1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
|
Upstream-Status: Pending
This patch makes assembly syntax compatible to the x32 toolchain.
This fixes these kinds of compilations errors with x32 gcc.
| pulsecore/svolume_mmx.c: Assembler messages:
| pulsecore/svolume_mmx.c:107: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:135: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:161: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:162: Error: `8(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:180: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:210: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:244: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:245: Error: `8(%esi,%rdi,4)' is not a valid base/index expression
| make[3]: *** [libpulsecore_1.1_la-svolume_mmx.lo] Error 1
Orignally these assembly lines are written for x86_64 ABI, now they are also compatible with
X32 ABI.
Signed-Off-By: Nitin A Kamble <nitin.a.kamble@intel.com> 2011/12/03
Index: pulseaudio-1.1/src/pulsecore/svolume_mmx.c
===================================================================
--- pulseaudio-1.1.orig/src/pulsecore/svolume_mmx.c
+++ pulseaudio-1.1/src/pulsecore/svolume_mmx.c
@@ -107,7 +107,7 @@ static void pa_volume_s16ne_mmx(int16_t
" test $1, %2 \n\t" /* check for odd samples */
" je 2f \n\t"
- " movd (%1, %3, 4), %%mm0 \n\t" /* | v0h | v0l | */
+ " movd (%q1, %3, 4), %%mm0 \n\t" /* | v0h | v0l | */
" movw (%0), %w4 \n\t" /* .. | p0 | */
" movd %4, %%mm1 \n\t"
VOLUME_32x16 (%%mm1, %%mm0)
@@ -122,7 +122,7 @@ static void pa_volume_s16ne_mmx(int16_t
" je 4f \n\t"
"3: \n\t" /* do samples in groups of 2 */
- " movq (%1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
+ " movq (%q1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
" movd (%0), %%mm1 \n\t" /* .. | p1 | p0 | */
VOLUME_32x16 (%%mm1, %%mm0)
" movd %%mm0, (%0) \n\t" /* .. | p1*v1 | p0*v0 | */
@@ -135,8 +135,8 @@ static void pa_volume_s16ne_mmx(int16_t
" je 6f \n\t"
"5: \n\t" /* do samples in groups of 4 */
- " movq (%1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
- " movq 8(%1, %3, 4), %%mm2 \n\t" /* | v3h | v3l | v2h | v2l | */
+ " movq (%q1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
+ " movq 8(%q1, %3, 4), %%mm2 \n\t" /* | v3h | v3l | v2h | v2l | */
" movd (%0), %%mm1 \n\t" /* .. | p1 | p0 | */
" movd 4(%0), %%mm3 \n\t" /* .. | p3 | p2 | */
VOLUME_32x16 (%%mm1, %%mm0)
@@ -180,7 +180,7 @@ static void pa_volume_s16re_mmx(int16_t
" test $1, %2 \n\t" /* check for odd samples */
" je 2f \n\t"
- " movd (%1, %3, 4), %%mm0 \n\t" /* | v0h | v0l | */
+ " movd (%q1, %3, 4), %%mm0 \n\t" /* | v0h | v0l | */
" movw (%0), %w4 \n\t" /* .. | p0 | */
" rorw $8, %w4 \n\t"
" movd %4, %%mm1 \n\t"
@@ -197,7 +197,7 @@ static void pa_volume_s16re_mmx(int16_t
" je 4f \n\t"
"3: \n\t" /* do samples in groups of 2 */
- " movq (%1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
+ " movq (%q1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
" movd (%0), %%mm1 \n\t" /* .. | p1 | p0 | */
SWAP_16 (%%mm1)
VOLUME_32x16 (%%mm1, %%mm0)
@@ -212,8 +212,8 @@ static void pa_volume_s16re_mmx(int16_t
" je 6f \n\t"
