LU-1312 kernel: crash at boot time in isci driver

Restore SG_ALL to default value to avoid crash isci.

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I855ba8c7669b749fded51f3b0316f115d18e0fcd
Reviewed-on: http://review.whamcloud.com/2595
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1335 build: include lustre srpm in build

Add a support for building lustre-*.src.rpm

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: I301915c107e50ea5d1a3275ecc631f07aa2b78be
Reviewed-on: http://review.whamcloud.com/2576
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-447 add lctl --net XXX push

In order to clear out peer/conn data in the gnilnd for testing after
adding lnet_notify(see LU-446), we need an alternative to
lctl --net gni del_peer - as this nukes the peer and results in false
lnd_query failures.

Change-Id: Ie8029953a2881c0d6e3ac250101d2d4374bbf3c1
Signed-off-by: Wally Wang <wang@cray.com>
Reviewed-on: http://review.whamcloud.com/1311
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Liang Zhen <liang@whamcloud.com>
Reviewed-by: Lai Siyao <laisiyao@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>

LU-1340 release: bump version to 1.8.7.81-wc1

Yet another build which brings us closer to 1.8.8-wc1 RC1.

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: I7bac34ab4d29dc459171265319807c221a564c1a
Reviewed-on: http://review.whamcloud.com/2590

LU-995 utils: make lfs getstripe directory output consistent.

"lfs getstripe" should report the global default for any fields with
a value that means "use the default". This patch introduces the
following functionality:

1. If "lfs getstripe" is called on a directory and finds that striping
   EA is not set, the filesystem's defaults are looked up and printed.
2. If the striping EA is set, but the striping count and/or striping
   size has a value that means "use the default" (count = 0 and/or
   size = 0), the filesystem's default for that specific striping
   attribute is looked up and printed.
3. A new option to "lfs getstripe" is introduced; the "--raw" or
   "-R" option. If this option is specified, the previous two checks
   are skipped. In other words, if the striping EA is not set, 0, 0,
   -1, will be printed for the striping count, size, and offset
   respectively. Also, if the striping EA is set, the values will be
   printed without first converting them into their respective
   defaults.

This patch relies on the /proc filesystem to determine each
filesystem's default striping attributes, and a cache is maintained
which holds the default values for the last filesystem queried.

See Also:
Bugzilla #23802, https://bugzilla.lustre.org/show_bug.cgi?id=23802

Signed-off-by: Prakash Surya <surya1@llnl.gov>
Signed-off-by: Hongchao Zhang <hongchao.zhang@whamcloud.com>
Change-Id: Ic5b616dd83775c0a15be5c060b11cfbba05c4fbb
Reviewed-on: http://review.whamcloud.com/2117
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-891 test: waiting import state for next step.

Anyway, There still has a rare chance that the request meet
a invalid import after recovery. So we should waiting import
restore to a certain state and then doing next operation.

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I20bed347a16755ccaf102d4c67b0a7e87b1318a1
Reviewed-on: http://review.whamcloud.com/2248
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-734 tests: add sub-tests into recovery-*-scale tests

This patch adds sub-tests into the recovery-*-scale tests
so that test results and logs could be gathered properly
and uploaded to Maloo.

The patch also does some cleanup works on the test scripts
and moves some common functions into test-framework.sh.

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I514143e1fa29aad289d215174dbc68d2740da73b
Reviewed-on: http://review.whamcloud.com/2508
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Cliff White <cliffw@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-577 tests: FAIL replay-single test_70b rundbench load

Test 70b for replay-single assumes that lustre is mounted on
/mnt/lustre which is not the case for us. This patch passes
the proper MOUNT. The test also was not using the standard
DIR/tdir setup which had generated data files not being
cleaned up. Increased the sleep period to match dbench's
warm up period. This gives dbench a change to start up when
using many clients.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I3a793db35aa21d57220d0de1a9e92486e65ae21a
Reviewed-on: http://review.whamcloud.com/2518
Tested-by: Hudson
Reviewed-by: Wei Liu <sarah@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-1062 tests: incorrect path to configuration file

this is a bit modified port of the
http://review.whamcloud.com/#change,1877
(author Andreas Dilger)

Xyratex-bug-id: MRP-480

Reviewed-by: Sergey Glushchenko <Sergey_Glushchenko@xyratex.com>
Reviewed-by: Andriy Skulysh <andriy_skulysh@xyratex.com>
Signed-off-by: Iurii Golovach <iurii_golovach@xyratex.com>
Change-Id: I10b42de2c2d453e23142c01290742153dff7262a
Reviewed-on: http://review.whamcloud.com/2419
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-427 test: Test failure on test suite lfsck

- Reset $MDSDB & $OSTDB in generate_db(). Else they will
  stale if user redefine $SHARED_DIRECTORY.
- Add a function check_shared_dir() to ensure
  $SHARED_DIRECTORY is shared among tests nodes.
- Fix check_logdir() and check_write_access() to avoid using
  node.$(hostname).yml files which should not be deleted.

