1
2@ ====================================================================
3@ Written by Andy Polyakov <appro@openssl.org> for the OpenSSL
4@ project. The module is, however, dual licensed under OpenSSL and
5@ CRYPTOGAMS licenses depending on where you obtain it. For further
6@ details see http://www.openssl.org/~appro/cryptogams/.
7@ ====================================================================
8
9@ AES for ARMv4
10
11@ January 2007.
12@
13@ Code uses single 1K S-box and is >2 times faster than code generated
14@ by gcc-3.4.1. This is thanks to unique feature of ARMv4 ISA, which
15@ allows to merge logical or arithmetic operation with shift or rotate
16@ in one instruction and emit combined result every cycle. The module
17@ is endian-neutral. The performance is ~42 cycles/byte for 128-bit
18@ key [on single-issue Xscale PXA250 core].
19
20@ May 2007.
21@
22@ AES_set_[en|de]crypt_key is added.
23
24@ July 2010.
25@
26@ Rescheduling for dual-issue pipeline resulted in 12% improvement on
27@ Cortex A8 core and ~25 cycles per byte processed with 128-bit key.
28
29@ February 2011.
30@
31@ Profiler-assisted and platform-specific optimization resulted in 16%
32@ improvement on Cortex A8 core and ~21.5 cycles per byte.
33
34#if defined(__arm__)
35#ifndef __KERNEL__
36# include "arm_arch.h"
37#else
38# define __ARM_ARCH__ __LINUX_ARM_ARCH__
39#endif
40
41.text
42#if __ARM_ARCH__<7
43.code	32
44#else
45.syntax	unified
46# if defined(__thumb2__) && !defined(__APPLE__)
47.thumb
48# else
49.code	32
50# endif
51#endif
52
53.type	AES_Te,%object
54.align	5
55AES_Te:
56.word	0xc66363a5, 0xf87c7c84, 0xee777799, 0xf67b7b8d
57.word	0xfff2f20d, 0xd66b6bbd, 0xde6f6fb1, 0x91c5c554
58.word	0x60303050, 0x02010103, 0xce6767a9, 0x562b2b7d
59.word	0xe7fefe19, 0xb5d7d762, 0x4dababe6, 0xec76769a
60.word	0x8fcaca45, 0x1f82829d, 0x89c9c940, 0xfa7d7d87
61.word	0xeffafa15, 0xb25959eb, 0x8e4747c9, 0xfbf0f00b
62.word	0x41adadec, 0xb3d4d467, 0x5fa2a2fd, 0x45afafea
63.word	0x239c9cbf, 0x53a4a4f7, 0xe4727296, 0x9bc0c05b
64.word	0x75b7b7c2, 0xe1fdfd1c, 0x3d9393ae, 0x4c26266a
65.word	0x6c36365a, 0x7e3f3f41, 0xf5f7f702, 0x83cccc4f
66.word	0x6834345c, 0x51a5a5f4, 0xd1e5e534, 0xf9f1f108
67.word	0xe2717193, 0xabd8d873, 0x62313153, 0x2a15153f
68.word	0x0804040c, 0x95c7c752, 0x46232365, 0x9dc3c35e
69.word	0x30181828, 0x379696a1, 0x0a05050f, 0x2f9a9ab5
70.word	0x0e070709, 0x24121236, 0x1b80809b, 0xdfe2e23d
71.word	0xcdebeb26, 0x4e272769, 0x7fb2b2cd, 0xea75759f
72.word	0x1209091b, 0x1d83839e, 0x582c2c74, 0x341a1a2e
73.word	0x361b1b2d, 0xdc6e6eb2, 0xb45a5aee, 0x5ba0a0fb
74.word	0xa45252f6, 0x763b3b4d, 0xb7d6d661, 0x7db3b3ce
75.word	0x5229297b, 0xdde3e33e, 0x5e2f2f71, 0x13848497
76.word	0xa65353f5, 0xb9d1d168, 0x00000000, 0xc1eded2c
77.word	0x40202060, 0xe3fcfc1f, 0x79b1b1c8, 0xb65b5bed
78.word	0xd46a6abe, 0x8dcbcb46, 0x67bebed9, 0x7239394b
79.word	0x944a4ade, 0x984c4cd4, 0xb05858e8, 0x85cfcf4a
80.word	0xbbd0d06b, 0xc5efef2a, 0x4faaaae5, 0xedfbfb16
81.word	0x864343c5, 0x9a4d4dd7, 0x66333355, 0x11858594
82.word	0x8a4545cf, 0xe9f9f910, 0x04020206, 0xfe7f7f81
83.word	0xa05050f0, 0x783c3c44, 0x259f9fba, 0x4ba8a8e3
84.word	0xa25151f3, 0x5da3a3fe, 0x804040c0, 0x058f8f8a
85.word	0x3f9292ad, 0x219d9dbc, 0x70383848, 0xf1f5f504
86.word	0x63bcbcdf, 0x77b6b6c1, 0xafdada75, 0x42212163
87.word	0x20101030, 0xe5ffff1a, 0xfdf3f30e, 0xbfd2d26d
88.word	0x81cdcd4c, 0x180c0c14, 0x26131335, 0xc3ecec2f
89.word	0xbe5f5fe1, 0x359797a2, 0x884444cc, 0x2e171739
90.word	0x93c4c457, 0x55a7a7f2, 0xfc7e7e82, 0x7a3d3d47
91.word	0xc86464ac, 0xba5d5de7, 0x3219192b, 0xe6737395
92.word	0xc06060a0, 0x19818198, 0x9e4f4fd1, 0xa3dcdc7f
93.word	0x44222266, 0x542a2a7e, 0x3b9090ab, 0x0b888883
94.word	0x8c4646ca, 0xc7eeee29, 0x6bb8b8d3, 0x2814143c
95.word	0xa7dede79, 0xbc5e5ee2, 0x160b0b1d, 0xaddbdb76
96.word	0xdbe0e03b, 0x64323256, 0x743a3a4e, 0x140a0a1e
97.word	0x924949db, 0x0c06060a, 0x4824246c, 0xb85c5ce4
98.word	0x9fc2c25d, 0xbdd3d36e, 0x43acacef, 0xc46262a6
99.word	0x399191a8, 0x319595a4, 0xd3e4e437, 0xf279798b
100.word	0xd5e7e732, 0x8bc8c843, 0x6e373759, 0xda6d6db7
101.word	0x018d8d8c, 0xb1d5d564, 0x9c4e4ed2, 0x49a9a9e0
102.word	0xd86c6cb4, 0xac5656fa, 0xf3f4f407, 0xcfeaea25
103.word	0xca6565af, 0xf47a7a8e, 0x47aeaee9, 0x10080818
104.word	0x6fbabad5, 0xf0787888, 0x4a25256f, 0x5c2e2e72
105.word	0x381c1c24, 0x57a6a6f1, 0x73b4b4c7, 0x97c6c651
106.word	0xcbe8e823, 0xa1dddd7c, 0xe874749c, 0x3e1f1f21
107.word	0x964b4bdd, 0x61bdbddc, 0x0d8b8b86, 0x0f8a8a85
108.word	0xe0707090, 0x7c3e3e42, 0x71b5b5c4, 0xcc6666aa
109.word	0x904848d8, 0x06030305, 0xf7f6f601, 0x1c0e0e12
