We can use __exit for the driver remove function in plain dss panels
(ie. those that do not need i2c or spi).
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

omap_dss_register_device and omap_dss_unregister_device can only be
called from core.c, so we can make it static.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

VENC code was missing omap_dss_start/stop_device calls. This didn't
cause any problems as VENC could not be compiled as a module, but
nevertheless it's better to add the calls.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

Currently omapfb does an implicit display update (for manual update
displays) on unblank.

There is no guarantee that the framebuffer contains a valid image when
unblank is called. When using manual update displays it is the
responsibility of the user space to update the display, and so it should
be in this case also.

This patch removes the implicit display update on unblank.
Signed-off-by: NJani Nikula <ext-jani.1.nikula@nokia.com>
[tomi.valkeinen@ti.com: improved description]
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

Errors from the panel driver were ignored during panel initialization.
Handle the errors and fail accordingly.

Also move the display initialization to a separate function to make it
cleaner.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

Add get_dimensions() to struct omap_dss_driver. Use the call, if supported
by the driver, in OMAPFB.
Signed-off-by: NJani Nikula <ext-jani.1.nikula@nokia.com>
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

VENC code had 50ms sleep after enabling the output and 100ms sleep after
disabling the output. I don't see any reason for these sleeps.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

There's a 20ms sleep after VENC reset. It's unknown what bug this
circumvents and on what platforms. Add a Kconfig option to disable the
sleep.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

For some unknown reason we may get SYNC_LOST errors from the display
subsystem at initialization time if we don't sleep before resetting the
DSS. See the source (dss.c) for more comments.

However, 50ms is quite long time to sleep, and with some configurations
the SYNC_LOST may never happen, so this patch creates a Kconfig option
to disable the sleep.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

The sleep workaround for the problem described in the comment doesn't
really work. This patch removes the workaround, and improves the comment
about the bug, and proposes an userspace workaround.

omap_dss_set_manager() is called 6 times when loading omapdss and
omapfb, which means that 40ms * 6 = 240ms was spent sleeping when
booting up.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

Clock configuration was defined inside dssdev.phy.dsi struct. The clock
config doesn't really belong there, and so it's moved to dssdev.clock
struct.

Now the explicit clock configuration could also be used for other
interfaces than DSI, although there's no support for it currently.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

The DSI protocol engine has no interrupt for signalling the end of a Frame
transfer. The present approach is to send a BTA after DISPC generates a
FRAMEDONE interrupt, and unlock the dsi bus only when the BTA Ack is received.

The assumption made with this approach was that OMAP will send a BTA only after
the long packet corresponding to the last line is sent. However, it is possible
that on the DISPC FRAMEDONE interrupt there are 2 (or more) lines of pixel data
in the DSI line buffer. Hence, the BTA Ack could be received for the long packet
corresponding to the second last line (or the third last and so on..).
Therefore, the current method doesn't ensure that the complete frame data is
sent before we start a new transfer. A similar explanation holds valid if we
send a BTA in between multiple short/long command packets from the slave port.

Introduce dsi_sync_vc functions, based on Tomi Valkeinen's idea, which ensure
that the DSI Virtual Channel in use(update_channel) completes its previous work
before proceeding to the next Frame/Command.

For a frame update, the DSI driver now sends a callback to the Panel Driver
on the FRAMEDONE interrupt itself. The callback in the panel driver then unlocks
the bus. dsi_sync_vc() functions are placed in dsi_vc_config_l4() and
dsi_vc_config_vp() to ensure that the previous task of the Virtual Channel is
completed.
Signed-off-by: NArchit Taneja <archit@ti.com>
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

The following changes have changed from OMAP3 to OMAP4 DSI:
-The register field DSI_PLL_FREQSEL in DSI_PLL_CONFIGURATION2 has been removed.
-DCS_CMD_ENABLE and DCS_CMD_CODE bits have been moved from DSI_CTRL to
DSI_VC_CTRLi, hence the control of the bits is available per VC.
-DSI LDO powergood notification doesn't work on OMAP4. This is mentioned in
OMAP4 errata revision 1.8(Errata 1.76).
-OCP_WIDTH register field is included in DSI_VC_CTRL.
-The SCP clock is also required to access DSI PLL registers

Introduce dss features for these changes so that DSI runs on both OMAP3 and
OMAP4.
Signed-off-by: NArchit Taneja <archit@ti.com>
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

All DPI panels were missing dependency to OMAP2_DSS_DPI. Add the
dependency.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

OMAP3630 has a HW bug causing DSI PLL power command POWER_ON_DIV (0x3)
to not work properly. The bug prevents us from enabling DSI PLL power
only to HS divider block.

