Re: [PATCH] iio: dummy_evgen: check iio_evgen in iio_dummy_evgen_free()

[Jonathan Cameron <jic23@kernel.org>]

On Thu, 9 May 2019 10:04:47 +0800
Kefeng Wang <wangkefeng.wang@huawei.com> wrote:

> if iio_dummy_evgen_create() fails, iio_evgen should be NULL, when call
> iio_evgen_release() to cleanup, it throws some warning and could cause
> double free.
> 
> Reported-by: Hulk Robot <hulkci@huawei.com>
> Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Hi Kefeng,

I'm not seeing a path to be able to trigger this.
iio_dummy_evgen_create is called only in the module_init.
If it fails, then the init fails before the device
initialization call is made.

How would we then be running the device release call
in order to end up freeing this again?

So I think this is a false positive but perhaps there is
a path that I am missing.

Jonathan

> ---
>  drivers/iio/dummy/iio_dummy_evgen.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/iio/dummy/iio_dummy_evgen.c b/drivers/iio/dummy/iio_dummy_evgen.c
> index c6033e341963..2327b5f52086 100644
> --- a/drivers/iio/dummy/iio_dummy_evgen.c
> +++ b/drivers/iio/dummy/iio_dummy_evgen.c
> @@ -58,6 +58,7 @@ static int iio_dummy_evgen_create(void)
>  	ret = irq_sim_init(&iio_evgen->irq_sim, IIO_EVENTGEN_NO);
>  	if (ret < 0) {
>  		kfree(iio_evgen);
> +		iio_evgen = NULL;
>  		return ret;
>  	}
>  
> @@ -118,6 +119,9 @@ EXPORT_SYMBOL_GPL(iio_dummy_evgen_get_regs);
>  
>  static void iio_dummy_evgen_free(void)
>  {
> +	if (!iio_evgen)
> +		return;
> +
>  	irq_sim_fini(&iio_evgen->irq_sim);
>  	kfree(iio_evgen);
>  }