This stuff isn't yet in the MM tree, but it's in Greg's USB queue.
---------- Forwarded Message ----------
Subject: [patch 0/12] USB updates, mostly for suspend/resume
Date: Thursday 22 September 2005 10:29 pm
From: David Brownell <david-b@xxxxxxxxxxx>
To: linux-usb-devel@xxxxxxxxxxxxxxxxxxxxx
Cc: Greg KH <greg@xxxxxxxxx>
Here are a bunch of USB updates, many making power management
become simpler, for 2.6.15.early merging. They go on top of
2.6.14-rc2 with the patches already in Greg's queue [1].
The patches can be grouped roughly into three subsequences:
* Stuff that's pretty isolated; the first two listed here don't
actually affect anything else in this series.
- pm-shrink.patch ... saves a word per struct device, when
CONFIG_PM is enabled. Actually, safe for 2.6.14 ...
- quirk-ohci.patch ... builds on the PCI quirk patch, fixing
some problems in "early handoff" code for OHCI.
(Depends on the "quirk.patch" that needs some
minor build fixes.)
- ehci-mporter.patch ... Matt Porter's patch to split out
the PCI-specific bits from the ehci-hcd. Minor
updates included, making it build with CONFIG_PM.
- pm-hub.patch ... now with CONFIG_PM, all HCDs provide
access to their root hub suspend/resume methods.
Mostly an issue for some non-PCI OHCI code.
* PM updates for usbcore. This is the stuff that simplifies
things by making CONFIG_USB_SUSPEND mean a lot less.
- pm-root1.patch ... a handful of PM updates that don't
directly affect much else; e.g. a new method that
can make OHCI autosuspend work again.
- pm-root2.patch ... gets rid of lots of special cases
and merges most of the USB_SUSPEND logic in
with normal PM support.
- pm-rootpci.patch ... make the PCI bus glue catch up
to pm-root2.patch changes
- pm-usbsusp2.patch ... we don't want to export these
two routines any more
* HCD updates to match pm-root2.patch; mostly minor/obvious
removal of code. Also included: updates to make some more
swsusp special cases work.
- pm-ohci.patch
- pm-ehci.patch
- pm-uhci.patch
- pm-isp116x.patch
The stack has been tested on a handful of strategic OHCI and EHCI
configurations, involving remote wakeup (1..3 hubs involved),
swsusp disk snapshot/resume, and both USB_SUSPEND settings. But
I'd not be surprised to find a few configurations with bugs,
not necessarily involving UHCI or ISP116x hardware. :)
- Dave
[1] http://www.kernel.org/pub/linux/kernel/people/gregkh/gregkh-2.6/patches/
-------------------------------------------------------
