Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Message-ID: <69d5cef3-57c0-9bc7-a83b-a85ef1c4cf29@foss.st.com>
Date:   Mon, 23 May 2022 15:28:48 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.8.1
Subject: Re: [PATCH] ASoC: stm32: sai: Use of_device_get_match_data() to
 simplify code
Content-Language: en-US
To:     Tang Bin <tangbin@cmss.chinamobile.com>,
        <arnaud.pouliquen@foss.st.com>, <lgirdwood@gmail.com>,
        <broonie@kernel.org>, <perex@perex.cz>, <tiwai@suse.com>,
        <mcoquelin.stm32@gmail.com>, <alexandre.torgue@foss.st.com>
CC:     <alsa-devel@alsa-project.org>,
        <linux-stm32@st-md-mailman.stormreply.com>,
        <linux-arm-kernel@lists.infradead.org>,
        <linux-kernel@vger.kernel.org>
References: <20220519124235.21100-1-tangbin@cmss.chinamobile.com>
From:   Olivier MOYSAN <olivier.moysan@foss.st.com>
In-Reply-To: <20220519124235.21100-1-tangbin@cmss.chinamobile.com>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hello Tang,

Thanks for the patch. Unfortunately this patch introduces a regression.

In the SAI driver of_device_id struct the data is a simple enum cast to 
void* pointer.
static const struct of_device_id stm32_sai_sub_ids[] = {
	  .data = (void *)STM_SAI_A_ID},

This data is an ID which can be set to 0x0.
Here we have no way to know to discriminate between an error returned by 
of_device_get_match_data() or a data id set to 0x0.

The current patch requires a change in the driver.
Either changing STM_SAI_x_ID enums, or replacing data by a struct.
For instance:
struct stm32_sai_comp_data {
	unsigned int id;
}
struct stm32_sai_comp_data stm32_sai_comp_data_a = {
	.id = STM_SAI_A_ID;
}
struct of_device_id stm32_sai_sub_ids[] = {
	.data = &stm32_sai_comp_data_a},
}

Regards
Olivier

On 5/19/22 14:42, Tang Bin wrote:
> Retrieve of match data, it's better and cleaner to use
> 'of_device_get_match_data' over 'of_match_device'.
> 
> Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com>
> ---
>   sound/soc/stm/stm32_sai_sub.c | 6 ++----
>   1 file changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/sound/soc/stm/stm32_sai_sub.c b/sound/soc/stm/stm32_sai_sub.c
> index dd636af81..d300605a2 100644
> --- a/sound/soc/stm/stm32_sai_sub.c
> +++ b/sound/soc/stm/stm32_sai_sub.c
> @@ -1500,7 +1500,6 @@ static int stm32_sai_sub_parse_of(struct platform_device *pdev,
>   static int stm32_sai_sub_probe(struct platform_device *pdev)
>   {
>   	struct stm32_sai_sub_data *sai;
> -	const struct of_device_id *of_id;
>   	const struct snd_dmaengine_pcm_config *conf = &stm32_sai_pcm_config;
>   	int ret;
>   
> @@ -1508,10 +1507,9 @@ static int stm32_sai_sub_probe(struct platform_device *pdev)
>   	if (!sai)
>   		return -ENOMEM;
>   
> -	of_id = of_match_device(stm32_sai_sub_ids, &pdev->dev);
> -	if (!of_id)
> +	sai->id = (uintptr_t)of_device_get_match_data(&pdev->dev);
> +	if (!sai->id)
>   		return -EINVAL;
> -	sai->id = (uintptr_t)of_id->data;
>   
>   	sai->pdev = pdev;
>   	mutex_init(&sai->ctrl_lock);