Use blkid APIs to get device parition information into the metadata bundle, and leverage common/blkdev.* helpers to parse the device node when possible.
Signed-Off-By: joseph.t.handzik@hp.com
Prior to automake 1.13 the default behavior was serial-tests meaning
tests from make check were run one after the other and their output sent
to stdout/stderr. From automake 1.13 up the default became
parallel-tests which logs the output of each individual test in a
separate .log file and allows them to run in parallel.
http://lists.gnu.org/archive/html/automake/2012-12/msg00038.html
Enable parallel-tests so that tests can always run in parallel, even
with automake 1.11 which is the version found in CentOS 6.5 and Ubuntu
12.04 precise.
A nice side effect of always having the logs in separate files is that
tests do not need to worry about being too verbose because their output
will only be read for diagnostic purposes.
http://tracker.ceph.com/issues/11931Fixes: #11931http://tracker.ceph.com/issues/11906Fixes: #11906
Signed-off-by: Loic Dachary <ldachary@redhat.com>
The configure's --with-debug option builds also some of the tests
so this should be mentioned in its doc string in order to let users
know how to build all the tests.
Signed-off-by: Joaquim Rocha <joaquim.rocha@cern.ch>
The libexec path is different for different distributions.
systemd. This path is defined by a new variable on the
configure path.
This variable can be set with enviroment SYSTEMD_LIBEXEC_DIR.
The parameter --with-systemd-libexec-dir overrides the enviroment
variable.
Appropriate conditionals are set for SUSE and RHEL derivatives.
This is then used to template out systemd/ceph-osd@.service
Signed-off-by: Owen Synge <osynge@suse.com>
* do not require sphinx-build if --without-man-pages
* dist rst files for man pages
* build and dist .8 files if sphinx-build is found
Signed-off-by: Kefu Chai <kchai@redhat.com>
Currently, if you add the flag --disable-pgrefdebugging during configure
it enables pgrefdebugging. The only way for it to actually be disabled
is to remove the flag entirely. This patch makes the flag respect the
--enable and --disable options correctly. Therefore,
--enable-pgrefdebugging is the only way to enable this now and
--enable-pgrefdebugging=no and --disable-pgrefdebugging work as
expected.
Signed-off-by: William A. Kennington III <william@wkennington.com>
On RHEL 6 and CentOS 6 sphinx 1.0+ which is required to build man pages
is packaged as sphinx10 instead of sphinx and the binary is
sphinx-1.0-build instead of sphinx-build.
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Certain older systems (SLE11 in this case) do not provide the full
tcmalloc functionality, due to e.g. incomplete libunwind
pieces. Use --with-tcmalloc-minimal to enable the cut-down
version.
Here's how the various mem allocator switches interact now:
--with-jemalloc: overrides --with-tcmalloc & --with-tcmalloc-minimal
--with-tcmalloc-minimal: overrides --with-tcmalloc
--with-tcmalloc: the default. use --without-tcmalloc to disable
Signed-off-by: Thorsten Behrens <tbehrens@suse.com>
* move rst files from doc/man/8 into man/8
* extract man_pages setting for sphinx from doc/conf.py to man/conf.py
* generate all man pages in `make install`
* add python-sphinx to Build-Depends
* check for sphinx-build in `configure`
* run changed rst files through sphinx-build in `make check`
Signed-off-by: Kefu Chai <kchai@redhat.com>
During development each git commit cause currently a complete
rebuild of the source tree. By passing --disable-gitversion to
configure we set a generic version string for ceph to prevent
getting after each commit a new version set.
remove src/check_version and replace with a extended version
of src/make_version which allows detecting changes in .git_version
and ceph_ver.h to avoid not needed updates.
This change also forces to call make_version each time to make
sure that changes from configure or the NO_VERSION env variable
are processed.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
ARMv8 defines a set of optional CRC32/CRC32C instructions.
This patch defines an optimized function that uses these
instructions when available rather than table-based lookup.
Optimized function based on a Hadoop patch by Ed Nevill.
Autotools updated to check for compiler support.
Optimized function is selected at runtime based on HWCAP_CRC32.
Added crc32c "performance" unit test and arch unit test.
