[PATCH v5 net 0/3] fixes for mtk_eth_soc

[PATCH v5 net 0/3] fixes for mtk_eth_soc

[Bjørn Mork]

Changes since v4:
 - use same field order for kernel-doc and code in patch 1
 - cc'ing full maintainer list from get_maintainer.pl

Changes since v3:
 - fill hole in struct mtk_pcs with new interface field
 - improved patch 2 commit message
 - added fixes tags
 - updated review tags
 
Changes since v2:
 - use "true" for boolean
 - fix SoB typo
 - updated tags

Changes since v1:
 - only power down on changes, fix from Russel
 - dropped bogus uncondional in-band patch
 - added pcs poll patch from Alexander
 - updated tags


Fix mtk_eth_soc sgmii configuration.

This has been tested on a MT7986 with a Maxlinear GPY211C phy
permanently attached to the second SoC mac.


Alexander Couzens (2):
  net: mediatek: sgmii: ensure the SGMII PHY is powered down on
    configuration
  mtk_sgmii: enable PCS polling to allow SFP work

Bjørn Mork (1):
  net: mediatek: sgmii: fix duplex configuration

 drivers/net/ethernet/mediatek/mtk_eth_soc.h |  4 +-
 drivers/net/ethernet/mediatek/mtk_sgmii.c   | 46 ++++++++++++++-------
 2 files changed, 35 insertions(+), 15 deletions(-)

-- 
2.30.2

[PATCH v5 net 1/3] net: mediatek: sgmii: ensure the SGMII PHY is powered down on configuration

[Bjørn Mork]

From: Alexander Couzens <lynxis@fe80.eu>

The code expect the PHY to be in power down which is only true after reset.
Allow changes of the SGMII parameters more than once.

Only power down when reconfiguring to avoid bouncing the link when there's
no reason to - based on code from Russell King.

There are cases when the SGMII_PHYA_PWD register contains 0x9 which
prevents SGMII from working. The SGMII still shows link but no traffic
can flow. Writing 0x0 to the PHYA_PWD register fix the issue. 0x0 was
taken from a good working state of the SGMII interface.

Fixes: 42c03844e93d ("net-next: mediatek: add support for MediaTek MT7622 SoC")
Suggested-by: Russell King (Oracle) <linux@armlinux.org.uk>
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
[ bmork: rebased and squashed into one patch ]
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Bjørn Mork <bjorn@mork.no>
---
 drivers/net/ethernet/mediatek/mtk_eth_soc.h |  2 ++
 drivers/net/ethernet/mediatek/mtk_sgmii.c   | 39 +++++++++++++++------
 2 files changed, 30 insertions(+), 11 deletions(-)

diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
index 18a50529ce7b..01a38f5145b1 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
@@ -1036,11 +1036,13 @@ struct mtk_soc_data {
  * @regmap:            The register map pointing at the range used to setup
  *                     SGMII modes
  * @ana_rgc3:          The offset refers to register ANA_RGC3 related to regmap
+ * @interface:         Currently configured interface mode
  * @pcs:               Phylink PCS structure
  */
 struct mtk_pcs {
 	struct regmap	*regmap;
 	u32             ana_rgc3;
+	phy_interface_t	interface;
 	struct phylink_pcs pcs;
 };
 
diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
index 5c286f2c9418..0a06995099cf 100644
--- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
+++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
@@ -43,11 +43,6 @@ static int mtk_pcs_config(struct phylink_pcs *pcs, unsigned int mode,
 	int advertise, link_timer;
 	bool changed, use_an;
 
-	if (interface == PHY_INTERFACE_MODE_2500BASEX)
-		rgc3 = RG_PHY_SPEED_3_125G;
-	else
-		rgc3 = 0;
-
 	advertise = phylink_mii_c22_pcs_encode_advertisement(interface,
 							     advertising);
 	if (advertise < 0)
@@ -88,9 +83,22 @@ static int mtk_pcs_config(struct phylink_pcs *pcs, unsigned int mode,
 		bmcr = 0;
 	}
 
