1 | =head1 NAME |
1 | =head1 NAME |
2 | |
2 | |
3 | Crypt::Spritz - Crypt::CBC compliant Spritz encryption/hash/mac/aead/prng module |
3 | Crypt::Spritz - Spritz stream cipher/hash/MAC/AEAD/CSPRNG family |
4 | |
4 | |
5 | =head1 SYNOPSIS |
5 | =head1 SYNOPSIS |
6 | |
6 | |
7 | use Crypt::Spritz; |
7 | use Crypt::Spritz; |
8 | |
8 | |
9 | # keysize() is 32, but spritz accepts any key size |
9 | # see the commented examples in their respective classes, |
10 | # blocksize() is 16, but cna be anything |
10 | # but basically |
11 | |
11 | |
12 | $cipher = new Crypt::Twofish2 "a" x 32, Crypt::Twofish2::MODE_CBC; |
12 | my $cipher = new Crypt::Spritz::Cipher::XOR $key, $iv; |
|
|
13 | $ciphertext = $cipher->crypt ($cleartext); |
13 | |
14 | |
|
|
15 | my $cipher = new Crypt::Spritz::Cipher $key, $iv; |
14 | $crypted = $cipher->encrypt($plaintext); |
16 | $ciphertext = $cipher->encrypt ($cleartext); |
15 | # - OR - |
|
|
16 | $plaintext = $cipher->decrypt($crypted); |
17 | # $cleartext = $cipher->decrypt ($ciphertext); |
|
|
18 | |
|
|
19 | my $hasher = new Crypt::Spritz::Hash; |
|
|
20 | $hasher->add ($data); |
|
|
21 | $digest = $hasher->finish; |
|
|
22 | |
|
|
23 | my $hasher = new Crypt::Spritz::MAC $key; |
|
|
24 | $hasher->add ($data); |
|
|
25 | $mac = $hasher->finish; |
|
|
26 | |
|
|
27 | my $prng = new Crypt::Spritz::PRNG $entropy; |
|
|
28 | $prng->add ($additional_entropy); |
|
|
29 | $keydata = $prng->get (32); |
|
|
30 | |
|
|
31 | my $aead = new Crypt::Spritz::AEAD::XOR $key; |
|
|
32 | $aead->nonce ($counter); |
|
|
33 | $aead->associated_data ($header); |
|
|
34 | $ciphertext = $aead->crypt ($cleartext); |
|
|
35 | $mac = $aead->mac; |
|
|
36 | |
|
|
37 | my $aead = new Crypt::Spritz::AEAD $key; |
|
|
38 | $aead->nonce ($counter); |
|
|
39 | $aead->associated_data ($header); |
|
|
40 | $ciphertext = $aead->encrypt ($cleartext); |
|
|
41 | # $cleartext = $aead->decrypt ($ciphertext); |
|
|
42 | $mac = $aead->mac; |
17 | |
43 | |
18 | =head1 DESCRIPTION |
44 | =head1 DESCRIPTION |
19 | |
45 | |
20 | This module implements the spritz spongelike function. |
46 | This module implements the Spritz spongelike function (with N=256), the |
|
|
47 | spiritual successor of RC4 developed by Ron Rivest and Jacob Schuldt. |
21 | |
48 | |
22 | Although it is C<Crypt::CBC> compliant you usually gain nothing by using |
49 | Its strength is extreme versatility (you get a stream cipher, a hash, a |
23 | that module (except generality, which is often a good thing), since |
50 | MAC, a DRBG/CSPRNG, an authenticated encryption block/stream cipher and |
24 | C<Crypt::Twofish2> can work in either ECB or CBC mode itself. |
51 | more) and extremely simple and small code (encryption and authentication |
|
|
52 | can be had in 1KB of compiled code on amd64, which isn't an issue for most |
|
|
53 | uses in Perl, but is useful in embedded situations, or e.g. when doing |
|
|
54 | crypto using javascript in a browser and communicating with Perl). |
25 | |
55 | |
26 | =over 4 |
56 | Its weakness is its relatively slow speed (encryption is a few times |
|
|
57 | slower than RC4 or AES, hashing many times slower than SHA-3, although |
|
|
58 | this might be reversed on an 8-bit-cpu) and the fact that it is totally |
|
|
59 | unproven in the field (as of this writing, the cipher was just a few |
|
|
60 | months old), so it can't be called production-ready. |
|
|
61 | |
|
|
62 | All the usual caveats regarding stream ciphers apply - never repeat your |
|
|
63 | key, never repeat your nonce and so on - you should have some basic |
|
|
64 | understanding of cryptography before using this cipher in your own |
|
|
65 | designs. |
|
|
66 | |
|
|
67 | The Spritz base class is not meant for end users. To make usage simpler |
|
|
68 | and safer, a number of convenience classes are provided for typical |
|
|
69 | end-user tasks: |
|
|
70 | |
|
|
71 | random number generation - Crypt::Spritz::PRNG |
|
|
72 | hashing - Crypt::Spritz::Hash |
|
|
73 | message authentication - Crypt::Spritz::MAC |
|
|
74 | encryption - Crypt::Spritz::Cipher::XOR |
|
|
75 | encryption - Crypt::Spritz::Cipher |
|
|
76 | authenticated encryption - Crypt::Spritz::AEAD::XOR |
|
|
77 | authenticated encryption - Crypt::Spritz::AEAD |
27 | |
78 | |
