[media] sur40: set q->dev instead of allocating a context

Stop using alloc_ctx and just fill in the device pointer.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: Florian Echtler <floe@butterbrot.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l2-pci-skeleton: set q->dev instead of allocating a context

Stop using alloc_ctx as that is now no longer needed.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] vb2: add a dev field to use for the default allocation context

The allocation context is nothing more than a per-plane device pointer
to use when allocating buffers. So just provide a dev pointer in vb2_queue
for that purpose and drivers can skip allocating/releasing/filling in
the allocation context unless they require different per-plane device
pointers as used by some Samsung SoCs.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Sakari Ailus <sakari.ailus@iki.fi>
Cc: Florian Echtler <floe@butterbrot.org>
Cc: Federico Vaga <federico.vaga@gmail.com>
Cc: "Lad, Prabhakar" <prabhakar.csengg@gmail.com>
Cc: Scott Jiang <scott.jiang.linux@gmail.com>
Acked-by: Philipp Zabel <p.zabel@pengutronix.de>
Cc: Fabien Dessenne <fabien.dessenne@st.com>
Acked-by: Benoit Parrot <bparrot@ti.com>
Cc: Mikhail Ulyanov <mikhail.ulyanov@cogentembedded.com>
Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Ludovic Desroches <ludovic.desroches@atmel.com>
Cc: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Cc: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] vb2: move dma_attrs to vb2_queue

Make the dma attributes struct part of vb2_queue. This greatly simplifies
the remainder of the patch series since the dma_contig alloc context is
now (as before) just a struct device pointer.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] exynos4-is: Fix buffer release issue on fimc m2m video nodes

This fixes dropping ownership of buffers in the driver's stop_streaming
callback, so buffers on the memory-to-memory video nodes are properly
released, also in case when the driver has a buffer only on one of the
queues (OUTPUT, CAPTURE) before the video node close.
The issue was being reported by videobuf2 with a following warning while
checking q->owned_by_drv_count:

[ 2498.310766] WARNING: CPU: 0 PID: 9358 at drivers/media/v4l2-core/videobuf2-core.c:1818 __vb2_queue_cancel+0xe8/0x14c
[ 2498.320258] Modules linked in:
[ 2498.323212] CPU: 0 PID: 9358 Comm: v4l2_decode Not tainted 4.7.0-rc4-next-20160627 #1210
[ 2498.331284] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[ 2498.331327] [<c010d738>] (unwind_backtrace) from [<c010a4b0>] (show_stack+0x10/0x14)
[ 2498.331344] [<c010a4b0>] (show_stack) from [<c031a4ac>] (dump_stack+0x74/0x94)
[ 2498.331358] [<c031a4ac>] (dump_stack) from [<c011a52c>] (__warn+0xd4/0x100)
[ 2498.331369] [<c011a52c>] (__warn) from [<c011a578>] (warn_slowpath_null+0x20/0x28)
[ 2498.331381] [<c011a578>] (warn_slowpath_null) from [<c04ed420>] (__vb2_queue_cancel+0xe8/0x14c)
[ 2498.331395] [<c04ed420>] (__vb2_queue_cancel) from [<c04ee10c>] (vb2_core_queue_release+0x18/0x38)
[ 2498.331406] [<c04ee10c>] (vb2_core_queue_release) from [<c04eab50>] (v4l2_m2m_ctx_release+0x1c/0x28)
[ 2498.331420] [<c04eab50>] (v4l2_m2m_ctx_release) from [<c04fe184>] (fimc_m2m_release+0x24/0x78)
[ 2498.331437] [<c04fe184>] (fimc_m2m_release) from [<c04d76c8>] (v4l2_release+0x34/0x74)
[ 2498.331455] [<c04d76c8>] (v4l2_release) from [<c01dc8d4>] (__fput+0x80/0x1bc)
[ 2498.331469] [<c01dc8d4>] (__fput) from [<c0132edc>] (task_work_run+0xc0/0xe4)
[ 2498.331482] [<c0132edc>] (task_work_run) from [<c011d460>] (do_exit+0x304/0xa24)
[ 2498.331493] [<c011d460>] (do_exit) from [<c011dccc>] (do_group_exit+0x3c/0xbc)
[ 2498.331505] [<c011dccc>] (do_group_exit) from [<c0126cac>] (get_signal+0x200/0x65c)
[ 2498.331517] [<c0126cac>] (get_signal) from [<c010e928>] (do_signal+0x84/0x3c4)
[ 2498.331532] [<c010e928>] (do_signal) from [<c010a0ec>] (do_work_pending+0xa4/0xb4)
[ 2498.331545] [<c010a0ec>] (do_work_pending) from [<c0107954>] (slow_work_pending+0xc/0x20)

Reported-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: improve v4l2_capability driver and card fields

According to the V4L2 documentation the driver and card fields should be
used to identify the driver and the device but the s5p-mfc driver fills
those field using the platform device name, which in turn is the name of
the device DT node.

So not only the filled information isn't correct but also the same values
are used in all the fields for both the encoder and decoder video devices.

Before this patch:

Driver Info (not using libv4l2):
        Driver name   : 11000000.codec
        Card type     : 11000000.codec
        Bus info      : platform:11000000.codec
        Driver version: 4.7.0

Driver Info (not using libv4l2):
        Driver name   : 11000000.codec
        Card type     : 11000000.codec
        Bus info      : platform:11000000.codec
        Driver version: 4.7.0

After this patch:

Driver Info (not using libv4l2):
        Driver name   : s5p-mfc
        Card type     : s5p-mfc-dec
        Bus info      : platform:11000000.codec
        Driver version: 4.7.0

Driver Info (not using libv4l2):
        Driver name   : s5p-mfc
        Card type     : s5p-mfc-enc
        Bus info      : platform:11000000.codec
        Driver version: 4.7.0

Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: set capablity bus_info as required by VIDIOC_QUERYCAP

The driver doesn't set the struct v4l2_capability bus_info field so the
v4l2-compliance tool reports the following errors for VIDIOC_QUERYCAP:

Required ioctls:
                VIDIOC_QUERYCAP returned 0 (Success)
                fail: v4l2-compliance.cpp(304): string empty
                fail: v4l2-compliance.cpp(528): check_ustring(vcap.bus_info, sizeof(vcap.bus_info))
        test VIDIOC_QUERYCAP: FAIL

This patch fixes by setting the field in VIDIOC_QUERYCAP ioctl handler:

Required ioctls:
                VIDIOC_QUERYCAP returned 0 (Success)
        test VIDIOC_QUERYCAP: OK

Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: use vb2_is_streaming() to check vb2 queue status

The streaming field in struct vb2_queue is meant to be private and should
not be used by drivers directly, instead the vb2_is_streaming() function
should be used to check the videobuf2 queue streaming status.

Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: don't print errors on VIDIOC_REQBUFS unsupported mem type

The V4L2 documentation says that applications must call the VIDIOC_REQBUFS
ioctl to determine if a memory mapped, user pointer or DMABUF based I/O is
supported by the driver.

For example GStreamer does this by first calling VIDIOC_REQBUFS with count
zero for all the possible streaming I/O methods and then finally doing the
real VIDIOC_REQBUFS with count N using a known to be supported memory type.

But the driver prints an error on VIDIOC_REQBUFS if the memory type is not
supported which leads to the following errors that can confuse the users:

[  178.704390] vidioc_reqbufs:575: Only V4L2_MEMORY_MMAP is supported
[  178.704666] vidioc_reqbufs:575: Only V4L2_MEMORY_MMAP is supported
[  178.714956] vidioc_reqbufs:575: Only V4L2_MEMORY_MMAP is supported
[  178.715229] vidioc_reqbufs:575: Only V4L2_MEMORY_MMAP is supported

Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: fix typo in s5p_mfc_dec function comment

The function comment has an obvious typo error, so fix it.

Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: fix error path in driver probe

This patch fixes the error path in the driver probe, so in case of
any failure, the resources are not leaked.

Reported-by: Liviu Dudau <liviu.dudau@arm.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Acked-by: Liviu Dudau <Liviu.Dudau@arm.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: fix null pointer deference in clk_core_enable()

Fix null pointer deference in clk_core_enable() when driver unbind is run
when there is an application has an active pipeline playing.

s5p_mfc_release() gets called after s5p_mfc_final_pm() disables and does
clk_put() and s5p_mfc_release() attempts to enable clock and runs into
null pointer deference accessing invalid pointer.

[ 4869.434709] Unable to handle kernel NULL pointer dereference at virtual addr0
[ 4869.441312] pgd = e91ac000
[ 4869.443996] [00000010] *pgd=ba4f7835
[ 4869.447552] Internal error: Oops: 17 [#1] PREEMPT SMP ARM
[ 4869.452921] Modules linked in: cpufreq_userspace cpufreq_powersave cpufreq_ca
[ 4869.471728] CPU: 4 PID: 2965 Comm: lt-gst-launch-1 Not tainted 4.7.0-rc2-nex0
[ 4869.481778] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[ 4869.487844] task: e91f1e00 ti: ed650000 task.ti: ed650000
[ 4869.493227] PC is at clk_core_enable+0x4c/0x98
[ 4869.497637] LR is at clk_core_enable+0x40/0x98
[ 4869.502056] pc : [<c0559714>]    lr : [<c0559708>]    psr: 60060093
[ 4869.502056] sp : ed651f18  ip : 00000000  fp : 002641b4
[ 4869.513493] r10: e9088c08  r9 : 00000008  r8 : ed676d68
[ 4869.518692] r7 : ee3ac000  r6 : bf16b3c0  r5 : a0060013  r4 : ee37a8c0
[ 4869.525191] r3 : 00000000  r2 : 00000001  r1 : 00000004  r0 : 00000000
[ 4869.531692] Flags: nZCv  IRQs off  FIQs on  Mode SVC_32  ISA ARM  Segment noe
[ 4869.538883] Control: 10c5387d  Table: 691ac06a  DAC: 00000051
[ 4869.544603] Process lt-gst-launch-1 (pid: 2965, stack limit = 0xed650210)
[ 4869.551361] Stack: (0xed651f18 to 0xed652000)
[ 4869.555694] 1f00:                                                       ee373
[ 4869.563841] 1f20: bf16b3c0 c055a0e0 ee3ac004 ed676c10 bf16b3c0 bf1558e0 e9080
[ 4869.571986] 1f40: 00000000 ee98a510 ee502e40 bf047344 e9088c00 ee986938 00004
[ 4869.580132] 1f60: 00000000 00000000 e91f2204 00000000 c0b4658c e91f1e00 c0100
[ 4869.588277] 1f80: 00000000 c0135c58 ed650000 c0107904 ed651fb0 00000006 c0104
[ 4869.596423] 1fa0: 00229500 b6581000 b6f7b544 c0107794 00000000 00000002 b6f90
[ 4869.604568] 1fc0: 00229500 b6581000 b6f7b544 00000006 0017b600 0002c038 00264
[ 4869.612714] 1fe0: 00000000 bee56ef0 00000000 b6d49612 00060030 00000006 00000
[ 4869.620865] [<c0559714>] (clk_core_enable) from [<c055a0e0>] (clk_enable+0x2)
[ 4869.628509] [<c055a0e0>] (clk_enable) from [<bf1558e0>] (s5p_mfc_release+0x3)
[ 4869.637111] [<bf1558e0>] (s5p_mfc_release [s5p_mfc]) from [<bf047344>] (v4l2)
[ 4869.646706] [<bf047344>] (v4l2_release [videodev]) from [<c01e4274>] (__fput)
[ 4869.654745] [<c01e4274>] (__fput) from [<c0135c58>] (task_work_run+0x94/0xc8)
[ 4869.661852] [<c0135c58>] (task_work_run) from [<c010a9d4>] (do_work_pending+)
[ 4869.669735] [<c010a9d4>] (do_work_pending) from [<c0107794>] (slow_work_pend)
[ 4869.677878] Code: ebffffef e3500000 18bd8070 e5943004 (e5933010)

Signed-off-by: Shuah Khan <shuahkh@osg.samsung.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: fix memory leak in s5p_mfc_remove()

s5p_mfc_remove() fails to release encoder and decoder video devices.

