The attached is necessary to move the Amstrad Delta LCD driver over to
the new PWL defines that have been pushed.
Signed-off-by: Jonathan McDowell <noodles@earth.li>
Signed-off-by: Tony Lindgren <tony@atomide.com>

ARM: OMAP: LEDs driver support

OMAP GPIO LEDs support
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

This reverts commit da3f63af.

Header file got accidentally left out.

This patch adds pin mux info for the SPI master/slave interface on
OMAP16xx. Data from OMAP 1611/1612 TRM and errata. Works for me on my
1611/H2 with current git kernel.
Signed-off-by: Mark Howell <mhowell@northlink.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

ARM: OMAP: Fix SRAM static mapping for EMU devices.

Fix SRAM static mapping for EMU devices.
Signed-off-by: Kevin Hilman <khilman@deeprooted.net>
Signed-off-by: Tony Lindgren <tony@atomide.com>

This is an updated version of the LCD driver for the Amstrad Delta to
take into account the recent changes to the omapfb infrastructure. The
Delta features a 480x320 12 bit DSTN panel.
Signed-off-by: Jonathan McDowell <noodles@earth.li>
Signed-off-by: Tony Lindgren <tony@atomide.com>

The recent changes to the LCD driver have broken the 444 mode (where a
12 bit panel is driven from 16 bits in memory). This one liner fixes it
for me.
Signed-off-by: Jonathan McDowell <noodles@earth.li>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Fix missing channel parameter using omap_set_dma_priority() in omapfb_main.c:
drivers/video/omap/omapfb_main.c: In function 'omapfb_do_probe':
drivers/video/omap/omapfb_main.c:1441: error: too few
arguments to function
         'omap_set_dma_priority'

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Remove "16XX" from the names of PWL-related defines as they are valid
for other models as well.

ARM: OMAP: LEDs driver support

OMAP GPIO LEDs support
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Apollon board MMC supports on OMAP2

TODO:
        We have to check MMC on H4
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

McBSP does not need sys_ck or sys_clkout. If the devices connected
to McBSP need sys_clkout, they need to request it.
Signed-off-by: Tony Lindgren <tony@atomide.com>

Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Now there are three attributes per switch:

- state
- type
- direction

Also added support for an 'activity' type of switch.
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

We want to keep clocks static. Modify clock code accordingly.

Signed-off-by: Tony Lindgren <tony@atomide.com.

If we call clk_get() from arch_reset we get ugly messages before
reboot.
Signed-off-by: Tony Lindgren <tony@atomide.com>

Backport patch 68aa2c0d.
Without this patch sysrq does not work with 8250 serial console.

- pass dev_id to free_irq in dispc
- stop ESD poll only if it was started in lcd_mipid
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

"clocks" should be static. It is currenly needed only for
CONFIG_OMAP_RESET_CLOCKS. A follow-up patch will fix
CONFIG_OMAP_RESET_CLOCKS while keeping clocks static.
Signed-off-by: Tony Lindgren <tony@atomide.com>

Some time critical work may be put onto the keventd workqueue so let's
not block it with msleep, use a dedicated workqueue instead.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

ls041y3 needs the PCLK,HS,VS signals for a while after sending the sleep
in command, so sleep 50 msec.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Otherwise it will spin forever.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Add support for assigning memory regions dynamically to peripherals
attached to GPMC interface. Platform specific code should now call
gpmc_cs_request to get a free GPMC memory region instead of using
a fixed address.

Make the H4 and Apollon platform initialization use the new API.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Reprogramming takes places before putting the CPU into idle mode if
the dynamic tick option is enabled. The timer is then set to expire
at the next pending timer event. Because some time has already passed
since the last reported jiffy we have to wait less than the time
specified in jiffies.

Also make sure we don't set a load value of 0 whose outcome is
unspecified according to the TRM.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

FB kernel virtual address was left unset causing the FB console crash on
OMAP1.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>

The PWM block on the TI OMAP CPU can drive some external GPIOs.  LEDs
can be connected to these GPIOs.
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

On some LCDs leaving the Vref on provides much better readings.
Signed-off-by: Jarkko Oikarinen <jarkko.oikarinen@nokia.com>
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Some LCDs like the LS041Y3 require a customized filtering
logic for reliable readings, so make the filtering function
replacable through platform specific hooks.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Use hrtimer instead of the normal timer, since it provides better
sampling resolution. This will:

- avoid a problem where we have a 1 jiffy poll period and dynamic tick on
- utilize high resolution HW clocks when they are added to the hrtimer
  framework
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

We can't depend on the pressure value to determine when the pen was
lifted, so use the IRQ line state instead.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

DISPC DMA can, in certain conditions, underflow which will trigger an error
condition and halt the graphics pipeline. The only way to recover from
this would be to reset the related HW modules, which would abort the
current transfer and cause unwanted visual side-effects. The problem
arises mainly when enabling multiple planes with higher resolutions.

The arbitration priority of the DISPC DMA cannot be raised higher, thus -
so far - the only workaround is to keep the DMA FIFO full as much as
possible to minimize the risk of it being totally drained. The patch does
this for external LCD controllers connected to RFBI, where the new FIFO
setting doesn't degrade power consumption significantly.
Signed-off-by: Imre Deak <imre.deak@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Currently, IDLETIMER timer routine calls kobject_uevent, which might sleep.
Since we are in interrupt context, we now schedule a work when the timer
expires. This calls sysfs_notify() for notifying userspace.
Signed-off-by: Samuel Ortiz <samuel.ortiz@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

In ipt_IDLETIMER.c, we are using an old prototype for xt_target hooks. We need
to pass one additional argument, a pointer to xt_target.
We also need to specify the targetsize value, otherwise registration fails.
Signed-off-by: Samuel Ortiz <samuel.ortiz@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Signed-off-by: Kevin Hilman <khilman@deeprooted.net>

Signed-off-by: Samuel Ortiz <samuel.ortiz@solidboot.com>
Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>

Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>