110.word	0xc26161a3, 0x6a35355f, 0xae5757f9, 0x69b9b9d0
111.word	0x17868691, 0x99c1c158, 0x3a1d1d27, 0x279e9eb9
112.word	0xd9e1e138, 0xebf8f813, 0x2b9898b3, 0x22111133
113.word	0xd26969bb, 0xa9d9d970, 0x078e8e89, 0x339494a7
114.word	0x2d9b9bb6, 0x3c1e1e22, 0x15878792, 0xc9e9e920
115.word	0x87cece49, 0xaa5555ff, 0x50282878, 0xa5dfdf7a
116.word	0x038c8c8f, 0x59a1a1f8, 0x09898980, 0x1a0d0d17
117.word	0x65bfbfda, 0xd7e6e631, 0x844242c6, 0xd06868b8
118.word	0x824141c3, 0x299999b0, 0x5a2d2d77, 0x1e0f0f11
119.word	0x7bb0b0cb, 0xa85454fc, 0x6dbbbbd6, 0x2c16163a
120@ Te4[256]
121.byte	0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5
122.byte	0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76
123.byte	0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0
124.byte	0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0
125.byte	0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc
126.byte	0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15
127.byte	0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a
128.byte	0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75
129.byte	0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0
130.byte	0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84
131.byte	0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b
132.byte	0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf
133.byte	0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85
134.byte	0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8
135.byte	0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5
136.byte	0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2
137.byte	0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17
138.byte	0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73
139.byte	0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88
140.byte	0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb
141.byte	0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c
142.byte	0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79
143.byte	0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9
144.byte	0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08
145.byte	0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6
146.byte	0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a
147.byte	0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e
148.byte	0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e
149.byte	0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94
150.byte	0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf
151.byte	0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68
152.byte	0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16
153@ rcon[]
154.word	0x01000000, 0x02000000, 0x04000000, 0x08000000
155.word	0x10000000, 0x20000000, 0x40000000, 0x80000000
156.word	0x1B000000, 0x36000000, 0, 0, 0, 0, 0, 0
157.size	AES_Te,.-AES_Te
158
159@ void asm_AES_encrypt(const unsigned char *in, unsigned char *out,
160@ 		       const AES_KEY *key) {
161.globl	asm_AES_encrypt
162.hidden	asm_AES_encrypt
163.type	asm_AES_encrypt,%function
164.align	5
165asm_AES_encrypt:
166#if __ARM_ARCH__<7
167	sub	r3,pc,#8		@ asm_AES_encrypt
168#else
169	adr	r3,asm_AES_encrypt
170#endif
171	stmdb	sp!,{r1,r4-r12,lr}
172#ifdef	__APPLE__
173	adr	r10,AES_Te
174#else
175	sub	r10,r3,#asm_AES_encrypt-AES_Te	@ Te
176#endif
177	mov	r12,r0		@ inp
178	mov	r11,r2
179#if __ARM_ARCH__<7
180	ldrb	r0,[r12,#3]	@ load input data in endian-neutral
181	ldrb	r4,[r12,#2]	@ manner...
182	ldrb	r5,[r12,#1]
183	ldrb	r6,[r12,#0]
184	orr	r0,r0,r4,lsl#8
185	ldrb	r1,[r12,#7]
186	orr	r0,r0,r5,lsl#16
187	ldrb	r4,[r12,#6]
188	orr	r0,r0,r6,lsl#24
189	ldrb	r5,[r12,#5]
190	ldrb	r6,[r12,#4]
191	orr	r1,r1,r4,lsl#8
192	ldrb	r2,[r12,#11]
193	orr	r1,r1,r5,lsl#16
194	ldrb	r4,[r12,#10]
195	orr	r1,r1,r6,lsl#24
196	ldrb	r5,[r12,#9]
197	ldrb	r6,[r12,#8]
198	orr	r2,r2,r4,lsl#8
199	ldrb	r3,[r12,#15]
200	orr	r2,r2,r5,lsl#16
201	ldrb	r4,[r12,#14]
202	orr	r2,r2,r6,lsl#24
203	ldrb	r5,[r12,#13]
204	ldrb	r6,[r12,#12]
205	orr	r3,r3,r4,lsl#8
206	orr	r3,r3,r5,lsl#16
207	orr	r3,r3,r6,lsl#24
208#else
209	ldr	r0,[r12,#0]
210	ldr	r1,[r12,#4]
211	ldr	r2,[r12,#8]
212	ldr	r3,[r12,#12]
213#ifdef __ARMEL__
214	rev	r0,r0
215	rev	r1,r1
216	rev	r2,r2
217	rev	r3,r3
218#endif
219#endif
220	bl	_armv4_AES_encrypt
221
222	ldr	r12,[sp],#4		@ pop out
223#if __ARM_ARCH__>=7
224#ifdef __ARMEL__
225	rev	r0,r0
226	rev	r1,r1
227	rev	r2,r2
228	rev	r3,r3
229#endif
230	str	r0,[r12,#0]
231	str	r1,[r12,#4]
232	str	r2,[r12,#8]
233	str	r3,[r12,#12]
234#else
235	mov	r4,r0,lsr#24		@ write output in endian-neutral
236	mov	r5,r0,lsr#16		@ manner...
