aes-armv4.pl 33 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248
  1. #! /usr/bin/env perl
  2. # Copyright 2007-2020 The OpenSSL Project Authors. All Rights Reserved.
  3. #
  4. # Licensed under the Apache License 2.0 (the "License"). You may not use
  5. # this file except in compliance with the License. You can obtain a copy
  6. # in the file LICENSE in the source distribution or at
  7. # https://www.openssl.org/source/license.html
  8. # ====================================================================
  9. # Written by Andy Polyakov <appro@openssl.org> for the OpenSSL
  10. # project. The module is, however, dual licensed under OpenSSL and
  11. # CRYPTOGAMS licenses depending on where you obtain it. For further
  12. # details see http://www.openssl.org/~appro/cryptogams/.
  13. # ====================================================================
  14. # AES for ARMv4
  15. # January 2007.
  16. #
  17. # Code uses single 1K S-box and is >2 times faster than code generated
  18. # by gcc-3.4.1. This is thanks to unique feature of ARMv4 ISA, which
  19. # allows to merge logical or arithmetic operation with shift or rotate
  20. # in one instruction and emit combined result every cycle. The module
  21. # is endian-neutral. The performance is ~42 cycles/byte for 128-bit
  22. # key [on single-issue Xscale PXA250 core].
  23. # May 2007.
  24. #
  25. # AES_set_[en|de]crypt_key is added.
  26. # July 2010.
  27. #
  28. # Rescheduling for dual-issue pipeline resulted in 12% improvement on
  29. # Cortex A8 core and ~25 cycles per byte processed with 128-bit key.
  30. # February 2011.
  31. #
  32. # Profiler-assisted and platform-specific optimization resulted in 16%
  33. # improvement on Cortex A8 core and ~21.5 cycles per byte.
  34. # $output is the last argument if it looks like a file (it has an extension)
  35. # $flavour is the first argument if it doesn't look like a file
  36. $output = $#ARGV >= 0 && $ARGV[$#ARGV] =~ m|\.\w+$| ? pop : undef;
  37. $flavour = $#ARGV >= 0 && $ARGV[0] !~ m|\.| ? shift : undef;
  38. if ($flavour && $flavour ne "void") {
  39. $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
  40. ( $xlate="${dir}arm-xlate.pl" and -f $xlate ) or
  41. ( $xlate="${dir}../../perlasm/arm-xlate.pl" and -f $xlate) or
  42. die "can't locate arm-xlate.pl";
  43. open STDOUT,"| \"$^X\" $xlate $flavour \"$output\""
  44. or die "can't call $xlate: $!";
  45. } else {
  46. $output and open STDOUT,">$output";
  47. }
  48. $s0="r0";
  49. $s1="r1";
  50. $s2="r2";
  51. $s3="r3";
  52. $t1="r4";
  53. $t2="r5";
  54. $t3="r6";
  55. $i1="r7";
  56. $i2="r8";
  57. $i3="r9";
  58. $tbl="r10";
  59. $key="r11";
  60. $rounds="r12";
  61. $code=<<___;
  62. #ifndef __KERNEL__
  63. # include "arm_arch.h"
  64. #else
  65. # define __ARM_ARCH__ __LINUX_ARM_ARCH__
  66. #endif
  67. #if defined(__thumb2__) && !defined(__APPLE__)
  68. .syntax unified
  69. .thumb
  70. #else
  71. .code 32
  72. #undef __thumb2__
  73. #endif
  74. .text
  75. .type AES_Te,%object
  76. .align 5
  77. AES_Te:
  78. .word 0xc66363a5, 0xf87c7c84, 0xee777799, 0xf67b7b8d
  79. .word 0xfff2f20d, 0xd66b6bbd, 0xde6f6fb1, 0x91c5c554
  80. .word 0x60303050, 0x02010103, 0xce6767a9, 0x562b2b7d
  81. .word 0xe7fefe19, 0xb5d7d762, 0x4dababe6, 0xec76769a
  82. .word 0x8fcaca45, 0x1f82829d, 0x89c9c940, 0xfa7d7d87
  83. .word 0xeffafa15, 0xb25959eb, 0x8e4747c9, 0xfbf0f00b
  84. .word 0x41adadec, 0xb3d4d467, 0x5fa2a2fd, 0x45afafea
  85. .word 0x239c9cbf, 0x53a4a4f7, 0xe4727296, 0x9bc0c05b
  86. .word 0x75b7b7c2, 0xe1fdfd1c, 0x3d9393ae, 0x4c26266a
  87. .word 0x6c36365a, 0x7e3f3f41, 0xf5f7f702, 0x83cccc4f
  88. .word 0x6834345c, 0x51a5a5f4, 0xd1e5e534, 0xf9f1f108
  89. .word 0xe2717193, 0xabd8d873, 0x62313153, 0x2a15153f
  90. .word 0x0804040c, 0x95c7c752, 0x46232365, 0x9dc3c35e
  91. .word 0x30181828, 0x379696a1, 0x0a05050f, 0x2f9a9ab5
  92. .word 0x0e070709, 0x24121236, 0x1b80809b, 0xdfe2e23d
  93. .word 0xcdebeb26, 0x4e272769, 0x7fb2b2cd, 0xea75759f
  94. .word 0x1209091b, 0x1d83839e, 0x582c2c74, 0x341a1a2e
  95. .word 0x361b1b2d, 0xdc6e6eb2, 0xb45a5aee, 0x5ba0a0fb
  96. .word 0xa45252f6, 0x763b3b4d, 0xb7d6d661, 0x7db3b3ce
  97. .word 0x5229297b, 0xdde3e33e, 0x5e2f2f71, 0x13848497
  98. .word 0xa65353f5, 0xb9d1d168, 0x00000000, 0xc1eded2c
  99. .word 0x40202060, 0xe3fcfc1f, 0x79b1b1c8, 0xb65b5bed
  100. .word 0xd46a6abe, 0x8dcbcb46, 0x67bebed9, 0x7239394b
  101. .word 0x944a4ade, 0x984c4cd4, 0xb05858e8, 0x85cfcf4a
  102. .word 0xbbd0d06b, 0xc5efef2a, 0x4faaaae5, 0xedfbfb16
  103. .word 0x864343c5, 0x9a4d4dd7, 0x66333355, 0x11858594
  104. .word 0x8a4545cf, 0xe9f9f910, 0x04020206, 0xfe7f7f81
  105. .word 0xa05050f0, 0x783c3c44, 0x259f9fba, 0x4ba8a8e3
  106. .word 0xa25151f3, 0x5da3a3fe, 0x804040c0, 0x058f8f8a
  107. .word 0x3f9292ad, 0x219d9dbc, 0x70383848, 0xf1f5f504
  108. .word 0x63bcbcdf, 0x77b6b6c1, 0xafdada75, 0x42212163
  109. .word 0x20101030, 0xe5ffff1a, 0xfdf3f30e, 0xbfd2d26d
  110. .word 0x81cdcd4c, 0x180c0c14, 0x26131335, 0xc3ecec2f
  111. .word 0xbe5f5fe1, 0x359797a2, 0x884444cc, 0x2e171739
  112. .word 0x93c4c457, 0x55a7a7f2, 0xfc7e7e82, 0x7a3d3d47
  113. .word 0xc86464ac, 0xba5d5de7, 0x3219192b, 0xe6737395
  114. .word 0xc06060a0, 0x19818198, 0x9e4f4fd1, 0xa3dcdc7f
  115. .word 0x44222266, 0x542a2a7e, 0x3b9090ab, 0x0b888883
  116. .word 0x8c4646ca, 0xc7eeee29, 0x6bb8b8d3, 0x2814143c
  117. .word 0xa7dede79, 0xbc5e5ee2, 0x160b0b1d, 0xaddbdb76
  118. .word 0xdbe0e03b, 0x64323256, 0x743a3a4e, 0x140a0a1e
  119. .word 0x924949db, 0x0c06060a, 0x4824246c, 0xb85c5ce4
  120. .word 0x9fc2c25d, 0xbdd3d36e, 0x43acacef, 0xc46262a6
  121. .word 0x399191a8, 0x319595a4, 0xd3e4e437, 0xf279798b
