net: phy: tja11xx: Add IRQ support to the driver
Subsystems (4)
PATCH v2 1/1 net: phy: tja11xx: Add IRQ support to the driver
Add support for handling the TJA11xx PHY IRQ signal.
---
V2: - Define each bit of the MII_INTEN register and a mask
- Drop IRQ acking from tja11xx_config_intr()
--- drivers/net/phy/nxp-tja11xx.c | 48 +++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+)
drivers/net/phy/nxp-tja11xx.c +48 -0
@@ -40,6 +40,29 @@#define MII_INTSRC_TEMP_ERR BIT(1)#define MII_INTSRC_UV_ERR BIT(3) +#define MII_INTEN 22 +#define MII_INTEN_PWON_EN BIT(15) +#define MII_INTEN_WAKEUP_EN BIT(14) +#define MII_INTEN_PHY_INIT_FAIL_EN BIT(11) +#define MII_INTEN_LINK_STATUS_FAIL_EN BIT(10) +#define MII_INTEN_LINK_STATUS_UP_EN BIT(9) +#define MII_INTEN_SYM_ERR_EN BIT(8) +#define MII_INTEN_TRAINING_FAILED_EN BIT(7) +#define MII_INTEN_SQI_WARNING_EN BIT(6) +#define MII_INTEN_CONTROL_ERR_EN BIT(5) +#define MII_INTEN_UV_ERR_EN BIT(3) +#define MII_INTEN_UV_RECOVERY_EN BIT(2) +#define MII_INTEN_TEMP_ERR_EN BIT(1) +#define MII_INTEN_SLEEP_ABORT_EN BIT(0) +#define MII_INTEN_MASK \ + (MII_INTEN_PWON_EN | MII_INTEN_WAKEUP_EN | \ + MII_INTEN_PHY_INIT_FAIL_EN | MII_INTEN_LINK_STATUS_FAIL_EN | \ + MII_INTEN_LINK_STATUS_UP_EN | MII_INTEN_SYM_ERR_EN | \ + MII_INTEN_TRAINING_FAILED_EN | MII_INTEN_SQI_WARNING_EN | \ + MII_INTEN_CONTROL_ERR_EN | MII_INTEN_UV_ERR_EN | \ + MII_INTEN_UV_RECOVERY_EN | MII_INTEN_TEMP_ERR_EN | \ + MII_INTEN_SLEEP_ABORT_EN) +#define MII_COMMSTAT 23#define MII_COMMSTAT_LINK_UP BIT(15)
@@ -239,6 +262,25 @@ static int tja11xx_read_status(struct phy_device *phydev)return 0 ; } +static int tja11xx_config_intr (struct phy_device *phydev) +{ +int ret; + +if (phydev->interrupts ==PHY_INTERRUPT_ENABLED ) + ret =phy_write (phydev,MII_INTEN ,MII_INTEN_MASK ); +else + ret =phy_write (phydev,MII_INTEN ,0 ); + +return ret <0 ? ret :0 ; +} + +static int tja11xx_ack_interrupt (struct phy_device *phydev) +{ +int ret =phy_read (phydev,MII_INTSRC ); + +return ret <0 ? ret :0 ; +} +static int tja11xx_get_sset_count (struct phy_device *phydev) {return ARRAY_SIZE (tja11xx_hw_stats);
@@ -366,6 +408,9 @@ static struct phy_driver tja11xx_driver[] = { .suspend = genphy_suspend, .resume = genphy_resume, .set_loopback = genphy_loopback, +/* IRQ related */ + .config_intr = tja11xx_config_intr, + .ack_interrupt = tja11xx_ack_interrupt,/* Statistics */ .get_sset_count = tja11xx_get_sset_count, .get_strings = tja11xx_get_strings,
@@ -381,6 +426,9 @@ static struct phy_driver tja11xx_driver[] = { .suspend = genphy_suspend, .resume = genphy_resume, .set_loopback = genphy_loopback, +/* IRQ related */ + .config_intr = tja11xx_config_intr, + .ack_interrupt = tja11xx_ack_interrupt,/* Statistics */ .get_sset_count = tja11xx_get_sset_count, .get_strings = tja11xx_get_strings,
Signature · 1 line
2.20.1
Heiner Kallweit
maintainerQuoted 36 lines · Marek Vasut
Add support for handling the TJA11xx PHY IRQ signal.
