Reviewed-by Andy Shevchenko <[email protected]>

Signature · 2 lines

With Best Regards,
Andy Shevchenko

Peter Rosin

maintainer Acked-by on v10 1/2

2026-09-10 13:44 UTC #3 lore raw

Re: [PATCH v10 1/2] i2c: mux: Factor out channel node lookup

In reply to Ahmad Byagowi

Quoted 6 lines · Ahmad Byagowi

Move the existing Device Tree channel-node lookup into a helper in preparation for using generic firmware-node operations.
This is a pure refactoring with no functional change.
Signed-off-by Ahmad Byagowi <[email protected]>

Acked-by Peter Rosin <[email protected]>

Cheers,
Peter

Ahmad Byagowi

author

2026-09-08 06:26 UTC #4 lore raw

PATCH v10 2/2 i2c: mux: Propagate firmware nodes to channel adapters

In reply to Ahmad Byagowi

Device Tree channel nodes are associated with the adapters created by i2c-mux, but equivalent firmware-node descriptions are not.

Use generic firmware-node operations for the existing channel lookup and associate the returned node with the adapter. Do not restrict the lookup by firmware-node type, so Device Tree, software nodes, and ACPI descriptions all follow the same property traversal. The existing acpi_preset_companion() call remains in place for the standard ACPI channel association.

Keep a separate reference to the node returned by the generic lookup because acpi_preset_companion() may replace the device's primary firmware node. Release the saved reference after adapter deletion.

Signed-off-by Ahmad Byagowi <[email protected]>

---
 drivers/i2c/i2c-mux.c | 46 +++++++++++++++++++++----------------------
 1 file changed, 22 insertions(+), 24 deletions(-)

drivers/i2c/i2c-mux.c +22 ~~-24~~

index d207f0e81807..53428098ddc7 100644
@@ -24,7 +24,7 @@
 #include <linux/i2c-mux.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
-#include <linux/of.h>
+#include <linux/property.h>
 #include <linux/slab.h>
 #include <linux/sysfs.h>

@@ -33,6 +33,7 @@ struct i2c_mux_priv {
 	struct i2c_adapter adap;
 	struct i2c_algorithm algo;
 	struct i2c_mux_core *muxc;
+	struct fwnode_handle *channel_node;
 	u32 chan_id;
 };

@@ -264,48 +265,48 @@ static const struct i2c_lock_operations i2c_parent_lock_ops = {
 	.unlock_bus =  i2c_parent_unlock_bus,
 };
 
-static struct device_node *
+static struct fwnode_handle *
 i2c_mux_get_channel_node(struct i2c_mux_core *muxc, u32 chan_id)
 {
-	struct device_node *dev_node;
-	struct device_node *mux_node;
-	struct device_node *child;
+	struct fwnode_handle *dev_node;
+	struct fwnode_handle *mux_node;
+	struct fwnode_handle *child;
 	u32 reg;
 	int ret;
 
-	dev_node = dev_of_node(muxc->dev);
+	dev_node = dev_fwnode(muxc->dev);
 	if (!dev_node)
 		return NULL;
 
 	if (muxc->arbitrator)
-		mux_node = of_get_child_by_name(dev_node, "i2c-arb");
+		mux_node = fwnode_get_named_child_node(dev_node, "i2c-arb");
 	else if (muxc->gate)
-		mux_node = of_get_child_by_name(dev_node, "i2c-gate");
+		mux_node = fwnode_get_named_child_node(dev_node, "i2c-gate");
 	else
-		mux_node = of_get_child_by_name(dev_node, "i2c-mux");
+		mux_node = fwnode_get_named_child_node(dev_node, "i2c-mux");
 
 	if (mux_node) {
-		/* A "reg" property indicates an old-style DT entry */
-		if (!of_property_read_u32(mux_node, "reg", &reg)) {
-			of_node_put(mux_node);
+		/* A "reg" property indicates an old-style firmware entry. */
+		if (!fwnode_property_read_u32(mux_node, "reg", &reg)) {
+			fwnode_handle_put(mux_node);
 			mux_node = NULL;
 		}
 	}
 
 	if (!mux_node)
-		mux_node = of_node_get(dev_node);
+		mux_node = fwnode_handle_get(dev_node);
 	else if (muxc->arbitrator || muxc->gate)
 		return mux_node;
 
