• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1@ Copyright 2007-2020 The OpenSSL Project Authors. All Rights Reserved.
2@
3@ Licensed under the OpenSSL license (the "License").  You may not use
4@ this file except in compliance with the License.  You can obtain a copy
5@ in the file LICENSE in the source distribution or at
6@ https://www.openssl.org/source/license.html
7
8
9@ ====================================================================
10@ Written by Andy Polyakov <appro@openssl.org> for the OpenSSL
11@ project. The module is, however, dual licensed under OpenSSL and
12@ CRYPTOGAMS licenses depending on where you obtain it. For further
13@ details see http://www.openssl.org/~appro/cryptogams/.
14@ ====================================================================
15
16@ AES for ARMv4
17
18@ January 2007.
19@
20@ Code uses single 1K S-box and is >2 times faster than code generated
21@ by gcc-3.4.1. This is thanks to unique feature of ARMv4 ISA, which
22@ allows to merge logical or arithmetic operation with shift or rotate
23@ in one instruction and emit combined result every cycle. The module
24@ is endian-neutral. The performance is ~42 cycles/byte for 128-bit
25@ key [on single-issue Xscale PXA250 core].
26
27@ May 2007.
28@
29@ AES_set_[en|de]crypt_key is added.
30
31@ July 2010.
32@
33@ Rescheduling for dual-issue pipeline resulted in 12% improvement on
34@ Cortex A8 core and ~25 cycles per byte processed with 128-bit key.
35
36@ February 2011.
37@
38@ Profiler-assisted and platform-specific optimization resulted in 16%
39@ improvement on Cortex A8 core and ~21.5 cycles per byte.
40
41#ifndef __KERNEL__
42# include "arm_arch.h"
43#else
44# define __ARM_ARCH__ __LINUX_ARM_ARCH__
45#endif
46
47.text
48#if defined(__thumb2__) && !defined(__APPLE__)
49.syntax	unified
50.thumb
51#else
52.code	32
53#undef __thumb2__
54#endif
55
56.type	AES_Te,%object
57.align	5
58AES_Te:
59.word	0xc66363a5, 0xf87c7c84, 0xee777799, 0xf67b7b8d
60.word	0xfff2f20d, 0xd66b6bbd, 0xde6f6fb1, 0x91c5c554
61.word	0x60303050, 0x02010103, 0xce6767a9, 0x562b2b7d
62.word	0xe7fefe19, 0xb5d7d762, 0x4dababe6, 0xec76769a
63.word	0x8fcaca45, 0x1f82829d, 0x89c9c940, 0xfa7d7d87
64.word	0xeffafa15, 0xb25959eb, 0x8e4747c9, 0xfbf0f00b
65.word	0x41adadec, 0xb3d4d467, 0x5fa2a2fd, 0x45afafea
66.word	0x239c9cbf, 0x53a4a4f7, 0xe4727296, 0x9bc0c05b
67.word	0x75b7b7c2, 0xe1fdfd1c, 0x3d9393ae, 0x4c26266a
68.word	0x6c36365a, 0x7e3f3f41, 0xf5f7f702, 0x83cccc4f
69.word	0x6834345c, 0x51a5a5f4, 0xd1e5e534, 0xf9f1f108
70.word	0xe2717193, 0xabd8d873, 0x62313153, 0x2a15153f
71.word	0x0804040c, 0x95c7c752, 0x46232365, 0x9dc3c35e
72.word	0x30181828, 0x379696a1, 0x0a05050f, 0x2f9a9ab5
73.word	0x0e070709, 0x24121236, 0x1b80809b, 0xdfe2e23d
74.word	0xcdebeb26, 0x4e272769, 0x7fb2b2cd, 0xea75759f
75.word	0x1209091b, 0x1d83839e, 0x582c2c74, 0x341a1a2e
76.word	0x361b1b2d, 0xdc6e6eb2, 0xb45a5aee, 0x5ba0a0fb
77.word	0xa45252f6, 0x763b3b4d, 0xb7d6d661, 0x7db3b3ce
78.word	0x5229297b, 0xdde3e33e, 0x5e2f2f71, 0x13848497
79.word	0xa65353f5, 0xb9d1d168, 0x00000000, 0xc1eded2c
80.word	0x40202060, 0xe3fcfc1f, 0x79b1b1c8, 0xb65b5bed
81.word	0xd46a6abe, 0x8dcbcb46, 0x67bebed9, 0x7239394b
82.word	0x944a4ade, 0x984c4cd4, 0xb05858e8, 0x85cfcf4a
83.word	0xbbd0d06b, 0xc5efef2a, 0x4faaaae5, 0xedfbfb16
84.word	0x864343c5, 0x9a4d4dd7, 0x66333355, 0x11858594
85.word	0x8a4545cf, 0xe9f9f910, 0x04020206, 0xfe7f7f81
86.word	0xa05050f0, 0x783c3c44, 0x259f9fba, 0x4ba8a8e3
87.word	0xa25151f3, 0x5da3a3fe, 0x804040c0, 0x058f8f8a
88.word	0x3f9292ad, 0x219d9dbc, 0x70383848, 0xf1f5f504
89.word	0x63bcbcdf, 0x77b6b6c1, 0xafdada75, 0x42212163
90.word	0x20101030, 0xe5ffff1a, 0xfdf3f30e, 0xbfd2d26d
91.word	0x81cdcd4c, 0x180c0c14, 0x26131335, 0xc3ecec2f
92.word	0xbe5f5fe1, 0x359797a2, 0x884444cc, 0x2e171739
93.word	0x93c4c457, 0x55a7a7f2, 0xfc7e7e82, 0x7a3d3d47
94.word	0xc86464ac, 0xba5d5de7, 0x3219192b, 0xe6737395
95.word	0xc06060a0, 0x19818198, 0x9e4f4fd1, 0xa3dcdc7f
96.word	0x44222266, 0x542a2a7e, 0x3b9090ab, 0x0b888883
97.word	0x8c4646ca, 0xc7eeee29, 0x6bb8b8d3, 0x2814143c
98.word	0xa7dede79, 0xbc5e5ee2, 0x160b0b1d, 0xaddbdb76
99.word	0xdbe0e03b, 0x64323256, 0x743a3a4e, 0x140a0a1e
100.word	0x924949db, 0x0c06060a, 0x4824246c, 0xb85c5ce4
101.word	0x9fc2c25d, 0xbdd3d36e, 0x43acacef, 0xc46262a6
102.word	0x399191a8, 0x319595a4, 0xd3e4e437, 0xf279798b
103.word	0xd5e7e732, 0x8bc8c843, 0x6e373759, 0xda6d6db7
104.word	0x018d8d8c, 0xb1d5d564, 0x9c4e4ed2, 0x49a9a9e0
105.word	0xd86c6cb4, 0xac5656fa, 0xf3f4f407, 0xcfeaea25
106.word	0xca6565af, 0xf47a7a8e, 0x47aeaee9, 0x10080818
107.word	0x6fbabad5, 0xf0787888, 0x4a25256f, 0x5c2e2e72
108.word	0x381c1c24, 0x57a6a6f1, 0x73b4b4c7, 0x97c6c651
109.word	0xcbe8e823, 0xa1dddd7c, 0xe874749c, 0x3e1f1f21
