git://git.whamcloud.com - fs/lustre-release.git/log

LU-12842 utils: llog_print with snapshot name

The lsnapshot utility creates filesystems named with generated
hexadecimal strings. In some cases the filesystem name may start
with a number instead of a character, which causes "lctl llog_print"
(via llog_ioctl()) to consider the filesystem name invalid.

Allow filesystem names in llog_ioctl() to start with a digit.

Lustre-change: https://review.whamcloud.com/36414
Lustre-commit: 6e73b5705c6f1cda391b3a9cec8825eb9f914d38

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: Ib2054d5afbeaa3f661148fff834c29f83f5d98ad
Reviewed-by: Nathaniel Clark <nclark@whamcloud.com>
Reviewed-by: Ben Evans <bevans@cray.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36933
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12595 lnet: Return EHOSTUNREACH for unreachable gateway

Commit 43b35351e9ca258773e89c2d68047e939fb822fb contains a flaw in
that it shouldn't be a fatal error to encounter an unreachable
gateway when parsing routes. Parsing should continue in case there
are any valid, reachable routes that are being added. Returning EINAL
here will cause a failure to load the LNet module. lnet_parse_route()
explicitly allows for lnet_add_route() to return EHOSTUNREACH for
just this purpose.

Lustre-change: https://review.whamcloud.com/35630
Lustre-commit: 7c12c24c8a10be2d0ad005d6d99d97cee6bcde18

Test-parameters: trivial
Fixes: 43b35351e9 ("LU-12411 lnet: Do not allow gateways on remote nets")
Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: Ia0f28779a3505eff02dafdc23a6e01c1d0cbc84b
Reviewed-by: Alexey Lyashkov <c17817@cray.com>
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36871
Tested-by: Maloo <maloo@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12859 llite: clear flock when using localflock

When mounting a client with "-o localflock" or equivalent option in
/etc/fstab, it does not clear out the "flock" mount option flag from
the superblock.  This results in "flock" still being the option used
and it displays both options in the /proc/mounts output:

  10.0.0.1@o2ib:/lfs on /mnt/lfs type lustre (rw,flock,localflock)

Mount a client with both "flock,localflock" as mount options and
verify that the "flock" option is cleared by "localflock", and
vice versa.  Verify that "noflock" clears both options.

Remove the "remount_client()" helper in conf-sanity.sh, since this
shadows a helper function of the same name in test-framework.sh and
is confusing.  Instead, use "mount_client()" now that it can accept
mount options, and just pass "remount" explicitly in a few places.

Lustre-change: https://review.whamcloud.com/36452
Lustre-commit: 22ee4a1f64eca526ef34a3fd89dc4e95bb307732

Fixes: 3613af3e15cb ("LU-10885 llite: enable flock mount option by default")
Test-Parameters: trivial testlist=conf-sanity
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: Ie31b0c4f6674c99d3ed5b73caa39cfc23d3ebbe5
Reviewed-by: Ben Evans <bevans@cray.com>
Reviewed-by: Hongchao Zhang <hongchao@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36934
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12760 tests: stack_trap defaults to sigspec=EXIT

If the "sigspec" argument is not specified for stack_trap(), default
to "EXIT" as the signal, since this is what we use for all callers
of stack_trap() today anyway.

Lustre-change: https://review.whamcloud.com/36186
Lustre-commit: 5a911faae25784a91fc085debeb6dbe8512d80b6

Test-Parameters: trivial
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I2c8d986cdf8743e1d956cd7941a47bd4cd772592
Reviewed-by: James Nunez <jnunez@whamcloud.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36938
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12411 lnet: Do not allow gateways on remote nets

A gateway needs to be reachable over some local interface.

Lustre-change: https://review.whamcloud.com/35198
Lustre-commit: 43b35351e9ca258773e89c2d68047e939fb822fb

Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: Ib66d4f8fd48d8863097280c480648ab8e29d2767
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Reviewed-by: Sonia Sharma <sharmaso@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36870
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12691 ldlm: obd_max_recoverable_clients is not atomic

Originally obd_max_recoverable_clients is not increased
at the same moment. But because of LU-3540,
it will be increased by multiple processes.

The type of obd_max_recoverable_clients should be
atomic_t and be handled by atomic operations.

Lustre-change: https://review.whamcloud.com/35914
Lustre-commit: 01261e7b563adc97899d962f0ba2d1b430894bf7

Signed-off-by: Tatsushi Takamura <takamr.tatsushi@jp.fujitsu.com>
Change-Id: I9a67bbbfacab2e05858243f649e3a4e0d4b5d7f7
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36878
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12025 osp: allow OS_STATE_* flags from OSTs

Allow OS_STATE_* flags to be sent from the OST, so that the
OS_STATE_NOPRECREATE can be used to prevent a newly-added OST
from being used until it is ready.  Add the "no_precreate"
parameter on the OFD that can be set from userspace.

Close a race in the cached opd_statfs.os_state handling in
osp_pre_update_statfs().  It was being overwritten by the
new statfs data from the OST, but was globally visible for a
short time to the precreate threads before the OS_STATE_*
flags were set on the cached statfs data again.

Similarly, there was a race with updating the opd_pre_status
if the OST was out of space, where it would be cleared after
a successful statfs, and wouldn't be set to -ENOSPC until a
short time later.

Split osp_pre_update_status() into osp_pre_update_msfs() that
only copies the statfs data into the cache after all of the
flags are set.  Don't clear flags from the cache, they will
only be cleared when new statfs data is sent.

Add a test that the 'N'OPRECREATE flag appears in "lfs df".

Lustre-change: https://review.whamcloud.com/35029
Lustre-commit: 9b0ebf78f7919a144673edadc4a95bad84fae2d3

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I9c1c7a097f3de8edfdeef2b437f40936e73ebbe5
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-by: Hongchao Zhang <hongchao@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36872
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-11575 build: install systemd stuff only for debian with systemd enabled

Add a precheck for systemd, to avoid to try to package systemd
stuff intolustre-client/server-utils deb when building on debian
series without systemd supported.

Test-Parameters: clientdistro=ubuntu1804 trivial

Lustre-change: https://review.whamcloud.com/33492
Lustre-commit: 02b097440db37fe5e8054f983f8382dfa85f8e25

Change-Id: If58b64acc035e621594ab420a8b900b18a34a211
Signed-off-by: Gu Zheng <gzheng@ddn.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Li Dongyang <dongyangli@ddn.com>
Reviewed-by: Nathaniel Clark <nclark@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36405
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12236 tests: add tests for LNET network namespace

This patch adds tests for LNET for this feature.

Lustre-change: https://review.whamcloud.com/35666
Lustre-commit: b20704d5f63a07c54cfbea331df90e6ca765e79b

Test-Parameters: trivial testlist=sanity-lnet
Signed-off-by: Aurelien Degremont <degremoa@amazon.com>
Change-Id: I2320e5da1beef30be5dcca9529fa838fc9304876
Reviewed-on: https://review.whamcloud.com/36770
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12236 lnet: support non-default network namespace

Replace hard coded references to default root network namespace
(&init_net) in LNET code (LNET, socklnd and o2iblnd).

When a network interface is created, Lustre records the current
network namespace. This patch improves the LNET code to use
this reference namespace most of the time instead of the root
network namespace. When using lctl, lnetctl or insmod, we
use the current process network namespace.
When starting the listening acceptor, we use the namespace of the
process that triggers this start.

An additional patch is needed for RPCSEC GSS support.

Lustre-change: https://review.whamcloud.com/34768
Lustre-commit: 93b08edfb1c6ae8aec7e1009d3aca450416358d7

Signed-off-by: Aurelien Degremont <degremoa@amazon.com>
Change-Id: I56877ddcd7a27883662c86f245b196153211e7b2
Reviewed-on: https://review.whamcloud.com/36769
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11670 tests: do not fail the first half in sanityn test 103

There are two halves in sanityn test 103. The first half is to
reproduce the problem of incorrect size when using lockahead
and the second half is to verify that the fix works. Sometimes,
the problem cannot be reproduced in the first half test, so we
should not fail the whole test.

Test-Parameters: trivial fstype=zfs \
mdscount=2 mdtcount=4 testlist=sanityn,sanityn,sanityn

Test-Parameters: trivial fstype=ldiskfs \
mdscount=2 mdtcount=4 testlist=sanityn,sanityn,sanityn

Lustre-change: https://review.whamcloud.com/36303
Lustre-commit: a88d0aa76c62e3074a05e869c9f0ba7ac128300f

Change-Id: Ib6c82bfe512ac072104abfcb406e2ef1bd6a6a02
Signed-off-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Patrick Farrell <farr0186@gmail.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36407
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11670 osc: glimpse - search for active lock

When there are lock-ahead write locks on a file, the server
sends one glimpse AST RPC to each client having such (it
may have many) locks. This callback is sent to the lock
having the highest offset.

Client's glimpse callback goes up to the clio layers and
gets the global (not lock-specific) view of size.  The clio
layers are connected to the extent lock through the
l_ast_data (which points to the OSC object).

