LLVM  14.0.0git
SHA1.cpp
Go to the documentation of this file.
1 //====- SHA1.cpp - Private copy of the SHA1 implementation ---*- C++ -* ======//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 // This code is taken from public domain
10 // (http://oauth.googlecode.com/svn/code/c/liboauth/src/sha1.c and
11 // http://cvsweb.netbsd.org/bsdweb.cgi/src/common/lib/libc/hash/sha1/sha1.c?rev=1.6)
12 // and modified by wrapping it in a C++ interface for LLVM,
13 // and removing unnecessary code.
14 //
15 //===----------------------------------------------------------------------===//
16 
17 #include "llvm/Support/SHA1.h"
18 #include "llvm/ADT/ArrayRef.h"
19 #include "llvm/ADT/StringRef.h"
20 #include "llvm/Support/Endian.h"
21 #include "llvm/Support/Host.h"
22 #include <string.h>
23 
24 using namespace llvm;
25 
26 #if defined(BYTE_ORDER) && defined(BIG_ENDIAN) && BYTE_ORDER == BIG_ENDIAN
27 #define SHA_BIG_ENDIAN
28 #endif
29 
30 static inline uint32_t rol(uint32_t Number, int Bits) {
31  return (Number << Bits) | (Number >> (32 - Bits));
32 }
33 
34 static inline uint32_t blk0(uint32_t *Buf, int I) { return Buf[I]; }
35 
36 static inline uint32_t blk(uint32_t *Buf, int I) {
37  Buf[I & 15] = rol(Buf[(I + 13) & 15] ^ Buf[(I + 8) & 15] ^ Buf[(I + 2) & 15] ^
38  Buf[I & 15],
39  1);
40  return Buf[I & 15];
41 }
42 
43 static inline void r0(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D,
44  uint32_t &E, int I, uint32_t *Buf) {
45  E += ((B & (C ^ D)) ^ D) + blk0(Buf, I) + 0x5A827999 + rol(A, 5);
46  B = rol(B, 30);
47 }
48 
49 static inline void r1(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D,
50  uint32_t &E, int I, uint32_t *Buf) {
51  E += ((B & (C ^ D)) ^ D) + blk(Buf, I) + 0x5A827999 + rol(A, 5);
52  B = rol(B, 30);
53 }
54 
55 static inline void r2(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D,
56  uint32_t &E, int I, uint32_t *Buf) {
57  E += (B ^ C ^ D) + blk(Buf, I) + 0x6ED9EBA1 + rol(A, 5);
58  B = rol(B, 30);
59 }
60 
61 static inline void r3(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D,
62  uint32_t &E, int I, uint32_t *Buf) {
63  E += (((B | C) & D) | (B & C)) + blk(Buf, I) + 0x8F1BBCDC + rol(A, 5);
64  B = rol(B, 30);
65 }
66 
67 static inline void r4(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D,
68  uint32_t &E, int I, uint32_t *Buf) {
69  E += (B ^ C ^ D) + blk(Buf, I) + 0xCA62C1D6 + rol(A, 5);
70  B = rol(B, 30);
71 }
72 
73 /* code */
74 #define SHA1_K0 0x5a827999
75 #define SHA1_K20 0x6ed9eba1
76 #define SHA1_K40 0x8f1bbcdc
77 #define SHA1_K60 0xca62c1d6
78 
79 #define SEED_0 0x67452301
80 #define SEED_1 0xefcdab89
81 #define SEED_2 0x98badcfe
82 #define SEED_3 0x10325476
83 #define SEED_4 0xc3d2e1f0
84 
85 void SHA1::init() {
86  InternalState.State[0] = SEED_0;
87  InternalState.State[1] = SEED_1;
88  InternalState.State[2] = SEED_2;
89  InternalState.State[3] = SEED_3;
90  InternalState.State[4] = SEED_4;
91  InternalState.ByteCount = 0;
92  InternalState.BufferOffset = 0;
93 }
94 
95 void SHA1::hashBlock() {
96  uint32_t A = InternalState.State[0];
97  uint32_t B = InternalState.State[1];
98  uint32_t C = InternalState.State[2];
99  uint32_t D = InternalState.State[3];
100  uint32_t E = InternalState.State[4];
101 
102  // 4 rounds of 20 operations each. Loop unrolled.
