The Ubuntu package currently uses ld.so.conf.d fragments to add
/usr/lib/alsa-lib and /usr/lib{32,64}/alsa-lib to the dlopen search path.
These don't *belong* on the global search path, and it becomes much more
problematic to put them there with the advent of multiarch because each
architecture then needs its own distinct config file to add the separate
path... which is then also put in the global library namespace. Instead,
let ALSA make use of the already defined ALSA_PLUGIN_DIR to look up
plugins.
Signed-off-by: Jordi Mallach <jordi@debian.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Default to music mode filter for the HiFi use case on the Samsung ARM
Chromebook. This mode is better at 44.1k and 48k audio than the
"Voice" setting.
Signed-off-by: Dylan Reid <dgreid@chromium.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Commit 1989233449 (seq: Fix for snd_seq_parse_address()) removed the
ability to match sequencer client names by any of by their prefixes in
an attempt to avoid wrong matches when one client name is the prefix of
another.
However, the prefix match feature was documented and actually used.
Allow prefixes to match, but only if there is no exact match.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
The Scarlett 2i2 USB interface does not have any S/PDIF connections.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In particular, this adds definitions and descriptions for G.723 and
DSD types.
Signed-off-by: Daniel Mack <zonque@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
When multiple dlobj_cache items point to the same dlobj, dlclose() may
be called wrongly multiple times when these items are cleared, because
we manage the dlobj_cache list as a flat list. This results in a bad
segfault we've seen in openal-soft, for example.
For fixing this, we need the refcounting of dlobj itself. But, in
this case, we don't have to manage yet another list, since dlopen()
does a proper refcounting by itself. That is, we can just call always
dlopen() at each time a new function is assigned, and also call
dlclose() for each released dlobj_cache item at cleanup.
Bugzilla: https://bugzilla.novell.com/show_bug.cgi?id=814250
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cleaning up the dlobj cache seems crashing some cases when the library
is used from another plugin like openal-soft. A simple workaround is
to do the cleanup only when really no user is left, i.e. after all
close calls.
Bugzilla: https://bugzilla.novell.com/show_bug.cgi?id=814250
Signed-off-by: Takashi Iwai <tiwai@suse.de>
As reported dead-lock, do local lock counting and invoke abort() when
the lock counts do not match at close() time.
Reported-by: <mateen abdulmateen.shaikh@gmail.com>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Upon an interrupt, it is necessary to abort the wait loops with the EINTR
error code. Introduce snd_*_abort() functions to handle this case.
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Yet another headset without digital I/O.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The memcpy() function in snd_pcm_area_copy() should not be called
with the overlapped areas. Alex discovered - using own LD_PRELOAD checked
for memcpy() input - that the memcpy() is called with src == dst.
For some special plugin combos (rate+softvol+hw for example), the same
areas with same offsets can be asked to be copied (softvol). The collapse
check uses own areas created on heap, causing dst_area == src_area &&
dst_offset == src_offset check bypassed.
Two fixes are in this patch:
- use assert to check the memcpy() input for future triggers
- bypass the snd_pcm_area_copy() call for collapsed identical areas
Reported-by: Alexander Kruppa <akruppa@gmail.com>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
When both an SPDIF and an HDMI output are present on HD-audio, both
try to access IEC958 controls with index=0 although one of them must
be wrong. For avoiding this conflict, the recent kernel code (3.9 and
3.8 stable) moves the IEC958 controls of an SPDIF with index=16 once
when the conflict happens.
In this patch, the corresponding support is added in alsa-lib side.
The new "skip_rest" boolean flag is added to the hooked element
definition which indicates that the rest of element array will be
ignored once when this element is present and evaluated. With this
new flag, the HD-audio config takes index=16 primarily, then take
index=0 as fallback.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
snd_config_get_bool() was improved to parse also ASCII strings now,
so we don't have to open-code the boolean parser in
src/control/setup.c any longer.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This is needed by snd_pcm_format_silence* functions which
return u_int*_t. It was discovered while trying to compile ALSA
programs with eglibc 2.17.