Signed-off-by Marek Vasut <[email protected]>Cc Andrew Lunn <[email protected]>Cc Florian Fainelli <[email protected]>Cc Guenter Roeck <[email protected]>Cc Heiner Kallweit <[email protected]>Cc Jean Delvare <[email protected]>---
V2: - Define each bit of theMII_INTENregister and a mask
- Drop IRQ acking fromtja11xx_config_intr()--- drivers/net/phy/nxp-tja11xx.c | 48 +++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+)drivers/net/phy/nxp-tja11xx.c +12
-0@@ -40,6 +40,29 @@#define MII_INTSRC_TEMP_ERR BIT(1)#define MII_INTSRC_UV_ERR BIT(3) +#define MII_INTEN 22 +#define MII_INTEN_PWON_EN BIT(15) +#define MII_INTEN_WAKEUP_EN BIT(14) +#define MII_INTEN_PHY_INIT_FAIL_EN BIT(11) +#define MII_INTEN_LINK_STATUS_FAIL_EN BIT(10) +#define MII_INTEN_LINK_STATUS_UP_EN BIT(9) +#define MII_INTEN_SYM_ERR_EN BIT(8) +#define MII_INTEN_TRAINING_FAILED_EN BIT(7) +#define MII_INTEN_SQI_WARNING_EN BIT(6) +#define MII_INTEN_CONTROL_ERR_EN BIT(5) +#define MII_INTEN_UV_ERR_EN BIT(3) +#define MII_INTEN_UV_RECOVERY_EN BIT(2)
drivers/net/phy/nxp-tja11xx.c line 64
Quoted 10 lines
drivers/net/phy/nxp-tja11xx.c +10
-0+#define MII_INTEN_TEMP_ERR_EN BIT(1) +#define MII_INTEN_SLEEP_ABORT_EN BIT(0) +#define MII_INTEN_MASK \ + (MII_INTEN_PWON_EN | MII_INTEN_WAKEUP_EN | \ + MII_INTEN_PHY_INIT_FAIL_EN | MII_INTEN_LINK_STATUS_FAIL_EN | \ + MII_INTEN_LINK_STATUS_UP_EN | MII_INTEN_SYM_ERR_EN | \ + MII_INTEN_TRAINING_FAILED_EN | MII_INTEN_SQI_WARNING_EN | \ + MII_INTEN_CONTROL_ERR_EN | MII_INTEN_UV_ERR_EN | \ + MII_INTEN_UV_RECOVERY_EN | MII_INTEN_TEMP_ERR_EN | \ + MII_INTEN_SLEEP_ABORT_EN)
Why do you enable all these interrupt sources? As I said, phylib needs link change info only.
Quoted 8 lines
drivers/net/phy/nxp-tja11xx.c +1
-0+#define MII_COMMSTAT 23#define MII_COMMSTAT_LINK_UP BIT(15)@@ -239,6 +262,25 @@ static int tja11xx_read_status(struct phy_device *phydev)return 0 ; }
drivers/net/phy/nxp-tja11xx.c static int tja11xx_read_status(struct phy_device *phydev) · line 274
Quoted 10 lines
drivers/net/phy/nxp-tja11xx.c +10
-0+static int tja11xx_config_intr (struct phy_device *phydev) +{ +int ret; + +if (phydev->interrupts ==PHY_INTERRUPT_ENABLED ) + ret =phy_write (phydev,MII_INTEN ,MII_INTEN_MASK ); +else + ret =phy_write (phydev,MII_INTEN ,0 ); + +return ret <0 ? ret :0 ;
phy_write returns only 0 or negative errno. You don't need variable ret.
