[PATCH v6 2/3] Add default implementation of fenv.h and all methods

Jon Turney jon.turney@dronecode.org.uk
Mon Aug 12 20:37:00 GMT 2019


On 12/08/2019 19:50, Joel Sherrill wrote:
> On Mon, Aug 12, 2019 at 1:31 PM Jon Turney <jon.turney@dronecode.org.uk>
> wrote:
> 
>> On 08/08/2019 17:27, joel@rtems.org wrote:
>>> diff --git a/newlib/libm/fenv/Makefile.am b/newlib/libm/fenv/Makefile.am
>>> new file mode 100644
>>> index 0000000..fef5c36
>>> --- /dev/null
>>> +++ b/newlib/libm/fenv/Makefile.am
>> [...]
>>> +
>>> +CHEWOUT_FILES =      feclearexcept.def fe_dfl_env.def fegetenv.def \
>>> +     fegetexceptflag.def fegetround.def feholdexcept.def \
>>> +     feraiseexcept.def fesetenv.def fesetexceptflag.def fesetround.def \
>>> +     fetestexcept.def feupdateenv.def
>>> +
>>> +CHAPTERS =
>> [...]
>>
>> I'm not sure it totally makes sense to write this.
>>
>> This defines a bunch of source files which are chewed on with makedoc to
>> produce some texinfo fragments, but because there is no chapter which
>> includes them (not sure if that should be a new fenv.tex included by
>> libm.tex, or one of the existing chapters), the nodes defined by those
>> fragments are not navigable to from the libm root node (and I think

I think I should have written 'the nodes defined by those fragments are 
not included in the generated libm.info file'

>> likewise in the html documentation)
>>
> 
> Using complex.h as precedence, I would lean to making an fenv.tex file.
> 
> Do you mind adding it or do you want me to take a stab and it and email it
> to you privately to test?

Please go ahead.

I don't mind testing it, but it will just involve running 'make info' or 
'make html' and reading the generated files :)



More information about the Newlib mailing list