Re: svn commit: r1582438 - /subversion/trunk/subversion/tests/svn_test_main.c

[Branko Čibej <br...@wandisco.com>]

On 27.03.2014 19:42, rhuijben@apache.org wrote:
> Author: rhuijben
> Date: Thu Mar 27 18:42:30 2014
> New Revision: 1582438
>
> URL: http://svn.apache.org/r1582438
> Log:
> * subversion/tests/svn_test_main.c
>   (thread_local): Remove __GNUC__ check in an attempt to fix compilation
>     on OpenBSD.
>
> Modified:
>     subversion/trunk/subversion/tests/svn_test_main.c
>
> Modified: subversion/trunk/subversion/tests/svn_test_main.c
> URL: http://svn.apache.org/viewvc/subversion/trunk/subversion/tests/svn_test_main.c?rev=1582438&r1=1582437&r2=1582438&view=diff
> ==============================================================================
> --- subversion/trunk/subversion/tests/svn_test_main.c (original)
> +++ subversion/trunk/subversion/tests/svn_test_main.c Thu Mar 27 18:42:30 2014
> @@ -153,8 +153,14 @@ static svn_boolean_t skip_cleanup = FALS
>  #    define thread_local _Thread_local
>  #  elif defined(WIN32) && defined(_MSC_VER)
>  #    define thread_local __declspec(thread)
> -#  elif defined(__GNUC__)
> +#  elif defined(__thread)
> +     /* ### Might work somewhere? */
>  #    define thread_local __thread
> +#  else
> +     /* gcc defines __thread in some versions, but not all.
> +        ### Who knows how to check for this?
> +        ### stackoverflow recommends __GNUC__ but that breaks on
> +        ### openbsd. */
>  #  endif
>  #endif

Now this makes me wonder why we're not just using the apr_threadkey_*
functions instead of making guesses about compiler support for
thread-local data.


-- Brane


-- 
Branko Čibej | Director of Subversion
WANdisco // Non-Stop Data
e. brane@wandisco.com