I’ve posted doc 08-249 which also is a version of a proposal that adds link power management for SAS 2.1 (or whatever the number ends up being).

I had this in the works prior to seeing Mark Evan’s similar 08-206 proposal.  There are some overlaps, and some differences.

Please don’t consider this as a “competing” proposal… I’m just sharing different perspective.  I fully expect there will be a big group hug at some point and converge all the ideas into a single proposal going forward J

Also, and unfortunately, the power management call next week is scheduled when I can not attend, so I wanted to at least post some of my thoughts on the topic.

One general opinion of mine is to not overload the Idle states, and have dedicated power management states for the link and phy to handle it.  But I’ll fully admit this is more of a preference opinion than technical reason.

I’m just concerned with trying to keep track of too many various function all in an Idle state (which means it’s not very idle anymore).

I’ve also included a few thoughts at the end of the doc about how link multiplexing may intersect with power management as well.

Thanks…

Brian Day

LSI Corp.