1 .text
2 
3 .p2align 2,0x90
4 #ifdef __Darwin__
5 .globl __sha1block
6 __sha1block:
7 #else
8 .globl _sha1block
9 _sha1block:
10 #endif
11 
12 /* x = (wp[off-f] ^ wp[off-8] ^ wp[off-14] ^ wp[off-16]) <<< 1;
13  * wp[off] = x;
14  * x += A <<< 5;
15  * E += 0xca62c1d6 + x;
16  * x = FN(B,C,D);
17  * E += x;
18  * B >>> 2
19  */
20 #define BSWAPDI	BYTE $0x0f; BYTE $0xcf;
21 
22 #define BODY(off,FN,V,A,B,C,D,E)\
23 	movl (off-64)(%ebp), %edi;\
24 	xorl (off-56)(%ebp), %edi;\
25 	xorl (off-32)(%ebp), %edi;\
26 	xorl (off-12)(%ebp), %edi;\
27 	roll $1, %edi;\
28 	movl %edi, off(%ebp);\
29 	leal V(%edi, E, 1), E;\
30 	movl A, %edi;\
31 	roll $5, %edi;\
32 	addl %edi, E;\
33 	FN(B,C,D)\
34 	addl %edi, E;\
35 	rorl $2, B;\
36 
37 #define BODY0(off,FN,V,A,B,C,D,E)\
38 	movl off(%ebx), %edi;\
39 	bswap %edi;\
40 	movl %edi, off(%ebp);\
41 	leal V(%edi,E,1), E;\
42 	movl A, %edi;\
43 	roll $5,%edi;\
44 	addl %edi,E;\
45 	FN(B,C,D)\
46 	addl %edi,E;\
47 	rorl $2,B;\
48 
49 /*
50  * fn1 = (((C^D)&B)^D);
51  */
52 #define FN1(B,C,D)\
53 	movl C, %edi;\
54 	xorl D, %edi;\
55 	andl B, %edi;\
56 	xorl D, %edi;\
57 
58 /*
59  * fn24 = B ^ C ^ D
60  */
61 #define FN24(B,C,D)\
62 	movl B, %edi;\
63 	xorl C, %edi;\
64 	xorl D, %edi;\
65 
66 /*
67  * fn3 = ((B ^ C) & (D ^= B)) ^ B
68  * D ^= B to restore D
69  */
70 #define FN3(B,C,D)\
71 	movl B, %edi;\
72 	xorl C, %edi;\
73 	xorl B, D;\
74 	andl D, %edi;\
75 	xorl B, %edi;\
76 	xorl B, D;\
77 
78 /*
79  * stack offsets
80  * void sha1block(uchar *DATA, int LEN, ulong *STATE)
81  */
82 #define STACKSIZE (48+80*4)
83 #define	DATA	(STACKSIZE+8)
84 #define	LEN	(STACKSIZE+12)
85 #define	STATE	(STACKSIZE+16)
86 
87 /*
88  * stack offsets for locals
89  * ulong w[80];
90  * uchar *edata;
91  * ulong *w15, *w40, *w60, *w80;
92  * register local
93  * ulong *wp = %ebp
94  * ulong a = eax, b = ebx, c = ecx, d = edx, e = esi
95  * ulong tmp = edi
96  */
97 #define WARRAY	(STACKSIZE-4-(80*4))
98 #define TMP1	(STACKSIZE-8-(80*4))
99 #define TMP2	(STACKSIZE-12-(80*4))
100 #define W15	(STACKSIZE-16-(80*4))
101 #define W40	(STACKSIZE-20-(80*4))
102 #define W60	(STACKSIZE-24-(80*4))
103 #define W80	(STACKSIZE-28-(80*4))
104 #define EDATA	(STACKSIZE-32-(80*4))
105 #define OLDEBX	(STACKSIZE-36-(80*4))
106 #define OLDESI	(STACKSIZE-40-(80*4))
107 #define OLDEDI	(STACKSIZE-44-(80*4))
108 
109 	/* Prelude */
110 	pushl %ebp
111 	subl $(STACKSIZE), %esp
112 
113 	mov %ebx, OLDEBX(%esp)
114 	mov %esi, OLDESI(%esp)
115 	mov %edi, OLDEDI(%esp)
116 
117 	movl DATA(%esp), %eax
118 	addl LEN(%esp), %eax
119 	movl %eax, EDATA(%esp)
120 
121 	leal (WARRAY+15*4)(%esp), %edi	/* aw15 */
122 	movl %edi, W15(%esp)
123 	leal (WARRAY+40*4)(%esp), %edx	/* aw40 */
124 	movl %edx, W40(%esp)
125 	leal (WARRAY+60*4)(%esp), %ecx	/* aw60 */
