Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(6)

Issues Search
    My Issues | Starred     Open | Closed | All

Side by Side Diff: third_party/boringssl/linux-aarch64/crypto/modes/ghashv8-armx64.S

Issue 2219933002: Land BoringSSL roll on master (Closed) Base URL: git@github.com:dart-lang/sdk.git@master
Patch Set: Created 4 years, 4 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch
« no previous file with comments | « third_party/boringssl/linux-aarch64/crypto/chacha/chacha-armv8.S ('k') | third_party/boringssl/linux-aarch64/crypto/sha/sha1-armv8.S » ('j') | no next file with comments »
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
OLDNEW
1 #if defined(__aarch64__) 1 #if defined(__aarch64__)
2 #include "arm_arch.h" 2 #include <openssl/arm_arch.h>
3 3
4 .text 4 .text
5 #if !defined(__clang__) 5 #if !defined(__clang__)
6 .arch armv8-a+crypto 6 .arch armv8-a+crypto
7 #endif 7 #endif
8 .globl gcm_init_v8 8 .globl gcm_init_v8
9 .hidden gcm_init_v8
9 .type gcm_init_v8,%function 10 .type gcm_init_v8,%function
10 .align 4 11 .align 4
11 gcm_init_v8: 12 gcm_init_v8:
12 ld1 {v17.2d},[x1] //load input H 13 ld1 {v17.2d},[x1] //load input H
13 movi v19.16b,#0xe1 14 movi v19.16b,#0xe1
14 shl v19.2d,v19.2d,#57 //0xc2.0 15 shl v19.2d,v19.2d,#57 //0xc2.0
15 ext v3.16b,v17.16b,v17.16b,#8 16 ext v3.16b,v17.16b,v17.16b,#8
16 ushr v18.2d,v19.2d,#63 17 ushr v18.2d,v19.2d,#63
17 dup v17.4s,v17.s[1] 18 dup v17.4s,v17.s[1]
18 ext v16.16b,v18.16b,v19.16b,#8 //t0=0xc2....01 19 ext v16.16b,v18.16b,v19.16b,#8 //t0=0xc2....01
(...skipping 30 matching lines...) Expand all
49 eor v22.16b,v0.16b,v18.16b 50 eor v22.16b,v0.16b,v18.16b
50 51
51 ext v17.16b,v22.16b,v22.16b,#8 //Karatsuba pre-processi ng 52 ext v17.16b,v22.16b,v22.16b,#8 //Karatsuba pre-processi ng
52 eor v17.16b,v17.16b,v22.16b 53 eor v17.16b,v17.16b,v22.16b
53 ext v21.16b,v16.16b,v17.16b,#8 //pack Karatsuba pre-pro cessed 54 ext v21.16b,v16.16b,v17.16b,#8 //pack Karatsuba pre-pro cessed
54 st1 {v21.2d,v22.2d},[x0] //store Htable[1..2] 55 st1 {v21.2d,v22.2d},[x0] //store Htable[1..2]
55 56
56 ret 57 ret
57 .size gcm_init_v8,.-gcm_init_v8 58 .size gcm_init_v8,.-gcm_init_v8
58 .globl gcm_gmult_v8 59 .globl gcm_gmult_v8
60 .hidden gcm_gmult_v8
59 .type gcm_gmult_v8,%function 61 .type gcm_gmult_v8,%function
60 .align 4 62 .align 4
61 gcm_gmult_v8: 63 gcm_gmult_v8:
62 ld1 {v17.2d},[x0] //load Xi 64 ld1 {v17.2d},[x0] //load Xi
63 movi v19.16b,#0xe1 65 movi v19.16b,#0xe1
64 ld1 {v20.2d,v21.2d},[x1] //load twisted H, ... 66 ld1 {v20.2d,v21.2d},[x1] //load twisted H, ...
