Whamcloud - gitweb
Dmitry Eremin [Tue, 17 May 2016 13:22:15 +0000 (09:22 -0400)]
LU-6215 lnet: split struct ib_send_wr
In v4.4 Linux kernel split up struct ib_send_wr so that all non-trivial
verbs use their own structure which embeds struct ib_send_wr.
Linux-commit:
e622f2f4ad2142d2a613a57fb85f8cf737935ef5
Change-Id: Iea2599ea57a576a4c061841be47a989aba097cc6
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: http://review.whamcloud.com/19168
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Li Dongyang <dongyang.li@anu.edu.au>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
HemaHarish [Wed, 25 Nov 2015 11:25:54 +0000 (16:55 +0530)]
LU-7469 test: add missing init to lustre-rsync-test:test_1a
Target directory was not created for first time, so failed to write to
target location.
Seagate-bug-id: MRP-2382
Signed-off-by: Hema Yarramilli <hema.yarramilli@seagate.com>
Change-Id: I1192c413c8c173abdea3929dbaec19f1f5d20c92
Reviewed-on: http://review.whamcloud.com/17355
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Saurabh Tandan <saurabh.tandan@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Oleg Drokin [Mon, 23 May 2016 01:58:05 +0000 (01:58 +0000)]
Revert "LU-5050 libcfs: default CPT matches NUMA topology"
This causes rhel7 to crash on start in libcfs init
100% of the time.
This reverts commit
6b929502063ac0fa8e75a3e09259e9b2b0d0ee6e.
Change-Id: I9b6276380b580e9d76ae21bf389f5e7edc1a4f18
Reviewed-on: http://review.whamcloud.com/20376
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
Saurabh Tandan [Thu, 12 Nov 2015 19:41:26 +0000 (12:41 -0700)]
LU-7346 tests: Reintroduce SLOW tests to review process
Tests running for five minutes and more are considered to
be eligible to be marked as SLOW. After running tests
multiple times it was found, there are certain tests
which take under five minutes and hence can be
reintroduced in the normal review process.
Signed-off-by: Saurabh Tandan <saurabh.tandan@intel.com>
Change-Id: If638e41a8bf6c9418d69257b00cad964d0810a02
Reviewed-on: http://review.whamcloud.com/17143
Tested-by: Jenkins
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Christopher J. Morrone [Sat, 7 May 2016 01:53:07 +0000 (18:53 -0700)]
LU-8116 build: Shrink multi-line pclmulqdq warnings
The multi-line pclmulqdq warning in the configure script
are excessive. Reduce them to a single line each.
Change-Id: Ifb92d257e41099204bda1392fa6fcfd044f4a5c1
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/20054
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Alexander Boyko <alexander.boyko@seagate.com>
James Nunez [Tue, 2 Feb 2016 17:50:47 +0000 (10:50 -0700)]
LU-7667 tests: Write to unique files in $TMP
sanity uses $TMP/files to store the output of 'lfs getstripe'
and does not clean up the $TMP directory. When conf-sanity tries
to create $TMP/files, it fails. These tests need to use a unique
file, $tfile, or directory, $tdir, name in $TMP and should
clean $TMP out before test exits.
Tests should use $TMP instead of /tmp. $TMP defaults to /tmp.
Remove sanity test 51d from the ALWAYS_EXCEPT list.
Signed-off-by: James Nunez <james.a.nunez@intel.com>
Change-Id: I09a0a2be1d88c3f4a0042e3956bbc54f9df61f55
Reviewed-on: http://review.whamcloud.com/18261
Tested-by: Jenkins
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Saurabh Tandan <saurabh.tandan@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Liang Zhen [Fri, 29 Apr 2016 17:45:39 +0000 (13:45 -0400)]
LU-5050 libcfs: default CPT matches NUMA topology
Change default value of CPT pattern and make it match NUMA topology
Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Change-Id: Ia837ee4b1c52046fcac4a5c05d1c159001fbe810
Reviewed-on: http://review.whamcloud.com/17824
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Fri, 6 Nov 2015 17:17:54 +0000 (11:17 -0600)]
LU-7403 dlm: const qualify struct lustre_handle * params
Add a const qualifier to several struct lustre_handle * parameters in
the LDLM interface.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I65bc949af4ee78118ce9118801434b906dddf49b
Reviewed-on: http://review.whamcloud.com/17071
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
John L. Hammond [Fri, 6 Nov 2015 16:43:37 +0000 (10:43 -0600)]
LU-7403 llite: change it_data to it_request
Change the void *it_data member of struct lookup_intent to struct
ptlrpc_request *it_request.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: If3eb215620dfd376a18e4b15d7ab9f346468add0
Reviewed-on: http://review.whamcloud.com/17070
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Di Wang [Mon, 25 Apr 2016 21:34:39 +0000 (17:34 -0400)]
LU-8097 mdt: remove assertion for lock resent
Do not assert for missing lock for resent, because the lock
might be destroyed by other thread in the mean time.
Signed-off-by: Di Wang <di.wang@intel.com>
Change-Id: If184684ec734e7162e570d3044d9ece5f3ac3567
Reviewed-on: http://review.whamcloud.com/19974
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Emoly Liu [Thu, 12 May 2016 01:57:15 +0000 (09:57 +0800)]
LU-8098 doc: add "lfs mdts" to lfs main man page
Add "lfs mdts" to lfs.1.
Test-Parameters: trivial
Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Change-Id: I0da79a56025b58bd32bdcee35c36e7021d69194f
Reviewed-on: http://review.whamcloud.com/20080
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Mon, 16 May 2016 17:46:00 +0000 (11:46 -0600)]
LU-8139 tests: skip sanity test 101g until fix is landed
The landing of 16MB RPC and lprocfs value parsing were tested
independently, but had conflicting changes that are causing
sanity.sh test_101g to fail continuously. Disable the test
until a fix is made.
Test-Parameters: trivial
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I93e82cb5606400f39b6da7a6e35c580826c2d106
Reviewed-on: http://review.whamcloud.com/20241
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
Niu Yawei [Tue, 3 May 2016 07:37:41 +0000 (03:37 -0400)]
LU-8091 mount: error out properly in server_lsi2mti()
server_lsi2mti() should return an error if no proper NID is
initialized, otherwise, target will register successfully with
an invalid empty NID.
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: Ia5fd78e4df6344764109486ce1082099054ebf1e
Reviewed-on: http://review.whamcloud.com/19933
Tested-by: Jenkins
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Henri Doreau [Thu, 28 Apr 2016 11:01:16 +0000 (13:01 +0200)]
LU-8079 llog: Remove llog_cat_init_and_process
Unlike name suggests, this function does no longer process
the log but only initialize it. Replace it by direct calls
to the function it wraps.
Signed-off-by: Henri Doreau <henri.doreau@cea.fr>
Change-Id: Ib48e7d4cb15a2433cc07c6760e25450a9aa0a3b3
Reviewed-on: http://review.whamcloud.com/19850
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Li Dongyang [Wed, 4 May 2016 17:11:15 +0000 (13:11 -0400)]
LU-6215 o2iblnd: port to new fast reg API introduced in 4.4
Remove the allocation of fastreg page list, as the page
vector is now private to the provider. Just pass tx_frags
to ib_map_mr_sg() and construct ib_reg_wr.
