[PATCH 04/13] efi_loader: signature: fix a size check against revocation list

Heinrich Schuchardt xypron.glpk at gmx.de
Sat May 30 08:42:50 CEST 2020


On 5/29/20 8:41 AM, AKASHI Takahiro wrote:
> Since the size check against an entry in efi_search_siglist() is
> incorrect, this function will never find out a to-be-matched certificate
> and its associated revocation time in signature list.

%s/in signature/in the signature/

>
> Signed-off-by: AKASHI Takahiro <takahiro.akashi at linaro.org>
> ---
>  lib/efi_loader/efi_signature.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/lib/efi_loader/efi_signature.c b/lib/efi_loader/efi_signature.c
> index be6491c6e255..35f678de057e 100644
> --- a/lib/efi_loader/efi_signature.c
> +++ b/lib/efi_loader/efi_signature.c
> @@ -432,10 +432,11 @@ static bool efi_search_siglist(struct x509_certificate *cert,
>  		 *	time64_t revocation_time;
>  		 * };
>  		 */
> -		if ((sig_data->size == SHA256_SUM_LEN) &&
> -		    !memcmp(sig_data->data, hash, SHA256_SUM_LEN)) {
> +		if ((sig_data->size >= SHA256_SUM_LEN + sizeof(time64_t)) &&
> +		    !memcmp(sig_data->data, msg, SHA256_SUM_LEN)) {
>  			memcpy(revoc_time, sig_data->data + SHA256_SUM_LEN,
>  			       sizeof(*revoc_time));
> +			debug("revocation time: 0x%llx\n", *revoc_time);

Since this is seconds since 1970 wouldn't it be reasonable to use
decimal output (%llu)?

Otherwise:
Reviewed-by: Heinrich Schuchardt <xypron.glpk at gmx.de>

>  			found = true;
>  			goto out;
>  		}
>


More information about the U-Boot mailing list