/external/chromium_org/third_party/skia/src/opts/ |
H A D | memset32_neon.S | 32 vdup.32 q0, r1 33 vmov q1, q0 44 vst1.64 {q0, q1}, [r0]! 45 vst1.64 {q0, q1}, [r0]! 46 vst1.64 {q0, q1}, [r0]! 47 vst1.64 {q0, q1}, [r0]! 48 vst1.64 {q0, q1}, [r0]! 49 vst1.64 {q0, q1}, [r0]! 50 vst1.64 {q0, q1}, [r0]! 51 vst1.64 {q0, q [all...] |
H A D | memset16_neon.S | 77 * build q0 and q1 depending on if we'll need it, so that's 86 vmov q1, q0 92 vst1.64 {q0, q1}, [r0]! 93 vst1.64 {q0, q1}, [r0]! 94 vst1.64 {q0, q1}, [r0]! 95 vst1.64 {q0, q1}, [r0]! 105 vst1.64 {q0, q1}, [r0]! 114 vst1.32 {q0}, [r0]!
|
/external/skia/src/opts/ |
H A D | memset32_neon.S | 32 vdup.32 q0, r1 33 vmov q1, q0 44 vst1.64 {q0, q1}, [r0]! 45 vst1.64 {q0, q1}, [r0]! 46 vst1.64 {q0, q1}, [r0]! 47 vst1.64 {q0, q1}, [r0]! 48 vst1.64 {q0, q1}, [r0]! 49 vst1.64 {q0, q1}, [r0]! 50 vst1.64 {q0, q1}, [r0]! 51 vst1.64 {q0, q [all...] |
H A D | memset16_neon.S | 77 * build q0 and q1 depending on if we'll need it, so that's 86 vmov q1, q0 92 vst1.64 {q0, q1}, [r0]! 93 vst1.64 {q0, q1}, [r0]! 94 vst1.64 {q0, q1}, [r0]! 95 vst1.64 {q0, q1}, [r0]! 105 vst1.64 {q0, q1}, [r0]! 114 vst1.32 {q0}, [r0]!
|
/external/valgrind/main/none/tests/arm/ |
H A D | neon128.c | 355 TESTINSN_imm("vmov.i32 q0", q0, 0x7); 365 TESTINSN_imm("vmov.f32 q0", q0, 0.328125); 366 TESTINSN_imm("vmov.f32 q0", q0, -0.328125); 369 TESTINSN_imm("vmvn.i32 q0", q0, 0x7); 381 TESTINSN_imm("vorr.i32 q0", q0, [all...] |
H A D | neon128.stdout.exp | 2 vmov.i32 q0, #0x7 :: Qd 0x00000007 0x00000007 0x00000007 0x00000007 3 vmov.i32 q0, #0x7 :: Qd 0x00000007 0x00000007 0x00000007 0x00000007 22 vmov.f32 q0, #0.328125 :: Qd 0x3ea80000 0x3ea80000 0x3ea80000 0x3ea80000 23 vmov.f32 q0, #0.328125 :: Qd 0x3ea80000 0x3ea80000 0x3ea80000 0x3ea80000 24 vmov.f32 q0, #-0.328125 :: Qd 0xbea80000 0xbea80000 0xbea80000 0xbea80000 25 vmov.f32 q0, #-0.328125 :: Qd 0xbea80000 0xbea80000 0xbea80000 0xbea80000 27 vmvn.i32 q0, #0x7 :: Qd 0xfffffff8 0xfffffff8 0xfffffff8 0xfffffff8 28 vmvn.i32 q0, #0x7 :: Qd 0xfffffff8 0xfffffff8 0xfffffff8 0xfffffff8 48 vorr.i32 q0, #0x7 :: Qd 0x55555557 0x55555557 0x55555557 0x55555557 49 vorr.i32 q0, # [all...] |
/external/libvpx/libvpx/vp8/common/arm/neon/ |
