A return statement with an expression in a function returning void is
a constraint violation.
Signed-off-by: Michael Forney <mforney@mforney.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
snd_pcm_hw_sw_params() in pcm_hw.c tries to abuse the reserved bits
for passing period_Event flag. In this hackish way, we clear the
reserved bits at beginning, and restore before returning. However,
the code paths that return earlier don't restore the value, hence when
user calls this function twice, it may pass an unexpected value.
This patch fixes the failure, restoring the value always before
returning from the function.
Reported-by: Jamey Sharp <jamey@minilop.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A series of SNDRV_CTL_TLVO_XXX macro was introduced for position offset
of TLV data. This commit applies a code optimization.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Without this fix the application pointer would be reseted
whenever an application opens a device with SND_PCM_APPEND.
This would result in an Xrun if the device is already opened and
in running state and the appl_ptr is use.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The multiply defined versioned symbols have to be called with
INTERNAL() wrapper.
Add the missing declarations of versioned timer API functions in the
local header, and use them in the callers in PCM.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
SNDRV_PCM_IOCTL_SYNC_PTR command was introduced to PCM protocol/interface
in its version 2.0.7, however this command is used in a branch for the
newer version protocol/interface in snd_pcm_hw_forward().
This commit removes the superfluous code as a part of work for code
refactoring.
Fixes: eafb492512 ("- added SYNC_PTR ioctl support for pcm_hw plugin")
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
At present, applications can change avail_min parameter of PCM substream
by two ways; via mapped control data, and by executing ioctl(2) with
SNDRV_PCM_IOCTL_SYNC_PTR. The former is available in a case that the
applications map the data successfully.
When utilizing alsa-lib API, the above is done by a call of
'snd_pcm_sw_params()' to hw PCM plugin. In current implementation, this
call has an side-effect to issue appl_ptr unexpectedly.
This commit adds a helper function to issue avail_min without the
side-effect.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
After starting, PCM substream shift its state to running and applications
can move appl_ptr by several ways. When status and control data of runtime
of the PCM substream is not mapped, the applications should issue appl_ptr
to kernel land. In this case, when any PCM frames is handled by mmap
operation, the applications should issue appl_ptr to update.
This commit adds a helper function for this purpose. To avoid unexpected
change of avail_min, this commit uses a flag just to update appl_ptr.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
SNDRV_PCM_IOCTL_SYNC_PTR command for ioctl(2) with
SNDRV_PCM_SYNC_PTR_HWSYNC flag has an effect to update hw_ptr.
This is an alternative of SNDRV_PCM_IOCTL_HWSYNC but caller can get
current status simultaneously.
This commit adds a helper function just to issue hwsync. To avoid
side-effect to change appl_ptr and avail_min, this commit uses
SNDRV_PCM_SYNC_PTR_APPL and SNDRV_PCM_SYNC_PTR_AVAIL_MIN flags.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
When mapping status data successfully, mapped page includes status data
for applications. In this case, applications have no need to call ioctl(2)
with SNDRV_PCM_IOCTL_SYNC_PTR. However, in current implementation, when
map of control data is unavailable, applications execute the ioctl(2).
This is inconvenient for some cases that applications require to query
status only.
This commit adds a helper function to query status data without issuing
the control in fallback mode from failure of status mmap.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
When executing ioctl(2) with some commands, applications can request
ALSA PCM core to change appl_ptr in kernel space. Below is a list of
such operations:
- SNDRV_PCM_IOCTL_PREPARE
- SNDRV_PCM_IOCTL_RESET
- SNDRV_PCM_IOCTL_REWIND
- SNDRV_PCM_IOCTL_FORWARD
- SNDRV_PCM_IOCTL_WRITEI_FRAMES
- SNDRV_PCM_IOCTL_WRITEN_FRAMES
- SNDRV_PCM_IOCTL_READI_FRAMES
- SNDRV_PCM_IOCTL_READN_FRAMES
After these operations, the value of appl_ptr should be synchronized
between kernel/user spaces.
This commit adds a helper function to query status and control data
without issuing the control data just in fallback from failure of control
mapping.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
At failure of control data mapping, alsa-lib goes to fallback mode. In
this mode, a buffer is kept in user space and executes ioctl(2) with
SNDRV_PCM_IOCTL_SYNC_PTR for the buffer to synchronize the control data.
In current implementation, no helper function is used for initialize
the control data. This commit use an proper helper function instead of
a direct call of ioctl(2).
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit is to fix below warning.
pcm_hw.c: In function ‘snd1_pcm_hw_open_fd’:
pcm_hw.c:955:33: warning: ‘mmap_control’ may be used uninitialized in this function [-Wmaybe-uninitialized]
if (mmap_control == MAP_FAILED || mmap_control == NULL) {
^
pcm_hw.c:946:31: note: ‘mmap_control’ was declared here
struct snd_pcm_mmap_control *mmap_control;
^~~~~~~~~~~~
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Up from the new PCM protocol 2.0.14, user-space can inform the
protocol version it supports to kernel, so that the kernel may switch
its behavior depending on it. Add this ioctl call in the PCM hw
plugin at opening.
