From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <4A5222E9.2070105@domain.hid> Date: Mon, 06 Jul 2009 18:14:33 +0200 From: Gilles Chanteperdrix MIME-Version: 1.0 References: <4A4DE69D.1010105@domain.hid> In-Reply-To: <4A4DE69D.1010105@domain.hid> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: [Xenomai-core] Commit 6346c046b1 List-Id: Xenomai life and development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Jan Kiszka Cc: xenomai-core Jan Kiszka wrote: > Hi Philippe, > > what problem did you had to address with commit 6346c046b1? I'm asking > as it most probably breaks what e7d889f56c tried to fix (e.g. double > definitions of PTHREAD_PRIO_* with recent glibcs). The problem was that with uclibc, pthread_mutexattr_set_prio_inherit existed, but PTHREAD_PRIO_INHERIT did not exit, so the posix mutex-torture test was not compiling. Which is why, we replaced with: #ifndef PTHREAD_PRIO_INHERIT #define PTHREAD_PRIO_INHERIT #endif which should work as long as this piece of code is put after the #include_next But maybe we forgot that... Besides, I use recent libc (and I believe with -Wall -W) and do not see any warning . Would you be compiling with g++ or with special flags? -- Gilles