All of lore.kernel.org
 help / color / mirror / Atom feed
From: Johannes Sixt <j6t@kdbg.org>
To: Johannes Schindelin <johannes.schindelin@gmx.de>,
	Junio C Hamano <gitster@pobox.com>
Cc: git@vger.kernel.org, "Jeff King" <peff@peff.net>,
	"Torsten Bögershausen" <tboegi@web.de>
Subject: Re: [PATCH v2] Mark win32's pthread_exit() as NORETURN
Date: Tue, 1 Mar 2016 20:34:58 +0100	[thread overview]
Message-ID: <56D5EEE2.3080100@kdbg.org> (raw)
In-Reply-To: <d584d8bdaa8645a406c96f2a11f04febf57b2c25.1456841593.git.johannes.schindelin@gmx.de>

Am 01.03.2016 um 15:13 schrieb Johannes Schindelin:
> The pthread_exit() function is not expected to return. Ever. On Windows,
> we call ExitThread() whose documentation claims: "This function does not
> return a value.":
>
> 	https://msdn.microsoft.com/en-us/library/windows/desktop/ms682659

This is misleading: MSDN marks all functions declared void as "does not 
return a value," for example, look at EnterCriticalSection:

https://msdn.microsoft.com/en-us/library/windows/desktop/ms682608

For this reason, I actually prefer your version 1 patch without the 
explanation.

>
> Pointed out by Jeff King.
>
> Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
> ---
>
> 	Relative to v1, only the commit message changed (to clarify that
> 	ExitThread() indeed never returns).
>
>   compat/win32/pthread.h | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/compat/win32/pthread.h b/compat/win32/pthread.h
> index 20b35a2..148db60 100644
> --- a/compat/win32/pthread.h
> +++ b/compat/win32/pthread.h
> @@ -78,7 +78,7 @@ extern int win32_pthread_join(pthread_t *thread, void **value_ptr);
>   #define pthread_equal(t1, t2) ((t1).tid == (t2).tid)
>   extern pthread_t pthread_self(void);
>
> -static inline int pthread_exit(void *ret)
> +static inline int NORETURN pthread_exit(void *ret)

I would have written it as

#ifdef __GNUC__
__attribute__((__noreturn__))
#endif
static inline int pthread_exit(void *ret) ...

but I can live with your version as long as it compiles.

Your solution is pragmatic: NORETURN is defined in git-compat-util.h, 
and by using it here, we depend on that pthread.h is included 
sufficiently late that the macro is available at this point. The 
instance in compat/nedmalloc/malloc.c.h is bracketed with #ifndef WIN32 
so that it is not compiled on Windows, all other instances are after 
git-compat-util.h or cache.h or in headers that are to be included only 
after git-compat-util.h or cache.h per convention. Looks like we are safe.

>   {
>   	ExitThread((DWORD)(intptr_t)ret);
>   }
>

  parent reply	other threads:[~2016-03-01 19:35 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-01 13:53 [PATCH] Mark win32's pthread_exit() as NORETURN Johannes Schindelin
2016-03-01 13:57 ` Jeff King
2016-03-01 14:11   ` Johannes Schindelin
2016-03-01 14:13 ` [PATCH v2] " Johannes Schindelin
2016-03-01 14:38   ` stefan.naewe
2016-03-01 17:57     ` Junio C Hamano
2016-03-02 19:11     ` Johannes Schindelin
2016-03-01 19:34   ` Johannes Sixt [this message]
2016-03-01 19:44     ` Junio C Hamano
2016-03-02 19:10       ` Johannes Schindelin
2016-03-02 19:11   ` [PATCH v3] " Johannes Schindelin
2016-03-02 20:33     ` Junio C Hamano

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=56D5EEE2.3080100@kdbg.org \
    --to=j6t@kdbg.org \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=johannes.schindelin@gmx.de \
    --cc=peff@peff.net \
    --cc=tboegi@web.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.