237	mov	r6,r0,lsr#8
238	strb	r4,[r12,#0]
239	strb	r5,[r12,#1]
240	mov	r4,r1,lsr#24
241	strb	r6,[r12,#2]
242	mov	r5,r1,lsr#16
243	strb	r0,[r12,#3]
244	mov	r6,r1,lsr#8
245	strb	r4,[r12,#4]
246	strb	r5,[r12,#5]
247	mov	r4,r2,lsr#24
248	strb	r6,[r12,#6]
249	mov	r5,r2,lsr#16
250	strb	r1,[r12,#7]
251	mov	r6,r2,lsr#8
252	strb	r4,[r12,#8]
253	strb	r5,[r12,#9]
254	mov	r4,r3,lsr#24
255	strb	r6,[r12,#10]
256	mov	r5,r3,lsr#16
257	strb	r2,[r12,#11]
258	mov	r6,r3,lsr#8
259	strb	r4,[r12,#12]
260	strb	r5,[r12,#13]
261	strb	r6,[r12,#14]
262	strb	r3,[r12,#15]
263#endif
264#if __ARM_ARCH__>=5
265	ldmia	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,pc}
266#else
267	ldmia	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
268	tst	lr,#1
269	moveq	pc,lr			@ be binary compatible with V4, yet
270.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
271#endif
272.size	asm_AES_encrypt,.-asm_AES_encrypt
273
274.type	_armv4_AES_encrypt,%function
275.align	2
276_armv4_AES_encrypt:
277	str	lr,[sp,#-4]!		@ push lr
278	ldmia	r11!,{r4,r5,r6,r7}
279	eor	r0,r0,r4
280	ldr	r12,[r11,#240-16]
281	eor	r1,r1,r5
282	eor	r2,r2,r6
283	eor	r3,r3,r7
284	sub	r12,r12,#1
285	mov	lr,#255
286
287	and	r7,lr,r0
288	and	r8,lr,r0,lsr#8
289	and	r9,lr,r0,lsr#16
290	mov	r0,r0,lsr#24
291.Lenc_loop:
292	ldr	r4,[r10,r7,lsl#2]	@ Te3[s0>>0]
293	and	r7,lr,r1,lsr#16	@ i0
294	ldr	r5,[r10,r8,lsl#2]	@ Te2[s0>>8]
295	and	r8,lr,r1
296	ldr	r6,[r10,r9,lsl#2]	@ Te1[s0>>16]
297	and	r9,lr,r1,lsr#8
298	ldr	r0,[r10,r0,lsl#2]	@ Te0[s0>>24]
299	mov	r1,r1,lsr#24
300
301	ldr	r7,[r10,r7,lsl#2]	@ Te1[s1>>16]
302	ldr	r8,[r10,r8,lsl#2]	@ Te3[s1>>0]
303	ldr	r9,[r10,r9,lsl#2]	@ Te2[s1>>8]
304	eor	r0,r0,r7,ror#8
305	ldr	r1,[r10,r1,lsl#2]	@ Te0[s1>>24]
306	and	r7,lr,r2,lsr#8	@ i0
307	eor	r5,r5,r8,ror#8
308	and	r8,lr,r2,lsr#16	@ i1
309	eor	r6,r6,r9,ror#8
310	and	r9,lr,r2
311	ldr	r7,[r10,r7,lsl#2]	@ Te2[s2>>8]
312	eor	r1,r1,r4,ror#24
313	ldr	r8,[r10,r8,lsl#2]	@ Te1[s2>>16]
314	mov	r2,r2,lsr#24
315
316	ldr	r9,[r10,r9,lsl#2]	@ Te3[s2>>0]
317	eor	r0,r0,r7,ror#16
318	ldr	r2,[r10,r2,lsl#2]	@ Te0[s2>>24]
319	and	r7,lr,r3		@ i0
320	eor	r1,r1,r8,ror#8
321	and	r8,lr,r3,lsr#8	@ i1
322	eor	r6,r6,r9,ror#16
323	and	r9,lr,r3,lsr#16	@ i2
324	ldr	r7,[r10,r7,lsl#2]	@ Te3[s3>>0]
325	eor	r2,r2,r5,ror#16
326	ldr	r8,[r10,r8,lsl#2]	@ Te2[s3>>8]
327	mov	r3,r3,lsr#24
328
329	ldr	r9,[r10,r9,lsl#2]	@ Te1[s3>>16]
330	eor	r0,r0,r7,ror#24
331	ldr	r7,[r11],#16
332	eor	r1,r1,r8,ror#16
333	ldr	r3,[r10,r3,lsl#2]	@ Te0[s3>>24]
334	eor	r2,r2,r9,ror#8
335	ldr	r4,[r11,#-12]
336	eor	r3,r3,r6,ror#8
337
338	ldr	r5,[r11,#-8]
339	eor	r0,r0,r7
340	ldr	r6,[r11,#-4]
341	and	r7,lr,r0
342	eor	r1,r1,r4
343	and	r8,lr,r0,lsr#8
344	eor	r2,r2,r5
345	and	r9,lr,r0,lsr#16
346	eor	r3,r3,r6
347	mov	r0,r0,lsr#24
348
349	subs	r12,r12,#1
350	bne	.Lenc_loop
351
352	add	r10,r10,#2
353
354	ldrb	r4,[r10,r7,lsl#2]	@ Te4[s0>>0]
355	and	r7,lr,r1,lsr#16	@ i0
356	ldrb	r5,[r10,r8,lsl#2]	@ Te4[s0>>8]
357	and	r8,lr,r1
358	ldrb	r6,[r10,r9,lsl#2]	@ Te4[s0>>16]
359	and	r9,lr,r1,lsr#8
360	ldrb	r0,[r10,r0,lsl#2]	@ Te4[s0>>24]
361	mov	r1,r1,lsr#24
362
363	ldrb	r7,[r10,r7,lsl#2]	@ Te4[s1>>16]
364	ldrb	r8,[r10,r8,lsl#2]	@ Te4[s1>>0]
365	ldrb	r9,[r10,r9,lsl#2]	@ Te4[s1>>8]
366	eor	r0,r7,r0,lsl#8
367	ldrb	r1,[r10,r1,lsl#2]	@ Te4[s1>>24]
368	and	r7,lr,r2,lsr#8	@ i0
369	eor	r5,r8,r5,lsl#8
370	and	r8,lr,r2,lsr#16	@ i1
371	eor	r6,r9,r6,lsl#8
372	and	r9,lr,r2
373	ldrb	r7,[r10,r7,lsl#2]	@ Te4[s2>>8]
374	eor	r1,r4,r1,lsl#24
375	ldrb	r8,[r10,r8,lsl#2]	@ Te4[s2>>16]
376	mov	r2,r2,lsr#24
377
378	ldrb	r9,[r10,r9,lsl#2]	@ Te4[s2>>0]
379	eor	r0,r7,r0,lsl#8
380	ldrb	r2,[r10,r2,lsl#2]	@ Te4[s2>>24]
381	and	r7,lr,r3		@ i0
382	eor	r1,r1,r8,lsl#16
383	and	r8,lr,r3,lsr#8	@ i1
384	eor	r6,r9,r6,lsl#8
385	and	r9,lr,r3,lsr#16	@ i2
386	ldrb	r7,[r10,r7,lsl#2]	@ Te4[s3>>0]
387	eor	r2,r5,r2,lsl#24
388	ldrb	r8,[r10,r8,lsl#2]	@ Te4[s3>>8]
389	mov	r3,r3,lsr#24
390
391	ldrb	r9,[r10,r9,lsl#2]	@ Te4[s3>>16]
392	eor	r0,r7,r0,lsl#8
393	ldr	r7,[r11,#0]
394	ldrb	r3,[r10,r3,lsl#2]	@ Te4[s3>>24]
395	eor	r1,r1,r8,lsl#8
396	ldr	r4,[r11,#4]
397	eor	r2,r2,r9,lsl#16
398	ldr	r5,[r11,#8]
399	eor	r3,r6,r3,lsl#24
400	ldr	r6,[r11,#12]
401
402	eor	r0,r0,r7
403	eor	r1,r1,r4
404	eor	r2,r2,r5
405	eor	r3,r3,r6
406
407	sub	r10,r10,#2
408	ldr	pc,[sp],#4		@ pop and return
409.size	_armv4_AES_encrypt,.-_armv4_AES_encrypt
410
411.globl	asm_AES_set_encrypt_key
412.hidden	asm_AES_set_encrypt_key
413.type	asm_AES_set_encrypt_key,%function
414.align	5
415asm_AES_set_encrypt_key:
416_armv4_AES_set_encrypt_key:
417#if __ARM_ARCH__<7
418	sub	r3,pc,#8		@ asm_AES_set_encrypt_key
419#else
420	adr	r3,asm_AES_set_encrypt_key
421#endif
422	teq	r0,#0
423#if __ARM_ARCH__>=7
424	itt	eq			@ Thumb2 thing, sanity check in ARM
425#endif
426	moveq	r0,#-1
427	beq	.Labrt
428	teq	r2,#0
429#if __ARM_ARCH__>=7
430	itt	eq			@ Thumb2 thing, sanity check in ARM
431#endif
432	moveq	r0,#-1
433	beq	.Labrt
434
435	teq	r1,#128
436	beq	.Lok
437	teq	r1,#192
438	beq	.Lok
439	teq	r1,#256
440#if __ARM_ARCH__>=7
441	itt	ne			@ Thumb2 thing, sanity check in ARM
442#endif
443	movne	r0,#-1
444	bne	.Labrt
445
446.Lok:	stmdb	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
447	mov	r12,r0		@ inp
448	mov	lr,r1			@ bits
449	mov	r11,r2			@ key
450
451#ifdef	__APPLE__
452	adr	r10,AES_Te+1024				@ Te4
453#else
454	sub	r10,r3,#_armv4_AES_set_encrypt_key-AES_Te-1024	@ Te4
455#endif
456
457#if __ARM_ARCH__<7
458	ldrb	r0,[r12,#3]	@ load input data in endian-neutral
459	ldrb	r4,[r12,#2]	@ manner...