Speculative locks (AGL, lockahead) do not have l_ast_data
initialised until an IO happens under the lock. Thus, some
speculative locks may not have l_ast_data initialized.

It is possible for the client to do a write using one lock
(changing file size), but for the glimpse AST to be sent to
another lock without l_ast_data initialized.  Currently, a
lock with no l_ast_data set returns ELDLM_NO_LOCK_DATA to
the server.  In this case, this means we do not return the
updated size.

The solution is to search the granted lock tree for any lock
with initialized l_ast_data (it points to the OSC object
which is the same for all the extent locks) and to reach the
clio layers for the size through this lock instead.

Lustre-change: https://review.whamcloud.com/33660
Lustre-commit: b3461d11dcb04670cc2e1bfbb99306cfd3f645ef

cray-bug-id: LUS-6747
Signed-off-by: Patrick Farrell <pfarrell@whamcloud.com>
Change-Id: I6c60f4133154a3d6652315f155af24bbc5752dd2
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36406
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12103 ldiskfs: don't search large block range if disk full

Block allocator tries to find:
1) group with the same range as required
2) group with the same average range as required
3) group with required amount of space
4) any group

For quite full disk step 1 is failed with higth
probability, but takes a lot of time.

Skip 1st step if disk space < 25%
Skip 2d step if disk space < 15%
Skip 3d step if disk space < 5%
Also check if group has any free space on step 4.

This three thresholds can be adjusted through added interface.

Variables added which counts unsuccessfull group processing loops.
This can show allocator effectiveness in different circumstances.

This statistics output through mb_alloc file. This file is
useful to track allocator activity.

Lustre-change: https://review.whamcloud.com/35180
Lustre-commit: 95f8ae5677491508ae7182b4f61ead3d413434ae

Signed-off-by: Artem Blagodarenko <c17828@cray.com>
Change-Id: I18c7147e32951c49e12a2444803aa2995bb4ae2d
Cray-bug-id: LUS-6746
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36681
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12568 lnet: Defer rspt cleanup when MD queued for unlink

When an MD is queued for unlink its lnet_libhandle is invalidated so
that future lookups of the MD fail. As a result, the monitor thread
cannot detach the response tracker from such an MD, and instead must
wait for the remaining operations on the MD to complete before it can
safely free the response tracker and remove it from the list. Freeing
the memory while there are pending operations on the MD can result
in a use after free situation when the final operation on the MD
completes and we attempt to remove the response tracker from the MD
via the lnet_msg_detach_md()->lnet_detach_rsp_tracker() call chain.

Here we introduce zombie lists for such response trackers. This will
allow us to also handle the case where there are response trackers
on the monitor queue during LNet shutdown. In this instance the
zombie response trackers will be freed when either all the operations
on the MD have completed (this free'ing is performed by
lnet_detach_rsp_tracker()) or after the LND Nets have shutdown since
we are ensured there will not be any more operations on the
associated MDs (this free'ing is performed by
lnet_clean_zombie_rstqs()).

Three other small changes are included in this patch:
- When deleting the response tracker from the monitor's list we
   should use list_del() rather than list_del_init() since we'll
   be freeing the response tracker after removing it from the list.
- Perform a single ktime_get() call for each local queue.
- Move the check of whether the local queue is empty outside of
   the net lock.

Signed-off-by: Chris Horn <hornc@cray.com>
Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: If7e3fa53ae8585fb3e0e4aed29f0e1d97e85d356
Reviewed-on: https://review.whamcloud.com/36635
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Serguei Smirnov <ssmirnov@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12441 lnet: Detach rspt when md_threshold is infinite

MDs for pings use the infinite threshold on MD operations.
As such they aren't normally unlinkable as determined by
lnet_md_unlinkable(). We can cover this case by checking whether the
refcount is zero and threshold is LNET_MD_THRESH_INF.

Signed-off-by: Chris Horn <hornc@cray.com>
Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: Ib51f84d85005dd2d13dadca059a1d6c42ff3bf59
Reviewed-on: https://review.whamcloud.com/36634
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Serguei Smirnov <ssmirnov@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12530 utils: narrow l_tunedisk udev rule

Narrow the udev rule so that it runs l_tunedisk only for ext4 block
devices formatted for Lustre.

Devices which are members of ZFS pools do not need such tunings to
be provided by lustre - they are handled by ZFS.

There are currently no other OSD types in the tree. Sites/Vendors which
support other OSDs will need to adjust the rule appropriately.

Lustre-change: https://review.whamcloud.com/36599
Lustre-commit: 7b2cb54858daa60d560fd6374c4ecba552a10d27

Change-Id: Iba8b20fc705da0259ab71ee33b92193cae7e8eae
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Nathaniel Clark <nclark@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36776
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12703 utils: reset rootpath in llapi_search_rootpath()

as get_root_path() can use it as a source and fail if
passed pathname contains garbage (on stack);

Lustre-change: https://review.whamcloud.com/36335
Lustre-commit: 3e2e0025d1e929763f9d4de48746c3433d3684d5

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I9f628353c872afc82a582b0a6ca960cd0e8cffcb
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36482
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1365 utils: allow set block size for ldiskfs backend

Add “-b” option to mkfs.lustre that allows to set backend block size.

Lustre-change: https://review.whamcloud.com/33757
Lustre-commit: 5f674667bfd1ab9a0e47d9f03f3e7eab37eb8e17

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Artem Blagodarenko <c17828@cray.com>
Change-Id: I83fc76f64ce2a0b4bf500841b695d64d3dea78de
Reviewed-by: Li Dongyang <dongyangli@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36778
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12593 osd: zeroing a freshly allocated block buffer

Ldiskfs zeroes new buffer only when it is not uptodate.
In rare case we can get a new buffer head with uptodate flag.
This may cause a file corruption for non zero offset writes,
especially for internal Lustre files like update_log, CATALOGS,
lov_objid.

od_fld_lookup()) lustre-MDT0001-mdtlov: invalid FID [0x0:0x50:0x0]

The patch adds zeroing under i_mutex for unmaped blocks.

The performance results, since the patch adds mutex to a creation
path (lov_objid file).
40 tasks, 2000000 files
SUMMARY: (of 5 iterations)
Operation       Max           Min           Mean    Std Dev
---------       ---           ---           ----    -------
without fix
File creation: 39990.601   19020.238     27443.823  6909.605
With fix
File creation: 37958.809   21708.187     27065.855  5900.961

Lustre-change: https://review.whamcloud.com/35629
Lustre-commit: f832a7dc33c69fae9af199f0317e6385deeaeccf

Cray-bug-id: LUS-6132
Signed-off-by: Alexander Boyko <c17825@cray.com>
Change-Id: Ica8fbe29b5a7253d553b41a41ffe5d8d8b4b2e55
Reviewed-by: Shaun Tancheff <stancheff@cray.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36709
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12893 lnet: fix peer_ni selection

When selecting a peer-ni we must use the same peer NID
through all the messages which belong to the same RPC.
This is necessary in order to ensure we do the RDMA over
the optimal interface.

Lustre-change: https://review.whamcloud.com/36552
Lustre-commit: 94ee26738884e3f5b241698bc2e7a8da9702d264

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I0391537da32bc6ac7a8a3d92e207bf172d111981
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Serguei Smirnov <ssmirnov@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36643
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-4398 llite: do not cache write open lock for exec file

This is to avoid the problem that the MDT needs an extra lock
revocation to make the file be able to execute.

Lustre-change: https://review.whamcloud.com/32265
Lustre-commit: 6dd9d57bc006a37731d34409ce43de13c192e0cc

Signed-off-by: Jinshan Xiong <jinshan.xiong@uber.com>
Signed-off-by: Gu Zheng <gzheng@ddn.com>
Change-Id: Ibb42a9a8cb56a9bf48a6e972b72d3d71ed7fbaf5
Reviewed-by: Lai Siyao <lai.siyao@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36680
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12533 llite: Improve readahead RPC issuance

lov_io_submit receives a range of pages, then adds pages in
to a batch until it hits a page which is not in the stripe
associated with this lov object.  This means that if a
readahead page range hits the same stripe more than once,
we will issue multiple I/Os, even if the pages would fit in
one RPC.

