[PATCH net v2 0/7] Fixes for mv88e6xxx (mainly 6320 family)

[PATCH net v2 0/7] Fixes for mv88e6xxx (mainly 6320 family)

[Marek Behún]

Hello Andrew et al.,

this is v2 of fixes for the mv88e6xxx driver.
As requested, I left only the fixes that are needed for our board
to be backported, the rest will be sent to net-next.

Marek

v1 at:
  https://patchwork.kernel.org/project/netdevbpf/cover/20250313134146.27087-1-kabel@kernel.org/

Marek Behún (7):
  net: dsa: mv88e6xxx: fix VTU methods for 6320 family
  net: dsa: mv88e6xxx: fix atu_move_port_mask for 6341 family
  net: dsa: mv88e6xxx: enable PVT for 6321 switch
  net: dsa: mv88e6xxx: enable .port_set_policy() for 6320 family
  net: dsa: mv88e6xxx: enable STU methods for 6320 family
  net: dsa: mv88e6xxx: fix internal PHYs for 6320 family
  net: dsa: mv88e6xxx: workaround RGMII transmit delay erratum for 6320
    family

 drivers/net/dsa/mv88e6xxx/chip.c | 44 ++++++++++++++++++++++++++------
 1 file changed, 36 insertions(+), 8 deletions(-)

-- 
2.48.1

[PATCH net v2 1/7] net: dsa: mv88e6xxx: fix VTU methods for 6320 family

[Marek Behún]

The VTU registers of the 6320 family use the 6352 semantics, not 6185.
Fix it.

Fixes: b8fee9571063 ("net: dsa: mv88e6xxx: add VLAN Get Next support")
Signed-off-by: Marek Behún <kabel@kernel.org>
Cc: <stable@vger.kernel.org> # 5.15.x
---
 drivers/net/dsa/mv88e6xxx/chip.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 5db96ca52505..06b17c3b2205 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -5169,8 +5169,8 @@ static const struct mv88e6xxx_ops mv88e6320_ops = {
 	.hardware_reset_pre = mv88e6xxx_g2_eeprom_wait,
 	.hardware_reset_post = mv88e6xxx_g2_eeprom_wait,
 	.reset = mv88e6352_g1_reset,
-	.vtu_getnext = mv88e6185_g1_vtu_getnext,
-	.vtu_loadpurge = mv88e6185_g1_vtu_loadpurge,
+	.vtu_getnext = mv88e6352_g1_vtu_getnext,
+	.vtu_loadpurge = mv88e6352_g1_vtu_loadpurge,
 	.gpio_ops = &mv88e6352_gpio_ops,
 	.avb_ops = &mv88e6352_avb_ops,
 	.ptp_ops = &mv88e6352_ptp_ops,
@@ -5217,8 +5217,8 @@ static const struct mv88e6xxx_ops mv88e6321_ops = {
 	.hardware_reset_pre = mv88e6xxx_g2_eeprom_wait,
 	.hardware_reset_post = mv88e6xxx_g2_eeprom_wait,
 	.reset = mv88e6352_g1_reset,
-	.vtu_getnext = mv88e6185_g1_vtu_getnext,
-	.vtu_loadpurge = mv88e6185_g1_vtu_loadpurge,
+	.vtu_getnext = mv88e6352_g1_vtu_getnext,
+	.vtu_loadpurge = mv88e6352_g1_vtu_loadpurge,
 	.gpio_ops = &mv88e6352_gpio_ops,
 	.avb_ops = &mv88e6352_avb_ops,
 	.ptp_ops = &mv88e6352_ptp_ops,
-- 
2.48.1

Re: [PATCH net v2 1/7] net: dsa: mv88e6xxx: fix VTU methods for 6320 family

[Andrew Lunn]

On Mon, Mar 17, 2025 at 06:32:44PM +0100, Marek Behún wrote:
> The VTU registers of the 6320 family use the 6352 semantics, not 6185.
> Fix it.
> 
> Fixes: b8fee9571063 ("net: dsa: mv88e6xxx: add VLAN Get Next support")
> Signed-off-by: Marek Behún <kabel@kernel.org>
> Cc: <stable@vger.kernel.org> # 5.15.x

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net v2 2/7] net: dsa: mv88e6xxx: fix atu_move_port_mask for 6341 family

[Marek Behún]

The atu_move_port_mask for 6341 family (Topaz) is 0xf, not 0x1f. The
PortVec field is 8 bits wide, not 11 as in 6390 family. Fix this.