H A D | buildintrapredictorsmby_neon.asm | 45 vdup.u8 q0, r12 126 vdup.u8 q0, r5 129 vst1.u8 {q0}, [r1]! 130 vst1.u8 {q0}, [r1]! 131 vst1.u8 {q0}, [r1]! 132 vst1.u8 {q0}, [r1]! 133 vst1.u8 {q0}, [r1]! 134 vst1.u8 {q0}, [r1]! 135 vst1.u8 {q0}, [r1]! 136 vst1.u8 {q0}, [r [all...] |
/external/libvpx/libvpx/vp9/common/ |
H A D | vp9_loopfilter_filters.c | 24 uint8_t q0, uint8_t q1, 30 mask |= (abs(q1 - q0) > limit) * -1; 33 mask |= (abs(p0 - q0) * 2 + abs(p1 - q1) / 2 > blimit) * -1; 40 uint8_t q0, uint8_t q1, 44 mask |= (abs(q1 - q0) > thresh) * -1; 46 mask |= (abs(q2 - q0) > thresh) * -1; 48 mask |= (abs(q3 - q0) > thresh) * -1; 55 uint8_t p0, uint8_t q0, 58 int8_t mask = ~flat_mask4(thresh, p3, p2, p1, p0, q0, q1, q2, q3); 60 mask |= (abs(q4 - q0) > thres 21 filter_mask(uint8_t limit, uint8_t blimit, uint8_t p3, uint8_t p2, uint8_t p1, uint8_t p0, uint8_t q0, uint8_t q1, uint8_t q2, uint8_t q3) argument 37 flat_mask4(uint8_t thresh, uint8_t p3, uint8_t p2, uint8_t p1, uint8_t p0, uint8_t q0, uint8_t q1, uint8_t q2, uint8_t q3) argument 52 flat_mask5(uint8_t thresh, uint8_t p4, uint8_t p3, uint8_t p2, uint8_t p1, uint8_t p0, uint8_t q0, uint8_t q1, uint8_t q2, uint8_t q3, uint8_t q4) argument 65 hev_mask(uint8_t thresh, uint8_t p1, uint8_t p0, uint8_t q0, uint8_t q1) argument 115 const uint8_t q0 = s[0 * p], q1 = s[1 * p], q2 = s[2 * p], q3 = s[3 * p]; local 135 const uint8_t q0 = s[0], q1 = s[1], q2 = s[2], q3 = s[3]; local 151 const uint8_t q0 = *oq0, q1 = *oq1, q2 = *oq2, q3 = *oq3; local 176 const uint8_t q0 = s[0 * p], q1 = s[1 * p], q2 = s[2 * p], q3 = s[3 * p]; local 197 const uint8_t q0 = s[0], q1 = s[1], q2 = s[2], q3 = s[3]; local 222 const uint8_t q0 = *oq0, q1 = *oq1, q2 = *oq2, q3 = *oq3, local 270 const uint8_t q0 = s[0 * p], q1 = s[1 * p], q2 = s[2 * p], q3 = s[3 * p]; local 296 const uint8_t q0 = s[0], q1 = s[1], q2 = s[2], q3 = s[3]; local [all...] |
/external/libvpx/libvpx/vp9/common/mips/dspr2/ |
H A D | vp9_loopfilter_masks_dspr2.h | 26 uint32_t q0, uint32_t q1, 61 /* mask |= (abs(q1 - q0) > limit) 62 * hev |= (abs(q1 - q0) > thresh) 64 "subu_s.qb %[c], %[q1], %[q0] \n\t" 65 "subu_s.qb %[r_k], %[q0], %[q1] \n\t" 90 [p1] "r" (p1), [p0] "r" (p0), [q1] "r" (q1), [q0] "r" (q0), 95 /* abs(p0 - q0) */ 96 "subu_s.qb %[c], %[p0], %[q0] \n\t" 97 "subu_s.qb %[r_k], %[q0], 23 vp9_filter_hev_mask_dspr2(uint32_t limit, uint32_t flimit, uint32_t