This is unnecessary - Just submit all these pages at once.

mpirun -n 2 $IOR -s 2000 -t 47K -b 47K -k -r -E -o $FILE

Without patch:
osc.lustre-OST0001-osc-ffff8fe82c952000.rpc_stats=

                        read                    write
pages per rpc         rpcs   % cum % |       rpcs   % cum %
1:                     118  56  56   |          0   0   0
2:                       0   0  56   |          0   0   0
4:                       0   0  56   |          0   0   0
8:                       0   0  56   |          0   0   0
16:                      5   2  58   |          0   0   0
32:                      0   0  58   |          0   0   0
64:                      0   0  58   |          0   0   0
128:                    21  10  68   |          0   0   0
256:                    25  11  80   |          0   0   0
512:                    10   4  85   |          0   0   0
1024:                   31  14 100   |          0   0   0

osc.lustre-OST0002-osc-ffff8fe82c952000.rpc_stats=
                        read                    write
pages per rpc         rpcs   % cum % |       rpcs   % cum %
1:                       5   6   6   |          0   0   0
2:                       0   0   6   |          0   0   0
4:                       0   0   6   |          0   0   0
8:                       0   0   6   |          0   0   0
16:                      0   0   6   |          0   0   0
32:                      0   0   6   |          0   0   0
64:                      0   0   6   |          0   0   0
128:                    19  23  29   |          0   0   0
256:                    19  23  52   |          0   0   0
512:                     5   6  58   |          0   0   0
1024:                   34  41 100   |          0   0   0

With patch:
osc.lustre-OST0001-osc-ffff8fe7a7227800.rpc_stats=
                        read                    write
pages per rpc         rpcs   % cum % |       rpcs   % cum %
1:                      12  17  17   |          0   0   0
2:                       0   0  17   |          0   0   0
4:                       0   0  17   |          0   0   0
8:                       0   0  17   |          0   0   0
16:                      5   7  24   |          0   0   0
32:                      0   0  24   |          0   0   0
64:                      5   7  31   |          0   0   0
128:                     6   8  40   |          0   0   0
256:                     1   1  42   |          0   0   0
512:                     2   2  44   |          0   0   0
1024:                   38  55 100   |          0   0   0

osc.lustre-OST0002-osc-ffff8fe7a7227800.rpc_stats=
                        read                    write
pages per rpc         rpcs   % cum % |       rpcs   % cum %
1:                       0   0   0   |          0   0   0
2:                       0   0   0   |          0   0   0
4:                       0   0   0   |          0   0   0
8:                       0   0   0   |          0   0   0
16:                      0   0   0   |          0   0   0
32:                      0   0   0   |          0   0   0
64:                      4   7   7   |          0   0   0
128:                     7  13  21   |          0   0   0
256:                     0   0  21   |          0   0   0
512:                     3   5  26   |          0   0   0
1024:                   38  73 100   |          0   0   0

Note the much larger # of smaller RPC issued without the patch.

Lustre-change: https://review.whamcloud.com/35458
Lustre-commit: 05b9da4fd124c61fd41d4b560773c0552a1ee5d7

Signed-off-by: Patrick Farrell <pfarrell@whamcloud.com>
Change-Id: Ic10c138628c269afe57fbc57ec8c91ce990717f9
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36342
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12844 lnet: fix strncpy bound error

This patch fixes the following error while using gcc 8:

liblnetconfig.c: In function ‘lustre_lnet_parse_nids’:
liblnetconfig.c:320:3: error: ‘strncpy’ specified bound depends on
the length of the source argument [-Werror=stringop-overflow=]
   strncpy(entry, cur, len - 1);
   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
liblnetconfig.c:310:10: note: length computed here
    len = strlen(cur) + 1;
          ^~~~~~~~~~~
cc1: all warnings being treated as errors

This patch is back-ported from the following one:
Lustre-commit: cebda7a478f9943f10b9a3388377c61a54957a87
Lustre-change: https://review.whamcloud.com/36417

Change-Id: I2d5840fd58c7b7d27ef1b2aa12f1f187d30abbfd
Signed-off-by: Jian Yu <yujian@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36418
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-12745 build: Account for optional SPL for ZFS 0.8+

With ZFS 0.8.0 and later, SPL is not longer present.
Some zfs packages provide vestigial spl package contents, but zfs-dkms
does not. This makes testing SPL directories optional depending on
version of ZFS, this also accounts for the new location of the spl
include directory under the zfs include directory.

Lustre-change: https://review.whamcloud.com/36161
Lustre-commit: a245dde23a9fdbdff7d09a783bcbe3349f68a908

Test-Parameters: trivial
Signed-off-by: Nathaniel Clark <nclark@whamcloud.com>
Change-Id: I8afcff079f25543a3c86df0c404146a859b226aa
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36408
Reviewed-by: James Simmons <jsimmons@infradead.org>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-12925 test: assign right initial value for test_61

This patch snip from commit:591a9b4cebc510ff5. The test_62
would be failed since test_61 leave a failover state in some
case.

Signed-off-by: Patrick Farrell <pfarrell@whamcloud.com>
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Change-Id: If46a6d435bcaafb9000abb032ac561c5453776ee
Reviewed-on: https://review.whamcloud.com/36660
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12803 libcfs: bump module version

The linux client version of libcfs is further ahead in its
cleanup so its module version is higher. While this is the
case it does prevent the OpenSFS version of libcfs from
loading and since OpenSFS is current ahead of the linux
client we prefere to use it at this time. Lets just increase
the OpenSFS libcfs module to be just slightly ahead of the
linux client.

Test-Parameters: trivial

Lustre-change: https://review.whamcloud.com/36488
Lustre-commit: 4b25d733342bc6f3a424ecfb0db80f1c175a8986

Change-Id: Ie57d93529bf25d908099f7dab06d2960f9923d58
Signed-off-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Shaun Tancheff <stancheff@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36642
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12328 flr: avoid reading unhealthy mirror

* Fix an error in lov_io_mirror_init() which would wait unnecessarily
  if we're retrying the last mirror of the file.

* In osc_io_iter_init() we'd check its OSC import status so that the
  read path can quickly switch another mirror.
  sanity-flr test_33b is added to test this case.

* And with all mirrors have been tried, we'd turn off the quick switch
  so that when all mirrors contain bad OSTs, the read will still try
  its best to get partial data from a component before trying another
  mirror.
  sanity-flr test_33c is added to test this case.

Lustre-change: https://review.whamcloud.com/34952
Lustre-commit: 39da3c06275e04e2a6e7f055cb27ee9dff1ea576

Test-Parameters: envdefinitions=ONLY="33" testlist=sanity-flr,sanity-flr,sanity-flr,sanity-flr,sanity-flr,sanity-flr,sanity-flr,sanity-flr,sanity-flr,sanity-flr
Fixes: 5a6ceb664f07 ("LU-7236 ptlrpc: idle connections can disconnect")
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: I5621a834e58ee1bfccf6c407d2c68357b5c3eb3b
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36550
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11221 osd: allow concurrent bulks from pagecache

drop page lock earlier, once IO is complete so that page can be
read by few clients simultanously.

Lustre-change: https://review.whamcloud.com/33521
Lustre-commit: 0a92632538d8c985e024def73512d18d1570d5ca

Change-Id: Iee28f578e937744f07f7c5be7eae99e59e625e6e
Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36570
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11367 som: integrate LSOM with lfs find

The patch integrates LSOM functionality with lfs find so that it
is possible to use LSOM functionality directly on the client. The
MDS fills in the mbo_size and mbo_blocks fields from the LSOM
xattr, if the actual size/blocks are not available, and then set
new OBD_MD_FLLSIZE and OBD_MD_FLLBLOCKS flags in the reply so that
the client knows these fields are valid.

The lfs find command adds "-l|--lazy" option to allow the use of
LSOM data from the MDS.

Add a new version of ioctl(LL_IOC_MDC_GETINFO) call that also returns
valid flags from the MDS RPC to userspace in struct lov_user_mds_data
so that it is possible to determine whether the size and blocks are
returned by the call. The old LL_IOC_MDC_GETINFO ioctl number is
renamed to LL_IOC_MDC_GETINFO_OLD and is binary compatible, but
newly-compiled applications will use the new struct lov_user_mds_data.

New llapi interfaces llapi_get_lum_file(), llapi_get_lum_dir(),
llapi_get_lum_file_fd(), llapi_get_lum_dir_fd() are added to fetch
valid stat() attributes and LOV info to the user.

Lustre-change: https://review.whamcloud.com/35167
Lustre-commit: 11aa7f8704c490b011f60f234c3ac9929ce76948

Signed-off-by: Qian Yingjin <qian@ddn.com>
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I21dfae7c2633dead5d83b438ec340fea4d3ebbe5
Reviewed-by: Li Xi <lixi@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36553
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12824 o2ib: Record rc in debug log on startup failure

Since kiblnd_startup() return -ENETDOWN on failure, let's record the
rc value for the failure case in the debug log.

Lustre-change: https://review.whamcloud.com/36325
Lustre-commit: 99f85541a685df82265f18167e91c161c523ce50

Cray-bug-id: LUS-7935
Test-Parameters: trivial
Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: Ied934642bc567b8d3f51293d7dd095d47ff134df
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36547
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12824 o2ib: Fix whitespace in kiblnd_startup

Convert whitespace to tabs where appropriate in kiblnd_startup()

Lustre-change: https://review.whamcloud.com/36324
Lustre-commit: 50300e83e4cab3157149107eb735825cc4c3aff1

Cray-bug-id: LUS-7935
Test-Parameters: trivial
Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: I11aaaa8e47d754b219fb773d74e37190476e4eeb
Reviewed-by: Shaun Tancheff <stancheff@cray.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Neil Brown <neilb@suse.de>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36546
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12824 o2ib: Reintroduce kiblnd_dev_search

If we add an interface to multiple nets then we need to re-use the
struct ib_dev object for each of the nets.

