Whamcloud - gitweb
Amir Shehata [Mon, 10 Nov 2014 19:56:52 +0000 (11:56 -0800)]
LU-5734 lnet: improve clean up code and API
This patch addresses a set of related issues: LU-5734, LU-5839,
LU-5849, LU-5850.
Create the local lnet_startup_lndni() API. This function starts
up one LND. lnet_startup_lndnis() calls this function in a loop
on every ni in the list passed in. lnet_startup_lndni() is
responsible for cleaning up after itself in case of failure.
It calls lnet_free_ni() if the ni fails to start. It calls
lnet_shutdown_lndni() if it successfully called the
lnd startup function, but fails later on.
lnet_startup_lndnis() also cleans up after itself.
If lnet_startup_lndni() fails then lnet_shutdown_lndnis() is
called to clean up all nis that might have been
started, and then free the rest of the nis on the list
which have not been started yet.
To facilitate the above changes lnet_dyn_del_ni() now
manages the ping info. It calls lnet_shutdown_lndni(),
to shutdown the NI. lnet_shutdown_lndni() is no longer
an exposed API and doesn't manage the ping info, making
it callable from lnet_startup_lndni() as well.
There are two scenarios for calling lnet_startup_lndni()
1. from lnet_startup_lndnis()
If lnet_startup_lndni() fails it requires to shutdown the ni
without doing anything with the ping information as it hasn't
been created yet.
2. from lnet_dyn_add_ni()
As above it will shutdown the ni, and then lnet_dyn_add_ni() will
take care of managing the ping info
The second part of this change is to ensure that the LOLND is not
added by lnet_parse_networks(), but the caller which needs to do
it (IE: LNetNIInit)
This change ensures that lnet_dyn_add_ni() need only check if there is
only one net that's being added, if not then it frees everything,
otherwise it proceeds to startup the requested net.
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: I8bf1d93b7afa31c6f7f25733516f6a5c175c6c9d
Reviewed-on: http://review.whamcloud.com/12658
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-by: Isaac Huang <he.huang@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bruno Faccini [Fri, 19 Dec 2014 22:42:05 +0000 (23:42 +0100)]
LU-6056 tests: fix -hsm wrong usage of bash's local built-in
In sanity-hsm test script, there were several places where the
bash's local built-in was used with assignments from a command
substitution and command's exit code was tested after.
This is a wrong usage since the 'local' builtin does not propagate
the exit code.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I4257e55ec61ed0c4e48812407b5cf82ab316b701
Reviewed-on: http://review.whamcloud.com/13159
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Thu, 18 Dec 2014 17:41:39 +0000 (09:41 -0800)]
LU-6048 kernel: kernel update [RHEL7 3.10.0-123.13.2.el7]
update RHEL7 kernel to 3.10.0-123.13.2.el7
Test-Parameters: clientdistro=el7 mdsfilesystemtype=ldiskfs\
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I0dba7339527054c731fd7887cb2d01fde681cfad
Reviewed-on: http://review.whamcloud.com/13153
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 15 Dec 2014 23:18:40 +0000 (17:18 -0600)]
LU-5971 llite: rename ccc_object to vvp_object
Rename struct ccc_object to struct vvp_object and merge the CCC object
methods into the VVP object methods.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Iadb4ef84b430f6423312d27e4e95fbcd77af33b1
Reviewed-on: http://review.whamcloud.com/13077
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 15 Dec 2014 20:25:41 +0000 (14:25 -0600)]
LU-5971 llite: rename ccc_device to vvp_device
Rename struct ccc_device to struct vvp_device and merge the CCC device
methods into the VVP device methods.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ifb7c8c0fc7416c56b8bbafa969a52dc4d4db654e
Reviewed-on: http://review.whamcloud.com/13075
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Tue, 30 Dec 2014 15:01:58 +0000 (09:01 -0600)]
LU-4178 tests: increase sanity-hsm wait_request_state tiemout
In the sanity-hsm.sh utility function wait_request_state() increase
the wait_result timeout from 100 seconds to 200.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I4fd2734d7d178cbfa3b881b725e3b5aff7ab603e
Reviewed-on: http://review.whamcloud.com/13206
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 29 Dec 2014 16:42:57 +0000 (10:42 -0600)]
LU-6073 tests: use sys/xattr.h in tests/*.c
In lustre/tests/{multiop,orphan_linkea_check}.c include sys/xattr.h
rather than attr/xattr.h since the former is more likely to be
installed that the latter and provides the needed declarations.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I79b49233ab42566c8084c2070d0d75a6508656dc
Reviewed-on: http://review.whamcloud.com/13197
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Blake Caldwell <blakec@ornl.gov>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 29 Dec 2014 16:18:07 +0000 (10:18 -0600)]
LU-2675 build: remove obsolete Kernelenv files
Remove the obsolete Kernelenv.{in,mk} files from libcfs/ and lnet/.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I778b00c3c4b315d9811f594d2b99457a83b696c7
Reviewed-on: http://review.whamcloud.com/13196
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Thu, 27 Nov 2014 23:48:33 +0000 (16:48 -0700)]
LU-5946 lprocfs: cleanup stats locking code
Add comment blocks on lprocfs_stats_lock() and lprocfs_stats_unlock().
