The PCM hw_params core function tries to clear up the PCM buffer
before actually using for avoiding the information leak from the
previous usages or the usage before a new allocation.  It performs the
memset() with runtime->dma_bytes, but this might still leave some
remaining bytes untouched; namely, the PCM buffer size is aligned in
page size for mmap, hence runtime->dma_bytes doesn't necessarily cover
all PCM buffer pages, and the remaining bytes are exposed via mmap.

This patch changes the memory clearance to cover the all buffer pages
if the stream is supposed to be mmap-ready (that guarantees that the
buffer size is aligned in page size).
Reviewed-by: NLars-Peter Clausen <lars@metafoo.de>
Link: https://lore.kernel.org/r/20201218145625.2045-3-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Kernel-doc markups should use this format:
        identifier - description

There is a common comment marked, instead, with kernel-doc
notation.

Some identifiers have different names between their prototypes
and the kernel-doc markup.
Signed-off-by: NMauro Carvalho Chehab <mchehab+huawei@kernel.org>
Acked-by: NMark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/535182d6f55d7a7de293dda9676df68f5f60afc6.1603469755.git.mchehab+huawei@kernel.orgSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Replace the existing /* fall through */ comments and its variants with
the new pseudo-keyword macro fallthrough[1]. Also, remove unnecessary
fall-through markings when it is the case.

[1] https://www.kernel.org/doc/html/latest/process/deprecated.html?highlight=fallthrough#implicit-switch-case-fall-throughSigned-off-by: NGustavo A. R. Silva <gustavoars@kernel.org>
Link: https://lore.kernel.org/r/20200708203236.GA5112@embeddedorSigned-off-by: NTakashi Iwai <tiwai@suse.de>

We avoided the explicit use of dma_mmap_coherent() on x86 because of a
spurious warning in x86 APT code in the past.  However, this blindly
assumes that the pages allocated via dma_alloc_coherent() on x86 are
the ones convertible via virt_to_page() (that is used in the default
mmap handler), and it's no longer true; with the indirect DMA ops,
this can be handled differently.  The only certain way for doing mmap
such pages is the dma_mmap_coherent(), and the warning seems already
gone in the recent code, so let's use it consistently.

Link: https://lore.kernel.org/r/20200615160045.2703-2-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Prevent SNDRV_PCM_IOCTL_LINK linking stream to itself - the code
can't handle it. Fixed commit is not where bug was introduced, but
changes the context significantly.

Cc: stable@vger.kernel.org
Fixes: 0888c321 ("pcm_native: switch to fdget()/fdput()")
Signed-off-by: NMichał Mirosław <mirq-linux@rere.qmqm.pl>
Link: https://lore.kernel.org/r/89c4a2487609a0ed6af3ecf01cc972bdc59a7a2d.1591634956.git.mirq-linux@rere.qmqm.plSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Add and use snd_pcm_stream_lock_nested() in snd_pcm_link/unlink
implementation.  The code is fine, but generates a lockdep complaint:

============================================
WARNING: possible recursive locking detected
5.7.1mq+ #381 Tainted: G           O
--------------------------------------------
pulseaudio/4180 is trying to acquire lock:
ffff888402d6f508 (&group->lock){-...}-{2:2}, at: snd_pcm_common_ioctl+0xda8/0xee0 [snd_pcm]

but task is already holding lock:
ffff8883f7a8cf18 (&group->lock){-...}-{2:2}, at: snd_pcm_common_ioctl+0xe4e/0xee0 [snd_pcm]

other info that might help us debug this:
 Possible unsafe locking scenario:

       CPU0
       ----
  lock(&group->lock);
  lock(&group->lock);

 *** DEADLOCK ***

 May be due to missing lock nesting notation

2 locks held by pulseaudio/4180:
 #0: ffffffffa1a05190 (snd_pcm_link_rwsem){++++}-{3:3}, at: snd_pcm_common_ioctl+0xca0/0xee0 [snd_pcm]
 #1: ffff8883f7a8cf18 (&group->lock){-...}-{2:2}, at: snd_pcm_common_ioctl+0xe4e/0xee0 [snd_pcm]
[...]