p1, uint32_t p0, uint32_t p3, uint32_t p2, uint32_t q0, uint32_t q1, uint32_t q2, uint32_t q3, uint32_t thresh, uint32_t *hev, uint32_t *mask) argument 128 vp9_filter_hev_mask_flatmask4_dspr2(uint32_t limit, uint32_t flimit, uint32_t thresh, uint32_t p1, uint32_t p0, uint32_t p3, uint32_t p2, uint32_t q0, uint32_t q1, uint32_t q2, uint32_t q3, uint32_t *hev, uint32_t *mask, uint32_t *flat) argument 278 vp9_flatmask5(uint32_t p4, uint32_t p3, uint32_t p2, uint32_t p1, uint32_t p0, uint32_t q0, uint32_t q1, uint32_t q2, uint32_t q3, uint32_t q4, uint32_t *flat2) argument [all...] |
H A D | vp9_mbloop_loopfilter_dspr2.c | 36 uint32_t p3, p2, p1, p0, q0, q1, q2, q3; local 73 "lw %[q0], (%[sq0]) \n\t" 79 [q3] "=&r" (q3), [q2] "=&r" (q2), [q1] "=&r" (q1), [q0] "=&r" (q0) 85 p1, p0, p3, p2, q0, q1, q2, q3, 89 vp9_filter1_dspr2(mask, hev, p1, p0, q0, q1, 121 "sw %[q0], (%[sq0]) \n\t" 127 [q0] "r" (q0), [q1] "r" (q1), [q2] "r" (q2), 133 vp9_filter1_dspr2(mask, hev, p1, p0, q0, q 334 uint32_t p3, p2, p1, p0, q3, q2, q1, q0; local [all...] |
/external/chromium_org/third_party/libwebp/dsp/ |
H A D | dec_sse2.c | 281 #define GET_NOTHEV(p1, p0, q0, q1, hev_thresh, not_hev) { \ 284 const __m128i t_2 = MM_ABS(q1, q0); \ 288 const __m128i t_4 = _mm_subs_epu8(t_2, h); /* abs(q1 - q0) - hev_tresh */ \ 294 #define GET_BASE_DELTA(p1, p0, q0, q1, o) { \ 295 const __m128i qp0 = _mm_subs_epi8(q0, p0); /* q0 - p0 */ \ 297 o = _mm_adds_epi8(o, qp0); /* p1 - q1 + 1 * (q0 - p0) */ \ 298 o = _mm_adds_epi8(o, qp0); /* p1 - q1 + 2 * (q0 - p0) */ \ 299 o = _mm_adds_epi8(o, qp0); /* p1 - q1 + 3 * (q0 - p0) */ \ 302 #define DO_SIMPLE_FILTER(p0, q0, f 328 NeedsFilter(const __m128i* p1, const __m128i* p0, const __m128i* q0, const __m128i* q1, int thresh, __m128i *mask) argument 348 DoFilter2(const __m128i* p1, __m128i* p0, __m128i* q0, const __m128i* q1, int thresh) argument 369 DoFilter4(__m128i* p1, __m128i *p0, __m128i* q0, __m128i* q1, const __m128i* mask, int hev_thresh) argument 416 DoFilter6(__m128i *p2, __m128i* p1, __m128i *p0, __m128i* q0, __m128i* q1, __m128i *q2, const __m128i* mask, int hev_thresh) argument 514 Load16x4(const uint8_t* r0, const uint8_t* r8, int stride, __m128i* p1, __m128i* p0, __m128i* q0, __m128i* q1) argument 558 Store16x4(uint8_t* r0, uint8_t* r8, int