Fixes: e606ca36bbf2 ("net: dsa: mv88e6xxx: rework ATU Remove")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
 drivers/net/dsa/mv88e6xxx/chip.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 06b17c3b2205..b989123cdbeb 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -5818,7 +5818,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
-		.atu_move_port_mask = 0x1f,
+		.atu_move_port_mask = 0xf,
 		.g1_irqs = 9,
 		.g2_irqs = 10,
 		.stats_type = STATS_TYPE_BANK0 | STATS_TYPE_BANK1,
@@ -6296,7 +6296,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.global1_addr = 0x1b,
 		.global2_addr = 0x1c,
 		.age_time_coeff = 3750,
-		.atu_move_port_mask = 0x1f,
+		.atu_move_port_mask = 0xf,
 		.g1_irqs = 9,
 		.g2_irqs = 10,
 		.stats_type = STATS_TYPE_BANK0 | STATS_TYPE_BANK1,
-- 
2.48.1

Re: [PATCH net v2 2/7] net: dsa: mv88e6xxx: fix atu_move_port_mask for 6341 family

[Andrew Lunn]

On Mon, Mar 17, 2025 at 06:32:45PM +0100, Marek Behún wrote:
> The atu_move_port_mask for 6341 family (Topaz) is 0xf, not 0x1f. The
> PortVec field is 8 bits wide, not 11 as in 6390 family. Fix this.
> 
> Fixes: e606ca36bbf2 ("net: dsa: mv88e6xxx: rework ATU Remove")
> Signed-off-by: Marek Behún <kabel@kernel.org>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net v2 3/7] net: dsa: mv88e6xxx: enable PVT for 6321 switch

[Marek Behún]

Commit f36456522168 ("net: dsa: mv88e6xxx: move PVT description in
info") did not enable PVT for 6321 switch. Fix it.

Fixes: f36456522168 ("net: dsa: mv88e6xxx: move PVT description in info")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
 drivers/net/dsa/mv88e6xxx/chip.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index b989123cdbeb..3cddc8f084b3 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -6274,6 +6274,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.g2_irqs = 10,
 		.stats_type = STATS_TYPE_BANK0 | STATS_TYPE_BANK1,
 		.atu_move_port_mask = 0xf,
+		.pvt = true,
 		.multi_chip = true,
 		.edsa_support = MV88E6XXX_EDSA_SUPPORTED,
 		.ptp_support = true,
-- 
2.48.1

Re: [PATCH net v2 3/7] net: dsa: mv88e6xxx: enable PVT for 6321 switch

[Andrew Lunn]

On Mon, Mar 17, 2025 at 06:32:46PM +0100, Marek Behún wrote:
> Commit f36456522168 ("net: dsa: mv88e6xxx: move PVT description in
> info") did not enable PVT for 6321 switch. Fix it.
> 
> Fixes: f36456522168 ("net: dsa: mv88e6xxx: move PVT description in info")
> Signed-off-by: Marek Behún <kabel@kernel.org>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net v2 4/7] net: dsa: mv88e6xxx: enable .port_set_policy() for 6320 family

[Marek Behún]

Commit f3a2cd326e44 ("net: dsa: mv88e6xxx: introduce .port_set_policy")
did not add the .port_set_policy() method for the 6320 family. Fix it.