Cc: stable@vger.kernel.org
Fixes: f57f3df0 ("ALSA: pcm: More fine-grained PCM link locking")
Signed-off-by: NMichał Mirosław <mirq-linux@rere.qmqm.pl>
Link: https://lore.kernel.org/r/37252c65941e58473b1219ca9fab03d48f47e3e3.1591610330.git.mirq-linux@rere.qmqm.plSigned-off-by: NTakashi Iwai <tiwai@suse.de>

... and no, __put_user() doesn't help here - skipping
access_ok() on the second call does not remove the
possibility of page having become unmapped or r/o
in the meanwhile
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Call cpu_latency_qos_add/update/remove_request() and
cpu_latency_qos_request_active() instead of
pm_qos_add/update/remove_request() and pm_qos_request_active(),
respectively, because the latter are going to be dropped.

No intentional functional impact.
Signed-off-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Reviewed-by: NUlf Hansson <ulf.hansson@linaro.org>
Acked-by: NMark Brown <broonie@kernel.org>
Acked-by: NTakashi Iwai <tiwai@suse.de>
Reviewed-by: NAmit Kucheria <amit.kucheria@linaro.org>
Tested-by: NAmit Kucheria <amit.kucheria@linaro.org>

The commit 66f2d19f ("ALSA: pcm: Fix memory leak at closing a
stream without hw_free") tried to fix the regression wrt the missing
hw_free call at closing without SNDRV_PCM_IOCTL_HW_FREE ioctl.
However, the code change dropped mistakenly the state check, resulting
in calling hw_free twice when SNDRV_PCM_IOCTL_HW_FRE got called
beforehand.  For most drivers, this is almost harmless, but the
drivers like SOF show another regression now.

This patch adds the state condition check before calling do_hw_free()
at releasing the stream for avoiding the double hw_free calls.

Fixes: 66f2d19f ("ALSA: pcm: Fix memory leak at closing a stream without hw_free")
Reported-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Reported-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Tested-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/s5hd0ajyprg.wl-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

The parameter bit mask needs often explicit cast with __force,
e.g. for the PCM subformat type.  Instead of adding __force at each
place, which is error prone, this patch introduces a new macro and
replaces the all bit shift with it.  This fixes the sparse warnings
like the following:
  sound/core/pcm_native.c:2508:30: warning: restricted snd_pcm_access_t degrades to integer

No functional changes, just sparse warning fixes.

Link: https://lore.kernel.org/r/20200206163945.6797-7-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Simplify the code with the new macros for PCM format type iterations.
This fixes the sparse warnings nicely:
  sound/core/pcm_native.c:2302:26: warning: restricted snd_pcm_format_t degrades to integer
  sound/core/pcm_native.c:2306:54: warning: incorrect type in argument 1 (different base types)
  sound/core/pcm_native.c:2306:54:    expected restricted snd_pcm_format_t [usertype] format
  sound/core/pcm_native.c:2306:54:    got unsigned int [assigned] k
  ....

No functional changes, just sparse warning fixes.

Link: https://lore.kernel.org/r/20200206163945.6797-6-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Since we have a bitwise definition of snd_pcm_state_t and use it for
certain struct fields, a few new (and years old) sparse warnings came
up.  This patch is an attempt to cover them.

- The state fields in snd_pcm_mmap_status* and co are all defined as
  snd_pcm_state_t type now

- The PCM action callbacks take snd_pcm_state_t argument as well;
  some actions taking special values got the explicit cast and
  comments

- For the PCM action that doesn't need an extra argument receives
  ACTION_ARG_IGNORE instead of ambiguous 0

While we're at it, the boolean argument is also properly changed to
bool and true/false, as well as a slight refactoring of PCM pause
helper function to make easier to read.

No functional changes, just shutting up chatty sparse.

Fixes: 46b770f7 ("ALSA: uapi: Fix sparse warning")
Reported-by: Nkbuild test robot <lkp@intel.com>
Link: https://lore.kernel.org/r/20200131152214.11698-1-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

ALSA PCM core recently introduced a new managed PCM buffer allocation
mode that does allocate / free automatically at hw_params and
hw_free.  However, it overlooked the code path directly calling
hw_free PCM ops at releasing the PCM substream, and it may result in a
memory leak as spotted by syzkaller when no buffer preallocation is
used (e.g. vmalloc buffer).

This patch papers over it with a slight refactoring.  The hw_free ops
call and relevant tasks are unified in a new helper function, and call
it from both places.

