Sakari Ailus [Fri, 26 Aug 2022 11:53:58 +0000 (13:53 +0200)]
media: v4l: subdev: Fail graciously when getting try data for NULL state
The state argument for the functions for obtaining various parts of the
state is NULL if it is called by drivers for active state. Fail graciously
in that case instead of dereferencing a NULL pointer.
Suggested-by: Bingbu Cao <bingbu.cao@intel.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Sakari Ailus [Thu, 25 Aug 2022 18:36:37 +0000 (20:36 +0200)]
media: ipu3-imgu: Fix NULL pointer dereference in active selection access
What the IMGU driver did was that it first acquired the pointers to active
and try V4L2 subdev state, and only then figured out which one to use.
The problem with that approach and a later patch (see Fixes: tag) is that
as sd_state argument to v4l2_subdev_get_try_crop() et al is NULL, there is
now an attempt to dereference that.
Fix this.
Also rewrap lines a little.
Fixes: 0d346d2a6f54 ("media: v4l2-subdev: add subdev-wide state struct") Cc: stable@vger.kernel.org # for v5.14 and later Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Bingbu Cao <bingbu.cao@intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Alexander Stein [Fri, 22 Jul 2022 07:11:31 +0000 (09:11 +0200)]
media: v4l2: Fix v4l2_i2c_subdev_set_name function documentation
The doc says the I²C device's name is used if devname is NULL, but
actually the I²C device driver's name is used.
Fixes: 0658293012af ("media: v4l: subdev: Add a function to set an I²C sub-device's name") Signed-off-by: Alexander Stein <alexander.stein@ew.tq-group.com> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
media: staging: ipu3-imgu: Fix BNR wb gain documentation
The documentation states that the BNR factor is a multiplier coded as
u3.13 and with a range of (0, 8). This is not correct, as the isp is
adding 1.0 to the gain applied, ie Pout = { Pin * (1 + Gx) }.
media: hantro: Allows luma and chroma depth to be different
Luma and chroma depth are set on different hardware registers.
Even if they aren't identical the bitstream can be compliant
to HEVC specifications and decoded by the hardware.
With this patch TSUNEQBD_A_MAIN10_Technicolor_2 conformance test
is successfully decoded.
Signed-off-by: Benjamin Gaignard <benjamin.gaignard@collabora.com> Reviewed-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
media: MAINTAINERS: rectify entry in SAA7146 VIDEO4LINUX-2 DRIVER
Commit e33fdb5a0249 ("media: saa7146: deprecate hexium_gemini/orion, mxb
and ttpci") moves some media drivers to the staging subsystem, and
unfortunately only partially adjusts the entry in MAINTAINERS.
Hence, ./scripts/get_maintainer.pl --self-test=patterns complains about a
broken reference.
As the files matching include/media/drv-intf/saa7146* are moved to
drivers/staging/media/deprecated/saa7146/common, this directory is already
covered by the existing file entry
drivers/staging/media/deprecated/saa7146/.
Repair this file reference in SAA7146 VIDEO4LINUX-2 DRIVER.
media: MAINTAINERS: adjust entry to zoran driver movement
Commit 2a0c28063de2 ("media: zoran: move to mainline") moves the zoran
driver from the staging to the media subsystem, but does not adjust the
entry in MAINTAINERS.
Hence, ./scripts/get_maintainer.pl --self-test=patterns complains about a
broken reference.
Repair this file reference in ZR36067 VIDEO FOR LINUX DRIVER.
Sun Ke [Fri, 2 Sep 2022 08:58:19 +0000 (10:58 +0200)]
media: platform: mtk-mdp3: fix PM reference leak in mdp_comp_clock_on()
mdp_comp_clock_on will increase runtime PM usage counter,
and mdp_comp_clock_off will decrease the runtime PM usage counter.
so, if mdp_comp_clock_on failed after increment runtime PM usage
counter, it should decrease it before return a error code.
pm_runtime_get_sync will increment pm usage counter even it failed.
Forgetting to putting operation will result in reference leak here.
