As Clemens Ladisch kindly explained:
 "Please note that there are two methods to identify alternate settings:
  the number, which is the value in bAlternateSetting, and the index,
  which is the index in the descriptor array.  There might be some wording
  in the USB spec that these two values must be the same, but in reality,
  [insert standard rant about firmware writers], bAlternateSetting
  must be treated as a random ID value."

This patch changes the name to express the correct usage semantics.
No functional change.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

If setting the interface fails, the SUBSTREAM_FLAG_SYNC_EP_STARTED
should be cleared.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The only call site for deactivate_endpoints() at snd_usb_hw_free().
The return value is not checked there, as it is irrelevant if it
fails on hw_free.
This patch moves the deactivation of the endpoints directly into
snd_usb_hw_free().
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This patch changes the way URBs are allocated and their sizes are
determined for PCM playback in the snd-usb-audio driver.  Currently
the driver allocates too few URBs for endpoints that don't use
implicit sync, making underruns more likely to occur.  This may be a
holdover from before I/O delays could be measured accurately; in any
case, it is no longer necessary.

The patch allocates as many URBs as possible, subject to four
limitations:

	The total number of URBs for the endpoint is not allowed to
	exceed MAX_URBS (which the patch increases from 8 to 12).

	The total number of packets per URB is not allowed to exceed
	MAX_PACKS (or MAX_PACKS_HS for high-speed devices), which is
	decreased from 20 to 6.

	The total duration of queued data is not allowed to exceed
	MAX_QUEUE, which is decreased from 24 ms to 18 ms.

	The total number of ALSA frames in the output queue is not
	allowed to exceed the ALSA buffer size.

The last requirement is the hardest to implement.  Currently the
number of URBs needed to fill a buffer cannot be determined in
advance, because a buffer contains a fixed number of frames whereas
the number of frames in an URB varies to match shifts in the device's
clock rate.  To solve this problem, the patch changes the logic for
deciding how many packets an URB should contain.  Rather than using as
many as possible without exceeding an ALSA period boundary, now the
driver uses only as many packets as needed to transfer a predetermined
number of frames.  As a result, unless the device's clock has an
exceedingly variable rate, the number of URBs making up each period
(and hence each buffer) will remain constant.

The overall effect of the patch is that playback works better in
low-latency settings.  The user can still specify values for
frames/period and periods/buffer that exceed the capabilities of the
hardware, of course.  But for values that are within those
capabilities, the performance will be improved.  For example, testing
shows that a high-speed device can handle 32 frames/period and 3
periods/buffer at 48 KHz, whereas the current driver starts to get
glitchy at 64 frames/period and 2 periods/buffer.

A side effect of these changes is that the "nrpacks" module parameter
is no longer used.  The patch removes it.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
CC: Clemens Ladisch <clemens@ladisch.de>
Tested-by: NDaniel Mack <zonque@gmail.com>
Tested-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Since the quirks all apply to implicit feedback (the source endpoint
is always a data endpoint), there's no need to set and check
a flag for it.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

An implicit feedback endpoint can only be a capture source. The
consumer (sink) of the implicit feedback endpoint is therefore limited
to playback EPs.
Check if the target endpoint is a playback first and remove redundant
checks.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Since implicit_fb is not changed, !implicit_fb will always
be true - it is set only after these checks.
Similarly, there's also no need to set it at the top of the function.

Change the type of implicit_fb to bool (more appropriate).
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Reverse logic on the conditions required to qualify for a sync endpoint
and remove one level of indendation.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Separate setting implicit feedback quirks from setting
a sync endpoint (which may also be explicit feedback or async).
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Setting the sync endpoint currently takes up about half of set_format().
Move it to a dedicated function.
No functional change.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Following general kernel style.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Code block does not compile when enabled.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

All the Roland/Edirol/BOSS USB audio devices that need implicit feedback
show this unambiguously in their descriptors, so it might be a good idea
to let the driver detect this.

This should make playback work correctly (at least with Jack) with the
following devices:
- BOSS GT-100
- BOSS JS-8 Jam Station
- Edirol M-16DX
- Roland GAIA SH-01
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Instead of reading bInterfaceProtocol from the descriptor whenever it's
needed, store this value in the audioformat structure.  Besides
simplifying some code, this will allow us to correctly handle vendor-
specific devices where the descriptors are marked with other values.
Signed-off-by: NClemens Ladisch <clemens@ladisch.de>

Add a function to handle conversion from snd_pcm_format_t
to bitwise with proper typing.