65 shl v19.2d,v19.2d,#57 67 shl v19.2d,v19.2d,#57
66 #ifndef __ARMEB__ 68 #ifndef __ARMEB__
67 rev64 v17.16b,v17.16b 69 rev64 v17.16b,v17.16b
68 #endif 70 #endif
69 ext v3.16b,v17.16b,v17.16b,#8 71 ext v3.16b,v17.16b,v17.16b,#8
70 72
71 » pmull» v0.1q,v20.1d,v3.1d» » //H.lo·Xi.lo 73 » pmull» v0.1q,v20.1d,v3.1d» » //H.lo·Xi.lo
72 eor v17.16b,v17.16b,v3.16b //Karatsuba pre-processing 74 eor v17.16b,v17.16b,v3.16b //Karatsuba pre-processing
73 » pmull2» v2.1q,v20.2d,v3.2d» » //H.hi·Xi.hi 75 » pmull2» v2.1q,v20.2d,v3.2d» » //H.hi·Xi.hi
74 » pmull» v1.1q,v21.1d,v17.1d» » //(H.lo+H.hi)·(Xi.lo+Xi.hi) 76 » pmull» v1.1q,v21.1d,v17.1d» » //(H.lo+H.hi)·(Xi.lo+Xi.hi)
75 77
76 ext v17.16b,v0.16b,v2.16b,#8 //Karatsuba post-process ing 78 ext v17.16b,v0.16b,v2.16b,#8 //Karatsuba post-process ing
77 eor v18.16b,v0.16b,v2.16b 79 eor v18.16b,v0.16b,v2.16b
78 eor v1.16b,v1.16b,v17.16b 80 eor v1.16b,v1.16b,v17.16b
79 eor v1.16b,v1.16b,v18.16b 81 eor v1.16b,v1.16b,v18.16b
80 pmull v18.1q,v0.1d,v19.1d //1st phase of reduction 82 pmull v18.1q,v0.1d,v19.1d //1st phase of reduction
81 83
82 ins v2.d[0],v1.d[1] 84 ins v2.d[0],v1.d[1]
83 ins v1.d[1],v0.d[0] 85 ins v1.d[1],v0.d[0]
84 eor v0.16b,v1.16b,v18.16b 86 eor v0.16b,v1.16b,v18.16b
85 87
86 ext v18.16b,v0.16b,v0.16b,#8 //2nd phase of reduction 88 ext v18.16b,v0.16b,v0.16b,#8 //2nd phase of reduction
87 pmull v0.1q,v0.1d,v19.1d 89 pmull v0.1q,v0.1d,v19.1d
88 eor v18.16b,v18.16b,v2.16b 90 eor v18.16b,v18.16b,v2.16b
89 eor v0.16b,v0.16b,v18.16b 91 eor v0.16b,v0.16b,v18.16b
90 92
91 #ifndef __ARMEB__ 93 #ifndef __ARMEB__
92 rev64 v0.16b,v0.16b 94 rev64 v0.16b,v0.16b
93 #endif 95 #endif
94 ext v0.16b,v0.16b,v0.16b,#8 96 ext v0.16b,v0.16b,v0.16b,#8
95 st1 {v0.2d},[x0] //write out Xi 97 st1 {v0.2d},[x0] //write out Xi
96 98
97 ret 99 ret
98 .size gcm_gmult_v8,.-gcm_gmult_v8 100 .size gcm_gmult_v8,.-gcm_gmult_v8
99 .globl gcm_ghash_v8 101 .globl gcm_ghash_v8
102 .hidden gcm_ghash_v8
100 .type gcm_ghash_v8,%function 103 .type gcm_ghash_v8,%function
101 .align 4 104 .align 4
102 gcm_ghash_v8: 105 gcm_ghash_v8:
103 ld1 {v0.2d},[x0] //load [rotated] Xi 106 ld1 {v0.2d},[x0] //load [rotated] Xi
104 //"[rotated]" means that 107 //"[rotated]" means that
105 //loaded value would have 108 //loaded value would have
106 //to be rotated in order to 109 //to be rotated in order to