Fixes: 0dba808e ("ALSA: pcm: Introduce managed buffer allocation mode")
Reported-by: syzbot+30edd0f34bfcdc548ac4@syzkaller.appspotmail.com
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20200129195907.12197-1-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Both snd_pcm_hw_constraints_init() and _complete() functions are
called only from pcm_native.c, hence they can be static for further
optimization.

Link: https://lore.kernel.org/r/20200116162825.24792-1-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

The current PCM code doesn't initialize explicitly the buffers
allocated for PCM streams, hence it might leak some uninitialized
kernel data or previous stream contents by mmapping or reading the
buffer before actually starting the stream.

Since this is a common problem, this patch simply adds the clearance
of the buffer data at hw_params callback.  Although this does only
zero-clear no matter which format is used, which doesn't mean the
silence for some formats, but it should be OK because the intention is
just to clear the previous data on the buffer.
Reported-by: NLionel Koenig <lionel.koenig@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20191211155742.3213-1-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

The snd_pcm_mmap_status and snd_pcm_mmap_control interfaces are one of the
trickiest areas to get right when moving to 64-bit time_t in user space.

The snd_pcm_mmap_status structure layout is incompatible with user space
that uses a 64-bit time_t, so we need a new layout for it. Since the
SNDRV_PCM_IOCTL_SYNC_PTR ioctl combines it with snd_pcm_mmap_control
into snd_pcm_sync_ptr, we need to change those two as well.

Both structures are also exported via an mmap() operation on certain
architectures, and this suffers from incompatibility between 32-bit
and 64-bit user space. As we have to change both structures anyway,
this is a good opportunity to fix the mmap() problem as well, so let's
standardize on the existing 64-bit layout of the structure where possible.

The downside is that we lose mmap() support for existing 32-bit x86 and
powerpc applications, adding that would introduce very noticeable runtime
overhead and complexity. My assumption here is that not too many people
will miss the removed feature, given that:

- Almost all x86 and powerpc users these days are on 64-bit kernels,
the majority of today's 32-bit users are on architectures that never
supported mmap (ARM, MIPS, ...).
- It never worked in compat mode (it was intentionally disabled there)
- The application already needs to work with a fallback to
SNDRV_PCM_IOCTL_SYNC_PTR, which will keep working with both the old
and new structure layout.

Both the ioctl() and mmap() based interfaces are changed at the same
time, as they are based on the same structures. Unlike other interfaces,
we change the uapi header to export both the traditional structure and
a version that is portable between 32-bit and 64-bit user space code
and that corresponds to the existing 64-bit layout. We further check the
__USE_TIME_BITS64 macro that will be defined by future C library versions
whenever we use the new time_t definition, so any existing user space
source code will not see any changes until it gets rebuilt against a new
C library. However, the new structures are all visible in addition to the
old ones, allowing applications to explicitly request the new structures.

In order to detect the difference between the old snd_pcm_mmap_status and
the new __snd_pcm_mmap_status64 structure from the ioctl command number,
we rely on one quirk in the structure definition: snd_pcm_mmap_status
must be aligned to alignof(time_t), which leads the compiler to insert
four bytes of padding in struct snd_pcm_sync_ptr after 'flags' and a
corresponding change in the size of snd_pcm_sync_ptr itself. On x86-32
(and only there), the compiler doesn't use 64-bit alignment in structure,
so I'm adding an explicit pad in the structure that has no effect on the
existing 64-bit architectures but ensures that the layout matches for x86.

The snd_pcm_uframes_t type compatibility requires another hack: we can't
easily make that 64 bit wide, so I leave the type as 'unsigned long',
but add padding before and after it, to ensure that the data is properly
aligned to the respective 64-bit field in the in-kernel structure.

For the SNDRV_PCM_MMAP_OFFSET_STATUS/CONTROL constants that are used
as the virtual file offset in the mmap() function, we also have to
introduce new constants that depend on hte __USE_TIME_BITS64 macro:
The existing macros are renamed to SNDRV_PCM_MMAP_OFFSET_STATUS_OLD
and SNDRV_PCM_MMAP_OFFSET_CONTROL_OLD, they continue to work fine on
64-bit architectures, but stop working on native 32-bit user space.
The replacement _NEW constants are now used by default for user space
built with __USE_TIME_BITS64, those now work on all new kernels for x86,
ppc and alpha (32 and 64 bit, native and compat). It might be a good idea
for a future alsa-lib to support both the _OLD and _NEW macros and use
the corresponding structures directly. Unmodified alsa-lib source code
will retain the current behavior, so it will no longer be able to use
mmap() for the status/control structures on 32-bit systems, until either
the C library gets updated to 64-bit time_t or alsa-lib gets updated to
support both mmap() layouts.