Fix it by replacing it with pm_runtime_resume_and_get to keep usage
counter balanced.
And if failed to enable clk, add pm_runtime_put() to decrease the
runtime PM usage counter.
Fixes: 61890ccaefaf ("media: platform: mtk-mdp3: add MediaTek MDP3 driver") Signed-off-by: Sun Ke <sunke32@huawei.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Wed, 31 Aug 2022 08:56:04 +0000 (10:56 +0200)]
media: platform: mtk-mdp3: add pointer checks and use devm_kfree
Fix two errors reported by smatch:
drivers/media/platform/mediatek/mdp3/mtk-mdp3-core.c:292
mdp_probe() error: we previously assumed 'mdp' could be null
drivers/media/platform/mediatek/mdp3/mtk-mdp3-cmdq.c:460
mdp_cmdq_send() error: we previously assumed 'cmd' could be null
Also, avoid warnings reported by smatch:
drivers/media/platform/mediatek/mdp3/mtk-mdp3-comp.c:872
mdp_comp_create() warn: passing devm_ allocated variable to kfree. 'comp'
[hverkuil: fix devm_kfree call]
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans Verkuil [Tue, 30 Aug 2022 05:59:24 +0000 (07:59 +0200)]
media: dvb-frontends/drxk: initialize err to 0
Fix a compiler warning:
drivers/media/dvb-frontends/drxk_hard.c: In function 'drxk_read_ucblocks':
drivers/media/dvb-frontends/drxk_hard.c:6673:21: warning: 'err' may be used uninitialized [-Wmaybe-uninitialized]
6673 | *ucblocks = (u32) err;
| ^~~~~~~~~
drivers/media/dvb-frontends/drxk_hard.c:6663:13: note: 'err' was declared here
6663 | u16 err;
| ^~~
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Hans Verkuil [Mon, 29 Aug 2022 14:05:14 +0000 (16:05 +0200)]
media: v4l2-ctrls: drop 'elems' argument from control type ops.
The type ops for equal, init and validate have an elems argument,
but this can be taken from struct v4l2_ctrl: ctrl->elems for
equal and init, and ctrl->new_elems for validate (since you are
validating a new control value).
So drop this argument and update all callers.
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Laurent Pinchart [Tue, 23 Aug 2022 15:11:36 +0000 (17:11 +0200)]
media: rkisp1: Zero v4l2_subdev_format fields in when validating links
The local sd_fmt variable in rkisp1_capture_link_validate() has
uninitialized fields, which causes random failures when calling the
subdev .get_fmt() operation. Fix it by initializing the variable when
declaring it, which zeros all other fields.
Laurent Pinchart [Tue, 16 Aug 2022 23:53:57 +0000 (01:53 +0200)]
media: rkisp1: Configure LSC after enabling the ISP
The ISP8000Nano v18.02 (found in the i.MX8MP) requires the ISP to be
enabled (as indicated by the ISP_CTRL.ISP_ENABLE bit) to configure the
lens shading table in internal RAM. The driver currently configures all
ISP initial parameters before enabling the ISP, which causes the LSC RAM
to not be initialized properly.
To fix this, split the rkisp1_params_configure() function into a
rkisp1_params_pre_configure() and a rkisp1_params_post_configure(). The
former configures all ISP parameters but LSC, while the latter
configures LSC. To implement this, the rkisp1_params_apply_params_cfg()
function is deconstructed, with two small helpers created to deal with
the parameters buffers, which are then used in rkisp1_params_isr(),
rkisp1_params_pre_configure() and rkisp1_params_post_configure().
While this initialization ordering is only needed for the ISP8000Nano
v18.02, it doesn't affect other ISP versions negatively, and can thus be
followed unconditionally.