Defer the conversion of tx_frags to tx_pages and only do
it when the new API is not available.
Linux-commit:
4c67e2bfc8b7121d51434362fa7c2d012f8bcf1b
Linux-commit:
39bfc271bd687be2c8e396e976c0fb9a97963400
Signed-off-by: Li Dongyang <dongyang.li@anu.edu.au>
Change-Id: I5c62b0370ad6ddcc93102a29343491968e4446d0
Reviewed-on: http://review.whamcloud.com/19186
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Wed, 13 Apr 2016 09:37:18 +0000 (17:37 +0800)]
LU-7888 obdclass: not hold global lock when lwp callback
Originally, the lwp_notify_main() thread handled the every
registered lwp callback in turn with holding the global
lwp_register_list_lock. Because some lwp callback, such as
mdt_register_lwp_callback() may be blocked, then it will
cause others to be blocked when try to obtain such lock for
lustre_register_lwp_item(), and then block the mount.
This patch introduces reference count on the lwp_register_item
that will prevent the lwp item being freed during the callback,
so allow the lwp_notify_main() thread to do the real callback
without holding the global lock lwp_register_list_lock.
Intel-bug-Id: LDEV-323
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I1061205d5781e0699478d293e223567a0ebea286
Reviewed-on: http://review.whamcloud.com/19034
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Fri, 5 Feb 2016 20:47:48 +0000 (12:47 -0800)]
LU-7733 ptlrpc: print times in microseconds
report times and time diffs in seconds.microseconds instead of seconds
Signed-off-by: Bob Glosman <bob.glossman@intel.com>
Change-Id: I272626012a7162072fb3ad73912a56403afb6b39
Reviewed-on: http://review.whamcloud.com/18335
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bruno Faccini [Thu, 21 Apr 2016 13:56:28 +0000 (15:56 +0200)]
LU-8010 mdt: fix orphan layout_lock cases for restore
Previously to this patch layout was not unlocked when a restore
was failed before being sent to CT, leading to a situation where
other requestors hang and also to an orphan restore_handle to be
kept on CDT's list of registered restore actions.
Only way to clear situation then, was to stop CDT.
This situation could at least occur if a restore was canceled
but the CT does not handle cancel operation, allowing the restore
to complete but also to have new restore requests to be
registered in the mean time and then to be failed due to
incompatible (no longer released) file state.
Also fix similar deadlock cases where layout lock was taken for
previously started restore requests upon CDT restart, by forcing
their replay/restart. This should also strengthen overall HSM
recovery process.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: Ib1ba9156793a230d256ff80d74372813f10b0321
Reviewed-on: http://review.whamcloud.com/19710
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Reviewed-by: Aurelien Degremont <aurelien.degremont@cea.fr>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Tue, 19 Apr 2016 06:12:00 +0000 (14:12 +0800)]
LU-8127 lfsck: remove XATTR_NAME_LFSCK_NAMESPACE_OLD
We do not support the LFSCK compatibility between Lustre-2.9
and Lustre-2.6 any more. So drop related conditional compile
logic for XATTR_NAME_LFSCK_NAMESPACE_OLD.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I3374ed357fc675b382fbb6703d0ba65c2d1420d9
Reviewed-on: http://review.whamcloud.com/20112
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Wed, 4 May 2016 15:17:02 +0000 (08:17 -0700)]
LU-8093 kernel: kernel update RHEL 6.7 [2.6.32-573.26.1.el6]
Update RHEL 6.7 kernel to 2.6.32-573.26.1.el6
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I1ff675f734cb651d10860956c9d2a7edf191cbf2
Reviewed-on: http://review.whamcloud.com/19984
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Tested-by: Jenkins
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Wed, 4 May 2016 01:27:04 +0000 (18:27 -0700)]
LU-8094 ldlm: restore missing newlines
Recent commit f28cc2 landed for LU-7521 removed too many newlines
in debug messages, leading to lots of noise in syslog.
This mod puts some back in.
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I4d3eb59a5db3ed548b03aa29ec036b4b2f899a86
Reviewed-on: http://review.whamcloud.com/19960
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Niu Yawei [Tue, 3 May 2016 07:18:17 +0000 (03:18 -0400)]
LU-8041 mdd: increasing only atime update on close
The atime update on close should be increasing only, otherwise,
the atime could be overwritten by stale client's atime.
This patch also changed client to clear out the ATTR_ATIME
flag fro close RPC when atime == 0, so that patch client won't
overwrite the unpatched server with 0 atime.
A test case sanity 39p is added to verify that close won't
overwrite atime with zero, and sanity 39l is adjusted as well.
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I40d1103ddfea184d82170a6795f89ce303a5c14e
Reviewed-on: http://review.whamcloud.com/19932
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
John L. Hammond [Fri, 29 Apr 2016 12:00:33 +0000 (07:00 -0500)]
LU-8035 obd: rename md_getstatus() to md_get_root()
Finish the partial renaming of the the OBD MD method md_getstatus() to
md_get_root().
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Iadd785ce774377d3dba40f64dbddaaf68cfc8087
Reviewed-on: http://review.whamcloud.com/19824
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Wang Shilong <wshilong@ddn.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Yang Sheng [Thu, 14 Apr 2016 20:05:24 +0000 (04:05 +0800)]
LU-8027 llite: ensure obd is effective in onu_upcall
The watched obd device may still not setup while onu_upcall
invoked. So we need verify it in cl_ocd_update.
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Change-Id: Iddda6837127f1066ede8430f355d8fd04ed3b71c
Reviewed-on: http://review.whamcloud.com/19597
Tested-by: Jenkins
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Tue, 15 Mar 2016 15:37:32 +0000 (11:37 -0400)]
LU-7870 osd-zfs: Build against SPL/ZFS 0.6.5.6
Build against SPL/ZFS 0.6.5.6.
Fix issue with rpm's debugedit, where it can't account for path
canonicalization causes the path to shrink by only one characater. So
a trailing slash will cause it choke.
ZFS Changelogs since 0.6.4.3:
https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.6.5.6
https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.6.5.5
https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.6.5.4
https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.6.5.3
https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.6.5.2
https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.6.5.1
https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.6.5
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I26569a038df605b7b6a37e2806d477029d68fca0
Reviewed-on: http://review.whamcloud.com/18920
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Noopur Maheshwari [Tue, 1 Mar 2016 08:25:13 +0000 (13:55 +0530)]
LU-7656 tests: tar fix for replay-single/70c
The tar process runs in an infinite loop, and the removal/cleanup
of files interferes in the process and causes tar to fail.
The code flow could take two paths:
1. Mdt recovery has completed, catch exit signal.
In this case, put a check and confirm if the mdt recovery has
completed, if yes, then exit and perform cleanup
commands (remove files).
2. Main process code flow.
Wait for tar process to complete and then perform cleanup commands.
Also, instead of removing files within the tar process, keep the
removal of files in the cleanup part.
