[media] V4L2: sh_mobile_ceu_camera: remove CEU specific data from generic functions
Several functions in the sh_mobile_ceu_camera driver implement generic
algorithms and can be re-used by other V4L2 camera host drivers too. These
functions attempt to optimise scaling and cropping functions of the
subdevice, e.g. a camera sensor. This patch makes those functions generic
for future re-use by other camera host drivers.
[media] V4L2: soc-camera: fix uninitialised use compiler warning
In scan_async_group() if the size parameter is negative, the sasd pointer
will be used uninitialised:
drivers/media/platform/soc_camera/soc_camera.c: In function "soc_camera_host_register":
drivers/media/platform/soc_camera/soc_camera.c:1514:55: warning: "sasd" may
be used uninitialized in this function [-Wmaybe-uninitialized]
sasd->asd.match.i2c.adapter_id, sasd->asd.match.i2c.address);
^
drivers/media/platform/soc_camera/soc_camera.c:1464:34: note: "sasd" was
declared here
struct soc_camera_async_subdev *sasd;
Fix this by making "size" and the array, from which it is assigned unsigned.
Add support for asynchronous subdevice probing, using the v4l2-async API.
The legacy synchronous mode is still supported too, which allows to
gradually update drivers and platforms. The selected approach adds a
notifier for each struct soc_camera_device instance, i.e. for each video
device node, even when there are multiple such instances registered with a
single soc-camera host simultaneously.
[media] soc-camera: switch I2C subdevice drivers to use v4l2-clk
Instead of centrally enabling and disabling subdevice master clocks in
soc-camera core, let subdevice drivers do that themselves, using the
V4L2 clock API and soc-camera convenience wrappers.
[media] V4L2: support asynchronous subdevice registration
Currently bridge device drivers register devices for all subdevices
synchronously, typically, during their probing. E.g. if an I2C CMOS sensor
is attached to a video bridge device, the bridge driver will create an I2C
device and wait for the respective I2C driver to probe. This makes linking
of devices straight forward, but this approach cannot be used with
intrinsically asynchronous and unordered device registration systems like
the Flattened Device Tree. To support such systems this patch adds an
asynchronous subdevice registration framework to V4L2. To use it respective
(e.g. I2C) subdevice drivers must register themselves with the framework.
A bridge driver on the other hand must register notification callbacks,
that will be called upon various related events.
[media] V4L2: add a device pointer to struct v4l2_subdev
It is often useful to have simple means to get from a subdevice to the
underlying physical device. This patch adds such a pointer to struct
v4l2_subdev and sets it accordingly in the I2C and SPI cases.
Typical video devices like camera sensors require an external clock source.
Many such devices cannot even access their hardware registers without a
running clock. These clock sources should be controlled by their consumers.
This should be performed, using the generic clock framework. Unfortunately
so far only very few systems have been ported to that framework. This patch
adds a set of temporary helpers, mimicking the generic clock API, to V4L2.
Platforms, adopting the clock API, should switch to using it. Eventually
this temporary API should be removed.
[media] soc-camera: make .clock_{start,stop} compulsory, .add / .remove optional
All existing soc-camera host drivers use .clock_start() and .clock_stop()
callbacks to activate and deactivate their camera interfaces, whereas
.add() and .remove() callbacks are usually dummy. Make the former two
compulsory and the latter two optional.
[media] sh-mobile-ceu-camera: move interface activation and deactivation to clock callbacks
When adding and removing a client, the sh-mobile-ceu-camera driver activates
and, respectively, deactivates its camera interface and, if necessary, the
CSI2 controller. Only handling of the CSI2 interface is client-specific and
is only needed, when a data-exchange with the client is taking place. Move
the rest to .clock_start() and .clock_stop() callbacks.
[media] mx1-camera: move interface activation and deactivation to clock callbacks
When adding and removing a client, the mx1-camera driver only activates
and deactivates its camera interface respectively, which doesn't include
any client-specific actions. Move this functionality into .clock_start()
and .clock_stop() callbacks.
[media] mx2-camera: move interface activation and deactivation to clock callbacks
When adding and removing a client, the mx2-camera driver only activates
and deactivates its camera interface respectively, which doesn't include
any client-specific actions. Move this functionality into .clock_start()
and .clock_stop() callbacks.
