Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | Add AVX optimized versions for some x86-64 math functions | Ulrich Drepper | 2011-10-25 | 1 | -58/+41 | |
| | ||||||
* | Use math_force_eval in more places | Ulrich Drepper | 2011-10-25 | 20 | -310/+235 | |
| | ||||||
* | Cleanup FMA4 patch | Ulrich Drepper | 2011-10-25 | 24 | -197/+418 | |
| | | | | | Move the FMA4 code into its own section. Avoid some of the duplication of data resulting from the double use of source files. | |||||
* | Optimize accurate 64-bit routines for FMA4 on x86-64 | Ulrich Drepper | 2011-10-24 | 12 | -176/+205 | |
| | ||||||
* | Finish renamed DLA_FMA -> DLA_FMS | Ulrich Drepper | 2011-10-24 | 8 | -9/+9 | |
| | ||||||
* | Optimized remquo for 64-bit platforms | Ulrich Drepper | 2011-10-23 | 1 | -0/+112 | |
| | ||||||
* | Optimize f_mod | Ulrich Drepper | 2011-10-23 | 2 | -6/+109 | |
| | | | | | Branch prediction for the 32-bit implementation and a new optimized 64-bit implementation. | |||||
* | Clean up FMA use | Ulrich Drepper | 2011-10-23 | 1 | -3/+3 | |
| | | | | | The macro's name should reflect that subtraction is being done. And use __builtin_fma, it seems to work after all. | |||||
* | Clean up last dla.h change | Ulrich Drepper | 2011-10-23 | 10 | -21/+9 | |
| | ||||||
* | Add missing _finite aliases | Andreas Schwab | 2011-10-23 | 1 | -0/+2 | |
| | ||||||
* | Improve DLA_FMA for FMA4 | Ulrich Drepper | 2011-10-22 | 1 | -2/+2 | |
| | ||||||
* | Start using fma in the libm implementation | Ulrich Drepper | 2011-10-22 | 9 | -150/+192 | |
| | ||||||
* | Fix compilation due to __nan defines | Andreas Schwab | 2011-10-22 | 1 | -0/+1 | |
| | ||||||
* | Optimization to some complex math functions | Ulrich Drepper | 2011-10-22 | 3 | -3/+3 | |
| | | | | Also, change last reference to __sqrt into __sqrt to easily. | |||||
* | Add some __builtin_expect to generic __ieee754_log | Ulrich Drepper | 2011-10-21 | 1 | -3/+3 | |
| | ||||||
* | Use new internal libc_fe* interfaces in more functions | Ulrich Drepper | 2011-10-18 | 5 | -51/+33 | |
| | ||||||
* | Provide combined internal feholdexcept/fesetround interface | Ulrich Drepper | 2011-10-18 | 1 | -5/+1 | |
| | ||||||
* | Start optimizing the use of the fenv interfaces in libm itself | Ulrich Drepper | 2011-10-18 | 2 | -31/+16 | |
| | ||||||
* | Small optimization of generic exp2 | Ulrich Drepper | 2011-10-17 | 1 | -11/+12 | |
| | ||||||
* | Add optimized 64-bit frexp | Ulrich Drepper | 2011-10-17 | 1 | -0/+67 | |
| | ||||||
* | Use rounds{s,d} for x86 rint, ceil, floor | Ulrich Drepper | 2011-10-16 | 9 | -120/+58 | |
| | ||||||
* | Optimized floor for 64-bit architectures | Ulrich Drepper | 2011-10-15 | 1 | -0/+79 | |
| | ||||||
* | Add cosh optimized for 64-bit architectures | Ulrich Drepper | 2011-10-15 | 2 | -4/+82 | |
| | ||||||
* | Optimize exp | Ulrich Drepper | 2011-10-15 | 7 | -154/+122 | |
| | | | | Add __exp*_finite optimizations and rewrite some wrappers. | |||||
