**diff options**

author | nsz <nsz@port70.net> | 2012-06-20 23:25:58 +0200 |
---|---|---|

committer | nsz <nsz@port70.net> | 2012-06-20 23:25:58 +0200 |

commit | e5fb6820a42a1f675ba09c15273953e1ace65777 (patch) | |

tree | 19d075252fdb22352fd147e7c4333cfa015d3b21 /arch/arm | |

parent | ac4fb51dde2f1fe9371a4cb51ff309d4868245ec (diff) | |

download | musl-e5fb6820a42a1f675ba09c15273953e1ace65777.tar.gz |

math: fix fma bug on x86 (found by Bruno Haible with gnulib)

The long double adjustment was wrong:
The usual check is
mant_bits & 0x7ff == 0x400
before doing a mant_bits++ or mant_bits-- adjustment since
this is the only case when rounding an inexact ld80 into
double can go wrong. (only in nearest rounding mode)
After such a check the ++ and -- is ok (the mantissa will end
in 0x401 or 0x3ff).
fma is a bit different (we need to add 3 numbers with correct
rounding: hi_xy + lo_xy + z so we should survive two roundings
at different places without precision loss)
The adjustment in fma only checks for zero low bits
mant_bits & 0x3ff == 0
this way the adjusted value is correct when rounded to
double or *less* precision.
(this is an important piece in the fma puzzle)
Unfortunately in this case the -- is not a correct adjustment
because mant_bits might underflow so further checks are needed
and this was the source of the bug.

Diffstat (limited to 'arch/arm')

0 files changed, 0 insertions, 0 deletions