Re: [PATCH 06/15] cxl/core: Update devm_cxl_add_endpoint() for cxl_cachedevs

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] cxl: implicit cast from void* to struct bypassing container_of in schedule_detach
--

commit e4fe9279629e13947567c92e9190b5eeaa65a71a
Author: Ben Cheatham <Benjamin.Cheatham@amd.com>

cxl/core: Update devm_cxl_add_endpoint() for cxl_cachedevs

Update devm_cxl_add_endpoint() to allow for cxl_cachedevs as well as
cxl_memdevs. Add cxl_cachedevs to the port heirarchy.

> diff --git a/drivers/cxl/port.c b/drivers/cxl/port.c
> index 99cf77b6b6990..44d978d4205a2 100644
> --- a/drivers/cxl/port.c
> +++ b/drivers/cxl/port.c
> @@ -26,9 +27,13 @@
>   * PCIe topology.
>   */
>  
> -static void schedule_detach(void *cxlmd)
> +static void schedule_detach(void *ep_dev)
>  {
> -	schedule_cxl_memdev_detach(cxlmd);
> +	if (is_cxl_memdev(ep_dev))
> +		schedule_cxl_memdev_detach(ep_dev);
> +
> +	if (is_cxl_cachedev(ep_dev))
> +		schedule_cxl_cachedev_detach(ep_dev);

[Severity: Low]
In schedule_detach(), does this code implicitly cast the void pointer
representing a struct device directly to incompatible struct types?

Because ep_dev is a void pointer, passing it to schedule_cxl_memdev_detach()
and schedule_cxl_cachedev_detach() casts it to struct cxl_memdev * and
struct cxl_cachedev * without using the proper container_of() wrappers
like to_cxl_memdev() and to_cxl_cachedev().

While this works currently because struct device is the first member of these
structures, could this regression cause memory corruption if the struct
layouts ever change?

>  }

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260923173342.5584-1-Benjamin.Cheatham@amd.com?part=6