| /external/libvpx/vp8/encoder/x86/ |
| variance_impl_ssse3.asm | 68 movdqu xmm1, XMMWORD PTR [rsi+1]
71 punpcklbw xmm0, xmm1
72 punpckhbw xmm2, xmm1
92 movdqu xmm1, XMMWORD PTR [rsi]
94 movdqa xmm3, xmm1
96 punpcklbw xmm1, xmm2
98 pmaddubsw xmm1, [rax]
101 paddw xmm1, [GLOBAL(xmm_bi_rd)]
103 psraw xmm1, xmm_filter_shift
105 packuswb xmm1, xmm
[all...] |
| variance_impl_sse2.asm | 38 movdqa xmm1, [rax+16]
42 pmaddwd xmm1, xmm1
46 paddd xmm0, xmm1
129 movdqu xmm1, XMMWORD PTR [rsi]
135 movdqa xmm3, xmm1
139 punpcklbw xmm1, xmm0
146 psubw xmm1, xmm2
149 paddw xmm7, xmm1
150 pmaddwd xmm1, xmm
[all...] |
| subtract_sse2.asm | 96 movdqa xmm1, XMMWORD PTR [rax] ; pred
99 psubb xmm0, xmm1
101 pxor xmm1, [GLOBAL(t80)] ;convert to signed values
103 pcmpgtb xmm1, xmm2 ; obtain sign information
106 movdqa xmm3, xmm1
107 punpcklbw xmm0, xmm1 ; put sign back to subtraction
171 movdqa xmm1, XMMWORD PTR [rax] ; pred
175 psubb xmm0, xmm1 ; subtraction with sign missed
177 pxor xmm1, [GLOBAL(t80)] ;convert to signed values
179 pcmpgtb xmm1, xmm2 ; obtain sign informatio
[all...] |
| quantize_ssse3.asm | 39 movdqa xmm1, xmm0
45 pabsw xmm1, xmm1
48 paddw xmm1, xmm2
51 pmulhw xmm1, [rsi]
58 pxor xmm1, xmm0
60 psubw xmm1, xmm0
63 movdqa [rdi], xmm1
70 pmullw xmm2, xmm1
73 pcmpeqw xmm1, xmm4 ;non zero mas
[all...] |
| quantize_sse2.asm | 69 movdqa xmm1, xmm0
77 pxor xmm1, xmm0
81 psubw xmm1, xmm0
93 psubw xmm1, xmm2
95 movdqa [rsp + abs_minus_zbin], xmm1
99 paddw xmm1, xmm2
109 paddw xmm1, xmm2
113 pmulhw xmm3, xmm1
117 paddw xmm1, xmm3
120 movdqa [rsp + temp_qcoeff], xmm1
[all...] |
| fwalsh_sse2.asm | 32 movq xmm1, MMWORD PTR [rsi + rdx]
37 punpcklwd xmm0, xmm1
40 movdqa xmm1, xmm0
42 punpckhdq xmm1, xmm2 ; ip[3] ip[2]
45 paddw xmm0, xmm1
46 psubw xmm2, xmm1
51 movdqa xmm1, xmm0
53 punpckhqdq xmm1, xmm2 ; c1 d1
62 paddw xmm0, xmm1 ; b1+c1 a1+d1
63 psubw xmm2, xmm1 ; b1-c1 a1-d
[all...] |
| dct_sse2.asm | 70 movq xmm1, MMWORD PTR[input ] ;23 22 21 20
74 punpcklqdq xmm1, xmm3 ;33 32 31 30 23 22 21 20
77 punpckldq xmm0, xmm1 ;23 22 03 02 21 20 01 00
78 punpckhdq xmm2, xmm1 ;33 32 13 12 31 30 11 10
79 movdqa xmm1, xmm0
81 pshufhw xmm1, xmm1, 0b1h ;22 23 02 03 xx xx xx xx
84 punpckhdq xmm1, xmm2 ;32 33 22 23 12 13 02 03
86 paddw xmm0, xmm1 ;b1 a1 b1 a1 b1 a1 b1 a1
87 psubw xmm3, xmm1 ;c1 d1 c1 d1 c1 d1 c1 d
[all...] |
| sad_sse4.asm | 17 movq xmm1, MMWORD PTR [rdi]
20 punpcklqdq xmm1, xmm3
23 movdqa xmm2, xmm1
24 mpsadbw xmm1, xmm0, 0x0
33 paddw xmm1, xmm2
34 paddw xmm1, xmm3
35 paddw xmm1, xmm4
58 paddw xmm1, xmm5
83 paddw xmm1, xmm5
89 movq xmm1, MMWORD PTR [rdi
[all...] |
| encodeopt.asm | 222 movd xmm1, dword ptr arg(2) ;dc