Signed-off-by: Yang Sheng <ys@whamcloud.com>
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Ie7d1d08c0d2c701fa9fb74ef8b252fa8b31bf111
Reviewed-on: http://review.whamcloud.com/2498
Tested-by: Hudson
Reviewed-by: Wei Liu <sarah@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-313 tests: re-enable lfsck test to run by default

Due to bug 13698, the lfsck part of the lfsck.sh test script was
disabled by default. After the fixes in LU-113 were landed, lfsck
should work again. Remove SKIP_LFSCK checks so lfsck.sh actually runs
lfsck instead of silently skipping it unless SKIP_LFSCK=no is set.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I430f7398b2d21db0d0755726fdcb6053f25b4b10
Reviewed-on: http://review.whamcloud.com/2497
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Wei Liu <sarah@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-1052 kernel: Kernel update [RHEL5.8 2.6.18-308.1.1.el5]

Update RHEL5.8 kernel to 2.6.18-308.1.1.el5.

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: Id8394bc1cb57d9555e4cda1312200a23b0915c84
Reviewed-on: http://review.whamcloud.com/2175
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>

LU-935 quota: break early when b/i_unit_sz exceeded upper limit

While expanding b/i_unit_sz in dquot_create_oqaq(), we'd break the loop
early when the b/i_unit_sz exceeded upper limit, otherwise, qaq_b/iunit_sz
could be overflow and result in endless loop.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I6046c8b86cad00c8c2fa2d24d0e21821a3c02209
Reviewed-on: http://review.whamcloud.com/1887
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-1241 kernel: Kernel update [RHEL6.2 2.6.32-220.7.1.el6]

Update RHEL6.2 kernel to 2.6.32-220.7.1.el6.(patchless client)

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I0933095b4e412dbb4ff826f45196d3b6aa80aecc
Reviewed-on: http://review.whamcloud.com/2401
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-847 quota: client retrieve quota usage directly

Current 'lfs quota' sends getquota RPC to MDS, and MDS is responsible
for retrieving disk usage from all targets, this scheme will be
changed to client retrieving disk usage from all targets directly.

This patch addresses the compatibility issue as well: If the getquota
returned by MDS has QIF_SPACE, client just trust the disk usage
returned by MDS, otherwise, client has to issue RPCs to collect disk
usage by itself.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: I35526c453033809d2f23b1806b4783f5011e3fa2
Reviewed-on: http://review.whamcloud.com/1570
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-974 protocol: change OBD_CONNECT_GRANT_PARAM

Change the OBD_CONNECT_GRANT_PARAM flag value to avoid conflict
with the OBD_CONNECT_UMASK flag from LU-974.  While that patch is
not yet landed to our release tree, it is in use in production at
some customers.  While the risk of conflict is currently low, it
is easier to change the GRANT_PARAM value since only in use on the
orion branch, and isn't even handled by the client there yet.

Add (hopefully) clear comments for OBD_CONNECT and obd_connect_data
to ensure that they are not modified in some incompatible way across
branches.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I503892c3b595c0272b0941fa58a16a496318cab0
Reviewed-on: http://review.whamcloud.com/2298
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-649 io: DIO doesn't need lock i_mutex

Revert "LU-649 io: DIO doesn't need lock i_mutex"
This reverts commit 4ebec99e309fcafd1be8a0b10673d69a083ae87f.

There is a concern that this patch is wrong, so i'm reverting it until
the situation is clarified.

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: Ia7e12eaeeb87ef30d073e28391a25ffcd8ad146f
Reviewed-on: http://review.whamcloud.com/2398

LU-249 mds: Increase the maximum number of threads on MDS

Currently the largest number of the threads that can exist on
the MDS is 512. For very large system is those threads can
easily be exhausted and end up back logging the system. This
patch allows the user to define the upper limit of how many
threads are created on the MDS. If not selected the default
behavior of 512 is kept.

Change-Id: Ie48eb00f48ddcd98cc518178ba5e1da9ea1e40eb
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Reviewed-on: http://review.whamcloud.com/2354
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-1018 test: Reduce compilebench run time

We decided to reduce the run time for compilebench
so it would run under 1h on VM nodes. This is similar
to master

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: Ie664ad131e4fbce9443eb78d06cbe36d6002b881
Reviewed-on: http://review.whamcloud.com/2005
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Cliff White <cliffw@whamcloud.com>
Reviewed-by: Wei Liu <sarah@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-554 LNet: add gnilnd awareness to LNet

This allows servers on any network to talk to gnilnd routers.
See Oracle 23884 attachment 31892

Change-Id: I6989c987b2bd0d60058a290567f8ef0bee56420b
Signed-off-by: Chris Horn <hornc@cray.com>
Reviewed-on: http://review.whamcloud.com/1179
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-1221 ldiskfs: Replace sysname with nodename in MMP

sysname holds "Linux" by default, i.e. what appears when
doing a "uname -s"; nodename should be used to print the
machine's hostname, i.e. what is returned when doing a
"uname -n" or "hostname", and what
gethostname(2)/sethostname(2) manipulate, in order to
notify the administrator of the node which is contending
to mount the filesystem.

Andreas says this was introduced when porting the MMP
patches from RHEL5 to RHEL6, and then also pushed upstream
to ext4; a patch for upstream ext4 has already been
submitted.

MRP-222
Signed-off-by: Nikitas Angelinas <nikitas_angelinas@xyratex.com>
Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: Ie5c8e99ae536c2239134140589d678d3a4f92639
Reviewed-on: http://review.whamcloud.com/2337
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-1198 idl: move FID VER to DLM resource name[1]

Until Lustre 1.8.7/2.1.1 the FID version was packed into name[2].

