[prev in list] [next in list] [prev in thread] [next in thread] 

List:       botan-devel
Subject:    [Botan-devel] Re: Stream cipher operations
From:       carlton () cs ! mcgill ! ca (Carlton Davis)
Date:       2005-10-13 21:47:24
Message-ID: 20051013174406.N62339 () axon ! CS ! McGill ! CA
[Download RAW message or body]


On Thu, 13 Oct 2005, Carlton Davis wrote:

Ah, I didn't even notice that. It is basically due to random chance - as I
> recall, sizeof(SecureVector<byte>) is 16 on 32-bit platforms (as the
> member
> data consists of 2 pointers + 2 unsigned ints), or 128 bits, so it just
> so
> happens that sizeof(bits) == bits.size().

> That said I would recommend changing that one as well, since while it
> works,
> it's more than a bit fragile.

> -Jack

This makes perfect sense.

Thanks much.

-Carlton


On Thu, 13 Oct 2005, Carlton Davis wrote:
>
> Thanks again Jack for your prompt response. What you
> suggested did fix the problem as usual. I am curious
> to know though, the reason why
> SecureVector<byte> ciphertext = enc->encrypt(bits,sizeof(bits));
> where bits is a SecureVector<byte>
> works, whereas for the decryption process
> SecureVector<byte> size() function needs to be used instead of
> sizeof(SecureVector<byte>).
>
> Best regards,
>
> -Carlton
>
>
> On Thu, 13 Oct 2005, Jack wrote:
>
> This should be ciphertext.size(), not sizeof(ciphertext). In the call, the
> ciphertext is truncated so you get a basically random number as the
> output. Since you're using Raw encoding here, this wasn't detected as an
> encoding error as it would if you were using OAEP or PKCS #1 v1.5. So the
> value
> of plaintext iends up being a random integer less than n, which will
> usually
> end up being around the modulus size (in this case, 1024 bits, or 128
> bytes),
> and WiderWAKE only supports 128-bit keys so it blows up when you try to
> pass
> this to set_key.
>
> On Thu, 13 Oct 2005, Carlton Davis wrote:
>
> > I ran into another issue during some additional test
> > involving some stream cipher operations. Essentially,
> > I want to generate a symmetric key or octet string
> > on a network node, encrypts the symmetric key with
> > nodeB public key, and sends the encrypted key to nodeB.
> > I use the code below to test the above operation
> > and I got the following:
> > ---
> >
> > "terminate called after throwing an instance of
> > 'Botan::Invalid_Key_Length'
> >   what():  Botan: WiderWake4+1-BE cannot accept a key of length 128
> > Abort trap"
> >
> > --
> >
> > Here is the code:
> >
> >   std::string passphrase = "testing";
> >   //Just testing here: a quick insecure way of getting a 128-bit key
> >   HashFunction* hash = get_hash("MD5");
> >   SymmetricKey skey = hash->process(passphrase);
> >
> >   unsigned short flag = 4;
> >   for (i = 0; i != sizeof(flag); i++)
> >     buf[i] = get_byte(i, flag);
> >   fprintf(stderr, "\nThe number is: %i\n", make_u16bit(buf[0], buf[1]));
> >
> >   StreamCipher* cipher = get_stream_cipher("WiderWake4+1-BE");
> >   cipher->set_key(skey);  //This is fine
> >   cipher->encrypt(buf, 2);
> >
> >   SecureVector<byte> bits = skey.bits_of();
> >   PK_Encryptor* enc = get_pk_encryptor(*pubkey, "Raw");
> >   SecureVector<byte> ciphertext = enc->encrypt(bits,sizeof(bits));
> >
> >   PK_Decryptor* dec = get_pk_decryptor(*privkey, "Raw");
> >   SecureVector<byte> plaintext = dec->decrypt(ciphertext, sizeof(ciphertext));
> >
> >   StreamCipher* cipher2 = get_stream_cipher("WiderWake4+1-BE");
> >   //The prblem is this set_key operation
> >   cipher2->set_key(SymmetricKey(plaintext));
> >   cipher2->decrypt(buf, 2);
> >
> >   fprintf(stderr, "\nAfter decryption, the byte values are %x and %x\n",
> > 	  buf[0], buf[1]);
> >   unsigned short decryptflag = make_u16bit(buf[0], buf[1]);
> >   fprintf(stderr, "\nThe decrypted number is: %i\n", decryptflag);
> > ---
> >
> > Any idea why WiderWake4+1 is not accepting the SymmetricKey object
> > I created?
> >
> > Thanks and best regards,
> >
> > -Carlton
> >
> >
> >
> >
>


[prev in list] [next in list] [prev in thread] [next in thread]