OSDN Git Service

tomoyo/tomoyo-test1.git
5 years agomedia: imx-csi: Double crop height for alternate fields at sink
Steve Longerbeam [Wed, 9 Jan 2019 18:30:08 +0000 (13:30 -0500)]
media: imx-csi: Double crop height for alternate fields at sink

If the incoming sink field type is alternate, the reset crop height
and crop height bounds must be set to twice the incoming height,
because in alternate field mode, upstream will report only the
lines for a single field, and the CSI captures the whole frame.

Signed-off-by: Steve Longerbeam <slongerbeam@gmail.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: imx: Fix field negotiation
Steve Longerbeam [Wed, 9 Jan 2019 18:30:07 +0000 (13:30 -0500)]
media: imx: Fix field negotiation

IDMAC interlaced scan, a.k.a. interweave, should be enabled in the
IDMAC output channels only if the IDMAC output pad field type is
'seq-bt' or 'seq-tb', and field type at the capture interface is
'interlaced*'.

V4L2_FIELD_HAS_BOTH() macro should not be used on the input to determine
enabling interlaced/interweave scan. That macro includes the 'interlaced'
field types, and in those cases the data is already interweaved with
top/bottom field lines.

The CSI will capture whole frames when the source specifies alternate
field mode. So the CSI also enables interweave for alternate input
field type and the field type at capture interface is interlaced.

Fix the logic for setting field type in try_fmt in CSI entity.
The behavior should be:

- No restrictions on field type at sink pad.

- At the output pads, allow sequential fields in TB order, if the sink pad
  field type is sequential or alternate. Otherwise passthrough the field
  type from sink to source pad.

Move this logic to new function csi_try_field().

These changes result in the following allowed field transformations
from CSI sink -> source pads (all other field types at sink are passed
through to source):

seq-tb -> seq-tb
seq-bt -> seq-tb
alternate -> seq-tb

In a future patch, the CSI sink -> source will allow:

seq-tb -> seq-bt
seq-bt -> seq-bt
alternate -> seq-bt

This will require supporting interweave with top/bottom line swapping.
Until then seq-bt is not allowed at the CSI source pad because there is
no way to swap top/bottom lines when interweaving to INTERLACED_BT --
note that despite the name, INTERLACED_BT is top-bottom order in memory.
The BT in this case refers to field dominance: the bottom lines are
older in time than the top lines.

The capture interface device allows selecting IDMAC interweave by
choosing INTERLACED_TB if the CSI/PRPENCVF source pad is seq-tb and
INTERLACED_BT if the source pad is seq-bt (for future support of seq-bt).

Signed-off-by: Steve Longerbeam <slongerbeam@gmail.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gpu: ipu-v3: Add planar support to interlaced scan
Steve Longerbeam [Wed, 9 Jan 2019 18:30:06 +0000 (13:30 -0500)]
media: gpu: ipu-v3: Add planar support to interlaced scan

To support interlaced scan with planar formats, cpmem SLUV must
be programmed with the correct chroma line stride. For full and
partial planar 4:2:2 (YUV422P, NV16), chroma line stride must
be doubled. For full and partial planar 4:2:0 (YUV420, YVU420, NV12),
chroma line stride must _not_ be doubled, since a single chroma line
is shared by two luma lines.

Signed-off-by: Steve Longerbeam <slongerbeam@gmail.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gpu: ipu-csi: Swap fields according to input/output field types
Steve Longerbeam [Wed, 9 Jan 2019 18:30:05 +0000 (13:30 -0500)]
media: gpu: ipu-csi: Swap fields according to input/output field types

The function ipu_csi_init_interface() was inverting the F-bit for
NTSC case, in the CCIR_CODE_1/2 registers. The result being that
for NTSC bottom-top field order, the CSI would swap fields and
capture in top-bottom order.

Instead, base field swap on the field order of the input to the CSI,
and the field order of the requested output. If the input/output
fields are sequential but different, swap fields, otherwise do
not swap. This requires passing both the input and output mbus
frame formats to ipu_csi_init_interface().

Move this code to a new private function ipu_csi_set_bt_interlaced_codes()
that programs the CCIR_CODE_1/2 registers for interlaced BT.656 (and
possibly interlaced BT.1120 in the future).

When detecting input video standard from the input frame width/height,
make sure to double height if input field type is alternate, since
in that case input height only includes lines for one field.

Signed-off-by: Steve Longerbeam <slongerbeam@gmail.com>
Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: videodev2.h: Add more field helper macros
Steve Longerbeam [Wed, 9 Jan 2019 18:30:04 +0000 (13:30 -0500)]
media: videodev2.h: Add more field helper macros

Adds two helper macros:

V4L2_FIELD_IS_SEQUENTIAL: returns true if the given field type is
'sequential', that is a full frame is transmitted, or exists in
memory, as all top field lines followed by all bottom field lines,
or vice-versa.

V4L2_FIELD_IS_INTERLACED: returns true if the given field type is
'interlaced', that is a full frame is transmitted, or exists in
memory, as top field lines interlaced with bottom field lines.

Signed-off-by: Steve Longerbeam <slongerbeam@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: rcar-vin: Allow independent VIN link enablement
Steve Longerbeam [Tue, 15 Jan 2019 01:10:19 +0000 (20:10 -0500)]
media: rcar-vin: Allow independent VIN link enablement

There is a block of code in rvin_group_link_notify() that prevents
enabling a link to a VIN node if any entity in the media graph is
in use. This prevents enabling a VIN link even if there is an in-use
entity somewhere in the graph that is independent of the link's
pipeline.

For example, the code block will prevent enabling a link from
the first rcar-csi2 receiver to a VIN node even if there is an
enabled link somewhere far upstream on the second independent
rcar-csi2 receiver pipeline.

If this code block is meant to prevent modifying a link if any entity
in the graph is actively involved in streaming (because modifying
the CHSEL register fields can disrupt any/all running streams), then
the entities stream counts should be checked rather than the use counts.

(There is already such a check in __media_entity_setup_link() that verifies
the stream_count of the link's source and sink entities are both zero,
but that is insufficient, since there should be no running streams in
the entire graph).

Modify the code block to check the entity stream_count instead of the
use_count (and elaborate on the comment). VIN node links can now be
enabled even if there are other independent in-use entities that are
not streaming.

Fixes: c0cc5aef31 ("media: rcar-vin: add link notify for Gen3")

Signed-off-by: Steve Longerbeam <slongerbeam@gmail.com>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: coda: Add control for h.264 chroma qp index offset
Philipp Zabel [Thu, 10 Jan 2019 16:56:12 +0000 (11:56 -0500)]
media: coda: Add control for h.264 chroma qp index offset

Allow to set a fixed quantization parameter offset between luma and
chroma in the h.264 encoder.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: coda: Add control for h.264 constrained intra prediction
Philipp Zabel [Thu, 10 Jan 2019 16:56:11 +0000 (11:56 -0500)]
media: coda: Add control for h.264 constrained intra prediction

Allow to enable constrained intra prediction in the h.264 encoder.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: v4l2-ctrl: Add control for h.264 chroma qp offset
Philipp Zabel [Thu, 10 Jan 2019 16:56:10 +0000 (11:56 -0500)]
media: v4l2-ctrl: Add control for h.264 chroma qp offset

Allow to add fixed quantization parameter offset between luma and
chroma quantization parameters. This control directly corresponds
to the chroma_qp_index_offset field of the h.264 picture parameter
set.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: v4l2-ctrl: Add control to enable h.264 constrained intra prediction
Philipp Zabel [Thu, 10 Jan 2019 16:56:09 +0000 (11:56 -0500)]
media: v4l2-ctrl: Add control to enable h.264 constrained intra prediction

Allow to enable h.264 constrained intra prediction (macroblocks using
intra prediction modes are not allowed to use residual data and decoded
samples of neighboring macroblocks coded using inter prediction modes).
This control directly corresponds to the constrained_intra_pred_flag
field in the h.264 picture parameter set.

Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: ov534-ov772x: remove unnecessary COM3 initialization
Philipp Zabel [Fri, 14 Dec 2018 16:40:31 +0000 (11:40 -0500)]
media: gspca: ov534-ov772x: remove unnecessary COM3 initialization

The COM3 register at address 0x0c already defaults to 0x10, the two bits
COM3[7:6] are set according to V4L2 controls by sethvflip later.
There is no need to set it multiple times during bridge initialization.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: ov534-ov722x: remove camera clock setup from bridge_init
Philipp Zabel [Fri, 14 Dec 2018 16:40:30 +0000 (11:40 -0500)]
media: gspca: ov534-ov722x: remove camera clock setup from bridge_init

This register is later overwritten by set_frame_rate anyway.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: ov534-ov722x: remove mode specific video data registers from bridge_init
Philipp Zabel [Fri, 14 Dec 2018 16:40:29 +0000 (11:40 -0500)]
media: gspca: ov534-ov722x: remove mode specific video data registers from bridge_init

The video format, payload size, and frame size setup is video format
and frame size specific. Those registers are overwritten during
bridge_start anyway.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: ov534-ov772x: add SGBRG8 bayer mode support
Philipp Zabel [Fri, 14 Dec 2018 16:40:28 +0000 (11:40 -0500)]
media: gspca: ov534-ov772x: add SGBRG8 bayer mode support

Add support to pass through the sensor's native SGBRG8 bayer pattern,
allowing to cut the required USB bandwidth in half.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: ov543-ov772x: move video format specific registers into bridge_start
Philipp Zabel [Fri, 14 Dec 2018 16:40:27 +0000 (11:40 -0500)]
media: gspca: ov543-ov772x: move video format specific registers into bridge_start

In preparation for adding SGBRG8 as a second video format besides YUYV,
move video format specific register settings from the bridge_init array
into the bridge_start arrays.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: support multiple pixel formats in TRY_FMT
Philipp Zabel [Fri, 14 Dec 2018 16:40:26 +0000 (11:40 -0500)]
media: gspca: support multiple pixel formats in TRY_FMT

If a driver supports multiple pixel formats with the same frame size,
TRY_FMT will currently always return the first pixel format.
Fix this by adding pixelformat support to wxh_to_nearest_mode().

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: support multiple pixel formats in ENUM_FRAMEINTERVALS
Philipp Zabel [Fri, 14 Dec 2018 16:40:25 +0000 (11:40 -0500)]
media: gspca: support multiple pixel formats in ENUM_FRAMEINTERVALS

If a driver supports multiple pixel formats with the same frame size,
ENUM_FRAMEINTERVALS will currently only work for the first pixel format.
Fix this by adding pixelformat support to wxh_to_mode().

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: ov534: replace msleep(10) with usleep_range
Philipp Zabel [Fri, 14 Dec 2018 16:40:24 +0000 (11:40 -0500)]
media: gspca: ov534: replace msleep(10) with usleep_range

For short waits, usleep_range should be used instead of msleep,
see Documentation/timers/timers-howto.txt.

Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vim2m: the v4l2_m2m_buf_copy_data args were swapped
Hans Verkuil [Fri, 11 Jan 2019 11:43:12 +0000 (06:43 -0500)]
media: vim2m: the v4l2_m2m_buf_copy_data args were swapped

The buffer arguments to v4l2_m2m_buf_copy_data args were swapped.

The reason is confusing naming conventions in vim2m. It certainly
could be improved.

Fixes: 7aca565ee3d0 ("media: vim2m: use v4l2_m2m_buf_copy_data")

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vivid: do not implement VIDIOC_S_PARM for output streams
Hans Verkuil [Fri, 11 Jan 2019 11:37:03 +0000 (06:37 -0500)]
media: vivid: do not implement VIDIOC_S_PARM for output streams

v4l2_compliance gave a warning for the S_PARM test for output streams:

warn: v4l2-test-formats.cpp(1235): S_PARM is supported for buftype 2, but not for ENUM_FRAMEINTERVALS

The reason is that vivid mapped s_parm for output streams to g_parm. But if
S_PARM doesn't actually change anything, then it shouldn't be enabled at all.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vivid: take data_offset into account for video output
Hans Verkuil [Mon, 14 Jan 2019 16:23:52 +0000 (11:23 -0500)]
media: vivid: take data_offset into account for video output

The video output sizeimage calculation did not take data_offset into account.

This can cause problems with video loopback or exporting output buffers for
use as dmabuf import buffers since the output buffer size is now too small.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: v4l: uAPI: V4L2_BUF_TYPE_META_OUTPUT is an output buffer type
Sakari Ailus [Mon, 14 Jan 2019 14:01:54 +0000 (09:01 -0500)]
media: v4l: uAPI: V4L2_BUF_TYPE_META_OUTPUT is an output buffer type

V4L2_BUF_TYPE_META_OUTPUT was added by commit 72148d1a57e7
("media: v4l: Add support for V4L2_BUF_TYPE_META_OUTPUT") but the patch
missed adding the type to the macro telling whether a given type is an
output type or not. Do that now. Getting this wrong leads to handling the
buffer as a capture buffer in a lot of places.

Fixes: 72148d1a57e7 ("media: v4l: Add support for V4L2_BUF_TYPE_META_OUTPUT")

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vivid: disable VB2_USERPTR if dma_contig was configured
Hans Verkuil [Mon, 14 Jan 2019 13:50:54 +0000 (08:50 -0500)]
media: vivid: disable VB2_USERPTR if dma_contig was configured

It makes no sense to support the USERPTR memory model if the vivid instance was
configured as dma_contig. Disable it if this is the case.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: ipu3-cio2, dw9714: Remove Jian Xu's e-mail
Sakari Ailus [Mon, 7 Jan 2019 11:07:05 +0000 (06:07 -0500)]
media: ipu3-cio2, dw9714: Remove Jian Xu's e-mail

Jian Xu has left the company. Remove his e-mail address that no longer
works.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: MAINTAINERS: Update reviewers for ipu3-cio2
Sakari Ailus [Mon, 7 Jan 2019 11:06:18 +0000 (06:06 -0500)]
media: MAINTAINERS: Update reviewers for ipu3-cio2

Remove Jian Xu from the driver's reviewers.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: sun6i: Add H3 compatible
Chen-Yu Tsai [Fri, 30 Nov 2018 07:58:45 +0000 (02:58 -0500)]
media: sun6i: Add H3 compatible

The CSI controller found on the H3 (and H5) is a reduced version of the
one found on the A31. It only has 1 channel, instead of 4 channels for
time-multiplexed BT.656. Since the H3 is a reduced version, it cannot
"fallback" to a compatible that implements more features than it
supports.