Laurent Pinchart [Tue, 16 Aug 2022 23:53:57 +0000 (01:53 +0200)]
media: rkisp1: Use correct macro for gradient registers
The rkisp1_lsc_config() function incorrectly uses the
RKISP1_CIF_ISP_LSC_SECT_SIZE() macro for the gradient registers. Replace
it with the correct macro, and rename it from
RKISP1_CIF_ISP_LSC_GRAD_SIZE() to RKISP1_CIF_ISP_LSC_SECT_GRAD() as the
corresponding registers store the gradients for each sector, not a size.
This doesn't cause any functional change as the two macros are defined
identically (the size and gradient registers store fields in the same
number of bits at the same positions).
Laurent Pinchart [Tue, 16 Aug 2022 23:53:57 +0000 (01:53 +0200)]
media: rkisp1: Simplify LSC x/y size and grad register macros
The LSC module x/y size and grad configuration is stored in a set of 4
indexed registers each. The rkisp1-regs.h header defines all those
registers, but only the first one in each set is used, with manual
calculation of addresses of subsequent registers. Simplifies this by
merging all 4 register macros into one that takes the index as a
parameter. No functional change intended.
Laurent Pinchart [Tue, 16 Aug 2022 23:53:57 +0000 (01:53 +0200)]
media: rkisp1: Store LSC register values in u32 variables
Use the u32 type instead of unsigned int to store register values in the
LSC configuration code, to make the variables' size more explicit. No
functional change intended.
Laurent Pinchart [Tue, 16 Aug 2022 23:53:57 +0000 (01:53 +0200)]
media: rkisp1: Clean up LSC configuration code
Clean up the LSC configuration code to improve its readability by
shortening lines, using extra local variables and renaming long
variables. No functional change intended.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Allow setting color space on resizer sink pad
The resizer doesn't deal with color spaces, so it can accept any color
space on its input, and propagates it unchanged to its output. When
operating with a Bayer input format (in pass-through mode) further
restrict the YCbCr encoding and quantization to Rec 601 and full range
respectively, as for raw data the former ought to be ignored and the
latter is always full range.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Initialize color space on resizer sink and source pads
Initialize the four color space fields on the sink and source video pads
of the resizer in the .init_cfg() operation. The resizer can't perform
any color space conversion, so set the sink and source color spaces to
the same defaults, which match the ISP source video pad default.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Configure CSM based on YCbCr encoding
The driver currently only implements the Rec. 601 YCbCr encoding, extend
it with support for the other encodings defined by V4L2 (Rec. 709, Rec.
2020 and SMPTE240m). The coefficients have been calculated by rounding
the floating point values to the nearest Q1.7 fixed-point value,
adjusting the rounding to ensure that the sum of each line in the matrix
is preserved to avoid overflows.
At the hardware level, the RGB to YUV conversion matrix is fully
configurable, custom encoding could be supported by extending the ISP
parameters if desired.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Don't pass the quantization to rkisp1_csm_config()
The rkisp1_csm_config() function takes a pointer to the rkisp1_params
structure which contains the quantization value. There's no need to pass
it separately to the function. Drop it from the function parameters.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Configure quantization using ISP source pad
The rkisp1_config_isp() function uses the format on the sink pad of the
ISP to configure quantization at the output of the ISP. This is
incorrect, as hinted by the src_frm variable name that stores the
format. Fix it by using the source pad.
Laurent Pinchart [Fri, 12 Aug 2022 22:56:22 +0000 (00:56 +0200)]
media: rkisp1: Allow setting all color space fields on ISP source pad
The ISP output color space is configured through the ISP source pad. At
the moment, only the quantization can be set. Extend it to the three
other color space fields:
- The ycbcr_enc field will be used to configure the RGB to YUV matrix
(currently hardcoded to Rec. 601).
- The colorspace (which controls the color primaries) and xfer_func
fields will not be used to configure the ISP, as the corresponding
hardware blocks (the cross-talk RGB to RGB matrix and the tone mapping
curve) are programmed directly by applications through ISP parameters.
