When si2165_init fails, the clk values in state are still at zero.
But the dvb-core ignores the return value of init will call tune
afterwards.
This will trigger a division by zero when tuning.
At least check for the variables to be non-zero before dividing.
This happened for a system with WinTV HVR-4400 PCIe-card after suspend-to-disk.
Do suspend-to-disk without accessing the DVB device before.
After wakeup try to tune.
si2165_init fails at checking the chip_mode and aborts.
Then si2165_set_if_freq_shift will fail with div-by-zero.
inside lgdt3306a_search we reading demod status 20 times with 50 msec sleep after each read.
This gives us more than 1 sec of delay. Removing this delay should not affect demod functionality.
Signed-off-by: Abylay Ospan <aospan@netup.ru> Acked-by: Michael Ira Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The workqueue "_irq_audio_queues" runs the audio upstream handler.
It has a single work item(&dev->_audio_work_entry) and hence doesn't
require ordering. Also, it is not being used on a memory reclaim path.
Hence, the singlethreaded workqueue has been replaced with the use of
system_wq.
System workqueues have been able to handle high level of concurrency
for a long time now and hence it's not required to have a singlethreaded
workqueue just to gain concurrency. Unlike a dedicated per-cpu workqueue
created with create_singlethread_workqueue(), system_wq allows multiple
work items to overlap executions even on the same CPU; however, a
per-cpu workqueue doesn't have any CPU locality or global ordering
guarantee unless the target CPU is explicitly specified and thus the
increase of local concurrency shouldn't make any difference.
The work item has been flushed in cx25821_stop_upstream_audio() to ensure
that nothing is pending when the driver is disconnected.
Workqueues shouldn't be freed. destroy_workqueue should be used instead.
destroy_workqueue safely destroys a workqueue and ensures that all pending
work items are done before destroying the workqueue.
[media] dvb-usb: avoid link error with dib3000m{b,c|
Tha ARM randconfig builds came up with another rare build failure
for the dib3000mc driver, when dvb-usb-dibusb-mb is built-in and
dib3000mc is a loadable module:
Apparently this used to be a valid configuration (build-time, not
run-time), but broke as part of a cleanup.
I tried reverting the cleanup, but saw that the code was still wrong
then. This version adds a dependency for dib3000mb, to ensure that
dib3000mb does not force the dibusb_dib3000mc_frontend_attach function
to be built-in when dib3000mc is a loadable module.
I have also checked the two other files that were changed in the original
cleanup, and found them to be correct in either version, so I do not
touch that part.
As this is a rather obscure bug, there is no need for backports.
Tha ARM randconfig builds came up with another rare build failure
for the dib3000mc driver, when dvb-usb-dibusb-mb is built-in and
dib3000mc is a loadable module:
Apparently this used to be a valid configuration (build-time, not
run-time), but broke as part of a cleanup.
I tried reverting the cleanup, but saw that the code was still wrong
then. This tries to fix the code properly, by moving the problematic
functions into a new file that now is built as a loadable module or
built-in, whichever is correct for a particular configuration. It fixes
the regression as well as the runtime problem that already existed.
I have also checked the two other files that were changed in the original
cleanup, and found them to be correct in either version, so I do not
touch that part.
As this is a rather obscure bug, there is no need for backports.
Sean Young [Wed, 21 Sep 2016 09:54:19 +0000 (06:54 -0300)]
[media] rc: split nec protocol into its three variants
Currently we do not know what variant (bit length) of the nec protocol
is used, other than from guessing from the length of the scancode. Now
nec will be handled the same way as the sony protocol or the rc6 protocol;
one variant per bit length.
In the future we might want to expose the rc protocol type to userspace
and we don't want to be introducing this world of pain into userspace
too.
Signed-off-by: Sean Young <sean@mess.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Heiner Kallweit [Tue, 2 Aug 2016 05:45:30 +0000 (02:45 -0300)]
[media] media: rc: nuvoton: remove unneeded call to ir_raw_event_handle
ir_raw_event_handle is called anyway after the hw fifo content stored
in nvt->buf[] has been written to the kfifo. There is not really a
benefit in the potential additional call to ir_raw_event_handle
whilst nvt->buf[] is being processed.
Getting rid of this additional call allows to simplify the code.
