since we've dropped the support of GCC older than v8.0, there is no need
to detect <experimental/filesystem>
Signed-off-by: Kefu Chai <kchai@redhat.com>
for better C++17 support, for instance for a better std::filesystem
support.
the reason why 8.1 is required is that ubuntu focal provides GCC-8.1,
and RHEL/CentOS8 provides GCC-8.4.1. so we only test the build on
GCC-8.1 and up so far.
Signed-off-by: Kefu Chai <kchai@redhat.com>
based on recent observation, quite a few C++ source file take
around more than 3.0GiB to compile. for instance,
test_mock_HttpClient.cc could take up to 6270MiB memory to compile.
so increase MAX_{LINK,COMPILE}_MEM accordingly.
Signed-off-by: Kefu Chai <kchai@redhat.com>
this change make it more explicit how we override CC, CFLAGS and
CPPFLAGS:
* CC only for the compiler
* CFLAGS only for the flags consumed by C compiler
* CPPFLAGS only used for the preprocessor, it is used by both C and C++
compilers.
this change does not address any specific test failure. it just
improves the correctness and readability.
Signed-off-by: Kefu Chai <kchai@redhat.com>
the __asm__(".asmver ..") is a support provided by the compiler, so
would be better to detect it by either checking the compiler identifer
or just try it out.
in this change, instead of checking the building platform, we check this
feature using check_c_source_compiles().
in future, we could support versioned symbols using function attriubte
or symbol tables or version-script.
on platform where symbol versioning is not supported, we might need to
go with a different approach.
Signed-off-by: Kefu Chai <kchai@redhat.com>
the behavior of cmake 3.17 is a little bit different as it consider
NOT _library STREQUAL ""
as true, if ${_library} is "".
Signed-off-by: Kefu Chai <kchai@redhat.com>
to silence the warning like:
/home/jenkins-build/build/workspace/ceph-pull-requests/build/src/pybind/rados/rados.c: In function '__pyx_pymod_exec_rados':
/home/jenkins-build/build/workspace/ceph-pull-requests/build/src/pybind/rados/rados.c:78903:30: note: variable tracking size limit exceeded with '-fvar-tracking-assignments', retrying without
78903 | static CYTHON_SMALL_CODE int __pyx_pymod_exec_rados(PyObject *__pyx_pyinit_module)
| ^~~~~~~~~~~~~~~~~~~~~~
cc: warning: /usr/bin/cc: linker input file unused because linking not done
Signed-off-by: Kefu Chai <kchai@redhat.com>
since liburing's configure checks the CC and CXX env variables, let's
pass them when calling "configure". otherwise "make" still uses gcc and
g++ by default as they are specified in the "configure" script. this
also matches with the way how debian/rules calls "make".
Signed-off-by: Kefu Chai <kchai@redhat.com>
before this change, `pmem_VERSION_STRING` is not set if it is not able
to fulfill the specified version requirement. the intention was to check
if the version is able to satisfy the requirement. but actually, passing
an empty `pmem_VERSION_STRING` to `find_package_handle_standard_args()`
as the option of `VERSION_VAR` does not fail this check. on the
contrary, it prints
-- Found pmem: pmem_pmemobj_INCLUDE_DIR;pmem_pmem_INCLUDE_DIR (Required
is at least version "1.17")
if we requires pmem 1.17, while the found version is, for instance,
1.10.
if the required version is 1.7, and the found version is 1.10, the
output from cmake is:
-- Found pmem: pmem_pmemobj_INCLUDE_DIR;pmem_pmem_INCLUDE_DIR (found
suitable version "1.10", minimum required is "1.7")
in this change, the version spec is not specified when calling
`pkg_check_modules()`. so, `PKG_${component}_VERSION` is always set.
and we can always delegate the version checking to
`find_package_handle_standard_args()`. please note, we use the lower
version returned by pkg-config if multiple components are required and
both pkg-config settings return their versions.
Signed-off-by: Kefu Chai <kchai@redhat.com>
as FreeBSD also has msghdr but it does not have systemd, or
flags like MFD_ALLOW_SEALING, O_TMPFILE or F_SEAL_GROW. so
use WITH_SYSTEMD for enabling journald backend of logging system.
also move the option of "WITH_SYSTEMD" up so that the src/CMakeLists.txt
is able to see the variable of WITH_SYSTEMD defined by it.
Signed-off-by: Kefu Chai <kchai@redhat.com>
* cmake/modules/CephChecks.cmake: detect the existence of struct msghdr,
define HAVE_MSGHDR if it is found
* src/common/CMakeLists.txt: do not compile journald.cc if HAVE_MSGHDR
is FALSE. as in that case, we cannot use sendmsg() to write to
journald unix domain socket
* src/test/CMakeLists.txt, src/test/common/CMakeLists.txt: disable test
exercising journald logging backend if HAVE_MSGHDR is not defined
* src/common/Journald.h: define a dummy JournaldLogger and a dummy
JournaldClusterLogger when HAVE_MSGHDR is not defined, in order to
minimize the change in Log.h and Log.cc, otherwise the source code of
Log.h and Log.cc would be segmented into smaller chunks by
`ifdef HAVE_MSGHDR` macros.
