* [PATCH net v2 1/6] net: dsa: mv88e6xxx: Fix application of erratum 4.8 for 88E6393X
2021-11-30 17:01 [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) Marek Behún
@ 2021-11-30 17:01 ` Marek Behún
2021-11-30 17:01 ` [PATCH net v2 2/6] net: dsa: mv88e6xxx: Drop unnecessary check in mv88e6393x_serdes_erratum_4_6() Marek Behún
` (5 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Marek Behún @ 2021-11-30 17:01 UTC (permalink / raw)
To: netdev, Andrew Lunn
Cc: Russell King, Jakub Kicinski, davem, Vladimir Oltean,
Marek Behún
According to SERDES scripts for 88E6393X, erratum 4.8 has to be applied
every time before SerDes is powered on.
Split the code for erratum 4.8 into separate function and call it in
mv88e6393x_serdes_power().
Fixes: de776d0d316f ("net: dsa: mv88e6xxx: add support for mv88e6393x family")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
drivers/net/dsa/mv88e6xxx/serdes.c | 53 +++++++++++++++++++-----------
1 file changed, 33 insertions(+), 20 deletions(-)
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.c b/drivers/net/dsa/mv88e6xxx/serdes.c
index 6ea003678798..0658ee3b014c 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.c
+++ b/drivers/net/dsa/mv88e6xxx/serdes.c
@@ -1271,9 +1271,9 @@ void mv88e6390_serdes_get_regs(struct mv88e6xxx_chip *chip, int port, void *_p)
}
}
-static int mv88e6393x_serdes_port_errata(struct mv88e6xxx_chip *chip, int lane)
+static int mv88e6393x_serdes_erratum_4_6(struct mv88e6xxx_chip *chip, int lane)
{
- u16 reg, pcs;
+ u16 reg;
int err;
/* mv88e6393x family errata 4.6:
@@ -1300,11 +1300,32 @@ static int mv88e6393x_serdes_port_errata(struct mv88e6xxx_chip *chip, int lane)
if (err)
return err;
- err = mv88e6390_serdes_power_sgmii(chip, lane, false);
- if (err)
- return err;
+ return mv88e6390_serdes_power_sgmii(chip, lane, false);
}
+ return 0;
+}
+
+int mv88e6393x_serdes_setup_errata(struct mv88e6xxx_chip *chip)
+{
+ int err;
+
+ err = mv88e6393x_serdes_erratum_4_6(chip, MV88E6393X_PORT0_LANE);
+ if (err)
+ return err;
+
+ err = mv88e6393x_serdes_erratum_4_6(chip, MV88E6393X_PORT9_LANE);
+ if (err)
+ return err;
+
+ return mv88e6393x_serdes_erratum_4_6(chip, MV88E6393X_PORT10_LANE);
+}
+
+static int mv88e6393x_serdes_erratum_4_8(struct mv88e6xxx_chip *chip, int lane)
+{
+ u16 reg, pcs;
+ int err;
+
/* mv88e6393x family errata 4.8:
* When a SERDES port is operating in 1000BASE-X or SGMII mode link may
* not come up after hardware reset or software reset of SERDES core.
@@ -1334,29 +1355,21 @@ static int mv88e6393x_serdes_port_errata(struct mv88e6xxx_chip *chip, int lane)
MV88E6393X_ERRATA_4_8_REG, reg);
}
-int mv88e6393x_serdes_setup_errata(struct mv88e6xxx_chip *chip)
-{
- int err;
-
- err = mv88e6393x_serdes_port_errata(chip, MV88E6393X_PORT0_LANE);
- if (err)
- return err;
-
- err = mv88e6393x_serdes_port_errata(chip, MV88E6393X_PORT9_LANE);
- if (err)
- return err;
-
- return mv88e6393x_serdes_port_errata(chip, MV88E6393X_PORT10_LANE);
-}
-
int mv88e6393x_serdes_power(struct mv88e6xxx_chip *chip, int port, int lane,
bool on)
{
u8 cmode = chip->ports[port].cmode;
+ int err;
if (port != 0 && port != 9 && port != 10)
return -EOPNOTSUPP;
+ if (on) {
+ err = mv88e6393x_serdes_erratum_4_8(chip, lane);
+ if (err)
+ return err;
+ }
+
switch (cmode) {
case MV88E6XXX_PORT_STS_CMODE_SGMII:
case MV88E6XXX_PORT_STS_CMODE_1000BASEX:
--
2.32.0
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH net v2 2/6] net: dsa: mv88e6xxx: Drop unnecessary check in mv88e6393x_serdes_erratum_4_6()
2021-11-30 17:01 [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) Marek Behún
2021-11-30 17:01 ` [PATCH net v2 1/6] net: dsa: mv88e6xxx: Fix application of erratum 4.8 for 88E6393X Marek Behún
@ 2021-11-30 17:01 ` Marek Behún
2021-11-30 17:01 ` [PATCH net v2 3/6] net: dsa: mv88e6xxx: Save power by disabling SerDes trasmitter and receiver Marek Behún
` (4 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Marek Behún @ 2021-11-30 17:01 UTC (permalink / raw)
To: netdev, Andrew Lunn
Cc: Russell King, Jakub Kicinski, davem, Vladimir Oltean,
Marek Behún
The check for lane is unnecessary, since the function is called only
with allowed lane argument.