Seagate-bug-id: MRP-2925
Signed-off-by: Noopur Maheshwari <noopur.maheshwari@seagate.com>
Change-Id: Ia1c840dde554a2f1487aa9aebb70e5dcea29b806
Reviewed-on: http://review.whamcloud.com/18732
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Ujjwal Lanjewar <ujjwal@yahoo.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Giuseppe Di Natale [Thu, 1 Oct 2015 22:51:19 +0000 (15:51 -0700)]
LU-7334 lprocfs: Refactored string to value helpers
Refactored the string parsing helpers to correct
inconsistent behavior and added handling for s64
overflow/underflow. Removed the old helper
functions and refactored functions that called
them.
Signed-off-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Change-Id: I84f3b5d46d4e20bd09fe223589e1be96dc8f7468
Reviewed-on: http://review.whamcloud.com/16930
Reviewed-by: Christopher J. Morrone <morrone2@llnl.gov>
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Sun, 3 Apr 2016 13:10:48 +0000 (21:10 +0800)]
LU-8048 mgc: use the same lu_env for mgc_fs_{set,clean}up
Originally, the mgc_fs_setup() and mgc_fs_cleanup use each own
lu_env. Sometimes the mgc_fs_cleanup() may fail to prepare the
lu_env. Such failure will cause local_oid_storage_fini() to be
skipped as to object leak found when osd_device_free().
This patch prepares the lu_env before calling mgs_fs_setup(),
then mgc_fs_setup() and mgc_fs_cleanup() can share the lu_env
and no need to worry about the failures in mgc_fs_cleanup().
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I7062abd670c53e93f456a8d8f1f6131d7ee92320
Reviewed-on: http://review.whamcloud.com/19692
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Mon, 18 Apr 2016 10:45:02 +0000 (18:45 +0800)]
LU-8127 lfsck: compile XATTR_NAME_LFSCK_NAMESPACE_OLD
The logic related with XATTR_NAME_LFSCK_NAMESPACE_OLD is only
compiled for the version less than OBD_OCD_VERSION(2, 8, 53, 0).
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ie0f9c0901e69465fe73051c16fa5e7f5997ef082
Reviewed-on: http://review.whamcloud.com/20109
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Oleg Drokin [Wed, 11 May 2016 02:48:18 +0000 (22:48 -0400)]
New tag 2.8.53
Change-Id: I3f5aa724e3c629a97f51bdd885ca9b28db5de6ae
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Tue, 10 May 2016 14:36:42 +0000 (09:36 -0500)]
LU-6937 test: use ps in get_agent_uuid()
Use 'ps -C $HSMTOOL_BASE -o args=' rather than 'pgrep -fl
$HSMTOOL_BASE' to get the copytool command line.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Idb1238966de1a8bea5d5c225632936d8c0a88529
Reviewed-on: http://review.whamcloud.com/20093
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Fri, 6 May 2016 05:03:54 +0000 (23:03 -0600)]
LU-6844 tests: disable DNE testing of dbench
Temporarily disable DNE recovery testing with dbench until LU-6844
and related failures are fixed.
Test-Parameters: testlist=replay-single,replay-single,replay-single,replay-single,replay-single,replay-single
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I628642acb119121457ffb83d654f02e7ac40d056
Reviewed-on: http://review.whamcloud.com/20022
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Fri, 29 Apr 2016 20:48:52 +0000 (14:48 -0600)]
LU-6215 llite: remove obsolete conditional code
Remove conditional flock/aops code that was only for out-of-tree
vendor kernels but is not relevant anymore.
Linux-commit:
3746e6f93bbf28a25d2d69350ab6bfba02e14654
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change-Id: If119a2da2bbc30a48d872029bd048ea90a3ebbe5
Reviewed-on: http://review.whamcloud.com/19890
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 11 Apr 2016 14:07:34 +0000 (09:07 -0500)]
LU-7152 hsm: sync volatile file before setting times
In the restore case of llapi_hsm_action_end() call fsync() on the
volatile file before updating its timestamps. This prevents spurious
{a,m,c}times from fsync(). Correct a related comment in sanity-hsm
test_24a().
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I982dfd49cbb7d6f9cd7d569e834e4e0581e4bdc9
Reviewed-on: http://review.whamcloud.com/19441
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Fri, 29 Apr 2016 16:47:38 +0000 (12:47 -0400)]
LU-8017 obd: report correct health state of a node
The patch 16933 that changed the handling of seq_printf
to work with newer kernels introduced a bug in the proc
file that reports the health of a node. The original
code returned immediately after a healthy state was
reported but the bug prevented the immediate return
so a false unhealthy state was always reported instead.
Change-Id: I414d43d59f0df413ea955acf74c91d2402543228
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: http://review.whamcloud.com/19537
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Christopher J. Morrone [Tue, 26 Apr 2016 21:54:51 +0000 (14:54 -0700)]
LU-8073 build: Eliminate lustre-source binary package
The lustre source code is already available in tarball form
and on rpm systems in .src.rpm form. The source code is
also incidentally available in the lustre-debuginfo package.
The reasons for ever having a lustre-source binary package
are probably lost to history.
But this is a really odd thing to do, so we remove that package
from the build system.
Change-Id: I4a175c556d70a380eb550e5ec2830f4a8d3cd04a
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/19806
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Brian J. Murrell <brian.murrell@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andrew Perepechko [Sat, 23 Apr 2016 05:03:05 +0000 (08:03 +0300)]
LU-8059 ldiskfs: wrong ldiskfs patch
Something went wrong with rebasing the ldiskfs patch in
http://review.whamcloud.com/#/c/16360/ LU-7132.
This patch includes the corrected version.
Change-Id: Ice4319f3ebedcdea781bd2198e8b62d010a304de
Signed-off-by: Andrew Perepechko <andrew.perepechko@seagate.com>
Seagate-bug-id: MRP-2688
Reviewed-on: http://review.whamcloud.com/19747
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Herbert Xu [Mon, 25 Apr 2016 16:49:34 +0000 (12:49 -0400)]
LU-4423 libcfs: use ahash for libcfs crypto layer
This patch replaces uses of the long obsolete hash interface with
ahash.
Linux-commit :
6dae10001e84253992142c7343fdae16a13f3d73
Linux-commit :
9dffdb38d864ae89e16ff7b3a09451270736e35b
Change-Id: I9fba47cc05f755ebfd2350c7881708745283d666
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-on: http://review.whamcloud.com/19716
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alexander Boyko <alexander.boyko@seagate.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Thu, 4 Sep 2014 16:56:25 +0000 (09:56 -0700)]
LU-7990 clio: revise readahead to support 16MB IO
The major work is to adjust read ahead policy to issue
read ahead RPC by underlying RPC size
Test case is backport from Andrew Perepechko's
"LU-7140 llite: too few 4 MiB RPCs from readahead"
http://review.whamcloud.com/16374
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: Gu Zheng <gzheng@ddn.com>
Change-Id: Ib25ff37712389cca7fcee86996124e7c9dd4ec9a
Reviewed-on: http://review.whamcloud.com/19368
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Alex Zhuravlev [Wed, 23 Mar 2016 12:37:57 +0000 (15:37 +0300)]
LU-7904 osd: honor LOC_F_NEW
the caller passes this flag to ensure OSD that the object being
initialized is new so OSD can skip save on lookup in OI.
