| /external/libvpx/libvpx/vpx_dsp/x86/ |
| D | highbd_subpel_variance_impl_sse2.asm | 329 ; slightly faster because of pmullw latency. It would also cut our rodata
331 pmullw m1, filter_y_a
332 pmullw m5, filter_y_b
334 pmullw m0, filter_y_a
335 pmullw m4, filter_y_b
359 pmullw m1, filter_y_a
360 pmullw m5, filter_y_b
362 pmullw m0, filter_y_a
363 pmullw m4, filter_y_b
562 pmullw m1, filter_y_a
[all …]
|
| D | subpel_variance_sse2.asm | 396 ; slightly faster because of pmullw latency. It would also cut our rodata
398 pmullw m2, filter_y_a
399 pmullw m3, filter_y_b
401 pmullw m0, filter_y_a
402 pmullw m4, filter_y_b
440 pmullw m0, filter_y_a
441 pmullw m1, m2, filter_y_b
444 pmullw m2, filter_y_a
445 pmullw m4, filter_y_b
839 pmullw m2, filter_x_a
[all …]
|
| D | variance_impl_mmx.asm | 455 pmullw mm1, [rax] ;
458 pmullw mm3, [rax+8] ;
479 pmullw mm1, [rax] ;
482 pmullw mm3, [rax+8] ;
491 pmullw mm3, [rdx] ;
493 pmullw mm1, [rdx+8] ;
596 pmullw mm1, [rax] ;
598 pmullw mm2, [rax] ;
602 pmullw mm3, [rax+8] ;
604 pmullw mm4, [rax+8] ;
[all …]
|
| D | vpx_subpixel_bilinear_sse2.asm | 39 pmullw xmm0, xmm4 ;multiply the filter factors
87 pmullw xmm0, xmm6
88 pmullw xmm1, xmm7
110 pmullw xmm0, xmm6
111 pmullw xmm1, xmm7
112 pmullw xmm2, xmm6
113 pmullw xmm3, xmm7
|
| D | vpx_subpixel_8t_sse2.asm | 61 pmullw xmm0, k0k1 ;multiply the filter factors
62 pmullw xmm6, k6k7
63 pmullw xmm2, k2k3
64 pmullw xmm5, k5k4
152 pmullw xmm0, k0
153 pmullw xmm1, k1
154 pmullw xmm6, k6
155 pmullw xmm7, k7
156 pmullw xmm2, k2
157 pmullw xmm5, k5
[all …]
|
| D | quantize_avx_x86_64.asm | 141 pmullw m8, m3 ; dqc[i] = qc[i] * q
143 pmullw m13, m3 ; dqc[i] = qc[i] * q
323 pmullw m8, m3 ; dqc[i] = qc[i] * q
325 pmullw m13, m3 ; dqc[i] = qc[i] * q
445 pmullw m14, m3 ; dqc[i] = qc[i] * q
446 pmullw m13, m3 ; dqc[i] = qc[i] * q
|
| D | quantize_ssse3_x86_64.asm | 125 pmullw m8, m3 ; dqc[i] = qc[i] * q
127 pmullw m13, m3 ; dqc[i] = qc[i] * q
228 pmullw m14, m3 ; dqc[i] = qc[i] * q
229 pmullw m13, m3 ; dqc[i] = qc[i] * q
|
| /external/libvpx/libvpx/vp8/common/x86/ |
| D | subpixel_sse2.asm | 77 pmullw xmm3, XMMWORD PTR [rdx] ; x[-2] * H[-2]; Tap 1
81 pmullw xmm4, XMMWORD PTR [rdx+16] ; x[-1] * H[-1]; Tap 2
87 pmullw xmm5, [rdx+32] ; x[ 0] * H[ 0]; Tap 3
92 pmullw xmm6, [rdx+48] ; x[ 1] * h[ 1] ; Tap 4
98 pmullw xmm7, [rdx+64] ; x[ 2] * h[ 2] ; Tap 5
101 pmullw xmm1, [rdx+80] ; x[ 3] * h[ 3] ; Tap 6
202 pmullw xmm3, XMMWORD PTR [rdx] ; x[-2] * H[-2]; Tap 1
206 pmullw xmm4, XMMWORD PTR [rdx+16] ; x[-1] * H[-1]; Tap 2
212 pmullw xmm5, [rdx+32] ; x[ 0] * H[ 0]; Tap 3
217 pmullw xmm6, [rdx+48] ; x[ 1] * h[ 1] ; Tap 4
[all …]
|
| D | subpixel_mmx.asm | 59 pmullw mm3, mm1 ; mm3 *= kernel 1 modifiers.
