Re: [PATCH v2] tty: vt: keyboard: initialize "kbs" so that kfree(kbs) runs fine even if kbs is not kmalloced.

[Pavel Skripkin <paskripkin@gmail.com>]

Hi, Ajay!

On 11/6/21 13:40, Ajay Garg wrote:
> 
> v1 patch at :
> https://lore.kernel.org/linux-serial/YYZN30qfaKMskVwE@kroah.com/T/#t
> 
> 
> Changes in v2 :
> 
>          * Changes as required by scripts/checkpatch.pl
> 
>          * Checking whether kbs is not NULL before kfree is not required,
>            as kfree(NULL) is safe. So, dropped the check.
> 
> 
> For brevity, here is the background :
> 
> 

Please, don't put change log into commit message. It should go under ---

> In "vt_do_kdgkb_ioctl", kbs is kmalloced, if cmd is one of KDGKBSENT or
> KDSKBSENT.
> 
> If cmd is none of the above, kbs is not kmalloced, and runs
> direct to kfree(kbs).
> 
> Values of local-variables on the stack can take indeterminate values,
> so we initialize kbs to NULL. Then, if kbs is not kmalloced, we have
> kfree(NULL) at the last.
> 
> Note that kfree(NULL) is safe.
> 
> 

These is only one caller of vt_do_kdgkb_ioctl, which simple does:


	case KDGKBSENT:
	case KDSKBSENT:
		return vt_do_kdgkb_ioctl(cmd, up, perm);

It means, that cmd can not be different from KDGKBSENT and KDSKBSENT.

I guess, you found this "issue" via static analysis tool like smatch or 
smth similar, but this bug is impossible right now.

> 
> Signed-off-by: Ajay Garg <ajaygargnsit@gmail.com>
> ---
<--- here

>   drivers/tty/vt/keyboard.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/tty/vt/keyboard.c b/drivers/tty/vt/keyboard.c
> index dfef7de8a057..54155fc91cd2 100644
> --- a/drivers/tty/vt/keyboard.c
> +++ b/drivers/tty/vt/keyboard.c
> @@ -2049,7 +2049,7 @@ int vt_do_kdgkb_ioctl(int cmd, struct kbsentry __user *user_kdgkb, int perm)
>   {
>   	unsigned char kb_func;
>   	unsigned long flags;
> -	char *kbs;
> +	char *kbs = NULL;
>   	int ret;
>   
>   	if (get_user(kb_func, &user_kdgkb->kb_func))
> 


With regards,
Pavel Skripkin