test/libradosstriper/striping.cc: fix wrong indentation

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

msg/async/Event.cc: use static_cast instead of c-style cast

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

rgw/rgw_formats.cc: fix realloc memory leak

Fix handling of realloc. If realloc() fails it returns NULL, assigning
the return value of realloc() directly to the pointer without checking
for the result will lead to a memory leak in error case.

Use a temporary pointer to hold the result of realloc(). In error case
print error and exit, otherwise assign it to the pointer we want to realloc.

Fix also error checks for malloc to fail as soon as malloc fails and
don't try to run realloc again after failed malloc.

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

arch/neon.c: check result of fread()

Fix for:

CID 1128384 (#1 of 1): Ignoring number of bytes read (CHECKED_RETURN)
 check_return: fread(void * restrict, size_t, size_t, FILE * restrict)
 returns the number of bytes read, but it is ignored.

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

cls_rgw_client.cc: fix potential null pointer deref

Fix for:

CID 1054853 (#1 of 1): Dereference before null check (REVERSE_INULL)
 check_after_deref: Null-checking is_truncated suggests that it may
 be null, but it has already been dereferenced on all paths leading
 to the check.

Add vim line to file.

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

SyntheticClient.cc: check return value of getdir()

Check return value as done in all other places. Fix error messages
to print correct function name getdir and not read_dir/readdir since
the error isn't necessarily raised by read_dir().

Fix for:
CID 1219463 (#1 of 1): Unchecked return value (CHECKED_RETURN)
 check_return: Calling getdir without checking return value (as
 is done elsewhere 4 out of 5 times).

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

SyntheticClient.cc: remove empty lines in if/else blocks

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

test_rbd_replay.cc: fix va_list parameter handling

Fix for:

[src/test/test_rbd_replay.cc:193]: (portability) Passing NULL after the
last typed argument to a variadic function leads to undefined behaviour.

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

AsyncConnection.cc: prefer ++operator for non-primitive iterators

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

mon_store_converter.cc: move convert() call out of assert()

The convert() call is changing data. To prevent Heisenbugs
depending in NDEBUG definition (e.g. if debug is off).

Move convert() call out of the assert call and use assert
only on the result value of the call.

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

test_str_map.cc: remove unused variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

rgw/rgw_rest_user.cc: reduce scope of key_type

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

rgw/rgw_rest_swift.cc: use empty() instead of size()

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

rgw/rgw_rest_s3.cc: use !empty() instead of size() > 0

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

rgw/rgw_rados.cc: remove unused variable 'objs'

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

osd/ReplicatedPG.cc: use !empty() instead of size() > 0

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

PGLog.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

common/escape.c: add cppcheck-suppress

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

fuse_ll.cc: use reinterpret_cast instead of c-style cast

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

JournalScanner.cc: reduce scope of local variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

test_disk_bw.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

objectstore/store_test.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

TestErasureCodeIsa.cc: reduce scope of 'err' variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

TestCrushWrapper.cc: reduce scope of variabele

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

test_sharedptr_registry.cc: reduce scope of local variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

test_shared_cache.cc: reduce scope of local variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

TestSignalHandlers.cc: add cppcheck-suppress nullPointer

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

rgw_rest_user.cc: reduce scope of key_type in several places

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

rbd-replay-prep.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

osd/ReplicatedPG.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

osd/PGLog.cc: reduce scope of variable, remove not needed code

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

osd/PG.h: add cppcheck-suppress

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

os/chain_xattr.cc: reduce scope of some variables

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

os/KeyValueStore.cc: reduce scope of variables

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

os/HashIndex.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

mon/OSDMonitor.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

mds/Server.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

mds/MDS.cc: remove unused local variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

ErasureCodeLrc.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

ErasureCodeIsa.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

ErasureCodePlugin.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

cls_rbd_client.cc: reduce scope of variable

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

Merge pull request #2753 from dachary/wip-mailmap

mailmap updates

doc/release-notes: v0.87 giant (draft)

Signed-off-by: Sage Weil <sage@redhat.com>

Merge pull request #2763 from ceph/wip-da-SCA-20141010

Some fixes from SCA runs

Reviewed-by: Sage Weil <sage@redhat.com>

Paxos.cc: use reinterpret_cast instead of c-style cast

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

fuse_ll.cc: use reinterpret_cast instead of c-style cast

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

librados.cc: use static_cast instead of c-style cast

Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>

Merge pull request #2789 from ceph/wip-rgw-explicit-reason

rgw: send http status reason explicitly in fastcgi

Merge remote-tracking branch 'gh/giant'

Revert "Enforce cache size on read requests"

This reverts commit 4fc9fffc494abedac0a9b1ce44706343f18466f1.

Signed-off-by: Sage Weil <sage@redhat.com>

Revert "osdc: Fix ObjectCounter::_readx perf counters"

This reverts commit 830373ef018a78c3fe9d1b9ade208249fe8ccc6a.