28 | =cut |
79 | =cut |
29 | |
80 | |
30 | package Crypt::Spritz; |
81 | package Crypt::Spritz; |
31 | |
82 | |
32 | use XSLoader; |
83 | use XSLoader; |
33 | |
84 | |
34 | $VERSION = '0.0'; |
85 | $VERSION = 0.2; |
35 | |
86 | |
36 | XSLoader::load __PACKAGE__, $VERSION; |
87 | XSLoader::load __PACKAGE__, $VERSION; |
37 | |
88 | |
38 | @Crypt::Spritz::CipherBase::ISA = |
89 | @Crypt::Spritz::ISA = Crypt::Spritz::Base::; |
|
|
90 | |
39 | @Crypt::Spritz::Hash::ISA = |
91 | @Crypt::Spritz::Hash::ISA = |
|
|
92 | @Crypt::Spritz::PRNG::ISA = |
|
|
93 | @Crypt::Spritz::Cipher::ISA = |
40 | @Crypt::Spritz::PRNG::ISA = Crypt::Spritz::; |
94 | @Crypt::Spritz::AEAD::ISA = Crypt::Spritz::Base::; |
41 | |
95 | |
42 | @Crypt::Spritz::MAC::ISA = Crypt::Spritz::Hash::; |
96 | @Crypt::Spritz::MAC::ISA = Crypt::Spritz::Hash::; |
43 | |
97 | |
44 | @Crypt::Spritz::Cipher::XOR::ISA = |
98 | @Crypt::Spritz::Cipher::XOR::ISA = Crypt::Spritz::Cipher::; |
45 | @Crypt::Spritz::Cipher::ISA = |
|
|
46 | @Crypt::Spritz::AEAD::ISA = |
|
|
47 | @Crypt::Spritz::AEAD::XOR::ISA = Crypt::Spritz::CipherBase::; |
99 | @Crypt::Spritz::AEAD::XOR::ISA = Crypt::Spritz::AEAD::; |
48 | |
100 | |
49 | sub Crypt::Spritz::CipherBase::keysize () { 32 } |
101 | sub Crypt::Spritz::Cipher::keysize () { 32 } |
50 | sub Crypt::Spritz::CipherBase::blocksize () { 64 } |
102 | sub Crypt::Spritz::Cipher::blocksize () { 64 } |
|
|
103 | |
|
|
104 | *Crypt::Spritz::Hash::new = \&Crypt::Spritz::new; |
51 | |
105 | |
52 | *Crypt::Spritz::Hash::add = |
106 | *Crypt::Spritz::Hash::add = |
53 | *Crypt::Spritz::PRNG::add = \&Crypt::Spritz::absorb; |
107 | *Crypt::Spritz::PRNG::add = \&Crypt::Spritz::absorb; |
54 | |
108 | |
55 | *Crypt::Spritz::PRNG::get = \&Crypt::Spritz::squeeze; |
109 | *Crypt::Spritz::PRNG::get = \&Crypt::Spritz::squeeze; |
56 | |
110 | |
57 | *Crypt::Spritz::AEAD::XOR::finish = |
111 | *Crypt::Spritz::AEAD::new = \&Crypt::Spritz::MAC::new; |
58 | *Crypt::Spritz::AEAD::finish = \&Crypt::Spritz::Hash::finish; |
112 | *Crypt::Spritz::AEAD::finish = \&Crypt::Spritz::Hash::finish; |
59 | |
113 | |
60 | *Crypt::Spritz::AEAD::XOR::associated_data = |
114 | *Crypt::Spritz::AEAD::associated_data = |
61 | *Crypt::Spritz::AEAD::associated_data = |
|
|
62 | *Crypt::Spritz::AEAD::XOR::nonce = |
|
|
63 | *Crypt::Spritz::AEAD::nonce = \&Crypt::Spritz::absborb_and_stop; |
115 | *Crypt::Spritz::AEAD::nonce = \&Crypt::Spritz::absorb_and_stop; |
64 | |
116 | |
65 | =item keysize |
|
|
66 | |
117 | |
67 | Returns the keysize, which is 32 (bytes). The Twofish2 cipher actually |
118 | =head2 THE Crypt::Spritz CLASS |
68 | supports keylengths of 16, 24 or 32 bytes, but there is no way to |
|
|
69 | communicate this to C<Crypt::CBC>. |
|
|
70 | |
119 | |
71 | =item blocksize |
120 | This class implements most of the Spritz primitives. To use it effectively |
|
|
121 | you should understand them, for example, by reading the L<Spritz |
|
|
122 | paper|http://people.csail.mit.edu/rivest/pubs/RS14.pdf>, especially |
|
|
123 | pp. 5-6. |
72 | |
124 | |
73 | The blocksize for Twofish2 is 16 bytes (128 bits), which is somewhat |
125 | The Spritz primitive corresponding to the Perl method is given as |
74 | unique. It is also the reason I need this module myself ;) |
126 | comment. |
75 | |
127 | |
76 | =item $cipher = new $key [, $mode] |
128 | =over 4 |
77 | |
129 | |
78 | Create a new C<Crypt::Twofish2> cipher object with the given key (which |
130 | =item $spritz = new Crypt::Spritz # InitializeState |
79 | must be 128, 192 or 256 bits long). The additional C<$mode> argument is |
|
|
80 | the encryption mode, either C<MODE_ECB> (electronic cookbook mode, the |
|
|
81 | default), C<MODE_CBC> (cipher block chaining, the same that C<Crypt::CBC> |
|
|
82 | does) or C<MODE_CFB1> (1-bit cipher feedback mode). |
|
|
83 | |
131 | |
84 | ECB mode is very insecure (read a book on cryptography if you don't know |
132 | Creates and returns a new, initialised Spritz state. |
85 | why!), so you should probably use CBC mode. CFB1 mode is not tested and is |
|
|
86 | most probably broken, so do not try to use it. |
|
|
87 | |
133 | |
88 | In ECB mode you can use the same cipher object to encrypt and decrypt |
134 | =item $spritz->init # InitializeState |
