Jarod Wilson [Wed, 7 Jul 2010 14:29:44 +0000 (11:29 -0300)]
V4L/DVB: IR/lirc_dev: fix locking in lirc_dev_fop_read
On Wed, Jul 07, 2010 at 02:52:58PM +0200, Jiri Slaby wrote:
> Hi,
>
> stanse found a locking error in lirc_dev_fop_read:
> if (mutex_lock_interruptible(&ir->irctl_lock))
> return -ERESTARTSYS;
> ...
> while (written < length && ret == 0) {
> if (mutex_lock_interruptible(&ir->irctl_lock)) { #1
> ret = -ERESTARTSYS;
> break;
> }
> ...
> }
>
> remove_wait_queue(&ir->buf->wait_poll, &wait);
> set_current_state(TASK_RUNNING);
> mutex_unlock(&ir->irctl_lock); #2
>
> If lock at #1 fails, it beaks out of the loop, with the lock unlocked,
> but there is another "unlock" at #2.
This should do the trick. Completely untested beyond compiling, but its
not exactly a complicated fix, and in practice, I'm not aware of anyone
ever actually tripping that locking bug, so there's zero functional change
in typical use here.
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Since commit b4028437876866aba4747a655ede00f892089e14
the 'driver_data' field resides in device's struct device_private
which may be allocated by dev_set_drvdata() if device_private
struct was not allocated previously.
dev_set_drvdata() is used in video_set_drvdata() to set
the driver's private data pointer in v4l2 drivers. Setting
the private data _before_ registering the v4l2 device results
in a memory leak since __video_register_device() also calls
video_set_drvdata(), but after zeroing the device structure.
Thus, the reference to the previously allocated device_private
struct goes lost and a new device_private will be allocated.
All v4l drivers which call video_set_drvdata() _before_
calling video_register_device() are affected. The patch fixes
__video_register_device() to preserve previously allocated
device_private reference.
V4L/DVB: tda10048: fix the uncomplete function tda10048_read_ber
Completes the bit-error-rate read function with the CBER register (before
Viterbi decoder). The returned value is 1e8*actual_ber to be positive.
Also includes some typo mistakes.
Jarod Wilson [Sun, 4 Jul 2010 01:42:14 +0000 (22:42 -0300)]
V4L/DVB: IR/mceusb: unify and simplify different gen device init
Started out as an effort to try to tackle the last remaining issue I'm
having with this damned pinnacle device getting wedged the first time
its plugged in after an indeterminate length of not being plugged in.
Didn't get that solved yet, but did streamline the init code a bit more
and remove some superfluous gunk. Nukes a completely unneeded call to
usb_device_init() and several lines of overly complex crap in the gen1
device init path.
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
V4L/DVB: v4l: Add MPC5121e VIU video capture driver
Adds support for Video-In (VIU) unit of Freescale
MPC5121e. The driver supports RGB888/RGB565 formats,
capture and overlay on MPC5121e DIU frame buffer.
Laurent Pinchart [Fri, 18 Jun 2010 14:31:24 +0000 (11:31 -0300)]
V4L/DVB: uvcvideo: Don't use stack-based buffers for USB transfers
Data buffers on the stack are not allowed for USB I/O. Use dynamically
allocated buffers instead when querying control length and control
capabilities.
The control capabilities are now also stored in the uvc_control
structure.
Laurent Pinchart [Thu, 18 Feb 2010 19:38:52 +0000 (16:38 -0300)]
V4L/DVB: uvcvideo: Support menu controls in the control mapping API
The UVCIOC_CTRL_MAP ioctl doesn't support menu entries for menu
controls. As the uvc_xu_control_mapping structure has no reserved
fields, this can't be fixed while keeping ABI compatibility.
Modify the UVCIOC_CTRL_MAP ioctl to add menu entries support, and define
UVCIOC_CTRL_MAP_OLD that supports the old ABI without any ability to add
menu controls.
Hans de Goede [Wed, 19 May 2010 23:15:00 +0000 (20:15 -0300)]
V4L/DVB: uvcvideo: Make button controls work properly
According to the v4l2 spec, writing any value to a button control should
result in the action belonging to the button control being triggered.
