[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <56B1BEDB.6060709@linaro.org>
Date: Wed, 3 Feb 2016 08:48:27 +0000
From: Srinivas Kandagatla <srinivas.kandagatla@...aro.org>
To: Patrick Lai <plai@...eaurora.org>, alsa-devel@...a-project.org,
Banajit Goswami <bgoswami@...eaurora.org>,
Liam Girdwood <lgirdwood@...il.com>,
linux-arm-msm@...r.kernel.org, linux-kernel@...r.kernel.org,
Takashi Iwai <tiwai@...e.com>, Mark Brown <broonie@...nel.org>
Subject: Re: [alsa-devel] [PATCH RFC 01/15] ASoC: qcom: use
snd_dma_alloc/free* apis
On 03/02/16 00:35, Kenneth Westfield wrote:
> On Mon, Feb 01, 2016 at 09:27:59AM -0800, Srinivas Kandagatla wrote:
>> diff --git a/sound/soc/qcom/lpass-platform.c
>> b/sound/soc/qcom/lpass-platform.c
>> index 4aeb8e1..a6dce1b 100644
>> --- a/sound/soc/qcom/lpass-platform.c
>> +++ b/sound/soc/qcom/lpass-platform.c
>> @@ -439,39 +439,6 @@ static irqreturn_t lpass_platform_lpaif_irq(int irq,
>> void *data)
>> return IRQ_HANDLED;
>> }
>>
>> -static int lpass_platform_alloc_buffer(struct snd_pcm_substream
>> *substream,
>> - struct snd_soc_pcm_runtime *rt)
>> -{
>> - struct snd_dma_buffer *buf = &substream->dma_buffer;
>> - size_t size = lpass_platform_pcm_hardware.buffer_bytes_max;
>> -
>> - buf->dev.type = SNDRV_DMA_TYPE_DEV;
>> - buf->dev.dev = rt->platform->dev;
>> - buf->private_data = NULL;
>> - buf->area = dma_alloc_coherent(rt->platform->dev, size,
>> &buf->addr,
>> - GFP_KERNEL);
>> - if (!buf->area) {
>> - dev_err(rt->platform->dev, "%s: Could not allocate DMA
>> buffer\n",
>> - __func__);
>> - return -ENOMEM;
>> - }
>
> ...
>
>> @@ -499,7 +467,8 @@ static int lpass_platform_pcm_new(struct
>> snd_soc_pcm_runtime *soc_runtime)
>>
>> snd_soc_pcm_set_drvdata(soc_runtime, data);
>>
>> - ret = lpass_platform_alloc_buffer(substream, soc_runtime);
>> + ret = snd_dma_alloc_pages(SNDRV_DMA_TYPE_DEV, pcm->card->dev,
>> + size, &substream->dma_buffer);
>> if (ret)
>> return ret;
>>
>
> Is there a particular reason for using the soundcard device (pcm-card->dev)
> rather than the platform device (rt->platform->dev) for memory
> allocation? Especially considering you posted a fix for this several
> weeks ago (ASoC: qcom: use correct device pointer in dma allocation).
Thanks for spotting this, I will fix it and resend it with platform->dev.
>
Powered by blists - more mailing lists