63 pmullw mm4, mm7 ; mm5 *= kernel 4 modifiers
69 pmullw mm5, mm2 ; mm5 *= kernel 2 modifiers
75 pmullw mm4, mm6 ; mm5 *= kernel 3 modifiers
81 pmullw mm4, [rdx+80] ; mm5 *= kernel 0 modifiers
85 pmullw mm5, [rdx] ; mm5 *= kernel 5 modifiers
158 pmullw mm3, mm1 ; mm3 *= kernel 1 modifiers.
162 pmullw mm4, mm7 ; mm4 *= kernel 4 modifiers.
166 pmullw mm4, mm2 ; mm4 *= kernel 2 modifiers.
170 pmullw mm4, [rbx] ; mm4 *= kernel 0 modifiers.
[all …]
|
| D | dequantize_mmx.asm | 30 pmullw mm1, [rax+0] ; mm4 *= kernel 0 modifiers.
34 pmullw mm1, [rax+8] ; mm4 *= kernel 0 modifiers.
38 pmullw mm1, [rax+16] ; mm4 *= kernel 0 modifiers.
42 pmullw mm1, [rax+24] ; mm4 *= kernel 0 modifiers.
72 pmullw mm0, [rdx]
75 pmullw mm1, [rdx +8]
78 pmullw mm2, [rdx+16]
81 pmullw mm3, [rdx+24]
|
| D | mfqe_sse2.asm | 57 pmullw xmm2, xmm0
58 pmullw xmm3, xmm0
64 pmullw xmm4, xmm1
65 pmullw xmm5, xmm1
132 pmullw xmm2, xmm0
136 pmullw xmm3, xmm1
|
| D | idctllm_sse2.asm | 39 pmullw xmm4, xmm5
140 pmullw xmm0, [rdx]
141 pmullw xmm2, [rdx+16]
142 pmullw xmm1, [rdx]
143 pmullw xmm3, [rdx+16]
472 pmullw xmm0, [rdx]
473 pmullw xmm2, [rdx+16]
474 pmullw xmm1, [rdx]
475 pmullw xmm3, [rdx+16]
|
| D | postproc_mmx.asm | 100 pmullw mm1, mm1 ;
125 pmullw mm2, mm2
134 pmullw mm1, mm1
151 pmullw mm1, mm1
|
| D | postproc_sse2.asm | 319 pmullw xmm1, xmm1 ;
344 pmullw xmm2, xmm2
353 pmullw xmm1, xmm1
370 pmullw xmm1, xmm1
|
| /external/llvm/test/CodeGen/X86/ |
| D | vec_shift6.ll | 8 ; Check that we produce a SSE2 packed integer multiply (pmullw) instead.
15 ; CHECK: pmullw
24 ; CHECK: pmullw
55 ; into two pmullw instructions. With AVX2, the test case below would produce
63 ; SSE: pmullw
64 ; SSE-NEXT: pmullw
87 ; parts and then we convert each part into a pmullw.