  122. .word 0xd5e7e732, 0x8bc8c843, 0x6e373759, 0xda6d6db7
  123. .word 0x018d8d8c, 0xb1d5d564, 0x9c4e4ed2, 0x49a9a9e0
  124. .word 0xd86c6cb4, 0xac5656fa, 0xf3f4f407, 0xcfeaea25
  125. .word 0xca6565af, 0xf47a7a8e, 0x47aeaee9, 0x10080818
  126. .word 0x6fbabad5, 0xf0787888, 0x4a25256f, 0x5c2e2e72
  127. .word 0x381c1c24, 0x57a6a6f1, 0x73b4b4c7, 0x97c6c651
  128. .word 0xcbe8e823, 0xa1dddd7c, 0xe874749c, 0x3e1f1f21
  129. .word 0x964b4bdd, 0x61bdbddc, 0x0d8b8b86, 0x0f8a8a85
  130. .word 0xe0707090, 0x7c3e3e42, 0x71b5b5c4, 0xcc6666aa
  131. .word 0x904848d8, 0x06030305, 0xf7f6f601, 0x1c0e0e12
  132. .word 0xc26161a3, 0x6a35355f, 0xae5757f9, 0x69b9b9d0
  133. .word 0x17868691, 0x99c1c158, 0x3a1d1d27, 0x279e9eb9
  134. .word 0xd9e1e138, 0xebf8f813, 0x2b9898b3, 0x22111133
  135. .word 0xd26969bb, 0xa9d9d970, 0x078e8e89, 0x339494a7
  136. .word 0x2d9b9bb6, 0x3c1e1e22, 0x15878792, 0xc9e9e920
  137. .word 0x87cece49, 0xaa5555ff, 0x50282878, 0xa5dfdf7a
  138. .word 0x038c8c8f, 0x59a1a1f8, 0x09898980, 0x1a0d0d17
  139. .word 0x65bfbfda, 0xd7e6e631, 0x844242c6, 0xd06868b8
  140. .word 0x824141c3, 0x299999b0, 0x5a2d2d77, 0x1e0f0f11
  141. .word 0x7bb0b0cb, 0xa85454fc, 0x6dbbbbd6, 0x2c16163a
  142. @ Te4[256]
  143. .byte 0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5
  144. .byte 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76
  145. .byte 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0
  146. .byte 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0
  147. .byte 0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc
  148. .byte 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15
  149. .byte 0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a
  150. .byte 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75
  151. .byte 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0
  152. .byte 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84
  153. .byte 0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b
  154. .byte 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf
  155. .byte 0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85
  156. .byte 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8
  157. .byte 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5
  158. .byte 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2
  159. .byte 0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17
  160. .byte 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73
  161. .byte 0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88
  162. .byte 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb
  163. .byte 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c
  164. .byte 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79
  165. .byte 0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9
  166. .byte 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08
  167. .byte 0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6
  168. .byte 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a
  169. .byte 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e
  170. .byte 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e
  171. .byte 0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94
  172. .byte 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf
  173. .byte 0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68
  174. .byte 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16
  175. @ rcon[]
  176. .word 0x01000000, 0x02000000, 0x04000000, 0x08000000
  177. .word 0x10000000, 0x20000000, 0x40000000, 0x80000000
  178. .word 0x1B000000, 0x36000000, 0, 0, 0, 0, 0, 0
  179. .size AES_Te,.-AES_Te
  180. @ void AES_encrypt(const unsigned char *in, unsigned char *out,
  181. @ const AES_KEY *key) {
  182. .global AES_encrypt
  183. .type AES_encrypt,%function
  184. .align 5
  185. AES_encrypt:
  186. #ifndef __thumb2__
  187. sub r3,pc,#8 @ AES_encrypt
  188. #else
  189. adr r3,.
  190. #endif
  191. stmdb sp!,{r1,r4-r12,lr}
  192. #if defined(__thumb2__) || defined(__APPLE__)
  193. adr $tbl,AES_Te
  194. #else
  195. sub $tbl,r3,#AES_encrypt-AES_Te @ Te
  196. #endif
  197. mov $rounds,r0 @ inp
  198. mov $key,r2
  199. #if __ARM_ARCH__<7
  200. ldrb $s0,[$rounds,#3] @ load input data in endian-neutral
  201. ldrb $t1,[$rounds,#2] @ manner...
  202. ldrb $t2,[$rounds,#1]
  203. ldrb $t3,[$rounds,#0]
  204. orr $s0,$s0,$t1,lsl#8
  205. ldrb $s1,[$rounds,#7]
  206. orr $s0,$s0,$t2,lsl#16
  207. ldrb $t1,[$rounds,#6]
  208. orr $s0,$s0,$t3,lsl#24
  209. ldrb $t2,[$rounds,#5]
  210. ldrb $t3,[$rounds,#4]
  211. orr $s1,$s1,$t1,lsl#8
  212. ldrb $s2,[$rounds,#11]
  213. orr $s1,$s1,$t2,lsl#16
  214. ldrb $t1,[$rounds,#10]
  215. orr $s1,$s1,$t3,lsl#24
  216. ldrb $t2,[$rounds,#9]
  217. ldrb $t3,[$rounds,#8]
  218. orr $s2,$s2,$t1,lsl#8
  219. ldrb $s3,[$rounds,#15]
  220. orr $s2,$s2,$t2,lsl#16
  221. ldrb $t1,[$rounds,#14]
  222. orr $s2,$s2,$t3,lsl#24
  223. ldrb $t2,[$rounds,#13]
  224. ldrb $t3,[$rounds,#12]
  225. orr $s3,$s3,$t1,lsl#8
  226. orr $s3,$s3,$t2,lsl#16
  227. orr $s3,$s3,$t3,lsl#24
  228. #else
  229. ldr $s0,[$rounds,#0]
  230. ldr $s1,[$rounds,#4]
  231. ldr $s2,[$rounds,#8]
  232. ldr $s3,[$rounds,#12]
  233. #ifdef __ARMEL__
  234. rev $s0,$s0
  235. rev $s1,$s1
  236. rev $s2,$s2
  237. rev $s3,$s3
  238. #endif
  239. #endif
  240. bl _armv4_AES_encrypt
  241. ldr $rounds,[sp],#4 @ pop out
  242. #if __ARM_ARCH__>=7
  243. #ifdef __ARMEL__
  244. rev $s0,$s0
  245. rev $s1,$s1
  246. rev $s2,$s2
  247. rev $s3,$s3
  248. #endif
  249. str $s0,[$rounds,#0]
  250. str $s1,[$rounds,#4]
  251. str $s2,[$rounds,#8]
  252. str $s3,[$rounds,#12]
  253. #else
  254. mov $t1,$s0,lsr#24 @ write output in endian-neutral
  255. mov $t2,$s0,lsr#16 @ manner...