460	ldrb	r5,[r12,#1]
461	ldrb	r6,[r12,#0]
462	orr	r0,r0,r4,lsl#8
463	ldrb	r1,[r12,#7]
464	orr	r0,r0,r5,lsl#16
465	ldrb	r4,[r12,#6]
466	orr	r0,r0,r6,lsl#24
467	ldrb	r5,[r12,#5]
468	ldrb	r6,[r12,#4]
469	orr	r1,r1,r4,lsl#8
470	ldrb	r2,[r12,#11]
471	orr	r1,r1,r5,lsl#16
472	ldrb	r4,[r12,#10]
473	orr	r1,r1,r6,lsl#24
474	ldrb	r5,[r12,#9]
475	ldrb	r6,[r12,#8]
476	orr	r2,r2,r4,lsl#8
477	ldrb	r3,[r12,#15]
478	orr	r2,r2,r5,lsl#16
479	ldrb	r4,[r12,#14]
480	orr	r2,r2,r6,lsl#24
481	ldrb	r5,[r12,#13]
482	ldrb	r6,[r12,#12]
483	orr	r3,r3,r4,lsl#8
484	str	r0,[r11],#16
485	orr	r3,r3,r5,lsl#16
486	str	r1,[r11,#-12]
487	orr	r3,r3,r6,lsl#24
488	str	r2,[r11,#-8]
489	str	r3,[r11,#-4]
490#else
491	ldr	r0,[r12,#0]
492	ldr	r1,[r12,#4]
493	ldr	r2,[r12,#8]
494	ldr	r3,[r12,#12]
495#ifdef __ARMEL__
496	rev	r0,r0
497	rev	r1,r1
498	rev	r2,r2
499	rev	r3,r3
500#endif
501	str	r0,[r11],#16
502	str	r1,[r11,#-12]
503	str	r2,[r11,#-8]
504	str	r3,[r11,#-4]
505#endif
506
507	teq	lr,#128
508	bne	.Lnot128
509	mov	r12,#10
510	str	r12,[r11,#240-16]
511	add	r6,r10,#256			@ rcon
512	mov	lr,#255
513
514.L128_loop:
515	and	r5,lr,r3,lsr#24
516	and	r7,lr,r3,lsr#16
517	ldrb	r5,[r10,r5]
518	and	r8,lr,r3,lsr#8
519	ldrb	r7,[r10,r7]
520	and	r9,lr,r3
521	ldrb	r8,[r10,r8]
522	orr	r5,r5,r7,lsl#24
523	ldrb	r9,[r10,r9]
524	orr	r5,r5,r8,lsl#16
525	ldr	r4,[r6],#4			@ rcon[i++]
526	orr	r5,r5,r9,lsl#8
527	eor	r5,r5,r4
528	eor	r0,r0,r5			@ rk[4]=rk[0]^...
529	eor	r1,r1,r0			@ rk[5]=rk[1]^rk[4]
530	str	r0,[r11],#16
531	eor	r2,r2,r1			@ rk[6]=rk[2]^rk[5]
532	str	r1,[r11,#-12]
533	eor	r3,r3,r2			@ rk[7]=rk[3]^rk[6]
534	str	r2,[r11,#-8]
535	subs	r12,r12,#1
536	str	r3,[r11,#-4]
537	bne	.L128_loop
538	sub	r2,r11,#176
539	b	.Ldone
540
541.Lnot128:
542#if __ARM_ARCH__<7
543	ldrb	r8,[r12,#19]
544	ldrb	r4,[r12,#18]
545	ldrb	r5,[r12,#17]
546	ldrb	r6,[r12,#16]
547	orr	r8,r8,r4,lsl#8
548	ldrb	r9,[r12,#23]
549	orr	r8,r8,r5,lsl#16
550	ldrb	r4,[r12,#22]
551	orr	r8,r8,r6,lsl#24
552	ldrb	r5,[r12,#21]
553	ldrb	r6,[r12,#20]
554	orr	r9,r9,r4,lsl#8
555	orr	r9,r9,r5,lsl#16
556	str	r8,[r11],#8
557	orr	r9,r9,r6,lsl#24
558	str	r9,[r11,#-4]
559#else
560	ldr	r8,[r12,#16]
561	ldr	r9,[r12,#20]
562#ifdef __ARMEL__
563	rev	r8,r8
564	rev	r9,r9
565#endif
566	str	r8,[r11],#8
567	str	r9,[r11,#-4]
568#endif
569
570	teq	lr,#192
571	bne	.Lnot192
572	mov	r12,#12
573	str	r12,[r11,#240-24]
574	add	r6,r10,#256			@ rcon
575	mov	lr,#255
576	mov	r12,#8
577
578.L192_loop:
579	and	r5,lr,r9,lsr#24
580	and	r7,lr,r9,lsr#16
581	ldrb	r5,[r10,r5]
582	and	r8,lr,r9,lsr#8
583	ldrb	r7,[r10,r7]
584	and	r9,lr,r9
585	ldrb	r8,[r10,r8]
586	orr	r5,r5,r7,lsl#24
587	ldrb	r9,[r10,r9]
588	orr	r5,r5,r8,lsl#16
589	ldr	r4,[r6],#4			@ rcon[i++]
590	orr	r5,r5,r9,lsl#8
591	eor	r9,r5,r4
592	eor	r0,r0,r9			@ rk[6]=rk[0]^...