Tested on AMD Seattle.
Passes all crc32c unit tests.
Unit test shows ~4x performance increase versus sctp.
Signed-off-by: Yazen Ghannam <yazen.ghannam@linaro.org>
Reviewed-by: Steve Capper <steve.capper@linaro.org>
Google Testing Framework is included by default within the Google
C++ Mocking Framework. Update makefiles to use new gmock/gtest
libraries and remove old gtest source code.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Now, 2-4 async op thread can fully meet a OSD's network demand with SSD
backend. So we can bind limited thread to special cores, it can improve
async event loop performance because most of structure and method will
processed within thread.
For example,
ms_async_op_threads = 2
ms_async_affinity_cores = 0,3
Signed-off-by: Haomai Wang <haomaiwang@gmail.com>
Replace ceph-helpers.sh check for ms_nocrc with the new formula
for this. Fixes make check for default build.
Additionally, fix linkage of several unittests when building with
--enable-xio.
xio: add missing noinst headers
The common/address_helper.h file was not mentioned, also
msg/xio/XioSubmit.h.
Fix for Message.cc compilation error when Xio disabled.
Mention simple_dispatcher.h and xio_dispatcher.h in noinst_HEADERS.
xio: require boost-regex.
Make address_helper conditional on Xio.
This carries over to simple_client/simple_server,
for convenience.
Signed-off-by: Matt Benjamin <matt@cohortfs.com>
Prior to this commit, when ./configure can't find libkeyutils, it would
bail out with a terse error message.
Some of the other library checks helpfully print the DEB and RPM package
names in parentheses. Add the DEB and RPM package names to the
libkeyutils check.
Reported-by: Pankaj Garg <Pankaj.Garg@caviumnetworks.com>
Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
The --enable-docker logic was that each test that needs privileges to
run uses a container to do so. The problem with this approach and
make -j8 check is that such tests compete for a single container.
The --enable-root-make-check activates tests that require privileges and
assume it is ok to use sudo to acquire privileges. The decision to run
in a container is then taken by the caller who will run something like:
docker-tests.sh ./configure --enable-root-make-check
docker-tests.sh make -j8 check
Signed-off-by: Loic Dachary <ldachary@redhat.com>
If --with-jemalloc option is not specified, then it defaults to tcmalloc.
If --with-jemalloc is specified, then --without-tcmalloc is required.
Also if a rocksdb static linking is desired --with-librocksdb-static,
do not use --with-jemalloc, as rocksdb uses tcmalloc for builds.
Support for jemalloc in package build and perf stats/heap profiler is
underway
Signed-off-by: shishir gowda <shishir.gowda@sandisk.com>
The google/ headers location has been deprecated as of gperftools 2.0.
As of gperftools 2.2rc, the google/ headers will now give deprecation
warnings, and they will probably disappear in a future gperftools
update.
Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
Docker based tests should be explicit instead of auto-detected. It is
good that they do not run if docker is not available. It would be bad if
they run when the developer does not expect them to create docker
containers.
Signed-off-by: Loic Dachary <loic-201408@dachary.org>
AsyncMessenger will select event driver following epoll, kqueue and
select(now not exists) sequence
Fix#9926
Signed-off-by: Haomai Wang <haomaiwang@gmail.com>
Detect leveldb, but do not let autoconf blindly link it with everything on the
planet.
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Sighed-off-by: Sage Weil <sage@redhat.com>
Implement the KeyValueDB interface using libkinetic_client,
and allow it to be configured as the backend for the KeyValueStore,
running the entire OSD on it.
This prototype implementation has no transaction safety, and is
only suitable as a proof of concept. Since the libkinetic_client
API does not provide reverse iteration over keys without also reading
the value off disk, it implements iterators in a very slow but correct way.
These are used heavily by the KeyValueDB callers, so this is a bottleneck
in performance.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
Add libkrbd libtool convenience library to provide an interface for
mapping and unmapping rbd images programmatically. This will be used
by the rbd binary itself and the librbd_fsx testing tool.
libkrbd takes care of the kernel module stuff (common/module.h) and
makes use of libudev to be able to properly wait for block device
creation and deletion and tell which block device got assigned by the
kernel to the newly created mapping.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
strerror_r is not portable; on Gnu libc it returns char * and sometimes
does not fill in the supplied buffer. Use autoconf to test which
version this platform uses and adapt.
