34 static void lpc_apply_welch_window_sse2(
const int32_t *
data,
int len,
37 double c = 2.0 / (
len-1.0);
42 "movsd %4, %%xmm7 \n\t"
43 "movapd "MANGLE(pd_1)
", %%xmm6 \n\t"
44 "movapd "MANGLE(pd_2)
", %%xmm5 \n\t"
45 "movlhps %%xmm7, %%xmm7 \n\t"
46 "subpd %%xmm5, %%xmm7 \n\t"
47 "addsd %%xmm6, %%xmm7 \n\t"
50 #define WELCH(MOVPD, offset)\
52 "movapd %%xmm7, %%xmm1 \n\t"\
53 "mulpd %%xmm1, %%xmm1 \n\t"\
54 "movapd %%xmm6, %%xmm0 \n\t"\
55 "subpd %%xmm1, %%xmm0 \n\t"\
56 "pshufd $0x4e, %%xmm0, %%xmm1 \n\t"\
57 "cvtpi2pd (%3,%0), %%xmm2 \n\t"\
58 "cvtpi2pd "#offset"*4(%3,%1), %%xmm3 \n\t"\
59 "mulpd %%xmm0, %%xmm2 \n\t"\
60 "mulpd %%xmm1, %%xmm3 \n\t"\
61 "movapd %%xmm2, (%2,%0,2) \n\t"\
62 MOVPD" %%xmm3, "#offset"*8(%2,%1,2) \n\t"\
63 "subpd %%xmm5, %%xmm7 \n\t"\
74 :
"r"(w_data+n2),
"r"(
data+n2),
"m"(
c),
"r"(
len)
77 "%xmm5",
"%xmm6",
"%xmm7")
82 static void lpc_compute_autocorr_sse2(
const double *
data,
int len,
int lag,
90 for(j=0; j<lag; j+=2){
94 "movsd "MANGLE(pd_1)
", %%xmm0 \n\t"
95 "movsd "MANGLE(pd_1)
", %%xmm1 \n\t"
96 "movsd "MANGLE(pd_1)
", %%xmm2 \n\t"
98 "movapd (%2,%0), %%xmm3 \n\t"
99 "movupd -8(%3,%0), %%xmm4 \n\t"
100 "movapd (%3,%0), %%xmm5 \n\t"
101 "mulpd %%xmm3, %%xmm4 \n\t"
102 "mulpd %%xmm3, %%xmm5 \n\t"
103 "mulpd -16(%3,%0), %%xmm3 \n\t"
104 "addpd %%xmm4, %%xmm1 \n\t"
105 "addpd %%xmm5, %%xmm0 \n\t"
106 "addpd %%xmm3, %%xmm2 \n\t"
109 "movhlps %%xmm0, %%xmm3 \n\t"
110 "movhlps %%xmm1, %%xmm4 \n\t"
111 "movhlps %%xmm2, %%xmm5 \n\t"
112 "addsd %%xmm3, %%xmm0 \n\t"
113 "addsd %%xmm4, %%xmm1 \n\t"
114 "addsd %%xmm5, %%xmm2 \n\t"
115 "movsd %%xmm0, (%1) \n\t"
116 "movsd %%xmm1, 8(%1) \n\t"
117 "movsd %%xmm2, 16(%1) \n\t"
125 "movsd "MANGLE(pd_1)
", %%xmm0 \n\t"
126 "movsd "MANGLE(pd_1)
", %%xmm1 \n\t"
128 "movapd (%3,%0), %%xmm3 \n\t"
129 "movupd -8(%4,%0), %%xmm4 \n\t"
130 "mulpd %%xmm3, %%xmm4 \n\t"
131 "mulpd (%4,%0), %%xmm3 \n\t"
132 "addpd %%xmm4, %%xmm1 \n\t"
133 "addpd %%xmm3, %%xmm0 \n\t"
136 "movhlps %%xmm0, %%xmm3 \n\t"
137 "movhlps %%xmm1, %%xmm4 \n\t"
138 "addsd %%xmm3, %%xmm0 \n\t"
139 "addsd %%xmm4, %%xmm1 \n\t"
140 "movsd %%xmm0, %1 \n\t"
141 "movsd %%xmm1, %2 \n\t"
142 :
"+&r"(
i),
"=m"(autoc[j]),
"=m"(autoc[j+1])
158 c->lpc_apply_welch_window = lpc_apply_welch_window_sse2;
159 c->lpc_compute_autocorr = lpc_compute_autocorr_sse2;
__asm__(".macro parse_r var r\n\t" "\\var = -1\n\t" _IFC_REG(0) _IFC_REG(1) _IFC_REG(2) _IFC_REG(3) _IFC_REG(4) _IFC_REG(5) _IFC_REG(6) _IFC_REG(7) _IFC_REG(8) _IFC_REG(9) _IFC_REG(10) _IFC_REG(11) _IFC_REG(12) _IFC_REG(13) _IFC_REG(14) _IFC_REG(15) _IFC_REG(16) _IFC_REG(17) _IFC_REG(18) _IFC_REG(19) _IFC_REG(20) _IFC_REG(21) _IFC_REG(22) _IFC_REG(23) _IFC_REG(24) _IFC_REG(25) _IFC_REG(26) _IFC_REG(27) _IFC_REG(28) _IFC_REG(29) _IFC_REG(30) _IFC_REG(31) ".iflt \\var\n\t" ".error \"Unable to parse register name \\r\"\n\t" ".endif\n\t" ".endm")
Macro definitions for various function/variable attributes.
static atomic_int cpu_flags
int av_get_cpu_flags(void)
Return the flags which specify extensions supported by the CPU.
#define XMM_CLOBBERS_ONLY(...)
#define NAMED_CONSTRAINTS_ARRAY_ADD(...)
Memory handling functions.
#define INLINE_SSE2(flags)
#define INLINE_SSE2_SLOW(flags)
av_cold void ff_lpc_init_x86(LPCContext *c)
DECLARE_ASM_CONST(16, double, pd_1)[2]