  256. mov $t3,$s0,lsr#8
  257. strb $t1,[$rounds,#0]
  258. strb $t2,[$rounds,#1]
  259. mov $t1,$s1,lsr#24
  260. strb $t3,[$rounds,#2]
  261. mov $t2,$s1,lsr#16
  262. strb $s0,[$rounds,#3]
  263. mov $t3,$s1,lsr#8
  264. strb $t1,[$rounds,#4]
  265. strb $t2,[$rounds,#5]
  266. mov $t1,$s2,lsr#24
  267. strb $t3,[$rounds,#6]
  268. mov $t2,$s2,lsr#16
  269. strb $s1,[$rounds,#7]
  270. mov $t3,$s2,lsr#8
  271. strb $t1,[$rounds,#8]
  272. strb $t2,[$rounds,#9]
  273. mov $t1,$s3,lsr#24
  274. strb $t3,[$rounds,#10]
  275. mov $t2,$s3,lsr#16
  276. strb $s2,[$rounds,#11]
  277. mov $t3,$s3,lsr#8
  278. strb $t1,[$rounds,#12]
  279. strb $t2,[$rounds,#13]
  280. strb $t3,[$rounds,#14]
  281. strb $s3,[$rounds,#15]
  282. #endif
  283. #if __ARM_ARCH__>=5
  284. ldmia sp!,{r4-r12,pc}
  285. #else
  286. ldmia sp!,{r4-r12,lr}
  287. tst lr,#1
  288. moveq pc,lr @ be binary compatible with V4, yet
  289. bx lr @ interoperable with Thumb ISA:-)
  290. #endif
  291. .size AES_encrypt,.-AES_encrypt
  292. .type _armv4_AES_encrypt,%function
  293. .align 2
  294. _armv4_AES_encrypt:
  295. str lr,[sp,#-4]! @ push lr
  296. ldmia $key!,{$t1-$i1}
  297. eor $s0,$s0,$t1
  298. ldr $rounds,[$key,#240-16]
  299. eor $s1,$s1,$t2
  300. eor $s2,$s2,$t3
  301. eor $s3,$s3,$i1
  302. sub $rounds,$rounds,#1
  303. mov lr,#255
  304. and $i1,lr,$s0
  305. and $i2,lr,$s0,lsr#8
  306. and $i3,lr,$s0,lsr#16
  307. mov $s0,$s0,lsr#24
  308. .Lenc_loop:
  309. ldr $t1,[$tbl,$i1,lsl#2] @ Te3[s0>>0]
  310. and $i1,lr,$s1,lsr#16 @ i0
  311. ldr $t2,[$tbl,$i2,lsl#2] @ Te2[s0>>8]
  312. and $i2,lr,$s1
  313. ldr $t3,[$tbl,$i3,lsl#2] @ Te1[s0>>16]
  314. and $i3,lr,$s1,lsr#8
  315. ldr $s0,[$tbl,$s0,lsl#2] @ Te0[s0>>24]
  316. mov $s1,$s1,lsr#24
  317. ldr $i1,[$tbl,$i1,lsl#2] @ Te1[s1>>16]
  318. ldr $i2,[$tbl,$i2,lsl#2] @ Te3[s1>>0]
  319. ldr $i3,[$tbl,$i3,lsl#2] @ Te2[s1>>8]
  320. eor $s0,$s0,$i1,ror#8
  321. ldr $s1,[$tbl,$s1,lsl#2] @ Te0[s1>>24]
  322. and $i1,lr,$s2,lsr#8 @ i0
  323. eor $t2,$t2,$i2,ror#8
  324. and $i2,lr,$s2,lsr#16 @ i1
  325. eor $t3,$t3,$i3,ror#8
  326. and $i3,lr,$s2
  327. ldr $i1,[$tbl,$i1,lsl#2] @ Te2[s2>>8]
  328. eor $s1,$s1,$t1,ror#24
  329. ldr $i2,[$tbl,$i2,lsl#2] @ Te1[s2>>16]
  330. mov $s2,$s2,lsr#24
  331. ldr $i3,[$tbl,$i3,lsl#2] @ Te3[s2>>0]
  332. eor $s0,$s0,$i1,ror#16
  333. ldr $s2,[$tbl,$s2,lsl#2] @ Te0[s2>>24]
  334. and $i1,lr,$s3 @ i0
  335. eor $s1,$s1,$i2,ror#8
  336. and $i2,lr,$s3,lsr#8 @ i1
  337. eor $t3,$t3,$i3,ror#16
  338. and $i3,lr,$s3,lsr#16 @ i2
  339. ldr $i1,[$tbl,$i1,lsl#2] @ Te3[s3>>0]
  340. eor $s2,$s2,$t2,ror#16
  341. ldr $i2,[$tbl,$i2,lsl#2] @ Te2[s3>>8]
  342. mov $s3,$s3,lsr#24
  343. ldr $i3,[$tbl,$i3,lsl#2] @ Te1[s3>>16]
  344. eor $s0,$s0,$i1,ror#24
  345. ldr $i1,[$key],#16
  346. eor $s1,$s1,$i2,ror#16
  347. ldr $s3,[$tbl,$s3,lsl#2] @ Te0[s3>>24]
  348. eor $s2,$s2,$i3,ror#8
  349. ldr $t1,[$key,#-12]
  350. eor $s3,$s3,$t3,ror#8
  351. ldr $t2,[$key,#-8]
  352. eor $s0,$s0,$i1
  353. ldr $t3,[$key,#-4]
  354. and $i1,lr,$s0
  355. eor $s1,$s1,$t1
  356. and $i2,lr,$s0,lsr#8
  357. eor $s2,$s2,$t2
  358. and $i3,lr,$s0,lsr#16
  359. eor $s3,$s3,$t3
  360. mov $s0,$s0,lsr#24
  361. subs $rounds,$rounds,#1
  362. bne .Lenc_loop
  363. add $tbl,$tbl,#2
  364. ldrb $t1,[$tbl,$i1,lsl#2] @ Te4[s0>>0]
  365. and $i1,lr,$s1,lsr#16 @ i0
  366. ldrb $t2,[$tbl,$i2,lsl#2] @ Te4[s0>>8]
  367. and $i2,lr,$s1
  368. ldrb $t3,[$tbl,$i3,lsl#2] @ Te4[s0>>16]
  369. and $i3,lr,$s1,lsr#8
  370. ldrb $s0,[$tbl,$s0,lsl#2] @ Te4[s0>>24]
  371. mov $s1,$s1,lsr#24
  372. ldrb $i1,[$tbl,$i1,lsl#2] @ Te4[s1>>16]
  373. ldrb $i2,[$tbl,$i2,lsl#2] @ Te4[s1>>0]
  374. ldrb $i3,[$tbl,$i3,lsl#2] @ Te4[s1>>8]
  375. eor $s0,$i1,$s0,lsl#8
  376. ldrb $s1,[$tbl,$s1,lsl#2] @ Te4[s1>>24]
  377. and $i1,lr,$s2,lsr#8 @ i0
  378. eor $t2,$i2,$t2,lsl#8
  379. and $i2,lr,$s2,lsr#16 @ i1
  380. eor $t3,$i3,$t3,lsl#8
  381. and $i3,lr,$s2
  382. ldrb $i1,[$tbl,$i1,lsl#2] @ Te4[s2>>8]
  383. eor $s1,$t1,$s1,lsl#24
  384. ldrb $i2,[$tbl,$i2,lsl#2] @ Te4[s2>>16]
  385. mov $s2,$s2,lsr#24
  386. ldrb $i3,[$tbl,$i3,lsl#2] @ Te4[s2>>0]
  387. eor $s0,$i1,$s0,lsl#8
  388. ldrb $s2,[$tbl,$s2,lsl#2] @ Te4[s2>>24]
  389. and $i1,lr,$s3 @ i0
  390. eor $s1,$s1,$i2,lsl#16
  391. and $i2,lr,$s3,lsr#8 @ i1
  392. eor $t3,$i3,$t3,lsl#8
  393. and $i3,lr,$s3,lsr#16 @ i2
  394. ldrb $i1,[$tbl,$i1,lsl#2] @ Te4[s3>>0]
  395. eor $s2,$t2,$s2,lsl#24
  396. ldrb $i2,[$tbl,$i2,lsl#2] @ Te4[s3>>8]
  397. mov $s3,$s3,lsr#24
  398. ldrb $i3,[$tbl,$i3,lsl#2] @ Te4[s3>>16]
  399. eor $s0,$i1,$s0,lsl#8
  400. ldr $i1,[$key,#0]
  401. ldrb $s3,[$tbl,$s3,lsl#2] @ Te4[s3>>24]
  402. eor $s1,$s1,$i2,lsl#8
  403. ldr $t1,[$key,#4]
  404. eor $s2,$s2,$i3,lsl#16
  405. ldr $t2,[$key,#8]
  406. eor $s3,$t3,$s3,lsl#24
  407. ldr $t3,[$key,#12]
  408. eor $s0,$s0,$i1
  409. eor $s1,$s1,$t1
  410. eor $s2,$s2,$t2
  411. eor $s3,$s3,$t3
  412. sub $tbl,$tbl,#2
  413. ldr pc,[sp],#4 @ pop and return
  414. .size _armv4_AES_encrypt,.-_armv4_AES_encrypt
  415. .global AES_set_encrypt_key
  416. .type AES_set_encrypt_key,%function
  417. .align 5
  418. AES_set_encrypt_key:
  419. _armv4_AES_set_encrypt_key:
  420. #ifndef __thumb2__
  421. sub r3,pc,#8 @ AES_set_encrypt_key
  422. #else
  423. adr r3,.
