With the conversion to videobuf2, some unnecessary calls of
em28xx_set_alternate() have been removed. It is now called at analog streaming
start only.
This has unveiled a bug that causes USB bulk transfers to fail with all urbs
having status -EVOERFLOW.
The reason is, that for bulk transfers usb_set_interface() needs to be called
even if the previous alt setting was the same (side note: bulk transfers seem
to work only with alt=0).
While it seems to be NOT necessary for isoc transfers, it's reasonable to just
call usb_set_interface() unconditionally in em28xx_set_alternate().
Also add a comment that explains the issue to prevent regressions in the future.

Cc: stable@vger.kernel.org # for 3.8
Signed-off-by: NFrank Schäfer <fschaefer.oss@googlemail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The address of a variable is impossible to be null, so we remove the check.
Signed-off-by: NCong Ding <dinggnu@gmail.com>
Signed-off-by: NMichael Krufky <mkrufky@linuxtv.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Return early, if we already tuned to a frequency.
Signed-off-by: NMichael Buesch <m@bues.ch>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Add some sanity checks to the calculations and make the REG_16 register write consistent
with the other ones.
Signed-off-by: NMichael Buesch <m@bues.ch>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Fix the xin value clamping and use clamp_t().
Signed-off-by: NMichael Buesch <m@bues.ch>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch was applied by mistake. Michael thinks that it
it needs more work than simply moving the soft reset.

So, it should be held back for further review.

This reverts commit 0a323770.

Requested by: Michael Krufky <mkrufky@linuxtv.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Assign the maximum instead of masking with the maximum on value overflow.
Signed-off-by: NMichael Buesch <m@bues.ch>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

tm6000_poll could use an uninitialized buf pointer. Move the buf-handling
code inside the 'if' that initializes the buf pointer.
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch is driven by a static checker warning.
The ttusb_process_muxpack() function is only called from
ttusb_process_frame().  Before calling, it verifies that len >= 2.  The
problem is that len == 2 is not valid and would lead to an array
underflow.
Odd number values for len are also invalid and would lead to reading
past the end of the array.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Without this patch we get the following build error:
drivers/media/platform/s5p-tv/mixer_video.c:
In function ‘find_and_register_subdev’:
drivers/media/platform/s5p-tv/mixer_video.c:42:34: error:
‘platform_bus_type’ undeclared (first use in this function)
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Without this patch we get the following compilation errors:
drivers/media/platform/s5p-tv/mixer.h:345:13: error: Expected ; at end of declaration
drivers/media/platform/s5p-tv/mixer.h:345:13: error: got mxr_irq_handler
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The bytesperline calculation was incorrect: it used the old width instead of
the provided width. Fixed.
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Acked-by: NFrank Schäfer <fschaefer.oss@googlemail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

While running checkpatch.pl after the last patch, I noticed
lots of those:
	WARNING: please, no space before tabs
	#151: FILE: drivers/media/common/tveeprom.c:99:
	+^I{ TUNER_ABSENT,        ^I^I"None" },$

(together with other checkpatch.pl errors/warnings)

While I won't be fixing everything, as I have already an
script to fix the above, let's do it, in order to clean it
a little bit.

While here, also drop cmacs-specific format text at the end.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The tveeprom module is a helper module for Hauppauge-based eeproms.
It's used by many drivers and the i2c part is actually optional, so this
driver is better placed in the media/common directory.
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The btcx-risc module is a helper module for bttv/conexant based TV cards.
It isn't an i2c module at all, instead it should be in common since it is
used by 4 pci drivers.
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The cx2341x module is a helper module for conexant-based MPEG encoders.
It isn't an i2c module at all, instead it should be in common since it is
used by 7 pci and usb drivers to handle the MPEG setup.
It also shouldn't be visible in the config menu as it is always
selected automatically by those drivers that need it.
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The ttusb2 module is already updated to recognize the TechnoTrend CT-3650
CI DVB C/T USB2.0 receiver in addition to the Pinnacle 400e. But if
MEDIA_SUBDRV_AUTOSELECT is used, the required tuner and demodulator
modules are not automatically selected. Here's a patch to fix that and add a
note of the CT-3650 to the online help of the ttusb2 module.
This patch applies cleanly to 3.7.6 and other 3.7.x kernels.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch move the soft reset to the attach function because with dual
tuners, when one tuner do reset, the other one is perturbed, and the
stream has errors.
Signed-off-by: NJose Alberto Reguero <jareguero@telefonica.net>
Reviewed-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