Add a compatible string entry for the H3.

Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Reviewed-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dt-bindings: media: sun6i: Separate H3 compatible from A31
Chen-Yu Tsai [Fri, 30 Nov 2018 07:58:44 +0000 (02:58 -0500)]
media: dt-bindings: media: sun6i: Separate H3 compatible from A31

The CSI controller found on the H3 (and H5) is a reduced version of the
one found on the A31. It only has 1 channel, instead of 4 channels for
time-multiplexed BT.656. Since the H3 is a reduced version, it cannot
"fallback" to a compatible that implements more features than it
supports.

Split out the H3 compatible as a separate entry, with no fallback.

Fixes: b7eadaa3a02a ("media: dt-bindings: media: sun6i: Add A31 and H3 compatibles")

Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Reviewed-by: Jagan Teki <jagan@amarulasolutions.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: ov9640: Wrap long and unwrap short lines, align wrapped lines correctly
Sakari Ailus [Thu, 13 Dec 2018 20:29:34 +0000 (15:29 -0500)]
media: ov9640: Wrap long and unwrap short lines, align wrapped lines correctly

Some little style fixup work.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: ov9640: fix missing error handling in probe
Petr Cvek [Thu, 13 Dec 2018 15:39:19 +0000 (10:39 -0500)]
media: i2c: ov9640: fix missing error handling in probe

Control handlers registration lacked error path with
v4l2_ctrl_handler_free() call. Fix it by using goto to alread existing
v4l2_ctrl_handler_free() call.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: ov9640: make array of supported formats constant
Petr Cvek [Thu, 13 Dec 2018 15:39:18 +0000 (10:39 -0500)]
media: i2c: ov9640: make array of supported formats constant

An array which defines sensor's supported formats is not written anywhere,
so it can be constant.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: ov9640: add space before return for better clarity
Petr Cvek [Thu, 13 Dec 2018 15:39:17 +0000 (10:39 -0500)]
media: i2c: ov9640: add space before return for better clarity

Some returns were adjoined to unrelated code blocks. This patch adds
a space inbetween.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: ov9640: change array index or length variables to unsigned
Petr Cvek [Thu, 13 Dec 2018 15:39:16 +0000 (10:39 -0500)]
media: i2c: ov9640: change array index or length variables to unsigned

The driver uses variables to store frame resolutions and to indexing
various arrays. These should be unsigned.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: ov9640: add missing SPDX identifiers
Petr Cvek [Thu, 13 Dec 2018 15:39:15 +0000 (10:39 -0500)]
media: i2c: ov9640: add missing SPDX identifiers

Add missing SPDX identifiers to .c and .h files of the sensor driver.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: MAINTAINERS: add Petr Cvek as a maintainer for the ov9640 driver
Petr Cvek [Thu, 13 Dec 2018 15:39:14 +0000 (10:39 -0500)]
media: MAINTAINERS: add Petr Cvek as a maintainer for the ov9640 driver

The soc_camera drivers are marked as orphaned. Add Petr Cvek as a new
maintainer for ov9640 driver after its switch from the soc_camera.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: ov9640: drop soc_camera code and switch to v4l2_async
Petr Cvek [Thu, 13 Dec 2018 15:39:13 +0000 (10:39 -0500)]
media: i2c: ov9640: drop soc_camera code and switch to v4l2_async

This patch removes the dependency on an obsoleted soc_camera from ov9640
driver and changes the code to be a standalone v4l2 async subdevice.
It also adds GPIO allocations for power and reset signals (as they are not
handled by soc_camera now).

The values for waiting on GPIOs (reset and power) settling down were taken
from the datasheet (> 1 ms after HW/SW reset). The upper limit was chosen
as an arbitrary value. Also one occurrence of mdelay() was changed to
msleep(). The delays were successfully tested on a real hardware.

The patch makes ov9640 sensor again compatible with the pxa_camera driver.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: soc_camera: ov9640: move ov9640 out of soc_camera
Petr Cvek [Thu, 13 Dec 2018 15:39:12 +0000 (10:39 -0500)]
media: soc_camera: ov9640: move ov9640 out of soc_camera

Initial part of ov9640 transition from soc_camera subsystem to a standalone
v4l2 subdevice. The soc_camera version seems to be used only in Palm Zire72
and in (the future) HTC Magician. On these two devices the support is
broken as pxa_camera driver doesn't use soc_camera anymore. The other
mentions from git grep are "TODOs" (in board-osk.c) or chip names for
unsupported sensors on HW which doesn't use soc_camera at all (irelevant).

Copy the driver files from soc_camera and mark the original ones in the
Kconfig description as obsoleted.

Add config option VIDEO_OV9640 to the build files in drivers/media/i2c.

Signed-off-by: Petr Cvek <petrcvekcz@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: v4l2: i2c: ov7670: Fix PLL bypass register values
Jacopo Mondi [Fri, 29 Dec 2017 12:22:26 +0000 (07:22 -0500)]
media: v4l2: i2c: ov7670: Fix PLL bypass register values

The following commits:
commit f6dd927f34d6 ("[media] media: ov7670: calculate framerate properly for ov7675")
commit 04ee6d92047e ("[media] media: ov7670: add possibility to bypass pll for ov7675")
introduced the ability to bypass PLL multiplier and use input clock (xvclk)
as pixel clock output frequency for ov7675 sensor.

PLL is bypassed using register DBLV[7:6], according to ov7670 and ov7675
sensor manuals. Macros used to set DBLV register seem wrong in the
driver, as their values do not match what reported in the datasheet.

Fix by changing DBLV_* macros to use bits [7:6] and set bits [3:0] to
default 0x0a reserved value (according to datasheets).

While at there, remove a write to DBLV register in
"ov7675_set_framerate()" that over-writes the previous one to the same
register that takes "info->pll_bypass" flag into account instead of setting PLL
multiplier to 4x unconditionally.

And, while at there, since "info->pll_bypass" is only used in
set/get_framerate() functions used by ov7675 only, it is not necessary
to check for the device id at probe time to make sure that when using
ov7670 "info->pll_bypass" is set to false.

Fixes: f6dd927f34d6 ("[media] media: ov7670: calculate framerate properly for ov7675")

Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: ov5640: Add RAW bayer format support
Loic Poulain [Fri, 2 Nov 2018 16:38:43 +0000 (12:38 -0400)]
media: ov5640: Add RAW bayer format support

OV5640 sensor supports raw image output (bayer).
Configure ISP mux/format registers accordingly.

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dt-bindings: media: i2c: Fix i2c address for OV5645 camera sensor
Manivannan Sadhasivam [Fri, 9 Nov 2018 07:56:43 +0000 (02:56 -0500)]
media: dt-bindings: media: i2c: Fix i2c address for OV5645 camera sensor

The i2c address for the Omnivision OV5645 camera sensor is 0x3c. It is
incorrectly mentioned as 0x78 in binding. Hence fix that.