[media] mx3-camera: move interface activation and deactivation to clock callbacks
When adding and removing a client, the mx3-camera driver only activates
and deactivates its camera interface respectively, which doesn't include
any client-specific actions. Move this functionality into .clock_start()
and .clock_stop() callbacks.
[media] atmel-isi: move interface activation and deactivation to clock callbacks
When adding and removing a client, the atmel-isi camera host driver only
activates and deactivates its camera interface respectively, which doesn't
include any client-specific actions. Move this functionality into
.clock_start() and .clock_stop() callbacks.
[media] omap1-camera: move interface activation and deactivation to clock callbacks
When adding and removing a client, the omap1-camera driver only activates
and deactivates its camera interface respectively, which doesn't include
any client-specific actions. Move this functionality into .clock_start()
and .clock_stop() callbacks.
[media] pxa-camera: move interface activation and deactivation to clock callbacks
When adding and removing a client, the pxa-camera driver only activates
and deactivates its camera interface respectively, which doesn't include
any client-specific actions. Move this functionality into .clock_start()
and .clock_stop() callbacks.
[media] soc-camera: add host clock callbacks to start and stop the master clock
Currently soc-camera uses a single camera host callback to activate the
interface master clock and to configure the interface for a specific
client. However, during probing we might not have the information about
a client, we just need to activate the clock. Add new camera host driver
callbacks to only start and stop the clock without and client-specific
configuration.
[media] soc-camera: move common code to soc_camera.c
All soc-camera host drivers include a pointer to an soc-camera device in
their host private struct to check, that only one client is connected.
Move this common code to soc_camera.c.
[media] ttusb-budget: fix memory leak in ttusb_probe()
If something goes wrong starting from i2c_add_adapter(),
ttusb->iso_urb[] and ttusb itself are not deallocated.
Found by Linux Driver Verification project (linuxtesting.org).
[media] go7007: fix return 0 for unsupported devices in go7007_usb_probe()
probe() should not return 0 for unsupported devices, but go7007_usb_probe() does.
The patch fixes it to return -ENODEV.
Found by Linux Driver Verification project (linuxtesting.org).
Lubomir Rintel [Mon, 17 Jun 2013 13:29:06 +0000 (10:29 -0300)]
[media] usbtv: Add driver for Fushicai USBTV007 video frame grabber
Reverse-engineered driver for cheapo video digitizer, made from observations of
Windows XP driver. The protocol is not yet completely understood, so far we
don't provide any controls, only support a single format out of three and don't
support the audio device.
Lad, Prabhakar [Sat, 15 Jun 2013 15:34:10 +0000 (12:34 -0300)]
[media] media: i2c: ths7303: remove unused member driver_data
This patch removes the driver_data member from ths7303_state structure.
The driver_data member was intended to differentiate between ths7303 and
ths7353 chip and get the g_chip_ident, But as of now g_chip_ident is
obsolete, so there is no need of driver_data.
Ondrej Zary [Fri, 14 Jun 2013 20:01:39 +0000 (17:01 -0300)]
[media] radio-sf16fmi: Set frequency during init
Set freqency during initialization to fix v4l2-compliance error.
This also fixes VIDIOC_G_FREQUENCY always returning zero (broken by me during LM7000 conversion).
Signed-off-by: Ondrej Zary <linux@rainbow-software.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
[media] usbvision-video: fix memory leak of alt_max_pkt_size
1. usbvision->alt_max_pkt_size is not deallocated anywhere.
2. if allocation of usbvision->alt_max_pkt_size fails,
there is no proper deallocation of already acquired resources.
The patch adds kfree(usbvision->alt_max_pkt_size) to
usbvision_release() as soon as other deallocations happen there.
It calls usbvision_release() if allocation of
usbvision->alt_max_pkt_size fails as soon as usbvision_release()
is safe to work with incompletely initialized usbvision structure.
Found by Linux Driver Verification project (linuxtesting.org).
