pthread_rwlockattr_setkind_np() is a GNU extension of libpthread. and
Tianshan Qu pointed out, we cannot use
ifdef(PTHREAD_RWLOCK_PREFER_WRITER_NONRECURSIVE_NP) to detect the
availability of this function, because it's an enum not a macro. so,
like other *_np() extensions, we check this one also using cmake at
the configure phase.
Reported-by: Tianshan Qu <tianshan@xsky.com>
Signed-off-by: Kefu Chai <kchai@redhat.com>
bluestore,cmake: enable building bluestore without aio
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: Willem Jan Withagen <wjw@digiware.nl>
Reviewed-by: Mykola Golub <to.my.trociny@gmail.com>
KernelDevice is tightly coupled with libaio. more work is needed to
decouple aio from it. but by guarding KernelDevice with HAVE_LIBAIO, we
can enable bluestore on platforms w/o libaio.
Signed-off-by: Kefu Chai <kchai@redhat.com>
* move the check of `USE_CRYPTOPP` to $top_srcdir/CMakeLists.txt
* remove reference of DPDK_LIBRARY, it's defined nowhere
* move the dpdk code to a single place
Signed-off-by: Kefu Chai <kchai@redhat.com>
For CONFIG_BASE_FULL Linux kernels, the maximum number of proc/thread
IDs is set to 32768 by default. This default limit can be quite easily
hit during recovery on nodes with high OSD counts.
To avoid hitting the pid_max default limit, attempt to configure it to
4194304, which corresponds to the maximum limit possible on 64-bit
CONFIG_BASE_FULL kernels.
Fixes: http://tracker.ceph.com/issues/21929
Signed-off-by: David Disseldorp <ddiss@suse.de>
It is not really our business to debug python, boost, or our other
dependencies. Mark them as system includes.
Signed-off-by: Adam C. Emerson <aemerson@redhat.com>
the second param of enable_launage() is not used, we should call it
multiple times to enable more than one language. switch to project()
command for simplicity.
Signed-off-by: Kefu Chai <kchai@redhat.com>
we've updated the rockdb wrapper on ceph side to be compatible with
the latest version of rocksdb upstream. so ceph is not compatible with
older version of rocksdb.
Signed-off-by: Kefu Chai <kchai@redhat.com>
Tracing with LTTng doesn't work if an application clone()s other
threads. This is an LTTng-related issue and the fix is LD_PRELOAD'ing
liblttng-ust-fork, or linking against it.
We don't have this issue with "regular" tracepoints (see WITH_LTTNG)
because the tracepoints are loaded using dlopen() at runtime by each
Ceph process.
Signed-off-by: Mohamad Gebai <mgebai@suse.com>
All options to en/disable inclusion of libraries or other software
are of the format WITH_ so that the Cmake commaind like ahs all
WITH_* options. The WITH_=ON option will result in a HAVE_ setting
in CMAKE so that tests can use that variable.
Last "abusers" to actually follow this format.
- HAVE_BABELTRACE
- HAVE_ZFSLIB is fixed in #15907
Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>
* do not REQUIRE libzfs if it is enabled, this follows the way how we
handle `WITH_XFS` option.
* also refactor the cmake script related to libxfs backend support a
little bit.
Signed-off-by: Kefu Chai <kchai@redhat.com>
