Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

If we accidentally pass an incorrect bpp value to pixel_to_pat(),
it panics. This is pretty useless, as we generally have the various
console locks held at that point, so nothing will be displayed,
and there is no reason to make this a fatal event.

Let's WARN instead.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

LCD blink is observed during suspend/resume and blank/unblank
operations as backlight is ON during LCDC disable and enable.
So make sure to turn OFF backlight before disabling and turn
it ON after enabling.

Signed-off-by: Manjunathappa, Prakash <prakash.pm@ti.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Commit 15dede88

 added support for
horizontal panning but accidentally computes the Y pan step value
incorrectly for NV12/21 and NV16/61 formats. Fix this.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

When FB_BLANK_UNLANK event occured, exynos mipi dsi regulators have to turn on.

Signed-off-by: Donghwa Lee <dh09.lee@samsung.com>
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

As in specification software reset should be applied for several
cycles before bringing it out of reset. Without this patch
particularly during suspend and resume clock reset is not guaranteed
to happen.

Signed-off-by: Manjunathappa, Prakash <prakash.pm@ti.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Configure fifo threshold in raster dma_control register to
512 bytes. This reduces LCDC underflow errors.

Signed-off-by: Manjunathappa, Prakash <prakash.pm@ti.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Patch works around the below silicon errata:
During LCDC initialization, there is the potential for a FIFO
underflow condition to occur. A FIFO underflow condition
occurs when the input FIFO is completely empty and the LCDC
raster controller logic that drives data to the output pins
attempts to fetch data from the FIFO. When a FIFO underflow
condition occurs, incorrect data will be driven out on the
LCDC data pins.

Software should poll the FUF bit field in the LCD_STAT register
to check if an error condition has occurred or service the
interrupt if FUF_EN is enabled when FUF occurs. If the FUF bit
field has been set to 1, this will indicate an underflow
condition has occurred and then the software should execute a
reset of the LCDC via the LPSC.

This problem may occur if the LCDC FIFO threshold size
(LCDDMA_CTRL[TH_FIFO_READY]) is left at its default value after
reset. Increasing the FIFO threshold size will reduce or
eliminate underflows. Setting the threshold size to 256 double
words or larger is recommended.

Above issue is described in section 2.1.3 of silicon errata
http://www.ti.com/lit/er/sprz313e/sprz313e.pdf



Signed-off-by: Rajashekhara, Sudhakar <sudhakar.raj@ti.com>
Signed-off-by: Manjunathappa, Prakash <prakash.pm@ti.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Flicker/tearing effect is observed with current FB driver.
Issue is because of 2 active DMA channels ping ponging among them
along with usage of 2 DDR ping pong buffers in driver. Application
unaware of active DMA channel keeps updating frame being displayed,
this leads to tearing effect.
Below steps describes the issue:
1)Initially assume both buffers FB0 and FB1 are programmed for buffer-0.
2)On EOF0: Program FB0 for buffer-1, indicate(wake up) application
 to fill up buffer-0. As FB1 is active and continues to DMA buffer-0
(which is being filled), leading to tearing/flickering issue.
3)On EOF1: Program FB1 for buffer-0, indicate(wake up) application to
 fill up buffer-1. As FB0 is active and continues to DMA buffer-1(which
 is being filled), leading to tearing/flickering issue.
4)On EOF0: Program FB0 for buffer-1, indicate(wake up) application to
fill up buffer-0. As FB1 is active and continues to DMA buffer-0(which is
being filled), leading to tearing/flickering issue.
...
Above steps depict that issue is because of 1 frame delay in frame
panned by application.

Patch fixes the issue by keeping track free DMA channel and configures
it in drivers PAN callback so that panned frame from application gets
displayed in next frame period.

Wiki below describes the issue in detail and it also has link to
application with which issue can be reproduced.
http://processors.wiki.ti.com/index.php/DA8xx_LCDC_Linux_FB_FAQs



Signed-off-by: Nellutla, Aditya <aditya.n@ti.com>
Signed-off-by: Manjunathappa, Prakash <prakash.pm@ti.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Writing '1' to particular bit of IRQENABLE_CLEAR register disables the
corresponding interrupt on revision 2 LCDC. This register was wrongly
configured to disable all previous enabled interrupts instead of
disabling only palette completion interrupt. Patch fixes it by clearing
only palette completion interrupt bit.

Signed-off-by: Manjunathappa, Prakash <prakash.pm@ti.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

add FB_BLANK_NORMAL, FB_BLANK_VSYNC_SUSPEND and FB_BLANK_HSYNC_SUSPEND
modes (copy drivers/video/omap2/omapfb/omapfb-main.c implementation).
Otherwise X-server will complain about invalid parameter.

Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Conflicts:
	drivers/video/omap2/dss/core.c
	drivers/video/omap2/dss/dispc.c

This patch replaces udelay and mdelay with usleep_range to remove
the busy loop waiting.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

The only INTERLANE_ALIGN_DONE bit should be checked for channel equalization
during Link Training. Previously, the other bits such as LINK_STATUS_UPDATED
were checked, and channel equalization procedure was repeated unnecessarily.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

This patch fixes "section mismatch" warning in the epson1355fb driver.

WARNING: vmlinux.o(.devinit.text+0x184): Section mismatch in reference from the function epson1355fb_probe() to the function .init.text:fetch_hw_state()
The function __devinit epson1355fb_probe() references
a function __init fetch_hw_state().
If fetch_hw_state is only used by epson1355fb_probe then
annotate fetch_hw_state with a matching annotation.

Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Acked-by: Christopher Hoover <ch@murgatroid.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Wrong DPCD addresses were used for clock recovery during Link Training.
The training pattern should be set by TRAINING_PATTERN_SET (0x102), while
voltage swing and pre-emphasis should be set by TRAINING_LANE0_SET (0x103).

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

Line 0 and 1 were both written to line 0 (on the display) and all subsequent
lines had an offset of -1. The result was that the last line on the display
was never overwritten by writes to /dev/fbN.

The origin of this bug seems to have been udlfb.

Signed-off-by: Alexander Holler <holler@ahsoftware.de>
Cc: stable@vger.kernel.org
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

This patch cleans up some coding style issues.

-Some lines are indented with 4 spaces, most of this code
 is not used but it should be correctly indented anyway.

-I also fixed some long lines exceeding the 80 char limit.

Signed-off-by: Emil Goode <emilgoode@gmail.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

The chroma plane offset in memory is equal to the luma plane maximum
size. Fix offset computations.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Updating overlay registers require switching to overlay update mode.
This was correctly done when configuring the overlay format and size,
but not when updating the base address registers during pan operation.
Fix it.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Copy the x and y virtual resolutions from the channel information
instead of recomputing them.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Line pitch depends on the virtual horizontal resolution, compute it
accordingly.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Add a missing mutex_destroy() call when the driver is unbound from the
device.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

The API can be used to allocate and free MERAM blocks directly, without
going through ICBs.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

There's no reason to use abstract operation pointers to implement the
MERAM API. Replace them by direct function calls.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

The MERAM operations meram_register, meram_unregister and meram_update
handle LCDC cache. In preparation for "raw" MERAM allocation, rename
them to more appropriate names.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

Add support for Virge/MX (86C260) chip. Although this is a laptop chip,
there's an AGP card with this chip too.

Tested with AGP card, will probably not work correctly with laptops.
DDC does not work on this card (even in DOS or Windows).

Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

This patch fixes two problems with the error handling in the
grvga_probe function and simplifies it making the code
easier to read.

- If the call to grvga_parse_custom on line 370 fails we use
  the wrong label so that release_mem_region will be called
  without a call to request_mem_region being made.

- If the call to ioremap on line 436 fails we should not try
  to call iounmap in the error handling code.

This patch introduces the following changes:

- Converts request_mem_region into its devm_ equivalent
  which simplifies the otherwise messy clean up code.

- Changes the labels for correct error handling and their
  names to make the code easier to read.

Signed-off-by: Emil Goode <emilgoode@gmail.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

In 2006 and 2007 the handhelds.org kernel w100fb driver was patched to
reduce sleep mode battery discharge. Unfortunately those two patches
never migrated to the mainline kernel.

Fortunately the function affected - w100_suspend() - has not changed
since; thus those patches still apply cleanly.

Applying those patches to linux-3.4-rc3 running on an HP iPAQ hx4700
reduces the sleep mode battery discharge from approximately 26 mA to
approximately 11 mA.

This patch combines those two patches into a single unified patch.

Signed-off-by: Paul Parsons <lost.distance@yahoo.com>
Cc: Matt Reimer <mreimer@sdgsystems.com>
Cc: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

If runtime PM is not enabled in the kernel config, pm_runtime_get_sync()
will always return 1 and pm_runtime_put_sync() will always return
-ENOSYS. pm_runtime_get_sync() returning 1 presents no problem to the
driver, but -ENOSYS from pm_runtime_put_sync() causes the driver to
print a warning.