Heiner Kallweit [Tue, 2 Aug 2016 05:44:44 +0000 (02:44 -0300)]
[media] media: rc: nuvoton: remove usage of b_idx in nvt_get_rx_ir_data
The call to nvt_get_rx_ir_data and nvt_process_rx_ir_data from
the ISR is protected with spinlock nvt->lock. Therefore it's
guaranteed that nvt->pkts is 0 when entering nvt_get_rx_ir_data
(as nvt->pkts is set to 0 at the end of nvt_process_rx_ir_data).
Having said that we can remove b_idx.
[media] media: rc: nuvoton: ignore spurious interrupt when logical device is being disabled
When removing module nuvoton-cir I get a fifo overrun warning.
It turned out to be caused by a spurious interrupt when the logical CIR
device is being disabled (although no interrupt source bit being set).
Reading the interrupt status register returns 0xff, therefore the fifo
overrun bit is mistakenly interpreted as being set.
Fix this by ignoring interrupts when interrupt source and status register
reads return 0xff.
[media] media: rc: fix deadlock when module ir_lirc_codec is removed
When removing module ir_lirc_codec I got this deadlock warning.
Fix this by introducing a separate mutex to protect access
to available_protocols instead of using ir_raw_handler_lock
for this purpose.
======================================================
[ INFO: possible circular locking dependency detected ]
4.7.0-next-20160729 #1 Not tainted
-------------------------------------------------------
rmmod/2542 is trying to acquire lock:
(&dev->lock){+.+.+.}, at: [<ffffffffa03b1267>]
ir_raw_handler_unregister+0x77/0xd0 [rc_core]
but task is already holding lock:
(ir_raw_handler_lock){+.+.+.}, at: [<ffffffffa03b1212>]
ir_raw_handler_unregister+0x22/0xd0 [rc_core]
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
Sakari Ailus [Tue, 3 May 2016 21:42:13 +0000 (18:42 -0300)]
[media] media: Refactor copying IOCTL arguments from and to user space
Refactor copying the IOCTL argument structs from the user space and back,
in order to reduce code copied around and make the implementation more
robust.
As a result, the copying is done while not holding the graph mutex.
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Acked-by: Hans Verkuil <hans.verkuil@cisco.com> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
[media] exynos4-is: add of_platform_populate() call for FIMC-IS child devices
Instead of relying on the "simple-bus" compatible and the driver core
populating FIMC-IS child devices make the fimc-is driver populating
its child devices. This prevents issues related to accessing ISP_I2C
clock registers with corresponding power domain switched off, which
popped out after applying some pending IOMMU driver patches. Now the
I2C_ISP child devices will be instantiated only when required parent
device drivers are initialized and ready.
[media] exynos4-is: Clear isp-i2c adapter power.ignore_children flag
Since commit 04f59143b571161d25315dd52d7a2ecc022cb71a
("i2c: let I2C masters ignore their children for PM")
the power.ignore_children flag is set when registering an I2C
adapter. Since I2C transfers are not managed by the fimc-isp-i2c
driver its clients use pm_runtime_* calls directly to communicate
required power state of the bus controller.
However, when the power.ignore_children flag is set that doesn't
work, so clear that flag back after registering the adapter.
While at it drop pm_runtime_enable() call on the i2c_adapter
as it is already done by the I2C subsystem when registering
I2C adapter.
This patch is meant as a minimal change to fix the regression,
eventually the I2C_ISPx clock handling will be moved to the
top level fimc-is driver and whole runtime PM code is going to
be dropped from the fimc-is-i2c module.
[media] exynos4-is: Add missing entity function initialization
Suppresses following warnings:
s5p-fimc-md camera: Entity type for entity FIMC.0 was not initialized!
s5p-fimc-md camera: Entity type for entity FIMC.1 was not initialized!
s5p-fimc-md camera: Entity type for entity s5p-mipi-csis.0 was not initialized!
s5p-fimc-md camera: Entity type for entity s5p-mipi-csis.1 was not initialized!
s5p-fimc-md camera: Entity type for entity FIMC-LITE.0 was not initialized!
s5p-fimc-md camera: Entity type for entity FIMC-LITE.1 was not initialized!
s5p-fimc-md camera: Entity type for entity FIMC-IS-ISP was not initialized!
Marek Szyprowski [Wed, 31 Aug 2016 13:25:17 +0000 (10:25 -0300)]
[media] exynos4-is: Improve clock management
There is no need to keep all clocks prepared all the time. Call to
clk_prepare/unprepare can be done on demand from runtime pm callbacks
(it is allowed to call sleeping functions from that context).