110.word	0x964b4bdd, 0x61bdbddc, 0x0d8b8b86, 0x0f8a8a85
111.word	0xe0707090, 0x7c3e3e42, 0x71b5b5c4, 0xcc6666aa
112.word	0x904848d8, 0x06030305, 0xf7f6f601, 0x1c0e0e12
113.word	0xc26161a3, 0x6a35355f, 0xae5757f9, 0x69b9b9d0
114.word	0x17868691, 0x99c1c158, 0x3a1d1d27, 0x279e9eb9
115.word	0xd9e1e138, 0xebf8f813, 0x2b9898b3, 0x22111133
116.word	0xd26969bb, 0xa9d9d970, 0x078e8e89, 0x339494a7
117.word	0x2d9b9bb6, 0x3c1e1e22, 0x15878792, 0xc9e9e920
118.word	0x87cece49, 0xaa5555ff, 0x50282878, 0xa5dfdf7a
119.word	0x038c8c8f, 0x59a1a1f8, 0x09898980, 0x1a0d0d17
120.word	0x65bfbfda, 0xd7e6e631, 0x844242c6, 0xd06868b8
121.word	0x824141c3, 0x299999b0, 0x5a2d2d77, 0x1e0f0f11
122.word	0x7bb0b0cb, 0xa85454fc, 0x6dbbbbd6, 0x2c16163a
123@ Te4[256]
124.byte	0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5
125.byte	0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76
126.byte	0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0
127.byte	0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0
128.byte	0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc
129.byte	0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15
130.byte	0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a
131.byte	0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75
132.byte	0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0
133.byte	0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84
134.byte	0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b
135.byte	0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf
136.byte	0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85
137.byte	0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8
138.byte	0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5
139.byte	0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2
140.byte	0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17
141.byte	0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73
142.byte	0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88
143.byte	0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb
144.byte	0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c
145.byte	0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79
146.byte	0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9
147.byte	0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08
148.byte	0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6
149.byte	0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a
150.byte	0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e
151.byte	0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e
152.byte	0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94
153.byte	0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf
154.byte	0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68
155.byte	0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16
156@ rcon[]
157.word	0x01000000, 0x02000000, 0x04000000, 0x08000000
158.word	0x10000000, 0x20000000, 0x40000000, 0x80000000
159.word	0x1B000000, 0x36000000, 0, 0, 0, 0, 0, 0
160.size	AES_Te,.-AES_Te
161
162@ void AES_encrypt(const unsigned char *in, unsigned char *out,
163@ 		 const AES_KEY *key) {
164.global AES_encrypt
165.type   AES_encrypt,%function
166.align	5
167AES_encrypt:
168#ifndef	__thumb2__
169	sub	r3,pc,#8		@ AES_encrypt
170#else
171	adr	r3,.
172#endif
173	stmdb   sp!,{r1,r4-r12,lr}
174#if defined(__thumb2__) || defined(__APPLE__)
175	adr	r10,AES_Te
176#else
177	sub	r10,r3,#AES_encrypt-AES_Te	@ Te
178#endif
179	mov	r12,r0		@ inp
180	mov	r11,r2
181#if __ARM_ARCH__<7
182	ldrb	r0,[r12,#3]	@ load input data in endian-neutral
183	ldrb	r4,[r12,#2]	@ manner...
184	ldrb	r5,[r12,#1]
185	ldrb	r6,[r12,#0]
186	orr	r0,r0,r4,lsl#8
187	ldrb	r1,[r12,#7]
188	orr	r0,r0,r5,lsl#16
189	ldrb	r4,[r12,#6]
190	orr	r0,r0,r6,lsl#24
191	ldrb	r5,[r12,#5]
192	ldrb	r6,[r12,#4]
193	orr	r1,r1,r4,lsl#8
194	ldrb	r2,[r12,#11]
195	orr	r1,r1,r5,lsl#16
196	ldrb	r4,[r12,#10]
197	orr	r1,r1,r6,lsl#24
198	ldrb	r5,[r12,#9]
199	ldrb	r6,[r12,#8]
200	orr	r2,r2,r4,lsl#8
201	ldrb	r3,[r12,#15]
202	orr	r2,r2,r5,lsl#16
203	ldrb	r4,[r12,#14]
204	orr	r2,r2,r6,lsl#24
205	ldrb	r5,[r12,#13]
206	ldrb	r6,[r12,#12]
207	orr	r3,r3,r4,lsl#8
208	orr	r3,r3,r5,lsl#16
209	orr	r3,r3,r6,lsl#24
210#else
211	ldr	r0,[r12,#0]
212	ldr	r1,[r12,#4]
213	ldr	r2,[r12,#8]
214	ldr	r3,[r12,#12]
215#ifdef __ARMEL__
216	rev	r0,r0
217	rev	r1,r1
218	rev	r2,r2
219	rev	r3,r3
220#endif
221#endif
222	bl	_armv4_AES_encrypt
223
224	ldr	r12,[sp],#4		@ pop out
225#if __ARM_ARCH__>=7
226#ifdef __ARMEL__
227	rev	r0,r0
228	rev	r1,r1
229	rev	r2,r2
230	rev	r3,r3
231#endif
232	str	r0,[r12,#0]
233	str	r1,[r12,#4]
234	str	r2,[r12,#8]
235	str	r3,[r12,#12]
236#else
237	mov	r4,r0,lsr#24		@ write output in endian-neutral
238	mov	r5,r0,lsr#16		@ manner...
