The 0003 patch that links the data_source test carrier binary without
the host runtime also adds -fno-stack-protector to the object compile
line, needed because host gcc toolchains default to
-fstack-protector-strong and the -nostdlib link then fails with:
data_source.c:212: undefined reference to `__stack_chk_fail'
However patch 0001 (DEBUG_PREFIX_MAP) runs first in SRC_URI and rewrites
the same compile line to insert ${DEBUG_PREFIX_MAP_LIST}, so 0003's
first hunk no longer matched the tree and the -fno-stack-protector flag
was silently dropped from the object compile (it only landed on the
link line). Rebase 0003's first hunk onto the post-0001 context so the
flag is applied where it is needed.
Signed-off-by: Khem Raj <khem.raj@oss.qualcomm.com>
The build host HOME directory check recently added to OE-Core's
buildpaths QA test flags the ptest gtest binary:
WARNING: bpftrace-0.25.1-r0 do_package_qa: QA Issue: File
/usr/lib/bpftrace/ptest/tests/bpftrace_test in package bpftrace-ptest
contains a reference to the build host HOME directory. [buildpaths]
WARNING: bpftrace-0.25.1-r0 do_package_qa: QA Issue: File
/usr/lib/bpftrace/ptest/tests/.debug/bpftrace_test in package
bpftrace-dbg contains a reference to the build host HOME directory.
[buildpaths]
cmake/BuildBPF.cmake links the intermediate DWARF carrier binary
data_source_exe with the plain build host gcc, which bakes the host
dynamic loader's absolute path into the PT_INTERP segment. With a
distro gcc that is an innocuous /lib64/ld-linux-x86-64.so.2, but when
the host toolchain is a buildtools-extended/SDK gcc installed under
the build user's home directory (as on the autobuilder workers), it
is a $HOME-prefixed path like
/srv/pokybuild/.../buildtools/sysroots/x86_64-pokysdk-linux/lib/ld-linux-x86-64.so.2
The executable is then embedded byte-for-byte into bpftrace_test via
embed()/xxd (tests/data/CMakeLists.txt), so the path ends up in the
test binary's .rodata, triggering the bpftrace-ptest warning. And
because cmake/Embed.tmpl declares the embedded blob as a constexpr
array, gcc additionally copies its contents into .debug_info as
DW_AT_const_value, which is how the same string survives
objcopy --only-keep-debug into the split debug file, triggering the
bpftrace-dbg warning. This is the same embedding mechanism that
previously leaked TMPDIR paths, fixed for the compile step by the
DEBUG_PREFIX_MAP patch; the link step was still leaking.
The tests only write this binary to a temp file and use it as a
uprobe target (tests/dwarf_common.h), parsing its DWARF and symbols;
it is never executed. So link it with -nostdlib -no-pie -Wl,--entry=0:
no interpreter, no host crt objects, no dynamic segment, and thus
nothing host-specific in the embedded blob. Function symbols, DWARF
type information and the pahole -J BTF encoding step are unaffected,
and the field_analyser_dwarf ptests still pass.
Because the object is now linked -nostdlib it must not reference the
libc stack-protector runtime. Host gcc toolchains that default to
-fstack-protector-strong (e.g. Debian/Ubuntu and the Yocto
autobuilder workers) otherwise leave an undefined reference and the
link fails:
ld: data_source.o: in function `main':
data_source.c:212: undefined reference to `__stack_chk_fail'
collect2: error: ld returned 1 exit status
Compile the object and link the binary with -fno-stack-protector; the
canary is irrelevant to a DWARF/BTF carrier that is never run.
Signed-off-by: Khem Raj <khem.raj@oss.qualcomm.com>
0001-Add-ARM64-syscall-prefix-detection-in-C-API.patch
0002-Add-riscv-syscall-prefix-detection-in-C-API.patch
0003-folly-tracing-Remove-x86-specific-naming-from-tracin.patch
0004-folly-tracing-Add-ARM-and-AArch64-support-to-static-.patch
0001-Fix-build-with-LLVM-22.patch
removed since they're included in 0.37.0
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Khem Raj <khem.raj@oss.qualcomm.com>
With the current version, I had the QA issues below:
WARNING: bpftrace-0.25.1-r0 do_package_qa: QA Issue:
File /usr/lib/bpftrace/ptest/tests/.debug/bpftrace_test in package bpftrace-dbg contains reference to TMPDIR [buildpaths]
WARNING: bpftrace-0.25.1-r0 do_package_qa: QA Issue:
File /usr/lib/bpftrace/ptest/tests/bpftrace_test in package bpftrace-ptest contains reference to TMPDIR [buildpaths]
A bit of find + grep + strings show that this originally comes from:
field_analyser.cpp.o: [my tmpdir...]/bpftrace/0.25.1/build/tests/data
field_analyser.cpp.o: [my tmpdir...]/bpftrace/0.25.1/sources/bpftrace-0.25.1/tests/data/data_source.c
field_analyser.cpp.o: [my tmpdir...]/bpftrace/0.25.1/sources/bpftrace-0.25.1/tests/data
The .rodata section is infected with these, and after some more digging,
the paths come from debug symbols (i.e., .debug_line_str) in
data_source.o:
build/tests/data/data_source.o: /src/update-2026.01/build/tmp/work/riscv64.../bpftrace/0.25.1/build/tests/data
build/tests/data/data_source.o: /src/update-2026.01/build/tmp/work/riscv64.../bpftrace/0.25.1/sources/bpftrace-0.25.1/tests/data/data_source.c
build/tests/data/data_source.o: /src/update-2026.01/build/tmp/work/riscv64.../bpftrace/0.25.1/build/tests/data
build/tests/data/data_source.o: /src/update-2026.01/build/tmp/work/riscv64.../bpftrace/0.25.1/sources/bpftrace-0.25.1/tests/data
I'm not familiar with the magic performed by BuildBPF.cmake, but as far
as I can tell, gcc is being invoked with "-g" and no prefix mapping,
so there is no way of stopping the source path(s) from ending in data_source.o.