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Acked-by: Krzysztof Kozlowski <k.kozlowski@samsung.com> Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Marek Szyprowski [Wed, 31 Aug 2016 13:25:16 +0000 (10:25 -0300)]
[media] exynos4-is: Add support for all required clocks
This patch adds 3 more clocks to Exynos4 ISP driver. Enabling them is
needed to make the hardware operational. Till now it worked only because
those clocks were registered with IGNORE_UNUSED flag and were enabled
by default after SoC reset.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Acked-by: Krzysztof Kozlowski <k.kozlowski@samsung.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
[media] get rid of a number of problems at the cross references
As warned by linuxdoc[1] tool, using:
$ for i in $(git grep kernel-doc Documentation/media/kapi/|cut -d: -f4); do kernel-lintdoc --sloppy $i; done
include/media/v4l2-dev.h:118 :WARN: function name from comment differs: v4l2_prio_close <--> v4l2_prio_check
include/media/v4l2-mc.h:56 [kernel-doc WARN] : enum name from comment differs: if_vid_dec_index <--> if_vid_dec_pad_index
include/media/v4l2-mc.h:71 [kernel-doc WARN] : enum name from comment differs: if_aud_dec_index <--> if_aud_dec_pad_index
include/media/v4l2-mem2mem.h:396 [kernel-doc WARN] : function name from comment differs: v4l2_m2m_num_src_bufs_ready <--> v4l2_m2m_num_dst_bufs_ready
drivers/media/dvb-core/dvb_math.h:28 [kernel-doc WARN] : function name from comment differs: cintlog2 <--> intlog2
include/media/v4l2-subdev.h:215 [kernel-doc WARN] : struct name from comment differs: s_radio <--> v4l2_subdev_tuner_ops
include/media/v4l2-subdev.h:890 [kernel-doc WARN] : function name from comment differs: v4l2_set_subdevdata <--> v4l2_set_subdev_hostdata
include/media/v4l2-subdev.h:901 [kernel-doc WARN] : function name from comment differs: v4l2_get_subdevdata <--> v4l2_get_subdev_hostdata
drivers/media/dvb-core/dvb_ringbuffer.h:196 [kernel-doc WARN] : function name from comment differs: dvb_ringbuffer_writeuser <--> dvb_ringbuffer_write_user
include/media/videobuf2-core.h:399 [kernel-doc WARN] : struct name from comment differs: vb2_ops <--> vb2_buf_ops
include/media/media-entity.h:132 [kernel-doc ERROR] : duplicate parameter definition 'source'
include/media/media-entity.h:477 [kernel-doc WARN] : function name from comment differs: media_entity_enum_test <--> media_entity_enum_test_and_set
include/media/media-entity.h:535 [kernel-doc WARN] : function name from comment differs: gobj_to_entity <--> gobj_to_pad
include/media/media-entity.h:544 [kernel-doc WARN] : function name from comment differs: gobj_to_entity <--> gobj_to_link
include/media/media-entity.h:553 [kernel-doc WARN] : function name from comment differs: gobj_to_entity <--> gobj_to_intf
include/media/media-entity.h:562 [kernel-doc WARN] : function name from comment differs: gobj_to_entity <--> intf_to_devnode
include/media/rc-core.h:234 [kernel-doc WARN] : function name from comment differs: rc_open <--> rc_close
include/media/v4l2-ctrls.h:397 [kernel-doc WARN] : missing initial short description of 'v4l2_ctrl_handler_init'
include/media/v4l2-dev.h:118 [kernel-doc WARN] : function name from comment differs: v4l2_prio_close <--> v4l2_prio_check
include/media/v4l2-event.h:225 [kernel-doc WARN] : missing initial short description of 'v4l2_src_change_event_subscribe'
for line in buf.readlines():
if line.find('.. flat-table::') != -1:
in_table = True
table = ''
elif in_table and not re.match('^[\t\n]|( )', line):
in_table = False
output += process_table(fname, table)
if in_table:
table += line
else:
output += line
if in_table:
in_table = False
output += process_table(fname, table)
return output
fname = sys.argv[1]
data = file(fname, 'rb').read().decode('utf-8')
data = process_file(fname, data)
file(fname, 'wb').write(data.encode('utf-8'))
Merge tag 'docs-next' of git://git.lwn.net/linux.git into patchwork
Merge back from docs-next in order to get the cdomain extension.
