Major issues:  None

Minor issues:

The mechanism described in this document is fairly simple.  I recommend that the specific semantics of the use of the parameter options should be explained with greater clarity to ensure correct and interoperable implementations.  For example, I found the description of LMA behavior in section 5.1 to be quite convoluted and confusing.  Putting the "if...then...else" construct in two bullets seemed obtuse.  In the first bullet, the LMA "SHOULD include" the sub-option.  Are there circumstances under which the LMA would not include the sub-option and, if so, what are those circumstances?  Can the LMA decide, perhaps for efficiency, to return the sub-option in only, say, one of ten responses to the MAG?

Is there a specific reason for encoding the LAM Controlled MAG Session Parameters as sub-options under the LAM-Controlled-MAG-Parameters option?  Will additional sub-options be defined in the future?

Editorial issues.

For clarity, the document should use acronyms and names for system components in a consistent way: use acronyms throughout and expand the acronym on first use.  For example, LMA and "local mobility anchor" are used interchangeably throughout the document, which this reviewer found to be distracting.

What is the expansion for "PBU"?

The use of the "Configuration Variables" defined in section 4 is repeated in section 5.1.  To avoid internal inconsistency, I recommend that the use of the variable be described only once, with internal pointers to that text from other places in the document.

In section 6, it would help the reader to include the name of the registry to be modified in the first bullet.