Searched refs:ld128 (Results 1 – 9 of 9) sorted by relevance
/external/XNNPACK/scripts/ |
D | generate-f32-gemm.sh | 57 …s/xngen src/f32-gemm/4x8-aarch64-neonfma-ld128.S.in -D INC=0 -o src/f32-gemm/gen/4x8-aarch64… 58 …xngen src/f32-gemm/4x8-aarch64-neonfma-ld128.S.in -D INC=1 -o src/f32-gemm/gen-inc/4x8-aarch… 84 …s/xngen src/f32-gemm/6x8-aarch64-neonfma-ld128.S.in -D INC=0 -o src/f32-gemm/gen/6x8-aarch64… 85 …xngen src/f32-gemm/6x8-aarch64-neonfma-ld128.S.in -D INC=1 -o src/f32-gemm/gen-inc/6x8-aarch… 110 …ls/xngen src/f32-gemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=0 -D INC=0 -D DUP=0 -o src/f32-g… 111 …/xngen src/f32-gemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=0 -D INC=1 -D DUP=0 -o src/f32-gem… 112 …s/xngen src/f32-gemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=1 -D INC=0 -D DUP=0 -o src/f32-ge… 113 …xngen src/f32-gemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=1 -D INC=1 -D DUP=0 -o src/f32-gemm… 114 …ls/xngen src/f32-gemm/neon-ld128.c.in -D MR=6 -D NR=8 -D FMA=0 -D INC=0 -D DUP=0 -o src/f32-g… 115 …/xngen src/f32-gemm/neon-ld128.c.in -D MR=6 -D NR=8 -D FMA=0 -D INC=1 -D DUP=0 -o src/f32-gem… [all …]
|
D | generate-f32-igemm.sh | 46 tools/xngen src/f32-igemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=0 -D DUP=0 -o src/f32-igemm/ge… 47 …ools/xngen src/f32-igemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=1 -D DUP=0 -o src/f32-igemm/ge… 48 tools/xngen src/f32-igemm/neon-ld128.c.in -D MR=6 -D NR=8 -D FMA=0 -D DUP=0 -o src/f32-igemm/ge… 49 …ools/xngen src/f32-igemm/neon-ld128.c.in -D MR=6 -D NR=8 -D FMA=1 -D DUP=0 -o src/f32-igemm/ge… 61 tools/xngen src/f32-igemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=0 -D DUP=1 -o src/f32-igemm/ge… 62 tools/xngen src/f32-igemm/neon-ld128.c.in -D MR=4 -D NR=8 -D FMA=1 -D DUP=1 -o src/f32-igemm/ge… 63 tools/xngen src/f32-igemm/neon-ld128.c.in -D MR=6 -D NR=8 -D FMA=0 -D DUP=1 -o src/f32-igemm/ge… 64 tools/xngen src/f32-igemm/neon-ld128.c.in -D MR=6 -D NR=8 -D FMA=1 -D DUP=1 -o src/f32-igemm/ge…
|
/external/XNNPACK/ |
D | Android.bp | 347 "src/f32-gemm/gen/4x8-neon-lane-ld128.c", 351 "src/f32-gemm/gen/6x8-neon-lane-ld128.c", 353 "src/f32-gemm/gen/4x8-neon-dup-ld128.c", 356 "src/f32-gemm/gen/6x8-neon-dup-ld128.c", 362 "src/f32-gemm/gen-inc/4x8-neon-lane-ld128.c", 366 "src/f32-gemm/gen-inc/6x8-neon-lane-ld128.c", 368 "src/f32-gemm/gen-inc/4x8-neon-dup-ld128.c", 371 "src/f32-gemm/gen-inc/6x8-neon-dup-ld128.c", 381 "src/f32-igemm/gen/4x8-neon-lane-ld128.c", 384 "src/f32-igemm/gen/6x8-neon-lane-ld128.c", [all …]
|
D | BUILD.bazel | 468 "src/f32-gemm/gen/4x8-neon-lane-ld128.c", 472 "src/f32-gemm/gen/6x8-neon-lane-ld128.c", 474 "src/f32-gemm/gen/4x8-neon-dup-ld128.c", 477 "src/f32-gemm/gen/6x8-neon-dup-ld128.c", 483 "src/f32-gemm/gen-inc/4x8-neon-lane-ld128.c", 487 "src/f32-gemm/gen-inc/6x8-neon-lane-ld128.c", 489 "src/f32-gemm/gen-inc/4x8-neon-dup-ld128.c", 492 "src/f32-gemm/gen-inc/6x8-neon-dup-ld128.c", 502 "src/f32-igemm/gen/4x8-neon-lane-ld128.c", 505 "src/f32-igemm/gen/6x8-neon-lane-ld128.c", [all …]
|
D | CMakeLists.txt | 472 src/f32-gemm/gen/4x8-neon-lane-ld128.c 476 src/f32-gemm/gen/6x8-neon-lane-ld128.c 478 src/f32-gemm/gen/4x8-neon-dup-ld128.c 481 src/f32-gemm/gen/6x8-neon-dup-ld128.c 487 src/f32-gemm/gen-inc/4x8-neon-lane-ld128.c 491 src/f32-gemm/gen-inc/6x8-neon-lane-ld128.c 493 src/f32-gemm/gen-inc/4x8-neon-dup-ld128.c 496 src/f32-gemm/gen-inc/6x8-neon-dup-ld128.c 506 src/f32-igemm/gen/4x8-neon-lane-ld128.c 509 src/f32-igemm/gen/6x8-neon-lane-ld128.c [all …]
|
/external/XNNPACK/src/f32-gemm/gen-inc/ |
D | 6x8-aarch64-neonfma-ld128.S | 128 # 48 FMA + 6 ld128 A + 4 LDP B
|
/external/XNNPACK/src/f32-gemm/gen/ |
D | 6x8-aarch64-neonfma-ld128.S | 132 # 48 FMA + 6 ld128 A + 4 LDP B
|
/external/XNNPACK/src/f32-gemm/ |
D | 6x8-aarch64-neonfma-ld128.S.in | 155 # 48 FMA + 6 ld128 A + 4 LDP B
|
/external/neon_2_sse/ |
D | NEON_2_SSE.h | 9848 __m128i ld128; variable 9849 ld128 = vld1q_u8(ptr); //merge two 64-bits in 128 bit 9850 ld128 = _mm_shuffle_epi8(ld128, *(__m128i*)mask8_16_even_odd); 9851 vst1q_u8((v.val), ld128); // v.val[1] = _mm_shuffle_epi32(v.val[0], _SWAP_HI_LOW32); 9859 __m128i ld128; variable 9860 ld128 = vld1q_u16(ptr); //merge two 64-bits in 128 bit 9861 ld128 = _mm_shuffle_epi8(ld128, *(__m128i*) mask8_32_even_odd); 9862 vst1q_u16((v.val), ld128); 9870 __m128i ld128; variable 9871 ld128 = vld1q_u32(ptr); //merge two 64-bits in 128 bit [all …]
|