Change such conversions to use this function and silence sparse
warnings.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

There is quite some confusion around the bit-ordering in DSD samples,
and no general agreement that defines whether hardware is supposed to
expect the oldest sample in the MSB or the LSB of a byte.

ALSA will hence set the rule that on the software API layer, bytes
always carry the oldest bit in the most significant bit of a byte, and
the driver has to translate that at runtime in order to match the
hardware layout.

This patch adds support for this by adding a boolean flag to the
audio format struct.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

In order to provide a compatibility way for pushing DSD
samples through ordinary PCM channels, the "DoP open Standard" was
invented. See http://www.dsd-guide.com for the official document.

The host is required to stuff DSD marker bytes (0x05, 0xfa,
alternating) in the MSB of 24 bit wide samples on the bus, in addition
to the 16 bits of actual DSD sample payload.

To support this, the hardware and software stride logic in the driver
has to be tweaked a bit, as we make the userspace believe we're
operating on 16 bit samples, while we in fact push one more byte per
channel down to the hardware.

The DOP runtime information is stored in struct snd_usb_substream, so
we can keep track of our state across multiple calls to
prepare_playback_urb_dsd_dop().
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

For normal PCM transfer, this change has no effect, as the endpoint's
stride is always frame_bits/8. For DSD DOP streams, however, which is
added later, the hardware stride differs from the software stride, and
the endpoint has the correct information in these cases.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

When recording at 176.2KHz or 192Khz, the device adds a 32-bit length
header to the capture packets, which obviously needs to be ignored for
recording to work properly.

Userspace expected:  L0 L1 L2 R0 R1 R2
...but actually got: R2 L0 L1 L2 R0 R1

Also, the last byte of the length header being interpreted as L0 of
the first sample caused spikes every 0.5ms, resulting in a loud 16KHz
tone (about the highest 'B' on a piano) being present throughout
captures.

Tested at all sample rates on an E-Mu 0404USB, and tested for
regressions on a generic USB headset.
Signed-off-by: NCalvin Owens <jcalvinowens@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

It turns out the devices from Playback Design need the delay quirk
after usb_set_interface from clocks.c as well. Make it a proper
quirks function and factor out the code to quirks.c.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Correct spelling of snd_usb_endpoint_implict_feedback_sink in all
occurances.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Change occurances of list_for_each into list_for_each_entry where
applicable.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

"Playback Design" products need a 50ms delay after setting the USB
interface.
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Reported-by: NAndreas Koch <andreas@akdesigninc.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Adds quirks and mixer support for the M-Audio Fast Track C600 USB
audio interface. This device is very similar to the C400 - the C600
simply has some more inputs and outputs, so the existing C400 support
is extended to support this device as well.
Signed-off-by: NMatt Gruskin <matthew.gruskin@gmail.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Because currently snd_printd() and snd_printdd() macros are expanded
to empty when CONFIG_SND_DEBUG=n, a compile warning like below
appears sometimes, and we had to covert it by ugly ifdefs:
  sound/pci/hda/patch_sigmatel.c: In function ‘stac92hd71bxx_fixup_hp’:
  sound/pci/hda/patch_sigmatel.c:2434:24: warning: unused variable ‘spec’ [-Wunused-variable]

For "fixing" these issues better, this patch replaces snd_printd() and
snd_printdd() definitions with empty inline functions instead of
macros.  This should have the same effect but shut up warnings like
above.

But since we had already put ifdefs, changing to inline functions
would trigger compile errors.  So, such ifdefs is removed in this
patch.

In addition, snd_pci_quirk name field is defined only when
CONFIG_SND_DEBUG_VERBOSE is set, and the reference to it in
snd_printdd() argument triggers the build errors, too.  For avoiding
these errors, introduce a new macro snd_pci_quirk_name() that is
defined no matter how the debug option is set.
Reported-by: NStratos Karafotis <stratosk@semaphore.gr>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The commit [0d9741c0: ALSA: usb-audio: sync ep init fix for
audioformat mismatch] introduced the correction of parameters to be
set for sync EP.  But since the new code assumes that the sync EP is
always paired with the data EP of another direction, it triggers Oops
when a device only with a single direction is used.

This patch adds a proper check of sync EP type and the presence of the
paired substream for avoiding the crash.
Reported-and-tested-by: NJens Axboe <axboe@kernel.dk>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Enable delay report on capture path. The delay is reset when an
URB is retired and increment at each call to .pointer based
on frame counter changes. The precision of the delay
information is limited to 1ms as in the playback case.

