mpeg_users is always read or written with core->lock held except
in mpeg_release (where it looks like a bug). A plain int is simpler
and faster.
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
It should not be possible to enter mpeg_open and acquire core->lock
without the blackbird driver being registered, so just error out if it
is not. This makes the code more readable and should prevent the bug
fixed by the patch "hold device lock during sub-driver initialization"
from resurfacing again.
Similarly, if we enter mpeg_release and acquire core->lock then either
the blackbird driver is registered (since open files keep it loaded)
or the sysadmin forced the driver's removal. In the latter case the
state will be inconsistent and this is worth a loud warning.
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Add and use a mutex to protect the cx88-mpeg device list. Previously
the BKL prevented races.
Based on work by Ben Hutchings <ben@decadent.org.uk>.
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
cx8802_blackbird_probe makes a device node for the mpeg sub-device
before it has been added to dev->drvlist. If the device is opened
during that time, the open succeeds but request_acquire cannot be
called, so the reference count remains zero. Later, when the device
is closed, the reference count becomes negative --- uh oh.
Close the race by holding core->lock during probe and not releasing
until the device is in drvlist and initialization finished.
Previously the BKL prevented this race.
Reported-by: Andreas Huber <hobrom@gmx.at>
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Cc: stable@kernel.org
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The BKL conversion of this driver seems to have gone wrong.
Loading the cx88-blackbird driver deadlocks.
The cause: mpeg_ops::open in the cx2388x blackbird driver acquires the
device lock and calls the sub-driver's request_acquire, which tries to
acquire the lock again. Fix it by clarifying the semantics of
request_acquire, request_release, advise_acquire, and advise_release:
now all will rely on the caller to acquire the device lock.
Based on work by Ben Hutchings <ben@decadent.org.uk>.
Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=31962
Reported-by: Andi Huber <hobrom@gmx.at>
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Cc: stable@kernel.org
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The BKL conversion of this driver seems to have gone wrong. Various
uses of the sub-device and driver lists appear to be subject to race
conditions.
In particular, some functions access drvlist without a relevant lock
held, which will race against removal of drivers. Let's start with
that --- clean up by consistently protecting dev->drvlist with
dev->core->lock, noting driver functions that require the device lock
to be held or not to be held.
After this patch, there are still some races --- e.g.,
cx8802_blackbird_remove can run between the time the blackbird driver
is acquired and the time it is used in mpeg_release, and there's a
similar race in cx88_dvb_bus_ctrl. Later patches will address the
remaining known races and the deadlock noticed by Andi. This patch
just makes the semantics clearer in preparation for those later
changes.
Based on work by Ben Hutchings <ben@decadent.org.uk>.
Tested-by: Andi Huber <hobrom@gmx.at>
Tested-by: Marlon de Boer <marlon@hyves.nl>
Cc: stable@kernel.org
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Enabling force_testpattern module parameter in usbvision causes kernel panic.
Things like that does not belong to the kernel anyway so the fix is easy.
Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Add Nogatech USB MicroCam PAL (NV3001P) and NTSC (NV3000N) support to
usbvision driver.
PAL version is tested, NTSC untested.
Data captured using usbsnoop, init_values are listed in the INF file along
with image dimensions, offsets and frame rates.
Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Improve PID filtering and program register 20 correctly.
Make sure stream_on message is sent if streaming is off, otherwise
PIDs are not registered.
Move mutex outside lme2510_enable_pid.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
usb_buffer not inside mutex lock, waiting caller can alter buffer.
Static added to lme2510_exit and lme2510_exit_int.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Gate could close after first I2C message. On stv0288 it does.
Keep 2nd and 3rd message I2C gate control alive.
Remove unnecessary gate closing in this module.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
In drivers/media/dvb/siano/smsusb.c we have this code:
...
kfree(dev);
sms_info("device %p destroyed", dev);
...
at least one static analysis tool (Coverity Prevent) complains about this
as a use-after-free bug.
While it's true that we do use the pointer variable after freeing it, the
only use is to print the value of the pointer, so there's not actually any
problem here. But still, silencing the complaint is trivial by just moving
the kfree() call below the sms_info(), so why not just do it?. It doesn't
change the workings of the code in any way, but it makes the tool shut up.
The patch below also removes a rather pointless blank line.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
On some situations, it is desired to use eeprom data to detect
the board type. This patch adds a logic for it and fixes 2 detection
issues:
1) 10Moons UT-821 uses a generic Trident ID. Other boards
also share the same ID. So, better to use an alternative way for
it;
2) Sometimes, HVR-900H is loaded with the default Trident
ID. This seems to be some hardware bug or race condition.
The new logic will only be enabled if the device is detected as
having a generic ID.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Return device errors to demuxer from on/off streamming and
pid filtering.
