[PATCHSET 1/3] implement ata_link, take 2

[Tejun Heo <htejun@gmail.com>]

Hello, all.

This is part of patchset series described in [T].

This is the second take of implement-ata_link patchset.  This patchset
contains 17 patches which can be categorized as follows.

#01-02: prep
#03-04: implement ata_link
#05-08: make libata deal with link instead of port
#09-11: misc link stuff (link init, reset_tries, ata_link_abort())
#12-17: add ap->pmp_link[] and update libata to deal with multiple
	links

ata_link abstracts PHY and sits between ata_port and ata_device.  The
following attributes are moved to ata_link from ata_port.

- active command state (active_tag, sactive)
- [hw_]sata_spd_limit
- eh_info and eh_context
- device array

With above and a few extra fields, a link can fully host attached
devices including qc management and EH/hotplug.  This patchset makes
libata ready to handle PM links.

Changes from the last take[L] are.

* updated to fit new #upstream
* ata_eh_handle_dev_fail() is separated out from ata_eh_recover() to
  ease making EH batch later.
* Link resume handling udpated.
* Power Management is linkified.

This patchset is against

  upstream (309bade002e9226781c2d7a015340d0089e399b5)
  + [1] hp-poll patchset, take #2

Thanks.

--
tejun

[T] http://article.gmane.org/gmane.linux.ide/11927
[L] http://article.gmane.org/gmane.linux.ide/10106
[1] http://article.gmane.org/gmane.linux.ide/11862