From: Takashi Iwai Subject: Re: [PATCH v2 2/2] ALSA: hda: fix to wait for RIRB & CORB DMA to set Date: Sun, 08 May 2016 11:45:50 +0200 Message-ID: References: <1462427683-3646-1-git-send-email-vinod.koul@intel.com> <1462427683-3646-2-git-send-email-vinod.koul@intel.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 AC558260522 for ; Sun, 8 May 2016 11:45:50 +0200 (CEST) In-Reply-To: <1462427683-3646-2-git-send-email-vinod.koul@intel.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: Vinod Koul Cc: liam.r.girdwood@linux.intel.com, patches.audio@intel.com, alsa-devel@alsa-project.org, broonie@kernel.org, Jeeja KP List-Id: alsa-devel@alsa-project.org On Thu, 05 May 2016 07:54:43 +0200, Vinod Koul wrote: > > From: Jeeja KP > > If the DMAs are not being quiesced properly, it may lead to > stability issues, so the recommendation is to wait till DMAs are > stopped. > > After setting the stop bit of RIRB/CORB DMA, we should wait for > stop bit to be set. > > Signed-off-by: Jeeja KP > Signed-off-by: Vinod Koul > --- > changes in v2: > - use common wait routine as suggested by Takashi > > sound/hda/hdac_controller.c | 17 +++++++++++++++++ > 1 file changed, 17 insertions(+) > > diff --git a/sound/hda/hdac_controller.c b/sound/hda/hdac_controller.c > index 8c486235c905..9fee464e5d49 100644 > --- a/sound/hda/hdac_controller.c > +++ b/sound/hda/hdac_controller.c > @@ -80,6 +80,22 @@ void snd_hdac_bus_init_cmd_io(struct hdac_bus *bus) > } > EXPORT_SYMBOL_GPL(snd_hdac_bus_init_cmd_io); > > +/* wait for cmd dmas till they are stopped */ > +static void hdac_wait_for_cmd_dmas(struct hdac_bus *bus) > +{ > + unsigned long timeout; > + > + timeout = jiffies + msecs_to_jiffies(100); > + while ((snd_hdac_chip_readb(bus, RIRBCTL) & AZX_RBCTL_DMA_EN) > + && time_before(jiffies, timeout)) > + udelay(10); > + > + timeout = jiffies + msecs_to_jiffies(100); > + while ((snd_hdac_chip_readb(bus, CORBCTL) & AZX_CORBCTL_RUN) > + && time_before(jiffies, timeout)) > + udelay(10); This will end up with doubled timeout. But it doesn't matter so much in practice. So I applied the patch as is. thanks, Takashi > +} > + > /** > * snd_hdac_bus_stop_cmd_io - clean up CORB/RIRB buffers > * @bus: HD-audio core bus > @@ -90,6 +106,7 @@ void snd_hdac_bus_stop_cmd_io(struct hdac_bus *bus) > /* disable ringbuffer DMAs */ > snd_hdac_chip_writeb(bus, RIRBCTL, 0); > snd_hdac_chip_writeb(bus, CORBCTL, 0); > + hdac_wait_for_cmd_dmas(bus); > /* disable unsolicited responses */ > snd_hdac_chip_updatel(bus, GCTL, AZX_GCTL_UNSOL, 0); > spin_unlock_irq(&bus->reg_lock); > -- > 1.9.1 >