-	/* Configure the underlying interface speed */
-	regmap_update_bits(mpcs->regmap, mpcs->ana_rgc3,
-			   RG_PHY_SPEED_3_125G, rgc3);
+	if (mpcs->interface != interface) {
+		/* PHYA power down */
+		regmap_update_bits(mpcs->regmap, SGMSYS_QPHY_PWR_STATE_CTRL,
+				   SGMII_PHYA_PWD, SGMII_PHYA_PWD);
+
+		if (interface == PHY_INTERFACE_MODE_2500BASEX)
+			rgc3 = RG_PHY_SPEED_3_125G;
+		else
+			rgc3 = 0;
+
+		/* Configure the underlying interface speed */
+		regmap_update_bits(mpcs->regmap, mpcs->ana_rgc3,
+				   RG_PHY_SPEED_3_125G, rgc3);
+
+		mpcs->interface = interface;
+	}
 
 	/* Update the advertisement, noting whether it has changed */
 	regmap_update_bits_check(mpcs->regmap, SGMSYS_PCS_ADVERTISE,
@@ -108,9 +116,17 @@ static int mtk_pcs_config(struct phylink_pcs *pcs, unsigned int mode,
 	regmap_update_bits(mpcs->regmap, SGMSYS_PCS_CONTROL_1,
 			   SGMII_AN_RESTART | SGMII_AN_ENABLE, bmcr);
 
-	/* Release PHYA power down state */
-	regmap_update_bits(mpcs->regmap, SGMSYS_QPHY_PWR_STATE_CTRL,
-			   SGMII_PHYA_PWD, 0);
+	/* Release PHYA power down state
+	 * Only removing bit SGMII_PHYA_PWD isn't enough.
+	 * There are cases when the SGMII_PHYA_PWD register contains 0x9 which
+	 * prevents SGMII from working. The SGMII still shows link but no traffic
+	 * can flow. Writing 0x0 to the PHYA_PWD register fix the issue. 0x0 was
+	 * taken from a good working state of the SGMII interface.
+	 * Unknown how much the QPHY needs but it is racy without a sleep.
+	 * Tested on mt7622 & mt7986.
+	 */
+	usleep_range(50, 100);
+	regmap_write(mpcs->regmap, SGMSYS_QPHY_PWR_STATE_CTRL, 0);
 
 	return changed;
 }
@@ -171,6 +187,7 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *r, u32 ana_rgc3)
 			return PTR_ERR(ss->pcs[i].regmap);
 
 		ss->pcs[i].pcs.ops = &mtk_pcs_ops;
+		ss->pcs[i].interface = PHY_INTERFACE_MODE_NA;
 	}
 
 	return 0;
-- 
2.30.2

[PATCH v5 net 2/3] net: mediatek: sgmii: fix duplex configuration

[Bjørn Mork]

The logic of the duplex bit is inverted.  Setting it means half
duplex, not full duplex.

Fix and rename macro to avoid confusion.

Fixes: 7e538372694b ("net: ethernet: mediatek: Re-add support SGMII")
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Bjørn Mork <bjorn@mork.no>
---
 drivers/net/ethernet/mediatek/mtk_eth_soc.h | 2 +-
 drivers/net/ethernet/mediatek/mtk_sgmii.c   | 6 +++---
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
index 01a38f5145b1..2d9186d32bc0 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
@@ -519,7 +519,7 @@
 #define SGMII_SPEED_10			FIELD_PREP(SGMII_SPEED_MASK, 0)
 #define SGMII_SPEED_100			FIELD_PREP(SGMII_SPEED_MASK, 1)
 #define SGMII_SPEED_1000		FIELD_PREP(SGMII_SPEED_MASK, 2)
-#define SGMII_DUPLEX_FULL		BIT(4)
+#define SGMII_DUPLEX_HALF		BIT(4)
 #define SGMII_IF_MODE_BIT5		BIT(5)
 #define SGMII_REMOTE_FAULT_DIS		BIT(8)
 #define SGMII_CODE_SYNC_SET_VAL		BIT(9)
diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
index 0a06995099cf..c4261069b521 100644
--- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
+++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
@@ -154,11 +154,11 @@ static void mtk_pcs_link_up(struct phylink_pcs *pcs, unsigned int mode,
 		else
 			sgm_mode = SGMII_SPEED_1000;
 
-		if (duplex == DUPLEX_FULL)
-			sgm_mode |= SGMII_DUPLEX_FULL;
+		if (duplex != DUPLEX_FULL)
+			sgm_mode |= SGMII_DUPLEX_HALF;
 
 		regmap_update_bits(mpcs->regmap, SGMSYS_SGMII_MODE,
-				   SGMII_DUPLEX_FULL | SGMII_SPEED_MASK,
+				   SGMII_DUPLEX_HALF | SGMII_SPEED_MASK,
 				   sgm_mode);
 	}
 }
-- 
2.30.2

[PATCH v5 net 3/3] mtk_sgmii: enable PCS polling to allow SFP work

[Bjørn Mork]

From: Alexander Couzens <lynxis@fe80.eu>

Currently there is no IRQ handling (even the SGMII supports it).
Enable polling to support SFP ports.