107 //make it appear as in 110 //make it appear as in
108 //alorithm specification 111 //alorithm specification
109 subs x3,x3,#32 //see if x3 is 32 or larger 112 subs x3,x3,#32 //see if x3 is 32 or larger
(...skipping 18 matching lines...) Expand all
128 rev64 v0.16b,v0.16b 131 rev64 v0.16b,v0.16b
129 #endif 132 #endif
130 ext v3.16b,v16.16b,v16.16b,#8 //rotate I[0] 133 ext v3.16b,v16.16b,v16.16b,#8 //rotate I[0]
131 b.lo .Lodd_tail_v8 //x3 was less than 32 134 b.lo .Lodd_tail_v8 //x3 was less than 32
132 ld1 {v17.2d},[x2],x12 //load [rotated] I[1] 135 ld1 {v17.2d},[x2],x12 //load [rotated] I[1]
133 #ifndef __ARMEB__ 136 #ifndef __ARMEB__
134 rev64 v17.16b,v17.16b 137 rev64 v17.16b,v17.16b
135 #endif 138 #endif
136 ext v7.16b,v17.16b,v17.16b,#8 139 ext v7.16b,v17.16b,v17.16b,#8
137 eor v3.16b,v3.16b,v0.16b //I[i]^=Xi 140 eor v3.16b,v3.16b,v0.16b //I[i]^=Xi
138 » pmull» v4.1q,v20.1d,v7.1d» » //H·Ii+1 141 » pmull» v4.1q,v20.1d,v7.1d» » //H·Ii+1
139 eor v17.16b,v17.16b,v7.16b //Karatsuba pre-processing 142 eor v17.16b,v17.16b,v7.16b //Karatsuba pre-processing
140 pmull2 v6.1q,v20.2d,v7.2d 143 pmull2 v6.1q,v20.2d,v7.2d
141 b .Loop_mod2x_v8 144 b .Loop_mod2x_v8
142 145
143 .align 4 146 .align 4
144 .Loop_mod2x_v8: 147 .Loop_mod2x_v8:
145 ext v18.16b,v3.16b,v3.16b,#8 148 ext v18.16b,v3.16b,v3.16b,#8
146 subs x3,x3,#32 //is there more data? 149 subs x3,x3,#32 //is there more data?
147 » pmull» v0.1q,v22.1d,v3.1d» » //H^2.lo·Xi.lo 150 » pmull» v0.1q,v22.1d,v3.1d» » //H^2.lo·Xi.lo
148 csel x12,xzr,x12,lo //is it time to zero x12? 151 csel x12,xzr,x12,lo //is it time to zero x12?
149 152
150 pmull v5.1q,v21.1d,v17.1d 153 pmull v5.1q,v21.1d,v17.1d
151 eor v18.16b,v18.16b,v3.16b //Karatsuba pre-processing 154 eor v18.16b,v18.16b,v3.16b //Karatsuba pre-processing
152 » pmull2» v2.1q,v22.2d,v3.2d» » //H^2.hi·Xi.hi 155 » pmull2» v2.1q,v22.2d,v3.2d» » //H^2.hi·Xi.hi
153 eor v0.16b,v0.16b,v4.16b //accumulate 156 eor v0.16b,v0.16b,v4.16b //accumulate
154 » pmull2» v1.1q,v21.2d,v18.2d» » //(H^2.lo+H^2.hi)·(Xi.lo+Xi.hi) 157 » pmull2» v1.1q,v21.2d,v18.2d» » //(H^2.lo+H^2.hi)·(Xi.lo+Xi.hi)
155 ld1 {v16.2d},[x2],x12 //load [rotated] I[i+2] 158 ld1 {v16.2d},[x2],x12 //load [rotated] I[i+2]
156 159
157 eor v2.16b,v2.16b,v6.16b 160 eor v2.16b,v2.16b,v6.16b
158 csel x12,xzr,x12,eq //is it time to zero x12? 161 csel x12,xzr,x12,eq //is it time to zero x12?