89 | data. However, every change of "direction" causes an internal reordering |
|
|
90 | of key data, which is quite slow, so if you want ECB mode and |
|
|
91 | encryption/decryption at the same time you should create two seperate |
|
|
92 | C<Crypt::Twofish2> objects with the same key. |
|
|
93 | |
135 | |
94 | In CBC mode you have to use seperate objects for encryption/decryption in |
136 | Initialises the Spritz state again, throwing away the previous state. |
95 | any case. |
|
|
96 | |
137 | |
97 | The C<MODE_*>-constants are not exported by this module, so you must |
138 | =item $another_spritz = $spritz->clone |
98 | specify them as C<Crypt::Twofish2::MODE_CBC> etc. (sorry for that). |
|
|
99 | |
139 | |
100 | =item $cipher->encrypt($data) |
140 | Make an exact copy of the spritz state. This method can be called on all |
|
|
141 | of the objects in this module, but is documented separately to give some |
|
|
142 | cool usage examples. |
101 | |
143 | |
102 | Encrypt data. The size of C<$data> must be a multiple of C<blocksize> (16 |
144 | =item $spritz->update # Update |
103 | bytes), otherwise this function will croak. Apart from that, it can be of |
|
|
104 | (almost) any length. |
|
|
105 | |
145 | |
106 | =item $cipher->decrypt($data) |
146 | =item $spritz->whip ($r) # Whip |
107 | |
147 | |
108 | The pendant to C<encrypt> in that it I<de>crypts data again. |
148 | =item $spritz->crush # Crush |
109 | |
149 | |
|
|
150 | =item $spritz->shuffle # Shuffle |
|
|
151 | |
|
|
152 | =item $spritz->output # Output |
|
|
153 | |
|
|
154 | Calls the Spritz primitive ovf the same name - these are not normally |
|
|
155 | called manually. |
|
|
156 | |
|
|
157 | =item $spritz->absorb ($I) # Absorb |
|
|
158 | |
|
|
159 | Absorbs the given data into the state (usually used for key material, |
|
|
160 | nonces, IVs messages to be hashed and so on). |
|
|
161 | |
|
|
162 | =item $spritz->absorb_stop # AbsorbStop |
|
|
163 | |
|
|
164 | Absorbs a special stop symbol - this is usually used as delimiter between |
|
|
165 | multiple strings to be absorbed, to thwart extension attacks. |
|
|
166 | |
|
|
167 | =item $spritz->absorb_and_stop ($I) |
|
|
168 | |
|
|
169 | This is a convenience function that simply calls C<absorb> followed by |
|
|
170 | C<absorb_stop>. |
|
|
171 | |
|
|
172 | =item $octet = $spritz->drip # Drip |
|
|
173 | |
|
|
174 | Squeezes out a single byte from the state. |
|
|
175 | |
|
|
176 | =item $octets = $spritz->squeeze ($len) # Squeeze |
|
|
177 | |
|
|
178 | Squeezes out the requested number of bytes from the state - this is usually |
|
|
179 | |
110 | =back |
180 | =back |
|
|
181 | |
|
|
182 | |
|
|
183 | =head2 THE Crypt::Spritz::PRNG CLASS |
|
|
184 | |
|
|
185 | This class implements a Pseudorandom Number Generatore (B<PRNG>), |
|
|
186 | sometimes also called a Deterministic Random Bit Generator (B<DRBG>). In |
|
|
187 | fact, it is even cryptographically secure, making it a B<CSPRNG>. |
|
|
188 | |
|
|
189 | Typical usage as a random number generator involves creating a PRNG |
|
|
190 | object with a seed of your choice, and then fetching randomness via |
|
|
191 | C<get>: |
|
|
192 | |
|
|
193 | # create a PRNG object, use a seed string of your choice |
|
|
194 | my $prng = new Crypt::Spritz::PRNG $seed; |
|
|
195 | |
|
|
196 | # now call get as many times as you wish to get binary randomness |
|
|
197 | my $some_randomness = $prng->get (17); |
|
|
198 | my moree_randomness = $prng->get (5000); |
|
|
199 | ... |
|
|
200 | |
|
|
201 | Typical usage as a cryptographically secure random number generator is to |
|
|
202 | feed in some secret entropy (32 octets/256 bits are commonly considered |
|
|
203 | enough), for example from C</dev/random> or C</dev/urandom>, and then |
|
|
204 | generate some key material. |
|
|
205 | |
|
|
206 | # create a PRNG object |
|
|
207 | my $prng = new Crypt::Spritz::PRNG; |
|
|
208 | |
|
|
209 | # seed some entropy (either via ->add or in the constructor) |
|
|
210 | $prng->add ($some_secret_highly_entropic_string); |
|
|
211 | |
|
|
212 | # now call get as many times as you wish to get |
|
|
213 | # hard to guess binary randomness |
|
|
214 | my $key1 = $prng->get (32); |
|
|
215 | my $key2 = $prng->get (16); |
|
|
216 | ... |
|
|
217 | |
|
|