Change-Id: Iaebc23071420f812571c7652a9f29a7e9fc7b67e
Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-on: http://review.whamcloud.com/19089
Tested-by: Jenkins
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Mel Gorman [Fri, 1 Apr 2016 23:37:57 +0000 (19:37 -0400)]
LU-4423 mem: remove GFP_IOFS
GFP_IOFS was intended to be shorthand for clearing two flags, not a set of
allocation flags. There is only one user of this flag combination now and
there appears to be no reason why Lustre had to be protected from reclaim
stalls. As none of the sites appear to be atomic, this patch simply
deletes GFP_IOFS and converts Lustre to using GFP_KERNEL, GFP_NOFS or
GFP_NOIO as appropriate.
Linux-commit :
40113370836e8e79befa585277296ed42781ef31
Change-Id: Ie17d225583f9b48ec2558a85bb6e4e9957010c14
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: http://review.whamcloud.com/18956
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Christopher J. Morrone [Tue, 26 Apr 2016 21:41:02 +0000 (14:41 -0700)]
LU-8072 build: Restore module debuginfo
In commit
3ce96f162a weaccidentally dropped the chmod that makes kernel
modules available to the find-debuginfo.sh.
Reintroducing this chmod allows the lustre debuginfo package to contain
debuginfo for the kenrel modules.
Change-Id: I57e2a6dc248aaccf573e68a49a981bf92c7fbb62
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/19805
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Mon, 11 Apr 2016 21:42:00 +0000 (17:42 -0400)]
LU-6245 libcfs: Remove kp30.h
The header kp30.h contains various wrappers to handle things
like module parameter handling. We can now remove this
abstraction.
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Change-Id: I7f5fec4372023146215c740e287529a05dfceba6
Reviewed-on: http://review.whamcloud.com/19460
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Tue, 26 Apr 2016 00:54:50 +0000 (20:54 -0400)]
LU-6245 libcfs: remove sysctl module handling
Remove the sysctl handling we have for module parameters.
This code only gets built if CFS_SYSFS_MODULE_PARM is not
1 which is never the case. So this code is never built.
Plus this code has long been replaced by the sysfs module
parameters handling.
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Change-Id: I5635bfd75a53bdd25af51c65fb1c2550d3cbbbaa
Reviewed-on: http://review.whamcloud.com/19459
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Fri, 8 Apr 2016 15:32:34 +0000 (18:32 +0300)]
LU-7999 utils: fix double free the pointer
Object 'uuids' was freed at line 1895 after being freed by
calling 'realloc' at line 1892.
Change-Id: I2549f2cb0d2e1a2e4707c4d54d37c90ac099f375
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: http://review.whamcloud.com/19409
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Wed, 30 Mar 2016 15:03:20 +0000 (08:03 -0700)]
LU-7957 kernel: kernel update [SLES11 SP4 3.0.101-71]
Update SLES11 SP4 kernel to 3.0.101-71
Test-Parameters: mdsdistro=sles11sp4 ossdistro=sles11sp4 \
clientdistro=sles11sp4 mdsfilesystemtype=ldiskfs \
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs \
testgroup=review-ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I1e7ff0e2abb2274176a04e8a16a33fe298e77bc5
Reviewed-on: http://review.whamcloud.com/19215
Tested-by: Jenkins
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Alex Zhuravlev [Mon, 7 Dec 2015 06:42:24 +0000 (09:42 +0300)]
LU-7521 ldlm: LDLM_DEBUG() shouldn't be passed \n
as it adds own \n, so any extra \n break log format.
Change-Id: Ie78d0969421b744db6789d7bd46505ba7097a46b
Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-on: http://review.whamcloud.com/17494
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Emoly Liu [Mon, 25 Apr 2016 07:41:40 +0000 (15:41 +0800)]
LU-8064 lctl: fix a typo in lustre_cfg.c
When checking overflow issue in param_dispaly(), we should use rc2
not rc, because rc2 is the return value of pathname string size.
Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Change-Id: I35820b00c7a4e264444e5574967d56c54b4a469b
Reviewed-on: http://review.whamcloud.com/19759
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Kirtankumar Krishna Shetty [Thu, 21 Apr 2016 08:48:01 +0000 (14:18 +0530)]
LU-8051 tests: Skipping sanity tests 300[a-d] for old servers.
The tests 300[a-i] were added in version 2.7.0 for the newer
version of the code to test striped directory related features
and hence the tests 300[a,b,c,d] used to fail for the v2.7.0
and above client with older server versions. Added skip
statement for the tests if server is of an older versions.
Seagate-bug-id: MRP-3393
Change-Id: I26f271c1a4aa62c8f41c391e0c595f1ccaf18f8b
Signed-off-by: Kirtankumar Krishna Shetty <kirtan.shetty@seagate.com>
Reviewed-on: http://review.whamcloud.com/19698
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Jinshan Xiong [Mon, 11 Apr 2016 17:56:37 +0000 (10:56 -0700)]
LU-8005 osc: set lock data for readahead lock
If osc_io_readahead() finds a lock that belongs to the previous
instance of osc_object, the lock data pointer will be null. It has
to instantiate with new instance otherwise those pages won't be
destroyed at lock cancel, and then finally hit the assertion in
osc_req_attr_set().
This patch revised dlmlock_at_pgoff() to call osc_match_base() to
find caching locks for readahead. And new osc_object will be set
to the lock if it doesn't have one yet.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I6e95499bf0c1e501f3f55d25e0c2b3fbc61ed0da
Reviewed-on: http://review.whamcloud.com/19453
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Emoly Liu [Thu, 21 Apr 2016 04:00:31 +0000 (12:00 +0800)]
LU-7972 tests: use $REFORMAT in test-framework.sh correctly
$REFORMAT is always used incorrectly in test-framework.sh, because
as long as $REFORMAT is not null, format will be done, even if
REFORMAT=no.
This patch fixes this issue by setting/judging $REFORMAT=yes/no
explicitly.
Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Change-Id: Iaabea34e0ad453b3f9f986009fa1f8bdec8a98ff
Reviewed-on: http://review.whamcloud.com/19276
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Nathaniel Clark [Wed, 9 Mar 2016 22:14:15 +0000 (17:14 -0500)]
LU-7859 tests: Use getent instead of /etc/passwd
Use getent passwd instead of directly reading form /etc/passwd
to prevent missing user checks if users come from ldap or another
auth source that isn't /etc/passwd.
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ia2831c777c7c3f8bae5a634fa897744954d14c1e
Reviewed-on: http://review.whamcloud.com/18846
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Parinay Kondekar [Mon, 18 Jan 2016 03:29:35 +0000 (08:59 +0530)]
LU-6512 llite:ll_dir_ioctl cleanup of redundant comparisons
In ll_dir_ioctl() two identical comparisions are present for
return code (rc) of ll_dir_getstripe(). This patch removes
the other inside if( ) condition which is not necessary.