Lad, Prabhakar [Mon, 17 Jun 2013 14:20:48 +0000 (11:20 -0300)]
[media] media: davinci: vpif_display: move the freeing of irq and global variables to remove()
Ideally the freeing of irq's and the global variables needs to be
done in the remove() rather than module_exit(), this patch moves
the freeing up of irq's and freeing the memory allocated to channel
objects to remove() callback of struct platform_driver.
Lad, Prabhakar [Mon, 17 Jun 2013 14:20:44 +0000 (11:20 -0300)]
[media] media: davinci: vpif_capture: move the freeing of irq and global variables to remove()
Ideally the freeing of irq's and the global variables needs to be
done in the remove() rather than module_exit(), this patch moves
the freeing up of irq's and freeing the memory allocated to channel
objects to remove() callback of struct platform_driver.
drivers/media/usb/pvrusb2/pvrusb2-v4l2.c: In function 'pvr2_v4l2_dev_init':
drivers/media/usb/pvrusb2/pvrusb2-v4l2.c:1268:21: warning: variable 'usbdev' set but not used [-Wunused-but-set-variable]
This warning is due to changeset a28fbd04facb, with removed the
usage of usbdev inside pvr2_v4l2_dev_init().
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com> Cc: Hans Verkuil <hans.verkuil@cisco.com> Cc: Mike Isely <isely@pobox.com>
Hans Verkuil [Wed, 12 Jun 2013 14:23:44 +0000 (11:23 -0300)]
[media] cx88: set dev_parent to the correct parent PCI bus
The cx88 driver has one v4l2_device, but the video nodes are owned by two
different PCI busses. So the dev_parent pointer should be set to the correct
parent bus, otherwise sysfs won't show the correct device hierarchy.
This broke starting in 3.6 after a driver change, so this patch resurrects
the correct behavior.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Wed, 12 Jun 2013 14:15:12 +0000 (11:15 -0300)]
[media] v4l2: always require v4l2_dev, rename parent to dev_parent
The last set of drivers still using the parent field of video_device instead
of the v4l2_dev field have been converted, so v4l2_dev is now always set.
A proper pointer to v4l2_dev is necessary these days otherwise the advanced
debugging ioctls will not work when addressing sub-devices. It also ensures
that the core can always go from a video_device struct to the top-level
v4l2_device struct.
There is still one single use case for the parent pointer: if there are
multiple busses, each being the parent of one or more video nodes, and if
they all share the same v4l2_device struct. In that case one still needs a
parent pointer since the v4l2_device struct can only refer to a single
parent device. The cx88 driver is one such case. Unfortunately, the cx88
failed to set the parent pointer since 3.6. The next patch will correct this.
In order to support this use-case the parent pointer is only renamed to
dev_parent, not removed altogether. It has been renamed to ensure that the
compiler will catch any (possibly out-of-tree) drivers that were missed during
the conversion.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Acked-by: Sakari Ailus <sakari.ailus@iki.fi> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Wed, 12 Jun 2013 09:03:27 +0000 (06:03 -0300)]
[media] omap24xxcam: add v4l2_device and replace parent with v4l2_dev
This driver did not yet support struct v4l2_device, so add it. This
make it possible to replace the deprecated parent field with the
v4l2_dev field, allowing the eventual removal of the parent field.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Acked-by: Sakari Ailus <sakari.ailus@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Wed, 12 Jun 2013 09:02:38 +0000 (06:02 -0300)]
[media] f_uvc: add v4l2_device and replace parent with v4l2_dev
This driver did not yet support struct v4l2_device, so add it. This
make it possible to replace the deprecated parent field with the
v4l2_dev field, allowing the eventual removal of the parent field.
Hans Verkuil [Wed, 12 Jun 2013 08:49:50 +0000 (05:49 -0300)]
[media] saa7164: add v4l2_device and replace parent with v4l2_dev
This driver did not yet support struct v4l2_device, so add it. This
make it possible to replace the deprecated parent field with the
v4l2_dev field, allowing the eventual removal of the parent field.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Wed, 12 Jun 2013 08:46:30 +0000 (05:46 -0300)]
[media] sn9c102_core: add v4l2_device and replace parent with v4l2_dev
This driver did not yet support struct v4l2_device, so add it. This
make it possible to replace the deprecated parent field with the
v4l2_dev field, allowing the eventual removal of the parent field.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Mon, 10 Jun 2013 12:27:41 +0000 (09:27 -0300)]
[media] soc_camera: replace vdev->parent by vdev->v4l2_dev
The parent field will eventually disappear to be replaced by v4l2_dev.
soc_camera does provide a v4l2_device struct but did not point to it in
struct video_device. This is now fixed.