With such extension, the number of warnings when building docs
in nitpick mode reduced from 22 to 2 warnings.
* docs-next/docs-next:
docs/driver-model: fix typo
DMA-API-HOWTO: <asm/generic/scatterlist.h> is no more
doc-rst:c-domain: function-like macros arguments
doc-rst:c-domain: fix sphinx version incompatibility
Documentation/filesystems: Fixed typo
docs: Don't format internal MPT docs
docs: split up serial-interfaces.rst
docs: Pull the HSI documentation together
docs: Special-case function-pointer parameters in kernel-doc
docs: make kernel-doc handle varargs properly
x86: fix memory ranges in mm documentation
documentation/scsi: Remove nodisconnect parameter
doc: ioctl: Add some clarifications to botching-up-ioctls
docs: split up the driver book
Docs: sphinxify device-drivers.tmpl
[media] media: rcar-vin: add support for V4L2_FIELD_ALTERNATE
The HW can capture both ODD and EVEN fields in separate buffers so it's
possible to support V4L2_FIELD_ALTERNATE. This patch add support for
this mode.
At probe time and when S_STD is called the driver will default to use
V4L2_FIELD_INTERLACED if the subdevice reports V4L2_FIELD_ALTERNATE. The
driver will only change the field type if the subdevice implements
G_STD, if not it will keep the default at V4L2_FIELD_ALTERNATE.
The user can always explicitly ask for V4L2_FIELD_ALTERNATE in S_FMT and
the driver will use that field format.
It was not possible to scale beyond the image size of the video source
limitation. The output frame would be bigger but the captured image was
limited to the size of the video source.
The error was that the crop boundary was set after the requested frame
size and not the video source size. This patch breaks out the resetting
of the crop, compose and format to separate functions so the error wont
creep back.
[media] media: rcar-vin: make V4L2_FIELD_INTERLACED standard dependent
The field V4L2_FIELD_INTERLACED is standard dependent and should not
unconditionally be equivalent to V4L2_FIELD_INTERLACED_TB.
This patch adds a check to see if the video standard can be obtained and
if it's a 60 Hz format. If the condition is met V4L2_FIELD_INTERLACED
is treated as V4L2_FIELD_INTERLACED_BT if not as
V4L2_FIELD_INTERLACED_TB.
[media] adv7180: rcar-vin: change mbus format to UYVY
The media bus format reported by the adv7180 is wrong. Steve Longerbeam
posted a patch which changed the format to UYVY8_2X8 with the commit
message:
Change the media bus format from YUYV8_2X8 to UYVY8_2X8. Colors
now look correct when capturing with the i.mx6 backend. The other
option is to set the SWPC bit in register 0x27 to swap the Cr and Cb
output samples.
The rcar-vin driver was developed and tested with the adv7180 and
therefor suffers from the same issue, looking for the wrong media bus
format. The two errors corrected each other.
This patch takes Steve's patch and merge it with a fix for rcar-vin
driver. The rcar-vin driver is used used in together with the adv7180
och Koelsch and this ensures it will not break while fixing the adv7180
issue. I checked wit Steve and he was fine with me merging the patches.
ADV7180 parts:
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> Suggested-by: Steve Longerbeam <steve_longerbeam@mentor.com> Signed-off-by: Steve Longerbeam <steve_longerbeam@mentor.com> Acked-by: Lars-Peter Clausen <lars@metafoo.de> Tested-by: Tim Harvey <tharvey@gateworks.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
This driver has been replaced by the non-soc-camera rcar-vin driver.
The soc-camera framework is being deprecated, so drop this older
rcar-vin driver in favor of the newer version that does not rely on
this deprecated framework.
Hans Verkuil [Fri, 16 Sep 2016 08:23:53 +0000 (05:23 -0300)]
[media] hva: fix sparse warnings
drivers/media/platform/sti/hva/hva-v4l2.c:43:22: warning: symbol 'hva_encoders' was not declared. Should it be static?
drivers/media/platform/sti/hva/hva-v4l2.c:1401:24: warning: symbol 'hva_driver' was not declared. Should it be static?
