Closed edoapra closed 1 year ago
This commit avoid using dcopy() when BLAS uses 32-bit integers to avoid any issue between 64-bit and 32-bit integers. The previous code resulted in erroneous results on big endian architectures. https://github.com/nwchemgit/nwchem/issues/864#issuecomment-1770549014
dcopy()
I'm good with the changes.
This commit avoid using
dcopy()
when BLAS uses 32-bit integers to avoid any issue between 64-bit and 32-bit integers.The previous code resulted in erroneous results on big endian architectures.
https://github.com/nwchemgit/nwchem/issues/864#issuecomment-1770549014