UVC cams however want to see a 1 written, this patch fixes this by
overriding whatever value user space passed in with -1 (0xffffffff) when
the control is a button control.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jean Delvare [Mon, 28 Jun 2010 15:59:49 +0000 (12:59 -0300)]
V4L/DVB: cx88: Let the user disable IR support
It might be useful to be able to disable the IR support, either for
debugging purposes, or just for users who know they won't use the IR
remote control anyway. On many cards, IR support requires expensive
polling/sampling which is better avoided if never needed.
Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jean Delvare [Mon, 28 Jun 2010 15:55:43 +0000 (12:55 -0300)]
V4L/DVB: cx88: Move I2C IR initialization
Move I2C IR initialization from just after I2C bus setup to right
before non-I2C IR initialization. This is the same as was done for
the bttv driver several months ago. Might solve bugs which have not yet
been reported for some cards. It makes both drivers consistent, and
makes it easier to disable IR support (coming soon.)
Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
drivers/media/common/tuners/xc5000.c: In function ‘xc_write_reg’:
drivers/media/common/tuners/xc5000.c:298: warning: passing argument 3 of ‘xc5000_readreg’ from incompatible pointer type
drivers/media/common/tuners/xc5000.c:235: note: expected ‘u16 *’ but argument is of type ‘u8 *’
drivers/media/common/tuners/xc5000.c: At top level:
drivers/media/common/tuners/xc5000.c:223: warning: ‘xc_read_i2c_data’ defined but not used
Andy Walls [Mon, 28 Jun 2010 02:15:35 +0000 (23:15 -0300)]
V4L/DVB: cx23885: Convert cx23885-input to use new in kernel IR pulse decoders
Convert the cx23885 driver to use the new in kernel IR pulse
decoders for the integrated CX2388[578] IR controllers. Rip out a lot
of RC-5 decoding related code in the process and rename some variables
for clarity or to more accurately describe their usage.
Signed-off-by: Andy Walls <awalls@md.metrocast.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Andy Walls [Sun, 27 Jun 2010 20:17:06 +0000 (17:17 -0300)]
V4L/DVB: cx23885: Convert from struct card_ir to struct cx23885_ir_input for IR Rx
Move from the generic, shared card_ir state structure to a cx23885 driver
specific IR state structure in anticipation of moving to the new IR pulse
decoders in the IR core.
Fix up the card name truncation in the dmesg log while we're at it,
by avoiding using fixed length string storage in our new IR state
structure.
Signed-off-by: Andy Walls <awalls@md.metrocast.net> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The image pointer and its length are now in the main structure instead
of in the frame buffer. They are updated on application vidioc_qbuf and
in the URB interrupt function when ending an image.
Vaibhav Hiremath [Sat, 12 Jun 2010 12:09:56 +0000 (09:09 -0300)]
V4L/DVB: vpfe_capture: Create separate Kconfig file for davinci devices
Currently VPFE Capture driver and DM6446 CCDC driver is being
reused for AM3517. So this patch is preparing the Kconfig/makefile
for re-use of such IP's.
V4L/DVB: tuners:tuner-simple Fix warning: variable 'tun' set but not used
Resend due to a whitespace issue I created by mistake.
The below patch fixes a warning message create by gcc 4.6.0
CC [M] drivers/media/common/tuners/tuner-simple.o
drivers/media/common/tuners/tuner-simple.c: In function 'simple_set_tv_freq':
drivers/media/common/tuners/tuner-simple.c:548:20: warning: variable 'tun' set but not used
Signed-off-by: Justin P. Mattock <justinmattock@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Dan Carpenter [Fri, 4 Jun 2010 15:39:03 +0000 (12:39 -0300)]
V4L/DVB: dvb_ca_en50221: return -EFAULT on copy_to_user errors
copy_to_user() returns the number of bytes remaining to be copied which
isn't the right thing to return here. The comments say that these
functions in dvb_ca_en50221.c should return the number of bytes copied or
an error return. I've changed it to return -EFAULT.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Dan Carpenter [Fri, 4 Jun 2010 10:34:40 +0000 (07:34 -0300)]
V4L/DVB: media/radio: fix copy_to_user to user handling
copy_to/from_user() returns the number of bytes remaining to be copied
but the code here was testing for negative returns. I modified it to
return -EFAULT. These functions are called from si4713_s_ext_ctrls() and
that only tests for negative error codes.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
In the hdpvr_probe () function, when an error occurs while probing the device,
the workqueue created by the create_single_thread () call is not properly
destroyed.