Signed-off-by: Shuah Khan <shuahkh@osg.samsung.com>
Reviewed-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] s5p-mfc: fix video device release double release in probe error path

Fix Decoder and encoder video device double release in probe error path.
video_device_release(dev->vfd_dec) get called twice if decoder register
fails. Also, video_device_release(dev->vfd_enc) get called twice if encoder
register fails.

Signed-off-by: Shuah Khan <shuahkh@osg.samsung.com>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] arm64: dts: mediatek: Add Video Encoder for MT8173

Add video encoder node for MT8173

Signed-off-by: Tiffany Lin <tiffany.lin@mediatek.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] vcodec: mediatek: Add Mediatek H264 Video Encoder Driver

Add h264 encoder driver for MT8173

Signed-off-by: PoChun Lin <pochun.lin@mediatek.com>
Signed-off-by: Tiffany Lin <tiffany.lin@mediatek.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] vcodec: mediatek: Add Mediatek VP8 Video Encoder Driver

Add vp8 encoder driver for MT8173

Signed-off-by: PoChun Lin <pochun.lin@mediatek.com>
Signed-off-by: Tiffany Lin <tiffany.lin@mediatek.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] vcodec: mediatek: Add Mediatek V4L2 Video Encoder Driver

Add v4l2 layer encoder driver for MT8173

Signed-off-by: Tiffany Lin <tiffany.lin@mediatek.com>
[hans.verkuil@cisco.com: drop unnecessary ARM || ARM64 dependency]
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] dt-bindings: Add a binding for Mediatek Video Encoder

Add a DT binding documentation of Video Encoder for the
MT8173 SoC from Mediatek.

Signed-off-by: Tiffany Lin <tiffany.lin@mediatek.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] arm64: dts: mediatek: Add node for Mediatek Video Processor Unit

Add VPU drivers for MT8173

Signed-off-by: Andrew-CT Chen <andrew-ct.chen@mediatek.com>
Signed-off-by: Tiffany Lin <tiffany.lin@mediatek.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] VPU: mediatek: support Mediatek VPU

The VPU driver for hw video codec embedded in Mediatek's MT8173 SOCs.
It is able to handle video decoding/encoding of in a range of formats.
The driver provides with VPU firmware download, memory management and
the communication interface between CPU and VPU.
For VPU initialization, it will create virtual memory for CPU access and
IOMMU address for vcodec hw device access. When a decode/encode instance
opens a device node, vpu driver will download vpu firmware to the device.
A decode/encode instant will decode/encode a frame using VPU
interface to interrupt vpu to handle decoding/encoding jobs.

Signed-off-by: Andrew-CT Chen <andrew-ct.chen@mediatek.com>
Signed-off-by: Tiffany Lin <tiffany.lin@mediatek.com>
[hans.verkuil@cisco.com: drop unnecessary ARM || ARM64 dependency]
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] videodev2.h: Fix V4L2_PIX_FMT_YUV411P description

YUV 4:1:1 uses 12 bits per pixel on average, not 16.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] videodev2.h: Group YUV 3 planes formats together

The formats are interleaved with the YUV packed and miscellaneous
formats, making the result confusing especially with the YUV444 format
being packed and not planar like YUV410 or YUV420. Move them to their
own group as the 2 planes or 3 non-contiguous planes formats to clarify
the header.

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media: rcar-vin: add DV timings support

Adds ioctls DV_TIMINGS_CAP, ENUM_DV_TIMINGS, G_DV_TIMINGS, S_DV_TIMINGS,
and QUERY_DV_TIMINGS.

Signed-off-by: Ulrich Hecht <ulrich.hecht+renesas@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media: rcar_vin: Use correct pad number in try_fmt

Fix rcar_vin_try_fmt's use of an inappropriate pad number when calling
the subdev set_fmt function - for the ADV7612, IDs should be non-zero.

Signed-off-by: William Towle <william.towle@codethink.co.uk>
Reviewed-by: Rob Taylor <rob.taylor@codethink.co.uk>
[ulrich.hecht+renesas@gmail.com: adapted to rcar-vin rewrite]
Signed-off-by: Ulrich Hecht <ulrich.hecht+renesas@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media: rcar-vin: pad-aware driver initialisation

Add detection of source pad number for drivers aware of the media controller
API, so that rcar-vin can create device nodes to support modern drivers such
as adv7604.c (for HDMI on Lager) and the converted adv7180.c (for composite)
underneath.

Building rcar_vin gains a dependency on CONFIG_MEDIA_CONTROLLER, in
line with requirements for building the drivers associated with it.

Signed-off-by: William Towle <william.towle@codethink.co.uk>
Signed-off-by: Rob Taylor <rob.taylor@codethink.co.uk>
[ulrich.hecht+renesas@gmail.com: adapted to rcar-vin rewrite]
Signed-off-by: Ulrich Hecht <ulrich.hecht+renesas@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l2-ctrl.h: fix comments

The comments for the unlocked v4l2_ctrl_s_ctrl* functions were wrong (copy
and pasted from the locked variants). Fix this, since it is confusing.

Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] radio-maxiradio: fix memory leak when device is removed

Memory allocated for maxiradio device is not deallocated when
the device is removed.

Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] stk1160: Check *nplanes in queue_setup

If *nplanes is not zero, it should use the requested size if valid

Signed-off-by: Helen Koike <helen.koike@collabora.co.uk>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] solo6x10: Simplify solo_enum_ext_input

Additionally, now it specifies which channels it's showing.

Signed-off-by: Ismael Luceno <ismael@iodev.co.uk>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] adv7604: Don't ignore pad number in subdev DV timings pad operations

The dv_timings_cap() and enum_dv_timings() pad operations take a pad
number as an input argument and return the DV timings capabilities and
list of supported DV timings for that pad.

