GitHub/LineageOS/G12/android_kernel_amlogic_linux-4.9.git
8 years agoALSA: timer: Introduce disconnect op to snd_timer_instance
Takashi Iwai [Thu, 21 Jan 2016 16:43:08 +0000 (17:43 +0100)]
ALSA: timer: Introduce disconnect op to snd_timer_instance

Instead of the previous ugly hack, introduce a new op, disconnect, to
snd_timer_instance object for handling the wake up of pending tasks
more cleanly.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=109431
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: timer: Handle disconnection more safely
Takashi Iwai [Thu, 21 Jan 2016 16:19:31 +0000 (17:19 +0100)]
ALSA: timer: Handle disconnection more safely

Currently ALSA timer device doesn't take the disconnection into
account very well; it merely unlinks the timer device at disconnection
callback but does nothing else.  Because of this, when an application
accessing the timer device is disconnected, it may release the
resource before actually closed.  In most cases, it results in a
warning message indicating a leftover timer instance like:
   ALSA: timer xxxx is busy?
But basically this is an open race.

This patch tries to address it.  The strategy is like other ALSA
devices: namely,
- Manage card's refcount at each open/close
- Wake up the pending tasks at disconnection
- Check the shutdown flag appropriately at each possible call

Note that this patch has one ugly hack to handle the wakeup of pending
tasks.  It'd be cleaner to introduce a new disconnect op to
snd_timer_instance ops.  But since it would lead to internal ABI
breakage and it eventually increase my own work when backporting to
stable kernels, I took a different path to implement locally in
timer.c.  A cleanup patch will follow at next for 4.5 kernel.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=109431
Cc: <stable@vger.kernel.org> # v3.15+
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - Flush the pending probe work at remove
Takashi Iwai [Wed, 20 Jan 2016 16:19:02 +0000 (17:19 +0100)]
ALSA: hda - Flush the pending probe work at remove

As HD-audio driver does deferred probe internally via workqueue, the
driver might go into the mixed state doing both probe and remove when
the module gets unloaded during the probe work.  This eventually
triggers an Oops, unsurprisingly.

For avoiding this race, we just need to flush the pending probe work
explicitly before actually starting the resource release.

Bugzilla: https://bugzilla.opensuse.org/show_bug.cgi?id=960710
Cc: <stable@vger.kernel.org> # v3.17+
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - Fix missing module loading with model=generic option
Takashi Iwai [Wed, 20 Jan 2016 14:46:44 +0000 (15:46 +0100)]
ALSA: hda - Fix missing module loading with model=generic option

When the generic codec driver is specified via model option or such,
the hda driver doesn't try to load the generic driver module but still
loads the codec-specific driver, and this ends up with the binding
failure.

This patch fixes it by moving the generic module request in the common
helper code.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=111021
Cc: <stable@vger.kernel.org> # v4.4+
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - Degrade i915 binding failure message
Takashi Iwai [Wed, 20 Jan 2016 14:00:26 +0000 (15:00 +0100)]
ALSA: hda - Degrade i915 binding failure message

Currently HD-audio driver on Intel Skylake or Broxteon gives an error
message when binding with i915 audio component fails.  However, this
isn't any serious error on a system without Intel graphics.  Indeed
there are such systems, where a third-party codec (e.g. Creative) is
put on the mobo while using other discrete GPU (e.g. Nvidia).
Printing a kernel "error" message is overreaction in such a case.

This patch downgrades the print level for that message.  For systems
that mandate the i915 binding (e.g. Haswell or Broadwell HDMI/DP),
another kernel error message is shown in addition to make clear what
went wrong.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=111021
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: at73c213: manage SSC clock
Mans Rullgard [Sun, 17 Jan 2016 22:31:02 +0000 (22:31 +0000)]
ALSA: at73c213: manage SSC clock

Since commit 49af54ff0fc7 ("misc: atmel-ssc: prepare clock only when
request") the SSC driver expects clients to manage the SSC clock.
This updates the at73c213 driver to enable the SSC clock whenever it is
needed.

Signed-off-by: Mans Rullgard <mans@mansr.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: control: Avoid kernel warnings from tlv ioctl with numid 0
Takashi Iwai [Mon, 18 Jan 2016 13:12:40 +0000 (14:12 +0100)]
ALSA: control: Avoid kernel warnings from tlv ioctl with numid 0

When a TLV ioctl with numid zero is handled, the driver may spew a
kernel warning with a stack trace at each call.  The check was
intended obviously only for a kernel driver, but not for a user
interaction.  Let's fix it.

This was spotted by syzkaller fuzzer.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: seq: Fix snd_seq_call_port_info_ioctl in compat mode
Nicolas Boichat [Mon, 18 Jan 2016 13:35:01 +0000 (21:35 +0800)]
ALSA: seq: Fix snd_seq_call_port_info_ioctl in compat mode

This reverts one hunk of
commit ef44a1ec6eee ("ALSA: sound/core: use memdup_user()"), which
replaced a number of kmalloc followed by memcpy with memdup calls.

In this case, we are copying from a struct snd_seq_port_info32 to a
struct snd_seq_port_info, but the latter is 4 bytes longer than the
32-bit version, so we need to separate kmalloc and copy calls.

Fixes: ef44a1ec6eee ('ALSA: sound/core: use memdup_user()')
Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: pcm: Fix snd_pcm_hw_params struct copy in compat mode
Nicolas Boichat [Mon, 18 Jan 2016 13:35:00 +0000 (21:35 +0800)]
ALSA: pcm: Fix snd_pcm_hw_params struct copy in compat mode

This reverts one hunk of
commit ef44a1ec6eee ("ALSA: sound/core: use memdup_user()"), which
replaced a number of kmalloc followed by memcpy with memdup calls.

In this case, we are copying from a struct snd_pcm_hw_params32 to
a struct snd_pcm_hw_params, but the latter is 4 bytes longer than
the 32-bit version, so we need to separate kmalloc and copy calls.

This actually leads to an out-of-bounds memory access later on
in sound/soc/soc-pcm.c:soc_pcm_hw_params() (detected using KASan).

Fixes: ef44a1ec6eee ('ALSA: sound/core: use memdup_user()')
Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hrtimer: Fix stall by hrtimer_cancel()
Takashi Iwai [Mon, 18 Jan 2016 12:52:47 +0000 (13:52 +0100)]
ALSA: hrtimer: Fix stall by hrtimer_cancel()

hrtimer_cancel() waits for the completion from the callback, thus it
must not be called inside the callback itself.  This was already a
problem in the past with ALSA hrtimer driver, and the early commit
[fcfdebe70759: ALSA: hrtimer - Fix lock-up] tried to address it.