Signed-off-by: Sage Weil <sage@redhat.com>

Revert "osdc: Fix compiler warning"

This reverts commit cb290a117ae17961c9e43e5fe76bb3edb2ed7efa.

Signed-off-by: Sage Weil <sage@redhat.com>

Revert "rbd: ObjectCacher reads can hang when reading sparse files"

This reverts commit cdb7675a21c9107e3596c90c2b1598def3c6899f.

Signed-off-by: Sage Weil <sage@redhat.com>

Revert "Fix read performance regression in ObjectCacher"

This reverts commit 65be257e9295619b960b49f6aa80ecdf8ea4d16a.

Signed-off-by: Sage Weil <sage@redhat.com>

.gitmodules: ignoring changes in rocksdb submodule

Signed-off-by: Federico Gimenez <fgimenez@coit.es>
(cherry picked from commit 60eaeca4ddccc79b29b17ad433c6569cb2a89500)

Merge pull request #2797 from ceph/wip-rbd-revert

rbd/objectcacher: revert recent changes for giant

Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Revert "Enforce cache size on read requests"

This reverts commit 4fc9fffc494abedac0a9b1ce44706343f18466f1.

Signed-off-by: Sage Weil <sage@redhat.com>

Revert "rbd: ObjectCacher reads can hang when reading sparse files"

This reverts commit cdb7675a21c9107e3596c90c2b1598def3c6899f.

Signed-off-by: Sage Weil <sage@redhat.com>

Revert "Fix read performance regression in ObjectCacher"

This reverts commit 65be257e9295619b960b49f6aa80ecdf8ea4d16a.

Too late for giant.

Signed-off-by: Sage Weil <sage@redhat.com>

Merge pull request #2768 from ceph/wip-msgr

msgr: merge AsyncMessenger

Lightly reviewed.  Won't impact build or runtime unless explicitly enabled with ms_type = async (or random).

Merge pull request #2795 from ceph/wip-9873

objecter: fix tick_event shutdown race (9873)

Reviewed-by: David Zafman <dzafman@redhat.com>

Merge pull request #2793 from dachary/wip-dead-code-lrc

erasure-code: remove LRC dead code

osdc/Objecter: fix tick_event handling in shutdown vs tick race

If we fail to cancel the tick_event, we rely on tick() itself to clear
tick_event.  I'm not quite sure how we got this wrong in the previous
commit, but this boils down to two cases:

1) shutdown() successfully cancels the event and clears tick_event.  tick()
   never runs.  tick_event == NULL when we finish.
2) shutdown() fails to cancel the event because it has already started.  In
   this case tick itself is blocking (or about to block) waiting on the
   rlock.  When it does run it will clear tick_event itself, then see
   initiazed == 0 and exit without rescheduling.

Fixes: #9873
Signed-off-by: Sage Weil <sage@redhat.com>

common/Timer: recheck stopping before sleep if we dropped the lock

If we have safe_callbacks==false, the stopping flag may have changed while
we were doing our callback. Recheck it and exit to avoid a deadlock on
shutdown.

Signed-off-by: Sage Weil <sage@redhat.com>

erasure-code: remove LRC dead code

Signed-off-by: Loic Dachary <loic-201408@dachary.org>

Merge pull request #2790 from fgimenez/git-ignore-src-rocksdb

.gitmodules: ignoring changes in rocksdb submodule

Merge pull request #2791 from majianpeng/fix1

ECTransaction: Remove unused typedef.

Reviewed-by: Sage Weil <sage@redhat.com>

ECTransaction: Remove unused typedef.

Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>

.gitmodules: ignoring changes in rocksdb submodule

Signed-off-by: Federico Gimenez <fgimenez@coit.es>

Merge pull request #2783 from xanpeng/patch-1

Remove unnecessary expressions about conf_journal_sz

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #2787 from ceph/fix-fstat-mode

java: fill in stat structure correctly

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #2784 from xinxinsh/master

check rocksdb library when '--with-librocksdb' option is enabled

Reviewed-by: Sage Weil <sage@redhat.com>

rgw: send http status reason explicitly in fastcgi

There are issues in certain versions of apache 2.4, where the reason is
not sent back. Instead, just provide the reason explicitly.

Backport: firefly, giant

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>

Merge remote-tracking branch 'gh/giant'

java: fill in stat structure correctly

Added stat filling helper function but only stat and lstat were updated.
This patch makes fstat use it. Crucially the fstat wasn't updating the
mode flags.

