Once allocated, isochronous resources are available for packet
streaming, even if the streaming is cancelled. For this reason,
current implementation handles allocation of the resources and
starting packet streaming at the same time. However, this brings
complicated procedure to start packet streaming.

This commit separates the allocation and starting. The allocation is
done in pcm.hw_params callback and available till pcm.hw_free callback.
Even if any XRUN occurs, pcm.prepare callback is done to restart
packet streaming for allocated the resources.

There are two points to stop packet streaming; in pcm.hw_params and
pcm.prepare callbacks.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a part of preparation to perform allocation/release
of isochronous channels in pcm.hw_params/hw_free callbacks.

This commit implements allocate_resources callback for the protocol
specific to latter models. The encoded values of constant table is
split into several condition statements to separate the operation
to configure sampling transfer frequency from the operation to
configure the number of data channels in rx packet.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a part of preparation to perform allocation/release
of isochronous channels in pcm.hw_params/hw_free callbacks.

This commit implements allocate_resources callback for ff400 protocol.
In this callback, sampling transfer frequency is configured to the
device as well.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a part of preparation to perform allocation/release
of isochronous channels in pcm.hw_params/hw_free callbacks.

This commit implements allocate_resources callback for ff800 protocol.
As I noted in commit fc716397 ("ALSA: fireface: add support for
packet streaming on Fireface 800"), this unit allocates isochronous
resources for tx stream voluntarily. Therefore, this commit is to
maintain isochronous rsources for rx stream. In the callback,
sampling transfer frequency is configured to the device as well.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a part of preparation to perform allocation/release
of isochronous channels in pcm.hw_params/hw_free callbacks.

In ALSA fireface driver, the allocation of isochronous resources is
programmed in each implementation of protocol. This commit adds
protocol-specific operation for the allocation separated from
the operation to begin session.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The pairs of pcm.hw_params callbacks and .hw_free callbacks for both
direction have no differences.

This commit unifies the pairs.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The operation to finish packet streaming corresponds to stopping
isochronous contexts.

This commit applies code refactoring to move codes to stop into
a helper function to finish the session.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

After bus reset, isochronous resource manager releases all of allocated
isochronous resources. The nodes to restart packet streaming should
request reallocation of the resources.

However, between the bus-reset and invocation of 'struct fw_driver.update'
handler, ALSA PCM application can detect this situation by XRUN because
the target device cancelled to transmit packets once bus-reset occurs.

Due to the above mechanism, ALSA firewire-tascam driver just stops
packet streaming in the update handler, thus pcm.prepare handler
should request the reallocation.

This commit requests the reallocation in pcm.prepare callback when
bus generation is changed.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Once allocated, isochronous resources are available for packet
streaming, even if the streaming is cancelled. For this reason,
current implementation handles allocation of the resources and
starting packet streaming at the same time. However, this brings
complicated procedure to start packet streaming.

This commit separates the allocation and starting. The allocation is
done in pcm.hw_params callback and available till pcm.hw_free callback.
Even if any XRUN occurs, pcm.prepare callback is done to restart
packet streaming for allocated the resources.

There are two points to stop packet streaming; in pcm.hw_params and
pcm.prepare callbacks.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a part of preparation to perform allocation/release
of isochronous channels in pcm.hw_params/hw_free callbacks.

This commit obsoletes a helper function to release isochronous resources
for both direction.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a part of preparation to perform allocation/release
of isochronous channels in pcm.hw_params/hw_free callbacks.

This commit applies minor code refactoring for a helper function to
allocate isochronous resources.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a part of preparation to perform allocation/release
of isochronous channels in pcm.hw_params/hw_free callbacks.

The registration of isochronous channels is done just after allocation
of isochronous resources. This commit separates the registration just
before starting packet streaming.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Fixes gcc '-Wunused-but-set-variable' warning:

sound/pci/lx6464es/lx_core.c: In function 'lx_interrupt_handle_async_events':
sound/pci/lx6464es/lx_core.c:990:6: warning:
 variable 'urun_mask' set but not used [-Wunused-but-set-variable]
sound/pci/lx6464es/lx_core.c:989:6: warning:
 variable 'orun_mask' set but not used [-Wunused-but-set-variable]

They are never used, so can be removed.
Signed-off-by: NYueHaibing <yuehaibing@huawei.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

For code simplification and safety, use struct_size() macro for
calculating the dsp_image_seg object size.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

For code simplification and safety, use struct_size() macro for
calculating the hda_conn_list object size with the variable array.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

For code simplification and safety, use struct_size() macro for
calculating the snd_kcontrol object size with the variable array.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

A follow-up fix for the HD-audio previous polling mode changes.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The patch is to fix commit 5e13cf6c
(ALSA: hda: add polling mode in snd_hdac_bus_get_response)
spin_lock_irq should be called before snd_hdac_bus_update_rirb.