  424. #endif
  425. teq r0,#0
  426. #ifdef __thumb2__
  427. itt eq @ Thumb2 thing, sanity check in ARM
  428. #endif
  429. moveq r0,#-1
  430. beq .Labrt
  431. teq r2,#0
  432. #ifdef __thumb2__
  433. itt eq @ Thumb2 thing, sanity check in ARM
  434. #endif
  435. moveq r0,#-1
  436. beq .Labrt
  437. teq r1,#128
  438. beq .Lok
  439. teq r1,#192
  440. beq .Lok
  441. teq r1,#256
  442. #ifdef __thumb2__
  443. itt ne @ Thumb2 thing, sanity check in ARM
  444. #endif
  445. movne r0,#-1
  446. bne .Labrt
  447. .Lok: stmdb sp!,{r4-r12,lr}
  448. mov $rounds,r0 @ inp
  449. mov lr,r1 @ bits
  450. mov $key,r2 @ key
  451. #if defined(__thumb2__) || defined(__APPLE__)
  452. adr $tbl,AES_Te+1024 @ Te4
  453. #else
  454. sub $tbl,r3,#_armv4_AES_set_encrypt_key-AES_Te-1024 @ Te4
  455. #endif
  456. #if __ARM_ARCH__<7
  457. ldrb $s0,[$rounds,#3] @ load input data in endian-neutral
  458. ldrb $t1,[$rounds,#2] @ manner...
  459. ldrb $t2,[$rounds,#1]
  460. ldrb $t3,[$rounds,#0]
  461. orr $s0,$s0,$t1,lsl#8
  462. ldrb $s1,[$rounds,#7]
  463. orr $s0,$s0,$t2,lsl#16
  464. ldrb $t1,[$rounds,#6]
  465. orr $s0,$s0,$t3,lsl#24
  466. ldrb $t2,[$rounds,#5]
  467. ldrb $t3,[$rounds,#4]
  468. orr $s1,$s1,$t1,lsl#8
  469. ldrb $s2,[$rounds,#11]
  470. orr $s1,$s1,$t2,lsl#16
  471. ldrb $t1,[$rounds,#10]
  472. orr $s1,$s1,$t3,lsl#24
  473. ldrb $t2,[$rounds,#9]
  474. ldrb $t3,[$rounds,#8]
  475. orr $s2,$s2,$t1,lsl#8
  476. ldrb $s3,[$rounds,#15]
  477. orr $s2,$s2,$t2,lsl#16
  478. ldrb $t1,[$rounds,#14]
  479. orr $s2,$s2,$t3,lsl#24
  480. ldrb $t2,[$rounds,#13]
  481. ldrb $t3,[$rounds,#12]
  482. orr $s3,$s3,$t1,lsl#8
  483. str $s0,[$key],#16
  484. orr $s3,$s3,$t2,lsl#16
  485. str $s1,[$key,#-12]
  486. orr $s3,$s3,$t3,lsl#24
  487. str $s2,[$key,#-8]
  488. str $s3,[$key,#-4]
  489. #else
  490. ldr $s0,[$rounds,#0]
  491. ldr $s1,[$rounds,#4]
  492. ldr $s2,[$rounds,#8]
  493. ldr $s3,[$rounds,#12]
  494. #ifdef __ARMEL__
  495. rev $s0,$s0
  496. rev $s1,$s1
  497. rev $s2,$s2
  498. rev $s3,$s3
  499. #endif
  500. str $s0,[$key],#16
  501. str $s1,[$key,#-12]
  502. str $s2,[$key,#-8]
  503. str $s3,[$key,#-4]
  504. #endif
  505. teq lr,#128
  506. bne .Lnot128
  507. mov $rounds,#10
  508. str $rounds,[$key,#240-16]
  509. add $t3,$tbl,#256 @ rcon
  510. mov lr,#255
  511. .L128_loop:
  512. and $t2,lr,$s3,lsr#24
  513. and $i1,lr,$s3,lsr#16
  514. ldrb $t2,[$tbl,$t2]
  515. and $i2,lr,$s3,lsr#8
  516. ldrb $i1,[$tbl,$i1]
  517. and $i3,lr,$s3
  518. ldrb $i2,[$tbl,$i2]
  519. orr $t2,$t2,$i1,lsl#24
  520. ldrb $i3,[$tbl,$i3]
  521. orr $t2,$t2,$i2,lsl#16
  522. ldr $t1,[$t3],#4 @ rcon[i++]
  523. orr $t2,$t2,$i3,lsl#8
  524. eor $t2,$t2,$t1
  525. eor $s0,$s0,$t2 @ rk[4]=rk[0]^...
  526. eor $s1,$s1,$s0 @ rk[5]=rk[1]^rk[4]
  527. str $s0,[$key],#16
  528. eor $s2,$s2,$s1 @ rk[6]=rk[2]^rk[5]
  529. str $s1,[$key,#-12]
  530. eor $s3,$s3,$s2 @ rk[7]=rk[3]^rk[6]
  531. str $s2,[$key,#-8]
  532. subs $rounds,$rounds,#1
  533. str $s3,[$key,#-4]
  534. bne .L128_loop
  535. sub r2,$key,#176
  536. b .Ldone
  537. .Lnot128:
  538. #if __ARM_ARCH__<7
  539. ldrb $i2,[$rounds,#19]
  540. ldrb $t1,[$rounds,#18]
  541. ldrb $t2,[$rounds,#17]
  542. ldrb $t3,[$rounds,#16]
  543. orr $i2,$i2,$t1,lsl#8
  544. ldrb $i3,[$rounds,#23]
  545. orr $i2,$i2,$t2,lsl#16
  546. ldrb $t1,[$rounds,#22]
  547. orr $i2,$i2,$t3,lsl#24
  548. ldrb $t2,[$rounds,#21]
  549. ldrb $t3,[$rounds,#20]
  550. orr $i3,$i3,$t1,lsl#8
  551. orr $i3,$i3,$t2,lsl#16
  552. str $i2,[$key],#8
  553. orr $i3,$i3,$t3,lsl#24
  554. str $i3,[$key,#-4]
  555. #else
  556. ldr $i2,[$rounds,#16]
  557. ldr $i3,[$rounds,#20]
  558. #ifdef __ARMEL__
  559. rev $i2,$i2
  560. rev $i3,$i3
  561. #endif
  562. str $i2,[$key],#8
  563. str $i3,[$key,#-4]
  564. #endif
  565. teq lr,#192
  566. bne .Lnot192
  567. mov $rounds,#12
  568. str $rounds,[$key,#240-24]
  569. add $t3,$tbl,#256 @ rcon
  570. mov lr,#255
  571. mov $rounds,#8
  572. .L192_loop:
  573. and $t2,lr,$i3,lsr#24
  574. and $i1,lr,$i3,lsr#16
  575. ldrb $t2,[$tbl,$t2]
  576. and $i2,lr,$i3,lsr#8
  577. ldrb $i1,[$tbl,$i1]
  578. and $i3,lr,$i3
  579. ldrb $i2,[$tbl,$i2]
  580. orr $t2,$t2,$i1,lsl#24
  581. ldrb $i3,[$tbl,$i3]
  582. orr $t2,$t2,$i2,lsl#16
  583. ldr $t1,[$t3],#4 @ rcon[i++]
  584. orr $t2,$t2,$i3,lsl#8
  585. eor $i3,$t2,$t1
  586. eor $s0,$s0,$i3 @ rk[6]=rk[0]^...