"5: \n\t" /* do samples in groups of 4 */
- " movq (%1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
- " movq 8(%1, %3, 4), %%mm2 \n\t" /* | v3h | v3l | v2h | v2l | */
+ " movq (%q1, %3, 4), %%mm0 \n\t" /* | v1h | v1l | v0h | v0l | */
+ " movq 8(%q1, %3, 4), %%mm2 \n\t" /* | v3h | v3l | v2h | v2l | */
" movd (%0), %%mm1 \n\t" /* .. | p1 | p0 | */
" movd 4(%0), %%mm3 \n\t" /* .. | p3 | p2 | */
SWAP_16_2 (%%mm1, %%mm3)
Index: pulseaudio-1.1/src/pulsecore/svolume_sse.c
===================================================================
--- pulseaudio-1.1.orig/src/pulsecore/svolume_sse.c
+++ pulseaudio-1.1/src/pulsecore/svolume_sse.c
@@ -92,7 +92,7 @@ static void pa_volume_s16ne_sse2(int16_t
" test $1, %2 \n\t" /* check for odd samples */
" je 2f \n\t"
- " movd (%1, %3, 4), %%xmm0 \n\t" /* | v0h | v0l | */
+ " movd (%q1, %3, 4), %%xmm0 \n\t" /* | v0h | v0l | */
" movw (%0), %w4 \n\t" /* .. | p0 | */
" movd %4, %%xmm1 \n\t"
VOLUME_32x16 (%%xmm1, %%xmm0)
@@ -107,7 +107,7 @@ static void pa_volume_s16ne_sse2(int16_t
" je 4f \n\t"
"3: \n\t" /* do samples in groups of 2 */
- " movq (%1, %3, 4), %%xmm0 \n\t" /* | v1h | v1l | v0h | v0l | */
+ " movq (%q1, %3, 4), %%xmm0 \n\t" /* | v1h | v1l | v0h | v0l | */
" movd (%0), %%xmm1 \n\t" /* .. | p1 | p0 | */
VOLUME_32x16 (%%xmm1, %%xmm0)
" movd %%xmm0, (%0) \n\t" /* .. | p1*v1 | p0*v0 | */
@@ -123,7 +123,7 @@ static void pa_volume_s16ne_sse2(int16_t
* that the array is 16 bytes aligned, we probably have to do the odd values
* after this then. */
"5: \n\t" /* do samples in groups of 4 */
- " movdqu (%1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
+ " movdqu (%q1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
" movq (%0), %%xmm1 \n\t" /* .. | p3 .. p0 | */
VOLUME_32x16 (%%xmm1, %%xmm0)
" movq %%xmm0, (%0) \n\t" /* .. | p3*v3 .. p0*v0 | */
@@ -136,8 +136,8 @@ static void pa_volume_s16ne_sse2(int16_t
" je 8f \n\t"
"7: \n\t" /* do samples in groups of 8 */
- " movdqu (%1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
- " movdqu 16(%1, %3, 4), %%xmm2 \n\t" /* | v7h | v7l .. v4h | v4l | */
+ " movdqu (%q1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
+ " movdqu 16(%q1, %3, 4), %%xmm2 \n\t" /* | v7h | v7l .. v4h | v4l | */
" movq (%0), %%xmm1 \n\t" /* .. | p3 .. p0 | */
" movq 8(%0), %%xmm3 \n\t" /* .. | p7 .. p4 | */
VOLUME_32x16 (%%xmm1, %%xmm0)
@@ -176,7 +176,7 @@ static void pa_volume_s16re_sse2(int16_t
" test $1, %2 \n\t" /* check for odd samples */
" je 2f \n\t"
- " movd (%1, %3, 4), %%xmm0 \n\t" /* | v0h | v0l | */
+ " movd (%q1, %3, 4), %%xmm0 \n\t" /* | v0h | v0l | */
" movw (%0), %w4 \n\t" /* .. | p0 | */
" rorw $8, %w4 \n\t"
" movd %4, %%xmm1 \n\t"
@@ -193,7 +193,7 @@ static void pa_volume_s16re_sse2(int16_t
" je 4f \n\t"
"3: \n\t" /* do samples in groups of 2 */
- " movq (%1, %3, 4), %%xmm0 \n\t" /* | v1h | v1l | v0h | v0l | */
+ " movq (%q1, %3, 4), %%xmm0 \n\t" /* | v1h | v1l | v0h | v0l | */
" movd (%0), %%xmm1 \n\t" /* .. | p1 | p0 | */
SWAP_16 (%%xmm1)
VOLUME_32x16 (%%xmm1, %%xmm0)
@@ -211,7 +211,7 @@ static void pa_volume_s16re_sse2(int16_t
* that the array is 16 bytes aligned, we probably have to do the odd values