Make these static.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
[media] staging: media: omap4iss: mark omap4iss_flush() static
We get 1 warning when building kernel with W=1:
drivers/staging/media/omap4iss/iss.c:64:6: warning: no previous prototype for 'omap4iss_flush' [-Wmissing-prototypes]
In fact, this function is only used in the file in which it is
declared and don't need a declaration, but can be made static.
so this patch marks this function with 'static'.
Trivial fix to spelling mistake in pr_warn message.
Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Trivial fix to spelling mistake in dev_err message.
Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Julia Lawall [Mon, 29 Aug 2016 13:12:01 +0000 (10:12 -0300)]
[media] constify i2c_algorithm structures
These i2c_algorithm structures are only stored in the alg field of an
i2c_adapter structure, which is declared as const. This declare the
structures as const as well.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
Building the UVC gadget into the kernel fails to build when
CONFIG_VIDEO_V4L2 is a loadable module:
drivers/usb/gadget/function/usb_f_uvc.o: In function `uvc_function_ep0_complete':
uvc_configfs.c:(.text.uvc_function_ep0_complete+0x84): undefined reference to `v4l2_event_queue'
drivers/usb/gadget/function/usb_f_uvc.o: In function `uvc_function_disable':
uvc_configfs.c:(.text.uvc_function_disable+0x34): undefined reference to `v4l2_event_queue'
Adding a dependency in USB_CONFIGFS_F_UVC (which is a bool symbol)
make the 'select USB_F_UVC' statement turn the USB_F_UVC into 'm'
whenever CONFIG_VIDEO_V4L2=m too, avoiding the link failure.
[media] atmel-isc: mark PM functions as __maybe_unused
The newly added atmel-isc driver uses SET_RUNTIME_PM_OPS() to
refer to its suspend/resume functions, causing a warning when
CONFIG_PM is not set:
media/platform/atmel/atmel-isc.c:1477:12: error: 'isc_runtime_resume' defined but not used [-Werror=unused-function]
media/platform/atmel/atmel-isc.c:1467:12: error: 'isc_runtime_suspend' defined but not used [-Werror=unused-function]
This adds __maybe_unused annotations to avoid the warning without
adding an error-prone #ifdef around it.
[media] ad5820: use __maybe_unused for PM functions
The new ad5820 driver uses #ifdef to hide the suspend/resume functions,
but gets it wrong when CONFIG_PM_SLEEP is disabled:
drivers/media/i2c/ad5820.c:286:12: error: 'ad5820_resume' defined but not used [-Werror=unused-function]
drivers/media/i2c/ad5820.c:274:12: error: 'ad5820_suspend' defined but not used [-Werror=unused-function]
This replaces the #ifdef with a __maybe_unused annotation that is
simpler and harder to get wrong, avoiding the warning.
Fixes: bee3d5115611 ("[media] ad5820: Add driver for auto-focus coil") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Pavel Machek <pavel@ucw.cz> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The newly added debug mode for the synaptics-rmi4 driver relies on
the v4l2 interface and vb2_vmalloc, but those might be configured
as loadable modules when the driver itself is built-in, resulting
in a link failure:
drivers/input/rmi4/rmi_core.o: In function `rmi_f54_remove':
rmi_f54.c:(.text.rmi_f54_remove+0x14): undefined reference to `video_unregister_device'
rmi_f54.c:(.text.rmi_f54_remove+0x20): undefined reference to `v4l2_device_unregister'
drivers/input/rmi4/rmi_core.o: In function `rmi_f54_vidioc_s_input':
rmi_f54.c:(.text.rmi_f54_vidioc_s_input+0x10): undefined reference to `video_devdata'
drivers/input/rmi4/rmi_core.o: In function `rmi_f54_vidioc_g_input':
rmi_f54.c:(.text.rmi_f54_vidioc_g_input+0x10): undefined reference to `video_devdata'
drivers/input/rmi4/rmi_core.o: In function `rmi_f54_vidioc_fmt':
rmi_f54.c:(.text.rmi_f54_vidioc_fmt+0x10): undefined reference to `video_devdata'
drivers/input/rmi4/rmi_core.o: In function `rmi_f54_vidioc_enum_input':
rmi_f54.c:(.text.rmi_f54_vidioc_enum_input+0x10): undefined reference to `video_devdata'
drivers/input/rmi4/rmi_core.o: In function `rmi_f54_vidioc_querycap':
...
The best workaround I could come up with is to disallow the debug
mode unless it's actually possible to call it.