Commit bd3e275f3ec0 ("[media] media: i2c: adv7604: Use v4l2-dv-timings
helpers") broke this as it started ignoring the pad number, always
returning the information associated with the currently selected input.
Fix it.

Fixes: bd3e275f3ec0 ("[media] media: i2c: adv7604: Use v4l2-dv-timings helpers")

Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l: mt9v032: Add V4L2 controls for AEC and AGC

This patch adds V4L2 controls for Auto Exposure Control and Auto Gain
Control settings. These settings include low pass filter, update
frequency of these settings and the update interval for those units.

[Avoid forward declarations]
[Fix 80 columns limit violation]
[Rename controls to avoid underscores in names]
[Fix the AEC maximum shutter width on MT9V032]

Signed-off-by: Markus Pargmann <mpa@pengutronix.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l: mt9v032: Do not unset master_mode

The power_on function of the driver resets the chip and sets the
CHIP_CONTROL register to 0. This switches the operating mode to slave.
The s_stream function sets the correct mode. But this caused problems on
a board where the camera chip is operated as master. The camera started
after a random amount of time streaming an image, I observed between 10
and 300 seconds.

The STRFM_OUT and STLN_OUT pins are not connected on this board which
may cause some issues in slave mode. I could not find any documentation
about this.

Keeping the chip in master mode after the reset helped to fix this
issue for me.

Signed-off-by: Markus Pargmann <mpa@pengutronix.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l: mt9v032: Remove duplicate test for I2C_FUNC_SMBUS_WORD_DATA functionality

Since commit b42261078a91 ("regmap: i2c: fallback to SMBus if the adapter
does not support standard I2C"), regmap-i2c will check the
I2C_FUNC_SMBUS_[BYTE|WORD]_DATA functionality based on the regmap_config
setting if the adapter does not support standard I2C.

So remove the I2C_FUNC_SMBUS_WORD_DATA functionality check in the driver code.

Signed-off-by: Axel Lin <axel.lin@ingics.com>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l: mt9v032: Remove unneeded header

The of_gpio.h header isn't needed, don't include it.

Reported-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l: mt9t001: fix clean up in case of power-on failures

If the driver fails to reset the camera or to set up control handlers, it
has to power the camera back off.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] v4l: mt9t001: constify v4l2_subdev_internal_ops structure

This v4l2_subdev_internal_ops structure is never modified.  All other
v4l2_subdev_internal_ops structures are declared as const.

Done with the help of Coccinelle.

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] smiapp: Remove useless rval assignment in smiapp_get_pdata()

The value is not used after the assignment.

Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] tw686x: be explicit about the possible dma_mode options

Users won't know what to put in this module option if it isn't
described.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] tw686x: audio: Prevent hw param changes while busy

Audio hw params are shared across all DMA channels,
so if the user changes any of these while any DMA channel is
enabled, it will impact the enabled channels, potentially causing
serious instability issues.

This commit avoids such situation, by preventing any hw param
change (on any DMA channel) if any other DMA audio channel is capturing.

Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] tw686x: audio: Allow to configure the period size

Currently, the driver has a fixed period size of 4096 bytes
(2048 frames). Since this hardware can configure the audio
capture size, this commit allows a period size range of [512-4096].

This is very useful to reduce the audio latency.

Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] tw686x: audio: Implement non-memcpy capture

Now that we've introduced the dma_mode parameter to pick the
DMA operation, let's use it to also select the audio DMA
operation.

When dma_mode != memcpy, the driver will avoid using memcpy
in the audio capture path, and the DMA hardware operation
will act directly on the ALSA buffers.

Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] tw686x: Add support for DMA scatter-gather mode

Now that the driver has the infrastructure to support more
DMA modes, let's add the DMA scatter-gather mode.

In this mode, the device delivers sequential top-bottom
frames. The scatter-gather logic is based on staging's
tw686x-kh driver (by Krzysztof Ha?asa).

Tested-by: Tim Harvey <tharvey@gateworks.com>
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] tw686x: Add support for DMA contiguous interlaced frame mode

Now that the driver has the infrastructure to support more
DMA modes, let's add the DMA contiguous interlaced frame mode.

In this mode, the DMA P and B buffers are programmed with
the user-provided buffers. When a P (or B) frame is ready,
a new buffer is dequeued into P (or B).

In addition to interlaced fields, the device can also be
programmed to deliver alternate fields. Only interlaced
mode is supported for now.

Tested-by: Tim Harvey <tharvey@gateworks.com>
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] tw686x: Introduce an interface to support multiple DMA modes

Let's set the corner stone to support all the DMA modes
available on this device.

For stability reasons, the driver is currently setting DMA frame
mode, and using single DMA buffers to get the P and B buffers.
Each frame is then memcpy'ed into the user buffer.

However, other platforms might be interested in avoiding this
memcpy, or in taking advantage of the chip's DMA scatter-gather
capabilities.

To achieve this, this commit introduces a "dma_mode" module parameter,
and a tw686x_dma_ops struct. This will allow to define functions to
alloc/free DMA buffers, and to return the frames to userspace.

The memcpy-based method described above is named as dma_mode="memcpy".
Current alloc/free functions are renamed as tw686x_memcpy_xxx,
and are now used through a memcpy_dma_ops.

Tested-by: Tim Harvey <tharvey@gateworks.com>
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

bdisp: move the V/H filter spec to bdisp-hw.c

Those structs are used only at bdisp-hw, so they shouldn't be
there in a header file that it is used elsewhere.