Now the video nodes can be found under the correct platform bus, and
the advanced debug ioctls work correctly as well (the core implementation
of those ioctls requires that v4l2_dev is set correctly).
Hans Verkuil [Wed, 12 Jun 2013 09:04:04 +0000 (06:04 -0300)]
[media] v4l2-device: check if already unregistered
It was possible to unregister an already unregistered v4l2_device struct.
Add a check whether that already happened and just return if that was
the case.
Also refuse to register a v4l2_device if both the dev and name fields are
empty. A warning was already produced in that case, but since the name field
is now used to detect whether or not the v4l2_device was already unregistered
this particular combination should be rejected.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This function is no longer needed since it is now the responsibility of the
v4l2 core to check if the DBG_G/S_REGISTER and DBG_G_CHIP_INFO ioctls are
called for the bridge driver or not.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Hans Verkuil [Sun, 26 May 2013 10:56:06 +0000 (07:56 -0300)]
[media] cx88: remove g_chip_ident
Remove g_chip_ident from cx88. Also remove the v4l2-chip-ident.h include.
The board code used defines from v4l2-chip-ident.h to tell the driver which
audio chip is used. Replace this with a cx88-specific enum.
[media] Fix build when drivers are builtin and frontend modules
There are a large number of reports that the media build is
not compiling when some drivers are compiled as builtin, while
the needed frontends are compiled as module.
On the last one of such reports:
From: kbuild test robot <fengguang.wu@intel.com>
Subject: saa7134-dvb.c:undefined reference to `zl10039_attach'
The .config file has:
CONFIG_VIDEO_SAA7134=y
CONFIG_VIDEO_SAA7134_DVB=y
# CONFIG_MEDIA_ATTACH is not set
CONFIG_DVB_ZL10039=m
And it produces all those errors:
drivers/built-in.o: In function `set_type':
tuner-core.c:(.text+0x2f263e): undefined reference to `tea5767_attach'
tuner-core.c:(.text+0x2f273e): undefined reference to `tda9887_attach'
drivers/built-in.o: In function `tuner_probe':
tuner-core.c:(.text+0x2f2d20): undefined reference to `tea5767_autodetection'
drivers/built-in.o: In function `av7110_attach':
av7110.c:(.text+0x330bda): undefined reference to `ves1x93_attach'
av7110.c:(.text+0x330bf7): undefined reference to `stv0299_attach'
av7110.c:(.text+0x330c63): undefined reference to `tda8083_attach'
av7110.c:(.text+0x330d09): undefined reference to `ves1x93_attach'
av7110.c:(.text+0x330d33): undefined reference to `tda8083_attach'
av7110.c:(.text+0x330d5d): undefined reference to `stv0297_attach'
av7110.c:(.text+0x330dbe): undefined reference to `stv0299_attach'
drivers/built-in.o: In function `tuner_attach_dtt7520x':
ngene-cards.c:(.text+0x3381cb): undefined reference to `dvb_pll_attach'
drivers/built-in.o: In function `demod_attach_lg330x':
ngene-cards.c:(.text+0x33828a): undefined reference to `lgdt330x_attach'
drivers/built-in.o: In function `demod_attach_stv0900':
ngene-cards.c:(.text+0x3383d5): undefined reference to `stv090x_attach'
drivers/built-in.o: In function `cineS2_probe':
ngene-cards.c:(.text+0x338b7f): undefined reference to `drxk_attach'
drivers/built-in.o: In function `configure_tda827x_fe':
saa7134-dvb.c:(.text+0x346ae7): undefined reference to `tda10046_attach'
drivers/built-in.o: In function `dvb_init':
saa7134-dvb.c:(.text+0x347283): undefined reference to `mt352_attach'
saa7134-dvb.c:(.text+0x3472cd): undefined reference to `mt352_attach'
saa7134-dvb.c:(.text+0x34731c): undefined reference to `tda10046_attach'
saa7134-dvb.c:(.text+0x34733c): undefined reference to `tda10046_attach'
saa7134-dvb.c:(.text+0x34735c): undefined reference to `tda10046_attach'
saa7134-dvb.c:(.text+0x347378): undefined reference to `tda10046_attach'
saa7134-dvb.c:(.text+0x3473db): undefined reference to `tda10046_attach'
drivers/built-in.o:saa7134-dvb.c:(.text+0x347502): more undefined references to `tda10046_attach' follow
drivers/built-in.o: In function `dvb_init':
saa7134-dvb.c:(.text+0x347812): undefined reference to `mt352_attach'
saa7134-dvb.c:(.text+0x347951): undefined reference to `mt312_attach'
saa7134-dvb.c:(.text+0x3479a9): undefined reference to `mt312_attach'
>> saa7134-dvb.c:(.text+0x3479c1): undefined reference to `zl10039_attach'
This is happening because a builtin module can't use directly a symbol
found on a module. By enabling CONFIG_MEDIA_ATTACH, the configuration
becomes valid, as dvb_attach() macro loads the module if needed, making
the symbol available to the builtin module.