Change-Id: Ia32dd5ac5937d30dd3676094ef99a623e5cca608
Signed-off-by: Parinay Kondekar <parinay.kondekar@seagate.com>
Reviewed-on: http://review.whamcloud.com/18027
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Jenkins
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Fri, 6 Nov 2015 16:05:32 +0000 (10:05 -0600)]
LU-7403 llite: flatten struct lookup_intent
Replace the union in struct lookup_intent with the members of struct
lustre_indent_data. Remove the then unused struct lustre_intent_data.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I47d815f7c8e51bcff33ea7dba6efbc08337ca509
Reviewed-on: http://review.whamcloud.com/17069
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Patrick Farrell [Tue, 5 Apr 2016 15:48:57 +0000 (10:48 -0500)]
LU-7981 llite: take trunc_sem only at vvp layer
The lli_trunc_sem is taken in 'read' mode in both
ll_page_mkwrite and vvp_io_fault_start. This can lead to a
deadlock with another thread which asks for the semaphore
in write mode between thse two read calls.
Since all users of lli_trunc_sem are in the vvp layer, we
can satisfy the requirement to exclude truncate by taking
the semaphore only in vvp_io_fault_start.
Signed-off-by: Patrick Farrell <paf@cray.com>
Change-Id: I7341d179f550835ccebfb1bb2d90260c1d5d0ff8
Reviewed-on: http://review.whamcloud.com/19315
Tested-by: Jenkins
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andriy Skulsyh
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Alexander Boyko [Fri, 22 Apr 2016 13:00:40 +0000 (16:00 +0300)]
LU-6722 ldiskfs: fix credits at ldiskfs_delete_inode
extra_credits was added to take account of quota, but later
journal extend still use 3 blocks. The patch changes
this 3 credits to extra_credits.
Signed-off-by: Alexander Boyko <alexander.boyko@seagate.com>
Seagate-bug-id: MRP-3441
Change-Id: Ib098f3181790367cbc384f25072588b08ea624e4
Reviewed-on: http://review.whamcloud.com/19732
Tested-by: Jenkins
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Christopher J. Morrone [Mon, 25 Apr 2016 14:36:49 +0000 (10:36 -0400)]
LU-7699 build: Convert version underscores to dashes for dpkg
For rpm, we can't have dashes in the version string so we use
underscores. For dpkg, the requirements are reversed.
In the "debs" make target, we convert underscores in the version
string to dashes. We also trim down the comment that no longer
really makes sense with the new versioning scheme.
Change-Id: I6d600fc4a16be60c664ef0662b482c31877fbfea
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/19488
Tested-by: Jenkins
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Mikhal Pershin [Mon, 4 Apr 2016 10:56:35 +0000 (13:56 +0300)]
LU-7558 import: don't reconnect during connect interpret
The import connect flags might be cleared by ptlrpc_connect_import()
wrongly if there is still connect interpret function is running.
Use imp_connected boolean variable to indicate that we are still
interpretting connect reply and don't try to reconnect until it ends.
Signed-off-by: Mikhal Pershin <mike.pershin@intel.com>
Change-Id: I77cf70820835dc06b33a75353c8ef71ba4ed4fad
Reviewed-on: http://review.whamcloud.com/19312
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Oleg Drokin [Tue, 26 Apr 2016 17:46:09 +0000 (17:46 +0000)]
Revert "LU-7791 ldlm: signal vs CP callback race"
This causes frequent failures tracked in LU-8067
This reverts commit
a880f385164547837a34df92acd6cdc8e9a9698d.
Change-Id: Ib247bb924dabe4635edd17f9a2dca391135f558f
Reviewed-on: http://review.whamcloud.com/19798
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Jenkins
Li Xi [Tue, 21 Jul 2015 09:12:16 +0000 (17:12 +0800)]
LU-7982 libcfs: memory allocation without CPT for binheap
In order to use heap in cases which are not tied to a specific CPT,
memory allocation should support normal memory allocation. This
patch adds support for it.
Signed-off-by: Li Xi <lixi@ddn.com>
Change-Id: I1daeb76b708489b3adf09251fe1fbf2ddeab35aa
Reviewed-on: http://review.whamcloud.com/19317
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Nikitas Angelinas <nikitasangelinas@gmail.com>
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Vitaly Fertman [Sun, 21 Feb 2016 01:35:20 +0000 (04:35 +0300)]
LU-7434 ptlrpc: Early Reply vs Reply MDunlink
A race between unregister_reply & early reply.
When buffers are busy for the early transfer, they cannon be unlinked
by unregister_reply, so the RPC gets into UNREGISTERING state. The
coming reply_in_callback for the early RPC already has unlinked flag
set due to previous mdunlink attempt, but we handle it properly only
for UNILNK event, whereas this is PUT in this case.
Signed-off-by: Vitaly Fertman <vitaly.fertman@seagate.com>
Change-Id: I98a939e54d5d9298f217761c8c7289843e46162d
Seagate-bug-id: MRP-3323
Reviewed-by: Alexey Leonidovich Lyashkov <alexey.lyashkov@seagate.com>
Reviewed-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Tested-by: Parinay Vijayprakash Kondekar <parinay.kondekar@seagate.com>
Reviewed-on: http://review.whamcloud.com/18934
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bruno Faccini [Wed, 6 Apr 2016 14:38:05 +0000 (16:38 +0200)]
LU-7980 ldiskfs: always pre-allocate max depth for path
Upon very specific meta-data and I/O pattern a situation can be
encountered where ext_depth() will concurrently (due to i_data_sem
release) grow outside of currently used ext4_ext_path[] array
boundary.
This can lead to potential Slab overrun if current ext_depth()
is used to index previously sized+allocated ext4_ext_path[] array.
This patch fixes this by already pre-allocating _ext_path[]
array of the max possible depth.
This is an alternate way to avoid racy cases, that can lead to
slab overrun, with current and unsafe implementation that causes
ext4_ext_path[] array re-[sizing,allocation] based on current depth.
Now unnecessay free an realloc upon depth change detection prior to
reuse already allocated path space have also been removed.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I0ae955c5eccf032cb02ef55c6381964a5eaf82c2
Reviewed-on: http://review.whamcloud.com/19349
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Li Dongyang [Sat, 2 Apr 2016 00:26:46 +0000 (20:26 -0400)]
LU-6215 o2iblnd: rdma_create_id() takes extra parameter
4.4 kernel added network namespace parameter to rdma_create_id().
This patch handles the API change.
Linux-commit:
fa20105e09e97e81aadf02f722c31195e4a75c84
Signed-off-by: Li Dongyang <dongyang.li@anu.edu.au>
Change-Id: I3f3792e70e91ac3b3d9fefdd58cc9e7e4d1c25c6
Reviewed-on: http://review.whamcloud.com/18759
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Li Dongyang [Tue, 29 Mar 2016 03:57:57 +0000 (14:57 +1100)]
LU-6215 o2iblnd: ib_alloc_fast_reg_mr() removed in 4.3
ib_alloc_fast_reg_mr() is replaced by a more generic
ib_alloc_mr() since kernel 4.3
Linux-commit:
d9f272c523db47a56a64942eb6f25361c400de66
Signed-off-by: Li Dongyang <dongyang.li@anu.edu.au>
Change-Id: I53f4e7c395c83d731aaa342f308f878b07b54d61
Reviewed-on: http://review.whamcloud.com/19185
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Li Dongyang [Thu, 21 Apr 2016 14:25:28 +0000 (10:25 -0400)]
LU-6215 gss: key->payload.data is an array in 4.4+ kernels
Kernel 4.4 merged the type-specific data with the payload data
for keys, as a result, payload.data is now an array.
This patch handles the code change.