Signed-off-by: Marek Behún <kabel@kernel.org>
---
drivers/net/dsa/mv88e6xxx/serdes.c | 28 +++++++++++-----------------
1 file changed, 11 insertions(+), 17 deletions(-)
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.c b/drivers/net/dsa/mv88e6xxx/serdes.c
index 0658ee3b014c..3a6244596a67 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.c
+++ b/drivers/net/dsa/mv88e6xxx/serdes.c
@@ -1284,26 +1284,20 @@ static int mv88e6393x_serdes_erratum_4_6(struct mv88e6xxx_chip *chip, int lane)
* It seems that after this workaround the SERDES is automatically
* powered up (the bit is cleared), so power it down.
*/
- if (lane == MV88E6393X_PORT0_LANE || lane == MV88E6393X_PORT9_LANE ||
- lane == MV88E6393X_PORT10_LANE) {
- err = mv88e6390_serdes_read(chip, lane,
- MDIO_MMD_PHYXS,
- MV88E6393X_SERDES_POC, ®);
- if (err)
- return err;
+ err = mv88e6390_serdes_read(chip, lane, MDIO_MMD_PHYXS,
+ MV88E6393X_SERDES_POC, ®);
+ if (err)
+ return err;
- reg &= ~MV88E6393X_SERDES_POC_PDOWN;
- reg |= MV88E6393X_SERDES_POC_RESET;
+ reg &= ~MV88E6393X_SERDES_POC_PDOWN;
+ reg |= MV88E6393X_SERDES_POC_RESET;
- err = mv88e6390_serdes_write(chip, lane, MDIO_MMD_PHYXS,
- MV88E6393X_SERDES_POC, reg);
- if (err)
- return err;
-
- return mv88e6390_serdes_power_sgmii(chip, lane, false);
- }
+ err = mv88e6390_serdes_write(chip, lane, MDIO_MMD_PHYXS,
+ MV88E6393X_SERDES_POC, reg);
+ if (err)
+ return err;
- return 0;
+ return mv88e6390_serdes_power_sgmii(chip, lane, false);
}
int mv88e6393x_serdes_setup_errata(struct mv88e6xxx_chip *chip)
--
2.32.0
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH net v2 3/6] net: dsa: mv88e6xxx: Save power by disabling SerDes trasmitter and receiver
2021-11-30 17:01 [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) Marek Behún
2021-11-30 17:01 ` [PATCH net v2 1/6] net: dsa: mv88e6xxx: Fix application of erratum 4.8 for 88E6393X Marek Behún
2021-11-30 17:01 ` [PATCH net v2 2/6] net: dsa: mv88e6xxx: Drop unnecessary check in mv88e6393x_serdes_erratum_4_6() Marek Behún
@ 2021-11-30 17:01 ` Marek Behún
2021-11-30 17:01 ` [PATCH net v2 4/6] net: dsa: mv88e6xxx: Add fix for erratum 5.2 of 88E6393X family Marek Behún
` (3 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Marek Behún @ 2021-11-30 17:01 UTC (permalink / raw)
To: netdev, Andrew Lunn
Cc: Russell King, Jakub Kicinski, davem, Vladimir Oltean,
Marek Behún
Save power on 88E6393X by disabling SerDes receiver and transmitter
after SerDes is SerDes is disabled.