218 | # for long running programs, it is advisable to |
|
|
219 | # reseed the PRNG from time to time with new entropy |
|
|
220 | $prng->add ($some_more_entropy); |
|
|
221 | |
|
|
222 | =over 4 |
|
|
223 | |
|
|
224 | =item $prng = new Crypt::Spritz::PRNG [$seed] |
|
|
225 | |
|
|
226 | Creates a new random number generator object. If C<$seed> is given, then |
|
|
227 | the C<$seed> is added to the internal state as if by a call to C<add>. |
|
|
228 | |
|
|
229 | =item $prng->add ($entropy) |
|
|
230 | |
|
|
231 | Adds entropy to the internal state, thereby hopefully making it harder |
|
|
232 | to guess. Good sources for entropy are irregular hardware events, or |
|
|
233 | randomness provided by C</dev/urandom> or C</dev/random>. |
|
|
234 | |
|
|
235 | The design of the Spritz PRNG should make it strong against attacks where |
|
|
236 | the attacker controls all the entropy, so it should be safe to add entropy |
|
|
237 | from untrusted sources - more is better than less if you need a CSPRNG. |
|
|
238 | |
|
|
239 | For use as PRNG, of course, this matters very little. |
|
|
240 | |
|
|
241 | =item $octets = $prng->get ($length) |
|
|
242 | |
|
|
243 | Generates and returns C<$length> random octets as a string. |
|
|
244 | |
|
|
245 | =back |
|
|
246 | |
|
|
247 | |
|
|
248 | =head2 THE Crypt::Spritz::Hash CLASS |
|
|
249 | |
|
|
250 | This implements the Spritz digest/hash algorithm. It works very similar to |
|
|
251 | other digest modules on CPAN, such as L<Digest::SHA3>. |
|
|
252 | |
|
|
253 | Typical use for hashing: |
|
|
254 | |
|
|
255 | # create hasher object |
|
|
256 | my $hasher = new Crypt::Spritz::Hash; |
|
|
257 | |
|
|
258 | # now feed data to be hashed into $hasher |
|
|
259 | # in as few or many calls as required |
|
|
260 | $hasher->add ("Some data"); |
|
|
261 | $hasher->add ("Some more"); |
|
|
262 | |
|
|
263 | # extract the hash - the object is not usable afterwards |
|
|
264 | my $digest = $hasher->finish (32); |
|
|
265 | |
|
|
266 | =over 4 |
|
|
267 | |
|
|
268 | =item $hasher = new Crypt::Spritz::Hash |
|
|
269 | |
|
|
270 | Creates a new hasher object. |
|
|
271 | |
|
|
272 | =item $hasher->add ($data) |
|
|
273 | |
|
|
274 | Adds data to be hashed into the hasher state. It doesn't matter whether |
|
|
275 | you pass your data in in one go or split it up, the hash will be the same. |
|
|
276 | |
|
|
277 | =item $digest = $hasher->finish ($length) |
|
|
278 | |
|
|
279 | Calculates a hash digest of the given length and return it. The object |
|
|
280 | cannot sensibly be used for further hashing afterwards. |
|
|
281 | |
|
|
282 | Typical digest lengths are 16 and 32, corresponding to 128 and 256 bit |
|
|
283 | digests, respectively. |
|
|
284 | |
|
|
285 | =item $another_hasher = $hasher->clone |
|
|
286 | |
|
|
287 | Make an exact copy of the hasher state. This can be useful to generate |
|
|
288 | incremental hashes, for example. |
|
|
289 | |
|
|
290 | Example: generate a hash for the data already fed into the hasher, by keeping |
|
|
291 | the original hasher for further C<add> calls and calling C<finish> on a C<clone>. |
|
|
292 | |
|
|
293 | my $intermediate_hash = $hasher->clone->finish; |
|
|
294 | |
|
|
295 | Example: hash 64KiB of data, and generate a hash after every kilobyte that |
|
|
296 | is over the full data. |
|
|
297 | |
|
|
298 | my $hasher = new Crypt::Spritz::Hash; |
|
|
299 | |
|
|
300 | for (0..63) { |
|
|
301 | my $kib = "x" x 1024; # whatever data |
|
|
302 | |
|
|
303 | $hasher->add ($kib); |
|
|
304 | |
|
|
305 | my $intermediate_hash = $hasher->clone->finish; |
|
|
306 | ... |
|
|
307 | } |
|
|
308 | |
|
|
309 | These kind of intermediate hashes are sometimes used in communications |
|
|
310 | protocols to protect the integrity of the data incrementally, e.g. to |
|
|
311 | detect errors early, while still having a complete hash at the end of a |
|
|
312 | transfer. |
|
|
313 | |
|
|
314 | =back |
|
|
315 | |
|
|
316 | |
|
|
317 | =head2 THE Crypt::Spritz::MAC CLASS |
|
|
318 | |
|
|
319 | This implements the Spritz Message Authentication Code algorithm. It works |
|
|
320 | very similar to other digest modules on CPAN, such as L<Digest::SHA3>, but |
|
|
321 | implements an authenticated digest (like L<Digest::HMAC>). |
|
|
322 | |
|
|