Fixes: 14a44ab0330d ("net: mtk_eth_soc: partially convert to phylink_pcs")
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Alexander Couzens <lynxis@fe80.eu>
[ bmork: changed "1" => "true" ]
Signed-off-by: Bjørn Mork <bjorn@mork.no>
---
 drivers/net/ethernet/mediatek/mtk_sgmii.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/ethernet/mediatek/mtk_sgmii.c b/drivers/net/ethernet/mediatek/mtk_sgmii.c
index c4261069b521..bb00de1003ac 100644
--- a/drivers/net/ethernet/mediatek/mtk_sgmii.c
+++ b/drivers/net/ethernet/mediatek/mtk_sgmii.c
@@ -187,6 +187,7 @@ int mtk_sgmii_init(struct mtk_sgmii *ss, struct device_node *r, u32 ana_rgc3)
 			return PTR_ERR(ss->pcs[i].regmap);
 
 		ss->pcs[i].pcs.ops = &mtk_pcs_ops;
+		ss->pcs[i].pcs.poll = true;
 		ss->pcs[i].interface = PHY_INTERFACE_MODE_NA;
 	}
 
-- 
2.30.2

Re: [PATCH v5 net 0/3] fixes for mtk_eth_soc

[Daniel Golle]

On Wed, Feb 01, 2023 at 07:23:28PM +0100, Bjørn Mork wrote:
> Changes since v4:
>  - use same field order for kernel-doc and code in patch 1
>  - cc'ing full maintainer list from get_maintainer.pl
> 
> Changes since v3:
>  - fill hole in struct mtk_pcs with new interface field
>  - improved patch 2 commit message
>  - added fixes tags
>  - updated review tags
>  
> Changes since v2:
>  - use "true" for boolean
>  - fix SoB typo
>  - updated tags
> 
> Changes since v1:
>  - only power down on changes, fix from Russel
>  - dropped bogus uncondional in-band patch
>  - added pcs poll patch from Alexander
>  - updated tags
> 
> 
> Fix mtk_eth_soc sgmii configuration.
> 
> This has been tested on a MT7986 with a Maxlinear GPY211C phy
> permanently attached to the second SoC mac.

for the whole series:
Acked-by: Daniel Golle <daniel@makrotopia.org>
Tested-by: Daniel Golle <daniel@makrotopia.org>

> 
> 
> Alexander Couzens (2):
>   net: mediatek: sgmii: ensure the SGMII PHY is powered down on
>     configuration
>   mtk_sgmii: enable PCS polling to allow SFP work
> 
> Bjørn Mork (1):
>   net: mediatek: sgmii: fix duplex configuration
> 
>  drivers/net/ethernet/mediatek/mtk_eth_soc.h |  4 +-
>  drivers/net/ethernet/mediatek/mtk_sgmii.c   | 46 ++++++++++++++-------
>  2 files changed, 35 insertions(+), 15 deletions(-)
> 
> -- 
> 2.30.2
> 

Re: [PATCH v5 net 0/3] fixes for mtk_eth_soc

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net.git (master)
by Jakub Kicinski <kuba@kernel.org>:

On Wed,  1 Feb 2023 19:23:28 +0100 you wrote:
> Changes since v4:
>  - use same field order for kernel-doc and code in patch 1
>  - cc'ing full maintainer list from get_maintainer.pl
> 
> Changes since v3:
>  - fill hole in struct mtk_pcs with new interface field
>  - improved patch 2 commit message
>  - added fixes tags
>  - updated review tags
> 
> [...]

Here is the summary with links:
  - [v5,net,1/3] net: mediatek: sgmii: ensure the SGMII PHY is powered down on configuration
    https://git.kernel.org/netdev/net/c/7ff82416de82
  - [v5,net,2/3] net: mediatek: sgmii: fix duplex configuration
    https://git.kernel.org/netdev/net/c/9d32637122de
  - [v5,net,3/3] mtk_sgmii: enable PCS polling to allow SFP work
    https://git.kernel.org/netdev/net/c/3337a6e04ddf

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html