However, the previous fix is still insufficient: it may still cause a
lockup when the ALSA timer instance reprograms itself in its callback.
Then it invokes the start function even in snd_timer_interrupt() that
is called in hrtimer callback itself, results in a CPU stall.  This is
no hypothetical problem but actually triggered by syzkaller fuzzer.

This patch tries to fix the issue again.  Now we call
hrtimer_try_to_cancel() at both start and stop functions so that it
won't fall into a deadlock, yet giving some chance to cancel the queue
if the functions have been called outside the callback.  The proper
hrtimer_cancel() is called in anyway at closing, so this should be
enough.

Reported-and-tested-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - Fix bass pin fixup for ASUS N550JX
Takashi Iwai [Mon, 18 Jan 2016 08:17:30 +0000 (09:17 +0100)]
ALSA: hda - Fix bass pin fixup for ASUS N550JX

ASUS N550JX (PCI SSID 1043:13df) requires the same fixup for a bass
speaker output pin as other N550 models.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=110001
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: timer: Code cleanup
Takashi Iwai [Thu, 14 Jan 2016 16:01:46 +0000 (17:01 +0100)]
ALSA: timer: Code cleanup

This is a minor code cleanup without any functional changes:
- Kill keep_flag argument from _snd_timer_stop(), as all callers pass
  only it false.
- Remove redundant NULL check in _snd_timer_stop().

Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: timer: Harden slave timer list handling
Takashi Iwai [Thu, 14 Jan 2016 15:30:58 +0000 (16:30 +0100)]
ALSA: timer: Harden slave timer list handling

A slave timer instance might be still accessible in a racy way while
operating the master instance as it lacks of locking.  Since the
master operation is mostly protected with timer->lock, we should cope
with it while changing the slave instance, too.  Also, some linked
lists (active_list and ack_list) of slave instances aren't unlinked
immediately at stopping or closing, and this may lead to unexpected
accesses.

This patch tries to address these issues.  It adds spin lock of
timer->lock (either from master or slave, which is equivalent) in a
few places.  For avoiding a deadlock, we ensure that the global
slave_active_lock is always locked at first before each timer lock.

Also, ack and active_list of slave instances are properly unlinked at
snd_timer_stop() and snd_timer_close().

Last but not least, remove the superfluous call of _snd_timer_stop()
at removing slave links.  This is a noop, and calling it may confuse
readers wrt locking.  Further cleanup will follow in a later patch.

Actually we've got reports of use-after-free by syzkaller fuzzer, and
this hopefully fixes these issues.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - Add fixup for Dell Latitidue E6540
Takashi Iwai [Fri, 15 Jan 2016 11:59:25 +0000 (12:59 +0100)]
ALSA: hda - Add fixup for Dell Latitidue E6540

Another Dell model, another fixup entry: Latitude E6540 needs the same
fixup as other Latitude E series as workaround for noise problems.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=104341
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: timer: Fix race among timer ioctls
Takashi Iwai [Wed, 13 Jan 2016 16:48:01 +0000 (17:48 +0100)]
ALSA: timer: Fix race among timer ioctls

ALSA timer ioctls have an open race and this may lead to a
use-after-free of timer instance object.  A simplistic fix is to make
each ioctl exclusive.  We have already tread_sem for controlling the
tread, and extend this as a global mutex to be applied to each ioctl.

The downside is, of course, the worse concurrency.  But these ioctls
aren't to be parallel accessible, in anyway, so it should be fine to
serialize there.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Tested-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - add codec support for Kabylake display audio codec
Libin Yang [Thu, 14 Jan 2016 06:09:00 +0000 (14:09 +0800)]
ALSA: hda - add codec support for Kabylake display audio codec

This patch adds codec ID (0x8086280b) for Kabylake display codec
and apply the hsw fix-ups to Kabylake.

Signed-off-by: Libin Yang <libin.yang@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: timer: Fix double unlink of active_list
Takashi Iwai [Wed, 13 Jan 2016 20:35:06 +0000 (21:35 +0100)]
ALSA: timer: Fix double unlink of active_list

ALSA timer instance object has a couple of linked lists and they are
unlinked unconditionally at snd_timer_stop().  Meanwhile
snd_timer_interrupt() unlinks it, but it calls list_del() which leaves
the element list itself unchanged.  This ends up with unlinking twice,
and it was caught by syzkaller fuzzer.

The fix is to use list_del_init() variant properly there, too.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Tested-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: usb-audio: Fix mixer ctl regression of Native Instrument devices
Takashi Iwai [Wed, 13 Jan 2016 06:20:13 +0000 (07:20 +0100)]
ALSA: usb-audio: Fix mixer ctl regression of Native Instrument devices

The commit [da6d276957ea: ALSA: usb-audio: Add resume support for
Native Instruments controls] brought a regression where the Native
Instrument audio devices don't get the correct value at update due to
the missing shift at writing.  This patch addresses it.

Fixes: da6d276957ea ('ALSA: usb-audio: Add resume support for Native Instruments controls')
Reported-and-tested-by: Owen Williams <owilliams@mixxx.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - fix the headset mic detection problem for a Dell laptop
Hui Wang [Wed, 13 Jan 2016 03:51:38 +0000 (11:51 +0800)]
ALSA: hda - fix the headset mic detection problem for a Dell laptop

The machine uses codec alc255, and the pin configuration value for
pin 0x14 on this machine is 0x90171130 which is not in the pin quirk
table yet.

Cc: stable@vger.kernel.org
BugLink: https://bugs.launchpad.net/bugs/1533461
Signed-off-by: Hui Wang <hui.wang@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda - Fix white noise on Dell Latitude E5550
Takashi Iwai [Tue, 12 Jan 2016 20:06:39 +0000 (21:06 +0100)]
ALSA: hda - Fix white noise on Dell Latitude E5550

Dell Latitude E5550 (1028:062c) has a white noise problem like other
Latitude E models, and it gets fixed by the very same quirk as well.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=110591
Cc: <stable@vger.kernel.org> # v4.1+
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: hda_intel: add card number to irq description
Heiner Kallweit [Tue, 22 Dec 2015 18:09:05 +0000 (19:09 +0100)]
ALSA: hda_intel: add card number to irq description

Currently the info in /proc/interrupts doesn't allow to figure out which
interrupt belongs to which card (HDMI, PCH, ..).
Therefore add card details to the interrupt description.
With the patch the info in /proc/interrupts looks like this:

PCI-MSI 442368-edge      snd_hda_intel:card1
PCI-MSI 49152-edge      snd_hda_intel:card0

NOTE: this patch adds the new irq_descr field snd_card struct that is
filled automatically at a card object creation.  This can be used
generically for other drivers as well.  The changes for others will
follow later -- tiwai

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: seq: Fix race at timer setup and close
Takashi Iwai [Tue, 12 Jan 2016 14:36:27 +0000 (15:36 +0100)]
ALSA: seq: Fix race at timer setup and close

ALSA sequencer code has an open race between the timer setup ioctl and
the close of the client.  This was triggered by syzkaller fuzzer, and
a use-after-free was caught there as a result.