However, this leaves very little room in the LDLM resource name
for other uses.  The upcoming quota code needs to store another
FID into the LDLM resource to allow directory tree quotas, and
managed by the DLM.

The 32-bit VER, which is currently always 0, is moved into the high
bits of name[1] along with the 32-bit OID, to avoid consuming the
name[2] field.  Since future use of the FID version (including
snapshots, pools, etc) will need changes on the client side anyway,
there will never be non-zero VER on an existing client.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: If1e500cfb277dfc25bc056bb0c5763e48e7d500c
Reviewed-on: http://review.whamcloud.com/2275
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-164 Update testframework to restart sessions at next suite

Update the test framework so that it can append to yaml files rather
than begin a new series and allow auster to begin at a selected
test_suite.

auster has a new switch '-S suite-name' which allows the caller to
start a test group part way through.

The idea behind all of this is that a test session that fails at a
test suite can be restarted at the next suite.

Change-Id: Ifc84184506cff5ee11e5dc570ce17baab4d439a4
Signed-off-by: Chris Gearing <chris@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/538
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Robert Read <rread@whamcloud.com>
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>

LU-983 llite: align readahead to 1M after ra_max adjustment

Align the readahead pages only if ria_start != 0, otherwise the
readahead pages will be cut to zero. And it can be checked by
checking (ret > ((ria->ria_start + ret) % PTLRPC_MAX_BRW_PAGES))

Add test_101e for checking small reads for small size files.

Put 101a to Exception list, because it will do random small
reads(64k) on the file, then checking pages being discarded.
But in current implementation, we will always extend small size
read to 1M during read-ahead, which might cause too much unused
pages being read, then being discarded later.

Remove big rpc check in 101c, because the test will do random small
size(64K) read, so it only need make sure there are no small size
RPC during the read. Besides, the original big RPC check is wrong,
since it can not make sure all of RPC exists in this case.

Signed-off-by: Wang Di <di.wang@whamcloud.com>
Change-Id: If86975274f36050b9252304c447f057e99f0c5ea
Reviewed-on: http://review.whamcloud.com/1983
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-432 multiop test needs $(PTHREAD_LIBS)

Added $(PTHREAD_LIBS) to multiop_LDADD for multiop test.

Change-Id: I95ce209fc033122ac89bb4ee74f8e095e5727b17
Signed-off-by: Christos Triantafyllidis <ctria@grid.auth.gr>
Reviewed-on: http://review.whamcloud.com/978
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Johann Lombardi <johann@whamcloud.com>

LU-1116 kernel: Support RHEL6.2 patchless client.

Update RHEL6.2 kernel to 2.6.32-220.4.2(client).

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I654d8d56b14196acf8d27b31a5af8c34812d8b99
Reviewed-on: http://review.whamcloud.com/2171
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-987 build: Fail to create ldisk rpms

The autoMakefile.am in ldsikfs does not define the
BUILD_SERVER flag so make rpms fails. This patch
simply set the flag to true since ldiskfs will most
likely be used only on servers
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: Ic42eb543658f97bbff9c2f360d8693da840a8a36
Reviewed-on: http://review.whamcloud.com/2113
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-746 obdfilter-survey FAIL: test_1b ost4: hndls expected > 8, have 2

obdfilter-survey.sh checks jbd proc stats after survey, but it
doesn't take obd cleanup time into account, so the stats
collected may not reflect the data of survey test. Save run time
of survey test, and collect stats of (run_time/4) ago to fix this.

Change-Id: Ib0b55693d13da172ca476ca79013a8a2e9dd81ca
Signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1534
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Reviewed-by: Liang Zhen <liang@whamcloud.com>

LU-322 tests: Pool test fixes for large number of OSTS

This patch fixes issues with large numbers of OSTs with the
ost-pools test. we need to use hexidecimal numbers for
the pool args since the OSTs UUID are named with hex numbers.
Currently the test does pass but erros can be seen in the logs
when more than 9 OSTs exist.

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: I5db7540ee710c528af33d768a105a55bd09229b3
Reviewed-on: http://review.whamcloud.com/1841
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-952 quota: follow locking order of quota code

The locking order of quota code is: i_mutex > dqonoff_sem >
journal_lock > dqptr_sem > dquot->dq_lock > dqio_mutex, so we
should call the ll_vfs_dq_init() after journal started to avoid
deadlock.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I7f805abc3105e338dd60bc632d860be5389d87ec
Reviewed-on: http://review.whamcloud.com/1919
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>

LU-358 Add the branch and commiti-id to the yml.

yaml.sh updated to provide the information requested in the JIRA the
change is simple and requires that LUSTRE_BRANCH and LUSTRE_REVISION
are defined, if they are not defined then the elements are left blank

Signed-off-by: Chris Gearing <chris@whamcloud.com>
Change-Id: Id18f3d5c68d3703af3b76628d7d43289c57bf485
Reviewed-on: http://review.whamcloud.com/1400
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-534 test: nfsread_orphan_file test

Reproducer of bug17764.

In order to hit the assertion the lustre file has to be open twice
from lustre clients, and once from nfs mount.  Then unlink and close
opened lustre file and read the file opened on nfs.

