LU-1905 build: Always create libcfs.a

Lustre and LNET utilites have become dependent on libcfs.a but
one can easily disable libcfs.a from being built by disabling
liblustre creation. This patch have Lustre always build libcfs.a.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: Id80f9baa16da238eca708409e0b5d5248e437c5a
Reviewed-on: http://review.whamcloud.com/3953
Tested-by: Hudson
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Reviewed-by: Artem Blagodarenko <artem_blagodarenko@xyratex.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1856 build: fix 'out-of-bounds write' errors

Fix 'out-of-bounds write' defects found by Coverity version 6.0.3:
Out-of-bounds write (OVERRUN_DYNAMIC or OVERRUN_STATIC)
Overrunning array at too high offset.

Signed-off-by: Sebastien Buisson <sebastien.buisson@bull.net>
Change-Id: Idb2ed5f814bb0c4c9418cedffb8e01efae0f4bd7
Reviewed-on: http://review.whamcloud.com/3903
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Keith Mannthey <keith@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1854 ptlrpc: track culled request seq

Ptlrpc service should track highest culled request-seq by storing
it in ptlrpc_servce_part::scp_hist_seq_culled, however, there is
a missed case in ptlrpc_server_drop_request(): if rq_reply_state of
the request is pre-allocated and it should be freed immediately,
ptlrpc_server_drop_request() will unlink the request from history
list w/o track sequence number, and it can be the reason of
hitting LASSERT while reading req_history.

Signed-off-by: Liang Zhen <liang@whamcloud.com>
Change-Id: I74ed9c7f00e77707628d628792bf9583c4bc4259
Reviewed-on: http://review.whamcloud.com/3916
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: wangdi <di.wang@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1301 lu: local objects library

set of functions working on top of OSD API to create/access
local objects by name.
the library maintains own top device to be able to work in
multi-service environment (mds + mgs).

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I26cc47b866bb0925be4f4419ac663a1d42520e02
Reviewed-on: http://review.whamcloud.com/3665
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>

LU-1919 mount: do not access vfs mnt directly

from obd_mount.c - it's under OSD control now and at umount time
the structure can disappear with OSD, sooner then lustre umount
is completed.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I91bae528bc229cd4e0e191e61eae8a7854494c8f
Reviewed-on: http://review.whamcloud.com/3982
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1303 osp: code to update per-OST last_id in OSP

Store and Update last_id - per OST counter of created objects.

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: Iedd84afceb692f1b545c6dfe8e035de184952b18
Reviewed-on: http://review.whamcloud.com/3961
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-1303 osp: Basic OSP device implementation

Initial OSP code: osp_device, osp_object and structures

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: I21ea3d0cdfe07634278d8e7fee9a7c3fef79f81e
Reviewed-on: http://review.whamcloud.com/3958
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-1881 mdd: cleanup partial modification

Usually, the modification triggered by client will be split into
several sub-operations in MDD layer and processed one by one. If
some step failed, we should rollback the former sub-operation(s)
to avoid partial modification the target(s) in RAM or on disk.

Signed-off-by: Fan Yong <yong.fan@whamcloud.com>
Change-Id: I02559618d252f65cd3ae3b26bd186b77c9f42dcc
Reviewed-on: http://review.whamcloud.com/3981
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-by: Niu Yawei <niu@whamcloud.com>

LU-1877 mdt: initialize lock and expiration

for statfs cache.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ief24f6a855c3ed14a7afc142921b9b249ab11604
Reviewed-on: http://review.whamcloud.com/3946
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1302 llog: llog test update and fixes

- port llog_test fixed from Orion
- remove unneeded include files
- remove llog init/fini from lmv, it is not used
- add new test_7 for llog_test
- add unlink64 handling

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: I4793bb02748f2aaab33d67d0f04c8767e7f5ffb9
Reviewed-on: http://review.whamcloud.com/3825
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1903 test: disable lfsck for ZFS tests

lfsck test does not apply to ZFS backend filesystem

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: If332b5335b1e444c94669a5e8b5899d896dfb18a
Reviewed-on: http://review.whamcloud.com/3963
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1842 ldlm: support for sending GL ASTs to multiple locks

Traditionally, a glimpse callback is sent to one single lock which
is the highest PW extent lock.
The new quota code uses glimpse callbacks in a different manner since
it can issue multiple glimpse ASTs to different slaves (as done for
blocking callbacks for instance) to notify those latters of the new
qunit value.

This patch modifies the glimpse callback path to use
ldlm_run_ast_work() to issue multiple glimpse ASTs in one shot (as
done currently for other lock callbacks).

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: I972630fcf3158dd1d361392913f8086f1ebaa39f
Reviewed-on: http://review.whamcloud.com/3887
Reviewed-by: Niu Yawei <niu@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1881 oi: not shrink the last entry in OI index node

Shrinking OI index node to recycle idle leaf for the last entry
will cause subsequent lookup/insert ops to access invaild space.
So just keep the last entry there, which can be reused directly
by next new node.

Other fixes:

1) The recycled empty OI blocks should be recorded on divice and
be re-loaded after the device remounted. Then they can be reused
when needs new OI blocks.

2) Need not check iam_container::ic_idle_failed in iam_new_node().

3) Clear iam_frame::at_shifted when iam_path_release().

Signed-off-by: Fan Yong <yong.fan@whamcloud.com>
Change-Id: I46611c208563a943a0980110b2c416186e6d1249
Reviewed-on: http://review.whamcloud.com/3931
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Tested-by: Oleg Drokin <green@whamcloud.com>
Reviewed-by: wangdi <di.wang@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1923 lov: verify stripe is on given device

When restarting FIEMAP from a file with many extents, the first
supplied fiemap extent contains the offset of the previous last
extent found.  If this data is invalid, it may cause the LOV code
to acccess out-of-bounds array indices.