Linux-commit:
146aa8b1453bd8f1ff2304ffb71b4ee0eb9acdcc
Signed-off-by: Li Dongyang <dongyang.li@anu.edu.au>
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Change-Id: I1036ab927514f0efa073d9f0136e299be8950461
Reviewed-on: http://review.whamcloud.com/18731
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Sebastien Buisson <sbuisson@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Thu, 21 Apr 2016 17:18:16 +0000 (10:18 -0700)]
LU-8052 test: stop using unknonwn sed option
Avoid the use of the -E command line option for sed.
It isn't known in all versions.
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I66f58aa04ecc751f4b3d61475cafe0c478ce6206
Reviewed-on: http://review.whamcloud.com/19714
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Dmitry Eremin [Tue, 19 Apr 2016 15:18:41 +0000 (18:18 +0300)]
LU-8023 lbuild: add find-requires parameter to rpmbuild
If kernel-devel unpacked in not default directory we need to
pass modified version of find-requires which can find it.
Test-Parameters: mdsdistro=sles11sp4 ossdistro=sles11sp4 \
clientdistro=sles11sp4 mdsfilesystemtype=ldiskfs \
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs \
testgroup=review-ldiskfs
Change-Id: Id30744fe596b1bf690611e9f66dd804219b20fe3
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: http://review.whamcloud.com/19650
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Oleg Drokin [Tue, 26 Apr 2016 00:35:34 +0000 (00:35 +0000)]
Revert "LU-7434 ptlrpc: lost bulk leads to a hang"
This causes widespread testing failures reported in LU-8062
This reverts commit
55f8520817a31dabf19fe0a8ac2492b85d039c38.
Change-Id: Ifcdf0bf569f6ea43cf625095e77ba347649c37df
Reviewed-on: http://review.whamcloud.com/19778
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>
Frank Zago [Tue, 12 Apr 2016 21:31:07 +0000 (17:31 -0400)]
LU-8014 hsm: remove kuc_ispayload
That function takes a pointer, decrements it and dereference the new
address. However we have no idea if the page it's in is
readable or even exists. Such dereference would cause an oops.
That function is only called from mdt_hsm_agent_send, which has itself
2 callers. In the most common case, from mdt_coordinator(), the buffer
passed is already a KUC, but in the less used case, from
hsm_cancel_all_actions(), it is not. Allocating the KUC buffer in
mdt_coordinator() doesn't buy us anything. So remove that code from
the coordinator, which means mdt_hsm_agent_send() is now always called
with a regular buffer, and kuc_ispayload can go.
Signed-off-by: frank zago <fzago@cray.com>
Change-Id: I99c29d4223c14b6fb778215032cd07b6e4676990
Reviewed-on: http://review.whamcloud.com/19494
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Henri Doreau <henri.doreau@cea.fr>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Fri, 9 May 2014 01:24:19 +0000 (18:24 -0700)]
LU-7990 ofd: add a parameter to adjust preferred brw size
Add parameter ofd_device::ofd_brw_size to set the
preferred BRW size of an OST.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: Gu Zheng <gzheng@ddn.com>
Change-Id: Iab9b26fb1e1c1af8de10d2c96da69151a0a654f5
Reviewed-on: http://review.whamcloud.com/19367
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Lokesh Nagappa Jaliminche [Thu, 14 Apr 2016 08:12:01 +0000 (13:42 +0530)]
LU-7922 ldiskfs: correction in ext4_kzalloc
ext4_kzalloc uses kmalloc instead of kzalloc, this was
introduced by commit 9933fc0. Due to this get_orlov_stats
was not working properly, making orlov allocator completely
random. kmalloc is replaced with kzalloc to get correct
stats.This is corrected in kernel as part of following
commit v3.0-7217-gdb9481c04
Added __GFP_NOWARN to k.alloc calls with v.alloc fallbacks
as per this commit v3.11-rc2-221-g8be04b937.
Seagate-bug-id: MRP-3412
Signed-off-by: Lokesh Nagappa Jaliminche <lokesh.jaliminche@seagate.com>
Change-Id: I15f1a28537cb5e4f32d63527c99508383e12c97f
Reviewed-on: http://review.whamcloud.com/19541
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Bobi Jam [Sat, 26 Mar 2016 14:36:34 +0000 (22:36 +0800)]
LU-7198 clio: remove mtime check in vvp_io_fault_start()
In fault IO initialization, inode's mtime is saved, and after
getting locks, when the IO is about to start, vvp_io_fault_start()
checks the mtime's intactness.
It's a false alarm, since the timestamp from MDS could be stale,
we maintain mtime mainly on OST objects, and if the check in
vvp_io_fault_start() happens before mtime on OST objects are merged,
it will get wrong timestamp from the inode, even the timestamp it
fetched in vvp_io_fault_init() could be wrong in the first place.
This patch remove the mtime check in vvp_io_fault_start().
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: Iff0a7f3ae789b443927fe2f3487829f357a93a4a
Reviewed-on: http://review.whamcloud.com/19162
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Andreas Dilger [Sat, 13 Feb 2016 10:11:50 +0000 (03:11 -0700)]
LU-7746 tests: some fixes for sanity.sh with upstream kernel
sanity.sh:
- delete is_patchless() check, since this is always true today.
- remove test for "lfs getstripe --raw" added in 2.5.57
- minor style cleanups
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I6c808febe078fdc7afd89b82a04be622879cab07
Reviewed-on: http://review.whamcloud.com/19663
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Yang Sheng [Tue, 19 Apr 2016 03:56:26 +0000 (11:56 +0800)]
LU-8038 tests: test case for lfs migrate --block
Add a test case to cover 'lfs migrate --block'.
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Change-Id: I59bd0e9f881a37f3416ac07063921985f0313342
Reviewed-on: http://review.whamcloud.com/19645
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Yang Sheng [Mon, 18 Apr 2016 18:43:51 +0000 (02:43 +0800)]
LU-6601 tests: test case validate OSP status op
Add a test case to verify status change is effective for OSP.
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Change-Id: I11a6968756f66989ba87063424f6618475814901
Reviewed-on: http://review.whamcloud.com/19637
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jian Yu [Fri, 15 Apr 2016 07:54:20 +0000 (00:54 -0700)]
LU-8015 utils: fix lr_link() to get correct src and dest
The current codes in lr_link() of lustre_rsync.c can not
find out the correct info->src and info->dest in the
following situations:
1) the source name ends with the destination name
2) the file has two links with the same name but
in different directories
This patch fixes the above issue by using the path of the
parent FID and target name to construct info->dest, and
comparing the path of the target FID with info->dest to
get the correct info->src.
Test-Parameters: testlist=lustre-rsync-test
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: Ia92b01c069bde01bd0f094017bd2502292afc7f8
Reviewed-on: http://review.whamcloud.com/19599
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bruno Faccini [Fri, 1 Apr 2016 12:31:55 +0000 (14:31 +0200)]
LU-7973 llite: lock i_lock/dcache_lock before __d_drop()
There has been several Lustre Client crashes reported by sites
running with Lustre versions 2.1/2.5, all showing the same
dentry->d_hash->next corrupted pointer cause.