Dean Anderson [Tue, 1 Jun 2010 22:12:07 +0000 (19:12 -0300)]
V4L/DVB: s2255drv: cleanup of device structure
s2255drv: cleanup of device structure
cleanup of device structure.
single channel array instead of multiple arrays in device for
each channel property.
simplifies open callback by removing search for channel index.
Signed-off-by: Dean Anderson <linux-dev@sensoray.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Dan Carpenter [Mon, 31 May 2010 19:27:39 +0000 (16:27 -0300)]
V4L/DVB: remove unneeded null check in anysee_probe()
Smatch complained because "d" is dereferenced first and then checked for
null later . The only code path where "d" could be a invalid pointer is
if this is a cold device in dvb_usb_device_init(). I consulted Antti
Palosaari and he explained that anysee is always a warm device.
I have added a comment and removed the unneeded null check.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Amerigo Wang [Mon, 31 May 2010 06:16:17 +0000 (03:16 -0300)]
V4L/DVB: Remove obsolete zc0301 v4l driver
On 05/29/10 01:30, Jean-Francois Moine wrote:
> On Fri, 28 May 2010 13:03:28 -0400
> Amerigo Wang<amwang@redhat.com> wrote:
>
>> Subject: [PATCH 6/6] Remove obsolete zc0301 v4l driver
>>
>> Duplicate functionality with the gspca_zc3xx driver, zc0301 only
>> supports 2 USB-ID's (because it only supports a limited set of
>> sensors) wich are also supported by the gspca_zc3xx driver
>> (which supports 53 USB-ID's in total).
>
> You forgot to remove the conditionnal compilation in the gspca_zc3xx
> driver (USB_DEVICE(0x046d, 0x08ae) in gspca/zc3xx.c)
>
Right, thanks for pointing this out!
Attached is the updated patch, please use this one instead.
Thanks!
Duplicate functionality with the gspca_zc3xx driver, zc0301 only
supports 2 USB-ID's (because it only supports a limited set of
sensors) wich are also supported by the gspca_zc3xx driver
(which supports 53 USB-ID's in total).
Signed-off-by: Amerigo Wang <amwang@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Laurent Pinchart [Tue, 11 May 2010 13:36:34 +0000 (10:36 -0300)]
V4L/DVB: videobuf: Rename vmalloc fields to vaddr
The videobuf_dmabuf and videobuf_vmalloc_memory fields have a vmalloc
field to store the kernel virtual address of vmalloc'ed buffers. Rename
the field to vaddr.
Laurent Pinchart [Tue, 11 May 2010 13:36:31 +0000 (10:36 -0300)]
V4L/DVB: Remove videobuf_sg_alloc abuse
The cx88 and cx25821 drivers abuse videobuf_buffer to handle audio data.
Remove the abuse by creating private audio buffer structures with a
videobuf_dmabuf field.
Laurent Pinchart [Tue, 11 May 2010 13:36:30 +0000 (10:36 -0300)]
V4L/DVB: videobuf: Remove the videobuf_sg_dma_map/unmap functions
Instead of creating dirty wrappers around videobuf_dma_map/unmap that
create a dummy videobuf_queue structure, modify videobuf_dma_map/unmap
to take a device pointer argument and use it directly. The
videobuf_sg_dma_map/unmap then become unused and can be removed.
Pawel Osciak [Tue, 11 May 2010 13:36:29 +0000 (10:36 -0300)]
V4L/DVB: videobuf: rename videobuf_mmap_free and add sanity checks
This function is not specific to mmap, hence the rename.
Add a check whether we are not streaming or reading (for read mode that
uses the stream queue) before freeing anything.
Jarod Wilson [Sat, 3 Jul 2010 04:10:09 +0000 (01:10 -0300)]
V4L/DVB: IR/lirc: add docbook info covering lirc device interface
First ever crack at creating docbook documentation... Contains a bevy of
information on the various lirc device interface ioctls, as well as a
bit about the read and write interfaces.
