From: Mark Brown <broonie@kernel.org> To: Barry Song <21cnbao@gmail.com> Cc: alsa-devel@alsa-project.org, lgirdwood@gmail.com, Workgroup.Linux@csr.com, Rongjun Ying <Rongjun.Ying@csr.com>, Barry Song <Baohua.Song@csr.com>, linux-arm-kernel@lists.infradead.org Subject: Re: [PATCH 2/6] ASoC: sirf: add I2S CPU DAI driver Date: Fri, 19 Jul 2013 17:52:46 +0100 [thread overview] Message-ID: <20130719165246.GL9858@sirena.org.uk> (raw) In-Reply-To: <1374232042-26088-3-git-send-email-Baohua.Song@csr.com> [-- Attachment #1.1: Type: text/plain, Size: 2919 bytes --] On Fri, Jul 19, 2013 at 07:07:18PM +0800, Barry Song wrote: Pretty good overall - there's a few things below but should all be fairly small. > +static int sirf_i2s_startup(struct snd_pcm_substream *substream, > + struct snd_soc_dai *dai) > +{ > + > + struct sirf_i2s *si2s = snd_soc_dai_get_drvdata(dai); > + > + if (si2s->master_mode) > + pwm_enable(si2s->mclk_pwm); > + clk_prepare_enable(si2s->clk); > + > + device_reset(dai->dev); > + snd_soc_dai_set_dma_data(dai, substream, > + &sirf_i2s_dai_dma_data[substream->stream]); > + return 0; > +} device_reset() sounds like it's not going to work for simultaneous playback and capture. > +static int sirf_i2s_trigger(struct snd_pcm_substream *substream, > + int cmd, struct snd_soc_dai *dai) > +{ > + struct sirf_i2s *si2s = snd_soc_dai_get_drvdata(dai); > + int playback = (substream->stream == SNDRV_PCM_STREAM_PLAYBACK); > + > + switch (cmd) { > + case SNDRV_PCM_TRIGGER_START: > + case SNDRV_PCM_TRIGGER_RESUME: > + case SNDRV_PCM_TRIGGER_PAUSE_RELEASE: > + spin_lock(&si2s->lock); > + > + if (playback) { > + /* First start the FIFO, then enable the tx/rx */ > + writel(AUDIO_FIFO_RESET, > + si2s->base + AUDIO_CTRL_EXT_TXFIFO1_OP); > + mdelay(1); > + writel(AUDIO_FIFO_START, > + si2s->base + AUDIO_CTRL_EXT_TXFIFO1_OP); > + mdelay(1); Do we really need these 1ms delays? They're very long for the context... > +static int sirf_i2s_prepare(struct snd_pcm_substream *substream, > + struct snd_soc_dai *dai) > +{ > + struct snd_pcm_runtime *runtime = substream->runtime; > + struct sirf_i2s *si2s = snd_soc_dai_get_drvdata(dai); > + u32 ctrl = readl(si2s->base + AUDIO_CTRL_I2S_CTRL); > + > + if (runtime->channels == 2) > + ctrl &= ~I2S_SIX_CHANNELS; > + else > + ctrl |= I2S_SIX_CHANNELS; A switch statement for the number of channels would make me happier here. > + switch (fmt & SND_SOC_DAIFMT_MASTER_MASK) { > + case SND_SOC_DAIFMT_CBM_CFM: > + ctrl = readl(si2s->base + AUDIO_CTRL_I2S_CTRL); > + ctrl |= I2S_SLAVE_MODE; > + writel(ctrl, si2s->base + AUDIO_CTRL_I2S_CTRL); > + si2s->master_mode = 0; > + break; Do we need locking on all these register updates? > + case SND_SOC_DAIFMT_CBS_CFS: > + si2s->master_mode = 1; > + return -EINVAL; Should this be undoing the work done for _CBM_CFM? > +#ifdef CONFIG_PM > +static int sirf_i2s_suspend(struct platform_device *pdev, > + pm_message_t state) > +{ > + struct sirf_i2s *si2s = platform_get_drvdata(pdev); > + > + si2s->i2s_ctrl = readl(si2s->base+AUDIO_CTRL_I2S_CTRL); > + > + clk_disable_unprepare(si2s->clk); > + > + if (si2s->master_mode) > + pwm_disable(si2s->mclk_pwm); > + return 0; > +} Should these be runtime PM calls as well? Seems like the bus could be unclocked whenever audio is not playing. If you need the clock to write conifguraiton perhaps regmap-mmio could help? > + spin_lock_init(&si2s->lock); What is this lock actually protecting? [-- Attachment #1.2: Digital signature --] [-- Type: application/pgp-signature, Size: 836 bytes --] [-- Attachment #2: Type: text/plain, Size: 0 bytes --]
WARNING: multiple messages have this Message-ID (diff)
From: broonie@kernel.org (Mark Brown) To: linux-arm-kernel@lists.infradead.org Subject: [PATCH 2/6] ASoC: sirf: add I2S CPU DAI driver Date: Fri, 19 Jul 2013 17:52:46 +0100 [thread overview] Message-ID: <20130719165246.GL9858@sirena.org.uk> (raw) In-Reply-To: <1374232042-26088-3-git-send-email-Baohua.Song@csr.com> On Fri, Jul 19, 2013 at 07:07:18PM +0800, Barry Song wrote: Pretty good overall - there's a few things below but should all be fairly small. > +static int sirf_i2s_startup(struct snd_pcm_substream *substream, > + struct snd_soc_dai *dai) > +{ > + > + struct sirf_i2s *si2s = snd_soc_dai_get_drvdata(dai); > + > + if (si2s->master_mode) > + pwm_enable(si2s->mclk_pwm); > + clk_prepare_enable(si2s->clk); > + > + device_reset(dai->dev); > + snd_soc_dai_set_dma_data(dai, substream, > + &sirf_i2s_dai_dma_data[substream->stream]); > + return 0; > +} device_reset() sounds like it's not going to work for simultaneous playback and capture. > +static int sirf_i2s_trigger(struct snd_pcm_substream *substream, > + int cmd, struct snd_soc_dai *dai) > +{ > + struct sirf_i2s *si2s = snd_soc_dai_get_drvdata(dai); > + int playback = (substream->stream == SNDRV_PCM_STREAM_PLAYBACK); > + > + switch (cmd) { > + case SNDRV_PCM_TRIGGER_START: > + case SNDRV_PCM_TRIGGER_RESUME: > + case SNDRV_PCM_TRIGGER_PAUSE_RELEASE: > + spin_lock(&si2s->lock); > + > + if (playback) { > + /* First start the FIFO, then enable the tx/rx */ > + writel(AUDIO_FIFO_RESET, > + si2s->base + AUDIO_CTRL_EXT_TXFIFO1_OP); > + mdelay(1); > + writel(AUDIO_FIFO_START, > + si2s->base + AUDIO_CTRL_EXT_TXFIFO1_OP); > + mdelay(1); Do we really need these 1ms delays? They're very long for the context... > +static int sirf_i2s_prepare(struct snd_pcm_substream *substream, > + struct snd_soc_dai *dai) > +{ > + struct snd_pcm_runtime *runtime = substream->runtime; > + struct sirf_i2s *si2s = snd_soc_dai_get_drvdata(dai); > + u32 ctrl = readl(si2s->base + AUDIO_CTRL_I2S_CTRL); > + > + if (runtime->channels == 2) > + ctrl &= ~I2S_SIX_CHANNELS; > + else > + ctrl |= I2S_SIX_CHANNELS; A switch statement for the number of channels would make me happier here. > + switch (fmt & SND_SOC_DAIFMT_MASTER_MASK) { > + case SND_SOC_DAIFMT_CBM_CFM: > + ctrl = readl(si2s->base + AUDIO_CTRL_I2S_CTRL); > + ctrl |= I2S_SLAVE_MODE; > + writel(ctrl, si2s->base + AUDIO_CTRL_I2S_CTRL); > + si2s->master_mode = 0; > + break; Do we need locking on all these register updates? > + case SND_SOC_DAIFMT_CBS_CFS: > + si2s->master_mode = 1; > + return -EINVAL; Should this be undoing the work done for _CBM_CFM? > +#ifdef CONFIG_PM > +static int sirf_i2s_suspend(struct platform_device *pdev, > + pm_message_t state) > +{ > + struct sirf_i2s *si2s = platform_get_drvdata(pdev); > + > + si2s->i2s_ctrl = readl(si2s->base+AUDIO_CTRL_I2S_CTRL); > + > + clk_disable_unprepare(si2s->clk); > + > + if (si2s->master_mode) > + pwm_disable(si2s->mclk_pwm); > + return 0; > +} Should these be runtime PM calls as well? Seems like the bus could be unclocked whenever audio is not playing. If you need the clock to write conifguraiton perhaps regmap-mmio could help? > + spin_lock_init(&si2s->lock); What is this lock actually protecting? -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 836 bytes Desc: Digital signature URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130719/727ea2f8/attachment.sig>
next prev parent reply other threads:[~2013-07-19 16:52 UTC|newest] Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top 2013-07-19 11:07 [PATCH 0/6] ASoC: add CSR SiRFSoC sound drivers Barry Song 2013-07-19 11:07 ` Barry Song 2013-07-19 11:07 ` [PATCH 1/6] ASoC: sirf: add sirf platform driver which provides DMA Barry Song 2013-07-19 11:07 ` Barry Song 2013-07-19 15:08 ` Lars-Peter Clausen 2013-07-19 15:08 ` [alsa-devel] " Lars-Peter Clausen 2013-07-19 16:12 ` Mark Brown 2013-07-19 16:12 ` Mark Brown 2013-07-22 0:06 ` Barry Song 2013-07-22 0:06 ` Barry Song 2013-07-19 11:07 ` [PATCH 2/6] ASoC: sirf: add I2S CPU DAI driver Barry Song 2013-07-19 11:07 ` Barry Song 2013-07-19 16:52 ` Mark Brown [this message] 2013-07-19 16:52 ` Mark Brown 2013-07-22 9:07 ` Barry Song 2013-07-22 9:07 ` Barry Song 2013-07-19 11:07 ` [PATCH 3/6] ASoC: usp-pcm: add CPU DAI driver for PCM simulated from USP Barry Song 2013-07-19 11:07 ` Barry Song 2013-07-19 18:13 ` Mark Brown 2013-07-19 18:13 ` Mark Brown 2013-07-25 9:32 ` Barry Song 2013-07-25 9:32 ` Barry Song 2013-07-25 10:24 ` Mark Brown 2013-07-25 10:24 ` Mark Brown 2013-07-19 11:07 ` [PATCH 4/6] ASoC: sirf-soc-inner: add drivers for both CPU and Codec DAIs Barry Song 2013-07-19 11:07 ` Barry Song 2013-07-19 18:35 ` Mark Brown 2013-07-19 18:35 ` Mark Brown 2013-07-25 9:11 ` Barry Song 2013-07-25 9:11 ` Barry Song 2013-07-19 11:07 ` [PATCH 5/6] ASoC: sirf-inner: add mach driver for SiRFSoC internal codec Barry Song 2013-07-19 11:07 ` Barry Song 2013-07-19 18:51 ` Mark Brown 2013-07-19 18:51 ` Mark Brown 2013-07-25 8:57 ` Barry Song 2013-07-25 8:57 ` Barry Song 2013-07-19 11:07 ` [PATCH 6/6] arm: prima2: defconfig: enable sound components Barry Song 2013-07-19 11:07 ` Barry Song
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=20130719165246.GL9858@sirena.org.uk \ --to=broonie@kernel.org \ --cc=21cnbao@gmail.com \ --cc=Baohua.Song@csr.com \ --cc=Rongjun.Ying@csr.com \ --cc=Workgroup.Linux@csr.com \ --cc=alsa-devel@alsa-project.org \ --cc=lgirdwood@gmail.com \ --cc=linux-arm-kernel@lists.infradead.org \ /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: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.