C1 vs odd-word violations need improvement

[pkimpel]

The C1 (low-order) bit of a command does not always signify a double precision operation. In some operations it is not relevant (e.g., copy to/from destinations < 24 with characteristic=0) and in others it has a different purpose (e.g., MARK RETURN, MARK EXIT, and Slow-Out I/O). The current test for (C1 && !CE) in transferDriver() is too restrictive and probably needs to be broken up and distributed to the appropriate places in the code.

[pkimpel]

As of version 1.00, the CODE VIOLATION feature has been suppressed, so this issue has been placed on indefinite hold.