Move common NOPERCPU code out of the switch() statements to reduce
code size and complexity, since it doesn't depend on the opc at all.
Replace switch() in lprocfs_stats_unlock() with a simple if/else,
since the lock opc was already checked in lprocfs_stats_lock().
Add an enum for the lprocfs_stats_lock() operations to make it clear
what the valid values are and allow compiler checking.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I1906d44981c025964d6a510000b217f10ca23031
Reviewed-on: http://review.whamcloud.com/12872
Tested-by: Jenkins
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Wed, 5 Nov 2014 01:39:12 +0000 (20:39 -0500)]
LU-5577 osc: change cl_extent_tax and *grants to unsigned
Change the type accordant usage and remove warnings.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: Iaaf45b3634a6d3667c0481c9e76c5ee69964c48c
Reviewed-on: http://review.whamcloud.com/12386
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Liang Zhen [Thu, 11 Dec 2014 20:53:31 +0000 (15:53 -0500)]
LU-5570 lnet: check router aliveness timestamp
In current LNet, peer aliveness timestamp is only recorded for peers
on routers, this patch will change this and also record aliveness of
routers on regular nodes, so lnet can check aliveness timestamp of
routers before sending message, avoid to choose router which has old
aliveness timestamp and can be potentially dead or congested.
Another change in this patch is, when lnet got a remote message from
router, it will update remote NI status of router, in this case, lnet
can get up-to-date aliveness informaiton without waiting to the next
RC ping.
Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Change-Id: I305ba684ced231a87656458382966aeb8ff7ba48
Reviewed-on: http://review.whamcloud.com/11748
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Tue, 23 Dec 2014 13:43:59 +0000 (16:43 +0300)]
LU-5928 obdclass: fix export of nodemap to procfs
Rewrite export code of nodemaps to use nid_stats instead of
exp directly like is done for other members.
Clean up redundant code.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: Ib3ddeabcb1c132eb1b91a0cf9583cc0f74a61e5d
Reviewed-on: http://review.whamcloud.com/13175
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Amir Shehata [Thu, 18 Dec 2014 02:34:00 +0000 (18:34 -0800)]
LU-6043 lnet: Fix YAML FSM and other cleanup
1. Fixed the FSM to be able to parse lnetctl YAML output properly
2. Don't print out elements which are not configured
3. Minor static analysis cleanup (LU-6021)
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: I0e7812775a3c718882f0845c71ac7fb2fa46d1b0
Reviewed-on: http://review.whamcloud.com/13115
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Thu, 6 Nov 2014 18:00:32 +0000 (10:00 -0800)]
LU-5781 osc: osc_lock_weight endless loop fix
With huge number of pages to scan by osc_lock_weight() it is likely
CLP_GANG_RESCHED is returned from osc_page_gang_lookup() and the scan
will be repeated again from the start. To be sure that the scan is
progressing across those restarts, next scan should be started from
the last scanned page index plus one.
Xyratex-bug-id: MRP-2145
Change-Id: I60775f2d8f688029d97ad4fe64f9b2698ed278c8
Signed-off-by: Alexander Zarochentsev <alexander.zarochentsev@seagate.com>
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-on: http://review.whamcloud.com/12362
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Tue, 16 Dec 2014 15:11:09 +0000 (10:11 -0500)]
LU-4259 osd-zfs: Add brw_stats collection and display
Collect stats consistant with brw_stats where possible for osd-zfs.
The supported stats are: (prefixed with BRW_R and BRW_W)
* PAGES
* RPC_HIST
* IO_TIME (reads only)
* DISK_IOSIZE
* DISCONTINOUS PAGES (only meaningful for writes, always single
buffer for reads)
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Icd0d703d3cf31487cdfa524613cfc550db228dba
Reviewed-on: http://review.whamcloud.com/11467
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Oleg Drokin [Fri, 2 Jan 2015 17:54:25 +0000 (12:54 -0500)]
new tag 2.6.92
Change-Id: If6d3f1f589179aa930d3bb59f38b86ceaf583f9f
Bob Glossman [Wed, 24 Dec 2014 22:35:41 +0000 (14:35 -0800)]
LU-6069 kernel: kernel update [SLES11 SP3 3.0.101-0.46]
Update target and config files for new version
Revision of the ldiskfs patch series for only sles11sp3 is included.