Quoted 33 lines
drivers/net/phy/nxp-tja11xx.c +15
-0+} + +static int tja11xx_ack_interrupt (struct phy_device *phydev) +{ +int ret =phy_read (phydev,MII_INTSRC ); + +return ret <0 ? ret :0 ; +} +static int tja11xx_get_sset_count (struct phy_device *phydev) {return ARRAY_SIZE (tja11xx_hw_stats);@@ -366,6 +408,9 @@ static struct phy_driver tja11xx_driver[] = { .suspend = genphy_suspend, .resume = genphy_resume, .set_loopback = genphy_loopback, +/* IRQ related */ + .config_intr = tja11xx_config_intr, + .ack_interrupt = tja11xx_ack_interrupt,/* Statistics */ .get_sset_count = tja11xx_get_sset_count, .get_strings = tja11xx_get_strings,@@ -381,6 +426,9 @@ static struct phy_driver tja11xx_driver[] = { .suspend = genphy_suspend, .resume = genphy_resume, .set_loopback = genphy_loopback, +/* IRQ related */ + .config_intr = tja11xx_config_intr, + .ack_interrupt = tja11xx_ack_interrupt,/* Statistics */ .get_sset_count = tja11xx_get_sset_count, .get_strings = tja11xx_get_strings,
Quoted 42 lines · Heiner Kallweit
Quoted 41 lines · Marek Vasut
Add support for handling the TJA11xx PHY IRQ signal.
Signed-off-by Marek Vasut <[email protected]>Cc Andrew Lunn <[email protected]>Cc Florian Fainelli <[email protected]>Cc Guenter Roeck <[email protected]>Cc Heiner Kallweit <[email protected]>Cc Jean Delvare <[email protected]>---
V2: - Define each bit of theMII_INTENregister and a mask
- Drop IRQ acking fromtja11xx_config_intr()--- drivers/net/phy/nxp-tja11xx.c | 48 +++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+)drivers/net/phy/nxp-tja11xx.c +17
-0@@ -40,6 +40,29 @@#define MII_INTSRC_TEMP_ERR BIT(1)#define MII_INTSRC_UV_ERR BIT(3) +#define MII_INTEN 22 +#define MII_INTEN_PWON_EN BIT(15) +#define MII_INTEN_WAKEUP_EN BIT(14) +#define MII_INTEN_PHY_INIT_FAIL_EN BIT(11) +#define MII_INTEN_LINK_STATUS_FAIL_EN BIT(10) +#define MII_INTEN_LINK_STATUS_UP_EN BIT(9) +#define MII_INTEN_SYM_ERR_EN BIT(8) +#define MII_INTEN_TRAINING_FAILED_EN BIT(7) +#define MII_INTEN_SQI_WARNING_EN BIT(6) +#define MII_INTEN_CONTROL_ERR_EN BIT(5) +#define MII_INTEN_UV_ERR_EN BIT(3) +#define MII_INTEN_UV_RECOVERY_EN BIT(2) +#define MII_INTEN_TEMP_ERR_EN BIT(1) +#define MII_INTEN_SLEEP_ABORT_EN BIT(0) +#define MII_INTEN_MASK \ + (MII_INTEN_PWON_EN | MII_INTEN_WAKEUP_EN | \ + MII_INTEN_PHY_INIT_FAIL_EN | MII_INTEN_LINK_STATUS_FAIL_EN | \
Quoted 8 lines
drivers/net/phy/nxp-tja11xx.c line 64
Quoted 5 lines
drivers/net/phy/nxp-tja11xx.c +5
-0+MII_INTEN_LINK_STATUS_UP_EN |MII_INTEN_SYM_ERR_EN | \ +MII_INTEN_TRAINING_FAILED_EN |MII_INTEN_SQI_WARNING_EN | \ +MII_INTEN_CONTROL_ERR_EN |MII_INTEN_UV_ERR_EN | \ +MII_INTEN_UV_RECOVERY_EN |MII_INTEN_TEMP_ERR_EN | \ + MII_INTEN_SLEEP_ABORT_EN)Why do you enable all these interrupt sources? As I said, phylib needs link change info only.
Because I need them to reliably detect that the link state changed.