While this bug started to appear after the patches that use IS_DEFINED
macro (like changeset 7b34be71db533f3e0cf93d53cf62d036cdb5418a), this
bug is a way ancient than that.
The thing is that, before the IS_DEFINED() patches, the logic used to be:
The above code, with the .config file used, was evoluting to FALSE
(instead of TRUE as it should be, as CONFIG_DVB_ZL10039 is 'm'),
and were adding the static inline code at saa7134-dvb, instead
of the external call. So, while it weren't producing any compilation
error, the code weren't working either.
So, as the overhead for using CONFIG_MEDIA_ATTACH is minimal, just
enable it, if MODULES is defined.
[media] s5p makefiles: don't override other selections on obj-[ym]
The $obj-m/$obj-y vars should be adding new modules to build, not
overriding it. So, it should never use
$obj-y := foo.o
instead, it should use:
$obj-y += foo.o
Failing to do that is very bad, as it will suppress needed modules.
* linus: (1465 commits)
ARM: tegra30: clocks: Fix pciex clock registration
lseek(fd, n, SEEK_END) does *not* go to eof - n
Linux 3.10-rc6
smp.h: Use local_irq_{save,restore}() in !SMP version of on_each_cpu().
powerpc: Fix missing/delayed calls to irq_work
powerpc: Fix emulation of illegal instructions on PowerNV platform
powerpc: Fix stack overflow crash in resume_kernel when ftracing
snd_pcm_link(): fix a leak...
use can_lookup() instead of direct checks of ->i_op->lookup
move exit_task_namespaces() outside of exit_notify()
fput: task_work_add() can fail if the caller has passed exit_task_work()
xfs: don't shutdown log recovery on validation errors
xfs: ensure btree root split sets blkno correctly
xfs: fix implicit padding in directory and attr CRC formats
xfs: don't emit v5 superblock warnings on write
mei: me: clear interrupts on the resume path
mei: nfc: fix nfc device freeing
mei: init: Flush scheduled work before resetting the device
sctp: fully initialize sctp_outq in sctp_outq_init
netiucv: Hold rtnl between name allocation and device registration.
...
Soeren Moch [Thu, 6 Jun 2013 00:26:23 +0000 (21:26 -0300)]
[media] media: dmxdev: remove dvb_ringbuffer_flush() on writer side
In dvb_ringbuffer lock-less synchronizationof reader and writer threads is done
with separateread and write pointers. Sincedvb_ringbuffer_flush() modifies the
read pointer, this function must not be called from the writer thread.
This patch removes the dvb_ringbuffer_flush() calls in the dmxdev ringbuffer
write functions, this fixes Oopses "Unable to handle kernel paging request"
I could observe for the call chaindvb_demux_read ->dvb_dmxdev_buffer_read ->
dvb_ringbuffer_read_user -> __copy_to_user (the reader side of the ringbuffer).
