Date: Wed, 17 Sep 2003 12:52:24 +0400 (MSD) From: "lg" <zevlg@yandex.ru> To: Hajimu UMEMOTO <ume@FreeBSD.org> Cc: hackers@freebsd.org Subject: Re: possible rijndael bug Message-ID: <3F6820C8.000001.18006@tide.yandex.ru>
next in thread | raw e-mail | index | archive | help
> vlm> Absolutely NOT. > vlm> According to RFC1423 and FIPS81, the padding length may be somewhere > vlm> in between 1 to 16 bytes, which translated into > vlm> if(padLen < 0 || padLen >= 16) > vlm> for this particular code. > Ah, yes. Then, `assert(padLen > 0 && padLen <= 16)'; should be wrong. The thing is that our padLen is not real padlen it is the (BlockSize-realpadlen), and since BlockSize is 16 and realpadlen is in [0, 16) range(as vlm writes) than our padLen will be in (0, 16] range. > Sincerely,
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3F6820C8.000001.18006>