  587. eor $s1,$s1,$s0 @ rk[7]=rk[1]^rk[6]
  588. str $s0,[$key],#24
  589. eor $s2,$s2,$s1 @ rk[8]=rk[2]^rk[7]
  590. str $s1,[$key,#-20]
  591. eor $s3,$s3,$s2 @ rk[9]=rk[3]^rk[8]
  592. str $s2,[$key,#-16]
  593. subs $rounds,$rounds,#1
  594. str $s3,[$key,#-12]
  595. #ifdef __thumb2__
  596. itt eq @ Thumb2 thing, sanity check in ARM
  597. #endif
  598. subeq r2,$key,#216
  599. beq .Ldone
  600. ldr $i1,[$key,#-32]
  601. ldr $i2,[$key,#-28]
  602. eor $i1,$i1,$s3 @ rk[10]=rk[4]^rk[9]
  603. eor $i3,$i2,$i1 @ rk[11]=rk[5]^rk[10]
  604. str $i1,[$key,#-8]
  605. str $i3,[$key,#-4]
  606. b .L192_loop
  607. .Lnot192:
  608. #if __ARM_ARCH__<7
  609. ldrb $i2,[$rounds,#27]
  610. ldrb $t1,[$rounds,#26]
  611. ldrb $t2,[$rounds,#25]
  612. ldrb $t3,[$rounds,#24]
  613. orr $i2,$i2,$t1,lsl#8
  614. ldrb $i3,[$rounds,#31]
  615. orr $i2,$i2,$t2,lsl#16
  616. ldrb $t1,[$rounds,#30]
  617. orr $i2,$i2,$t3,lsl#24
  618. ldrb $t2,[$rounds,#29]
  619. ldrb $t3,[$rounds,#28]
  620. orr $i3,$i3,$t1,lsl#8
  621. orr $i3,$i3,$t2,lsl#16
  622. str $i2,[$key],#8
  623. orr $i3,$i3,$t3,lsl#24
  624. str $i3,[$key,#-4]
  625. #else
  626. ldr $i2,[$rounds,#24]
  627. ldr $i3,[$rounds,#28]
  628. #ifdef __ARMEL__
  629. rev $i2,$i2
  630. rev $i3,$i3
  631. #endif
  632. str $i2,[$key],#8
  633. str $i3,[$key,#-4]
  634. #endif
  635. mov $rounds,#14
  636. str $rounds,[$key,#240-32]
  637. add $t3,$tbl,#256 @ rcon
  638. mov lr,#255
  639. mov $rounds,#7
  640. .L256_loop:
  641. and $t2,lr,$i3,lsr#24
  642. and $i1,lr,$i3,lsr#16
  643. ldrb $t2,[$tbl,$t2]
  644. and $i2,lr,$i3,lsr#8
  645. ldrb $i1,[$tbl,$i1]
  646. and $i3,lr,$i3
  647. ldrb $i2,[$tbl,$i2]
  648. orr $t2,$t2,$i1,lsl#24
  649. ldrb $i3,[$tbl,$i3]
  650. orr $t2,$t2,$i2,lsl#16
  651. ldr $t1,[$t3],#4 @ rcon[i++]
  652. orr $t2,$t2,$i3,lsl#8
  653. eor $i3,$t2,$t1
  654. eor $s0,$s0,$i3 @ rk[8]=rk[0]^...
  655. eor $s1,$s1,$s0 @ rk[9]=rk[1]^rk[8]
  656. str $s0,[$key],#32
  657. eor $s2,$s2,$s1 @ rk[10]=rk[2]^rk[9]
  658. str $s1,[$key,#-28]
  659. eor $s3,$s3,$s2 @ rk[11]=rk[3]^rk[10]
  660. str $s2,[$key,#-24]
  661. subs $rounds,$rounds,#1
  662. str $s3,[$key,#-20]
  663. #ifdef __thumb2__
  664. itt eq @ Thumb2 thing, sanity check in ARM
  665. #endif
  666. subeq r2,$key,#256
  667. beq .Ldone
  668. and $t2,lr,$s3
  669. and $i1,lr,$s3,lsr#8
  670. ldrb $t2,[$tbl,$t2]
  671. and $i2,lr,$s3,lsr#16
  672. ldrb $i1,[$tbl,$i1]
  673. and $i3,lr,$s3,lsr#24
  674. ldrb $i2,[$tbl,$i2]
  675. orr $t2,$t2,$i1,lsl#8
  676. ldrb $i3,[$tbl,$i3]
  677. orr $t2,$t2,$i2,lsl#16
  678. ldr $t1,[$key,#-48]
  679. orr $t2,$t2,$i3,lsl#24
  680. ldr $i1,[$key,#-44]
  681. ldr $i2,[$key,#-40]
  682. eor $t1,$t1,$t2 @ rk[12]=rk[4]^...
  683. ldr $i3,[$key,#-36]
  684. eor $i1,$i1,$t1 @ rk[13]=rk[5]^rk[12]
  685. str $t1,[$key,#-16]
  686. eor $i2,$i2,$i1 @ rk[14]=rk[6]^rk[13]
  687. str $i1,[$key,#-12]
  688. eor $i3,$i3,$i2 @ rk[15]=rk[7]^rk[14]
  689. str $i2,[$key,#-8]
  690. str $i3,[$key,#-4]
  691. b .L256_loop
  692. .align 2
  693. .Ldone: mov r0,#0
  694. ldmia sp!,{r4-r12,lr}
  695. .Labrt:
  696. #if __ARM_ARCH__>=5
  697. ret @ bx lr
  698. #else
  699. tst lr,#1
  700. moveq pc,lr @ be binary compatible with V4, yet
  701. bx lr @ interoperable with Thumb ISA:-)
  702. #endif
  703. .size AES_set_encrypt_key,.-AES_set_encrypt_key
  704. .global AES_set_decrypt_key
  705. .type AES_set_decrypt_key,%function
  706. .align 5
  707. AES_set_decrypt_key:
  708. str lr,[sp,#-4]! @ push lr
  709. bl _armv4_AES_set_encrypt_key
  710. teq r0,#0
  711. ldr lr,[sp],#4 @ pop lr
  712. bne .Labrt
  713. mov r0,r2 @ AES_set_encrypt_key preserves r2,
  714. mov r1,r2 @ which is AES_KEY *key
  715. b _armv4_AES_set_enc2dec_key
  716. .size AES_set_decrypt_key,.-AES_set_decrypt_key
  717. @ void AES_set_enc2dec_key(const AES_KEY *inp,AES_KEY *out)
  718. .global AES_set_enc2dec_key
  719. .type AES_set_enc2dec_key,%function
  720. .align 5
  721. AES_set_enc2dec_key:
  722. _armv4_AES_set_enc2dec_key:
  723. stmdb sp!,{r4-r12,lr}
  724. ldr $rounds,[r0,#240]
  725. mov $i1,r0 @ input
  726. add $i2,r0,$rounds,lsl#4
  727. mov $key,r1 @ output
  728. add $tbl,r1,$rounds,lsl#4
  729. str $rounds,[r1,#240]
  730. .Linv: ldr $s0,[$i1],#16
  731. ldr $s1,[$i1,#-12]
  732. ldr $s2,[$i1,#-8]
  733. ldr $s3,[$i1,#-4]
  734. ldr $t1,[$i2],#-16
  735. ldr $t2,[$i2,#16+4]
  736. ldr $t3,[$i2,#16+8]
  737. ldr $i3,[$i2,#16+12]
  738. str $s0,[$tbl],#-16
  739. str $s1,[$tbl,#16+4]
  740. str $s2,[$tbl,#16+8]
  741. str $s3,[$tbl,#16+12]
  742. str $t1,[$key],#16
  743. str $t2,[$key,#-12]
  744. str $t3,[$key,#-8]
  745. str $i3,[$key,#-4]
  746. teq $i1,$i2
  747. bne .Linv
  748. ldr $s0,[$i1]
  749. ldr $s1,[$i1,#4]
  750. ldr $s2,[$i1,#8]
  751. ldr $s3,[$i1,#12]
  752. str $s0,[$key]
  753. str $s1,[$key,#4]
  754. str $s2,[$key,#8]
  755. str $s3,[$key,#12]
  756. sub $key,$key,$rounds,lsl#3
