This adds a backport patch that fixes a problem in subinterpreters related
to the garbagecollector. Without the patch, there are random segfaults in
several Kodi addons that use python3-sqlite3. Presumably there are real world
issues in other programs as well.
(From OE-Core rev: 39d57ef7e341a048b94bc5dd9c29d58f57e06f19)
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There are two issues inside importlib. Firstly, the modules are accessed in
on disk order. This means behaviour seen on one system might not reproduce
on another and is a real headache.
Secondly, empty directories left behind by previous modules might be looked
at. This has caused a long string of different issues for us.
As a result, patch this to a behaviour which works for us. Upstream discussion
can follow later, this is breaking builds for too many people to leave unpatched.
[YOCTO #14816]
(From OE-Core rev: cdb6879b49d4f6dc2bea8ff064a5b1e62db34781)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit e5944a38db513e033c3a3e9313267055f7254be7)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In the native sysroot we should never have paths to the python3-native
build directory. These may or may not exist at the time some dependency
is building and nothing should rely upon them.
I suspect nothing is relying on this at the moment but clean up
just to be sure.
The various config copies are adjusted to be modified consistently as some
copies were and some were not. The Makefile has the "bad" ${B} paths
replaced with a dummy placeholder too.
(From OE-Core rev: 2f93d30fee0698e3cf13be17f1691e85d2122b0e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit ae9e6249ded8fc063d6333231c391cfa2d594567)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
traceback.cpython-310.pyc is non-deterministic due to 'frozenset'
being written without strict ordering. For now let's just not
install the problematic file.
(From OE-Core rev: 00b62965bd4c9eb6fd8c016f83126da2a107a5c5)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 4b1f0f7542abcb8606688c974695a6c8a142e7a2)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
CVE-2015-20107 describes an arbitrary command execution in the mailcap
module, but this is by design in mailcap and needs to be worked around
by the calling application.
Upstream Python will be documenting this flaw in the library reference,
and it is likely that the mailcap module will be deprecated and removed
in the future.
(From OE-Core rev: f525745af38b0e5ea26693849cd4f19c627efd46)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(cherry picked from commit 85fac8408baf92d8b71946f5bfea92952b7eab01)
Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This was a long standing problem seen on aarch64 build hosts when
compiling python3 with clang cross compiler. The issue is not seen with
gcc because native glibc headers are still compatible with gcc cross compiler
(From OE-Core rev: 407744b00d702e3133304e1b43064a5634ca02cf)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Cc: Ross Burton <ross.burton@arm.com>
Cc: Jon Mason <jdmason@kudzu.us>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
