linux-kernel - Re: [PATCH v2] iio: light: tsl2583: Fix module unloading

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: <20220828173327.7949ad73@jic23-huawei>
Date:   Sun, 28 Aug 2022 17:33:27 +0100
From:   Jonathan Cameron <jic23@...nel.org>
To:     Shreeya Patel <shreeya.patel@...labora.com>
Cc:     lars@...afoo.de, krisman@...labora.com,
        dmitry.osipenko@...labora.com, kernel@...labora.com,
        linux-iio@...r.kernel.org, linux-kernel@...r.kernel.org,
        stable@...r.kernel.org
Subject: Re: [PATCH v2] iio: light: tsl2583: Fix module unloading

On Fri, 26 Aug 2022 17:53:52 +0530
Shreeya Patel <shreeya.patel@...labora.com> wrote:

> tsl2583 uses devm_iio_device_register() function and
> calling iio_device_unregister() in remove breaks the
> module unloading.
> Fix this by using iio_device_register() instead of
> devm_iio_device_register() function in probe.
Not sure why you are wrapping at 55 chars. I rewrapped this whilst applying.

Reworded it a little too as I was touching it anyway.

Applied to the fixes-togreg branch of iio.git.

> 
> Cc: stable@...r.kernel.org
> Fixes: 371894f5d1a0 ("iio: tsl2583: add runtime power management support")

I took a look at this patch and it introduces the issue I just pointed
out in replying to your v1 by dropping the 
/* Make sure the chip is on */
Which was correct even with runtime pm because it covered the case of
runtime_pm being disabled.   We probably need to bring that back as well,
perhaps as part of a cleanup patch taking this fully devm_

This driver has another issue for working if runtime PM isn't built into
the kernel which is that it checks the return of pm_runtime_put_autosuspend()
which calls 

static inline int __pm_runtime_suspend(struct device *dev, int rpmflags)
{
	return -ENOSYS;
}

I've been meaning to do an audit for drivers that have this problem for
a while, but not yet gotten to it.

An ideal IIO driver needs to work correctly whether or not CONFIG_PM is
enabled.

Jonathan


> Signed-off-by: Shreeya Patel <shreeya.patel@...labora.com>
> ---
> Changes in v2
>   - Use iio_device_register() instead of devm_iio_device_register()
>   - Add fixes and stable tags
> 
>  drivers/iio/light/tsl2583.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/iio/light/tsl2583.c b/drivers/iio/light/tsl2583.c
> index 82662dab87c0..94d75ec687c3 100644
> --- a/drivers/iio/light/tsl2583.c
> +++ b/drivers/iio/light/tsl2583.c
> @@ -858,7 +858,7 @@ static int tsl2583_probe(struct i2c_client *clientp,
>  					 TSL2583_POWER_OFF_DELAY_MS);
>  	pm_runtime_use_autosuspend(&clientp->dev);
>  
> -	ret = devm_iio_device_register(indio_dev->dev.parent, indio_dev);
> +	ret = iio_device_register(indio_dev);
>  	if (ret) {
>  		dev_err(&clientp->dev, "%s: iio registration failed\n",
>  			__func__);