stride, __m128i* p1, __m128i* p0, __m128i* q0, __m128i* q1) argument 603 __m128i q0 = _mm_loadu_si128((__m128i*)&p[0]); local 614 __m128i p1, p0, q0, q1; local 693 __m128i p2, p1, p0, q0, q1, q2; local 718 __m128i p3, p2, p1, p0, q0, q1, q2, q3; local 739 __m128i t1, t2, p1, p0, q0, q1; local 768 __m128i t1, t2, p1, p0, q0, q1; local 793 __m128i t1, p2, p1, p0, q0, q1, q2; local 818 __m128i p3, p2, p1, p0, q0, q1, q2, q3; local 838 __m128i t1, t2, p1, p0, q0, q1; local 864 __m128i t1, t2, p1, p0, q0, q1; local [all...] |
H A D | dec_neon.c | 25 #define QRegs "q0", "q1", "q2", "q3", "q4", "q5", "q6", "q7", \ 36 #define NEEDS_FILTER(p1, p0, q0, q1, thresh, mask) \ 37 "vabd.u8 q15," #p0 "," #q0 " \n" /* abs(p0 - q0) */ \ 39 "vqadd.u8 q15, q15, q15 \n" /* abs(p0 - q0) * 2 */ \ 41 "vqadd.u8 q15, q15, q14 \n" /* abs(p0 - q0) * 2 + abs(p1 - q1) / 2 */ \ 45 #define GET_BASE_DELTA(p1, p0, q0, q1, o) \ 46 "vqsub.s8 q15," #q0 "," #p0 " \n" /* (q0 - p0) */ \ 48 "vqadd.s8 " #o "," #o ", q15 \n" /* (p1 - q1) + 1 * (p0 - q0) */ \ [all...] |
/external/webp/src/dsp/ |
H A D | dec_sse2.c | 281 #define GET_NOTHEV(p1, p0, q0, q1, hev_thresh, not_hev) { \ 284 const __m128i t_2 = MM_ABS(q1, q0); \ 288 const __m128i t_4 = _mm_subs_epu8(t_2, h); /* abs(q1 - q0) - hev_tresh */ \ 294 #define GET_BASE_DELTA(p1, p0, q0, q1, o) { \ 295 const __m128i qp0 = _mm_subs_epi8(q0, p0); /* q0 - p0 */ \ 297 o = _mm_adds_epi8(o, qp0); /* p1 - q1 + 1 * (q0 - p0) */ \ 298 o = _mm_adds_epi8(o, qp0); /* p1 - q1 + 2 * (q0 - p0) */ \ 299 o = _mm_adds_epi8(o, qp0); /* p1 - q1 + 3 * (q0 - p0) */ \ 302 #define DO_SIMPLE_FILTER(p0, q0, f 328 NeedsFilter(const __m128i* p1, const __m128i* p0, const __m128i* q0, const __m128i* q1, int thresh, __m128i *mask) argument 348 DoFilter2(const __m128i* p1, __m128i* p0, __m128i* q0, const __m128i* q1, int thresh) argument 369 DoFilter4(__m128i* p1, __m128i *p0, __m128i* q0, __m128i* q1, const __m128i* mask, int hev_thresh) argument 416 DoFilter6(__m128i *p2, __m128i* p1, __m128i *p0, __m128i* q0, __m128i* q1, __m128i *q2, const __m128i* mask, int hev_thresh) argument 514 Load16x4(const uint8_t* r0, const uint8_t* r8, int stride, __m128i* p1, __m128i* p0, __m128i* q0, __m128i* q1) argument 558 Store16x4(uint8_t* r0, uint8_t* r8, int stride, __m128i* p1, __m128i* p0, __m128i* q0, __m128i* q1) argument 