Re: [RFC PATCH net] net: dsa: mt7530: move PLL setup out of port 6 pad configuration

From: Arınç ÜNAL
Date: Mon Mar 06 2023 - 08:20:26 EST


On 4.03.2023 15:54, arinc9.unal@xxxxxxxxx wrote:
From: Arınç ÜNAL <arinc.unal@xxxxxxxxxx>

Move the PLL setup of the MT7530 switch out of the pad configuration of
port 6 to mt7530_setup, after reset.

This fixes the improper initialisation of the switch when only port 5 is
used as a CPU port.

Add supported phy modes of port 5 on the PLL setup.

Remove now incorrect comment regarding P5 as GMAC5.

Fixes: b8f126a8d543 ("net-next: dsa: add dsa support for Mediatek MT7530 switch")
Fixes: 38f790a80560 ("net: dsa: mt7530: Add support for port 5")
Signed-off-by: Arınç ÜNAL <arinc.unal@xxxxxxxxxx>
---

I'm trying to mimic this change by Alexander [0] for the MT7530 switch.
This is already the case for MT7530 and MT7531 on the MediaTek ethernet
driver on U-Boot [1] [2].

[0] https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=42bc4fafe359ed6b73602b7a2dba0dd99588f8ce
[1] https://github.com/u-boot/u-boot/blob/a94ab561e2f49a80d8579930e840b810ab1a1330/drivers/net/mtk_eth.c#L729
[2] https://github.com/u-boot/u-boot/blob/a94ab561e2f49a80d8579930e840b810ab1a1330/drivers/net/mtk_eth.c#L903

There are some parts I couldn't figure out myself with my limited C
knowledge. I've pointed them out on the code. Vladimir, could you help?

There is a lot of code which is only needed for port 6 or trgmii on port 6,
but runs whether port 6 or trgmii is used or not. For now, the best I can
do is to fix port 5 so it works without port 6 being used.

The U-Boot driver seems to be much more organised so it could be taken as
a reference to sort out this DSA driver further.

Also, now that the pad setup for mt7530 is also moved somewhere else, can
we completely get rid of @pad_setup?

Arınç

---
drivers/net/dsa/mt7530.c | 46 +++++++++++++++++++++++++++++-----------
1 file changed, 34 insertions(+), 12 deletions(-)

diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c
index 0e99de26d159..fb20ce4f443e 100644
--- a/drivers/net/dsa/mt7530.c
+++ b/drivers/net/dsa/mt7530.c
@@ -395,9 +395,8 @@ mt7530_fdb_write(struct mt7530_priv *priv, u16 vid,
/* Setup TX circuit including relevant PAD and driving */
static int
-mt7530_pad_clk_setup(struct dsa_switch *ds, phy_interface_t interface)
+mt7530_pad_clk_setup(struct mt7530_priv *priv, phy_interface_t interface)
{
- struct mt7530_priv *priv = ds->priv;
u32 ncpo1, ssc_delta, trgint, i, xtal;
xtal = mt7530_read(priv, MT7530_MHWTRAP) & HWTRAP_XTAL_MASK;
@@ -409,9 +408,32 @@ mt7530_pad_clk_setup(struct dsa_switch *ds, phy_interface_t interface)
return -EINVAL;
}
+ /* Is setting trgint to 0 really needed for the !trgint check? */
+ trgint = 0;
+
+ /* FIXME: run this switch if p5 is defined on the devicetree */
+ /* and change interface to the phy-mode of port 5 */
+ switch (interface) {
+ case PHY_INTERFACE_MODE_GMII:
+ /* PLL frequency: 125MHz */
+ ncpo1 = 0x0c80;
+ break;
+ case PHY_INTERFACE_MODE_MII:
+ break;
+ case PHY_INTERFACE_MODE_RGMII:
+ /* PLL frequency: 125MHz */
+ ncpo1 = 0x0c80;
+ break;
+ default:
+ dev_err(priv->dev, "xMII interface %d not supported\n",
+ interface);
+ return -EINVAL;
+ }
+
+ /* FIXME: run this switch if p6 is defined on the devicetree */
+ /* and change interface to the phy-mode of port 6 */
switch (interface) {
case PHY_INTERFACE_MODE_RGMII:
- trgint = 0;
/* PLL frequency: 125MHz */
ncpo1 = 0x0c80;
break;
@@ -2172,7 +2194,11 @@ mt7530_setup(struct dsa_switch *ds)
SYS_CTRL_PHY_RST | SYS_CTRL_SW_RST |
SYS_CTRL_REG_RST);
- /* Enable Port 6 only; P5 as GMAC5 which currently is not supported */
+ /* Setup switch core pll */
+ /* FIXME: feed the phy-mode of port 5 and 6, if the ports are defined on the devicetree */
+ mt7530_pad_clk_setup(priv, interface);

Unlike mt7531_pll_setup, there are return codes on mt7530_pad_clk_setup so we may need to check for the return code here.

Arınç