239	mov	r6,r0,lsr#8
240	strb	r4,[r12,#0]
241	strb	r5,[r12,#1]
242	mov	r4,r1,lsr#24
243	strb	r6,[r12,#2]
244	mov	r5,r1,lsr#16
245	strb	r0,[r12,#3]
246	mov	r6,r1,lsr#8
247	strb	r4,[r12,#4]
248	strb	r5,[r12,#5]
249	mov	r4,r2,lsr#24
250	strb	r6,[r12,#6]
251	mov	r5,r2,lsr#16
252	strb	r1,[r12,#7]
253	mov	r6,r2,lsr#8
254	strb	r4,[r12,#8]
255	strb	r5,[r12,#9]
256	mov	r4,r3,lsr#24
257	strb	r6,[r12,#10]
258	mov	r5,r3,lsr#16
259	strb	r2,[r12,#11]
260	mov	r6,r3,lsr#8
261	strb	r4,[r12,#12]
262	strb	r5,[r12,#13]
263	strb	r6,[r12,#14]
264	strb	r3,[r12,#15]
265#endif
266#if __ARM_ARCH__>=5
267	ldmia	sp!,{r4-r12,pc}
268#else
269	ldmia   sp!,{r4-r12,lr}
270	tst	lr,#1
271	moveq	pc,lr			@ be binary compatible with V4, yet
272	.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
273#endif
274.size	AES_encrypt,.-AES_encrypt
275
276.type   _armv4_AES_encrypt,%function
277.align	2
278_armv4_AES_encrypt:
279	str	lr,[sp,#-4]!		@ push lr
280	ldmia	r11!,{r4-r7}
281	eor	r0,r0,r4
282	ldr	r12,[r11,#240-16]
283	eor	r1,r1,r5
284	eor	r2,r2,r6
285	eor	r3,r3,r7
286	sub	r12,r12,#1
287	mov	lr,#255
288
289	and	r7,lr,r0
290	and	r8,lr,r0,lsr#8
291	and	r9,lr,r0,lsr#16
292	mov	r0,r0,lsr#24
293.Lenc_loop:
294	ldr	r4,[r10,r7,lsl#2]	@ Te3[s0>>0]
295	and	r7,lr,r1,lsr#16	@ i0
296	ldr	r5,[r10,r8,lsl#2]	@ Te2[s0>>8]
297	and	r8,lr,r1
298	ldr	r6,[r10,r9,lsl#2]	@ Te1[s0>>16]
299	and	r9,lr,r1,lsr#8
300	ldr	r0,[r10,r0,lsl#2]	@ Te0[s0>>24]
301	mov	r1,r1,lsr#24
302
303	ldr	r7,[r10,r7,lsl#2]	@ Te1[s1>>16]
304	ldr	r8,[r10,r8,lsl#2]	@ Te3[s1>>0]
305	ldr	r9,[r10,r9,lsl#2]	@ Te2[s1>>8]
306	eor	r0,r0,r7,ror#8
307	ldr	r1,[r10,r1,lsl#2]	@ Te0[s1>>24]
308	and	r7,lr,r2,lsr#8	@ i0
309	eor	r5,r5,r8,ror#8
310	and	r8,lr,r2,lsr#16	@ i1
311	eor	r6,r6,r9,ror#8
312	and	r9,lr,r2
313	ldr	r7,[r10,r7,lsl#2]	@ Te2[s2>>8]
314	eor	r1,r1,r4,ror#24
315	ldr	r8,[r10,r8,lsl#2]	@ Te1[s2>>16]
316	mov	r2,r2,lsr#24
317
318	ldr	r9,[r10,r9,lsl#2]	@ Te3[s2>>0]
319	eor	r0,r0,r7,ror#16
320	ldr	r2,[r10,r2,lsl#2]	@ Te0[s2>>24]
321	and	r7,lr,r3		@ i0
322	eor	r1,r1,r8,ror#8
323	and	r8,lr,r3,lsr#8	@ i1
324	eor	r6,r6,r9,ror#16
325	and	r9,lr,r3,lsr#16	@ i2
326	ldr	r7,[r10,r7,lsl#2]	@ Te3[s3>>0]
327	eor	r2,r2,r5,ror#16
328	ldr	r8,[r10,r8,lsl#2]	@ Te2[s3>>8]
329	mov	r3,r3,lsr#24
330
331	ldr	r9,[r10,r9,lsl#2]	@ Te1[s3>>16]
332	eor	r0,r0,r7,ror#24
333	ldr	r7,[r11],#16
334	eor	r1,r1,r8,ror#16
335	ldr	r3,[r10,r3,lsl#2]	@ Te0[s3>>24]
336	eor	r2,r2,r9,ror#8
337	ldr	r4,[r11,#-12]
338	eor	r3,r3,r6,ror#8
339
340	ldr	r5,[r11,#-8]
341	eor	r0,r0,r7
342	ldr	r6,[r11,#-4]
343	and	r7,lr,r0
344	eor	r1,r1,r4
345	and	r8,lr,r0,lsr#8
346	eor	r2,r2,r5
347	and	r9,lr,r0,lsr#16
348	eor	r3,r3,r6
349	mov	r0,r0,lsr#24
350
351	subs	r12,r12,#1
352	bne	.Lenc_loop
353
354	add	r10,r10,#2
355
356	ldrb	r4,[r10,r7,lsl#2]	@ Te4[s0>>0]
357	and	r7,lr,r1,lsr#16	@ i0
358	ldrb	r5,[r10,r8,lsl#2]	@ Te4[s0>>8]
359	and	r8,lr,r1
360	ldrb	r6,[r10,r9,lsl#2]	@ Te4[s0>>16]
361	and	r9,lr,r1,lsr#8
362	ldrb	r0,[r10,r0,lsl#2]	@ Te4[s0>>24]
363	mov	r1,r1,lsr#24
364
365	ldrb	r7,[r10,r7,lsl#2]	@ Te4[s1>>16]
366	ldrb	r8,[r10,r8,lsl#2]	@ Te4[s1>>0]
367	ldrb	r9,[r10,r9,lsl#2]	@ Te4[s1>>8]
368	eor	r0,r7,r0,lsl#8
369	ldrb	r1,[r10,r1,lsl#2]	@ Te4[s1>>24]
370	and	r7,lr,r2,lsr#8	@ i0
371	eor	r5,r8,r5,lsl#8
372	and	r8,lr,r2,lsr#16	@ i1
373	eor	r6,r9,r6,lsl#8
374	and	r9,lr,r2
375	ldrb	r7,[r10,r7,lsl#2]	@ Te4[s2>>8]
376	eor	r1,r4,r1,lsl#24
377	ldrb	r8,[r10,r8,lsl#2]	@ Te4[s2>>16]
378	mov	r2,r2,lsr#24
379
380	ldrb	r9,[r10,r9,lsl#2]	@ Te4[s2>>0]
381	eor	r0,r7,r0,lsl#8
382	ldrb	r2,[r10,r2,lsl#2]	@ Te4[s2>>24]
383	and	r7,lr,r3		@ i0
384	eor	r1,r1,r8,lsl#16
385	and	r8,lr,r3,lsr#8	@ i1
386	eor	r6,r9,r6,lsl#8
387	and	r9,lr,r3,lsr#16	@ i2
388	ldrb	r7,[r10,r7,lsl#2]	@ Te4[s3>>0]
389	eor	r2,r5,r2,lsl#24
390	ldrb	r8,[r10,r8,lsl#2]	@ Te4[s3>>8]
391	mov	r3,r3,lsr#24
392
393	ldrb	r9,[r10,r9,lsl#2]	@ Te4[s3>>16]
394	eor	r0,r7,r0,lsl#8
395	ldr	r7,[r11,#0]
396	ldrb	r3,[r10,r3,lsl#2]	@ Te4[s3>>24]
397	eor	r1,r1,r8,lsl#8
398	ldr	r4,[r11,#4]
399	eor	r2,r2,r9,lsl#16
400	ldr	r5,[r11,#8]
401	eor	r3,r6,r3,lsl#24
402	ldr	r6,[r11,#12]
403
404	eor	r0,r0,r7
405	eor	r1,r1,r4
406	eor	r2,r2,r5
407	eor	r3,r3,r6
408
409	sub	r10,r10,#2
410	ldr	pc,[sp],#4		@ pop and return
411.size	_armv4_AES_encrypt,.-_armv4_AES_encrypt
412
413.global AES_set_encrypt_key
414.type   AES_set_encrypt_key,%function
415.align	5
416AES_set_encrypt_key:
417_armv4_AES_set_encrypt_key:
418#ifndef	__thumb2__
419	sub	r3,pc,#8		@ AES_set_encrypt_key
420#else
421	adr	r3,.