Co-developed-with: Baolin Wang <baolin.wang@linaro.org>
Signed-off-by: NBaolin Wang <baolin.wang@linaro.org>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

This is a preparation patch, moving the compat handler for
snd_pcm_ioctl_sync_ptr_compat from pcm_compat.c to pcm_native.c.
No other changes are indented.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NBaolin Wang <baolin.wang@linaro.org>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

The struct snd_pcm_status will use 'timespec' type variables to record
timestamp, which is not year 2038 safe on 32bits system.

Userspace will use SNDRV_PCM_IOCTL_STATUS and SNDRV_PCM_IOCTL_STATUS_EXT
as commands to issue ioctl() to fill the 'snd_pcm_status' structure in
userspace. The command number is always defined through _IOR/_IOW/IORW,
so when userspace changes the definition of 'struct timespec' to use
64-bit types, the command number also changes.

Thus in the kernel, we now need to define two versions of each such ioctl
and corresponding ioctl commands to handle 32bit time_t and 64bit time_t
in native mode:
struct snd_pcm_status32 {
	......

	s32 trigger_tstamp_sec;
	s32 trigger_tstamp_nsec;

	......

	s32 audio_tstamp_sec;
	s32 audio_tstamp_nsec;

	......
};

struct snd_pcm_status64 {
	......

	s32 trigger_tstamp_sec;
	s32 trigger_tstamp_nsec;

	......

	s32 audio_tstamp_sec;
	s32 audio_tstamp_nsec;

	......
};

Moreover in compat file, we renamed or introduced new structures to handle
32bit/64bit time_t in compatible mode. The 'struct snd_pcm_status32' and
snd_pcm_status_user32() are used to handle 32bit time_t in compat mode.
'struct compat_snd_pcm_status64' and snd_pcm_status_user_compat64() are used
to handle 64bit time_t.

The implicit padding before timespec is made explicit to avoid incompatible
structure layout between 32-bit and 64-bit x86 due to the different
alignment requirements, and the snd_pcm_status structure is now hidden
from the kernel to avoid relying on the timespec definitio definitionn

Finally we can replace SNDRV_PCM_IOCTL_STATUS and SNDRV_PCM_IOCTL_STATUS_EXT
with new commands and introduce new functions to fill new 'struct snd_pcm_status64'
instead of using unsafe 'struct snd_pcm_status'. Then in future, the new
commands can be matched when userspace changes 'timespec' to 64bit type
to make a size change of 'struct snd_pcm_status'. When glibc changes time_t
to 64-bit, any recompiled program will issue ioctl commands that the kernel
does not understand without this patch.
Signed-off-by: NBaolin Wang <baolin.wang@linaro.org>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

Since timespec is not year 2038 safe on 32bit system, and we need to
convert all timespec variables to timespec64 type for sound subsystem.

This patch is used to do preparation for following patches, that will
convert all structures defined in uapi/sound/asound.h to use 64-bit
time_t.
Signed-off-by: NBaolin Wang <baolin.wang@linaro.org>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>

Many PCI and other drivers performs snd_pcm_period_elapsed() simply in
its interrupt handler, so the sync_stop operation is just to call
synchronize_irq().  Instead of putting this call multiple times,
introduce the common card->sync_irq field.  When this field is set,
PCM core performs synchronize_irq() for sync-stop operation.  Each
driver just needs to copy its local IRQ number to card->sync_irq, and
that's all we need.

Link: https://lore.kernel.org/r/20191117085308.23915-8-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

The standard programming model of a PCM sound driver is to process
snd_pcm_period_elapsed() from an interrupt handler.  When a running
stream is stopped, PCM core calls the trigger-STOP PCM ops, sets the
stream state to SETUP, and moves on to the next step.  This is
performed in an atomic manner -- this could be called from the interrupt
context, after all.