323 | I<Authenticated> means that, unlike L<Crypt::Spritz::Hash>, where |
|
|
324 | everybody can verify and recreate the hash value for some data, with a |
|
|
325 | MAC, knowledge of the (hopefully) secret key is required both to create |
|
|
326 | and to verify the digest. |
|
|
327 | |
|
|
328 | Typical use for hashing is almost the same as with L<Crypt::Spritz::MAC>, |
|
|
329 | except a key (typically 16 or 32 octets) is provided to the constructor: |
|
|
330 | |
|
|
331 | # create hasher object |
|
|
332 | my $hasher = new Crypt::Spritz::Mac $key; |
|
|
333 | |
|
|
334 | # now feed data to be hashed into $hasher |
|
|
335 | # in as few or many calls as required |
|
|
336 | $hasher->add ("Some data"); |
|
|
337 | $hasher->add ("Some more"); |
|
|
338 | |
|
|
339 | # extract the mac - the object is not usable afterwards |
|
|
340 | my $mac = $hasher->finish (32); |
|
|
341 | |
|
|
342 | =over 4 |
|
|
343 | |
|
|
344 | =item $hasher = new Crypt::Spritz::MAC $key |
|
|
345 | |
|
|
346 | Creates a new hasher object. The C<$key> can be of any length, but 16 and |
|
|
347 | 32 (128 and 256 bit) are customary. |
|
|
348 | |
|
|
349 | =item $hasher->add ($data) |
|
|
350 | |
|
|
351 | Adds data to be hashed into the hasher state. It doesn't matter whether |
|
|
352 | you pass your data in in one go or split it up, the hash will be the same. |
|
|
353 | |
|
|
354 | =item $mac = $hasher->finish ($length) |
|
|
355 | |
|
|
356 | Calculates a message code of the given length and return it. The object |
|
|
357 | cannot sensibly be used for further hashing afterwards. |
|
|
358 | |
|
|
359 | Typical digest lengths are 16 and 32, corresponding to 128 and 256 bit |
|
|
360 | digests, respectively. |
|
|
361 | |
|
|
362 | =item $another_hasher = $hasher->clone |
|
|
363 | |
|
|
364 | Make an exact copy of the hasher state. This can be useful to |
|
|
365 | generate incremental macs, for example. |
|
|
366 | |
|
|
367 | See the description for the C<Crypt::Spritz::Hash::clone> method for some |
|
|
368 | examples. |
|
|
369 | |
|
|
370 | =back |
|
|
371 | |
|
|
372 | |
|
|
373 | =head2 THE Crypt::Spritz::Cipher::XOR CLASS |
|
|
374 | |
|
|
375 | This class implements stream encryption/decryption. It doesn't implement |
|
|
376 | the standard Spritz encryption but the XOR variant (called B<spritz-xor> |
|
|
377 | in the paper). |
|
|
378 | |
|
|
379 | The XOR variant should be as secure as the standard variant, but |
|
|
380 | doesn't have separate encryption and decryaption functions, which saves |
|
|
381 | codesize. IT is not compatible with standard Spritz encryption, however - |
|
|
382 | drop me a note if you want that implemented as well. |
|
|
383 | |
|
|
384 | Typical use for encryption I<and> decryption (code is identical for |
|
|
385 | decryption, you simply pass the encrypted data to C<crypt>): |
|
|
386 | |
|
|
387 | # create a cipher - $salt can be a random string you send |
|
|
388 | # with your message, in clear, a counter (best), or empty if |
|
|
389 | # you only want to encrypt one message with the given key. |
|
|
390 | # 16 or 32 octets are typical sizes for the key, for the salt, |
|
|
391 | # use whatever you need to give a unique salt for every |
|
|
392 | # message you encrypt with the same key. |
|
|
393 | |
|
|
394 | my $cipher = Crypt::Spritz::Cipher::XOR $key, $salt; |
|
|
395 | |
|
|
396 | # encrypt a message in one or more calls to crypt |
|
|
397 | |
|
|
398 | my $encrypted; |
|
|
399 | |
|
|
400 | $encrypted .= $cipher->crypt ("This is"); |
|
|
401 | $encrypted .= $cipher->crypt ("all very"); |
|
|
402 | $encrypted .= $cipher->crypt ("secret"); |
|
|
403 | |
|
|
404 | # that's all |
|
|
405 | |
|
|
406 | =over 4 |
|
|
407 | |
|
|
408 | =item $cipher = new Crypt::Spritz::Cipher::XOR $key[, $iv] |
|
|
409 | |
|
|
410 | Creates a new cipher object usable for encryption and decryption. The |
|
|
411 | C<$key> must be provided, the initial vector C<$IV> is optional. |
|
|
412 | |
|
|
413 | Both C<$key> and C<$IV> can be of any length. Typical lengths for the |
|
|
414 | C<$key> are 16 (128 bit) or 32 (256 bit), while the C<$IV> simply needs to |
|
|
415 | be long enough to distinguish repeated uses of tghe same key. |
|
|
416 | |
|
|
417 | =item $encrypted = $cipher->crypt ($cleartext) |