Bugzilla: 17764
Author: Elena Gryanova <elena.gryaznova@oracle.com>
Signed-off-by: Elena Gryanova <elena.gryaznova@oracle.com>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: I05d8f04a2bd7c64864e1b82a07ef26a170887714
Reviewed-on: http://review.whamcloud.com/1895
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-534 mds: correct assertion

Orphan inode can be reached on mds_open when opening by fid which
takes place on accessing files via nfs correct the assertion
correspondingly.

Bugzilla: 17764
Author: Vladimir Saveliev <vladimir.saveliev@oracle.com>
Signed-off-by: Vladimir Saveliev <vladimir.saveliev@oracle.com>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: Id6cde949003c722b56fe7ccec43b2d8b8a0667bb
Reviewed-on: http://review.whamcloud.com/1894
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-955 build: fix bad lustre-backend-fs dependency

Fix an incorrect RPM package dependency if Lustre RPMs are built
with "make rpms" with client only
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: I1e00a2d2710ec96036cbad0a3539f704251f880a
Reviewed-on: http://review.whamcloud.com/1931
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-805 quota: lfs quota doesn't print grace time correctly

Lustre always trigger grace time when the allocated qunit exceeding
softlimit, however, user tools 'lfs quota' only print grace time
when the total usage greater than softlimit, so sometimes user can't
tell if the softlimit is already exceeded from 'lfs quota' output.

This patch changes the 'lfs quota' to use the data get from kernel
instead of comparing usage with softlimit.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I320cd0fafcb92e8b91f688b2c8bb50bc44442a09
Reviewed-on: http://review.whamcloud.com/1667
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-649 io: DIO doesn't need lock i_mutex

There is a dead lock if a file is written with with normal IO, and
read with DIO at the mean time:
regular write: extent lock -> i_mutex
DIO read: i_mutex -> server take extent lock

i_mutex taken at DIO read is redundant, because server will take
extent lock to serialize with write and truncate. Removing it can
avoid the deadlock.

Change-Id: I280559cebae5d04e4fd943bf3fbe82d67e07834b
Signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1829
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <niu@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-882 quota: Quota code compares unsigned < 0

Port from b23858.

In check_cur_qunit(), it checks "if (limit + record < 0)", however,
the limit is unsigned, so this check will be always false, and when
limit is smaller than -record, following "limit += record" will make
limit a unreasonable large value.

This patch also fixed a similar defect in dqacq_handler().

Signed-off-by: Vladimir Saveliev <vladimir.saveliev@oracle.com>
Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: Ibe1d643ea3b310b2e55c05a8c200ba5e0137ee27
Reviewed-on: http://review.whamcloud.com/1750
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-620 llite: add delete/remove_from_page_cache check

Later 2.6.32 kernel use memory cgroup feature but does not export
truncate_complete_page but export delete_from_page_cache or
remove_from_page_cache, we need properly use them for pachless client
code.

Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: I4b95da2b4cac7f2c2f63b69896dd91c3544a5473
Reviewed-on: http://review.whamcloud.com/1649
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-792 lbuild-rhel5 should use redhat's SRPM repo

Download EL5 kernel .src.rpm packages from redhat's repo instead
of trying to use CentOS's often-outdated repo.

Change-Id: I45c59894c649f3290a8ce1f0d3f1f033395567a4
Signed-off-by: Michael MacDonald <mjmac@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1595
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-424 tests: conf-sanity test 55, 56, 58 failure fixes

The MGS service was not started in conf-sanity test 55, 56 and 58
with separate MGS and MDT configuration. This patch fix the issue.

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Id82121f645d1903934bcf00328627947e943c45c
Reviewed-on: http://review.whamcloud.com/955
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-780 test: improve parallel-scale to support hyperion run

We need to add support for srun/slurm, and a few tests
from hyperion-sanity script that has been used for hyperion
testing
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: I30d5c9c84fed6f6533ec766e6278a9c86046e4a9
Reviewed-on: http://review.whamcloud.com/1662
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Cliff White <cliffw@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-604 open non-exist object should return ENOENT

when call ll_intent_file_open() against non-exist object, we cannot obtain
RPC request refcount after the call, so subsequent operation should not use
such internal RPC request to avoid to access freed memory.

Change-Id: Ic8cd2a7798d2d454cbe5931eb93b92b1a44cca4b
Signed-off-by: Hongchao Zhang<hongchao.zhang@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1271
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-715 lov: fix procfs reporting for qos values

When writing to
/proc/fs/lustre/lov/<fsname>-mdtlov/{qos_prio_free,qos_threshold_rr},
the values read back are often one less than the values written.
This happens because internally the value is stored as a number from
0-255 but accessed by the user with 0-100. Integer truncation in the
storage and retrieval stages causes this to often show lower. Adding
255 to an internal step causes the bit-shift to "round up".

Signed-off-by: Matt Ezell <ezell@nics.utk.edu>
Change-Id: I3bb1bb4b048d40eb3882c7cb22799320f90496c7
Reviewed-on: http://review.whamcloud.com/1522
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-841 tests: sanity.sh 27q does not create a testing directory

sanity.sh 27q does not create a testing directory which causes
ENOENT errors from "ONLY=27q bash sanity.sh"
Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: Ied574286fe8ee9c1859da05cc7737fb325c5ef28
Reviewed-on: http://review.whamcloud.com/1807
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Wei Liu <sarah@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-880 mds: Fix mds_open counter.

Remove duplicate mds_open counter.

