[PATCH] net: dsa: lantiq_gswip: Add 200ms assert delay

The delay is especially needed by the xRX300 and xRX330 SoCs. Without
this patch, some phys are sometimes not properly detected.

Fixes: a09d042b086202735c4ed64 ("net: dsa: lantiq: allow to use all GPHYs on xRX300 and xRX330")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Tested-by: Aleksander Jan Bajkowski <redacted> # tested on DWR966, HH5A
---
 drivers/net/dsa/lantiq_gswip.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/net/dsa/lantiq_gswip.c b/drivers/net/dsa/lantiq_gswip.c
index e78026ef6d8c..9eaf013b82a3 100644
--- a/drivers/net/dsa/lantiq_gswip.c
+++ b/drivers/net/dsa/lantiq_gswip.c

@@ -1884,6 +1884,12 @@ static int gswip_gphy_fw_load(struct gswip_priv *priv, struct gswip_gphy_fw *gph
 
 	reset_control_assert(gphy_fw->reset);
 
+	/* The vendor BSP uses a 200ms delay after asserting the reset line.
+	 * Without this some users are observing that the PHY is not coming up
+	 * on the MDIO bus.
+	 */
+	msleep(200);
+
 	ret = request_firmware(&fw, gphy_fw->fw_name, dev);
 	if (ret) {
 		dev_err(dev, "failed to load firmware: %s, error: %i\n",

-- 
2.30.2

On Tue, Aug 17, 2021 at 09:32:07PM +0200, Aleksander Jan Bajkowski wrote:

The delay is especially needed by the xRX300 and xRX330 SoCs. Without
this patch, some phys are sometimes not properly detected.

Fixes: a09d042b086202735c4ed64 ("net: dsa: lantiq: allow to use all GPHYs on xRX300 and xRX330")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Tested-by: Aleksander Jan Bajkowski <redacted> # tested on DWR966, HH5A
---

Generally the convention is:

From: Patch Author <redacted>

Commit description

Signed-off-by: Patch Author <redacted>
Signed-off-by: Patch Carrier 1 <redacted>
Signed-off-by: Patch Carrier 2 <redacted>
Signed-off-by: Patch Carrier 3 <redacted>
Signed-off-by: Patch Submitter <redacted>

This patch is clearly not following this model for more than one reason.

On Tue, Aug 17, 2021 at 10:44:48PM +0300, Vladimir Oltean wrote:

On Tue, Aug 17, 2021 at 09:32:07PM +0200, Aleksander Jan Bajkowski wrote:

quoted

The delay is especially needed by the xRX300 and xRX330 SoCs. Without
this patch, some phys are sometimes not properly detected.

Fixes: a09d042b086202735c4ed64 ("net: dsa: lantiq: allow to use all GPHYs on xRX300 and xRX330")
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Tested-by: Aleksander Jan Bajkowski <redacted> # tested on DWR966, HH5A
---

Generally the convention is:

From: Patch Author <redacted>

Commit description

Signed-off-by: Patch Author <redacted>
Signed-off-by: Patch Carrier 1 <redacted>
Signed-off-by: Patch Carrier 2 <redacted>
Signed-off-by: Patch Carrier 3 <redacted>
Signed-off-by: Patch Submitter <redacted>

This patch is clearly not following this model for more than one reason.

Let's not even talk about the kilometer-long commit sha1sum.
This is not even my pet peeve, if this patch gets merged as-is you'll
get an email titled "linux-next: Fixes tag needs some work in the net tree"
(google it if you want examples).

Stick this in your ~/.gitconfig and thank me later:

[core]
	abbrev = 12
[pretty]
	fixes = Fixes: %h (\"%s\")

Now run:

git show a09d042b086202735c4ed64 --pretty=fixes
Fixes: a09d042b0862 ("net: dsa: lantiq: allow to use all GPHYs on xRX300 and xRX330")

Voila!