platform/kernel/linux-starfive.git
3 years agomedia: meson-ge2d: fix rotation parameters
Neil Armstrong [Mon, 12 Apr 2021 13:48:33 +0000 (15:48 +0200)]
media: meson-ge2d: fix rotation parameters

With these settings, 90deg and 270deg rotation leads to inverted
vertical, fix them to have correct rotation.

Fixes: 59a635327ca7 ("media: meson: Add M2M driver for the Amlogic GE2D Accelerator Unit")
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: v4l2-ctrls: fix reference to freed memory
Hans Verkuil [Mon, 12 Apr 2021 11:51:23 +0000 (13:51 +0200)]
media: v4l2-ctrls: fix reference to freed memory

When controls are used together with the Request API, then for
each request a v4l2_ctrl_handler struct is allocated. This contains
the controls that can be set in a request. If a control is *not* set in
the request, then the value used in the most recent previous request
must be used, or the current value if it is not found in any outstanding
requests.

The framework tried to find such a previous request and it would set
the 'req' pointer in struct v4l2_ctrl_ref to the v4l2_ctrl_ref of the
control in such a previous request. So far, so good. However, when that
previous request was applied to the hardware, returned to userspace, and
then userspace would re-init or free that request, any 'ref' pointer in
still-queued requests would suddenly point to freed memory.

This was not noticed before since the drivers that use this expected
that each request would always have the controls set, so there was
never any need to find a control in older requests. This requirement
was relaxed, and now this bug surfaced.

It was also made worse by changeset
2fae4d6aabc8 ("media: v4l2-ctrls: v4l2_ctrl_request_complete() should always set ref->req")
which increased the chance of this happening.

The use of the 'req' pointer in v4l2_ctrl_ref was very fragile, so
drop this entirely. Instead add a valid_p_req bool to indicate that
p_req contains a valid value for this control. And if it is false,
then just use the current value of the control.

Note that VIDIOC_G_EXT_CTRLS will always return -EACCES when attempting
to get a control from a request until the request is completed. And in
that case, all controls in the request will have the control value set
(i.e. valid_p_req is true). This means that the whole 'find the most
recent previous request containing a control' idea is pointless, and
the code can be simplified considerably.

The v4l2_g_ext_ctrls_common() function was refactored a bit to make
it more understandable. It also avoids updating volatile controls
in a completed request since that was already done when the request
was completed.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Fixes: 2fae4d6aabc8 ("media: v4l2-ctrls: v4l2_ctrl_request_complete() should always set ref->req")
Fixes: 6fa6f831f095 ("media: v4l2-ctrls: add core request support")
Cc: <stable@vger.kernel.org> # for v5.9 and up
Tested-by: Alexandre Courbot <acourbot@chromium.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus : hfi: add venus image info into smem
Dikshita Agarwal [Mon, 12 Apr 2021 08:05:08 +0000 (10:05 +0200)]
media: venus : hfi: add venus image info into smem

Fill fw version info into smem to be printed as part of
soc info.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: Fix internal buffer size calculations for v6.
Dikshita Agarwal [Wed, 17 Mar 2021 08:35:28 +0000 (09:35 +0100)]
media: venus: Fix internal buffer size calculations for v6.

- Update persist buffer size for encoder to 204800.
- Update persist buffer size calculation for h264 decoder.
- h264d level 6 support needs update in internal buffer size.
  update below buffers size
  - h264 decoder colocated motion vector buffer.
  - h264 decoder VPP command buffer.
  - h265 decoder VPP command buffer.
- Update VP9_NUM_FRAME_INFO_BUF to 32.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: helpers: keep max bandwidth when mbps exceeds the supported range
Vikash Garodia [Wed, 31 Mar 2021 09:28:22 +0000 (11:28 +0200)]
media: venus: helpers: keep max bandwidth when mbps exceeds the supported range

When the video usecase have macro blocks per sec which is  more than
supported, keep the required bus bandwidth as the maximum supported.

Signed-off-by: Vikash Garodia <vgarodia@codeaurora.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: fix hw overload error log condition
Mansur Alisha Shaik [Thu, 8 Apr 2021 13:35:00 +0000 (15:35 +0200)]
media: venus: fix hw overload error log condition

In current video driver, frequency is calculated for all the
running video instances and check aganist maximum supported frequency.
If both calculated frequency and maximum supported frequency are same,
even then HW overload error is printed.
Fix this by printing error log only when frequency is greater than
maximum supported frequency.

Signed-off-by: Mansur Alisha Shaik <mansur@codeaurora.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: correct firmware name for sm8250
Dmitry Baryshkov [Fri, 9 Apr 2021 00:33:55 +0000 (02:33 +0200)]
media: venus: core: correct firmware name for sm8250

Firmware name for venus should be qcom/vpu-1.0/venus.mdt, not
qcom/sm8250/venus.mdt.

Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core,pm: fix potential infinite loop
Colin Ian King [Wed, 7 Apr 2021 13:48:31 +0000 (15:48 +0200)]
media: venus: core,pm: fix potential infinite loop

The for-loop iterates with a u8 loop counter i and compares this
with the loop upper limit of res->resets_num which is an unsigned
int type.  There is a potential infinite loop if res->resets_num
is larger than the u8 loop counter i. Fix this by making the loop
counter the same type as res->resets_num.

Addresses-Coverity: ("Infinite loop")

Fixes: 3bca43585e22 ("media: venus: core,pm: Add handling for resets")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: Fix kerneldoc warnings
Stanimir Varbanov [Wed, 7 Apr 2021 09:28:45 +0000 (11:28 +0200)]
media: venus: core: Fix kerneldoc warnings

Fix kerneldoc warnings in description of venus_core and venus_inst
structures

Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: gscpa/stv06xx: fix memory leak
Hans Verkuil [Thu, 8 Apr 2021 10:31:20 +0000 (12:31 +0200)]
media: gscpa/stv06xx: fix memory leak

For two of the supported sensors the stv06xx driver allocates memory which
is stored in sd->sensor_priv. This memory is freed on a disconnect, but if
the probe() fails, then it isn't freed and so this leaks memory.

Add a new probe_error() op that drivers can use to free any allocated
memory in case there was a probe failure.

Thanks to Pavel Skripkin <paskripkin@gmail.com> for discovering the cause
of the memory leak.

Reported-and-tested-by: syzbot+e7f4c64a4248a0340c37@syzkaller.appspotmail.com
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: cx25821: remove unused including <linux/version.h>
Tian Tao [Thu, 8 Apr 2021 06:04:13 +0000 (08:04 +0200)]
media: cx25821: remove unused including <linux/version.h>

Remove including <linux/version.h> that don't need it.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Zhiqi Song <songzhiqi1@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media/meson: remove redundant dev_err call
Muhammad Usama Anjum [Wed, 7 Apr 2021 16:09:21 +0000 (18:09 +0200)]
media: staging: media/meson: remove redundant dev_err call

devm_ioremap_resource() prints error message in itself. Remove the
dev_err call to avoid redundant error message.

Signed-off-by: Muhammad Usama Anjum <musamaanjum@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: adv7842: support 1 block EDIDs, fix clearing EDID
Hans Verkuil [Wed, 7 Apr 2021 14:16:18 +0000 (16:16 +0200)]
media: adv7842: support 1 block EDIDs, fix clearing EDID

Add support for EDIDs consisting of one EDID block.

Related to this, improve CEC physical address handling.

Clearing the EDID caused a bug since v4l2_calc_aspect_ratio() was
called with a NULL pointer.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: adv7842: configure all pads
Hans Verkuil [Wed, 7 Apr 2021 14:16:17 +0000 (16:16 +0200)]
media: adv7842: configure all pads

Only the first pad was configured, but there are 4. This causes
set_fmt to fail with -EINVAL since the passed pad is > 0.

Configure all three sink pads and the source pad.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: allegro: change kernel-doc comment blocks to normal comments
Hans Verkuil [Wed, 7 Apr 2021 07:36:06 +0000 (09:36 +0200)]
media: allegro: change kernel-doc comment blocks to normal comments

The /** at the start indicates that these are kernel-doc comments, but
really these are just regular comments.