159 eor v1.16b,v1.16b,v5.16b 162 eor v1.16b,v1.16b,v5.16b
160 163
161 ext v17.16b,v0.16b,v2.16b,#8 //Karatsuba post-process ing 164 ext v17.16b,v0.16b,v2.16b,#8 //Karatsuba post-process ing
162 eor v18.16b,v0.16b,v2.16b 165 eor v18.16b,v0.16b,v2.16b
163 eor v1.16b,v1.16b,v17.16b 166 eor v1.16b,v1.16b,v17.16b
164 ld1 {v17.2d},[x2],x12 //load [rotated] I[i+3] 167 ld1 {v17.2d},[x2],x12 //load [rotated] I[i+3]
165 #ifndef __ARMEB__ 168 #ifndef __ARMEB__
166 rev64 v16.16b,v16.16b 169 rev64 v16.16b,v16.16b
167 #endif 170 #endif
168 eor v1.16b,v1.16b,v18.16b 171 eor v1.16b,v1.16b,v18.16b
169 pmull v18.1q,v0.1d,v19.1d //1st phase of reduction 172 pmull v18.1q,v0.1d,v19.1d //1st phase of reduction
170 173
171 #ifndef __ARMEB__ 174 #ifndef __ARMEB__
172 rev64 v17.16b,v17.16b 175 rev64 v17.16b,v17.16b
173 #endif 176 #endif
174 ins v2.d[0],v1.d[1] 177 ins v2.d[0],v1.d[1]
175 ins v1.d[1],v0.d[0] 178 ins v1.d[1],v0.d[0]
176 ext v7.16b,v17.16b,v17.16b,#8 179 ext v7.16b,v17.16b,v17.16b,#8
177 ext v3.16b,v16.16b,v16.16b,#8 180 ext v3.16b,v16.16b,v16.16b,#8
178 eor v0.16b,v1.16b,v18.16b 181 eor v0.16b,v1.16b,v18.16b
179 » pmull» v4.1q,v20.1d,v7.1d» » //H·Ii+1 182 » pmull» v4.1q,v20.1d,v7.1d» » //H·Ii+1
180 eor v3.16b,v3.16b,v2.16b //accumulate v3.16b early 183 eor v3.16b,v3.16b,v2.16b //accumulate v3.16b early
181 184
182 ext v18.16b,v0.16b,v0.16b,#8 //2nd phase of reduction 185 ext v18.16b,v0.16b,v0.16b,#8 //2nd phase of reduction
183 pmull v0.1q,v0.1d,v19.1d 186 pmull v0.1q,v0.1d,v19.1d
184 eor v3.16b,v3.16b,v18.16b 187 eor v3.16b,v3.16b,v18.16b
185 eor v17.16b,v17.16b,v7.16b //Karatsuba pre-processing 188 eor v17.16b,v17.16b,v7.16b //Karatsuba pre-processing
186 eor v3.16b,v3.16b,v0.16b 189 eor v3.16b,v3.16b,v0.16b
187 pmull2 v6.1q,v20.2d,v7.2d 190 pmull2 v6.1q,v20.2d,v7.2d
188 b.hs .Loop_mod2x_v8 //there was at least 32 more bytes 191 b.hs .Loop_mod2x_v8 //there was at least 32 more bytes
189 192
190 eor v2.16b,v2.16b,v18.16b 193 eor v2.16b,v2.16b,v18.16b
191 ext v3.16b,v16.16b,v16.16b,#8 //re-construct v3.16b 194 ext v3.16b,v16.16b,v16.16b,#8 //re-construct v3.16b
192 adds x3,x3,#32 //re-construct x3 195 adds x3,x3,#32 //re-construct x3
193 eor v0.16b,v0.16b,v2.16b //re-construct v0.16b 196 eor v0.16b,v0.16b,v2.16b //re-construct v0.16b
194 b.eq .Ldone_v8 //is x3 zero? 197 b.eq .Ldone_v8 //is x3 zero?