* src/include/config-h.in.cmake: define a new macro named
HAVE_MSGHDR.
Signed-off-by: Kefu Chai <kchai@redhat.com>
detect the libpmem libraries' version using pkg-config
please note, the version is not identical to the ones defined in
libpmem.h or libpmemobj/base.h
Signed-off-by: Kefu Chai <kchai@redhat.com>
since we are moving the test nodes from bionic to focal, we are able to
use the prebuilt libzstd libraries when running "make check". to speed
up the build and test, in this change:
* add FindZstd.cmake which allows us to use the libzstd in system
* extract BuildZstd.cmake for better readability
* add an option named "WITH_SYSTEM_ZSTD", which defaults to "OFF",
so user can enable it on demand.
Signed-off-by: Kefu Chai <kchai@redhat.com>
this allows us to pass extra argument to the script or executable to
further customize its behavior when testing.
also fix all the callers of add_ceph_unittest(), as there is no need to
repeat the executable, which is also used as the name of the test.
Signed-off-by: Kefu Chai <kchai@redhat.com>
this change should allow us to decouple "ninja tests" from "ctest".
in other words, we can just run
ctest -R run-tox-python-common -V
without running "ninja tests" first. before this change
${name}-venv is added as a dependency of "tests" target.
after this change,
setup-venv-for-${name} is added as a test, which is in turn a test of
run-tox-${name}, so we can just
ctest -R run-tox-${name}
now for preparing the venv and then testing the tox test of ${name}.
Signed-off-by: Kefu Chai <kchai@redhat.com>
to limit the parallel level of compile jobs which consumes lots of memory
to compile and/or link.
compile following targets with heavy_compile_pool.
* unittest_librbd
* crimson*
* ceph-dencoder
some of the .cc files in the target above use up to 6 GiB memory to
compile.
Signed-off-by: Kefu Chai <kchai@redhat.com>
This is a complete rewrite on top of SimpleRADOSStriper as well as the
API. The VFS is now a loadable extension as well.
Fixes: https://tracker.ceph.com/issues/40609
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
the vanilla FindBoost.cmake pulled from cmake has couple assumptions
which do not hold in our environment. so address them case by case.
Signed-off-by: Kefu Chai <kchai@redhat.com>
git warns us when the repo is in 'detached HEAD' state, like:
You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in
this state without impacting any branches by switching back to a branch.
...
Turn off this advice by setting config variable advice.detachedHead to
false
Signed-off-by: Kefu Chai <kchai@redhat.com>
libpmemobj should link against libpmem, but, in CMake, imported library
does not allow PRIVATE linkage. so pmem::pmem is added to the list of
INTERFACE_LINK_LIBRARIES.
Signed-off-by: Kefu Chai <kchai@redhat.com>
add two components: pmem and pmemobj to this package. so we can find
them and link against them in a more intuitive way.
before this change the COMPONENTS parameter passed to
find_package(pmem ...)
is dropped on the floor and ignored.
after this change, it is checked and taken into consideration.
also, in this change, the exposed variables are renamed from
PMEM_* to pmem_*
to be consistent with the package name. it's encouraged to be consistent
with the package name when it comes to the INCLUDE_DIR and LIBRARIES
variable names.
Signed-off-by: Kefu Chai <kchai@redhat.com>
to address the warning from cmake, like
CMake Warning (dev) at /usr/share/cmake-3.18/Modules/FindPackageHandleStandardArgs.cmake:273 (message):
The package name passed to `find_package_handle_standard_args` (numa) does
not match the name of the calling package (NUMA). This can lead to
problems in calling code that expects `find_package` result variables
(e.g., `_FOUND`) to follow a certain pattern.
Signed-off-by: Kefu Chai <kchai@redhat.com>
RocksDB's cmake has a similar logic to turn on ccache as the main cmake.
Currently BuildRocksDB.cmake explicitly passes ccache to RocksDB cmake
arguments as CMAKE_CXX_COMPILER_LAUNCHER. This causes a compilation error
on ccache 4.0, especially in make -j mode.
ccache: error: Recursive invocation (the name of the ccache binary must
be "ccache")
This commit lets cmake from RocksDB decide RULE_LAUNCH_COMPILE by itself.
Signed-off-by: Sheng Mao <shngmao@gmail.com>
if nasm is able to emit AVX512 instructions, we can assume that it's
able to generate AVX2 instructions as well. so no need to print "Found
nasm" multiple times.
and by chaining mutual exclusive branches with "elseif", we can have
better readability.
Signed-off-by: Kefu Chai <kchai@redhat.com>