Fixes: f3a2cd326e44 ("net: dsa: mv88e6xxx: introduce .port_set_policy")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
 drivers/net/dsa/mv88e6xxx/chip.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 3cddc8f084b3..f886a69d7a3c 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -5145,6 +5145,7 @@ static const struct mv88e6xxx_ops mv88e6320_ops = {
 	.port_set_rgmii_delay = mv88e6320_port_set_rgmii_delay,
 	.port_set_speed_duplex = mv88e6185_port_set_speed_duplex,
 	.port_tag_remap = mv88e6095_port_tag_remap,
+	.port_set_policy = mv88e6352_port_set_policy,
 	.port_set_frame_mode = mv88e6351_port_set_frame_mode,
 	.port_set_ucast_flood = mv88e6352_port_set_ucast_flood,
 	.port_set_mcast_flood = mv88e6352_port_set_mcast_flood,
@@ -5194,6 +5195,7 @@ static const struct mv88e6xxx_ops mv88e6321_ops = {
 	.port_set_rgmii_delay = mv88e6320_port_set_rgmii_delay,
 	.port_set_speed_duplex = mv88e6185_port_set_speed_duplex,
 	.port_tag_remap = mv88e6095_port_tag_remap,
+	.port_set_policy = mv88e6352_port_set_policy,
 	.port_set_frame_mode = mv88e6351_port_set_frame_mode,
 	.port_set_ucast_flood = mv88e6352_port_set_ucast_flood,
 	.port_set_mcast_flood = mv88e6352_port_set_mcast_flood,
-- 
2.48.1

Re: [PATCH net v2 4/7] net: dsa: mv88e6xxx: enable .port_set_policy() for 6320 family

[Andrew Lunn]

On Mon, Mar 17, 2025 at 06:32:47PM +0100, Marek Behún wrote:
> Commit f3a2cd326e44 ("net: dsa: mv88e6xxx: introduce .port_set_policy")
> did not add the .port_set_policy() method for the 6320 family. Fix it.
> 
> Fixes: f3a2cd326e44 ("net: dsa: mv88e6xxx: introduce .port_set_policy")
> Signed-off-by: Marek Behún <kabel@kernel.org>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net v2 5/7] net: dsa: mv88e6xxx: enable STU methods for 6320 family

[Marek Behún]

Commit c050f5e91b47 ("net: dsa: mv88e6xxx: Fill in STU support for all
supported chips") introduced STU methods, but did not add them to the
6320 family. Fix it.

Fixes: c050f5e91b47 ("net: dsa: mv88e6xxx: Fill in STU support for all supported chips")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
 drivers/net/dsa/mv88e6xxx/chip.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index f886a69d7a3c..74b8bae226e4 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -5172,6 +5172,8 @@ static const struct mv88e6xxx_ops mv88e6320_ops = {
 	.reset = mv88e6352_g1_reset,
 	.vtu_getnext = mv88e6352_g1_vtu_getnext,
 	.vtu_loadpurge = mv88e6352_g1_vtu_loadpurge,
+	.stu_getnext = mv88e6352_g1_stu_getnext,
+	.stu_loadpurge = mv88e6352_g1_stu_loadpurge,
 	.gpio_ops = &mv88e6352_gpio_ops,
 	.avb_ops = &mv88e6352_avb_ops,
 	.ptp_ops = &mv88e6352_ptp_ops,
@@ -5221,6 +5223,8 @@ static const struct mv88e6xxx_ops mv88e6321_ops = {
 	.reset = mv88e6352_g1_reset,
 	.vtu_getnext = mv88e6352_g1_vtu_getnext,
 	.vtu_loadpurge = mv88e6352_g1_vtu_loadpurge,
+	.stu_getnext = mv88e6352_g1_stu_getnext,
+	.stu_loadpurge = mv88e6352_g1_stu_loadpurge,
 	.gpio_ops = &mv88e6352_gpio_ops,
 	.avb_ops = &mv88e6352_avb_ops,
 	.ptp_ops = &mv88e6352_ptp_ops,
@@ -6241,6 +6245,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.num_gpio = 15,
 		.max_vid = 4095,
+		.max_sid = 63,
 		.port_base_addr = 0x10,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
@@ -6267,6 +6272,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_internal_phys = 5,
 		.num_gpio = 15,
 		.max_vid = 4095,
+		.max_sid = 63,
 		.port_base_addr = 0x10,
 		.phy_base_addr = 0x0,
 		.global1_addr = 0x1b,
-- 
2.48.1