This fixes the following Gcc 6.1 warnings:

In file included from drivers/media/platform/sti/bdisp/bdisp-debug.c:11:0:
drivers/media/platform/sti/bdisp/bdisp-filter.h:207:65: warning: ‘bdisp_v_spec’ defined but not used [-Wunused-const-variable=]
 static const struct __maybe_unused bdisp_filter_v_spec bdisp_v_spec[] = {
                                                                 ^~~~~~~~~
In file included from drivers/media/platform/sti/bdisp/bdisp-debug.c:11:0:
drivers/media/platform/sti/bdisp/bdisp-filter.h:23:65: warning: ‘bdisp_h_spec’ defined but not used [-Wunused-const-variable=]
 static const struct __maybe_unused bdisp_filter_h_spec bdisp_h_spec[] = {
                                                                 ^~~~~~~~~

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

adv7842: comment out a table useful for debug

Gcc 6.1 warns about an unused table:

drivers/media/i2c/adv7842.c:2400:27: warning: 'prim_mode_txt' defined but not used [-Wunused-const-variable=]
 static const char * const prim_mode_txt[] = {
                           ^~~~~~~~~~~~~

That seems to be useful for debug, and likely were used before.
While we could simply remove, let's comment it out, for now.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

em28xx-dvb: remove some left over

Gcc 6.1 warns about an unused table:

drivers/media/usb/em28xx/em28xx-dvb.c:907:38: warning: 'pctv_461e_m88ds3103_config' defined but not used [-Wunused-const-variable=]
 static const struct m88ds3103_config pctv_461e_m88ds3103_config = {
                                      ^~~~~~~~~~~~~~~~~~~~~~~~~~

That's a left over of patch 76b91be3d360a ('em28xx: PCTV 461e use I2C
client for demod and SEC').

Remove the dead code.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

vivid: remove some unused vars

Gcc 6.1 warns about some unused vars. Remove them:
drivers/media/platform/vivid/vivid-vid-cap.c:40:2: warning: 'tpf_default' defined but not used [-Wunused-const-variable=]
  tpf_default = {.numerator = 1,  .denominator = 30};
  ^~~~~~~~~~~
drivers/media/platform/vivid/vivid-sdr-cap.c:54:27: warning: 'NUM_FORMATS' defined but not used [-Wunused-const-variable=]
 static const unsigned int NUM_FORMATS = ARRAY_SIZE(formats);
                           ^~~~~~~~~~~

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

zr36016: remove some unused tables

Gcc 6.1 warns about some unused tables:

drivers/media/pci/zoran/zr36016.c:251:18: warning: 'zr016_yoff' defined but not used [-Wunused-const-variable=]
 static const int zr016_yoff[] = { 8, 9, 7 };
                  ^~~~~~~~~~
drivers/media/pci/zoran/zr36016.c:250:18: warning: 'zr016_xoff' defined but not used [-Wunused-const-variable=]
 static const int zr016_xoff[] = { 20, 20, 20 };
                  ^~~~~~~~~~

Those tables aren't used anywere. So, remove them.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

r820t: comment out two ancillary tables

As Gcc6.1 warned, those tables are currently unused:
drivers/media/tuners/r820t.c:349:18: warning: 'r820t_mixer_gain_steps' defined but not used [-Wunused-const-variable=]
 static const int r820t_mixer_gain_steps[]  = {
                  ^~~~~~~~~~~~~~~~~~~~~~
drivers/media/tuners/r820t.c:345:18: warning: 'r820t_lna_gain_steps' defined but not used [-Wunused-const-variable=]
 static const int r820t_lna_gain_steps[]  = {
                  ^~~~~~~~~~~~~~~~~~~~

They're actually used only by a routine that it is currently
commented out. So, move those tables to be together with such
code and comment them out.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

dib0090: comment out the unused tables

Those tables are currently unused, so comment them out:

drivers/media/dvb-frontends/dib0090.c:852:18: warning: 'rf_ramp_pwm_sband' defined but not used [-Wunused-const-variable=]
 static const u16 rf_ramp_pwm_sband[] = {
                  ^~~~~~~~~~~~~~~~~
drivers/media/dvb-frontends/dib0090.c:800:18: warning: 'bb_ramp_pwm_boost' defined but not used [-Wunused-const-variable=]
 static const u16 bb_ramp_pwm_boost[] = {
                  ^~~~~~~~~~~~~~~~~

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

drxj: comment out the unused nicam_presc_table_val table

Avoid this warning:

drivers/media/dvb-frontends/drx39xyj/drxj.c:1243:18: warning: 'nicam_presc_table_val' defined but not used [-Wunused-const-variable=]
 static const u16 nicam_presc_table_val[43] = {
                  ^~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

cx25821-alsa: shutup a Gcc 6.1 warning

The PCI device ID table is only used if compiled with modules
support. When compiled with modules disabled, this is now
producing this bogus warning:

drivers/media/pci/cx25821/cx25821-alsa.c:696:35: warning: 'cx25821_audio_pci_tbl' defined but not used [-Wunused-const-variable=]
 static const struct pci_device_id cx25821_audio_pci_tbl[] = {
                                   ^~~~~~~~~~~~~~~~~~~~~

Fix it by annotating that the function may not be used.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

m5602_ov7660: move skeletons to the .c file

The header file has some private static structures that
are used only by the C file. Move those structures to the C file,
in order to shut up gcc 6.1 warnings.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

m5602_po1030: move skeletons to the .c file

The header file has some private static structures that
are used only by the C file. Move those structures to the C file,
in order to shut up gcc 6.1 warnings.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

m5602_s5k83a: move skeletons to the .c file

The header file has some private static structures that
are used only by the C file. Move those structures to the C file,
in order to shut up gcc 6.1 warnings.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

m5602_ov9650: move skeletons to the .c file

The header file has some private static structures that
are used only by the C file. Move those structures to the C file,
in order to shut up gcc 6.1 warnings.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

m5602_mt9m111: move skeletons to the .c file

The header file has some private static structures that
are used only by the C file. Move those structures to the C file,
in order to shut up gcc 6.1 warnings.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

m5602_s5k4aa: move skeletons to the .c file

The mc5602_s5k4aa.h has some private static structures that
are used only by the C file. Move those structures to the C file,
in order to shut up gcc 6.1 warnings.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

m5602_core: move skeletons to the .c file

The mc5602_brigde.h is included at m5602 submodules. This
causes Gcc 6.1 to complain:

drivers/media/usb/gspca/m5602/m5602_bridge.h:124:28: warning: 'sensor_urb_skeleton' defined but not used [-Wunused-const-variable=]
 static const unsigned char sensor_urb_skeleton[] = {
                            ^~~~~~~~~~~~~~~~~~~
drivers/media/usb/gspca/m5602/m5602_bridge.h:119:28: warning: 'bridge_urb_skeleton' defined but not used [-Wunused-const-variable=]
 static const unsigned char bridge_urb_skeleton[] = {
                           ^~~~~~~~~~~~~~~~~~~