This patch adds a "bridge" between this .cmake file and bpftrace's
recipe, where I export DEBUG_PREFIX_MAP, so gcc is invoked with the
-ffile-prefix-map flags from OE Core.
The end result is:
$ strings ./build/tests/data/data_source.o | grep "debug\/bpftrace"
/usr/src/debug/bpftrace/0.25.1/tests/data/data_source.c
/usr/src/debug/bpftrace/0.25.1/tests/data
/usr/src/debug/bpftrace/0.25.1/tests/data
/usr/src/debug/bpftrace/0.25.1/tests/data
Signed-off-by: João Marcos Costa <joaomarcos.costa@bootlin.com>
Signed-off-by: Khem Raj <khem.raj@oss.qualcomm.com>
Relocate the recipe from dynamic-layers/selinux/ to the main
meta-oe/recipes-devtools/ directory.
The android-tools recipe (version 29.0.6.r14) was previously restricted
to the selinux dynamic-layer.
Investigation shows that version 29.0.6.r14 does not have a hard dependency
on libselinux for core tool functionality.
(adb, fastboot, and sparse image tools).
- Basic runable test was done for binaries in android-tools-native
- Checked for selinux absense by looking for selinx using $ ldd binary-name
Changes:
- Relocate recipe from dynamic-layers/selinux/ to recipes-devtools/
Signed-off-by: AshishKumar Mishra <ashishkumar.mishra@bmwtechworks.in>
Signed-off-by: Khem Raj <khem.raj@oss.qualcomm.com>
bpftrace-ptest requires bpftool which is not compatible with powerpc64.
So remove sub-package bpftrace-ptest for powerpc64 & powerpc64le.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Khem Raj <khem.raj@oss.qualcomm.com>
* bash and python3 are only needed by the ptest package.
* xz appears to not be needed at all.
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
- Labeled adb binary
- Moved adb shell from initrc_t to unconfined_t
- meta-selinux does not provide adb domain added policy in meta-oe
instead of refpolicy: SELinuxProject/refpolicy#1085
Signed-off-by: Gargi Misra <gmisra@qti.qualcomm.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Without xz present, compilation fails with the following error:
| <...>/x86_64-oe-linux-ld: warning: liblzma.so.5, needed by <...>/libbcc.so, not found (try using -rpath or -rpath-link)
| <...>/x86_64-oe-linux-ld: <...>/libbcc.so: undefined reference to `lzma_code@XZ_5.0'
| <...>/x86_64-oe-linux-ld: <...>/libbcc.so: undefined reference to `lzma_end@XZ_5.0'
| <...>/x86_64-oe-linux-ld: <...>/libbcc.so: undefined reference to `lzma_stream_decoder@XZ_5.0'
| collect2: error: ld returned 1 exit status
Signed-off-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Building android-tools v29.0.6.r14 with glibc 2.43 fails due to ISO C23
changes to strchr(). Add a patch to update the affected libunwind sources
to use const pointer types to fix this build failure.
Signed-off-by: Viswanath Kraleti <viswanath.kraleti@oss.qualcomm.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
This resolves USDT probe test failures on ARM64 platforms.
Without these changes, the .note.stapsdt section containing probe
information was missing entirely on ARM64, causing test failures when
attempting to find and attach to USDT probes in the BCC test suite.
Upstream-Status: Submitted [https://github.com/iovisor/bcc/pull/5491]
Signed-off-by: Harish Sadineni <Harish.Sadineni@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
For the boards with multiple UDC ports, when compiling with meta-selinux
layer adb funtion will be impacted due to below error. Remove the recipe
under dynamic-layers/selinux/ as it's not required anymore, and ensure
adb work normally on the boards with multiple UDC ports.
ls /sys/class/udc/ > /sys/kernel/config/usb_gadget/adb/UDC
ls: write error: Device or resource busy
Signed-off-by: Shoudi Li <shoudil@qti.qualcomm.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
2. Pyrwmem and static-libc build option has been removed in 2.0.
Signed-off-by: Liu Yiding <liuyd.fnst@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
With thingsboard-gateway 3.8.0, the source URI must be set to the
project Git repository, as source releases are no longer provided via
PyPi for this project.
