In JPEG mode, the size of image is variable due to different JPEG compression
rate. We only can get the pix->sizeimage from the user.

If we clear pix->sizeimage in soc_camera_try_fmt() then we will get it from:
	ret = soc_mbus_image_size(xlate->host_fmt, pix->bytesperline,
				pix->height);
	if (ret < 0)
		return ret;

	pix->sizeimage = max_t(u32, pix->sizeimage, ret);

In general, this sizeimage will be larger than the actul JPEG image size.

But vb2 will check the buffer and size of image in __qbuf_userptr():
	/* Check if the provided plane buffer is large enough */
	if (planes[plane].length < q->plane_sizes[plane])

So we shouldn't clear the pix->sizeimage and also shouldn't re-calculate
the pix->sizeimage in soc_mbus_image_size() in JPEG mode

We also shouldn't re-calculate pix->bytesperline:
	ret = soc_mbus_bytes_per_line(pix->width, xlate->host_fmt);
	if (ret < 0)
		return ret;

	pix->bytesperline = max_t(u32, pix->bytesperline, ret);

pix->bytesperline also should be set by the user or by the driver's
try_fmt() implementation.

Change-Id: I700690a2287346127a624b5260922eaa5427a596
Signed-off-by: NAlbert Wang <twang13@marvell.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch contains code change only to use the present macro-
MX3_CAMERA_DATAWIDTH_MASK to check valid camera platform data
bus width instead of enumerating every possible data bus width.
Signed-off-by: NLiu Ying <Ying.Liu@freescale.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Driver for mt9v022 camera sensor is fully compatible for mt9v024 camera sensor
with the exception of several registers which have been changed addresses.
mt9v024 also has improved and additional features, but they are currently not in use.
Signed-off-by: NAlex Gershgorin <alexg@meprolight.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

On i.MX27 two clocks are required: emma-ipg and emma-ahb. The ahb clock
has to be requested using both a device and a connection ID.
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
[g.liakhovetski@gmx.de: rebase to the current media tree]
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch checks the state of the buffer when calling .buf_init() method.
This is needed for the USERPTR buffer type, because in that case
.buf_init() is called every time a buffer is queued, and not only once
during the preparation stage, like in the MMAP case. Without this check
buffers get initialised repeatedly, which also leads to the allocation
of new DMA descriptors, of which there is only a final relatively small
number available. Both MMAP and USERPTR methods were successfully tested.
Signed-off-by: NAlex Gershgorin <alexg@meprolight.com>
[g.liakhovetski@gmx.de: remove mx3_camera_buffer::state completely]
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

All these seems to survive .reset_resume.
Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

All these seems to survive .reset_resume.
Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Add .reset_resume() support.
Also some other small changes for suspend / resume.
Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Move initial suspend / resume support from dvb_usb_v2 to dvb_frontend
as it is dvb general feature that could be used all dvb devices.
Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Put device full sleep on suspend, wake-up it on resume and acquire
retune in order to return same television channel.
Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

We need to retune when resume from suspend.
Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The commit 98d7bbb9 changed *of_device_id.data to const
which introduced warnings in various places that have mostly
been fixed. This patch fixes one such warning by introducing
two const qualifiers.

GCC warning:
drivers/media/platform/coda.c:1785:16: warning:
        assignment discards ‘const’ qualifier
        from pointer target type [enabled by default]
Signed-off-by: NEmil Goode <emilgoode@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

debug_dump macro was defined twice when CONFIG_DVB_USB_DEBUG was
not set. Move debug_dump macro to correct place.
Reported-by: NRandy Dunlap <rdunlap@xenotime.net>
Signed-off-by: NAntti Palosaari <crope@iki.fi>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

[media] drivers/media/usb/{s2255drv.c, tm6000/tm6000-alsa.c, tm6000/tm6000-input.c}: Remove potential NULL dereferences

If the NULL test is necessary, the initialization involving a dereference of
the tested value should be moved after the NULL test.