Please test this patch with all dvb-usb devices.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
There are a couple of places where mutex_unlock() is called even
if mutex_lock_interruptible() failed. The patch fixes the issue.
Found by Linux Driver Verification project (linuxtesting.org).
Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Avoid unnecessary DVB TS 188 sized packet copying from DMA buffer into stack.
Backtrack one 188 sized packet just after some garbage bytes when possible.
This obsoletes patch https://patchwork.kernel.org/patch/118147/
Signed-off-by: Marko Ristola <marko.ristola@kolumbus.fi>
Acked-by: Andreas Oberritter <obi@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
We leak the memory allocated to 'fw' (the firmware) when the variable goes
out of scope.
Fix the leak by calling release_firmware(fw) before 'fw' goes out of
scope.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
If the second memory allocation in dib9000_attach() fails, we'll leak the
memory allocated by the first.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Cc: Patrick Boettcher <patrick.boettcher@dibcom.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Add support for mmap method streaming of raw YUV video on cx18-based
hardware, in addition to the existing support for read() streaming of
raw YUV and MPEG-2 encoded video.
[simon.farnsworth@onelan.co.uk: I forward-ported this from Steven's original work,
done under contract to ONELAN. The original code is at
http://www.kernellabs.com/hg/~stoth/cx18-videobuf]
Signed-off-by: Steven Toth <stoth@kernellabs.com>
Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
A greatly simplified version of the PID Filter now added
back to the Driver.
The driver allows for the PID filter to be turned off.
applied after patch 683781.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Register 42 bits 2,3,6 and 7 should be set to 0.
This gives difficult locking on some channels and may be compensated
for by other methods.
This affects any driver using the stv0288 frontend on the default
or earda inittab.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
stv0288 Register 42 bits 6 & 7 should be set to 0.
This is causing intermittent lock, the dvb-usb-lmedm04 driver uses
register 50 (auto fine mode) to correct for this, this register is
now returned to its default setting.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
An usb_submit_urb should be atomic in a completion handler. This fixes
"BUG: scheduling while atomic" messages.
Signed-off-by: Pete Eberlein <pete@sensoray.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Add support for TT S-1500 with BSBE1-D01A tuner.
Signed-off-by: Oliver Endriss <o.endriss@gmx.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Seems like 'adap->fe' test for NULL was meant to be before we dereference
that pointer.
Signed-off-by: Mariusz Kozlowski <mk@lab.zgora.pl>
Acked-by: Patrick Boettcher <patrick.boettcher@dibcom.fr>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The generic_standards_cnt define is only used in one place and it's
more readable to just call ARRAY_SIZE(generic_standards) directly.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This function returns NULL on failure so lets do that if kzalloc()
fails. There is a separate problem that the caller for this function
doesn't check for errors...
Signed-off-by: Dan Carpenter <error27@gmail.com>
Acked-By: Mike Isely <isely@pobox.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This card has a TD18271 silicon tuner, and uses TDA8290 and LGS8G75 to
demodulate analog and digital broadcast respectively. GPIO configurations
were derived using DScaler regspy.
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The ability to force the encoder or decoder chip was broken by commit
0ab6e1c38d in February 2009. As nobody
complained for over 2 years, I take it that these parameters were no
longer used so we can simply drop them.
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Tested-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Florian Mickler <florian@mickler.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Signed-off-by: Florian Mickler <florian@mickler.org>
Acked-by: Antti Palosaari <crope@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Tested-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Signed-off-by: Florian Mickler <florian@mickler.org>
Acked-by: Antti Palosaari <crope@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Tested-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Signed-off-by: Florian Mickler <florian@mickler.org>
Acked-by: Antti Palosaari <crope@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Tested-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Note: This change is tested to compile only as I don't have the
hardware.
Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
If we need a bigger buffer, we reallocte a new buffer and free the old
one.
Note: This change is tested to compile only as I don't have the
hardware.
Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Note: This change is tested to compile only as I don't have the
hardware.
Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Otherwise it is not obvious that vp702x_usb_in_op or vp702x_usb_out_op
will not interfere with any vp702x_usb_inout_op.
Note: This change is tested to compile only, as I don't have the
hardware.
Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Note: This change is tested to compile only, as I don't have the hardware.
Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
This sets up a buffer and a mutex protecting that buffer in
the struct vp702x_device_state.
The definition of struct vp702x_device_state is moved into the header
in order to use the buffer also in the frontend.
Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
We need a state struct for the dvb_usb_device.
In order to reduce confusion we rename the vp702x_state struct.
Signed-off-by: Florian Mickler <florian@mickler.org>
Cc: Patrick Boettcher <pb@linuxtv.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Jonathan Nieder