/dports/lang/gcc10/gcc-10.3.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/devel/arm-none-eabi-gcc/gcc-8.4.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/devel/riscv64-none-elf-gcc/gcc-8.4.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/devel/aarch64-none-elf-gcc/gcc-8.4.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc9-aux/gcc-9.1.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc9-devel/gcc-9-20211007/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/devel/riscv32-unknown-elf-gcc/gcc-8.4.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc11-devel/gcc-11-20211009/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc8/gcc-8.5.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/devel/riscv64-gcc/gcc-8.3.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/devel/avr-gcc/gcc-10.2.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc11/gcc-11.2.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc9/gcc-9.4.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/misc/cxx_atomics_pic/gcc-11.2.0/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc10-devel/gcc-10-20211008/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/lang/gcc12-devel/gcc-12-20211205/gcc/testsuite/gcc.target/arm/ |
H A D | armv8_2-fp16-neon-3.c | 19 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_2() 25 return vsub_f16 (vmul_f16 (vneg_f16 (a), vneg_f16 (b)), vneg_f16 (c)); in test_vfma_3() 31 return vsub_f16 (vmul_f16 (a, b), vneg_f16 (c)); in test_vfma_4() 63 return vsub_f16 (c, vmul_f16 (a, b)); in test_vfms_1() 69 return vsub_f16 (a, vmul_f16 (b, c)); in test_vfms_2()
|
/dports/graphics/realsr-ncnn-vulkan/realsr-ncnn-vulkan-20210210/src/ncnn/src/layer/arm/ |
H A D | convolution_3x3_pack4_fp16s.h | 327 … float16x4_t _tmp0m = vfma_n_f16(vsub_f16(_r00, _r06), vsub_f16(_r04, _r02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 328 … float16x4_t _tmp7m = vfma_n_f16(vsub_f16(_r07, _r01), vsub_f16(_r03, _r05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 342 float16x4_t _tmp2m = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 356 float16x4_t _tmp4m = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 370 float16x4_t _tmp6m = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 400 … float16x4_t _r0tm0 = vfma_n_f16(vsub_f16(_tmp00, _tmp06), vsub_f16(_tmp04, _tmp02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 401 … float16x4_t _r0tm7 = vfma_n_f16(vsub_f16(_tmp07, _tmp01), vsub_f16(_tmp03, _tmp05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 413 float16x4_t _r0tm2 = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 425 float16x4_t _r0tm4 = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 437 float16x4_t _r0tm6 = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() [all …]
|
H A D | neon_mathfun_fp16s.h | 90 x = vsub_f16(x, one); in log_ps() 91 e = vsub_f16(e, vreinterpret_f16_u16(vand_u16(vreinterpret_u16_f16(one), mask))); in log_ps() 121 y = vsub_f16(y, tmp); in log_ps() 232 fx = vsub_f16(tmp, vreinterpret_f16_u16(mask)); in exp_ps() 236 x = vsub_f16(x, tmp); in exp_ps() 237 x = vsub_f16(x, z); in exp_ps() 420 y1 = vsub_f16(y1, vmul_f16(z, vdup_n_f16(0.5f))); in sincos_ps() 553 float16x4_t y0 = vsub_f16(_one, vdiv_f16(_two, vadd_f16(exp_x_x, _one))); in tanh_ps()
|
/dports/graphics/vapoursynth-waifu2x-ncnn-vulkan/vapoursynth-waifu2x-ncnn-vulkan-r4/deps/ncnn/src/layer/arm/ |
H A D | convolution_3x3_pack4_fp16s.h | 327 … float16x4_t _tmp0m = vfma_n_f16(vsub_f16(_r00, _r06), vsub_f16(_r04, _r02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 328 … float16x4_t _tmp7m = vfma_n_f16(vsub_f16(_r07, _r01), vsub_f16(_r03, _r05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 342 float16x4_t _tmp2m = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 356 float16x4_t _tmp4m = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 370 float16x4_t _tmp6m = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 400 … float16x4_t _r0tm0 = vfma_n_f16(vsub_f16(_tmp00, _tmp06), vsub_f16(_tmp04, _tmp02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 401 … float16x4_t _r0tm7 = vfma_n_f16(vsub_f16(_tmp07, _tmp01), vsub_f16(_tmp03, _tmp05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 413 float16x4_t _r0tm2 = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 425 float16x4_t _r0tm4 = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 437 float16x4_t _r0tm6 = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() [all …]
|
H A D | neon_mathfun_fp16s.h | 93 x = vsub_f16(x, one); in log_ps() 94 e = vsub_f16(e, vreinterpret_f16_u16(vand_u16(vreinterpret_u16_f16(one), mask))); in log_ps() 124 y = vsub_f16(y, tmp); in log_ps() 235 fx = vsub_f16(tmp, vreinterpret_f16_u16(mask)); in exp_ps() 239 x = vsub_f16(x, tmp); in exp_ps() 240 x = vsub_f16(x, z); in exp_ps() 423 y1 = vsub_f16(y1, vmul_f16(z, vdup_n_f16(0.5f))); in sincos_ps() 556 float16x4_t y0 = vsub_f16(_one, vdiv_f16(_two, vadd_f16(exp_x_x, _one))); in tanh_ps()