Signed-off-by: Wang Di <di.wang@whamcloud.com>
Change-Id: I229929e9c1635704a668364f72cf17428ed70ef1
Reviewed-on: http://review.whamcloud.com/1747
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-824 corrupted ldiskfs after md rebuild (bz24264)

Pick up a patch from upstream to fix the md bug may
cause a corruption issue after rebuild.

Change-Id: I764a7c5ac4595aa73d821fc8e66f55598d1ff099
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1672
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-553 build: improve checks for commit-msg

Improve the checks done by the commit-msg script.  It now ensures
that all the parts of the commit message are present.
- validate that the Change-Id: generated from 'git hash-object' is
  not empty, since this can happen if git is unhappy with the options
- check for only one Change-Id: line (multiple Signed-off-by: OK)
- describe the "component:" field better, with examples

If there was an error committing the message, save a copy to a
temporary file, so that it can be edited and re-used, instead of
having to recreate it each time, or fetch it from .git/COMMIT_MSG.

Add a simple regression test with good & bad commit messages, so
it is easier to verify that any changes made to the script will
continue to both detect errors, and pass valid commit messages.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I9f954cd142a4b5cd9d596c7221cd3d6f73d8cf17
Merge-from: I15cb3690560400a591598997424cf79dee3a039d
Reviewed-on: http://review.whamcloud.com/1689
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-737 utils: check device name for digit

We need to check the whole string for digit
not only the first character

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: Ib3002f2f8dd9c657d7bcfcfa9166488c080152c2
Reviewed-on: http://review.whamcloud.com/1660
Reviewed-by: wangdi <di.wang@whamcloud.com>
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-645 Avoid unnecessary dentry rehashing

In patchless case the sequence __d_drop() then d_rehash_cond() creates
race window if dentry is already hashed, in this case rehashing should
be avoided.

Change-Id: Ib891e7d2e5ae7efe19ef2ac0abb111e9f9681b7b
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1435
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-695 clean up for OBD_FAIL_CHECK_ONCE

Clean up for OBD_FAIL_CHECK_ONCE. Just use OBD_FAIL_CHECK.

Signed-off-by: Chris Horn <hornc@cray.com>
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Change-Id: I86bb314a052d74da6f3be1d4b85aa287f706ed36
Reviewed-on: http://review.whamcloud.com/1424
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-691 fix OST index errors in test suite - sanity 133c defect

Several test run do_facet ost which assumes ost is ost0 which does not exist
according to the way the test suite works. This patch address several areas
where the wrong ost index is used. This patch also introduces some short hand
functions to get OST properties from the index. Those functions ensure ost1,
ost10 and ost100 are seen as different which is not always the case.

The following patch was also ported:
b=24504 sanity test_133* and check_stats() fix
i=Vladimir.Saveliev

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Ia7016e3c152acd8aa75cda6f838275bee316f639
Reviewed-on: http://review.whamcloud.com/1450
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

Change version to 1.8.7.80 for 1.8.8-wc1

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: I64aa0df0c2f40f884b3ccd5db38d8bbbf022bf86
Reviewed-on: http://review.whamcloud.com/1648

LU-692 ptlrpc: don't LASSERT on larger ptlrpc_body

Don't LASSERT() on a larger ptlrpc_body that might be sent from
the server.  The lustre_msg_size() function is called against
a lustre_msg buffer sent from the server in case of an early
reply, and it shouldn't LASSERT() on data received from the network.

Reserve OBD_CONNECT_* flags and obd_connect_data fields from master.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I1f68c20d1048c379ae33e328995673d3cfbe8145
Reviewed-on: http://review.whamcloud.com/1584
Tested-by: Hudson
Reviewed-by: Niu Yawei <niu@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-762 ldiskfs: don't drop directory nlink to 0

When landing the nlink patch for ext4, for an unknown reason the
logic in ext4_dec_count() was changed from the ext3 version of
the patch.  It now drops the nlink = 0 temporarily and then
if it is a directory with nlink == 0 it increases nlink again.

Instead, only drop nlink if it is larger than 2.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: Id4dd16d3764efdf6a7ea29da92de51398b300370
Reviewed-on: http://review.whamcloud.com/1579
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson

LU-553 build: Lustre style commit hooks for Git

Adds recommended commit hooks for Lustre commit messages, that
can/should be installed into lustre/.git/hooks for all Lustre trees.

prepare-commit-msg: adds the output of checkpatch.pl as a
comment to the commit message, to "soft" validate the new
patch is matching the Lustre coding style. It doesn't check
100% of the Lustre coding style, but at least finds most of
the obvious errors so that they can be fixed locally before
any manual inspection time is spent on the patch.

commit-msg: verifies the format of the commit message (somewhat).
Checks that there is a valid-looking Signed-off-by: line.
Adds the Gerrit Change-Id: line for change tracking, or
validates that the existing Change-Id: line is well formed.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I1fb7db8b29d01b6820a5468f59e4df55bb5fe00e
Reviewed-on: http://review.whamcloud.com/1555
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-745 kernel: ost-pools test_23 hung

It might be caused by a jbd2 bug which result in forever sleep
in the do_get_write_access().
http://www.spinics.net/lists/linux-ext4/msg24689.html

Apply the fix from Jan Kara to see if it can resolve the problem.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I5dc74ab9fc49686fce9fdcff3c94ef8f0c35b38f
Reviewed-on: http://review.whamcloud.com/1541
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-553 add checkpatch.pl for Lustre coding style

Add a modified version of the Linux kernel checkpatch.pl script (from 2.6.38)
to Lustre for validating patches against the Lustre coding style. The Lustre
coding style is largely the same as the kernel, except using spaces instead of
tabs for indenting.