This patch fixes a regression that has been introduced since a
long time by commit :
d10c9e05ab593f361fdfd27842766a1924e63e58
(LU-506 kernel: FC15 - support dcache scalability changes.)
where i_lock/dcache_lock protection usage has been removed and
that is likely to cause racy condition during dentry [un]hashing
and to be the root cause of these crashes.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: Ie1239058222f15f7e094d8fae4d37e4a36449707
Reviewed-on: http://review.whamcloud.com/19287
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Christopher J. Morrone [Wed, 30 Mar 2016 01:26:00 +0000 (21:26 -0400)]
LU-5725 ofd: Expose OFD site_stats through proc
While exposed for the MDT the lu_object cache site stats are not
available for the OFD. This small patch makes those same stats
available under /proc/fs/lustre/obdfilter/lustre-OST0001/site_stats.
Change-Id: Ic5207d5471cd0f8eacc5a3363c12975c8880224d
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/19247
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Tue, 29 Mar 2016 15:09:47 +0000 (18:09 +0300)]
LU-7937 utils: Fix NULL pointer dereference
Check the error of malloc() function and handle it properly.
Change-Id: I2763d1b6d9c97c137a2593077ee31eaa3758ad81
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: http://review.whamcloud.com/19194
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Alex Zhuravlev [Wed, 23 Mar 2016 12:32:54 +0000 (15:32 +0300)]
LU-7902 osd: no need to set same [acm]time
osd_attr_set() has [acm]time cached in the objects. so if the caller
wants to set them to the same value, there is no point to call very
expensive SA update procedure.
Change-Id: I4e06306198496170f5d9c486a8d245c0978c0f79
Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-on: http://review.whamcloud.com/19088
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Yang Sheng [Thu, 25 Feb 2016 16:16:31 +0000 (00:16 +0800)]
LU-1118 llite: Invoke file_update_time in page_mkwrite
Upstream commit
41c4d25f78c01ede13efee1f2e979f3f35dd26f6
Only update file times if page_mkwrite is not set. So we
need call file_update_time by ourselves. Since all we
supported kernel have included this change. Then config
check is unnecessary.
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Change-Id: I1477ea6c6576df01af28afdf0dbaa7b012a7238e
Reviewed-on: http://review.whamcloud.com/18683
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Fri, 6 Nov 2015 15:39:17 +0000 (09:39 -0600)]
LU-7402 llite: correct request handling after ll_lookup_it()
In the FIFO cases of ll_atomic_open() and ll_lookup_nd() remove
spurious calls to ptlrpc_req_finished(). Explain that these cases are
unreachable in practice anyway.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ic7d677674ed7f9e52d039bb81259f60bc75ba71b
Reviewed-on: http://review.whamcloud.com/17068
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Kit Westneat [Fri, 15 Apr 2016 03:05:58 +0000 (23:05 -0400)]
LU-8001 nodemap: fix null deref when reclassifying
There are some situations in which an export's connection can be
NULL. This patch adds a NULL check before dereferencing in
nm_member_reclassify_nodemap. If the connection is null, keep export
in current nodemap since it should get reclassified on reconnect.
Signed-off-by: Kit Westneat <kit.westneat@gmail.com>
Change-Id: I795afdac3b388e883bafe5a79c8237794bc01ef6
Reviewed-on: http://review.whamcloud.com/19595
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tatsushi Takamura [Fri, 15 Apr 2016 01:04:46 +0000 (10:04 +0900)]
LU-8009 utils: Lustre_rsync does not sync files on MDT1
This patch fixes the 3rd parameter of llapi_changelog_start()
to refer a changelog correctly.
Signed-off-by: Tatsushi Takamura <takamr.tatsushi@jp.fujitsu.com>
Change-Id: I84ff553b299648159b8107fd9be486db44829ef2
Reviewed-on: http://review.whamcloud.com/19594
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Tue, 23 Feb 2016 03:33:55 +0000 (11:33 +0800)]
LU-5070 osd-zfs: get object's FID from its LMA EA
Do not assume that the FID-in-dirent is always valid. If the
FID-in-dirent is absent or invalid, then the osd_dir_lookup
needs to find out the FID from the target object's LMA EA.
It is the similar logic as osd-ldiskfs does.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: If75ff4ed03bd75a3385666842e94838ece3f4344
Reviewed-on: http://review.whamcloud.com/18595
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Artem Blagodarenko [Tue, 22 Dec 2015 09:15:15 +0000 (12:15 +0300)]
LU-7592 osd-ldiskfs: increase supported ldiskfs fs size limit
Change "force_over_128tb" mount option to "force_over_256tb".
Signed-off-by: Alexander Zarochentsev <alexander.zarochentsev@seagate.com>
Signed-off-by: Artem Blagodarenko <artem.blagodarenko@seagate.com>
Seagate-bug-id: MRP-2840
Reviewed-by: Artem Blagodarenko <artem.blagodarenko@seagate.com>
Reviewed-by: Alexey Leonidovich Lyashkov <alexey.lyashkov@seagate.com>
Tested-by: Elena V. Gryaznova <elena.gryaznova@seagate.com>
Change-Id: Ib53c4375a170823997a5a6b9ef6d832e8b3e0c72
Reviewed-on: http://review.whamcloud.com/17702
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Wed, 21 Oct 2015 02:24:58 +0000 (10:24 +0800)]
LU-5070 utils: mkfs.lustre code cleanup
Re-order the options to increase the code readability.
Remove unused code, redundant logic, white spaces,
blank lines, and so on.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I88ee52fb2425e9e256b7751bad4b6cfcd88011ec
Reviewed-on: http://review.whamcloud.com/15303
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Ben Evans <bevans@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Hongchao Zhang [Tue, 15 Dec 2015 21:04:13 +0000 (05:04 +0800)]
LU-7851 mdt: skip SOM related configuration
the SOM is deprecated, then the old SOM configuration should be
skipped by the newer version when upgrading Lustre.
Change-Id: Ibb32e43501479aac75c8e02a652bea8d8a727445
Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Reviewed-on: http://review.whamcloud.com/18834
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Wang Shilong [Tue, 19 Apr 2016 04:19:30 +0000 (12:19 +0800)]
LU-8037 fileset: reject dotdot as a path component
Administrator might access Lustre internal files by
using dotdot path for example:
mount -t lustre ip@tcp:/lustre/.. /mnt/lustre
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Change-Id: I137e12aa3f2feacb8549eb85f966c9dc8426f3af
Reviewed-on: http://review.whamcloud.com/19646
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Li Xi <lixi@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Thu, 14 Apr 2016 19:32:34 +0000 (12:32 -0700)]
LU-8024 kernel: kernel update [SLES12 SP1 3.12.57-60.35]
Update target and kernel_config files for new version
Test-Parameters: clientdistro=sles12 testgroup=review-ldiskfs \
mdsdistro=sles12 ossdistro=sles12 mdsfilesystemtype=ldiskfs \
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: Ic9f33c4877f93249b37726d8ad60c98ee624f719
Reviewed-on: http://review.whamcloud.com/19593
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Fri, 5 Jun 2015 21:39:23 +0000 (16:39 -0500)]
LU-7986 hsm: update actions llog in place
Update HSM actions llog entries in place rather than doing an append
and cancel.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I5c5282d199e43fbc6709d5dc9bc050e0b98fb889
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-on: http://review.whamcloud.com/19447
Tested-by: Jenkins
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Tested-by: Maloo <hpdd-maloo@intel.com>
John L. Hammond [Thu, 4 Jun 2015 14:48:44 +0000 (09:48 -0500)]
LU-7992 hsm: close KUC pipe before unregistering
In llapi_hsm_copytool_unregister() close the read side of the KUC pipe
before unregistering the copytool to avoid deadlocking with a ldlm_cb
thread.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I844086b32a3b2958ecaec391accded0ca06f60a2
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-on: http://review.whamcloud.com/19442
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Frank Zago <fzago@cray.com>
Dmitry Eremin [Fri, 8 Apr 2016 15:20:01 +0000 (18:20 +0300)]
LU-7998 utils: fix fd lost
A file descriptor for "/dev/loop-control" was never closed.