The problem is that, if the stream goes further and reaches to the
CLOSE state immediately, the stream might be still being processed in
snd_pcm_period_elapsed() in the interrupt context, and hits a NULL
dereference.  Such a crash happens because of the atomic operation,
and we can't wait until the stream-stop finishes.

For addressing such a problem, this commit adds a new PCM ops,
sync_stop.  This gets called at the appropriate places that need a
sync with the stream-stop, i.e. at hw_params, prepare and hw_free.

Some drivers already have a similar mechanism implemented locally, and
we'll refactor the code later.

Link: https://lore.kernel.org/r/20191117085308.23915-7-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Currently PCM ioctl ops is a mandatory field but almost all drivers
simply pass snd_pcm_lib_ioctl.  For simplicity, allow to set NULL in
the field and call snd_pcm_lib_ioctl() as default.

Link: https://lore.kernel.org/r/20191117085308.23915-4-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

This patch adds the support for the feature to automatically allocate
and free PCM buffers, so called "managed buffer allocation" mode.
It's set up via new PCM helpers, snd_pcm_set_managed_buffer() and
snd_pcm_set_managed_buffer_all(), both of which correspond to the
existing preallocator helpers, snd_pcm_lib_preallocate_pages() and
snd_pcm_lib_preallocate_pages_for_all().  When the new helper is used,
it not only performs the pre-allocation of buffers, but also it
manages to call snd_pcm_lib_malloc_pages() before the PCM hw_params
ops and snd_lib_pcm_free() after the PCM hw_free ops inside PCM core,
respectively.  This allows drivers to drop the explicit calls of the
memory allocation / release functions, and it will be a good amount of
code reduction in the end of this patch series.

When the PCM substream is set to the managed buffer allocation mode,
the managed_buffer_alloc flag is set in the substream object.  Since
some drivers want to know when a buffer is newly allocated or
re-allocated at hw_params callback (e.g. want to set up the additional
stuff for the given buffer only at allocation time), now PCM core
turns on buffer_changed flag when the buffer has changed.

The standard conversions to use the new API will be straightforward:
- Replace snd_pcm_lib_preallocate*() calls with the corresponding
  snd_pcm_set_managed_buffer*(); the arguments should be unchanged
- Drop superfluous snd_pcm_lib_malloc() and snd_pcm_lib_free() calls;
  the check of snd_pcm_lib_malloc() returns should be replaced with
  the check of runtime->buffer_changed flag.
- If hw_params or hw_free becomes empty, drop them from PCM ops

Link: https://lore.kernel.org/r/20191117085308.23915-2-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

For non-x86 architectures, SNDRV_DMA_TYPE_DEV_UC should be treated
equivalent with SNDRV_DMA_TYPE_DEV, where the default mmap handler
still checks only about SNDRV_DMA_TYPE_DEV.  Make the check more
proper.

Note that all existing users of *_UC buffer types are x86-only, so
this doesn't fix any bug, but just for consistency.

Fixes: 42e748a0 ("ALSA: memalloc: Add non-cached buffer type")
Link: https://lore.kernel.org/r/20191108165626.5947-1-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

When a driver needs to deal with a special buffer like a SG or a
vmalloc buffer, it has to set up the PCM page ops explicitly for the
corresponding helper function.  This is rather error-prone and many
people forgot or incorrectly used it.

For simplifying the call patterns and avoiding such a potential bug,
this patch enhances the PCM default mmap handler to check the
(pre-)allocated buffer type and handles the page gracefully depending
on the buffer type.  If the PCM page ops is given, the ops is still
used in a higher priority.  The new code path is only for the default
(NULL page ops) case.

Link: https://lore.kernel.org/r/20191105080138.1260-4-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

The check for the mmap support via hw_support_mmap() function misses
the case where the device is with SNDRV_DMA_TYPE_DEV_UC, which should
have been treated equally as SNDRV_DMA_TYPE_DEV.  Let's fix it.

Note that this bug doesn't hit any practical problem, because
SNDRV_DMA_TYPE_DEV_UC is used only for x86-specific drivers
(snd-hda-intel and snd-intel8x0) for the specific platforms that need
the non-cached buffers.  And, on such platforms, hw_support_mmap()
already returns true in anyway.  That's the reason I didn't put
Cc-to-stable mark here.  This is only for any theoretical future
extension.