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jarod Wilson [Sat, 3 Jul 2010 04:07:53 +0000 (01:07 -0300)]
V4L/DVB: IR: add ir-core to lirc userspace decoder bridge driver
v2: copy of buffer data from userspace done inside this plugin/driver,
keeping the actual drivers minimal, and more flexible in what we can
deliver to them later on (they may be fed from within kernelspace later
on, by an in-kernel IR encoder).
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Dan Carpenter [Thu, 1 Jul 2010 04:50:04 +0000 (01:50 -0300)]
V4L/DVB: dvb-usb-init.c: white space changes in dvb-usb-init
I started fixing one or two lines, but after a while I got into a groove
and started changing everything. I left the lines longer than 80
characters because that seemed to be the style in this file.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: Douglas Schilling Landgraf <dougsland@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Some (North American) providers use a non-standard mode called
"8psk turbo fec". Since there is no flag in the driver that
would allow an application to determine whether a particular
device can handle "turbo fec", the attached patch introduces
FE_CAN_TURBO_FEC.
Since there is no flag in the SI data that would indicate
that a transponder uses "turbo fec", VDR will assume that
all 8psk transponders on DVB-S use "turbo fec".
Tested-by: Derek Kelly <user.vdr@gmail.com> Signed-off-by: Klaus Schmidinger <Klaus.Schmidinger@tvdr.de> Signed-off-by: Douglas Schilling Landgraf <dougsland@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Add a mutex_unlock missing on the error path. In the other functions in
the same file the locks and unlocks of this mutex appear to be balanced,
so it would seem that the same should hold in this case.
The semantic match that finds this problem is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@
expression E1;
@@
* mutex_lock(E1,...);
<+... when != E1
if (...) {
... when != E1
* return ...;
}
...+>
* mutex_unlock(E1,...);
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk> Signed-off-by: Douglas Schilling Landgraf <dougsland@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Randy Dunlap [Tue, 29 Jun 2010 04:31:40 +0000 (01:31 -0300)]
V4L/DVB: vivi: fix depends again
My previous patch to depend on FONTS was not sufficient since
FONTS is boolean. VIDEO_VIVI needs to depend on a tristate so that
it won't be enabled as =y when framebuffer is built as modular, so
modify it to depend on the same symbols that FONTS depends on, which
are FRAMEBUFFER_CONSOLE || STI_CONSOLE.
Fixes this build error when VIDEO_VIVI=y and FRAMEBUFFER_CONSOLE=m:
vivi.c:(.init.text+0x7205): undefined reference to `find_font'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Douglas Schilling Landgraf <dougsland@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jarod Wilson [Wed, 16 Jun 2010 20:55:52 +0000 (17:55 -0300)]
V4L/DVB: IR/mceusb: add tx callback functions and wire up
mchehab: merged with IR/mceusb: userspace buffer copy moved out of driver
Userspace buffer copy moved out of driver and into lirc bridge driver
[mchehab@redhat.com: merged the patch to avoid compilation errors with allyesconfig ] Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
I have pinnacle hardware now. None of this pinnacle-specific crap is at
all necessary (in fact, some of it needed to be removed to actually make
it work). The only thing unique about this device is that it often
transfers inbound data w/a header of 0x90, meaning 16 bytes of IR data
following it, so I had to make adjustments for that, and now its working
perfectly fine.
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jarod Wilson [Wed, 16 Jun 2010 20:10:05 +0000 (17:10 -0300)]
V4L/DVB: IR/mceusb: misc cleanups and init fixes
The first-gen mceusb device init code, while mostly functional, had a few
issues in it. This patch does the following:
1) removes use of magic numbers
2) eliminates mapping of memory from stack
3) makes debug spew translator functional
Additionally, this clean-up revealed that we cannot read the proper default
tx blaster bitmask from the device, we do actually have to initialize it
ourselves, which requires use of a somewhat gross list-based mask inversion
check.
This patch also removes the entirely unnecessary use of struct ir_input_state.