|
/dports/graphics/waifu2x-ncnn-vulkan/waifu2x-ncnn-vulkan-20210521/src/ncnn/src/layer/arm/ |
H A D | convolution_3x3_pack4_fp16s.h | 327 … float16x4_t _tmp0m = vfma_n_f16(vsub_f16(_r00, _r06), vsub_f16(_r04, _r02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 328 … float16x4_t _tmp7m = vfma_n_f16(vsub_f16(_r07, _r01), vsub_f16(_r03, _r05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 342 float16x4_t _tmp2m = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 356 float16x4_t _tmp4m = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 370 float16x4_t _tmp6m = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 400 … float16x4_t _r0tm0 = vfma_n_f16(vsub_f16(_tmp00, _tmp06), vsub_f16(_tmp04, _tmp02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 401 … float16x4_t _r0tm7 = vfma_n_f16(vsub_f16(_tmp07, _tmp01), vsub_f16(_tmp03, _tmp05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 413 float16x4_t _r0tm2 = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 425 float16x4_t _r0tm4 = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 437 float16x4_t _r0tm6 = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() [all …]
|
H A D | neon_mathfun_fp16s.h | 93 x = vsub_f16(x, one); in log_ps() 94 e = vsub_f16(e, vreinterpret_f16_u16(vand_u16(vreinterpret_u16_f16(one), mask))); in log_ps() 124 y = vsub_f16(y, tmp); in log_ps() 235 fx = vsub_f16(tmp, vreinterpret_f16_u16(mask)); in exp_ps() 239 x = vsub_f16(x, tmp); in exp_ps() 240 x = vsub_f16(x, z); in exp_ps() 423 y1 = vsub_f16(y1, vmul_f16(z, vdup_n_f16(0.5f))); in sincos_ps() 556 float16x4_t y0 = vsub_f16(_one, vdiv_f16(_two, vadd_f16(exp_x_x, _one))); in tanh_ps()
|
/dports/benchmarks/vkpeak/vkpeak-20210430/ncnn/src/layer/arm/ |
H A D | convolution_3x3_pack4_fp16s.h | 327 … float16x4_t _tmp0m = vfma_n_f16(vsub_f16(_r00, _r06), vsub_f16(_r04, _r02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 328 … float16x4_t _tmp7m = vfma_n_f16(vsub_f16(_r07, _r01), vsub_f16(_r03, _r05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 342 float16x4_t _tmp2m = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 356 float16x4_t _tmp4m = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 370 float16x4_t _tmp6m = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 400 … float16x4_t _r0tm0 = vfma_n_f16(vsub_f16(_tmp00, _tmp06), vsub_f16(_tmp04, _tmp02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 401 … float16x4_t _r0tm7 = vfma_n_f16(vsub_f16(_tmp07, _tmp01), vsub_f16(_tmp03, _tmp05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 413 float16x4_t _r0tm2 = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 425 float16x4_t _r0tm4 = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 437 float16x4_t _r0tm6 = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() [all …]
|
H A D | neon_mathfun_fp16s.h | 93 x = vsub_f16(x, one); in log_ps() 94 e = vsub_f16(e, vreinterpret_f16_u16(vand_u16(vreinterpret_u16_f16(one), mask))); in log_ps() 124 y = vsub_f16(y, tmp); in log_ps() 235 fx = vsub_f16(tmp, vreinterpret_f16_u16(mask)); in exp_ps() 239 x = vsub_f16(x, tmp); in exp_ps() 240 x = vsub_f16(x, z); in exp_ps() 423 y1 = vsub_f16(y1, vmul_f16(z, vdup_n_f16(0.5f))); in sincos_ps() 556 float16x4_t y0 = vsub_f16(_one, vdiv_f16(_two, vadd_f16(exp_x_x, _one))); in tanh_ps()
|
/dports/misc/ncnn/ncnn-20211208/src/layer/arm/ |
H A D | convolution_3x3_pack4_fp16s.h | 327 … float16x4_t _tmp0m = vfma_n_f16(vsub_f16(_r00, _r06), vsub_f16(_r04, _r02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 328 … float16x4_t _tmp7m = vfma_n_f16(vsub_f16(_r07, _r01), vsub_f16(_r03, _r05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 342 float16x4_t _tmp2m = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 356 float16x4_t _tmp4m = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 370 float16x4_t _tmp6m = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 400 … float16x4_t _r0tm0 = vfma_n_f16(vsub_f16(_tmp00, _tmp06), vsub_f16(_tmp04, _tmp02), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 401 … float16x4_t _r0tm7 = vfma_n_f16(vsub_f16(_tmp07, _tmp01), vsub_f16(_tmp03, _tmp05), 5.25f); in conv3x3s1_winograd64_pack4_fp16sa_neon() 413 float16x4_t _r0tm2 = vsub_f16(_tmp12a, _tmp12b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 425 float16x4_t _r0tm4 = vsub_f16(_tmp34a, _tmp34b); in conv3x3s1_winograd64_pack4_fp16sa_neon() 437 float16x4_t _r0tm6 = vsub_f16(_tmp56a, _tmp56b); in conv3x3s1_winograd64_pack4_fp16sa_neon() [all …]
|