This patch papers over it by adding a proper queue->timer_mutex lock
around the timer-related calls in the relevant code path.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Tested-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: seq: Fix missing NULL check at remove_events ioctl
Takashi Iwai [Tue, 12 Jan 2016 11:38:02 +0000 (12:38 +0100)]
ALSA: seq: Fix missing NULL check at remove_events ioctl

snd_seq_ioctl_remove_events() calls snd_seq_fifo_clear()
unconditionally even if there is no FIFO assigned, and this leads to
an Oops due to NULL dereference.  The fix is just to add a proper NULL
check.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Tested-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: usb-audio: Avoid calling usb_autopm_put_interface() at disconnect
Takashi Iwai [Tue, 12 Jan 2016 13:03:33 +0000 (14:03 +0100)]
ALSA: usb-audio: Avoid calling usb_autopm_put_interface() at disconnect

ALSA PCM may still have a leftover instance after disconnection and
it delays its release.  The problem is that the PCM close code path of
USB-audio driver has a call of snd_usb_autosuspend().  This involves
with the call of usb_autopm_put_interface() and it may lead to a
kernel Oops due to the NULL object like:

 BUG: unable to handle kernel NULL pointer dereference at 0000000000000190
 IP: [<ffffffff815ae7ef>] usb_autopm_put_interface+0xf/0x30 PGD 0
 Call Trace:
  [<ffffffff8173bd94>] snd_usb_autosuspend+0x14/0x20
  [<ffffffff817461bc>] snd_usb_pcm_close.isra.14+0x5c/0x90
  [<ffffffff8174621f>] snd_usb_playback_close+0xf/0x20
  [<ffffffff816ef58a>] snd_pcm_release_substream.part.36+0x3a/0x90
  [<ffffffff816ef6b3>] snd_pcm_release+0xa3/0xb0
  [<ffffffff816debb0>] snd_disconnect_release+0xd0/0xe0
  [<ffffffff8114d417>] __fput+0x97/0x1d0
  [<ffffffff8114d589>] ____fput+0x9/0x10
  [<ffffffff8109e452>] task_work_run+0x72/0x90
  [<ffffffff81088510>] do_exit+0x280/0xa80
  [<ffffffff8108996a>] do_group_exit+0x3a/0xa0
  [<ffffffff8109261f>] get_signal+0x1df/0x540
  [<ffffffff81040903>] do_signal+0x23/0x620
  [<ffffffff8114c128>] ? do_readv_writev+0x128/0x200
  [<ffffffff810012e1>] prepare_exit_to_usermode+0x91/0xd0
  [<ffffffff810013ba>] syscall_return_slowpath+0x9a/0x120
  [<ffffffff817587cd>] ? __sys_recvmsg+0x5d/0x70
  [<ffffffff810d2765>] ? ktime_get_ts64+0x45/0xe0
  [<ffffffff8115dea0>] ? SyS_poll+0x60/0xf0
  [<ffffffff818d2327>] int_ret_from_sys_call+0x25/0x8f

We have already a check of disconnection in snd_usb_autoresume(), but
the check is missing its counterpart.  The fix is just to put the same
check in snd_usb_autosuspend(), too.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=109431
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoMerge tag 'asoc-v4.4-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie...
Takashi Iwai [Mon, 11 Jan 2016 15:24:58 +0000 (16:24 +0100)]
Merge tag 'asoc-v4.4-rc8' of git://git./linux/kernel/git/broonie/sound into for-next

ASoC: Last updates for v4.5

A bunch more updates for v4.5, mainly driver work:

 - More topology API enhancements from Mengdong Lin working towards
   making everything more component based and being able to specify PCM
   links via topology.
 - Large sets driver updates from Cirrus, Intel (mainly more Skylake
   support) and Renesas.
 - New driver for AMD ACP
 - Rename PCM1792a driver to be generic pcm179x

8 years agoMerge remote-tracking branch 'asoc/topic/wm8960' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:40 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/wm8960' into asoc-next

8 years agoMerge remote-tracking branch 'asoc/topic/rt5651' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:39 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/rt5651' into asoc-next

8 years agoMerge remote-tracking branches 'asoc/topic/rcar' and 'asoc/topic/rockchip' into asoc...
Mark Brown [Mon, 11 Jan 2016 13:54:38 +0000 (13:54 +0000)]
Merge remote-tracking branches 'asoc/topic/rcar' and 'asoc/topic/rockchip' into asoc-next

8 years agoMerge remote-tracking branches 'asoc/topic/fsl-asrc' and 'asoc/topic/fsl-ssi' into...
Mark Brown [Mon, 11 Jan 2016 13:54:36 +0000 (13:54 +0000)]
Merge remote-tracking branches 'asoc/topic/fsl-asrc' and 'asoc/topic/fsl-ssi' into asoc-next

8 years agoMerge remote-tracking branches 'asoc/topic/da7218' and 'asoc/topic/da7219' into asoc...
Mark Brown [Mon, 11 Jan 2016 13:54:35 +0000 (13:54 +0000)]
Merge remote-tracking branches 'asoc/topic/da7218' and 'asoc/topic/da7219' into asoc-next

8 years agoMerge remote-tracking branches 'asoc/topic/bcm2835' and 'asoc/topic/cleanup' into...
Mark Brown [Mon, 11 Jan 2016 13:54:34 +0000 (13:54 +0000)]
Merge remote-tracking branches 'asoc/topic/bcm2835' and 'asoc/topic/cleanup' into asoc-next

8 years agoMerge remote-tracking branches 'asoc/topic/adsp' and 'asoc/topic/amd' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:32 +0000 (13:54 +0000)]
Merge remote-tracking branches 'asoc/topic/adsp' and 'asoc/topic/amd' into asoc-next

8 years agoMerge remote-tracking branch 'asoc/topic/rt5645' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:32 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/rt5645' into asoc-next

8 years agoMerge remote-tracking branch 'asoc/topic/pcm1792a' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:31 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/pcm1792a' into asoc-next

8 years agoMerge remote-tracking branch 'asoc/topic/pcm-list' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:31 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/pcm-list' into asoc-next

8 years agoMerge remote-tracking branch 'asoc/topic/pcm' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:30 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/pcm' into asoc-next

8 years agoMerge remote-tracking branch 'asoc/topic/arizona' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:29 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/arizona' into asoc-next

8 years agoMerge tag 'asoc-v4.5' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:29 +0000 (13:54 +0000)]
Merge tag 'asoc-v4.5' into asoc-next

ASoC: Updates for v4.5

This is quite a busy release on the driver front with a lot of new
drivers being added but comparatively quiet on the core side with only
one big change going in and that a fairly straightforward refactoring.