103  r0(A, B, C, D, E, 0, InternalState.Buffer.L);
104  r0(E, A, B, C, D, 1, InternalState.Buffer.L);
105  r0(D, E, A, B, C, 2, InternalState.Buffer.L);
106  r0(C, D, E, A, B, 3, InternalState.Buffer.L);
107  r0(B, C, D, E, A, 4, InternalState.Buffer.L);
108  r0(A, B, C, D, E, 5, InternalState.Buffer.L);
109  r0(E, A, B, C, D, 6, InternalState.Buffer.L);
110  r0(D, E, A, B, C, 7, InternalState.Buffer.L);
111  r0(C, D, E, A, B, 8, InternalState.Buffer.L);
112  r0(B, C, D, E, A, 9, InternalState.Buffer.L);
113  r0(A, B, C, D, E, 10, InternalState.Buffer.L);
114  r0(E, A, B, C, D, 11, InternalState.Buffer.L);
115  r0(D, E, A, B, C, 12, InternalState.Buffer.L);
116  r0(C, D, E, A, B, 13, InternalState.Buffer.L);
117  r0(B, C, D, E, A, 14, InternalState.Buffer.L);
118  r0(A, B, C, D, E, 15, InternalState.Buffer.L);
119  r1(E, A, B, C, D, 16, InternalState.Buffer.L);
120  r1(D, E, A, B, C, 17, InternalState.Buffer.L);
121  r1(C, D, E, A, B, 18, InternalState.Buffer.L);
122  r1(B, C, D, E, A, 19, InternalState.Buffer.L);
123 
124  r2(A, B, C, D, E, 20, InternalState.Buffer.L);
125  r2(E, A, B, C, D, 21, InternalState.Buffer.L);
126  r2(D, E, A, B, C, 22, InternalState.Buffer.L);
127  r2(C, D, E, A, B, 23, InternalState.Buffer.L);
128  r2(B, C, D, E, A, 24, InternalState.Buffer.L);
129  r2(A, B, C, D, E, 25, InternalState.Buffer.L);
130  r2(E, A, B, C, D, 26, InternalState.Buffer.L);
131  r2(D, E, A, B, C, 27, InternalState.Buffer.L);
132  r2(C, D, E, A, B, 28, InternalState.Buffer.L);
133  r2(B, C, D, E, A, 29, InternalState.Buffer.L);
134  r2(A, B, C, D, E, 30, InternalState.Buffer.L);
135  r2(E, A, B, C, D, 31, InternalState.Buffer.L);
136  r2(D, E, A, B, C, 32, InternalState.Buffer.L);
137  r2(C, D, E, A, B, 33, InternalState.Buffer.L);
138  r2(B, C, D, E, A, 34, InternalState.Buffer.L);
139  r2(A, B, C, D, E, 35, InternalState.Buffer.L);
140  r2(E, A, B, C, D, 36, InternalState.Buffer.L);
141  r2(D, E, A, B, C, 37, InternalState.Buffer.L);
142  r2(C, D, E, A, B, 38, InternalState.Buffer.L);
143  r2(B, C, D, E, A, 39, InternalState.Buffer.L);
144 
145  r3(A, B, C, D, E, 40, InternalState.Buffer.L);
146  r3(E, A, B, C, D, 41, InternalState.Buffer.L);
147  r3(D, E, A, B, C, 42, InternalState.Buffer.L);
148  r3(C, D, E, A, B, 43, InternalState.Buffer.L);
149  r3(B, C, D, E, A, 44, InternalState.Buffer.L);
150  r3(A, B, C, D, E, 45, InternalState.Buffer.L);
151  r3(E, A, B, C, D, 46, InternalState.Buffer.L);
