SPI: Seems faulty

[robert-hh1]

The hard-SPI mode with device # 0 seems faulty. Observed aspects:

• the setting of phase and polarity does not have the expected result. Looking into the sources, the respective code seems wrong. || instead && in logical expressions, and phase and polarity swapped.
• The bytes of a message are written in reverse order, compared to soft-SPI. Changing the firstbit parameter swaps the byte order. Maybe it is only the default setting which is wrong.
• the method write_readinto() does not work. Data is written, but not read. Looking into the code it seems, that write and read is performed sequential instead of simultaneous. I did not test yet if read works at all.

Soft-SPI (device # -1) works as expected.

[robert-hh]

Tested, and read works. However there is still an problem with SDK and MicroPython with devices, which require sequential commands without de-assigning CS. the way it works now, CS gets active and inactive for each of the SPI transactions. Is there a method to deactivate this automatism? Besides just assigning a dummy pin.
B.t.w.: I added a function to the SDK which offers simultaneous write & read.

[robert-hh]

PR #27 fixing the findings