Message-ID: <8f827e42-e5d8-1190-b5ae-2f99c71fbeee@linux.intel.com>
Date:   Mon, 30 Jan 2023 17:50:23 -0600
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Firefox/102.0 Thunderbird/102.4.2
Subject: Re: [RFC PATCH v2 11/22] ASoC: qcom: Add USB backend ASoC driver for
 Q6
Content-Language: en-US
To:     Wesley Cheng <quic_wcheng@quicinc.com>,
        srinivas.kandagatla@linaro.org, mathias.nyman@intel.com,
        perex@perex.cz, lgirdwood@gmail.com, andersson@kernel.org,
        krzysztof.kozlowski+dt@linaro.org, gregkh@linuxfoundation.org,
        Thinh.Nguyen@synopsys.com, broonie@kernel.org,
        bgoswami@quicinc.com, tiwai@suse.com, robh+dt@kernel.org,
        agross@kernel.org
Cc:     devicetree@vger.kernel.org, alsa-devel@alsa-project.org,
        linux-arm-msm@vger.kernel.org, linux-usb@vger.kernel.org,
        linux-kernel@vger.kernel.org, quic_jackp@quicinc.com,
        quic_plai@quicinc.com
References: <20230126031424.14582-1-quic_wcheng@quicinc.com>
 <20230126031424.14582-12-quic_wcheng@quicinc.com>
 <7f471960-0909-4680-e192-261d1fdfe6d7@linux.intel.com>
 <4a255518-6fbb-122a-ba97-1e995785c35f@quicinc.com>
From:   Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
In-Reply-To: <4a255518-6fbb-122a-ba97-1e995785c35f@quicinc.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk


>>> +static struct snd_soc_dai_driver q6usb_be_dais[] = {
>>> +    {
>>> +        .playback = {
>>> +            .stream_name = "USB BE RX",
>>> +            .rates = SNDRV_PCM_RATE_8000 | SNDRV_PCM_RATE_11025 |
>>> +                SNDRV_PCM_RATE_16000 | SNDRV_PCM_RATE_22050 |
>>> +                SNDRV_PCM_RATE_32000 | SNDRV_PCM_RATE_44100 |
>>> +                SNDRV_PCM_RATE_48000 | SNDRV_PCM_RATE_96000 |
>>> +                SNDRV_PCM_RATE_192000,
>>> +            .formats = SNDRV_PCM_FMTBIT_S16_LE |
>>> SNDRV_PCM_FMTBIT_S16_BE |
>>> +                SNDRV_PCM_FMTBIT_U16_LE | SNDRV_PCM_FMTBIT_U16_BE |
>>> +                SNDRV_PCM_FMTBIT_S24_LE | SNDRV_PCM_FMTBIT_S24_BE |
>>> +                SNDRV_PCM_FMTBIT_U24_LE | SNDRV_PCM_FMTBIT_U24_BE,
>>> +            .channels_min = 1,
>>> +            .channels_max = 2,
>>> +            .rate_max =     192000,
>>> +            .rate_min =    8000,
>>> +        },
>>> +        .id = USB_RX,
>>> +        .name = "USB_RX_BE",
>>> +        .ops = &q6usb_ops,
>>> +    },
>>> +};
>>
>> ... here I see a single DAI, so presumably ONE endpoint can be supported?
>>
> 
> One USB audio device can be supported.
> one AFE port = one USB audio device
> 
>> I didn't see in the rest of the code how a card with multiple endpoint
>> would be rejected, nor how the capabilities are checked?
>>
> 
> Need to take a look at this query a bit more.  Let me try to pass in a
> format that can't be supported by the audio DSP, and see if the formats
> specified in this structure will not allow userspace to start the session.
> 
> When you say a "card with multiple endpoints" are you referring to a USB
> device that exposes multiple data (ISOC let's say) eps for its data
> interface?  I haven't run into a device like that.

A headset will typically only have two isoc endpoints for playback and
capture respectively, but while that's a very large market USB audio is
far from restricted to this configuration. It's not uncommon for the pro
or prosumer market to see devices with multiple input/output
capabilities and run-time mixing on the host.