Whamcloud - gitweb
Dmitry Eremin [Mon, 28 Oct 2013 18:38:58 +0000 (22:38 +0400)]
LU-4069 build: cleanup from GOTO(label, -ERRNO)
Cleanup the code from GOTO(label, -ERRNO) and other bad GOTOs.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I58df89810c112a6856da83c56c9f29dc22ffdf99
Reviewed-on: http://review.whamcloud.com/8080
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Fri, 15 Nov 2013 17:40:42 +0000 (12:40 -0500)]
LU-2800 autoconf: block device release changed in newer kernels.
For kernels at 3.10 and above the release method for block
devices no longer return a integer. This impacts the lloop
block device we use on the lustre client side.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: Ib63e2897a48989f1cf5185a1c88a5ddbf989c0a9
Reviewed-on: http://review.whamcloud.com/5380
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Nunez [Fri, 1 Nov 2013 18:18:05 +0000 (12:18 -0600)]
LU-3867 utils: llapi_create_volatile_idx uses 0 instead idx
If llapi_create_volatile_idx() is passed an MDT index other than
-1, it always uses 0 as the MDT index rather than the one provided.
This patch removes the use of LUSTRE_VOLATILE_IDX and uses
LUSTRE_VOLATILE_HDR with the MDT index instead, uses the provided
MDT index in the volatile file name and uses base 16 when reading
the MDT index from the file name using strtoul.
Signed-off-by: James Nunez <james.a.nunez@intel.com>
Signed-off-by: Bobbie Lind <bobbie.j.lind@intel.com>
Change-Id: I27e2beb8524683aba07c48297614dd02d7957af4
Reviewed-on: http://review.whamcloud.com/7613
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
James Simmons [Wed, 30 Oct 2013 12:26:33 +0000 (08:26 -0400)]
LU-2800 autoconf: clean up sysctl table handling
Before linux kernel version 2.6.24 came out lustre
like other subsystems registering with sysctl had
to assign unquie numbers for ctl_name in struct
ctl_table. After 2.6.24 this requirement changed
and lustre started to use CTL_UNNUMBERED as the
default value. Since we no longer support 2.6.24
and earlier kernels we can remove this test and
simply the code. We also simplifed the handling
of the strategy field in struct ctl_table. The
final fix is that several LND drivers never handled
that case of 2.6.33 and above kernels which removed
the ctl_name and strategy fields altogether. This
patch addresses this so LND drivers can be handled
better on distros using newer kernels.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I0f5b64552e7d54abf4b73bfe285109573725edaf
Reviewed-on: http://review.whamcloud.com/8102
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
James Simmons [Tue, 5 Nov 2013 18:47:26 +0000 (13:47 -0500)]
LU-4015 ldiskfs: access external journal using device path
Currently ldiskfs when configured with an external journal
uses the device number to access it. The problem with this
approach is that the device number can change across
reboots. This patch adds a new mount option -o journal_path
to tell ldiskfs where the external journal is instead.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I43fff89c2bebeff4f0305cdc24cefa534fb9fe67
Reviewed-on: http://review.whamcloud.com/7777
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Sebastien Buisson [Wed, 2 Oct 2013 15:09:35 +0000 (17:09 +0200)]
LU-4049 build: fix 'NULL pointer dereference' errors
Fix 'NULL pointer dereference' defects found by Coverity version
6.6.1:
Dereference after null check (FORWARD_NULL)
For instance, Passing null pointer to a function which dereferences
it.
Dereference before null check (REVERSE_INULL)
Null-checking variable suggests that it may be null, but it has
already been dereferenced on all paths leading to the check.
Dereference null return value (NULL_RETURNS)
Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Change-Id: I1db4e80fd2c048cc20459e8da4cb61a97b3171cb
Reviewed-on: http://review.whamcloud.com/7827
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Wed, 13 Nov 2013 15:41:59 +0000 (10:41 -0500)]
LU-2880 ldiskfs: rename patches for consistency
Rename the ext4_data_in_dirent.patch, ext4_pdirop.patch, and
ext4_extra_isize.patch files to be ext4-data-in-dirent.patch,
ext4-extra-isize.patch and ext4-pdirop.patch respectively.
This is more consistent with all of the other patches in the
series, and avoids missing these files with tab completion.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I92125f09284e23ab08d5e9f8658e68416bde025f
Reviewed-on: http://review.whamcloud.com/6411
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>
James Simmons [Fri, 15 Nov 2013 13:16:01 +0000 (08:16 -0500)]
LU-3319 procfs: provide framework for seq_file handling
In linux 3.10 kernels and above the proc file system interface
changed. To handle this change we need to move from the current
method of using proc_read and proc_write function pointers to
using the seq_file. This patch provides the base to enable moving
all the lustre proc handling gradually over to this new method
of proc data handling.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I06d941398bf248bffab851f9f25bea6bbe438a30
Reviewed-on: http://review.whamcloud.com/7135
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
wang di [Tue, 26 Nov 2013 14:24:54 +0000 (06:24 -0800)]
LU-4316 mdt: Do not grant Layout lock for remote file
Do not grant layout lock for remote file, because only
the MDT, where the file/dir resides, should grant the
LAYOUT lock.
Signed-off-by: wang di <di.wang@intel.com>
Change-Id: I174d54f3ce5b7c69cb35aa63f4322b743ea677e8
Reviewed-on: http://review.whamcloud.com/8406
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Bobi Jam [Fri, 15 Nov 2013 10:21:40 +0000 (18:21 +0800)]
LU-4245 ldlm: do not fill empty lvbo buffer
Don't fill empty lvbo buffer if server does not prepare it.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I22a2ddd016dcc43c00b2b3d6be5d18921bf26d67
Reviewed-on: http://review.whamcloud.com/8285
Tested-by: Jenkins
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Ned Bass [Mon, 4 Nov 2013 23:07:11 +0000 (15:07 -0800)]
LU-4208 osd-zfs: hold pool config lock to register property
- Hold the DSL pool configuration lock when calling
dsl_prop_register(). Failure to do so will panic the node if
assertions are enabled in ZFS. This change requires a build of ZFS
on Linux that exports symbols dsl_pool_config_enter and
dsl_pool_config_exit, which was done in commit 40a806d [1], and will
appear in ZFS release 0.6.3.
- Fix up variable declaration alignment in osd_mount().
- Add check for exported symbols in autoconf
[1] https://github.com/zfsonlinux/zfs/commit/40a806d
Signed-off-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ib2df05301ca73847d49ca2b183d72b9e2d957c10
Reviewed-on: http://review.whamcloud.com/8172
Tested-by: Jenkins
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Li Wei <wei.g.li@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Christopher J. Morrone [Fri, 15 Nov 2013 21:40:19 +0000 (13:40 -0800)]
LU-4194 ldlm: Make OBD_[ALLOC|FREE]_LARGE use consistent
struct ldlm_lock's l_lvb_data field is freed in ldlm_lock_put()
using OBD_FREE. However, some other code paths can attach
a buffer to l_lvb_data that was allocated using OBD_ALLOC_LARGE.
This can lead to a kfree() of a vmalloc()ed buffer, which can
trigger a kernel Oops.