|
|
418 | |
|
|
419 | =item $cleartext = $cipher->crypt ($encrypted) |
|
|
420 | |
|
|
421 | Encrypt or decrypt a piece of a message. This can be called as many times |
|
|
422 | as you want, and the message can be split into as few or many pieces as |
|
|
423 | required without affecting the results. |
|
|
424 | |
|
|
425 | =item $cipher->crypt_inplace ($cleartext_or_ciphertext) |
|
|
426 | |
|
|
427 | Same as C<crypt>, except it I<modifies the argument in-place>. |
|
|
428 | |
|
|
429 | =item $another_cipher = $cipher->clone |
|
|
430 | |
|
|
431 | Make an exact copy of the cipher state. This can be useful to cache states |
|
|
432 | for reuse later, for example, to avoid expensive key setups. |
|
|
433 | |
|
|
434 | While there might be use cases for this feature, it makes a lot more sense |
|
|
435 | for C<Crypt::Spritz::AEAD> and C<Crypt::Spritz::AEAD::XOR>, as they allow |
|
|
436 | you to specify the IV/nonce separately. |
|
|
437 | |
|
|
438 | =item $constant_32 = $cipher->keysize |
|
|
439 | |
|
|
440 | =item $constant_64 = $cipher->blocksize |
|
|
441 | |
|
|
442 | These methods are provided for L<Crypt::CBC> compatibility and simply |
|
|
443 | return C<32> and C<64>, respectively. |
|
|
444 | |
|
|
445 | Note that it is pointless to use Spritz with L<Crypt::CBC>, as Spritz is |
|
|
446 | not a block cipher and already provides an appropriate mode. |
|
|
447 | |
|
|
448 | =back |
|
|
449 | |
|
|
450 | |
|
|
451 | =head2 THE Crypt::Spritz::Cipher CLASS |
|
|
452 | |
|
|
453 | This class is pretty much the same as the C<Crypt::Spritz::Cipher::XOR> |
|
|
454 | class, with two differences: first, it implements the "standard" Spritz |
|
|
455 | encryption algorithm, and second, while this variant is easier to analyze |
|
|
456 | mathematically, there is little else to recommend it for, as it is slower, |
|
|
457 | and requires lots of code duplication code. |
|
|
458 | |
|
|
459 | So unless you need to be compatible with another implementation that does |
|
|
460 | not offer the XOR variant, stick to C<Crypt::Spritz::Cipher::XOR>. |
|
|
461 | |
|
|
462 | All the methods from C<Crypt::Spritz::Cipher::XOR> are available, except |
|
|
463 | C<crypt>, which has been replaced by separate C<encrypt> and C<decrypt> |
|
|
464 | methods: |
|
|
465 | |
|
|
466 | =over 4 |
|
|
467 | |
|
|
468 | =item $encrypted = $cipher->encrypt ($cleartext) |
|
|
469 | |
|
|
470 | =item $cleartext = $cipher->decrypt ($encrypted) |
|
|
471 | |
|
|
472 | Really the same as C<Crypt::Spritz::Cipher::XOR>, except you need separate |
|
|
473 | calls and code for encryption and decryption. |
|
|
474 | |
|
|
475 | =back |
|
|
476 | |
|
|
477 | |
|
|
478 | =head2 THE Crypt::Spritz::AEAD::XOR CLASS |
|
|
479 | |
|
|
480 | This is the most complicated class - it combines encryption and |
|
|
481 | message authentication into a single "authenticated encryption |
|
|
482 | mode". It is similar to using both L<Crypt::Spritz::Cipher::XOR> and |
|
|
483 | L<Crypt::Spritz::MAC>, but makes it harder to make mistakes in combining |
|
|
484 | them. |
|
|
485 | |
|
|
486 | You can additionally provide cleartext data that will not be encrypted or |
|
|
487 | decrypted, but that is nevertheless authenticated using the MAC, which |
|
|
488 | is why this mode is called I<AEAD>, I<Authenticated Encryption with |
|
|
489 | Associated Data>. Associated data is usually used to any header data that |
|
|
490 | is in cleartext, but should nevertheless be authenticated. |
|
|
491 | |
|
|
492 | This implementation implements the XOR variant. Just as with |
|
|
493 | L<Crypt::Spritz::Cipher::XOR>, this means it is not compatible with |
|
|
494 | the standard mode, but uses less code and doesn't distinguish between |
|
|
495 | encryption and decryption. |
|
|
496 | |
|
|
497 | Typical usage is as follows: |
|
|
498 | |
|
|
499 | # create a new aead object |
|
|
500 | # you use one object per message |
|
|
501 | # key length customarily is 16 or 32 |
|
|
502 | my $aead = new Crypt::Spritz::AEAD::XOR $key; |
|
|
503 | |
|
|
504 | # now you must feed the nonce. if you do not need a nonce, |
|
|
505 | # you can provide the empty string, but you have to call it |
|
|
506 | # after creating the object, before calling associated_data. |
|
|