Fixes: 09c716af36e6 [media] media: i2c/ov5645: add the device tree binding document

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dt-bindings: media: i2c: Fix external clock frequency for OV5645
Manivannan Sadhasivam [Wed, 14 Nov 2018 12:13:38 +0000 (07:13 -0500)]
media: dt-bindings: media: i2c: Fix external clock frequency for OV5645

Commit "4adb0a0432f4 media: ov5645: Supported external clock is 24MHz"
modified the external clock frequency to be 24MHz instead of the
23.88MHz in driver. Hence, modify the frequency value in binding.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: ov2640: set all mbus format field when G_FMT and S_FMT ioctls
Akinobu Mita [Sun, 9 Dec 2018 05:20:33 +0000 (00:20 -0500)]
media: ov2640: set all mbus format field when G_FMT and S_FMT ioctls

This driver doesn't set all members of mbus format field when the
VIDIOC_SUBDEV_{S,G}_FMT ioctls are called.

This is detected by v4l2-compliance.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: ov2640: make VIDIOC_SUBDEV_G_FMT ioctl work with V4L2_SUBDEV_FORMAT_TRY
Akinobu Mita [Sun, 9 Dec 2018 05:20:32 +0000 (00:20 -0500)]
media: ov2640: make VIDIOC_SUBDEV_G_FMT ioctl work with V4L2_SUBDEV_FORMAT_TRY

The VIDIOC_SUBDEV_G_FMT ioctl for this driver doesn't recognize
V4L2_SUBDEV_FORMAT_TRY and always works as if V4L2_SUBDEV_FORMAT_ACTIVE
is specified.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: ov2640: set default window and format code at probe time
Akinobu Mita [Sun, 9 Dec 2018 05:20:31 +0000 (00:20 -0500)]
media: ov2640: set default window and format code at probe time

Set default window and format code at probe time instead of always checking
if they have not been set yet when VIDIOC_SUBDEV_G_FMT ioctl is called.

This change simplifies the next patch (make VIDIOC_SUBDEV_G_FMT ioctl work
with V4L2_SUBDEV_FORMAT_TRY).

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: imx274: fix wrong order in test pattern menus
Luca Ceresoli [Wed, 5 Dec 2018 16:32:21 +0000 (11:32 -0500)]
media: imx274: fix wrong order in test pattern menus

The description of test patterns 11 and 12 are swapped.

Checked against the live sensor.

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: ipu3-cio2: Allow probe to succeed if there are no sensors connected
Sakari Ailus [Fri, 30 Nov 2018 22:30:48 +0000 (17:30 -0500)]
media: ipu3-cio2: Allow probe to succeed if there are no sensors connected

The device won't be powered off on systems that have no sensors connected
unless it has a driver bound to it. Allow that to happen even if there are
no sensors connected to cio2.

Reviewed-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@linux.intel.com>
Tested-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@linux.intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: s5p-jpeg: Correct step and max values for V4L2_CID_JPEG_RESTART_INTERVAL
Pawe? Chmiel [Wed, 9 Jan 2019 18:00:41 +0000 (13:00 -0500)]
media: s5p-jpeg: Correct step and max values for V4L2_CID_JPEG_RESTART_INTERVAL

This commit corrects max and step values for v4l2 control for
V4L2_CID_JPEG_RESTART_INTERVAL. Max should be 0xffff and step should be 1.
It was found by using v4l2-compliance tool and checking result of
VIDIOC_QUERY_EXT_CTRL/QUERYMENU test.
Previously it was complaining that step was bigger than difference
between max and min.

Fixes: 15f4bc3b1f42 ("[media] s5p-jpeg: Add JPEG controls support")

Signed-off-by: Pawe? Chmiel <pawel.mikolaj.chmiel@gmail.com>
Reviewed-by: Jacek Anaszewski <jacek.anaszewski@gmail.com>
Reviewed-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: saa7146: make use of i2c_8bit_addr_from_msg
Peter Rosin [Wed, 9 Jan 2019 11:24:56 +0000 (06:24 -0500)]
media: saa7146: make use of i2c_8bit_addr_from_msg

Because it looks neater.

Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: rcar-csi2: Add support for RZ/G2E
Fabrizio Castro [Thu, 13 Dec 2018 20:24:12 +0000 (15:24 -0500)]
media: rcar-csi2: Add support for RZ/G2E

According to the RZ/G2 User's manual, RZ/G2E and R-Car E3 CSI-2
blocks are identical, therefore use R-Car E3 definitions to add
RZ/G2E support.

Signed-off-by: Fabrizio Castro <fabrizio.castro@bp.renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: rcar-vin: Add support for RZ/G2E
Fabrizio Castro [Thu, 13 Dec 2018 20:24:04 +0000 (15:24 -0500)]
media: rcar-vin: Add support for RZ/G2E

According to the RZ/G2 User's manual, RZ/G2E and R-Car E3 VIN
blocks are identical, therefore use R-Car E3 definitions to add
RZ/G2E support.

Signed-off-by: Fabrizio Castro <fabrizio.castro@bp.renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dt-bindings: rcar-vin: Add R8A774C0 support
Fabrizio Castro [Thu, 13 Dec 2018 20:21:59 +0000 (15:21 -0500)]
media: dt-bindings: rcar-vin: Add R8A774C0 support

Add the compatible string for RZ/G2E (a.k.a. R8A774C0) to the list
of SoCs supported by rcar-vin driver.

Signed-off-by: Fabrizio Castro <fabrizio.castro@bp.renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dt-bindings: rcar-csi2: Add r8a774c0
Fabrizio Castro [Thu, 13 Dec 2018 20:21:49 +0000 (15:21 -0500)]
media: dt-bindings: rcar-csi2: Add r8a774c0

Add the compatible string for RZ/G2E (a.k.a. R8A774C0) to the
list of supported SoCs.

Signed-off-by: Fabrizio Castro <fabrizio.castro@bp.renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: adv748x: configure number of lanes used for TXA CSI-2 transmitter
Niklas Söderlund [Thu, 29 Nov 2018 02:01:47 +0000 (21:01 -0500)]
media: i2c: adv748x: configure number of lanes used for TXA CSI-2 transmitter

The driver fixed the TXA CSI-2 transmitter in 4-lane mode while it could
operate using 1-, 2- and 4-lanes. Update the driver to support all
available modes.

The driver makes use of large tables of static register/value writes
when powering up/down the TXA and TXB transmitters which include the
write to the NUM_LANES register. By converting the tables into functions
and using parameters the power up/down functions for TXA and TXB power
up/down can be merged and used for both transmitters.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: adv748x: store number of CSI-2 lanes described in device tree
Niklas Söderlund [Thu, 29 Nov 2018 02:01:46 +0000 (21:01 -0500)]
media: i2c: adv748x: store number of CSI-2 lanes described in device tree

The adv748x CSI-2 transmitters TXA and TXB can use different number of
lanes to transmit data. In order to be able to configure the device
correctly this information need to be parsed from device tree and stored
in each TX private data structure.

TXA supports 1, 2 and 4 lanes while TXB supports 1 lane.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: i2c: adv748x: reuse power up sequence when initializing CSI-2
Niklas Söderlund [Thu, 29 Nov 2018 02:01:45 +0000 (21:01 -0500)]
media: i2c: adv748x: reuse power up sequence when initializing CSI-2

Extend the MIPI CSI-2 power up sequence to match the power up sequence
in the hardware manual chapter "9.5.1 Power Up Sequence". This change
allows the power up functions to be reused when initializing the
hardware reducing code duplicating as well aligning with the
documentation.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Tested-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dt-bindings: adv748x: make data-lanes property mandatory for CSI-2 endpoints
Niklas Söderlund [Thu, 29 Nov 2018 02:01:44 +0000 (21:01 -0500)]
media: dt-bindings: adv748x: make data-lanes property mandatory for CSI-2 endpoints

The CSI-2 transmitters can use a different number of lanes to transmit
data. Make the data-lanes mandatory for the endpoints that describe the
transmitters as no good default can be set to fallback on.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dvb: Add check on sp8870_readreg
Aditya Pakki [Thu, 27 Dec 2018 18:58:01 +0000 (13:58 -0500)]
media: dvb: Add check on sp8870_readreg

In sp8870_set_frontend_parameters, the function sp8870_readreg
may return an error when i2c_transfer fails. The fix checks for
this error and returns upstream consistent with other invocations.

