Re: [PATCH] usb: gadget: f_hid: do not copy_from_user() under a spinlock

[David Laight]

On Mon, 17 Aug 2026 14:11:41 +0800
Linkai Gong <gonglinkai@xxxxxxxxxx> wrote:

> f_hidg_get_report() already copied the report from userspace into a
> new entry, then called copy_from_user() again under
> get_report_spinlock. That can fault and sleep in atomic context.
> 
> Update the existing entry with memcpy() from the copy already taken.
> 
> Fixes: a139c98f760e ("USB: gadget: f_hid: Add GET_REPORT via userspace IOCTL")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Linkai Gong <gonglinkai@xxxxxxxxxx>
> ---
>  drivers/usb/gadget/function/f_hid.c | 11 +++--------
>  1 file changed, 3 insertions(+), 8 deletions(-)
> diff --git a/drivers/usb/gadget/function/f_hid.c b/drivers/usb/gadget/function/f_hid.c
> index 3c6b43d06a6d..e4621e5a0b69 100644
> --- a/drivers/usb/gadget/function/f_hid.c
> +++ b/drivers/usb/gadget/function/f_hid.c
> @@ -667,14 +667,9 @@ static int f_hidg_get_report(struct file *file, struct usb_hidg_report __user *b
>  	ptr = f_hidg_search_for_report(hidg, report_id);
>  
>  	if (ptr) {
> -		/* Report already exists in list - update it */
> -		if (copy_from_user(&ptr->report_data, buffer,
> -				sizeof(struct usb_hidg_report))) {
> -			spin_unlock_irqrestore(&hidg->get_report_spinlock, flags);
> -			ERROR(cdev, "copy_from_user error\n");

How about a patch to remove the ERROR() from the earlier copy_from_user()
error path?

	David

> -			kfree(entry);
> -			return -EINVAL;
> -		}
> +		/* Report already exists; data was copied before taking the lock. */
> +		memcpy(&ptr->report_data, &entry->report_data,
> +		       sizeof(struct usb_hidg_report));
>  		kfree(entry);
>  	} else {
>  		/* Report does not exist in list - add it */