Re: [PATCH net v2 5/7] net: dsa: mv88e6xxx: enable STU methods for 6320 family

[Andrew Lunn]

On Mon, Mar 17, 2025 at 06:32:48PM +0100, Marek Behún wrote:
> Commit c050f5e91b47 ("net: dsa: mv88e6xxx: Fill in STU support for all
> supported chips") introduced STU methods, but did not add them to the
> 6320 family. Fix it.
> 
> Fixes: c050f5e91b47 ("net: dsa: mv88e6xxx: Fill in STU support for all supported chips")
> Signed-off-by: Marek Behún <kabel@kernel.org>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net v2 6/7] net: dsa: mv88e6xxx: fix internal PHYs for 6320 family

[Marek Behún]

Fix internal PHYs definition for the 6320 family, which has only 2
internal PHYs (on ports 3 and 4).

Fixes: bc3931557d1d ("net: dsa: mv88e6xxx: Add number of internal PHYs")
Signed-off-by: Marek Behún <kabel@kernel.org>
Cc: <stable@vger.kernel.org> # 6.6.x
---
 drivers/net/dsa/mv88e6xxx/chip.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 74b8bae226e4..88f479dc328c 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -6242,7 +6242,8 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_databases = 4096,
 		.num_macs = 8192,
 		.num_ports = 7,
-		.num_internal_phys = 5,
+		.num_internal_phys = 2,
+		.internal_phys_offset = 3,
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.max_sid = 63,
@@ -6269,7 +6270,8 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = {
 		.num_databases = 4096,
 		.num_macs = 8192,
 		.num_ports = 7,
-		.num_internal_phys = 5,
+		.num_internal_phys = 2,
+		.internal_phys_offset = 3,
 		.num_gpio = 15,
 		.max_vid = 4095,
 		.max_sid = 63,
-- 
2.48.1

Re: [PATCH net v2 6/7] net: dsa: mv88e6xxx: fix internal PHYs for 6320 family

[Andrew Lunn]

On Mon, Mar 17, 2025 at 06:32:49PM +0100, Marek Behún wrote:
> Fix internal PHYs definition for the 6320 family, which has only 2
> internal PHYs (on ports 3 and 4).
> 
> Fixes: bc3931557d1d ("net: dsa: mv88e6xxx: Add number of internal PHYs")
> Signed-off-by: Marek Behún <kabel@kernel.org>
> Cc: <stable@vger.kernel.org> # 6.6.x

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

[PATCH net v2 7/7] net: dsa: mv88e6xxx: workaround RGMII transmit delay erratum for 6320 family

[Marek Behún]

Implement the workaround for erratum
  3.3 RGMII timing may be out of spec when transmit delay is enabled
for the 6320 family, which says:

  When transmit delay is enabled via Port register 1 bit 14 = 1, duty
  cycle may be out of spec. Under very rare conditions this may cause
  the attached device receive CRC errors.

Signed-off-by: Marek Behún <kabel@kernel.org>
Cc: <stable@vger.kernel.org> # 5.4.x
---
 drivers/net/dsa/mv88e6xxx/chip.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c
index 88f479dc328c..901929f96b38 100644
--- a/drivers/net/dsa/mv88e6xxx/chip.c
+++ b/drivers/net/dsa/mv88e6xxx/chip.c
@@ -3674,6 +3674,21 @@ static int mv88e6xxx_stats_setup(struct mv88e6xxx_chip *chip)
 	return mv88e6xxx_g1_stats_clear(chip);
 }
 