Fixes: 425da159 ("ALSA: pcm: use dma_can_mmap() to check if a device supports dma_mmap_*")
Fixes: 42e748a0 ("ALSA: memalloc: Add non-cached buffer type")
Link: https://lore.kernel.org/r/20191104101115.27311-1-tiwai@suse.deSigned-off-by: NTakashi Iwai <tiwai@suse.de>

Replace the local hack with the dma_can_mmap helper to check if
a given device supports mapping DMA allocations to userspace.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NTakashi Iwai <tiwai@suse.de>

Most of the modern codecs supports 352.8KHz and 384KHz sample rates.
Currenlty HW params fails to set 352.8Kz and 384KHz sample rate
as these are not in known rates list.
Add these new rates to known list to allow them.

This patch also adds defines in pcm.h so that drivers can use it.
Signed-off-by: NVidyakumar Athota <vathota@codeaurora.org>
Signed-off-by: NBanajit Goswami <bgoswami@codeaurora.org>
Signed-off-by: NSrinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20190822095653.7200-2-srinivas.kandagatla@linaro.orgSigned-off-by: NMark Brown <broonie@kernel.org>

lost wakeup can occur after enabling irq, therefore put task
into interruptible before enabling interrupts,

without this change, task can be put to sleep and snd_pcm_drain
will delay

Fixes: f2b3614c ("ALSA: PCM - Don't check DMA time-out too shortly")
Signed-off-by: NYuki Tsunashima <ytsunashima@jp.adit-jv.com>
Signed-off-by: NSuresh Udipi <sudipi@jp.adit-jv.com>
[ported from 4.9]
Signed-off-by: NAdam Miartus <amiartus@de.adit-jv.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The recent rewrite of PCM link lock management introduced the refcount
in snd_pcm_group object, managed by the kernel refcount_t API.  This
caused unexpected kernel warnings when the kernel is built with
CONFIG_REFCOUNT_FULL=y.  As the warning line indicates, the problem is
obviously that we start with refcount=0 and do refcount_inc() for
adding each PCM link, while refcount_t API doesn't like refcount_inc()
performed on zero.

For adapting the proper refcount_t usage, this patch changes the logic
slightly:
- The initial refcount is 1, assuming the single list entry
- The refcount is incremented / decremented at each PCM link addition
  and deletion
- ... which allows us concentrating only on the refcount as a release
  condition

Fixes: f57f3df0 ("ALSA: pcm: More fine-grained PCM link locking")
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=204221Reported-and-tested-by: NDuncan Overbruck <kernel@duncano.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Based on 1 normalized pattern(s):

  this program is free software you can redistribute it and or modify
  it under the terms of the gnu general public license as published by
  the free software foundation either version 2 of the license or at
  your option any later version this program is distributed in the
  hope that it will be useful but without any warranty without even
  the implied warranty of merchantability or fitness for a particular
  purpose see the gnu general public license for more details you
  should have received a copy of the gnu general public license along
  with this program if not write to the free software foundation inc
  59 temple place suite 330 boston ma 02111 1307 usa

extracted by the scancode license scanner the SPDX license identifier

  GPL-2.0-or-later

has been chosen to replace the boilerplate/reference in 1334 file(s).
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NAllison Randal <allison@lohutok.net>
Reviewed-by: NRichard Fontana <rfontana@redhat.com>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190527070033.113240726@linutronix.deSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Currently PCM core sets each opened stream forcibly to SUSPENDED state
via snd_pcm_suspend_all() call, and the user-space is responsible for
re-triggering the resume manually either via snd_pcm_resume() or
prepare call.  The scheme works fine usually, but there are corner
cases where the stream can't be resumed by that call: the streams
still in OPEN state before finishing hw_params.  When they are
suspended, user-space cannot perform resume or prepare because they
haven't been set up yet.  The only possible recovery is to re-open the
device, which isn't nice at all.  Similarly, when a stream is in
DISCONNECTED state, it makes no sense to change it to SUSPENDED
state.  Ditto for in SETUP state; which you can re-prepare directly.

So, this patch addresses these issues by filtering the PCM streams to
be suspended by checking the PCM state.  When a stream is in either
OPEN, SETUP or DISCONNECTED as well as already SUSPENDED, the suspend
action is skipped.