Signed-off-by: Marek Behún <kabel@kernel.org>
Cc: stable@vger.kernel.org # de776d0d316f ("net: dsa: mv88e6xxx: add support for mv88e6393x family")
---
drivers/net/dsa/mv88e6xxx/serdes.c | 46 +++++++++++++++++++++++++++---
drivers/net/dsa/mv88e6xxx/serdes.h | 3 ++
2 files changed, 45 insertions(+), 4 deletions(-)
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.c b/drivers/net/dsa/mv88e6xxx/serdes.c
index 3a6244596a67..ceb63d7f1f97 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.c
+++ b/drivers/net/dsa/mv88e6xxx/serdes.c
@@ -1271,6 +1271,28 @@ void mv88e6390_serdes_get_regs(struct mv88e6xxx_chip *chip, int port, void *_p)
}
}
+static int mv88e6393x_serdes_power_lane(struct mv88e6xxx_chip *chip, int lane,
+ bool on)
+{
+ u16 reg;
+ int err;
+
+ err = mv88e6390_serdes_read(chip, lane, MDIO_MMD_PHYXS,
+ MV88E6393X_SERDES_CTRL1, ®);
+ if (err)
+ return err;
+
+ if (on)
+ reg &= ~(MV88E6393X_SERDES_CTRL1_TX_PDOWN |
+ MV88E6393X_SERDES_CTRL1_RX_PDOWN);
+ else
+ reg |= MV88E6393X_SERDES_CTRL1_TX_PDOWN |
+ MV88E6393X_SERDES_CTRL1_RX_PDOWN;
+
+ return mv88e6390_serdes_write(chip, lane, MDIO_MMD_PHYXS,
+ MV88E6393X_SERDES_CTRL1, reg);
+}
+
static int mv88e6393x_serdes_erratum_4_6(struct mv88e6xxx_chip *chip, int lane)
{
u16 reg;
@@ -1297,7 +1319,11 @@ static int mv88e6393x_serdes_erratum_4_6(struct mv88e6xxx_chip *chip, int lane)
if (err)
return err;
- return mv88e6390_serdes_power_sgmii(chip, lane, false);
+ err = mv88e6390_serdes_power_sgmii(chip, lane, false);
+ if (err)
+ return err;
+
+ return mv88e6393x_serdes_power_lane(chip, lane, false);
}
int mv88e6393x_serdes_setup_errata(struct mv88e6xxx_chip *chip)
@@ -1362,17 +1388,29 @@ int mv88e6393x_serdes_power(struct mv88e6xxx_chip *chip, int port, int lane,
err = mv88e6393x_serdes_erratum_4_8(chip, lane);
if (err)
return err;
+
+ err = mv88e6393x_serdes_power_lane(chip, lane, true);
+ if (err)
+ return err;
}
switch (cmode) {
case MV88E6XXX_PORT_STS_CMODE_SGMII:
case MV88E6XXX_PORT_STS_CMODE_1000BASEX:
case MV88E6XXX_PORT_STS_CMODE_2500BASEX:
- return mv88e6390_serdes_power_sgmii(chip, lane, on);
+ err = mv88e6390_serdes_power_sgmii(chip, lane, on);
+ break;
case MV88E6393X_PORT_STS_CMODE_5GBASER:
case MV88E6393X_PORT_STS_CMODE_10GBASER:
- return mv88e6390_serdes_power_10g(chip, lane, on);
+ err = mv88e6390_serdes_power_10g(chip, lane, on);
+ break;
}
- return 0;
+ if (err)
+ return err;
+
+ if (!on)
+ err = mv88e6393x_serdes_power_lane(chip, lane, false);
+
+ return err;
}
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.h b/drivers/net/dsa/mv88e6xxx/serdes.h
index cbb3ba30caea..e9292c8beee4 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.h
+++ b/drivers/net/dsa/mv88e6xxx/serdes.h
@@ -93,6 +93,9 @@
#define MV88E6393X_SERDES_POC_PCS_MASK 0x0007
#define MV88E6393X_SERDES_POC_RESET BIT(15)
#define MV88E6393X_SERDES_POC_PDOWN BIT(5)
+#define MV88E6393X_SERDES_CTRL1 0xf003
+#define MV88E6393X_SERDES_CTRL1_TX_PDOWN BIT(9)
+#define MV88E6393X_SERDES_CTRL1_RX_PDOWN BIT(8)
#define MV88E6393X_ERRATA_4_8_REG 0xF074
#define MV88E6393X_ERRATA_4_8_BIT BIT(14)
--
2.32.0
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH net v2 4/6] net: dsa: mv88e6xxx: Add fix for erratum 5.2 of 88E6393X family
2021-11-30 17:01 [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) Marek Behún
` (2 preceding siblings ...)