Signed-off-by: Aditya Pakki <pakki001@umn.edu>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: dvb: add return value check on Write16
Aditya Pakki [Thu, 27 Dec 2018 18:47:20 +0000 (13:47 -0500)]
media: dvb: add return value check on Write16

Write16 can return an error code -1 when the i2c_write fails. The
fix checks for these failures and returns the error upstream

Signed-off-by: Aditya Pakki <pakki001@umn.edu>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: secocec: fix ir address shift
Ettore Chimenti [Thu, 27 Dec 2018 13:06:35 +0000 (08:06 -0500)]
media: secocec: fix ir address shift

The actual value of the RC5 System Number (address) is stored in the
IR_READ_DATA common register masked with 0x1F00 so it have to be shifted
by 8 bits.

Signed-off-by: Ettore Chimenti <ek5.chimenti@gmail.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: cxd2880-spi: fix two memory leaks of dvb_spi
Colin Ian King [Sat, 22 Dec 2018 14:12:26 +0000 (09:12 -0500)]
media: cxd2880-spi: fix two memory leaks of dvb_spi

There are two return paths that do not kfree dvb_spi. Fix the memory
leaks by returning via the exit label fail_adapter that will free
dvi_spi.

Detected by CoverityScan, CID#1475991 ("Resource Leak")

Fixes: cb496cd472af ("media: cxd2880-spi: Add optional vcc regulator")

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Yasunari Takiguchi <Yasunari.Takiguchi@sony.com>?
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: mt312: fix a missing check of mt312 reset
Kangjie Lu [Fri, 21 Dec 2018 07:07:20 +0000 (02:07 -0500)]
media: mt312: fix a missing check of mt312 reset

mt312_reset() may fail. Although it is called in the end of
mt312_set_frontend(), we better check its status and return its error
code upstream instead of 0.

Signed-off-by: Kangjie Lu <kjlu@umn.edu>
Reviewed-by: Matthias Schwarzott <zzam@gentoo.org>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: lgdt3306a: fix a missing check of return value
Kangjie Lu [Thu, 20 Dec 2018 07:48:42 +0000 (02:48 -0500)]
media: lgdt3306a: fix a missing check of return value

If lgdt3306a_read_reg() fails, the read data in "val" is incorrect, thus
shouldn't be further used. The fix inserts a check for the return value
of lgdt3306a_read_reg(). If it fails, goto fail.

Signed-off-by: Kangjie Lu <kjlu@umn.edu>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: seco-cec: fix RC_CORE dependency
Arnd Bergmann [Mon, 10 Dec 2018 20:41:40 +0000 (15:41 -0500)]
media: seco-cec: fix RC_CORE dependency

All other drivers that need RC_CORE have a dependency rather than using
'select', so we should do the same here to avoid circular dependencies
as well as this warning about missing dependencies:

WARNING: unmet direct dependencies detected for RC_CORE
  Depends on [n]: INPUT [=n]
  Selected by [y]:
  - VIDEO_SECO_RC [=y] && MEDIA_SUPPORT [=y] && CEC_PLATFORM_DRIVERS [=y] && VIDEO_SECO_CEC [=y]

Fixes: daef95769b3a ("media: seco-cec: add Consumer-IR support")

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: rcar-vin: remove unneeded locking in async callbacks
Niklas Söderlund [Fri, 4 Jan 2019 16:12:18 +0000 (11:12 -0500)]
media: rcar-vin: remove unneeded locking in async callbacks

The locking implemented in the async notifier callbacks are unnecessary
as the global list_lock in v4l2-async.c is held whenever one of the
callbacks are called.

The locking in itself is not harmful however it produces a LOCKDEP
warning between the global v4l2-async list_lock and the rcar-vin local
locking schema. Remove the rcar-vin locking for the async callbacks to
reduce complexity and silent the false LOCKDEP warning.

Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: staging: bcm2835-camera: use V4L2_FRACT_COMPARE
Akinobu Mita [Sun, 30 Dec 2018 13:20:16 +0000 (08:20 -0500)]
media: staging: bcm2835-camera: use V4L2_FRACT_COMPARE

Now the equivalent of FRACT_CMP() is added in v4l2 common internal API
header.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: tw9910: add helper function for setting gpiod value
Alexey Khoroshilov [Sun, 30 Dec 2018 11:41:41 +0000 (06:41 -0500)]
media: tw9910: add helper function for setting gpiod value

tw9910 driver tries to sleep for the same period of time
after each gpiod_set_value(). The patch moves duplicated code
to a helper function.

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: tw9910: fix failure handling in tw9910_power_on()
Alexey Khoroshilov [Sun, 30 Dec 2018 11:41:40 +0000 (06:41 -0500)]
media: tw9910: fix failure handling in tw9910_power_on()

If gpiod_get_optional() fails in tw9910_power_on(), clk is left undisabled.

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

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: s5p-jpeg: Check for fmt_ver_flag when doing fmt enumeration
Pawe? Chmiel [Sat, 29 Dec 2018 15:46:01 +0000 (10:46 -0500)]
media: s5p-jpeg: Check for fmt_ver_flag when doing fmt enumeration

Previously when doing format enumeration, it was returning all
 formats supported by driver, even if they're not supported by hw.
Add missing check for fmt_ver_flag, so it'll be fixed and only those
 supported by hw will be returned. Similar thing is already done
 in s5p_jpeg_find_format.

It was found by using v4l2-compliance tool and checking result
 of VIDIOC_ENUM_FMT/FRAMESIZES/FRAMEINTERVALS test
and using v4l2-ctl to get list of all supported formats.

Tested on s5pv210-galaxys (Samsung i9000 phone).

Fixes: bb677f3ac434 ("[media] Exynos4 JPEG codec v4l2 driver")

Signed-off-by: Pawe? Chmiel <pawel.mikolaj.chmiel@gmail.com>
Reviewed-by: Jacek Anaszewski <jacek.anaszewski@gmail.com>
[hverkuil-cisco@xs4all.nl: fix a few alignment issues]
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: mt9m111: Check write_bridge for timeout
Aditya Pakki [Fri, 28 Dec 2018 18:51:10 +0000 (13:51 -0500)]
media: gspca: mt9m111: Check write_bridge for timeout

In mt9m111_probe, m5602_write_bridge can timeout and return a negative
error value. The fix checks for this error and passes it upstream.

Signed-off-by: Aditya Pakki <pakki001@umn.edu>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: gspca: Check the return value of write_bridge for timeout
Aditya Pakki [Fri, 28 Dec 2018 18:37:36 +0000 (13:37 -0500)]
media: gspca: Check the return value of write_bridge for timeout

In po1030_probe(), m5602_write_bridge() can timeout and return an error
value. The fix checks for the return value and propagates upstream
consistent with other usb drivers.

