On 09/19/2014 07:47 AM, Sascha Hauer wrote:
> On Wed, Sep 17, 2014 at 10:22:38PM +0200, Sebastian Hesselbarth wrote:
>> This is v2 of the Multi-SoC series for MVEBU. It has been tested
>> on Kirkwood (Guruplug), Dove (CuBox), and Armada 370 (Mirabox)
>> with the very same non-EABI binary. It should
Sebastian, Ezequiel,
On Wed, Sep 17, 2014 at 10:22:38PM +0200, Sebastian Hesselbarth wrote:
> This is v2 of the Multi-SoC series for MVEBU. It has been tested
> on Kirkwood (Guruplug), Dove (CuBox), and Armada 370 (Mirabox)
> with the very same non-EABI binary. It should also work on Armada
> XP b
On Wed, Sep 17, 2014 at 12:25:08PM +0200, Steffen Trumtrar wrote:
> Hi!
>
> Changes since v1:
>
> - removed clocks.imxcfg -> actually not needed
> - removed magic values in flash-header-tx6dl-1g -> not needed
> - removed barebox-usdhc-environment
> - include imx6qdl.dtsi f
Signed-off-by: Michael Olbrich
---
arch/efi/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/efi/Makefile b/arch/efi/Makefile
index af28085..85cb655 100644
--- a/arch/efi/Makefile
+++ b/arch/efi/Makefile
@@ -24,7 +24,7 @@ cmd_barebox__ ?= $(LD) $(LDFLAGS) $(LDFLAG
Using 'dev' as name for variables which are not of type
struct device(_d) is bad habit and leads to confusions. Use phydev
instead.
Signed-off-by: Sascha Hauer
---
drivers/net/phy/phy.c | 78 +--
1 file changed, 39 insertions(+), 39 deletions(-)
d
No need to call kzalloc for small allocations, xzalloc will do without
the need for an additional check.
Signed-off-by: Sascha Hauer
---
drivers/net/phy/phy.c | 9 ++---
1 file changed, 2 insertions(+), 7 deletions(-)
diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c
index 7604e1d.
This series gets direct phy register accesses out of the generic code
path. This is necessary for phys which have a different register
layout not handled correctly by the generic code. Also the phy link
status check is moved to a poller callback. With this we get periodic
link status updates printe
phy_wait_aneg_done() is directly called by the network code, so it
should not read phy registers directly. Introduce phy_aneg_done to
give phy drivers the chance to do something different to poll for
autonegotiation completion.
Signed-off-by: Sascha Hauer
---
drivers/net/phy/phy.c | 57 +
This continuously updates the link status in the background. The networking
code no longer has to periodically update the link status itself but instead
can only check for phydev->link.
With this we also always have link status changes printed to the console.
Signed-off-by: Sascha Hauer
---
driv