4 Dec
2006
4 Dec
'06
1:45 p.m.
Vladimir Prus wrote:
I can reproduce this with just
#include <cmath>
int main() { long double d = std::tan(-0.53682426991417537756471137222469281L); return d; }
So I'll file a bug report with Intel.
Thanks! As soon as Intel confirms this is their bug, or if you sufficiently sure it is, can you please update status/expected_failures.xml?
Will do so shortly: Intel have reproduced the issue BTW, looks like a specific glibc or kernel version issue as it doesn't show up on all machines. John.