Fixes: 3a762dbd5347 ("[media] Input: synaptics-rmi4 - add support for F54 diagnostics") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Nick Dyer <nick@shmanahar.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
The newnly added debug mode for the atmel_mxt_ts driver relies on
the v4l2 interface and vb2_vmalloc, but those might be configured
as loadable modules when the driver itself is built-in, resulting
in a link failure:
drivers/input/touchscreen/atmel_mxt_ts.o: In function `mxt_vidioc_querycap':
atmel_mxt_ts.c:(.text.mxt_vidioc_querycap+0x10): undefined reference to `video_devdata'
drivers/input/touchscreen/atmel_mxt_ts.o: In function `mxt_buffer_queue':
atmel_mxt_ts.c:(.text.mxt_buffer_queue+0x20): undefined reference to `vb2_plane_vaddr'
atmel_mxt_ts.c:(.text.mxt_buffer_queue+0x164): undefined reference to `vb2_buffer_done'
drivers/input/touchscreen/atmel_mxt_ts.o: In function `mxt_free_object_table':
atmel_mxt_ts.c:(.text.mxt_free_object_table+0x18): undefined reference to `video_unregister_device'
atmel_mxt_ts.c:(.text.mxt_free_object_table+0x20): undefined reference to `v4l2_device_unregister'
The best workaround I could come up with is to disallow the debug
mode unless it's actually possible to call it.
Fixes: ecfdd7e2660e ("[media] Input: atmel_mxt_ts - output diagnostic debug via V4L2 device") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Nick Dyer <nick@shmanahar.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Julia Lawall [Sun, 11 Sep 2016 13:05:55 +0000 (10:05 -0300)]
[media] constify local structures
For structure types defined in the same file or local header files, find
top-level static structure declarations that have the following
properties:
1. Never reassigned.
2. Address never taken
3. Not passed to a top-level macro call
4. No pointer or array-typed field passed to a function or stored in a
variable.
Declare structures having all of these properties as const.
Done using Coccinelle.
Based on a suggestion by Joe Perches <joe@perches.com>.
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Julia Lawall [Thu, 8 Sep 2016 23:59:18 +0000 (20:59 -0300)]
[media] pci: constify vb2_ops structures
Check for vb2_ops structures that are only stored in the ops field of a
vb2_queue structure. That field is declared const, so vb2_ops structures
that have this property can be declared as const also.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
Julia Lawall [Thu, 8 Sep 2016 23:59:10 +0000 (20:59 -0300)]
[media] platform: constify vb2_ops structures
Check for vb2_ops structures that are only stored in the ops field of a
vb2_queue structure. That field is declared const, so vb2_ops structures
that have this property can be declared as const also.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
Julia Lawall [Thu, 8 Sep 2016 23:59:01 +0000 (20:59 -0300)]
[media] usb: constify vb2_ops structures
Check for vb2_ops structures that are only stored in the ops field of a
vb2_queue structure. That field is declared const, so vb2_ops structures
that have this property can be declared as const also.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
Julia Lawall [Thu, 8 Sep 2016 00:59:06 +0000 (21:59 -0300)]
[media] usb: constify snd_pcm_ops structures
Check for snd_pcm_ops structures that are only stored in the ops field of a
snd_soc_platform_driver structure or passed as the third argument to
snd_pcm_set_ops. The corresponding field or parameter is declared const,
so snd_pcm_ops structures that have this property can be declared as const
also.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
Julia Lawall [Thu, 8 Sep 2016 00:44:39 +0000 (21:44 -0300)]
[media] pci: constify snd_pcm_ops structures
Check for snd_pcm_ops structures that are only stored in the ops field of a
snd_soc_platform_driver structure or passed as the third argument to
snd_pcm_set_ops. The corresponding field or parameter is declared const,
so snd_pcm_ops structures that have this property can be declared as const
also.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
[media] v4l: vsp1: Support multiple partitions per frame
Adapt vsp1_video_pipeline_run() such that it can iterate each partition
required for constructing this frame's display list chain in the event
that multiple display lists are required to process in hardware.
The first display list is held as the head list object, whilst any
following parition display lists are linked to the head by means of
vsp1_dl_list_add_chain().
Linking the chained display list headers to process using the auto start
mechanism of the hardware is performed during the vsp1_dl_list_commit().
