On 11/22/21 11:36, Ramon Fried wrote:
On Mon, Nov 22, 2021 at 9:40 AM Michal Simek michal.simek@xilinx.com wrote:
On 11/21/21 20:11, Ramon Fried wrote:
On Thu, Nov 18, 2021 at 2:06 PM Michal Simek michal.simek@xilinx.com wrote:
There are flying two configurations around.
Can you rephrase ? I don't understand.
The first configuration is eth { phy-handle = <&phy0>; phy0: ethernet-phy@21 { ... }; };
The second is (based on mdio.yaml) eth { phy-handle = <&phy0>; mdio { phy0: ethernet-phy@21 { ... }; }; };
Xilinx dts files are using both ways. But the second case is not fully when you have one mdio bus shared with at least 2 phys where one phy is connected to another gem controller. This code is updating logic which finds which gen controller is mdio bus owner.
Please let me know if you want me to update commit message.
Thanks, Michal
The first (already supported) has phy as subnode of gem node. The second has mdio subnode (with mdio name) which has phy subnode.
This patch adds support for the second case where mdio subnode is found driver will look at its parent to find out which gem is handling MDIO bus.
Signed-off-by: Michal Simek michal.simek@xilinx.com
drivers/net/zynq_gem.c | 3 +++ 1 file changed, 3 insertions(+)
diff --git a/drivers/net/zynq_gem.c b/drivers/net/zynq_gem.c index 033021f1cbfc..91957757727d 100644 --- a/drivers/net/zynq_gem.c +++ b/drivers/net/zynq_gem.c @@ -804,6 +804,9 @@ static int zynq_gem_of_to_plat(struct udevice *dev) SPEED_1000);
parent = ofnode_get_parent(phandle_args.node);
if (ofnode_name_eq(parent, "mdio"))parent = ofnode_get_parent(parent);addr = ofnode_get_addr(parent); if (addr != FDT_ADDR_T_NONE) { debug("MDIO bus not found %s\n", dev->name);-- 2.33.1
"There are flying two configurations around." The above sentence is semantically incorrect, just fix that.
v2 sent.
Thanks, Michal