When device_add is called in rc_register_device, the rc sysfs nodes show
up, and there's a window in which ir-keytable can be launched via udev
and trigger a show_protocols call, which runs without various rc_dev
fields filled in yet. Add some locking around registration and
store/show_protocols to prevent that from happening.

The problem manifests thusly:

[64692.957872] BUG: unable to handle kernel NULL pointer dereference at 0000000000000090
[64692.957878] IP: [<ffffffffa036a4c1>] show_protocols+0x47/0xf1 [rc_core]
[64692.957890] PGD 19cfc7067 PUD 19cfc6067 PMD 0
[64692.957894] Oops: 0000 [#1] SMP
[64692.957897] last sysfs file: /sys/devices/pci0000:00/0000:00:03.1/usb3/3-1/3-1:1.0/rc/rc2/protocols
[64692.957902] CPU 3
[64692.957903] Modules linked in: redrat3(+) ir_lirc_codec lirc_dev ir_sony_decoder ir_jvc_decoder ir_rc6_decoder ir_rc5_decoder rc_hauppauge ir_nec
_decoder rc_core ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 ip6table_filter ip6_tables snd_emu10k1_synth snd_emux_synth snd_seq_virmidi snd_seq_mi
di_event snd_seq_midi_emul snd_emu10k1 snd_rawmidi snd_ac97_codec ac97_bus snd_seq snd_pcm snd_seq_device snd_timer snd_page_alloc snd_util_mem pcsp
kr tg3 snd_hwdep emu10k1_gp snd amd64_edac_mod gameport edac_core soundcore edac_mce_amd k8temp shpchp i2c_piix4 lm63 e100 mii uinput ipv6 raid0 rai
d1 ata_generic firewire_ohci pata_acpi firewire_core crc_itu_t sata_svw pata_serverworks floppy radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core
[last unloaded: redrat3]
[64692.957949] [64692.957952] Pid: 12265, comm: ir-keytable Tainted: G   M    W   2.6.39-rc6+ #2 empty empty/TYAN Thunder K8HM S3892
[64692.957957] RIP: 0010:[<ffffffffa036a4c1>]  [<ffffffffa036a4c1>] show_protocols+0x47/0xf1 [rc_core]
[64692.957962] RSP: 0018:ffff880194509e38  EFLAGS: 00010202
[64692.957964] RAX: 0000000000000000 RBX: ffffffffa036d1e0 RCX: ffffffffa036a47a
[64692.957966] RDX: ffff88019a84d000 RSI: ffffffffa036d1e0 RDI: ffff88019cf2f3f0
[64692.957969] RBP: ffff880194509e68 R08: 0000000000000002 R09: 0000000000000000
[64692.957971] R10: 0000000000000002 R11: 0000000000001617 R12: ffff88019a84d000
[64692.957973] R13: 0000000000001000 R14: ffff8801944d2e38 R15: ffff88019ce5f190
[64692.957976] FS:  00007f0a30c9a720(0000) GS:ffff88019fc00000(0000) knlGS:0000000000000000
[64692.957979] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[64692.957981] CR2: 0000000000000090 CR3: 000000019a8e0000 CR4: 00000000000006e0
[64692.957983] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[64692.957986] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[64692.957989] Process ir-keytable (pid: 12265, threadinfo ffff880194508000, task ffff88019a9fc720)
[64692.957991] Stack:
[64692.957992]  0000000000000002 ffffffffa036d1e0 ffff880194509f58 0000000000001000
[64692.957997]  ffff8801944d2e38 ffff88019ce5f190 ffff880194509e98 ffffffff8131484b
[64692.958001]  ffffffff8118e923 ffffffff810e9b2f ffff880194509e98 ffff8801944d2e18
[64692.958005] Call Trace:
[64692.958014]  [<ffffffff8131484b>] dev_attr_show+0x27/0x4e
[64692.958014]  [<ffffffff8118e923>] ? sysfs_read_file+0x94/0x172
[64692.958014]  [<ffffffff810e9b2f>] ? __get_free_pages+0x16/0x52
[64692.958014]  [<ffffffff8118e94c>] sysfs_read_file+0xbd/0x172
[64692.958014]  [<ffffffff8113205e>] vfs_read+0xac/0xf3
[64692.958014]  [<ffffffff8113347b>] ? fget_light+0x3a/0xa1
[64692.958014]  [<ffffffff811320f2>] sys_read+0x4d/0x74
[64692.958014]  [<ffffffff814c19c2>] system_call_fastpath+0x16/0x1b

