Recently, the atanh implementation was changed, but the man seems not changed yet

Recently, the atanh implementation was changed, but the man seems not changed yet

[KOSAKI Motohiro]

Hi Michael,

I recently found little mismatch between the impmentation and the man in
atanh function.

current man page says
http://www.kernel.org/doc/man-pages/online/pages/man3/atanh.3.html

---------------------------------------------------------
BUGS
       For a pole error, errno is set to EDOM; POSIX.1 says it should be set to
       ERANGE.
---------------------------------------------------------

but following test programs "34: Numerical result out of range".
It mean atanh return ERANGE rightly.


atanh_test.c
---------------------------------------------------------
#include <math.h>
#include <stdio.h>
#include <sys/errno.h>
#include <string.h>

int main(void)
{
        double d;

        d = atanh(1.0);
        if (d == HUGE_VAL)
                printf("huge %d: %s\n", errno, strerror(errno));
        else
                printf("Oops no error\n");

        return 0;
}
---------------------------------------------------------


I dug in glibc's ChangeLog and I found following change.
So, can we consider to change the man too?


---------------------------------------------------------
2009-04-26  Ulrich Drepper  <drepper-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>

        * sysdeps/ieee754/dbl-64/s_expm1.c: Set errno for overflow.
        * sysdeps/ieee754/flt-32/s_expm1f.c: Likewise.
        * sysdeps/x86_64/fpu/s_expm1l.S: Likewise.

        * sysdeps/ieee754/k_standard.c (__kernel_standard): Pole errors in
        atanh should set ERANGE.

--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Recently, the atanh implementation was changed, but the man seems not changed yet

[Michael Kerrisk]

Hello Motohiro,

On Thu, Feb 18, 2010 at 5:19 AM, KOSAKI Motohiro
<kosaki.motohiro-+CUm20s59erQFUHtdCDX3A@public.gmane.org> wrote:
> Hi Michael,
>
> I recently found little mismatch between the impmentation and the man in
> atanh function.
>
> current man page says
> http://www.kernel.org/doc/man-pages/online/pages/man3/atanh.3.html
>
> ---------------------------------------------------------
> BUGS
>       For a pole error, errno is set to EDOM; POSIX.1 says it should be set to
>       ERANGE.
> ---------------------------------------------------------
>
> but following test programs "34: Numerical result out of range".
> It mean atanh return ERANGE rightly.
>
>
> atanh_test.c
> ---------------------------------------------------------
> #include <math.h>
> #include <stdio.h>
> #include <sys/errno.h>
> #include <string.h>
>
> int main(void)
> {
>        double d;
>
>        d = atanh(1.0);
>        if (d == HUGE_VAL)
>                printf("huge %d: %s\n", errno, strerror(errno));
>        else
>                printf("Oops no error\n");
>
>        return 0;
> }
> ---------------------------------------------------------
>
>
> I dug in glibc's ChangeLog and I found following change.
> So, can we consider to change the man too?
>
>
> ---------------------------------------------------------
> 2009-04-26  Ulrich Drepper  <drepper-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
>
>        * sysdeps/ieee754/dbl-64/s_expm1.c: Set errno for overflow.
>        * sysdeps/ieee754/flt-32/s_expm1f.c: Likewise.
>        * sysdeps/x86_64/fpu/s_expm1l.S: Likewise.
>
>        * sysdeps/ieee754/k_standard.c (__kernel_standard): Pole errors in
>        atanh should set ERANGE.

Thanks for your detailed report. I applied the following change for
man-pages-2.37.

Cheers,

Michael


--- a/man3/atanh.3
+++ b/man3/atanh.3
@@ -129,16 +129,16 @@ The variant returning
 also conforms to
 SVr4, 4.3BSD, C89.
 .SH BUGS
-For a pole error,
-.\"
-.\" FIXME . this is as at glibc 2.8; check later if this bug is fixed
+In glibc 2.9 and earlier,
+.\" Bug: http://sources.redhat.com/bugzilla/show_bug.cgi?id=6759
 .\" This can be seen in sysdeps/ieee754/k_standard.c
-.\" Bug raised: http://sources.redhat.com/bugzilla/show_bug.cgi?id=6759
+when a pole error occurs,
 .I errno
-is set to
-.BR EDOM ;
-POSIX.1 says it should be set to
+as set to
+.BR EDOM
+instead of the POSIX-mandated
 .BR ERANGE .
+Since version 2.10, glibc does the right thing.
 .SH "SEE ALSO"
 .BR acosh (3),
 .BR asinh (3),




-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Author of "The Linux Programming Interface"; http://man7.org/tlpi/
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Recently, the atanh implementation was changed, but the man seems not changed yet

[KOSAKI Motohiro]

> > ---------------------------------------------------------
> > 2009-04-26  Ulrich Drepper  <drepper-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> >
> >        * sysdeps/ieee754/dbl-64/s_expm1.c: Set errno for overflow.
> >        * sysdeps/ieee754/flt-32/s_expm1f.c: Likewise.
> >        * sysdeps/x86_64/fpu/s_expm1l.S: Likewise.
> >
> >        * sysdeps/ieee754/k_standard.c (__kernel_standard): Pole errors in
> >        atanh should set ERANGE.
> 
> Thanks for your detailed report. I applied the following change for
> man-pages-2.37.
>
> --- a/man3/atanh.3
> +++ b/man3/atanh.3
> @@ -129,16 +129,16 @@ The variant returning
>  also conforms to
>  SVr4, 4.3BSD, C89.
>  .SH BUGS
> -For a pole error,
> -.\"
> -.\" FIXME . this is as at glibc 2.8; check later if this bug is fixed
> +In glibc 2.9 and earlier,
> +.\" Bug: http://sources.redhat.com/bugzilla/show_bug.cgi?id=6759
>  .\" This can be seen in sysdeps/ieee754/k_standard.c
> -.\" Bug raised: http://sources.redhat.com/bugzilla/show_bug.cgi?id=6759
> +when a pole error occurs,
>  .I errno
> -is set to
> -.BR EDOM ;
> -POSIX.1 says it should be set to
> +as set to
> +.BR EDOM
> +instead of the POSIX-mandated
>  .BR ERANGE .
> +Since version 2.10, glibc does the right thing.
>  .SH "SEE ALSO"
>  .BR acosh (3),
>  .BR asinh (3),

Great! This looks goo to me :)

Thanks.




--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html