| Message ID | 1465037919-8497-1-git-send-email-hdegoede@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi, On Sat, Jun 4, 2016 at 6:58 PM, Hans de Goede <hdegoede@redhat.com> wrote: > It seems that recent kernels have a shorter timeout when scanning for > ethernet phys causing us to hit a timeout on boards where the phy's > regulator gets enabled just before scanning, which leads to non working > ethernet. > > A 10ms startup delay seems to be enough to fix it, this commit adds a > 20ms startup delay just to be safe. I couldn't find any power sequencing information on the RTL8201CP, which is what I assume these boards are using. The RTL8211E does mention after power ramp up, there's a 20ms delay before the core is ready. If 20ms works for you, I've nothing against it though. Just putting the information out there. > This has been tested on a sun4i-a10-a1000 and sun5i-a10s-wobo-i5 board, > both of which have non-working ethernet on recent kernels without this > fix. > > Cc: stable@vger.kernel.org > Signed-off-by: Hans de Goede <hdegoede@redhat.com> > --- > Changes in v2: > -Add Cc stable > --- > arch/arm/boot/dts/sun4i-a10-a1000.dts | 1 + > arch/arm/boot/dts/sun4i-a10-hackberry.dts | 1 + > arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts | 1 + > arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts | 1 + > 4 files changed, 4 insertions(+) > > diff --git a/arch/arm/boot/dts/sun4i-a10-a1000.dts b/arch/arm/boot/dts/sun4i-a10-a1000.dts > index c92a1ae..fa70b8f 100644 > --- a/arch/arm/boot/dts/sun4i-a10-a1000.dts > +++ b/arch/arm/boot/dts/sun4i-a10-a1000.dts > @@ -84,6 +84,7 @@ > regulator-name = "emac-3v3"; > regulator-min-microvolt = <3300000>; > regulator-max-microvolt = <3300000>; > + startup-delay-us = <20000>; > enable-active-high; > gpio = <&pio 7 15 GPIO_ACTIVE_HIGH>; > }; > diff --git a/arch/arm/boot/dts/sun4i-a10-hackberry.dts b/arch/arm/boot/dts/sun4i-a10-hackberry.dts > index 2b17c51..6de83a6 100644 > --- a/arch/arm/boot/dts/sun4i-a10-hackberry.dts > +++ b/arch/arm/boot/dts/sun4i-a10-hackberry.dts > @@ -66,6 +66,7 @@ > regulator-name = "emac-3v3"; > regulator-min-microvolt = <3300000>; > regulator-max-microvolt = <3300000>; > + startup-delay-us = <20000>; > enable-active-high; > gpio = <&pio 7 19 GPIO_ACTIVE_HIGH>; > }; > diff --git a/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts b/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts > index 7afc7a6..e28f080 100644 > --- a/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts > +++ b/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts > @@ -80,6 +80,7 @@ > regulator-name = "emac-3v3"; > regulator-min-microvolt = <3300000>; > regulator-max-microvolt = <3300000>; > + startup-delay-us = <20000>; > enable-active-high; > gpio = <&pio 7 19 GPIO_ACTIVE_HIGH>; /* PH19 */ > }; > diff --git a/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts b/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts > index 52d18ca..b5de75f 100644 > --- a/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts > +++ b/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts > @@ -79,6 +79,7 @@ > regulator-name = "emac-3v3"; > regulator-min-microvolt = <3300000>; > regulator-max-microvolt = <3300000>; > + startup-delay-us = <20000>; > enable-active-high; > gpio = <&pio 0 2 GPIO_ACTIVE_HIGH>; > }; > -- > 2.7.4 >
Hi, On 06-06-16 11:32, Chen-Yu Tsai wrote: > Hi, > > On Sat, Jun 4, 2016 at 6:58 PM, Hans de Goede <hdegoede@redhat.com> wrote: >> It seems that recent kernels have a shorter timeout when scanning for >> ethernet phys causing us to hit a timeout on boards where the phy's >> regulator gets enabled just before scanning, which leads to non working >> ethernet. >> >> A 10ms startup delay seems to be enough to fix it, this commit adds a >> 20ms startup delay just to be safe. > > I couldn't find any power sequencing information on the RTL8201CP, > which is what I assume these boards are using. > > The RTL8211E does mention after power ramp up, there's a 20ms > delay before the core is ready. > > If 20ms works for you, I've nothing against it though. Just putting the > information out there. Some of them are using the icplus IP101A phy. As said 10 ms works in my testing and I've bumped it to 20 to add some safety margin. Regards, Hans > >> This has been tested on a sun4i-a10-a1000 and sun5i-a10s-wobo-i5 board, >> both of which have non-working ethernet on recent kernels without this >> fix. >> >> Cc: stable@vger.kernel.org >> Signed-off-by: Hans de Goede <hdegoede@redhat.com> >> --- >> Changes in v2: >> -Add Cc stable >> --- >> arch/arm/boot/dts/sun4i-a10-a1000.dts | 1 + >> arch/arm/boot/dts/sun4i-a10-hackberry.dts | 1 + >> arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts | 1 + >> arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts | 1 + >> 4 files changed, 4 insertions(+) >> >> diff --git a/arch/arm/boot/dts/sun4i-a10-a1000.dts b/arch/arm/boot/dts/sun4i-a10-a1000.dts >> index c92a1ae..fa70b8f 100644 >> --- a/arch/arm/boot/dts/sun4i-a10-a1000.dts >> +++ b/arch/arm/boot/dts/sun4i-a10-a1000.dts >> @@ -84,6 +84,7 @@ >> regulator-name = "emac-3v3"; >> regulator-min-microvolt = <3300000>; >> regulator-max-microvolt = <3300000>; >> + startup-delay-us = <20000>; >> enable-active-high; >> gpio = <&pio 7 15 GPIO_ACTIVE_HIGH>; >> }; >> diff --git a/arch/arm/boot/dts/sun4i-a10-hackberry.dts b/arch/arm/boot/dts/sun4i-a10-hackberry.dts >> index 2b17c51..6de83a6 100644 >> --- a/arch/arm/boot/dts/sun4i-a10-hackberry.dts >> +++ b/arch/arm/boot/dts/sun4i-a10-hackberry.dts >> @@ -66,6 +66,7 @@ >> regulator-name = "emac-3v3"; >> regulator-min-microvolt = <3300000>; >> regulator-max-microvolt = <3300000>; >> + startup-delay-us = <20000>; >> enable-active-high; >> gpio = <&pio 7 19 GPIO_ACTIVE_HIGH>; >> }; >> diff --git a/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts b/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts >> index 7afc7a6..e28f080 100644 >> --- a/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts >> +++ b/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts >> @@ -80,6 +80,7 @@ >> regulator-name = "emac-3v3"; >> regulator-min-microvolt = <3300000>; >> regulator-max-microvolt = <3300000>; >> + startup-delay-us = <20000>; >> enable-active-high; >> gpio = <&pio 7 19 GPIO_ACTIVE_HIGH>; /* PH19 */ >> }; >> diff --git a/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts b/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts >> index 52d18ca..b5de75f 100644 >> --- a/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts >> +++ b/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts >> @@ -79,6 +79,7 @@ >> regulator-name = "emac-3v3"; >> regulator-min-microvolt = <3300000>; >> regulator-max-microvolt = <3300000>; >> + startup-delay-us = <20000>; >> enable-active-high; >> gpio = <&pio 0 2 GPIO_ACTIVE_HIGH>; >> }; >> -- >> 2.7.4 >>
On Sat, Jun 04, 2016 at 12:58:39PM +0200, Hans de Goede wrote: > It seems that recent kernels have a shorter timeout when scanning for > ethernet phys causing us to hit a timeout on boards where the phy's > regulator gets enabled just before scanning, which leads to non working > ethernet. > > A 10ms startup delay seems to be enough to fix it, this commit adds a > 20ms startup delay just to be safe. > > This has been tested on a sun4i-a10-a1000 and sun5i-a10s-wobo-i5 board, > both of which have non-working ethernet on recent kernels without this > fix. > > Cc: stable@vger.kernel.org > Signed-off-by: Hans de Goede <hdegoede@redhat.com> Applied, thanks! Maxime
diff --git a/arch/arm/boot/dts/sun4i-a10-a1000.dts b/arch/arm/boot/dts/sun4i-a10-a1000.dts index c92a1ae..fa70b8f 100644 --- a/arch/arm/boot/dts/sun4i-a10-a1000.dts +++ b/arch/arm/boot/dts/sun4i-a10-a1000.dts @@ -84,6 +84,7 @@ regulator-name = "emac-3v3"; regulator-min-microvolt = <3300000>; regulator-max-microvolt = <3300000>; + startup-delay-us = <20000>; enable-active-high; gpio = <&pio 7 15 GPIO_ACTIVE_HIGH>; }; diff --git a/arch/arm/boot/dts/sun4i-a10-hackberry.dts b/arch/arm/boot/dts/sun4i-a10-hackberry.dts index 2b17c51..6de83a6 100644 --- a/arch/arm/boot/dts/sun4i-a10-hackberry.dts +++ b/arch/arm/boot/dts/sun4i-a10-hackberry.dts @@ -66,6 +66,7 @@ regulator-name = "emac-3v3"; regulator-min-microvolt = <3300000>; regulator-max-microvolt = <3300000>; + startup-delay-us = <20000>; enable-active-high; gpio = <&pio 7 19 GPIO_ACTIVE_HIGH>; }; diff --git a/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts b/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts index 7afc7a6..e28f080 100644 --- a/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts +++ b/arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts @@ -80,6 +80,7 @@ regulator-name = "emac-3v3"; regulator-min-microvolt = <3300000>; regulator-max-microvolt = <3300000>; + startup-delay-us = <20000>; enable-active-high; gpio = <&pio 7 19 GPIO_ACTIVE_HIGH>; /* PH19 */ }; diff --git a/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts b/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts index 52d18ca..b5de75f 100644 --- a/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts +++ b/arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts @@ -79,6 +79,7 @@ regulator-name = "emac-3v3"; regulator-min-microvolt = <3300000>; regulator-max-microvolt = <3300000>; + startup-delay-us = <20000>; enable-active-high; gpio = <&pio 0 2 GPIO_ACTIVE_HIGH>; };
It seems that recent kernels have a shorter timeout when scanning for ethernet phys causing us to hit a timeout on boards where the phy's regulator gets enabled just before scanning, which leads to non working ethernet. A 10ms startup delay seems to be enough to fix it, this commit adds a 20ms startup delay just to be safe. This has been tested on a sun4i-a10-a1000 and sun5i-a10s-wobo-i5 board, both of which have non-working ethernet on recent kernels without this fix. Cc: stable@vger.kernel.org Signed-off-by: Hans de Goede <hdegoede@redhat.com> --- Changes in v2: -Add Cc stable --- arch/arm/boot/dts/sun4i-a10-a1000.dts | 1 + arch/arm/boot/dts/sun4i-a10-hackberry.dts | 1 + arch/arm/boot/dts/sun4i-a10-jesurun-q5.dts | 1 + arch/arm/boot/dts/sun5i-a10s-wobo-i5.dts | 1 + 4 files changed, 4 insertions(+)