Verify data passed from userspace is within bounds.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I8e70e891b9f23c8f72aa78a4807369584ac2b04f
Reviewed-on: http://review.whamcloud.com/3962
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Reviewed-by: wangdi <di.wang@whamcloud.com>

LU-1900 test: check for lustre version in sanity-scrub

Since OI scrub only available on lustre 2.3.0, we need
to check if the mds runs at least 2.3.0

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: I5ce0462a1deb5bddfec88177a75a1c17b69b5217
Reviewed-on: http://review.whamcloud.com/3947
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1799 o2iblnd: debug patch for o2iblnd

IBM reported kernel panic on their BGQ IO node when loading the
ptlrpc module with an o2ib network. The IB interface had an IPv4
and IPv6 address. Removing the IPv6 address avoided the crash.

I suspect rdma_bind_addr can't associate any RDMA device in this
case, this patch will check if there's attached IB device on cmid
even returned value is ZERO, it will also output more information.

Test-Parameters: nettypes=o2ib
Signed-off-by: Liang Zhen <liang@whamcloud.com>
Change-Id: Id44110fcf56b199b1504ab4e6b0157d87bc2d270
Reviewed-on: http://review.whamcloud.com/3815
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Isaac Huang <he.huang@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1808 build: Lustre build does not support FC15.

Change lbuild to support for FC15 and calm down warning message.

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I365f708f82c01b0f72aeffa3c09e46800ca770d5
Reviewed-on: http://review.whamcloud.com/3849
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Chris Gearing <chris.gearing@intel.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1489 test: lfs data_version is not available prior to 2.3

We need to check lustre version to make sure the test for
new feature does not run on unsupported version

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: Iae808413ba4965033d17d53977f647efc533eecc
Reviewed-on: http://review.whamcloud.com/3948
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1872 ptlrpc: cleanup jobstats in thread context

Since cfs_hash function isn't safe in interrupt context, we should
move the jobstats cleanup work into each stats logging function.

If there are any activities, cleanup will be triggerred on logging
time, and the hash table will be shrinked automatically; If there
isn't any activity, the hash table will stop growing, and it can
be cleared manually.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I1c60e123f67c0ad3b959bbb330282cd07c63a28c
Reviewed-on: http://review.whamcloud.com/3945
Tested-by: Hudson
Reviewed-by: Liang Zhen <liang@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1301 mgs: remove compatibility code

which is inteded to support 1.6-originated profiles

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I64a74e1f139fb3d94e7d272febc31f0306d01bab
Reviewed-on: http://review.whamcloud.com/3671
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1301 mgs: improve error handling

taken as is from orion

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ib26c5f519f686a77fa1057af214a4823d0c16359
Reviewed-on: http://review.whamcloud.com/3670
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1301 mgs: introduce mgs_thread_info

use it to save stack

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ibefb8a7adf5ac9c6b12212d27d890467c889e5d9
Reviewed-on: http://review.whamcloud.com/3669
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1644 recovery: to not check client version for mne swap

data->ocd_version in obd_connect_data can be overwritten by server
so that it'll be wrong if client is reconnected after server recovery.
Wrong clienv version caused wrong exp_need_mne_swab was set and then
has interop issue between 2.2 clients and 2.3 server.

In this patch, the following fixes are made:
1. do not check the client version on the MGS;
2. reset client version on each import connect;
3. client side fix will be removed after lustre 3.2

Signed-off-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Change-Id: Ia2899d1b8385fd4799ca6fe260fc4c84444247cb
Reviewed-on: http://review.whamcloud.com/3897
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Tested-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1869 build: gcc/ar utility can vary

- ar utility can vary depeding on build platform.
- do not call gcc directly, use $(CC)

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ia60752b40531886d805a095c9f571e565bcd0e88
Reviewed-on: http://review.whamcloud.com/3917
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Liu Xuezhao <xuezhao.liu@emc.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1711 mount: use OSD API to export statfs data

so that server mount points (like /mnt/ostXXX) show
valid data on any backend (ldiskfs, zfs)

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I6ad0546d7630769083e8aea69547a6bb3256784f
Reviewed-on: http://review.whamcloud.com/3652
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>

LU-1711 mount: lu_site is provided by OSD

previously it was created by MDT, which is started after MGS.
so we relocate lu_site in OSD and now it's created before
any service like MDS/MGS.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I43312218c293df509064cc6a854074e652a6e691
Reviewed-on: http://review.whamcloud.com/3651
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>

LU-1711 mount: obd_mount to start osd

... instead of mounting underlying ldiskfs directly.
so, now OSD can be any depending on osd= mount option.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Iab1ca6ca9f84ba6c708e05cd85c4de0a51d34de6
Reviewed-on: http://review.whamcloud.com/3650
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>

LU-1301 mgs: env and mgs_device to pass around

a trivial part of orion changes:
 - introduce mgs_device (old mgs_obd)
 - pass lu_env around

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I53d55c9f0a91c16c9eb65ab41d9660c3699bac45
Reviewed-on: http://review.whamcloud.com/3668
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>

LU-1823 mdd: Don't provide buffer for xattr when we just need size

mdd_declare_attr_set incorrectly passed uninitialized buffer to mdo_xattr_get
where as it is just interested in knowing if xattr exists at all.
The buffer happens to point to other people's data sometimes which leads
to random memory corruptions.

Change-Id: I28b477d385427292ad38455d495ab546f757c7d5
Signed-off-by: Oleg Drokin <green@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/3928
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Mike Pershin <tappro@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-521 lnet: Fix endian issue introduced by change 1338

In change 1338 done for LU-521, an endian issue was introduced
when the magic number for the RPC header was swapped before
the body has been swapped. We use the magic number to
determine if the body needs swapping.

