The compiler-rt project intructs to build out of tree:
https://compiler-rt.llvm.org/
Otherwise, the dependency chain declared in the include/CMakelLists.txt
causes a circular dependency of source files upon themselves:
make[4]: Circular include/sanitizer/allocator_interface.h <- include/sanitizer/allocator_interface.h dependency dropped.
make[4]: Circular include/sanitizer/asan_interface.h <- include/sanitizer/asan_interface.h dependency dropped.
make[4]: Circular include/sanitizer/common_interface_defs.h <- include/sanitizer/common_interface_defs.h dependency dropped.
[--snip 19 other files--]
This is because include/CMakeLists.txt unconctional declares
dependencies in this manner:
set(SANITIZER_HEADERS
sanitizer/allocator_interface.h
sanitizer/asan_interface.h
sanitizer/common_interface_defs.h
[...]
)
set(COMPILER_RT_HEADERS
${SANITIZER_HEADERS}
[...])
set(output_dir ${COMPILER_RT_OUTPUT_DIR}/include)
foreach( f ${COMPILER_RT_HEADERS} )
set( src ${CMAKE_CURRENT_SOURCE_DIR}/${f} )
set( dst ${output_dir}/${f} )
add_custom_command(OUTPUT ${dst}
[...]
)
[...]
endforeach( f )
The froeach() loop creates dependency rules between the files in
output_dir and CMAKE_CURRENT_SOURCE_DIR, without provision for the
case they are both the same directories, thus in-source builds are
not supported.
With the Makefiles backend, this only triggers the above warning
from make, because make arbitrarily breaks circluar dependencies, and
in this case it makes it work. But when we switch to the ninja backend,
this is going to be a hard error.
Anyway, ninja or make, compiler-rt does not support in-source builds.
Signed-off-by: Thomas Devoogdt <thomas@devoogdt.com>
[yann.morin.1998@free.fr: drastically expand commit log]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Since it was iontroduced in commit d4ecefa3cc (package/perftest: new
package), perftest was missing the reverse dependency on the kernel
headers version inherited from rdma-core.
In commit 3279d8b9c4 (package/rdma-core: needs NPTL), the dependency
of rdma-core on NTPL was not propagated to its reverse dependencies,
notably perftests.
Fixes:
- http://autobuild.buildroot.net/results/00c/00ce19a1589c6028ee77c333ed9107e2c8c8d6de
- d4ecefa3cc
- 3279d8b9c4
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
[yann.morin.1998@free.fr:
- also add misisng headers dependency
- expand commit log
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Previously, gen-bootlin-toolchains did not add a `depends` guard to
limit the available toolchains based on the minimum required GCC version
for the user selected CPU tuning.
Now, the proper BR2_ARCH_NEEDS_GCC_AT_LEAST_X guard will be added based
on the version of GCC provided by the toolchain.
Signed-off-by: Vincent Fazio <vfazio@gmail.com>
[yann.morin.1998@free.fr: regenerate the toolchain list]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Previously, it was possible to select an external toolchain that did not
support the GCC arch tuning the user had selected. This is problematic
because it can lead to confusing error messages during builds [0].
Now, external toolchain selections will be filtered to only those that
support the required GCC version specified by the target arch tuning.
Note: this patch does not touch the Bootlin toolchain config file as it
is generated by a script.
Additional note: there is "soft" support for toolchains prior to GCC 4.8
but there are no accompanying BR2_ARCH_NEEDS_GCC_AT_LEAST_X symbols.
Instead of adding those, just use BR2_ARCH_NEEDS_GCC_AT_LEAST_4_8 which
is the minimum GCC version with claimed support [1].
[0]: https://lists.buildroot.org/pipermail/buildroot/2023-August/671877.html
[1]: https://buildroot.org/downloads/manual/manual.html#requirement-mandatory
Signed-off-by: Vincent Fazio <vfazio@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Fix the following build failure:
>>> toolchain-external-codescape-img-mips 2018.09-02 Copying gdbserver
Could not find gdbserver in external toolchain
Fixes:
- http://autobuild.buildroot.org/results/b0786965e0b249c8168df855682e54cfe95fa0cc
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Fix the following build failure:
>>> toolchain-external-codescape-mti-mips 2018.09-02 Copying gdbserver
Could not find gdbserver in external toolchain
Fixes:
- http://autobuild.buildroot.org/results/ead304285a9cf32bf6b9dcc7fc30acece25ec680
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Add a custom case to make sure that a random configuration with an empty
configuration file for ubi doesn't fail.
ubinize: error!: no sections found the ini-file "/home/buildroot/autobuild/instance-2/output-1/build/ubinize.cfg"
Fixes:
- http://autobuild.buildroot.org/results/f678c17cc4df06fb2737467e769cd8f72a3ea420
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Use full relative pathes for rootfs and kernel, like in other
readme's.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Urcu does not like -O0 it needs gcc optimizations.