2021-11-30 17:01 ` [PATCH net v2 3/6] net: dsa: mv88e6xxx: Save power by disabling SerDes trasmitter and receiver Marek Behún
@ 2021-11-30 17:01 ` Marek Behún
2021-11-30 17:01 ` [PATCH net v2 5/6] net: dsa: mv88e6xxx: Fix inband AN for 2500base-x on " Marek Behún
` (2 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Marek Behún @ 2021-11-30 17:01 UTC (permalink / raw)
To: netdev, Andrew Lunn
Cc: Russell King, Jakub Kicinski, davem, Vladimir Oltean,
Marek Behún
Add fix for erratum 5.2 of the 88E6393X (Amethyst) family: for 10gbase-r
mode, some undocumented registers need to be written some special
values.
Fixes: de776d0d316f ("net: dsa: mv88e6xxx: add support for mv88e6393x family")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
drivers/net/dsa/mv88e6xxx/serdes.c | 48 ++++++++++++++++++++++++++++++
1 file changed, 48 insertions(+)
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.c b/drivers/net/dsa/mv88e6xxx/serdes.c
index ceb63d7f1f97..9e4f18a4adc2 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.c
+++ b/drivers/net/dsa/mv88e6xxx/serdes.c
@@ -1375,6 +1375,50 @@ static int mv88e6393x_serdes_erratum_4_8(struct mv88e6xxx_chip *chip, int lane)
MV88E6393X_ERRATA_4_8_REG, reg);
}
+static int mv88e6393x_serdes_erratum_5_2(struct mv88e6xxx_chip *chip, int lane,
+ u8 cmode)
+{
+ static const struct {
+ u16 dev, reg, val, mask;
+ } fixes[] = {
+ { MDIO_MMD_VEND1, 0x8093, 0xcb5a, 0xffff },
+ { MDIO_MMD_VEND1, 0x8171, 0x7088, 0xffff },
+ { MDIO_MMD_VEND1, 0x80c9, 0x311a, 0xffff },
+ { MDIO_MMD_VEND1, 0x80a2, 0x8000, 0xff7f },
+ { MDIO_MMD_VEND1, 0x80a9, 0x0000, 0xfff0 },
+ { MDIO_MMD_VEND1, 0x80a3, 0x0000, 0xf8ff },
+ { MDIO_MMD_PHYXS, MV88E6393X_SERDES_POC,
+ MV88E6393X_SERDES_POC_RESET, MV88E6393X_SERDES_POC_RESET },
+ };
+ int err, i;
+ u16 reg;
+
+ /* mv88e6393x family errata 5.2:
+ * For optimal signal integrity the following sequence should be applied
+ * to SERDES operating in 10G mode. These registers only apply to 10G
+ * operation and have no effect on other speeds.
+ */
+ if (cmode != MV88E6393X_PORT_STS_CMODE_10GBASER)
+ return 0;
+
+ for (i = 0; i < ARRAY_SIZE(fixes); ++i) {
+ err = mv88e6390_serdes_read(chip, lane, fixes[i].dev,
+ fixes[i].reg, ®);
+ if (err)
+ return err;
+
+ reg &= ~fixes[i].mask;
+ reg |= fixes[i].val;
+
+ err = mv88e6390_serdes_write(chip, lane, fixes[i].dev,
+ fixes[i].reg, reg);
+ if (err)
+ return err;
+ }
+
+ return 0;
+}
+
int mv88e6393x_serdes_power(struct mv88e6xxx_chip *chip, int port, int lane,
bool on)
{
@@ -1389,6 +1433,10 @@ int mv88e6393x_serdes_power(struct mv88e6xxx_chip *chip, int port, int lane,
if (err)
return err;
+ err = mv88e6393x_serdes_erratum_5_2(chip, lane, cmode);
+ if (err)
+ return err;
+
err = mv88e6393x_serdes_power_lane(chip, lane, true);
if (err)
return err;
--
2.32.0
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH net v2 5/6] net: dsa: mv88e6xxx: Fix inband AN for 2500base-x on 88E6393X family
2021-11-30 17:01 [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) Marek Behún
` (3 preceding siblings ...)