To be noted, this problem was originally reported for the PCM runtime
PM on HD-audio.  And, the runtime PM problem itself was already
addressed (although not intended) by the code refactoring commits
3d21ef0b ("ALSA: pcm: Suspend streams globally via device type PM
ops") and 17bc4815 ("ALSA: pci: Remove superfluous
snd_pcm_suspend*() calls").  These commits eliminated the
snd_pcm_suspend*() calls from the runtime PM suspend callback code
path, hence the racy OPEN state won't appear while runtime PM.
(FWIW, the race window is between snd_pcm_open_substream() and the
first power up in azx_pcm_open().)

Although the runtime PM issue was already "fixed", the same problem is
still present for the system PM, hence this patch is still needed.
And for stable trees, this patch alone should suffice for fixing the
runtime PM problem, too.
Reported-and-tested-by: NJon Hunter <jonathanh@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

BE dai links only have internal PCM's and their substream ops may
not be set. Suspending these PCM's will result in their
 ops->trigger() being invoked and cause a kernel oops.
So skip suspending PCM's if their ops are NULL.

[ NOTE: this change is required now for following the recent PCM core
  change to get rid of snd_pcm_suspend() call.  Since DPCM BE takes
  the runtime carried from FE while keeping NULL ops, it can hit this
  bug.  See details at:
     https://github.com/thesofproject/linux/pull/582
  -- tiwai ]
Signed-off-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Declaration of snd_pcm_drop() in sound/core/pcm_native.c is superfluous
since the function isn't called before being defined. Remove the
declaration.
Signed-off-by: NGuennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

It's assigned but nowhere used.  Let's remove it.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

After the previous code refactoring, the PCM stream locking code
became nothing but the PCM group lock with self_group object.  Use the
existing helper function for simplifying the code.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Remove the hackish down_write_nonfifo() that was introduced as a
workaround of rwsem deadlock.

It used to be a problem for non-atomic PCM streams that take the rwsem
for the locking and hit the high lock contention.  Since the current
PCM locking refactoring, we'll no longer hit it as the hot code-paths
don't take global locks.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

We have currently two global locks, a rwlock and a rwsem, that are
used for managing linking the PCM streams.  Due to these global locks,
once when a linked stream is used, the lock granularity suffers a
lot.

This patch attempts to eliminate the former global lock for atomic
ops.  The latter rwsem needs remaining because of the loosy way of the
loop calls in snd_pcm_action_nonatomic(), as well as for avoiding the
deadlock at linking.  However, these are used far rarely, actually
only by two actions (prepare and  reset), where both are no timing
critical ones.  So this can be still seen as a good improvement.

The basic strategy to eliminate the rwlock is to assure group->lock at
adding or removing a stream to / from the group.  Since we already
takes the group lock whenever taking the all substream locks under the
group, this shouldn't be a big problem.  The reference to group
pointer in snd_pcm_substream object is protected by the stream lock
itself.

However, there are still pitfalls: a race window at re-locking and the
lifecycle of group object.  The former is a small race window for
dereferencing the substream group object opened while snd_pcm_action()
performs re-locking to avoid ABBA deadlocks.  This includes the unlink
of group during that window, too.  And the latter is the kfree
performed after all streams are removed from the group while it's
still dereferenced.

For addressing these corner cases, two new tricks are introduced:
- After re-locking, the group assigned to the stream is checked again;
  if the group is changed, we retry the whole procedure.
- Introduce a refcount to snd_pcm_group object, so that it's freed
  only when it's empty and really no one refers to it.

(Some readers might wonder why not RCU for the latter.  RCU in this
case would cost more than refcounting, unfortunately.  We take the
group lock sooner or later, hence the performance improvement by RCU
would be negligible.  Meanwhile, because we need to deal with
schedulable context depending on the pcm->nonatomic flag, it'll become
dynamic RCU/SRCU switch, and the grace period may become too long.)

Along with these changes, there are a significant amount of code
refactoring.  The complex group re-lock & ref code is factored out to
snd_pcm_stream_group_ref() function, for example.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The snd_pcm_group_for_each_entry() loop found in snd_pcm_unlink() is
only for taking the first list entry.  Use list_first_entry() to make
clearer.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Make a common helper to re-assign the PCM link using list_move() instead
of open code with manual list_del() and list_add_tail().  This assures
the consistency and we can get rid of snd_pcm_group.count field -- its
purpose is only to check whether the list is singular, and we can know
it by list_is_singular() call now.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>