The drain and drain_notify callback were blocked by low level driver untill the
draining was complete. Due to this being invoked with big fat mutex held, others
ops like reading timestamp, calling pause, drop were blocked.

So to fix this we add a new snd_compr_drain_notify() API. This would be required
to be invoked by low level driver when drain or partial drain has been completed
by the DSP. Thus we make the drain and partial_drain callback as non blocking
and driver returns immediately after notifying DSP.
The waiting is done while relasing the lock so that other ops can go ahead.
Signed-off-by: NVinod Koul <vinod.koul@intel.com>
CC: stable@vger.kernel.org
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

It turned out that we can't use gen_pool_*() functions on archs
without CONFIG_GENERIC_ALLOCATOR (resulting in missing symbols), since
linux/genalloc.h doesn't provide dummy functions for all.  We'd be
able to fix linux/genalloc.h size, but I take an easier path for
now...
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

When HP laptops with mute and mic-record LEDs go to runtime suspend,
these LEDs are turned on forcibly no matter whether GPIO pis are on or
off.  This strange behavior seems triggered by resetting the HD-audio
bus link at azx_rutime_suspend().  So, just add a new hda_bus flag to
avoid the link reset at runtime suspend and set it for these HP
machines.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Yet another variants of ALC269.
Signed-off-by: NKailang Yang <kailang@realtek.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

It's just another variant of ALC269 & co.
Signed-off-by: NKailang Yang <kailang@realtek.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Emitting an OOM message isn't necessary after input_allocate_device
as there's a generic OOM and a dump_stack already done.
Signed-off-by: NJoe Perches <joe@perches.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Now it's quite common that an SoC contains its on-chip internal RAM.
By using this RAM space for DMA buffer during audio playback/record,
we can shutdown the voltage for external RAM to save power.

So add new DEV type with iram malloc()/free() and accordingly modify
current default mmap() for the iram circumstance.
Signed-off-by: NNicolin Chen <b42378@freescale.com>
Reviewed-by: NLars-Peter Clausen <lars@metafoo.de>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

... which was introduced by the previous commit a4e9a38b, causing
build errors without CONFIG_PROC_FS.
Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This patch adds codec ID (0x80862882) and module alias for
Valleyview2 display codec.
Signed-off-by: NMengdong Lin <mengdong.lin@intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

At the moment, this driver supports only playback, while FFADO supports
(only) full-duplex devices.  So, prevent conflicts by not claiming
devices that would be better handled by FFADO.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

While most DICE devices keep TCAT's default category ID of 0x04, Weiss
devices identify themselves with 0x00.
Reported-by: NRolf Anderegg <rolf.anderegg@weiss.ch>
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Remove a wrong typecast that resulted from a copy-and-paste error.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Doing accesses without quadlet alignment is a bad idea because the
firmware's byte-swapping would garble the data; clarify this in the
documentation.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

For easier debugging, add a proc file to show the device's capabilities
and current status.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Output a warning if the wait for the clock change notification times
out.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Instead of relying of some control panel application to configure some
fixed sample rate, allow applications to set it automatically.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

In preparation for sample rate selection support, read the stream
parameters that might change when running at different sample rates.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Reorganize the initialization order so that the driver can receive
notifications earlier.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

In preparation for sample rate selection support, ensure that the driver
knows about the device's clock capabilities.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Avoid a lock inversion between dice->mutex and pcm->open_mutex.
Signed-off-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

The pcm field was not actually used.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Since commit f2b3614c (Don't check DMA time-out too shortly), we
need no longer to restrict the period length to less than 10 s.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

When a bus reset happens, the enable register is automatically cleared,
so we do not need to clear it manually when stopping the stream.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Add a flag to snd_fw_transaction() to allow it to abort when a bus reset
happens.  This removes most of the duplicated error handling loops that
were required around calls to the low-level fw_run_transaction().

Also add a flag to suppress error messages; errors are expected when we
attempt to clean up after the device was unplugged.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Instead of reading two consecutive register with two quadlet requests,
use one block read request.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Change the AMDTP streaming code to handle the non-standard stream format
that DICE devices use at sample rates greater than 96 kHz.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

DICE devices do not have a unique specifier ID in their unit directory
(it's always the same as the device vendor's ID), so rely on just the
version ID for driver loading, and use a heuristic in the probe callback
to detect actual DICE devices.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Move the DICE interface symbols into a separate header file, and add
more documentation.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Introduce the helper function amdtp_out_stream_running().  This makes
many checks in amdtp.c clearer and frees the device drivers from having
to track this with a separate variable.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Ensure that misbehaving or aborted userspace programs do not
accidentally keep the lock.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Implement the hwdep locking and notification mechanisms.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

When aborting a PCM stream, the xrun is signaled only if the stream is
running.  When disconnecting a PCM stream, calling snd_card_disconnect()
too early would change the stream into a non-running state and thus
prevent the xrun from being noticed by user space.

To prevent this, move the snd_card_disconnect() call after the xrun.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Allow AMDTP output streams to use blocking mode.

Use it for DICE devices, because the old DICE-II chip will in some cases
not be able to lock to non-blocking streams (erratum E7).
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

The notification bits are not of general interest; log them only when
debugging.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Instead of forcing a constant 44.1 kHz, read the current sample rate
from the device when opening the PCM device.

Actually changing the sample rate requires some separate controller
application.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

After a bus reset, do not stop the stream completely to avoid having to
reconfigure the device when restarting the stream.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Partially restructures _snd_emu10k1_audigy_init_efx() and
_snd_emu10k1_init_efx() functions.

Be noted that the cast is demanded to use '__user'. So, in these cases,
avoid patches based on the coccinelle 'drop_kmalloc_cast' semantic patch.
Signed-off-by: NGeyslan G. Bem <geyslan@gmail.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

As a start point for further development, this is an incomplete driver
for DICE devices:
- only playback (so no clock source except the bus clock)
- only 44.1 kHz
- no MIDI
- recovery after bus reset is slow
- hwdep device is created, but not actually implemented

Contains compilation fixes by Stefan Richter.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Since the lock is used primarily in patch_hdmi.c, it's better to move
it in the local struct instead of exporting in hda_eld.  The only
functions requiring the lock in hda_eld.c are proc accessors.  So in
this patch, the proc entry and its creation/deletion/accessors are
moved into patch_hdmi.c, together with the mutex lock to pin_spec
struct.

The former proc info functions are exported so that they can be called
from patch_hdmi.c.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>