On 2014-06-20 19:52, H Hartley Sweeten wrote:
Im not sure what was intended with this boardinfo. It appears to be trying to handle the sink/source output type of the boards supported by the driver. But, some of the boardinfo does not match the actual hardware descriptions in the manual. The manual also does not mention anything about the port setting vs. the actual output level. Also, the (*auto_attach) is a bit goofy. If the boardinfo indicates to 'invert_outputs' only the first channel of the port is set. Simplify the driver a bit by just removing this boardinfo. The values written to the "ports" will then be what is passed to comedi by the user and values returned to the user will be the value in the "ports". Signed-off-by: H Hartley Sweeten <hsweeten@xxxxxxxxxxxxxxxxxxx> Cc: Ian Abbott <abbotti@xxxxxxxxx> Cc: Greg Kroah-Hartman <gregk@xxxxxxxxxxxxxxxxxxx>
I've been pondering the best way to handle this. Although the 'invert_outputs' is wrong, it's been wrong for years and people have worked around it in userspace code. Now suddenly their outputs are no longer inverted after updating the kernel.
I was thinking of adding a module parameter to decide whether to honour the 'invert_outputs' for those boards that set it, with the default being to do the right thing (ignoring the 'invert_outputs' bit).
In any case, the auto_attach ought to set all the outputs to the same value to match the internal state.
-- -=( Ian Abbott @ MEV Ltd. E-mail: <abbotti@xxxxxxxxx> )=- -=( Tel: +44 (0)161 477 1898 FAX: +44 (0)161 718 3587 )=- _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