223 por xmm1, xmm2
225 pcmpeqw xmm1, xmm7
240 pand xmm3, xmm1
259 movdqa xmm1, xmm0
262 paddd xmm0, xmm1
349 movdqa xmm1, [rsi]
352 psubw xmm1, xmm2
353 pmaddwd xmm1, xmm1
[all...] |
| temporal_filter_apply_sse2.asm | 81 movq xmm1, [rsi] ; second row
83 punpcklbw xmm1, xmm7 ; src[ 8-15]
89 movdqa xmm1, xmm0
91 punpckhbw xmm1, xmm7 ; src[ 8-15]
101 psubw xmm1, xmm3 ; src - pred[ 8-15]
105 pmullw xmm1, xmm1 ; modifer[ 8-15]^2
109 pmullw xmm1, [GLOBAL(_const_3w)]
113 paddw xmm1, [rsp + rounding_bit]
117 psrlw xmm1, [rsp + strength
[all...] |
| /external/llvm/test/CodeGen/X86/ |
| sse-minmax.ll | 19 ; CHECK-NEXT: maxsd %xmm1, %xmm0
22 ; UNSAFE-NEXT: maxsd %xmm1, %xmm0
25 ; FINITE-NEXT: maxsd %xmm1, %xmm0
34 ; CHECK-NEXT: minsd %xmm1, %xmm0
37 ; UNSAFE-NEXT: minsd %xmm1, %xmm0
40 ; FINITE-NEXT: minsd %xmm1, %xmm0
49 ; CHECK-NEXT: minsd %xmm0, %xmm1
50 ; CHECK-NEXT: movap{{[sd]}} %xmm1, %xmm0
53 ; UNSAFE-NEXT: minsd %xmm0, %xmm1
54 ; UNSAFE-NEXT: movap{{[sd]}} %xmm1, %xmm
[all...] |
| /external/llvm/test/MC/X86/ |
| x86-32-fma3.s | 3 // CHECK: vfmadd132pd %xmm2, %xmm5, %xmm1
5 vfmadd132pd %xmm2, %xmm5, %xmm1
7 // CHECK: vfmadd132pd (%eax), %xmm5, %xmm1
9 vfmadd132pd (%eax), %xmm5, %xmm1
11 // CHECK: vfmadd132ps %xmm2, %xmm5, %xmm1
13 vfmadd132ps %xmm2, %xmm5, %xmm1
15 // CHECK: vfmadd132ps (%eax), %xmm5, %xmm1
17 vfmadd132ps (%eax), %xmm5, %xmm1
19 // CHECK: vfmadd213pd %xmm2, %xmm5, %xmm1
21 vfmadd213pd %xmm2, %xmm5, %xmm1
[all...] |
| x86-32-avx.s | 274 // CHECK: vunpckhps %xmm1, %xmm2, %xmm4
276 vunpckhps %xmm1, %xmm2, %xmm4
278 // CHECK: vunpckhpd %xmm1, %xmm2, %xmm4
280 vunpckhpd %xmm1, %xmm2, %xmm4
282 // CHECK: vunpcklps %xmm1, %xmm2, %xmm4
284 vunpcklps %xmm1, %xmm2, %xmm4
286 // CHECK: vunpcklpd %xmm1, %xmm2, %xmm4
288 vunpcklpd %xmm1, %xmm2, %xmm4
306 // CHECK: vcmpps $0, %xmm0, %xmm6, %xmm1
308 vcmpps $0, %xmm0, %xmm6, %xmm1
[all...] |
| /dalvik/vm/mterp/x86-atom/ |
| OP_NEG_LONG.S | 20 %include "x86-atom/unopWide.S" {"preinstr":"xorps %xmm1, %xmm1", "instr":"psubq %xmm0, %xmm1", "result":"%xmm1"}
|
| OP_USHR_LONG.S | 36 movsd (rFP, %edx, 4), %xmm1 # %xmm1<- vBB
37 psrlq %xmm0, %xmm1 # %xmm1<- shifted vBB
38 movsd %xmm1, (rFP, rINST, 4) # vAA<- shifted vBB
|
| /frameworks/compile/libbcc/runtime/lib/i386/ |
| floatdidf.S | 22 cvtsi2sd 8(%esp), %xmm1
26 mulsd REL_ADDR(twop32), %xmm1 // a_hi as a double (without rounding)
28 subsd %xmm2, %xmm1 // a_hi - 0x1p52 (no rounding occurs)
30 addsd %xmm1, %xmm0 // a_hi + a_lo (round happens here)
|
| ashrdi3.S | 18 movd 8(%esp), %xmm1
19 punpckldq %xmm1, %xmm0 // Load input
31 pcmpeqb %xmm1, %xmm1 // -1ULL
32 psrlq $58, %xmm1 // 0x3f