[media] v4l: vsp1: Determine partition requirements for scaled images
The partition algorithm needs to determine the capabilities of each
entity in the pipeline to identify the correct maximum partition width.
Extend the vsp1 entity operations to provide a max_width operation and
use this call to calculate the number of partitions that will be
processed by the algorithm.
Gen 2 hardware does not require multiple partitioning, and as such
will always return a single partition.
When display lists are linked in a chain, they will be processed
automatically by the hardware, with each list linking to the next. Only
on the last display list will the frame end interrupt be fired to mark
the completion event.
Upon frame-end, the chain will be iterated to release each display list
back to the free list.
The chained lists use case (image partitioning) can require up to 64
lists per frame in the worst case scenario, bump up the number of
preallocated lists.
[media] v4l: vsp1: Replace .set_memory() with VSP1_ENTITY_PARAMS_PARTITION
The new VSP1_ENTITY_PARAMS_PARTITION configuration parameters type
covers all registers that need to be configured for every partition.
This prepares for support of image partitioning, and replaces the
.set_memory() operation as the memory registers take different values
for every partition.
[media] v4l: vsp1: Pass parameter type to entity configuration operation
Replace the current boolean parameter (full / !full) with an explicit
enum.
- VSP1_ENTITY_PARAMS_INIT for parameters to be configured at pipeline
initialization time only (V4L2 stream on or DRM atomic update)
- VSP1_ENTITY_PARAMS_RUNTIME for all parameters that can be freely
modified at runtime (through V4L2 controls)
This will allow future extensions when implementing image partitioning
support.
[media] v4l: vsp1: Disable cropping on WPF sink pad
Cropping on the WPF sink pad restricts the left and top coordinates to
0-255. The same result can be obtained by cropping on the RPF without
any such restriction, this feature isn't useful. Disable it.
[media] v4l: vsp1: Use DFE instead of FRE for frame end
The DFE and FRE interrupts are both fired at frame completion, as each
display list processes a single frame. This won't be true anymore when
using image partitioning, switch to DFE in preparation.
[media] v4l: vsp1: Repair race between frame end and qbuf handler
The frame-end function releases and completes the buffers on the input
and output entities of the pipe before marking the pipe->state as
'STOPPED'. This introduces a race whereby with the pipe->state still
'RUNNING', a QBUF handler can commence processing a frame before the
frame_end function has completed.
In the event that this happens, a frame queued by QBUF hangs due to the
incorrect pipe->state setting which prevents vsp1_pipeline_run from
issuing a CMD_STRCMD.
By locking the entire function we prevent this from occurring, but we
also change the locking state of the buffer release code. This has been
analysed visually as acceptable, but it must be considered that this now
causes the video->irqlock to be taken under the pipe->irqlock context.
[media] v4l: vsp1: Prevent pipelines from running when not streaming
Pipelines can only be run if all their video nodes are streaming. Commit b4dfb9b35a19 ("[media] v4l: vsp1: Stop the pipeline upon the first
STREAMOFF") fixed the pipeline stop sequence, but introduced a race
condition that makes it possible to run a pipeline after stopping the
stream on a video node by queuing a buffer on the other side of the
pipeline.
Fix this by clearing the buffers ready flag when stopping the stream,
which will prevent the QBUF handler from running the pipeline.
Fixes: b4dfb9b35a19 ("[media] v4l: vsp1: Stop the pipeline upon the first STREAMOFF") Reported-by: Kieran Bingham <kieran@bingham.xyz> Tested-by: Kieran Bingham <kieran@bingham.xyz> Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>
Changeset ab6343956f9c ("[media] V4L2: Add documentation for SDI timings
and related flags") added documentation for new V4L2 defines, but
it forgot to update videodev2.h.rst.exceptions to point to where
the documentation for those new values will be inside the book,
causing those warnings:
Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-dv-bt-std-sdi (if the link has no caption the label must precede a section header)
Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-dv-fl-first-field-extra-line (if the link has no caption the label must precede a section header)
Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-in-st-no-v-lock (if the link has no caption the label must precede a section header)
Documentation/output/videodev2.h.rst:6: WARNING: undefined label: v4l2-in-st-no-std-lock (if the link has no caption the label must precede a section header)
Fixes: ab6343956f9c ("[media] V4L2: Add documentation for SDI timings and related flags") Cc: Charles-Antoine Couret <charles-antoine.couret@nexvision.fr> Cc: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com>