422#endif
423	teq	r0,#0
424#ifdef	__thumb2__
425	itt	eq			@ Thumb2 thing, sanity check in ARM
426#endif
427	moveq	r0,#-1
428	beq	.Labrt
429	teq	r2,#0
430#ifdef	__thumb2__
431	itt	eq			@ Thumb2 thing, sanity check in ARM
432#endif
433	moveq	r0,#-1
434	beq	.Labrt
435
436	teq	r1,#128
437	beq	.Lok
438	teq	r1,#192
439	beq	.Lok
440	teq	r1,#256
441#ifdef	__thumb2__
442	itt	ne			@ Thumb2 thing, sanity check in ARM
443#endif
444	movne	r0,#-1
445	bne	.Labrt
446
447.Lok:	stmdb   sp!,{r4-r12,lr}
448	mov	r12,r0		@ inp
449	mov	lr,r1			@ bits
450	mov	r11,r2			@ key
451
452#if defined(__thumb2__) || defined(__APPLE__)
453	adr	r10,AES_Te+1024				@ Te4
454#else
455	sub	r10,r3,#_armv4_AES_set_encrypt_key-AES_Te-1024	@ Te4
456#endif
457
458#if __ARM_ARCH__<7
459	ldrb	r0,[r12,#3]	@ load input data in endian-neutral
460	ldrb	r4,[r12,#2]	@ manner...
461	ldrb	r5,[r12,#1]
462	ldrb	r6,[r12,#0]
463	orr	r0,r0,r4,lsl#8
464	ldrb	r1,[r12,#7]
465	orr	r0,r0,r5,lsl#16
466	ldrb	r4,[r12,#6]
467	orr	r0,r0,r6,lsl#24
468	ldrb	r5,[r12,#5]
469	ldrb	r6,[r12,#4]
470	orr	r1,r1,r4,lsl#8
471	ldrb	r2,[r12,#11]
472	orr	r1,r1,r5,lsl#16
473	ldrb	r4,[r12,#10]
474	orr	r1,r1,r6,lsl#24
475	ldrb	r5,[r12,#9]
476	ldrb	r6,[r12,#8]
477	orr	r2,r2,r4,lsl#8
478	ldrb	r3,[r12,#15]
479	orr	r2,r2,r5,lsl#16
480	ldrb	r4,[r12,#14]
481	orr	r2,r2,r6,lsl#24
482	ldrb	r5,[r12,#13]
483	ldrb	r6,[r12,#12]
484	orr	r3,r3,r4,lsl#8
485	str	r0,[r11],#16
486	orr	r3,r3,r5,lsl#16
487	str	r1,[r11,#-12]
488	orr	r3,r3,r6,lsl#24
489	str	r2,[r11,#-8]
490	str	r3,[r11,#-4]
491#else
492	ldr	r0,[r12,#0]
493	ldr	r1,[r12,#4]
494	ldr	r2,[r12,#8]
495	ldr	r3,[r12,#12]
496#ifdef __ARMEL__
497	rev	r0,r0
498	rev	r1,r1
499	rev	r2,r2
500	rev	r3,r3
501#endif
502	str	r0,[r11],#16
503	str	r1,[r11,#-12]
504	str	r2,[r11,#-8]
505	str	r3,[r11,#-4]
506#endif
507
508	teq	lr,#128
509	bne	.Lnot128
510	mov	r12,#10
511	str	r12,[r11,#240-16]
512	add	r6,r10,#256			@ rcon
513	mov	lr,#255
514
515.L128_loop:
516	and	r5,lr,r3,lsr#24
517	and	r7,lr,r3,lsr#16
518	ldrb	r5,[r10,r5]
519	and	r8,lr,r3,lsr#8
520	ldrb	r7,[r10,r7]
521	and	r9,lr,r3
522	ldrb	r8,[r10,r8]
523	orr	r5,r5,r7,lsl#24
524	ldrb	r9,[r10,r9]
525	orr	r5,r5,r8,lsl#16
526	ldr	r4,[r6],#4			@ rcon[i++]
527	orr	r5,r5,r9,lsl#8
528	eor	r5,r5,r4
529	eor	r0,r0,r5			@ rk[4]=rk[0]^...
530	eor	r1,r1,r0			@ rk[5]=rk[1]^rk[4]
531	str	r0,[r11],#16
532	eor	r2,r2,r1			@ rk[6]=rk[2]^rk[5]
533	str	r1,[r11,#-12]
534	eor	r3,r3,r2			@ rk[7]=rk[3]^rk[6]
535	str	r2,[r11,#-8]
536	subs	r12,r12,#1
537	str	r3,[r11,#-4]
538	bne	.L128_loop
539	sub	r2,r11,#176
540	b	.Ldone
541
542.Lnot128:
543#if __ARM_ARCH__<7
544	ldrb	r8,[r12,#19]
545	ldrb	r4,[r12,#18]
546	ldrb	r5,[r12,#17]
547	ldrb	r6,[r12,#16]
548	orr	r8,r8,r4,lsl#8
549	ldrb	r9,[r12,#23]
550	orr	r8,r8,r5,lsl#16
551	ldrb	r4,[r12,#22]
552	orr	r8,r8,r6,lsl#24
553	ldrb	r5,[r12,#21]
554	ldrb	r6,[r12,#20]
555	orr	r9,r9,r4,lsl#8
556	orr	r9,r9,r5,lsl#16
557	str	r8,[r11],#8
558	orr	r9,r9,r6,lsl#24
559	str	r9,[r11,#-4]
560#else
561	ldr	r8,[r12,#16]
562	ldr	r9,[r12,#20]
563#ifdef __ARMEL__
564	rev	r8,r8
565	rev	r9,r9
566#endif
567	str	r8,[r11],#8
568	str	r9,[r11,#-4]
569#endif
570
571	teq	lr,#192
572	bne	.Lnot192
573	mov	r12,#12
574	str	r12,[r11,#240-24]
575	add	r6,r10,#256			@ rcon
576	mov	lr,#255
577	mov	r12,#8
578
579.L192_loop:
580	and	r5,lr,r9,lsr#24
581	and	r7,lr,r9,lsr#16
582	ldrb	r5,[r10,r5]
583	and	r8,lr,r9,lsr#8
584	ldrb	r7,[r10,r7]
585	and	r9,lr,r9
586	ldrb	r8,[r10,r8]
587	orr	r5,r5,r7,lsl#24
588	ldrb	r9,[r10,r9]
589	orr	r5,r5,r8,lsl#16
590	ldr	r4,[r6],#4			@ rcon[i++]
591	orr	r5,r5,r9,lsl#8
592	eor	r9,r5,r4
593	eor	r0,r0,r9			@ rk[6]=rk[0]^...