603 __m128i q0 = _mm_loadu_si128((__m128i*)&p[0]); local 614 __m128i p1, p0, q0, q1; local 693 __m128i p2, p1, p0, q0, q1, q2; local 718 __m128i p3, p2, p1, p0, q0, q1, q2, q3; local 739 __m128i t1, t2, p1, p0, q0, q1; local 768 __m128i t1, t2, p1, p0, q0, q1; local 793 __m128i t1, p2, p1, p0, q0, q1, q2; local 818 __m128i p3, p2, p1, p0, q0, q1, q2, q3; local 838 __m128i t1, t2, p1, p0, q0, q1; local 864 __m128i t1, t2, p1, p0, q0, q1; local [all...] |
H A D | dec_neon.c | 25 #define QRegs "q0", "q1", "q2", "q3", "q4", "q5", "q6", "q7", \ 36 #define NEEDS_FILTER(p1, p0, q0, q1, thresh, mask) \ 37 "vabd.u8 q15," #p0 "," #q0 " \n" /* abs(p0 - q0) */ \ 39 "vqadd.u8 q15, q15, q15 \n" /* abs(p0 - q0) * 2 */ \ 41 "vqadd.u8 q15, q15, q14 \n" /* abs(p0 - q0) * 2 + abs(p1 - q1) / 2 */ \ 45 #define GET_BASE_DELTA(p1, p0, q0, q1, o) \ 46 "vqsub.s8 q15," #q0 "," #p0 " \n" /* (q0 - p0) */ \ 48 "vqadd.s8 " #o "," #o ", q15 \n" /* (p1 - q1) + 1 * (p0 - q0) */ \ [all...] |
/external/llvm/test/MC/AArch64/ |
H A D | elf-reloc-ldstunsimm.s | 8 str q0, [sp, #:lo12:some_label]
|
/external/fdlibm/ |
H A D | k_rem_pio2.c | 104 * q0 the corresponding exponent of q[0]. Note that the 105 * exponent for q[i] would be q0-24*i. 170 int jz,jx,jv,jp,jk,carry,n,iq[20],i,j,k,m,q0,ih; local 177 /* determine jx,jv,q0, note that 3>q0 */ 180 q0 = e0-24*(jv+1); 201 z = ieee_scalbn(z,q0); /* actual value of z */ 206 if(q0>0) { /* need iq[jz-1] to determine n */ 207 i = (iq[jz-1]>>(24-q0)); n += i; 208 iq[jz-1] -= i<<(24-q0); [all...] |
/external/libvpx/libvpx/vp9/common/arm/neon/ |
H A D | vp9_avg_neon.asm | 36 vld1.8 {q0-q1}, [r0]! 41 vrhadd.u8 q0, q0, q8 45 vst1.8 {q0-q1}, [r2@128]! 52 vld1.8 {q0-q1}, [r0], r1 57 vrhadd.u8 q0, q0, q8 64 vst1.8 {q0-q1}, [r2@128], r3 71 vld1.8 {q0}, [r0], r1 77 vrhadd.u8 q0, q [all...] |
H A D | vp9_short_idct32x32_add_neon.asm | 358 ; into q0-q7 and the second one into q8-q15. There is a stride of 64, 361 vld1.s16 {q0}, [r3]! 397 vtrn.32 q0, q2 405 vtrn.16 q0, q1 420 vst1.16 {q0}, [r0]! 481 vadd.s16 q4, q0, q1 482 vsub.s16 q13, q0, q1 518 vsub.s16 q14, q1, q0 519 vadd.s16 q2, q1, q0 538 vadd.s16 q9, q5, q0 [all...] |
H A D | vp9_copy_neon.asm | 35 vld1.8 {q0-q1}, [r0]! 37 vst1.8 {q0-q1}, [r2@128]! 45 vld1.8 {q0-q1}, [r0], r1 48 vst1.8 {q0-q1}, [r2@128], r3 56 vld1.8 {q0}, [r0], r1 59 vst1.8 {q0}, [r2@128], r3
