match the config_activate_children() sequences found in other usb host
controller drivers.

remove prototypes with no matching function
ok mpi@

Constify struct cfattach.

ok visa@ a long time ago, ok patrick@

remove uneeded includes in md armv7 files

based on include-what-you-use suggestions

spelling

use correct node when enabling phy

ok kettenis@

Use PHY framework to replace hardcoded board-specific GPIO poking.

ok jsg@

Dynamically attach omehci using the FDT.
ok kettenis@ patrick@

There is no need to remember which usb(4) device is the child of an USB
host controller because autoconf(9) already does it.

Breakage reported by todd@

After factorizing armv7_machdep code here is an effort to factorize all
the code present in omap/omap.c imx/imx.c and sunxi/sunxi.c

All this code looks quite the same, so we move it in a generic arm

After factorizing armv7_machdep code here is an effort to factorize all
the code present in omap/omap.c imx/imx.c and sunxi/sunxi.c

All this code looks quite the same, so we move it in a generic armv7/armv7.c

This is a step closer to only one kernel for all armv7 boards.

ok patrick@, rapha@

show more ...

In the future, we shouldn't have one port port ARM SoC, that's just
ridiculous. This is the first step for a common and generic ARM port
for ARMv7 SoCs.