594	eor	r1,r1,r0			@ rk[7]=rk[1]^rk[6]
595	str	r0,[r11],#24
596	eor	r2,r2,r1			@ rk[8]=rk[2]^rk[7]
597	str	r1,[r11,#-20]
598	eor	r3,r3,r2			@ rk[9]=rk[3]^rk[8]
599	str	r2,[r11,#-16]
600	subs	r12,r12,#1
601	str	r3,[r11,#-12]
602#ifdef	__thumb2__
603	itt	eq				@ Thumb2 thing, sanity check in ARM
604#endif
605	subeq	r2,r11,#216
606	beq	.Ldone
607
608	ldr	r7,[r11,#-32]
609	ldr	r8,[r11,#-28]
610	eor	r7,r7,r3			@ rk[10]=rk[4]^rk[9]
611	eor	r9,r8,r7			@ rk[11]=rk[5]^rk[10]
612	str	r7,[r11,#-8]
613	str	r9,[r11,#-4]
614	b	.L192_loop
615
616.Lnot192:
617#if __ARM_ARCH__<7
618	ldrb	r8,[r12,#27]
619	ldrb	r4,[r12,#26]
620	ldrb	r5,[r12,#25]
621	ldrb	r6,[r12,#24]
622	orr	r8,r8,r4,lsl#8
623	ldrb	r9,[r12,#31]
624	orr	r8,r8,r5,lsl#16
625	ldrb	r4,[r12,#30]
626	orr	r8,r8,r6,lsl#24
627	ldrb	r5,[r12,#29]
628	ldrb	r6,[r12,#28]
629	orr	r9,r9,r4,lsl#8
630	orr	r9,r9,r5,lsl#16
631	str	r8,[r11],#8
632	orr	r9,r9,r6,lsl#24
633	str	r9,[r11,#-4]
634#else
635	ldr	r8,[r12,#24]
636	ldr	r9,[r12,#28]
637#ifdef __ARMEL__
638	rev	r8,r8
639	rev	r9,r9
640#endif
641	str	r8,[r11],#8
642	str	r9,[r11,#-4]
643#endif
644
645	mov	r12,#14
646	str	r12,[r11,#240-32]
647	add	r6,r10,#256			@ rcon
648	mov	lr,#255
649	mov	r12,#7
650
651.L256_loop:
652	and	r5,lr,r9,lsr#24
653	and	r7,lr,r9,lsr#16
654	ldrb	r5,[r10,r5]
655	and	r8,lr,r9,lsr#8
656	ldrb	r7,[r10,r7]
657	and	r9,lr,r9
658	ldrb	r8,[r10,r8]
659	orr	r5,r5,r7,lsl#24
660	ldrb	r9,[r10,r9]
661	orr	r5,r5,r8,lsl#16
662	ldr	r4,[r6],#4			@ rcon[i++]
663	orr	r5,r5,r9,lsl#8
664	eor	r9,r5,r4
665	eor	r0,r0,r9			@ rk[8]=rk[0]^...
666	eor	r1,r1,r0			@ rk[9]=rk[1]^rk[8]
667	str	r0,[r11],#32
668	eor	r2,r2,r1			@ rk[10]=rk[2]^rk[9]
669	str	r1,[r11,#-28]
670	eor	r3,r3,r2			@ rk[11]=rk[3]^rk[10]
671	str	r2,[r11,#-24]
672	subs	r12,r12,#1
673	str	r3,[r11,#-20]
674#ifdef	__thumb2__
675	itt	eq				@ Thumb2 thing, sanity check in ARM
676#endif
677	subeq	r2,r11,#256
678	beq	.Ldone
679
680	and	r5,lr,r3
681	and	r7,lr,r3,lsr#8
682	ldrb	r5,[r10,r5]
683	and	r8,lr,r3,lsr#16
684	ldrb	r7,[r10,r7]
685	and	r9,lr,r3,lsr#24
686	ldrb	r8,[r10,r8]
687	orr	r5,r5,r7,lsl#8
688	ldrb	r9,[r10,r9]
689	orr	r5,r5,r8,lsl#16
690	ldr	r4,[r11,#-48]
691	orr	r5,r5,r9,lsl#24
692
693	ldr	r7,[r11,#-44]
694	ldr	r8,[r11,#-40]
695	eor	r4,r4,r5			@ rk[12]=rk[4]^...
696	ldr	r9,[r11,#-36]
697	eor	r7,r7,r4			@ rk[13]=rk[5]^rk[12]
698	str	r4,[r11,#-16]
699	eor	r8,r8,r7			@ rk[14]=rk[6]^rk[13]
700	str	r7,[r11,#-12]
701	eor	r9,r9,r8			@ rk[15]=rk[7]^rk[14]
702	str	r8,[r11,#-8]
703	str	r9,[r11,#-4]
704	b	.L256_loop
705
706.align	2
707.Ldone:	mov	r0,#0
708	ldmia   sp!,{r4-r12,lr}
709.Labrt:
710#if __ARM_ARCH__>=5
711	bx	lr				@ .word	0xe12fff1e
712#else
713	tst	lr,#1
714	moveq	pc,lr			@ be binary compatible with V4, yet
715	.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
716#endif
717.size	AES_set_encrypt_key,.-AES_set_encrypt_key
718
719.global AES_set_decrypt_key
720.type   AES_set_decrypt_key,%function
721.align	5
722AES_set_decrypt_key:
723	str	lr,[sp,#-4]!            @ push lr
724	bl	_armv4_AES_set_encrypt_key
725	teq	r0,#0
726	ldr	lr,[sp],#4              @ pop lr
727	bne	.Labrt
728
729	mov	r0,r2			@ AES_set_encrypt_key preserves r2,
730	mov	r1,r2			@ which is AES_KEY *key
731	b	_armv4_AES_set_enc2dec_key
732.size	AES_set_decrypt_key,.-AES_set_decrypt_key
733
734@ void AES_set_enc2dec_key(const AES_KEY *inp,AES_KEY *out)
735.global	AES_set_enc2dec_key
736.type	AES_set_enc2dec_key,%function
737.align	5
738AES_set_enc2dec_key:
739_armv4_AES_set_enc2dec_key:
740	stmdb   sp!,{r4-r12,lr}
741
742	ldr	r12,[r0,#240]
743	mov	r7,r0			@ input
744	add	r8,r0,r12,lsl#4
745	mov	r11,r1			@ output
746	add	r10,r1,r12,lsl#4
747	str	r12,[r1,#240]
748
749.Linv:	ldr	r0,[r7],#16
750	ldr	r1,[r7,#-12]
751	ldr	r2,[r7,#-8]
752	ldr	r3,[r7,#-4]
753	ldr	r4,[r8],#-16
754	ldr	r5,[r8,#16+4]
755	ldr	r6,[r8,#16+8]
756	ldr	r9,[r8,#16+12]
757	str	r0,[r10],#-16
758	str	r1,[r10,#16+4]
759	str	r2,[r10,#16+8]
760	str	r3,[r10,#16+12]
761	str	r4,[r11],#16
762	str	r5,[r11,#-12]
763	str	r6,[r11,#-8]
764	str	r9,[r11,#-4]
765	teq	r7,r8
766	bne	.Linv
767
768	ldr	r0,[r7]
769	ldr	r1,[r7,#4]
770	ldr	r2,[r7,#8]
771	ldr	r3,[r7,#12]
772	str	r0,[r11]
773	str	r1,[r11,#4]
774	str	r2,[r11,#8]
775	str	r3,[r11,#12]
776	sub	r11,r11,r12,lsl#3
777	ldr	r0,[r11,#16]!		@ prefetch tp1
778	mov	r7,#0x80