2021-11-30 17:01 ` [PATCH net v2 4/6] net: dsa: mv88e6xxx: Add fix for erratum 5.2 of 88E6393X family Marek Behún
@ 2021-11-30 17:01 ` Marek Behún
2021-11-30 17:01 ` [PATCH net v2 6/6] net: dsa: mv88e6xxx: Link in pcs_get_state() if AN is bypassed Marek Behún
2021-12-01 15:00 ` [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) patchwork-bot+netdevbpf
6 siblings, 0 replies; 9+ messages in thread
From: Marek Behún @ 2021-11-30 17:01 UTC (permalink / raw)
To: netdev, Andrew Lunn
Cc: Russell King, Jakub Kicinski, davem, Vladimir Oltean,
Marek Behún
Inband AN is broken on Amethyst in 2500base-x mode when set by standard
mechanism (via cmode).
(There probably is some weird setting done by default in the switch for
this mode that make it cycle in some state or something, because when
the peer is the mvneta controller, it receives link change interrupts
every ~0.3ms, but the link is always down.)
Get around this by configuring the PCS mode to 1000base-x (where inband
AN works), and then changing the SerDes frequency while SerDes
transmitter and receiver are disabled, before enabling SerDes PHY. After
disabling SerDes PHY, change the PCS mode back to 2500base-x, to avoid
confusing the device (if we leave it at 1000base-x PCS mode but with
different frequency, and then change cmode to sgmii, the device won't
change the frequency because it thinks it already has the correct one).
The register which changes the frequency is undocumented. I discovered
it by going through all registers in the ranges 4.f000-4.f100 and
1e.8000-1e.8200 for all SerDes cmodes (sgmii, 1000base-x, 2500base-x,
5gbase-r, 10gbase-r, usxgmii) and filtering out registers that didn't
make sense (the value was the same for modes which have different
frequency). The result of this was:
reg sgmii 1000base-x 2500base-x 5gbase-r 10gbase-r usxgmii
04.f002 005b 0058 0059 005c 005d 005f
04.f076 3000 0000 1000 4000 5000 7000
04.f07c 0950 0950 1850 0550 0150 0150
1e.8000 0059 0059 0058 0055 0051 0051
1e.8140 0e20 0e20 0e28 0e21 0e42 0e42
Register 04.f002 is the documented Port Operational Confiuration
register, it's last 3 bits select PCS type, so changing this register
also changes the frequency to the appropriate value.
Registers 04.f076 and 04.f07c are not writable.
Undocumented register 1e.8000 was the one: changing bits 3:0 from 9 to 8
changed SerDes frequency to 3.125 GHz, while leaving the value of PCS
mode in register 04.f002.2:0 at 1000base-x. Inband autonegotiation
started working correctly.
(I didn't try anything with register 1e.8140 since 1e.8000 solved the
problem.)
Since I don't have documentation for this register 1e.8000.3:0, I am
using the constants without names, but my hypothesis is that this
register selects PHY frequency. If in the future I have access to an
oscilloscope able to handle these frequencies, I will try to test this
hypothesis.
Fixes: de776d0d316f ("net: dsa: mv88e6xxx: add support for mv88e6393x family")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
drivers/net/dsa/mv88e6xxx/serdes.c | 61 +++++++++++++++++++++++++++++-
drivers/net/dsa/mv88e6xxx/serdes.h | 1 +
2 files changed, 61 insertions(+), 1 deletion(-)
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.c b/drivers/net/dsa/mv88e6xxx/serdes.c
index 9e4f18a4adc2..6f60376b932c 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.c
+++ b/drivers/net/dsa/mv88e6xxx/serdes.c
@@ -1419,6 +1419,54 @@ static int mv88e6393x_serdes_erratum_5_2(struct mv88e6xxx_chip *chip, int lane,
return 0;
}
+static int mv88e6393x_serdes_fix_2500basex_an(struct mv88e6xxx_chip *chip,
+ int lane, u8 cmode, bool on)
+{
+ u16 reg;
+ int err;
+
+ if (cmode != MV88E6XXX_PORT_STS_CMODE_2500BASEX)
+ return 0;
+
+ /* Inband AN is broken on Amethyst in 2500base-x mode when set by
+ * standard mechanism (via cmode).
