Now that we've added MEDIA_ENT_T_V4L2_SUBDEV_TUNER at the API,
document it.
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Add MEDIA_ENT_T_V4L2_SUBDEV_TUNER to represent the V4L2
(and dvb) tuner subdevices.
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

The DVB API is actually several different APIs bundled together, each
using its own device node.

Fix the media controller DVB API to actually reflect what's there at the
DVB subsystem.
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Most of the DVB subdevs have already their own devnode.

Add support for them at the media controller API.
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

The media-ctl userspace application assumes that all device nodes
are uniquelly defined via major,minor, just like v4l and fb.

That's ok for those types of devices, but, as we're adding support
for DVB at the API, what's written there at the DocBook is wrong.

So, fix it.

While here, fix the size of the reserved space inside the union,
with is 184, and not 180.
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

The previous provision for DVB media controller support were to
define an ID (likely meaning the adapter number) for the DVB
devnodes.

This is just plain wrong. Just like V4L, DVB devices (and any other
device node)) are uniquely identified via a (major, minor) tuple.

This is enough to uniquely identify a devnode, no matter what
API it implements.

So, before we go too far, let's mark the old v4l, fb, dvb and alsa
"devnode" info as deprecated, and just call it as "dev".

We can latter add fields specific to each API if needed.

As we don't want to break compilation on already existing apps,
let's just keep the old definitions as-is, adding a note that
those are deprecated at media-entity.h.
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Commit 9439eb3a ("asm-generic: io: implement relaxed
accessor macros as conditional wrappers") has added
{read,write}{b,w,l,q}_relaxed to include/asm-generic/io.h
Signed-off-by: Zhangfei Gao <zhangfei.gao@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

As reported by Peter Kovar, there's a potential risk of a division by
zero on calls to jpeg_set_qual() when quality is zero.

As quality can't be 0 or lower than that, add an extra clause to cover
this special case.
Signed-off-by: Luis de Bethencourt <luis.bg@samsung.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Use the generic facility instead of a home-grown one.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Convert dprintk to netdev_dbg where appropriate.
Remove dvb_net_debug module_param.
Remove __func__ from output as that can be added by dynamic_debug.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

No need for more macros, so remove them and use the kernel extension.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Adjusts boolean assignments from 0/1 to false/true.

And accordingly, it also adjusts the if conditions.
Signed-off-by: Heba Aamer <heba93aamer@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Instead of implement its own logic, use the already-defined one.
Signed-off-by: Martin Kepplinger <martink@posteo.de>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

This patch is based on Antti's silabs branch.

Add support for 1.7 MHz bandwidth. Supported in all versions of Si2168 according to short data sheets.

[mchehab@osg.samsung.com: Fix CodingStyle]
Signed-off-by: Olli Salonen <olli.salonen@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

This patch should is based on Antti's silabs branch.

According to dvb-frontend.h set_frontend may be called with bandwidth_hz set to 0 if automatic bandwidth is required. Si2168 does not support automatic bandwidth and does not declare FE_CAN_BANDWIDTH_AUTO in caps.

This patch will change the behaviour in a way that EINVAL is returned if bandwidth_hz is 0.
Signed-off-by: Olli Salonen <olli.salonen@iki.fi>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

We have to check pointer for NULL and then dereference it.
Signed-off-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Signed-off-by: Benjamin Larsson <benjamin@southpole.se>
Reviewed-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

symbol-rate report was wrongly removed off by the commit:906aaf5a .
Signed-off-by: Akihiro Tsukada <tskd08@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

This allows calling the original functions providing the streaming is off.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Create dm04_read_status to check lock through either interrupt values
or directly by the call back.

When the device is not streaming the frontends original call back is
used. When streaming has started it turns off I2C messaging by
setting st->i2c_talk_onoff to zero. I2C can only be turn on again
by one of the other allowed frontend calls.

All old code is removed from lme2510_msg and this function only needs
to set st->i2c_talk_onoff to 1.

The lock status is saved and when the frondend is locked is maintained
by lme2510_int_response who will now just kill the lock.

The call back for rs2000 tuner is nologer required.

All frontend types have been tested.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Create call backs dm04_read_signal_strength, dm04_read_snr and
move dm04_read_ber and dm04_read_ucblocks for all frontends

Removing the I2C filtering from lme2510_msg and the old rs2000 callbacks.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

A quirk of some older firmwares that report endpoint pipe type as PIPE_BULK
but the endpoint otheriwse functions as interrupt.

Check if usb_endpoint_type is USB_ENDPOINT_XFER_BULK and set as usb_rcvbulkpipe.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Ocassionally the device fails to report back an interrupt urb status which
results in false no lock trigger on the RS2000 demodulator.