Let's shut up gcc 6.1 warnings by moving those data structures
to the core, as they're used only there.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

cx18: use macros instead of static const vars

Gcc 6.1 now complains about unused vars:

drivers/media/pci/cx18/cx18-driver.h:497:18: warning: 'vbi_hblank_samples_50Hz' defined but not used [-Wunused-const-variable=]
 static const u32 vbi_hblank_samples_50Hz = 284; /* 4 byte EAV + 280 anc/fill */
                  ^~~~~~~~~~~~~~~~~~~~~~~
drivers/media/pci/cx18/cx18-driver.h:496:18: warning: 'vbi_hblank_samples_60Hz' defined but not used [-Wunused-const-variable=]
 static const u32 vbi_hblank_samples_60Hz = 272; /* 4 byte EAV + 268 anc/fill */
                  ^~~~~~~~~~~~~~~~~~~~~~~
In file included from drivers/media/pci/cx18/cx18-cards.c:25:0:
drivers/media/pci/cx18/cx18-driver.h:497:18: warning: 'vbi_hblank_samples_50Hz' defined but not used [-Wunused-const-variable=]
 static const u32 vbi_hblank_samples_50Hz = 284; /* 4 byte EAV + 280 anc/fill */
                  ^~~~~~~~~~~~~~~~~~~~~~~
drivers/media/pci/cx18/cx18-driver.h:496:18: warning: 'vbi_hblank_samples_60Hz' defined but not used [-Wunused-const-variable=]
 static const u32 vbi_hblank_samples_60Hz = 272; /* 4 byte EAV + 268 anc/fill */
                  ^~~~~~~~~~~~~~~~~~~~~~~
drivers/media/pci/cx18/cx18-driver.h:495:18: warning: 'vbi_active_samples' defined but not used [-Wunused-const-variable=]
 static const u32 vbi_active_samples = 1444; /* 4 byte SAV + 720 Y + 720 U/V */
                  ^~~~~~~~~~~~~~~~~~

In this specific case, this is somewhat intentional, as those
values are actually used in parts of the driver. The code assumes
that gcc optimizer it and not actually create any var, but convert
it to immediate access at the routines.

Yet, as we want to shut up gcc warnings, let's use #define, with
is the standard way to store values that will use assembler's
immediate access code.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

exynos4-is: remove some unused vars

Gcc 6.1 warns about some unused vars and functions. Remove them:

drivers/media/platform/exynos4-is/mipi-csis.c:665:46: warning: 's5pcsis_sd_internal_ops' defined but not used [-Wunused-const-variable=]
 static const struct v4l2_subdev_internal_ops s5pcsis_sd_internal_ops = {
                                              ^~~~~~~~~~~~~~~~~~~~~~~

drivers/media/platform/exynos4-is/mipi-csis.c:652:12: warning: 's5pcsis_open' defined but not used [-Wunused-function]
 static int s5pcsis_open(struct v4l2_subdev *sd, struct v4l2_subdev_fh *fh)
            ^~~~~~~~~~~~

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

usbvision: remove some unused vars

Gcc 6.1 warns about some unused vars. Remove them:

drivers/media/usb/usbvision/usbvision-core.c:94:18: warning: 'min_imgheight' defined but not used [-Wunused-const-variable=]
 static const int min_imgheight = MIN_FRAME_HEIGHT;
                  ^~~~~~~~~~~~~
drivers/media/usb/usbvision/usbvision-core.c:93:18: warning: 'min_imgwidth' defined but not used [-Wunused-const-variable=]
 static const int min_imgwidth = MIN_FRAME_WIDTH;
                  ^~~~~~~~~~~~
drivers/media/usb/usbvision/usbvision-core.c:92:18: warning: 'max_imgheight' defined but not used [-Wunused-const-variable=]
 static const int max_imgheight = MAX_FRAME_HEIGHT;
                  ^~~~~~~~~~~~~
drivers/media/usb/usbvision/usbvision-core.c:91:18: warning: 'max_imgwidth' defined but not used [-Wunused-const-variable=]
 static const int max_imgwidth = MAX_FRAME_WIDTH;
                  ^~~~~~~~~~~~

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media-devnode.h: Fix documentation

Two parameters were documented with a wrong name, and a struct
device pointer description was missing.

That caused the following warnings, when building documentation:

include/media/media-devnode.h:102: warning: No description found for parameter 'media_dev'
include/media/media-devnode.h:126: warning: No description found for parameter 'mdev'
include/media/media-devnode.h:126: warning: Excess function parameter 'media_dev' description in 'media_devnode_register'

Rename the description, to match the function parameter and fix
Documentation.

No funcional changes.

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] rcar-vin: get rid of an unused var

drivers/media/platform/rcar-vin/rcar-core.c: In function 'rvin_graph_notify_complete':
drivers/media/platform/rcar-vin/rcar-core.c:65:22: warning: variable 'sd' set but not used [-Wunused-but-set-variable]
  struct v4l2_subdev *sd;
                      ^

Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] rcar-vin: add Renesas R-Car VIN driver

A V4L2 driver for Renesas R-Car VIN driver that do not depend on
soc_camera. The driver is heavily based on its predecessor and aims to
replace it.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] videobuf2-v4l2: Verify planes array in buffer dequeueing

When a buffer is being dequeued using VIDIOC_DQBUF IOCTL, the exact buffer
which will be dequeued is not known until the buffer has been removed from
the queue. The number of planes is specific to a buffer, not to the queue.

This does lead to the situation where multi-plane buffers may be requested
and queued with n planes, but VIDIOC_DQBUF IOCTL may be passed an argument
struct with fewer planes.

__fill_v4l2_buffer() however uses the number of planes from the dequeued
videobuf2 buffer, overwriting kernel memory (the m.planes array allocated
in video_usercopy() in v4l2-ioctl.c)  if the user provided fewer
planes than the dequeued buffer had. Oops!