+ * We can get around this by configuring the PCS mode to 1000base-x
+ * and then writing value 0x58 to register 1e.8000. (This must be done
+ * while SerDes receiver and transmitter are disabled, which is, when
+ * this function is called.)
+ * It seem that when we do this configuration to 2500base-x mode (by
+ * changing PCS mode to 1000base-x and frequency to 3.125 GHz from
+ * 1.25 GHz) and then configure to sgmii or 1000base-x, the device
+ * thinks that it already has SerDes at 1.25 GHz and does not change
+ * the 1e.8000 register, leaving SerDes at 3.125 GHz.
+ * To avoid this, change PCS mode back to 2500base-x when disabling
+ * SerDes from 2500base-x mode.
+ */
+ err = mv88e6390_serdes_read(chip, lane, MDIO_MMD_PHYXS,
+ MV88E6393X_SERDES_POC, ®);
+ if (err)
+ return err;
+
+ reg &= ~(MV88E6393X_SERDES_POC_PCS_MASK | MV88E6393X_SERDES_POC_AN);
+ if (on)
+ reg |= MV88E6393X_SERDES_POC_PCS_1000BASEX |
+ MV88E6393X_SERDES_POC_AN;
+ else
+ reg |= MV88E6393X_SERDES_POC_PCS_2500BASEX;
+ reg |= MV88E6393X_SERDES_POC_RESET;
+
+ err = mv88e6390_serdes_write(chip, lane, MDIO_MMD_PHYXS,
+ MV88E6393X_SERDES_POC, reg);
+ if (err)
+ return err;
+
+ err = mv88e6390_serdes_write(chip, lane, MDIO_MMD_VEND1, 0x8000, 0x58);
+ if (err)
+ return err;
+
+ return 0;
+}
+
int mv88e6393x_serdes_power(struct mv88e6xxx_chip *chip, int port, int lane,
bool on)
{
@@ -1437,6 +1485,11 @@ int mv88e6393x_serdes_power(struct mv88e6xxx_chip *chip, int port, int lane,
if (err)
return err;
+ err = mv88e6393x_serdes_fix_2500basex_an(chip, lane, cmode,
+ true);
+ if (err)
+ return err;
+
err = mv88e6393x_serdes_power_lane(chip, lane, true);
if (err)
return err;
@@ -1457,8 +1510,14 @@ int mv88e6393x_serdes_power(struct mv88e6xxx_chip *chip, int port, int lane,
if (err)
return err;
- if (!on)
+ if (!on) {
err = mv88e6393x_serdes_power_lane(chip, lane, false);
+ if (err)
+ return err;
+
+ err = mv88e6393x_serdes_fix_2500basex_an(chip, lane, cmode,
+ false);
+ }
return err;
}
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.h b/drivers/net/dsa/mv88e6xxx/serdes.h
index e9292c8beee4..8dd8ed225b45 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.h
+++ b/drivers/net/dsa/mv88e6xxx/serdes.h
@@ -93,6 +93,7 @@
#define MV88E6393X_SERDES_POC_PCS_MASK 0x0007
#define MV88E6393X_SERDES_POC_RESET BIT(15)
#define MV88E6393X_SERDES_POC_PDOWN BIT(5)
+#define MV88E6393X_SERDES_POC_AN BIT(3)
#define MV88E6393X_SERDES_CTRL1 0xf003
#define MV88E6393X_SERDES_CTRL1_TX_PDOWN BIT(9)
#define MV88E6393X_SERDES_CTRL1_RX_PDOWN BIT(8)
--
2.32.0
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH net v2 6/6] net: dsa: mv88e6xxx: Link in pcs_get_state() if AN is bypassed
2021-11-30 17:01 [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) Marek Behún
` (4 preceding siblings ...)
2021-11-30 17:01 ` [PATCH net v2 5/6] net: dsa: mv88e6xxx: Fix inband AN for 2500base-x on " Marek Behún
@ 2021-11-30 17:01 ` Marek Behún
2021-12-01 15:00 ` [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) patchwork-bot+netdevbpf
6 siblings, 0 replies; 9+ messages in thread
From: Marek Behún @ 2021-11-30 17:01 UTC (permalink / raw)
To: netdev, Andrew Lunn
Cc: Russell King, Jakub Kicinski, davem, Vladimir Oltean,
Marek Behún
Function mv88e6xxx_serdes_pcs_get_state() currently does not report link
up if AN is enabled, Link bit is set, but Speed and Duplex Resolved bit
is not set, which testing shows is the case for when auto-negotiation
was bypassed (we have AN enabled but link partner does not).