Quoted 8 lines
Quoted 8 lines
drivers/net/phy/nxp-tja11xx.c +1
-0+#define MII_COMMSTAT 23#define MII_COMMSTAT_LINK_UP BIT(15)@@ -239,6 +262,25 @@ static int tja11xx_read_status(struct phy_device *phydev)return 0 ; }
Quoted 13 lines
drivers/net/phy/nxp-tja11xx.c static int tja11xx_read_status(struct phy_device *phydev) · line 274
Quoted 10 lines
drivers/net/phy/nxp-tja11xx.c +10
-0+static int tja11xx_config_intr (struct phy_device *phydev) +{ +int ret; + +if (phydev->interrupts ==PHY_INTERRUPT_ENABLED ) + ret =phy_write (phydev,MII_INTEN ,MII_INTEN_MASK ); +else + ret =phy_write (phydev,MII_INTEN ,0 ); + +return ret <0 ? ret :0 ;
phy_writereturns only 0 or negative errno. You don't need variable ret.
OK
Signature · 2 lines
Best regards, Marek Vasut
Heiner Kallweit
maintainerQuoted 52 lines · Marek Vasut
Quoted 42 lines · Heiner Kallweit
Quoted 41 lines · Marek Vasut
Add support for handling the TJA11xx PHY IRQ signal.
Signed-off-by Marek Vasut <[email protected]>Cc Andrew Lunn <[email protected]>Cc Florian Fainelli <[email protected]>Cc Guenter Roeck <[email protected]>Cc Heiner Kallweit <[email protected]>Cc Jean Delvare <[email protected]>---
V2: - Define each bit of theMII_INTENregister and a mask
- Drop IRQ acking fromtja11xx_config_intr()--- drivers/net/phy/nxp-tja11xx.c | 48 +++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+)drivers/net/phy/nxp-tja11xx.c +17
-0@@ -40,6 +40,29 @@#define MII_INTSRC_TEMP_ERR BIT(1)#define MII_INTSRC_UV_ERR BIT(3) +#define MII_INTEN 22 +#define MII_INTEN_PWON_EN BIT(15) +#define MII_INTEN_WAKEUP_EN BIT(14) +#define MII_INTEN_PHY_INIT_FAIL_EN BIT(11) +#define MII_INTEN_LINK_STATUS_FAIL_EN BIT(10) +#define MII_INTEN_LINK_STATUS_UP_EN BIT(9) +#define MII_INTEN_SYM_ERR_EN BIT(8) +#define MII_INTEN_TRAINING_FAILED_EN BIT(7) +#define MII_INTEN_SQI_WARNING_EN BIT(6) +#define MII_INTEN_CONTROL_ERR_EN BIT(5) +#define MII_INTEN_UV_ERR_EN BIT(3) +#define MII_INTEN_UV_RECOVERY_EN BIT(2) +#define MII_INTEN_TEMP_ERR_EN BIT(1) +#define MII_INTEN_SLEEP_ABORT_EN BIT(0) +#define MII_INTEN_MASK \ + (MII_INTEN_PWON_EN | MII_INTEN_WAKEUP_EN | \ + MII_INTEN_PHY_INIT_FAIL_EN | MII_INTEN_LINK_STATUS_FAIL_EN | \Quoted 8 lines
drivers/net/phy/nxp-tja11xx.c line 64
Quoted 5 lines
drivers/net/phy/nxp-tja11xx.c +5
-0+MII_INTEN_LINK_STATUS_UP_EN |MII_INTEN_SYM_ERR_EN | \ +MII_INTEN_TRAINING_FAILED_EN |MII_INTEN_SQI_WARNING_EN | \ +MII_INTEN_CONTROL_ERR_EN |MII_INTEN_UV_ERR_EN | \ +MII_INTEN_UV_RECOVERY_EN |MII_INTEN_TEMP_ERR_EN | \ + MII_INTEN_SLEEP_ABORT_EN)Why do you enable all these interrupt sources? As I said, phylib needs link change info only.
Quoted 2 lines
Because I need them to reliably detect that the link state changed.
Hmm, e.g. this one MII_INTEN_TEMP_ERR_EN doesn't seem to be related to a link status change. Name sounds like it just reports exceeding a temperature threshold.