Lustre-change: https://review.whamcloud.com/36326
Lustre-commit: e25e45c612a061031e8b4b5233137fbb57b50cc4

Cray-bug-id: LUS-7935
Fixes: 75ab841 ("LU-11893 lnet: consoldate secondary IP address handling")
Test-Parameters: trivial
Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: I1790e24458f47d632fd137b78de076d408fe5260
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36545
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12355 llite: vfs atomic_open change with FMODE_CREATED

Kernel 4.19 introduced FMODE_CREATED and switched to it while
the last argument to vfs atomic_open was removed and the f_mode
flags are used to indicate the created state on return.

Linux-commit: 73a09dd94377e4b186b300bd5461920710c7c3d5

Lustre-change: https://review.whamcloud.com/35020
Lustre-commit: 4decb4c2da6053066f10cbe419e2db212de8e4aa

Test-Parameters: trivial
Change-Id: I26d4aadb123bb1d1bc0aa1d78a64a75b94276ffb
Signed-off-by: Shaun Tancheff <stancheff@cray.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Thomas Stibor <t.stibor@gsi.de>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36415
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12734 misc: add bash completion for lctl set/get_param

Add some start of bash completion for lctl, mainly set_param and
get_param, and modify build system to install it.

Lustre-change: https://review.whamcloud.com/36105
Lustre-commit: f87a7f2656ceff174a00933a170032f093ecc72d

Test-Parameters: trivial
Signed-off-by: Dominique Martinet <dominique.martinet@cea.fr>
Change-Id: I16d2698e782702375c7fa3edf3bfde2e3b197297
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Quentin Bouget <quentin.bouget@cea.fr>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36483
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-11526 rpc: support maximum 64MB I/O RPC

On newer systems, some block drivers allow max_hw_sector_kb to
be up to 65536KB (64MB) to the underlying storage. To maximize
driver efficiency, Lustre should also have bump up maximum I/O
RPC size to 64MB.
Clamp max_read_ahead_whold_mb not to exceed
max_read_ahead_per_file_mb

Lustre-change: https://review.whamcloud.com/34042
Lustre-commit: 1a9be0046b1f1772d3f934c2146dc5233c391377

Signed-off-by: Qian Yingjin <qian@ddn.com>
Change-Id: Icbf78742f8210d82dc310af7d05b7c32b93af34f
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/35369
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12043 llite,readahead: don't always use max RPC size

Since 64M RPC landed, @PTLRPC_MAX_BRW_PAGES will be 64M.
And we always try to use this max possible RPC size to check
whether we should avoid fast IO and trigger real context IO.

This is not good for following reasons:

(1) Since current default RPC size is still 4M,
most of system won't use 64M for most of time.

(2) Currently default readahead size per file is still 64M,
which makes fast IO always run out of all readahead pages
before next IO. This breaks what users really want for readahead
grapping pages in advance.

To fix this problem, we use 16M as a balance value if RPC smaller
than 16M, patch also fix the problem that @ras_rpc_size could not
grow bigger which is possibe in the following case:

1) set RPC to 16M
2) Set RPC to 64M

In the current logic ras->ras_rpc_size will be kept as 16M which is wrong.

Lustre-change: https://review.whamcloud.com/35033
Lustre-commit: 7864a6854c3dfe3319dcf6809e728eed9a37b9b4

Change-Id: Ida9f839f7c692cd88d32dc0909503f6ae991d909
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/35559
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11933 mdt: clear sp_cr_flags in migrate unpack

mdt_thread_info.mti_spec is not cleared after operation handling, so
mdt_migrate_unpack() should clear it in case the old values are used.

Lustre-change: https://review.whamcloud.com/36154
Lustre-commit: d4da3b55a8303d937828e74341b3ab5c4dfd52b2

Signed-off-by: Lai Siyao <lai.siyao@whamcloud.com>
Change-Id: Ib3d5d39a4a072621c8da8b6ef7869cb4d8178aac
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-on: https://review.whamcloud.com/36399
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11626 mdc: hold obd while processing changelog

During read/write changelog, the corresponding obd_device should
be held to protect it from being released by umount.

Lustre-change: https://review.whamcloud.com/35784
Lustre-commit: d7bb6647cd4dd26949bceb6a099cd606623aff2b

Change-Id: Ib5b528f178edcf73425587ea60335df640c1696d
Signed-off-by: Hongchao Zhang <hongchao@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Emoly Liu <emoly@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36338
Reviewed-by: James Simmons <jsimmons@infradead.org>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-11743 utils: allow lctl pool_list on separate MGS

Change lctl pool_list command to parse the configuration log directly
when run on a standalone MGS node. This also allows the pool commands
to be run when only the MGS is started.

Also, those test scripts from the patch of LU-9899 to mount a client
on the standalone MGS to allow OST pools to work properly are cleared.

Lustre-change: https://review.whamcloud.com/35895
Lustre-commit: d908fe9686bc1e583da7434856d9c06e6cbbc4fd

Change-Id: Ic25931d49c2cf747da2a3f2ac3c25a21f6878991
Test-Parameters: standalonemgs=true testlist=ost-pools.sh,sanity.sh,conf-sanity.sh
Signed-off-by: Emoly Liu <emoly@whamcloud.com>
Signed-off-by: Vladimir Saveliev <c17830@cray.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36314
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12131 tests: properly handle GSS in server failover

In case of server failover, a number of aspects must be handled when
GSS based features (SSK or Kerberos) are activated:
- lsvcgssd daemon must be restarted;
- targets must be mounted with proper skpath option;
- permissions on keys must be adjusted.
When service is initially started, all that is managed in setupall().
fail() and facet_failover() have to be improved to take GSS aspects
into account.

Lustre-change: https://review.whamcloud.com/35041
Lustre-commit: 1cbfb44fb59945da62acbb672330fde5c75ddc98

Test-Parameters: envdefinitions=SHARED_KEY=true testlist=sanity,recovery-small,sanity-sec
Signed-off-by: Sebastien Buisson <sbuisson@ddn.com>
Change-Id: I8db686f406629c7eec655496cf83c0539c1bfb33
Reviewed-on: https://review.whamcloud.com/35534
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11204 obdclass: remove unprotected access to lu_object

The check of lu_object_is_dying() is done after reference
drop and without lock, so can access freed object if concurrent
thread did final put.

The patch saves object state right before atomic_dec_and_lock()
and checks it after check, so object is not being accessed

Lustre-change: https://review.whamcloud.com/34960
Lustre-commit: 336cf0f2f3a9ce5b11a34aeaeec062a5d5144213

Signed-off-by: Mikhail Pershin <mpershin@whamcloud.com>
Change-Id: I926991f465e7913e5fc150095425bfb5bf07f57f
Reviewed-on: https://review.whamcloud.com/36217
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12719 obdclass: serialize lwp list access

lustre_sb_info.lsi_lwp_list should be acessed with lock, and
some place may sleep, change lsi_lwp_lock from spinlock to mutex.

Lustre-change: https://review.whamcloud.com/36003
Lustre-commit: 875252d59924ad09db8de9f0fbb611788a0b9c78

Signed-off-by: Lai Siyao <lai.siyao@whamcloud.com>
Change-Id: Ifc3622eb28cd6cf49661b14fc10e98aa689a58dc
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36349
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12622 tests: skip sanity test_816 with SSK

sanity test_816 is incompatible with SSK, so skip it
in case SHARED_KEY is true.

Lustre-change: https://review.whamcloud.com/35664
Lustre-commit: b4ab1084cc0f1947ab65c32075d579edac5ab547

Whamcloud-bug-id: ATM-1283
Test-Parameters: trivial
Signed-off-by: Sebastien Buisson <sbuisson@ddn.com>
Change-Id: I019cf6a4bab8b0cf9825b7c49f364225d5156dfa
Reviewed-on: https://review.whamcloud.com/36401
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: James Nunez <jnunez@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11157 obd: round values to nearest MiB for *_mb syfs files

Several sysfs files report their settings with the functions
lprocfs_[seq]_read_frac_helper() which has the intent of showing
fractional values i.e 1.5 MiB. This approach has caused problems
with shells which don't handle fractional representation and the
values reported don't faithfully represent the original value the
configurator passed into the sysfs file. To resolve this lets
instead always round up the value the configurator passed into
the sysfs file to the nearest MiB value. This way it is always
guaranteed the values reported are always exactly some MiB value.

Lustre-change: https://review.whamcloud.com/34317
Lustre-commit: ba2817fe3ead1b8e32be6d6c6ce25b490626118a

Change-Id: Ia2e8cf8421784853aa33d4bb87c54aee00953835
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Ben Evans <bevans@cray.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36393
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12674 osp: handle -EINPROGRESS on llog objects

if llog object is corrupted and OI doesn't allow access to that
OSP panics being asked to declare new llog record (e.g. for unlink).
Instead OSP should complain in the logs, skip llogging and suggest
to run LFSCK to fix orphans.

