linux-kernel - Re: [PATCH] reset: meson: add level reset support for GX SoC family

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]

Message-ID: <1508234898.6854.6.camel@pengutronix.de>
Date:   Tue, 17 Oct 2017 12:08:18 +0200
From:   Philipp Zabel <p.zabel@...gutronix.de>
To:     Neil Armstrong <narmstrong@...libre.com>
Cc:     linux-amlogic@...ts.infradead.org,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] reset: meson: add level reset support for GX SoC family

Hi Neil,

On Mon, 2017-10-16 at 17:26 +0200, Neil Armstrong wrote:
> The Amlogic GX SoC family embeds alternate registers to drive the reset
> levels next to the pulse registers.
> 
> This patch adds support for level reset handling on the GX family only.
> 
> The Meson8 family has an alternate way to handle level reset.
> 
> Signed-off-by: Neil Armstrong <narmstrong@...libre.com>

thank you for the patch, comments below:

> ---
>  drivers/reset/reset-meson.c | 57 +++++++++++++++++++++++++++++++++++++++++----
>  1 file changed, 53 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/reset/reset-meson.c b/drivers/reset/reset-meson.c
> index a8b915e..d55e440 100644
> --- a/drivers/reset/reset-meson.c
> +++ b/drivers/reset/reset-meson.c
> @@ -62,9 +62,11 @@
>  #include <linux/reset-controller.h>
>  #include <linux/slab.h>
>  #include <linux/types.h>
> +#include <linux/of_device.h>
>  
>  #define REG_COUNT	8
>  #define BITS_PER_REG	32
> +#define LEVEL_OFFSET	0x7c
>  
>  struct meson_reset {
>  	void __iomem *reg_base;
> @@ -88,18 +90,61 @@ static int meson_reset_reset(struct reset_controller_dev *rcdev,
>  	return 0;
>  }
>  
> -static const struct reset_control_ops meson_reset_ops = {
> +static int meson_reset_level(struct reset_controller_dev *rcdev,
> +			    unsigned long id, bool assert)
> +{
> +	struct meson_reset *data =
> +		container_of(rcdev, struct meson_reset, rcdev);
> +	unsigned int bank = id / BITS_PER_REG;
> +	unsigned int offset = id % BITS_PER_REG;
> +	void __iomem *reg_addr = data->reg_base + LEVEL_OFFSET + (bank << 2);
> +	u32 reg;
> +
> +	if (bank >= REG_COUNT)
> +		return -EINVAL;

This check is not necessary. The same check is in meson_reset_reset,
which I didn't notice last time.

of_reset_simple_xlate, the default rcdev->of_xlate implementation,
already guarantees id < rcdev->nr_resets. And since nr_resets is set to
REG_COUNT * BITS_PER_REG, we know that id < REG_COUNT * BITS_PER_REG and
thus bank <= id / BITS_PER_REG < REG_COUNT.

> +	reg = readl(reg_addr);
> +	if (assert)
> +		writel(reg & ~BIT(offset), reg_addr);
> +	else
> +		writel(reg | BIT(offset), reg_addr);

These read-modify-write operations must be protected by a spinlock.

> +
> +	return 0;
> +}
> +
> +static int meson_reset_assert(struct reset_controller_dev *rcdev,
> +			      unsigned long id)
> +{
> +	return meson_reset_level(rcdev, id, true);
> +}
> +
> +static int meson_reset_deassert(struct reset_controller_dev *rcdev,
> +				unsigned long id)
> +{
> +	return meson_reset_level(rcdev, id, false);
> +}
> +
> +static const struct reset_control_ops meson_reset_meson8_ops = {
> +	.reset		= meson_reset_reset,
> +};
> +
> +static const struct reset_control_ops meson_reset_gx_ops = {
>  	.reset		= meson_reset_reset,
> +	.assert		= meson_reset_assert,
> +	.deassert	= meson_reset_deassert,
>  };
>  
>  static const struct of_device_id meson_reset_dt_ids[] = {
> -	 { .compatible = "amlogic,meson8b-reset", },
> -	 { .compatible = "amlogic,meson-gxbb-reset", },
> +	 { .compatible = "amlogic,meson8b-reset",
> +	   .data = (void *) &meson_reset_meson8_ops, },
> +	 { .compatible = "amlogic,meson-gxbb-reset",
> +	   .data = (void *) &meson_reset_gx_ops, },
>  	 { /* sentinel */ },
>  };

of_device_id.data ist const void *, so there is no need to cast here.

>  static int meson_reset_probe(struct platform_device *pdev)
>  {
> +	const struct reset_control_ops *ops;
>  	struct meson_reset *data;
>  	struct resource *res;
>  
> @@ -107,6 +152,10 @@ static int meson_reset_probe(struct platform_device *pdev)
>  	if (!data)
>  		return -ENOMEM;
>  
> +	ops = of_device_get_match_data(&pdev->dev);
> +	if (!ops)
> +		return -EINVAL;
> +
>  	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>  	data->reg_base = devm_ioremap_resource(&pdev->dev, res);
>  	if (IS_ERR(data->reg_base))
> @@ -116,7 +165,7 @@ static int meson_reset_probe(struct platform_device *pdev)
>  
>  	data->rcdev.owner = THIS_MODULE;
>  	data->rcdev.nr_resets = REG_COUNT * BITS_PER_REG;
> -	data->rcdev.ops = &meson_reset_ops;
> +	data->rcdev.ops = ops;
>  	data->rcdev.of_node = pdev->dev.of_node;
>  
>  	return devm_reset_controller_register(&pdev->dev, &data->rcdev);

regards
Philipp