Signed-off-by: Doug Oucharek <doug.s.oucharek@intel.com>
Change-Id: I0a64ca13dbdc04169d59a1bd531b4877e8736993
Reviewed-on: http://review.whamcloud.com/3831
Reviewed-by: Isaac Huang <he.huang@intel.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Liang Zhen <liang@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>
Tested-by: Oleg Drokin <green@whamcloud.com>

LU-1824 obdfilter: reset lnb[n].rc in filter_grant_check()

After patch http://review.whamcloud.com/3446 was committed,
filter_grant_check() would be called a second time if there
was no grant space for the write operation in the first time
it was called.

The issue was that all of the lnb[n].rc values were set to
-ENOSPC inside filter_grant_check() the first time it was
called, but at the second time, although there was some space
and the function returned 0 not -ENOSPC, the lnb[n].rc values
were not reset, all of them were still -ENOSPC, which caused
the assertion failure inside filter_commitrw_write().

This patch fixes the above issue.

Test-Parameters: envdefinitions=SLOW=yes testlist=obdfilter-survey
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Ia0ec57a7d329136b4e1f2f82a337c6cf43f7f66c
Reviewed-on: http://review.whamcloud.com/3913
Tested-by: Hudson
Reviewed-by: hongchao.zhang <hongchao.zhang@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1813 osc: osc_lock_unuse() race

The lock may have been released by upper layer so the cl_unuse_try()
in osc_lock_upcall() will release the dlm lock instead of crashing.

Signed-off-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Change-Id: Ie1c3f8d2ba8b227e42d08b1631651c5834cbac25
Reviewed-on: http://review.whamcloud.com/3895
Tested-by: Hudson
Reviewed-by: Lai Siyao <laisiyao@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: wangdi <di.wang@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1821 build: fix build error about VLA at file scope

Fix build error:
drivers/incore/fs_incore.c:340:54: error: variably modified
‘incore_dir_template’ at file scope [-Werror]

C99 does not allow VLA at file scope. Allocate incore_dir_template
dynamically to avoid build error with newer gcc(>4.6.3) that checks
this.

Signed-off-by: Peng Tao <tao.peng@emc.com>
Change-Id: Icc81c649f58772e76e1cd3dcd42348a53a9a942b
Reviewed-on: http://review.whamcloud.com/3856
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Yang Sheng <ys@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1842 quota: define quota records for glb/slv indexes

This patch defines lquota_glb_rec & lquota_slv_rec which are the
record which will be used with the global & slave index files.
It also introduces the global quota lu_context_key which will be used
by all quota-level handlers.
The lquota module init/exit function are now defined in lquota_lib.c
which calls the old quota init/exit function for the time being.
Finally, this patch also modifies lquotactl_slv() to collect
enforcement information.

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: I2b261bd8433a5e6bd64c2de2710bca32bf606122
Reviewed-on: http://review.whamcloud.com/3886
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Niu Yawei <niu@whamcloud.com>

LU-1302 llog: introduce llog_open

- llog_open is pair to llog_close. Llog handle is allocated inside
  llog_open while llog_free_handle is part of llog_close. Each llog
  even non-existent should be opened and closed.
- llog_exist() is added to API to check if llog file exists
- llog_cat_put is renamed to llog_cat_close with cleaning up empty
  llogs like llog_obd_origin_cleanup did.

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: I6360b4c089ec1dde50289563447eefd3dfa4365a
Reviewed-on: http://review.whamcloud.com/3740
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1818 quota: en/disable quota enforcement via conf_param

In the new quota architecture, quota is enabled/disabled
per-filesystem, by 'lctl conf_param':

lctl conf_param $FSNAME.quota.$POOLNAME=$VAL

$FSNAME  : filesystem name;
$POOLNAME: 'mdt' (for meta pool) or 'ost' (for data pool);
$VAL     : 'none' - disable both,
           'u'    - enable user quota,
           'g'    - enable group quota,
           'ug'   - enable both;

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I89ef9fc1a067981cf7caac29c8311b75054c91fe
Reviewed-on: http://review.whamcloud.com/3850
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1830 quota: fix ldiskfs quota accouning iterator

There were defects in the ldiskfs quota accounting iterator:
- leaf blocks could be re-processed;
- entry index was updated incorrectly, so some entries could be
  skipped;

Fix the defect in osd_diskfs_read(), the 32bit 'size' could be
overflowed.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: Icfd96fdaf38c57bc63a5fcfd1f96c02d0ddeb74d
Reviewed-on: http://review.whamcloud.com/3877
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1842 quota: quota identifier unification

This patch defines a generic lquota_id structure which is an union
of all the possible identifier types that can be used with quota,
this includes:
- 64-bit user ID
- 64-bit group ID
- a FID which can be used for per-directory quota in the future

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: Idead35f27ee9d07b9b25f39b791a1ad8778e0f2d
Reviewed-on: http://review.whamcloud.com/3888
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1581 mount: get rid of ldd in the kernel

all ldd parsing is now done outside of the kernel
by mount.lustre utility.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Iea6257b75f7573b6594c6bc7a2fa3a86b1c91700
Reviewed-on: http://review.whamcloud.com/3614
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1789 protocol: reserve connect flag for lightweight conn

Reserve connection flag for lightweight connection support to avoid
conflicts. The main property of a lightweight connection is that
there is no entry in the last_rcvd file for this export, so no
recovery is possible. This type of connection will be used by quota
(between master & slave), FIDonOST and likely by other features in the
future.

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: I4367f5bfaef0207e7f307c4ea3d35e97ca99dc5b
Reviewed-on: http://review.whamcloud.com/3794
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <niu@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-812 utils: remove obsolete jt_dbg_modules_2_4 code

jt_dbg_modules_2_4 is for 2.4 kernel and it breaks "lctl modules"
for 3.x kernel.

