On 2011-08-02 18:29, Tanu Kaskinen wrote:
On Tue, 2011-08-02 at 14:59 +0200, David Henningsson wrote:
On 2011-08-02 14:34, David Henningsson wrote:
This patch introduces some extra protocol information, so protocol
version is bumped. This functionality is primarily needed to solve
a long standing issue in alsa-plugins, which should ignore underruns
if and only if it is obsolete, i e, if more data has been written to
the pipe in the meantime (which will automatically end the underrun).
BugLink: http://bugs.launchpad.net/bugs/805940
Signed-off-by: David Henningsson<[email protected]>
For reference, here's a preliminary (it has two flaws, but works for
testing) alsa-plugins patch that makes use of the new functionality.
pcm->written is always zero - is that one of the flaws, and how can it
work for testing?
Thanks for the proofreading, attaching fixed patch.
(It was accidentally lost when I removed all the debug stuff I had in
there during development.)
The two flaws are
1) there is no way to configure handle_underrun=2
2) would make sense to make the patch fall back to old behaviour, if the
new PA functionality is not available (perhaps both compile-time and
run-time).
--
David Henningsson, Canonical Ltd.
http://launchpad.net/~diwic
>From 4e31ba395b751a6ab3254256dc8a227b3be67932 Mon Sep 17 00:00:00 2001
From: David Henningsson <[email protected]>
Date: Tue, 2 Aug 2011 14:49:04 +0200
Subject: [PATCH] alsa-plugins: Pulse: only underrun if no more data has been written
If more data has already been written after the underrun, the underrun
will automatically end and therefore we should not report it or
restart the stream.
Signed-off-by: David Henningsson <[email protected]>
---
pulse/pcm_pulse.c | 10 +++++++---
1 files changed, 7 insertions(+), 3 deletions(-)
diff --git a/pulse/pcm_pulse.c b/pulse/pcm_pulse.c
index d6c6792..9c4a7e5 100644
--- a/pulse/pcm_pulse.c
+++ b/pulse/pcm_pulse.c
@@ -39,9 +39,10 @@ typedef struct snd_pcm_pulse {
size_t last_size;
size_t ptr;
int underrun;
- int handle_underrun;
+ int handle_underrun; /* can be 0=never, 1=always or 2=only if more data has not been written */
size_t offset;
+ int64_t written;
pa_stream *stream;
@@ -459,6 +460,7 @@ static snd_pcm_sframes_t pulse_write(snd_pcm_ioplug_t * io,
}
/* Make sure the buffer pointer is in sync */
+ pcm->written += writebytes;
pcm->last_size -= writebytes;
ret = update_ptr(pcm);
if (ret < 0)
@@ -594,7 +596,8 @@ static void stream_underrun_cb(pa_stream * p, void *userdata)
if (!pcm->p)
return;
- pcm->underrun = 1;
+ if (pcm->handle_underrun == 1 || pcm->written <= pa_stream_get_underflow_index(p))
+ pcm->underrun = 1;
}
static void stream_latency_cb(pa_stream *p, void *userdata) {
@@ -691,6 +694,7 @@ static int pulse_prepare(snd_pcm_ioplug_t * io)
goto finish;
}
+ pcm->written = 0;
pa_stream_set_state_callback(pcm->stream, stream_state_cb, pcm);
pa_stream_set_latency_update_callback(pcm->stream, stream_latency_cb, pcm);
@@ -983,7 +987,7 @@ SND_PCM_PLUGIN_DEFINE_FUNC(pulse)
const char *server = NULL;
const char *device = NULL;
const char *fallback_name = NULL;
- int handle_underrun = 0;
+ int handle_underrun = 2;
int err;
snd_pcm_pulse_t *pcm;
--
1.7.4.1
_______________________________________________
pulseaudio-discuss mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/pulseaudio-discuss