Replace /** by /* to avoid a large number of warnings from kernel-doc.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Suggested-by: Michael Tretter <m.tretter@pengutronix.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: camss: ispif: Remove redundant dev_err call in msm_ispif_subdev_init()
Yang Yingliang [Fri, 9 Apr 2021 05:06:33 +0000 (07:06 +0200)]
media: camss: ispif: Remove redundant dev_err call in msm_ispif_subdev_init()

There is a error message within devm_ioremap_resource
already, so remove the dev_err call to avoid redundant
error message.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: i2c: rdamc21: Fix warning on u8 cast
Jacopo Mondi [Fri, 19 Mar 2021 16:41:39 +0000 (17:41 +0100)]
media: i2c: rdamc21: Fix warning on u8 cast

Sparse reports a warning on a cast to u8 of a 16 bits constant.

drivers/media/i2c/rdacm21.c:348:62: warning: cast truncates bits
from constant value (300a becomes a)

Even if the behaviour is intended, silence the sparse warning replacing
the cast with a bitwise & operation.

Fixes: a59f853b3b4b ("media: i2c: Add driver for RDACM21 camera module")
Reported-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: zoran: Rename 'He' to 'he'
Zhansaya Bagdauletkyzy [Sat, 3 Apr 2021 18:09:07 +0000 (20:09 +0200)]
media: staging: media: zoran: Rename 'He' to 'he'

Rename variable 'He' to 'he' to eliminate camelcase.
Reported by checkpatch.pl.

Signed-off-by: Zhansaya Bagdauletkyzy <zhansayabagdaulet@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: zoran: Rename 'We' to 'we'
Zhansaya Bagdauletkyzy [Sat, 3 Apr 2021 18:09:05 +0000 (20:09 +0200)]
media: staging: media: zoran: Rename 'We' to 'we'

Rename variable 'We' to 'we' to eliminate camelcase.
Reported by checkpatch.pl.

Signed-off-by: Zhansaya Bagdauletkyzy <zhansayabagdaulet@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: zoran: Rename 'VidWinHt' to 'vid_win_ht'
Zhansaya Bagdauletkyzy [Sat, 3 Apr 2021 18:09:04 +0000 (20:09 +0200)]
media: staging: media: zoran: Rename 'VidWinHt' to 'vid_win_ht'

Rename 'VidWinHt' to 'vid_win_ht' to eliminate camelcase.
Reported by checkpatch.pl.

Signed-off-by: Zhansaya Bagdauletkyzy <zhansayabagdaulet@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: zoran: Rename 'VidWinWid' to 'vid_win_wid'
Zhansaya Bagdauletkyzy [Sat, 3 Apr 2021 18:09:02 +0000 (20:09 +0200)]
media: staging: media: zoran: Rename 'VidWinWid' to 'vid_win_wid'

Rename variable 'VidWinWid' to 'vid_win_wid' to eliminate camelcase.
Reported by checkpatch.pl.

Signed-off-by: Zhansaya Bagdauletkyzy <zhansayabagdaulet@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: zoran: Rename 'DispMode' to 'disp_mode'
Zhansaya Bagdauletkyzy [Sat, 3 Apr 2021 18:09:01 +0000 (20:09 +0200)]
media: staging: media: zoran: Rename 'DispMode' to 'disp_mode'

Rename variable 'DispMode' to 'disp_mode' to eliminate camelcase.
Reported by checkpatch.pl.

Signed-off-by: Zhansaya Bagdauletkyzy <zhansayabagdaulet@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: zoran: Rename 'VEnd' to 'v_end'
Zhansaya Bagdauletkyzy [Sat, 3 Apr 2021 18:09:00 +0000 (20:09 +0200)]
media: staging: media: zoran: Rename 'VEnd' to 'v_end'

Rename variable 'VEnd' to 'v_end' to eliminate camelcase.
Reported by checkpatch.pl.

Signed-off-by: Zhansaya Bagdauletkyzy <zhansayabagdaulet@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: zoran: Rename 'HEnd' to 'h_end'
Zhansaya Bagdauletkyzy [Sat, 3 Apr 2021 18:08:59 +0000 (20:08 +0200)]
media: staging: media: zoran: Rename 'HEnd' to 'h_end'

Rename variable 'HEnd' to 'h_end' to eliminate camelcase
Reported by checkpatch.pl.

Signed-off-by: Zhansaya Bagdauletkyzy <zhansayabagdaulet@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging: media: omap4iss: align arguments with open parenthesis
Beatriz Martins de Carvalho [Thu, 1 Apr 2021 15:07:39 +0000 (17:07 +0200)]
media: staging: media: omap4iss: align arguments with open parenthesis

Cleans up checks of "Alignment should match open parenthesis"
in iss.c:96.

Signed-off-by: Beatriz Martins de Carvalho <martinsdecarvalhobeatriz@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging/intel-ipu3: Fix race condition during set_fmt
Ricardo Ribalda [Fri, 9 Apr 2021 08:41:35 +0000 (10:41 +0200)]
media: staging/intel-ipu3: Fix race condition during set_fmt

Do not modify imgu_pipe->nodes[inode].vdev_fmt.fmt.pix_mp, until the
format has been correctly validated.

Otherwise, even if we use a backup variable, there is a period of time
where imgu_pipe->nodes[inode].vdev_fmt.fmt.pix_mp might have an invalid
value that can be used by other functions.

Cc: stable@vger.kernel.org
Fixes: ad91849996f9 ("media: staging/intel-ipu3: Fix set_fmt error handling")
Reviewed-by: Tomasz Figa <tfiga@chromium.org>
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: staging/intel-ipu3: Fix memory leak in imu_fmt
Ricardo Ribalda [Mon, 15 Mar 2021 12:34:05 +0000 (13:34 +0100)]
media: staging/intel-ipu3: Fix memory leak in imu_fmt

We are losing the reference to an allocated memory if try. Change the
order of the check to avoid that.

Cc: stable@vger.kernel.org
Fixes: 6d5f26f2e045 ("media: staging/intel-ipu3-v4l: reduce kernel stack usage")
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: mxl692: remove impossible condition
Jiapeng Chong [Thu, 25 Mar 2021 08:39:56 +0000 (09:39 +0100)]
media: mxl692: remove impossible condition

Fix the following coverity warning:

This greater-than-or-equal-to-zero comparison of an unsigned value is
always true. "opcode >= 0".

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: dvb-usb: avoid -Wempty-body warnings
Arnd Bergmann [Mon, 22 Mar 2021 10:07:25 +0000 (11:07 +0100)]
media: dvb-usb: avoid -Wempty-body warnings

Building with 'make W=1' shows many warnings -Wempty-body warnings like

drivers/media/usb/dvb-usb/vp702x-fe.c: In function 'vp702x_fe_set_frontend':
drivers/media/usb/dvb-usb/vp702x-fe.c:190:46: error: suggest braces around empty body in an 'else' statement [-Werror=empty-body]
  190 |                 deb_fe("tuning succeeded.\n");
drivers/media/usb/dvb-usb/dtt200u.c: In function 'dtt200u_rc_query':
drivers/media/usb/dvb-usb/dtt200u.c:124:58: error: suggest braces around empty body in an 'if' statement [-Werror=empty-body]
  124 |                 deb_info("st->data: %*ph\n", 5, st->data);
drivers/media/usb/dvb-usb/m920x.c: In function 'm920x_rc_query':
drivers/media/usb/dvb-usb/m920x.c:207:58: error: suggest braces around empty body in an 'if' statement [-Werror=empty-body]
  207 |                 deb("Unknown rc key %02x\n", rc_state[1]);

Change the empty dprintk() macros to no_printk(), which avoids this
warning and adds format string checking.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: flexcop: avoid -Wempty-body warning
Arnd Bergmann [Mon, 22 Mar 2021 10:07:24 +0000 (11:07 +0100)]
media: flexcop: avoid -Wempty-body warning

Building with 'make W=1' shows many warnings -Wempty-body warnings like

drivers/media/common/b2c2/flexcop-misc.c: In function 'flexcop_determine_revision':
drivers/media/common/b2c2/flexcop-misc.c:35:85: error: suggest braces around empty body in an 'else' statement [-Werror=empty-body]
   35 |                 deb_info("this FlexCop has the 6 basic main hardware pid filter.\n");

drivers/media/usb/b2c2/flexcop-usb.c: In function 'flexcop_usb_process_frame':
drivers/media/usb/b2c2/flexcop-usb.c:357:79: error: suggest braces around empty body in an 'else' statement [-Werror=empty-body]
  357 |                                         deb_ts("not ts packet %*ph\n", 4, b+2);
      |                                                                               ^
drivers/media/common/b2c2/flexcop-misc.c: In function 'flexcop_determine_revision':
drivers/media/common/b2c2/flexcop-misc.c:35:85: error: suggest braces around empty body in an 'else' statement [-Werror=empty-body]
   35 |                 deb_info("this FlexCop has the 6 basic main hardware pid filter.\n");
      |                                                                                     ^

Change the empty dprintk() macros to no_printk(), which avoids this
warning and adds format string checking.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: ite-cir: probe of ITE8708 on ASUS PN50 fails
Sean Young [Thu, 8 Apr 2021 08:33:22 +0000 (10:33 +0200)]
media: ite-cir: probe of ITE8708 on ASUS PN50 fails

The Asus PN50 has 16 byte io region for the ITE8708 in its DSDT, which
causes the probe fail. So, accept larger regions.

Link: https://www.spinics.net/lists/linux-media/msg177725.html
Cc: Nikolaos Beredimas <beredim@gmail.com>
Reported-by: Michael Zimmermann <sigmaepsilon92@gmail.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: dt-bindings: media: IR: Add H616 IR compatible string
Andre Przywara [Wed, 27 Jan 2021 17:24:51 +0000 (18:24 +0100)]
media: dt-bindings: media: IR: Add H616 IR compatible string

Add the obvious compatible name to the existing IR binding, and pair
it with the existing A31 fallback compatible string, as the devices
are compatible.

On the way use enums to group all compatible devices together.

Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: rc: ir-hix5hd2: use the correct HiSilicon copyright
Hao Fang [Tue, 30 Mar 2021 06:50:00 +0000 (08:50 +0200)]
media: rc: ir-hix5hd2: use the correct HiSilicon copyright

s/Hisilicon/HiSilicon/g.
It should use capital S, according to
https://www.hisilicon.com/en/terms-of-use.

Signed-off-by: Hao Fang <fanghao11@huawei.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: rc: remove zte zx ir driver
Arnd Bergmann [Wed, 20 Jan 2021 14:25:42 +0000 (15:25 +0100)]
media: rc: remove zte zx ir driver

The zte zx platform is getting removed, so this driver is no
longer needed.

Cc: Jun Nie <jun.nie@linaro.org>
Cc: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: dvb-usb: fix memory leak in dvb_usb_adapter_init
Pavel Skripkin [Sun, 28 Mar 2021 19:32:19 +0000 (21:32 +0200)]
media: dvb-usb: fix memory leak in dvb_usb_adapter_init

syzbot reported memory leak in dvb-usb. The problem was
in invalid error handling in dvb_usb_adapter_init().

for (n = 0; n < d->props.num_adapters; n++) {
....
if ((ret = dvb_usb_adapter_stream_init(adap)) ||
(ret = dvb_usb_adapter_dvb_init(adap, adapter_nrs)) ||
(ret = dvb_usb_adapter_frontend_init(adap))) {
return ret;
}
...
d->num_adapters_initialized++;
...
}

In case of error in dvb_usb_adapter_dvb_init() or
dvb_usb_adapter_dvb_init() d->num_adapters_initialized won't be
incremented, but dvb_usb_adapter_exit() relies on it:

for (n = 0; n < d->num_adapters_initialized; n++)

So, allocated objects won't be freed.

Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Reported-by: syzbot+3c2be7424cea3b932b0e@syzkaller.appspotmail.com
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: rc: add keymaps for mecool-kii-pro/kiii-pro remotes
Christian Hewitt [Mon, 29 Mar 2021 15:12:26 +0000 (17:12 +0200)]
media: rc: add keymaps for mecool-kii-pro/kiii-pro remotes

Add keymaps and bindings for the simple IR (NEC) remotes used with
the MeCool KII-Pro and MeCool KIII-Pro Android STB devices.

Tested-by: Drazen Spio <drazsp@gmail.com>
Signed-off-by: Christian Hewitt <christianshewitt@gmail.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: dvb-usb-remote: fix dvb_usb_nec_rc_key_to_event type mismatch
Arnd Bergmann [Mon, 22 Mar 2021 16:46:00 +0000 (17:46 +0100)]
media: dvb-usb-remote: fix dvb_usb_nec_rc_key_to_event type mismatch

gcc-11 warns about the prototype not exactly matching the function
definition:

drivers/media/usb/dvb-usb/dvb-usb-remote.c:363:20: error: argument 2 of type â€˜u8[5]’ {aka â€˜unsigned char[5]’} with mismatched bound [-Werror=array-parameter=]
  363 |                 u8 keybuf[5], u32 *event, int *state)
      |                 ~~~^~~~~~~~~
In file included from drivers/media/usb/dvb-usb/dvb-usb-common.h:13,
                 from drivers/media/usb/dvb-usb/dvb-usb-remote.c:9:
drivers/media/usb/dvb-usb/dvb-usb.h:490:65: note: previously declared as â€˜u8[]’ {aka â€˜unsigned char[]’}
  490 | extern int dvb_usb_nec_rc_key_to_event(struct dvb_usb_device *, u8[], u32 *, int *);
      |                                                                 ^~~~

Fixes: 776338e121b9 ("[PATCH] dvb: Add generalized dvb-usb driver")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: dvb-frontends: Remove redundant error check on variable ret
Colin Ian King [Tue, 23 Mar 2021 14:00:47 +0000 (15:00 +0100)]
media: dvb-frontends: Remove redundant error check on variable ret

An earlier commit removed a call to lgdt3306a_spectral_inversion and
omitted to remove the error return check. The check on ret is now
redundant and can be removed.

Addresses-Coverity: ("Logically dead code")

Fixes: d4a3fa6652e3 ("media: dvb-frontends: lgdt3306a.c: remove dead code")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: cobalt: drop static for sd_fmt
Hans Verkuil [Wed, 7 Apr 2021 14:11:57 +0000 (16:11 +0200)]
media: cobalt: drop static for sd_fmt

The struct v4l2_subdev_format sd_fmt cannot be static since it can be
written back by the subdev. Just have it on the stack.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: sun8i-di: Fix runtime PM imbalance in deinterlace_start_streaming
Dinghao Liu [Wed, 7 Apr 2021 05:46:06 +0000 (07:46 +0200)]
media: sun8i-di: Fix runtime PM imbalance in deinterlace_start_streaming

pm_runtime_get_sync() will increase the runtime PM counter
even it returns an error. Thus a pairing decrement is needed
to prevent refcount leak. Fix this by replacing this API with
pm_runtime_resume_and_get(), which will not change the runtime
PM counter on error.

Signed-off-by: Dinghao Liu <dinghao.liu@zju.edu.cn>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: platform: sti: Fix runtime PM imbalance in regs_show
Dinghao Liu [Wed, 7 Apr 2021 05:43:13 +0000 (07:43 +0200)]
media: platform: sti: Fix runtime PM imbalance in regs_show

pm_runtime_get_sync() will increase the runtime PM counter
even it returns an error. Thus a pairing decrement is needed
to prevent refcount leak. Fix this by replacing this API with
pm_runtime_resume_and_get(), which will not change the runtime
PM counter on error.

Signed-off-by: Dinghao Liu <dinghao.liu@zju.edu.cn>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: imx-jpeg: Fix double free in mxc_jpeg_remove
Mirela Rabulea [Tue, 6 Apr 2021 17:34:28 +0000 (19:34 +0200)]
media: imx-jpeg: Fix double free in mxc_jpeg_remove

The video_unregister_device already calls video_device_release,
so remove video_device_release, to avoid a double free, when removing
the module. This showed up in a repeated rmmod/insmod scenario.

Signed-off-by: Mirela Rabulea <mirela.rabulea@nxp.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: i2c: adv7842: fix possible use-after-free in adv7842_remove()
Yang Yingliang [Tue, 6 Apr 2021 13:50:53 +0000 (15:50 +0200)]
media: i2c: adv7842: fix possible use-after-free in adv7842_remove()

This driver's remove path calls cancel_delayed_work(). However, that
function does not wait until the work function finishes. This means
that the callback function may still be running after the driver's
remove function has finished, which would result in a use-after-free.

Fix by calling cancel_delayed_work_sync(), which ensures that
the work is properly cancelled, no longer running, and unable
to re-schedule itself.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: i2c: tda1997: Fix possible use-after-free in tda1997x_remove()
Yang Yingliang [Tue, 6 Apr 2021 13:49:45 +0000 (15:49 +0200)]
media: i2c: tda1997: Fix possible use-after-free in tda1997x_remove()

This driver's remove path calls cancel_delayed_work(). However, that
function does not wait until the work function finishes. This means
that the callback function may still be running after the driver's
remove function has finished, which would result in a use-after-free.

Fix by calling cancel_delayed_work_sync(), which ensures that
the work is properly cancelled, no longer running, and unable
to re-schedule itself.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: i2c: adv7511-v4l2: fix possible use-after-free in adv7511_remove()
Yang Yingliang [Tue, 6 Apr 2021 13:48:12 +0000 (15:48 +0200)]
media: i2c: adv7511-v4l2: fix possible use-after-free in adv7511_remove()

This driver's remove path calls cancel_delayed_work(). However, that
function does not wait until the work function finishes. This means
that the callback function may still be running after the driver's
remove function has finished, which would result in a use-after-free.

Fix by calling cancel_delayed_work_sync(), which ensures that
the work is properly cancelled, no longer running, and unable
to re-schedule itself.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: adv7604: fix possible use-after-free in adv76xx_remove()
Yang Yingliang [Tue, 6 Apr 2021 13:42:46 +0000 (15:42 +0200)]
media: adv7604: fix possible use-after-free in adv76xx_remove()

This driver's remove path calls cancel_delayed_work(). However, that
function does not wait until the work function finishes. This means
that the callback function may still be running after the driver's
remove function has finished, which would result in a use-after-free.

Fix by calling cancel_delayed_work_sync(), which ensures that
the work is properly cancelled, no longer running, and unable
to re-schedule itself.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: tc358743: fix possible use-after-free in tc358743_remove()
Yang Yingliang [Tue, 6 Apr 2021 13:39:29 +0000 (15:39 +0200)]
media: tc358743: fix possible use-after-free in tc358743_remove()

This driver's remove path calls cancel_delayed_work(). However, that
function does not wait until the work function finishes. This means
that the callback function may still be running after the driver's
remove function has finished, which would result in a use-after-free.

Fix by calling cancel_delayed_work_sync(), which ensures that
the work is properly cancelled, no longer running, and unable
to re-schedule itself.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: don't de-reference NULL pointers at IRQ time
Mauro Carvalho Chehab [Thu, 8 Apr 2021 06:49:12 +0000 (08:49 +0200)]
media: venus: don't de-reference NULL pointers at IRQ time

Smatch is warning that:
drivers/media/platform/qcom/venus/hfi_venus.c:1100 venus_isr() warn: variable dereferenced before check 'hdev' (see line 1097)

The logic basically does:
hdev = to_hfi_priv(core);

with is translated to:
hdev = core->priv;

If the IRQ code can receive a NULL pointer for hdev, there's
a bug there, as it will first try to de-reference the pointer,
and then check if it is null.

After looking at the code, it seems that this indeed can happen:
Basically, the venus IRQ thread is started with:
devm_request_threaded_irq()
So, it will only be freed after the driver unbinds.

In order to prevent the IRQ code to work with freed data,
the logic at venus_hfi_destroy() sets core->priv to NULL,
which would make the IRQ code to ignore any pending IRQs.

There is, however a race condition, as core->priv is set
to NULL only after being freed. So, we need also to move the
core->priv = NULL to happen earlier.

Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: use NULL instead of zero for pointers
Mauro Carvalho Chehab [Thu, 8 Apr 2021 06:35:30 +0000 (08:35 +0200)]
media: venus: use NULL instead of zero for pointers

As reported by sparse:

drivers/media/platform/qcom/venus/core.c:227:41: warning: Using plain integer as NULL pointer
drivers/media/platform/qcom/venus/core.c:228:34: warning: Using plain integer as NULL pointer

Two vars are using zero instead of NULL for pointers. Not really
an issue, but using NULL makes it clearer that the init data is
expecting a pointer.

Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agoMAINTAINERS: update ovti,ov2680.yaml reference
Mauro Carvalho Chehab [Wed, 7 Apr 2021 07:36:25 +0000 (09:36 +0200)]
MAINTAINERS: update ovti,ov2680.yaml reference

The file name: Documentation/devicetree/bindings/media/i2c/ov2680.yaml
should be, instead: Documentation/devicetree/bindings/media/i2c/ovti,ov2680.yaml.

Update its cross-reference accordingly.

Fixes: 57226cd8c8bf ("media: dt-bindings: ov2680: convert bindings to yaml")
Reviewed-by: Rui Miguel Silva <rmfrfs@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agoMAINTAINERS: update imi,rdacm2x-gmsl.yaml reference
Mauro Carvalho Chehab [Wed, 7 Apr 2021 07:36:41 +0000 (09:36 +0200)]
MAINTAINERS: update imi,rdacm2x-gmsl.yaml reference

The file name: Documentation/devicetree/bindings/media/i2c/rdacm2x-gmsl.yaml
should be, instead: Documentation/devicetree/bindings/media/i2c/imi,rdacm2x-gmsl.yaml.

Update its cross-reference accordingly.

Fixes: 34009bffc1c6 ("media: i2c: Add RDACM20 driver")
Fixes: e9f817689789 ("media: dt-bindings: media: i2c: Add bindings for IMI RDACM2x")
Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: s5p-mfc: Fix kernel-doc entries in s5p_mfc_common.h
Sylwester Nawrocki [Fri, 26 Mar 2021 17:20:40 +0000 (18:20 +0100)]
media: s5p-mfc: Fix kernel-doc entries in s5p_mfc_common.h

Fixes scripts/kernel-doc warnings:

s5p_mfc_common.h:343: warning: Function parameter or member 'fw_buf' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'mem_size' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'mem_base' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'mem_bitmap' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'mem_virt' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'dma_base' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'watchdog_timer' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'fw_get_done' not described in 's5p_mfc_dev'
s5p_mfc_common.h:343: warning: Function parameter or member 'risc_on' not described in 's5p_mfc_dev'
s5p_mfc_common.h:714: warning: Function parameter or member 'src_bufs_cnt' not described in 's5p_mfc_ctx'
s5p_mfc_common.h:714: warning: Function parameter or member 'dst_bufs_cnt' not described in 's5p_mfc_ctx'
s5p_mfc_common.h:714: warning: Function parameter or member 'pb_count' not described in 's5p_mfc_ctx'
s5p_mfc_common.h:714: warning: Function parameter or member 'force_frame_type' not described in 's5p_mfc_ctx'
s5p_mfc_common.h:714: warning: Function parameter or member 'slice_mode' not described in 's5p_mfc_ctx'
s5p_mfc_common.h:714: warning: Function parameter or member 'slice_size' not described in 's5p_mfc_ctx'
s5p_mfc_common.h:714: warning: Function parameter or member 'frame_tag' not described in 's5p_mfc_ctx'
s5p_mfc_common.h:714: warning: Function parameter or member 'scratch_buf_size' not described in 's5p_mfc_ctx'

Reported-by: Hans verkuil <hverkuil@xs4all.nl>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: exynos4-is: Fix kernel-doc entries in fimc-is.h
Sylwester Nawrocki [Fri, 26 Mar 2021 17:17:25 +0000 (18:17 +0100)]
media: exynos4-is: Fix kernel-doc entries in fimc-is.h

Fixes scripts/kernel-doc warnings:

fimc-is.h:286: warning: Function parameter or member 'fw' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'memory' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'f_w' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'isp' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'sensor' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'setfile' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'ctrl_handler' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'irq' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'sensor_index' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'i2h_cmd' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'h2i_cmd' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'fd_header' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'config' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'config_index' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'is_p_region' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'is_dma_p_region' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'is_shared_region' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'af' not described in 'fimc_is'
fimc-is.h:286: warning: Function parameter or member 'debugfs_entry' not described in 'fimc_is'

The f_w field is unused so remove it.