It is no longer the same as the sp2 series.
Test-Parameters: envdefinitions=SANITY_EXCEPT=170\
mdsdistro=sles11sp3 ossdistro=sles11sp3\
clientdistro=sles11sp3 mdsfilesystemtype=ldiskfs\
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I9a20662339ca9c95a4bfabd182d43aba3c4443ea
Reviewed-on: http://review.whamcloud.com/13186
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Wed, 31 Dec 2014 17:06:32 +0000 (12:06 -0500)]
LU-6076 tests: skip sanity/124a if necessary
If prior tests haven't been run, this test will fail, so safely skip
it.
Test-Parameters: testlist=sanity envdefinitions=ONLY=124a
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I13d7a56b96eda5084627edd8c912782bee20a27c
Reviewed-on: http://review.whamcloud.com/13222
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Lai Siyao [Fri, 26 Dec 2014 14:15:35 +0000 (22:15 +0800)]
LU-5914 lfsck: dt_lookup() LBUG
This patch contains two fixes:
1. object can be remote in lfsck_namespace_rebuild_linkea(), remove
assert on this.
2. initialize dt_index_ops for parent in l_n_verify_stripe_slave()
before dt_lookup() on it.
Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Change-Id: I0755f471470cc0805db222640bba1d04f4128388
Reviewed-on: http://review.whamcloud.com/13188
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bobi Jam [Mon, 22 Dec 2014 04:36:59 +0000 (12:36 +0800)]
LU-6042 osc: osc_object_ast_clear() LBUG
An OSC object could be destroyed with AGL locks waiting for granted,
so we'd get rid of the osc_object_ast_clear() assertion that its
dlm locks all getting granted.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: I47f6c9a5f5b57291e8452713af417310f25a5e03
Reviewed-on: http://review.whamcloud.com/13163
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bruno Faccini [Wed, 5 Nov 2014 19:38:36 +0000 (20:38 +0100)]
LU-5851 build: split lustre-osd RPMs
Since Lustre Server DKMS builds osd Kernel modules and Provides
lustre-osd, we need to find a way to also install the
osd-specific dso that is also shipped as part of lustre-osd RPM.
To fix this, now ship the osd-specific dso as part of its own RPM.
And also create its new lustre-osd-mount dependency to ensure
the lustre-osd -> lustre-osd-mount -> lustre coherency and
ordering for Server installs.
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I25d498590212e696f74018c0af694bfee4d1cc81
Reviewed-on: http://review.whamcloud.com/12589
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Kit Westneat [Sat, 13 Dec 2014 17:36:29 +0000 (12:36 -0500)]
LU-6025 utils: reset optind before calls to getopt_long()
The nodemap commands in lctl were not properly resetting optind
between calls to getopt_long(). This patch sets optind to 0 before
calling any subcommands.
Signed-off-by: Kit Westneat <kit.westneat@gmail.com>
Change-Id: I14a1f34000c678799298dae3c248c839a1f32d52
Reviewed-on: http://review.whamcloud.com/13063
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Fan Yong [Thu, 9 Oct 2014 09:36:11 +0000 (17:36 +0800)]
LU-6033 tests: inject failure before counting the objects
Some test cases need to inject some failure stub on OST that will
be triggered when creating the OST objects. For that, we need to
calculate how many objects to be created to guarantee that all the
pre-created OST-objects on the MDT can be exhausted. Under some
race cases, the MDT may be in pre-creating OST-objects and finished
the pre-creation just after the precreated_ost_obj_count() returning
zero to the caller but before the caller injecting the failure stub
on the OST. Then the subsequent failure injection will be useless.
To resolve such trouble, the caller of precreated_ost_obj_count()
should inject related failure stub before calculating the pre-created
OST-objects count.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I01c98c0eb67aa0974c3223356557ae76f23ddb77
Reviewed-on: http://review.whamcloud.com/13098
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Hongchao Zhang [Mon, 20 Oct 2014 16:28:17 +0000 (00:28 +0800)]
LU-3455 mdt: sync when evicting client
If some client is evicted from the MDT, then the underlying OSD
should be synchronized, otherwise the recovery will be failed
for the replay requests were dropped alongside with the client.
in test_20b in replay-single.sh, the occupied disk space will be
released only after DMUs are committed for ZFS, then it should
wait some more time before calling 'df' to get the disk usage info.