Increase time from 60 msecs to 200 msecs.
Signed-off-by: Malcolm Priestley <tvboxspy@gmail.com>
Cc: <stable@vger.kernel.org> # v3.17+
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

X-Patchwork-Delegate: m.chehab@samsung.com
Fix the following smatch warning:
	drivers/media/pci/cx88//cx88-dvb.c:1508 dvb_register() warn: if statement not indented
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

As complained by smatch:
	drivers/media/usb/dvb-usb-v2/rtl28xxu.c:1159 rtl2832u_tuner_attach() info: 'pdata' is not actually initialized (unreached code).

Cc: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

drivers/media/dvb-frontends/rtl2832.c:157:5: warning: no previous prototype for ‘rtl2832_bulk_write’ [-Wmissing-prototypes]
 int rtl2832_bulk_write(struct i2c_client *client, unsigned int reg,
     ^
drivers/media/dvb-frontends/rtl2832.c:169:5: warning: no previous prototype for ‘rtl2832_update_bits’ [-Wmissing-prototypes]
 int rtl2832_update_bits(struct i2c_client *client, unsigned int reg,
     ^
drivers/media/dvb-frontends/rtl2832.c:181:5: warning: no previous prototype for ‘rtl2832_bulk_read’ [-Wmissing-prototypes]
 int rtl2832_bulk_read(struct i2c_client *client, unsigned int reg, void *val,

Cc: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

drivers/media/dvb-frontends/rtl2830.c:21:5: warning: no previous prototype for ‘rtl2830_bulk_write’ [-Wmissing-prototypes]
 int rtl2830_bulk_write(struct i2c_client *client, unsigned int reg,
     ^
drivers/media/dvb-frontends/rtl2830.c:33:5: warning: no previous prototype for ‘rtl2830_update_bits’ [-Wmissing-prototypes]
 int rtl2830_update_bits(struct i2c_client *client, unsigned int reg,
     ^
drivers/media/dvb-frontends/rtl2830.c:45:5: warning: no previous prototype for ‘rtl2830_bulk_read’ [-Wmissing-prototypes]
 int rtl2830_bulk_read(struct i2c_client *client, unsigned int reg, void *val,
     ^

Cc: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Add kernel-doc comments for platform_data configuration structure.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Add kernel-doc comments for platform_data configuration structure.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Add kernel-doc comments for platform_data configuration structure.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Introduce own lock to silence lockdep warning. lockdep validator
makes wrong decision when two similar (&map->mutex) locks were taken
recursively, even those are different mutexes in a two different
drivers. After that patch, functionality remains same, but mutex names
are different. That is a temporary hack, proper solution is make
regmap aware of locked nested locking rules.

=============================================
[ INFO: possible recursive locking detected ]
3.18.0-rc4+ #4 Tainted: G           O
---------------------------------------------
kdvb-ad-0-fe-0/2814 is trying to acquire lock:
 (&map->mutex){+.+.+.}, at: [<ffffffff814ec90f>] regmap_lock_mutex+0x2f/0x40

but task is already holding lock:
 (&map->mutex){+.+.+.}, at: [<ffffffff814ec90f>] regmap_lock_mutex+0x2f/0x40

other info that might help us debug this:
 Possible unsafe locking scenario:
       CPU0
       ----
  lock(&map->mutex);
  lock(&map->mutex);

 *** DEADLOCK ***
 May be due to missing lock nesting notation
1 lock held by kdvb-ad-0-fe-0/2814:
 #0:  (&map->mutex){+.+.+.}, at: [<ffffffff814ec90f>] regmap_lock_mutex+0x2f/0x40

