Re: [PATCH] KEYS: encrypted: Replace deprecated strcpy and improve get_derived_key

[Jarkko Sakkinen <jarkko@kernel.org>]

On Thu, Nov 13, 2025 at 02:58:31PM +0100, Thorsten Blum wrote:
> Determine 'key_name' before allocating memory for 'derived_buf' to only
> allocate as many bytes as needed. Currently, we potentially allocate one

Who is "we"?

> more byte than necessary when 'key_name' is "ENC_KEY".
> 
> strcpy() is deprecated and uses an additional strlen() internally; use
> memcpy() directly to copy 'key_name' since we already know its length
> and that it is guaranteed to be NUL-terminated.
> 
> Also reuse 'key_name_len' when copying 'master_key' instead of calling
> strlen() again.
> 
> Link: https://github.com/KSPP/linux/issues/88
> Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
> ---
>  security/keys/encrypted-keys/encrypted.c | 22 +++++++++-------------
>  1 file changed, 9 insertions(+), 13 deletions(-)
> 
> diff --git a/security/keys/encrypted-keys/encrypted.c b/security/keys/encrypted-keys/encrypted.c
> index 15841466b5d4..b16a5b8b935b 100644
> --- a/security/keys/encrypted-keys/encrypted.c
> +++ b/security/keys/encrypted-keys/encrypted.c
> @@ -12,6 +12,7 @@
>   */
>  
>  #include <linux/uaccess.h>
> +#include <linux/minmax.h>
>  #include <linux/module.h>
>  #include <linux/init.h>
>  #include <linux/slab.h>
> @@ -330,23 +331,18 @@ static int get_derived_key(u8 *derived_key, enum derived_key_type key_type,
>  			   const u8 *master_key, size_t master_keylen)
>  {
>  	u8 *derived_buf;
> -	unsigned int derived_buf_len;
> -
> -	derived_buf_len = strlen("AUTH_KEY") + 1 + master_keylen;
> -	if (derived_buf_len < HASH_SIZE)
> -		derived_buf_len = HASH_SIZE;
> +	size_t derived_buf_len;
> +	const char *key_name;
> +	size_t key_name_len;
>  
> +	key_name = key_type ? "AUTH_KEY" : "ENC_KEY";
> +	key_name_len = strlen(key_name) + 1;
> +	derived_buf_len = max(key_name_len + master_keylen, HASH_SIZE);
>  	derived_buf = kzalloc(derived_buf_len, GFP_KERNEL);
>  	if (!derived_buf)
>  		return -ENOMEM;
> -
> -	if (key_type)
> -		strcpy(derived_buf, "AUTH_KEY");
> -	else
> -		strcpy(derived_buf, "ENC_KEY");
> -
> -	memcpy(derived_buf + strlen(derived_buf) + 1, master_key,
> -	       master_keylen);
> +	memcpy(derived_buf, key_name, key_name_len);
> +	memcpy(derived_buf + key_name_len, master_key, master_keylen);
>  	sha256(derived_buf, derived_buf_len, derived_key);
>  	kfree_sensitive(derived_buf);
>  	return 0;
> -- 
> 2.51.1
> 

I don't see much value in this other than potentially causing merge
conflicts when backporting bug fixes.

BR, Jarkko