linux-toradex.git/drivers/media, branch v3.2.59

uvcvideo: Do not use usb_set_interface on bulk EP

2014-04-30T15:23:20+00:00

commit b1e43f232698274871e1358c276d7b0242a7d607 upstream.

The UVC specification uses alternate setting selection to notify devices
of stream start/stop. This breaks when using bulk-based devices, as the
video streaming interface has a single alternate setting in that case,
making video stream start and video stream stop events to appear
identical to the device. Bulk-based devices are thus not well supported
by UVC.

The webcam built in the Asus Zenbook UX302LA ignores the set interface
request and will keep the video stream enabled when the driver tries to
stop it. If USB autosuspend is enabled the device will then be suspended
and will crash, requiring a cold reboot.

USB trace capture showed that Windows sends a CLEAR_FEATURE(HALT)
request to the bulk endpoint when stopping the stream instead of
selecting alternate setting 0. The camera then behaves correctly, and
thus seems to require that behaviour.

Replace selection of alternate setting 0 with clearing of the endpoint
halt feature at video stream stop for bulk-based devices. Let's refrain
from blaming Microsoft this time, as it's not clear whether this
Windows-specific but USB-compliant behaviour was specifically developed
to handle bulkd-based UVC devices, or if the camera just took advantage
of it.

Signed-off-by: Oleksij Rempel 
Signed-off-by: Laurent Pinchart 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.2: adjust filename]
Signed-off-by: Ben Hutchings

media: gspca: sn9c20x: add ID for Genius Look 1320 V2

2014-04-30T15:23:20+00:00

commit 61f0319193c44adbbada920162d880b1fdb3aeb3 upstream.

Signed-off-by: Wolfram Sang 
Signed-off-by: Hans de Goede 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.2: adjust filename]
Signed-off-by: Ben Hutchings

saa7134: Fix unlocked snd_pcm_stop() call

2014-04-01T23:59:01+00:00

commit e6355ad7b1c6f70e2f48ae159f5658b441ccff95 upstream.

snd_pcm_stop() must be called in the PCM substream lock context.

Signed-off-by: Takashi Iwai 
[wml: Backported to 3.4: Adjust filename]
Signed-off-by: Weng Meiling 
Signed-off-by: Ben Hutchings

mxl111sf: Fix compile when CONFIG_DVB_USB_MXL111SF is unset

2014-04-01T23:58:47+00:00

commit 13e1b87c986100169b0695aeb26970943665eda9 upstream.

Fix the following build error:

drivers/media/usb/dvb-usb-v2/
mxl111sf-tuner.h:72:9: error: expected ‘;’, ‘,’ or ‘)’ before ‘struct’
         struct mxl111sf_tuner_config *cfg)

Signed-off-by: Dave Jones 
Signed-off-by: Michael Krufky 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.2: adjust filename]
Signed-off-by: Ben Hutchings

dib8000: make 32 bits read atomic

2014-04-01T23:58:39+00:00

commit 5ac64ba12aca3bef18e61c866583155a3bbf81c4 upstream.

As the dvb-frontend kthread can be called anytime, it can race
with some get status ioctl. So, it seems better to avoid one to
race with the other while reading a 32 bits register.
I can't see any other reason for having a mutex there at I2C, except
to provide such kind of protection, as the I2C core already has a
mutex to protect I2C transfers.

Note: instead of this approach, it could eventually remove the dib8000
specific mutex for it, and either group the 4 ops into one xfer or
to manually control the I2C mutex. The main advantage of the current
approach is that the changes are smaller and more puntual.

Signed-off-by: Mauro Carvalho Chehab 
Signed-off-by: Mauro Carvalho Chehab 
Acked-by: Patrick Boettcher 
[bwh: Backported to 3.2: adjust filename]
Signed-off-by: Ben Hutchings

saa7164: fix return value check in saa7164_initdev()

2014-01-03T04:33:27+00:00

commit 89f4d45b2752df5d222b5f63919ce59e2d8afaf4 upstream.

In case of error, the function kthread_run() returns ERR_PTR()
and never returns NULL. The NULL test in the return value check
should be replaced with IS_ERR().

Signed-off-by: Wei Yongjun 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.2: adjust filename]
Signed-off-by: Ben Hutchings

hdpvr: fix iteration over uninitialized lists in hdpvr_probe()

2013-10-26T20:06:02+00:00

commit 2e923a0527ac439e135b9961e58d3acd876bba10 upstream.

free_buff_list and rec_buff_list are initialized in the middle of hdpvr_probe(),
but if something bad happens before that, error handling code calls hdpvr_delete(),
which contains iteration over the lists (via hdpvr_free_buffers()).
The patch moves the lists initialization to the beginning and by the way fixes
goto label in error handling of registering videodev.
Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.2: adjust filename]
Signed-off-by: Ben Hutchings

hdpvr: register the video node at the end of probe

2013-10-26T20:06:01+00:00

commit 280847b532433ffe7a22795f926327805a127162 upstream.

Video nodes can be used at once after registration, so make sure the full
initialization is done before registering them.

Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.2: adjust filename, context]
Signed-off-by: Ben Hutchings

media: dmxdev: remove dvb_ringbuffer_flush() on writer side

2013-07-27T04:34:13+00:00

commit 414abbd2cd4c2618895f02ed3a76ec6647281436 upstream.

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.

Signed-off-by: Soeren Moch 
Reviewed-by: Sakari Ailus 
Signed-off-by: Mauro Carvalho Chehab 
[bwh: Backported to 3.2: adjust filename]
Signed-off-by: Ben Hutchings

[media] mantis: fix silly crash case

2013-05-30T13:35:15+00:00

commit e1d45ae10aea8e8a403e5d96bf5902ee670007ff upstream.

If we set mantis->fe to NULL on an error its not a good idea to then try
passing NULL to the unregister paths and oopsing really.

Resolves-bug: https://bugzilla.kernel.org/show_bug.cgi?id=16473

Signed-off-by: Alan Cox 
Signed-off-by: Mauro Carvalho Chehab 
Signed-off-by: Ben Hutchings