| Commit message (Expand) | Author | Age | Files | Lines |
* | Use a better method of scaling by 2**k. Instead of adding to the | bde | 2008-02-07 | 4 | -33/+28 |
* | As for the float trig functions and logf, use a minimax polynomial | bde | 2008-02-06 | 1 | -6/+7 |
* | Adjust the exponent before converting the result from double to | das | 2008-01-28 | 1 | -16/+10 |
* | Fix a harmless type error in 1.9. | bde | 2008-01-25 | 1 | -1/+1 |
* | Fix cutoffs. This is just a cleanup and an optimization for unusual | bde | 2008-01-21 | 1 | -2/+2 |
* | Oops, when merging from the float version to the double versions, don't | bde | 2008-01-20 | 1 | -1/+1 |
* | Remove the float version of the kernel of arg reduction for pi/2, since | bde | 2008-01-19 | 1 | -198/+0 |
* | Do an ordinary assignment in STRICT_ASSIGN() except for floats until | bde | 2008-01-19 | 1 | -2/+6 |
* | Use STRICT_ASSIGN() for exp2f() and exp2() instead of a volatile | bde | 2008-01-19 | 2 | -4/+5 |
* | Use STRICT_ASSIGN() for _kernel_rem_pio2f() and _kernel_rem_pio2f() | bde | 2008-01-19 | 2 | -7/+10 |
* | Use STRICT_ASSIGN() for log1pf() and log1p() instead of a volatile cast | bde | 2008-01-19 | 2 | -8/+10 |
* | Use STRICT_ASSIGN() instead of assorted direct volatile hacks to work | bde | 2008-01-19 | 2 | -6/+8 |
* | Use volatile hacks to make sure these functions generate an underflow | das | 2008-01-18 | 3 | -3/+6 |
* | Implement exp2l(). There is one version for machines with 80-bit | das | 2008-01-18 | 2 | -0/+8 |
* | Add a macro STRICT_ASSIGN() to help avoid the compiler bug that | bde | 2008-01-17 | 1 | -0/+16 |
* | Optimize this a bit better. | das | 2008-01-15 | 1 | -13/+18 |
* | Implement rintl(), nearbyintl(), lrintl(), and llrintl(). | das | 2008-01-14 | 6 | -5/+105 |
* | - Correct the range check in the double version to catch negative values | das | 2008-01-11 | 2 | -19/+22 |
* | Grumble. DO declare logbl(), DON'T declare logl() just yet. | das | 2007-12-20 | 1 | -0/+2 |
* | Remove the declaration of logl(). The relevant bits haven't been | das | 2007-12-20 | 1 | -2/+0 |
* | Since nan() is supposed to work the same as strtod("nan(...)", NULL), | das | 2007-12-18 | 2 | -18/+67 |
* | Remove z_abs(). The z_*() functions were in libf77, and for some reason | das | 2007-12-18 | 1 | -7/+0 |
* | Add logbl(3) to libm. | das | 2007-12-17 | 2 | -0/+61 |
* | Implement and document nan(), nanf(), and nanl(). This commit | das | 2007-12-16 | 2 | -1/+71 |
* | Implement and document csqrt(3) and csqrtf(3). | das | 2007-12-15 | 2 | -0/+192 |
* | Implement carg(3) and cargf(3). | das | 2007-12-12 | 2 | -0/+76 |
* | Oops, back out previous commit since it was backwards to a wrong branch. | bde | 2007-06-14 | 1 | -1/+1 |
* | MFC: 1.11: fix the threshold for (not) using the simple Taylor approximation. | bde | 2007-06-14 | 1 | -1/+1 |
* | Fix an aliasing bug which was finally detected by gcc-4.2. fdlibm has | bde | 2007-06-11 | 1 | -1/+1 |
* | Merge the relevant part of rev.1.14 of s_cbrt.c (a micro-optimization | bde | 2007-05-29 | 1 | -3/+3 |
* | Don't assume that int is signed 32-bits in one place. Keep assuming | bde | 2007-05-02 | 2 | -8/+11 |
* | Implement modfl(). | das | 2007-01-07 | 2 | -1/+102 |
* | Correctly handle NaN. | das | 2007-01-06 | 1 | -0/+2 |
* | Correctly handle inf/nan. This routine is currently unused because we | das | 2007-01-06 | 1 | -0/+4 |
* | Fixed the threshold for using the simple Taylor approximation. | bde | 2006-07-07 | 2 | -2/+2 |
* | Fixed tanh(-0.0) on ia64 and optimizeed tanh(x) for 2**-55 <= |x| < | bde | 2006-07-05 | 1 | -10/+10 |
* | Backed out rev.1.10. It tried to implement ldexpf() as a weak reference | bde | 2006-07-05 | 1 | -2/+0 |
* | Oops, on amd64 (and probably on all non-i386 systems), the previous | bde | 2006-01-05 | 1 | -6/+11 |
* | Use double precision internally to optimize cbrtf(), and change the | bde | 2006-01-05 | 1 | -28/+13 |
* | Extract the high and low words together. With gcc-3.4 on uniformly | bde | 2005-12-20 | 1 | -8/+6 |
* | Use a minimax polynomial approximation instead of a Pade rational | bde | 2005-12-19 | 2 | -37/+30 |
* | Fixed code to match comments and the algorithm: | bde | 2005-12-18 | 2 | -20/+42 |
* | Added comments about the apparently-magic rational function used in | bde | 2005-12-15 | 1 | -1/+15 |
* | Optimize by not doing excessive conversions for handling the sign bit. | bde | 2005-12-13 | 2 | -15/+9 |
* | Fixed some especially horrible style bugs (indentation that is neither | bde | 2005-12-13 | 2 | -12/+14 |
* | Added comments about the magic behind | bde | 2005-12-11 | 2 | -14/+25 |
* | Fixed the unexpectedly large maximum error after the previous commit. | bde | 2005-12-11 | 1 | -2/+2 |
* | Fixed all 502518670 errors of more than 1 ulp for cbrtf() on amd64. | bde | 2005-12-11 | 1 | -1/+13 |
* | Fixed some magic numbers. | bde | 2005-12-11 | 1 | -8/+8 |
* | "Create" ldexpf for non-i386 architectures. | obrien | 2005-12-06 | 1 | -0/+2 |