Signed-off-by: Martin Schwan <m.schwan@phytec.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Some post-processing options require an argument, otherwise a segfault
will occur:
root@qemux86-64:~# rasdaemon -p --status --ipid
Segmentation fault (core dumped) rasdaemon -p --status --ipid
Backport a patch to fix this issue.
Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Compilation fails due to undefined types 'uint16_t' and 'uint32_t' in
zip_writer.h. These types are used in struct members but are not recognized
because the required header <cstdint> is not included. This results in
errors such as:
error: 'uint16_t' does not name a type
error: 'uint32_t' does not name a type
Added '#include <cstdint>' to zip_writer.h to ensure fixed-width integer
types are properly defined.
Signed-off-by: Sasi Kumar Maddineni <quic_sasikuma@quicinc.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The build fails due to a macro redefinition conflict for `unreachable()`.
GCC 15.2.0 defines `unreachable()` in `stddef.h`, and `libunwind_i.h` also
defines it based on the presence of `HAVE__BUILTIN_UNREACHABLE`. This causes
a redefinition error when building with `-Werror`.
Added a guard around the `unreachable()` macro definition in `libunwind_i.h`
to ensure it is only defined if not already present.
Signed-off-by: Sasi Kumar Maddineni <quic_sasikuma@quicinc.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The build fails with a compiler error due to a macro redefinition of 'write'
to '___xxx_write'. This macro substitution interferes with the standard C++
library's definition of std::basic_ostream::write, leading to a mismatch
between function declarations and definitions in <ostream.tcc>.
Wrap the `#undef write` and `#define write ___xxx_write` directives with
`#if defined(__ANDROID__)` to ensure they are only applied when building
for
Android.
Signed-off-by: Sasi Kumar Maddineni <quic_sasikuma@quicinc.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Update the systemd service file to reflect the changes of the source
repository.
Add a user and group "thingsboard_gateway" which the service runs under,
instead of root.
Store temporary files, like logs, in /var and make use of the newly
created user and group.
Additionally, the service should start after and wants a network
connection to be online, not just be activated.
Signed-off-by: Martin Schwan <m.schwan@phytec.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Replace the outdated configuration files with the ones found the source
repository of thingsboard-gateway. This makes sure, the configuration
file examples stay up-to-date in future versions.
Signed-off-by: Martin Schwan <m.schwan@phytec.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Add all missing runtime dependencies that are required by
thingsboard-gateway. Since thingsboard-gateway tries to install missing
dependencies via pip and fails to do so, install the dependencies
system-wide beforehand.
Signed-off-by: Martin Schwan <m.schwan@phytec.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
There's no need to inherit python3native as there are no python3- native
dependencies.
Remove RDEPENDS_${PN}, these are all libraries that are linked to so the
relevant RDEPENDS are created. The fact that this recipe is still using
the old syntax suggests that they're not needed.
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
meson's rtti detection logic fails especially with
clang which disables rtti by default. The test is
triggerred in one of taisei's submodules especially
this commit [1], I think it should be something in
meson to fix in its rtti detection logic
Similarily LTO is only enabled when it is in distro
features, clang disables support for LTO in toolchain
when its not in distro features and linking fails since
it can not find linker plugin.
[1] https://github.com/taisei-project/basis_universal/commit/851bfc63fd0f9d00dd1a21fb542a9b8c0f7d06ec
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Stable release with fixes
Fixed the macOS build crashing on startup.
Fixed audio distortion when the audio device uses a sample rate other than 48 kHz.
Fixed the internal mixer_chunksize setting being ignored. This resulted in a larger audio buffer than intended, increasing latency.
Fixed some minor SDL3 migration issues, particularly in handling of IO errors.
Fixed mimalloc being built incorrectly as a subproject.
Debugging symbols for the official builds are now available as a separate download.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Remove unsupported '-mcpu=octeontx2+crypto' from BINDGEN_EXTRA_CLANG_ARGS
as clang does not recognize 'octeontx2' as a valid target CPU, causing
bindgen to fail when generating Rust bindings.
Since bindgen only parses headers using Clang, CPU-specific options
like -mcpu are generally unnecessary.
Fixes build failure:
| error: unsupported argument 'octeontx2+crypto' to option '-mcpu='
| error: unknown target CPU 'octeontx2'
Signed-off-by: Bo Sun <bo.sun.cn@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Khem Raj
Wang Mingyu
Joao Marcos Costa
AshishKumar Mishra
Gargi Misra
Yi Zhao
Swamil Jain
Abhilasha Manna
Kai Kang
Peter Kjellerstedt
Gyorgy Sarvari
Viswanath Kraleti
Harish Sadineni
Shoudi Li
Khem Raj
Liu Yiding
Martin Schwan
Weisser, Pascal
Sasi Kumar Maddineni
Changqing Li
Ross Burton
Bo Sun