Reported-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: adv7511-v4l2: correctly report EDIDs of 1 block
Hans Verkuil [Fri, 26 Mar 2021 09:03:32 +0000 (10:03 +0100)]
media: adv7511-v4l2: correctly report EDIDs of 1 block

If the EDID has an odd number of blocks (usually just 1, but the
same problem occurs with 3 blocks), then VIDIOC_G_EDID reported 2
(or 4) blocks. Fix this.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: adv7604: support EDIDs up to 4 blocks
Hans Verkuil [Thu, 25 Mar 2021 10:39:37 +0000 (11:39 +0100)]
media: adv7604: support EDIDs up to 4 blocks

While the adv7604/11/12 hardware supported EDIDs up to 4 blocks, the
driver didn't. This patch adds support for this. It also improves support
for EDIDs that do not have a Source Physical Address: in that case the
spa location is set to the first byte of the second block, and the
'physical address' is just the two bytes at that location. This is per
the suggestion in the adv76xx documentation for such EDIDs.

Tested with an adv7604 and adv7612.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: adv7604: fix HPD support for adv7611/12
Hans Verkuil [Wed, 24 Mar 2021 07:56:42 +0000 (08:56 +0100)]
media: adv7604: fix HPD support for adv7611/12

For the adv7604 the hotplug detect pin is controlled through gpio pins from
the SoC, but the adv7611 and adv7612 control the hotplug detect pin
themselves.

But the driver had no support for this, so the HPD was always high, even
when changing the EDID. Add proper support for this to the driver.

Tested with an adv7612.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: rkvdec: Do not require all controls to be present in every request
Daniel Almeida [Tue, 23 Mar 2021 18:57:09 +0000 (19:57 +0100)]
media: rkvdec: Do not require all controls to be present in every request

According to the v4l2 api, it is allowed to skip
setting a control if its contents haven't changed for performance
reasons: userspace should only update the controls that changed from
last frame rather then updating them all. Still some ancient code
that checks for mandatory controls has been left in this driver.

Remove it.

Fixes: cd33c830448b ("media: rkvdec: Add the rkvdec driver")
Signed-off-by: Daniel Almeida <daniel.almeida@collabora.com>
Reviewed-by: Ezequiel Garcia <ezequiel@collabora.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: VIDEO_IMX8_JPEG should depend on ARCH_MXC and not default to m
Geert Uytterhoeven [Tue, 6 Apr 2021 09:32:26 +0000 (11:32 +0200)]
media: VIDEO_IMX8_JPEG should depend on ARCH_MXC and not default to m

The i.MX8 QXP/QM integrated JPEG encoder/decoder is only present on
Freescale/NXP i.MX8 QXP and QM SoCs.  Hence add a dependency on
ARCH_MXC, to prevent asking the user about this driver when configuring
a kernel without i.MX8 support.

Drop the "default m" (which means "default y" if CONFIG_MODULES is not
enabled), as merely enabling CONFIG_COMPILE_TEST should not enable
additional code.

Fixes: 2db16c6ed72ce644 ("media: imx-jpeg: Add V4L2 driver for i.MX8 JPEG Encoder/Decoder")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: em28xx/em28xx.h: Remove duplicate struct declaration
Wan Jiabing [Tue, 6 Apr 2021 08:57:37 +0000 (10:57 +0200)]
media: em28xx/em28xx.h: Remove duplicate struct declaration

struct em28xx is declared at 219th line.
The declaration here is unnecessary. Remove it.

Signed-off-by: Wan Jiabing <wanjiabing@vivo.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: cx231xx/cx231xx.h: Remove repeated struct declaration
Wan Jiabing [Tue, 6 Apr 2021 08:56:52 +0000 (10:56 +0200)]
media: cx231xx/cx231xx.h: Remove repeated struct declaration

struct cx231xx is declared at 146th line.
The declaration here is unnecessary. Remove it.

Signed-off-by: Wan Jiabing <wanjiabing@vivo.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: camss: csid: Remove redundant dev_err call in msm_csid_subdev_init()
Yang Yingliang [Fri, 2 Apr 2021 09:32:55 +0000 (11:32 +0200)]
media: camss: csid: Remove redundant dev_err call in msm_csid_subdev_init()

There is a error message within devm_ioremap_resource
already, so remove the dev_err call to avoid redundant
error message.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: camss: csiphy: Remove redundant dev_err call in msm_csiphy_subdev_init()
Yang Yingliang [Thu, 1 Apr 2021 10:35:00 +0000 (12:35 +0200)]
media: camss: csiphy: Remove redundant dev_err call in msm_csiphy_subdev_init()

There is an error message within devm_ioremap_resource
already, so remove the dev_err call to avoid redundant
error message.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Robert Foss <robert.foss@linaro.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: ti-vpe: sc: remove redundant dev_err call in sc_create()
Yang Yingliang [Thu, 1 Apr 2021 10:30:15 +0000 (12:30 +0200)]
media: ti-vpe: sc: remove redundant dev_err call in sc_create()

There is an error message within devm_ioremap_resource
already, so remove the dev_err call to avoid redundant
error message.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: ti-vpe: csc: remove redundant dev_err call in csc_create()
Yang Yingliang [Thu, 1 Apr 2021 10:28:50 +0000 (12:28 +0200)]
media: ti-vpe: csc: remove redundant dev_err call in csc_create()

There is an error message within devm_ioremap_resource
already, so remove the dev_err call to avoid redundant
error message.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Benoit Parrot <bparrot@ti.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: imx-pxp: remove redundant dev_err call in pxp_probe()
Yang Yingliang [Thu, 1 Apr 2021 10:26:07 +0000 (12:26 +0200)]
media: imx-pxp: remove redundant dev_err call in pxp_probe()

There is an error message within devm_ioremap_resource
already, so remove the dev_err call to avoid redundant
error message.

Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.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+huawei@kernel.org>
3 years agomedia: cec/core: clarify rx-arb-lost usage message
Hans Verkuil [Wed, 31 Mar 2021 08:32:12 +0000 (10:32 +0200)]
media: cec/core: clarify rx-arb-lost usage message

The rx-arb-lost error injection command only works with <op> set to 'any'.
Explicitly say so in the usage message.

Also use [] to indicate that the <poll> argument is optional.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: imx-pxp: Remove unneeded of_match_ptr()
Fabio Estevam [Mon, 29 Mar 2021 12:21:43 +0000 (14:21 +0200)]
media: imx-pxp: Remove unneeded of_match_ptr()

i.MX is a DT-only platform, so of_match_ptr() can be safely
removed.

Remove the unneeded of_match_ptr().

Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: tw68: switch from 'pci_' to 'dma_' API
Christophe JAILLET [Sun, 28 Mar 2021 18:31:01 +0000 (20:31 +0200)]
media: tw68: switch from 'pci_' to 'dma_' API

The wrappers in include/linux/pci-dma-compat.h should go away.

