This is the mail archive of the
newlib@sourceware.org
mailing list for the newlib project.
Re: [PATCH] Don't do double divide in powf.
On 12/12/2017 08:06 AM, Craig Howland wrote:
No to either of the above. The 0/0 construct is to cause a floating
point "invalid" exception, which the functions do not do. (The divide
has to be done at runtime to cause the exception.) The double divide
can be made float by changing it to 0.0F/0.0F.
After double checking the ISO C standard Annex F, I have to agree with you.
However, it looks like there are other places that already make this
mistake. wf_log.c for instance calls nan, and the ISO C Annex F says
this case should generate an invalid exception too. There may also be
other places that are broken in the same way. This is no excuse for me
to make it worse though, so I will post a new patch.
Jim