| commit | 39c2f59709f454f04cc13151301ca19c4ba9c152 | |
| author | Peter Klausler <35819229+klausler@users.noreply.github.com> | |
| Tue, 26 Dec 2023 23:28:36 +0000 (26 15:28 -0800) | ||
| committer | GitHub <noreply@github.com> | |
| Tue, 26 Dec 2023 23:28:36 +0000 (26 15:28 -0800) | ||
| tree | 0d25402ab3b412f1239865615768747b831e1e9e | treesnapshot (tar.gz zip) |
| parent | 933882f73971546d529ab225cb9bb982ed5ff47b | commitdiff |
[flang][runtime] Fix NEAREST() when exponent decreases (#75368)
When the result of NEAREST() has an exponent less than that of the
argument (e.g., NEAREST(1.,-1.) and NEAREST(-1.,1.)), the result was
wrong, because the increment value uses the result of SPACING() in terms
of the argument. Fix by just calling into the C runtime routine
std::nextafter().
When the result of NEAREST() has an exponent less than that of the
argument (e.g., NEAREST(1.,-1.) and NEAREST(-1.,1.)), the result was
wrong, because the increment value uses the result of SPACING() in terms
of the argument. Fix by just calling into the C runtime routine
std::nextafter().
| flang/runtime/numeric.cpp | diffblobblamehistory | |
| flang/unittests/Runtime/Numeric.cpp | diffblobblamehistory |