Change-Id: Ic75a67530862eeb4d065c14bbbac80939bff5731
Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Reviewed-on: http://review.whamcloud.com/8298
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Bruno Faccini [Thu, 14 Nov 2013 16:20:00 +0000 (17:20 +0100)]
LU-4194 ldlm: set l_lvb_type coherent when layout is returned
In case layout has been packed into server reply when not
requested, lock l_lvb_type must be set accordingly.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: Iaf54c9ba27785e529f4f2fb967d2fad4fc1dfbcb
Reviewed-on: http://review.whamcloud.com/8270
Tested-by: Jenkins
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Mikhail Pershin [Mon, 11 Nov 2013 17:42:43 +0000 (21:42 +0400)]
LU-793 ptlrpc: allow client to reconnect with RPC in progress
Since not letting clients with pending RPCs to reconnect caused a lot
of problems, this patch allows such a reconnection and would only
refuse to process requests with the same xids that are already being
processed.
Bulk requests are aborted upon reconnection by comparing connection
count of request and export.
Signed-off-by: Mikhail Pershin <mike.pershin@intel.com>
Change-Id: Ic4b23a71fd288df02d1040d98867373ae06b60f6
Reviewed-on: http://review.whamcloud.com/4960
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Reviewed-by: Alexander Boyko <alexander_boyko@xyratex.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
wang di [Wed, 27 Nov 2013 19:03:51 +0000 (11:03 -0800)]
LU-4322 tests: disable sanity 101a in DNE
Disable sanity 101a in DNE test for now.
Signed-off-by: wang di <di.wang@intel.com>
Change-Id: I67531bf9f1fb75b8cd6684d864793a612e1d1cf4
Reviewed-on: http://review.whamcloud.com/8428
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Jian Yu [Wed, 27 Nov 2013 12:34:34 +0000 (20:34 +0800)]
LU-14 tests: add version check code to conf-sanity test 69
This patch adds Lustre version check codes into conf-sanity
test 69 to make the test work with servers which do not have
the following patch:
Lustre-commit:
db6613f5bed1606cc8f97b46d1b298746af03a75
Lustre-change: http://review.whamcloud.com/7443
Test-Parameters: envdefinitions=SLOW=yes,ENABLE_QUOTA=yes,ONLY=69 \
ossjob=lustre-b2_4 mdsjob=lustre-b2_4 ossbuildno=45 mdsbuildno=45 \
testlist=conf-sanity
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I4e736fd5d760872f88364f30a39430f961419c64
Reviewed-on: http://review.whamcloud.com/8411
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Niu Yawei [Tue, 26 Nov 2013 10:04:43 +0000 (05:04 -0500)]
LU-4310 kernel: missed unlock in quota patch
The quota-replace-dqptr-sem.patch introduced a regression which
could miss unlock inode->i_lock in __dquot_alloc_space().
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: If0f240fe76d661909c13b11e3d384e77da1662b1
Reviewed-on: http://review.whamcloud.com/8400
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Tested-by: Jenkins
Reviewed-by: Andrew Perepechko <andrew_perepechko@xyratex.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Jian Yu [Mon, 25 Nov 2013 03:55:58 +0000 (11:55 +0800)]
LU-3738 tests: improve posix.sh to support BASELINE_FS=zfs
This patch improves posix.sh to support BASELINE_FS=zfs. It also
adds create_zpool(), create_zfs() and destroy_zpool() common
functions into test-framework.sh.
Test-Parameters: envdefinitions=SLOW=yes \
mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs \
ostcount=2 testlist=posix
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I7e59181ef5f3926c6a8799ac4f6c775d04350a99
Reviewed-on: http://review.whamcloud.com/8180
Tested-by: Jenkins
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Mikhail Pershin [Fri, 16 Aug 2013 05:10:33 +0000 (09:10 +0400)]
LU-3467 ptlrpc: initialize request session early
Initialize request session early to make it available in
high-priority handlers
Signed-off-by: Mikhail Pershin <mike.pershin@intel.com>
Change-Id: Ic4f5967d3ba8e723bf90accf20742cb4c6bc801c
Reviewed-on: http://review.whamcloud.com/7350
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Jenkins
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Mikhail Pershin [Wed, 14 Aug 2013 09:14:51 +0000 (13:14 +0400)]
LU-3467 target: unified transaction callbacks
Use common transaction callbacks to update last_rcvd
in target
Signed-off-by: Mikhail Pershin <mike.pershin@intel.com>
Change-Id: Ib10d0b958fd9f09753246c85f272e1f917069d01
Reviewed-on: http://review.whamcloud.com/7330
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Lai Siyao [Fri, 15 Nov 2013 09:22:57 +0000 (17:22 +0800)]
LU-3286 llog: local storage doesn't need vfsmount
After introduction of local storage device vfsmount is redundant,
and servers don't need this struct to access local storage based
llog.
* remove lsi_srv_mnt from lustre_sb_info.
* remove lmi_mnt from lustre_mount_info.
* remove ddp_mnt from dt_device_param.
Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Change-Id: I8cfdda8c871301ec9a332a01fd0a2523fe1a09c2
Reviewed-on: http://review.whamcloud.com/8286
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Wed, 20 Nov 2013 12:51:21 +0000 (07:51 -0500)]
LU-1199 rpm: don't open-code %_smp_mflags
In some distant past, %_smp_mflags didn't exist on all
supported platforms.
We can use it now. The only difference is the lack of an
8-CPU limit, which can be overridden by explicitly defining
RPM_BUILD_NCPUS anyway.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I0b5c2f879da9ba6e8aa3103bfe59e6b43f8d72bc
Reviewed-on: http://review.whamcloud.com/5492
Tested-by: Jenkins
Reviewed-by: Christopher J. Morrone <chris.morrone.llnl@gmail.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Niu Yawei [Tue, 19 Nov 2013 06:56:10 +0000 (01:56 -0500)]
LU-4270 test: fix sanity test_209
Fix the connect_flags checking in test_209 of sanity.sh
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I6c34b9dbe6d3b7475d85588e7adb3acb762fab32
Reviewed-on: http://review.whamcloud.com/8326
Tested-by: Jenkins
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Bruno Faccini [Tue, 19 Nov 2013 11:00:51 +0000 (12:00 +0100)]
LU-4093 tests: fix typo and inverted test in original patch
1st/original patch for LU-4093 ticket, patch-set #3 of Change at
http://review.whamcloud.com/8157/ with Id
I3274e5864469251e6f2040377b84429e4f218f47, contains a typo and
an inverted test that prevent it to work as expected but rather
with reversed logic.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I3f8031b85db116711312b8f7decf2c33761b89e6
Reviewed-on: http://review.whamcloud.com/8329
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
James Simmons [Tue, 17 Sep 2013 17:50:59 +0000 (13:50 -0400)]
LU-2800 compat: remove crypto shims
Now that we've removed the tests for the crypto APIs, we can
remove all of the ll_* prefixed compatibility shims from the
rest of the code.
Change-Id: I9a10b57d48614f892dd61f31ee4bee8fbea93ea8
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: Thomas Stibor <thomas@stibor.net>
Reviewed-on: http://review.whamcloud.com/5347
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Bob Glossman [Tue, 19 Nov 2013 22:51:09 +0000 (14:51 -0800)]
LU-4276 ldiskfs: enable read/write access by default
Add build time config option to allow read/write access by default.
While the new CONFIG_LDISKFS_FS_RW only matters in SLES11 builds,
it's easiest to just add the flag to all builds unconditionally.
It will be ignored in builds where it doesn't matter.
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: Icaab7a4fd003dd29a49af9222852bded7ec344ff
Reviewed-on: http://review.whamcloud.com/8335
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
wang di [Sun, 24 Nov 2013 21:05:21 +0000 (13:05 -0800)]
LU-4305 tests: mkdir instead of test_mkdir in 39P
In sanity 39p, it should use mkdir, instead of
test_mkdir to avoid creating remote directory before
the real "lfs mkdir -i".