Change-Id: Ia278370e0867c03e5f0814ef16109f64bd2bb9b3
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1167
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Johann Lombardi <johann@whamcloud.com>

LU-754 Skip permission check for NFS open after create

For open(create) operation, NFS client will trigger separate open RPC request after create one,
then we only need to check permission for create operation in such case.

Change-Id: I958f190398305c3bcc1b1740e31c64855dc631e8
Signed-off-by: Hongchao Zhang <hongchao.zhang@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1509
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-610 fix the wrong data type in find_value_cmp

in function find_value_cmp(lustre/utils/liblustreapi.c), the data type for
parameters "file" and "limit" is converted from "unsigned long long" type
into "unsigned int" type by mistake, and the file size should be defined
as "unsigned long long" type too.

Change-Id: I83326b50860bdfeb4cd31509f99f7ab1b6a8ab13
Signed-off-by: Hongchao Zhang <hongchao.zhang@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1401
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

Bump version to 1.8.7-wc1 for RC1 and update changelog

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: Ib644ed108560715d2837601a749c2a49de3fb2c8

LU-443 LNet: Only squawk when md->start is NULL on non-zero length

Only squawk when md->start is NULL on non-zero length.
The md->start == NULL check prevents anyone from creating a ME/MD
with no buffer. These are used as backstop buffers to generate events
when traffic has exceeded the local buffer space.
See Oracle bug 21103 attachment 32922.

Change-Id: Id4fdf26b047df56cdf034952fea15f9a4e0dba62
Signed-off-by: Wally Wang <wang@cray.com>

LU-606 Add alternative servers start order

When newly created mdt registers itself, mgs erases all the logs for
that particular fs, which fails client mount.

Add alternative server start orders workaround to enable
MGS->MDT(s)>OST(s) as well as MGS>OST(s)->MDT(s) order.

Bugzilla: 24050,24464
Author: E.Gryaznova <elena.gryaznova@oracle.com>
Author: Wang Yibin <wang.yibin@oracle.com>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I3a342636efdb7efefe8377add0e4dc808f7733c1
Reviewed-on: http://review.whamcloud.com/1272
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-721 Parallel writes to same file results in a file of zeroes

Canceling lock may contain data being sent to OSTs. Change
find_cbdata iterator to take that into account.

Change-Id: Ifdb21472780331cba452db68a52fd15bb6cbe35c
Author: Vladimir Saveliev <vladimir.saveliev@oracle.com>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1422
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-690 test: wait_osc_import_state() fixes

b=24498 wait_osc_import_state() fixes

i=Vladimir.Saveliev

-- increase maxtime to wait the timeout of 1st request;
   take into account at_min value;
-- cleanup wait_osc_import_state() to use _wait_import_state();
-- ost-pools test_1 fix: use local var instead of global NAME

Signed-off-by: Elena Gryaznova <elena.gryaznova@oracle.com>
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I580152df654c9ba5048ae54ff933611b832ef3d9
Reviewed-on: http://review.whamcloud.com/1498
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

bump version to 1.8.6.81-wc1

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: I3cdfe619963741cbe24ab6e929d562bca51569f1

LU-624  Kernel update [RHEL6.1 2.6.32-131.12.1.el6] (client only)

Change-Id: Ia7473c717c0862fd4b2167ed61400f5493217ce4
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1438
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-146 mds_open() may deadlock

* Get open lock inside mds_get_parent_child_locked() to avoid
  deadlock.
* Never get open lock if child is newly created to avoid deadlock.

Signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Change-Id: I1a1dadafef364f36ad3b892ba7a2d669f6027f10
Reviewed-on: http://review.whamcloud.com/1259
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-674 kernel update RHEL5.7 [2.6.18-274.3.1.el5].

Test-Paramaters-Start:
serverdistro=el5
clientdistro=el5
testgroup=regression
Test-Paramaters-End:

Change-Id: I4d080a22395ce11c43ba52d6bb3089761180a08a
Signed-off-by: Yang Sheng <ys@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1213
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-478 Set tunables on lvm device

- The set_blockdev_tunables() shouldn't chop the tail digit on
  the /dev/mapper/xxx device.
- Sometimes, the /sys/block/dm-x doesn't have the max_[hw]_sectors_kb,
  such errors should be tolerated.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I3c3737be561302adc93ab699701649e71c0c1321
Reviewed-on: http://review.whamcloud.com/1439
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

Revert "LU-607 Avoid race between ldlm_pools_shrink and ldlm_namespace_free"

This reverts commit ba79e90a7028e2637e64367535715c81729f4cb2.
This patch is causing conf-sanity subtest 23a to fail.

Change-Id: I35f27ba55bb12a3d231f34e4764aa48a71025b4b
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

LU-448 add lst stat --count

This adds lst stat --count, making it easier to script up data
collection for performance tests.