* | Optimize libm | Ulrich Drepper | 2011-10-12 | 94 | -2638/+1459 | |
| | | | | | libm is now somewhat integrated with gcc's -ffinite-math-only option and lots of the wrapper functions have been optimized. | |||||
* | Implement __isinf_nsl for IBM long double | Andreas Schwab | 2011-10-10 | 1 | -0/+16 | |
| | ||||||
* | Fix some nit warnings. | Roland McGrath | 2011-10-08 | 1 | -1/+1 | |
| | ||||||
* | Optimize use of isnan, isinf, finite | Ulrich Drepper | 2011-10-08 | 11 | -1/+104 | |
| | ||||||
* | finite for 64-bit platforms | Ulrich Drepper | 2011-09-15 | 1 | -0/+32 | |
| | ||||||
* | Optimize logb code for 64-bit machines | Ulrich Drepper | 2011-09-12 | 1 | -0/+44 | |
| | ||||||
* | Fix jn precision | Petr Baudis | 2011-09-09 | 5 | -5/+50 | |
| | ||||||
* | Fix lround() loses precision | Paul Pluzhnikov | 2011-09-08 | 1 | -1/+1 | |
| | ||||||
* | Check for finite/infinity parms in IBM Long Double 128 fmal( ) | Ulrich Drepper | 2011-09-07 | 1 | -0/+39 | |
| | | | | | | | | | | | This patch addresses some IBM Long Double 128 fmal () test-ldouble.out and test-ildoubl.out failures. If the ‘x’ and ‘y’ parameters are finite values and ‘z’ is infinity, the result of fmal () should be ‘z’ not NaN. Conflicts: ChangeLog | |||||
* | Fix up __kernel_rem_pio2 for FLT_EVAL_METHOD != 0 architectures | Jakub Jelinek | 2011-08-04 | 1 | -6/+12 | |
| | ||||||
* | ldbl-128 fmal compile fix | Jakub Jelinek | 2010-11-22 | 1 | -1/+1 | |
| | ||||||
* | Fix comparison in sqrtl for IBM long double 128. | Luis Machado | 2010-11-10 | 1 | -2/+2 | |
| | ||||||
* | IEEE quad fmal fixes | Jakub Jelinek | 2010-10-18 | 1 | -2/+4 | |
| | ||||||
* | Implement fmal, some fma bugfixes | Jakub Jelinek | 2010-10-15 | 5 | -3/+467 | |
| | ||||||
* | Fix some more dbl-64/s_fma.c issue | Jakub Jelinek | 2010-10-15 | 1 | -15/+88 | |
| | ||||||
* | Implement accurate fma. | Jakub Jelinek | 2010-10-13 | 5 | -2/+268 | |
| | ||||||
* | Correct implementation of fmaf. | Jakub Jelinek | 2010-10-11 | 1 | -0/+50 | |
| | ||||||
* | Fix cproj implmentation. | Ulrich Drepper | 2010-04-08 | 1 | -17/+5 | |
| | ||||||
* | sysdeps/ieee754/ldbl-128/ fixes | Jakub Jelinek | 2010-01-16 | 5 | -15/+13 | |
| | ||||||
* | Readd definition of __expl in ldbl-128 | Joseph S. Myers | 2009-10-28 | 1 | -0/+1 | |
| | ||||||
* | Correct errno handling in expm1. | Andreas Schwab | 2009-10-19 | 3 | -0/+3 | |
| | ||||||
* | configure tweaks, support $libc_add_on_config_subdirs | Roland McGrath | 2009-09-15 | 1 | -27/+24 | |
| | ||||||
* | Add isinf optimized for 64-bit. | Ulrich Drepper | 2009-08-26 | 1 | -0/+30 | |
| | ||||||
* | Optimized isnan for 64-bit machines. | Ulrich Drepper | 2009-08-26 | 1 | -0/+39 | |
| | ||||||
* | Make llround an alias for lround on 64-bit machines. | Ulrich Drepper | 2009-08-25 | 2 | -0/+14 | |
| | ||||||
* | Optimized nearbyint for 64-bit. | Ulrich Drepper | 2009-08-25 | 1 | -0/+86 | |
| |