Signed-off-by: wang di <di.wang@intel.com>
Change-Id: If5a09bf4bdadd1f27d66e9d7b252ae0f4ee071fb
Reviewed-on: http://review.whamcloud.com/8383
Tested-by: Jenkins
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Bob Glossman [Mon, 4 Nov 2013 15:46:20 +0000 (07:46 -0800)]
LU-4204 tests: fix test_69 in conf-sanity
Recent commit added new test 69 to conf-sanity,
but calls test_68 by mistake. This patch fixes that typo.
In addition the test in its original form could never have worked.
There were several functional problems in it.
This patch fixes these up too.
Test-Parmeters: testlist=conf-sanity envdefinitions=SLOW=yes,ONLY=69
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I44d1e5f620a3e4cc14b59842abdd2f56ed409872
Reviewed-on: http://review.whamcloud.com/8158
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Nathaniel Clark [Wed, 7 Aug 2013 22:18:05 +0000 (18:18 -0400)]
LU-3689 tests: Support ZFS in large-lun.sh and run_llverdev
This adds support to "check" zfs and run llrevfs against zfs backed
nodes. This fixes some regressions added by LU-3264, specifically
that the stop function also exports the zfs pool.
The zfs "check" is really a non-entity since there is no functional
zfs fsck at this time.
Test-Parameters: mdsfilesystemtype=zfs mdtfilesystemtype=zfs ostfilesystemtype=zfs testlist=large-lun ostcount=2 envdefinitions=SLOW=yes
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Idc35d35152f1cbdc4afe6a0f627a5c1ba40b3faf
Reviewed-on: http://review.whamcloud.com/7395
Tested-by: Jenkins
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Nathaniel Clark [Mon, 18 Nov 2013 13:48:14 +0000 (08:48 -0500)]
LU-2059 tests: Skip insanity/10 on ZFS
This test fails in the same as all the other LU-2059 skipped tests,
trying to start OST with MDS down causes OST to hang.
Test-Parameters: osscount=2 mdscount=2 mdsfilesystemtype=zfs mdtfilesystemtype=zfs ostfilesystemtype=zfs testlist=insanity
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I29cec88c0f3f5eb4ffef0117363b0862f44dd61a
Reviewed-on: http://review.whamcloud.com/8316
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Fan Yong [Sat, 26 Oct 2013 20:56:28 +0000 (04:56 +0800)]
LU-4165 tests: skip sanity-lfsck test_2c for 2.4 or older
It makes no sense to run sanity-lfsck test_2c against 2.4 or older.
Test-Parameters: mdsjob=lustre-b2_4 ossjob=lustre-b2_4 mdsbuildno=58 ossbuildno=58 testlist=sanity-lfsck
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I96792b4325a69f880e326dc8963cf3e6bd09bf87
Reviewed-on: http://review.whamcloud.com/8386
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Fan Yong [Wed, 23 Oct 2013 06:19:03 +0000 (14:19 +0800)]
LU-4058 tests: skip sanity-scrub test_15 for 2.4.1 or older
Lustre-2.4.1 or older does not support dry-run mode OI scrub yet.
Test-Parameters: mdsjob=lustre-b2_4 ossjob=lustre-b2_4 \
mdsbuildno=54 ossbuildno=54 testlist=sanity-scrub
Signed-off-by: Fan Yong <fan.yong@intel.com>
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I1a97f850ab39d7d42ef7ed6e191ce66b3b093415
Reviewed-on: http://review.whamcloud.com/7868
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
James Nunez [Wed, 20 Nov 2013 17:49:17 +0000 (10:49 -0700)]
LU-3879 test: Check for e2fsprogs on server
sanity-quota checks if e2fsprogs is loaded on the client,
but should be checking for e2fsprogs on the server. This
patch changes checking for dumpe2fs from the client to the MDS.
Signed-off-by: James Nunez <james.a.nunez@intel.com>
Change-Id: I83b0b5b99deec376f4ee2d59e30810373e330174
Reviewed-on: http://review.whamcloud.com/8346
Tested-by: Jenkins
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Jian Yu [Wed, 13 Nov 2013 12:43:24 +0000 (20:43 +0800)]
LU-3906 tests: fix space estimation codes in run_compilebench()
While running compilebench test, the actual space required for each
kernel directory is about 1GB instead of 680MB. This patch fixes this
number in run_compilebench() so as to make the space estimation
more correct.
Test-Parameters: envdefinitions=SLOW=yes,ONLY=compilebench \
testlist=parallel-scale,parallel-scale-nfsv3,parallel-scale-nfsv4
Test-Parameters: envdefinitions=SLOW=yes,ONLY=compilebench \
mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs \
ostcount=2 mdtcount=1 \
testlist=parallel-scale,parallel-scale-nfsv3,parallel-scale-nfsv4
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I8aa14673b8df99f6b50980ee091acda76da7cd9a
Reviewed-on: http://review.whamcloud.com/8258
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Tue, 5 Nov 2013 03:39:58 +0000 (19:39 -0800)]
LU-3321 obdclass: Add a preallocated percpu cl_env
This change adds support for a single preallocated cl_env per CPU
which can be used in circumstances where reschedule is not possible.
Currently this interface is only used by the ll_releasepage function.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: Prakash Surya <surya1@llnl.gov>
Change-Id: I14a06294f0c2caae8806d7da134a8076f75ddc81
Reviewed-on: http://review.whamcloud.com/8174
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Tue, 18 Dec 2012 19:05:55 +0000 (14:05 -0500)]
LU-2469 tests: use MDS/OSTDEV and MDS/OSTDEVBASE with zfs
Add support for using MDSDEVn, OSTDEVn, MDSDEVBASE and OSTDEVBASE
variables when testing with zfs.
Add *ZFSDEVn variable for specifying ZFS format devices.
Remove use of *DEVn use outside of test-framework.sh
MDSDEV et al. specify physical device formatted.
Formerly Change-Id I89115ce935294cdbbd6620fec3f9dc3eb76f8960
Test-Parameters: mdsfilesystemtype=zfs ostfilesystemtype=zfs mdtfilesystemtype=zfs testlist=sanity
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: If6506995d9d61f9c97827f45071f0d317c7a09f9
Reviewed-on: http://review.whamcloud.com/5016
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Li Wei <wei.g.li@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Tue, 15 Oct 2013 18:25:53 +0000 (12:25 -0600)]
LU-1538 tests: run sanity-lfsck instead of lfsck
Run sanity-lfsck in the default acceptance-small test list instead of
the old lfsck test, which will be deprecated and removed in the 2.6
release.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I5903d854dc16d3ed8bca1139355a73230e500c1e
Reviewed-on: http://review.whamcloud.com/7952
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Stephen Champion [Tue, 29 Oct 2013 04:13:49 +0000 (12:13 +0800)]
LU-4123 tests: use nodename in check_write_access
Check for a file create using the remote nodename.
When the node name of a host does not exactly match the host name
passed via xxx_HOST parameters, check_write_access can check for
creation of a filename different from what the remote node created.
By querying and using the nodename from the remote host, the node
initiating the tests can check for the same file name that the remote
node creates.