593	eor	r1,r1,r0			@ rk[7]=rk[1]^rk[6]
594	str	r0,[r11],#24
595	eor	r2,r2,r1			@ rk[8]=rk[2]^rk[7]
596	str	r1,[r11,#-20]
597	eor	r3,r3,r2			@ rk[9]=rk[3]^rk[8]
598	str	r2,[r11,#-16]
599	subs	r12,r12,#1
600	str	r3,[r11,#-12]
601#if __ARM_ARCH__>=7
602	itt	eq				@ Thumb2 thing, sanity check in ARM
603#endif
604	subeq	r2,r11,#216
605	beq	.Ldone
606
607	ldr	r7,[r11,#-32]
608	ldr	r8,[r11,#-28]
609	eor	r7,r7,r3			@ rk[10]=rk[4]^rk[9]
610	eor	r9,r8,r7			@ rk[11]=rk[5]^rk[10]
611	str	r7,[r11,#-8]
612	str	r9,[r11,#-4]
613	b	.L192_loop
614
615.Lnot192:
616#if __ARM_ARCH__<7
617	ldrb	r8,[r12,#27]
618	ldrb	r4,[r12,#26]
619	ldrb	r5,[r12,#25]
620	ldrb	r6,[r12,#24]
621	orr	r8,r8,r4,lsl#8
622	ldrb	r9,[r12,#31]
623	orr	r8,r8,r5,lsl#16
624	ldrb	r4,[r12,#30]
625	orr	r8,r8,r6,lsl#24
626	ldrb	r5,[r12,#29]
627	ldrb	r6,[r12,#28]
628	orr	r9,r9,r4,lsl#8
629	orr	r9,r9,r5,lsl#16
630	str	r8,[r11],#8
631	orr	r9,r9,r6,lsl#24
632	str	r9,[r11,#-4]
633#else
634	ldr	r8,[r12,#24]
635	ldr	r9,[r12,#28]
636#ifdef __ARMEL__
637	rev	r8,r8
638	rev	r9,r9
639#endif
640	str	r8,[r11],#8
641	str	r9,[r11,#-4]
642#endif
643
644	mov	r12,#14
645	str	r12,[r11,#240-32]
646	add	r6,r10,#256			@ rcon
647	mov	lr,#255
648	mov	r12,#7
649
650.L256_loop:
651	and	r5,lr,r9,lsr#24
652	and	r7,lr,r9,lsr#16
653	ldrb	r5,[r10,r5]
654	and	r8,lr,r9,lsr#8
655	ldrb	r7,[r10,r7]
656	and	r9,lr,r9
657	ldrb	r8,[r10,r8]
658	orr	r5,r5,r7,lsl#24
659	ldrb	r9,[r10,r9]
660	orr	r5,r5,r8,lsl#16
661	ldr	r4,[r6],#4			@ rcon[i++]
662	orr	r5,r5,r9,lsl#8
663	eor	r9,r5,r4
664	eor	r0,r0,r9			@ rk[8]=rk[0]^...
665	eor	r1,r1,r0			@ rk[9]=rk[1]^rk[8]
666	str	r0,[r11],#32
667	eor	r2,r2,r1			@ rk[10]=rk[2]^rk[9]
668	str	r1,[r11,#-28]
669	eor	r3,r3,r2			@ rk[11]=rk[3]^rk[10]
670	str	r2,[r11,#-24]
671	subs	r12,r12,#1
672	str	r3,[r11,#-20]
673#if __ARM_ARCH__>=7
674	itt	eq				@ Thumb2 thing, sanity check in ARM
675#endif
676	subeq	r2,r11,#256
677	beq	.Ldone
678
679	and	r5,lr,r3
680	and	r7,lr,r3,lsr#8
681	ldrb	r5,[r10,r5]
682	and	r8,lr,r3,lsr#16
683	ldrb	r7,[r10,r7]
684	and	r9,lr,r3,lsr#24
685	ldrb	r8,[r10,r8]
686	orr	r5,r5,r7,lsl#8
687	ldrb	r9,[r10,r9]
688	orr	r5,r5,r8,lsl#16
689	ldr	r4,[r11,#-48]
690	orr	r5,r5,r9,lsl#24
691
692	ldr	r7,[r11,#-44]
693	ldr	r8,[r11,#-40]
694	eor	r4,r4,r5			@ rk[12]=rk[4]^...
695	ldr	r9,[r11,#-36]
696	eor	r7,r7,r4			@ rk[13]=rk[5]^rk[12]
697	str	r4,[r11,#-16]
698	eor	r8,r8,r7			@ rk[14]=rk[6]^rk[13]
699	str	r7,[r11,#-12]
700	eor	r9,r9,r8			@ rk[15]=rk[7]^rk[14]
701	str	r8,[r11,#-8]
702	str	r9,[r11,#-4]
703	b	.L256_loop
704
705.align	2
706.Ldone:	mov	r0,#0
707	ldmia	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
708.Labrt:
709#if __ARM_ARCH__>=5
710	bx	lr				@ .word	0xe12fff1e
711#else
712	tst	lr,#1
713	moveq	pc,lr			@ be binary compatible with V4, yet
714.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
715#endif
716.size	asm_AES_set_encrypt_key,.-asm_AES_set_encrypt_key
717
718.globl	asm_AES_set_decrypt_key
719.hidden	asm_AES_set_decrypt_key
720.type	asm_AES_set_decrypt_key,%function
721.align	5
722asm_AES_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			@ asm_AES_set_encrypt_key preserves r2,
730	mov	r1,r2			@ which is AES_KEY *key
731	b	_armv4_AES_set_enc2dec_key
732.size	asm_AES_set_decrypt_key,.-asm_AES_set_decrypt_key
733
734@ void AES_set_enc2dec_key(const AES_KEY *inp,AES_KEY *out)
735.globl	AES_set_enc2dec_key
736.hidden	AES_set_enc2dec_key
737.type	AES_set_enc2dec_key,%function
738.align	5
739AES_set_enc2dec_key:
740_armv4_AES_set_enc2dec_key:
741	stmdb	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
742
743	ldr	r12,[r0,#240]
744	mov	r7,r0			@ input
745	add	r8,r0,r12,lsl#4
746	mov	r11,r1			@ ouput
747	add	r10,r1,r12,lsl#4
748	str	r12,[r1,#240]
749
750.Linv:	ldr	r0,[r7],#16
751	ldr	r1,[r7,#-12]
752	ldr	r2,[r7,#-8]
753	ldr	r3,[r7,#-4]
754	ldr	r4,[r8],#-16
755	ldr	r5,[r8,#16+4]
756	ldr	r6,[r8,#16+8]
757	ldr	r9,[r8,#16+12]
758	str	r0,[r10],#-16
759	str	r1,[r10,#16+4]
760	str	r2,[r10,#16+8]
761	str	r3,[r10,#16+12]
762	str	r4,[r11],#16
763	str	r5,[r11,#-12]
764	str	r6,[r11,#-8]
765	str	r9,[r11,#-4]
766	teq	r7,r8
767	bne	.Linv
768
769	ldr	r0,[r7]
770	ldr	r1,[r7,#4]
771	ldr	r2,[r7,#8]
772	ldr	r3,[r7,#12]
773	str	r0,[r11]
774	str	r1,[r11,#4]
775	str	r2,[r11,#8]
776	str	r3,[r11,#12]
777	sub	r11,r11,r12,lsl#3
778	ldr	r0,[r11,#16]!		@ prefetch tp1
779	mov	r7,#0x80
780	mov	r8,#0x1b