126 	movl %ecx, W60(%esp)
127 	leal (WARRAY+80*4)(%esp), %edi	/* aw80 */
128 	movl %edi, W80(%esp)
129 
130 0:
131 	leal WARRAY(%esp), %ebp		/* warray */
132 
133 	movl STATE(%esp), %edi		/* state */
134 	movl (%edi),%eax
135 	movl 4(%edi),%ebx
136 	movl %ebx, TMP1(%esp)		/* tmp1 */
137 	movl 8(%edi), %ecx
138 	movl 12(%edi), %edx
139 	movl 16(%edi), %esi
140 
141 	movl DATA(%esp), %ebx		/* data */
142 
143 1:
144 	BODY0(0,FN1,0x5a827999,%eax,TMP1(%esp),%ecx,%edx,%esi)
145 	movl %esi,TMP2(%esp)
146 	BODY0(4,FN1,0x5a827999,%esi,%eax,TMP1(%esp),%ecx,%edx)
147 	movl TMP1(%esp),%esi
148 	BODY0(8,FN1,0x5a827999,%edx,TMP2(%esp),%eax,%esi,%ecx)
149 	BODY0(12,FN1,0x5a827999,%ecx,%edx,TMP2(%esp),%eax,%esi)
150 	movl %esi,TMP1(%esp)
151 	BODY0(16,FN1,0x5a827999,%esi,%ecx,%edx,TMP2(%esp),%eax)
152 	movl TMP2(%esp),%esi
153 
154 	addl $20, %ebx
155 	addl $20, %ebp
156 	cmpl W15(%esp), %ebp	/* w15 */
157 	jb 1b
158 
159 	BODY0(0,FN1,0x5a827999,%eax,TMP1(%esp),%ecx,%edx,%esi)
160 	addl $4, %ebx
161 	MOVL %ebx, DATA(%esp)	/* data */
162 	MOVL TMP1(%esp),%ebx
163 
164 	BODY(4,FN1,0x5a827999,%esi,%eax,%ebx,%ecx,%edx)
165 	BODY(8,FN1,0x5a827999,%edx,%esi,%eax,%ebx,%ecx)
166 	BODY(12,FN1,0x5a827999,%ecx,%edx,%esi,%eax,%ebx)
167 	BODY(16,FN1,0x5a827999,%ebx,%ecx,%edx,%esi,%eax)
168 
169 	addl $20, %ebp
170 
171 2:
172 	BODY(0,FN24,0x6ed9eba1,%eax,%ebx,%ecx,%edx,%esi)
173 	BODY(4,FN24,0x6ed9eba1,%esi,%eax,%ebx,%ecx,%edx)
174 	BODY(8,FN24,0x6ed9eba1,%edx,%esi,%eax,%ebx,%ecx)
175 	BODY(12,FN24,0x6ed9eba1,%ecx,%edx,%esi,%eax,%ebx)
176 	BODY(16,FN24,0x6ed9eba1,%ebx,%ecx,%edx,%esi,%eax)
177 
178 	addl $20,%ebp
179 	cmpl W40(%esp), %ebp
180 	jb 2b
181 
182 3:
183 	BODY(0,FN3,0x8f1bbcdc,%eax,%ebx,%ecx,%edx,%esi)
184 	BODY(4,FN3,0x8f1bbcdc,%esi,%eax,%ebx,%ecx,%edx)
185 	BODY(8,FN3,0x8f1bbcdc,%edx,%esi,%eax,%ebx,%ecx)
186 	BODY(12,FN3,0x8f1bbcdc,%ecx,%edx,%esi,%eax,%ebx)
187 	BODY(16,FN3,0x8f1bbcdc,%ebx,%ecx,%edx,%esi,%eax)
188 
189 	addl $20, %ebp
190 	cmpl W60(%esp), %ebp 	/* w60 */
191 	jb 3b
192 
193 4:
194 	BODY(0,FN24,0xca62c1d6,%eax,%ebx,%ecx,%edx,%esi)
195 	BODY(4,FN24,0xca62c1d6,%esi,%eax,%ebx,%ecx,%edx)
196 	BODY(8,FN24,0xca62c1d6,%edx,%esi,%eax,%ebx,%ecx)
197 	BODY(12,FN24,0xca62c1d6,%ecx,%edx,%esi,%eax,%ebx)
198 	BODY(16,FN24,0xca62c1d6,%ebx,%ecx,%edx,%esi,%eax)
199 
200 	addl $20, %ebp
201 	cmpl W80(%esp), %ebp 	/* w80 */
202 	jb 4b
203 
204 	movl STATE(%esp), %edi	/* state */
205 	addl %eax, 0(%edi)
206 	addl %ebx, 4(%edi)
207 	addl %ecx, 8(%edi)
208 	addl %edx, 12(%edi)
209 	addl %esi, 16(%edi)
210 
211 	movl EDATA(%esp), %edi	/* edata */
212 	cmpl %edi, DATA(%esp)	/* data */
213 	jb 0b
214 
215 	/* Postlude */
216 	mov OLDEBX(%esp), %ebx
217 	mov OLDESI(%esp), %esi
218 	mov OLDEDI(%esp), %edi
219 	addl $(STACKSIZE), %esp
220 	popl %ebp
221 	ret