 - Conversion of the array of DAI links to a list by Mengdong Lin,
   supporting dynamically adding and removing DAI links.
 - Some more fixes for the topology code, though it is still not final
   and ready for enabling in production.  We really need to get to the
   point where that can be done.
 - A pile of changes for Intel SkyLake drivers which hopefully deliver
   some useful initial functionality for systems with this chipset,
   though there is more work still to come.
 - New drivers for a number of Imagination Technologies IPs.
 - Lots of new features and cleanups for the Renesas drivers.
 - ANC support for WM5110.
 - New driver for Atmel class D speaker drivers.
 - New drivers for Cirrus CS47L24 and WM1831.
 - New driver for Dialog DA7128.
 - New drivers for Realtek RT5659 and RT56156.
 - New driver for Rockchip RK3036.
 - New driver for TI PC3168A

# gpg: Signature made Wed 23 Dec 2015 00:42:40 GMT using RSA key ID 5D5487D0
# gpg: Good signature from "Mark Brown <broonie@sirena.org.uk>"
# gpg:                 aka "Mark Brown <broonie@debian.org>"
# gpg:                 aka "Mark Brown <broonie@kernel.org>"
# gpg:                 aka "Mark Brown <broonie@tardis.ed.ac.uk>"
# gpg:                 aka "Mark Brown <broonie@linaro.org>"
# gpg:                 aka "Mark Brown <Mark.Brown@linaro.org>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 3F25 68AA C269 98F9 E813  A1C5 C3F4 36CA 30F5 D8EB
#      Subkey fingerprint: ADE6 68AA 6757 18B5 9FE2  9FEA 24D6 8B72 5D54 87D0

8 years agoMerge remote-tracking branch 'asoc/topic/intel' into asoc-next
Mark Brown [Mon, 11 Jan 2016 13:54:28 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/topic/intel' into asoc-next

8 years agoMerge remote-tracking branch 'asoc/fix/compress' into asoc-linus
Mark Brown [Mon, 11 Jan 2016 13:54:27 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/fix/compress' into asoc-linus

8 years agoMerge remote-tracking branch 'asoc/fix/arizona' into asoc-linus
Mark Brown [Mon, 11 Jan 2016 13:54:26 +0000 (13:54 +0000)]
Merge remote-tracking branch 'asoc/fix/arizona' into asoc-linus

8 years agoMerge tag 'asoc-fix-v4.4-rc8' into asoc-linus
Mark Brown [Mon, 11 Jan 2016 13:54:26 +0000 (13:54 +0000)]
Merge tag 'asoc-fix-v4.4-rc8' into asoc-linus

ASoC: Last minute fixes for v4.4

A few final fixes for v4.4, the main one being the two patches to the
new Sky Lake drivers which fix a previous incorrect fix that went in
during an earlier -rc.

# gpg: Signature made Wed 06 Jan 2016 17:12:57 GMT using RSA key ID 5D5487D0
# gpg: Good signature from "Mark Brown <broonie@sirena.org.uk>"
# gpg:                 aka "Mark Brown <broonie@debian.org>"
# gpg:                 aka "Mark Brown <broonie@kernel.org>"
# gpg:                 aka "Mark Brown <broonie@tardis.ed.ac.uk>"
# gpg:                 aka "Mark Brown <broonie@linaro.org>"
# gpg:                 aka "Mark Brown <Mark.Brown@linaro.org>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 3F25 68AA C269 98F9 E813  A1C5 C3F4 36CA 30F5 D8EB
#      Subkey fingerprint: ADE6 68AA 6757 18B5 9FE2  9FEA 24D6 8B72 5D54 87D0

8 years agoASoC: hdac_hdmi: remove unused hdac_hdmi_query_pin_connlist
Arnd Bergmann [Mon, 11 Jan 2016 12:09:55 +0000 (13:09 +0100)]
ASoC: hdac_hdmi: remove unused hdac_hdmi_query_pin_connlist

A recent rework removed the only user of the hdac_hdmi_query_pin_connlist
function, so we now get a warning when building the hdac_hdmi driver:

hdac_hdmi.c:313:12: warning: 'hdac_hdmi_query_pin_connlist' defined but not used [-Wunused-function]

This removes the function, which makes the file build cleanly again.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 15b914476bf2 ("ASoC: hdac_hdmi: Use list to add pins and converters")
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: AMD: Add missing include file
Guenter Roeck [Mon, 11 Jan 2016 10:41:05 +0000 (02:41 -0800)]
ASoC: AMD: Add missing include file

arm:allmodconfig, s390:allmodconfig, sparc64:allmodconfig, and probably
other builds fail with

sound/soc/amd/acp-pcm-dma.c:83:2: error:
implicit declaration of function ‘readl’
sound/soc/amd/acp-pcm-dma.c:88:2: error:
implicit declaration of function ‘writel’

Include linux/io.h explicitly to fix the problem.

Fixes: 7c31335a03b6a ("ASoC: AMD: add AMD ASoC ACP 2.x DMA driver")
Cc: Maruthi Srinivas Bayyavarapu <Maruthi.Bayyavarapu@amd.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoALSA: hda - Fixup inverted internal mic for Lenovo E50-80
David Henningsson [Mon, 11 Jan 2016 08:33:14 +0000 (09:33 +0100)]
ALSA: hda - Fixup inverted internal mic for Lenovo E50-80

Inform userspace that one channel of the internal mic has reversed
polarity, so it does not attempt to add both channels together and
end up with silence.

Cc: stable@vger.kernel.org
Reported-by: Andrzej Mendel <andrzej.mendel@gmail.com>
Alsa-info: http://www.alsa-project.org/db/?f=3088f82a0cf977855f92af9db8ad406c04f71efa
BugLink: https://bugs.launchpad.net/bugs/1529624
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: usb: Add native DSD support for Oppo HA-1
Jurgen Kramer [Mon, 11 Jan 2016 07:16:58 +0000 (08:16 +0100)]
ALSA: usb: Add native DSD support for Oppo HA-1

This patch adds native DSD support for the Oppo HA-1. It uses a XMOS chipset
but they use their own vendor ID.