-	for_each_child_of_node(mux_node, child) {
-		ret = of_property_read_u32(child, "reg", &reg);
+	fwnode_for_each_child_node(mux_node, child) {
+		ret = fwnode_property_read_u32(child, "reg", &reg);
 		if (ret)
 			continue;
 		if (chan_id == reg)
 			break;
 	}
 
-	of_node_put(mux_node);
+	fwnode_handle_put(mux_node);
 	return child;
 }

@@ -368,11 +369,9 @@ int i2c_mux_add_adapter(struct i2c_mux_core *muxc,
 	else
 		priv->adap.lock_ops = &i2c_parent_lock_ops;
 
-	/*
-	 * Try to populate the mux adapter's of_node, expands to
-	 * nothing if !CONFIG_OF.
-	 */
-	priv->adap.dev.of_node = i2c_mux_get_channel_node(muxc, chan_id);
+	/* Associate the mux adapter with its channel firmware node. */
+	priv->channel_node = i2c_mux_get_channel_node(muxc, chan_id);
+	device_set_node(&priv->adap.dev, priv->channel_node);
 
 	/*
 	 * Associate the mux channel with an ACPI node.

@@ -416,7 +415,7 @@ int i2c_mux_add_adapter(struct i2c_mux_core *muxc,
 	return 0;
 
 err_free_priv:
-	of_node_put(priv->adap.dev.of_node);
+	fwnode_handle_put(priv->channel_node);
 	kfree(priv);
 	return ret;
 }

@@ -429,7 +428,6 @@ void i2c_mux_del_adapters(struct i2c_mux_core *muxc)
 	while (muxc->num_adapters) {
 		struct i2c_adapter *adap = muxc->adapter[--muxc->num_adapters];
 		struct i2c_mux_priv *priv = adap->algo_data;
-		struct device_node *np = adap->dev.of_node;
 
 		muxc->adapter[muxc->num_adapters] = NULL;

@@ -439,7 +437,7 @@ void i2c_mux_del_adapters(struct i2c_mux_core *muxc)
 
 		sysfs_remove_link(&priv->adap.dev.kobj, "mux_device");
 		i2c_del_adapter(adap);
-		of_node_put(np);
+		fwnode_handle_put(priv->channel_node);
 		kfree(priv);
 	}
 }

Signature · 1 line

2.50.1 (Apple Git-155)

Andy Shevchenko

Reviewed-by on v10 2/2

2026-09-08 10:43 UTC #5 lore raw

Quoted 10 lines · Ahmad Byagowi

Device Tree channel nodes are associated with the adapters created by i2c-mux, but equivalent firmware-node descriptions are not.
Use generic firmware-node operations for the existing channel lookup and associate the returned node with the adapter. Do not restrict the lookup by firmware-node type, so Device Tree, software nodes, and ACPI descriptions all follow the same property traversal. The existing acpi_preset_companion() call remains in place for the standard ACPI channel association.

Quoted 3 lines

Keep a separate reference to the node returned by the generic lookup because acpi_preset_companion() may replace the device's primary firmware node. Release the saved reference after adapter deletion.

Reviewed-by Andy Shevchenko <[email protected]>

...

drivers/i2c/i2c-mux.c static const struct i2c_lock_operations i2c_parent_lock_ops = { · line 273

Quoted 6 lines

drivers/i2c/i2c-mux.c +3 -3

-	struct device_node *dev_node;
-	struct device_node *mux_node;
-	struct device_node *child;
+	struct fwnode_handle *dev_node;
+	struct fwnode_handle *mux_node;
+	struct fwnode_handle *child;

When the local variable changes the type (usually this is related to the pointers) the best practice is to rename. In such a case it will avoid the cases when some API accepts void * and the change would not have been noticed until run-time crash or subtle issues.

