diff options
author | Rich Felker <dalias@aerifal.cx> | 2020-02-06 11:34:54 -0500 |
---|---|---|
committer | Rich Felker <dalias@aerifal.cx> | 2020-02-06 13:18:00 -0500 |
commit | 1c9afd69051a64cf085c6fb3674a444ff9a43857 (patch) | |
tree | cf65884fe30ea7acb77644e07ce41997067e29bf /src/math/i386/atan2f.s | |
parent | b4712ba445a5cb589d1ac37785c29164cd3cf1f9 (diff) | |
download | musl-1c9afd69051a64cf085c6fb3674a444ff9a43857.tar.gz |
fix excess precision in return value of i386 atan[2][f]
for functions implemented in C, this is a requirement of C11 (F.6);
strictly speaking that text does not apply to standard library
functions, but it seems to be intended to apply to them, and C2x is
expected to make it a requirement.
failure to drop excess precision is particularly bad for inverse trig
functions, where a value with excess precision can be outside the
range of the function (entire range, or range for a particular
subdomain), breaking reasonable invariants a caller may expect.
Diffstat (limited to 'src/math/i386/atan2f.s')
-rw-r--r-- | src/math/i386/atan2f.s | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/src/math/i386/atan2f.s b/src/math/i386/atan2f.s index c9408a90..3908c86d 100644 --- a/src/math/i386/atan2f.s +++ b/src/math/i386/atan2f.s @@ -4,7 +4,8 @@ atan2f: flds 4(%esp) flds 8(%esp) fpatan - fsts 4(%esp) + fstps 4(%esp) + flds 4(%esp) mov 4(%esp),%eax add %eax,%eax cmp $0x01000000,%eax |