The patch has been generated with the coccinelle script below and has been
hand modified to replace GFP_ with a correct flag.
It has been compile tested.

When memory is allocated in 'tw68_risc_buffer()' (tw68-risc.c) GFP_KERNEL
can be used because this function is only called from a vb2_ops buf_prepare
function.
The call chain is:
  tw68_video_qops.buf_prepare          (tw68-video.c)
    --> tw68_buf_prepare               (tw68-video.c)
      --> tw68_risc_buffer

@@
@@
-    PCI_DMA_BIDIRECTIONAL
+    DMA_BIDIRECTIONAL

@@
@@
-    PCI_DMA_TODEVICE
+    DMA_TO_DEVICE

@@
@@
-    PCI_DMA_FROMDEVICE
+    DMA_FROM_DEVICE

@@
@@
-    PCI_DMA_NONE
+    DMA_NONE

@@
expression e1, e2, e3;
@@
-    pci_alloc_consistent(e1, e2, e3)
+    dma_alloc_coherent(&e1->dev, e2, e3, GFP_)

@@
expression e1, e2, e3;
@@
-    pci_zalloc_consistent(e1, e2, e3)
+    dma_alloc_coherent(&e1->dev, e2, e3, GFP_)

@@
expression e1, e2, e3, e4;
@@
-    pci_free_consistent(e1, e2, e3, e4)
+    dma_free_coherent(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_map_single(e1, e2, e3, e4)
+    dma_map_single(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_single(e1, e2, e3, e4)
+    dma_unmap_single(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4, e5;
@@
-    pci_map_page(e1, e2, e3, e4, e5)
+    dma_map_page(&e1->dev, e2, e3, e4, e5)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_page(e1, e2, e3, e4)
+    dma_unmap_page(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_map_sg(e1, e2, e3, e4)
+    dma_map_sg(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_sg(e1, e2, e3, e4)
+    dma_unmap_sg(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_single_for_cpu(e1, e2, e3, e4)
+    dma_sync_single_for_cpu(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_single_for_device(e1, e2, e3, e4)
+    dma_sync_single_for_device(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_sg_for_cpu(e1, e2, e3, e4)
+    dma_sync_sg_for_cpu(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_sg_for_device(e1, e2, e3, e4)
+    dma_sync_sg_for_device(&e1->dev, e2, e3, e4)

@@
expression e1, e2;
@@
-    pci_dma_mapping_error(e1, e2)
+    dma_mapping_error(&e1->dev, e2)

@@
expression e1, e2;
@@
-    pci_set_dma_mask(e1, e2)
+    dma_set_mask(&e1->dev, e2)

@@
expression e1, e2;
@@
-    pci_set_consistent_dma_mask(e1, e2)
+    dma_set_coherent_mask(&e1->dev, e2)

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: tw686x: switch from 'pci_' to 'dma_' API
Christophe JAILLET [Sun, 28 Mar 2021 17:02:18 +0000 (19:02 +0200)]
media: tw686x: switch from 'pci_' to 'dma_' API

The wrappers in include/linux/pci-dma-compat.h should go away.

The patch has been generated with the coccinelle script below and has been
hand modified to replace GFP_ with a correct flag.
It has been compile tested.

When memory is allocated in 'tw686x_audio_dma_alloc()' (tw686x-audio.c)
GFP_KERNEL can be used because it is only called from a probe function
and no spinlock is taken in the between.
The call chain is:
  tw686x_probe (tw686x-core.c)
    --> tw686x_audio_init (tw686x-audio.c)
      --> tw686x_audio_dma_alloc (tw686x-audio.c)

When memory is allocated in 'tw686x_memcpy_dma_alloc()' and in
'tw686x_sg_dma_alloc()' (tw686x-video.c) GFP_KERNEL can be used because
these functions are .alloc functions from a tw686x_dma_ops structure.

@@
@@
-    PCI_DMA_BIDIRECTIONAL
+    DMA_BIDIRECTIONAL

@@
@@
-    PCI_DMA_TODEVICE
+    DMA_TO_DEVICE

@@
@@
-    PCI_DMA_FROMDEVICE
+    DMA_FROM_DEVICE

@@
@@
-    PCI_DMA_NONE
+    DMA_NONE

@@
expression e1, e2, e3;
@@
-    pci_alloc_consistent(e1, e2, e3)
+    dma_alloc_coherent(&e1->dev, e2, e3, GFP_)

@@
expression e1, e2, e3;
@@
-    pci_zalloc_consistent(e1, e2, e3)
+    dma_alloc_coherent(&e1->dev, e2, e3, GFP_)

@@
expression e1, e2, e3, e4;
@@
-    pci_free_consistent(e1, e2, e3, e4)
+    dma_free_coherent(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_map_single(e1, e2, e3, e4)
+    dma_map_single(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_single(e1, e2, e3, e4)
+    dma_unmap_single(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4, e5;
@@
-    pci_map_page(e1, e2, e3, e4, e5)
+    dma_map_page(&e1->dev, e2, e3, e4, e5)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_page(e1, e2, e3, e4)
+    dma_unmap_page(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_map_sg(e1, e2, e3, e4)
+    dma_map_sg(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_sg(e1, e2, e3, e4)
+    dma_unmap_sg(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_single_for_cpu(e1, e2, e3, e4)
+    dma_sync_single_for_cpu(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_single_for_device(e1, e2, e3, e4)
+    dma_sync_single_for_device(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_sg_for_cpu(e1, e2, e3, e4)
+    dma_sync_sg_for_cpu(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_sg_for_device(e1, e2, e3, e4)
+    dma_sync_sg_for_device(&e1->dev, e2, e3, e4)

@@
expression e1, e2;
@@
-    pci_dma_mapping_error(e1, e2)
+    dma_mapping_error(&e1->dev, e2)

@@
expression e1, e2;
@@
-    pci_set_dma_mask(e1, e2)
+    dma_set_mask(&e1->dev, e2)

@@
expression e1, e2;
@@
-    pci_set_consistent_dma_mask(e1, e2)
+    dma_set_coherent_mask(&e1->dev, e2)

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: cx23885: switch from 'pci_' to 'dma_' API
Christophe JAILLET [Sun, 28 Mar 2021 08:08:19 +0000 (10:08 +0200)]
media: cx23885: switch from 'pci_' to 'dma_' API

The wrappers in include/linux/pci-dma-compat.h should go away.

The patch has been generated with the coccinelle script below and has been
hand modified to replace GFP_ with a correct flag.
It has been compile tested.

When memory is allocated in 'cx23885_risc_buffer()' GFP_KERNEL can be used
because this function is only called from a vb2_ops buf_prepare function.
The call chain is:
  cx23885_video_qops.buf_prepare        (cx23885-video.c)
    --> buffer_prepare                  (cx23885-video.c)
      --> cx23885_risc_buffer

When memory is allocated in 'cx23885_risc_databuffer()' GFP_KERNEL can be
used because this function is only called from a function that already uses
GFP_KERNEL or from a vb2_ops buf_prepare
function.
The call chains are:
  snd_cx23885_hw_params                 (cx23885-alsa.c) --> use GFP_KERNEL
    --> cx23885_risc_databuffer

  cx23885_qops.buffer_prepare           (cx23885-417.c)
     or
  dvb_qops.buffer_prepare               (cx23885-dvb.c)
    --> cx23885_buf_prepare
      --> cx23885_risc_databuffer

When memory is allocated in 'cx23885_risc_vbibuffer()' GFP_KERNEL can be
used because this function is only called from a vb2_ops buf_prepare
function.
The call chains are:
  cx23885_vbi_qops.buffer_prepare       (cx23885-vbi.c)
    --> cx23885_risc_vbibuffer

@@
@@
-    PCI_DMA_BIDIRECTIONAL
+    DMA_BIDIRECTIONAL

@@
@@
-    PCI_DMA_TODEVICE
+    DMA_TO_DEVICE

@@
@@
-    PCI_DMA_FROMDEVICE
+    DMA_FROM_DEVICE

@@
@@
-    PCI_DMA_NONE
+    DMA_NONE

@@
expression e1, e2, e3;
@@
-    pci_alloc_consistent(e1, e2, e3)
+    dma_alloc_coherent(&e1->dev, e2, e3, GFP_)

