On 27 May 2014 02:59, Hongxu Jia <hongxu....@windriver.com> wrote: > Signed-off-by: Hongxu Jia <hongxu....@windriver.com> > --- > .../recipes-connectivity/socat/socat/compile.patch | 163 > +++++++++++----------
Oh wow this patch is ... special. 99% of it is adding the extra arguments to AC_DEFINE so that recent autoheader actually works. This is essential upstream and I'm surprised they still don't have it. Remove all of these and we'll left with these two chunks: [AC_TRY_COMPILE([#include <termios.h>], -[struct termios t; t.c_ispeed=0;], +[struct termios t; t.c_ispeed=ISPEED_OFFSET; t.c_ospeed=OSPEED_OFFSET;], I can't see a reason why that is obviously wrong but if it's right, it needs to be explained in the patch header and pushed upstream. -AC_DEFINE_UNQUOTED($1_SHIFT, ${$2}) +SHIFT_NAME="$1"_SHIFT +AC_DEFINE_UNQUOTED($SHIFT_NAME, ${$2}, [foo]) This introduces a change that is subsequently removed by fix-xxx-SHIFT-autoheader.patch, which again is basically fixing the AC_DEFINE autoheader use, should be merged with the majority of compile.patch, and submitted upstream. Finally there's a do_configure_prepend that does yet more alternations to the configure script. This needs to be a patch instead of a sed. > --- a/meta/recipes-connectivity/socat/socat_1.7.2.3.bb > +++ b/meta/recipes-connectivity/socat/socat_1.7.2.4.bb > @@ -8,16 +8,18 @@ SECTION = "console/network" > DEPENDS = "openssl readline" > > LICENSE = "GPL-2.0+-with-OpenSSL-exception" > -LIC_FILES_CHKSUM = "file://COPYING;md5=18810669f13b87348459e611d31ab760 \ > - > file://README;beginline=252;endline=282;md5=79246f11a1db0b6ccec54d1fb711c01e" > +LIC_FILES_CHKSUM = "file://COPYING;md5=b234ee4d69f5fce4486a80fdaf4a4263 \ > + > file://README;beginline=257;endline=287;md5=79246f11a1db0b6ccec54d1fb711c01e" Please explain why the license checksum changed. Ross -- _______________________________________________ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core