The patch contains also the addition of SNDRV_PCM_INFO_SYNC_APPLPTR
carried from the upstream kernel commit 42f945970af9 ("ALSA: pcm: Add
the explicit appl_ptr sync support"), as well as the trivial change
(an addition of comma) to sync with the kernel asound.h.
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Tested-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Currently, failures of status/control data mapping are handled dependently.
However, it's not sure that one of the operations is failed when another
is failed.
This commit adds a member into private data structure to maintain fallback
mode for control data mapping, independently of status data mapping. As a
result, we have four cases to handle status/control data:
1. both of status/control data are mapped.
Nothing changed. A structure with alias of 'snd_pcm_hw_t' already has two
members to point the mapped area and in application runtime they're used
to refer/set status/control data. No need to call ioctl(2) with
SNDRV_PCM_IOCTL_SYNC_PTR to issue/query the data.
2. both of status/control data are unmapped.
The two members point to allocated memory for fallback buffer. In
application runtime, the buffer is given as an argument for ioctl(2) with
SNDRV_PCM_IOCTL_SYNC_PTR to issue/query the data.
3. status data is mapped only.
One of the two members is used to point the mapped area. Another points to
allocated memory for fallback buffer. In application runtime, the buffer
is used as an argument to execute ioctl(2) with SNDRV_PCM_IOCTL_SYNC_PTR
for the latter data, but the former data is already synchronized.
4. control data is mapped only.
The same as the above.
In design of ALSA PCM interface, userspace applications are not expected
to map the status data as writable. On the other hand, expected to map
the control data as writable. In a focus on the differences, we could
achieve to reduce calls of the ioctl(2) in a case that one of the
status/control data is successfully mapped and another is failed (case 3
and 4). Especially, in current alsa-lib implementation, application
runtime queries state of runtime of PCM substream so often.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In current implementation, results to map status/control data are not
maintained separately. It's handled as a fatal error that mapping of status
data is successful and mapping of control data is failed. However, it's
possible to handle this case by utilizing fallback buffer.
This commit adds a member into a local structure to maintain fallback mode
just for the mapping of status data as a preparation of later commit, in
which mapping results are maintained separately for each of status/control
data.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
When allowing failure of map operation for both of status/control data
for runtime of PCM substream, applications need to use fallback buffer
for an alternative ioctl. However, in current implementation, status
mapping is dominant to the allocation.
This commit moves code for the allocation outside of the mapping
operation for status data.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In current implementation, deallocation of fallback buffer is done at
several places.
This commit unifies these deallocations in one place.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Regardless of success/failure mapping of control/status data for runtime of
PCM substream, appl_ptr/avail_min parameters are initialized. In current
implementation, they are initialized in case-dependent, different places.
It's possible to collect them to one place.
This commit unifies relevant code in a place after all of trials for the
mappings are finished.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Handling mapping operation for status/control data includes some
supplemental operations for fallback mode. It's better to have helper
functions for this purpose.
This commit adds the helper functions.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
There is a call of snd_pcm_set_appl_ptr() in snd_pcm_hw_hw_params()
only for the capture direction. This must be a leftover from the
ancient code. Although it's harmless for now, it's superfluous and
confusing. Let's kill it.
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Tested-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
When error occurs, return value from ioctl(2) is -1 and error code can
be got thread local variable, errno. It's OK just to check the return
value without any assignment.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A function, snd_pcm_hw_open_fd(), is just for internal use. This function
has an obsoleted parameter and we can remove it without any compatibility
issue.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
I noticed that some plugin codes have no proper PCM state checks and
it results in expected outcomes. For example, when snd_pcm_drain() is
called for a dmix PCM after calling snd_pcm_drop(), it stalls
unexpectedly. It's just because its drain callback doesn't expect the
SND_PCM_SETUP state.
We can fix such a bug in each place one by one, but a safer way would
be to filter out all such cases commonly in the PCM API functions
themselves. This patch adds the PCM state sanity checks to major API
functions so that they return -EBADFD when called in the unexpected
PCM states.
As well as for the thread-safety extension, it'd be a question of the
performance; again at this time, the hw PCM is considered as an
exception, and it has pcm->own_state_check flag set, which means that
the common PCM state checks are skipped.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The newly added locking code seems to have confused quite a few
people, as "thread_safe=1" may be considered as if the thread-safety
lock has to be turned on. (It meant that the plugin _is_ thread-safe,
i.e. it needs no extra locking.)
For avoiding such a misunderstanding, this commit renames the relevant
pcm fields and give more comments to explain what is for what.
The former single pcm->thread_safe flag is now split to two boolean
flags, pcm->need_lock and pcm->lock_enabled. It consumes a few more
bytes, but this would be (hopefully) better understandable.