Nonetheless, those two fields should be set by applications to match
the ISP configuration, in order to propagate the correct color space
down the pipeline up to the capture video nodes.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Fix source pad format configuration
The ISP converts Bayer data to YUV when operating normally, and can also
operate in pass-through mode where the input and output formats must
match. Converting from YUV to Bayer isn't possible. If such an invalid
configuration is attempted, adjust it by copying the sink pad media bus
code to the source pad.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Allow setting color space on ISP sink pad
The ISP accepts different color spaces on its input: for YUV input, it
doesn't set any restrictions, and for Bayer inputs, any color primaries
or transfer function can be accepted (YCbCr encoding isn't applicable
there, and quantization range can only be full).
Allow setting a color space on the ISP sink pad, with the aforementioned
restrictions. The settings don't influence hardware yet (only the YUV
quantization range will, anything else has no direct effect on the ISP
configuration), but can already be set to allow color space information
to be coherent across the ISP sink link.
Laurent Pinchart [Fri, 12 Aug 2022 22:44:14 +0000 (00:44 +0200)]
media: rkisp1: Initialize color space on ISP sink and source pads
Initialize the four color space fields on the sink and source video pads
of the ISP in the .init_cfg() operation. As the main use case for the
ISP is to convert Bayer data to YUV, select a raw color space on the
sink pad and a limited range quantization of SYCC on the source pad by
default.
media: rockchip: rkisp1: Mask invalid bits in DPCC parameters
Restrict the DPCC configuration that can be set by userspace to valid
register bits. To do so, reorganize the related register macros to
define valid bitmasks, as well as bits of the DPCC mode register.
media: rockchip: rkisp1: Set DPCC methods enable bits inside loop
The rkisp1_dpcc_config() function looks over methods sets to configure
them, but sets the RKISP1_CIF_ISP_DPCC_METHODS_SET_* registers outside
of the loop with hand-unrolled code. Move this to the loop to simplify
the code.
Sean Young [Fri, 2 Sep 2022 10:32:21 +0000 (12:32 +0200)]
media: mceusb: set timeout to at least timeout provided
By rounding down, the actual timeout can be lower than requested. As a
result, long spaces just below the requested timeout can be incorrectly
reported as timeout and truncated.
Fixes: 877f1a7cee3f ("media: rc: mceusb: allow the timeout to be configurable") Cc: stable@vger.kernel.org Signed-off-by: Sean Young <sean@mess.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
The reason for the warning is clear enough; the driver sends an
unusual read request on endpoint 0 but does not set the USB_DIR_IN bit
in the bRequestType field.
More importantly, the whole situation can be avoided and the driver
simplified by converting it over to the relatively new
usb_control_msg_recv() and usb_control_msg_send() routines. That's
what this fix does.
media: remove reference to CONFIG_EMBEDDED in MEDIA_SUPPORT_FILTER
The config EMBEDDED selects EXPERT, i.e., when EMBEDDED is enabled, EXPERT
is usually also enabled. Hence, it sufficient to have the option
MEDIA_SUPPORT_FILTER set to y if !EXPERT.
This way, MEDIA_SUPPORT_FILTER does not refer to CONFIG_EMBEDDED anymore
and allows us to remove CONFIG_EMBEDDED in the close future.
Remove the reference to CONFIG_EMBEDDED in MEDIA_SUPPORT_FILTER.
This patch adds driver for MediaTek's Media Data Path ver.3 (MDP3).
It provides the following functions:
color transform, format conversion, resize, crop, rotate, flip
and additional image quality enhancement.
The MDP3 driver is mainly used for Google Chromebook products to
import the new architecture to set the HW settings as shown below:
User -> V4L2 framework
-> MDP3 driver -> SCP (setting calculations)
-> MDP3 driver -> CMDQ (GCE driver) -> HW
Each modules' related operation control is sited in mtk-mdp3-comp.c
Each modules' register table is defined in file with "mdp_reg_" prefix
GCE related API, operation control sited in mtk-mdp3-cmdq.c
V4L2 m2m device functions are implemented in mtk-mdp3-m2m.c
Probe, power, suspend/resume, system level functions are defined in
mtk-mdp3-core.c
[hverkuil: add 'depends on REMOTEPROC']
Signed-off-by: Ping-Hsun Wu <ping-hsun.wu@mediatek.com> Signed-off-by: daoyuan huang <daoyuan.huang@mediatek.com> Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Tue, 23 Aug 2022 02:38:01 +0000 (04:38 +0200)]
media: dt-binding: mediatek: add bindings for MediaTek CCORR and WDMA
This patch adds DT binding documentation for MediaTek's CCORR and
WDMA components.
