Inspired by
c250c61cc3
"GCC 15 defaults to C23. The last release of this package was over a
decade ago, and it is no longer maintained, therefore it should not be
expected to compile to the latest standards."
Signed-off-by: Bernd Kuhls <bernd@kuhls.net>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit c1d422edde)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
This commit adds the same patch that was already added to GDB 14.x,
15.x and 16.x to fix a GCC 15.x build issue.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit dc0691f038)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
This commit backports an upstream patch fixing the bundled readline
library so that it builds with GCC 15.x.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 56c834400e)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
bash uses K&R function declarations which have been removed in C23.
Since part of the build process (like the mkbuiltins helper) is written
in C, building bash now fails on hosts with GCC 15 (which defaults to
C23).
Since properly fixing this on the source code level is a larger
endeavor, just set the C standard to an old enough version for now.
Signed-off-by: Florian Larysch <fl@n621.de>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 6d09b25d08)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
On systems running GCC 15, host-cpio will fail to build with errors like
copyout.c:646:12: error: too many arguments to function 'xstat'; expected 0, have 2
and
main.c:407:13: error: assignment to 'int (*)(void)' from incompatible pointer type 'int (*)(const char * restrict, struct stat * restrict)' [-Wincompatible-pointer-types]
This was reported[1] and fixed upstream, but there is no new release
yet. Import the upstream patch for now.
Fixes:
https://autobuild.buildroot.org/results/a10c5f2b0f9cb05b2550fe97f1133deaaac97277/
(and many more)
[1] https://lists.gnu.org/archive/html/bug-cpio/2025-05/msg00000.html
Signed-off-by: Florian Larysch <fl@n621.de>
[Julien:
- add missing "Signed-off-by:" in patch to fix check-package error
- change "Upstream:" link to use the commitdiff in patch
- add "Fixes:" in commit log
]
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit e3cae9e1ca)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
When using host gcc 15, host-gmp fails at configure time with
error message:
configure: error: could not find a working compiler, see config.log for details
The error is due to the macro GMP_PROG_CC_WORKS in the file
acinclude.m4 containing C-code declaring functions without declaring
its parameters. This construct is now an error in C23, which is the
new default in gcc 15. See:
https://gcc.gnu.org/gcc-15/porting_to.html#c23
This commit fixes the issue by adding a package patch from
upstream changesets, not yet published in a release.
Even if gcc 15 is not yet included in Buildroot, this commit also
sets _AUTORECONF = YES for the target package in order to have this
package already fixed.
Fixes:
https://autobuild.buildroot.org/results/623634fa7bbeceeb6d90b15ce0abb1b9b4b24045/
Tested-by: Brigham Campbell <me@brighamcampbell.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 31569bcc1b)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
When compiling host-m4 1.4.19 with a host gcc 15 (which is the version
included in Fedora 42, released on 2025-04-15), compilation fails with
error:
In file included from gl_avltree_oset.h:21,
from gl_avltree_oset.c:21:
gl_oset.h:275:1: warning: 'nodiscard' attribute ignored [-Wattributes]
275 | GL_OSET_INLINE _GL_ATTRIBUTE_NODISCARD int
| ^~~~~~~~~~~~~~
gl_oset.h:275:40: error: expected identifier or '(' before 'int'
275 | GL_OSET_INLINE _GL_ATTRIBUTE_NODISCARD int
| ^~~
This error is due to the gnulib copy included in m4 1.4.19, which does
not detect properly the default C language standard of gcc 15 which
has been changed from "gnu17" to "gnu23". See [1]. Note that m4 1.4.19
is the latest version available at the time of this commit, and was
released in May 2021. The issue is tracked upstream in [2].
Upcoming m4 release is expected to fix this issue, by updating its
gnulib copy. See [3], which states: "Update to comply with newer C
standards, and inherit portability improvements from gnulib".
Until this new m4 version is released, this commit fixes the issue by
forcing the C langage standard to "-std=gnu17" (the previous gcc
default) when host-gcc 15 is detected.
Note that the "-std=gnu17" option was introduced in gcc 8. See [4].
This is the reason why this patch adds this option only when the
problematic gcc 15 version is detected.
See also the discussions around this patch at [5].