781	orr	r7,r7,#0x8000
782	orr	r8,r8,#0x1b00
783	orr	r7,r7,r7,lsl#16
784	orr	r8,r8,r8,lsl#16
785	sub	r12,r12,#1
786	mvn	r9,r7
787	mov	r12,r12,lsl#2	@ (rounds-1)*4
788
789.Lmix:	and	r4,r0,r7
790	and	r1,r0,r9
791	sub	r4,r4,r4,lsr#7
792	and	r4,r4,r8
793	eor	r1,r4,r1,lsl#1	@ tp2
794
795	and	r4,r1,r7
796	and	r2,r1,r9
797	sub	r4,r4,r4,lsr#7
798	and	r4,r4,r8
799	eor	r2,r4,r2,lsl#1	@ tp4
800
801	and	r4,r2,r7
802	and	r3,r2,r9
803	sub	r4,r4,r4,lsr#7
804	and	r4,r4,r8
805	eor	r3,r4,r3,lsl#1	@ tp8
806
807	eor	r4,r1,r2
808	eor	r5,r0,r3		@ tp9
809	eor	r4,r4,r3		@ tpe
810	eor	r4,r4,r1,ror#24
811	eor	r4,r4,r5,ror#24	@ ^= ROTATE(tpb=tp9^tp2,8)
812	eor	r4,r4,r2,ror#16
813	eor	r4,r4,r5,ror#16	@ ^= ROTATE(tpd=tp9^tp4,16)
814	eor	r4,r4,r5,ror#8	@ ^= ROTATE(tp9,24)
815
816	ldr	r0,[r11,#4]		@ prefetch tp1
817	str	r4,[r11],#4
818	subs	r12,r12,#1
819	bne	.Lmix
820
821	mov	r0,#0
822#if __ARM_ARCH__>=5
823	ldmia	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,pc}
824#else
825	ldmia	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
826	tst	lr,#1
827	moveq	pc,lr			@ be binary compatible with V4, yet
828.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
829#endif
830.size	AES_set_enc2dec_key,.-AES_set_enc2dec_key
831
832.type	AES_Td,%object
833.align	5
834AES_Td:
835.word	0x51f4a750, 0x7e416553, 0x1a17a4c3, 0x3a275e96
836.word	0x3bab6bcb, 0x1f9d45f1, 0xacfa58ab, 0x4be30393
837.word	0x2030fa55, 0xad766df6, 0x88cc7691, 0xf5024c25
838.word	0x4fe5d7fc, 0xc52acbd7, 0x26354480, 0xb562a38f
839.word	0xdeb15a49, 0x25ba1b67, 0x45ea0e98, 0x5dfec0e1
840.word	0xc32f7502, 0x814cf012, 0x8d4697a3, 0x6bd3f9c6
841.word	0x038f5fe7, 0x15929c95, 0xbf6d7aeb, 0x955259da
842.word	0xd4be832d, 0x587421d3, 0x49e06929, 0x8ec9c844
843.word	0x75c2896a, 0xf48e7978, 0x99583e6b, 0x27b971dd
844.word	0xbee14fb6, 0xf088ad17, 0xc920ac66, 0x7dce3ab4
845.word	0x63df4a18, 0xe51a3182, 0x97513360, 0x62537f45
846.word	0xb16477e0, 0xbb6bae84, 0xfe81a01c, 0xf9082b94
847.word	0x70486858, 0x8f45fd19, 0x94de6c87, 0x527bf8b7
848.word	0xab73d323, 0x724b02e2, 0xe31f8f57, 0x6655ab2a
849.word	0xb2eb2807, 0x2fb5c203, 0x86c57b9a, 0xd33708a5
850.word	0x302887f2, 0x23bfa5b2, 0x02036aba, 0xed16825c
851.word	0x8acf1c2b, 0xa779b492, 0xf307f2f0, 0x4e69e2a1
852.word	0x65daf4cd, 0x0605bed5, 0xd134621f, 0xc4a6fe8a
853.word	0x342e539d, 0xa2f355a0, 0x058ae132, 0xa4f6eb75
854.word	0x0b83ec39, 0x4060efaa, 0x5e719f06, 0xbd6e1051
855.word	0x3e218af9, 0x96dd063d, 0xdd3e05ae, 0x4de6bd46
856.word	0x91548db5, 0x71c45d05, 0x0406d46f, 0x605015ff
857.word	0x1998fb24, 0xd6bde997, 0x894043cc, 0x67d99e77
858.word	0xb0e842bd, 0x07898b88, 0xe7195b38, 0x79c8eedb
859.word	0xa17c0a47, 0x7c420fe9, 0xf8841ec9, 0x00000000
860.word	0x09808683, 0x322bed48, 0x1e1170ac, 0x6c5a724e
861.word	0xfd0efffb, 0x0f853856, 0x3daed51e, 0x362d3927
862.word	0x0a0fd964, 0x685ca621, 0x9b5b54d1, 0x24362e3a
863.word	0x0c0a67b1, 0x9357e70f, 0xb4ee96d2, 0x1b9b919e
864.word	0x80c0c54f, 0x61dc20a2, 0x5a774b69, 0x1c121a16
865.word	0xe293ba0a, 0xc0a02ae5, 0x3c22e043, 0x121b171d
866.word	0x0e090d0b, 0xf28bc7ad, 0x2db6a8b9, 0x141ea9c8
867.word	0x57f11985, 0xaf75074c, 0xee99ddbb, 0xa37f60fd
868.word	0xf701269f, 0x5c72f5bc, 0x44663bc5, 0x5bfb7e34
869.word	0x8b432976, 0xcb23c6dc, 0xb6edfc68, 0xb8e4f163
870.word	0xd731dcca, 0x42638510, 0x13972240, 0x84c61120
871.word	0x854a247d, 0xd2bb3df8, 0xaef93211, 0xc729a16d
872.word	0x1d9e2f4b, 0xdcb230f3, 0x0d8652ec, 0x77c1e3d0
873.word	0x2bb3166c, 0xa970b999, 0x119448fa, 0x47e96422
874.word	0xa8fc8cc4, 0xa0f03f1a, 0x567d2cd8, 0x223390ef
875.word	0x87494ec7, 0xd938d1c1, 0x8ccaa2fe, 0x98d40b36
876.word	0xa6f581cf, 0xa57ade28, 0xdab78e26, 0x3fadbfa4
877.word	0x2c3a9de4, 0x5078920d, 0x6a5fcc9b, 0x547e4662
878.word	0xf68d13c2, 0x90d8b8e8, 0x2e39f75e, 0x82c3aff5
879.word	0x9f5d80be, 0x69d0937c, 0x6fd52da9, 0xcf2512b3
880.word	0xc8ac993b, 0x10187da7, 0xe89c636e, 0xdb3bbb7b
881.word	0xcd267809, 0x6e5918f4, 0xec9ab701, 0x834f9aa8
882.word	0xe6956e65, 0xaaffe67e, 0x21bccf08, 0xef15e8e6
883.word	0xbae79bd9, 0x4a6f36ce, 0xea9f09d4, 0x29b07cd6
884.word	0x31a4b2af, 0x2a3f2331, 0xc6a59430, 0x35a266c0
885.word	0x744ebc37, 0xfc82caa6, 0xe090d0b0, 0x33a7d815
886.word	0xf104984a, 0x41ecdaf7, 0x7fcd500e, 0x1791f62f
887.word	0x764dd68d, 0x43efb04d, 0xccaa4d54, 0xe49604df
888.word	0x9ed1b5e3, 0x4c6a881b, 0xc12c1fb8, 0x4665517f