Signed-off-by: Liu Xuezhao <xuezhao.liu@emc.com>
Change-Id: If2953c71707ae200a4095786150b1dee05547fff
Reviewed-on: http://review.whamcloud.com/3778
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-812 llite: O_LOV_DELAY_CREATE conflicts with FMODE_NONOTIFY

Kernel 2.6.36 introduces FMODE_NONOTIFY and defines it as 0x1000000
(commit 12ed2e36c98aec6c41559222e311f4aa15d254b6), this causes the
confliction with O_LOV_DELAY_CREATE which is now defined as octal
0100000000.
This confliction causes kernel remove this flag in build_open_flags,
then "lfs setstripe" a new file will fail with -EEXIST(stripe
already set).

This patch changes O_LOV_DELAY_CREATE to 0120000000(0x1400000)
to make it works for new kernel and compatible with old version
statically linked binary.

Another change is starting to build dynamic linked version of
liblustreapi.so.

Signed-off-by: Liu Xuezhao <xuezhao.liu@emc.com>
Change-Id: I643d3734cfdc2f45099fc6be663a370c2dcc8ea3
Reviewed-on: http://review.whamcloud.com/3779
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Iurii Golovach <Iurii_Golovach@xyratex.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>

LU-1203 mdt: map old params to new ones by using an array

This patch improves mdt_process_config() to use an array for
mapping old params to new ones.

The patch also adds a common function class_find_old_param()
to check whether a proc param is an old one or not, and return
new one if it's an old one.

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I122e4e83a3f4e84a3fdbd3fe6a69b7c1b5e08d3f
Reviewed-on: http://review.whamcloud.com/3836
Tested-by: Hudson
Reviewed-by: Niu Yawei <niu@whamcloud.com>
Reviewed-by: Liang Zhen <liang@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1816 scrub: OI scrub skips new created objects for once

For accelerating OI scrub, it will skip the objects which are new
created after the latest MDT mount and marked as I_LUSTRE_NOSCRUB.
But if the object with I_LUSTRE_NOSCRUB is used/cached in RAM for
very long time, then OI scrub has no chance to process it even if
there has been inconsistency happened until the MDT remount.

So as compromise, the OI scrub will skip the new created objects
for once only when the first accessing.

Signed-off-by: Fan Yong <yong.fan@whamcloud.com>
Change-Id: I560d7d851e3430326d24ec74154bb2a99f80b737
Reviewed-on: http://review.whamcloud.com/3848
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Mike Pershin <tappro@whamcloud.com>

LU-1804 tests: skip OI scrub speed test under UP env

The OI scrub speed depends on CPU/RAM schedule. There may be more
schedule delay under UP environment. So skip related test cases.

Signed-off-by: Fan Yong <yong.fan@whamcloud.com>
Change-Id: Iebc514857c6d88601598a7da58a340296655dcf1
Reviewed-on: http://review.whamcloud.com/3846
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1755 osc: solve a race between fsync and truncate

If an OSC extent is being truncated when fsync is called, it will
have oe_fsync_wait set but no oe_urgent or oe_hp set. This causes
problem because when the extent changes OES_CACHE later, it won't
be written out immediately because urgent bit is not set.

This problem can be fixed by checking oe_fsync_wait bit and set urgent
bit correspondingly when changing osc extent's state from OES_TRUNC to
OES_CACHE at the end of truncate.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I755baac066375a92730b14de1c470c66baad5320
Reviewed-on: http://review.whamcloud.com/3699
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-1678 mdd: don't create LOV EA by going thru .lustre path

When a file is created with O_LOV_DELAY_CREAT, we should not create
its LOV EA data by going through .lustre path.

Add a test to sanity.sh open by fid test suite (test_154).

Signed-off-by: jc lafoucriere <jacques-charles.lafoucriere@cea.fr>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: Ib223da44d16bf3a541269d5cfb60826e43345493
Reviewed-on: http://review.whamcloud.com/3483
Tested-by: Hudson
Reviewed-by: Lai Siyao <laisiyao@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1779 tests: fix run_one_logged() to log SKIP status

In the current test framework, only those tests which are in the
$ALWAYS_EXCEPT list are logged with SKIP status, other skipped
tests are all logged with PASS status.

This patch fixes the above issue by setting the SKIP status in
pass() and logging the status in run_one_logged().

Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Ia4e8f51040034abdec9c67aaf0a128b63baaa688
Reviewed-on: http://review.whamcloud.com/3810
Tested-by: Hudson
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1735 ptlrpc: cleanup jobid code

Some cleanups for the jobid code:
- if obd_jobid_var is too large, only print an error message once
- in linux cfs_get_environ():
-- cache strlen(key) since it doesn't change and is used often
-- remove unnecessary typecasts of void pointers
-- use "rc" instead of "ret"
-- balance ENTRY and RETURN/GOTO calls
- add cfs_get_environ() for liblustre and remove inline #ifdef
- use strcmp() to compare strings that are known NUL-terminated
- use strlcpy() to ensure NUL-terminated strings in target buffer
-- add strlcpy() wrapper for liblustre, it isn't in Glibc on RHEL5

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I22c3d5c1755c1d6aab666a769df38218b954cab0
Reviewed-on: http://review.whamcloud.com/3713
Tested-by: Hudson
Reviewed-by: Niu Yawei <niu@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1304 mdt: use slab for mdt_objects

should save a bit of memory and cpu

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I13601aba3c2cdb5e7c0f4c3a45ff5fb6e428b38f
Reviewed-on: http://review.whamcloud.com/3781
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>

LU-1304 mdt: cache statfs in mdt_device

... and save few cycles

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ib3727c7419cfdfa393a2c1c72770ef960b1d05e6
Reviewed-on: http://review.whamcloud.com/3780
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1425 build: make Lustre build ready for gcov

gcov requires profiled code to be built with compiler options
-fprofile-arcs -ftest-coverage. To profile all project this
options should be passed during every source file compilation.
This option enable debug logging. No all Lustre parts are ready
for debug logging now.

To patch adds debug librarias where needed and resolve
linking conflicts. Do not influence when no
-fprofile-arcs -ftest-coverage compile options are
passed.