Change-Id: I0dc1cb2453f49322315f00266308c829c04b4792
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-on: http://review.whamcloud.com/19408
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Ashley Pittman [Wed, 2 Mar 2016 12:24:12 +0000 (12:24 +0000)]
LU-7835 checkpatch: Update gerrit_checkpatch for new features.
Allow checking of a single patch per invocation rather than running as
daemon. Update wiki URL for coding standards. Supply --notree option
allowing upstream checkpatch.pl to be used. Add linebreaks to output
when reporting multiple errors per line of source
Signed-off-by: Ashley Pittman <ashley.m.pittman@intel.com>
Change-Id: I3823db55073b606fb84b84f520e598e8427b22e7
Signed-off-by: Ashley Pittman <ashley.m.pittman@intel.com>
Reviewed-on: http://review.whamcloud.com/18748
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Andriy Skulysh [Tue, 1 Mar 2016 22:15:55 +0000 (00:15 +0200)]
LU-7791 ldlm: signal vs CP callback race
a lock isn't canceled on the server on reception
of a signal on the client when it is blocked waiting
for a lock to be granted, thus client is evicted for
not responding to LDLM_FL_AST_SENT.
Seagate-bug-id: MRP-3314
Change-Id: Ie025bb58b13fb2d741119f74c87439f917983268
Signed-off-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Reviewed-on: http://review.whamcloud.com/18498
Tested-by: Jenkins
Reviewed-by: Vitaly Fertman <vitaly.fertman@seagate.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Vitaly Fertman [Tue, 1 Mar 2016 23:46:31 +0000 (02:46 +0300)]
LU-7434 ptlrpc: lost bulk leads to a hang
The reverse order of request_out_callback() and reply_in_callback()
puts the RPC into UNREGISTERING state, which is waiting for RPC &
bulk md unlink, whereas only RPC md unlink has been called so far.
If bulk is lost, even expired_set does not check for UNREGISTERING
state.
The same for write if server returns an error.
This phase is ambiguous, split to UNREG_RPC and UNREG_BULK.
Signed-off-by: Vitaly Fertman <vitaly.fertman@seagate.com>
Change-Id: Ib1eeb1777ad1ab4c7ea1c83fe95dc9ae82c1894c
Seagate-bug-id: MRP-2953, MRP-3206
Reviewed-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Reviewed-by: Alexey Leonidovich Lyashkov <alexey.lyashkov@seagate.com>
Tested-by: Elena V. Gryaznova <elena.gryaznova@seagate.com>
Reviewed-on: http://review.whamcloud.com/17221
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Chris Horn <hornc@cray.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Parinay Kondekar [Wed, 6 Apr 2016 04:43:45 +0000 (10:13 +0530)]
LU-7323 test: Unknown symbol mb_cache_create (err 0)" on RHEL7
The error message is seen if mbcache module isn't loaded.
Before RHEL7 it was part of ext4, but EL7 have mbcache as
separate module and needs an extra `modprobe mbcache` command
before tests are run from local build dir.
This fixes the function load_modules_local() in test-framework.sh
Seagate-bug-id: MRP-3018
Change-Id: I7f1317355785c191287788bcb177d78b12231c97
Signed-off-by: Alexey Lyashkov <alexey.lyashkov@seagate.com>
Signed-off-by: Parinay Kondekar <parinay.kondekar@seagate.com>
Reviewed-on: http://review.whamcloud.com/16900
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Kyrylo Shatskyy [Wed, 13 Apr 2016 06:29:32 +0000 (11:59 +0530)]
LU-7130 test: sanity/162 doesn't handle DIR values properly
The test fails when path values have trailing slash,
e.g. DIR=/mnt/lustre/
The patch adds validation which at the moment processes
these parameters-DIR, DIR1, DIR2, MOUNT, MOUNT1, MOUNT2.
Change-Id: Ief85e8798911cb67b4dafce985a9e4ef56cc26a4
Signed-off-by: Kyrylo Shatskyy <kyrylo.shatskyy@seagate.com>
Signed-off-by: Parinay Kondekar <parinay.kondekar@seagate.com>
Seagate-bug-id: MRP-334
Reviewed-on: http://es-gerrit.xyus.xyratex.com:8080/#/c/269
Tested-by: Jenkins
Reviewed-by: Andrew Perepechko <Andrew_Perepechko@xyratex.com>
Reviewed-by: Alexander Lezhoev <Alexander_Lezhoev@xyratex.com>
Tested-by: Alexander Lezhoev <Alexander_Lezhoev@xyratex.com>
Reviewed-by: Vitaly Fertman <Vitaly_Fertman@xyratex.com>
Reviewed-on: http://review.whamcloud.com/16358
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Parinay Kondekar [Wed, 9 Sep 2015 12:09:39 +0000 (17:39 +0530)]
LU-7121 tests: Extra items shown in the stack trace in t-f
The call to functions error_exit(), error_noexit(), error()
in turn call function log_trace_dump().
This call along with important information about where the
particular test failed shows some extra information in the
dump trace.
This changes fix the problem noticed and dumps only the relevant
information instead.
Signed-off-by: Parinay Kondekar <parinay.kondekar@seagate.com>
Change-Id: I5fba23f6e296e77feb0339c5e20815c288b8b751
Seagate-Bug-Id: MRP-891
Reviewed-on: http://es-gerrit.xyus.xyratex.com:8080/721/
Tested-by: Jenkins
Reviewed-by: Elena V. Gryaznova <elena.gryaznova@seagate.com>
Tested-by: Elena V. Gryaznova <elena.gryaznova@seagate.com>
Reviewed-by: Kyrylo Shatskyy <kyrylo_shatskyy@xyratex.com>
Signed-off-by: Kyrylo Shatskyy <kyrylo_shatskyy@xyratex.com>
Reviewed-on: http://review.whamcloud.com/16338
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Mon, 11 Apr 2016 20:01:00 +0000 (16:01 -0400)]
LU-6245 libcfs: replace CFS_MODULE_PARAM with linux kernel module api
Remove the CFS_MODULE_PARM wrapper and use the linux macros
module_param and MODULE_PARM_DESC instead.
Change-Id: If551518b34195caf92435189f98b01895a0b0d24
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-on: http://review.whamcloud.com/19383
Reviewed-by: Frank Zago <fzago@cray.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
git://git.whamcloud.com / fs / lustre-release.git / log