Signed-off-by: Jurgen Kramer <gtmkramer@xs4all.nl>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoASoC: Make aux_dev more like a generic component
Mengdong Lin [Wed, 6 Jan 2016 05:29:31 +0000 (13:29 +0800)]
ASoC: Make aux_dev more like a generic component

aux_dev is mainly used by the machine driver to specify analog devices,
which are registered as codecs. Making it more like a generic component
can help the machine driver to use it to specify any component with
topology info by name.

Details:
- Remove the stub 'rtd_aux' array from the soc card.
- Add a list 'aux_comp_list' to store the components of aux_devs.
  And add a list head 'list_aux' to struct snd_soc_component, for adding
  such components to the above list.
- Add a 'init' ops to a component for machine specific init.
  soc_bind_aux_dev() will set it to be aux_dev's init. And it will be
  called when probing the component.
- soc_bind_aux_dev() will also search components by name of an aux_dev,
  since it may not be a codec.
- Move probing of aux_devs before checking new DAI links brought by
  topology.
- Move removal of aux_devs later than removal of links. Because topology
  of aux components may register DAIs and the DAI drivers will go with
  removal of the aux components, we want soc_remove_link_dais() to remove
  the DAIs at first.

Signed-off-by: Mengdong Lin <mengdong.lin@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: bcm2835: cleanup includes by ordering them alphabetically
Martin Sperl [Sat, 9 Jan 2016 09:25:53 +0000 (09:25 +0000)]
ASoC: bcm2835: cleanup includes by ordering them alphabetically

Cleanup of includes so that they are ordered alphabetically.

Signed-off-by: Martin Sperl <kernel@martin.sperl.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: AMD: Manage ACP 2.x SRAM banks power
Maruthi Srinivas Bayyavarapu [Fri, 8 Jan 2016 23:22:11 +0000 (18:22 -0500)]
ASoC: AMD: Manage ACP 2.x SRAM banks power

ACP SRAM banks gets turned on when ACP is powered on.
Not all banks are used for playback/capture. So, power on
required banks during audio device open and power off during
audio device close.

Signed-off-by: Maruthi Bayyavarapu <maruthi.bayyavarapu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: AMD: add pm ops
Maruthi Srinivas Bayyavarapu [Fri, 8 Jan 2016 23:22:10 +0000 (18:22 -0500)]
ASoC: AMD: add pm ops

genpd will power off/on ACP to manage runtime ACP PM. ACP runtime PM
hooks are added to get it deinitialized and initialized respectively,
after it is powered off/on.

When system goes to suspend when audio usecase is active, ACP will
be powered off through genpd. When it resumes, ACP needs to be
initialized and reconfigured.

Signed-off-by: Maruthi Bayyavarapu <maruthi.bayyavarapu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: AMD: add AMD ASoC ACP 2.x DMA driver
Maruthi Srinivas Bayyavarapu [Fri, 8 Jan 2016 23:22:09 +0000 (18:22 -0500)]
ASoC: AMD: add AMD ASoC ACP 2.x DMA driver

ACP IP has internal DMA controller with multiple channels which
can be programmed in cyclic/non cyclic manner. ACP can generate
interrupt upon completion of DMA transfer, if required.
The PCM driver provides the platform DMA component to ALSA core.

Signed-off-by: Maruthi Bayyavarapu <maruthi.bayyavarapu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Murali Krishna Vemuri <murali-krishna.vemuri@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: AMD : add ACP 2.2 register headers
Maruthi Srinivas Bayyavarapu [Fri, 8 Jan 2016 23:22:08 +0000 (18:22 -0500)]
ASoC: AMD : add ACP 2.2 register headers

These are register headers for the ACP (Audio CoProcessor) v2.2

Signed-off-by: Maruthi Bayyavarapu <maruthi.bayyavarapu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: fsl_ssi: mark some registers precious
Maciej S. Szmigiero [Sun, 20 Dec 2015 20:31:48 +0000 (21:31 +0100)]
ASoC: fsl_ssi: mark some registers precious

Mark some registers precious since their
reads have side effects (like clearing flags).

Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
Reviewed-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: fsl_ssi: mark SACNT register volatile
Maciej S. Szmigiero [Sun, 20 Dec 2015 20:30:25 +0000 (21:30 +0100)]
ASoC: fsl_ssi: mark SACNT register volatile

SACNT register should be marked volatile since
its WR and RD bits are cleared by SSI after
completing the relevant operation.
This unbreaks AC'97 register access.

Fixes: 05cf237972fe ("ASoC: fsl_ssi: Add driver suspend and resume to support MEGA Fast")

Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
Reviewed-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Skylake: Add Resume capability in PCM info.
Jeeja KP [Fri, 18 Dec 2015 09:42:09 +0000 (15:12 +0530)]
ASoC: Intel: Skylake: Add Resume capability in PCM info.

This patch adds pcm capability to support Resume.

Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Skylake: Reconfigure Link stream on suspend/resume
Jeeja KP [Fri, 18 Dec 2015 09:42:08 +0000 (15:12 +0530)]
ASoC: Intel: Skylake: Reconfigure Link stream on suspend/resume

On suspend the link register are lost so we need to reconfigure
them in resume. This patch adds the reconfiguration of the link
register in trigger resume.

Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Skylake: Add DMA resume position in Trigger resume/suspend
Jeeja KP [Fri, 18 Dec 2015 09:42:07 +0000 (15:12 +0530)]
ASoC: Intel: Skylake: Add DMA resume position in Trigger resume/suspend

Use the DMA resume capability to resume the DMA position when
stream is suspended/resumed.

In suspend we save the position and when stream is resumed the stream needs
to be started from the position when the stream was suspended using the new
DMA resume capabilities

Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Skylake: enable interrupt as wake source in active suspend
Jeeja KP [Fri, 18 Dec 2015 09:42:06 +0000 (15:12 +0530)]
ASoC: Intel: Skylake: enable interrupt as wake source in active suspend

In active suspend, any HDA interrupt should wake the system. When device
enters active suspend, we need to enable HDA controller interrupt as wake
source. Similarly disable HDA controller interrupt as wake source when
exiting active suspend.

Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Skylake: manage link power in active suspend
Vinod Koul [Fri, 18 Dec 2015 09:42:05 +0000 (15:12 +0530)]
ASoC: Intel: Skylake: manage link power in active suspend

When device enters active suspend, we should turn off the links
as they are not in use. Similarly we need to bring back links
when we exit active suspend.