Lustre-change: https://review.whamcloud.com/35844
Lustre-commit: a3ec8ff69fceb53a467a80c2e6008869f25f72b4

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I18d4d68811833c08cdc1937d147ac6e8c3408a30
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Reviewed-by: Lai Siyao <lai.siyao@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36348
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-11967 mdt: reint layout_change in standard way

Layout_change is a reint operation, and it should be handled the
same as other reint operations, so that resent and replay can
work correctly.

Also replace the lock passed in ldlm_handle_enqueue0() with the
lock taken in mdt_layout_change(). This avoids taking lock again
in ldlm_handle_enqueue0(), and also makes replay eaiser. Note,
before replacing, the mode is downgraded from EX to CR, because
client only needs this mode, as can avoid unnecessary lock cancel
later.

Add missing resent reconstructor for REINT_RESYNC.

Lustre-change: https://review.whamcloud.com/35465
Lustre-commit: e1bebbdf53c8490a3be35793070f45f1a68721b1

Signed-off-by: Lai Siyao <lai.siyao@whamcloud.com>
Change-Id: I328044dacbf18d03232c9bbb51271f6202e9b939
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36376
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11739 lod: subdir under ROOT should honor default layout

Though sub directories under ROOT don't inherit ROOT default
layout, they should hornor ROOT default layout in creation.

Add sub test in sanity.sh 65n to verify this.

Fixes: 0a988cae95 ("LU-11739 lod: don't inherit default
layout from root directory")

Lustre-change: https://review.whamcloud.com/35204
Lustre-commit: 693fb63ac777eab426f1b618316a5649534759ad

Signed-off-by: Lai Siyao <lai.siyao@whamcloud.com>
Change-Id: I1edf91da7944a8871652df7bca2104d00f66163a
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36370
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11768 test: limit at_max to timeout in time

In test_6 of sanity-quota, if the AT is enabled, the timeout of
the QUOTA_DQACQ request could be longer than OBD_TIMEOUT*2, which
cause the watchdog to be triggered.

This is a backport to Lustre b2_12 of
Lustre-change: https://review.whamcloud.com/35651
Lustre-commit: d8226b9353dbc1448af8d23c13cae5f21cbe3a86

Change-Id: I7e3a976a004259f5c956fc48f4d8d63c751ee2c0
Signed-off-by: Hongchao Zhang <hongchao@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36365
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12639 tests: initialize variable sanity 317

sanity test 317 checks the file system type of $facet,
but facet is not initiaized in the test.

Replace the check of the facet file system type with the
variable ost1_FSTYPE. The call to return after skip is
removed.

This is a backport to Lustre b2_12 of
Lustre-change: https://review.whamcloud.com/35716
Lustre-commit: c00be06a1f8f27eb5bd8bb47086d0f1e5b5f5f50

Fixes: 6115eb7fd55a ("LU-10370 ofd: truncate does not update blocks count on client")
Test-Parameters: trivial
Test-Parameters: fstype=zfs envdefinitions=ONLY=317 testlist=sanity

Signed-off-by: James Nunez <jnunez@whamcloud.com>
Change-Id: If67c4d786e4d23752effd1aaebc82bb1be8aceb5
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Arshad Hussain <arshad.super@gmail.com>
Reviewed-on: https://review.whamcloud.com/36366
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11239 lfs: fix mirror resync error handling

This patch returns error for partially successful mirror resync.

Lustre-change: https://review.whamcloud.com/33537
Lustre-commit: 0f670d1ca9dd5af697bfbf3b95a301c61a8b4447

Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: I9d6c9ef5aca1674ceb7a9cbc6b790f3f7276ff5d
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36341
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12100 tests: Use minimum soft qunit limits

Ensure that we don't create limits that are too small, which would
cause all writes to fail.
Wait for grace period to timeout.

Lustre-change: https://review.whamcloud.com/35667
Lustre-commit: 37e28b7e05a5b1f77fe663f9407436aea312b3b2

Test-Parameters: trivial
Test-Parameters: testlist=sanity-quota
Test-Parameters: testlist=sanity-quota fstype=zfs
Signed-off-by: Nathaniel Clark <nclark@whamcloud.com>
Change-Id: I9342272615ca9c252fcc7f77ed8a61030fc9672a
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36346
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12707 obdecho: avoid panic with partially object init

in some cases (like ENOMEM) init function can't called, so
any init related code should placed in the object delete handler,
not in the object free.

Lustre-change: https://review.whamcloud.com/35950
Lustre-commit: 1a9ca8417c60f04a9aa719b7254372e2d18a6b0a

Signed-off-by: Alexey Lyashkov <c17817@cray.com>
Change-Id: I1fca56423de9a045aac2c495fbc45069c3bbc97c
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36317
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

New release 2.12.3

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I0901824abf874ecf6124907adad23fbc486cab07

New RC 2.12.3-RC1

Change-Id: I24e7f7622654332ce3e29a292ce87591cd0e09be

LU-12553 mdc: polling mode for changelog reader

this allows the user (like lsom_sync and similar) to follow
the changelog and don't rescan getting duplicates.

Lustre-change: https://review.whamcloud.com/35262
Lustre-commit: e215002883d5620f43615013452935da8e7e3f8c

Change-Id: I78dc163838c1b88f9447a4731ad4bfe00fec7eff
Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36362
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11426 llog: changelog records reordering

Changelog records can get reordered because of a race
window between cr_index generation and llog file
space allocation. This can lead to llog records
loss.

llog_write() holds loghandle->lgh_lock semaphore,
so it seems an appropriate place to generate a
new changelog index.

Lustre-change: https://review.whamcloud.com/36187
Lustre-commit: 1fa0a984c5c3863d8f40b3b0d63c3d08cfa1a9f0

Change-Id: I034d1a696bde1d0f780e494ab65073e4018ceec9
Signed-off-by: Andrew Perepechko <c17827@cray.com>
Reviewed-by: Alexander Boyko <c17825@cray.com>
Reviewed-by: Alexander Zarochentsev <c17826@cray.com>
Cray-bug-id: LUS-7691
Reviewed-by: Alexandr Boyko <c17825@cray.com>
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Reviewed-by: Yingjin Qian <qian@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36316
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>

LU-12790 obdclass: print jobid error message properly

Modify unlikely() condition to print error message properly when
(rc == -EOVERFLOW).

Lustre-change: https://review.whamcloud.com/36272
Lustre-commit: df21a3b9eb01621de92940e441cd557913d1cd05

Signed-off-by: Emoly Liu <emoly@whamcloud.com>
Change-Id: I19bfb353c71b55a0dfb6eec78c1af915494acd71
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Ben Evans <bevans@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36378
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-11756 o2iblnd: kib_conn leak

A new tx can be queued while kiblnd_finalise_conn()
aborts txs. Thus a reference from new tx will
prevent connection from moving into kib_connd_zombies.

Insert new tx after IBLND_CONN_DISCONNECTED into
ibc_zombie_txs list and abort it during
kiblnd_destroy_conn().

Lustre-change: https://review.whamcloud.com/33828
Lustre-commit: a155c3fca38d2a3092f9b5d116ad7877d51d1db1

Change-Id: Ib92d8d02e6e3f66f7140041a330fc00b7ad44ae3
Cray-bug-id: LUS-6412
Signed-off-by: Andriy Skulysh <c17819@cray.com>
Reviewed-by: Alexey Lyashkov <c17817@cray.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36347
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12802 tests: speedup cleanup of racer

After racer test survives for a given time, it starts to cleanup.
And the parent racer.sh script waits the child racer/racer.sh
to exit. However sometimes, somehow, this stucks for a long time.
Sending a signal to remaining dd(or other) processes will wake up
the wait in parent racer.sh script immediately.

Lustre-change: https://review.whamcloud.com/36289
Lustre-commit: fcf219db6d2cfb692f0b987945e85953a5b07de7

Test-Parameter: trivial testlist=racer
DDN-Bug-ID: DDN-256
Signed-off-by: Li Xi <lixi@ddn.com>
Reviewed-by: Gu Zheng <gzheng@ddn.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Change-Id: I2ff2784b76faa0532c39af29b1586a48f2b90a21
Reviewed-by: Shilong Wang <wshilong@ddn.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36381
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12777 test: fix to pass facet to facet_fstype

Function facet_fstype() expect mgs1 mds1 etc as its
argument, and we used it wrong to pass $mds1 which will
cause following error.

line 1192: lustre-ost1/ost1_FSTYPE: bad substitution

And we fail to detect this is ZFS based OSD, and pool
reimporting will be missed thus failed to mount.

Lustre-change: https://review.whamcloud.com/36298
Lustre-commit: 38c8fdfde3953f239bd3d86a91a3213737231ce5

Test-Parameters: trivial clientdistro=el8 testlist=conf-sanity \
fstype=zfs envdefinitions=ONLY=103
Test-Parameters: trivial clientdistro=el8 testlist=conf-sanity \
fstype=ldiskfs envdefinitions=ONLY=103
Change-Id: Id8fd5b9f17e666614e83e5c1a2399fde8b91b023
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: James Nunez <jnunez@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36379
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12229 tests: fix "bad substitution" error

In newer bash version, the special characters is invalid in the
usage of indirect variable expansion {!word}. For example,
# a=lustre,pool
# echo ${!a}
-bash: lustre,pool: bad substitution
To avoid "bad sustitution" error, pool_new command is used in
test_1j and test_1k directly.

