Re: [glibc] Move generic nan-pseudo-number.h to ldbl-96

Re: [glibc] Move generic nan-pseudo-number.h to ldbl-96

[Florian Weimer]

* Siddhesh Poyarekar via Glibc-cvs:

> https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=fee3b889d88193eb3cebaab2bcf60e39ce84d369
>
> commit fee3b889d88193eb3cebaab2bcf60e39ce84d369
> Author: Siddhesh Poyarekar <siddhesh@sourceware.org>
> Date:   Fri Jan 1 06:27:11 2021 +0530
>
>     Move generic nan-pseudo-number.h to ldbl-96
>     
>     The concept of pseudo number formats only exists in the realm of the
>     96 bit long double format.
>
> Diff:
> ---
>  sysdeps/{generic => ieee754/ldbl-96}/nan-pseudo-number.h | 0
>  1 file changed, 0 insertions(+), 0 deletions(-)
>
> diff --git a/sysdeps/generic/nan-pseudo-number.h b/sysdeps/ieee754/ldbl-96/nan-pseudo-number.h
> similarity index 100%
> rename from sysdeps/generic/nan-pseudo-number.h
> rename to sysdeps/ieee754/ldbl-96/nan-pseudo-number.h

This does not work because libm-test-driver.c includes
<nan-pseudo-number.h> unconditionally, but ldbl-96 is only used on a
fraction of the targets.

Thanks,
Florian
-- 
Red Hat GmbH, https://de.redhat.com/ , Registered seat: Grasbrunn,
Commercial register: Amtsgericht Muenchen, HRB 153243,
Managing Directors: Charles Cachera, Brian Klemm, Laurie Krebs, Michael O'Neill

Re: [glibc] Move generic nan-pseudo-number.h to ldbl-96

[Siddhesh Poyarekar]

On 1/4/21 4:54 PM, Florian Weimer via Libc-alpha wrote:
> This does not work because libm-test-driver.c includes
> <nan-pseudo-number.h> unconditionally, but ldbl-96 is only used on a
> fraction of the targets.

... and those were the targets I had tested, i.e. m68k, x86 and ia64 :/

Sorry, I'll post a fix in a moment; I suppose TEST_COND_intel96 is a 
good filter for this too.

Siddhesh