Signed-off-by: Jayachandran B <jayachandran.b@intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Skylake: Use CGCTL.MISCBDCGE for Phrase detection notification
Dharageswari.R [Fri, 18 Dec 2015 09:42:04 +0000 (15:12 +0530)]
ASoC: Intel: Skylake: Use CGCTL.MISCBDCGE for Phrase detection notification

Per HW recommendation, SW shall clear the CGCTL.MISCBDCGE and set
it back once data is transferred. So clear this when we get the
IPC and track using a driver flag, and set back on closure

Signed-off-by: Dharageswari.R <dharageswari.r@intel.com>
Signed-off-by: Jayachandran B <jayachandran.b@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Skylake: fix reset controller sequencing
Jayachandran B [Fri, 18 Dec 2015 09:42:03 +0000 (15:12 +0530)]
ASoC: Intel: Skylake: fix reset controller sequencing

MISCBDCGE is a new register for Misc Backbone clock gate control
which is useful to control while resetting the link and ensuring
controller is in required state so add API to control it

HW recommends that we reset with CGCTL.MISCBDCGE disabled, so add
that while doing init chip and reset sequence.

Signed-off-by: Jayachandran B <jayachandran.b@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoMerge branch 'topic/hdac' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie...
Mark Brown [Sun, 10 Jan 2016 12:13:34 +0000 (12:13 +0000)]
Merge branch 'topic/hdac' of git://git./linux/kernel/git/broonie/sound into asoc-intel

8 years agoASoC: dwc: add quirk to override COMP_PARAM_1 register
Maruthi Srinivas Bayyavarapu [Fri, 8 Jan 2016 23:22:05 +0000 (18:22 -0500)]
ASoC: dwc: add quirk to override COMP_PARAM_1 register

DWC for capture in ACP 2.x IP reports playback and capture capabilities
though it supports only capture. Added a quirk to override default value
to represent capture capability only.

Signed-off-by: Maruthi Bayyavarapu <maruthi.bayyavarapu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Support registering a DAI dynamically
Mengdong Lin [Thu, 31 Dec 2015 08:40:43 +0000 (16:40 +0800)]
ASoC: Support registering a DAI dynamically

Define API snd_soc_register_dai() to add a DAI dynamically and
create the DAI widgets. Topology can use this API to register DAIs
when probing a component with topology info. These DAIs's playback
& capture widgets will be freed when the sound card is unregistered
and the DAIs will be freed when cleaning up the component.

And a dobj is embedded into the struct snd_soc_dai_driver. Topology
can use the dobj to find the DAI drivers created by it and free them
when the topology component is removed.

Signed-off-by: Mengdong Lin <mengdong.lin@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Define soc_add_dai() to add a DAI to a component
Mengdong Lin [Thu, 31 Dec 2015 08:40:20 +0000 (16:40 +0800)]
ASoC: Define soc_add_dai() to add a DAI to a component

Define soc_add_dai() as a wrapper to add a single DAI to a component.
It can be reused to register a DAI dynamically by topology.

Signed-off-by: Mengdong Lin <mengdong.lin@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoALSA: hdac: add snd_hdac_ext_bus_link_power_up_all
Vinod Koul [Fri, 18 Dec 2015 09:42:02 +0000 (15:12 +0530)]
ALSA: hdac: add snd_hdac_ext_bus_link_power_up_all

We have an API for powering down all links, we need a similar one
for powering up links, so add for power up as well

Signed-off-by: Jayachandran B <jayachandran.b@intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoALSA: hdac: Increase timeout value for link power check
Jayachandran B [Fri, 18 Dec 2015 09:42:01 +0000 (15:12 +0530)]
ALSA: hdac: Increase timeout value for link power check

HW recommends 180us for worst case values for link power up
delay, so change the current delay value from 50 (150us) to 150
(450us)

Signed-off-by: Jayachandran B <jayachandran.b@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoALSA: hdac: couple the hda DMA stream in cleanup
Jeeja KP [Fri, 18 Dec 2015 09:42:00 +0000 (15:12 +0530)]
ALSA: hdac: couple the hda DMA stream in cleanup

A stream is by default in coupled mode, in DSP operation we move
it to decoupled mode. On cleanup HW expects that we leave it back
to default state so couple the DMA on cleanup.

Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoALSA: hdac: Add support for hda DMA Resume capability
Jeeja KP [Fri, 18 Dec 2015 09:41:59 +0000 (15:11 +0530)]
ALSA: hdac: Add support for hda DMA Resume capability

Skylake sports new capability of DMA resume, DRSM where we can
resume the DMA. This capability is defined by presence of
AZX_DRSM_CAP_ID.

If this capability is present, we use this capability.
So we add:

snd_hdac_ext_stream_drsm_enable() - DMA resume caps
snd_hdac_ext_stream_set_dpibr() - set the DMA position
snd_hdac_ext_stream_set_lpib() - set the lpib

Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: pcm1792a: Rename pcm1792a to pcm179x
Michael Trimarchi [Sat, 9 Jan 2016 23:38:04 +0000 (00:38 +0100)]
ASoC: pcm1792a: Rename pcm1792a to pcm179x

pcm1792a is compatible with pcm1795 and pcm1796 so it's
better to have them under the common name pcm179x

Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: pcm1792a: Rename internal data and function to pcm179x
Michael Trimarchi [Sat, 9 Jan 2016 23:38:03 +0000 (00:38 +0100)]
ASoC: pcm1792a: Rename internal data and function to pcm179x

Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: rockchip: i2s: Add SNDRV_PCM_FMTBIT_S32_LE support
Michael Trimarchi [Sat, 9 Jan 2016 22:47:58 +0000 (23:47 +0100)]
ASoC: rockchip: i2s: Add SNDRV_PCM_FMTBIT_S32_LE support

Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: compress: Fix compress device direction check
Vinod Koul [Thu, 7 Jan 2016 16:18:14 +0000 (21:48 +0530)]
ASoC: compress: Fix compress device direction check

The detection of direction for compress was only taking into account codec
capabilities and not CPU ones. Fix this by checking the CPU side capabilities
as well

Cc: <stable@vger.kernel.org>
Tested-by: Ashish Panwar <ashish.panwar@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: hdac_hdmi: Use list to add pins and converters
Subhransu S. Prusty [Wed, 9 Dec 2015 16:16:10 +0000 (21:46 +0530)]
ASoC: hdac_hdmi: Use list to add pins and converters

Future platforms may have a different set of pins/converters.
So use lists to add pins and converters based on enumeration.

