From: <twischer@de.adit-jv.com>
To: tiwai@suse.de
Cc: Timo Wischer <twischer@de.adit-jv.com>, alsa-devel@alsa-project.org
Subject: [PATCH - IOPLUG DRAIN 1/1] ioplug: drain: Wait with pollwhen EAGAIN in blocking mode
Date: Thu, 22 Mar 2018 14:48:56 +0100 [thread overview]
Message-ID: <1521726537-7651-2-git-send-email-twischer@de.adit-jv.com> (raw)
In-Reply-To: <1521726537-7651-1-git-send-email-twischer@de.adit-jv.com>
From: Timo Wischer <twischer@de.adit-jv.com>
With this changes an IO plugin should not block by it self in the
its own snd_pcm_drain() implementation. It should return -EAGAIN and
inform the ALSA library via the poll_descriptors that draining is
done.
In non-blocking mode this implementation expects that the user blocks on
the poll_descriptors whenever snd_pcm_drain() returns -EAGAIN. In
addition the user has to call snd_pcm_drain() again to check if draining
is really done.
This change will not harm existing IO plugins which are blocking in its
snd_pcm_drain() callback.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
diff --git a/include/pcm_ioplug.h b/include/pcm_ioplug.h
index c1310e3..aa08962 100644
--- a/include/pcm_ioplug.h
+++ b/include/pcm_ioplug.h
@@ -169,6 +169,12 @@ struct snd_pcm_ioplug_callback {
int (*prepare)(snd_pcm_ioplug_t *io);
/**
* drain; optional
+ * This function should never block. It should return -EAGAIN in case of
+ * it is not yet done. In this case it will be called again by the ALSA
+ * library (in blocking mode) or by the user (in non blocking mode).
+ * In case of -EAGAIN the poll_descriptors have to be used to inform the
+ * ALSA library that draining is possibly done and this callback has to
+ * be called again.
*/
int (*drain)(snd_pcm_ioplug_t *io);
/**
diff --git a/src/pcm/pcm_ioplug.c b/src/pcm/pcm_ioplug.c
index 8c0ed48..7682dcd 100644
--- a/src/pcm/pcm_ioplug.c
+++ b/src/pcm/pcm_ioplug.c
@@ -492,17 +492,37 @@ static int snd_pcm_ioplug_drop(snd_pcm_t *pcm)
static int snd_pcm_ioplug_drain(snd_pcm_t *pcm)
{
ioplug_priv_t *io = pcm->private_data;
- int err;
+ int err = 0;
if (io->data->state == SND_PCM_STATE_OPEN)
return -EBADFD;
io->data->state = SND_PCM_STATE_DRAINING;
- if (io->data->callback->drain)
- io->data->callback->drain(io->data);
- snd_pcm_lock(pcm);
- err = snd_pcm_ioplug_drop(pcm);
- snd_pcm_unlock(pcm);
+ if (io->data->callback->drain) {
+ err = io->data->callback->drain(io->data);
+
+ /* in blocking mode wait unti draining is done or
+ * an issue was detected.
+ */
+ if (!io->data->nonblock) {
+ while (err == -EAGAIN) {
+ snd_pcm_lock(pcm);
+ err = snd_pcm_wait_nocheck(pcm, -1);
+ snd_pcm_unlock(pcm);
+ if (err < 0)
+ break;
+ err = io->data->callback->drain(io->data);
+ }
+ }
+ }
+
+ /* only drop if draining is done or there was an issue */
+ if (err != -EAGAIN) {
+ snd_pcm_lock(pcm);
+ err = snd_pcm_ioplug_drop(pcm);
+ snd_pcm_unlock(pcm);
+ }
+
return err;
}
--
2.7.4
next prev parent reply other threads:[~2018-03-22 13:49 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-03-22 13:48 [PATCH - IOPLUG DRAIN 0/2] twischer
2018-03-22 13:48 ` twischer [this message]
2018-03-22 13:48 ` [PATCH - IOPLUG DRAIN 1/1] jack: Support snd_pcm_drain() twischer
2018-03-22 14:28 ` [PATCH - IOPLUG DRAIN 0/2] Takashi Iwai
2018-03-22 14:50 ` Wischer, Timo (ADITG/ESB)
2018-03-22 14:55 ` Takashi Iwai
2018-03-22 15:17 ` Wischer, Timo (ADITG/ESB)
2018-03-22 16:22 ` Takashi Iwai
2018-03-23 7:23 ` Wischer, Timo (ADITG/ESB)
2018-03-23 7:28 ` Takashi Iwai
2018-03-23 7:43 ` Wischer, Timo (ADITG/ESB)
2018-03-23 8:01 ` Takashi Iwai
2018-03-23 8:08 ` Takashi Iwai
2018-03-23 9:03 ` Takashi Iwai
2018-03-28 8:42 ` Wischer, Timo (ADITG/ESB)
2018-03-28 16:09 ` Takashi Iwai
2018-03-29 6:39 ` Wischer, Timo (ADITG/ESB)
2018-03-29 7:25 ` Takashi Iwai
2018-03-29 7:38 ` Wischer, Timo (ADITG/ESB)
2018-03-23 8:21 ` Wischer, Timo (ADITG/ESB)
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1521726537-7651-2-git-send-email-twischer@de.adit-jv.com \
--to=twischer@de.adit-jv.com \
--cc=alsa-devel@alsa-project.org \
--cc=tiwai@suse.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).