507 | # the nonce must be different for each usage of the $key. |
|
|
508 | # a counter of some kind is good enough. |
|
|
509 | # reusing a nonce with the same key completely |
|
|
510 | # destroys security! |
|
|
511 | $aead->nonce ($counter); |
|
|
512 | |
|
|
513 | # then you must feed any associated data you have. if you |
|
|
514 | # do not have associated cleartext data, you can provide the empty |
|
|
515 | # string, but you have to call it after nonce and before crypt. |
|
|
516 | $aead->associated_data ($header); |
|
|
517 | |
|
|
518 | # next, you call crypt one or more times with your data |
|
|
519 | # to be encrypted (opr decrypted). |
|
|
520 | # all except the last call must use a length that is a |
|
|
521 | # multiple of 64. |
|
|
522 | # the last block can have any length. |
|
|
523 | my $encrypted; |
|
|
524 | |
|
|
525 | $encrypted .= $aead->crypt ("1" x 64); |
|
|
526 | $encrypted .= $aead->crypt ("2" x 64); |
|
|
527 | $encrypted .= $aead->crypt ("3456"); |
|
|
528 | |
|
|
529 | # finally you can calculate the MAC for all of the above |
|
|
530 | my $mac = $aead->finish; |
|
|
531 | |
|
|
532 | =over 4 |
|
|
533 | |
|
|
534 | =item $aead = new Crypt::Spritz::AEAD::XOR $key |
|
|
535 | |
|
|
536 | Creates a new cipher object usable for encryption and decryption. |
|
|
537 | |
|
|
538 | The C<$key> can be of any length. Typical lengths for the C<$key> are 16 |
|
|
539 | (128 bit) or 32 (256 bit). |
|
|
540 | |
|
|
541 | After creation, you have to call C<nonce> next. |
|
|
542 | |
|
|
543 | =item $aead->nonce ($nonce) |
|
|
544 | |
|
|
545 | Provide the nonce value (nonce means "value used once"), a value the is |
|
|
546 | unique between all uses with the same key. This method I<must> be called |
|
|
547 | I<after> C<new> and I<before> C<associated_data>. |
|
|
548 | |
|
|
549 | If you only ever use a given key once, you can provide an empty nonce - |
|
|
550 | but you still have to call the method. |
|
|
551 | |
|
|
552 | Common strategies to provide a nonce are to implement a persistent counter |
|
|
553 | or to generate a random string of sufficient length to guarantee that it |
|
|
554 | differs each time. |
|
|
555 | |
|
|
556 | The problem with counters is that you might get confused and forget |
|
|
557 | increments, and thus reuse the same sequence number. The problem with |
|
|
558 | random strings i that your random number generator might be hosed and |
|
|
559 | generate the same randomness multiple times (randomness can be very hard |
|
|
560 | to get especially on embedded devices). |
|
|
561 | |
|
|
562 | =item $aead->associated_data ($data) |
|
|
563 | |
|
|
564 | Provide the associated data (cleartext data to be authenticated but not |
|
|
565 | encrypted). This method I<must> be called I<after> C<nonce> and I<before> |
|
|
566 | C<crypt>. |
|
|
567 | |
|
|
568 | If you don't have any associated data, you can provide an empty string - |
|
|
569 | but you still have to call the method. |
|
|
570 | |
|
|
571 | Associated data is typically header data - data anybody is allowed to |
|
|
572 | see in cleartext, but that should nevertheless be protected with an |
|
|
573 | authentication code. Typically such data is used to identify where to |
|
|
574 | forward a message to, how to find the key to decrypt the message or in |
|
|
575 | general how to interpret the encrypted part of a message. |
|
|
576 | |
|
|
577 | =item $encrypted = $cipher->crypt ($cleartext) |
|
|
578 | |
|
|
579 | =item $cleartext = $cipher->crypt ($encrypted) |
|
|
580 | |
|
|
581 | Encrypt or decrypt a piece of a message. This can be called as many times |
|
|
582 | as you want, and the message can be split into as few or many pieces as |
|
|
583 | required without affecting the results, with one exception: All except the |
|
|
584 | last call to C<crypt> needs to pass in a multiple of C<64> octets. The |
|
|
585 | last call to C<crypt> does not have this limitation. |
|
|
586 | |
|
|
587 | =item $cipher->crypt_inplace ($cleartext_or_ciphertext) |
|
|
588 | |
|
|
589 | Same as C<crypt>, except it I<modifies the argument in-place>. |
|
|
590 | |
|
|
591 | =item $another_cipher = $cipher->clone |
|
|
592 | |
|
|
593 | Make an exact copy of the cipher state. This can be useful to cache states |
