Commit 503ed9c3 authored by Max Kellermann's avatar Max Kellermann

Merge tag 'release-0.18.9'

parents cb63189f 2784d656
......@@ -39,13 +39,18 @@ ver 0.19 (not yet released)
* install systemd unit for socket activation
* Android port
ver 0.18.9 (not yet released)
ver 0.18.9 (2014/03/02)
* protocol
- "findadd" requires the "add" permission
* output
- alsa: improved workaround for noise after manual song change
* decoder
- vorbis: fix linker failure when libvorbis/libogg are static
* encoder
- vorbis: fix another linker failure
* output
- pipe: fix hanging child process due to blocked signals
* fix build failure due to missing signal.h include
ver 0.18.8 (2014/02/07)
* decoder
......
......@@ -99,7 +99,7 @@ static const struct command commands[] = {
{ "enableoutput", PERMISSION_ADMIN, 1, 1, handle_enableoutput },
#ifdef ENABLE_DATABASE
{ "find", PERMISSION_READ, 2, -1, handle_find },
{ "findadd", PERMISSION_READ, 2, -1, handle_findadd},
{ "findadd", PERMISSION_ADD, 2, -1, handle_findadd},
#endif
{ "idle", PERMISSION_READ, 0, -1, handle_idle },
{ "kill", PERMISSION_ADMIN, -1, -1, handle_kill },
......
......@@ -105,12 +105,16 @@ struct AlsaOutput {
snd_pcm_uframes_t period_position;
/**
* Set to non-zero when the Raspberry Pi workaround has been
* activated in alsa_recover(); decremented by each write.
* This will avoid activating it again, leading to an endless
* loop. This problem was observed with a "RME Digi9636/52".
* Do we need to call snd_pcm_prepare() before the next write?
* It means that we put the device to SND_PCM_STATE_SETUP by
* calling snd_pcm_drop().
*
* Without this flag, we could easily recover after a failed
* optimistic write (returning -EBADFD), but the Raspberry Pi
* audio driver is infamous for generating ugly artefacts from
* this.
*/
unsigned pi_workaround;
bool must_prepare;
/**
* This buffer gets allocated after opening the ALSA device.
......@@ -671,8 +675,6 @@ alsa_open(AudioOutput *ao, AudioFormat &audio_format, Error &error)
{
AlsaOutput *ad = (AlsaOutput *)ao;
ad->pi_workaround = 0;
int err = snd_pcm_open(&ad->pcm, alsa_device(ad),
SND_PCM_STREAM_PLAYBACK, ad->mode);
if (err < 0) {
......@@ -694,6 +696,8 @@ alsa_open(AudioOutput *ao, AudioFormat &audio_format, Error &error)
ad->in_frame_size = audio_format.GetFrameSize();
ad->out_frame_size = ad->pcm_export->GetFrameSize(audio_format);
ad->must_prepare = false;
return true;
}
......@@ -731,29 +735,6 @@ alsa_recover(AlsaOutput *ad, int err)
case SND_PCM_STATE_XRUN:
ad->period_position = 0;
err = snd_pcm_prepare(ad->pcm);
if (err == 0 && ad->pi_workaround == 0) {
/* this works around a driver bug observed on
the Raspberry Pi: after snd_pcm_drop(), the
whole ring buffer must be invalidated, but
the snd_pcm_prepare() call above makes the
driver play random data that just happens
to be still in the buffer; by adding and
cancelling some silence, this bug does not
occur */
alsa_write_silence(ad, ad->period_frames);
/* cancel the silence data right away to avoid
increasing latency; even though this
function call invalidates the portion of
silence, the driver seems to avoid the
bug */
snd_pcm_reset(ad->pcm);
/* disable the workaround for some time */
ad->pi_workaround = 8;
}
break;
case SND_PCM_STATE_DISCONNECTED:
break;
......@@ -796,6 +777,7 @@ alsa_cancel(AudioOutput *ao)
AlsaOutput *ad = (AlsaOutput *)ao;
ad->period_position = 0;
ad->must_prepare = true;
snd_pcm_drop(ad->pcm);
}
......@@ -817,6 +799,16 @@ alsa_play(AudioOutput *ao, const void *chunk, size_t size,
assert(size % ad->in_frame_size == 0);
if (ad->must_prepare) {
ad->must_prepare = false;
int err = snd_pcm_prepare(ad->pcm);
if (err < 0) {
error.Set(alsa_output_domain, err, snd_strerror(-err));
return 0;
}
}
chunk = ad->pcm_export->Export(chunk, size, size);
assert(size % ad->out_frame_size == 0);
......@@ -829,9 +821,6 @@ alsa_play(AudioOutput *ao, const void *chunk, size_t size,
ad->period_position = (ad->period_position + ret)
% ad->period_frames;
if (ad->pi_workaround > 0)
--ad->pi_workaround;
size_t bytes_written = ret * ad->out_frame_size;
return ad->pcm_export->CalcSourceSize(bytes_written);
}
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment