Home
last modified time | relevance | path

Searched refs:fma (Results 1 – 25 of 257) sorted by relevance

1234567891011

/glibc-2.36/math/
Dlibm-test-narrow-fma.inc1 /* Test narrowing fma.
23 TEST_aaa_f (fma, arg_qnan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
24 TEST_aaa_f (fma, -arg_qnan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
25 TEST_aaa_f (fma, arg_snan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
26 TEST_aaa_f (fma, -arg_snan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
27 TEST_aaa_f (fma, 1.0, arg_qnan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
28 TEST_aaa_f (fma, 1.0, -arg_qnan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
29 TEST_aaa_f (fma, 1.0, arg_snan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
30 TEST_aaa_f (fma, 1.0, -arg_snan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
31 TEST_aaa_f (fma, 1.0, 2.0, arg_qnan_value, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
[all …]
Dlibm-test-fma.inc1 /* Test fma.
23 TEST_fff_f (fma, qnan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
24 TEST_fff_f (fma, -qnan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
25 TEST_fff_f (fma, snan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
26 TEST_fff_f (fma, -snan_value, 2.0, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
27 TEST_fff_f (fma, 1.0, qnan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
28 TEST_fff_f (fma, 1.0, -qnan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
29 TEST_fff_f (fma, 1.0, snan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
30 TEST_fff_f (fma, 1.0, -snan_value, 3.0, qnan_value, NO_INEXACT_EXCEPTION|INVALID_EXCEPTION),
31 TEST_fff_f (fma, 1.0, 2.0, qnan_value, qnan_value, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
[all …]
Dauto-libm-test-out-fma1 fma 1.0 2.0 3.0
2 = fma downward binary32 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
3 = fma tonearest binary32 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
4 = fma towardzero binary32 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
5 = fma upward binary32 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
6 = fma downward binary64 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
7 = fma tonearest binary64 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
8 = fma towardzero binary64 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
9 = fma upward binary64 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
10 = fma downward intel96 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
[all …]
Dauto-libm-test-out-narrow-fma1 fma 1.0 2.0 3.0
2 = fma downward binary32:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
3 = fma tonearest binary32:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
4 = fma towardzero binary32:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
5 = fma upward binary32:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
6 = fma downward binary64:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
7 = fma tonearest binary64:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
8 = fma towardzero binary64:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
9 = fma upward binary64:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
10 = fma downward intel96:arg_fmt(1,2,0,2) 0x1p+0 0x2p+0 0x3p+0 : 0x5p+0 :
[all …]
/glibc-2.36/sysdeps/x86_64/fpu/multiarch/
DMakefile12 libm-sysdep_routines += e_exp-fma e_log-fma e_pow-fma s_atan-fma \
13 e_asin-fma e_atan2-fma s_sin-fma s_tan-fma \
14 s_sincos-fma
16 CFLAGS-e_asin-fma.c = -mfma -mavx2
17 CFLAGS-e_atan2-fma.c = -mfma -mavx2
18 CFLAGS-e_exp-fma.c = -mfma -mavx2
19 CFLAGS-e_log-fma.c = -mfma -mavx2
20 CFLAGS-e_pow-fma.c = -mfma -mavx2
21 CFLAGS-s_atan-fma.c = -mfma -mavx2
22 CFLAGS-s_sin-fma.c = -mfma -mavx2
[all …]
/glibc-2.36/sysdeps/ia64/fpu/
Dw_tgammal_compat.S1261 fma.s1 FR_p_2 = f1, f1, f1 // 2.0
1328 fma.s1 FR_l_Q0 = f1,FR_l_Y0,f0 // Q0 = Y0
1366 (p15) fma.s1 FR_n_XS = FR_l_AbsX , f1, FR_n_XNS // xs = x - int(x)
1378 fma.s1 FR_l_E2 = FR_l_E0,FR_l_E0,FR_l_E0 // e2 = e+e^2
1383 fma.s1 FR_l_E1 = FR_l_E0,FR_l_E0,f0 // e1 = e^2
1400 fma.s1 FR_l_CXL = FR_l_AbsX, f1, FR_l_CXH // CXL = |X|+CXH
1412 (p15) fma.s1 FR_n_XS2 = FR_n_XS, FR_n_XS, f0 // xs^2 = xs*xs
1424 fma.s1 FR_l_Y1 = FR_l_Y0,FR_l_E2,FR_l_Y0 // y1 = y+y*e2
1429 fma.s1 FR_l_E3 = FR_l_E1,FR_l_E1,FR_l_E0 // e3 = e+e1^2
1454 (p15) fma.s1 FR_n_TT = FR_n_A2L, FR_n_XS2, f0 // T=A2L*x^2
[all …]
Dlibm_lgamma.S1340 fma.s1 FR_2 = f1,f1,f1
1395 (p14) fma.s1 FR_5 = FR_2,FR_2,f1
1406 (p14) fma.s1 FR_Xp1 = f1,f1,FR_NormX // x+1
1412 (p6) fma.s1 FR_x = f0,f0,FR_NormX
1422 (p14) fma.s1 FR_3 = f1,f1,FR_2
1427 (p14) fma.s1 FR_Xp4 = FR_2,FR_2,FR_NormX
1437 (p14) fma.s1 FR_r2 = FR_r,FR_r,f0 // log(x)
1447 (p14) fma.s1 FR_Xp5 = FR_5,f1,FR_NormX
1457 (p14) fma.s1 FR_Xp10 = FR_5,FR_2,FR_NormX
1463 fma.s1 FR_C01 = FR_x,f1,FR_C01
[all …]
Ds_atan.S277 fma.s1 atan2_two = f1,f1,f1
289 fma.s1 atan2_ysq = atan2_Y,atan2_Y,f0
322 fma.s1 atan2_wp_Y = atan2_u1_Y, atan2_u1_Y, f0
329 fma.s1 atan2_z1_X = atan2_u1_X, atan2_Y, f0
341 fma.s1 atan2_z2_X = atan2_u1_X, atan2_ysq, f0
346 (p10) fma.d.s0 f8 = atan2_Y,atan2_X,f0 // If y=nan, result quietized y
367 fma.s1 atan2_E = atan2_u1_Y, atan2_B1Y, atan2_Y
372 fma.s1 atan2_B1sq_Y = atan2_B1Y, atan2_B1Y, f0
379 (p7) fma.s1 atan2_wp_X = atan2_z1_X, atan2_z1_X, f0
384 (p7) fma.s1 atan2_B1sq_X = atan2_B1X, atan2_B1X, f0
[all …]
De_acosl.S961 fma.s1 F_S2T2 = f8, f1, F_T
970 fma.s1 F_S2T2 = F_S2T2, F_D, f0
979 fma.s1 F_INV_1T2 = F_INV_1T2, F_2M64, f0
987 fma.s1 F_SQRT_1T2 = F_SQRT_1T2, F_2M64, f0
1007 fma.s1 F_TSQRT = F_T, F_SQRT_1T2, f0;;
1014 fma.s1 F_S89 = F_CS9, F_X, F_CS8
1021 fma.s1 F_S67 = F_CS7, F_X, F_CS6
1028 fma.s1 F_S45 = F_CS5, F_X, F_CS4
1035 fma.s1 F_X2 = F_X, F_X, f0
1072 fma.s1 F_S69 = F_S89, F_X2, F_S67
[all …]
Ds_erf.S316 fma.s1 fArgSqr = f8, f8, f0 // x^2
360 (p7) fma.d.s0 f8 = f8,f1,f8 // NaN or +/-0
425 fma.s1 fTSqr = fArgAbsNorm, fArgAbsNorm, f0
481 fma.s1 fTDeg3 = fArgAbsNorm, fTSqr, f0
486 fma.s1 fTQuadr = fTSqr, fTSqr, f0
493 fma.s1 fArgAbsNormSgn = fArgAbsNorm, fSignumX, f0
504 fma.s1 fRes = fA19, fArgAbsNorm, fA18
515 fma.s1 fA17 = fA17, fArgAbsNorm, fA16
521 fma.s1 fA15 = fA15, fArgAbsNorm, fA14
527 fma.s1 fTDeg7 = fTDeg3, fTQuadr, f0
[all …]
De_asinl.S955 fma.s1 F_S2T2 = f8, f1, F_T
964 fma.s1 F_S2T2 = F_S2T2, F_D, f0
973 fma.s1 F_INV_1T2 = F_INV_1T2, F_2M64, f0
981 fma.s1 F_SQRT_1T2 = F_SQRT_1T2, F_2M64, f0
999 fma.s1 F_TSQRT = F_T, F_SQRT_1T2, f0
1007 fma.s1 F_S89 = F_CS9, F_X, F_CS8
1014 fma.s1 F_S67 = F_CS7, F_X, F_CS6
1021 fma.s1 F_S45 = F_CS5, F_X, F_CS4
1028 fma.s1 F_X2 = F_X, F_X, f0
1051 fma.s1 F_S69 = F_S89, F_X2, F_S67
[all …]
De_asin.S281 fma.s1 f1pX = f1, f1, f8
292 fma.s1 fXSqr = f8, f8, f0 // x^2
319 (p8) fma.s1 fR = f1, f1, f8
357 fma.s1 fXCube = fXSqr, f8, f0 // x^3
369 fma.s1 fXQuadr = fXSqr, fXSqr, f0 // x^4
375 fma.s1 fRSqr = fR, fR, f0 // R^2
387 (p9) fma.s1 fH = fHalf, f1mXRcp, f0 // H0 for x > 0
392 (p9) fma.s1 fS = f1mX, f1mXRcp, f0 // S0 for x > 0
398 (p8) fma.s1 fH = fHalf, f1pXRcp, f0 // H0 for x < 0
403 (p8) fma.s1 fS = f1pX, f1pXRcp, f0 // S0 for x > 0
[all …]
Dw_tgamma_compat.S724 fma.s1 FR_1m2X = f8,f1,f8 // 2x
734 fma.s1 FR_2 = f1,f1,f1 // 2
752 fma.s1 FR_r02 = f8,f1,f0
757 fma.s1 FR_NormX = f8,f1,f0
772 (p14) fma.s1 FR_1m2X = f1,f1,FR_1m2X // RB=1-2|x|
783 fma.s1 FR_4 = FR_2,FR_2,f0
788 fma.s1 FR_6 = FR_2,FR_2,FR_2
799 fma.s1 FR_5 = FR_2,FR_2,f1
821 (p15) fma.s1 FR_X2pX = f8,f8,f8 // RA=x^2+x
827 fma.s1 FR_Rq2 = FR_2,f1,FR_1m2X // 2 + B
[all …]
De_atan2.S346 fma.s1 atan2_two = f1,f1,f1
358 fma.s1 atan2_xsq = atan2_X,atan2_X,f0
370 fma.s1 atan2_ysq = atan2_Y,atan2_Y,f0
391 (p10) fma.d.s0 f8 = atan2_X,atan2_Y,f0 // If y=nan, result quietized y
399 fma.s1 atan2_z1_X = atan2_u1_X, atan2_Y, f0
411 fma.s1 atan2_z1_Y = atan2_u1_Y, atan2_X, f0
423 fma.s1 atan2_z2_X = atan2_u1_X, atan2_ysq, f0
428 fma.s1 atan2_z2_Y = atan2_u1_Y, atan2_xsq, f0
442 (p12) fma.d.s0 f8 = atan2_X,atan2_Y,f0 // If x nan, result quiet x
458 fma.s1 atan2_wp_X = atan2_z1_X, atan2_z1_X, f0
[all …]
Ds_atanf.S192 fma.s1 atanf_t = f8,f8,f0
234 fma.s1 atanf_zsq = atanf_z,atanf_z,f0
241 fma.s1 atanf_xcub = f8,atanf_t,f0
246 fma.s1 atanf_tsq = atanf_t,atanf_t,f0
290 fma.s1 atanf_bsq = atanf_b,atanf_b,f0
295 fma.s1 atanf_z4 = atanf_zsq,atanf_zsq,f0
302 fma.s1 atanf_x6 = atanf_t,atanf_tsq,f0
307 fma.s1 atanf_t4 = atanf_tsq,atanf_tsq,f0
314 fma.s1 atanf_x5 = atanf_t,atanf_xcub,f0
321 fma.s1 atanf_poly_r1 = atanf_b,atanf_coeff_R1,f1
[all …]
Dlibm_lgammal.S2137 fma.s1 fTwo = f1, f1, f1 // 2.0
2239 fma.s1 fSix = fTwo, fTwo, fTwo // 6.0
2251 fma.s1 fRoot = f8, f1, f1 // x + 1
2295 fma.s1 fRcpX = fInvX, fRcpX, fRcpX // end of 1st NR iteration
2334 fma.s1 fThirteen = fSix, fTwo, f1 // 13.0
2349 fma.s1 fDxSqr = fDx, fDx, f0 // deltaX^2
2428 fma.s1 fXSqr = FR_FracX, FR_FracX, f0 // y^2
2435 fma.s1 fRcpX = fInvX, fRcpX, fRcpX // end of 2nd NR iteration
2477 fma.s1 fDelX4 = fDxSqr, fDxSqr, f0 // deltaX^4
2489 fma.s1 fLnSin20 = fLnSin20, fDxSqr, fLnSin18
[all …]
Dlibm_lgammaf.S780 fma.s1 FR_Xp1 = f8,f1,f1
829 fma.s1 FR_Xp2 = f1,f1,FR_Xp1 // (x+2)
853 (p6) fma.s1 FR_x = f0,f0,FR_NormX
873 fma.s1 FR_r2 = FR_r,FR_r,f0
883 fma.s1 FR_Xp3 = FR_2,f1,FR_Xp1 // (x+3)
888 fma.s1 FR_Xp4 = FR_2,FR_2,FR_NormX // (x+4)
901 fma.s1 FR_Xp5 = FR_2,FR_2,FR_Xp1 // (x+5)
907 fma.s1 FR_P32 = FR_P3,FR_r,FR_P2
912 fma.s1 FR_x2 = FR_x,FR_x,f0
925 (p12) fma.s1 FR_P10 = FR_P1,FR_r,f1
[all …]
De_acos.S284 fma.s1 f1pX = f1, f1, f8
295 fma.s1 fXSqr = f8, f8, f0 // x^2
322 (p8) fma.s1 fR = f1, f1, f8
360 fma.s1 fXCube = fXSqr, f8, f0 // x^3
372 fma.s1 fXQuadr = fXSqr, fXSqr, f0 // x^4
378 fma.s1 fRSqr = fR, fR, f0 // R^2
390 (p9) fma.s1 fH = fHalf, f1mXRcp, f0 // H0 for x > 0
395 (p9) fma.s1 fS = f1mX, f1mXRcp, f0 // S0 for x > 0
401 (p8) fma.s1 fH = fHalf, f1pXRcp, f0 // H0 for x < 0
406 (p8) fma.s1 fS = f1pX, f1pXRcp, f0 // S0 for x > 0
[all …]
Ds_cosl.S907 fma.s1 FR_N_float_signif = FR_Input_X, FR_inv_pi_2to63, FR_rshf_2to64
925 fma.s1 FR_rsq = FR_norm_x, FR_norm_x, f0 // rsq = x*x, in case |x| < pi/4
978 (p6) fma.s1 FR_r_cubed = FR_r, FR_rsq, f0 // r^3 if |x| < pi/4
1018 fma.s1 FR_w = FR_N_float, FR_P_2, f0 // w = N * P_2
1052 fma.s1 FR_rsq = FR_r, FR_r, f0 // rsq = r * r, for |s| >= 2^-33
1059 (p7) fma.s1 FR_w = FR_N_float, FR_P_3, f0
1080 (p7) fma.s1 FR_U_1 = FR_N_float, FR_P_2, FR_w
1097 (p9) fma.s1 FR_poly = FR_rsq, FR_PP_8, FR_PP_7 // poly = rsq*PP_8+PP_7 if i_1=0
1102 (p10) fma.s1 FR_poly = FR_rsq, FR_QQ_8, FR_QQ_7 // poly = rsq*QQ_8+QQ_7 if i_1=1
1116 (p6) fma.s1 FR_r_cubed = FR_r, FR_rsq, f0 // rcubed = r * rsq
[all …]
Dlibm_sincosl.S957 fma.s1 FR_N_float_signif = FR_Input_X, FR_inv_pi_2to63, FR_rshf_2to64
974 fma.s1 FR_rsq = FR_norm_x, FR_norm_x, f0 // rsq = x*x, in case |x| < pi/4
1022 (p6) fma.s1 FR_r_cubed = FR_r, FR_rsq, f0 // r^3 if |x| < pi/4