* after this then. */
"5: \n\t" /* do samples in groups of 4 */
- " movdqu (%1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
+ " movdqu (%q1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
" movq (%0), %%xmm1 \n\t" /* .. | p3 .. p0 | */
SWAP_16 (%%xmm1)
VOLUME_32x16 (%%xmm1, %%xmm0)
@@ -226,8 +226,8 @@ static void pa_volume_s16re_sse2(int16_t
" je 8f \n\t"
"7: \n\t" /* do samples in groups of 8 */
- " movdqu (%1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
- " movdqu 16(%1, %3, 4), %%xmm2 \n\t" /* | v7h | v7l .. v4h | v4l | */
+ " movdqu (%q1, %3, 4), %%xmm0 \n\t" /* | v3h | v3l .. v0h | v0l | */
+ " movdqu 16(%q1, %3, 4), %%xmm2 \n\t" /* | v7h | v7l .. v4h | v4l | */
" movq (%0), %%xmm1 \n\t" /* .. | p3 .. p0 | */
" movq 8(%0), %%xmm3 \n\t" /* .. | p7 .. p4 | */
SWAP_16_2 (%%xmm1, %%xmm3)
Index: pulseaudio-1.1/src/pulsecore/sconv_sse.c
===================================================================
--- pulseaudio-1.1.orig/src/pulsecore/sconv_sse.c
+++ pulseaudio-1.1/src/pulsecore/sconv_sse.c
@@ -54,8 +54,8 @@ static void pa_sconv_s16le_from_f32ne_ss
" je 2f \n\t"
"1: \n\t"
- " movups (%2, %0, 2), %%xmm0 \n\t" /* read 8 floats */
- " movups 16(%2, %0, 2), %%xmm2 \n\t"
+ " movups (%q2, %0, 2), %%xmm0 \n\t" /* read 8 floats */
+ " movups 16(%q2, %0, 2), %%xmm2 \n\t"
" minps %%xmm5, %%xmm0 \n\t" /* clamp to 1.0 */
" minps %%xmm5, %%xmm2 \n\t"
" maxps %%xmm6, %%xmm0 \n\t" /* clamp to -1.0 */
@@ -72,8 +72,8 @@ static void pa_sconv_s16le_from_f32ne_ss
" packssdw %%mm1, %%mm0 \n\t" /* pack parts */
" packssdw %%mm3, %%mm2 \n\t"
- " movq %%mm0, (%3, %0) \n\t"
- " movq %%mm2, 8(%3, %0) \n\t"
+ " movq %%mm0, (%q3, %0) \n\t"
+ " movq %%mm2, 8(%q3, %0) \n\t"
" add $16, %0 \n\t"
" dec %1 \n\t"
@@ -85,12 +85,12 @@ static void pa_sconv_s16le_from_f32ne_ss
" je 4f \n\t"
"3: \n\t"
- " movss (%2, %0, 2), %%xmm0 \n\t"
+ " movss (%q2, %0, 2), %%xmm0 \n\t"
" minss %%xmm5, %%xmm0 \n\t"
" maxss %%xmm6, %%xmm0 \n\t"
" mulss %%xmm7, %%xmm0 \n\t"
" cvtss2si %%xmm0, %4 \n\t"
- " movw %w4, (%3, %0) \n\t"
+ " movw %w4, (%q3, %0) \n\t"
" add $2, %0 \n\t"
" dec %1 \n\t"
" jne 3b \n\t"
@@ -119,8 +119,8 @@ static void pa_sconv_s16le_from_f32ne_ss
" je 2f \n\t"
"1: \n\t"
- " movups (%2, %0, 2), %%xmm0 \n\t" /* read 8 floats */
- " movups 16(%2, %0, 2), %%xmm2 \n\t"
+ " movups (%q2, %0, 2), %%xmm0 \n\t" /* read 8 floats */
+ " movups 16(%q2, %0, 2), %%xmm2 \n\t"
" minps %%xmm5, %%xmm0 \n\t" /* clamp to 1.0 */
" minps %%xmm5, %%xmm2 \n\t"
" maxps %%xmm6, %%xmm0 \n\t" /* clamp to -1.0 */
@@ -132,7 +132,7 @@ static void pa_sconv_s16le_from_f32ne_ss
" cvtps2dq %%xmm2, %%xmm2 \n\t"
" packssdw %%xmm2, %%xmm0 \n\t"
- " movdqu %%xmm0, (%3, %0) \n\t"
+ " movdqu %%xmm0, (%q3, %0) \n\t"
" add $16, %0 \n\t"
" dec %1 \n\t"
@@ -144,12 +144,12 @@ static void pa_sconv_s16le_from_f32ne_ss
" je 4f \n\t"
"3: \n\t"
- " movss (%2, %0, 2), %%xmm0 \n\t"
+ " movss (%q2, %0, 2), %%xmm0 \n\t"
" minss %%xmm5, %%xmm0 \n\t"
" maxss %%xmm6, %%xmm0 \n\t"
" mulss %%xmm7, %%xmm0 \n\t"
" cvtss2si %%xmm0, %4 \n\t"
- " movw %w4, (%3, %0) \n\t"
+ " movw %w4, (%q3, %0) \n\t"
" add $2, %0 \n\t"
" dec %1 \n\t"
" jne 3b \n\t"
|