195 .Lodd_tail_v8: 198 .Lodd_tail_v8:
196 ext v18.16b,v0.16b,v0.16b,#8 199 ext v18.16b,v0.16b,v0.16b,#8
197 eor v3.16b,v3.16b,v0.16b //inp^=Xi 200 eor v3.16b,v3.16b,v0.16b //inp^=Xi
198 eor v17.16b,v16.16b,v18.16b //v17.16b is rotated inp^Xi 201 eor v17.16b,v16.16b,v18.16b //v17.16b is rotated inp^Xi
199 202
200 » pmull» v0.1q,v20.1d,v3.1d» » //H.lo·Xi.lo 203 » pmull» v0.1q,v20.1d,v3.1d» » //H.lo·Xi.lo
201 eor v17.16b,v17.16b,v3.16b //Karatsuba pre-processing 204 eor v17.16b,v17.16b,v3.16b //Karatsuba pre-processing
202 » pmull2» v2.1q,v20.2d,v3.2d» » //H.hi·Xi.hi 205 » pmull2» v2.1q,v20.2d,v3.2d» » //H.hi·Xi.hi
203 » pmull» v1.1q,v21.1d,v17.1d» » //(H.lo+H.hi)·(Xi.lo+Xi.hi) 206 » pmull» v1.1q,v21.1d,v17.1d» » //(H.lo+H.hi)·(Xi.lo+Xi.hi)
204 207
205 ext v17.16b,v0.16b,v2.16b,#8 //Karatsuba post-process ing 208 ext v17.16b,v0.16b,v2.16b,#8 //Karatsuba post-process ing
206 eor v18.16b,v0.16b,v2.16b 209 eor v18.16b,v0.16b,v2.16b
207 eor v1.16b,v1.16b,v17.16b 210 eor v1.16b,v1.16b,v17.16b
208 eor v1.16b,v1.16b,v18.16b 211 eor v1.16b,v1.16b,v18.16b
209 pmull v18.1q,v0.1d,v19.1d //1st phase of reduction 212 pmull v18.1q,v0.1d,v19.1d //1st phase of reduction
210 213
211 ins v2.d[0],v1.d[1] 214 ins v2.d[0],v1.d[1]
212 ins v1.d[1],v0.d[0] 215 ins v1.d[1],v0.d[0]
213 eor v0.16b,v1.16b,v18.16b 216 eor v0.16b,v1.16b,v18.16b
214 217
215 ext v18.16b,v0.16b,v0.16b,#8 //2nd phase of reduction 218 ext v18.16b,v0.16b,v0.16b,#8 //2nd phase of reduction
216 pmull v0.1q,v0.1d,v19.1d 219 pmull v0.1q,v0.1d,v19.1d
217 eor v18.16b,v18.16b,v2.16b 220 eor v18.16b,v18.16b,v2.16b
218 eor v0.16b,v0.16b,v18.16b 221 eor v0.16b,v0.16b,v18.16b
219 222
220 .Ldone_v8: 223 .Ldone_v8:
221 #ifndef __ARMEB__ 224 #ifndef __ARMEB__
222 rev64 v0.16b,v0.16b 225 rev64 v0.16b,v0.16b
223 #endif 226 #endif
224 ext v0.16b,v0.16b,v0.16b,#8 227 ext v0.16b,v0.16b,v0.16b,#8
225 st1 {v0.2d},[x0] //write out Xi 228 st1 {v0.2d},[x0] //write out Xi
226 229
227 ret 230 ret
228 .size gcm_ghash_v8,.-gcm_ghash_v8 231 .size gcm_ghash_v8,.-gcm_ghash_v8
229 .byte 71,72,65,83,72,32,102,111,114,32,65,82,77,118,56,44,32,67,82,89,80,84,79 ,71,65,77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115,115,108,4 6,111,114,103,62,0 232 .byte 71,72,65,83,72,32,102,111,114,32,65,82,77,118,56,44,32,67,82,89,80,84,79 ,71,65,77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115,115,108,4 6,111,114,103,62,0
230 .align 2 233 .align 2
231 .align 2 234 .align 2
232 #endif 235 #endif
OLDNEW
« no previous file with comments | « third_party/boringssl/linux-aarch64/crypto/chacha/chacha-armv8.S ('k') | third_party/boringssl/linux-aarch64/crypto/sha/sha1-armv8.S » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698