889.word	0x9d5eea04, 0x018c355d, 0xfa877473, 0xfb0b412e
890.word	0xb3671d5a, 0x92dbd252, 0xe9105633, 0x6dd64713
891.word	0x9ad7618c, 0x37a10c7a, 0x59f8148e, 0xeb133c89
892.word	0xcea927ee, 0xb761c935, 0xe11ce5ed, 0x7a47b13c
893.word	0x9cd2df59, 0x55f2733f, 0x1814ce79, 0x73c737bf
894.word	0x53f7cdea, 0x5ffdaa5b, 0xdf3d6f14, 0x7844db86
895.word	0xcaaff381, 0xb968c43e, 0x3824342c, 0xc2a3405f
896.word	0x161dc372, 0xbce2250c, 0x283c498b, 0xff0d9541
897.word	0x39a80171, 0x080cb3de, 0xd8b4e49c, 0x6456c190
898.word	0x7bcb8461, 0xd532b670, 0x486c5c74, 0xd0b85742
899@ Td4[256]
900.byte	0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38
901.byte	0xbf, 0x40, 0xa3, 0x9e, 0x81, 0xf3, 0xd7, 0xfb
902.byte	0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87
903.byte	0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb
904.byte	0x54, 0x7b, 0x94, 0x32, 0xa6, 0xc2, 0x23, 0x3d
905.byte	0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e
906.byte	0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2
907.byte	0x76, 0x5b, 0xa2, 0x49, 0x6d, 0x8b, 0xd1, 0x25
908.byte	0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16
909.byte	0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92
910.byte	0x6c, 0x70, 0x48, 0x50, 0xfd, 0xed, 0xb9, 0xda
911.byte	0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84
912.byte	0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a
913.byte	0xf7, 0xe4, 0x58, 0x05, 0xb8, 0xb3, 0x45, 0x06
914.byte	0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02
915.byte	0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b
916.byte	0x3a, 0x91, 0x11, 0x41, 0x4f, 0x67, 0xdc, 0xea
917.byte	0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73
918.byte	0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85
919.byte	0xe2, 0xf9, 0x37, 0xe8, 0x1c, 0x75, 0xdf, 0x6e
920.byte	0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89
921.byte	0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b
922.byte	0xfc, 0x56, 0x3e, 0x4b, 0xc6, 0xd2, 0x79, 0x20
923.byte	0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4
924.byte	0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31
925.byte	0xb1, 0x12, 0x10, 0x59, 0x27, 0x80, 0xec, 0x5f
926.byte	0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d
927.byte	0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef
928.byte	0xa0, 0xe0, 0x3b, 0x4d, 0xae, 0x2a, 0xf5, 0xb0
929.byte	0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61
930.byte	0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26
931.byte	0xe1, 0x69, 0x14, 0x63, 0x55, 0x21, 0x0c, 0x7d
932.size	AES_Td,.-AES_Td
933
934@ void asm_AES_decrypt(const unsigned char *in, unsigned char *out,
935@ 		       const AES_KEY *key) {
936.globl	asm_AES_decrypt
937.hidden	asm_AES_decrypt
938.type	asm_AES_decrypt,%function
939.align	5
940asm_AES_decrypt:
941#if __ARM_ARCH__<7
942	sub	r3,pc,#8		@ asm_AES_decrypt
943#else
944	adr	r3,asm_AES_decrypt
945#endif
946	stmdb	sp!,{r1,r4-r12,lr}
947#ifdef	__APPLE__
948	adr	r10,AES_Td
949#else
950	sub	r10,r3,#asm_AES_decrypt-AES_Td	@ Td
951#endif
952	mov	r12,r0		@ inp
953	mov	r11,r2
954#if __ARM_ARCH__<7
955	ldrb	r0,[r12,#3]	@ load input data in endian-neutral
956	ldrb	r4,[r12,#2]	@ manner...
957	ldrb	r5,[r12,#1]
958	ldrb	r6,[r12,#0]
959	orr	r0,r0,r4,lsl#8
960	ldrb	r1,[r12,#7]
961	orr	r0,r0,r5,lsl#16
962	ldrb	r4,[r12,#6]
963	orr	r0,r0,r6,lsl#24
964	ldrb	r5,[r12,#5]
965	ldrb	r6,[r12,#4]
966	orr	r1,r1,r4,lsl#8
967	ldrb	r2,[r12,#11]
968	orr	r1,r1,r5,lsl#16
969	ldrb	r4,[r12,#10]
970	orr	r1,r1,r6,lsl#24
971	ldrb	r5,[r12,#9]
972	ldrb	r6,[r12,#8]
973	orr	r2,r2,r4,lsl#8
974	ldrb	r3,[r12,#15]
975	orr	r2,r2,r5,lsl#16
976	ldrb	r4,[r12,#14]
977	orr	r2,r2,r6,lsl#24
978	ldrb	r5,[r12,#13]
979	ldrb	r6,[r12,#12]
980	orr	r3,r3,r4,lsl#8
981	orr	r3,r3,r5,lsl#16
982	orr	r3,r3,r6,lsl#24
983#else
984	ldr	r0,[r12,#0]
985	ldr	r1,[r12,#4]
986	ldr	r2,[r12,#8]
987	ldr	r3,[r12,#12]
988#ifdef __ARMEL__
989	rev	r0,r0
990	rev	r1,r1
991	rev	r2,r2
992	rev	r3,r3
993#endif
994#endif
995	bl	_armv4_AES_decrypt
996
997	ldr	r12,[sp],#4		@ pop out
998#if __ARM_ARCH__>=7
999#ifdef __ARMEL__
1000	rev	r0,r0
1001	rev	r1,r1
1002	rev	r2,r2
1003	rev	r3,r3
1004#endif
1005	str	r0,[r12,#0]
1006	str	r1,[r12,#4]
1007	str	r2,[r12,#8]
1008	str	r3,[r12,#12]
1009#else
1010	mov	r4,r0,lsr#24		@ write output in endian-neutral
1011	mov	r5,r0,lsr#16		@ manner...