779	mov	r8,#0x1b
780	orr	r7,r7,#0x8000
781	orr	r8,r8,#0x1b00
782	orr	r7,r7,r7,lsl#16
783	orr	r8,r8,r8,lsl#16
784	sub	r12,r12,#1
785	mvn	r9,r7
786	mov	r12,r12,lsl#2	@ (rounds-1)*4
787
788.Lmix:	and	r4,r0,r7
789	and	r1,r0,r9
790	sub	r4,r4,r4,lsr#7
791	and	r4,r4,r8
792	eor	r1,r4,r1,lsl#1	@ tp2
793
794	and	r4,r1,r7
795	and	r2,r1,r9
796	sub	r4,r4,r4,lsr#7
797	and	r4,r4,r8
798	eor	r2,r4,r2,lsl#1	@ tp4
799
800	and	r4,r2,r7
801	and	r3,r2,r9
802	sub	r4,r4,r4,lsr#7
803	and	r4,r4,r8
804	eor	r3,r4,r3,lsl#1	@ tp8
805
806	eor	r4,r1,r2
807	eor	r5,r0,r3		@ tp9
808	eor	r4,r4,r3		@ tpe
809	eor	r4,r4,r1,ror#24
810	eor	r4,r4,r5,ror#24	@ ^= ROTATE(tpb=tp9^tp2,8)
811	eor	r4,r4,r2,ror#16
812	eor	r4,r4,r5,ror#16	@ ^= ROTATE(tpd=tp9^tp4,16)
813	eor	r4,r4,r5,ror#8	@ ^= ROTATE(tp9,24)
814
815	ldr	r0,[r11,#4]		@ prefetch tp1
816	str	r4,[r11],#4
817	subs	r12,r12,#1
818	bne	.Lmix
819
820	mov	r0,#0
821#if __ARM_ARCH__>=5
822	ldmia	sp!,{r4-r12,pc}
823#else
824	ldmia   sp!,{r4-r12,lr}
825	tst	lr,#1
826	moveq	pc,lr			@ be binary compatible with V4, yet
827	.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
828#endif
829.size	AES_set_enc2dec_key,.-AES_set_enc2dec_key
830
831.type	AES_Td,%object
832.align	5
833AES_Td:
834.word	0x51f4a750, 0x7e416553, 0x1a17a4c3, 0x3a275e96
835.word	0x3bab6bcb, 0x1f9d45f1, 0xacfa58ab, 0x4be30393
836.word	0x2030fa55, 0xad766df6, 0x88cc7691, 0xf5024c25
837.word	0x4fe5d7fc, 0xc52acbd7, 0x26354480, 0xb562a38f
838.word	0xdeb15a49, 0x25ba1b67, 0x45ea0e98, 0x5dfec0e1
839.word	0xc32f7502, 0x814cf012, 0x8d4697a3, 0x6bd3f9c6
840.word	0x038f5fe7, 0x15929c95, 0xbf6d7aeb, 0x955259da
841.word	0xd4be832d, 0x587421d3, 0x49e06929, 0x8ec9c844
842.word	0x75c2896a, 0xf48e7978, 0x99583e6b, 0x27b971dd
843.word	0xbee14fb6, 0xf088ad17, 0xc920ac66, 0x7dce3ab4
844.word	0x63df4a18, 0xe51a3182, 0x97513360, 0x62537f45
845.word	0xb16477e0, 0xbb6bae84, 0xfe81a01c, 0xf9082b94
846.word	0x70486858, 0x8f45fd19, 0x94de6c87, 0x527bf8b7
847.word	0xab73d323, 0x724b02e2, 0xe31f8f57, 0x6655ab2a
848.word	0xb2eb2807, 0x2fb5c203, 0x86c57b9a, 0xd33708a5
849.word	0x302887f2, 0x23bfa5b2, 0x02036aba, 0xed16825c
850.word	0x8acf1c2b, 0xa779b492, 0xf307f2f0, 0x4e69e2a1
851.word	0x65daf4cd, 0x0605bed5, 0xd134621f, 0xc4a6fe8a
852.word	0x342e539d, 0xa2f355a0, 0x058ae132, 0xa4f6eb75
853.word	0x0b83ec39, 0x4060efaa, 0x5e719f06, 0xbd6e1051
854.word	0x3e218af9, 0x96dd063d, 0xdd3e05ae, 0x4de6bd46
855.word	0x91548db5, 0x71c45d05, 0x0406d46f, 0x605015ff
856.word	0x1998fb24, 0xd6bde997, 0x894043cc, 0x67d99e77
857.word	0xb0e842bd, 0x07898b88, 0xe7195b38, 0x79c8eedb
858.word	0xa17c0a47, 0x7c420fe9, 0xf8841ec9, 0x00000000
859.word	0x09808683, 0x322bed48, 0x1e1170ac, 0x6c5a724e
860.word	0xfd0efffb, 0x0f853856, 0x3daed51e, 0x362d3927
861.word	0x0a0fd964, 0x685ca621, 0x9b5b54d1, 0x24362e3a
862.word	0x0c0a67b1, 0x9357e70f, 0xb4ee96d2, 0x1b9b919e
863.word	0x80c0c54f, 0x61dc20a2, 0x5a774b69, 0x1c121a16
864.word	0xe293ba0a, 0xc0a02ae5, 0x3c22e043, 0x121b171d
865.word	0x0e090d0b, 0xf28bc7ad, 0x2db6a8b9, 0x141ea9c8
866.word	0x57f11985, 0xaf75074c, 0xee99ddbb, 0xa37f60fd
867.word	0xf701269f, 0x5c72f5bc, 0x44663bc5, 0x5bfb7e34
868.word	0x8b432976, 0xcb23c6dc, 0xb6edfc68, 0xb8e4f163
869.word	0xd731dcca, 0x42638510, 0x13972240, 0x84c61120
870.word	0x854a247d, 0xd2bb3df8, 0xaef93211, 0xc729a16d
871.word	0x1d9e2f4b, 0xdcb230f3, 0x0d8652ec, 0x77c1e3d0
872.word	0x2bb3166c, 0xa970b999, 0x119448fa, 0x47e96422
873.word	0xa8fc8cc4, 0xa0f03f1a, 0x567d2cd8, 0x223390ef
874.word	0x87494ec7, 0xd938d1c1, 0x8ccaa2fe, 0x98d40b36
875.word	0xa6f581cf, 0xa57ade28, 0xdab78e26, 0x3fadbfa4
876.word	0x2c3a9de4, 0x5078920d, 0x6a5fcc9b, 0x547e4662
877.word	0xf68d13c2, 0x90d8b8e8, 0x2e39f75e, 0x82c3aff5
878.word	0x9f5d80be, 0x69d0937c, 0x6fd52da9, 0xcf2512b3
879.word	0xc8ac993b, 0x10187da7, 0xe89c636e, 0xdb3bbb7b
880.word	0xcd267809, 0x6e5918f4, 0xec9ab701, 0x834f9aa8
881.word	0xe6956e65, 0xaaffe67e, 0x21bccf08, 0xef15e8e6
882.word	0xbae79bd9, 0x4a6f36ce, 0xea9f09d4, 0x29b07cd6
883.word	0x31a4b2af, 0x2a3f2331, 0xc6a59430, 0x35a266c0
884.word	0x744ebc37, 0xfc82caa6, 0xe090d0b0, 0x33a7d815
885.word	0xf104984a, 0x41ecdaf7, 0x7fcd500e, 0x1791f62f
886.word	0x764dd68d, 0x43efb04d, 0xccaa4d54, 0xe49604df
887.word	0x9ed1b5e3, 0x4c6a881b, 0xc12c1fb8, 0x4665517f
888.word	0x9d5eea04, 0x018c355d, 0xfa877473, 0xfb0b412e