The sematic patch that fixes this problem is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
type T;
expression E;
identifier i,fld;
statement S;
@@

- T i = E->fld;
+ T i;
  ... when != E
      when != i
  if (E == NULL) S
+ i = E->fld;
// </smpl>
Signed-off-by: NJulia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

devm_* functions are device managed functions and make error handling
and cleanup cleaner and simpler.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Replace printk with pr_* functions to silence checkpatch warnings.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

devm_regulator_get() is a device managed function and makes the exit code
a bit simpler and cleaner.
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NSylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Using devm_kzalloc simplifies the code and ensures that the use of
devm_request_irq is safe.  When kzalloc and kfree were used, the interrupt
could be triggered after the handler's data argument had been freed.

This also introduces some missing initializations of the return variable
ret, and uses devm_request_and_ioremap instead of the combination of
devm_request_mem_region and devm_ioremap.

The problem of a free after a devm_request_irq was found using the
following semantic match (http://coccinelle.lip6.fr/)

// <smpl>
@r exists@
expression e1,e2,x,a,b,c,d;
identifier free;
position p1,p2;
@@

  devm_request_irq@p1(e1,e2,...,x)
  ... when any
      when != e2 = a
      when != x = b
  if (...) {
    ... when != e2 = c
        when != x = d
    free@p2(...,x,...);
    ...
    return ...;
  }
// </smpl>
Signed-off-by: NJulia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

More validation fixes as reported by xmllint.

There are still three xmllint errors after this remaining regarding SVG file support.
Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NHans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

"card" is a valid pointer here because we checked snd_card_create() for
error returns.  Checking after a dereference makes the static checkers
complain.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Acked-by: NEzequiel Garcia <elezegarcia@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

There are some unlocks missing on error.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Acked-by: NEzequiel Garcia <elezegarcia@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

We recently added locking to this function, but there was an error path
which accidentally returned holding a lock.  Also we returned zero on
failure on some paths instead of the error code.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

We need to move the unlock before the kfree(dev);
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This code looks suspicious, but it turns out that "nv" is an array of u8
so sizeof() is the same as ARRAY_SIZE().  Using ARRAY_SIZE() is more
readable though.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Cc: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Add .get_selection() and .set_selection() soc-camera host driver
operations. Additionally check, that the user is not trying to change the
output sizes during a running capture.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Add explicit conversions from UYVY and YUYV to YUYV so that
csicr1 configuration can be set properly for each format.
Signed-off-by: NJavier Martin <javier.martin@vista-silicon.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Prepare the clock before enabling it.

Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Cc: <linux-media@vger.kernel.org>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Prepare the clock before enabling it.

Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Cc: <linux-media@vger.kernel.org>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The sensor starts streaming video as soon as it gets powered or is
reset. Disable the output in the reset function.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Instead of hardcoding register arrays, compute the values at runtime.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
[g.liakhovetski@gmx.de: keep (Q)VGA_* macros for now]
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

And use them instead of calling SMBus access functions directly.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Conversion from i2c_client to ov772x_priv is only needed in a single
location, while conversion from v4l2_subdev to ov772x_priv is needed in
several locations. Perform the former manually, and use to_ov772x for
the later.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

If the requested format isn't supported, return a fixed default format
instead of the current format.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Select a default format instead.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The format and window size are only initialized during the first g_fmt
call. This leaves the device in an inconsistent state after
initialization, which will cause problems when implementing pad
operations. Move the format and window size initialization to probe
time.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The control handler isn't freed if its initialization fails. Fix it.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Several client drivers access the hardware at probe time, for instance
to read the probe chip ID. Such chips need to be powered up when being
probed.

soc-camera handles this by powering chips up in the soc-camera probe
implementation. However, this will break with non soc-camera hosts that
don't perform the same operations.

Fix the problem by pushing the power up/down from the soc-camera core
down to individual drivers on a needs basis.
Signed-off-by: NLaurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>