Fixes:
https://autobuild.buildroot.org/results/1c33ef0a710cfae13e496485787b351c8f951217/
(and many, many others)
[1] https://gcc.gnu.org/gcc-15/changes.html#c
[2] https://savannah.gnu.org/support/?111150
[3] https://git.savannah.gnu.org/cgit/m4.git/commit/?h=branch-1.4&id=a22c9802dd7e724eaefb21dc21d84ac2d3a49c89
[4] https://gcc.gnu.org/gcc-8/changes.html#c
[5] https://lore.kernel.org/buildroot/CAPWx8vsoJUt8YMJG1aUqFRK1=yizNbgjVjGL1Q1+9ygjJGnZLA@mail.gmail.com/
Signed-off-by: Joseph Zikusooka (ZIK) <zik@jambula.net>
Tested-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
[Julien:
- change mail url to lore.kernel.org for stable link
- reword, reflow and add extra info in the commit log
- force -std=gnu17 only when host gcc-15 is detected
- add a comment in .mk to remove the workaround at next bump
]
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 7a07a9d155)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
A basic Fedora 42 container does not have 'awk' installed, but it is
needed by Buildroot. First by check-host-python3.sh:
support/dependencies/check-host-python3.sh: line 6: awk: command not found
support/dependencies/check-host-python3.sh: line 19: awk: command not found
support/dependencies/check-host-python3.sh: line 19: awk: command not found
but then even building host-expat assumes awk is available:
config.status: creating Makefile
./config.status: line 1404: awk: command not found
config.status: creating expat.pc
./config.status: line 1404: awk: command not found
Since it's a pretty basic tool, make it part of the tools checked by
dependencies.sh. One minor annoyance is that check-host-python3.sh is
executed *before* dependencies.sh does its thing, so when 'awk' is not
available, we end up seeing:
support/dependencies/check-host-python3.sh: line 6: awk: command not found
support/dependencies/check-host-python3.sh: line 19: awk: command not found
support/dependencies/check-host-python3.sh: line 19: awk: command not found
which: no awk in (/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin)
You must install 'awk' on your build machine
make: *** [support/dependencies/dependencies.mk:27: dependencies] Error 1
It would be nice to have the awk check *before* it gets used in
check-host-python3.sh, but that's a topic for another patch.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 448ceefa78)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
Changes:
Set GDB version number to 15.2.
[gdb/python] Make sure python sys.exit makes gdb exit
[gdb/symtab] Revert "Change handling of DW_TAG_enumeration_type in DWARF scanner"
[gdb/testsuite] Add regression test for PR32158
[gdb/testsuite] Add gdb.dwarf2/enum-type-c++.exp, regression test for PR31900.
gdb-15-branch: Clear the X86_XSTATE_MPX bit in XCRO for x32
Recognize -2 as a tombstone value in .debug_line
[gdb] Handle ^C during disassembly
Mark unavailable bytes of limited-length arrays when allocating contents
gdb/solib-frv: move lm_info object to solib
Fix loading a saved recording
Bump GDB's version number to 15.1.90.DATE-git.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit e40bf89e40)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
Changes:
Set GDB version number to 16.3.
gstack: [downstream regression] Output file names and line numbers
Fix build failure for gdbserver's raw_compare self test
Fix gdbserver crashes on SVE/SME-enabled systems
gdb: allow selecting default fg/bg colors in tui mode
gdb: Fix assertion failure when inline frame #0 is duplicated
[gdb/tdep] Rewrite i386_canonicalize_syscall
[gdb/record] Fix out-of-bounds write in aarch64_record_asimd_load_store
gdb/dwarf: save DWARF version in dwarf2_loclist_baton, remove it from dwarf2_per_cu
Fix segfault if target_fileio_read_alloc fails
gdb/tui: use wrefresh if output is not surpressed
[gdb/corefiles] Fix segfault in core_target_open
Bump GDB's version number to 16.2.90.DATE-git.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit b793160964)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
This patch was commited upstream, and released as part of sqlite 3.49.1
However, the configuration system changed between sqlite 3.48 and 3.49
from autotools to autosetup, and this has proven challenging to support
in Buildroot (see `git log package/sqlite`), hence why we are still on
sqlite 3.48.
Therefore, until the package build infrastructure correctly supports
building sqlite 3.49, let's simply import the upstream patch to address
the CVE.
Note: the upstream patch is on the orignal sqlite sources. Buildroot is
using the sqlite "amalgamation" source archive, which basically
concatenate all the source files in a single "sqlite3.c" file. So the
patch was reformated to apply correctly on the sqlite release archive.