This patch adds a dss feature for the bug and converts POWER_ON_DIV
requests to POWER_ON_ALL (0x2).
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

dss.lcd_clk_source is set to the default value DSS_CLK_SRC_FCK at dss_init.
For OMAP2 and OMAP3, the dss.lcd_clk_source should always be the same as
dss.dispc_clk_source. The function dss_get_lcd_clk_source() always returns the
default value DSS_CLK_SRC_FCK for OMAP2/3. This leads to wrong clock dumps when
dispc_clk_source is not DSS_CLK_SRC_FCK.

Correct this function to always return dss.dispc_clk_source for OMAP2/3.
Signed-off-by: NArchit Taneja <archit@ti.com>
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

On OMAP4, reading DSI_PLL_CONFIGURATION2 register requires the L3 clock
(CIO_CLK_ICG) to PLL. Currently dsi_dump_clocks() tries to read that
register without enabling the L3 clock, leading to crash if DSI is not
in use.

The status of the bit being read from DSI_PLL_CONFIGURATION2 is
available from dsi_clock_info->use_sys_clk, so we can avoid the whole
problem by just using that.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

use_sys_clk and highfreq fields in dsi.current_cinfo were never set.
Luckily they weren't used anywhere so it didn't cause any problems.

This patch fixes those fields and they are now set at the same time as
the rest of the fields.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

arch/arm/plat-omap/include/plat/nokia-dsi-panel.h is an include for the
OMAP DSS panel driver for Nokia's DSI displays. A more logical place for
it is in include/video.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

arch/arm/plat-omap/include/plat/panel-generic-dpi.h is an include for
the OMAP DSS panel driver for generic DPI displays. A more logical place
for it is in include/video.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

arch/arm/plat-omap/include/plat/display.h is an include for the OMAP DSS
driver. A more logical place for it is in include/video.
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

If we're using vga switcheroo, the device may be turned off
and poking it can return random state. This provokes an OOPS fixed
separately by 8ff887c847 (drm/i915/dp: Be paranoid in case we disable a
DP before it is attached). Trying to use and respond to events on a
device that has been turned off by the user is in principle a silly thing
to do.
Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org
Signed-off-by: NKeith Packard <keithp@keithp.com>

Despite the fixes in 548f245b (drm/i915: fix per-pipe reads after
"cleanup"), we missed one neighbouring read that was mistakenly replaced
with the reg value in 9db4a9c7 (drm/i915: cleanup per-pipe reg usage).
This was preventing us from correctly determining the mode the BIOS left
the panel in for machines that neither have an OpRegion nor access to
the VBT, (e.g. the EeePC 700).
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Cc: stable@kernel.org
Reviewed-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NKeith Packard <keithp@keithp.com>

When enabling the plane, it is helpful to have already pointed that
plane to valid memory or else we may incur the wrath of a PGTBL_ER.
This code preserved the behaviour from the bad old days for unknown
reasons...

Found by assert_fb_bound_for_plane().

References: https://bugs.freedesktop.org/show_bug.cgi?id=36246Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NKeith Packard <keithp@keithp.com>

The ACPI notification we get from rfkill events on these machines gives
us all the information we need to identify the port that's changed. Do
so rather than assuming that it's always bus 1.
Signed-off-by: NMatthew Garrett <mjg@redhat.com>

The new style brightness control provides an operating range of 9 values
(seems consistent over a large number of models sharing the same
brightness control methods).
Read and use the minimum and maximum values to limit the backlight
interface between those boundaries.
Signed-off-by: NMattia Dongili <malattia@linux.it>
Signed-off-by: NMatthew Garrett <mjg@redhat.com>