152  r3(D, E, A, B, C, 47, InternalState.Buffer.L);
153  r3(C, D, E, A, B, 48, InternalState.Buffer.L);
154  r3(B, C, D, E, A, 49, InternalState.Buffer.L);
155  r3(A, B, C, D, E, 50, InternalState.Buffer.L);
156  r3(E, A, B, C, D, 51, InternalState.Buffer.L);
157  r3(D, E, A, B, C, 52, InternalState.Buffer.L);
158  r3(C, D, E, A, B, 53, InternalState.Buffer.L);
159  r3(B, C, D, E, A, 54, InternalState.Buffer.L);
160  r3(A, B, C, D, E, 55, InternalState.Buffer.L);
161  r3(E, A, B, C, D, 56, InternalState.Buffer.L);
162  r3(D, E, A, B, C, 57, InternalState.Buffer.L);
163  r3(C, D, E, A, B, 58, InternalState.Buffer.L);
164  r3(B, C, D, E, A, 59, InternalState.Buffer.L);
165 
166  r4(A, B, C, D, E, 60, InternalState.Buffer.L);
167  r4(E, A, B, C, D, 61, InternalState.Buffer.L);
168  r4(D, E, A, B, C, 62, InternalState.Buffer.L);
169  r4(C, D, E, A, B, 63, InternalState.Buffer.L);
170  r4(B, C, D, E, A, 64, InternalState.Buffer.L);
171  r4(A, B, C, D, E, 65, InternalState.Buffer.L);
172  r4(E, A, B, C, D, 66, InternalState.Buffer.L);
173  r4(D, E, A, B, C, 67, InternalState.Buffer.L);
174  r4(C, D, E, A, B, 68, InternalState.Buffer.L);
175  r4(B, C, D, E, A, 69, InternalState.Buffer.L);
176  r4(A, B, C, D, E, 70, InternalState.Buffer.L);
177  r4(E, A, B, C, D, 71, InternalState.Buffer.L);
178  r4(D, E, A, B, C, 72, InternalState.Buffer.L);
179  r4(C, D, E, A, B, 73, InternalState.Buffer.L);
180  r4(B, C, D, E, A, 74, InternalState.Buffer.L);
181  r4(A, B, C, D, E, 75, InternalState.Buffer.L);
182  r4(E, A, B, C, D, 76, InternalState.Buffer.L);
183  r4(D, E, A, B, C, 77, InternalState.Buffer.L);
184  r4(C, D, E, A, B, 78, InternalState.Buffer.L);
185  r4(B, C, D, E, A, 79, InternalState.Buffer.L);
186 
187  InternalState.State[0] += A;
188  InternalState.State[1] += B;
189  InternalState.State[2] += C;
190  InternalState.State[3] += D;
191  InternalState.State[4] += E;
192 }
193 
194 void SHA1::addUncounted(uint8_t Data) {
195 #ifdef SHA_BIG_ENDIAN
196  InternalState.Buffer.C[InternalState.BufferOffset] = Data;
197 #else
198  InternalState.Buffer.C[InternalState.BufferOffset ^ 3] = Data;
199 #endif
200 
201  InternalState.BufferOffset++;
202  if (InternalState.BufferOffset == BLOCK_LENGTH) {
203  hashBlock();
204  InternalState.BufferOffset = 0;
205  }
206 }
207 
208 void SHA1::writebyte(uint8_t Data) {
209  ++InternalState.ByteCount;
210  addUncounted(Data);
211 }
212 
214  InternalState.ByteCount += Data.size();
215 
216  // Finish the current block.