  757. ___
  758. $mask80=$i1;
  759. $mask1b=$i2;
  760. $mask7f=$i3;
  761. $code.=<<___;
  762. ldr $s0,[$key,#16]! @ prefetch tp1
  763. mov $mask80,#0x80
  764. mov $mask1b,#0x1b
  765. orr $mask80,$mask80,#0x8000
  766. orr $mask1b,$mask1b,#0x1b00
  767. orr $mask80,$mask80,$mask80,lsl#16
  768. orr $mask1b,$mask1b,$mask1b,lsl#16
  769. sub $rounds,$rounds,#1
  770. mvn $mask7f,$mask80
  771. mov $rounds,$rounds,lsl#2 @ (rounds-1)*4
  772. .Lmix: and $t1,$s0,$mask80
  773. and $s1,$s0,$mask7f
  774. sub $t1,$t1,$t1,lsr#7
  775. and $t1,$t1,$mask1b
  776. eor $s1,$t1,$s1,lsl#1 @ tp2
  777. and $t1,$s1,$mask80
  778. and $s2,$s1,$mask7f
  779. sub $t1,$t1,$t1,lsr#7
  780. and $t1,$t1,$mask1b
  781. eor $s2,$t1,$s2,lsl#1 @ tp4
  782. and $t1,$s2,$mask80
  783. and $s3,$s2,$mask7f
  784. sub $t1,$t1,$t1,lsr#7
  785. and $t1,$t1,$mask1b
  786. eor $s3,$t1,$s3,lsl#1 @ tp8
  787. eor $t1,$s1,$s2
  788. eor $t2,$s0,$s3 @ tp9
  789. eor $t1,$t1,$s3 @ tpe
  790. eor $t1,$t1,$s1,ror#24
  791. eor $t1,$t1,$t2,ror#24 @ ^= ROTATE(tpb=tp9^tp2,8)
  792. eor $t1,$t1,$s2,ror#16
  793. eor $t1,$t1,$t2,ror#16 @ ^= ROTATE(tpd=tp9^tp4,16)
  794. eor $t1,$t1,$t2,ror#8 @ ^= ROTATE(tp9,24)
  795. ldr $s0,[$key,#4] @ prefetch tp1
  796. str $t1,[$key],#4
  797. subs $rounds,$rounds,#1
  798. bne .Lmix
  799. mov r0,#0
  800. #if __ARM_ARCH__>=5
  801. ldmia sp!,{r4-r12,pc}
  802. #else
  803. ldmia sp!,{r4-r12,lr}
  804. tst lr,#1
  805. moveq pc,lr @ be binary compatible with V4, yet
  806. bx lr @ interoperable with Thumb ISA:-)
  807. #endif
  808. .size AES_set_enc2dec_key,.-AES_set_enc2dec_key
  809. .type AES_Td,%object
  810. .align 5
  811. AES_Td:
  812. .word 0x51f4a750, 0x7e416553, 0x1a17a4c3, 0x3a275e96
  813. .word 0x3bab6bcb, 0x1f9d45f1, 0xacfa58ab, 0x4be30393
  814. .word 0x2030fa55, 0xad766df6, 0x88cc7691, 0xf5024c25
  815. .word 0x4fe5d7fc, 0xc52acbd7, 0x26354480, 0xb562a38f
  816. .word 0xdeb15a49, 0x25ba1b67, 0x45ea0e98, 0x5dfec0e1
  817. .word 0xc32f7502, 0x814cf012, 0x8d4697a3, 0x6bd3f9c6
  818. .word 0x038f5fe7, 0x15929c95, 0xbf6d7aeb, 0x955259da
  819. .word 0xd4be832d, 0x587421d3, 0x49e06929, 0x8ec9c844
  820. .word 0x75c2896a, 0xf48e7978, 0x99583e6b, 0x27b971dd
  821. .word 0xbee14fb6, 0xf088ad17, 0xc920ac66, 0x7dce3ab4
  822. .word 0x63df4a18, 0xe51a3182, 0x97513360, 0x62537f45
  823. .word 0xb16477e0, 0xbb6bae84, 0xfe81a01c, 0xf9082b94
  824. .word 0x70486858, 0x8f45fd19, 0x94de6c87, 0x527bf8b7
  825. .word 0xab73d323, 0x724b02e2, 0xe31f8f57, 0x6655ab2a
  826. .word 0xb2eb2807, 0x2fb5c203, 0x86c57b9a, 0xd33708a5
  827. .word 0x302887f2, 0x23bfa5b2, 0x02036aba, 0xed16825c
  828. .word 0x8acf1c2b, 0xa779b492, 0xf307f2f0, 0x4e69e2a1
  829. .word 0x65daf4cd, 0x0605bed5, 0xd134621f, 0xc4a6fe8a
  830. .word 0x342e539d, 0xa2f355a0, 0x058ae132, 0xa4f6eb75
  831. .word 0x0b83ec39, 0x4060efaa, 0x5e719f06, 0xbd6e1051
  832. .word 0x3e218af9, 0x96dd063d, 0xdd3e05ae, 0x4de6bd46
  833. .word 0x91548db5, 0x71c45d05, 0x0406d46f, 0x605015ff
  834. .word 0x1998fb24, 0xd6bde997, 0x894043cc, 0x67d99e77
  835. .word 0xb0e842bd, 0x07898b88, 0xe7195b38, 0x79c8eedb
  836. .word 0xa17c0a47, 0x7c420fe9, 0xf8841ec9, 0x00000000
  837. .word 0x09808683, 0x322bed48, 0x1e1170ac, 0x6c5a724e
  838. .word 0xfd0efffb, 0x0f853856, 0x3daed51e, 0x362d3927
  839. .word 0x0a0fd964, 0x685ca621, 0x9b5b54d1, 0x24362e3a
  840. .word 0x0c0a67b1, 0x9357e70f, 0xb4ee96d2, 0x1b9b919e
  841. .word 0x80c0c54f, 0x61dc20a2, 0x5a774b69, 0x1c121a16
  842. .word 0xe293ba0a, 0xc0a02ae5, 0x3c22e043, 0x121b171d
  843. .word 0x0e090d0b, 0xf28bc7ad, 0x2db6a8b9, 0x141ea9c8
  844. .word 0x57f11985, 0xaf75074c, 0xee99ddbb, 0xa37f60fd
  845. .word 0xf701269f, 0x5c72f5bc, 0x44663bc5, 0x5bfb7e34
  846. .word 0x8b432976, 0xcb23c6dc, 0xb6edfc68, 0xb8e4f163
  847. .word 0xd731dcca, 0x42638510, 0x13972240, 0x84c61120
  848. .word 0x854a247d, 0xd2bb3df8, 0xaef93211, 0xc729a16d
  849. .word 0x1d9e2f4b, 0xdcb230f3, 0x0d8652ec, 0x77c1e3d0
  850. .word 0x2bb3166c, 0xa970b999, 0x119448fa, 0x47e96422
  851. .word 0xa8fc8cc4, 0xa0f03f1a, 0x567d2cd8, 0x223390ef
  852. .word 0x87494ec7, 0xd938d1c1, 0x8ccaa2fe, 0x98d40b36
  853. .word 0xa6f581cf, 0xa57ade28, 0xdab78e26, 0x3fadbfa4
  854. .word 0x2c3a9de4, 0x5078920d, 0x6a5fcc9b, 0x547e4662
  855. .word 0xf68d13c2, 0x90d8b8e8, 0x2e39f75e, 0x82c3aff5
  856. .word 0x9f5d80be, 0x69d0937c, 0x6fd52da9, 0xcf2512b3
  857. .word 0xc8ac993b, 0x10187da7, 0xe89c636e, 0xdb3bbb7b
  858. .word 0xcd267809, 0x6e5918f4, 0xec9ab701, 0x834f9aa8
  859. .word 0xe6956e65, 0xaaffe67e, 0x21bccf08, 0xef15e8e6
  860. .word 0xbae79bd9, 0x4a6f36ce, 0xea9f09d4, 0x29b07cd6
  861. .word 0x31a4b2af, 0x2a3f2331, 0xc6a59430, 0x35a266c0
  862. .word 0x744ebc37, 0xfc82caa6, 0xe090d0b0, 0x33a7d815
  863. .word 0xf104984a, 0x41ecdaf7, 0x7fcd500e, 0x1791f62f
  864. .word 0x764dd68d, 0x43efb04d, 0xccaa4d54, 0xe49604df
  865. .word 0x9ed1b5e3, 0x4c6a881b, 0xc12c1fb8, 0x4665517f