A recent commit broke compilation of 3 camera drivers: for PXA2x0, OMAP1
and MX1 by using a wrong pointer. Fix them.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch adds new capture board Hawell HW-9004V1. This card has 4 SAA71300 chips. In order to work it is needed to initialize its registers (gpio mask and value). The value of these registers were dumped under Windows using flytest.
Signed-off-by: NVadim Frolov <fralik@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This allocation had no error checking.  It didn't need to be under
the mutex so I moved it out form there. That makes the error handling
easier and is a potential speed up.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

via-camera and mcam-core were the only bridge drivers that used ov7670.
Since now they have been moved to use the ctrl framework, the old
legacy callbacks in the ov7670 can be removed.
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

And added a missing kfree to clean up the via_camera struct.
Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NHans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Some bridge drivers capture pixels during blanking periods if
pixclk is enabled. In order to avoid capturing bogus data we need to
disable pixclk in the sensor during those blanking periods.
Acked-by: NJonathan Corbet <corbet@lwn.net>
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

For a frame rate of 30 fps a pixclk of 24MHz is needed. For those
cases where the ov7670 has a clean 24MHz input (xvclk) the PLL
can be bypassed.
This will result in a value of clkrc of 1, which means that in practice
pixclk = xvclk (input clock)
Acked-by: NJonathan Corbet <corbet@lwn.net>
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

According to the datasheet ov7675 uses a formula to achieve
the desired framerate that is different from the operations
done in the current code.
In fact, this formula should apply to ov7670 too. This would
mean that current code is wrong but, in order to preserve
compatibility, the new formula will be used for ov7675 only.
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

'min_height' and 'min_width' are variables that allow to specify the minimum
resolution that the sensor will achieve. This patch make v4l2 fmt callbacks
consider this parameters in order to return valid data to user space.
Acked-by: NJonathan Corbet <corbet@lwn.net>
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

ov7675 and ov7670 share the same registers but there is no way
to distinguish them at runtime. However, they require different
tweaks to achieve the desired resolution. For this reason this
patch adds a new ov7675 entry to the ov7670_id table.
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

If more than 127 different lengths are transmitted then the driver causes
an overrun on sample_lens. Try to send as much as possible and return the
amount sent.
Signed-off-by: NSean Young <sean@mess.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This transceiver expects the set IR TX ports and IR data as seperate
packets, like the Windows driver does. Remove unnecessary kzalloc.
Signed-off-by: NSean Young <sean@mess.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The GET_REVISION command puts the device in an unresponsive state,
although it continues to report any IR activity. Note that GET_REVISION
command is not documented, nor is any possible response to it parsed.
Signed-off-by: NSean Young <sean@mess.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

spotted by sparse.
Signed-off-by: NSean Young <sean@mess.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

led_classdev_resume already sets the led.
Signed-off-by: NSean Young <sean@mess.org>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NAntonio Ospite <ao2@amarulasolutions.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Found with coccicheck.
The semantic patch that makes this change is available
in scripts/coccinelle/api/ptr_ret.cocci.
Signed-off-by: NPeter Huewe <peterhuewe@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Remove no longer supported __devinit, __devexit attributes.
Signed-off-by: NSylwester Nawrocki <sylvester.nawrocki@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

- move the bit shifting of width+height values inside the function
- fix the debug message format and output values
- add comment about the size limit (e.g. EM277x supports >2MPix)
- make void, because error checking is incomplete and we never check the
  returned value (we would continue anyway)
Signed-off-by: NFrank Schäfer <fschaefer.oss@googlemail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>