Source: h5py Version: 3.14.0-2 Severity: serious Tags: ftbfs patch User: [email protected] Usertags: ieee-long-double
Hi Maintainer We are working on https://wiki.debian.org/ToolChain/IEEELongDouble Please apply the attached patch, originally from Fedora, and applied in Ubuntu. It drops a workaround for float128 that is no longer needed. Regards Graham
Description: fix build after migration to IEEE 128-bit long double on ppc64le Origin: fedora, https://src.fedoraproject.org/rpms/h5py/c/4578df64e4c923d1c3edf1213a209500e7caf453 Author: Dan Horák <[email protected]> Last-Update: 2022-06-15 --- a/h5py/h5t.pyx +++ b/h5py/h5t.pyx @@ -293,18 +293,7 @@ nmant = finfo.nmant maxexp = finfo.maxexp minexp = finfo.minexp - # workaround for numpy's buggy finfo on float128 on ppc64 archs - if ftype_ == np.longdouble and _IS_PPC64: - # values reported by hdf5 - nmant = 116 - maxexp = 1024 - minexp = -1022 - elif ftype_ == np.longdouble and _IS_PPC64LE: - # values reported by hdf5 - nmant = 52 - maxexp = 1024 - minexp = -1022 - elif nmant == 63 and finfo.nexp == 15: + if nmant == 63 and finfo.nexp == 15: # This is an 80-bit float, correct mantissa size nmant += 1