  866. .word 0x9d5eea04, 0x018c355d, 0xfa877473, 0xfb0b412e
  867. .word 0xb3671d5a, 0x92dbd252, 0xe9105633, 0x6dd64713
  868. .word 0x9ad7618c, 0x37a10c7a, 0x59f8148e, 0xeb133c89
  869. .word 0xcea927ee, 0xb761c935, 0xe11ce5ed, 0x7a47b13c
  870. .word 0x9cd2df59, 0x55f2733f, 0x1814ce79, 0x73c737bf
  871. .word 0x53f7cdea, 0x5ffdaa5b, 0xdf3d6f14, 0x7844db86
  872. .word 0xcaaff381, 0xb968c43e, 0x3824342c, 0xc2a3405f
  873. .word 0x161dc372, 0xbce2250c, 0x283c498b, 0xff0d9541
  874. .word 0x39a80171, 0x080cb3de, 0xd8b4e49c, 0x6456c190
  875. .word 0x7bcb8461, 0xd532b670, 0x486c5c74, 0xd0b85742
  876. @ Td4[256]
  877. .byte 0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38
  878. .byte 0xbf, 0x40, 0xa3, 0x9e, 0x81, 0xf3, 0xd7, 0xfb
  879. .byte 0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87
  880. .byte 0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb
  881. .byte 0x54, 0x7b, 0x94, 0x32, 0xa6, 0xc2, 0x23, 0x3d
  882. .byte 0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e
  883. .byte 0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2
  884. .byte 0x76, 0x5b, 0xa2, 0x49, 0x6d, 0x8b, 0xd1, 0x25
  885. .byte 0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16
  886. .byte 0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92
  887. .byte 0x6c, 0x70, 0x48, 0x50, 0xfd, 0xed, 0xb9, 0xda
  888. .byte 0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84
  889. .byte 0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a
  890. .byte 0xf7, 0xe4, 0x58, 0x05, 0xb8, 0xb3, 0x45, 0x06
  891. .byte 0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02
  892. .byte 0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b
  893. .byte 0x3a, 0x91, 0x11, 0x41, 0x4f, 0x67, 0xdc, 0xea
  894. .byte 0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73
  895. .byte 0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85
  896. .byte 0xe2, 0xf9, 0x37, 0xe8, 0x1c, 0x75, 0xdf, 0x6e
  897. .byte 0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89
  898. .byte 0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b
  899. .byte 0xfc, 0x56, 0x3e, 0x4b, 0xc6, 0xd2, 0x79, 0x20
  900. .byte 0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4
  901. .byte 0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31
  902. .byte 0xb1, 0x12, 0x10, 0x59, 0x27, 0x80, 0xec, 0x5f
  903. .byte 0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d
  904. .byte 0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef
  905. .byte 0xa0, 0xe0, 0x3b, 0x4d, 0xae, 0x2a, 0xf5, 0xb0
  906. .byte 0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61
  907. .byte 0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26
  908. .byte 0xe1, 0x69, 0x14, 0x63, 0x55, 0x21, 0x0c, 0x7d
  909. .size AES_Td,.-AES_Td
  910. @ void AES_decrypt(const unsigned char *in, unsigned char *out,
  911. @ const AES_KEY *key) {
  912. .global AES_decrypt
  913. .type AES_decrypt,%function
  914. .align 5
  915. AES_decrypt:
  916. #ifndef __thumb2__
  917. sub r3,pc,#8 @ AES_decrypt
  918. #else
  919. adr r3,.
  920. #endif
  921. stmdb sp!,{r1,r4-r12,lr}
  922. #if defined(__thumb2__) || defined(__APPLE__)
  923. adr $tbl,AES_Td
  924. #else
  925. sub $tbl,r3,#AES_decrypt-AES_Td @ Td
  926. #endif
  927. mov $rounds,r0 @ inp
  928. mov $key,r2
  929. #if __ARM_ARCH__<7
  930. ldrb $s0,[$rounds,#3] @ load input data in endian-neutral
  931. ldrb $t1,[$rounds,#2] @ manner...
  932. ldrb $t2,[$rounds,#1]
  933. ldrb $t3,[$rounds,#0]
  934. orr $s0,$s0,$t1,lsl#8
  935. ldrb $s1,[$rounds,#7]
  936. orr $s0,$s0,$t2,lsl#16
  937. ldrb $t1,[$rounds,#6]
  938. orr $s0,$s0,$t3,lsl#24
  939. ldrb $t2,[$rounds,#5]
  940. ldrb $t3,[$rounds,#4]
  941. orr $s1,$s1,$t1,lsl#8
  942. ldrb $s2,[$rounds,#11]
  943. orr $s1,$s1,$t2,lsl#16
  944. ldrb $t1,[$rounds,#10]
  945. orr $s1,$s1,$t3,lsl#24
  946. ldrb $t2,[$rounds,#9]
  947. ldrb $t3,[$rounds,#8]
  948. orr $s2,$s2,$t1,lsl#8
  949. ldrb $s3,[$rounds,#15]
  950. orr $s2,$s2,$t2,lsl#16
  951. ldrb $t1,[$rounds,#14]
  952. orr $s2,$s2,$t3,lsl#24
  953. ldrb $t2,[$rounds,#13]
  954. ldrb $t3,[$rounds,#12]
  955. orr $s3,$s3,$t1,lsl#8
  956. orr $s3,$s3,$t2,lsl#16
  957. orr $s3,$s3,$t3,lsl#24
  958. #else
  959. ldr $s0,[$rounds,#0]
  960. ldr $s1,[$rounds,#4]
  961. ldr $s2,[$rounds,#8]
  962. ldr $s3,[$rounds,#12]
  963. #ifdef __ARMEL__
  964. rev $s0,$s0
  965. rev $s1,$s1
  966. rev $s2,$s2
  967. rev $s3,$s3
  968. #endif
  969. #endif
  970. bl _armv4_AES_decrypt
  971. ldr $rounds,[sp],#4 @ pop out
  972. #if __ARM_ARCH__>=7
  973. #ifdef __ARMEL__
  974. rev $s0,$s0
  975. rev $s1,$s1
  976. rev $s2,$s2
  977. rev $s3,$s3
  978. #endif
  979. str $s0,[$rounds,#0]
  980. str $s1,[$rounds,#4]
  981. str $s2,[$rounds,#8]
  982. str $s3,[$rounds,#12]
  983. #else
  984. mov $t1,$s0,lsr#24 @ write output in endian-neutral
  985. mov $t2,$s0,lsr#16 @ manner...