Credits to Richard Shaw, Gary Buhrmaster, Matthieu Baerts and
Adam Conrad for this fix.
BugLink: https://bugs.launchpad.net/bugs/1109298
BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=885306
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Fixes an issue where a variable is used undeclared, which can cause seg
faults on some systems if the configuration file is not formatted
properly.
Signed-off-by: Allan Wirth <allan@allanwirth.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit and its parent make the function reentrant.
Signed-off-by: Jerome Forissier <jerome@taodyne.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This is the first step towards making this function reentrant.
Signed-off-by: Jerome Forissier <jerome@taodyne.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This is required so we can make other functions reentrant (such as
snd_device_name_hint()).
The default error handling function snd_lib_error_default() now checks
if a local handler exists, and if so, calls it. Otherwise, the previous
behavior is unchanged.
Signed-off-by: Jerome Forissier <jerome@taodyne.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
On Ubuntu 11.04, configuring with --build=x86_64-unknown-linux-gnu
--host=x86_64-linux-gnu finds a wrong cross-compiler:
checking for cross-compiler... x86_64-x86_64-pc-linux-gnu-gcc
This happens because of a dash vs underscore inconsistency in configure.in:
host=x86_64-pc-linux-gnu
host_cpu=x86_64
host_os=linux-gnu
which ${host_cpu}-${host_os}-gcc >/dev/null 2>&1 && echo ${host_cpu}-${host-os}-gcc
This bug has been introduced in the initial --with-host support from
2002, commit eb267ade29
This configure command is about "cross-compiling for i586", where the
system compiler is used, which just -m32 additional options.
The --build value comes from config.guess.
SNDRV_LITTLE_ENDIAN and SNDRV_BIG_ENDIAN checks have been removed from
sound/asound.h during UAPI header move, and this resulted in a wrong
detected endian.
Move together with the similar check for SND_*_ENDIAN at the earlier
place in local.h.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Check BUILD_MODULES conditional not to compile simple_abst.c.
Also return -ENXIO to caller statically in that case.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Fix the call of _snd_pcm_hw_params() to _snd_pcm_hw_params_internal().
The build passed wrongly because of _snd_pcm_hw_params define.
Having the same function name as the struct name is is simply
confusing...
Also, to be sure, _snd_pcm_hw_params is undef'ed in pcm_local.h, too.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
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>
snd_pcm_write_areas() tries to wait until avail >= avail_min condition
is satisfied. This doesn't work always well when a rate plugin is in
the play.
When a partial data with a smaller size than a period is written, the
rate plugin doesn't transfer the data immediately to the slave PCM,
but kept in an internal buffer and it changes only the hwptr of the
plugin. Thus, the condition "avail < avail_min" is triggered for a
wait check although the underlying slave PCM has enough room. This
results in a call of snd_pcm_wait() which returns immediately after
poll() call, and the snd_pcm_write_areas() loop continues. As a
consequence, it falls into a CPU hog.
This patch fixes that busy loop by introducing a new fast_ops to check
the availability for wait of avail_min. Then a plugin can ask the
slave PCM whether the wait is required (or possible).
A few plugins like multi plugin need a special handling. Otherwise a
generic plugin function can be used.
Reported-by: Trent Piepho <tpiepho@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The init-only plugins do not have own pcm handle, so free the references
to open function immediately after open.
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Commit cf40ea169a (pcm: support for audio timestamps) added the new
audio_tstamp field to struct sndrv_pcm_status. However, struct timespec
requires 64-bit alignment, so the 64-bit compiler would insert
32 bits of padding before this field, which broke SNDRV_PCM_IOCTL_STATUS
with error messages like this:
kernel: unknown ioctl = 0x80984120
To solve this, insert the padding explicitly so that it can be taken
into account when calculating the ABI structure size.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