Lustre-change: https://review.whamcloud.com/36243
Lustre-commit: ac426d6f17b80ed36052f11b9780fa444cfa24aa

Test-Parameters:trivial clientdistro=el8 testlist=ost-pools
Signed-off-by: Emoly Liu <emoly@whamcloud.com>
Change-Id: Ifce4616cd7f314416fe5fa09f8fba846ae45bcef
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: James Nunez <jnunez@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36377
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11816 lnet: setup health timeout defaults

Enable health feature by default.
Setup transaction timeout to a default 10 seconds and
retry count to 3 when health is enabled. When health
is disabled set default transaction timeout to 50.
When toggling between health enabled/disabled the defaults
will always kick in.

Lustre-change: https://review.whamcloud.com/34252
Lustre-commit: 8632e94aeb7e62da07f342a9897d15dfd8251148

This is a new commit for the previous reverted of commit
https://review.whamcloud.com/#/c/36031/

Change-Id: I359f9cc6c93b5f7d0b58df1abdd29ae0bffd4faf
Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36382
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12344 lnet: handle remote health error

When a peer is dead set the health status to REMOTE_DROPPED
in order to handle health properly for the peer.
When dropping a routed message set REMOTE_ERROR. Routed messages
are dropped when the routing feature is turned off which could
be considered a configuration error if it happens in the middle
of traffic. Therefore, it's better to flag this issue at this
point without resending the message.

Lustre-change: https://review.whamcloud.com/34967
Lustre-commit: b45e3d96fc4d82ebf5b1bb3ef0b5a59e8ff86e75

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I131263215a68fc8607582643a47007ce4d04abbc
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36030
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11478 lnet: misleading discovery seqno.

There is a sequence number used when sending discovery messages. This
sequence number is intended to detect stale messages. However it
could be misleading if the peer reboots. In this case the peer's
sequence number will reset. The node will think that all information
being sent to it is stale, while in reality the peer might've
changed configuration.

There is no reliable why to know whether a peer rebooted, so we'll
always assume that the messages we're receiving are valid. So we'll
operate on first come first serve basis.

Lustre-change: https://review.whamcloud.com/33304
Lustre-commit: 42d999ed8f6113724b1ac103b832d5b74b878d55

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I421a00e47bc93ee60fa37c648d6d9a726d9def9c
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36041
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12825 build: change lbuild to support MOFED 4.7

* Remove 'alternate' name in MOFED tar
* use MLNX_LIBS to download rpms

Test-Parameters: trivial

Lustre-change: https://review.whamcloud.com/36333
Lustre-commit: 279c26466bff37dd25fe26e4bb56a16a9a797870

Change-Id: Ia5a4f51455be836a7df4fa6b3e9eccc17cffef2c
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: Nathaniel Clark <nclark@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36344
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12789 o2ib: fix configure checks

Fix configure checks for modern kernels / MOFED 4.7
1) sg_dma_address() and sg_dma_len() always have only one argument.
2) Make configure checks executed in proper enviroment

Lustre-change: https://review.whamcloud.com/36245
Lustre-commit: f44f657ee218303220f41182ced4fac290266b7f

Change-Id: I9910de888371776758376743ab4418778e1d85e4
Signed-off-by: Alexey Lyashkov <c17817@cray.com>
Reviewed-by: Shaun Tancheff <stancheff@cray.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36331
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11617 mdc: fix possible deadlock in chlg_open()

Lockdep reports a possible deadlock between chlg_open() and
mdc_changelog_cdev_init()

mdc_changelog_cdev_init() takes chlg_registered_dev_lock and then
calls misc_register() which takes misc_mtx.
chlg_open() is called while misc_mtx is held, and tries to take
chlg_registered_dev_lock.
If these two functions race, a deadlock can occur as each thread will
hold one of the locks while trying to take the other.

chlg_open() does not need to take a lock. It only uses the
lock to stablize a list while looking for the matching
chlg_registered_dev, and this can be found directly by examining
file->private_data.

So remove chlg_obd_get(), and use file->private_data to find the
obd_device.
Also ensure the device is fully initialized before calling
misc_register(). This means setting up some list linkage before the
call, and tearing it down if there is an error.

Lustre-change: https://review.whamcloud.com/33572
Lustre-commit: 206b21741b07a10269bbcfdac28743591b64ab2f

Change-Id: Icffdebcee656ee6199297ba2a28ba57dcbc51ae1
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Quentin Bouget <quentin.bouget@cea.fr>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36230
Reviewed-by: Neil Brown <neilb@suse.de>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12705 utils: cleanup unnecessary typecasting

There're a bunch of variables typeecasted in utils/lfs.c where
they are not needed, so cleanup them here.

Lustre-change: https://review.whamcloud.com/36224
Lustre-commit: d8135ad2fbe58a0fbe6984584816338542901c5c

Change-Id: I6c944f18137fd1ff1162d9b6567c9328dfa185eb
Test-Parameters: trivial
Signed-off-by: Gu Zheng <gzheng@ddn.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36313
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12400 lnet: Infiniband sg_dma changes for linux 5.1

IB/core: Remove ib_sg_dma_address() and ib_sg_dma_len()
Linux-commit: a163afc88556e099271a7b423295bc5176fcecce

This simplification can be applied to mainline 3.15 and later
however the test should remain for 3rd party ib driver support

Lustre-change: https://review.whamcloud.com/35497
Lustre-commit: bbc2cf593b83f5f1822889ef5c910906aadbe735

Test-Parameters: trivial
Cray-bug-id: LUS-7600
Signed-off-by: Shaun Tancheff <stancheff@cray.com>
Change-Id: I4824b3b737388a3fc0aec43b2d8e5d10f871ccdd
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Petros Koutoupis <pkoutoupis@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36330
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12690 llite: error handling of ll_och_fill()

The return error of ll_och_fill() should be handled.

Lustre-change: https://review.whamcloud.com/35913
Lustre-commit: 4d6d58575d3d957aa3dbf38f83f749259b580bf2

Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: I4e750001cb124104836fa24e39ec8ae203b51a83
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36315
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12570 mdt: request env for DT threads

as part of lock enqueue MDT thread can call ldlm_reclaim_full() to
cancel old unused LDLM locks and that scans all presented namespace
including OFD-originated (with extent locks). thus MDT ends with
calls into OFD code which needs own env marked with LCT_DT_THREAD.

Lustre-change: https://review.whamcloud.com/36179
Lustre-commit: 1f94d5eb2be4e921e909d8f18523dcab91bb6531

Test-Parameters: testlist=sanity,sanity,sanity,sanity envdefinitions=ONLY="134a",SHARED_KEY=true
Test-Parameters: testlist=sanity,sanity,sanity,sanity envdefinitions=ONLY="134a",SHARED_KEY=true
Test-Parameters: testlist=sanity,sanity,sanity,sanity envdefinitions=ONLY="134a",SHARED_KEY=true
Signed-off-by: Alexey Zhuravlev <bzzz@whamcloud.com>
Change-Id: I231b88159978bc3ce7a3fa0f27e57eb32137c343
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36312
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12355 lnet: ib_fmr_pool_unmap returns void

Historically ib_fmr_pool_unmap only ever returned 0
Linux kernel 4.20 changed the return for ib_fmr_pool_unmap to void.

Linux-commit: 3eeeb7a59acddaa326b03efdf6dce61c120449a3

Lustre-change: https://review.whamcloud.com/35017
Lustre-commit: 46298ffe0b436a8cf1c60aa3d7bde7ae52c78d00

Test-Parameters: trivial
Signed-off-by: Shaun Tancheff <stancheff@cray.com>
Change-Id: I49d91a49c452dad5c7d9b153fdbc011f2f25743a
Reviewed-by: Petros Koutoupis <pkoutoupis@cray.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36329
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12355 lnet: Adjust checks for ib_device_ops

RDMA/core: Introduce ib_device_ops

The ib_device_ops structure defines all the InfiniBand device
operations in one place

Linux-commit: 521ed0d92ab0db3edd17a5f4716b7f698f4fce61

Lustre-change: https://review.whamcloud.com/35016
Lustre-commit: 27572b0476b07b396174430940f184ed85088eeb

Test-Parameters: trivial
Change-Id: Ia2a617597c75ec819f485b93a1deb368d4b5e873
Signed-off-by: Shaun Tancheff <stancheff@cray.com>
Reviewed-by: Petros Koutoupis <pkoutoupis@cray.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36328
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12339 lnet: select LO interface for sending

In the following scenario

Lustre->LNetPrimaryNID with 0@lo
Discover is initiated on 0@lo
The peer is created with 0@lo and <addr>@<net>
The interface health of the peer's <addr>@<net> is decremented
LNetPut() to self
selection algorithm selects 0@lo to send to

This exposes an issue where we try and go through the peer credit
management algorithm, but because there are no credits associated with
0@lo we end up indefinitely queuing the message. ptlrpc will then get
stuck waiting for send completion on the message.