Quoted 24 lines
Quoted 8 lines
Quoted 8 lines
drivers/net/phy/nxp-tja11xx.c +1
-0+#define MII_COMMSTAT 23#define MII_COMMSTAT_LINK_UP BIT(15)@@ -239,6 +262,25 @@ static int tja11xx_read_status(struct phy_device *phydev)return 0 ; }Quoted 13 lines
drivers/net/phy/nxp-tja11xx.c static int tja11xx_read_status(struct phy_device *phydev) · line 274
Quoted 10 lines
drivers/net/phy/nxp-tja11xx.c +10
-0+static int tja11xx_config_intr (struct phy_device *phydev) +{ +int ret; + +if (phydev->interrupts ==PHY_INTERRUPT_ENABLED ) + ret =phy_write (phydev,MII_INTEN ,MII_INTEN_MASK ); +else + ret =phy_write (phydev,MII_INTEN ,0 ); + +return ret <0 ? ret :0 ;
phy_writereturns only 0 or negative errno. You don't need variable ret.OK
Quoted 53 lines · Heiner Kallweit
Quoted 52 lines · Marek Vasut
Quoted 50 lines · Heiner Kallweit
Quoted 36 lines · Marek Vasut
Add support for handling the TJA11xx PHY IRQ signal.
Signed-off-by Marek Vasut <[email protected]>Cc Andrew Lunn <[email protected]>Cc Florian Fainelli <[email protected]>Cc Guenter Roeck <[email protected]>Cc Heiner Kallweit <[email protected]>Cc Jean Delvare <[email protected]>---
V2: - Define each bit of theMII_INTENregister and a mask
- Drop IRQ acking fromtja11xx_config_intr()--- drivers/net/phy/nxp-tja11xx.c | 48 +++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+)drivers/net/phy/nxp-tja11xx.c +12
-0@@ -40,6 +40,29 @@#define MII_INTSRC_TEMP_ERR BIT(1)#define MII_INTSRC_UV_ERR BIT(3) +#define MII_INTEN 22 +#define MII_INTEN_PWON_EN BIT(15) +#define MII_INTEN_WAKEUP_EN BIT(14) +#define MII_INTEN_PHY_INIT_FAIL_EN BIT(11) +#define MII_INTEN_LINK_STATUS_FAIL_EN BIT(10) +#define MII_INTEN_LINK_STATUS_UP_EN BIT(9) +#define MII_INTEN_SYM_ERR_EN BIT(8) +#define MII_INTEN_TRAINING_FAILED_EN BIT(7) +#define MII_INTEN_SQI_WARNING_EN BIT(6) +#define MII_INTEN_CONTROL_ERR_EN BIT(5) +#define MII_INTEN_UV_ERR_EN BIT(3) +#define MII_INTEN_UV_RECOVERY_EN BIT(2)drivers/net/phy/nxp-tja11xx.c line 64
Quoted 10 lines
drivers/net/phy/nxp-tja11xx.c +10
-0+#define MII_INTEN_TEMP_ERR_EN BIT(1) +#define MII_INTEN_SLEEP_ABORT_EN BIT(0) +#define MII_INTEN_MASK \ + (MII_INTEN_PWON_EN | MII_INTEN_WAKEUP_EN | \ + MII_INTEN_PHY_INIT_FAIL_EN | MII_INTEN_LINK_STATUS_FAIL_EN | \ + MII_INTEN_LINK_STATUS_UP_EN | MII_INTEN_SYM_ERR_EN | \ + MII_INTEN_TRAINING_FAILED_EN | MII_INTEN_SQI_WARNING_EN | \ + MII_INTEN_CONTROL_ERR_EN | MII_INTEN_UV_ERR_EN | \ + MII_INTEN_UV_RECOVERY_EN | MII_INTEN_TEMP_ERR_EN | \ + MII_INTEN_SLEEP_ABORT_EN)Why do you enable all these interrupt sources? As I said, phylib needs link change info only.
Quoted 6 lines
Quoted 2 lines
Because I need them to reliably detect that the link state changed.
Hmm, e.g. this one
MII_INTEN_TEMP_ERR_ENdoesn't seem to be related to a link status change. Name sounds like it just reports exceeding a temperature threshold.
It's PHY over-temperature. Whether it tears the link down or not is not clear.