33 pandn %xmm1, %xmm2 // 63 - count
34 pcmpeqb %xmm1, %xmm1 // -1ULL
35 psubq %xmm1, %xmm2 // 64 - count
36 psllq %xmm2, %xmm1 // -1 << (64 - count) = leading sign bit
[all...] |
| /bionic/libc/arch-x86/string/ |
| ssse3-strcmp-latest.S | 189 movlpd (%eax), %xmm1
191 movhpd 8(%eax), %xmm1
193 pcmpeqb %xmm1, %xmm0
194 pcmpeqb %xmm2, %xmm1
195 psubb %xmm0, %xmm1
196 pmovmskb %xmm1, %ecx
273 movdqa (%eax), %xmm1
275 pcmpeqb %xmm1, %xmm0
276 pcmpeqb (%edx), %xmm1
277 psubb %xmm0, %xmm1
[all...] |
| /external/libvpx/vp8/common/x86/ |
| postproc_sse2.asm | 66 movdqa xmm1, xmm3 ; mm1 = p0..p3
74 movdqa xmm7, xmm1 ; mm7 = r0 p0..p3
76 psubusw xmm5, xmm1 ; mm5 = r1 p0..p3 - r0 p0..p3
85 movdqa xmm6, xmm1 ; mm6 = r0 p0..p3
87 psubusw xmm5, xmm1 ; mm5 = r2 p0..p3 - r2 p0..p3
99 movdqa xmm6, xmm1 ; mm6 = r0 p0..p3
101 psubusw xmm5, xmm1 ; mm5 = r-2 p0..p3 - p0..p3
111 movdqa xmm6, xmm1 ; mm6 = r0 p0..p3
113 psubusw xmm4, xmm1 ; mm5 = r-1 p0..p3 - p0..p3
122 pand xmm1, xmm7 ; mm1 select vals > thresh from sourc
[all...] |
| iwalsh_sse2.asm | 30 movdqa xmm1, [rsi + 16] ;ip[12] ip[8]
35 pshufd xmm2, xmm1, 4eh ;ip[8] ip[12]
46 movdqa xmm1, xmm4 ;c1 b1
48 psubw xmm0, xmm1 ;d1-c1 a1-b1 aka op[12] op[8]
62 movdqa xmm1, xmm4 ; 23 03 22 02 21 01 20 00
64 punpckhwd xmm1, xmm3 ; 33 23 13 03 32 22 12 02
66 pshufd xmm2, xmm1, 4eh ;ip[8] ip[12]
78 movdqa xmm1, xmm5 ;c1 b1
80 psubw xmm4, xmm1 ;d1-c1 a1-b1 aka op[12] op[8]
89 movdqa xmm1, xmm5 ; 23 03 22 02 21 01 20 0
[all...] |
| subpixel_ssse3.asm | 79 movdqa xmm1, xmm0
82 movdqa xmm2, xmm1
83 pshufb xmm1, [GLOBAL(shuf2bfrom1)]
86 pmaddubsw xmm1, xmm5
94 paddsw xmm0, xmm1
133 movq xmm1, MMWORD PTR [rsi + 3] ; 3 4 5 6 7 8 9 10
135 punpcklbw xmm0, xmm1 ; -2 3 -1 4 0 5 1 6 2 7 3 8 4 9 5 10
218 movdqa xmm1, xmm0
221 movdqa xmm2, xmm1
222 pshufb xmm1, [GLOBAL(shuf2bfrom1)
[all...] |
| postproc_mmx.c | 445 movdqa xmm1, xmm3 ;
458 movdqa xmm7, xmm1 ; local
462 psubusw xmm5, xmm1 ;
477 movdqa xmm6, xmm1 ; local
481 psubusw xmm5, xmm1 ;
500 movdqa xmm6, xmm1 ; local
504 psubusw xmm5, xmm1 ;
521 movdqa xmm6, xmm1 ; local
525 psubusw xmm4, xmm1 ;
539 pand xmm1, xmm7
543 paddusw xmm1, xmm7 ; local
546 packuswb xmm1, xmm0 ; local
548 movq QWORD PTR [edi], xmm1 ; local
592 movdqa xmm7, xmm1 ; local
611 movdqa xmm6, xmm1 ; local
633 movdqa xmm6, xmm1 ; local
654 movdqa xmm6, xmm1 ; local
675 paddusw xmm1, xmm7 ; local
678 packuswb xmm1, xmm0 ; local
913 movq xmm1, QWORD PTR [edi]; local
914 punpcklbw xmm1, xmm0 ; local
916 paddw xmm5, xmm1 ; local
917 pmullw xmm1, xmm1 ; local
919 movdqa xmm2, xmm1 ; local
920 punpcklwd xmm1, xmm0 ; local
923 paddd xmm6, xmm1 ; local