|
|
594 | for reuse later, for example, to avoid expensive key setups. |
|
|
595 | |
|
|
596 | Example: set up a cipher state with a key, then clone and use it to |
|
|
597 | encrypt messages with different nonces. |
|
|
598 | |
|
|
599 | my $cipher = new Crypt::Spritz::AEAD::XOR $key; |
|
|
600 | |
|
|
601 | my $message_counter; |
|
|
602 | |
|
|
603 | for my $message ("a", "b", "c") { |
|
|
604 | my $clone = $cipher->clone; |
|
|
605 | $clone->nonce (pack "N", ++$message_counter); |
|
|
606 | $clone->associated_data (""); |
|
|
607 | my $encrypted = $clone->crypt ($message); |
|
|
608 | ... |
|
|
609 | } |
|
|
610 | |
|
|
611 | =back |
|
|
612 | |
|
|
613 | |
|
|
614 | =head2 THE Crypt::Spritz::AEAD CLASS |
|
|
615 | |
|
|
616 | This class is pretty much the same as the C<Crypt::Spritz::AEAD::XOR> |
|
|
617 | class, with two differences: first, it implements the "standard" Spritz |
|
|
618 | encryption algorithm, and second, while this variant is easier to analyze |
|
|
619 | mathematically, there is little else to recommend it for, as it is slower, |
|
|
620 | and requires lots of code duplication code. |
|
|
621 | |
|
|
622 | So unless you need to be compatible with another implementation that does |
|
|
623 | not offer the XOR variant, stick to C<Crypt::Spritz::AEAD::XOR>. |
|
|
624 | |
|
|
625 | All the methods from C<Crypt::Spritz::AEAD::XOR> are available, except |
|
|
626 | C<crypt>, which has been replaced by separate C<encrypt> and C<decrypt> |
|
|
627 | methods: |
|
|
628 | |
|
|
629 | =over 4 |
|
|
630 | |
|
|
631 | =item $encrypted = $cipher->encrypt ($cleartext) |
|
|
632 | |
|
|
633 | =item $cleartext = $cipher->decrypt ($encrypted) |
|
|
634 | |
|
|
635 | Really the same as C<Crypt::Spritz::AEAD::XOR>, except you need separate |
|
|
636 | calls and code for encryption and decryption, but you have the same |
|
|
637 | limitations on usage. |
|
|
638 | |
|
|
639 | =back |
|
|
640 | |
|
|
641 | |
|
|
642 | =head1 SECURITY CONSIDERATIONS |
|
|
643 | |
|
|
644 | At the time of this writing, Spritz has not been through a lot of |
|
|
645 | cryptanalysis - it might get broken tomorrow. That's true for any crypto |
|
|
646 | algo, but the probability is quite a bit higher with Spritz. Having said |
|
|
647 | that, Spritz is almost certainly safer than RC4 at this time. |
|
|
648 | |
|
|
649 | Nevertheless, I wouldn't protect something very expensive with it. I also |
|
|
650 | would be careful about timing attacks. |
|
|
651 | |
|
|
652 | Regarding key lengths - as has been pointed out, traditional symmetric key |
|
|
653 | lengths (128 bit, 256 bit) work fine. Longer keys will be overkill, but |
|
|
654 | you can expect keys up to about a kilobit to be effective. Longer keys are |
|
|
655 | safe to use, they will simply be a waste of time. |
|
|
656 | |
|
|
657 | |
|
|
658 | =head1 PERFORMANCE |
|
|
659 | |
|
|
660 | As a cipher/prng, Spritz is reasonably fast (about 100MB/s on 2014 era |
|
|
661 | hardware, for comparison, AES will be more like 200MB/s). |
|
|
662 | |
|
|
663 | For key setup, ivs, hashing, nonces and so on, Spritz is very slow (about |
|
|
664 | 5MB/s on 2014 era hardware, which does SHA-256 at about 200MB/s). |
|
|
665 | |
111 | |
666 | |
112 | =head1 SEE ALSO |
667 | =head1 SEE ALSO |
113 | |
668 | |
114 | L<Crypt::CBC>, L<Digest::HMAC>, L<http://people.csail.mit.edu/rivest/pubs/RS14.pdf>. |
669 | L<http://people.csail.mit.edu/rivest/pubs/RS14.pdf>. |
115 | |
670 | |
116 | =head1 SECURITY CONSIDERATIONS |
671 | =head1 SECURITY CONSIDERATIONS |
117 | |
672 | |
118 | I also cannot guarantee for security. |
673 | I also cannot give any guarantees for security, Spritz is a very new |
|
|
674 | cryptographic algorithm, and when this module was written, almost |
|
|
675 | completely unproven. |
119 | |
676 | |
120 | =head1 AUTHOR |
677 | =head1 AUTHOR |
121 | |
678 | |
122 | Marc Lehmann <schmorp@schmorp.de> |
679 | Marc Lehmann <schmorp@schmorp.de> |
123 | http://home.schmorp.de/ |
680 | http://software.schmorp.de/pkg/Crypt-Spritz |
124 | |
|
|
125 | The actual twofish encryption is written in horribly microsoft'ish looking |
|
|
126 | almost ansi-c by Doug Whiting. |
|
|
127 | |
681 | |
128 | =cut |
682 | =cut |
129 | |
683 | |
130 | 1; |
684 | 1; |
131 | |
685 | |