linux-kernel - Re: [PATCH 2/3] ASoC: codecs: Add support for the Infineon PEB2466 codec

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 8 Feb 2023 08:10:48 +0100
From:   Herve Codina <herve.codina@...tlin.com>
To:     Christophe JAILLET <christophe.jaillet@...adoo.fr>
Cc:     Liam Girdwood <lgirdwood@...il.com>,
        Mark Brown <broonie@...nel.org>,
        Rob Herring <robh+dt@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Linus Walleij <linus.walleij@...aro.org>,
        Bartosz Golaszewski <brgl@...ev.pl>,
        Jaroslav Kysela <perex@...ex.cz>,
        Takashi Iwai <tiwai@...e.com>, alsa-devel@...a-project.org,
        devicetree@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-gpio@...r.kernel.org,
        Christophe Leroy <christophe.leroy@...roup.eu>,
        Thomas Petazzoni <thomas.petazzoni@...tlin.com>
Subject: Re: [PATCH 2/3] ASoC: codecs: Add support for the Infineon PEB2466
 codec

On Tue, 7 Feb 2023 22:17:39 +0100
Hi Christophe,

Christophe JAILLET <christophe.jaillet@...adoo.fr> wrote:

> Le 06/02/2023 à 15:49, Herve Codina a écrit :
> > The Infineon PEB2466 codec is a programmable DSP-based four channels
> > codec with filters capabilities.
> > It also provides signals as GPIOs.
> > 
> > Signed-off-by: Herve Codina <herve.codina@...tlin.com>
> > ---
> >   sound/soc/codecs/Kconfig   |   12 +
> >   sound/soc/codecs/Makefile  |    2 +
> >   sound/soc/codecs/peb2466.c | 2071 ++++++++++++++++++++++++++++++++++++
> >   3 files changed, 2085 insertions(+)
> >   create mode 100644 sound/soc/codecs/peb2466.c
> >   
> 
> [...]
> 
> > +static int peb2466_spi_probe(struct spi_device *spi)
> > +{
> > +	struct peb2466 *peb2466;
> > +	unsigned long mclk_rate;
> > +	int ret;
> > +	u8 xr5;
> > +
> > +	spi->bits_per_word = 8;
> > +	ret = spi_setup(spi);
> > +	if (ret < 0)
> > +		return ret;
> > +
> > +	peb2466 = devm_kzalloc(&spi->dev, sizeof(*peb2466), GFP_KERNEL);
> > +	if (!peb2466)
> > +		return -ENOMEM;
> > +
> > +	peb2466->spi = spi;
> > +
> > +	peb2466->regmap = devm_regmap_init(&peb2466->spi->dev, NULL, peb2466,
> > +					   &peb2466_regmap_config);
> > +	if (IS_ERR(peb2466->regmap))
> > +		return PTR_ERR(peb2466->regmap);
> > +
> > +	peb2466->reset_gpio = devm_gpiod_get_optional(&peb2466->spi->dev,
> > +						      "reset", GPIOD_OUT_LOW);
> > +	if (IS_ERR(peb2466->reset_gpio))
> > +		return PTR_ERR(peb2466->reset_gpio);
> > +
> > +	peb2466->mclk = devm_clk_get(&peb2466->spi->dev, "mclk");  
> 
> Hi,
> 
> Up to you to decide if it is a good idea or not, but using 
> devm_clk_get_enabled() would save the 'mclk' field in peb2466 ...
> 
> > +	if (IS_ERR(peb2466->mclk))
> > +		return PTR_ERR(peb2466->mclk);
> > +	ret = clk_prepare_enable(peb2466->mclk);
> > +	if (ret)
> > +		return ret;  
> 
> ... these 3 lines ...
> 
> > +
> > +	if (peb2466->reset_gpio) {
> > +		gpiod_set_value_cansleep(peb2466->reset_gpio, 1);
> > +		udelay(4);
> > +		gpiod_set_value_cansleep(peb2466->reset_gpio, 0);
> > +		udelay(4);
> > +	}
> > +
> > +	spi_set_drvdata(spi, peb2466);  
> 
> ... this spi_set_drvdata() call ...
> 
> > +
> > +	mclk_rate = clk_get_rate(peb2466->mclk);
> > +	switch (mclk_rate) {
> > +	case 1536000:
> > +		xr5 = PEB2466_XR5_MCLK_1536;
> > +		break;
> > +	case 2048000:
> > +		xr5 = PEB2466_XR5_MCLK_2048;
> > +		break;
> > +	case 4096000:
> > +		xr5 = PEB2466_XR5_MCLK_4096;
> > +		break;
> > +	case 8192000:
> > +		xr5 = PEB2466_XR5_MCLK_8192;
> > +		break;
> > +	default:
> > +		dev_err(&peb2466->spi->dev, "Unsupported clock rate %lu\n",
> > +			mclk_rate);
> > +		ret = -EINVAL;
> > +		goto failed;
> > +	}
> > +	ret = regmap_write(peb2466->regmap, PEB2466_XR5, xr5);
> > +	if (ret) {
> > +		dev_err(&peb2466->spi->dev, "Setting MCLK failed (%d)\n", ret);
> > +		goto failed;
> > +	}
> > +
> > +	ret = devm_snd_soc_register_component(&spi->dev, &peb2466_component_driver,
> > +					      &peb2466_dai_driver, 1);
> > +	if (ret)
> > +		goto failed;
> > +
> > +	if (IS_ENABLED(CONFIG_GPIOLIB)) {
> > +		ret = peb2466_gpio_init(peb2466);
> > +		if (ret)
> > +			goto failed;
> > +	}
> > +
> > +	return 0;
> > +
> > +failed:
> > +	clk_disable_unprepare(peb2466->mclk);
> > +	return ret;  
> 
> ... this error handling path ...
> 
> > +}
> > +
> > +static void peb2466_spi_remove(struct spi_device *spi)
> > +{
> > +	struct peb2466 *peb2466 = spi_get_drvdata(spi);
> > +
> > +	clk_disable_unprepare(peb2466->mclk);
> > +}  
> 
> ... and the remove function.
> 
> CJ
> 

Thanks for pointing this.
I will use devm_clk_get_enabled() in the next series iteration as suggested.

Best regards,
Hervé

-- 
Hervé Codina, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com