The flush calls at the write side are not necessary anyway since ringbuffer_flush
is also called in dvb_dmxdev_buffer_read() when an error condition is set in the
ringbuffer.
This patch should also be applied to stable kernels.
Frank Schaefer [Mon, 3 Jun 2013 17:12:05 +0000 (14:12 -0300)]
[media] em28xx: remove GPIO register caching
The GPIO register caching is the result of wrong assumptions and incomplete
knowledge about the GPIO registers and their functionality.
Today, we know that it is not needed.
It is also limited to a single register and therefore incomplete (newer chips
are using multiple registers).
Instead of extending the caching, get rid of it, because it has no real
benefits and just bloats/complicates the code.
Signed-off-by: Frank Schäfer <fschaefer.oss@googlemail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Frank Schaefer [Mon, 3 Jun 2013 17:12:03 +0000 (14:12 -0300)]
[media] em28xx: improve em2820-em2873/83 GPIO port register definitions and descriptions
- add definition for GPIO register 0x09 (reading/input)
- extend the information the chip variants that support GPIO registers 0x08/0x09
- rename EM28XX_R08_GPIO to EM2820_R08_GPIO_CTRL
Signed-off-by: Frank Schäfer <fschaefer.oss@googlemail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Frank Schaefer [Mon, 3 Jun 2013 17:12:02 +0000 (14:12 -0300)]
[media] em28xx: extend GPIO register definitions for the em25xx, em276x/7x/8x, em2874/174/84
The em25xx/em276x/7x/8x provides 4 GPIO register sets,
each of them consisting of separate read and a write registers.
The same registers are also used by the em2874/174/84.
Signed-off-by: Frank Schäfer <fschaefer.oss@googlemail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Remove the silly redefines of pr_err/info/debug.
This improves readability and it also gets rid of a bunch of warnings when
compiling this driver for older kernels using the compatibility media_build
system.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Michael Krufky <mkrufky@linuxtv.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Gianluca Gennari [Mon, 17 Jun 2013 23:32:45 +0000 (20:32 -0300)]
[media] r820t: fix imr calibration
The r820t_imr() calibration function of the Rafael Micro R820T tuner
generates this error at every tune attempt:
r820t 0-001a: No valid PLL values for 2252021 kHz!
The function was inspired by the original Realtek driver for rtl2832 devices
with the r820t tuner; anyway, in the original code the XTAL frequency of
the tuner was expressed in KHz, while in the kernel driver it is expressed
in Hz; so the calibration failed because of an out-of-range initial value.
The final result of the computation is then passed to the r820t_set_mux()
and r820t_set_pll() functions, but the conversion from KHz to Hz is already
correctly implemented.
[media] r820t: remove redundant initializations in r820t_attach()
fe->tuner_priv and fe->ops.tuner_ops are initialized twice in r820t_attach().
Remove the redundant initializations and also move fe->ops.tuner_ops
initialization outside of the mutex lock (as in the xc4000 tuner code for example).
The ISP clock register content is not preserved over the ISP power domain
off/on cycle. Instead of setting the clock frequencies once at probe time
the clock rates set up is moved to the runtime_resume handler, which is
invoked after the related power domain is already enabled, ensuring the
clocks are properly configured when the device is actively used.
This fixes the FIMC-IS malfunctions and STREAM ON timeout errors accuring
on some boards:
[ 59.860000] fimc_is_general_irq_handler:583 ISR_NDONE: 5: 0x800003e8, IS_ERROR_UNKNOWN
[ 59.860000] fimc_is_general_irq_handler:586 IS_ERROR_TIME_OUT
[media] V4L: Remove all links of the media entity when unregistering subdev
Remove all links of the subdev's media entity after internal_ops
'unregistered' call and right before unregistering the entity from
a media device.
It is assumed here that an unregistered (orphan) media entity cannot
have links to other entities registered to a media device.
It is also assumed the media links should be created/removed with
the media graph's mutex held.
The above implies that the caller of v4l2_device_unregister_subdev()
must not hold the graph's mutex.
Reviewed-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Acked-by: Sakari Ailus <sakari.ailus@iki.fi> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
[media] media: Add a function removing all links of a media entity
This function allows to remove all media entity's links to other
entities, leaving no references to a media entity's links array
at its remote entities.
