diff options
Diffstat (limited to 'debian/patches/0009-Skip-testing-nanmin-on-float64-arrays-on-mips-el.patch')
-rw-r--r-- | debian/patches/0009-Skip-testing-nanmin-on-float64-arrays-on-mips-el.patch | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/debian/patches/0009-Skip-testing-nanmin-on-float64-arrays-on-mips-el.patch b/debian/patches/0009-Skip-testing-nanmin-on-float64-arrays-on-mips-el.patch index 27a396a..f84e01a 100644 --- a/debian/patches/0009-Skip-testing-nanmin-on-float64-arrays-on-mips-el.patch +++ b/debian/patches/0009-Skip-testing-nanmin-on-float64-arrays-on-mips-el.patch @@ -2,8 +2,11 @@ From: Stefano Rivera <stefanor@debian.org> Date: Fri, 3 Feb 2023 15:34:10 -0400 Subject: Skip testing nanmin() on float64 arrays on mips*el -This fails on mipsel and mips64el, due to a numpy bug: -Bug-numpy: https://github.com/numpy/numpy/issues/23158 +There was an ABI break in NAN handling on MIPS that hasn't been fully +dealt with, yet. See: +https://sourceware.org/binutils/docs/as/MIPS-NaN-Encodings.html + +Discussion in the numpy bug: https://github.com/numpy/numpy/issues/23158 --- src/silx/math/test/test_combo.py | 7 +++++++ 1 file changed, 7 insertions(+) |