  986. mov $t3,$s0,lsr#8
  987. strb $t1,[$rounds,#0]
  988. strb $t2,[$rounds,#1]
  989. mov $t1,$s1,lsr#24
  990. strb $t3,[$rounds,#2]
  991. mov $t2,$s1,lsr#16
  992. strb $s0,[$rounds,#3]
  993. mov $t3,$s1,lsr#8
  994. strb $t1,[$rounds,#4]
  995. strb $t2,[$rounds,#5]
  996. mov $t1,$s2,lsr#24
  997. strb $t3,[$rounds,#6]
  998. mov $t2,$s2,lsr#16
  999. strb $s1,[$rounds,#7]
  1000. mov $t3,$s2,lsr#8
  1001. strb $t1,[$rounds,#8]
  1002. strb $t2,[$rounds,#9]
  1003. mov $t1,$s3,lsr#24
  1004. strb $t3,[$rounds,#10]
  1005. mov $t2,$s3,lsr#16
  1006. strb $s2,[$rounds,#11]
  1007. mov $t3,$s3,lsr#8
  1008. strb $t1,[$rounds,#12]
  1009. strb $t2,[$rounds,#13]
  1010. strb $t3,[$rounds,#14]
  1011. strb $s3,[$rounds,#15]
  1012. #endif
  1013. #if __ARM_ARCH__>=5
  1014. ldmia sp!,{r4-r12,pc}
  1015. #else
  1016. ldmia sp!,{r4-r12,lr}
  1017. tst lr,#1
  1018. moveq pc,lr @ be binary compatible with V4, yet
  1019. bx lr @ interoperable with Thumb ISA:-)
  1020. #endif
  1021. .size AES_decrypt,.-AES_decrypt
  1022. .type _armv4_AES_decrypt,%function
  1023. .align 2
  1024. _armv4_AES_decrypt:
  1025. str lr,[sp,#-4]! @ push lr
  1026. ldmia $key!,{$t1-$i1}
  1027. eor $s0,$s0,$t1
  1028. ldr $rounds,[$key,#240-16]
  1029. eor $s1,$s1,$t2
  1030. eor $s2,$s2,$t3
  1031. eor $s3,$s3,$i1
  1032. sub $rounds,$rounds,#1
  1033. mov lr,#255
  1034. and $i1,lr,$s0,lsr#16
  1035. and $i2,lr,$s0,lsr#8
  1036. and $i3,lr,$s0
  1037. mov $s0,$s0,lsr#24
  1038. .Ldec_loop:
  1039. ldr $t1,[$tbl,$i1,lsl#2] @ Td1[s0>>16]
  1040. and $i1,lr,$s1 @ i0
  1041. ldr $t2,[$tbl,$i2,lsl#2] @ Td2[s0>>8]
  1042. and $i2,lr,$s1,lsr#16
  1043. ldr $t3,[$tbl,$i3,lsl#2] @ Td3[s0>>0]
  1044. and $i3,lr,$s1,lsr#8
  1045. ldr $s0,[$tbl,$s0,lsl#2] @ Td0[s0>>24]
  1046. mov $s1,$s1,lsr#24
  1047. ldr $i1,[$tbl,$i1,lsl#2] @ Td3[s1>>0]
  1048. ldr $i2,[$tbl,$i2,lsl#2] @ Td1[s1>>16]
  1049. ldr $i3,[$tbl,$i3,lsl#2] @ Td2[s1>>8]
  1050. eor $s0,$s0,$i1,ror#24
  1051. ldr $s1,[$tbl,$s1,lsl#2] @ Td0[s1>>24]
  1052. and $i1,lr,$s2,lsr#8 @ i0
  1053. eor $t2,$i2,$t2,ror#8
  1054. and $i2,lr,$s2 @ i1
  1055. eor $t3,$i3,$t3,ror#8
  1056. and $i3,lr,$s2,lsr#16
  1057. ldr $i1,[$tbl,$i1,lsl#2] @ Td2[s2>>8]
  1058. eor $s1,$s1,$t1,ror#8
  1059. ldr $i2,[$tbl,$i2,lsl#2] @ Td3[s2>>0]
  1060. mov $s2,$s2,lsr#24
  1061. ldr $i3,[$tbl,$i3,lsl#2] @ Td1[s2>>16]
  1062. eor $s0,$s0,$i1,ror#16
  1063. ldr $s2,[$tbl,$s2,lsl#2] @ Td0[s2>>24]
  1064. and $i1,lr,$s3,lsr#16 @ i0
  1065. eor $s1,$s1,$i2,ror#24
  1066. and $i2,lr,$s3,lsr#8 @ i1
  1067. eor $t3,$i3,$t3,ror#8
  1068. and $i3,lr,$s3 @ i2
  1069. ldr $i1,[$tbl,$i1,lsl#2] @ Td1[s3>>16]
  1070. eor $s2,$s2,$t2,ror#8
  1071. ldr $i2,[$tbl,$i2,lsl#2] @ Td2[s3>>8]
  1072. mov $s3,$s3,lsr#24
  1073. ldr $i3,[$tbl,$i3,lsl#2] @ Td3[s3>>0]
  1074. eor $s0,$s0,$i1,ror#8
  1075. ldr $i1,[$key],#16
  1076. eor $s1,$s1,$i2,ror#16
  1077. ldr $s3,[$tbl,$s3,lsl#2] @ Td0[s3>>24]
  1078. eor $s2,$s2,$i3,ror#24
  1079. ldr $t1,[$key,#-12]
  1080. eor $s0,$s0,$i1
  1081. ldr $t2,[$key,#-8]
  1082. eor $s3,$s3,$t3,ror#8
  1083. ldr $t3,[$key,#-4]
  1084. and $i1,lr,$s0,lsr#16
  1085. eor $s1,$s1,$t1
  1086. and $i2,lr,$s0,lsr#8
  1087. eor $s2,$s2,$t2
  1088. and $i3,lr,$s0
  1089. eor $s3,$s3,$t3
  1090. mov $s0,$s0,lsr#24
  1091. subs $rounds,$rounds,#1
  1092. bne .Ldec_loop
  1093. add $tbl,$tbl,#1024
  1094. ldr $t2,[$tbl,#0] @ prefetch Td4
  1095. ldr $t3,[$tbl,#32]
  1096. ldr $t1,[$tbl,#64]
  1097. ldr $t2,[$tbl,#96]
  1098. ldr $t3,[$tbl,#128]
  1099. ldr $t1,[$tbl,#160]
  1100. ldr $t2,[$tbl,#192]
  1101. ldr $t3,[$tbl,#224]
  1102. ldrb $s0,[$tbl,$s0] @ Td4[s0>>24]
  1103. ldrb $t1,[$tbl,$i1] @ Td4[s0>>16]
  1104. and $i1,lr,$s1 @ i0
  1105. ldrb $t2,[$tbl,$i2] @ Td4[s0>>8]
  1106. and $i2,lr,$s1,lsr#16
  1107. ldrb $t3,[$tbl,$i3] @ Td4[s0>>0]
  1108. and $i3,lr,$s1,lsr#8
  1109. add $s1,$tbl,$s1,lsr#24
  1110. ldrb $i1,[$tbl,$i1] @ Td4[s1>>0]
  1111. ldrb $s1,[$s1] @ Td4[s1>>24]
  1112. ldrb $i2,[$tbl,$i2] @ Td4[s1>>16]
  1113. eor $s0,$i1,$s0,lsl#24
  1114. ldrb $i3,[$tbl,$i3] @ Td4[s1>>8]
  1115. eor $s1,$t1,$s1,lsl#8
  1116. and $i1,lr,$s2,lsr#8 @ i0
  1117. eor $t2,$t2,$i2,lsl#8
  1118. and $i2,lr,$s2 @ i1
  1119. ldrb $i1,[$tbl,$i1] @ Td4[s2>>8]
  1120. eor $t3,$t3,$i3,lsl#8
  1121. ldrb $i2,[$tbl,$i2] @ Td4[s2>>0]
  1122. and $i3,lr,$s2,lsr#16
  1123. add $s2,$tbl,$s2,lsr#24
  1124. ldrb $s2,[$s2] @ Td4[s2>>24]
  1125. eor $s0,$s0,$i1,lsl#8
  1126. ldrb $i3,[$tbl,$i3] @ Td4[s2>>16]
  1127. eor $s1,$i2,$s1,lsl#16
  1128. and $i1,lr,$s3,lsr#16 @ i0
  1129. eor $s2,$t2,$s2,lsl#16
  1130. and $i2,lr,$s3,lsr#8 @ i1
  1131. ldrb $i1,[$tbl,$i1] @ Td4[s3>>16]
  1132. eor $t3,$t3,$i3,lsl#16
  1133. ldrb $i2,[$tbl,$i2] @ Td4[s3>>8]
  1134. and $i3,lr,$s3 @ i2
  1135. add $s3,$tbl,$s3,lsr#24
  1136. ldrb $i3,[$tbl,$i3] @ Td4[s3>>0]
  1137. ldrb $s3,[$s3] @ Td4[s3>>24]
  1138. eor $s0,$s0,$i1,lsl#16
  1139. ldr $i1,[$key,#0]
  1140. eor $s1,$s1,$i2,lsl#8
  1141. ldr $t1,[$key,#4]
  1142. eor $s2,$i3,$s2,lsl#8
  1143. ldr $t2,[$key,#8]
  1144. eor $s3,$t3,$s3,lsl#24
  1145. ldr $t3,[$key,#12]
  1146. eor $s0,$s0,$i1
  1147. eor $s1,$s1,$t1
  1148. eor $s2,$s2,$t2
  1149. eor $s3,$s3,$t3
  1150. sub $tbl,$tbl,#1024
  1151. ldr pc,[sp],#4 @ pop and return
  1152. .size _armv4_AES_decrypt,.-_armv4_AES_decrypt
  1153. .asciz "AES for ARMv4, CRYPTOGAMS by <appro\@openssl.org>"
  1154. .align 2
  1155. ___
  1156. $code =~ s/\bbx\s+lr\b/.word\t0xe12fff1e/gm; # make it possible to compile with -march=armv4
  1157. $code =~ s/\bret\b/bx\tlr/gm;
  1158. open SELF,$0;
  1159. while(<SELF>) {
  1160. next if (/^#!/);
  1161. last if (!s/^#/@/ and !/^$/);
  1162. print;
  1163. }
  1164. close SELF;
  1165. print $code;
  1166. close STDOUT or die "error closing STDOUT: $!"; # enforce flush