These components exist in both MediaTek's Media Data Path 3(MDP3) and DRM,
and the bindings are placed under the folder "./soc/mediatek" to prevent
duplicate builds.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Moudy Ho [Tue, 23 Aug 2022 02:38:00 +0000 (04:38 +0200)]
media: dt-binding: mediatek: add bindings for MediaTek MDP3 components
This patch adds DT binding documents for Media Data Path 3 (MDP3)
a unit in multimedia system combined with several components and
used for scaling and color format convert.
Signed-off-by: Moudy Ho <moudy.ho@mediatek.com> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Miaoqian Lin [Wed, 1 Jun 2022 04:25:14 +0000 (06:25 +0200)]
media: xilinx: vipp: Fix refcount leak in xvip_graph_dma_init
of_get_child_by_name() returns a node pointer with refcount
incremented, we should use of_node_put() on it when not need anymore.
Add missing of_node_put() to avoid refcount leak.
Fixes: df3305156f98 ("[media] v4l: xilinx: Add Xilinx Video IP core") Signed-off-by: Miaoqian Lin <linmq006@gmail.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
media: staging: media: imx: imx7-media-csi: Increase video mem limit
Some high resolution (like Sony IMX492 47Mpix) sensors requre large amount space
for buffering. 64MB is far from sufficient so this patch increases the limit to
512MB.
media: uvcvideo: Use indexed loops in uvc_ctrl_init_ctrl()
As shown by the bug introduced in commit 86f7ef773156 ("media: uvcvideo:
Add support for per-device control mapping overrides"), the loop style
used by uvc_ctrl_init_ctrl() is error-prone. Rewrite the loops to use
indices instead.
Biju Das [Sun, 28 Aug 2022 08:13:34 +0000 (10:13 +0200)]
media: renesas: vsp1: Add support for RZ/G2L VSPD
The RZ/G2L VSPD provides a single VSPD instance. It has the following
sub modules MAU, CTU, RPF, DPR, LUT, BRS, WPF and LIF.
The VSPD block on RZ/G2L SoCs does not have a version register, so
added a new compatible string "renesas,r9a07g044-vsp2" with a data
pointer containing the info structure. Also the reset line is shared
with the DU module.
Biju Das [Sun, 28 Aug 2022 08:13:33 +0000 (10:13 +0200)]
media: renesas: vsp1: Add VSP1_HAS_NON_ZERO_LBA feature bit
As per HW manual V3M and RZ/G2L SoCs has nonzero LIF buffer
attributes. So, introduce a feature bit for handling the same.
This patch also adds separate device info structure for V3M and V3H
SoCs, as both these SoCs share the same model ID, but V3H does not
have VSP1_HAS_NON_ZERO_LBA feature bit.
Biju Das [Sun, 28 Aug 2022 08:13:31 +0000 (10:13 +0200)]
media: renesas: vsp1: Add support to deassert/assert reset line
As the resets DT property is mandatory, and is present in all .dtsi
in mainline, add support to perform deassert/assert using reference
counted reset handle.
Document VSPD found in RZ/G2L SoC. VSPD block is similar to VSP2-D
found on R-Car SoC's, but it does not have a version register and
it has 3 clocks compared to 1 clock on vsp1 and vsp2.
This patch introduces a new compatible 'renesas,r9a07g044-vsp2' to
handle these differences.
