From: Takashi Iwai Subject: Re: [PATCH 2/2] pcm:plugin: Fix sound capture via MMAP access Date: Wed, 05 Apr 2017 21:40:26 +0200 Message-ID: References: <1491287418-11297-1-git-send-email-sutar.mounesh@gmail.com> Mime-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka") Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mx2.suse.de (mx2.suse.de [195.135.220.15]) by alsa0.perex.cz (Postfix) with ESMTP id 0844C2671F1 for ; Wed, 5 Apr 2017 21:40:26 +0200 (CEST) In-Reply-To: <1491287418-11297-1-git-send-email-sutar.mounesh@gmail.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: alsa-devel-bounces@alsa-project.org Sender: alsa-devel-bounces@alsa-project.org To: sutar.mounesh@gmail.com Cc: Timo Wischer , alsa-devel@alsa-project.org, mounesh_sutar@mentor.com List-Id: alsa-devel@alsa-project.org On Tue, 04 Apr 2017 08:30:18 +0200, sutar.mounesh@gmail.com wrote: > > From: Timo Wischer > > Distorted sound is heard if the capture device of an ioplug plugin is used, > which is accessed through any ALSA plugin (like copy, linear, ...) via MMAP access. > E.g. aplay -> loop -> copy -> arecord -M -> aplay -> hw:0 > As mentioned in the ALSA API (see pcm/pcm.c:942): > The function #snd_pcm_avail_update() have to be called > before any mmap begin+commit operation. > > Signed-off-by: Timo Wischer > Signed-off-by: Mounesh Sutar Applied this one, too. Thanks. Takashi > > diff --git a/src/pcm/pcm_plugin.c b/src/pcm/pcm_plugin.c > index aaea2ab..ec57f6e 100644 > --- a/src/pcm/pcm_plugin.c > +++ b/src/pcm/pcm_plugin.c > @@ -490,7 +490,14 @@ static snd_pcm_sframes_t snd_pcm_plugin_avail_update(snd_pcm_t *pcm) > snd_pcm_uframes_t slave_offset; > snd_pcm_uframes_t slave_frames = ULONG_MAX; > snd_pcm_sframes_t result; > - > + /* As mentioned in the ALSA API (see pcm/pcm.c:942): > + * The function #snd_pcm_avail_update() > + * have to be called before any mmap begin+commit operation. > + * Otherwise the snd_pcm_areas_copy will not called a second time. > + * But this is needed, if the ring buffer wrap is reached and > + * there is more data available. > + */ > + slave_size = snd_pcm_avail_update(slave); > result = snd_pcm_mmap_begin(slave, &slave_areas, &slave_offset, &slave_frames); > if (result < 0) { > err = result; > -- > 2.7.4 >