On Thu, Nov 03, 2016 at 09:31:21AM +0100, Rafal Milecki wrote:

On 11/02/2016 06:08 PM, Jon Mason wrote:

quoted

Change the bgmac driver to allow for phy's defined by the device tree

This is a late review, I know, sorry... :(

quoted

+static int bcma_phy_direct_connect(struct bgmac *bgmac)
+{
+	struct fixed_phy_status fphy_status = {
+		.link = 1,
+		.speed = SPEED_1000,
+		.duplex = DUPLEX_FULL,
+	};
+	struct phy_device *phy_dev;
+	int err;
+
+	phy_dev = fixed_phy_register(PHY_POLL, &fphy_status, -1, NULL);
+	if (!phy_dev || IS_ERR(phy_dev)) {
+		dev_err(bgmac->dev, "Failed to register fixed PHY device\n");
+		return -ENODEV;
+	}
+
+	err = phy_connect_direct(bgmac->net_dev, phy_dev, bgmac_adjust_link,
+				 PHY_INTERFACE_MODE_MII);
+	if (err) {
+		dev_err(bgmac->dev, "Connecting PHY failed\n");
+		return err;
+	}
+
+	return err;
+}

This bcma specific function looks exactly the same as...

quoted

+static int platform_phy_direct_connect(struct bgmac *bgmac)
+{
+	struct fixed_phy_status fphy_status = {
+		.link = 1,
+		.speed = SPEED_1000,
+		.duplex = DUPLEX_FULL,
+	};
+	struct phy_device *phy_dev;
+	int err;
+
+	phy_dev = fixed_phy_register(PHY_POLL, &fphy_status, -1, NULL);
+	if (!phy_dev || IS_ERR(phy_dev)) {
+		dev_err(bgmac->dev, "Failed to register fixed PHY device\n");
+		return -ENODEV;
+	}
+
+	err = phy_connect_direct(bgmac->net_dev, phy_dev, bgmac_adjust_link,
+				 PHY_INTERFACE_MODE_MII);
+	if (err) {
+		dev_err(bgmac->dev, "Connecting PHY failed\n");
+		return err;
+	}
+
+	return err;
+}

This one.

Would that make sense to keep bgmac_phy_connect_direct and just use it in
bcma/platform code?

Yes, I was having the same internal debate.  I hate the duplication of
code, but I really wanted to keep the PHY logic out of the bgmac.c
file.  Do you think it is acceptable to make this an inline function
in bgmac.h?

Thanks,
Jon