Fixes: 5e13cf6c ("ALSA: hda: add polling mode in snd_hdac_bus_get_response")
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Pull LINE6 driver cleanups and fixes.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

... instead of unconditional cast.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Now all timer usages in line6 drivers are gone, we can get rid of some
helper macro and function that became superfluous.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Variax driver had a very complex and staged startup sequence using
multiple timers and a work.  This patch simplifies the procedure to a
single delayed work.

Now the startup stage consists of:
- VARIAX_STARTUP_VERSIONREQ:
  requesting the version and the message handler raises up to the next
  stage upon receiving the reply.  The request is repeated until a
  reply arrives.
- VARIAX_STARTUP_ACTIVATE:
  does activation, and queue for the next stage.
- VARIAX_STARTUP_SETUP:
  registers the card.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

POD HD driver had a complex staged startup sequence with both timer
and work.  This patch simplifies it to a single delayed work with a
single stage.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The POD driver had a complex staged startup procedure using both timer
and work.  This patch simplifies it via a single delayed work with the
reduced stages.

Now basically only two intermediate stages:
- POD_STARTUP_VERSIONREQ:
  requesting the version information and the process_message callback
  triggers the next stage,
- POD_STARTUP_SETUP:
  registering the actual card object.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The current code performs the cancel of a delayed work at the late
stage of disconnection procedure, which may lead to the access to the
already cleared state.

This patch assures to call cancel_delayed_work_sync() at the beginning
of the disconnection procedure for avoiding that race.  The delayed
work object is now assigned in the common line6 object instead of its
derivative, so that we can call cancel_delayed_work_sync().

Along with the change, the startup function is called via the new
callback instead.  This will make it easier to port other LINE6
drivers to use the delayed work for startup in later patches.

Reported-by: syzbot+5255458d5e0a2b10bbb9@syzkaller.appspotmail.com
Fixes: 7f84ff68 ("ALSA: line6: toneport: Fix broken usage of timer for delayed execution")
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Pull the HD-audio polling mode changes, moving the stuff into HD-audio
core for re-using it in ASoC.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

There is a workaround in legacy HDA codec for too long time respone
with CFL machine. We need the same workaround on SOF driver. The same
issue is also seen on CNL machine.
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

We observed the same issue as reported by commit a8d7bde2
("ALSA: hda - Force polling mode on CFL for fixing codec communication")
We don't have a better solution. So apply the same workaround to CNL.
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Polling mode is useful if a machine somehow missed an expected IRQ.
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

polling mode is a useful function in the get_response function. Move
polling_mode flag from struct azx to struct hdac_bus so people can
implement polling mode in their own get_response function without
adding a polling_mode flag in their local chip structure.
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

We will move the polling_mode flag from struct azx to struct hdac_bus,
and the flag should be assigned after bus init.
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

As a result of heavy refactoring based on IR context header, the packet
handler becomes simpler.

This commit merges the packet handler into function for IR context
callback. The logic to parse IR context header and tracepoints event is
split to a function.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

As a result of heavy refactoring based on IT packet header, the packet
handler becomes simpler.

This commit merges the packet handler into function for IT context
callback. The logic to build IT packet header and tracepoints event is
split to a function.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The amdtp_packet events have inverted node IDs for src/dst. This commit
fixes the bug.

Fixes: 8d3f1fdf ("ALSA: firewire-lib: unify tracing events to 'amdtp_packet' event")
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The value of data block counter is not calculated for incoming packet
without CIP header. This commit fixes the bug.

Fixes: 947b437e ("ALSA: firewire-lib: unify packet handler for IR context")
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

In Linux firewire subsystem, for IT context, some quadlets of isochronous
packet payload can be indicated as a part of packet header to queue to
the context.

This commit uses the packet header to split CIP headers from CIP
payload. As a result, regardless of CIP or non-CIP, context payload
includes data blocks only.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This commit is a preparation to queue IT packet with header. To enable
packet handler to fill the header, this commit uses kernel stack for
data structure of packet parameter in several part of this file.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The handlers for packet with CIP and without CIP include common codes.
This commit unifies them and remove an member for pointer to callback
function from data structure.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This is minor code refactoring to split a function to generate CIP
header.
Signed-off-by: NTakashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Since this LED is found on all Huawei laptops, we can hook it to
huawei-wmi platform driver to control it.

Also, some renames have been made to use product name instead of common
name to avoid confusions.

Fixes: 8ac51bbc ("ALSA: hda: fix front speakers on Huawei MBXP")
Signed-off-by: NAyman Bagabas <ayman.bagabas@gmail.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>