The patch is from Oracle bug 22638 attachment 33001

Change-Id: Ic60e6226ce85ebbc2d314e41c7e85ba5879ea64e
Signed-off-by: Wally Wang <wang@cray.com>
Reviewed-on: http://review.whamcloud.com/1312
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Liang Zhen <liang@whamcloud.com>
Reviewed-by: Lai Siyao <laisiyao@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-608 canonicalize the devices names

Perform a readlink on the device name if path is /dev/disk/by-id...
See Oracle bug 24487.

Change-Id: I52bc670181606663bfbd746d34f3c2bc710ecc72
Signed-off-by: Wally Wang <wang@cray.com>
Reviewed-on: http://review.whamcloud.com/1284
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-261 Renamed auster.sh to be auster

This makes it like b2_1 and following the convention that human commands don't
have extensions and modified Makefile.am which references auster[.sh]
Also remove libcfs from REQUIRED_DIRS in autogen.sh since libcfs is still
under lnet in lustre 1.8.

Change-Id: I29247005a6b02b039cf4a861af4dc5de30082c23
Signed-off-by: Chris Gearing <chris@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/541
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-646 port bz23485 (clarification of lustre fsync behavior)

Add directory fsync operation.

Signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Change-Id: I6bad2cc1df840bbf827b9fcb352361877dd493d1
Reviewed-on: http://review.whamcloud.com/1308
Tested-by: Hudson
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-607 Avoid race between ldlm_pools_shrink and ldlm_namespace_free

ldlm_namespace_free removes namespace from list and free memory
without checking namespace's refcount while ldlm_pools_shrink might
get namespace from the list and start ldlm_pool_shrink() for it.

Bugzila: 24419
Author: Vladimir Saveliev <vladimir.saveliev@oracle.com>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: I757204ad119927cacdb73db66f5e40ace50eeea8
Reviewed-on: http://review.whamcloud.com/1273
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-594 Make sure sanity 27v fallout clears

We need wait long enough for OST to wake up from
OBD_FAIL_TGT_DELAY_PRECREATE sleep.

Change-Id: Ie8b56e3f8bad79049946563353360dce58280ba4
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1263
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-544 speed up filesystem formatting time in conf-sanity.sh

Pass "-E lazy_itable_init" to mke2fs to speed up Lustre
filesystem formatting time in conf-sanity.sh.

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I2df0e47332b967deae668405786006c7ad7250bd
Reviewed-on: http://review.whamcloud.com/1223
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-541 Fix getstripe interop with 2.0 server

* 2.0 server has changed dir getstripe rule as follows:
          If the directory in question DOES NOT have it's EA set, AND it
          IS NOT the root directory of the filesystem, then no EA values
          are returned to userspace. Rather, an error is returned from
          the ioctl and errno is set to ENODATA.
  1.8 client needs adapt to this rule
* fix liblustreapi.c usage of errno and rc, llapi_xxx() functions should
  return negative value on error.

Change-Id: If81a456209d5985aae7f20569eba1ac45ef576b1
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1154
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-565 Update llite_nfc.c to 2.x interop fid API

It seems we forgot to update lustre NFS code when converting everything
to FID interop API. This lead to ESTALE errors trying to export NFS
from 1.8 client that ewas mounting 2.x MDT due to inode generation
and fid generation mismatch (inode generation always 0 for interop).

Signed-off-by: Oleg Drokin <green@whamcloud.com>
Change-Id: Ic88671826da73b19857a42ceb9ac8dd0fd35b244

LU-544 reduce the journal size for conf-sanity test 56

1) reduce the journal size from 16MB to 8MB for reformatting
   the MDT device which has a small size (40MB)
2) speed up filesystem initialization by passing
   "-E lazy_itable_init" to mke2fs

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I43eeaa1edf43420adeb101ae96b1d9faa36fcbbb
Reviewed-on: http://review.whamcloud.com/1171
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-550 incorrect quota hash function

Quota code uses 'uid' or '(1 << 32) + gid' as the hash key, however,
the quota hash function lqs_hash() always try to interpret such a
unsigned long long key as a 'quota_adjust_qunit' struct, then calculate
the hash value by quota_adjust_qunit->qaq_id.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I128f2835ff590d712b0a1641af7361146e8e3e81
Reviewed-on: http://review.whamcloud.com/1163
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-471 fix mkfs.lustre to avoid creating duplicate options

Also make user specified options overwrite default internal default
options.

Change-Id: I27d55e7490566e07c310e52ba7319fefcd0528d9
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1103
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-530 group quotas not enforced

In filter_quota_getflag(), we shouldn't break the loop when the file owner's
quota limit isn't set, otherwise, the check for the file group will be skipped
and the out of quota flag will not be returned to client at the end.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: Ibf6f1fe3a550ce1f4b3d627fe24ca4484e754a41
Reviewed-on: http://review.whamcloud.com/1153
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-278 build: Only warn for tag/version mismatch

The configure process should NOT abort just because the most
recent tag is not of the form that upstream uses to tag Lustre.
Downstream developers may use their own tags, or just add
extensions to upsteam's version tags.

Change-Id: I43b63e8438cc37588d5a5b91677102ded22917fb
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1105
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-413 limit used inodes for performance tests

If doing wide stripe mdsrate tests, don't try to consume all of
the inodes on the OSTs.  The MDS does not consume inodes on OSTs
completely uniformly, and this only introduces spurious failures
when the inodes run out unevenly.

