From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Korsgaard Date: Mon, 23 Aug 2010 07:42:51 +0200 Subject: [Buildroot] [PATCH 10/20] libglib2: add patch to rework clock_gettime() test In-Reply-To: <7dfc97b926288b402f07f49c2464cf9fd48edbdd.1282471311.git.thomas.petazzoni@free-electrons.com> (Thomas Petazzoni's message of "Sun, 22 Aug 2010 12:52:02 +0200") References: <7dfc97b926288b402f07f49c2464cf9fd48edbdd.1282471311.git.thomas.petazzoni@free-electrons.com> Message-ID: <8739u5khpg.fsf@macbook.be.48ers.dk> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net >>>>> "Thomas" == Thomas Petazzoni writes: Thomas> The test for clock_gettime() in configure.in doesn't work properly Thomas> when a previous package has loaded the shared configuration cache with Thomas> informations about the availability of clock_gettime. A package such Thomas> as ctorrent does so, which means that compiling ctorrent *then* Thomas> libglib2 currently fails. Thomas> According to people on the Autoconf mailing list, the libglib2 test is Thomas> likely the one that needs to be fixed. The problem is that the Thomas> AC_CHECK_FUNCS() test assumes that if it finds clock_gettime() it Thomas> means that there's no need to add any -lrt flag to the Thomas> build. Unfortunately, due to the shared configuration cache, this test Thomas> is already done with -lrt, so the test succeeds, and libglib2 does not Thomas> know that it needs to add -lrt to G_THREAD_LIBS and Thomas> G_THREAD_LIBS_FOR_GTHREAD. Thanks! It would be good to have a summary of the above in the patch header itself. Have you sent this upstream? -- Bye, Peter Korsgaard