Fixes:
https://www.cve.org/CVERecord?id=CVE-2025-29087
Signed-off-by: Titouan Christophe <titouan.christophe@mind.be>
[Julien:
- reformat patch to be applicable on amalgamated sqlite sources
- add comment in commit log about patch format
- add "Fixes:" in commit log
]
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 835b5659ea)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
Since its introduction in commit [1], the
friendlyarm_nanopi_r3s_defconfig is failing to build with error:
Incorrect selection of kernel headers: expected 6.12.x, got 6.13.x
The error happens because the defconfig has:
BR2_PACKAGE_HOST_LINUX_HEADERS_CUSTOM_6_12=y
and
BR2_LINUX_KERNEL_CUSTOM_VERSION_VALUE="6.13.4"
This commit fixes the issue by setting instead:
BR2_LINUX_KERNEL_CUSTOM_VERSION_VALUE="6.12.28"
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/9887216429
[1] 41734e00c0
Cc: Sergey Kuzminov <kuzminov.sergey81@gmail.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 74c4dd4150)
[Thomas:
Since this has been introduced in LTS branch I set the kernel
version to 6.12.x present in LTS instead of changing the linux header
version.
]
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
This commit adds BR2_DOWNLOAD_FORCE_CHECK_HASHES=y in the defconfig
and adds custom hash files. The exception entry in .checkpackageignore
is also removed.
Cc: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 683681261b)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
The 'bird' package with only the `BR2_PACKAGE_BIRD_BFD=y` protocol
enabled fails to build with the following error
```
bison -Dparse.lac=full -Dparse.error=verbose -dv -pcf_ -b obj/conf/cf-parse obj/conf/cf-parse.y
proto/bfd/config.Y:204.27-33: error: symbol 'ADDRESS' is used, but is not defined as a token and has no rules
204 | | bfd_show_sessions_args ADDRESS net_or_ipa { net_copy(&($$->address...
| ^~~~~~~
```
The `ADDRESS` token is defined only when certain protocols (e.g. OSPF,
RIP, RPKI, or BGP) are enabled. As a result, builds including any of
these protocols do not encounter the issue.
The issue can be reproduced with the commands:
cat >.config <<EOF
BR2_aarch64=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_PACKAGE_BIRD=y
BR2_PACKAGE_BIRD_BFD=y
# BR2_PACKAGE_BIRD_BGP is not set
EOF
make olddefconfig
make bird
This patch backports upstream commits that define the `ADDRESS` token
for the BFD protocol.
Fixes:
https://autobuild.buildroot.org/results/68c5dd84585a7018ad57ea3e7134748c08858ef7/
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
[Julien: add commands to reproduce the issue]
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 80cfdcb86b)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
This patch fixes the `S_IFMT` undeclared error in `statx.c` when musl
is used.
Signed-off-by: José Luis Salvador Rufo <salvador.joseluis@gmail.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 9a672635a1)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
This fixes a potential NULL pointer dereference
As a side note, this package has many opened CVEs, but upstream doesn't seem
to really consider them as security issues, see their disclaimer here:
https://github.com/yasm/yasm/blob/master/SECURITY.md
We could speculate that this disclaimer has been written as a consequence of
the many small CVEs opened in a short time, that don't have a substantial
security impact (besides the command line tool crashing). All of these small
CVEs have been opened for bug reports issued by a third party who used a
fuzzy tester to manipulate the assembler input
Signed-off-by: Titouan Christophe <titouan.christophe@mind.be>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 875f5670aa)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
This fixes the following CVE:
- CVE-2025-1492: The Bundle Protocol and CBOR dissectors could crash
It may be possible to make Wireshark crash by injecting a malformed
packet onto the wire or by convincing someone to read a malformed
packet trace file.
See https://www.wireshark.org/security/wnpa-sec-2025-01
Signed-off-by: Titouan Christophe <titouan.christophe@mind.be>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 032b268890)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
Annoyingly, using "--disable warning" does not disable the warnings
checks.
It turns out that we look for "warnings" (i.e. with an 's') to know if
we should disable the warnings check, so update the help text
accordingly.
Signed-off-by: Raphaël Mélotte <raphael.melotte@mind.be>
Signed-off-by: Julien Olivain <ju.o@free.fr>
(cherry picked from commit 92e7ab78d6)
Signed-off-by: Thomas Perale <thomas.perale@mind.be>