Check if we were successful in setting the requested brightness and
report failure in that case.
Signed-off-by: NMattia Dongili <malattia@linux.it>
Signed-off-by: NMatthew Garrett <mjg@redhat.com>

The newer Lenovo ThinkPads have HKEY HID of LEN0068 instead
of IBM0068. Added new HID so that thinkpad_acpi module will
auto load on these newer Lenovo ThinkPads.
Acked-by: NHenrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: stable@kernel.org
Signed-off-by: NManoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: NAndy Lutomirski <luto@mit.edu>
Signed-off-by: NMatthew Garrett <mjg@redhat.com>

Fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=34082

Reported by: Sampo Laaksonen <zhamahn@gmail.com>
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

 if object pin or object lookup in radeon_cursor_set fail, the function
 could leave inconsistent mouse width and hight values in radeon_crtc
 fixed by moving cursor width and height assignments after all
 checks have passed
Signed-off-by: NIlija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

The looping helper didn't do anything due to a superficial
semicolon. Furthermore one of the two dump functions suffered
from copy&paste fail.

While staring at the code I've also noticed that the replace
helper (currently unused) is a bit broken.
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NDave Airlie <airlied@redhat.com>

When we switch the display mux, also switch
the i2c mux.  Also use the start and finish
methods to let the sbios know that the switch
is happening.

Should fix:
https://bugs.freedesktop.org/show_bug.cgi?id=35398Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

NV40 and older cards (pre NV50) reserve a vram bo for the vga memory at
card init. This bo is then freed at card shutdown.  The problem is that
the ttm bo vram manager was already freed. So a crash occurs when the
vga bo is freed. The fix is to free the vga bo prior to freeing the ttm
bo vram manager. There might be other solutions but this seemed the
simplest to me.
Signed-off-by: NJimmy Rentz <jb17bsome@gmail.com>
Signed-off-by: NBen Skeggs <bskeggs@redhat.com>

Given that the hardware may be left in a random condition by the BIOS,
it is conceivable that we then attempt to clear the DP_PIPEB_SELECT bit
without us ever enabling/attaching the DP encoder to a pipe. Thus
causing a NULL deference when we attempt to wait for a vblank on that
crtc.
Reported-and-tested-by: NBryan Christ <bryan.christ@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=36314Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=36456Reported-and-tested-by: NBo Wang <bo.b.wang@intel.com>
Cc: stable@kernel.org
Signed-off-by: NKeith Packard <keithp@keithp.com>

Reported-by: NAlan Cox <alan@linux.intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org
Signed-off-by: NKeith Packard <keithp@keithp.com>

Out of the entire GART/VM subsystem, the hw designers changed
the location of 3 regs.

v2: airlied: add parameter for userspace to work from.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

This is the least-bad behaviour.  It means that we signal the
vblank event before it actually happens, but since we're disabling
vblanks there's no guarantee that it will *ever* happen otherwise.

This prevents GL applications which use WaitMSC from hanging
indefinitely.
Signed-off-by: NChristopher James Halse Rogers <christopher.halse.rogers@canonical.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Since fafcf94e introduced an edid size, it seems to have broken this path.

This manifest as oops on T500 Lenovo laptops with dual graphics primarily.

Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=33812

cc: stable@kernel.org
Reviewed-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

The recent commit closing the race window in device teardown:

commit 86cbfb56
Author: James Bottomley <James.Bottomley@suse.de>
Date:   Fri Apr 22 10:39:59 2011 -0500

    [SCSI] put stricter guards on queue dead checks

is causing a potential NULL deref in scsi_run_queue() because the
q->queuedata may already be NULL by the time this function is called.
Since we shouldn't be running a queue that is being torn down, simply
add a NULL check in scsi_run_queue() to forestall this.
Tested-by: NJim Schutt <jaschut@sandia.gov>
Cc: stable@kernel.org
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>

Hi us,
   When i was compiling kernel, a warning happened to me.
The warning said like following.

drivers/staging/wlan-ng/cfg80211.c:709: warning: initialization from
incompatible pointer type.

See http://s1202.photobucket.com/albums/bb364/harrywei/?action=view&current=patched2.png
for more details.

So i patch like following.
Signed-off-by: NHarry Wei <harryxiyou@gmail.com>
Acked-by: NRandy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>