94 ; SSE: pmullw
95 ; SSE-NEXT: pmullw
96 ; SSE-NEXT: pmullw
[all …]
|
| D | pmul.ll | 14 ; SSE2-NEXT: pmullw %xmm1, %xmm2
19 ; SSE2-NEXT: pmullw %xmm1, %xmm0
28 ; SSE41-NEXT: pmullw %xmm2, %xmm1
33 ; SSE41-NEXT: pmullw %xmm2, %xmm0
59 ; SSE-NEXT: pmullw {{.*}}(%rip), %xmm0
133 ; SSE2-NEXT: pmullw %xmm2, %xmm3
140 ; SSE2-NEXT: pmullw %xmm1, %xmm0
149 ; SSE41-NEXT: pmullw %xmm3, %xmm2
156 ; SSE41-NEXT: pmullw %xmm1, %xmm0
182 ; SSE-NEXT: pmullw %xmm1, %xmm0
|
| D | 2008-02-26-AsmDirectMemOp.ll | 15 …mm0 \0A\09movq 8($0, $3), %mm1 \0A\09pmullw %mm6, %mm0 \0A\09pmullw …
|
| /external/libvpx/libvpx/vp9/common/x86/ |
| D | vp9_mfqe_sse2.asm | 58 pmullw xmm2, xmm0
59 pmullw xmm3, xmm0
65 pmullw xmm4, xmm1
66 pmullw xmm5, xmm1
133 pmullw xmm2, xmm0
137 pmullw xmm3, xmm1
|
| /external/mesa3d/src/mesa/x86/ |
| D | read_rgba_span_x86.S | 561 pmullw %mm6, %mm0
562 pmullw %mm6, %mm2
596 pmullw %mm6, %mm0
597 pmullw %mm6, %mm2
634 pmullw %mm6, %mm0
635 pmullw %mm6, %mm2
665 pmullw %mm6, %mm0
|
| /external/libvpx/libvpx/vp8/encoder/x86/ |
| D | quantize_mmx.asm | 66 pmullw mm3, mm2
106 pmullw mm7, mm6
147 pmullw mm7, mm6
188 pmullw mm7, mm6
|
| /external/libvpx/libvpx/vp9/encoder/x86/ |
| D | vp9_quantize_ssse3_x86_64.asm | 78 pmullw m8, m3 ; r4[i] = r3[i] * q
80 pmullw m13, m3 ; r4[i] = r3[i] * q
132 pmullw m14, m3 ; r4[i] = r3[i] * q
133 pmullw m13, m3 ; r4[i] = r3[i] * q
|
| /external/valgrind/VEX/test/ |
| D | mmxtest.c | 286 #define pmullw_m2r(var, reg) mmx_m2r(pmullw, var, reg)
287 #define pmullw_r2r(regs, regd) mmx_r2r(pmullw, regs, regd)
288 #define pmullw(vars, vard) mmx_m2m(pmullw, vars, vard) macro
555 do_test("pmullw", pmullw(ma,mb)); in main()
|
| /external/libjpeg-turbo/simd/ |
| D | jdsample-sse2-64.asm | 268 pmullw xmm0,[rel PW_THREE]
269 pmullw xmm4,[rel PW_THREE]
|
| D | jdsample-mmx.asm | 295 pmullw mm0,[GOTOFF(ebx,PW_THREE)]
296 pmullw mm4,[GOTOFF(ebx,PW_THREE)]
|
| /external/libyuv/files/source/ |
| D | row_win.cc | 1574 __asm pmullw xmm3, kYToRgb \
1602 __asm pmullw xmm3, kYToRgb \
3037 pmullw xmm2, xmm3 // _r_b * alpha in ARGBBlendRow_SSE2()
3042 pmullw xmm1, xmm3 // _a_g * alpha in ARGBBlendRow_SSE2()
3068 pmullw xmm2, xmm3 // _r_b * alpha in ARGBBlendRow_SSE2()
3073 pmullw xmm1, xmm3 // _a_g * alpha in ARGBBlendRow_SSE2()
3099 pmullw xmm2, xmm3 // _r_b * alpha in ARGBBlendRow_SSE2()
3104 pmullw xmm1, xmm3 // _a_g * alpha in ARGBBlendRow_SSE2()
3169 pmullw xmm2, xmm3 // _r_b * alpha in ARGBBlendRow_SSSE3()
3174 pmullw xmm1, xmm3 // _a_g * alpha in ARGBBlendRow_SSSE3()
[all …]
|