I hope here it's just a mechanical stuff and we don't have side effects as per above.

Signature · 2 lines

With Best Regards,
Andy Shevchenko

Ahmad Byagowi

author

2026-09-08 13:12 UTC #6 lore raw

Andy,

Quoted 2 lines

I hope here it's just a mechanical stuff and we don't have side effects as per above.

Thanks for reviewing both patches and for the advice about renaming variables when their types change.

I checked all uses of dev_node, mux_node, and child in the converted helper. They use firmware-node APIs throughout, with no remaining OF calls or void * consumers for those pointers. I did not find a type-mismatch issue in this conversion.

Regards,
Ahmad

Peter Rosin

maintainer Acked-by Peter Rosin on v10 2/2

2026-09-10 13:44 UTC #7 lore raw

In reply to Ahmad Byagowi

Quoted 10 lines · Ahmad Byagowi

Device Tree channel nodes are associated with the adapters created by i2c-mux, but equivalent firmware-node descriptions are not.
Use generic firmware-node operations for the existing channel lookup and associate the returned node with the adapter. Do not restrict the lookup by firmware-node type, so Device Tree, software nodes, and ACPI descriptions all follow the same property traversal. The existing acpi_preset_companion() call remains in place for the standard ACPI channel association.

Quoted 5 lines

Keep a separate reference to the node returned by the generic lookup because acpi_preset_companion() may replace the device's primary firmware node. Release the saved reference after adapter deletion.
Signed-off-by Ahmad Byagowi <[email protected]>

Acked-by Peter Rosin <[email protected]>

Cheers,
Peter

Peter Rosin

maintainer

2026-09-10 13:47 UTC #8 lore raw

In reply to Ahmad Byagowi

Quoted 30 lines · Ahmad Byagowi

Device Tree channel nodes are associated with the adapters created by i2c-mux, but equivalent firmware-node descriptions are not. Convert the existing lookup to generic firmware-node operations so software-node and ACPI descriptions can use the same property traversal.
This is the remaining I2C prerequisite split out of the Time Card series. The channel-node leak fix from v7 was applied to i2c-next as commit 385c7af4e3b9 ("i2c: mux: Fix channel node leak on adapter add failure"), so it is no longer included here. The first patch factors the Device Tree lookup into a helper without changing behavior. The second converts the helper and node lifetime handling to generic firmware-node operations.
The existing acpi_preset_companion() call remains in place after the generic lookup. A separate reference to the node returned by that lookup is retained because ACPI companion setup may replace the adapter's primary firmware node. The dependent ptp_ocp board profiles will be resent to net-next after the LED and I2C prerequisites land.
Changes since v9:
- Declare dev_node, mux_node, and child on separate lines in both patches, as requested by Peter Rosin and agreed by Andy Shevchenko.
The broader __free(fwnode_handle) cleanup remains a separate follow-up.
Validation:
- git diff --check
- strict checkpatch with an 80-column limit
- x86-64 GCC W=1 object build of drivers/i2c/i2c-mux.o
- mailbox replay onto the declared base with an identical resulting tree

Quoted 8 lines

v9: https://lore.kernel.org/r/[email protected]/
Ahmad Byagowi (2):
i2c: mux: Factor out channel node lookup
i2c: mux: Propagate firmware nodes to channel adapters
 drivers/i2c/i2c-mux.c | 98 +++++++++++++++++++++++--------------------
 1 file changed, 52 insertions(+), 46 deletions(-)

Thanks!

Andi, you can pick this series now, thanks.

Cheers,
Peter

Andi Shyti

maintainer Applied

2026-09-16 22:07 UTC #9 lore raw

In reply to Ahmad Byagowi

Hi Ahmad,

Quoted 3 lines

Ahmad Byagowi (2):
i2c: mux: Factor out channel node lookup
i2c: mux: Propagate firmware nodes to channel adapters

pushed to i2c/i2c.

Thanks,
Andi