Signed-off-by: Aditya Pakki <pakki001@umn.edu>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: usb: gspca: add a missed check for goto_low_power
Kangjie Lu [Tue, 25 Dec 2018 08:24:41 +0000 (03:24 -0500)]
media: usb: gspca: add a missed check for goto_low_power

The fix checks if goto_low_power() fails, and if so, issues an error
message.

Signed-off-by: Kangjie Lu <kjlu@umn.edu>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: usb: gspca: add a missed return-value check for do_command
Kangjie Lu [Tue, 25 Dec 2018 06:31:21 +0000 (01:31 -0500)]
media: usb: gspca: add a missed return-value check for do_command

do_command() may fail. The fix adds the missed return value of
do_command(). If it fails, returns its error code.

Signed-off-by: Kangjie Lu <kjlu@umn.edu>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: sunxi: cedrus: Fix missing error message context
Ondrej Jirman [Fri, 21 Dec 2018 16:56:41 +0000 (11:56 -0500)]
media: sunxi: cedrus: Fix missing error message context

When cedrus_hw_probe is called, v4l2_dev is not yet initialized.
Use dev_err instead.

Signed-off-by: Ondrej Jirman <megous@megous.com>
Acked-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: v4l2-ctrls.c/uvc: zero v4l2_event
Hans Verkuil [Tue, 18 Dec 2018 13:37:08 +0000 (08:37 -0500)]
media: v4l2-ctrls.c/uvc: zero v4l2_event

Control events can leak kernel memory since they do not fully zero the
event. The same code is present in both v4l2-ctrls.c and uvc_ctrl.c, so
fix both.

It appears that all other event code is properly zeroing the structure,
it's these two places.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reported-by: syzbot+4f021cf3697781dbd9fb@syzkaller.appspotmail.com
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vicodec: bugfix - replace '=' with '|='
Dafna Hirschfeld [Mon, 17 Dec 2018 17:56:17 +0000 (12:56 -0500)]
media: vicodec: bugfix - replace '=' with '|='

In the fwht_encode_frame, 'encoding = encode_plane'
should be replaced with 'encoding |= encode_plane'
so existing flags won't be overwrriten.

Signed-off-by: Dafna Hirschfeld <dafna3@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: MAINTAINERS: Change s5p-jpeg maintainer information.
Andrzej Pietrasiewicz [Thu, 13 Dec 2018 13:36:38 +0000 (08:36 -0500)]
media: MAINTAINERS: Change s5p-jpeg maintainer information.

My @samsung.com address is going to cease existing soon, so change it to
an address which can actually be used to contact me.

Adding Sylwester Nawrocki, who still has access to a wide spectrum
of Exynos-based hardware.

Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: Change Andrzej Pietrasiewicz's e-mail address
Andrzej Pietrasiewicz [Thu, 13 Dec 2018 13:31:07 +0000 (08:31 -0500)]
media: Change Andrzej Pietrasiewicz's e-mail address

My @samsung.com address is going to cease existing soon, so change it to
an address which can actually be used to contact me.

Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: rcar-csi2: Fix PHTW table values for E3/V3M
Jacopo Mondi [Mon, 10 Dec 2018 14:53:55 +0000 (09:53 -0500)]
media: rcar-csi2: Fix PHTW table values for E3/V3M

The PHTW selection algorithm implemented in rcsi2_phtw_write_mbps() checks for
lower bound of the interval used to match the desired bandwidth. Use that
in place of the currently used upper bound.

Fixes: 10c08812fe60 ("media: rcar: rcar-csi2: Update V3M/E3 PHTW tables")

Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Acked-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vsp1: Fix trivial documentation
Kieran Bingham [Fri, 7 Dec 2018 16:31:34 +0000 (11:31 -0500)]
media: vsp1: Fix trivial documentation

In the partition sizing the term 'prevents' is inappropriately
pluralized.  Simplify to 'prevent'.

Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: MAINTAINERS: added include/trace/events/pwc.h
Hans Verkuil [Wed, 12 Dec 2018 17:36:43 +0000 (12:36 -0500)]
media: MAINTAINERS: added include/trace/events/pwc.h

Added include/trace/events/pwc.h to the list of files.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: usb: pwc: Don't use coherent DMA buffers for ISO transfer
Matwey V. Kornilov [Fri, 9 Nov 2018 19:03:27 +0000 (14:03 -0500)]
media: usb: pwc: Don't use coherent DMA buffers for ISO transfer

DMA cocherency slows the transfer down on systems without hardware
coherent DMA.
Instead we use noncocherent DMA memory and explicit sync at data receive
handler.

Based on previous commit the following performance benchmarks have been
carried out. Average memcpy() data transfer rate (rate) and handler
completion time (time) have been measured when running video stream at
640x480 resolution at 10fps.

x86_64 based system (Intel Core i5-3470). This platform has hardware
coherent DMA support and proposed change doesn't make big difference here.

 * kmalloc:            rate = (2.0 +- 0.4) GBps
                       time = (5.0 +- 3.0) usec
 * usb_alloc_coherent: rate = (3.4 +- 1.2) GBps
                       time = (3.5 +- 3.0) usec

We see that the measurements agree within error ranges in this case.
So theoretically predicted performance downgrade cannot be reliably
measured here.

armv7l based system (TI AM335x BeagleBone Black @ 300MHz). This platform
has no hardware coherent DMA support. DMA coherence is implemented via
disabled page caching that slows down memcpy() due to memory controller
behaviour.

 * kmalloc:            rate =  ( 94 +- 4) MBps
                       time =  (101 +- 4) usec
 * usb_alloc_coherent: rate = (28.1 +- 0.1) MBps
                       time =  (341 +- 2) usec

Note, that quantative difference leads (this commit leads to 3.3 times
acceleration) to qualitative behavior change in this case. As it was
stated before, the video stream cannot be successfully received at AM335x
platforms with MUSB based USB host controller due to performance issues
[1].

[1] https://www.spinics.net/lists/linux-usb/msg165735.html

Signed-off-by: Matwey V. Kornilov <matwey@sai.msu.ru>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: usb: pwc: Introduce TRACE_EVENTs for pwc_isoc_handler()
Matwey V. Kornilov [Fri, 9 Nov 2018 19:03:26 +0000 (14:03 -0500)]
media: usb: pwc: Introduce TRACE_EVENTs for pwc_isoc_handler()

There were reports that PWC-based webcams don't work at some
embedded ARM platforms. [1] Isochronous transfer handler seems to
work too long leading to the issues in MUSB USB host subsystem.
Also note, that urb->giveback() handlers are still called with
disabled interrupts. In order to be able to measure performance of
PWC driver, traces are introduced in URB handler section.

[1] https://www.spinics.net/lists/linux-usb/msg165735.html

Signed-off-by: Matwey V. Kornilov <matwey@sai.msu.ru>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: s5p-mfc: fix incorrect bus assignment in virtual child device
Marek Szyprowski [Mon, 7 Jan 2019 12:04:14 +0000 (07:04 -0500)]
media: s5p-mfc: fix incorrect bus assignment in virtual child device

Virtual MFC codec's child devices must not be assigned to platform bus,
because they are allocated as raw 'struct device' and don't have the
corresponding 'platform' part. This fixes NULL pointer access revealed
recently by commit a66d972465d1 ("devres: Align data[] to
ARCH_KMALLOC_MINALIGN").