Change-Id: I36243e0d3e990858d268d043012af40419936fe4
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/942
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Niu Yawei <yawei.niu@gmail.com>
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-525 llapi_file_get_stripe() fails with ENAMETOOLONG

Remove bogus check on fsname in llapi_file_get_stripe().

Change-Id: I6ea7c403ce3a1994e80bc039260fae1e881b892b
Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/1131
Tested-by: Hudson
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-387 mmp.sh test_9 failed

The mmp.sh test_9 often fail for the script killing the e2fsck process
on wrong timing, so we use some tricky way to make sure the MMP block
stay in 'e2fsck' state before mounting the device.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I35a2f9708fa7dd3ea0cad246b767578209feeee3
Reviewed-on: http://review.whamcloud.com/991
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-432 QUOTA_OK / NO_QUOTA definitions

Starting from kernel 2.6.34 the definitions of QUOTA_OK and NO_QUOTA
have been removed from linux kernel (linux/quota.h). This adds the
needed definitions to the relevant files.

Change-Id: I9b67ea491a34777c2ea5ff86630788cec18ebcf6
Signed-off-by: Christos Triantafyllidis <ctria@grid.auth.gr>
Reviewed-on: http://review.whamcloud.com/973
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-456 Force commit to reuse the just-deleted blocks

In ext4 non-writeback journal mode, the just-deleted blocks are not
useable until the transaction committed, this could cause filter
return -ENOSPC mistakenly when the just-deleted blocks are not committed.

Whenever block allocation fails for -ENOSPC on filter, we should wait
for the previous deleted blocks committed, then retry the allocation.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I50878d4459377d68a9c67575f137a3a44cb3625e
Reviewed-on: http://review.whamcloud.com/1039
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-452 : Don't export static functions

As with LU-102 in the master branch, recent versions of GCC on ia64
sensibly refuse to export static functions.  obd_device_alloc and
obd_device_free are not referenced externally, so resolve this error
by not exporting them.

Signed-off-by: Stephen Champion <schamp@sgi.com>
Change-Id: I784cfbe59f4685d0e975427c3095aa082de1a1c6
Reviewed-on: http://review.whamcloud.com/1007
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson

LU-419 increase default inode ratio for OST/MDT

Increase the default inode ratio for large OSTs to avoid creating
far too many inodes, which can slow down e2fsck dramatically.

Decrease the default inode ratio for the MDT, since it never uses
the space that is allocated to it.  Also make this calculation take
the expected stripe EA into account, which wasn't done before.

Change-Id: Ic827d5f7a012fa07106e7e053abc0e646aba30ac
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/959
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

LU-419 enable ext4 features by default

Enable the flex_bg, huge_file, and dir_nlink features from ext4 by
default for ext4-based ldiskfs filesystems.  The flex_bg feature
can significantly reduce e2fsck time as well as time-to-first-write
on newly mounted OSTs.

Enable the 64bit feature for ext4 filesystems if the device size is
larger than 2^32 blocks.  This is required in order to format OSTs
larger than 16TB in size.

Reduce the number of inodes created on larger OST filesystems, since
there are far too many inodes created by default, which wastes space
and slows down e2fsck.

Skip zeroing the journal at format time.  It will be overwritten
at first use, and avoids writing 400MB of zeroes to the filesystem
needlessly.  We can't skip zering the inode table by default, but
newer mke2fs will check if the kernel handles this itself, and
request it internally.

Cleanups to remove support for 2.4 kernels and reiserfs filesystem.

Based on LU-255 commit eb012d4a10208b26c2d3e795a90f1bb07dde6d91.

Change-Id: I614003cc4d471f5fc33c379f66407e995fe856e2
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/958
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>

Bump version to 1.8.6.60-wc1

Change-Id: I8622e6a1235a7bcfb8298a632201342965016435
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

LU-130 Add changelog entry

Page migration is now disabled, add changelog entry.

Change-Id: Ie417cac4a6b0a9c2a82cc217daa81e3d40eeff9a
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

Revert "LU-457 Enable slab debug for RHEL5/x86_64"

This reverts commit 529529a3dfbb8849f479c15f75e8f875ba5427b4.

Change-Id: Ic1955242ca260d9669a7e2ad4ba4d78f512739ce
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

LU-130 disable page migration

Disable page migration until we implement a proper ->migratepage handler.

Change-Id: I04eb78c0d04c828eabb3598a3e42ebf51daec3c2
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

LU-457 Enable slab debug for RHEL5/x86_64

Debug patch to turn on slab debugging in the RHEL5 kernel.

Change-Id: I6bcd6f6239b9f5b5953981bcb5fafec426b90005
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

LU-410 Revert LU-15 slow IO with read intense application

This reverts commit 12aae18c0ec495f459a1026ad2e86bbf0232b0fe.
This patch indirectly causes a performance drop with SRP.

Change-Id: If1af51ac8c2d45bc4d5cc5502dc59501678baaec
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

Update e2fsprogs version in changelog

Change-Id: Ib0c6cd31dfec23bf92548028d50655b2520b99d4
Signed-off-by: Johann Lombardi <johann@whamcloud.com>

LU-85 Fix missing bracket

Change-Id: I0c3d16320aab019ef0ef7994c34f6ea5f9ded2a2
Signed-off-by: Johann Lombardi <johann@whamcloud.com>