Clean up the random calls to strerror and strerror_r (along with all
their private little one-use buffers) and regularize the code to use
cpp_strerror almost everywhere. Where changed, any negation of the
error code is also removed, since cpp_strerror() will do that.
Note: some tools were using their own calls to strerror/strerror_r, so
will now get a (%d) in their output that wasn't there before; hence
the change to test/cli/monmaptool/print-nonexistent.t
Fixes: #8041
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Rename SIMD to INTEL for clarity.
Instead of agregating all flags in INTEL_FLAGS, create individual flags
for each feature (INTEL_SSE2_FLAGS etc.) for finer control in the
makefiles.
Signed-off-by: Loic Dachary <loic@dachary.org>
For each SSE feature supported by the compiler
* add the corresponding -msse* flag
* define HAVE_SSE*
Remove AX_EXT because it decides based on the CPU capabilities of the
machine compiling the binary which may or may not be the one running
them.
Signed-off-by: Loic Dachary <loic@dachary.org>
When configured with --without-libxfs, use GenericFileStoreBackend
instead of XfsFileStoreBackend for XFS. At this point this would only
impact the allocation hint op. The default is to compile with
--with-libxfs. (Previously it was unconditionally enabled on linux and
disabled for non-linux arches.)
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Introduce XfsFileStoreBackend class, currently the only filestore
backend implementing SETALLOCHINT op. This commit adds a build-time
dependency on libxfs as xfs-specific ioctl (XFS_IOC_FSSETXATTR /
XFS_XFLAG_EXTSIZE) is used to implement the new set_alloc_hint()
method.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Currently CEPH_HAVE_SETPIPE_SZ is not set even if F_SETPIPE_SZ is
available, because AC_COMPILE_IFELSE test program as written always
fails to compile. F_SETPIPE_SZ is a macro, so use AC_EGREP_CPP which
works on the preprocessor output instead of trying to compile.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Working around missing integer types is pretty easy. For example, the
__u32 family are Linux-specific types, and using these in Ceph
internally is fine because we can typedef them.
Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
Adds a ceph_spinlock_t implementation that will use pthread_spinlock_t
if available, and otherwise reverts to pthread_mutex_t. Note that this
spinlock is not intended to be used in process-shared memory.
Switches implementation in:
ceph_context
SimpleMessenger
atomic_t
Only ceph_context initialized its spinlock with PTHREAD_PROCESS_SHARED.
However, there does not appear to be any instance in which CephContext
is allocated in shared memory, and thus can use the default private
memory space behavior.
Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
get_linux_version() returns a version of the currently running kernel,
encoded as in int, and is contained in common/linux_version.[ch].
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Break up AC_CHECK_HEADERS macro into one header-file per line so it's
easier to read and make changes.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Currently the way 'rbd unmap' translates a user-provided block device
into an rbd id is it matches the major number of the specified device
against /sys/bus/rbd/devices/<id>/major for each rbd mapping and
declares success on the first match. This works for both entire disks
and partitions, because under the current device number allocation
scheme, each mapping means a new major number.
In preparation for support for single-major device number allocation
scheme, which would require matching both major and minor numbers, make
sure to always match against entire disk device numbers, by converting
the specified device major:minor pair into wholdedisk major:minor pair.
To achive that, use the libblkid library, which accomplishes this goal
by walking stable sysfs structures.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Checking for fdatasync uses the same approach as the qemu configure
script. The relevant commit is d1722a27f552a22561104210e0afad4577878e53.
Here is a copy of the commit message which explains the check:
Under Darwin, a symbol exists for the fdatasync() function, so that our
link test succeeds. However _POSIX_SYNCHRONIZED_IO is set to '-1'.
According to POSIX:2008, a value of -1 means the feature is not
supported.
A value of 0 means supported at compilation time, and a value greater 0
means supported at both compilation and run time.
Enable fdatasync() only if _POSIX_SYNCHRONIZED_IO is '>0'.
Signed-off-by: Noah Watkins <noahwatkins@gmail.com>