An example of such link partner is Marvell 88X3310 PHY, when put into
the mode where host interface changes between 10gbase-r, 5gbase-r,
2500base-x and sgmii according to copper speed. The 88X3310 does not
enable AN in 2500base-x, and so SerDes on mv88e6xxx currently does not
link with it.
Fix this.
Fixes: a5a6858b793f ("net: dsa: mv88e6xxx: extend phylink to Serdes PHYs")
Signed-off-by: Marek Behún <kabel@kernel.org>
---
drivers/net/dsa/mv88e6xxx/serdes.c | 48 ++++++++++++++++++++++++++----
1 file changed, 42 insertions(+), 6 deletions(-)
diff --git a/drivers/net/dsa/mv88e6xxx/serdes.c b/drivers/net/dsa/mv88e6xxx/serdes.c
index 6f60376b932c..55273013bfb5 100644
--- a/drivers/net/dsa/mv88e6xxx/serdes.c
+++ b/drivers/net/dsa/mv88e6xxx/serdes.c
@@ -50,11 +50,22 @@ static int mv88e6390_serdes_write(struct mv88e6xxx_chip *chip,
}
static int mv88e6xxx_serdes_pcs_get_state(struct mv88e6xxx_chip *chip,
- u16 status, u16 lpa,
+ u16 ctrl, u16 status, u16 lpa,
struct phylink_link_state *state)
{
+ state->link = !!(status & MV88E6390_SGMII_PHY_STATUS_LINK);
+
if (status & MV88E6390_SGMII_PHY_STATUS_SPD_DPL_VALID) {
- state->link = !!(status & MV88E6390_SGMII_PHY_STATUS_LINK);
+ /* The Spped and Duplex Resolved register is 1 if AN is enabled
+ * and complete, or if AN is disabled. So with disabled AN we
+ * still get here on link up. But we want to set an_complete
+ * only if AN was enabled, thus we look at BMCR_ANENABLE.
+ * (According to 802.3-2008 section 22.2.4.2.10, we should be
+ * able to get this same value from BMSR_ANEGCAPABLE, but tests
+ * show that these Marvell PHYs don't conform to this part of
+ * the specificaion - BMSR_ANEGCAPABLE is simply always 1.)
+ */
+ state->an_complete = !!(ctrl & BMCR_ANENABLE);
state->duplex = status &
MV88E6390_SGMII_PHY_STATUS_DUPLEX_FULL ?
DUPLEX_FULL : DUPLEX_HALF;
@@ -81,6 +92,18 @@ static int mv88e6xxx_serdes_pcs_get_state(struct mv88e6xxx_chip *chip,
dev_err(chip->dev, "invalid PHY speed\n");
return -EINVAL;
}
+ } else if (state->link &&
+ state->interface != PHY_INTERFACE_MODE_SGMII) {
+ /* If Speed and Duplex Resolved register is 0 and link is up, it
+ * means that AN was enabled, but link partner had it disabled
+ * and the PHY invoked the Auto-Negotiation Bypass feature and
+ * linked anyway.