Force -O1 in this case. Inspired by package fwts.
The issue got introduced in commit 2f8f69012e:
before this commit, xfsprogs builds fine at -O0, after this commit, it fails
to build at -O0. As this commit went into 2022.02, the fix needs to be
backported to stable branches.
Fixes:
- http://autobuild.buildroot.net/results/3bd/3bd7de784953ab1fa70bf5572eccd7b10e4a0e17
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
gcc 12 added a warning that triggers on access to low addresses. Add a
patch to allow access since this is normal for low level code.
Rebase our existing patch on top. While at it, add also a proper
Upstream tag.
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/4795673785
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Linux added mnt_id member to struct statx in commit:
fa2fcf4f1df1559a0a4ee0f46915b496cc2ebf60
Linux 5.4.0 was released on 24. November 2019, but it seems this change
never got backported to 5.4.252.
Upstream added a patch to guard the use of the member in commit:
c0136ac0c9
Add the patch to fix an autobuild failure.
Fixes:
- http://autobuild.buildroot.net/results/d50/d502bc9236b577e2470a30ffc39c21579b038a1c
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Compile of e2fsprogs is failing because of gcc bug
111001. Workaround it by forcing the usage of -Os.
Compile fails with:
CC rw_bitmaps.c
terminate called after throwing an instance of 'std::bad_alloc'
what(): std::bad_alloc
during RTL pass: sh_treg_combine2
rw_bitmaps.c: In function 'read_bitmaps_range_start':
rw_bitmaps.c:447:1: internal compiler error: Aborted
447 | }
| ^
0x5e7a6f ???
../sysdeps/aarch64/start.S:81
Please submit a full bug report, with preprocessed source (by using -freport-bug).
Please include the complete backtrace with any bug report.
See <http://bugs.buildroot.net/> for instructions.
No need to backport, because gcc 13.2.0 was introduced recently in
Buildroot commit: 3009095ba8
Fixes:
- http://autobuild.buildroot.net/results/fb7/fb7e9c8b02b979b3a75185c0afa4ab0ca654f6f6
- http://autobuild.buildroot.net/results/659/659313b1783db491653bcc85b37da40c7b077225
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
e2fsprogs package fails to build for the SH4 architecture due to a gcc
13.2.0 bug that leads to:
during RTL pass: sh_treg_combine2
rw_bitmaps.c: In function read_bitmaps_range_start:
internal compiler error: Aborted
Let's add gcc bug to avoid to deal with architectures in packages
when a gcc bug arises. Let's instead deal directly with gcc
bug number.
It's been reported upstream:
https://gcc.gnu.org/bugzilla//show_bug.cgi?id=111001
No need to backport as gcc 13.20 was introduced recently in
buildroot.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This is a new Buildroot package and this is the first build error
the autobuild servers found. It is not part of any Buildroot release
so no need to backport.
Patch removes dlfcn.h, because it is unused.
Fixes:
- http://autobuild.buildroot.net/results/db3/db301e9f556592828800fa663f2549f430adb58f
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The dependency is not provided by Buildroot, and it can potentially be
found by gdb's build system from the host system, which results in the
following build errors:
../../gdb/source-cache.c:37:10: fatal error: srchilite/sourcehighlight.h: No such file or directory
37 | #include <srchilite/sourcehighlight.h>
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
make[4]: *** [Makefile:1645: source-cache.o] Error 1
make[4]: *** Waiting for unfinished jobs....
Signed-off-by: Bartosz Bilas <b.bilas@grinn-global.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
screenfetch requires an "awk" implementation, but upstream would
only use an executable called "awk": pick up upstream patch that
can autodetect awk implementation, so that "mawk" can be used.
Signed-off-by: Nicolas Boichat <drinkcat@google.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Most of the users use HDMI console on RPi and with increasing popularity
of systemd they face a trouble.
systemd doesn't use /etc/inittab, enable getty@tty1.service instead.
Signed-off-by: Alexey Roslyakov <alexey.roslyakov@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
In Buildroot commit cc69dbf8fd
("package/libedit: bump version to 20221030-3.1"), the version of
libedit was bumped from 20210910-3.1 to 20221030-3.1. This broke the
build with uClibc:
./sys.h:96:1: error: unknown type name 'ssize_t'; did you mean 'size_t'?
