Mostly rewrite the imx i2c driver. This started out as an attempt to fix
one specific problem: the driver didn't check for ACK/NAK after writing a
slave address byte to the bus, and some slaves signal that they are busy
(such as when completing an internal write to flash memory) by sending a
NAK in response to being addressed.

While working on that problem I discovered that the driver's handling of
error conditions in general didn't match the state transition diagram in
the reference manual, and making that right resulted in a lot of code
reorganization.

Along the way various other changes also happened...

• Remove a mutex that wasn't protecting anything.
• Remove some mystery DELAY()s, document the few that remain.
• Use pause_sbt(9) to yield the processor for the bulk of the time it takes to transfer each byte rather than busy-polling the whole time.
• Disable the controller when no transfers are in progress; since we don't operate in slave mode, there's no reason to run the hardware.
• Remove a bunch of unecessary code from probe().