Also it may be required to connect any converter to any pin
dynamically as per different use cases (for example DP is
connected to pin 6 on skylake board). So this will help in
dynamically select and route.

Fix the dai map as well to use the pin/cvt from list. Not
enabling all dai maps for now.

Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: wm8960: boost switch should be closed when using L/RINPUT1
Zidan Wang [Fri, 8 Jan 2016 08:57:02 +0000 (16:57 +0800)]
ASoC: wm8960: boost switch should be closed when using L/RINPUT1

L/RINPUT1 can line to Left/Right Boost Mixer through boost switch.
If boost switch is open, there will be no voice when using L/RINPUT1.

Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: wm8960: add DAC mono mix kcontrol
Zidan Wang [Fri, 8 Jan 2016 08:57:01 +0000 (16:57 +0800)]
ASoC: wm8960: add DAC mono mix kcontrol

In normal operation, the left and right channel digital audio data is
converted to analogue in two separate DACs. There is a mono-mix mode
where the two audio channels are mixed together digitally and then
converted to analogue using only one DAC, while the other DAC is
switched off. The mono-mix signal can be selected to appear on both
analogue output channels. The mono mix is automatically attenuated by
6dB to prevent clipping.

Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: hdac_hdmi: Fix to warn instead of err for no connected nids
Subhransu S. Prusty [Wed, 9 Dec 2015 16:16:09 +0000 (21:46 +0530)]
ASoC: hdac_hdmi: Fix to warn instead of err for no connected nids

It is possible that some pin widget may return with no converter
connected. So don't throw error if none are found to be connected.
Instead print a warning and continue.

Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoMerge branch 'for-linus' into for-next
Takashi Iwai [Wed, 6 Jan 2016 19:58:13 +0000 (20:58 +0100)]
Merge branch 'for-linus' into for-next

 Conflicts:
drivers/gpu/drm/i915/intel_display.c
sound/soc/intel/skylake/skl.h

8 years agoMerge tag 'asoc-fix-v4.4-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/brooni...
Takashi Iwai [Wed, 6 Jan 2016 19:53:28 +0000 (20:53 +0100)]
Merge tag 'asoc-fix-v4.4-rc8' of git://git./linux/kernel/git/broonie/sound into for-linus

ASoC: Last minute fixes for v4.4

A few final fixes for v4.4, the main one being the two patches to the
new Sky Lake drivers which fix a previous incorrect fix that went in
during an earlier -rc.

8 years agoASoC: wm_adsp: Pull data through compressed read
Charles Keepax [Wed, 6 Jan 2016 12:33:19 +0000 (12:33 +0000)]
ASoC: wm_adsp: Pull data through compressed read

Data is read in blocks of up to one fragment is size from the circular
buffer on the DSP and is re-packed to remove the padding byte that
exists in the DSP memory map.

Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: wm_adsp: Add a handler for the compressed IRQ
Charles Keepax [Wed, 6 Jan 2016 12:33:18 +0000 (12:33 +0000)]
ASoC: wm_adsp: Add a handler for the compressed IRQ

Here support is added for responding to DSP IRQs that are used to
indicate data being available on the DSP. The idea is that we check the
amount of data available upon receipt of an IRQ and on subsequent calls
to the pointer callback we recheck once less than one fragment is
available (to avoid excessive SPI traffic), if there is truely less than
one fragment available we ack the last IRQ and wait for a new one.

Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: wm5110: Fix PGA clear when disabling DRE
Charles Keepax [Wed, 6 Jan 2016 15:15:37 +0000 (15:15 +0000)]
ASoC: wm5110: Fix PGA clear when disabling DRE

We don't want to use a bypassed write in wm5110_clear_pga_volume,
we might disable the DRE whilst the CODEC is powered down. A
normal regmap_write will always go to the hardware (when not on
cache_only) even if the written value matches the cache. As using
a normal write will still achieve the desired behaviour of bring
the cache and hardware in sync, this patch updates the function
to use a normal write, which avoids issues when the CODEC is
powered down.

Signed-off-by: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org
8 years agoASoC: rsnd: precedence error in rsnd_ssiu_init()
Dan Carpenter [Wed, 6 Jan 2016 09:38:41 +0000 (12:38 +0300)]
ASoC: rsnd: precedence error in rsnd_ssiu_init()

The bitwise OR has higher precedence than ?: so the val2 was always set
to 0x2.

Fixes: b4c83b171557 ('ASoC: rsnd: add Multi channel support')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoMerge branch 'for-linus' into for-next
Takashi Iwai [Wed, 6 Jan 2016 09:20:41 +0000 (10:20 +0100)]
Merge branch 'for-linus' into for-next

8 years agoALSA: emux: constify nrpn_conv_table structures
Julia Lawall [Mon, 4 Jan 2016 16:50:47 +0000 (17:50 +0100)]
ALSA: emux: constify nrpn_conv_table structures

The nrpn_conv_table structures are never modified, so declare them as
const.

Done with the help of Coccinelle.

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: dice: expand timeout to wait for Dice notification
Takashi Sakamoto [Thu, 31 Dec 2015 04:58:14 +0000 (13:58 +0900)]
ALSA: dice: expand timeout to wait for Dice notification

Some users have reported that their Dice based models generate ETIMEDOUT
when starting PCM playback. It means that current timeout (=100msec) is
not enough for their models to transfer notifications.

This commit expands the timeout up to 2 sec. As a result, in a worst case,
any operations to start AMDTP streams takes 2 sec or more. Then, in
userspace, snd_pcm_hw_params(), snd_pcm_prepare(), snd_pcm_recover(),
snd_rawmidi_open(), snd_seq_connect_from() and snd_seq_connect_to() may
take the time.

Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: dice: purge transaction initialization at timeout of Dice notification
Takashi Sakamoto [Thu, 31 Dec 2015 04:58:13 +0000 (13:58 +0900)]
ALSA: dice: purge transaction initialization at timeout of Dice notification

In previous commit, card registration is processed under situation
with few bus reset. There's no need to add a workaround of transaction
re-initialization at timeout.

This commit purges the re-initialization.

Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: dice: postpone card registration
Takashi Sakamoto [Thu, 31 Dec 2015 04:58:12 +0000 (13:58 +0900)]
ALSA: dice: postpone card registration

Some models based on ASIC for Dice II series (STD, CP) change their
hardware configurations after appearing on IEEE 1394 bus. This is due to
interactions of boot loader (RedBoot), firmwares (eCos) and vendor's
configurations. This causes current ALSA dice driver to get wrong
information about the hardware's capability because its probe function
runs just after detecting unit of the model.