Signature · 2 lines
Best regards, Marek Vasut
Andrew Lunn
maintainerQuoted 54 lines · Marek Vasut
Quoted 53 lines · Heiner Kallweit
Quoted 52 lines · Marek Vasut
Quoted 50 lines · Heiner Kallweit
Quoted 36 lines · Marek Vasut
Add support for handling the TJA11xx PHY IRQ signal.
Signed-off-by Marek Vasut <[email protected]>Cc Andrew Lunn <[email protected]>Cc Florian Fainelli <[email protected]>Cc Guenter Roeck <[email protected]>Cc Heiner Kallweit <[email protected]>Cc Jean Delvare <[email protected]>---
V2: - Define each bit of theMII_INTENregister and a mask
- Drop IRQ acking fromtja11xx_config_intr()--- drivers/net/phy/nxp-tja11xx.c | 48 +++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+)drivers/net/phy/nxp-tja11xx.c +12
-0@@ -40,6 +40,29 @@#define MII_INTSRC_TEMP_ERR BIT(1)#define MII_INTSRC_UV_ERR BIT(3) +#define MII_INTEN 22 +#define MII_INTEN_PWON_EN BIT(15) +#define MII_INTEN_WAKEUP_EN BIT(14) +#define MII_INTEN_PHY_INIT_FAIL_EN BIT(11) +#define MII_INTEN_LINK_STATUS_FAIL_EN BIT(10) +#define MII_INTEN_LINK_STATUS_UP_EN BIT(9) +#define MII_INTEN_SYM_ERR_EN BIT(8) +#define MII_INTEN_TRAINING_FAILED_EN BIT(7) +#define MII_INTEN_SQI_WARNING_EN BIT(6) +#define MII_INTEN_CONTROL_ERR_EN BIT(5) +#define MII_INTEN_UV_ERR_EN BIT(3) +#define MII_INTEN_UV_RECOVERY_EN BIT(2)drivers/net/phy/nxp-tja11xx.c line 64
Quoted 10 lines
drivers/net/phy/nxp-tja11xx.c +10
-0+#define MII_INTEN_TEMP_ERR_EN BIT(1) +#define MII_INTEN_SLEEP_ABORT_EN BIT(0) +#define MII_INTEN_MASK \ + (MII_INTEN_PWON_EN | MII_INTEN_WAKEUP_EN | \ + MII_INTEN_PHY_INIT_FAIL_EN | MII_INTEN_LINK_STATUS_FAIL_EN | \ + MII_INTEN_LINK_STATUS_UP_EN | MII_INTEN_SYM_ERR_EN | \ + MII_INTEN_TRAINING_FAILED_EN | MII_INTEN_SQI_WARNING_EN | \ + MII_INTEN_CONTROL_ERR_EN | MII_INTEN_UV_ERR_EN | \ + MII_INTEN_UV_RECOVERY_EN | MII_INTEN_TEMP_ERR_EN | \ + MII_INTEN_SLEEP_ABORT_EN)Why do you enable all these interrupt sources? As I said, phylib needs link change info only.
Quoted 1 line
Quoted 1 line
Quoted 1 line
Because I need them to reliably detect that the link state changed.
Hi Marek
That statement suggests you started with just bits 10 and 9 and it failed to detect some sort of link up/down event? What was missed?
Andrew
Quoted 55 lines · Andrew Lunn
Quoted 54 lines · Marek Vasut
Quoted 53 lines · Heiner Kallweit
Quoted 52 lines · Marek Vasut
Quoted 50 lines · Heiner Kallweit
Quoted 36 lines · Marek Vasut
Add support for handling the TJA11xx PHY IRQ signal.