Test-Parameters: testlist=lfsck
Signed-off-by: Stephen Champion <schamp@sgi.com>
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I57f081597395578c1a37e16531477e6fa4867c8d
Reviewed-on: http://review.whamcloud.com/8009
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
James Simmons [Thu, 14 Nov 2013 13:26:09 +0000 (08:26 -0500)]
LU-2800 autoconf: move LC_BLK_QUEUE_MAX_SEGMENTS test
This patch is a cosmetic in that we move the test for
blk_queue_max_segements out of the quota test section
and place it in the list were it makes sense.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: Ib2bc9515051bd4ba5dc72911fa1b770db041c5be
Reviewed-on: http://review.whamcloud.com/5383
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
James Simmons [Wed, 13 Nov 2013 17:03:34 +0000 (12:03 -0500)]
LU-3974 llite: fix mkdir endless loop
Running on 3.11-rc4 kernel, I got below endless loop. It
turns to be that Lustre always saves the first page of a
dir inode mapping at index ~0UL. And after commit
5a720394
(mm: teach truncate_inode_pages_range() to handle non page
aligned ranges), truncate_inode_pages_range() _NO LONGER_
truncates the page that is sitting at index ~0UL.
[16768.998006] mkdir R running task 0 2717 2716 0x00000080
[16768.998073]
000000000000000e 0000000000000000 0000000000000000 ffff88000be00460
[16768.998157]
ffff88000ea65908 ffffffff810fec3e ffff88000ea65968 ffff8800229e7750
[16768.998241]
ffff88000ea658b8 0000000000000000 0000000000000000 ffff88000ea65958
[16768.998326] Call Trace:
[16768.998401] [<
ffffffff810fc6ed>] ? rcu_read_unlock+0x1c/0x2d
[16768.998473] [<
ffffffff810fec3e>] ? find_get_pages+0xf5/0x11b
[16768.998530] [<
ffffffff811078f0>] ? pagevec_lookup+0x20/0x2a
[16768.998586] [<
ffffffff8110920e>] ? truncate_inode_pages_range.part.2+0x161/0x39a
[16768.998680] [<
ffffffffa02ad5dc>] ? ll_md_blocking_ast+0x338/0x62f [lustre]
[16768.998744] [<
ffffffff8110947f>] ? truncate_inode_pages_range+0x38/0x3f
[16768.998805] [<
ffffffff811094f8>] ? truncate_inode_pages+0x12/0x14
[16768.998871] [<
ffffffffa02ad6e8>] ? ll_md_blocking_ast+0x444/0x62f [lustre]
[16768.998948] [<
ffffffff810981b5>] ? arch_local_irq_save+0x9/0xc
[16768.999022] [<
ffffffffa07ee0e8>] ? ldlm_cancel_callback+0x67/0x12a [ptlrpc]
[16768.999100] [<
ffffffffa07f85b2>] ? ldlm_cli_cancel_local+0xf3/0x2bc [ptlrpc]
[16768.999176] [<
ffffffffa07f9163>] ? ldlm_cli_cancel_list_local+0x7e/0x1e4 [ptlrpc]
[16768.999268] [<
ffffffffa07f9473>] ? ldlm_cancel_resource_local+0x1aa/0x1b9 [ptlrpc]
[16768.999385] [<
ffffffffa0657bad>] ? mdc_resource_get_unused+0xf8/0x115 [mdc]
[16768.999472] [<
ffffffff8109c887>] ? trace_hardirqs_on+0xd/0xf
[16768.999533] [<
ffffffffa06583d8>] ? mdc_create+0x11e/0x4db [mdc]
[16768.999597] [<
ffffffff8152ed84>] ? mutex_unlock+0xe/0x10
[16768.999654] [<
ffffffffa0350e99>] ? lmv_create+0x355/0x3e9 [lmv]
[16768.999712] [<
ffffffff811553b7>] ? final_putname+0x35/0x39
[16768.999775] [<
ffffffffa02ae167>] ? ll_new_node+0x33b/0x3ff [lustre]
[16768.999841] [<
ffffffffa02ae62c>] ? ll_mkdir+0xf2/0x127 [lustre]
[16768.999897] [<
ffffffff81156996>] ? vfs_mkdir+0x84/0xc9
[16768.999961] [<
ffffffff81158cf8>] ? SyS_mkdirat+0x77/0xad
[16769.000014] [<
ffffffff81158d47>] ? SyS_mkdir+0x19/0x1b
[16769.000066] [<
ffffffff81538652>] ? system_call_fastpath+0x16/0x1b
Signed-off-by: Peng Tao <bergwolf@gmail.com>
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: Ib0898ce908271b155ee9ef46ea53059cb78d0c63
Reviewed-on: http://review.whamcloud.com/8237
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Peng Tao [Mon, 22 Jul 2013 06:04:19 +0000 (14:04 +0800)]
LU-1330 obdecho: splits metadata echo client from client-side
Client-side module needs not to support MDS echo client. Comment out
(using HAVE_SERVER_SUPPORT) it to facilitate obdclass' split/cleanup.
Signed-off-by: Liu Xuezhao <xuezhao.liu@emc.com>
Signed-off-by: Peng Tao <tao.peng@emc.com>
Change-Id: Ie70d9a9bd665878d4ce7327c288fb55abd60c878
Reviewed-on: http://review.whamcloud.com/2674
Tested-by: Jenkins
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@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>
Bobi Jam [Thu, 17 Oct 2013 07:25:35 +0000 (15:25 +0800)]
LU-4106 mdt: avoid recursive lu_object_find an object
LU-3240 (commit
762f2114d282a98ebfa4dbbeea9298a8088ad24e) set parent
dir fid the same as child fid in getattr by fid case we should not
lu_object_find() the object again, could lead to hung if there is a
concurrent unlink destroyed the object.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I75256c0fa684877cbd2e1f36f8ab2ac3faab2989
Reviewed-on: http://review.whamcloud.com/7990
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Thomas Stibor [Thu, 10 Oct 2013 13:33:33 +0000 (15:33 +0200)]
LU-4085 build: gss/krb5 is disabled despite functions found
The function available check for krb5int_derive_key and
krb5_derive_key by means of AC_CHECK_LIB($gssapi_lib,...)
does not set shell variables HAVE_KRB5_DERIVE_KEY,
HAVE_KRB5INT_DERIVE_KEY so the test
AS_IF([test "x$HAVE_KRB5INT_DERIVE_KEY" = "x1"
-o "x$HAVE_KRB5_DERIVE_KEY" = "x1"] gives false despite
the functions are found. As a side effect this fix
revealed that calls with real prototypes are required.
Thus prototypes of krb5_derive_key(...) and
krb5int_derive_key(...) are provided, and required
structures from krb5-int.h are copied.
Signed-off-by: Thomas Stibor <thomas@stibor.net>
Change-Id: I8c4cce5f861c9e4d01071024c4b852a2274c1b40
Reviewed-on: http://review.whamcloud.com/7913
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Sebastien Buisson [Wed, 2 Oct 2013 06:43:08 +0000 (08:43 +0200)]
LU-4044 build: fix 'resource leak' errors
Fix 'resource leak' defects found by Coverity version 6.6.1:
Resource leak (RESOURCE_LEAK)
Variable going out of scope leaks the storage it points to.
Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I1039fa8bd5e82acab0ffcb082d35fca28dfb694c
Reviewed-on: http://review.whamcloud.com/7820
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Nunez [Thu, 31 Oct 2013 22:07:54 +0000 (16:07 -0600)]
LU-4116 test: Correct error message search
replay-dual test_18 looks for "entering recovery in server" from
the local (client) console log. The problem is that this string
is split across multiple lines and the message should be searched
for on the MDS.
The error message to search for is changed to "not entering
recovery" and we are looking for it on the MDS.
Signed-off-by: James Nunez <james.a.nunez@intel.com>
Change-Id: Ia091f8c389f4a7321c2565cd85b6cead74ed68d7
Reviewed-on: http://review.whamcloud.com/8129
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Isaac Huang [Tue, 22 Oct 2013 06:21:37 +0000 (00:21 -0600)]
LU-4006 lnet: abort messages whose MD has been unlinked
If LNetMDUnlink has been called, all outgoing messages
on that MD should be aborted before lnet_ni_send() is
called.