@@
expression e1, e2, e3;
@@
-    pci_zalloc_consistent(e1, e2, e3)
+    dma_alloc_coherent(&e1->dev, e2, e3, GFP_)

@@
expression e1, e2, e3, e4;
@@
-    pci_free_consistent(e1, e2, e3, e4)
+    dma_free_coherent(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_map_single(e1, e2, e3, e4)
+    dma_map_single(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_single(e1, e2, e3, e4)
+    dma_unmap_single(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4, e5;
@@
-    pci_map_page(e1, e2, e3, e4, e5)
+    dma_map_page(&e1->dev, e2, e3, e4, e5)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_page(e1, e2, e3, e4)
+    dma_unmap_page(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_map_sg(e1, e2, e3, e4)
+    dma_map_sg(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_unmap_sg(e1, e2, e3, e4)
+    dma_unmap_sg(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_single_for_cpu(e1, e2, e3, e4)
+    dma_sync_single_for_cpu(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_single_for_device(e1, e2, e3, e4)
+    dma_sync_single_for_device(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_sg_for_cpu(e1, e2, e3, e4)
+    dma_sync_sg_for_cpu(&e1->dev, e2, e3, e4)

@@
expression e1, e2, e3, e4;
@@
-    pci_dma_sync_sg_for_device(e1, e2, e3, e4)
+    dma_sync_sg_for_device(&e1->dev, e2, e3, e4)

@@
expression e1, e2;
@@
-    pci_dma_mapping_error(e1, e2)
+    dma_mapping_error(&e1->dev, e2)

@@
expression e1, e2;
@@
-    pci_set_dma_mask(e1, e2)
+    dma_set_mask(&e1->dev, e2)

@@
expression e1, e2;
@@
-    pci_set_consistent_dma_mask(e1, e2)
+    dma_set_coherent_mask(&e1->dev, e2)

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: mtk-vcodec: Separating mtk encoder driver
Irui Wang [Thu, 25 Mar 2021 12:26:25 +0000 (13:26 +0100)]
media: mtk-vcodec: Separating mtk encoder driver

MTK H264 Encoder(VENC_SYS) and VP8 Encoder(VENC_LT_SYS) are two
independent hardware instance. They have their owner interrupt,
register mapping, and special clocks.

This patch separates them into two devices. This is a preparing
patch for adding device_link between the larbs and venc-device.
It's mainly for fixing the problem:
https://lkml.org/lkml/2019/9/3/316

Acked-by: Tiffany Lin <tiffany.lin@mediatek.com>
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Signed-off-by: Maoguang Meng <maoguang.meng@mediatek.com>
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+huawei@kernel.org>
3 years agomedia: dt-bindings: media: mtk-vcodec: Separating mtk vcodec encoder node
Irui Wang [Thu, 25 Mar 2021 12:26:23 +0000 (13:26 +0100)]
media: dt-bindings: media: mtk-vcodec: Separating mtk vcodec encoder node

Updates binding document since the avc and vp8 hardware encoder in
MT8173 are now separated. Separate "mediatek,mt8173-vcodec-enc" to
"mediatek,mt8173-vcodec-enc-vp8" and "mediatek,mt8173-vcodec-enc".

This patch is not a compatible change, but we must do this modifaction
because MediaTek IOMMU add the device_link between the smi-larb
device and venc_device, if the venc device call the pm_runtime_get_sync,
the smi-larb's pm_runtime_get_sync also be called automatically.

There is a bit of backward compatibility for avc encoder, the avc
encoder device node still has compatible "mediatek,mt8173-vcodec-enc".

Acked-by: Tiffany Lin <tiffany.lin@mediatek.com>
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Signed-off-by: Maoguang Meng <maoguang.meng@mediatek.com>
Signed-off-by: Irui Wang <irui.wang@mediatek.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: v4l2-ctrls.c: fix race condition in hdl->requests list
Hans Verkuil [Sat, 27 Mar 2021 11:27:40 +0000 (12:27 +0100)]
media: v4l2-ctrls.c: fix race condition in hdl->requests list

When a request is re-inited it will release all control handler
objects that are still in the request. It does that by unbinding
and putting all those objects. When the object is unbound the
obj->req pointer is set to NULL, and the object's unbind op is
called. When the object it put the object's release op is called
to free the memory.

For a request object that contains a control handler that means
that v4l2_ctrl_handler_free() is called in the release op.

A control handler used in a request has a pointer to the main
control handler that is created by the driver and contains the
current state of all controls. If the device is unbound (due to
rmmod or a forced unbind), then that main handler is freed, again
by calling v4l2_ctrl_handler_free(), and any outstanding request
objects that refer to that main handler have to be unbound and put
as well.

It does that by this test:

if (!hdl->req_obj.req && !list_empty(&hdl->requests)) {

I.e. the handler has no pointer to a request, so is the main
handler, and one or more request objects refer to this main
handler.

However, this test is wrong since hdl->req_obj.req is actually
NULL when re-initing a request (the object unbind will set req to
NULL), and the only reason this seemingly worked is that the
requests list is typically empty since the request's unbind op
will remove the handler from the requests list.

But if another thread is at the same time adding a new control
to a request, then there is a race condition where one thread
is removing a control handler object from the requests list and
another thread is adding one. The result is that hdl->requests
is no longer empty and the code thinks that a main handler is
being freed instead of a control handler that is part of a request.

There are two bugs here: first the test for hdl->req_obj.req: this
should be hdl->req_obj.ops since only the main control handler will
have a NULL pointer there.

The second is that adding or deleting request objects from the
requests list of the main handler isn't protected by taking the
main handler's lock.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Reported-by: John Cox <jc@kynesim.co.uk>
Fixes: 6fa6f831f095 ("media: v4l2-ctrls: add core request support")
Tested-by: John Cox <jc@kynesim.co.uk>
Reported-by: John Cox <jc@kynesim.co.uk>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: vivid: add read-only int32 control
Hans Verkuil [Tue, 18 Aug 2020 14:37:09 +0000 (16:37 +0200)]
media: vivid: add read-only int32 control

This read-only int32 control is used to test read-only controls in
combination with requests. It is set by the driver to the buffer sequence
counter module 256.

Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: vdec: Fix decoder cmd STOP issue
Stanimir Varbanov [Fri, 2 Apr 2021 10:06:48 +0000 (12:06 +0200)]
media: venus: vdec: Fix decoder cmd STOP issue

Fixes an issue when issuing a stop command to the controller, negating the
following firmware error.

"SFR message from FW: Exception: TID = Unknown IP = 0x3b7dc FA = 0x0
 cause = 0x6"

Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: helper: Decide work mode
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:47 +0000 (12:06 +0200)]
media: venus: helper: Decide work mode

Decide work mode for encoder and decoder based on different
use-cases.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi: Increase plat_buf_v6 o/p buffer count.
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:46 +0000 (12:06 +0200)]
media: venus: hfi: Increase plat_buf_v6 o/p buffer count.

Presently, minimum buffer count for output buffers for h264 &
hevc codec is kept as 8 which might not be sufficient for
few use cases so increasing the value to 18.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: helpers, hfi, vdec: Set actual plane constraints to FW
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:45 +0000 (12:06 +0200)]
media: venus: helpers, hfi, vdec: Set actual plane constraints to FW

Set actual plane alignments to FW with
HFI_PROPERTY_PARAM_UNCOMPRESSED_PLANE_ACTUAL_CONSTRAINTS_INFO to calculate
correct buffer size.

bod: Fixed fall-through error in pkt_session_set_property_6xx() switch
     Ensure setting format constraints on 6xx only

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: helpers: Add internal buffer list for v6
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:44 +0000 (12:06 +0200)]
media: venus: helpers: Add internal buffer list for v6

- Internal buffers required by v6 are different than v4,
  add new list of internal buffers for v6.
- Differentiate encoder/decoder buffers for 6xx

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: firmware: Do not toggle WRAPPER_A9SS_SW_RESET on 6xx
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:43 +0000 (12:06 +0200)]
media: venus: firmware: Do not toggle WRAPPER_A9SS_SW_RESET on 6xx

Do not toggle the WRAPPER_A9SS_SW_RESET on 6xx.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: pm: Toggle 6xx wrapper power in vcodec_control
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:42 +0000 (12:06 +0200)]
media: venus: pm: Toggle 6xx wrapper power in vcodec_control

When in vcodec_control_v4() on the 4xx path we select between one of two
vcodec cores to toggle power control bits. In 6xx there is only one core
which offsets the relevant control registers to a slightly different
address.