889.word	0xb3671d5a, 0x92dbd252, 0xe9105633, 0x6dd64713
890.word	0x9ad7618c, 0x37a10c7a, 0x59f8148e, 0xeb133c89
891.word	0xcea927ee, 0xb761c935, 0xe11ce5ed, 0x7a47b13c
892.word	0x9cd2df59, 0x55f2733f, 0x1814ce79, 0x73c737bf
893.word	0x53f7cdea, 0x5ffdaa5b, 0xdf3d6f14, 0x7844db86
894.word	0xcaaff381, 0xb968c43e, 0x3824342c, 0xc2a3405f
895.word	0x161dc372, 0xbce2250c, 0x283c498b, 0xff0d9541
896.word	0x39a80171, 0x080cb3de, 0xd8b4e49c, 0x6456c190
897.word	0x7bcb8461, 0xd532b670, 0x486c5c74, 0xd0b85742
898@ Td4[256]
899.byte	0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38
900.byte	0xbf, 0x40, 0xa3, 0x9e, 0x81, 0xf3, 0xd7, 0xfb
901.byte	0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87
902.byte	0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb
903.byte	0x54, 0x7b, 0x94, 0x32, 0xa6, 0xc2, 0x23, 0x3d
904.byte	0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e
905.byte	0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2
906.byte	0x76, 0x5b, 0xa2, 0x49, 0x6d, 0x8b, 0xd1, 0x25
907.byte	0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16
908.byte	0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92
909.byte	0x6c, 0x70, 0x48, 0x50, 0xfd, 0xed, 0xb9, 0xda
910.byte	0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84
911.byte	0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a
912.byte	0xf7, 0xe4, 0x58, 0x05, 0xb8, 0xb3, 0x45, 0x06
913.byte	0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02
914.byte	0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b
915.byte	0x3a, 0x91, 0x11, 0x41, 0x4f, 0x67, 0xdc, 0xea
916.byte	0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73
917.byte	0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85
918.byte	0xe2, 0xf9, 0x37, 0xe8, 0x1c, 0x75, 0xdf, 0x6e
919.byte	0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89
920.byte	0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b
921.byte	0xfc, 0x56, 0x3e, 0x4b, 0xc6, 0xd2, 0x79, 0x20
922.byte	0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4
923.byte	0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31
924.byte	0xb1, 0x12, 0x10, 0x59, 0x27, 0x80, 0xec, 0x5f
925.byte	0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d
926.byte	0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef
927.byte	0xa0, 0xe0, 0x3b, 0x4d, 0xae, 0x2a, 0xf5, 0xb0
928.byte	0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61
929.byte	0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26
930.byte	0xe1, 0x69, 0x14, 0x63, 0x55, 0x21, 0x0c, 0x7d
931.size	AES_Td,.-AES_Td
932
933@ void AES_decrypt(const unsigned char *in, unsigned char *out,
934@ 		 const AES_KEY *key) {
935.global AES_decrypt
936.type   AES_decrypt,%function
937.align	5
938AES_decrypt:
939#ifndef	__thumb2__
940	sub	r3,pc,#8		@ AES_decrypt
941#else
942	adr	r3,.
943#endif
944	stmdb   sp!,{r1,r4-r12,lr}
945#if defined(__thumb2__) || defined(__APPLE__)
946	adr	r10,AES_Td
947#else
948	sub	r10,r3,#AES_decrypt-AES_Td	@ Td
949#endif
950	mov	r12,r0		@ inp
951	mov	r11,r2
952#if __ARM_ARCH__<7
953	ldrb	r0,[r12,#3]	@ load input data in endian-neutral
954	ldrb	r4,[r12,#2]	@ manner...
955	ldrb	r5,[r12,#1]
956	ldrb	r6,[r12,#0]
957	orr	r0,r0,r4,lsl#8
958	ldrb	r1,[r12,#7]
959	orr	r0,r0,r5,lsl#16
960	ldrb	r4,[r12,#6]
961	orr	r0,r0,r6,lsl#24
962	ldrb	r5,[r12,#5]
963	ldrb	r6,[r12,#4]
964	orr	r1,r1,r4,lsl#8
965	ldrb	r2,[r12,#11]
966	orr	r1,r1,r5,lsl#16
967	ldrb	r4,[r12,#10]
968	orr	r1,r1,r6,lsl#24
969	ldrb	r5,[r12,#9]
970	ldrb	r6,[r12,#8]
971	orr	r2,r2,r4,lsl#8
972	ldrb	r3,[r12,#15]
973	orr	r2,r2,r5,lsl#16
974	ldrb	r4,[r12,#14]
975	orr	r2,r2,r6,lsl#24
976	ldrb	r5,[r12,#13]
977	ldrb	r6,[r12,#12]
978	orr	r3,r3,r4,lsl#8
979	orr	r3,r3,r5,lsl#16
980	orr	r3,r3,r6,lsl#24
981#else
982	ldr	r0,[r12,#0]
983	ldr	r1,[r12,#4]
984	ldr	r2,[r12,#8]
985	ldr	r3,[r12,#12]
986#ifdef __ARMEL__
987	rev	r0,r0
988	rev	r1,r1
989	rev	r2,r2
990	rev	r3,r3
991#endif
992#endif
993	bl	_armv4_AES_decrypt
994
995	ldr	r12,[sp],#4		@ pop out
996#if __ARM_ARCH__>=7
997#ifdef __ARMEL__
998	rev	r0,r0
999	rev	r1,r1
1000	rev	r2,r2
1001	rev	r3,r3
1002#endif
1003	str	r0,[r12,#0]
1004	str	r1,[r12,#4]
1005	str	r2,[r12,#8]
1006	str	r3,[r12,#12]
1007#else
1008	mov	r4,r0,lsr#24		@ write output in endian-neutral
1009	mov	r5,r0,lsr#16		@ manner...
