On 11/12/2009 05:56 PM, David Cournapeau wrote: > On Thu, Nov 12, 2009 at 11:34 PM, Michael Droettboom<[email protected]> wrote: > >> I'm happy to make these changes, since I've got access to the "finicky" >> platform, but want to confirm how you would prefer it done first. >> > Cool. The changes should be done for all platforms, and pointer cast > should be done through union, not (*(new type*)(&x)). > I've committed this change in r7731. It's been tested in x86/gcc, x86_64/gcc (both by "forcing" it to use the npymath version of nextafterl) and sparc/Sun Studio. It probably could use some testing on platforms with other "long double" types.
Mike _______________________________________________ NumPy-Discussion mailing list [email protected] http://mail.scipy.org/mailman/listinfo/numpy-discussion
