From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vicente Olivert Riera Date: Wed, 9 Mar 2016 14:00:19 +0000 Subject: [Buildroot] [PATCH] rtptools: disable for MIPS Codescape toolchains In-Reply-To: <20160308171149.7dff6ff4@free-electrons.com> References: <1457446983-17165-1-git-send-email-Vincent.Riera@imgtec.com> <20160308163256.764be70e@free-electrons.com> <56DEF7EC.20904@imgtec.com> <20160308171149.7dff6ff4@free-electrons.com> Message-ID: <56E02C73.40105@imgtec.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Hello Thomas, On 08/03/16 16:11, Thomas Petazzoni wrote: > Hello, > > On Tue, 8 Mar 2016 16:03:56 +0000, Vicente Olivert Riera wrote: > >>> Isn't the problem that the Codescape toolchain don't provide RPC >>> support at all ? >> >> Yes, that's the problem. > > Ah, OK. That's a different problem indeed. > >>> I don't really like to add random toolchain exclusions without a more >>> serious explanation than "header file is not in toolchain". >> >> Ok. Perhaps we should make rtptools package depeding on >> BR2_TOOLCHAIN_HAS_NATIVE_RPC? > > Well, I just built rtptools with > http://autobuild.buildroot.org/toolchains/configs/br-arm-basic.config, > which does *not* have RPC support, and it builds fine. So it is not > simply that rtptools unconditionally needs RPC support. Looking at the > code, it does: > > #ifndef __UCLIBC__ > #define HAVE_YP > #endif > > #ifdef HAVE_YP > #include /* YP */ > #endif > > > So it seems like it has special handling for uClibc. I think you should > replace this with an AC_CHECK_HEADERS test for rpcsvc/ypclnt.h, and use > the HAVE_ variable that will be defined by this test. the thing is that rtptools already includes its own ypclnt.h under the "nt/include/rpcsvc" directory, so we could replace the code you wrote above with something like this: #ifndef __UCLIBC__ #include "nt/include/rpcsvc/ypclnt.h" #endif CC'ing Peter as he wrote the patch which adds that uClibc stuff to host2ip.c. Regards, Vincent. > Best regards, > > Thomas >