[hurd,commited] htl: Add missing fork.h

  2b47727c68b6 ("posix: Consolidate register-atfork") introduced a fork.h
header to declare the atfork unregister hook, but was missing adding it
for htl.

This fixes tst-atfork2.
---
 sysdeps/htl/fork.h | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)
 create mode 100644 sysdeps/htl/fork.h
  

On 23/03/2021 21:20, Samuel Thibault wrote:
> 2b47727c68b6 ("posix: Consolidate register-atfork") introduced a fork.h
> header to declare the atfork unregister hook, but was missing adding it
> for htl.
> 
> This fixes tst-atfork2.

Both lowlevellock.h and register-atfork.h are common headers,
I think it would be better to just add them to generic fork.h header.

> ---
>   sysdeps/htl/fork.h | 20 ++++++++++++++++++++
>   1 file changed, 20 insertions(+)
>   create mode 100644 sysdeps/htl/fork.h
> 
> diff --git a/sysdeps/htl/fork.h b/sysdeps/htl/fork.h
> new file mode 100644
> index 0000000000..9d0d1d2b41
> --- /dev/null
> +++ b/sysdeps/htl/fork.h
> @@ -0,0 +1,20 @@
> +/* Copyright (C) 2002-2021 Free Software Foundation, Inc.
> +   This file is part of the GNU C Library.
> +   Contributed by Ulrich Drepper <drepper@redhat.com>, 2002.
> +
> +   The GNU C Library is free software; you can redistribute it and/or
> +   modify it under the terms of the GNU Lesser General Public
> +   License as published by the Free Software Foundation; either
> +   version 2.1 of the License, or (at your option) any later version.
> +
> +   The GNU C Library is distributed in the hope that it will be useful,
> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> +   Lesser General Public License for more details.
> +
> +   You should have received a copy of the GNU Lesser General Public
> +   License along with the GNU C Library; if not, see
> +   <https://www.gnu.org/licenses/>.  */
> +
> +#include <lowlevellock.h>
> +#include <register-atfork.h>
>
  

Adhemerval Zanella, le mer. 24 mars 2021 09:29:44 -0300, a ecrit:
> On 23/03/2021 21:20, Samuel Thibault wrote:
> > 2b47727c68b6 ("posix: Consolidate register-atfork") introduced a fork.h
> > header to declare the atfork unregister hook, but was missing adding it
> > for htl.
> > 
> > This fixes tst-atfork2.
> 
> Both lowlevellock.h and register-atfork.h are common headers,
> I think it would be better to just add them to generic fork.h header.

Ok but perhaps there was a reason why you didn't do it in your commit?

Samuel
  

On 24/03/2021 09:31, Samuel Thibault wrote:
> Adhemerval Zanella, le mer. 24 mars 2021 09:29:44 -0300, a ecrit:
>> On 23/03/2021 21:20, Samuel Thibault wrote:
>>> 2b47727c68b6 ("posix: Consolidate register-atfork") introduced a fork.h
>>> header to declare the atfork unregister hook, but was missing adding it
>>> for htl.
>>>
>>> This fixes tst-atfork2.
>>
>> Both lowlevellock.h and register-atfork.h are common headers,
>> I think it would be better to just add them to generic fork.h header.
> 
> Ok but perhaps there was a reason why you didn't do it in your commit?

I haven't touched the generic fork.h because I haven't see build
issues on i686-gnu.  Why exactly tst-atfork2 fails without
this includes?
  

Adhemerval Zanella, le mer. 24 mars 2021 09:40:36 -0300, a ecrit:
> Why exactly tst-atfork2 fails without this includes?

It misses the atfork unregister hook called at module unload.

Samuel
  

On 24/03/2021 09:45, Samuel Thibault wrote:
> Adhemerval Zanella, le mer. 24 mars 2021 09:40:36 -0300, a ecrit:
>> Why exactly tst-atfork2 fails without this includes?
> 
> It misses the atfork unregister hook called at module unload.

Ok, but *why* exactly? The sysdeps/htl/pt-atfork.c already
includes register-atfork.h after the fork.h inclusion.  What
the missing includes misses to define on hurd implementation?
  

Adhemerval Zanella, le mer. 24 mars 2021 09:56:47 -0300, a ecrit:
> On 24/03/2021 09:45, Samuel Thibault wrote:
> > Adhemerval Zanella, le mer. 24 mars 2021 09:40:36 -0300, a ecrit:
> > > Why exactly tst-atfork2 fails without this includes?
> > 
> > It misses the atfork unregister hook called at module unload.
> 
> Ok, but *why* exactly? The sysdeps/htl/pt-atfork.c already
> includes register-atfork.h after the fork.h inclusion.

It's not pt-atfork.c which needs UNREGISTER_ATFORK, but
stdlib/cxa_finalize, which gets it from fork.h

Samuel
  

On 24/03/2021 10:00, Samuel Thibault wrote:
> Adhemerval Zanella, le mer. 24 mars 2021 09:56:47 -0300, a ecrit:
>> On 24/03/2021 09:45, Samuel Thibault wrote:
>>> Adhemerval Zanella, le mer. 24 mars 2021 09:40:36 -0300, a ecrit:
>>>> Why exactly tst-atfork2 fails without this includes?
>>>
>>> It misses the atfork unregister hook called at module unload.
>>
>> Ok, but *why* exactly? The sysdeps/htl/pt-atfork.c already
>> includes register-atfork.h after the fork.h inclusion.
> 
> It's not pt-atfork.c which needs UNREGISTER_ATFORK, but
> stdlib/cxa_finalize, which gets it from fork.h

Thanks, this is a good indication that is better to just remove
the ifdef UNREGISTER_ATFORK from cxa_finalize (since now it is
defined on generic implementation).
  

On 24/03/2021 10:04, Adhemerval Zanella wrote:
> 
> 
> On 24/03/2021 10:00, Samuel Thibault wrote:
>> Adhemerval Zanella, le mer. 24 mars 2021 09:56:47 -0300, a ecrit:
>>> On 24/03/2021 09:45, Samuel Thibault wrote:
>>>> Adhemerval Zanella, le mer. 24 mars 2021 09:40:36 -0300, a ecrit:
>>>>> Why exactly tst-atfork2 fails without this includes?
>>>>
>>>> It misses the atfork unregister hook called at module unload.
>>>
>>> Ok, but *why* exactly? The sysdeps/htl/pt-atfork.c already
>>> includes register-atfork.h after the fork.h inclusion.
>>
>> It's not pt-atfork.c which needs UNREGISTER_ATFORK, but
>> stdlib/cxa_finalize, which gets it from fork.h
> 
> Thanks, this is a good indication that is better to just remove
> the ifdef UNREGISTER_ATFORK from cxa_finalize (since now it is
> defined on generic implementation).

And I think the fork.h header is not really required here, it is
being used solely to define the NPTL specific __fork_generation,
which Florian recent patch moves to libc.so.  So I think we can
move this declaration to another internal header and remove
fork.h.