No functional change by this commit.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Both of alloca() and automatic variables keeps storages on stack, while
the former generates more instructions than the latter. It's better to use
the latter if the size of storage is computable at pre-compile or compile
time; i.e. just for structures.
This commit obsolete usages of alloca() with automatic variables.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Both of alloca() and automatic variables keeps storages on stack, while
the former generates more instructions than the latter. It's better to use
the latter if the size of storage is computable at pre-compile or compile
time; i.e. just for structures.
This commit obsolete usages of alloca() with automatic variables.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Both of alloca() and automatic variables keeps storages on stack, while
the former generates more instructions than the latter. It's better to use
the latter if the size of storage is computable at pre-compile or compile
time; i.e. just for structures.
This commit obsolete usages of alloca() with automatic variables.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Both of alloca() and automatic variables keeps storages on stack, while
the former generates more instructions than the latter. It's better to use
the latter if the size of storage is computable at pre-compile or compile
time; i.e. just for structures.
This commit obsolete usages of alloca() with automatic variables.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit applies code format according to typical and moderate rule,
for snd_pcm_hw_change_timer().
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Traditionally, many of ALSA library functions are supposed to be
thread-unsafe, and applications are required to take care of thread
safety by themselves. However, people never be careful enough, and
almost all applications fail in this regard.
This patch is an attempt to harden the thread safety in exported PCM
functions in a simplistic way: just wrap some of exported functions
with the pthread mutex of each PCM object. Not all API functions are
wrapped by the mutex since it doesn't make sense. Instead, the
patchset covers only the functions that may be likely called
concurrently. The supposedly thread-safe API functions are marked in
the document.
For achieving the feature, two new fields are added snd_pcm_t when the
option is enabled: thread_safe and lock. The former indicates that
the plugin is thread-safe that doesn't need this workaround and the
latter is the pthread mutex. Currently only hw plugin have
thread_safe=1. So, the most of real-time sensitive apps won't be
influenced by this patchset.
Although the patch covers most of PCM ops, a few snd_pcm_fast_ops are
left without the extra mutex locking: namely, the ones that may have
blocking behavior, i.e. resume, drain, readi, writei, readn and
writen. These are supposed to handle own locking in the callbacks.
Also, if anyone wants to disable this new thread-safe API feature, it
can be still turned off via --disable-thread-safety configure option.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit fixes below compiler warnings.
pcm_hw.c: In function ‘_snd_pcm_hw_open’:
pcm_hw.c:1816:16: warning: ‘err’ may be used uninitialized in this function [-Wmaybe-uninitialized]
return err;
^
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This header isn't used in these files, so stop trying to include it.
This helps builds on systems that don't provide the header (such as
Android/Bionic).
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
use STATUS_EXT ioctl if PCM protocol is > 2.0.12
All audio timestamp configuration will be ignored with an
older protocol.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The current behavior of snd_pcm_rewindable and snd_pcm_forwardable means
that the returned value is only accurate to one period. Or maybe even
meaningless if period interrupts are off. Fetch the up-to-date position
of the hardware pointer, as that's what is wanted by callers.
Signed-off-by: Alexander E. Patrakov <patrakov@gmail.com>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Such negative values can happen when an underrun happens and xrun
detection is disabled. Another situation is if the device updated the
pointer before alsa-lib has a chance to detect the xrun.
The problem is that these negative values could propagate to the
snd_pcm_rewindable return value, where it is specified that negative
returns must be interpreted as error codes and not as negative amount of
samples.
Signed-off-by: Alexander E. Patrakov <patrakov@gmail.com>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Now all PCM plugins do support the proper timestamp type or pass it
over slaves. The internal monotonic flag is dropped and replaced with
tstamp_type in all places.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This patch implements the support for sw_params timestamp type in PCM
hw layer. As gettimestamp() is still unchanged, the resultant
timstamps may be still with CLOCK_MONOTONIC even if you pass monotonic
raw type. More fixes will follow.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
On most HDA Intel HDMI devices, the channel map list is approx. 500 ints
in size, making the 256-sized buffer used by
snd_pcm_query_chmaps_from_hw() too small and causing the query to fail
(NULL result to caller, ENOMEM).
Bump the buffer size to 2048 ints.
Signed-off-by: Anssi Hannula <anssi.hannula@iki.fi>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
We can't calculate memory storage in bytes, when we're supposed
to store ints in it!
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
These fields are not used, and their name similarity to other
fields are quite confusing when trying to debug alsa-lib.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Instead of keeping modified asound.h and asequencer.h, copy the files
in the kernel include/uapi/sound as is, and give some renames for
avoiding conflicts with alsa-lib's definitions.
Some structs, unions and typedefs in asound.h and asequencer.h are
once renamed in local.h before inclusion, then renamed back again.
A bonus by this action is that some local codes don't have to refer to
sndrv_xxx any longer. However, some codes like src/seq/seq.c need
explicit cast because of unavoidable conflicts of struct types.
Another significant change is that now snd_pcm_sw_params.period_event
field is removed. Instead, try to access the last reserved field
internally.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