+static int mv88e6320_setup_errata(struct mv88e6xxx_chip *chip)
+{
+	u16 dummy;
+	int err;
+
+	/* Workaround for erratum
+	 *   3.3 RGMII timing may be out of spec when transmit delay is enabled
+	 */
+	err = mv88e6xxx_port_hidden_write(chip, 0, 0xf, 0x7, 0xe000);
+	if (err)
+		return err;
+
+	return mv88e6xxx_port_hidden_read(chip, 0, 0xf, 0x7, &dummy);
+}
+
 /* Check if the errata has already been applied. */
 static bool mv88e6390_setup_errata_applied(struct mv88e6xxx_chip *chip)
 {
@@ -5130,6 +5145,7 @@ static const struct mv88e6xxx_ops mv88e6290_ops = {
 
 static const struct mv88e6xxx_ops mv88e6320_ops = {
 	/* MV88E6XXX_FAMILY_6320 */
+	.setup_errata = mv88e6320_setup_errata,
 	.ieee_pri_map = mv88e6085_g1_ieee_pri_map,
 	.ip_pri_map = mv88e6085_g1_ip_pri_map,
 	.irl_init_all = mv88e6352_g2_irl_init_all,
@@ -5182,6 +5198,7 @@ static const struct mv88e6xxx_ops mv88e6320_ops = {
 
 static const struct mv88e6xxx_ops mv88e6321_ops = {
 	/* MV88E6XXX_FAMILY_6320 */
+	.setup_errata = mv88e6320_setup_errata,
 	.ieee_pri_map = mv88e6085_g1_ieee_pri_map,
 	.ip_pri_map = mv88e6085_g1_ip_pri_map,
 	.irl_init_all = mv88e6352_g2_irl_init_all,
-- 
2.48.1

Re: [PATCH net v2 7/7] net: dsa: mv88e6xxx: workaround RGMII transmit delay erratum for 6320 family

[Andrew Lunn]

On Mon, Mar 17, 2025 at 06:32:50PM +0100, Marek Behún wrote:
> Implement the workaround for erratum
>   3.3 RGMII timing may be out of spec when transmit delay is enabled
> for the 6320 family, which says:
> 
>   When transmit delay is enabled via Port register 1 bit 14 = 1, duty
>   cycle may be out of spec. Under very rare conditions this may cause
>   the attached device receive CRC errors.
> 
> Signed-off-by: Marek Behún <kabel@kernel.org>
> Cc: <stable@vger.kernel.org> # 5.4.x

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCH net v2 0/7] Fixes for mv88e6xxx (mainly 6320 family)

[patchwork-bot+netdevbpf]

Hello:

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

On Mon, 17 Mar 2025 18:32:43 +0100 you wrote:
> Hello Andrew et al.,
> 
> this is v2 of fixes for the mv88e6xxx driver.
> As requested, I left only the fixes that are needed for our board
> to be backported, the rest will be sent to net-next.
> 
> Marek
> 
> [...]

Here is the summary with links:
  - [net,v2,1/7] net: dsa: mv88e6xxx: fix VTU methods for 6320 family
    https://git.kernel.org/netdev/net/c/f9a457722cf5
  - [net,v2,2/7] net: dsa: mv88e6xxx: fix atu_move_port_mask for 6341 family
    https://git.kernel.org/netdev/net/c/4ae01ec00771
  - [net,v2,3/7] net: dsa: mv88e6xxx: enable PVT for 6321 switch
    https://git.kernel.org/netdev/net/c/f85c69369854
  - [net,v2,4/7] net: dsa: mv88e6xxx: enable .port_set_policy() for 6320 family
    https://git.kernel.org/netdev/net/c/a2ef58e2c4ae
  - [net,v2,5/7] net: dsa: mv88e6xxx: enable STU methods for 6320 family
    https://git.kernel.org/netdev/net/c/1428a6109b20
  - [net,v2,6/7] net: dsa: mv88e6xxx: fix internal PHYs for 6320 family
    https://git.kernel.org/netdev/net/c/52fdc41c3278
  - [net,v2,7/7] net: dsa: mv88e6xxx: workaround RGMII transmit delay erratum for 6320 family
    https://git.kernel.org/netdev/net/c/1ebc8e1ef906

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