Ming Qian [Thu, 18 Aug 2022 03:18:21 +0000 (05:18 +0200)]
media: amphion: fix a bug that vpu core may not resume after suspend
driver will enable the vpu core when request the first instance
on the core.
one vpu core can only support 8 streaming instances in the same
time, the instance won't be added to core's list before streamon.
so the actual instance count may be greater then the number in
the core's list.
in pm resume callback, driver will resume the core immediately if
core's list is not empty.
but this check is not accurate,
if suspend during one instance is requested, but not streamon,
then after suspend, the core won't be resume, and led to instance failure.
use the request_count instead of the core's list to check
whether is the core needed to resume immediately after suspend.
And it can make the pm suspend and resume callback more clear.
Ming Qian [Tue, 26 Jul 2022 03:02:29 +0000 (05:02 +0200)]
media: amphion: don't change the colorspace reported by decoder.
decoder will report the colorspace information
which is parsed from the sequence header,
if they are unspecified, just let application to determine it,
don't change it in driver.
Ming Qian [Fri, 15 Jul 2022 07:38:00 +0000 (09:38 +0200)]
media: amphion: adjust the encoder's value range of gop size
adjust the value range of gop size from [0, 65535] to [1, 8000].
when the gop size is set to a too large value,
it may affect the encoded picture quality.
so constrain it to a reasonable range.
Ming Qian [Fri, 15 Jul 2022 07:15:49 +0000 (09:15 +0200)]
media: amphion: insert picture startcode after seek for vc1g format
For format vc1, the amphion vpu requires driver to
help insert some custom startcode before sequence and frame.
the startcode is different for vc1l and vc1g format.
But the sequence startcode is only needed at the beginning,
and it's not expected after seek.
driver need to treat the codec header and the first frame after seek
as a normal frame, and insert picture startcode for it.
In previous patch, I just fix it for vc1l format,
and should fix the similar issue for vc1g too.
Fixes: e670f5d672ef (media: amphion: only insert the first sequence startcode for vc1l format) Signed-off-by: Ming Qian <ming.qian@nxp.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Dmitry Osipenko [Thu, 18 Aug 2022 20:33:08 +0000 (22:33 +0200)]
media: cedrus: Fix endless loop in cedrus_h265_skip_bits()
The busy status bit may never de-assert if number of programmed skip
bits is incorrect, resulting in a kernel hang because the bit is polled
endlessly in the code. Fix it by adding timeout for the bit-polling.
This problem is reproducible by setting the data_bit_offset field of
the HEVC slice params to a wrong value by userspace.
Cc: stable@vger.kernel.org Fixes: 7678c5462680 (media: cedrus: Fix decoding for some HEVC videos) Reported-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Signed-off-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Dmitry Osipenko [Thu, 18 Aug 2022 20:33:07 +0000 (22:33 +0200)]
media: cedrus: Set the platform driver data earlier
The cedrus_hw_resume() crashes with NULL deference on driver probe if
runtime PM is disabled because it uses platform data that hasn't been
set up yet. Fix this by setting the platform data earlier during probe.
Cc: stable@vger.kernel.org Fixes: 50e761516f2b (media: platform: Add Cedrus VPU decoder driver) Signed-off-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Reviewed-by: Samuel Holland <samuel@sholland.org> Acked-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Nicolas Dufresne [Thu, 18 Aug 2022 20:33:06 +0000 (22:33 +0200)]
media: cedrus: Fix watchdog race condition
The watchdog needs to be scheduled before we trigger the decode
operation, otherwise there is a risk that the decoder IRQ will be
called before we have schedule the watchdog. As a side effect, the
watchdog would never be cancelled and its function would be called
at an inappropriate time.
This was observed while running Fluster with GStreamer as a backend.
Some programming error would cause the decoder IRQ to be call very
quickly after the trigger. Later calls into the driver would deadlock
due to the unbalanced state.
Cc: stable@vger.kernel.org Fixes: 7c38a551bda1 ("media: cedrus: Add watchdog for job completion") Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Reviewed-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
media: mediatek: vcodec: add decoder compatible to support mt8188
1: add mt8188's compatible name: mediatek,mt8188-vcodec-dec.
2: mt8188 is lat single core architecture, using mtk_lat_sig_core_pdata to
initialize private data.
