Re: NaN fixes in pow, powf, modf, modff

Corinna Vinschen via Newlib <[email protected]>
Newsgroups gmane.comp.lib.newlib
Message-ID <[email protected]>
On Mar 25 17:18, Keith Packard via Newlib wrote:
> The first patch makes modf set *iptr to qnan when provided an snan
> parameter, which wasn't happening on ARM softfp because GCC elided the
> * 1.0 operation.
> 
> The second patch cleans up modf/modff by not re-converting the float
> to bits before extracting the sign bit.
> 
> The third patch makes pow return qnan instead of 1.0 when the
> parameter is snan.
> 
> These functions now match glibc for all of my tests, which were
> derived from the IEEE requirements.
> 

Pushed.


Thanks,
Corinna

-- 
Corinna Vinschen
Cygwin Maintainer
Red Hat
signature.asc (application/pgp-signature, 833 B)
-----BEGIN PGP SIGNATURE-----
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=tFLn
-----END PGP SIGNATURE-----
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.