Signed-off-by: Isaac Huang <he.huang@intel.com>
Change-Id: I3ebe6d3f4ecc21cd6f00aebfb98b90109501df5a
Reviewed-on: http://review.whamcloud.com/8041
Tested-by: Jenkins
Reviewed-by: Liang Zhen <liang.zhen@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 14 Oct 2013 17:34:13 +0000 (12:34 -0500)]
LU-4101 mdt: protect internal xattrs
In mdt_reint_setxattr() require CAP_SYS_ADMIN to modify a trusted
xattr and silently disallow modification those trusted xattrs used by
Lustre internally.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ic616dca74a90da0aedb0ec2624618f91ac6fcaf4
Reviewed-on: http://review.whamcloud.com/7943
Reviewed-by: Andrew Perepechko <andrew_perepechko@xyratex.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Vladimir Saveliev [Sat, 2 Nov 2013 15:44:11 +0000 (19:44 +0400)]
LU-4030 tests: use free_fd() to allocate file descriptor
free_fd() lists /proc/self/fd to find the smallest unused file
descriptor
sanity test_31n is changed to use free_fd() instead of using hardcoded
173
sanity test_236 is changed to use free_fd() instead of using "{FD}<>"
which is not available on eariler bash
Signed-off-by: Vladimir Saveliev <vladimir_saveliev@xyratex.com>
Change-Id: Ie13da99e4a758d7169454c7a6b3b58c594b30ecc
Reviewed-on: http://review.whamcloud.com/8181
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Liang Zhen [Wed, 27 Feb 2013 07:54:52 +0000 (15:54 +0800)]
LU-2745 lnet: fix failure handle of create reply
error handler of lnet_create_reply_msg() didn't release lnet_res_lock
if lnet_msg_alloc() failed.
It can be fixed by moving validation check of msg out from lock.
Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Change-Id: I5a25b7e07d24f5c4f165a3266cca6e79dffc1461
Reviewed-on: http://review.whamcloud.com/5542
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Isaac Huang <he.huang@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
wang di [Fri, 15 Nov 2013 08:00:05 +0000 (00:00 -0800)]
LU-4223 tests: delete loop device when umount
Delete loop device with umount -d, to avoid loop device leaking.
Signed-off-by: wang di <di.wang@intel.com>
Change-Id: I55bafefa1d387116333aafc71b599ef888c57a2b
Reviewed-on: http://review.whamcloud.com/8296
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Stephen Champion [Fri, 18 Oct 2013 22:02:45 +0000 (15:02 -0700)]
LU-4124 build: make module installation directory flexible
Add --with-kmp-moddir option to configure.
Distributions vary in the installation directory for kernel modules.
The RHEL standard installation directory is
/lib/modules/$(uname -r)/extra
while the SLES standard is
/lib/modules/$(uname -r)/updates
Adding the option --with-kmp-moddir option to configure allows users
to select the appropriate installation target. With this change, it
is neccessary to support both options in the test framework.
Signed-off-by: Stephen Champion <schamp@sgi.com>
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: Iec3137e0e5039dd43622c2e285030a5339fa6fd3
Reviewed-on: http://review.whamcloud.com/8065
Reviewed-by: James Simmons <uja.ornl@gmail.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>
Swapnil Pimpale [Fri, 25 Oct 2013 12:20:11 +0000 (17:50 +0530)]
LU-4082 llite: Chk ret of ll_prep_md_op_data in ll_dir_filler
ll_dir_filler() now checks the returned pointer from
ll_prep_md_op_data() using IS_ERR()
Signed-off-by: Swapnil Pimpale <spimpale@ddn.com>
Change-Id: Ia4321db2ca9dddaa4487f62544de10239120bad3
Reviewed-on: http://review.whamcloud.com/8073
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bruno Faccini [Sat, 26 Oct 2013 05:59:39 +0000 (07:59 +0200)]
LU-4065 mdt: Ensure accurate coordinator state
Some control must be added in procfs interface to
ensure that the coordinator state will be kept accurate.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I41e5b42ddbdbcb652c202ee01fa120126e78f3d2
Reviewed-on: http://review.whamcloud.com/8074
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Fri, 3 May 2013 04:05:48 +0000 (22:05 -0600)]
LU-1993 llite: debugging for ll_file_open LASSERT
Add debugging for LASSERTF(it_disposition(it, DISP_ENQ_OPEN_REF)
in ll_file_open(), since this is a rarely hit failure under racer,
and it would be useful to get more information if this is hit
again. Print the full intent disposition, as well as the status,
in case Oleg's earlier comment about bailing out on any error is
actually the case here.
Test-Parameters: envdefinitions=DURATION=1800 testlist=racer
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Ie03f706bcb32b03d94c84c5d80dae197813ebbe5
Reviewed-on: http://review.whamcloud.com/6250
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@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>
Andriy Skulysh [Mon, 23 Sep 2013 06:31:57 +0000 (09:31 +0300)]
LU-3992 libcfs: Fix NUMA emulated mode
Kernel commit
c1c3443c9c5e9be92641029ed229a41563e44506
assigns all allowed cpus to emulated node.
End cpt initialization loop when all CPUs are assigned.
Change-Id: I4cf22e4cf58cd8377cf19af30217fec72503be69
Signed-off-by: Andriy Skulysh <Andriy_Skulysh@xyratex.com>
Reviewed-on: http://review.whamcloud.com/7724
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Sebastien Buisson [Thu, 6 Jun 2013 12:42:37 +0000 (14:42 +0200)]
LU-2744 build: fix 'data race condition' issues
Fix 'data race condition' defects found by Coverity version
6.5.0:
Data race condition (MISSING_LOCK)
Accessing variable without holding lock. Elsewhere,
this variable is accessed with lock held.
Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Change-Id: Ib5477c494feedd0b4ab052ccd9db9ea99f47101b
Reviewed-on: http://review.whamcloud.com/6575
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
wang di [Fri, 11 Oct 2013 07:00:08 +0000 (00:00 -0700)]
LU-3240 mdt: comments on LOOKUP and PERM lock
1. Add more comments for MDS_INODELOCK_PERM and
MDS_INODELOCK_LOOKUP.
2. Add sanity 39p for checking nlink of remote
directories.
Signed-off-by: wang di <di.wang@intel.com>
Change-Id: I58a697f2aee02924f716333d8619873131db84f7
Reviewed-on: http://review.whamcloud.com/7937
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Lai Siyao [Tue, 27 Aug 2013 09:58:08 +0000 (17:58 +0800)]
LU-3544 llite: simplify dentry revalidate
Lustre client dentry validation is protected by LDLM lock, so
any time a dentry is found, it's valid and no need to revalidate
from MDS, and even it does, there is race that it may be
invalidated after revalidation is finished.
Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Change-Id: I7700cbaddc4ec08e12c9f7d8021783a6135dd35a
Reviewed-on: http://review.whamcloud.com/7475
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Peng Tao <bergwolf@gmail.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Lai Siyao [Tue, 29 Oct 2013 02:37:06 +0000 (10:37 +0800)]
LU-3544 fid: add a connect flag for open by FID
Add OBD_CONNECT_OPEN_BY_FID for open by FID, if MDS supports this
, for open by FID, it won't retry with name if object with the FID
doesn't exist; while if client supports this, client won't pack name in
open request if FID is known.
Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Change-Id: I158b96e8f55fe357b3542d320e894aeca18fca94
Reviewed-on: http://review.whamcloud.com/8093
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Niu Yawei [Mon, 28 Oct 2013 16:03:15 +0000 (12:03 -0400)]
LU-4139 quota: improve write performance when over softlimit
To improve the write performance when over softlimit, we choose
to lose some grace time accuracy: shrinking the qunit size to
(4 * least_qunit_size) instead of least_qunit_size.
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I20e341d53979ac42227197ec6fa04dd38974073f
Reviewed-on: http://review.whamcloud.com/8078
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Nathaniel Clark [Mon, 11 Nov 2013 20:05:45 +0000 (15:05 -0500)]
LU-3889 tests: fix parallel-scale typo and allow overrides
This fixes a typo introduced by change #7778. This also allows
overriding all variables that are set lower due to SLOW=no, which is
forced for ZFS tests. This will allow values specified by caller for
all variables even when testing on ZFS.
Test-Parameters: envdefinitions=SLOW=yes mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs testlist=parallel-scale
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I5b585ffa28c9f7d729d5169ff96c383cc0af38d3
Reviewed-on: http://review.whamcloud.com/8234
Reviewed-by: Cliff White <cliff.white@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Jinshan Xiong [Mon, 30 Sep 2013 22:00:38 +0000 (15:00 -0700)]
LU-3321 clio: collapse layer of cl_page
Move radix tree to osc layer to for performance improvement.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I93e3cb8352f7be41c23465b12945874316aa1809
Reviewed-on: http://review.whamcloud.com/7892
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Mon, 30 Sep 2013 21:36:56 +0000 (14:36 -0700)]
LU-3321 osc: to drop LRU pages with cl_lru_work
This way we can drop it async.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: Id67c41b0f56201183c07bbbc0369ceb32fe3bbb3
Reviewed-on: http://review.whamcloud.com/7891
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Mon, 28 Oct 2013 12:35:55 +0000 (08:35 -0400)]
LU-2800 auotconf: remove linux key-type.h test
Since linux kernel version 2.6.24 key-type.h has
been available so we no longer need to test this.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I6ed82effb62a0dadabb8477e419c42e8540c91d9
Reviewed-on: http://review.whamcloud.com/5546
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Thomas Stibor <thomas@stibor.net>
Reviewed-by: Andrew Korty <ajk@iu.edu>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Niu Yawei [Mon, 4 Nov 2013 12:34:27 +0000 (07:34 -0500)]
LU-4028 quota: improve lfs quota output
- print the total allocated block/inode limit for all OSTs/MDTs
when the verbose option is specified;
- add 'h' option to print used space in human readable format;
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: Idc9cd3d92a50744885f65e82c9ce53b332c1959a
Reviewed-on: http://review.whamcloud.com/8152
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Bruno Faccini [Mon, 4 Nov 2013 13:59:55 +0000 (14:59 +0100)]
LU-4176 tests: do not wait for deleted HSM request record
restore_and_check_size() function of sanity-hsm was systematically
waiting for restore request status, even if already flagged as
succeed. This extra wait can become forever if request record has
already been deleted.
Also fix wrong restore timing msg.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I9709d9e665b66fb5d16e93c7139fa8d95fa92589
Reviewed-on: http://review.whamcloud.com/8154
Tested-by: Jenkins
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Swapnil Pimpale [Thu, 5 Sep 2013 12:27:00 +0000 (17:57 +0530)]
LU-3810 obd: Cleanup client import if client_obd_setup fails
client_obd_setup() allocates an obd_import which should be cleaned up
if there is any failure afterwards in callers of client_obd_setup().
This patch fixes the bug in osc_setup(), mgc_setup(), mdc_setup() and
lwp_setup(). The fix is to call obd_cleanup_client_import() before
calling client_obd_cleanup() in case of an error.
Signed-off-by: Swapnil Pimpale <spimpale@ddn.com>
Change-Id: Ib8c2194d80fd14538d4ca42fb8084c93dd575199
Reviewed-on: http://review.whamcloud.com/7561
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Dmitry Eremin [Thu, 31 Oct 2013 12:35:25 +0000 (16:35 +0400)]
LU-3885 ldlm: check all errors during ldlm_proc_setup()
Fix ignorig errors from lprocfs_add_vars() function.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: Ie2429f91ad5ac6de8d7e9027a0cd2907588294d3
Reviewed-on: http://review.whamcloud.com/8115
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Bobi Jam [Thu, 29 Aug 2013 00:00:08 +0000 (08:00 +0800)]
LU-3853 tests: mpiexec requires strict global/local args order
Put -machinefile option ahead of other mpiexec options, since some
version of mpiexec (at least 1.90) requests global options be put
before its local options, and -machinefile is a global options.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: Ib188881a78a2b345552e8c6070f7dfa7a086ce5d
Reviewed-on: http://review.whamcloud.com/7493
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Li Wei [Wed, 20 Mar 2013 12:06:20 +0000 (20:06 +0800)]
LU-2353 tests: Introduce back end mount/unmount helpers
Some tests need to access back end file systems directly. Currently,
for ldiskfs targets, they either mount the devices as ldiskfs or
invoke the ldiskfs-specific debugfs tool; for ZFS targets, they simply
skip forward. Many of them could actually work on both back end
types. Since the mount/unmount procedures diff for ldiskfs and ZFS
back ends, this patch introduces a pair of generic helpers,
mount_fstype()/unmount_fstype(), to hide the differences from tests.
Then, sanity 27z is adapted to work with ZFS targets using the new
helpers, as an example.
Test-Parameters: mdsfilesystemtype=zfs ostfilesystemtype=zfs mdtfilesystemtype=zfs
Change-Id: I0b6b358b147240519f28852ba2c8fea85c09aab3
Signed-off-by: Li Wei <wei.g.li@intel.com>
Reviewed-on: http://review.whamcloud.com/5785
Tested-by: Jenkins
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Mon, 4 Nov 2013 11:35:11 +0000 (06:35 -0500)]
LU-2829 tests: Decrease load on zfs for sanityn/33a-b
Decrease number of files when testing sanityn/33a and 33b when using
ZFS due to performance issues (cf LU-2887). ZFS performance when
commit_on_sharing is enabled is very slow.
Test-Parameters: ostcount=2 envdefinitions=SLOW=yes testlist=sanityn mdsfilesystemtype=zfs mdtfilesystemtype=zfs ostfilesystemtype=zfs
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I55b9267a3df68abc631b7d47886fb4b0f7e81594
Reviewed-on: http://review.whamcloud.com/8161
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Mikhail Pershin [Fri, 26 Jul 2013 11:29:22 +0000 (15:29 +0400)]
LU-3467 ofd: use unified handler for OST requests
Switch OST/OFD request processing to the unified request
handle.
Signed-off-by: Mikhail Pershin <mike.pershin@intel.com>
Change-Id: I27a2f8c2345b1aad3d714eab1e138d0c9d2ec258
Reviewed-on: http://review.whamcloud.com/7130
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
James Simmons [Wed, 30 Oct 2013 13:04:18 +0000 (09:04 -0400)]
LU-2800 autoconf: remove bio_endio args test
Since linux kernel version 2.6.24 the bio api has been
taking one less argument than in previous versions. Since
we no longer support kernels earlier than 2.6.32 we can
drop this test and simplify the code using the bio api.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: Ieedface5065d660284c5dfb9a61cb81e4174c61d
Reviewed-on: http://review.whamcloud.com/8104
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
James Simmons [Mon, 28 Oct 2013 12:15:46 +0000 (08:15 -0400)]
LU-2800 autoconf: remove LC_HAVE_QUOTAIO_H
fs/quota/quotaio_v2.h has been in the upstream kernel since v2.6.30.