Xyratex-bug-id: MRP-526
Reviewed-by: Alexey Lyashkov <alexey_lyashkov@xyratex.com>
Reviewed-by: Alexander Boyko <alexander_boyko@xyratex.com>
Signed-off-by: Artem Blagodarenko <artem_blagodarenko@xyratex.com>
Change-Id: If5c1742f3263faf48c2c5d8457c1af36a29408d4
Reviewed-on: http://review.whamcloud.com/2844
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Brian J. Murrell <brian@whamcloud.com>

LU-1581 mount: kernel do not parse ldd

all ldd parsing is done by mount.lustre utility now.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ice15cf4fa8e411144713d0276d006f8cfe9c56f1
Reviewed-on: http://review.whamcloud.com/3613
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1337 build: add missing include dir

arch/include/generated is added since kernel version v3.2.
while at it, also change "uname -m" in .m4 to $target_cpu,
because users may cross building.

Signed-off-by: Peng Tao <tao.peng@emc.com>
Change-Id: Ib9aa526a9318e5c93b05cceba9fc8cdfbf9fef0a
Reviewed-on: http://review.whamcloud.com/3745
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Yang Sheng <ys@whamcloud.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>

LU-1302 llog: pass lu_env as parametr in llog functions

Add lu_env as paramenter in llog API functions and helpers

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: I3c3c1878b7dec7059602a91790820b70f225ade5
Reviewed-on: http://review.whamcloud.com/3696
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1302 llog: pass lu_env to the llog callback

- Pass lu_env to he llog callback.
- Move cat_cancel_cb() to llog_cat.c to be with all
  other catalog functions.

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: Iedbc7953b461ba52b52680d7d709d27c6581f934
Reviewed-on: http://review.whamcloud.com/3649
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1302 llog: add env to llog_process functions

Add lu_env to llog processing function.
Remove using flags and mark fork is needed by parameter.

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: Ie38b3da4090df35281e51ef196d91e3f96f95bdc
Reviewed-on: http://review.whamcloud.com/3644
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1788 osc: don't print error msg for EINPROGRESS resend

Now that -EINPROGRESS can be legitimately returned during normal
operation (e.g. quota rebalancing in progress), we shouldn't print an
error message on the client each time the BRW is resent because of
-EINPROGRESS.

This patch also caps the resend delay for BRW to the current request
timeout.

Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Change-Id: Ie7447602756b0721351c7c90cbfb40ad8e3bb720
Reviewed-on: http://review.whamcloud.com/3792
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Niu Yawei <niu@whamcloud.com>

LU-1798 test: check connect_flags for jobstats

Check connect_flags before enabling jobstats in the t-f.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: Ib36575fe7d08b257a2a7dda86851806bf44d7a41
Reviewed-on: http://review.whamcloud.com/3806
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1772 lov: lock with lov lsm refcount taken may deadlock

lov_io_init() and lov_lock_init() has taken lsm refcount, no need to
lock lo_type_guard otherwise it may lead to deadlock.

Signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Change-Id: I3e953d41e181c3ef19f7bf8d92d9791753d4c58a
Reviewed-on: http://review.whamcloud.com/3807
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1795 lustre: remove LASSERT in dt_fiemap_get()

dt_fiemap_get should return EOPNOTSUPP instead of
issuing a LASSERT when it's not supported on the
underlying filesystem

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: I2cd5044472af3deeb77dbc54ced93f7e553586aa
Reviewed-on: http://review.whamcloud.com/3817
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1676 quota: missing vfs_dq_init() in write path

Calling vfs_dq_init() in the osd_write_commit(), otherwise space
won't be accounted for the existing inode.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I9671ee9865449dd298139d7668ab17d3d58f8866
Reviewed-on: http://review.whamcloud.com/3476
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <tappro@whamcloud.com>

LU-1302 llog: structures changes, llog_thread_info

Add changes in structures, introduce llog_thread_info,
protect llog header data with new lock.

Signed-off-by: Mikhail Pershin <tappro@whamcloud.com>
Change-Id: Id1438e69d8771880481e1a335d38d3e2d1635085
Reviewed-on: http://review.whamcloud.com/3631
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1592 ldlm: protect obd_export:exp_imp_reverse's change

* Protect obd_export::exp_imp_reverse from reconnect and destroy race.
* Add an assertion in class_import_put() to catch race in the first
  place.

Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: If0abf6717456931c567d8d41c1d20fe49452e959
Reviewed-on: http://review.whamcloud.com/3684
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: wangdi <di.wang@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1693 obdfilter: Set bi_rw before calling bio_add_page()

In filter_do_bio() and osd_do_io() set bi_rw before calling
bio_add_page() so that the correct bio direction is used in
mgrge_bvec_fn().

Signed-off-by: John L. Hammond <jhammond@tacc.utexas.edu>
Change-Id: Ife34d5d89337ea7fea41b5844fd724211b05c670
Reviewed-on: http://review.whamcloud.com/3501
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1518 mdd: Fixup mdd_{obf,dot_lustre}_obj_ops.

Define several missing md_object ops for .lustre/fid.  Unify
attribute handling for .lustre with that of normal md_objects.

Fixed several defects:
- use parent ops instead of child ops in mdo_create();
- skip version get for obf in mdt_obj_version_get();
- disable non-normal fid lookup in obf_lookup();

Signed-off-by: John L. Hammond <jhammond@tacc.utexas.edu>
Signed-off-by: Bob Glossman <bogl@whamcloud.com>
Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I89c5b37c822680b0ad1e23372e3e2be464f6a3d8
Reviewed-on: http://review.whamcloud.com/3726
Tested-by: Hudson
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1512 oi: reuse idle OI blocks

With the system running, some OI blocks may become empty because
related files have been removed. Originally, these empty blocks
almost cannot be reused beause it always allocates new block for
new leaf or index node. Such space waste is an important reason
for the OI file size increasing quickly.