3: Getting mt8188's chip name according to decoder compatible name.
Signed-off-by: Yunfei Dong <yunfei.dong@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Irui Wang [Wed, 20 Jul 2022 08:57:30 +0000 (10:57 +0200)]
media: mediatek: vcodec: Fix bitstream crop information error
Usually, the real bitstream width and height will set to driver
by vidioc_s_fmt, and vidioc_try_fmt() does align to get the
buffer width and height, driver calculate the encoded bitstream
crop information through them. The aligned resolution will be set
as real resolution now if user didn't set crop info by
V4L2_SEL_TGT_CROP, and the encoded bitstream may exist green line
because of crop information error.
Irui Wang [Wed, 20 Jul 2022 08:57:29 +0000 (10:57 +0200)]
media: mediatek: vcodec: Remove encoder driver get IRQ resource
The "platform_get_resource(pdev, IORESOURCE_IRQ, 0)" is no longer
used after commit a1a2b7125e107("of/platform: Drop static setup of
IRQ resource from DT core"), so just remove the function in
encoder driver to avoid driver probe failed.
Signed-off-by: Irui Wang <irui.wang@mediatek.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Irui Wang [Wed, 20 Jul 2022 08:57:26 +0000 (10:57 +0200)]
media: mediatek: vcodec: Add encoder driver support for 34-bit iova
Encoder driver got iova from IOMMU is 34-bit, for example:
Here is the sample code:
encoder input frame buffer dma address is:
frm_buf =
vb2_dma_contig_plane_dma_addr(&vb2_v4l2_buffer->vb2_buf, 0);
the value of frm_buf is 0x1_ff30_0000.
encoder driver got the frm_buf and send the iova to SCP firmware
through SCP IPI message, then write to encoder hardware in SCP.
The iova is stored in IPI message as uint32_t data type, so the
value will be truncated from *0x1_ff30_0000* to *0xff30_0000*,
and then *0xff30_0000* will be written to encoder hardware, but
IOMMU will help to add the high *0x1_* bit back, so IOMMU can
translate the iova to PA correctly, encoder hardware can access
the correct memory for encoding.
Another reason to do this is the encoder hardware can't access
the 34-bit iova, IOMMU will help to add the remaining high bits
of iova. But for mt8188, encoder hardware can access 34-bit iova
directly, and encoder driver need write all 34 bits because
IOMMU can't help driver do this if the hardware support access
34-bit iova.
For the reasons above, this patch is added to support transfer
34-bit iova between kernel and SCP encoder driver. Use uint64_t
data type to store the iova, for compatibility with old chipsets,
add some new struct definitions for 34-bit.
[hverkuil: initialize wb and wb_34 to NULL]
Signed-off-by: Irui Wang <irui.wang@mediatek.com> Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Tomi Valkeinen [Tue, 26 Apr 2022 07:06:18 +0000 (09:06 +0200)]
media: ti: cal: fix wdma irq for metadata
CAL HW interrupts are inherently racy. If we get both start and end
interrupts, we don't know what has happened: did the DMA for a single
frame start and end, or did one frame end and a new frame start?
Usually for normal pixel frames we get the interrupts separately. If
we do get both, we have to guess. The assumption in the code is that the
active vertical area is larger than the blanking vertical area, and thus
it is more likely that we get the end of the old frame and the start of
a new frame.
However, for embedded data, which is only a few lines high, we always
get both interrupts. Here the assumption is that we get both for the
same frame.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Tomi Valkeinen [Tue, 26 Apr 2022 07:06:17 +0000 (09:06 +0200)]
media: ti: cal: combine wdma irq handling
Instead of handling the WDMA START and END interrupts separately, we
need to handle both at the same time to better manage the inherent race
conditions related to CAL interrupts.
Change the code so that we have a single function,
cal_irq_handle_wdma(), which gets two booleans, start and end, as
parameters, which allows us to manage the race conditions in the
following patch.
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab@kernel.org>
projects / thirdparty / kernel / linux.git / log