Signed-off-by Marek Vasut <[email protected]>Cc Andrew Lunn <[email protected]>Cc Florian Fainelli <[email protected]>Cc Guenter Roeck <[email protected]>Cc Heiner Kallweit <[email protected]>Cc Jean Delvare <[email protected]>---
V2: - Define each bit of theMII_INTENregister and a mask
- Drop IRQ acking fromtja11xx_config_intr()--- drivers/net/phy/nxp-tja11xx.c | 48 +++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+)drivers/net/phy/nxp-tja11xx.c +12
-0@@ -40,6 +40,29 @@#define MII_INTSRC_TEMP_ERR BIT(1)#define MII_INTSRC_UV_ERR BIT(3) +#define MII_INTEN 22 +#define MII_INTEN_PWON_EN BIT(15) +#define MII_INTEN_WAKEUP_EN BIT(14) +#define MII_INTEN_PHY_INIT_FAIL_EN BIT(11) +#define MII_INTEN_LINK_STATUS_FAIL_EN BIT(10) +#define MII_INTEN_LINK_STATUS_UP_EN BIT(9) +#define MII_INTEN_SYM_ERR_EN BIT(8) +#define MII_INTEN_TRAINING_FAILED_EN BIT(7) +#define MII_INTEN_SQI_WARNING_EN BIT(6) +#define MII_INTEN_CONTROL_ERR_EN BIT(5) +#define MII_INTEN_UV_ERR_EN BIT(3) +#define MII_INTEN_UV_RECOVERY_EN BIT(2)drivers/net/phy/nxp-tja11xx.c line 64
Quoted 10 lines
drivers/net/phy/nxp-tja11xx.c +10
-0+#define MII_INTEN_TEMP_ERR_EN BIT(1) +#define MII_INTEN_SLEEP_ABORT_EN BIT(0) +#define MII_INTEN_MASK \ + (MII_INTEN_PWON_EN | MII_INTEN_WAKEUP_EN | \ + MII_INTEN_PHY_INIT_FAIL_EN | MII_INTEN_LINK_STATUS_FAIL_EN | \ + MII_INTEN_LINK_STATUS_UP_EN | MII_INTEN_SYM_ERR_EN | \ + MII_INTEN_TRAINING_FAILED_EN | MII_INTEN_SQI_WARNING_EN | \ + MII_INTEN_CONTROL_ERR_EN | MII_INTEN_UV_ERR_EN | \ + MII_INTEN_UV_RECOVERY_EN | MII_INTEN_TEMP_ERR_EN | \ + MII_INTEN_SLEEP_ABORT_EN)Why do you enable all these interrupt sources? As I said, phylib needs link change info only.
Quoted 6 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Because I need them to reliably detect that the link state changed.
Hi Marek
That statement suggests you started with just bits 10 and 9 and it failed to detect some sort of link up/down event? What was missed?
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Signature · 2 lines
Best regards, Marek Vasut
Andrew Lunn
maintainerQuoted 3 lines
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Then it sounds like you should just ignore interrupts and stay will polling for the TJA1100.
Andrew
Quoted 6 lines · Andrew Lunn
Quoted 3 lines
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Then it sounds like you should just ignore interrupts and stay will polling for the TJA1100.
Polling for the link status change is slow(er) than the IRQ driven operation, so I would much rather use the interrupts.
Signature · 2 lines
Best regards, Marek Vasut
Andrew Lunn
maintainerQuoted 10 lines · Marek Vasut
Quoted 6 lines · Andrew Lunn
Quoted 3 lines
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Then it sounds like you should just ignore interrupts and stay will polling for the TJA1100.
Polling for the link status change is slow(er) than the IRQ driven operation, so I would much rather use the interrupts.
I agree about the speed, but it seems like interrupts on this PHY are not so reliable. Polling always works. But unfortunately, you cannot have both interrupts and polling to fix up problems when interrupts fail. Your call, do you think interrupts really do work?
If you say that tja1101 works as expected, then please just use the link up/down bits for it.
Andrew
Quoted 9 lines · Andrew Lunn
Quoted 8 lines · Marek Vasut
Quoted 6 lines · Andrew Lunn
Quoted 3 lines
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Then it sounds like you should just ignore interrupts and stay will polling for the TJA1100.
Quoted 7 lines
Quoted 2 lines
Polling for the link status change is slow(er) than the IRQ driven operation, so I would much rather use the interrupts.
I agree about the speed, but it seems like interrupts on this PHY are not so reliable. Polling always works. But unfortunately, you cannot have both interrupts and polling to fix up problems when interrupts fail. Your call, do you think interrupts really do work?