Also supersedes two earlier patches that also touched on first-gen
cleanup, but were partially botched. This one actually compiles, works,
etc., I swear. ;)
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jarod Wilson [Wed, 16 Jun 2010 20:08:32 +0000 (17:08 -0300)]
V4L/DVB: IR/mceusb: use the proper ir-core device unregister function
Was using input_unregister_device directly, instead of using
ir_input_unregister, which tears down a bunch of other things in
addition to eventually calling input_unregister_device.
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jarod Wilson [Wed, 16 Jun 2010 20:07:39 +0000 (17:07 -0300)]
V4L/DVB: IR/imon: use the proper ir-core device unregister function
Was using input_unregister_device directly, instead of using
ir_input_unregister, which tears down a bunch of other things in
addition to eventually calling input_unregister_device.
Signed-off-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Randy Dunlap [Wed, 26 May 2010 17:08:51 +0000 (14:08 -0300)]
V4L/DVB: media/IR: nec-decoder needs to select BITREV
Fix ir-nec-decoder build: it uses bitrev library code, so
select BITREVERSE in its Kconfig.
ir-nec-decoder.c:(.text+0x1a2517): undefined reference to `byte_rev_table'
ir-nec-decoder.c:(.text+0x1a2526): undefined reference to `byte_rev_table'
ir-nec-decoder.c:(.text+0x1a2530): undefined reference to `byte_rev_table'
ir-nec-decoder.c:(.text+0x1a2539): undefined reference to `byte_rev_table'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
which is currently run for each decoder on each event in order
to get the client-specific decoding state data.
In addition, ir decoding modules and ir driver module load
order is now independent. Centralizing the data also allows
for a nice code reduction of about 30% per raw decoder as
client lists and client registration callbacks are no longer
necessary (but still kept around for the benefit of the lirc
decoder).
Out-of-tree modules can still use a similar trick to what
the raw decoders did before this patch until they are merged.
Signed-off-by: David Härdeman <david@hardeman.nu> Acked-by: Jarod Wilson <jarod@redhat.com> Tested-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
V4L/DVB: ir-core: Add support for disabling all protocols
Writing "none" to /dev/class/rc/rc*/protocols will disable all protocols.
This allows an easier setup, from userspace, as userspace applications don't
need to disable protocol per protocol, before enabling a different set of
protocols.
V4L/DVB: ir-core: Rename sysfs protocols nomenclature to rc-5 and rc-6
While rc-5 and rc-6 protocols are generally abreviated as "rc5" and "rc6",
previous sysfs nodes uses rc-5 and rc-6 for the Philips protocols. This is
consistent with the protocol nomenclature given by the original Philips
spec: "Remote control system RC-5" (doc. Nr. 9398 706 23011).
Also, rc5 is the name of a widely known cryptography protocol.
So, the better is to keep referring to those protocols as "rc-5" and "rc-6".
David Härdeman [Sun, 13 Jun 2010 20:29:31 +0000 (17:29 -0300)]
V4L/DVB: ir-core: centralize sysfs raw decoder enabling/disabling
With the current logic, each raw decoder needs to add a copy of the exact
same sysfs code. This is both unnecessary and also means that (re)loading
an IR driver after raw decoder modules have been loaded won't work as
expected.
This patch moves that logic into ir-raw-event and adds a single sysfs
file per device.
Reading that file returns something like:
"rc5 [rc6] nec jvc [sony]"
(with enabled protocols in [] brackets)
Writing either "+protocol" or "-protocol" to that file will
enable or disable the according protocol decoder.
An additional benefit is that the disabling of a decoder will be
remembered across module removal/insertion so a previously
disabled decoder won't suddenly be activated again. The default
setting is to enable all decoders.
This is also necessary for the next patch which moves even more decoder
state into the central raw decoding structs.
Signed-off-by: David Härdeman <david@hardeman.nu> Acked-by: Jarod Wilson <jarod@redhat.com> Tested-by: Jarod Wilson <jarod@redhat.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
V4L/DVB: em28xx-input: Don't generate one debug message for every get_key read
Instead of generating one printk for every IR read, prints it only when
count is different from the last count.
While here, as this code is called on every 100ms during the runtime
lifetime, do some performance optimization, assuming that, under normal
circumstances, it is unlikely that the driver would get a new key/key
repeat on every poll.
git://bbs.cooldavid.org / net-next-2.6.git / log