Header file uses ssize_t so sys/types.h must be included.
This commit adds a patch that fixes this issue.
Fixes:
http://autobuild.buildroot.net/results/439538b23776a58ff81c38754a19ee92c3590a1f
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Add patch taken from upstream pull request ([1]) to fix gcc-13.x compile
failure because of a missing csdtint include.
Fixes:
In file included from libserial/SerialStreamBuf.h:36,
from SerialStreamBuf.cpp:34:
./libserial/SerialPortConstants.h:93:37: error: ‘uint8_t’ was not declared in this scope
93 | using DataBuffer = std::vector<uint8_t> ;
| ^~~~~~~
[1] https://github.com/crayzeewulf/libserial/pull/186
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Commit 503252d8b0 (boot/lpc32xxcdl: remove package) forgot to remove
the special handling in genrandconfig. Since the package no longer
exists, we don't need to special-case its symbol anymore, so drop it
now.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
This package has dubious licensing conditions (not even documented in
the .mk file), and is a bootloader for very old platforms. The
defconfigs making use of it have been removed in Buildroot in 2014, in
commit c6a410964b ("configs: remove
lpc32xx defconfigs"), so let's get rid of the package.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Acked-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
[yann.morin.1998@free.fr: remove reference in test]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
xlib_libXpresent is needed with x11 to avoid the following build failure
raised since bump to version 0.35.1 in commit
61ba55e9cc and
3d459832a8:
Checking for X11: no ('x11 >= 1.0.0 xscrnsaver >= 1.0.0 xext >= 1.0.0 xinerama >= 1.0.0 xpresent >= 1.0.0 xrandr >= 1.2.0' not found)
You manually enabled the feature 'x11', but the autodetection check failed.
Fixes:
- http://autobuild.buildroot.org/results/1aace6210ba966bb4979d935c08507b2a34e0460
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Needed for mpv since version 0.35.0:
3d459832a8
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
[Thomas: improved thanks to feedback from a similar patch submitted by
Bernd Kuhls.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Until now, the Python build system was building a _dbm.cpython-*.so
native module when BR2_PACKAGE_PYTHON3_BERKELEYDB was enabled, but
since the bump of Python to 3.11.x, it was no longer building this
native module.
Investigation this, we found out that Python 3.11 changed how libdb
was detected [1] and that check now fails:
quoting configure log:
checking for libdb... no
quoting python config.log:
conftest.c:(.text.startup+0x8): undefined reference to `__db_ndbm_open'
However, in fact it turns that this problem is not new in Python
3.11. In Python 3.10, the build system was always producing the native
module, but it was in fact not working at runtime:
>>> from _dbm import *
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ImportError: /usr/lib/python3.10/lib-dynload/_dbm.cpython-310-arm-linux-gnueabihf.so: undefined symbol: __db_ndbm_delete
It was not really visible because when one does "import dbm", it
actually tries several "backends", including ndbm (which relies on
_dbm above), and if it doesn't work, it falls back silently on a pure
Python implementation.
So the issue was never noticed, but has already been there,
potentially forever.
In order for this _dbm native module to be built (Python >= 3.11) or
to work (Python < 3.11), the BerkeleyDB library need to be built with
its so-called "dbm" interface, which we do by selecting select
BR2_PACKAGE_BERKELEYDB_DBM.
And now:
>>> import _dbm
>>>
[1] d2340ef257/configure.ac (L4002)
Signed-off-by: Bernd Kuhls <bernd@kuhls.net>
[Thomas: did more research to have a better explanation of what is
happening, and realize the problem is not related to Python 3.11]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Berkeley DB support in python depends on dbm, when missing python fails
to detect libdb:
d2340ef257/configure.ac (L4002)
quoting python configure log:
checking for libdb... no
quoting python config.log:
conftest.c:(.text.startup+0x8): undefined reference to `__db_ndbm_open'
This patch provides a configure option for the historic dbm interface
to be used by the python package.
Signed-off-by: Bernd Kuhls <bernd@kuhls.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
While it's not strictly necessary to update go-bootstrap-stage2, we can
continue to update the package to the latest 1.19.x until Go 1.19.x is
discontinued, at which point we can leave it at the final 1.19.x
version.
This package is only used when bootstrapping host-go.
go1.19.11 includes a security fix to the net/http package, as well as
bug fixes to cgo, the cover tool, the go command, the runtime, and the
go/printer package.
https://go.dev/doc/devel/release#go1.19.minor
Signed-off-by: Christian Stewart <christian@aperture.us>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