It works fine for me this way. And mind you, it's only the TJA1100 that's flaky, the TJA1101 is better.
Quoted 2 lines
If you say that tja1101 works as expected, then please just use the link up/down bits for it.
I still don't know which bits really trigger link status changes, so I'd like to play it safe and just trigger on all of them.
Signature · 2 lines
Best regards, Marek Vasut
Quoted 18 lines · Marek Vasut
Quoted 9 lines · Andrew Lunn
Quoted 8 lines · Marek Vasut
Quoted 6 lines · Andrew Lunn
Quoted 3 lines
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Then it sounds like you should just ignore interrupts and stay will polling for the TJA1100.
Quoted 7 lines
Quoted 2 lines
Polling for the link status change is slow(er) than the IRQ driven operation, so I would much rather use the interrupts.
I agree about the speed, but it seems like interrupts on this PHY are not so reliable. Polling always works. But unfortunately, you cannot have both interrupts and polling to fix up problems when interrupts fail. Your call, do you think interrupts really do work?
Quoted 8 lines
It works fine for me this way. And mind you, it's only the TJA1100 that's flaky, the TJA1101 is better.
Quoted 2 lines
If you say that tja1101 works as expected, then please just use the link up/down bits for it.
I still don't know which bits really trigger link status changes, so I'd like to play it safe and just trigger on all of them.
So what do we do here ?
Signature · 2 lines
Best regards, Marek Vasut
Andrew Lunn
maintainerQuoted 22 lines · Marek Vasut
Quoted 21 lines · Marek Vasut
Quoted 9 lines · Andrew Lunn
Quoted 8 lines · Marek Vasut
Quoted 6 lines · Andrew Lunn
Quoted 3 lines
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Then it sounds like you should just ignore interrupts and stay will polling for the TJA1100.
Quoted 7 lines
Quoted 2 lines
Polling for the link status change is slow(er) than the IRQ driven operation, so I would much rather use the interrupts.
I agree about the speed, but it seems like interrupts on this PHY are not so reliable. Polling always works. But unfortunately, you cannot have both interrupts and polling to fix up problems when interrupts fail. Your call, do you think interrupts really do work?
It works fine for me this way. And mind you, it's only the TJA1100 that's flaky, the TJA1101 is better.
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
If you say that tja1101 works as expected, then please just use the link up/down bits for it.
I still don't know which bits really trigger link status changes, so I'd like to play it safe and just trigger on all of them.
So what do we do here ?
Hi Marek
My personal preference would be to just enable what is needed. But I won't block a patch which enables everything.
Andrew
Quoted 29 lines · Andrew Lunn
Quoted 28 lines · Marek Vasut
Quoted 26 lines · Marek Vasut
Quoted 9 lines · Andrew Lunn
Quoted 8 lines · Marek Vasut
Quoted 6 lines · Andrew Lunn
Quoted 3 lines
The link detection on the TJA1100 (not TJA1101) seems unstable at best, so I better use all the interrupt sources to nudge the PHY subsystem and have it check the link change.
Then it sounds like you should just ignore interrupts and stay will polling for the TJA1100.
Quoted 7 lines
Quoted 2 lines
Polling for the link status change is slow(er) than the IRQ driven operation, so I would much rather use the interrupts.
I agree about the speed, but it seems like interrupts on this PHY are not so reliable. Polling always works. But unfortunately, you cannot have both interrupts and polling to fix up problems when interrupts fail. Your call, do you think interrupts really do work?
It works fine for me this way. And mind you, it's only the TJA1100 that's flaky, the TJA1101 is better.
Quoted 2 lines
If you say that tja1101 works as expected, then please just use the link up/down bits for it.
I still don't know which bits really trigger link status changes, so I'd like to play it safe and just trigger on all of them.
Quoted 6 lines
Quoted 1 line
So what do we do here ?
Hi Marek
My personal preference would be to just enable what is needed. But I won't block a patch which enables everything.
Thanks. I don't know exactly what is needed , but I know that if I enable everything, it works fine. And I'm not getting an interrupt storm either, so it's probably OKish.
Signature · 2 lines
Best regards, Marek Vasut