Message ID | 1292756331-3735-1-git-send-email-jhautbois@gmail.com |
---|---|
State | Rejected, archived |
Delegated to: | David Miller |
Headers | show |
From: Jean-Michel Hautbois <jhautbois@gmail.com> Date: Sun, 19 Dec 2010 11:58:48 +0100 > When phy interface changes its status, it calls phy_change() function. > This function calls the interrupt disabling functions for the driver > registered, but if this driver doesn't implement it, there is no IRQ > disabling. After doing the work, we call enable_irq and not the > respective driver function. This fixes it, as it could lead to an > unbalanced IRQ. Error code changed to EOPNOTSUPP. > > Signed-off-by: Jean-Michel Hautbois <jhautbois@gmail.com> This is completely bogus. First of all, there are 5 call sites for phy_change_interrupt() but you've only implemented the new semantics for two of those. Therefore, if we even wanted this, we should implement the behavior in phy_change_interrupt() itself instead of duplicating the logic at each and every call site. But we don't want this. It's not appropriate at all. If a device lacks a way to turn interrupt off and on, using disable_irq() and enable_irq() is not necessarily correct. If the interrupt line is shared, for example, this will break everything. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
2010/12/23 David Miller <davem@davemloft.net>: > From: Jean-Michel Hautbois <jhautbois@gmail.com> > Date: Sun, 19 Dec 2010 11:58:48 +0100 > >> When phy interface changes its status, it calls phy_change() function. >> This function calls the interrupt disabling functions for the driver >> registered, but if this driver doesn't implement it, there is no IRQ >> disabling. After doing the work, we call enable_irq and not the >> respective driver function. This fixes it, as it could lead to an >> unbalanced IRQ. Error code changed to EOPNOTSUPP. >> >> Signed-off-by: Jean-Michel Hautbois <jhautbois@gmail.com> > > This is completely bogus. > > First of all, there are 5 call sites for phy_change_interrupt() but > you've only implemented the new semantics for two of those. > > Therefore, if we even wanted this, we should implement the behavior in > phy_change_interrupt() itself instead of duplicating the logic at > each and every call site. > > But we don't want this. OK, I understand that point. > It's not appropriate at all. If a device lacks a way to turn > interrupt off and on, using disable_irq() and enable_irq() is not > necessarily correct. > > If the interrupt line is shared, for example, this will break > everything. > OK, well, maybe is there at least one thing we could do : in phy_change, instead of calling phy_disable_interrupts(), balanced by enable_irq, we probably should use phy_enable_interrupts(). -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c index 7670aac..5f23e8e 100644 --- a/drivers/net/phy/phy.c +++ b/drivers/net/phy/phy.c @@ -89,7 +89,8 @@ static int phy_config_interrupt(struct phy_device *phydev, u32 interrupts) phydev->interrupts = interrupts; if (phydev->drv->config_intr) err = phydev->drv->config_intr(phydev); - + else + err = -EOPNOTSUPP; return err; } @@ -541,6 +542,10 @@ static int phy_enable_interrupts(struct phy_device *phydev) return err; err = phy_config_interrupt(phydev, PHY_INTERRUPT_ENABLED); + if (err == -EOPNOTSUPP) { + err = 0; + enable_irq(phydev->irq); + } return err; } @@ -556,7 +561,10 @@ static int phy_disable_interrupts(struct phy_device *phydev) /* Disable PHY interrupts */ err = phy_config_interrupt(phydev, PHY_INTERRUPT_DISABLED); - if (err) + if (err == -EOPNOTSUPP) { + err = 0; + disable_irq(phydev->irq); + } else if (err != 0) goto phy_err; /* Clear the interrupt */
When phy interface changes its status, it calls phy_change() function. This function calls the interrupt disabling functions for the driver registered, but if this driver doesn't implement it, there is no IRQ disabling. After doing the work, we call enable_irq and not the respective driver function. This fixes it, as it could lead to an unbalanced IRQ. Error code changed to EOPNOTSUPP. Signed-off-by: Jean-Michel Hautbois <jhautbois@gmail.com> --- drivers/net/phy/phy.c | 12 ++++++++++-- 1 files changed, 10 insertions(+), 2 deletions(-)