217  if (InternalState.BufferOffset > 0) {
218  const size_t Remainder = std::min<size_t>(
219  Data.size(), BLOCK_LENGTH - InternalState.BufferOffset);
220  for (size_t I = 0; I < Remainder; ++I)
221  addUncounted(Data[I]);
222  Data = Data.drop_front(Remainder);
223  }
224 
225  // Fast buffer filling for large inputs.
226  while (Data.size() >= BLOCK_LENGTH) {
227  assert(InternalState.BufferOffset == 0);
228  static_assert(BLOCK_LENGTH % 4 == 0, "");
229  constexpr size_t BLOCK_LENGTH_32 = BLOCK_LENGTH / 4;
230  for (size_t I = 0; I < BLOCK_LENGTH_32; ++I)
231  InternalState.Buffer.L[I] = support::endian::read32be(&Data[I * 4]);
232  hashBlock();
233  Data = Data.drop_front(BLOCK_LENGTH);
234  }
235 
236  // Finish the remainder.
237  for (uint8_t C : Data)
238  addUncounted(C);
239 }
240 
242  update(
243  ArrayRef<uint8_t>((uint8_t *)const_cast<char *>(Str.data()), Str.size()));
244 }
245 
246 void SHA1::pad() {
247  // Implement SHA-1 padding (fips180-2 5.1.1)
248 
249  // Pad with 0x80 followed by 0x00 until the end of the block
250  addUncounted(0x80);
251  while (InternalState.BufferOffset != 56)
252  addUncounted(0x00);
253 
254  // Append length in the last 8 bytes
255  addUncounted(0); // We're only using 32 bit lengths
256  addUncounted(0); // But SHA-1 supports 64 bit lengths
257  addUncounted(0); // So zero pad the top bits
258  addUncounted(InternalState.ByteCount >> 29); // Shifting to multiply by 8
259  addUncounted(InternalState.ByteCount >>
260  21); // as SHA-1 supports bitstreams as well as
261  addUncounted(InternalState.ByteCount >> 13); // byte.
262  addUncounted(InternalState.ByteCount >> 5);
263  addUncounted(InternalState.ByteCount << 3);
264 }
265 
267  // Pad to complete the last block
268  pad();
269 
270 #ifdef SHA_BIG_ENDIAN
271  // Just copy the current state
272  for (int i = 0; i < 5; i++) {
273  HashResult[i] = InternalState.State[i];
274  }
275 #else
276  // Swap byte order back
277  for (int i = 0; i < 5; i++) {
278  HashResult[i] = (((InternalState.State[i]) << 24) & 0xff000000) |
279  (((InternalState.State[i]) << 8) & 0x00ff0000) |
280  (((InternalState.State[i]) >> 8) & 0x0000ff00) |
281  (((InternalState.State[i]) >> 24) & 0x000000ff);
282  }
283 #endif
284 
285  // Return pointer to hash (20 characters)
286  return StringRef((char *)HashResult, HASH_LENGTH);
287 }
288 
290  auto StateToRestore = InternalState;
291 
292  auto Hash = final();
293 
294  // Restore the state
295  InternalState = StateToRestore;
296 
297  // Return pointer to hash (20 characters)
298  return Hash;
299 }
300 
301 std::array<uint8_t, 20> SHA1::hash(ArrayRef<uint8_t> Data) {
302  SHA1 Hash;
303  Hash.update(Data);
304  StringRef S = Hash.final();
305 
306  std::array<uint8_t, 20> Arr;
307  memcpy(Arr.data(), S.data(), S.size());
308  return Arr;
309 }
SEED_4
#define SEED_4
Definition: SHA1.cpp:83
i
i
Definition: README.txt:29
r1
static void r1(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D, uint32_t &E, int I, uint32_t *Buf)
Definition: SHA1.cpp:49
llvm
---------------------— PointerInfo ------------------------------------—
Definition: AllocatorList.h:23
r4
static void r4(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D, uint32_t &E, int I, uint32_t *Buf)
Definition: SHA1.cpp:67
StringRef.h
SEED_0
#define SEED_0
Definition: SHA1.cpp:79
Host.h
llvm::SHA1::final
StringRef final()
Return a reference to the current raw 160-bits SHA1 for the digested data since the last call to init...