Signed-off-by: Noah Watkins <noahwatkins@gmail.com>

Merge pull request #2785 from athanatos/wip-9821

PG:: reset_interval_flush and in set_last_peering_reset

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #2733 from dachary/wip-9372-injectargs

cli: do not parse injectargs arguments twice

Reviewed-by: Dan Mick <dan.mick@inktank.com>

Merge pull request #2759 from dachary/wip-9836-test-mon-id

tests: mon id is incorrectly skipped

Merge pull request #2766 from dachary/wip-9408-buffer-alignment-giant

erasure-code: buffer alignment (giant)

Reviewed-by: Samuel Just <sam.just@inktank.com>

Merge pull request #2781 from zhouyuan/lrc_layer_plugin_doc

EC: document the LRC per layer plugin configuration

Reviewed-by: Loic Dachary <loic-201408@dachary.org>

Remove unnecessary expressions about conf_journal_sz

Signed-off-by: Xan Peng xanpeng@gmail.com

EC: document the LRC per layer plugin configuration

LRC now uses Jerasure as the default EC backend. But it is actually
possible to switch to other backend like Isa using the low level
configuration. This commits Adds documents on how to specify the EC
backend in each LRC layer:

Signed-off-by: Yuan Zhou <yuan.zhou@intel.com>

check rocksdb library when '--with-librocksdb' option is enabled

Signed-off-by: xinxin shu <xinxin.shu@intel.com>

Merge pull request #2767 from derricks/more_verbose_error_output

Make better error reporting for client's call to rados.Rados

Reviewed-by: Loic Dachary <loic-201408@dachary.org>

qa: use sudo when rsyncing /usr so we can read everything

Signed-off-by: Greg Farnum <greg@inktank.com>
(cherry picked from commit fa07c04231db2d130de54647957ffab4a7a53733)

qa: use sudo when rsyncing /usr so we can read everything

Signed-off-by: Greg Farnum <greg@inktank.com>

Merge pull request #2777 from ceph/wip-9859

mon: Monitor: MMonGetMap doesn't require caps

Reviewed-by: Sage Weil <sage@redhat.com>

mon: Monitor: MMonGetMap doesn't require caps

We are dropping the requirement for MON_CAP_R for MMonGetMap.

Reason is simple enough: clients may need to contact the monitors and
obtain the latest monmap before authenticating.  This happens, for
instance, when a client calls MonClient::get_monmap_privately().  The
osd uses this function during mkfs, prior to initializing a keyring or
even so much as existing.

Fixes: #9859
Signed-off-by: Joao Eduardo Luis <joao@redhat.com>

Merge pull request #2770 from ceph/wip-9857

rbd: Correct readahead divide by zero exception

Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Make better error reporting for client's call to rados.Rados

Signed-off-by: Derrick Schneider derrick.schneider@gmail.com

PG:: reset_interval_flush and in set_last_peering_reset

If we have a change in the prior set, but not in the up/acting set, we go back
through Reset in order to reset peering state.  Previously, we would reset
last_peering_reset in the Reset constructor.  This did not, however, reset the
flush_interval, which caused the eventual flush event to be ignored and the
peering messages to not be sent.

Instead, we will always reset_interval_flush if we are actually changing the
last_peering_reset value.

Fixes: #9821
Backport: firefly
Signed-off-by: Samuel Just <sam.just@inktank.com>

Merge remote-tracking branch 'gh/giant'

objecter: Unlock in shutdown before waiting for timer thread

Fixes: #9845
Reviewed-by: Sage Weil <sage@redhat.com>
Signed-off-by: David Zafman <dzafman@redhat.com>

erasure-code: use ErasureCode::SIMD_ALIGN in ceph_erasure_code_benchmark

Signed-off-by: Loic Dachary <loic-201408@dachary.org>

erasure-code: add ErasureCode::encode unit test

Re-create and describe the situation that is fixed by
91a7e18f60bbc9acab3045baaa1b6505474ec4a9 which reworks the buffer
preparation function provided by ErasureCode::encode.

http://tracker.ceph.com/issues/9408 Refs: #9408

Signed-off-by: Loic Dachary <loic-201408@dachary.org>

erasure-code: expose ErasureCode::SIMD_ALIGN as a const

For test purposes and it will also be useful for plugins that must
ensure the chunk size is a multiple of SIMD_ALIGN.

Signed-off-by: Loic Dachary <loic-201408@dachary.org>

ceph_erasure_code_benchmark: use 32-byte aligned input

The benchmark is supposed to measure the encoding/decoding speed and
not the overhead of buffer realignments.

Signed-off-by: Janne Grunau <j@jannau.net>

erasure code: use 32-byte aligned buffers

Requiring page aligned buffers and realigning the input if necessary
creates measurable oberhead. ceph_erasure_code_benchmark is between
10-20% faster depending on the workload.

Also prevents a misaligned buffer when bufferlist::c_str(bufferlist)
has to allocate a new buffer to provide continuous one. See bug #9408

Signed-off-by: Janne Grunau <j@jannau.net>

erasure code: use a function for the chunk mapping index