ashinn
|
b16a214fee
|
sexp_bignum_sqrt is only needed if SEXP_USE_MATH is enabled.
|
2013-03-23 21:16:21 +01:00 |
|
Alex Shinn
|
7dfad8a293
|
Don't gc release until after normalizing ratios.
Fixes issue #171.
|
2013-01-23 22:07:33 +09:00 |
|
Alex Shinn
|
2f56d37f0c
|
Conservatively reducing an initial estimate.
|
2013-01-05 23:54:27 +09:00 |
|
Alex Shinn
|
c559ae11be
|
Fixing fixnum results with remainders in bignum sqrt.
|
2013-01-05 23:38:20 +09:00 |
|
Alex Shinn
|
b42379539d
|
Fixing termination condition in certain bignum sqrts, handling negative inputs.
|
2013-01-03 23:49:31 +09:00 |
|
Alex Shinn
|
1012f7e129
|
A followup patch for bignum addition carrying from Lorenzo.
|
2013-01-03 23:02:59 +09:00 |
|
Alex Shinn
|
c0b257db25
|
Fixing overflow checking in certain cases in bignum addition.
(patch from Lorenzo)
|
2012-12-31 10:51:42 +09:00 |
|
Alex Shinn
|
7ae254fc28
|
More accurate square roots for bignums - compute via iteration rather
than approximation via flonums for very large bignums.
|
2012-12-22 18:06:22 +09:00 |
|
Alex Shinn
|
d9f7c07f8e
|
It's not sufficient to call sexp_negate on a copy of complex numbers,
since the negation can mutate the components. We need to make a deep copy.
|
2012-12-12 10:55:02 +09:00 |
|
Alex Shinn
|
122d8b8a00
|
Properly handling +nan.0 and infinite comparisons.
|
2012-11-10 20:53:16 +09:00 |
|
Alex Shinn
|
ab42ea2cfb
|
Coerce flonums to bignums instead of vice versa in comparisons.
Need to do the same for flonums <=> ratios but the conversion is
a hack right now.
|
2012-11-07 21:40:53 +09:00 |
|
Alex Shinn
|
93ba0926a7
|
Build fixes for plan9. Moving individual object files out of opt/ since plan9 mkfiles have issues with these.
|
2012-08-16 16:04:02 +09:00 |
|