We can remove the test. The result is only used for another test
that will also be removed.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I4e8c4be20548cd546c5787deb8a75f36ea4f1c9f
Reviewed-on: http://review.whamcloud.com/5371
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
James Simmons [Mon, 28 Oct 2013 11:56:02 +0000 (07:56 -0400)]
LU-2800 autoconf: remove LC_PROCFS_DELETED
The procfs deleted member in the procfs entry struct
has been removed since kernel version 2.6.27. We can
removed this test since it is no longer needed.
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I1dbf8cabfa85d31ea5f1386710dfcb7012946b91
Reviewed-on: http://review.whamcloud.com/5374
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Bruno Faccini [Mon, 4 Nov 2013 14:28:39 +0000 (15:28 +0100)]
LU-4109 tests: raise out-of inode LOV EA detection threshold
Sanity sub-test 57b sometimes failed because the threshold (< 8K
of MDT space) it used to detect if LOV EA has been stored out-of
inode is too short (vs space also allocated for llog files, ChangeLog,
etc). So raising threshold to 16.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I0f59328d59da69911473433769ff683ff4fdf452
Reviewed-on: http://review.whamcloud.com/8156
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
yangsheng [Sat, 21 Sep 2013 16:52:44 +0000 (00:52 +0800)]
LU-3373 osd-ldiskfs: struct file add f_inode member
struct file has a f_inode member. Need initalize
before use it.
Signed-off-by: yang sheng <yang.sheng@intel.com>
Change-Id: I290aaea09c4ff9be7f9d044bf972eb43510933d2
Reviewed-on: http://review.whamcloud.com/7727
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Wed, 18 Sep 2013 13:47:16 +0000 (09:47 -0400)]
LU-1199 rpm: don't clean up $RPM_BUILD_ROOT in %build
The %clean macro cleans up the buildroot. Doing it manually
makes the SUSE build system complain and breaks short-circuiting.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: I4a6b5ebfa34d9fcfddfb0608d4f7dcce3f88a3ef
Reviewed-on: http://review.whamcloud.com/5491
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Sebastien Buisson [Wed, 2 Oct 2013 09:29:14 +0000 (11:29 +0200)]
LU-4047 build: fix 'memory' errors
Fix 'memory corruption' defects found by Coverity version 6.6.1:
Out-of-bounds write (OVERRUN)
Overrunning array at byte offset using index.
Fix 'memory illegal accesses' defects found by Coverity version 6.6.1:
Read from pointer after free (USE_AFTER_FREE)
Dereferencing freed pointer.
Readlink used insecurely (READLINK)
buffer[sizeof(buffer)] is an off-by-one error.
Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Change-Id: I91af9fae93f89275cf6d3ad97bdc37958c0d709e
Reviewed-on: http://review.whamcloud.com/7822
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
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>
Amir Shehata [Wed, 18 Sep 2013 16:07:56 +0000 (09:07 -0700)]
LU-2149 llog: lctl to work with llogs from command line
- mgs_list_logs() is introduced to get all of configs and
mgs_iocontrol handles that and also cancel/remove ioctl
- remove obsoleted code from mdc_iocontrol
- add handler to list llog catalogs from mdt_iocontrol too
- lctl llog_print outputs records in YAML format
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: Ic6e392403da18a2d7578aafdddf01007c3b3c30e
Reviewed-on: http://review.whamcloud.com/4254
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
wang di [Fri, 1 Nov 2013 07:00:11 +0000 (00:00 -0700)]
LU-4179 mdt: skip open lock enqueue during resent
Skip open lock enqueue, if the open lock has been
acquired(mdt_intent_fixup_resent) during resent.
Signed-off-by: wang di <di.wang@intel.com>
Signed-off-by: Keith Mannthey <keith.mannthey@intel.com>
Change-Id: I78228112b0ba8f93e05930fdd042763d77700626
Reviewed-on: http://review.whamcloud.com/8142
Tested-by: Jenkins
Reviewed-by: Jinshan Xiong <jinshan.xiong@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>
Jian Yu [Fri, 8 Nov 2013 07:18:34 +0000 (15:18 +0800)]
LU-4219 mdd: limit os_namelen to the max of NAME_MAX
While running pathconf(".", _PC_NAME_MAX) on ldiskfs-based
Lustre filesystem, it returns 255 (NAME_MAX). However, on
zfs-based Lustre filesystem, it returns 256 (MAXNAMELEN).
Since Lustre can not handle the name length large than
NAME_MAX, this patch limits the value of os_namelen in
mdd_statfs() to the max of NAME_MAX.
Test-Parameters: envdefinitions=SLOW=yes \
mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs \
ostcount=2 testlist=posix
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I13b5398eff5b15c74b0e08e2a3394b79a3ff6a44
Reviewed-on: http://review.whamcloud.com/8217
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>
Bruno Faccini [Fri, 8 Nov 2013 08:24:21 +0000 (09:24 +0100)]
LU-4093 tests: prevent zombie requests when stopping CT
Fix to prevent any current requests to become zombie when stopping
copytool and thus not to sometimes to induce failures during next
sanity-hsm sub-tests.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I3274e5864469251e6f2040377b84429e4f218f47
Reviewed-on: http://review.whamcloud.com/8157
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>
wang di [Thu, 31 Oct 2013 07:01:15 +0000 (00:01 -0700)]
LU-4181 tests: disable lnet_selftest.sh for DNE.
Disable lnet_selftest for DNE to make review pass with DNE.
Signed-off-by: wang di <di.wang@intel.com>
Change-Id: Ie4abd794d0184955af84752f0a9c64181b85f1ec
Reviewed-on: http://review.whamcloud.com/8130
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Oleg Drokin [Wed, 6 Nov 2013 07:02:57 +0000 (02:02 -0500)]
New tag 2.5.51
Change-Id: Ib4389ce656c878704b329472ab3a848e1aff6b53
wang di [Tue, 29 Oct 2013 07:01:22 +0000 (00:01 -0700)]
LU-4149 tests: disable sanity_scrub(7, 15) for dne test.
Disable sanity_scrub(7, 15) for dne test, so make dne pass
on maloon right now.
Signed-off-by: wang di <di.wang@intel.com>
Change-Id: Ie6553f44655c8f38451ed2cb1086e5848ca033a4
Reviewed-on: http://review.whamcloud.com/8101
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Andreas Dilger [Tue, 1 Oct 2013 20:20:07 +0000 (14:20 -0600)]
LU-20 kernel: remove obsolete kernel patches
The jbd2-jcberr and lustre_version patches have been obsoleted since
Lustre 2.2. They were kept around to allow transitioning between
Lustre 1.8/2.1 and 2.4/2.5 but can be removed from Lustre 2.6 since
this does not allow a direct upgrade path anyway.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I5406272fe22580e48679c18dea562001d23ebbe5
Reviewed-on: http://review.whamcloud.com/7881
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Vladimir Saveliev [Mon, 28 Oct 2013 19:20:21 +0000 (23:20 +0400)]
LU-3920 tests: check MDS version before testing HSM feature
Sanity tests 65y and 229 fail when MDS does not have HSM
support. Check MDS version and skip the tests in that case.
Xyratex-bug-id: MRP-1417
Signed-off-by: Vladimir Saveliev <vladimir_saveliev@xyratex.com>
Change-Id: I80496ac67ec1b6b8032f465a0b5914bb12839d3b
Reviewed-on: http://review.whamcloud.com/8121
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Fan Yong [Wed, 4 Sep 2013 14:32:51 +0000 (22:32 +0800)]
LU-3335 osd: self repair LMA by the RPC service thread
Current OI scrub uses the FID in LMA to locate the OI mapping and
repair the inconsistent OI mapping. Such inode table based system
scanning does not guarantee all OI inconsistency can be repaired,
such as inode without LMA or with crashed LMA, or cannot find out
dummy OI mappings (the object for the OI mapping does not exist).