+ */
+ state->duplex = DUPLEX_FULL;
+ if (state->interface == PHY_INTERFACE_MODE_2500BASEX)
+ state->speed = SPEED_2500;
+ else
+ state->speed = SPEED_1000;
} else {
state->link = false;
}
@@ -168,9 +191,15 @@ int mv88e6352_serdes_pcs_config(struct mv88e6xxx_chip *chip, int port,
int mv88e6352_serdes_pcs_get_state(struct mv88e6xxx_chip *chip, int port,
int lane, struct phylink_link_state *state)
{
- u16 lpa, status;
+ u16 lpa, status, ctrl;
int err;
+ err = mv88e6352_serdes_read(chip, MII_BMCR, &ctrl);
+ if (err) {
+ dev_err(chip->dev, "can't read Serdes PHY control: %d\n", err);
+ return err;
+ }
+
err = mv88e6352_serdes_read(chip, 0x11, &status);
if (err) {
dev_err(chip->dev, "can't read Serdes PHY status: %d\n", err);
@@ -183,7 +212,7 @@ int mv88e6352_serdes_pcs_get_state(struct mv88e6xxx_chip *chip, int port,
return err;
}
- return mv88e6xxx_serdes_pcs_get_state(chip, status, lpa, state);
+ return mv88e6xxx_serdes_pcs_get_state(chip, ctrl, status, lpa, state);
}
int mv88e6352_serdes_pcs_an_restart(struct mv88e6xxx_chip *chip, int port,
@@ -883,9 +912,16 @@ int mv88e6390_serdes_pcs_config(struct mv88e6xxx_chip *chip, int port,
static int mv88e6390_serdes_pcs_get_state_sgmii(struct mv88e6xxx_chip *chip,
int port, int lane, struct phylink_link_state *state)
{
- u16 lpa, status;
+ u16 lpa, status, ctrl;
int err;
+ err = mv88e6390_serdes_read(chip, lane, MDIO_MMD_PHYXS,
+ MV88E6390_SGMII_BMCR, &ctrl);
+ if (err) {
+ dev_err(chip->dev, "can't read Serdes PHY control: %d\n", err);
+ return err;
+ }
+
err = mv88e6390_serdes_read(chip, lane, MDIO_MMD_PHYXS,
MV88E6390_SGMII_PHY_STATUS, &status);
if (err) {
@@ -900,7 +936,7 @@ static int mv88e6390_serdes_pcs_get_state_sgmii(struct mv88e6xxx_chip *chip,
return err;
}
- return mv88e6xxx_serdes_pcs_get_state(chip, status, lpa, state);
+ return mv88e6xxx_serdes_pcs_get_state(chip, ctrl, status, lpa, state);
}
static int mv88e6390_serdes_pcs_get_state_10g(struct mv88e6xxx_chip *chip,
--
2.32.0
^ permalink raw reply related [flat|nested] 9+ messages in thread* Re: [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family)
2021-11-30 17:01 [PATCH net v2 0/6] mv88e6xxx fixes (mainly 88E6393X family) Marek Behún
` (5 preceding siblings ...)
2021-11-30 17:01 ` [PATCH net v2 6/6] net: dsa: mv88e6xxx: Link in pcs_get_state() if AN is bypassed Marek Behún
@ 2021-12-01 15:00 ` patchwork-bot+netdevbpf
6 siblings, 0 replies; 9+ messages in thread
From: patchwork-bot+netdevbpf @ 2021-12-01 15:00 UTC (permalink / raw)
To: =?utf-8?q?Marek_Beh=C3=BAn_=3Ckabel=40kernel=2Eorg=3E?=
Cc: netdev, andrew, rmk+kernel, kuba, davem, olteanv
Hello:
This series was applied to netdev/net.git (master)
by David S. Miller <davem@davemloft.net>:
On Tue, 30 Nov 2021 18:01:45 +0100 you wrote:
> Hello,
>
> sending v2 of these fixes.
>
> Original cover letter:
>
> So I managed to discovered how to fix inband AN for 2500base-x mode on
> 88E6393x (Amethyst) family.
>
> [...]
Here is the summary with links:
- [net,v2,1/6] net: dsa: mv88e6xxx: Fix application of erratum 4.8 for 88E6393X
https://git.kernel.org/netdev/net/c/21635d9203e1
- [net,v2,2/6] net: dsa: mv88e6xxx: Drop unnecessary check in mv88e6393x_serdes_erratum_4_6()
https://git.kernel.org/netdev/net/c/8c3318b4874e
- [net,v2,3/6] net: dsa: mv88e6xxx: Save power by disabling SerDes trasmitter and receiver
https://git.kernel.org/netdev/net/c/7527d66260ac
- [net,v2,4/6] net: dsa: mv88e6xxx: Add fix for erratum 5.2 of 88E6393X family
https://git.kernel.org/netdev/net/c/93fd8207bed8
- [net,v2,5/6] net: dsa: mv88e6xxx: Fix inband AN for 2500base-x on 88E6393X family
https://git.kernel.org/netdev/net/c/163000dbc772
- [net,v2,6/6] net: dsa: mv88e6xxx: Link in pcs_get_state() if AN is bypassed
https://git.kernel.org/netdev/net/c/ede359d8843a
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 9+ messages in thread