stack backtrace:
CPU: 3 PID: 2814 Comm: kdvb-ad-0-fe-0 Tainted: G           O 3.18.0-rc4+ #4
Hardware name: System manufacturer System Product Name/M5A78L-M/USB3, BIOS 2001    09/11/2014
 0000000000000000 00000000410c8772 ffff880293af3868 ffffffff817a6f82
 0000000000000000 ffff8800b3462be0 ffff880293af3968 ffffffff810e7f94
 ffff880293af3888 00000000410c8772 ffffffff82dfee60 ffffffff81ab8f89
Call Trace:
 [<ffffffff817a6f82>] dump_stack+0x4e/0x68
 [<ffffffff810e7f94>] __lock_acquire+0x1ea4/0x1f50
 [<ffffffff810e2a7d>] ? trace_hardirqs_off+0xd/0x10
 [<ffffffff817b01f3>] ? _raw_spin_lock_irqsave+0x83/0xa0
 [<ffffffff810e13e6>] ? up+0x16/0x50
 [<ffffffff810e2a7d>] ? trace_hardirqs_off+0xd/0x10
 [<ffffffff817af8bf>] ? _raw_spin_unlock_irqrestore+0x5f/0x70
 [<ffffffff810e9069>] lock_acquire+0xc9/0x170
 [<ffffffff814ec90f>] ? regmap_lock_mutex+0x2f/0x40
 [<ffffffff817ab50e>] mutex_lock_nested+0x7e/0x430
 [<ffffffff814ec90f>] ? regmap_lock_mutex+0x2f/0x40
 [<ffffffff814ec90f>] ? regmap_lock_mutex+0x2f/0x40
 [<ffffffff817a530b>] ? printk+0x70/0x86
 [<ffffffff8110d9e8>] ? mod_timer+0x168/0x240
 [<ffffffff814ec90f>] regmap_lock_mutex+0x2f/0x40
 [<ffffffff814f08d9>] regmap_update_bits+0x29/0x60
 [<ffffffffa03e9778>] rtl2832_select+0x38/0x70 [rtl2832]
 [<ffffffffa039b03d>] i2c_mux_master_xfer+0x3d/0x90 [i2c_mux]
 [<ffffffff815da493>] __i2c_transfer+0x73/0x2e0
 [<ffffffff815dbaba>] i2c_transfer+0x5a/0xc0
 [<ffffffff815dbb6e>] i2c_master_send+0x4e/0x70
 [<ffffffffa03ff25a>] regmap_i2c_write+0x1a/0x50 [regmap_i2c]
 [<ffffffff817ab713>] ? mutex_lock_nested+0x283/0x430
 [<ffffffff814f06b2>] _regmap_raw_write+0x862/0x880
 [<ffffffff814ec90f>] ? regmap_lock_mutex+0x2f/0x40
 [<ffffffff814f0744>] _regmap_bus_raw_write+0x74/0xa0
 [<ffffffff814ef3d2>] _regmap_write+0x92/0x140
 [<ffffffff814f0b7b>] regmap_write+0x4b/0x70
 [<ffffffffa032b090>] ? dvb_frontend_release+0x110/0x110 [dvb_core]
 [<ffffffffa05141d4>] e4000_init+0x34/0x210 [e4000]
 [<ffffffffa032a029>] dvb_frontend_init+0x59/0xc0 [dvb_core]
 [<ffffffff810bde30>] ? finish_task_switch+0x80/0x180
 [<ffffffff810bddf2>] ? finish_task_switch+0x42/0x180
 [<ffffffffa032b116>] dvb_frontend_thread+0x86/0x7b0 [dvb_core]
 [<ffffffff817a9203>] ? __schedule+0x343/0x930
 [<ffffffffa032b090>] ? dvb_frontend_release+0x110/0x110 [dvb_core]
 [<ffffffff810b826b>] kthread+0x10b/0x130
 [<ffffffff81020099>] ? sched_clock+0x9/0x10
 [<ffffffff810b8160>] ? kthread_create_on_node+0x250/0x250
 [<ffffffff817b063c>] ret_from_fork+0x7c/0xb0
 [<ffffffff810b8160>] ? kthread_create_on_node+0x250/0x250
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Use rtl28xxu_ prefix for all register access routine names.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

As all the callbacks are already same we could merge device
properties struct too and save space.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Merge all chip type specific prefixed (rtl2831u_ and rtl2832u_)
callback to top level callback prefixed as rtl28xxu_.

rtl2831u_foo() => rtl28xxu_foo()
rtl2832u_foo() => rtl28xxu_foo()
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Detect automatically whether chip is old RTL2831U or newer
RTL2832U/RTL2832P. Detection is based I2C command that is found only
from newer RTL2832U models.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Buffer needed for USB control message is small so move it to state
and get rid of alloc/free used for each control message.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

Pass correct device pointer to dev_* logging in order to print
module name and bus id correctly.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

I prefer dev over priv and I want keep all my drivers in line with
that.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

There was 2 muxed I2C adapters, one for demod tuner bus and one for
internal use. Idea of internal I2C adapter was to force I2C repeater
close when demod access its registers. Driver has also delayed work
queue based method to close I2C repeater.

After regmap conversion internal I2C adapter based repeater close
left unused - only work queue method was in use. We could not use
internal mux adapter method with regmap as it makes recursive regmap
call, which causes deadlock as regmap has own locking. Due to that
remove whole method totally.
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>

We are consumer of DVB frontend provided by rtl28xxu module. Due to
that we must use refcount to ensure none could remove rtl28xxu when
we are alive (or when we are streaming, if more fine-grained
refcounting is wanted).
Signed-off-by: Antti Palosaari <crope@iki.fi>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>