As long as I investigated, it takes a bit time (less than 1 second) to load
the firmware after bootstrap. Just after loaded, the driver can get
information about the unit. Then the hardware is initialized according to
vendor's configurations. After, the got information becomes wrong.
Between bootstrap, firmware loading and post configuration, some bus resets
are observed.

This commit offloads most processing of probe function into workqueue and
schedules the workqueue after successive bus resets. This has an effect to
get correct hardware information and avoid involvement to bus reset storm.

For code simplicity, this change effects all of Dice-based models, i.e.
Dice II, Dice Jr., Dice Mini and Dice III.

I use a loose strategy to manage a race condition between the work and the
bus reset. This is due to a specification of dice transaction. When bus
reset occurs, registered address for the transaction is cleared. Drivers
must re-register their own address again. While, this operation is required
for the work because the work includes to wait for the transaction. This
commit uses no lock primitives for the race condition. Instead, checking
'registered' member of 'struct snd_dice' avoid executing the work again.
If sound card is not registered, the work can be scheduled again by bus
reset handler.

When .remove callback is executed, the sound card is going to be released.
The work should not be pending or executed in the releasing. This commit
uses cancel_delayed_work_sync() in .remove callback and wait till the
pending work finished. After .remove callback, .update callback is not
executed, therefore no works are scheduled again.

Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoALSA: dice: split subaddress check from category check
Takashi Sakamoto [Thu, 31 Dec 2015 04:58:11 +0000 (13:58 +0900)]
ALSA: dice: split subaddress check from category check

Before allocating an instance of sound card, ALSA dice driver checks
chip_ID_hi in Bus information block of Config ROM, then also checks
subaddresses. The former operation reads cache of Config ROM in Linux
FireWire subsystem, while the latter operation sends read transaction.
The latter can be merged into initialization of transaction system.

This commit splits these two operations to reduce needless transactions
in probe processing.

Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
8 years agoMerge remote-tracking branch 'asoc/fix/intel' into asoc-linus
Mark Brown [Tue, 5 Jan 2016 23:07:33 +0000 (23:07 +0000)]
Merge remote-tracking branch 'asoc/fix/intel' into asoc-linus

8 years agoMerge remote-tracking branch 'asoc/fix/rt5645' into asoc-linus
Mark Brown [Tue, 5 Jan 2016 23:07:33 +0000 (23:07 +0000)]
Merge remote-tracking branch 'asoc/fix/rt5645' into asoc-linus

8 years agoMerge remote-tracking branch 'asoc/fix/dapm' into asoc-linus
Mark Brown [Tue, 5 Jan 2016 23:07:32 +0000 (23:07 +0000)]
Merge remote-tracking branch 'asoc/fix/dapm' into asoc-linus

8 years agoMerge remote-tracking branch 'asoc/fix/arizona' into asoc-linus
Mark Brown [Tue, 5 Jan 2016 23:07:32 +0000 (23:07 +0000)]
Merge remote-tracking branch 'asoc/fix/arizona' into asoc-linus

8 years agoASoC: hdac_hdmi: Fix to check num nodes correctly
Subhransu S. Prusty [Wed, 9 Dec 2015 16:16:08 +0000 (21:46 +0530)]
ASoC: hdac_hdmi: Fix to check num nodes correctly

commit 3c83ac23253c ("ASoC: hdac_hdmi: check error return") fixes
the static checker warning reported by Dan Carpenter:

sound/soc/codecs/hdac_hdmi.c:416 hdac_hdmi_parse_and_map_nid()
warn: unsigned 'hdac->num_nodes' is never less than zero.

But it doesn't fix the issue completely.

It's also a failure if no sub nodes found for an afg node. So modify
the return condition appropriately.

Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: da7218: Correct BCLK inversion for DSP DAI format mode
Adam Thomson [Tue, 5 Jan 2016 18:15:33 +0000 (18:15 +0000)]
ASoC: da7218: Correct BCLK inversion for DSP DAI format mode

By default the device latches data on the falling edge of the
BCLK in DSP mode, whereas the expectation for normal BCLK is to
latch on the rising edge. This updates the driver to invert the
BCLK configuration for DSP mode, to align with expected behaviour.

Signed-off-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoMerge branch 'fix/intel' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie...
Mark Brown [Tue, 5 Jan 2016 18:06:29 +0000 (18:06 +0000)]
Merge branch 'fix/intel' of git://git./linux/kernel/git/broonie/sound into asoc-intel

8 years agoASoC: pcm: allow delayed suspending request by users
Sanyog Kale [Tue, 5 Jan 2016 11:44:49 +0000 (17:14 +0530)]
ASoC: pcm: allow delayed suspending request by users

If a device would like to use delayed suspending then PM
recommendation is to set ‘power.use_autosuspend’ flag. To allow
users to do so we need to change runtime calls in core to use
autosuspend counterparts.

For user who do not wish to use delayed suspend not setting the
device's ‘power.use_autosuspend’ flag will result in non-delayed
suspend even with these APIs which incidentally is also the default
behaviour, so only users will be impacted who opt in for this.

Signed-off-by: Sanyog Kale <sanyog.r.kale@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: da7219: Correct BCLK inversion for DSP DAI format mode
Adam Thomson [Tue, 5 Jan 2016 15:05:36 +0000 (15:05 +0000)]
ASoC: da7219: Correct BCLK inversion for DSP DAI format mode

By default the device latches data on the falling edge of the
BCLK in DSP mode, whereas the expectation for normal BCLK is to
latch on the rising edge. This updates the driver to invert the
BCLK configuration for DSP mode, to align with expected behaviour.

Signed-off-by: Adam Thomson <Adam.Thomson.Opensource@diasemi.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: fsl_asrc: sound is wrong after suspend/resume
Zidan Wang [Thu, 24 Dec 2015 03:42:11 +0000 (11:42 +0800)]
ASoC: fsl_asrc: sound is wrong after suspend/resume

The register ASRCFG is volatile, but some bits need to be recovered
after suspend/resume.

Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
8 years agoASoC: Intel: Atom: Add support for HP ElitePad 1000 G2
Jorge Fernandez Monteagudo [Mon, 4 Jan 2016 23:20:30 +0000 (17:20 -0600)]
ASoC: Intel: Atom: Add support for HP ElitePad 1000 G2

The BIOS for the HP ElitePad 1000 G2 uses an unexpected HID,
(INTCCFFD), add it to the white list of knowns HIDs.

Signed-off-by: Jorge Fernandez Monteagudo <jorgefm@cirsa.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>