Its a bit difficult to reproduce, but I'm fairly confident this has
fixed the problem.
Signed-off-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The ene_ir driver was using a private define of MS_TO_NS, which is meant
to be microseconds to nanoseconds. The mceusb driver copied it,
intending to use is a milliseconds to microseconds. Lets move the
defines to a common location, expand and standardize them a touch, so
that we now have:

  MS_TO_NS - milliseconds to nanoseconds
  MS_TO_US - milliseconds to microseconds
  US_TO_NS - microseconds to nanoseconds
Reported-by: NDavid Härdeman <david@hardeman.nu>
CC: Maxim Levitsky <maximlevitsky@gmail.com>
Signed-off-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Fix some minor comments etc which are leftover from the old naming scheme.
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The struct that describes a rc mapping had an weird and long name.
We should properly name it, to make easier for developers to work
with it, and to avoid confusion.

Basically, generated by this script:

for i in `find drivers/staging -type f -name *.[ch]` `find include/media -type f -name *.[ch]` `find drivers/media -type f -name *.[ch]`; do sed s,ir_scancode_table,rc_map,g <$i >a && mv a $i; done
for i in `find drivers/staging -type f -name *.[ch]` `find include/media -type f -name *.[ch]` `find drivers/media -type f -name *.[ch]`; do sed s,rc_tab,rc_map,g <$i >a && mv a $i; done

(and manually fixed where needed)
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

for i in `find drivers/staging -type f -name *.[ch]` `find include/media -type f -name *.[ch]` `find drivers/media -type f -name *.[ch]`; do sed s,IR_TYPE,RC_TYPE,g <$i >a && mv a $i; done
for i in `find drivers/staging -type f -name *.[ch]` `find include/media -type f -name *.[ch]` `find drivers/media -type f -name *.[ch]`; do sed s,ir_type,rc_type,g <$i >a && mv a $i; done
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Those functions are not InfraRed specific. So, rename them to properly
reflect it.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The Remote Controller subsystem is meant to be used not only by Infra Red
but also for similar types of Remote Controllers. The core is not specific
to Infra Red. As such, rename:
	- ir-core.h to rc-core.h
	- IR_CORE to RC_CORE
	- namespace inside rc-core.c/rc-core.h

To be consistent with the other changes.

No functional change on this patch.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch merges the ir_input_dev and ir_dev_props structs into a single
struct called rc_dev. The drivers and various functions in rc-core used
by the drivers are also changed to use rc_dev as the primary interface
when dealing with rc-core.

This means that the input_dev is abstracted away from the drivers which
is necessary if we ever want to support multiple input devs per rc device.

The new API is similar to what the input subsystem uses, i.e:
rc_device_alloc()
rc_device_free()
rc_device_register()
rc_device_unregister()

[mchehab@redhat.com: Fix compilation on mceusb and cx231xx, due to merge conflicts]
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Acked-by: NJarod Wilson <jarod@redhat.com>
Tested-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

[mchehab@redhat.com: this patch were originally bundled with some renaming
stuff and with the file merges, as seen at:
https://patchwork.kernel.org/patch/291092/.

Instead of using the original approach, I wrote the rename patches and the
code merge as separate changes, then applied the difference on this patch.
This way, it is easier to see the real changes at the code, and will be easier
to merge upstream, especially if some conflict rises on the renaming patches]
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Acked-by: NJarod Wilson <jarod@redhat.com>
Tested-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

cx88 only depends on VIDEO_IR because it needs ir_extract_bits().
Move that function to ir-core.h and make it inline.

Lots of drivers had dependencies on VIDEO_IR when they really
wanted IR_CORE.

The only remaining drivers to depend on VIDEO_IR are bt8xx and
saa7134 (ir_rc5_timer_end is the only function exported by
ir-functions).