933 movq xmm1, QWORD PTR [esi] \/\/ [s-pitch*8] local
936 punpcklbw xmm1, xmm0 local
951 pmullw xmm1, xmm1 local
954 punpcklwd xmm1, xmm0 local
965 movdqa xmm1, xmm5 local
968 pmullw xmm1, xmm1 local
973 punpcklwd xmm1, xmm4 local
993 movq xmm1, QWORD PTR [esi+eax*8] local
996 punpcklbw xmm1, xmm0 local
998 paddw xmm1, xmm5 local
1004 paddw xmm1, xmm4 local
1006 psraw xmm1, 4 local
1008 packuswb xmm1, xmm0 local
1009 pand xmm1, xmm3 local
1012 por xmm1, xmm3 local
1199 movd xmm1, DWORD PTR [esi+ecx-8] \/\/ -8 -7 -6 -5 local
1202 punpcklbw xmm1, xmm0 \/\/ expanding local
1205 punpcklwd xmm1, xmm0 \/\/ expanding to dwords local
1209 paddd xmm1, xmm1 \/\/ -8*2 -7*2 -6*2 -5*2 local
1211 paddd xmm1, xmm2 \/\/ 7+-8 8+-7 9+-6 10+-5 local
1212 pmaddwd xmm1, xmm2 \/\/ squared of 7+-8 8+-7 9+-6 10+-5 local
1220 psrldq xmm1, 4 \/\/ 8--7 9--6 10--5 0000 local
1223 pshufd xmm3, xmm1, 3 \/\/ 0000 8--7 8--7 8--7 squared local
1229 pshufd xmm3, xmm1, 01011111b \/\/ 0000 0000 9--6 9--6 squared local
1235 pshufd xmm3, xmm1, 10111111b \/\/ 0000 0000 8--7 8--7 squared local
1256 movd xmm1, DWORD PTR [esi+ecx] local
1259 punpcklbw xmm1, xmm0 local
1260 punpcklwd xmm1, xmm0 local
1262 paddd xmm1, xmm6 local
1263 paddd xmm1, four8s local
1265 psrad xmm1, 4 local
1266 packssdw xmm1, xmm0 local
1268 packuswb xmm1, xmm0 local
1269 pand xmm1, xmm5 local
1490 movdqu xmm1, [esi+eax] \/\/ get the source local
1492 psubusb xmm1, blackclamp \/\/ clamp both sides so we don't outrange adding noise local
1493 paddusb xmm1, bothclamp local
1494 psubusb xmm1, whiteclamp local
1497 paddb xmm1, xmm2 \/\/ add it in local
[all...] |
| idctllm_sse2.asm | 60 movq xmm1, [rax+rcx]
66 punpcklbw xmm1, xmm7
75 paddw xmm1, xmm4
81 packuswb xmm1, xmm7
87 movq [rax + rdx], xmm1
122 ; note the transpose of xmm1 and xmm2, necessary for shuffle
126 movdqa xmm1, [rax+32]
138 pmullw xmm1, [rdx]
143 punpckldq xmm0, xmm1
144 punpckhdq xmm4, xmm1
[all...] |
| recon_sse2.asm | 29 movq xmm1, MMWORD PTR [rsi]
30 punpcklbw xmm1, xmm0
31 paddsw xmm1, XMMWORD PTR [rdx]
32 packuswb xmm1, xmm0 ; pack and unpack to saturate
33 movq MMWORD PTR [rdi], xmm1
81 movdqa xmm1, XMMWORD PTR [rsi]
82 movdqa xmm5, xmm1
83 punpcklbw xmm1, xmm0
85 paddsw xmm1, XMMWORD PTR [rdx]
87 packuswb xmm1, xmm5 ; pack and unpack to saturat
[all...] |
| loopfilter_sse2.asm | 20 movdqa xmm1, [rsi+2*rax] ; q2
26 movlps xmm1, [rsi + rcx] ; q2
31 movhps xmm1, [rdi + rcx]
38 movdqa XMMWORD PTR [rsp], xmm1 ; store q2
42 movdqa xmm6, xmm1 ; q2
45 psubusb xmm1, xmm2 ; q2-=q3
52 por xmm1, xmm2 ; abs(q3-q2)
55 pmaxub xmm1, xmm4
63 pmaxub xmm1, xmm5
89 pmaxub xmm1, xmm4 ; abs(p3 - p2
[all...] |
