development kernel tree: Changes to 'andy'

[git at git.openmoko.org]

Rebased ref, commits from common ancestor:
commit 455f4da9373af8d8234033342ee998db651bbdb5
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:17:15 2008 +0100

    fix-reote-install-for-ext3-only-sd.patch
    
    Signed-off-by: Andy Green <andy at openmoko.com>

commit 8d846e86d33f0413c2515069df339d08971a3cba
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:17:15 2008 +0100

    defconfig-nuke-cruft.patch
    
    Signed-off-by: Andy Green <andy at openmoko.com>

commit 2c64d189f6895f6bb3b69bbee94adf5d0baa0e3b
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:17:14 2008 +0100

    defconfig-kill-hxd8.patch
    
    Signed-off-by: Andy Green <andy at openmoko.com>

commit 4a41dbf54b697df60750db4c11ab4aa151054236
Author: Mike Westerhof <mwester at dls.net>
Date:   Tue Jul 22 13:17:14 2008 +0100

    fix-touchscreen-driver-gta01-missing-includes.patch
    
        Add missing initialization for the touchscreen driver for the
        gta01 platform.
    
        Signed-off-by: Mike Westerhof <mwester at dls.net>

commit 18da8f31da15bf0a23c4a75daacabc9bba0cbb43
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:17:13 2008 +0100

    test-touchscreen-median.patch
    
    Signed-off-by: Andy Green <andy at openmoko.com>

commit f0e56a2ee67b55ce1be4c58e27d0200d22d78b1b
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:17:13 2008 +0100

    config-touchscreen-filters.patch
    
    Signed-off-by: Andy Green <andy at openmoko.com>

commit 66a83c97c4545ce4f07e0d90998f906fae49caf2
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:16:29 2008 +0100

    fix-glamo-mci-set-default-drive-level-0.patch
    
    Tests on access to SD Card with Glamo drive level "0" show
    that it reduces SD_CLK energy at 1.5GHz by 24dBm compared to
    drive level 3.  This puts it only 6dB above the background
    noise floor compared to 30dB and should make a solution for
    GPS trouble with SD Card in.
    
    SD card communication seems unaffected so far on the Sandisk
    512MB card we ship.
    
    Signed-off-by: Andy Green <andy at openmoko.com>

commit 46159c9a3fba291d106625092fd62358548894e0
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:16:16 2008 +0100

    fix-bq27000-charger-state-tracking.patch
    
    Charger trigger stuff goes and asks for POWER_SUPPLY_PROP_STATUS
    to figure out what the charger state is.  But until now, we only
    reported there what we found out from HDQ, and the HDQ registers
    are not updated very often in the coulomb counter, it can be 4
    or more second lag before it tells us about what it experiences.
    
    When we react to USB insertion and only after 500ms debounce tell
    power_supply stuff that something changed, it most times will
    see old pre-USB-insertion state from bq27000 over HDQ at that time
    and will report it ain't charging, buggering up the LED trigger
    tracking.
    
    This patch maintains distance between bq27000 and pcf50633 by
    having platform callbacks in bq27000 that it can use to ask about
    definitive charger "online" presence and "activity", whether the
    charger says it is charging.  If these callbacks are implemented
    (and we implement them in this patch up in mach_gta02.c) then
    this information is used in preference to what is found from
    HDQ.
    
    Result is if you set the LED trigger like this:
    
    echo bat-charging > /sys/devices/platform/gta02-led.0/leds/gta02-aux:red/trigger
    
    then it lights up properly on USB insertion now, goes away on
    removal properly, as as far as I saw, when charging stops too.
    
    Signed-off-by: Andy Green <andy at openmoko.com>

commit 2eab6880861e27126d059251f88de92ac3e75eb1
Author: Andy Green <andy at openmoko.com>
Date:   Tue Jul 22 13:16:07 2008 +0100

    fix-pcf50633-move-charger-enable-decision-to-pcf50633_usb_curlim_set.patch
    
    At the moment we ended up enabling and disabling the charger
    at the time we service USBINS and USBREM events in the ISR.
    
    We did that in machine specific code, but really it is PMU
    specific decision, so this patch moves it into PMU code.
    
    We enabled and disabled the charge immediately we saw the
    interrupt event from PMU, yet we deferred setting charging
    state based on ID and USB enumeration until at least after a
    PMU ADC read that was handled in a second interrupt later.
    
    Now the timing is unified to after we determined the charger
    type and we enable and disable right there.
    
    Signed-off-by: Andy Green <andy at openmoko.com>