One option would be to ignore errors returned by pm_runtime_put_sync()
totally, as they only say that the call was unable to put the hardware
into suspend mode.

However, I chose to ignore the returned -ENOSYS explicitly, and print a
warning for other errors, as I think we should get notified if the HW
failed to go to suspend properly.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: Jassi Brar <jaswinder.singh@linaro.org>
Cc: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Archit Taneja <archit@ti.com>
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

The current way how omapdss handles system suspend and resume is that
omapdss device (a platform device, which is not part of the device
hierarchy of the DSS HW devices, like DISPC and DSI, or panels.) uses
the suspend and resume callbacks from platform_driver to handle system
suspend. It does this by disabling all enabled panels on suspend, and
resuming the previously disabled panels on resume.

This presents a few problems.

One is that as omapdss device is not related to the panel devices or the
DSS HW devices, there's no ordering in the suspend process. This means
that suspend could be first ran for DSS HW devices and panels, and only
then for omapdss device. Currently this is not a problem, as DSS HW
devices and panels do not handle suspend.

Another, more pressing problem, is that when suspending or resuming, the
runtime PM functions return -EACCES as runtime PM is disabled during
system suspend. This causes the driver to print warnings, and operations
to fail as they think that they failed to bring up the HW.

This patch changes the omapdss suspend handling to use PM notifiers,
which are called before suspend and after resume. This way we have a
normally functioning system when we are suspending and resuming the
panels.

This patch, I believe, creates a problem that somebody could enable or
disable a panel between PM_SUSPEND_PREPARE and the system suspend, and
similarly the other way around in resume. I choose to ignore the problem
for now, as it sounds rather unlikely, and if it happens, it's not
fatal.

In the long run the system suspend handling of omapdss and panels should
be thought out properly. The current approach feels rather hacky.
Perhaps the panel drivers should handle system suspend, or the users of
omapdss (omapfb, omapdrm) should handle system suspend.

Note that after this patch we could probably revert
0eaf9f52

 (OMAPDSS: use sync versions of
pm_runtime_put). But as I said, this patch may be temporary, so let's
leave the sync version still in place.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Reported-by: Jassi Brar <jaswinder.singh@linaro.org>
Tested-by: Jassi Brar <jaswinder.singh@linaro.org>
Tested-by: Joe Woodward <jw@terrafix.co.uk>
Signed-off-by: Archit Taneja <archit@ti.com>
[fts: fixed 2 brace coding style issues]
Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>

The LCD interface drivers(DPI, DSI, RFBI, SDI) do some direct DISPC register
writes to configure LCD manager related fields. This series groups these fields
into a single struct, and let's the interface driver apply these parameters.

This allows us to:

- Check the LCD manager related parameters before applying them.
- Remove some omap_dss_device references as APPLY holds the applied parameters.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>

Replication logic for an overlay depends on the color mode in which it is
configured and the video port width of the manager it is connected to.

video port width now held in dss_lcd_mgr_config in the manager's private
data in APPLY. Use this instead of referring to the omap_dss_device connected to
the manager.

Replication is enabled in the case of TV manager, the video_port_width is set to
a default value of 24 for TV manager.

Make the replication checking an overlay function since it's more of an overlay
characteristic than a display characteristic.

Signed-off-by: Archit Taneja <archit@ti.com>

The RFBI driver uses a direct DISPC register write to enable the overlay
manager. Replace this with dss_mgr_enable() which checks if the connected
overlay and managers are correctly configured, and configure DSS for
fifomerge.

Signed-off-by: Archit Taneja <archit@ti.com>

dss_mgr_is_lcd() available in dss.h does the same thing as dispc_mgr_is_lcd()
in dispc.c. Remove the function from dispc.c and replace it with the one in
dss.h.

Signed-off-by: Archit Taneja <archit@ti.com>

APPLY needs to know at certain places whether an overlay manager is in manual
or auto update mode. The caps of the connected omap_dss_device were used to
check that.

A LCD manager is in manual update if stallmode is enabled for that manager. TV
managers for now always auto update.

Return the value of stallmode parameter in the private data 'lcd_confg' in
mgr_manual_update() and ovl_manual_update(), for TV managers stallmode field
will be false by default.

Signed-off-by: Archit Taneja <archit@ti.com>

The LCD related manager configurations are a part of the manager's private data
in APPLY. Pass this to dss_lcd_mgr_config to dss_mgr_check and create a function
to check the validity of some of the configurations.

To check some of the configurations, we require information of interface to
which the manager output is connected. These can be added once interfaces are
represented as an entity.

Signed-off-by: Archit Taneja <archit@ti.com>