This was exposed via conf-sanity 32a

Lustre-change: https://review.whamcloud.com/34957
Lustre-commit: 69d1535ebdac139c6b19db2bca5f65663fe88467

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I98e9d3428b594a0d041d27d8e8d8de7596825edc
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36040
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12199 lnet: verify msg is commited for send/recv

Before performing a health check make sure the message
is committed for either send or receive. Otherwise we
can just finalize it.

Lustre-change: https://review.whamcloud.com/34797
Lustre-commit: fc6b321036f34c00d5b32b49c817dc0034fbad9e

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: Id7bd956f8e81e60a2d63059730973f851d4c7abe
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36039
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12080 lnet: clean mt_eqh properly

There is a scenario where you have a peer on your recovery queue
that's down. So you keep pinging it, but every ping times out
after 10 seconds. In the middle of these 10 seconds you perform a
shutdown. First you try to do the rsp_tracker_clean. It goes through
and calls MDUnlink on the MD related to that ping. But because the
message has a ref count on the MD, it doesn't go away. The MD gets
zombied. And just waits for lnet_md_unlink to be called in
lnet_finalize(). Then you hit clean_peer_ni_recovery. We see the peer
on the queue, we try to call Unlink on it, but when we lookup the
MD using lnet_handle2md() we can't find it. Afterwards we try to clean
up the EQ and it asserts. Even if we remove the assert we end up with
a resource leak since the EQ is not actually freed since we won't call
LNetEQFree() again.

The solution is to pull the EQ create in the LNetNIInit() and deletion
happens in lnet_unprepare. By this point all the remaining messages
would've been finalized and all references on the EQ are gone,
allowing us to clean it up properly

Lustre-change: https://review.whamcloud.com/34477
Lustre-commit: 1065c8888e96fef9e98676bd3a71b46f7910b085

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I7fd6018ee2e57f82c649fc3658352e89a4309986
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36029
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12199 lnet: Ensure md is detached when msg is not committed

It's possible for lnet_is_health_check() to return "true" when the
message has not hit the network. In this situation the message is
freed without detaching the MD. As a result, requests do not receive
their unlink events and these requests are stuck forever.

A little cleanup is included here:
- The value of lnet_is_health_check() is only used in one place, so
   we don't need to save the result of it in a variable.
- We don't need separate logic to detach the md when the send was
   successful. We'll fall through to the finalizing code after
   incrementing the health counters

Lustre-change: https://review.whamcloud.com/34885
Lustre-commit: b65f3a1767ae82c7f629320187b33eb8670da537

Cray-bug-id: LUS-7239
Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: I6301d491090b862d016eed3aac8afd7be8685e57
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36038
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12264 lnet: Protect lp_dc_pendq manipulation with lp_lock

Protect the peer discovery queue from concurrent manipulation by
acquiring the lp_lock.

Lustre-change: https://review.whamcloud.com/34798
Lustre-commit: dd16a31bf4ae874a69cc7dc5fe1f3197993630ae

Signed-off-by: Chris Horn <hornc@cray.com>
Change-Id: If43b877c1c7ea203f346a3d6ea846f00b8f9661f
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36037
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12254 lnet: correct discovery LNetEQFree()

The EQ needs to be freed after all the queues are cleaned to avoid
having non-processed events on the event queue on free. This will
prevent the memory from being freed.

Lustre-change: https://review.whamcloud.com/34796
Lustre-commit: a0879b5985b41f92dede96e7f27623eb72102b15

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: Ie38ec25e09bf6d7cf2aadc30edd91d298897c51b
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36036
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12249 lnet: fix list corruption

In shutdown the resend queues are cleared and freed. The monitor
thread state is set to shutdown. It is possible to get lnet_finalize()
called after the queues are freed. The code checks for ln_state to see
if we're shutting down. But in this case we should really be checking
ln_mt_state. The monitor thread is the one that matters in this case,
because it's the one which allocates and frees the resend queues.

Lustre-change: https://review.whamcloud.com/34778
Lustre-commit: d799ac910cd6c980b40c81b76eaefb65b88904d0

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: Ia077cec7a52ef5cd2e1b231437c6265ba9416b1b
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36035
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11297 lnet: invalidate recovery ping mdh

For cleanliness, ensure that recovery ping mdh is invalidated when
an peer ni or a local ni are allocated

Lustre-change: https://review.whamcloud.com/34771
Lustre-commit: d7b5f3114d51d5a9d1a34f5073e0bb2d0d63d302

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: If06448b1602b3680831244923b6b982a555159ea
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36034
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12620 kernel: kernel update RHEL 8.0 [4.18.0-80.7.1.el8_0]

Update RHEL 8.0 kernel to 4.18.0-80.7.1.el8_0 for Lustre client.

Test-Parameters: trivial clientdistro=el8 \
envdefinitions=SANITY_EXCEPT="421a" \
testlist=sanity

Change-Id: I9a78ad00d1503cc90f5975e349fe96d452b1174f
Signed-off-by: Jian Yu <yujian@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/35657
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Yang Sheng <ys@whamcloud.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12755 ldiskfs: fix project quota unpon unpatched kernel

The value of MAXQUOTAS is the number of quota types supported
by kernel. With project quotas patch applied, MAXQUOTAS is
equal to EXT4_MAXQUOTAS. However, on an unpatched kernel,
project quota type is not supported and MAXQUOTAS is one less
than EXT4_MAXQUOTAS.

In ldiskfs, we need to make sure that the loop in
ext4_quota_off_umount() is limiting the EXT4_MAXQUOTAS loop
to the kernel MAXQUOTAS value. Otherwise, it is trying to
dereference sb_dqopt(sb)->files[2] which is not an inode at all,
and cause the kernel stick on a spinlock in ext4_quota_off()
as follows during unmount:

Call Trace:
[<ffffffffb9d733c5>] queued_spin_lock_slowpath+0xb/0xf
[<ffffffffb9d81b30>] _raw_spin_lock+0x20/0x30
[<ffffffffb9865e2e>] igrab+0x1e/0x60
[<ffffffffc08a8c4b>] ldiskfs_quota_off+0x3b/0x130 [ldiskfs]
[<ffffffffc08abcdd>] ldiskfs_put_super+0x4d/0x400 [ldiskfs]
[<ffffffffb984b13d>] generic_shutdown_super+0x6d/0x100
[<ffffffffb984b5b7>] kill_block_super+0x27/0x70
[<ffffffffb984b91e>] deactivate_locked_super+0x4e/0x70
[<ffffffffb984c0a6>] deactivate_super+0x46/0x60
[<ffffffffb986abff>] cleanup_mnt+0x3f/0x80
[<ffffffffb986ac92>] __cleanup_mnt+0x12/0x20
[<ffffffffb96c1c0b>] task_work_run+0xbb/0xe0
[<ffffffffb962cc65>] do_notify_resume+0xa5/0xc0
[<ffffffffb9d8d23b>] int_signal+0x12/0x17

This patch is back-ported from the following one:
Lustre-commit: 4b013aa4cdc12647cb1aa9c93bdd72d741b83af4
Lustre-change: https://review.whamcloud.com/36203

Test-Parameters: clientdistro=el7.7 serverdistro=el7.7

Change-Id: I18a4d97656e2f8478754943424c0fac927f843ca
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Reviewed-on: https://review.whamcloud.com/36270
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Jian Yu <yujian@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11296 osc: speed up page cache cleanup during blocking ASTs

While we are cleaning a write lock, we don't need to check if
page cache pages under this lock are covered by another lock.

If a client needs to give up its lock, cleaning gigabytes of
page cache can take quite a long time.

Lustre-change: https://review.whamcloud.com/33090
Lustre-commit: b9ebb17277c78101018a0cf4a63f6beb93b9baf0

Signed-off-by: Andrew Perepechko <c17827@cray.com>
Cray-bug-id: LUS-6352
Change-Id: I576130216ed4de4e352ea697bddb5ff83046443a
Reviewed-by: Patrick Farrell <paf@cray.com>
Reviewed-by: Alexander Zarochentsev <c17826@cray.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@gmail.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/35831
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-12559 ptlrpc: Hold imp lock for idle reconnect

Idle reconnect sets import state to IMP_NEW, then releases
the import lock before calling ptlrpc_connect_import. This
creates a gap where an import in IMP_NEW state is exposed,
which can cause new requests to fail with EIO.

Hold the lock across the call so as not to expose imports
in this state.

Lustre-change: https://review.whamcloud.com/35530
Lustre-commit: e9472c54ac820c3a0db2318a6ef894c3971e6e0b

Signed-off-by: Patrick Farrell <pfarrell@whamcloud.com>
Change-Id: I9f8509d11c4d5a8917a313349534d98b964cd588
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36215
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11743 utils: allow lctl pool commands on separate MGS

The current lctl code checks for the presence of configured pools on
the client and MDS via /proc or /sys files. However, the MGS does
not parse the client/MDS configuration logs, so it does not create
the various files for the pools, which causes the pool commands to
fail verification.

