From: Alexander Sverdlin <redacted>
of_mdio: Honour "status=disabled" property of device
Currently of_mdiobus_register() function registers all PHY devices,
independetly from their status property in device tree. According to
"ePAPR 1.1" spec, device should only be registered if there is no
"status" property, or it has "ok" (or "okay") value (see
of_device_is_available()). In case of "platform devices",
of_platform_device_create_pdata() checks for "status" and ensures
that disabled devices are not pupulated. But such check for MDIO buses
was missing until now. Fix it.
Signed-off-by: Alexander Sverdlin <redacted>
---
--- linux.orig/drivers/of/of_mdio.c
+++ linux/drivers/of/of_mdio.c
@@ -53,7 +53,7 @@ int of_mdiobus_register(struct mii_bus *
return rc;
/* Loop over the child nodes and register a phy_device for each one */
- for_each_child_of_node(np, child) {
+ for_each_available_child_of_node(np, child) {
const __be32 *paddr;
u32 addr;
int len;
On 11/29/2012 01:45 AM, Alexander Sverdlin wrote:
From: Alexander Sverdlin <redacted>
of_mdio: Honour "status=disabled" property of device
Currently of_mdiobus_register() function registers all PHY devices,
independetly from their status property in device tree. According to
"ePAPR 1.1" spec, device should only be registered if there is no
"status" property, or it has "ok" (or "okay") value (see
of_device_is_available()). In case of "platform devices",
of_platform_device_create_pdata() checks for "status" and ensures
that disabled devices are not pupulated. But such check for MDIO buses
was missing until now. Fix it.
Signed-off-by: Alexander Sverdlin <redacted>
---
Applied.
Rob
quoted hunk
--- linux.orig/drivers/of/of_mdio.c
+++ linux/drivers/of/of_mdio.c
@@ -53,7 +53,7 @@ int of_mdiobus_register(struct mii_bus *
return rc;
/* Loop over the child nodes and register a phy_device for each one */
- for_each_child_of_node(np, child) {
+ for_each_available_child_of_node(np, child) {
const __be32 *paddr;
u32 addr;
int len;_______________________________________________
devicetree-discuss mailing list
devicetree-discuss@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/devicetree-discuss
On Thu, 29 Nov 2012 08:45:20 +0100, Alexander Sverdlin [off-list ref] wrote:
From: Alexander Sverdlin <redacted>
of_mdio: Honour "status=disabled" property of device
Currently of_mdiobus_register() function registers all PHY devices,
independetly from their status property in device tree. According to
"ePAPR 1.1" spec, device should only be registered if there is no
"status" property, or it has "ok" (or "okay") value (see
of_device_is_available()). In case of "platform devices",
of_platform_device_create_pdata() checks for "status" and ensures
that disabled devices are not pupulated. But such check for MDIO buses
was missing until now. Fix it.
Signed-off-by: Alexander Sverdlin <redacted>
Applied, thanks.
g.
quoted hunk
---
--- linux.orig/drivers/of/of_mdio.c
+++ linux/drivers/of/of_mdio.c
@@ -53,7 +53,7 @@ int of_mdiobus_register(struct mii_bus *
return rc;
/* Loop over the child nodes and register a phy_device for each one */
- for_each_child_of_node(np, child) {
+ for_each_available_child_of_node(np, child) {
const __be32 *paddr;
u32 addr;
int len;
--
Grant Likely, B.Sc, P.Eng.
Secret Lab Technologies, Ltd.