Fixes: c79667dd93b0 ("media: s5p-mfc: replace custom reserved memory handling code with generic one")

Reported-by: Paweł Chmiel <pawel.mikolaj.chmiel@gmail.com>
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Paweł Chmiel <pawel.mikolaj.chmiel@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: platform: Fix missing spin_lock_init()
Wei Yongjun [Mon, 17 Dec 2018 12:14:35 +0000 (07:14 -0500)]
media: platform: Fix missing spin_lock_init()

The driver allocates the spinlock but not initialize it.
Use spin_lock_init() on it to initialize it correctly.

This is detected by Coccinelle semantic patch.

Fixes: d2b4387f3bdf ("media: platform: Add Aspeed Video Engine driver")

Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: Documentation: staging/ipu3-imgu: Add license information
Sakari Ailus [Fri, 14 Dec 2018 21:53:43 +0000 (16:53 -0500)]
media: Documentation: staging/ipu3-imgu: Add license information

The driver documentation is under GPL v2 and the uAPI documentation under
GNU FDL 1.1+ (without invariant sections) or GPL v2.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: Documentation: staging/ipu3-imgu: Fix reference file name
Sakari Ailus [Fri, 14 Dec 2018 12:27:54 +0000 (07:27 -0500)]
media: Documentation: staging/ipu3-imgu: Fix reference file name

The intel-ipu3.h intended-to-be-uAPI header is currently under
drivers/staging/media/ipu3/include/, not include/uapi/linux.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: lgdt330x: fix lock status reporting
French, Nicholas A [Sun, 9 Dec 2018 07:11:18 +0000 (02:11 -0500)]
media: lgdt330x: fix lock status reporting

A typo in code cleanup commit db9c1007bc07 ("media: lgdt330x: do
some cleanups at status logic") broke the FE_HAS_LOCK reporting
for 3303 chips by inadvertently modifying the register mask.

The broken lock status is critial as it prevents video capture
cards from reporting signal strength, scanning for channels,
and capturing video.

Fix regression by reverting mask change.

Cc: stable@vger.kernel.org # Kernel 4.17+
Fixes: db9c1007bc07 ("media: lgdt330x: do some cleanups at status logic")
Signed-off-by: Nick French <naf@ou.edu>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Tested-by: Adam Stylinski <kungfujesus06@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: extended-controls.rst: update the mpeg2 compound controls
Hans Verkuil [Wed, 5 Dec 2018 10:30:55 +0000 (05:30 -0500)]
media: extended-controls.rst: update the mpeg2 compound controls

The layout of the compound controls has changed to fix
32/64 bit alignment issues and the use of timestamps instead of
buffer indices to refer to buffers.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: cedrus: identify buffers by timestamp
Hans Verkuil [Fri, 9 Nov 2018 09:16:21 +0000 (04:16 -0500)]
media: cedrus: identify buffers by timestamp

Use the new v4l2_m2m_buf_copy_data helper function and use
timestamps to refer to reference frames instead of using
buffer indices.

Also remove the padding fields in the structs, that's a bad
idea. Just use the right types to keep everything aligned.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Tested-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vb2: add vb2_find_timestamp()
Hans Verkuil [Wed, 24 Oct 2018 10:51:01 +0000 (06:51 -0400)]
media: vb2: add vb2_find_timestamp()

Use v4l2_timeval_to_ns instead of timeval_to_ns to ensure that
both kernelspace and userspace will use the same conversion
function.

Next add a new vb2_find_timestamp() function to find buffers
with a specific timestamp.

This function will only look at DEQUEUED and DONE buffers, i.e.
buffers that are already processed.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: videodev2.h: add v4l2_timeval_to_ns inline function
Hans Verkuil [Wed, 24 Oct 2018 10:50:34 +0000 (06:50 -0400)]
media: videodev2.h: add v4l2_timeval_to_ns inline function

We want to be able to uniquely identify buffers for stateless
codecs. The internal timestamp (a u64) as stored internally in the
kernel is a suitable candidate for that, but in struct v4l2_buffer
it is represented as a struct timeval.

Add a v4l2_timeval_to_ns() function that converts the struct timeval
into a u64 in the same way that the kernel does. This makes it possible
to use this u64 elsewhere as a unique identifier of the buffer.

Since timestamps are also copied from the output buffer to the
corresponding capture buffer(s) by M2M devices, the u64 can be
used to refer to both output and capture buffers.

The plan is that in the future we redesign struct v4l2_buffer and use
u64 for the timestamp instead of a struct timeval (which has lots of
problems with 32 vs 64 bit and y2038 layout changes), and then there
is no more need to use this function.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: buffer.rst: clean up timecode documentation
Hans Verkuil [Tue, 13 Nov 2018 08:32:25 +0000 (03:32 -0500)]
media: buffer.rst: clean up timecode documentation

V4L2_BUF_FLAG_TIMECODE is not video capture specific, so drop that
part.

The 'Timecodes' section was a bit messy, so that's cleaned up.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reviewed-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vicodec: use v4l2_m2m_buf_copy_data
Hans Verkuil [Wed, 24 Oct 2018 10:51:30 +0000 (06:51 -0400)]
media: vicodec: use v4l2_m2m_buf_copy_data

Use the new v4l2_m2m_buf_copy_data() function in vicodec.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: vim2m: use v4l2_m2m_buf_copy_data
Hans Verkuil [Wed, 24 Oct 2018 10:51:20 +0000 (06:51 -0400)]
media: vim2m: use v4l2_m2m_buf_copy_data

Use the new v4l2_m2m_buf_copy_data() function in vim2m.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agomedia: v4l2-mem2mem: add v4l2_m2m_buf_copy_data helper function
Hans Verkuil [Tue, 13 Nov 2018 08:52:18 +0000 (03:52 -0500)]
media: v4l2-mem2mem: add v4l2_m2m_buf_copy_data helper function

Memory-to-memory devices should copy various parts of
struct v4l2_buffer from the output buffer to the capture buffer.

Add a helper function that does that to simplify the driver code.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reviewed-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
Reviewed-by: Alexandre Courbot <acourbot@chromium.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agoipu3: add missing #include
Mauro Carvalho Chehab [Mon, 7 Jan 2019 18:08:30 +0000 (13:08 -0500)]
ipu3: add missing #include

Lots of warning due to non-static functions are generated because
the headers with define them were not included.

Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
5 years agoLinux 5.0-rc1 v5.0-rc1
Linus Torvalds [Mon, 7 Jan 2019 01:08:20 +0000 (17:08 -0800)]
Linux 5.0-rc1

5 years agoMerge tag 'kbuild-v4.21-3' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy...
Linus Torvalds [Mon, 7 Jan 2019 00:33:10 +0000 (16:33 -0800)]
Merge tag 'kbuild-v4.21-3' of git://git./linux/kernel/git/masahiroy/linux-kbuild

Pull more Kbuild updates from Masahiro Yamada:

 - improve boolinit.cocci and use_after_iter.cocci semantic patches

 - fix alignment for kallsyms

 - move 'asm goto' compiler test to Kconfig and clean up jump_label
   CONFIG option

 - generate asm-generic wrappers automatically if arch does not
   implement mandatory UAPI headers

 - remove redundant generic-y defines

 - misc cleanups

* tag 'kbuild-v4.21-3' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
  kconfig: rename generated .*conf-cfg to *conf-cfg
  kbuild: remove unnecessary stubs for archheader and archscripts
  kbuild: use assignment instead of define ... endef for filechk_* rules
  arch: remove redundant UAPI generic-y defines
  kbuild: generate asm-generic wrappers if mandatory headers are missing
  arch: remove stale comments "UAPI Header export list"
  riscv: remove redundant kernel-space generic-y
  kbuild: change filechk to surround the given command with { }
  kbuild: remove redundant target cleaning on failure
  kbuild: clean up rule_dtc_dt_yaml
  kbuild: remove UIMAGE_IN and UIMAGE_OUT
  jump_label: move 'asm goto' support test to Kconfig
  kallsyms: lower alignment on ARM
  scripts: coccinelle: boolinit: drop warnings on named constants
  scripts: coccinelle: check for redeclaration
  kconfig: remove unused "file" field of yylval union
  nds32: remove redundant kernel-space generic-y
  nios2: remove unneeded HAS_DMA define

5 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 7 Jan 2019 00:30:14 +0000 (16:30 -0800)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf tooling updates form Ingo Molnar:
 "A final batch of perf tooling changes: mostly fixes and small
  improvements"

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (29 commits)
  perf session: Add comment for perf_session__register_idle_thread()
  perf thread-stack: Fix thread stack processing for the idle task
  perf thread-stack: Allocate an array of thread stacks
  perf thread-stack: Factor out thread_stack__init()
  perf thread-stack: Allow for a thread stack array
  perf thread-stack: Avoid direct reference to the thread's stack
  perf thread-stack: Tidy thread_stack__bottom() usage
  perf thread-stack: Simplify some code in thread_stack__process()
  tools gpio: Allow overriding CFLAGS
  tools power turbostat: Override CFLAGS assignments and add LDFLAGS to build command
  tools thermal tmon: Allow overriding CFLAGS assignments
  tools power x86_energy_perf_policy: Override CFLAGS assignments and add LDFLAGS to build command
  perf c2c: Increase the HITM ratio limit for displayed cachelines
  perf c2c: Change the default coalesce setup
  perf trace beauty ioctl: Beautify USBDEVFS_ commands
  perf trace beauty: Export function to get the files for a thread
  perf trace: Wire up ioctl's USBDEBFS_ cmd table generator
  perf beauty ioctl: Add generator for USBDEVFS_ ioctl commands
  tools headers uapi: Grab a copy of usbdevice_fs.h
  perf trace: Store the major number for a file when storing its pathname
  ...

5 years agoChange mincore() to count "mapped" pages rather than "cached" pages
Linus Torvalds [Sun, 6 Jan 2019 01:50:59 +0000 (17:50 -0800)]
Change mincore() to count "mapped" pages rather than "cached" pages

The semantics of what "in core" means for the mincore() system call are
somewhat unclear, but Linux has always (since 2.3.52, which is when
mincore() was initially done) treated it as "page is available in page
cache" rather than "page is mapped in the mapping".

The problem with that traditional semantic is that it exposes a lot of
system cache state that it really probably shouldn't, and that users
shouldn't really even care about.

So let's try to avoid that information leak by simply changing the
semantics to be that mincore() counts actual mapped pages, not pages
that might be cheaply mapped if they were faulted (note the "might be"
part of the old semantics: being in the cache doesn't actually guarantee
that you can access them without IO anyway, since things like network
filesystems may have to revalidate the cache before use).

In many ways the old semantics were somewhat insane even aside from the
information leak issue.  From the very beginning (and that beginning is
a long time ago: 2.3.52 was released in March 2000, I think), the code
had a comment saying

  Later we can get more picky about what "in core" means precisely.

and this is that "later".  Admittedly it is much later than is really
comfortable.

NOTE! This is a real semantic change, and it is for example known to
change the output of "fincore", since that program literally does a
mmmap without populating it, and then doing "mincore()" on that mapping
that doesn't actually have any pages in it.

I'm hoping that nobody actually has any workflow that cares, and the
info leak is real.

We may have to do something different if it turns out that people have
valid reasons to want the old semantics, and if we can limit the
information leak sanely.

Cc: Kevin Easton <kevin@guarana.org>
Cc: Jiri Kosina <jikos@kernel.org>
Cc: Masatake YAMATO <yamato@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Greg KH <gregkh@linuxfoundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
5 years agoFix 'acccess_ok()' on alpha and SH
Linus Torvalds [Sun, 6 Jan 2019 19:15:04 +0000 (11:15 -0800)]
Fix 'acccess_ok()' on alpha and SH

Commit 594cc251fdd0 ("make 'user_access_begin()' do 'access_ok()'")
broke both alpha and SH booting in qemu, as noticed by Guenter Roeck.

It turns out that the bug wasn't actually in that commit itself (which
would have been surprising: it was mostly a no-op), but in how the
addition of access_ok() to the strncpy_from_user() and strnlen_user()
functions now triggered the case where those functions would test the
access of the very last byte of the user address space.

The string functions actually did that user range test before too, but
they did it manually by just comparing against user_addr_max().  But
with user_access_begin() doing the check (using "access_ok()"), it now
exposed problems in the architecture implementations of that function.

For example, on alpha, the access_ok() helper macro looked like this:

  #define __access_ok(addr, size) \
        ((get_fs().seg & (addr | size | (addr+size))) == 0)

and what it basically tests is of any of the high bits get set (the
USER_DS masking value is 0xfffffc0000000000).

And that's completely wrong for the "addr+size" check.  Because it's
off-by-one for the case where we check to the very end of the user
address space, which is exactly what the strn*_user() functions do.

Why? Because "addr+size" will be exactly the size of the address space,
so trying to access the last byte of the user address space will fail
the __access_ok() check, even though it shouldn't.  As a result, the
user string accessor functions failed consistently - because they
literally don't know how long the string is going to be, and the max
access is going to be that last byte of the user address space.

Side note: that alpha macro is buggy for another reason too - it re-uses
the arguments twice.

And SH has another version of almost the exact same bug:

  #define __addr_ok(addr) \
        ((unsigned long __force)(addr) < current_thread_info()->addr_limit.seg)

so far so good: yes, a user address must be below the limit.  But then:

  #define __access_ok(addr, size)         \
        (__addr_ok((addr) + (size)))

is wrong with the exact same off-by-one case: the case when "addr+size"
is exactly _equal_ to the limit is actually perfectly fine (think "one
byte access at the last address of the user address space")

The SH version is actually seriously buggy in another way: it doesn't
actually check for overflow, even though it did copy the _comment_ that
talks about overflow.

So it turns out that both SH and alpha actually have completely buggy
implementations of access_ok(), but they happened to work in practice
(although the SH overflow one is a serious serious security bug, not
that anybody likely cares about SH security).

This fixes the problems by using a similar macro on both alpha and SH.
It isn't trying to be clever, the end address is based on this logic:

        unsigned long __ao_end = __ao_a + __ao_b - !!__ao_b;

which basically says "add start and length, and then subtract one unless
the length was zero".  We can't subtract one for a zero length, or we'd
just hit an underflow instead.

For a lot of access_ok() users the length is a constant, so this isn't
actually as expensive as it initially looks.

Reported-and-tested-by: Guenter Roeck <linux@roeck-us.net>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>