Rename VIDEO_IR -> IR_LEGACY to give a hint to anyone writing or
converting drivers to IR_CORE that they do not want a dependency
on IR_LEGACY.
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Acked-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch removes the remaining usages of the ir_input_nokey() and
ir_input_keydown() functions provided by drivers/media/IR/ir-functions.c
by using the corresponding functionality in ir-core instead.
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Acked-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Add new event types for timeout & carrier report
Move timeout handling from ir_raw_event_store_with_filter to
ir-lirc-codec, where it is really needed.
Now lirc bridge ensures proper gap handling.
Extend lirc bridge for carrier & timeout reports

Note: all new ir_raw_event variables now should be initialized
like that: DEFINE_IR_RAW_EVENT(ev);

To clean an existing event, use init_ir_raw_event(&ev);
Signed-off-by: NMaxim Levitsky <maximlevitsky@gmail.com>
Acked-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The imon driver currently reimplements its own version of ir_keyup
(along with key release timer functionality also already present in the
core IR code). A follow-up imon patch will make use of ir_keyup and the
IR stack's key release code.

Trivial extraction from David Härdeman's pending rc-core merge and
device interface abstraction patchset to facilitate merging a patch
based on his imon input dev split patch ahead of the larger churn, which
is slated for post-2.6.37-rc1 (after Dmitry's large keycode patches are
merged in mainline).
Signed-off-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

LIRC: add new IOCTL that enables learning mode (wide band receiver)
Still missing features: carrier report & timeout reports.
Will need to pack these into ir_raw_event
Signed-off-by: NMaxim Levitsky <maximlevitsky@gmail.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Some ir input devices have small buffer, and interrupt the host
each time it is full (or half full)

Add a helper that automaticly handles timeouts, and also
automaticly merges samples of same time (space-space)
Such samples might be placed by hardware because size of
sample in the buffer is small (a byte for example).

Also remove constness from ir_dev_props, because it now contains timeout
settings that driver might want to change
Signed-off-by: NMaxim Levitsky <maximlevitsky@gmail.com>
Acked-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Currently, ir device registration fails if keymap requested by driver is not found.
Fix that by always compiling in the empty keymap, and using it as a failback.
Signed-off-by: NMaxim Levitsky <maximlevitsky@gmail.com>
Acked-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

incoming IR buffer now an int pointer, and not fed from userspace
Signed-off-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This patch implements the agreed upon 1:31 integer encoded pulse/duration
struct for ir-core raw decoders. All decoders have been tested after the
change. Comments are welcome.
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The imon driver I've previously submitted and have been porting to
use ir-core needs to use ir_g_keycode_from_table, as ir_keydown is
not sufficient, due to these things having really oddball hardware
decoders in them. This just moves the function declaration from
ir-core-priv.h over to ir-core.h.
Signed-off-by: NJarod Wilson <jarod@redhat.com>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

ir-core.h has the kABI to be used by the bridge drivers, when needing to register
IR protocols and pass IR events. However, the same file also contains IR subsystem
internal calls, meant to be used inside ir-core and between ir-core and the raw
decoders.

Better to move those functions to an internal header, for some reasons:

1) Header will be a little more cleaner;

2) It avoids the need of recompile everything (bridge/hardware drivers, etc),
   just because a new decoder were added, or some other internal change were needed;

3) Better organize the ir-core API, splitting the functions that are internal to
   IR core and the ancillary drivers (decoders, lirc_dev) from the features that
   should be exported to IR subsystem clients.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

drivers/media/IR/ir-raw-event.c is currently written with the assumption
that all "raw" hardware will generate events only on state change (i.e.
when a pulse or space starts).

However, some hardware (like mceusb, probably the most popular IR receiver
out there) only generates duration data (and that data is buffered so using
any kind of timing on the data is futile).

Furthermore, using signed int's to represent pulse/space durations is a
well-known approach when writing ir decoders.

With this patch:

- s64 int's are used to represent pulse/space durations in ns

- a workqueue is used to decode the ir protocols outside of interrupt context

- #defines are added to make decoders clearer

- decoder reset is implemented by passing a zero duration to the kfifo queue
  and decoders are updated accordingly
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Some devices have in-hardware Remote Controller decoder, while others
need a software decoder to get the IR code. As each software decoder
can be enabled/disabled individually, allowing multiple protocol
decoding capability.

On the other hand, hardware decoders have a limited protocol
support, often being able of decoding just one protocol each time.
So, each type needs a different set of capabilities to control the
supported protocol(s).
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

A few hardware Remote Controller decoders, even using a standard protocol,
aren't able to provide the entire scancode. Due to that, the capability
of using other IR's are limited on those hardware.