This patch accounts for the address offset for 6xx silicon.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Co-developed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi: Add 6xx AXI halt logic
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:41 +0000 (12:06 +0200)]
media: venus: hfi: Add 6xx AXI halt logic

This patch takes the downstream AXI halt routine and applies it when
IS_V6() is true.

bod: Converted to readl_poll_timeout()
     Removed poll timeout for LPI register, testing showed the value
     would always timeout and work, so the polling did nothing of value.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Co-developed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core,pm: Vote for min clk freq during venus boot
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:40 +0000 (12:06 +0200)]
media: venus: core,pm: Vote for min clk freq during venus boot

Vote for min clk frequency for core clks during prepare and enable clocks
at boot sequence. Without this the controller clock runs at very low value
(9.6MHz) which is not sufficient to boot venus.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: pm: Hook 6xx pm ops into 4xx pm ops
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:39 +0000 (12:06 +0200)]
media: venus: pm: Hook 6xx pm ops into 4xx pm ops

At this time there is no need to differentiate between the two, we can
reuse the 4xx pm ops callback structure for 6xx.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi, vdec: v6 Add IS_V6() to existing IS_V4() if locations
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:38 +0000 (12:06 +0200)]
media: venus: hfi, vdec: v6 Add IS_V6() to existing IS_V4() if locations

In various places in the venus codebase we have if (IS_V4()) which takes
the code down paths for 4xx silicon. This logic is broadly applicable to
6xx silicon also. In this patch we add IS_V6() to various IS_V4() decision
locations.

Co-developed-by: Dikshita Agarwal <dikshita@qti.qualcomm.com>
Signed-off-by: Dikshita Agarwal <dikshita@qti.qualcomm.com>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi: Read WRAPPER_TZ_CPU_STATUS_V6 on 6xx
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:37 +0000 (12:06 +0200)]
media: venus: hfi: Read WRAPPER_TZ_CPU_STATUS_V6 on 6xx

On 6xx we should read wrapper cpu status from the wrapper TZ register set.

Co-developed-by: Jonathan Marek <jonathan@marek.ca>
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Co-developed-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi: Add 6xx interrupt support
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:36 +0000 (12:06 +0200)]
media: venus: hfi: Add 6xx interrupt support

This commit adds 6xx specific interrupt support, some register offsets and
bitmasks differ from 4xx, which we account for in this commit.

[bod: Added commit log. Moved register definition into commit]

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Co-developed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi: Add a 6xx boot logic
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:35 +0000 (12:06 +0200)]
media: venus: hfi: Add a 6xx boot logic

This patch adds a 6xx specific boot logic. The goal is to share as much
code as possible between 3xx, 4xx and 6xx silicon.

We need to do a different write to WRAPPER_INTR_MASK with an additional
write to CPU_CS_H2XSOFTINTEN_V6 and CPU_CS_X2RPMh_V6.

The other writes are the same for 6xx and non-6xx silicon albeit at
different absolute relative locations to the base of the venus address
space.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: Hook to V6 base registers when appropriate
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:34 +0000 (12:06 +0200)]
media: venus: core: Hook to V6 base registers when appropriate

This commit points the IO base registers 6xx offsets when probing for 6xx
hardware.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: Add an io base for AON regs
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:33 +0000 (12:06 +0200)]
media: venus: core: Add an io base for AON regs

6xx silicon needs to access registers from a AON base address range.
This commit defines the necessary variable for later use.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: Add an io base for TZ wrapper regs
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:32 +0000 (12:06 +0200)]
media: venus: core: Add an io base for TZ wrapper regs

6xx silicon needs to access registers from a wrapper trust-zone base
address range.
This commit defines the necessary variable for later use.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: Add differentiator IS_V6(core)
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:31 +0000 (12:06 +0200)]
media: venus: core: Add differentiator IS_V6(core)

This commit adds the macro helper IS_V6() which will be used to
differentiate iris2/v6 silicon from previous versions.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi: Define additional 6xx registers
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:30 +0000 (12:06 +0200)]
media: venus: hfi: Define additional 6xx registers

- Add X2 RPMh registers and definitions from the downstream example.
- Add 6xx core power definitions
- Add 6xx AON definitions
- Add 6xx wrapper tz definitions
- Add 6xx wrapper interrupt definitions
- Add 6xx soft interrupt definitions
- Define wrapper LPI register offsets

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Co-developed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi: Define block offsets for V6 hardware
Dikshita Agarwal [Fri, 2 Apr 2021 10:06:29 +0000 (12:06 +0200)]
media: venus: hfi: Define block offsets for V6 hardware

This commit defines a range of new base addresses for already defined
blocks.

- CPU_BASE_V6
- CPU_CS_BASE_V6
- CPU_IC_BASE_V6
- WRAPPER_BASE_V6

The base addresses of the blocks are slightly different for 6xx but, aside
from that are layout and bit compatible.

New 6xx specific block addresses will be added in separate commits.

Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Co-developed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: hfi,pm,firmware: Convert to block relative addressing
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:28 +0000 (12:06 +0200)]
media: venus: hfi,pm,firmware: Convert to block relative addressing

An upcoming silicon change places a number of existing blocks within the
Venus at different relative offsets to the base address of IO region.

In order to handle this difference this patch changes the address offsets
of the registers to function as offsets relative to the relevant sub-block
of registers within the IO region not the base address of the IO region.

As a result of this change venus_readl() and venus_writel() are deleted.

Co-developed-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: Add io base variables for each block
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:27 +0000 (12:06 +0200)]
media: venus: core: Add io base variables for each block

New silicon means that the pre-determined offsets we have been using
in this driver no longer hold. Existing blocks of registers can exist at
different offsets relative to the IO base address.

This commit adds a routine to assign the IO base hooks a subsequent commit
will convert from absolute to relative addressing.

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core: add sm8250 DT compatible and resource data
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:26 +0000 (12:06 +0200)]
media: venus: core: add sm8250 DT compatible and resource data

Adds an sm8250 compatible binding to the venus core.

Co-developed-by: Jonathan Marek <jonathan@marek.ca>
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Co-developed-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Dikshita Agarwal <dikshita@codeaurora.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: core,pm: Add handling for resets
Stanimir Varbanov [Fri, 2 Apr 2021 10:06:25 +0000 (12:06 +0200)]
media: venus: core,pm: Add handling for resets

The Venus driver has to control two reset signals related to
gcc video_axi0 and videocc mvs0c for v6. Add it.

Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: Update v6 buffer descriptors
Bryan O'Donoghue [Fri, 2 Apr 2021 10:06:24 +0000 (12:06 +0200)]
media: venus: Update v6 buffer descriptors

Currently hfi_platform_v6.c adds a bunch of capability parameters.
Reviewing downstream we can see a number of these need to be updated.

techpack/video/msm/vidc/msm_vidc_platform.c :: kona_capabilities[]

Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
3 years agomedia: venus: venc: Add support for CLL and Mastering display controls
Stanimir Varbanov [Thu, 23 Apr 2020 08:09:21 +0000 (10:09 +0200)]
media: venus: venc: Add support for CLL and Mastering display controls

Create CLL and Mastering display colour volume v4l2 controls for
encoder, add handling of HDR10 PQ SEI packet payloads for v4.

Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>