Patchwork Doubts about the get_phy_id function.

login
register
mail settings
Submitter Giuseppe CAVALLARO
Date Oct. 29, 2008, 9:12 a.m.
Message ID <490828F6.4000307@st.com>
Download mbox | patch
Permalink /patch/6233/
State Rejected
Delegated to: Jeff Garzik
Headers show

Comments

Giuseppe CAVALLARO - Oct. 29, 2008, 9:12 a.m.
Hi All,
Using a bad PHY address, the physical device is erroneously attached to 
the MAC device driver.
This happens because the get_phy_id function reads zero from the 
PHYIR1/2 regs and never fails.
I'm facing this problem on some PHY devices (*) but I've not clear 
enough if this can actually depend on an HW issue (e.g.invalid  phy 
addresses have been left floating) or we need to apply the patch below. 
Indeed, it helps me to catch the problem and, as final result, the 
phy_connect always fails on wrong PHY address and the network interface 
cannot be opened.

(*) on other ones, the get_phy_id function reads 0xffff from PHYIR1/2 
regs and fails (as expected results) when I try to attach a wrong PHY 
address.

Welcome advice.
Best Regards,
Peppe

PS: if it makes sense for you, I can rework the patch and send it to the 
ML again.

        *phy_id = (phy_reg & 0xffff) << 16;
@@ -200,7 +200,7 @@ int get_phy_id(struct mii_bus *bus, int addr, u32 
*phy_id)
        /* Grab the bits from PHYIR2, and put them in the lower half */
        phy_reg = bus->read(bus, addr, MII_PHYSID2);
 
-       if (phy_reg < 0)
+       if (phy_reg <= 0)
                return -EIO;
 
        *phy_id |= (phy_reg & 0xffff);
--
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
Andy Fleming - Nov. 3, 2008, 8:24 p.m.
On Wed, Oct 29, 2008 at 3:12 AM, Giuseppe CAVALLARO
<peppe.cavallaro@st.com> wrote:
> Hi All,
> Using a bad PHY address, the physical device is erroneously attached to the
> MAC device driver.
> This happens because the get_phy_id function reads zero from the PHYIR1/2
> regs and never fails.
> I'm facing this problem on some PHY devices (*) but I've not clear enough if
> this can actually depend on an HW issue (e.g.invalid  phy addresses have
> been left floating) or we need to apply the patch below. Indeed, it helps me
> to catch the problem and, as final result, the phy_connect always fails on
> wrong PHY address and the network interface cannot be opened.
>
> (*) on other ones, the get_phy_id function reads 0xffff from PHYIR1/2 regs
> and fails (as expected results) when I try to attach a wrong PHY address.

Some people seem to have PHYs that report 0 as their ID, so that's not
going to work.  I'm pretty sure your bus needs to be floating high, so
you might want to check that out.

Andy
--
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

Patch

diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c
index 8ad602c..8a1baf9 100644
--- a/drivers/net/phy/phy_device.c
+++ b/drivers/net/phy/phy_device.c
@@ -192,7 +192,7 @@  int get_phy_id(struct mii_bus *bus, int addr, u32 
*phy_id)
         * in the upper half */
        phy_reg = bus->read(bus, addr, MII_PHYSID1);
 
-       if (phy_reg < 0)
+       if (phy_reg <= 0)
                return -EIO;