Adds a way to indicate to ir-core what are the bits that the hardware
provides, from a scancode, allowing the addition of a complete IR table
to the kernel and allowing a limited support for changing the Remote
Controller on those devices.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The keymaps don't need to be recompiled every time a change at ir-core.h
happens, since it only depends on rc-map defines. By moving those
definitions to the proper header, the code became cleaner, and avoids
needing to recompile all the RC maps every time a non-related change
is introduced.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Now that the decoders are state machine, there's no need to create
an ancillary buffer while decoding the protocol. Just call the decoders
code directly, event by event.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

This decoder is also based on a state machine, just like the NEC protocol
decoder. It is pedantic in the sense that accepts only 14 bits. As there
are some variants that outputs less bits, it needs to be improved to also
handle those.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Rewrites the keyup/keydown logic in drivers/media/IR/ir-keytable.c.

All knowledge of keystates etc is now internal to ir-keytable.c
and not scattered around ir-raw-event.c and ir-nec-decoder.c (where
it doesn't belong).

In addition, I've changed the API slightly so that ir_input_dev is
passed as the first argument rather than input_dev. If we're ever
going to support multiple keytables we need to move towards making
ir_input_dev the main interface from a driver POV and obscure away
the input_dev as an implementational detail in ir-core.
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

The attached patch rewrites much of the keytable code in
drivers/media/IR/ir-keytable.c.

The scancodes are now inserted into the array in sorted
order which allows for a binary search on lookup.

The code has also been shrunk by about 150 lines.

In addition it fixes the following bugs:

Any use of ir_seek_table() was racy.

ir_dev->driver_name is leaked between ir_input_register() and
ir_input_unregister().

ir_setkeycode() unconditionally does clear_bit() on dev->keybit
when removing a mapping, but there might be another mapping with
a different scancode and the same keycode.

This version has been updated to incorporate patch feedback from
Mauro Carvalho Chehab.

[mchehab@redhat.com: Fix a conflict with RC keytable breakup patches and input changes]
Signed-off-by: NDavid Härdeman <david@hardeman.nu>
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Now that the remote keymaps were broken into separate modules,
get rid of the keycode tables that were hardcoded into ir-common.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Instead of using the ugly keymap sequences, use the new rc-*.ko keymap
files. For now, it is still needed to have one keymap loaded, for the
RC code to work. Later patches will remove this depenency.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

A latter patch will reuse the ir_input_register with a different meaning.
Before it, change all occurrences to a temporary name.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Instead of having all RC tables hardcoded on one file with
all tables there, add infrastructure for registering and dynamically
load the table(s) when needed.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Instead of having one big file with lots of keytables, create one include
file for each IR keymap.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Especially when IR needs to do polling, it generates lots of wakeups per
second. This makes no sense, if the input event device is closed.

Adds a callback handler to the IR hardware driver, to allow registering
an open/close ops.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Some decoders and a lirc_dev interface may need some other operations to work.
For example: IR device register/unregister and ir_keydown events may need to
be tracked.

As some operations can occur in interrupt time, and a lock is needed to prevent
un-registering a decode while decoding a key, the lock needed to be convert
into a spin lock.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Instead of hardcoding the protocols into ir-core, add a register interface
for the IR protocol decoders, and convert ir-nec-decoder into a client of
ir-core.

With this approach, it is possible to dynamically load the needed IR protocols,
and to add a RAW IR interface module, registered as one IR raw protocol decoder.

This patch opens a way to register a lirc_dev interface to work as an userspace
IR protocol decoder.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

At raw_decode mode, the key is processed after the end of a timer. The
previous code resets the timer every time something is received at the IR
port. While this works fine with IR's that don't implement repeat, like
Avermedia RM-JX IR, it keeps waiting until keydown, on IR's that implement
NEC repeat command, like the Terratec yellow.

The solution is to change the behaviour to do the timeout after the first
received data.

The timeout is currently set to 15 ms, as it works fine with NEC protcocol.
It may need some adjustments to support other protocols and to better handle
spurious detections that may happen with some IR sensors.
Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>

Signed-off-by: NMauro Carvalho Chehab <mchehab@redhat.com>