Change lctl pool_new, pool_add, pool_remove and pool_destroy commands
to parse the configuration log directly when run on a standalone MGS
node. This also allows the pool commands to be run when only the MGS
is started.

Lustre-change: https://review.whamcloud.com/34110
Lustre-commit: 4a003a1f554602265630637080f65d9b4474f822

Test-Parameters: standalonemgs=true testlist=ost-pools.sh
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Emoly Liu <emoly@whamcloud.com>
Change-Id: Ib6fdb367c919f7b726fbf551dcfa6015593ebbe5
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/35804
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12612 osd: add lnb size down to osd

so that each OSD can check for lnb array overflow.
the patch isn't final - there will be proper
implementation in osd-zfs and a new test.

Lustre-change: https://review.whamcloud.com/35801
Lustre-commit: 8033f80de3d0db87f7e965078ceee62033adb58d

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I43683c84e48006b4075f9a8b3e87cdfeae28c02b
Reviewed-by: Mike Pershin <mpershin@whamcloud.com>
Reviewed-by: Lai Siyao <lai.siyao@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36273
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12569 o2iblnd: Make credits hiw connection aware

The IBLND_CREDITS_HIGHWATER mark check currently looks only
at the global peer credits tunable, ignoring the connection
specific queue depth when determining the threshold at
which to send a NOOP message to return credits.

This is incorrect because while connection queue depth
defaults to the same as peer credits, it can be less than
that global value for specific connections.

So we must check for this case when setting the threshold.

Lustre-change: https://review.whamcloud.com/35578
Lustre-commit: 1b87e8f61781e48c31b4da647214d66addf2b90c

Test-Parameter: nettype=o2ib
Signed-off-by: Patrick Farrell <pfarrell@whamcloud.com>
Change-Id: Ie028ae11cdbd0f75a38b265b7ab5830f92f08d90
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Amir Shehata <ashehata@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36254
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12385 lnet: update opa defaults

Testing reveals no significant performance improvements
when using peer_credits > 32. Adjusted the default
peer_credits, peer_credits_hiw and concurrent_sends
to take that into account.

This has the advantage of avoiding an issue observed
on multiple opa sites where the qp can not be created because
of large initial queue_depth. The queue depth is then
reduced gradually until the qp creation succeeds.

Lustre-change: https://review.whamcloud.com/36072
Lustre-commit: 7f199dbf0261b89afe0dc8185db4403ae0efdefa

Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I6036ec1da7063e30b567446e5db89040f21bc701
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36252
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12621 o2iblnd: cache max_qp_wr

When creating the device the maximum number of work requests per qp
which can be allocated is already known. Cache that internally,
and when creating the qp make sure the qp's max_send_wr does not
exceed that max. If it does then cap max_send_wr to max_qp_wr.
Recalculate the connection's queue depth based on the max_qp_wr.

Lustre-change: https://review.whamcloud.com/36073
Lustre-commit: 7ee319ed7f9dfa365a66b20b03f2141c54fb0293

Test-Parameter: nettype=o2ib
Signed-off-by: Amir Shehata <ashehata@whamcloud.com>
Change-Id: I6d9a642d03633264f5f14445a051dd14515709c1
Reviewed-by: Doug Oucharek <dougso@me.com>
Reviewed-by: Olaf Weber <olaf.weber@hpe.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36253
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11542 import: Fix missing spin_unlock()

A recent patch moved the spin_unlock() down into
each branch of an 'if', but missed the final 'else'.
Add the spin_unlock in the else.

Lustre-change: https://review.whamcloud.com/35999
Lustre-commit: 3dbdd38a6adcee63b6d89d4656e0099a0006f26c

Fixes: 29904135df67 ("LU-11542 import: fix race between imp_state & imp_invalid")
Signed-off-by: Mr NeilBrown <neilb@suse.com>
Change-Id: I6ee399050aad0fe9df9c0e3ddf8ec0be8eae1641
Reviewed-by: Yang Sheng <ys@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36251
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Neil Brown <neilb@suse.de>
Tested-by: Maloo <maloo@whamcloud.com>

LU-11542 import: fix race between imp_state & imp_invalid

We set import to LUSTRE_IMP_DISCON and then deactive when
it is unreplayable. Someone may set this import up between
those two operations. So we will get a invalid import with
FULL state.

Lustre-change: https://review.whamcloud.com/33395
Lustre-commit: 29904135df671c624b1e542fdda94b221d76e667

Signed-off-by: Yang Sheng <ys@whamcloud.com>
Change-Id: Ib4cec0bcaf6f4b221ba260edb94749a4e523f5e6
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/35796
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: James Simmons <jsimmons@infradead.org>

LU-12090 utils: lfs rmfid

a new RPC_REINT_RMFID has been introduced by the patch.
it's supposed to be used with corresponding llapi_rmfid()
to unlink a batch of MDS files by their FIDs. the caller
has to have permission to modify parent dir(s) and the objects
themselves.

Lustre-change: https://review.whamcloud.com/34449

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ib22379033aca92692e0e219671ca0c2ec7893c24
Reviewed-on: https://review.whamcloud.com/35595
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12705 build: fix building fail against Power9 little endian

We use "%ll[dux]" for __u64 variable as an input/output modifier,
this may cause building error on some architectures which use "long"
for 64-bit types, for example, Power9 little endian.
Here add necessary typecasting (long long/unsigned long long) to
make the build correct.

Lustre-change: https://review.whamcloud.com/36007
Lustre-commit: 4eddf36ac3607c66c172668b30eb5dcf921e3de4

Test-Parameters: trivial
Change-Id: I2e8569f4ac14f7d328a29d153ff57c7834cabc46
Signed-off-by: Gu Zheng <gzheng@ddn.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36207
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11729 obdclass: align to T10 sector size when generating guard

Otherwise the client and server would come up with
different checksum when the page size is different.

Improve test_810 to verify all available checksum types.

Test-Parameters: trivial envdefinitions=ONLY=810 testlist=sanity,sanity,sanity
Test-Parameters: clientarch=aarch64 envdefinitions=ONLY=810 testlist=sanity,sanity
Test-Parameters: clientarch=ppc64 envdefinitions=ONLY=810 testlist=sanity,sanity

Lustre-change: https://review.whamcloud.com/34043
Lustre-commit: 98ceaf854bb4738305769c5cd1df556ee99aa859

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Li Xi <lixi@ddn.com>
Signed-off-by: Li Dongyang <dongyangli@ddn.com>
Change-Id: I24117aebb277d4ddcb7787b715587e33023ebbe5
Reviewed-by: James Simmons <jsimmons@infradead.org>
Reviewed-by: Arshad Hussain <arshad.super@gmail.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36205
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-11485 lod: disallow setting the last non-stale mirror as stale

"lfs setstripe" allows setting stale flag on the last
non-stale mirror of a file, which makes the file have
no valid component to read and return IO error.

This patch fixes the above issue by disallowing that.
It also disallows "lfs mirror split" to destroy the
last non-stale mirror of a file.

This patch is back-ported from the following one:
Lustre-commit: 29be32a759f696006a539d3cff74ca55a281aa64
Lustre-change: https://review.whamcloud.com/36141

Change-Id: I6934cfe0190cd1ea83de1cf28ddf840b9f96193a
Signed-off-by: Jian Yu <yujian@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36195
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-11022 lfs: remove mirror by pool name

lfs mirror split --pool <poolname> <file>

This patch is back-ported from the following one:
Lustre-commit: 0c710a46cfb43366dc57ff6e83e414086b1d0e6c
Lustre-change: https://review.whamcloud.com/35329

Change-Id: I012e68729b94657236ba3fc530fc7b7485529ed2
Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36194
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-12602 mdt: more EA size check in mdt_getxattr_pack_reply()

While the RMF_EAVALS field size can be arbitrary length,
the RMF_EAVALS_LENS field definition specifies
the RMF_F_STRUCT_ARRAY flag, so the passed size must be a multiple
of sizeof(__u32) or the internal LBUG() will trigger.

Lustre-change: https://review.whamcloud.com/36103
Lustre-commit: 4d8bc239c2c30a47e8833cf23db6ccd39ff61705

Signed-off-by: Emoly Liu <emoly@whamcloud.com>
Change-Id: I767e1b1496298e9a66274fc324f9c34daaed4a09
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Patrick Farrell <pfarrell@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/36208
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>

LU-8130 libcfs: port working hash from upstream

The hash_[32|64] function in pre-4.6 kernels produce hashes
with poor distributions which result in high collision rates.
Backport those improvements for the pre-4.6 kernels Lustre
supports. Details can be read here:

https://lwn.net/Articles/687494

Lustre-change: https://review.whamcloud.com/33789
Lustre-commit: 1658ae30a0e97e7f4018d8cba67e459078470d1a

Test-Parameters: trivial
Change-Id: Id2436ba8be2d3ed482c5386b79710f594d5b3e59
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Alexey Lyashkov <c17817@cray.com>
Reviewed-by: Yang Sheng <ys@whamcloud.com>
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/35179
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>