This reverts commit 3f94fad0.
Signed-off-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Commit 947d2996 , "ALSA: snd-usb:
properly initialize the sync endpoint", while correcting the
initialization of the sync endpoint when opening just the data
endpoint, prevents devices that has a sync endpoint, with a channel
number different than that of the data endpoint, from functioning.
Due to a different channel and period bytes count, attempting to
initialize the sync endpoint will fail at the usb host driver.
For example, when using xhci:

 cannot submit urb 0, error -90: internal error

With this patch, if a sync endpoint has multiple audioformats, a
matching audioformat is preferred. An audioformat must be found
with at least one channel and support the requested sample rate
and PCM format, otherwise the stream will not be opened.

If the number of channels differ between the selected audioformat
and the requested format, adjust the period bytes count accordingly.
It is safe to perform the calculation on the basis of the channel
count, since the requested PCM audio format and the rate must be
supported by the selected audioformat.

Cc: Jeffrey Barish <jeff_barish@earthlink.net>
Cc: Daniel Mack <zonque@gmail.com>
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The playback endpoint uses implicit feedback mode, similar
to the M-Audio FTU. Like with the FTU, we need to associate
the sync pipe ourselves.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

In this context, 0x01 is USB_ENDPOINT_XFER_ISOC.
Signed-off-by: NEldad Zack <eldad@fogrefinery.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

When a playback stream is paused, the stream isn't actually stopped,
thus we still need to take care of the in-flight data amount for the
delay calculation.  Otherwise the value of subs->last_delay is no
longer reliable and can give a bogus value after resuming from pause.
This will result in "delay: estimated XX, actual YY" error messages.

Also, during pause after all in flight data are processed
(i.e. last_delay = 0), we don't have to calculate the actual delay
from the current frame.  Give a short path in such a case.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

It doesn't make sense to calculate the delay for capture streams in
the current implementation.  It's always zero, so we should skip the
computation in snd_usb_pcm_pointer() in the case of capture.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Jeffrey Barish reported an obvious bug in the pcm part of the usb-audio
driver which causes the code to not initialize the sync endpoint from
configure_endpoint().
Reported-by: NJeffrey Barish <jeff_barish@earthlink.net>
Signed-off-by: NDaniel Mack <zonque@gmail.com>
Cc: stable@kernel.org [3.5+]
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

PCM hw_free and close should wait until all the pending stop
operations have been finished.  Basically only PCM trigger callback
should use non-wait calls.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

As we are stopping the endpoints asynchronously now, it's better to
trigger the stop of both data and sync endpoints and wait for pending
stopping operations, instead of the sequential trigger-and-wait
procedure.

So the wait argument in snd_usb_endpoint_stop() is dropped, and it's
expected that the caller synchronizes explicitly by calling
snd_usb_endpoint_sync_pending_stop().  (Actually there is only one
place calling this, so it was safe to change.)
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Reduce the redundant arguments for snd_usb_endpoint_start() and
snd_usb_endpoint_stop().  Also replaced from int to bool.

No functional changes by this commit.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

There are bug reports of a crash with USB-audio devices when PCM
prepare is performed immediately after the stream is stopped via
trigger callback.  It turned out that the problem is that we don't
wait until all URBs are killed.

This patch adds a new function to synchronize the pending stop
operation on an endpoint, and calls in the prepare callback for
avoiding the crash above.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=49181Reported-and-tested-by: NArtem S. Tashkinov <t.artem@lycos.com>
Cc: <stable@vger.kernel.org> [v3.6]
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Replace mutex with rwsem for codec->shutdown protection so that
concurrent accesses are allowed.

Also add the protection to snd_usb_autosuspend() and
snd_usb_autoresume(), too.
Reported-by: NMatthieu CASTET <matthieu.castet@parrot.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Close some races at disconnection of a USB audio device by adding the
chip->shutdown_mutex and chip->shutdown check at appropriate places.

The spots to put bandaids are:
- PCM prepare, hw_params and hw_free
- where the usb device is accessed for communication or get speed, in
 mixer.c and others; the device speed is now cached in subs->speed
 instead of accessing to chip->dev

The accesses in PCM open and close don't need the mutex protection
because these are already handled in the core PCM disconnection code.

The autosuspend/autoresume codes are still uncovered by this patch
because of possible mutex deadlocks.  They'll be covered by the
upcoming change to rwsem.

Also the mixer codes are untouched, too.  These will be fixed in
another patch, too.
Reported-by: NMatthieu CASTET <matthieu.castet@parrot.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The variable ep is initialized but never used
otherwise, so remove the unused variable.

dpatch engine is used to auto generate this patch.
(https://github.com/weiyj/dpatch)
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>