|
/external/libvpx/libvpx/vp9/common/x86/ |
H A D | vp9_loopfilter_mmx.asm | 68 movq mm4, [rsi] ; q0 69 movq mm0, mm4 ; q0 70 psubusb mm4, mm3 ; q0-=q1 71 psubusb mm3, mm0 ; q1-=q0 72 por mm4, mm3 ; abs(q0-q1) 118 movq mm3, [rsi] ; q0 119 psubusb mm5, mm3 ; p0-=q0 120 psubusb mm3, mm6 ; q0-=p0 121 por mm5, mm3 ; abs(p0 - q0) 122 paddusb mm5, mm5 ; abs(p0-q0)* [all...] |
/external/libvpx/libvpx/vp8/common/x86/ |
H A D | loopfilter_mmx.asm | 68 movq mm4, [rsi] ; q0 69 movq mm0, mm4 ; q0 70 psubusb mm4, mm3 ; q0-=q1 71 psubusb mm3, mm0 ; q1-=q0 72 por mm4, mm3 ; abs(q0-q1) 118 movq mm3, [rsi] ; q0 119 psubusb mm5, mm3 ; p0-=q0 120 psubusb mm3, mm6 ; q0-=p0 121 por mm5, mm3 ; abs(p0 - q0) 122 paddusb mm5, mm5 ; abs(p0-q0)* [all...] |
/external/libvpx/libvpx/vp8/encoder/arm/neon/ |
H A D | fastquantizeb_neon.asm | 33 vld1.16 {q0, q1}, [r4@128] ; load z 37 vabs.s16 q4, q0 ; calculate x = abs(z) 41 vshr.s16 q2, q0, #15 ; sz 52 vld1.16 {q0, q1}, [r4@128] ; load z2 57 vabs.s16 q10, q0 ; calculate x2 = abs(z_2) 59 vshr.s16 q12, q0, #15 ; sz2 119 vand q0, q6, q14 ; get all valid numbers from scan array 127 vmax.u16 q0, q0, q1 ; find maximum value in q0, q [all...] |
H A D | shortfdct_neon.asm | 105 vst1.16 {q0, q1}, [r1@128] 116 vld1.16 {q0}, [r0@128], r2 124 ; transpose q0=ip[0], q1=ip[1], q2=ip[2], q3=ip[3] 125 vtrn.32 q0, q2 ; [A0|B0] 127 vtrn.16 q0, q1 ; [A2|B2] 130 vadd.s16 q11, q0, q3 ; a1 = ip[0] + ip[3] 133 vsub.s16 q14, q0, q3 ; d1 = ip[0] - ip[3] 140 vadd.s16 q0, q11, q12 ; [A0 | B0] = a1 + b1 165 ; transpose q0=ip[0], q1=ip[4], q2=ip[8], q3=ip[12] 166 vtrn.32 q0, q [all...] |
H A D | vp8_memcpy_neon.asm | 32 vld1.8 {q0, q1}, [r1]! ;load src data 35 vst1.8 {q0, q1}, [r0]! ;copy to dst_ptr 61 vld1.8 {q0}, [r1]! ;load src data 63 vst1.8 {q0}, [r0]!
|
/external/libvpx/libvpx/vp8/common/arm/armv6/ |
H A D | simpleloopfilter_v6.asm | 64 ldr r5, [src] ; q0 77 uqsub8 r10, r4, r5 ; p0 - q0 78 uqsub8 r11, r5, r4 ; q0 - p0 80 orr r10, r10, r11 ; abs(p0 - q0) 81 uqadd8 r10, r10, r10 ; abs(p0 - q0) * 2 83 uqadd8 r10, r10, r8 ; abs(p0 - q0)*2 + abs(p1 - q1)/2 95 eor r5, r5, r2 ; q0 offset to convert to a signed value 98 qsub8 r6, r5, r4 ; q0 - p0 99 qadd8 r3, r3, r6 ; += q0 - p0 101 qadd8 r3, r3, r6 ; += q0 [all...] |