Fixes: b0e0e1f83de3 ("[media] media: videobuf2: Prepare to divide videobuf2")

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: stable@vger.kernel.org # for v4.4 and later
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] vb2: core: Skip planes array verification if pb is NULL

An earlier patch fixing an input validation issue introduced another
issue: vb2_core_dqbuf() is called with pb argument value NULL in some
cases, causing a NULL pointer dereference. Fix this by skipping the
verification as there's nothing to verify.

Fixes: e7e0c3e26587 ("[media] videobuf2-core: Check user space planes array in dqbuf")

Signed-off-by: David R <david@unsolicited.net>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: stable@vger.kernel.org # for v4.4 and later
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media: fix media devnode ioctl/syscall and unregister race

Media devnode open/ioctl could be in progress when media device unregister
is initiated. System calls and ioctls check media device registered status
at the beginning, however, there is a window where unregister could be in
progress without changing the media devnode status to unregistered.

process 1 process 2
fd = open(/dev/media0)
media_devnode_is_registered()
(returns true here)

media_device_unregister()
(unregister is in progress
and devnode isn't
unregistered yet)
...
ioctl(fd, ...)
__media_ioctl()
media_devnode_is_registered()
(returns true here)
...
media_devnode_unregister()
...
(driver releases the media device
memory)

media_device_ioctl()
(By this point
devnode->media_dev does not
point to allocated memory.
use-after free in in mutex_lock_nested)

BUG: KASAN: use-after-free in mutex_lock_nested+0x79c/0x800 at addr
ffff8801ebe914f0

Fix it by clearing register bit when unregister starts to avoid the race.

process 1                               process 2
fd = open(/dev/media0)
media_devnode_is_registered()
        (could return true here)

                                        media_device_unregister()
                                                (clear the register bit,
 then start unregister.)
                                        ...
ioctl(fd, ...)
__media_ioctl()
media_devnode_is_registered()
        (return false here, ioctl
 returns I/O error, and
 will not access media
 device memory)
                                        ...
                                        media_devnode_unregister()
                                        ...
                                        (driver releases the media device
 memory)

Signed-off-by: Shuah Khan <shuahkh@osg.samsung.com>
Suggested-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reported-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Tested-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media: fix use-after-free in cdev_put() when app exits after driver unbind

When driver unbinds while media_ioctl is in progress, cdev_put() fails with
when app exits after driver unbinds.

Add devnode struct device kobj as the cdev parent kobject. cdev_add() gets
a reference to it and releases it in cdev_del() ensuring that the devnode
is not deallocated as long as the application has the device file open.

media_devnode_register() initializes the struct device kobj before calling
cdev_add(). media_devnode_unregister() does cdev_del() and then deletes the
device. devnode is released when the last reference to the struct device is
gone.

This problem is found on uvcvideo, em28xx, and au0828 drivers and fix has
been tested on all three.

kernel: [  193.599736] BUG: KASAN: use-after-free in cdev_put+0x4e/0x50
kernel: [  193.599745] Read of size 8 by task media_device_te/1851
kernel: [  193.599792] INFO: Allocated in __media_device_register+0x54
kernel: [  193.599951] INFO: Freed in media_devnode_release+0xa4/0xc0

kernel: [  193.601083] Call Trace:
kernel: [  193.601093]  [<ffffffff81aecac3>] dump_stack+0x67/0x94
kernel: [  193.601102]  [<ffffffff815359b2>] print_trailer+0x112/0x1a0
kernel: [  193.601111]  [<ffffffff8153b5e4>] object_err+0x34/0x40
kernel: [  193.601119]  [<ffffffff8153d9d4>] kasan_report_error+0x224/0x530
kernel: [  193.601128]  [<ffffffff814a2c3d>] ? kzfree+0x2d/0x40
kernel: [  193.601137]  [<ffffffff81539d72>] ? kfree+0x1d2/0x1f0
kernel: [  193.601154]  [<ffffffff8157ca7e>] ? cdev_put+0x4e/0x50
kernel: [  193.601162]  [<ffffffff8157ca7e>] cdev_put+0x4e/0x50
kernel: [  193.601170]  [<ffffffff815767eb>] __fput+0x52b/0x6c0
kernel: [  193.601179]  [<ffffffff8117743a>] ? switch_task_namespaces+0x2a
kernel: [  193.601188]  [<ffffffff815769ee>] ____fput+0xe/0x10
kernel: [  193.601196]  [<ffffffff81170023>] task_work_run+0x133/0x1f0
kernel: [  193.601204]  [<ffffffff8117746e>] ? switch_task_namespaces+0x5e
kernel: [  193.601213]  [<ffffffff8111b50c>] do_exit+0x72c/0x2c20
kernel: [  193.601224]  [<ffffffff8111ade0>] ? release_task+0x1250/0x1250
-
-
-
kernel: [  193.601360]  [<ffffffff81003587>] ? exit_to_usermode_loop+0xe7
kernel: [  193.601368]  [<ffffffff810035c0>] exit_to_usermode_loop+0x120
kernel: [  193.601376]  [<ffffffff810061da>] syscall_return_slowpath+0x16a
kernel: [  193.601386]  [<ffffffff82848b33>] entry_SYSCALL_64_fastpath+0xa6

Signed-off-by: Shuah Khan <shuahkh@osg.samsung.com>
Tested-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media-device: dynamically allocate struct media_devnode

struct media_devnode is currently embedded at struct media_device.

While this works fine during normal usage, it leads to a race
condition during devnode unregister. the problem is that drivers
assume that, after calling media_device_unregister(), the struct
that contains media_device can be freed. This is not true, as it
can't be freed until userspace closes all opened /dev/media devnodes.

In other words, if the media devnode is still open, and media_device
gets freed, any call to an ioctl will make the core to try to access
struct media_device, with will cause an use-after-free and even GPF.

Fix this by dynamically allocating the struct media_devnode and only
freeing it when it is safe.

Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] media-devnode: fix namespace mess

Along all media controller code, "mdev" is used to represent
a pointer to struct media_device, and "devnode" for a pointer
to struct media_devnode.

However, inside media-devnode.[ch], "mdev" is used to represent
a pointer to struct media_devnode.

This is very confusing and may lead to development errors.

So, let's change all occurrences at media-devnode.[ch] to
also use "devnode" for such pointers.

This patch doesn't make any functional changes.

Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

Update my main e-mails at the Kernel tree

For the third time in three years, I'm changing my e-mail at
Samsung. That's bad, as it may stop communications with me for
a while. So, this time, I'll also the mchehab@kernel.org e-mail,
as it remains stable since ever.

Cc: stable@vger.kernel.org
Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>

[media] m88rs2000: initialize status to zero

status is not initialized so it can contain garbage. The
check for status containing the FE_HAS_LOCK bit may randomly pass
or fail if the read of register 0x8c fails to set status after 25
read attempts.  Fix this by initializing status to 0.

Issue found with CoverityScan, CID#986738

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] em28xx-i2c: rt_mutex_trylock() returns zero on failure

The code is checking for negative returns but it should be checking for
zero.

Fixes: aab3125c43d8 ('[media] em28xx: add support for registering multiple i2c buses')

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] mt2063: use lib gcd

This patch removes the local MT2063_gcd function, uses lib gcd instead

Signed-off-by: Zhaoxiu Zeng <zhaoxiu.zeng@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] rtl2832: add support for slave ts pid filter

The rtl2832 demod has 2 sets of PID filters. This patch enables
the filter support when using a slave demod.

Signed-off-by: Benjamin Larsson <benjamin@southpole.se>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] af9035: fix logging

Remove __func__ and KBUILD_MODNAME from logging formatters and pass
USB interface device instead, so logging can be done correctly.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] af9035: fix for MXL5007T devices with I2C read issues

The MXL5007T tuner will lock-up on some devices after an I2C read
transaction.  This patch works around this issue by inhibiting such
operations and emulating a 0x00 response. The workaround is only
applied to USB devices known to exhibit this flaw.

Signed-off-by: Alessandro Radicati <alessandro@radicati.net>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] af9035: I2C combined write + read transaction fix

This patch will modify the af9035 driver to use the register address
fields of the I2C read command for the combined write/read transaction
case.  Without this change, the firmware issues just a I2C read
transaction without the preceding write transaction to select the
register.

Signed-off-by: Alessandro Radicati <alessandro@radicati.net>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] of: reserved_mem: restore old behavior when no region is defined

Change return value back to -ENODEV when no region is defined for given
device. This restores old behavior of this function, as some drivers rely
on such error code.

Fixes: 59ce4039727ef40 ("of: reserved_mem: add support for using more than
       one region for given device")

Reported-by: Liviu Dudau <liviu.dudau@arm.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Reviewed-by: Liviu Dudau <Liviu.Dudau@arm.com>
Acked-by: Rob Herring <robh@kernel.org>
Reviewed-by: Sumit Semwal <sumit.semwal@linaro.org>

[media] drivers/media/dvb-core/en50221: move code to dvb_ca_private_free()

Prepare for postponing the call until all file handles have been
closed.

[mchehab@osg.samsung.com: make checkpatch happy]
Signed-off-by: Max Kellermann <max@duempel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] ds3000: return error if invalid symbol rate is set

Return -EINVAL if ds3000_set_frontend is called with invalid parameters.

v1 of the patch series got incorrect subject lines.

Signed-off-by: Olli Salonen <olli.salonen@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] fix semicolon.cocci warnings

drivers/media/dvb-frontends/helene.c:750:2-3: Unneeded semicolon

 Remove unneeded semicolon.

Generated by: scripts/coccinelle/misc/semicolon.cocci

CC: Abylay Ospan <aospan@netup.ru>
Signed-off-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] rtl28xxu: sort the config symbols which are auto-selected

No functional changes.

Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] rtl28xxu: auto-select more DVB-frontends and tuners

This adds the missing auto-select bits for DVB-frontends and tuners
(if MEDIA_SUBDRV_AUTOSELECT is enabled) which are used by the various
rtl28xxu devices.
The driver itself probes for three more tuners, but it's not actually
using any of them:
- MEDIA_TUNER_MT2063
- MEDIA_TUNER_MT2266
- MEDIA_TUNER_MXL5007T

Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] mn88472: move out of staging to media

Move mn88472 DVB-T/T2/C demod driver out of staging to media.

v2: Fix build error reported by kbuild test robot:
drivers/staging/media/mn88472/Makefile: No such file or directory

Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] mn88472: finalize driver

Finalize driver in order to move out of staging.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] mn88472: fix typo

firmare -> firmware

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] rtl28xxu: increase failed I2C msg repeat count to 3

1 and 2 wasn't enough for mn88472 chip on Astrometa device,
so increase it to 3.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] mn88473: fix typo

firmare -> firmware

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] mn88473: fix error path on probe()

Latest, 3rd, regmap instance should be freed on error case.

Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] staging/media: remove deprecated timb driver

Remove this deprecated old driver.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: Richard Röjfors <richard@puffinpack.se>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] staging/media: remove deprecated omap1 driver

Remove this deprecated old driver.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] staging/media: remove deprecated mx3 driver

Remove this deprecated old driver.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] staging/media: remove deprecated mx2 driver

Remove this deprecated old driver.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] s5p-mfc: fix a typo in s5p_mfc_dec

It is a cosmetic commit.

Signed-off-by: ayaka <ayaka@soulik.info>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] s5p-mfc: remove unnecessary check in try_fmt

We don't need to request the sizeimage or num_planes
in try_fmt.

Signed-off-by: ayaka <ayaka@soulik.info>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

[media] s5p-mfc: Add handling of buffer freeing reqbufs request

The encoder forget the work to call hardware to release its buffers.
This patch came from chromium project. I just change its code
style and make the API match with new kernel.

Signed-off-by: ayaka <ayaka@soulik.info>
Signed-off-by: Kamil Debski <k.debski@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>