Hardik Shah [Fri, 24 Apr 2009 09:03:48 +0000 (14:33 +0530)]
DSS2: Rotation attrs for YUV need not to be reversed for 90 and 270
DSS2 supports horizontal mirroing. so only 180 and 0 degree attributes needs
to be reversed for mirroring.
Slight corrupt image is seen with 90 ad 270 degree rotation with mirroring
when the image height and widht is not 32-pixels aligned. To be specific
resolution is 168X192.
Hardik Shah [Fri, 24 Apr 2009 09:59:06 +0000 (15:29 +0530)]
DSS2: Added global alpha support.
global_alpha sysfs entry for video1 plane will always show 255 as
it does not support global_alpha.
Initliazied the global alpha field for all the overlays
to 255 (fully opaque).
Documentation also updated.
Imre Deak [Mon, 20 Apr 2009 12:55:33 +0000 (14:55 +0200)]
DSS2: add sysfs entry for seting the rotate type
This can help in utilizing VRAM memory better. Since with VRFB rotation
we waste a lot of physical memory due to the VRFB HW design, provide the
possibility to turn it off and free the extra memory for the use by other
planes for example.
Ville Syrjälä [Mon, 20 Apr 2009 14:26:20 +0000 (16:26 +0200)]
DSS2: Swap field 0 and field 1 registers
The values for the registers which have alternate values for each field
were reveresed to what the hardware expects. For the hardware field 0
is the even field or the bottom field, field 1 is the odd field or the
top field. So simply swap the register values.
Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com>
Ville Syrjälä [Mon, 20 Apr 2009 14:26:19 +0000 (16:26 +0200)]
DSS2: Don't enable fieldmode automatically
The only case where enabling fieldmode automatically seems reasonable
is when source and destination heights are equal. Some kind of user
controllable knob should be added so the user could enable field mode
when the source is interlaced.
Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com>
Hardik Shah [Thu, 16 Apr 2009 13:17:49 +0000 (18:47 +0530)]
DSS2: Added the function pointer for getting default color.
V4L2 Framework has a CID for getting/setting default color.
So added the function pointer for doing same.
SYSFS based getting the default color will remain same
Ville Syrjälä [Thu, 9 Apr 2009 13:04:43 +0000 (15:04 +0200)]
DSS2: Prefer 3-tap filter
The 5-tap filter seems rather unstable. With some scaling settings it
works and with some it doesn't even though the functional clock remains
within the TRM limits. So prefer the 3-tap filter unless the functional
clock required for it is too high.
Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com>
Imre Deak [Tue, 14 Apr 2009 12:50:11 +0000 (14:50 +0200)]
DSS2: do bootmem reserve for exclusive access
BOOTMEM_DEFAULT would allow multiple reservations for the same location,
we need to reserve the region for our exclusive use. Also check if the
reserve succeeded.
Hardik Shah [Thu, 9 Apr 2009 06:43:07 +0000 (12:13 +0530)]
DSS2: OMAPFB: Added support for the YUV VRFB rotation and mirroring.
DSS2 now requires roatation_type to be specified by driver.
Added support for that.
DSS2 OMAPFB: Modified to pass the dss mode to omap_vrfb_setup function.
VRFB size register requires the width to be halved when the
mode is YUV or UYVY. So VRFB is modifed to pass the mode to omap_vrfb_setup
function.
Few changes done by Tim Yamin Signed-off-by: Tim Yamin <plasm@roo.me.uk> Signed-off-by: Hardik Shah <hardik.shah@ti.com>
Tomi Valkeinen [Wed, 15 Apr 2009 11:06:54 +0000 (14:06 +0300)]
DSS2: DSI: sidlemode to noidle while sending frame
DISPC interrupts are not wake-up capable. Smart-idle in DISPC_SIDLEMODE
causes DSS interface to go to idle at the end of the frame, and the
FRAMEDONE interrupt is then delayed until something wakes up the DSS
interface.
So we set SIDLEMODE to no-idle when we start sending the frame, and
set it back to smart-idle after receiving FRAMEDONE.
Imre Deak [Wed, 8 Apr 2009 10:51:46 +0000 (12:51 +0200)]
DSS2: pass the default FB color format through board info
Add a field to the FB memory region platform data, so that board
init code can pass a default color format to the driver. Set this
format as an initial setting for the given FB.