1010	mov	r6,r0,lsr#8
1011	strb	r4,[r12,#0]
1012	strb	r5,[r12,#1]
1013	mov	r4,r1,lsr#24
1014	strb	r6,[r12,#2]
1015	mov	r5,r1,lsr#16
1016	strb	r0,[r12,#3]
1017	mov	r6,r1,lsr#8
1018	strb	r4,[r12,#4]
1019	strb	r5,[r12,#5]
1020	mov	r4,r2,lsr#24
1021	strb	r6,[r12,#6]
1022	mov	r5,r2,lsr#16
1023	strb	r1,[r12,#7]
1024	mov	r6,r2,lsr#8
1025	strb	r4,[r12,#8]
1026	strb	r5,[r12,#9]
1027	mov	r4,r3,lsr#24
1028	strb	r6,[r12,#10]
1029	mov	r5,r3,lsr#16
1030	strb	r2,[r12,#11]
1031	mov	r6,r3,lsr#8
1032	strb	r4,[r12,#12]
1033	strb	r5,[r12,#13]
1034	strb	r6,[r12,#14]
1035	strb	r3,[r12,#15]
1036#endif
1037#if __ARM_ARCH__>=5
1038	ldmia	sp!,{r4-r12,pc}
1039#else
1040	ldmia   sp!,{r4-r12,lr}
1041	tst	lr,#1
1042	moveq	pc,lr			@ be binary compatible with V4, yet
1043	.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
1044#endif
1045.size	AES_decrypt,.-AES_decrypt
1046
1047.type   _armv4_AES_decrypt,%function
1048.align	2
1049_armv4_AES_decrypt:
1050	str	lr,[sp,#-4]!		@ push lr
1051	ldmia	r11!,{r4-r7}
1052	eor	r0,r0,r4
1053	ldr	r12,[r11,#240-16]
1054	eor	r1,r1,r5
1055	eor	r2,r2,r6
1056	eor	r3,r3,r7
1057	sub	r12,r12,#1
1058	mov	lr,#255
1059
1060	and	r7,lr,r0,lsr#16
1061	and	r8,lr,r0,lsr#8
1062	and	r9,lr,r0
1063	mov	r0,r0,lsr#24
1064.Ldec_loop:
1065	ldr	r4,[r10,r7,lsl#2]	@ Td1[s0>>16]
1066	and	r7,lr,r1		@ i0
1067	ldr	r5,[r10,r8,lsl#2]	@ Td2[s0>>8]
1068	and	r8,lr,r1,lsr#16
1069	ldr	r6,[r10,r9,lsl#2]	@ Td3[s0>>0]
1070	and	r9,lr,r1,lsr#8
1071	ldr	r0,[r10,r0,lsl#2]	@ Td0[s0>>24]
1072	mov	r1,r1,lsr#24
1073
1074	ldr	r7,[r10,r7,lsl#2]	@ Td3[s1>>0]
1075	ldr	r8,[r10,r8,lsl#2]	@ Td1[s1>>16]
1076	ldr	r9,[r10,r9,lsl#2]	@ Td2[s1>>8]
1077	eor	r0,r0,r7,ror#24
1078	ldr	r1,[r10,r1,lsl#2]	@ Td0[s1>>24]
1079	and	r7,lr,r2,lsr#8	@ i0
1080	eor	r5,r8,r5,ror#8
1081	and	r8,lr,r2		@ i1
1082	eor	r6,r9,r6,ror#8
1083	and	r9,lr,r2,lsr#16
1084	ldr	r7,[r10,r7,lsl#2]	@ Td2[s2>>8]
1085	eor	r1,r1,r4,ror#8
1086	ldr	r8,[r10,r8,lsl#2]	@ Td3[s2>>0]
1087	mov	r2,r2,lsr#24
1088
1089	ldr	r9,[r10,r9,lsl#2]	@ Td1[s2>>16]
1090	eor	r0,r0,r7,ror#16
1091	ldr	r2,[r10,r2,lsl#2]	@ Td0[s2>>24]
1092	and	r7,lr,r3,lsr#16	@ i0
1093	eor	r1,r1,r8,ror#24
1094	and	r8,lr,r3,lsr#8	@ i1
1095	eor	r6,r9,r6,ror#8
1096	and	r9,lr,r3		@ i2
1097	ldr	r7,[r10,r7,lsl#2]	@ Td1[s3>>16]
1098	eor	r2,r2,r5,ror#8
1099	ldr	r8,[r10,r8,lsl#2]	@ Td2[s3>>8]
1100	mov	r3,r3,lsr#24
1101
1102	ldr	r9,[r10,r9,lsl#2]	@ Td3[s3>>0]
1103	eor	r0,r0,r7,ror#8
1104	ldr	r7,[r11],#16
1105	eor	r1,r1,r8,ror#16
1106	ldr	r3,[r10,r3,lsl#2]	@ Td0[s3>>24]
1107	eor	r2,r2,r9,ror#24
1108
1109	ldr	r4,[r11,#-12]
1110	eor	r0,r0,r7
1111	ldr	r5,[r11,#-8]
1112	eor	r3,r3,r6,ror#8
1113	ldr	r6,[r11,#-4]
1114	and	r7,lr,r0,lsr#16
1115	eor	r1,r1,r4
1116	and	r8,lr,r0,lsr#8
1117	eor	r2,r2,r5
1118	and	r9,lr,r0
1119	eor	r3,r3,r6
1120	mov	r0,r0,lsr#24
1121
1122	subs	r12,r12,#1
1123	bne	.Ldec_loop
1124
1125	add	r10,r10,#1024
1126
1127	ldr	r5,[r10,#0]		@ prefetch Td4
1128	ldr	r6,[r10,#32]
1129	ldr	r4,[r10,#64]
1130	ldr	r5,[r10,#96]
1131	ldr	r6,[r10,#128]
1132	ldr	r4,[r10,#160]
1133	ldr	r5,[r10,#192]
1134	ldr	r6,[r10,#224]
1135
1136	ldrb	r0,[r10,r0]		@ Td4[s0>>24]
1137	ldrb	r4,[r10,r7]		@ Td4[s0>>16]
1138	and	r7,lr,r1		@ i0
1139	ldrb	r5,[r10,r8]		@ Td4[s0>>8]
1140	and	r8,lr,r1,lsr#16
1141	ldrb	r6,[r10,r9]		@ Td4[s0>>0]
1142	and	r9,lr,r1,lsr#8
1143
1144	add	r1,r10,r1,lsr#24
1145	ldrb	r7,[r10,r7]		@ Td4[s1>>0]
1146	ldrb	r1,[r1]		@ Td4[s1>>24]
1147	ldrb	r8,[r10,r8]		@ Td4[s1>>16]
1148	eor	r0,r7,r0,lsl#24
1149	ldrb	r9,[r10,r9]		@ Td4[s1>>8]
1150	eor	r1,r4,r1,lsl#8
1151	and	r7,lr,r2,lsr#8	@ i0
1152	eor	r5,r5,r8,lsl#8
1153	and	r8,lr,r2		@ i1
1154	ldrb	r7,[r10,r7]		@ Td4[s2>>8]
1155	eor	r6,r6,r9,lsl#8
1156	ldrb	r8,[r10,r8]		@ Td4[s2>>0]
1157	and	r9,lr,r2,lsr#16
1158
1159	add	r2,r10,r2,lsr#24
1160	ldrb	r2,[r2]		@ Td4[s2>>24]
1161	eor	r0,r0,r7,lsl#8
1162	ldrb	r9,[r10,r9]		@ Td4[s2>>16]
1163	eor	r1,r8,r1,lsl#16
1164	and	r7,lr,r3,lsr#16	@ i0
1165	eor	r2,r5,r2,lsl#16
1166	and	r8,lr,r3,lsr#8	@ i1
1167	ldrb	r7,[r10,r7]		@ Td4[s3>>16]
1168	eor	r6,r6,r9,lsl#16
1169	ldrb	r8,[r10,r8]		@ Td4[s3>>8]
1170	and	r9,lr,r3		@ i2
1171
1172	add	r3,r10,r3,lsr#24
1173	ldrb	r9,[r10,r9]		@ Td4[s3>>0]
1174	ldrb	r3,[r3]		@ Td4[s3>>24]
1175	eor	r0,r0,r7,lsl#16
1176	ldr	r7,[r11,#0]
1177	eor	r1,r1,r8,lsl#8
1178	ldr	r4,[r11,#4]
1179	eor	r2,r9,r2,lsl#8
1180	ldr	r5,[r11,#8]
1181	eor	r3,r6,r3,lsl#24
1182	ldr	r6,[r11,#12]
1183
1184	eor	r0,r0,r7
1185	eor	r1,r1,r4
1186	eor	r2,r2,r5
1187	eor	r3,r3,r6
1188
1189	sub	r10,r10,#1024
1190	ldr	pc,[sp],#4		@ pop and return
1191.size	_armv4_AES_decrypt,.-_armv4_AES_decrypt
1192.asciz	"AES for ARMv4, CRYPTOGAMS by <appro@openssl.org>"
1193.align	2
1194