Currently, when a driver of some entity is removed it will free its
media entities links[] array, leaving dangling pointers at other
entities that are part of same media graph. This is troublesome when
drivers of a media device entities are in separate kernel modules,
removing only some modules will leave others in an incorrect state.
This function is intended to be used when an entity is being
unregistered from a media device.
With an assumption that normally the media links should be created
between media entities registered to a media device, with the graph
mutex held.
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com> Reviewed-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Linus Torvalds [Tue, 18 Jun 2013 16:25:08 +0000 (06:25 -1000)]
Merge branch 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media fixes from Mauro Carvalho Chehab:
"Series of fixes for 3.10. There are some usual driver fixes (mostly
on s5p/exynos playform drivers), plus some fixes at V4L2 core"
* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (40 commits)
[media] soc_camera: error dev remove and v4l2 call
[media] sh_veu: fix the buffer size calculation
[media] sh_veu: keep power supply until the m2m context is released
[media] sh_veu: invoke v4l2_m2m_job_finish() even if a job has been aborted
[media] v4l2-ioctl: don't print the clips list
[media] v4l2-ctrls: V4L2_CTRL_CLASS_FM_RX controls are also valid radio controls
[media] cx88: fix NULL pointer dereference
[media] DocBook/media/v4l: update version number
[media] exynos4-is: Remove "sysreg" clock handling
[media] exynos4-is: Fix reported colorspace at FIMC-IS-ISP subdev
[media] exynos4-is: Ensure fimc-is clocks are not enabled until properly configured
[media] exynos4-is: Prevent NULL pointer dereference when firmware isn't loaded
[media] s5p-mfc: Add NULL check for allocated buffer
[media] s5p-mfc: added missing end-of-lines in debug messages
[media] s5p-mfc: v4l2 controls setup routine moved to initialization code
[media] s5p-mfc: separate encoder parameters for h264 and mpeg4
[media] s5p-mfc: Remove special clock usage in driver
[media] s5p-mfc: Remove unused s5p_mfc_get_decoded_status_v6() function
[media] v4l2: mem2mem: save irq flags correctly
[media] coda: v4l2-compliance fix: add VIDIOC_CREATE_BUFS support
...
Linus Torvalds [Tue, 18 Jun 2013 16:23:51 +0000 (06:23 -1000)]
Merge tag 'clk-fixes-for-linus' of git://git.linaro.org/people/mturquette/linux
Pull clock framework fixes from Mike Turquette:
"Half of the fixes here are for Exynos5, fixing regressions in CPUfreq
due to the common clock framework conversion as well as one fix which
allows the platform to properly reboot again.
One core framework fix patches up a memory leak, another fixes a build
error for the SPEAr platform and finally a Tegra-specific fix allows
PCIe to initialize properly on that platform again"
* tag 'clk-fixes-for-linus' of git://git.linaro.org/people/mturquette/linux:
ARM: tegra30: clocks: Fix pciex clock registration
clk: exynos5250: Add CLK_IGNORE_UNUSED flag for pmu clock
clk: spear: fix build error for spear3xx
clk: samsung: Fix pll36xx_recalc_rate to handle kdiv properly
clk: exynos5250: Add sclk_mpll to the parent list of mout_cpu clock
clk: exynos5250: Update cpufreq related clocks for EXYNOS5250
clk: remove notifier from list before freeing it
[media] staging/media: lirc_imon: fix leaks in imon_probe()
Error handling of usb_submit_urb() is not as all others in imon_probe().
It just unlocks mutexes and returns nonzero leaving all already
allocated resources unfreed.
The patch makes sure all the resources are deallocated.
Found by Linux Driver Verification project (linuxtesting.org).
[media] rtl28xxu: fix buffer overflow when probing Rafael Micro r820t tuner
As suggested by Antti, this patch replaces:
https://patchwork.kernel.org/patch/2649861/
The buffer overflow is fixed by reading only the r820t ID register.
Signed-off-by: Gianluca Gennari <gennarone@gmail.com> Acked-by: Antti Palosaari <crope@iki.fi> Reviewed-by: Antti Palosaari <crope@iki.fi> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
projects / linux-beck.git / log