This is needed for an upcoming patch that adds detection of the
color format set by the bootloader.
Ville Syrjälä [Mon, 6 Apr 2009 15:32:04 +0000 (17:32 +0200)]
DSS2: Check fclk limits when configuring video planes
Check that the currect functional clock is fast enough to support
the requested scaling ratios. Also check if 5-tap filtering can be
used even though the downscaling ratio is less than 1:2 since the
functional clock rate required for 5-tap filtering can be less than
the requirement for 3-tap filtering, and 5-tap filtering should look
better.
Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com>
Paul Walmsley [Wed, 28 Jan 2009 02:13:12 +0000 (19:13 -0700)]
[ARM] OMAP: Make dpll4_m4_ck programmable with clk_set_rate()
Filling the set_rate and round_rate fields of dpll4_m4_ck makes
this clock programmable through clk_set_rate(). This is needed
to give omapfb control over the dss1_alwon_fck rate.
This patch includes a fix from Tomi Valkeinen <tomi.valkeinen@nokia.com>.
Signed-off-by: Måns Rullgård <mans@mansr.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@nokia.com> Signed-off-by: Paul Walmsley <paul@pwsan.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
David Brownell [Tue, 25 Nov 2008 07:11:03 +0000 (23:11 -0800)]
USB: gadget: fix rndis working at high speed
Fix a bug specific to highspeed mode in the recently updated RNDIS
support: it wasn't setting up the high speed notification endpoint,
which prevented high speed RNDIS links from working.
Richard Röjfors [Sun, 16 Nov 2008 03:53:24 +0000 (19:53 -0800)]
USB: gadget rndis: send notifications
It turns out that atomic_inc_return() returns the *new* value
not the original one, so the logic in rndis_response_available()
kept the first RNDIS response notification from getting out.
This prevented interoperation with MS-Windows (but not Linux).
Fix this to make RNDIS behave again.
Signed-off-by: Richard Röjfors <richard.rojfors@endian.se> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: stable <stable@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
David Brownell [Sun, 16 Nov 2008 03:53:21 +0000 (19:53 -0800)]
USB: gadget rndis: stop windows self-immolation
Somewhere in the conversion of the RNDIS gadget code to the new
framework, the descriptor of its data interface seems to have
been copied from the CDC Ethernet driver. Unfortunately that
means it got a nonzero altsetting ... which is incorrect. Issue
uncovered by Richard Röjfors <richard.rojfors@endian.se>.
This patch fixes that problem, and resolves at least some cases
of Windows XP bluescreening itself.
Tested-by: Richard Röjfors <richard.rojfors@endian.se>. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: stable <stable@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This allows the device to act as mouse or scrolling device.
Mode can be switched by writing to special file in
/proc/pandora ("absolute", "mouse" and "scroll").
San Mehat [Mon, 14 Apr 2008 22:22:49 +0000 (15:22 -0700)]
mmc: Add concept of an 'embedded' SDIO device.
This is required to support chips which use SDIO for signaling/
communication but do not implement the various card enumeration registers
as required for full SD / SDIO cards.
mmc: sdio: Fix bug where we're freeing the CIS tables we never allocated when using EMBEDDED_SDIO
mmc: Add max_blksize to embedded SDIO data
Adrian Hunter [Thu, 16 Oct 2008 13:37:23 +0000 (16:37 +0300)]
ARM: OMAP: calculate hsmmc data timeout accurately
The data timeout can be calculated from the clock
rate and timing values given by the card. At
present the timeout is set to the maximum value,
about 2.7 secs. When dealing with cards that do have
errors, this causes large unnecessary delays because
the correct timeout may be around 10 millisecs.
Signed-off-by: Adrian Hunter <ext-adrian.hunter@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
This patches moves the twl4030 power sequencer code to drivers/mfd and
updates the code for the new twl4030 framework. The code handles downloading
the scripts provided by the board configuration and configuring the chip
to call the relevant script for each event (processor group 1 and 2 sleep,
processor group 3 sleep, wakeup or warm reset).
Signed-off-by: Peter 'p2' De Schrijver <peter.de-schrijver@nokia.com> Signed-off-by: Tony Lindgren <tony@atomide.com>
projects / pandora-kernel.git / log