Definition: SHA1.cpp:266
llvm::tgtok::Bits
@ Bits
Definition: TGLexer.h:50
llvm::support::endian::read32be
uint32_t read32be(const void *P)
Definition: Endian.h:384
llvm::Data
@ Data
Definition: SIMachineScheduler.h:55
llvm::SHA1
A class that wrap the SHA1 algorithm.
Definition: SHA1.h:26
E
static GCRegistry::Add< CoreCLRGC > E("coreclr", "CoreCLR-compatible GC")
SEED_3
#define SEED_3
Definition: SHA1.cpp:82
llvm::SHA1::C
uint8_t C[BLOCK_LENGTH]
Definition: SHA1.h:63
C
(vector float) vec_cmpeq(*A, *B) C
Definition: README_ALTIVEC.txt:86
B
static GCRegistry::Add< OcamlGC > B("ocaml", "ocaml 3.10-compatible GC")
r0
static void r0(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D, uint32_t &E, int I, uint32_t *Buf)
Definition: SHA1.cpp:43
llvm::SHA1::hash
static std::array< uint8_t, 20 > hash(ArrayRef< uint8_t > Data)
Returns a raw 160-bit SHA1 hash for the given data.
Definition: SHA1.cpp:301
SEED_1
#define SEED_1
Definition: SHA1.cpp:80
SEED_2
#define SEED_2
Definition: SHA1.cpp:81
r3
static void r3(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D, uint32_t &E, int I, uint32_t *Buf)
Definition: SHA1.cpp:61
blk
static uint32_t blk(uint32_t *Buf, int I)
Definition: SHA1.cpp:36
D
static GCRegistry::Add< StatepointGC > D("statepoint-example", "an example strategy for statepoint")
I
#define I(x, y, z)
Definition: MD5.cpp:59
ArrayRef.h
r2
static void r2(uint32_t &A, uint32_t &B, uint32_t &C, uint32_t &D, uint32_t &E, int I, uint32_t *Buf)
Definition: SHA1.cpp:55
assert
assert(ImpDefSCC.getReg()==AMDGPU::SCC &&ImpDefSCC.isDef())
llvm::SHA1::init
void init()
Reinitialize the internal state.
Definition: SHA1.cpp:85
blk0
static uint32_t blk0(uint32_t *Buf, int I)
Definition: SHA1.cpp:34
memcpy
<%struct.s * > cast struct s *S to sbyte *< sbyte * > sbyte uint cast struct s *agg result to sbyte *< sbyte * > sbyte uint cast struct s *memtmp to sbyte *< sbyte * > sbyte uint ret void llc ends up issuing two memcpy or custom lower memcpy(of small size) to be ldmia/stmia. I think option 2 is better but the current register allocator cannot allocate a chunk of registers at a time. A feasible temporary solution is to use specific physical registers at the lowering time for small(<
llvm::SHA1::update
void update(ArrayRef< uint8_t > Data)
Digest more data.
Definition: SHA1.cpp:213
llvm::ArrayRef< uint8_t >
llvm::StringRef
StringRef - Represent a constant reference to a string, i.e.
Definition: StringRef.h:58
Number
uint32_t Number
Definition: Profile.cpp:47
uint32_t
S
add sub stmia L5 ldr r0 bl L_printf $stub Instead of a and a wouldn t it be better to do three moves *Return an aggregate type is even return S
Definition: README.txt:210
llvm::SHA1::result
StringRef result()
Return a reference to the current raw 160-bits SHA1 for the digested data since the last call to init...
Definition: SHA1.cpp:289
SHA1.h
Endian.h
rol
static uint32_t rol(uint32_t Number, int Bits)
Definition: SHA1.cpp:30