In fact, after the inode table based scanning, the OI scrub can
make another OI table based scanning to resolve above mentioned
issues. But such scanning will be nonlinear, so performance may
be bad.
Currently, we have some other relative simple solution although
they are not perfect. We can make the RPC service thread to fix
the found inconsistency by itself: when verify the LMA with the
given FID, and found no LMA, then generate LMA according to the
given FID and store the new LMA as the inode xattr. Usually, it
is rare case that the inode has no LMA. So such additional work
for the RPC service thread will not affect performance much.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: Ic6b8b533e56b6a0e388d5c38de7b113cb9bd6dad
Reviewed-on: http://review.whamcloud.com/7144
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Bruno Faccini [Thu, 31 Oct 2013 23:50:27 +0000 (00:50 +0100)]
LU-3973 tests: use -P option of df to avoid line breaks
Sometimes df command output requires to break lines to
allow for columns alignement. This causes fields parsing
to fail. Use -P option to select compatibility output
and ensure one single line format per file-system is used.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: Ibc9d4b8f50857e680a5590dbbfe312fce137bcb6
Reviewed-on: http://review.whamcloud.com/7915
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Vladimir Saveliev [Tue, 29 Oct 2013 16:01:23 +0000 (20:01 +0400)]
LU-2479 ldiskfs: do not check dir max size for regular files
ldiskfs_append() is used not only to extend directory but also to
maintain iam container. In later case ldiskfs_append() should not
check for max directory size. The iam container is distunguished as
regular file.
Signed-off-by: Vladimir Saveliev <vladimir_saveliev@xyratex.com>
Change-Id: I205b0b1149d7bd477b3965ce68a546dfd9dc6a6f
Reviewed-on: http://review.whamcloud.com/8137
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alexey Lyashkov <alexey_lyashkov@xyratex.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Andreas Dilger [Thu, 10 Oct 2013 18:32:34 +0000 (12:32 -0600)]
LU-2528 misc: quiet warning for 2.1/2.5 connections
The Lustre 2.5.4 client will print a warning about connections with
2.1.3 servers, yet they are still supposed to be interoperable.
Increase the window of warning to be up to Lustre 2.5.50, since we
do not intend to allow interoperability between 2.1 and 2.6 systems.
This was from the time when major releases like Lustre 1.4 and 1.8
were many years apart would have no chance for interoperation.
Only print this message once per client, to avoid flooding the console
for connections to many servers or frequent network reconnections.
Server versions should all be nearly the same in any case.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I503523f80144c00c5ea1d312cc9204d9123ebbe5
Reviewed-on: http://review.whamcloud.com/7916
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>
Li Xi [Sat, 31 Aug 2013 13:39:36 +0000 (06:39 -0700)]
LU-3858 test: fix sanity 27A failure caused by stripe delay
When clearing the stripe of root directory, the updating of stripe
might be delayed because of the processing delay of config log.
This will cause the failure of sanity 27A test.
This makes the test suit wait for the update of root directory
stripe.
Signed-off-by: Li Xi <lixi@ddn.com>
Change-Id: Ib4f96d9d4d0ae8a0b54e3e306a7abed758dbdf39
Reviewed-on: http://review.whamcloud.com/7495
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Hongchao Zhang [Sun, 21 Jul 2013 21:40:37 +0000 (05:40 +0800)]
LU-2613 recovery: free open/close request promptly
- For the non-create open or committed open, the open request
should be freed along with the close request as soon as the
close done, despite that the transno of open/close is
greater than the last committed transno known by client or not.
- Move the committed open request into another dedicated list,
that will avoid scanning a huge replay list on receiving each
reply (when there are many open files).
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Change-Id: I1a25a35fe7a16681368d92d16964680b6209b3ee
Reviewed-on: http://review.whamcloud.com/6665
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Fri, 30 Aug 2013 18:29:50 +0000 (22:29 +0400)]
LU-3968 lbuild: Extend script with build for Xeon Phi card
Automatically download, compile and produce Lustre client RPMs
for Xeon Phi(TM) card if "--mpss-version" option is specified
for contrib/lbuild/lbuild script.
Also try to compile with Xeon Phi(TM) OFED if it's available.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: Ida07d764dc824c13f22ffb53d24e2c6f79ce3573
Reviewed-on: http://review.whamcloud.com/7066
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Brian J. Murrell <brian.murrell@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Nunez [Thu, 31 Oct 2013 19:43:25 +0000 (13:43 -0600)]
LU-4016 tests: use user and group names in metadata-updates
pdsh eats %u passed to stat executed on remote nodes. Avoid this
problem by using the user and group names instead of numerical
identifiers in metadata-updates ownership checks.
Test-Parameters: testlist=metadata-updates
Signed-off-by: James Nunez <james.a.nunez@intel.com>
Signed-off-by: Stephen Champion <schamp@sgi.com>
Change-Id: Id0a283d58873bb64772f28cbc28c5a6abad9a48e
Reviewed-on: http://review.whamcloud.com/8052
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Thu, 26 Sep 2013 17:39:49 +0000 (13:39 -0400)]
LU-2887 tests: disable SLOW for ZFS in parallel-scale
Because all zfs pools are on the same physical device in autotest
ZFS backed OSTs run VERY slowly so disable SLOW. Decrease load on zfs
backed MDS.
Test-Parameters: envdefinitions=SLOW=yes,ENABLE_QUOTA=yes mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs testlist=parallel-scale
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I99d980adcf4259f8431c351ce6fecb1088419130
Reviewed-on: http://review.whamcloud.com/7778
Tested-by: Jenkins
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Emoly Liu [Wed, 9 Oct 2013 15:51:36 +0000 (23:51 +0800)]
LU-4060 test: skip sanityn.sh test_74 on older than 2.4.93 MDS
2.5 client 2.4 server interop testing shows that sanityn.sh test_74
needs to be skipped for Lustre MDS versions older than 2.4.93 .
Test-Parameters: testlist=sanityn clientdistro=el6 clientarch=x86_64 serverdistro=el6 serverarch=x86_64 serverjob=lustre-b2_4 serverbuildno=46
Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Change-Id: I92d29a8db3ed5d145bd7db7bd18f5593ba80576c
Reviewed-on: http://review.whamcloud.com/7911
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Jinshan Xiong [Mon, 30 Sep 2013 21:24:17 +0000 (14:24 -0700)]
LU-3321 osc: Adjustment on osc LRU for performance
Add and discard pages from LRU in batch.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I90287b73c05066d420a31bff21866dfa1ffec665
Reviewed-on: http://review.whamcloud.com/7890
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Mon, 30 Sep 2013 21:09:26 +0000 (14:09 -0700)]
LU-3321 clio: revert LU-2622 for removing global env list
Revert "LU-2622 obdclass: Remove the global cl_env list"
This reverts commit
d5390bbde9b525b13ab91f5610abe1c4bb98eff9.
The performance dropped a lot when memory reclaim process kicked
in as ll_releasepage() was called to destroy lustre pages. It turned
out that big overhead to allocate cl_env and keys on the fly so we
have to revert this patch.
The problem of LU-2622 will be solved in another patch later on.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: If5f9faeabbe78d3a88b25bf986a2c9a2ecc9ba12
Reviewed-on: http://review.whamcloud.com/7888
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
git://git.whamcloud.com / fs / lustre-release.git / log