Radix4FFT_v5.s revision e2e838afcf03e603a41a0455846eaf9614537c16 
1@/*
2@ ** Copyright 2003-2010, VisualOn, Inc.
3@ **
4@ ** Licensed under the Apache License, Version 2.0 (the "License");
5@ ** you may not use this file except in compliance with the License.
6@ ** You may obtain a copy of the License at
7@ **
8@ **     http://www.apache.org/licenses/LICENSE-2.0
9@ **
10@ ** Unless required by applicable law or agreed to in writing, software
11@ ** distributed under the License is distributed on an "AS IS" BASIS,
12@ ** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13@ ** See the License for the specific language governing permissions and
14@ ** limitations under the License.
15@ */
16
17@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
18@	File:		Radix4FFT_v5.s
19@
20@	Content:	Radix4FFT armv5 assemble
21@
22@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
23	.section .text
24	.global	Radix4FFT
25
26Radix4FFT:
27	stmdb     sp!, {r4 - r11, lr}
28	sub       sp, sp, #32
29
30	mov			r1, r1, asr #2
31	cmp     r1, #0
32	beq     Radix4FFT_END
33
34Radix4FFT_LOOP1:
35	mov     r14, r0          							@ xptr = buf@
36	mov		r10, r1 												@ i = num@
37	mov     r9, r2, lsl #3  							@ step = 2*bgn@
38	cmp     r10, #0
39	str		r0, [sp]
40	str		r1, [sp, #4]
41	str		r2, [sp, #8]
42	str		r3, [sp, #12]
43	beq     Radix4FFT_LOOP1_END
44
45Radix4FFT_LOOP2:
46	mov     r12, r3				        				@ csptr = twidTab@
47	mov		r11, r2												@ j = bgn
48	cmp     r11, #0
49	str		r10, [sp, #16]
50	beq     Radix4FFT_LOOP2_END
51
52Radix4FFT_LOOP3:
53	str			r11, [sp, #20]
54
55	ldrd		r0, [r14, #0]									@ r0 = xptr[0]@ r1 = xptr[1]@
56	add			r14, r14, r9 	 								@ xptr += step@
57
58	ldrd		r10,	[r14, #0]  					 			@ r2 = xptr[0]@ r3 = xptr[1]@
59	ldr			r8, [r12], #4									@ cosxsinx = csptr[0]@
60
61	smulwt	r4, r10, r8										@ L_mpy_wx(cosx, t0)
62	smulwt	r3, r11, r8										@ L_mpy_wx(cosx, t1)
63
64	smlawb	r2, r11, r8, r4								@ r2 = L_mpy_wx(cosx, t0) + L_mpy_wx(sinx, t1)@
65	smulwb	r5, r10, r8										@ L_mpy_wx(sinx, t0)
66
67	mov			r10, r0, asr #2								@ t0 = r0 >> 2@
68	mov			r11, r1, asr #2								@	t1 = r1 >> 2@
69
70	sub			r3, r3, r5										@ r3 = L_mpy_wx(cosx, t1) - L_mpy_wx(sinx, t0)@
71	add     r14, r14, r9 	 								@ xptr += step@
72
73	sub			r0, r10, r2										@ r0 = t0 - r2@
74	sub			r1, r11, r3									  @ r1 = t1 - r3@
75
76	add			r2, r10, r2										@ r2 = t0 + r2@
77	add			r3, r11, r3										@ r3 = t1 + r3@
78
79	str			r2, [sp, #24]
80	str			r3, [sp, #28]
81
82	ldrd		r10, [r14, #0]								@ r4 = xptr[0]@ r5 = xptr[1]@
83	ldr			r8, [r12], #4									@ cosxsinx = csptr[1]@
84
85	smulwt	r6, r10, r8										@ L_mpy_wx(cosx, t0)
86	smulwt	r5, r11, r8										@ L_mpy_wx(cosx, t1)
87
88	smlawb	r4, r11, r8, r6								@ r4 = L_mpy_wx(cosx, t0) + L_mpy_wx(sinx, t1)@
89	smulwb	r7, r10, r8										@ L_mpy_wx(sinx, t0)
90
91	add			r14, r14, r9									@ xptr += step@
92	sub			r5, r5, r7										@ r5 = L_mpy_wx(cosx, t1) - L_mpy_wx(sinx, t0)@
93
94	ldrd		r10, [r14]										@ r6 = xptr[0]@ r7 = xptr[1]@
95	ldr			r8, [r12], #4									@ cosxsinx = csptr[1]@
96
97	smulwt	r2, r10, r8										@ L_mpy_wx(cosx, t0)
98	smulwt	r7, r11, r8										@ L_mpy_wx(cosx, t1)
99
100	smlawb	r6, r11, r8, r2								@ r4 = L_mpy_wx(cosx, t0) + L_mpy_wx(sinx, t1)@
101	smulwb	r3, r10, r8										@ L_mpy_wx(sinx, t0)
102
103	mov			r10, r4												@ t0 = r4@
104	mov			r11, r5												@ t1 = r5@
105
106	sub			r7, r7, r3										@ r5 = L_mpy_wx(cosx, t1) - L_mpy_wx(sinx, t0)@
107
108
109	add			r4,  r10, r6									@	r4 = t0 + r6@
110	sub			r5, r7, r11										@ r5 = r7 - t1@
111
112	sub			r6, r10, r6										@ r6 = t0 - r6@
113	add			r7, r7, r11										@ r7 = r7 + t1@
114
115	ldr			r2, [sp, #24]
116	ldr			r3, [sp, #28]
117
118	add			r10, r0, r5										@ xptr[0] = r0 + r5@
119	add			r11, r1, r6										@ xptr[0] = r1 + r6
120
121	strd		r10, [r14]
122	sub			r14, r14, r9									@ xptr -= step@
123
124	sub			r10, r2, r4										@	xptr[0] = r2 - r4@
125	sub			r11, r3, r7										@ xptr[1] = r3 - r7@
126
127	strd		r10, [r14]
128	sub			r14, r14, r9									@ xptr -= step@
129
130	sub			r10, r0, r5										@ xptr[0] = r0 - r5@
131	sub			r11, r1, r6										@ xptr[0] = r1 - r6
132
133	strd		r10, [r14]
134	sub			r14, r14, r9									@ xptr -= step@
135
136	add			r10, r2, r4										@	xptr[0] = r2 - r4@
137	add			r11, r3, r7										@ xptr[1] = r3 - r7@
138
139	strd		r10, [r14]
140	add			r14, r14, #8									@ xptr += 2@
141
142	ldr			r11, [sp, #20]
143	subs		r11, r11, #1
144	bne			Radix4FFT_LOOP3
145
146Radix4FFT_LOOP2_END:
147	ldr			r10, [sp, #16]
148	ldr			r3, [sp, #12]
149	ldr			r2, [sp, #8]
150	rsb			r8, r9, r9, lsl #2
151	sub			r10, r10, #1
152	add			r14, r14, r8
153	cmp			r10, #0
154	bhi     Radix4FFT_LOOP2
155
156Radix4FFT_LOOP1_END:
157	ldr     r0, [sp]
158	ldr		r1, [sp, #4]
159	add     r3, r3, r8, asr #1
160	mov     r2, r2, lsl #2
161	movs    r1, r1, asr #2
162	bne     Radix4FFT_LOOP1
163
164Radix4FFT_END:
165	add     sp, sp, #32
166	ldmia   sp!, {r4 - r11, pc}
167
168	@ENDP  @ |Radix4FFT|
169	.end