Now, when the non-released OI leaf node becomes empty, it will be
recycled into a per-OI based idle blocks pool, which is organized
as a series of tables. For each table, the on-disk format is:

+---------+---------+---------+---------+------+---------+-------+
|  magic  |  count  |  next   |  logic  |      |  logic  | free  |
|(16 bits)|(16 bits)|  table  |  blk #  | .... |  blk #  | space |
|         |         |(32 bits)|(32 bits)|      |(32 bits)|       |
+---------+---------+---------+---------+------+---------+-------+

The logic blk# for the first table is stored as "idle_blocks" in
the OI root node. In fact, there is padding field in the OI root
node after the "dx_countlimit". We will append the "idle_blocks"
just after the "dx_countlimit" to reuse part of the padding. So
it will not cause compatibility issues with old OI files.

Adjust the strategy for OI file block allocation: reuse empty block
in its idle blocks pool with priority; if such pool is empty, then
allocate new block from system volume.

Signed-off-by: Fan Yong <yong.fan@whamcloud.com>
Change-Id: I1bccc1cd9a4e1a16e4f51ac18e6b995ec85a0cf8
Reviewed-on: http://review.whamcloud.com/3153
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: wangdi <di.wang@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Liang Zhen <liang@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1786 build: Fix build for crc32 pclmulqdq

Old versions of gcc don`t know pextrd instruction. User mode
libcfs build had wrong dependency with ARCH variable defined.
PEXTRD marco was added for old gcc, wrong dependency was fixed.

Signed-off-by: Alexander Boyko <alexander_boyko@xyratex.com>
Change-Id: I8821a1c25367f0f284a80192ed65d89808d1b2da
Reviewed-on: http://review.whamcloud.com/3801
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1579 osd: reserve credits for set version

In current osd_declare_xattr_set(), we assume that version change is
always accompanied with inode changes, so we don't reserve extra
credits for the setting verison.

However, in the new quota design, we may issue a sole version change
without any other updates. I think osd_declare_xattr_set() should not
make assumption on the caller's behavior and reserve at least 1 block
for version change.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I77a13173eea8007b5300155e0d726bd82ace8ede
Reviewed-on: http://review.whamcloud.com/3242
Tested-by: Hudson
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Mike Pershin <tappro@whamcloud.com>

LU-1769 ofd: Add a fault injection for EINPROGRESS

Add OBD_FAIL_OST_DQACQ_NET case for fault injection
for EINPROGRESS

Signed-off-by: Minh Diep <mdiep@whamcloud.com>
Change-Id: I902a527af96af19b417e232d0e9d52c4b625e8eb
Reviewed-on: http://review.whamcloud.com/3763
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Li Wei <liwei@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

LU-1661 tests: improve setup_posix_users()

This patch improves setup_posix_users() in posix.cfg to use
do_rpc_nodes to add groups and users on remote nodes,
so as to simplify add_group() and add_user() and make them
become common functions to be used by other test scripts.

Test-Parameters: clientdistro=el5 testlist=posix
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: Ie07f32dffe7d7ae2037d92bab372977efff2e5fe
Reviewed-on: http://review.whamcloud.com/3770
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1716 ptlrpc: Race in updating of connection flags on client.

Update obd_connect_data before setting import to FULL state

Xyratex-bug-id: MRP-577
Reviewed-by: Alexey Lyashkov <alexey_lyashkov@xyratex.com>
Reviewed-by: Alexander Boyko <alexander_boyko@xyratex.com>
Signed-off-by: Andriy Skulysh <Andriy_Skulysh@xyratex.com>
Change-Id: I4818c65bcc2fb8ee847921924fa1ca2469f79b9f
Reviewed-on: http://review.whamcloud.com/3555
Tested-by: Hudson
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1762 tests: get correct MMP update and check intervals

This patch fixes the get_mmp_update_interval() and
get_mmp_check_interval() in mmp.sh to get the correct
MMP update and check intervals from both the old and
new outputs of debugfs.

The patch also improves test_8() to increase the running
time of e2fsck to allow mount operation to be started
before e2fsck operation stops.

Test-Parameters: testlist=mmp
Signed-off-by: Yu Jian <yujian@whamcloud.com>
Change-Id: I47a8759c97bf09a73a6bc4ed7a03ff0929c7127f
Reviewed-on: http://review.whamcloud.com/3743
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>

LU-1773 lov: lov_delete_raid0() need heed lov_fini_raid0()

Add a sanity test case and handle failure of lov_init_raid0()
correctly.

LU-1480 is also due to the failure of lov_init_raid0().

Signed-off-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: If900cb854846f778c74d0368d64cb7a9d5189406
Reviewed-on: http://review.whamcloud.com/3732
Tested-by: Hudson
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1014 mountconf: MGS update timeout upon config update

MGS should update obd/ldlm_timeout upon MDS/OST config update. It
will be the largest value of all servers.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Signed-off-by: Lai Siyao <laisiyao@whamcloud.com>
Change-Id: Ia0786475fa5608e06f8c2606f87f98192a870df5
Reviewed-on: http://review.whamcloud.com/2881
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-657 obdfilter: fix bug in previous patch

in the merged patch http://review.whamcloud.com/#change,3446,
the usage of fsfilt_commit_wait is wrong, and it doesn't stop
the journal firstly.

Signed-off-by: Hongchao Zhang <hongchao.zhang@whamcloud.com>
Change-Id: I3a36edf7049466880c27c14bb7f99966aa75d4f1
Reviewed-on: http://review.whamcloud.com/3692
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>

LU-1400 test: obdfilter-survey tests fail but still green

obdfilter-survey tests are failing but appear green in Maloo.
ENOSPACE errors are not being properly reported.

Add the "set -e" flag to make obdfilter-survey fail on errors.