1012	mov	r6,r0,lsr#8
1013	strb	r4,[r12,#0]
1014	strb	r5,[r12,#1]
1015	mov	r4,r1,lsr#24
1016	strb	r6,[r12,#2]
1017	mov	r5,r1,lsr#16
1018	strb	r0,[r12,#3]
1019	mov	r6,r1,lsr#8
1020	strb	r4,[r12,#4]
1021	strb	r5,[r12,#5]
1022	mov	r4,r2,lsr#24
1023	strb	r6,[r12,#6]
1024	mov	r5,r2,lsr#16
1025	strb	r1,[r12,#7]
1026	mov	r6,r2,lsr#8
1027	strb	r4,[r12,#8]
1028	strb	r5,[r12,#9]
1029	mov	r4,r3,lsr#24
1030	strb	r6,[r12,#10]
1031	mov	r5,r3,lsr#16
1032	strb	r2,[r12,#11]
1033	mov	r6,r3,lsr#8
1034	strb	r4,[r12,#12]
1035	strb	r5,[r12,#13]
1036	strb	r6,[r12,#14]
1037	strb	r3,[r12,#15]
1038#endif
1039#if __ARM_ARCH__>=5
1040	ldmia	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,pc}
1041#else
1042	ldmia	sp!,{r4,r5,r6,r7,r8,r9,r10,r11,r12,lr}
1043	tst	lr,#1
1044	moveq	pc,lr			@ be binary compatible with V4, yet
1045.word	0xe12fff1e			@ interoperable with Thumb ISA:-)
1046#endif
1047.size	asm_AES_decrypt,.-asm_AES_decrypt
1048
1049.type	_armv4_AES_decrypt,%function
1050.align	2
1051_armv4_AES_decrypt:
1052	str	lr,[sp,#-4]!		@ push lr
1053	ldmia	r11!,{r4,r5,r6,r7}
1054	eor	r0,r0,r4
1055	ldr	r12,[r11,#240-16]
1056	eor	r1,r1,r5
1057	eor	r2,r2,r6
1058	eor	r3,r3,r7
1059	sub	r12,r12,#1
1060	mov	lr,#255
1061
1062	and	r7,lr,r0,lsr#16
1063	and	r8,lr,r0,lsr#8
1064	and	r9,lr,r0
1065	mov	r0,r0,lsr#24
1066.Ldec_loop:
1067	ldr	r4,[r10,r7,lsl#2]	@ Td1[s0>>16]
1068	and	r7,lr,r1		@ i0
1069	ldr	r5,[r10,r8,lsl#2]	@ Td2[s0>>8]
1070	and	r8,lr,r1,lsr#16
1071	ldr	r6,[r10,r9,lsl#2]	@ Td3[s0>>0]
1072	and	r9,lr,r1,lsr#8
1073	ldr	r0,[r10,r0,lsl#2]	@ Td0[s0>>24]
1074	mov	r1,r1,lsr#24
1075
1076	ldr	r7,[r10,r7,lsl#2]	@ Td3[s1>>0]
1077	ldr	r8,[r10,r8,lsl#2]	@ Td1[s1>>16]
1078	ldr	r9,[r10,r9,lsl#2]	@ Td2[s1>>8]
1079	eor	r0,r0,r7,ror#24
1080	ldr	r1,[r10,r1,lsl#2]	@ Td0[s1>>24]
1081	and	r7,lr,r2,lsr#8	@ i0
1082	eor	r5,r8,r5,ror#8
1083	and	r8,lr,r2		@ i1
1084	eor	r6,r9,r6,ror#8
1085	and	r9,lr,r2,lsr#16
1086	ldr	r7,[r10,r7,lsl#2]	@ Td2[s2>>8]
1087	eor	r1,r1,r4,ror#8
1088	ldr	r8,[r10,r8,lsl#2]	@ Td3[s2>>0]
1089	mov	r2,r2,lsr#24
1090
1091	ldr	r9,[r10,r9,lsl#2]	@ Td1[s2>>16]
1092	eor	r0,r0,r7,ror#16
1093	ldr	r2,[r10,r2,lsl#2]	@ Td0[s2>>24]
1094	and	r7,lr,r3,lsr#16	@ i0
1095	eor	r1,r1,r8,ror#24
1096	and	r8,lr,r3,lsr#8	@ i1
1097	eor	r6,r9,r6,ror#8
1098	and	r9,lr,r3		@ i2
1099	ldr	r7,[r10,r7,lsl#2]	@ Td1[s3>>16]
1100	eor	r2,r2,r5,ror#8
1101	ldr	r8,[r10,r8,lsl#2]	@ Td2[s3>>8]
1102	mov	r3,r3,lsr#24
1103
1104	ldr	r9,[r10,r9,lsl#2]	@ Td3[s3>>0]
1105	eor	r0,r0,r7,ror#8
1106	ldr	r7,[r11],#16
1107	eor	r1,r1,r8,ror#16
1108	ldr	r3,[r10,r3,lsl#2]	@ Td0[s3>>24]
1109	eor	r2,r2,r9,ror#24
1110
1111	ldr	r4,[r11,#-12]
1112	eor	r0,r0,r7
1113	ldr	r5,[r11,#-8]
1114	eor	r3,r3,r6,ror#8
1115	ldr	r6,[r11,#-4]
1116	and	r7,lr,r0,lsr#16
1117	eor	r1,r1,r4
1118	and	r8,lr,r0,lsr#8
1119	eor	r2,r2,r5
1120	and	r9,lr,r0
1121	eor	r3,r3,r6
1122	mov	r0,r0,lsr#24
1123
1124	subs	r12,r12,#1
1125	bne	.Ldec_loop
1126
1127	add	r10,r10,#1024
1128
1129	ldr	r5,[r10,#0]		@ prefetch Td4
1130	ldr	r6,[r10,#32]
1131	ldr	r4,[r10,#64]
1132	ldr	r5,[r10,#96]
1133	ldr	r6,[r10,#128]
1134	ldr	r4,[r10,#160]
1135	ldr	r5,[r10,#192]
1136	ldr	r6,[r10,#224]
1137
1138	ldrb	r0,[r10,r0]		@ Td4[s0>>24]
1139	ldrb	r4,[r10,r7]		@ Td4[s0>>16]
1140	and	r7,lr,r1		@ i0
1141	ldrb	r5,[r10,r8]		@ Td4[s0>>8]
1142	and	r8,lr,r1,lsr#16
1143	ldrb	r6,[r10,r9]		@ Td4[s0>>0]
1144	and	r9,lr,r1,lsr#8
1145
1146	add	r1,r10,r1,lsr#24
1147	ldrb	r7,[r10,r7]		@ Td4[s1>>0]
1148	ldrb	r1,[r1]		@ Td4[s1>>24]
1149	ldrb	r8,[r10,r8]		@ Td4[s1>>16]
1150	eor	r0,r7,r0,lsl#24
1151	ldrb	r9,[r10,r9]		@ Td4[s1>>8]
1152	eor	r1,r4,r1,lsl#8
1153	and	r7,lr,r2,lsr#8	@ i0
1154	eor	r5,r5,r8,lsl#8
1155	and	r8,lr,r2		@ i1
1156	ldrb	r7,[r10,r7]		@ Td4[s2>>8]
1157	eor	r6,r6,r9,lsl#8
1158	ldrb	r8,[r10,r8]		@ Td4[s2>>0]
1159	and	r9,lr,r2,lsr#16
1160
1161	add	r2,r10,r2,lsr#24
1162	ldrb	r2,[r2]		@ Td4[s2>>24]
1163	eor	r0,r0,r7,lsl#8
1164	ldrb	r9,[r10,r9]		@ Td4[s2>>16]
1165	eor	r1,r8,r1,lsl#16
1166	and	r7,lr,r3,lsr#16	@ i0
1167	eor	r2,r5,r2,lsl#16
1168	and	r8,lr,r3,lsr#8	@ i1
1169	ldrb	r7,[r10,r7]		@ Td4[s3>>16]
1170	eor	r6,r6,r9,lsl#16
1171	ldrb	r8,[r10,r8]		@ Td4[s3>>8]
1172	and	r9,lr,r3		@ i2
1173
1174	add	r3,r10,r3,lsr#24
1175	ldrb	r9,[r10,r9]		@ Td4[s3>>0]
1176	ldrb	r3,[r3]		@ Td4[s3>>24]
1177	eor	r0,r0,r7,lsl#16
1178	ldr	r7,[r11,#0]
1179	eor	r1,r1,r8,lsl#8
1180	ldr	r4,[r11,#4]
1181	eor	r2,r9,r2,lsl#8
1182	ldr	r5,[r11,#8]
1183	eor	r3,r6,r3,lsl#24
1184	ldr	r6,[r11,#12]
1185
1186	eor	r0,r0,r7
1187	eor	r1,r1,r4
1188	eor	r2,r2,r5
1189	eor	r3,r3,r6
1190
1191	sub	r10,r10,#1024
1192	ldr	pc,[sp],#4		@ pop and return
1193.size	_armv4_AES_decrypt,.-_armv4_AES_decrypt
1194.byte	65,69,83,32,102,111,114,32,65,82,77,118,52,44,32,67,82,89,80,84,79,71,65,77,83,32,98,121,32,60,97,112,112,114,111,64,111,112,101,110,115,115,108,46,111,114,103,62,0
1195.align	2
1196.align	2
1197
1198#endif
1199