On Thu, Dec 13, 2018 at 5:26 PM Richard Purdie <[email protected]> wrote: > > On Thu, 2018-12-13 at 15:13 +0100, Andrea Adami wrote: > > Add the missing triplets (not considering uclibc) for do_configure. > > > > Signed-off-by: Andrea Adami <[email protected]> > > --- > > .../python3/powerpc-musl-triplets.patch | 43 > > +++++++++++++++++++ > > meta/recipes-devtools/python/python3_3.5.6.bb | 1 + > > 2 files changed, 44 insertions(+) > > create mode 100644 meta/recipes-devtools/python/python3/powerpc- > > musl-triplets.patch > > > > diff --git a/meta/recipes-devtools/python/python3/powerpc-musl- > > triplets.patch b/meta/recipes-devtools/python/python3/powerpc-musl- > > triplets.patch > > new file mode 100644 > > index 0000000000..477d049858 > > --- /dev/null > > +++ b/meta/recipes-devtools/python/python3/powerpc-musl- > > triplets.patch > > @@ -0,0 +1,43 @@ > > +Using musl libc the task do_configure fails on powerpc*. > > + > > + checking build system type... x86_64-pc-linux-gnu > > + checking host system type... powerpc-oe-linux-musl > > + checking for python3.5... python3.5 > > + ... > > + checking for the platform triplet based on compiler > > characteristics... > > + powerpc-linux-gnu > > + configure: error: internal configure error for the platform > > triplet > > + > > +Fix this by adding the musl triplets. > > +Note: uclibc is purposedly not considered (unsupported in oe-core). > > + > > +Upstream-Status: Pending > > +Signed-off-by: Andrea Adami <[email protected]> > > + > > +--- a/configure.ac 2018-12-13 14:44:55.414309610 +0100 > > ++++ b/configure.ac 2018-12-13 13:22:37.546497547 +0100 > > +@@ -831,14 +831,22 @@ > > + or1k-linux-gnu > > + # elif defined(__powerpc__) && defined(__SPE__) > > + powerpc-linux-gnuspe > > ++# elif defined(__powerpc__) && !defined(__GLIBC__) > > ++ powerpc-linux-musl > > ++# elif defined(__powerpc__) > > ++ powerpc-linux-gnu > > ++# elif defined(__powerpc64__) && !defined(__GLIBC__) > > ++# if defined(__LITTLE_ENDIAN__) > > ++ powerpc64le-linux-musl > > ++# else > > ++ powerpc64-linux-musl > > ++# endif > > + # elif defined(__powerpc64__) > > + # if defined(__LITTLE_ENDIAN__) > > + powerpc64le-linux-gnu > > + # else > > + powerpc64-linux-gnu > > + # endif > > +-# elif defined(__powerpc__) > > +- powerpc-linux-gnu > > + # elif defined(__s390x__) > > + s390x-linux-gnu > > + # elif defined(__s390__) > > diff --git a/meta/recipes-devtools/python/python3_3.5.6.bb > > b/meta/recipes-devtools/python/python3_3.5.6.bb > > index cd7991e7f8..2e8ae711d5 100644 > > --- a/meta/recipes-devtools/python/python3_3.5.6.bb > > +++ b/meta/recipes-devtools/python/python3_3.5.6.bb > > @@ -44,6 +44,7 @@ SRC_URI += "\ > > file://0005-bpo-30714-ALPN-changes-for-OpenSSL-1.1.0f- > > 2305.patch \ > > file://run-ptest \ > > " > > +SRC_URI_append_libc-musl = " file://powerpc-musl-triplets.patch" > > > > inherit multilib_header python3native update-alternatives qemu ptest > > Do we need to make the patch application musl only? This tends to make > it more painful for recipe upgrading and we should only do that if the > patch would have an adverse affect on non-musl. Would also be good to > get tweaks like this submitted upstream... > > Cheers, > > Richard >
If you prefert I add it unconditionally. This error with musl is not seen for other archs, I could not test the ppc64 but imagined it would suffer as well. Does your autobuilder succeed? Cheers Andrea -- _______________________________________________ Openembedded-core mailing list [email protected] http://lists.openembedded.org/mailman/listinfo/openembedded-core