Test-Parameters: testlist=obdfilter-survey
Signed-off-by: Keith Mannthey <keith@whamcloud.com>
Change-Id: I83666d085136178737727080a1a3b55768fb13a7
Reviewed-on: http://review.whamcloud.com/3591
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Li Wei <liwei@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1407 msg: Wrong error messages on MDS and Client

Don't show error messages on console in some case.

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I0b8b1bb523f677813b2f3998509afba639398d9d
Reviewed-on: http://review.whamcloud.com/3076
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Lai Siyao <laisiyao@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1410 test: Allow sanity test 200 to work with one OST

The sanity test is failing test 200 with one OST. When
the test adds machines to the pool it is not passing
valid arguments.

Changing first_ost=0 is allowing the test to run.

Signed-off-by:Keith Mannthey <keith@whamcloud.com>
Change-Id: I45aef14e7814f170024c71f904b888423d8f8c90
Reviewed-on: http://review.whamcloud.com/3730
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Minh Diep <mdiep@whamcloud.com>
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1721 ptlrpc: AT race in drop request

It's introduced in by commit 07b8db220e48782369f48d86213c5d404a628ded
which makes ptlrpc_server_drop_request() not to hold at_lock for
checking req::rq_at_linked.
This change might race with ptlrpc_at_check_timed() if:

1) thread-1: call ptlrpc_at_check_timed() and remove the request from
   paa_reqs_array, before it set req::rq_at_linked to zero...
2) thread-2: call ptlrpc_server_drop_request() to release the last
   refcount, and it found req::rq_at_linked is non-zero, so it
   entered the condition "if (req->rq_at_linked) {...}"
3) thread-1: set req::rq_at_linked to zero
4) thread-2: take at_lock, and hit the assetion
   LASSERT(!cfs_list_empty(&req->rq_timed_list)) because thread-1 has
   already removed req::rq_at_linked from paa_reqs_array in step 1)

This patch fixed this issue and did some code cleanup.

Signed-off-by: Liang Zhen <liang@whamcloud.com>
Change-Id: If4469637a5f9d63c9253a9f4c4cac0bcd7f8b46e
Reviewed-on: http://review.whamcloud.com/3564
Tested-by: Hudson
Reviewed-by: wangdi <di.wang@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1763 test: enable jobstats by default

Enable jobstats in the auto-test by default.

Signed-off-by: Niu Yawei <niu@whamcloud.com>
Change-Id: I03bd22b144f4372922c6dc67511d09a91d17b8e4
Reviewed-on: http://review.whamcloud.com/3718
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1735 ptlrpc: only set jobid if not already set

The ptlrpc_set_add_req->lustre_get_jobid->cfs_access_process_vm()
callpath locks mm->mmap_sem to fetch environment variables, but if
this could deadlock in case of MMAP IO, which also holds mmap_sem.
If the mmap_sem is already held for write, just don't set the jobid.
This will make jobid stats inconsistent for mmap IO, but avoids a
lot of complexity or race conditions in the code otherwise.

If the caller already has fetched the jobid and saved it for this
inode in the OSC layer, so we don't need to fetch and reset the
pd_jobid field at all in this case.  This avoids doing extra work
to fetch the jobid if it is not needed, and avoids storing it
temporarily on the stack when it won't be used.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I22c3d5c1755c1d6aab666a769df38218b954650a
Reviewed-on: http://review.whamcloud.com/3604
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Niu Yawei <niu@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1303 build: bump warning on LOD/OSP workarounds

There are a number of generated compile warnings in the code
that need to be fixed due to LOD/OSP workarounds that need to
be removed.  I've bumped the version check to version 2.3.55,
to give the LOD/OSP patches a chance to land.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I41356848d7f1d0699bb0f01cb9302012ff386bc4
Reviewed-on: http://review.whamcloud.com/3741
Tested-by: Hudson
Reviewed-by: Oleg Drokin <green@whamcloud.com>

New tag 2.3.50

After branching off b2_3, master becomes 2.4.0 development branch.

Change-Id: If7b8c875cb23d3f673ec6185dd5de9f13352c0f0
Signed-off-by: Oleg Drokin <green@whamcloud.com>

LU-1754 kernel: Kernel update [RHEL 6.3 2.6.32-279.5.1.el6]

Update RHEL6.3 kernel to 2.6.32-279.5.1.el6.

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I82ddff59f801fce2335583516470297a4664bedc
Reviewed-on: http://review.whamcloud.com/3683
Tested-by: Hudson
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1644 mgs: swab nidtbl entries for 2.2 clients

LU-1252 is missed in 2.2 release so it caused a problem that 2.2
clients always swab nidtbl entries even if the server and client
are using the same endian.  Conversely, 2.3 clients would not
swab the nidtbl entries but the 2.2 server would always swab.

To make it work, 2.3 adds a temporary OBD_CONNECT_MNE_SWAB flag,
which will cause the server to swab the nidtbl entries for 2.2
clients, and 2.3 clients will undo the swab done by 2.2 servers
if the MNE_SWAB flag is not set.  This avoids problems with only
checking the client/server version, in case the LU-1252 fix is
ever applied to a 2.2 client or server build.

This workaround is set to auto-expire for 2.5.50+ clients and
servers, which is a reasonable upper limit for interoperability
with old unpatched 2.2 clients in a mixed-endian environment.
This is enough for 2.2.0 clients/servers to work with the 2.5.x
feature releases, after which the OBD_CONNECT flag can be reused.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: Id316f7e1c7ee2b2c1d1077e8c5dd916edca04d84
Reviewed-on: http://review.whamcloud.com/3548
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1581 utils: mkfs warn about missing --index

Using the --index option to mkfs.lustre is standard practice, and is
documented as required in the Lustre manual.  Add a warning for mkfs
usage in the 2.3 release, so that users with formatting scripts/tools
can be notified of this change in advance of the hard requirement in
Lustre 2.4.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I1f481a334163af1033da5296f9e71fb3b4cee07e
Reviewed-on: http://review.whamcloud.com/3703
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Li Wei <liwei@whamcloud.com>
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1540 osd: add NUL terminator for long symlink

Add NUL terminator for long symlink to ldiskfs inode on-disk data.

Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: Id7ce7829ec9b4c8eb72cf257df046a5288a5eb7b
Reviewed-on: http://review.whamcloud.com/3560
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-957 scrub: disable OI Scrub for osd-zfs

ZFS is not supported by OI Scrub.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I7a2f90298f37edab870edf28008a1a140260c99b
Reviewed-on: http://review.whamcloud.com/3594
Reviewed-by: Fan Yong <yong.fan@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1608 utils: require --fsname for mkfs.lustre

When formatting a filesystem, require that fsname is specified to
mkfs.lustre so that users have to think about this a bit.  In some
cases, two filesystems at the same site have used the same default
filesystem name, and this later caused problems when trying to mount
both filesystems on the same node.

Signed-off-by: Andreas Dilger <adilger@whamcloud.com>
Change-Id: I15a5ebb8d7d3afc6037587961e94dc66426bcf5d
Reviewed-on: http://review.whamcloud.com/3297
Tested-by: Hudson
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Li Wei <liwei@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1668 tests: fix for conf-sanity 53_a & 53_b

conf-sanity 53_a & 53_b are using local cpt number as increment or
decrement of threads number which is wrong, they should use cpt
number from target server.

Test-Parameters: nettype=o2ib
Signed-off-by: Liang Zhen <liang@whamcloud.com>
Change-Id: Ief25ecfefc291decd6e72067b1363942a5871f81
Reviewed-on: http://review.whamcloud.com/3595
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: wangdi <di.wang@whamcloud.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>

LU-1461 build: asm/cpufeature.h is only available on x86

Only include it if ARCH_X86 is defined.

Change-Id: I6af260d6d753853c015885f33892b789948f1cff
Signed-off-by: Oleg Drokin <green@whamcloud.com>
Reviewed-on: http://review.whamcloud.com/3510
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Alex Zhuravlev <bzzz@whamcloud.com>
Reviewed-by: Alexander Boyko <alexander_boyko@xyratex.com>

LU-1625 test: reduce test duration for nfs mode

There isn't much value to run long duration in
nfs mode. Cut down IOR test as well.
Based on original work by Minh Diep.

Test-Parameters: testgroup=parallel-scale-nfsv4
Signed-off-by: Keith Mannthey <keith@whamcloud.com>
Change-Id: Id0c2b23d41a7e0a2763a45747a0f98a24fd77fa7
Reviewed-on: http://review.whamcloud.com/3701
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1745 test: Fix recovery-small test 105

Update check_target_ir_state() to handle the old IR reported
state of ON/OFF as well as the new reported state of DISABLED/
ENABLED.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Change-Id: Ib2e10594cd2fb5d2e54d22187db32af9229e27ff
Reviewed-on: http://review.whamcloud.com/3667
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1548 tests: add test 63 in conf-sanity

Add test_63 in conf-sanity to verify each page can
at least hold 3 ldiskfs inode.

Signed-off-by: wang di <di.wang@whamcloud.com>
Change-Id: I281df5945a1e26736bb2b27c211a8239d25e0c7e
Reviewed-on: http://review.whamcloud.com/3605
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

LU-1581 mount: use a special separator in svname

with this new approach mkfs marks fresh filesystem
by a special separator (:) between fsname and service
name. then mount utility recognizes this separator
and passes 'writeconf' option to the kernel mount.
upon this first mount (and having a new configuration
profile on MGS) mount utility replaces ':' with regular
'-' in svname and sets label to that.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: I83e1d1e4f1fd67ae9a6453069b324696715d4fee
Reviewed-on: http://review.whamcloud.com/3612
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>

new tag 2.2.93

Change-Id: I6bb117a22febd7a59273ec14310e2e8b82bca081
Signed-off-by: Oleg Drokin <green@whamcloud.com>

LU-1581 utils: label methods

label methods for ldiskfs and zfs, used by mount utility
to update the label upon the first successful mount.

Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
Change-Id: Ibd5e14bd10e39d71264d15bbee42021b089091ea
Reviewed-on: http://review.whamcloud.com/3611
Tested-by: Hudson
Reviewed-by: Jinshan Xiong <jinshan.xiong@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Oleg Drokin <green@whamcloud.com>

LU-1727 lprocfs: fix an old assertion

lprocfs_counter_add() can be called in interrupt context, as a memory
allocation can trigger memory shrinker call ldlm_pool_shrink(), which
calls lprocfs_counter_add().

Signed-off-by: Bobi Jam <bobijam@whamcloud.com>
Change-Id: I0f5d3e07c234f7d615ca87761a7bd6367e72beac
Reviewed-on: http://review.whamcloud.com/3597
Tested-by: Hudson
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Liang Zhen <liang@whamcloud.com>

LU-1024 tests: Test replay-vbr.test_1c failed

Fixed execution of runmultiop_bg_pause from replay-vbr.sh.

The test-framework.sh wasn't able to to be included in
runmultiop_bg_from script, because since it was started via sh
and dirname didn't return correct path for LUSTRE initialization.

Running the script as executable allows to detect it's directory
correctly via dirname.

Xyratex-bug-id: MRP-299
Reviewed-by: Alexander Zarochentsev <alexander_zarochentsev@xyratex.com>
Reviewed-by: Andriy Skulysh <Andriy_Skulysh@xyratex.com>
Signed-off-by: Kyrylo Shatskyy <kyrylo_shatskyy@xyratex.com>
Change-Id: I0c1e88af405a87bf73ff3e2ac438bccba2b92afc
Reviewed-on: http://review.whamcloud.com/2003
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>
Reviewed-by: Li Wei <liwei@whamcloud.com>
Reviewed-by: Yu Jian <yujian@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>