This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCHv3 1/4] ldbl-128: Rename 'long double' to '_Float128'


On Wed, 24 Aug 2016, Paul E. Murphy wrote:

> 	* sysdeps/ieee754/ldbl-128/e_acoshl.c: Rename long double
> 	_Float128 and ensure math_private.h gets added as needed.

I think the bit about math_private.h should only be mentioned in ChangeLog 
entries for the affected files, i.e. s_signbitl.c and t_sincosl.c.

> diff --git a/sysdeps/ieee754/ldbl-128/s_copysignl.c b/sysdeps/ieee754/ldbl-128/s_copysignl.c
> index 19b6880..358ff5a 100644
> --- a/sysdeps/ieee754/ldbl-128/s_copysignl.c
> +++ b/sysdeps/ieee754/ldbl-128/s_copysignl.c
> @@ -18,7 +18,7 @@ static char rcsid[] = "$NetBSD: $";
>  #endif
>  
>  /*
> - * copysignl(long double x, long double y)
> + * copysignl(_Float128 x, _Float128 y)

Mostly in this patch you aren't changing comments.  For consistency I 
think you shouldn't change this comment either.

> diff --git a/sysdeps/ieee754/ldbl-128/s_scalblnl.c b/sysdeps/ieee754/ldbl-128/s_scalblnl.c
> index 1e24197..1060cfe 100644
> --- a/sysdeps/ieee754/ldbl-128/s_scalblnl.c
> +++ b/sysdeps/ieee754/ldbl-128/s_scalblnl.c
> @@ -19,7 +19,7 @@ static char rcsid[] = "$NetBSD: $";
>  #endif
>  
>  /*
> - * scalblnl (long double x, long int n)
> + * scalblnl (_Float128 x, long int n)

Or this one, similarly.

This patch is OK with those changes.  To be clear on files with long 
double references in code that are not changed in this patch: you're 
correctly not changing s_fma.c, s_nexttoward.c, s_nexttowardf.c (function 
implementations for other types).  For long double references in ieee754.h 
(an installed header), ldbl2mpn.c, math_ldbl.h, mpn2ldbl.c, 
strtod_nan_ldouble.h, strtold_l.c you'll need to come up with solutions in 
subsequent patches; likewise for printf_fphex.c (not needed for printf but 
might be relevant for strfromf128).  w_expl.c is being deferred because 
different wrapper implementations will be needed anyway at some point to 
avoid the obsolescent matherr interface becoming part of the interface to 
new *f128 functions.

-- 
Joseph S. Myers
joseph@codesourcery.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]