1057 fma.s1 FR_w = FR_N_float, FR_P_2, f0 // w = N * P_2
1086 fma.s1 FR_rsq = FR_r, FR_r, f0 // rsq = r * r, for |s| >= 2^-33
1092 (p7) fma.s1 FR_w = FR_N_float, FR_P_3, f0
1110 (p7) fma.s1 FR_U_1 = FR_N_float, FR_P_2, FR_w
1126 fma.s1 FR_polyS = FR_rsq, FR_PP_8, FR_PP_7 // poly = rsq*PP_8+PP_7
1131 fma.s1 FR_polyC = FR_rsq, FR_QQ_8, FR_QQ_7 // poly = rsq*QQ_8+QQ_7
1143 (p6) fma.s1 FR_r_cubed = FR_r, FR_rsq, f0 // rcubed = r * rsq
[all …]
Dlibm_sincos_large.S756 fma.s1 FR_SINCOS_ARG = f8, f1, f0 // Save argument for sin and cos
768 fma.s1 FR_SINCOS_RES_SIN = f8, f1, f0 // Save sin result
774 fma.s1 f8 = FR_SINCOS_ARG, f1, f0 // Arg for cos
780 fma.s1 f9 = FR_SINCOS_RES_SIN, f1, f0 // Out sin result
1014 fma.s1 FR_N_float = FR_Input_X, FR_Inv_pi_by_2, f0
1127 fma.s1 FR_w = FR_N_float, FR_P_2, f0
1158 (p7) fma.s1 FR_w = FR_N_float, FR_P_3, f0
1164 (p7) fma.s1 FR_U_1 = FR_N_float, FR_P_2, FR_w
1255 (p10) fma.s1 FR_poly = f0, f1, FR_Neg_Two_to_M67
1272 (p7) fma.s1 FR_rsq = FR_r, FR_r, f0
[all …]
De_atan2f.S364 fma.s1 atan2f_xsq = f9,f9,f0
376 fma.s1 atan2f_ysq = f8,f8,f0
387 fma.s1 atan2f_xy = f9,f8,f0
407 fma.s1 atan2f_T_denom = atan2f_Z0,atan2f_xsq,f8
412 fma.s1 atan2f_Z = atan2f_Z0,f9,f0
419 fma.s1 atan2f_G_denom = atan2f_A0,atan2f_ysq,f9
424 fma.s1 atan2f_A = atan2f_A0,f8,f0
455 (p8) fma.s1 atan2f_const_1 = atan2f_sgn_Y,f0,f0
460 (p9) fma.s1 atan2f_const_1 = atan2f_sgn_Y,f1,f0
472 (p6) fma.s1 atan2f_Usq = atan2f_Z,atan2f_Z,f0
[all …]
De_acosf.S262 fma.s1 acosf_x2 = f8,f8,f0
306 fma.s1 acosf_t2 = acosf_t,acosf_t,f0
311 fma.s1 acosf_x4 = acosf_x2,acosf_x2,f0
323 fma.s1 acosf_x3 = f8,acosf_x2,f0
335 (p8) fma.s.s0 f8 = f8,f1,f0
354 fma.s1 acosf_x8 = acosf_x4,acosf_x4,f0
359 fma.s1 acosf_t4 = acosf_t2,acosf_t2,f0
365 fma.s1 acosf_x5 = acosf_x2,acosf_x3,f0
370 fma.s1 acosf_yby2 = acosf_t,acosf_1by2,f0
377 fma.s1 acosf_Az = acosf_t,acosf_B,f0
[all …]
De_asinf.S265 fma.s1 asinf_x2 = f8,f8,f0
309 fma.s1 asinf_t2 = asinf_t,asinf_t,f0
314 fma.s1 asinf_x4 = asinf_x2,asinf_x2,f0
326 fma.s1 asinf_x3 = f8,asinf_x2,f0
338 (p8) fma.s.s0 f8 = f8,f1,f0
357 fma.s1 asinf_x8 = asinf_x4,asinf_x4,f0
362 fma.s1 asinf_t4 = asinf_t2,asinf_t2,f0
368 fma.s1 asinf_x5 = asinf_x2,asinf_x3,f0
373 fma.s1 asinf_yby2 = asinf_t,asinf_1by2,f0
380 fma.s1 asinf_Az = asinf_t,asinf_B,f0
[all …]
/glibc-2.36/sysdeps/ia64/
Dia64libgcc.S22 (p6) fma.s1 f12 = farg0, f10, f0
24 (p6) fma.s1 f13 = f11, f11, f0
25 (p6) fma.s1 f14 = f11, f11, f11
27 (p6) fma.s1 f11 = f13, f13, f11
28 (p6) fma.s1 f13 = f14, f10, f10
30 (p6) fma.s1 f10 = f13, f11, f10
33 (p6) fma.s1 f11 = f11, f10, f12
36 (p6) fma.s1 f10 = f12, f10, f10
39 (p6) fma.s0 fret0 = f12, f10, f11
58 (p6) fma.s1 f11 = f12, f11, f11
[all …]

1234567891011