Test-Parameters: alwaysuploadlogs \
envdefinitions=SLOW=yes,ENABLE_QUOTA=yes,ONLY=20b \
mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs \
clientcount=4 osscount=2 mdscount=2 \
mdssizegb=10 ostsizegb=10 austeroptions=-R failover=true iscsi=1 \
testlist=replay-single,replay-single,replay-single,replay-single,replay-single
Change-Id: Ic126bd3c58a6d2da1c69c7231e88a8977417d37b
Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Reviewed-on: http://review.whamcloud.com/11767
Tested-by: Jenkins
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Tue, 23 Dec 2014 10:45:04 +0000 (03:45 -0700)]
LU-6068 misc: update Intel copyright messages 2014
Update the copyright messages for changes made in 2014.
In some cases, changes made in 2013 did not generate an updated
copyright message, so these files are updated accordingly. In some
cases there were no licenses on newly-added files, so add them (either
with new copyrights, or including copyrights from the source files).
Some types of changes that are not actually changing the code
(e.g. whitespace, mass changes to add or remote prefixes, patches
updating the copyright messages, patch reversions and the reverted
patch, patches that only delete code, etc.) have been excluded.
The updatecw.sh script was updated to fix a number of problems:
- it now handles comments in nroff (man page) files properly
- remove need to handle non-Intel copyrights specially
- allow subdirectory to be specified on command-line
- improve script debugging when running with "sh -vx"
- print patches changing the most files, to catch mass replacements
There are still a few cases that need to be fixed up by hand (e.g.
files split up and "git log --follow" can't find the original).
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Ib4875400a4908dc79ddc8b82ab0fe3ee903ebbe5
Reviewed-on: http://review.whamcloud.com/13174
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Fri, 19 Dec 2014 14:49:06 +0000 (17:49 +0300)]
LU-5478 utils: get rid of obd_* typedefs
We have a bunch of typedefs for common things that made no sense
and hid the actual type from plain view.
Replace them with proper uXX or sXX types.
Exception is in lustre_idl.h and lustre_ioctl.h where
they are replaced with __uXX and __sXX to be able to be included
in userspace
patch 8 in series: modify utils and ptlrpc
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I618e6e8465c845951f4735ac7ff2743544534809
Reviewed-on: http://review.whamcloud.com/13150
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Dmitry Eremin [Fri, 19 Dec 2014 14:37:39 +0000 (17:37 +0300)]
LU-5478 osd: get rid of obd_* typedefs
We have a bunch of typedefs for common things that made no sense
and hid the actual type from plain view.
Replace them with proper uXX or sXX types.
Exception is in lustre_idl.h and lustre_ioctl.h where
they are replaced with __uXX and __sXX to be able to be included
in userspace
patch 7 in series: modify osd-ldiskfs/osd-zfs
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I9809d57ae8665fb7a932645cc2258d3adc1a0ff1
Reviewed-on: http://review.whamcloud.com/13149
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Dmitry Eremin [Fri, 19 Dec 2014 14:09:55 +0000 (17:09 +0300)]
LU-5478 obdclass: get rid of obd_* typedefs
We have a bunch of typedefs for common things that made no sense
and hid the actual type from plain view.
Replace them with proper uXX or sXX types.
Exception is in lustre_idl.h and lustre_ioctl.h where
they are replaced with __uXX and __sXX to be able to be included
in userspace
patch 5 in series: modify obdclass/obdecho/ofd
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I674972390c808001249ce3c23a9f58888c13d1be
Reviewed-on: http://review.whamcloud.com/13146
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Dmitry Eremin [Fri, 19 Dec 2014 13:58:09 +0000 (16:58 +0300)]
LU-5478 mdc: get rid of obd_* typedefs
We have a bunch of typedefs for common things that made no sense
and hid the actual type from plain view.
Replace them with proper uXX or sXX types.
Exception is in lustre_idl.h and lustre_ioctl.h where
they are replaced with __uXX and __sXX to be able to be included
in userspace
patch 4 in series: modify mdc/mdt/mgc
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I35b51c3e722f7463b7bab04a5fb2e19544b47178
Reviewed-on: http://review.whamcloud.com/13145
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Dmitry Eremin [Fri, 19 Dec 2014 10:11:30 +0000 (13:11 +0300)]
LU-5478 style: get rid of obd_* typedefs
We have a bunch of typedefs for common things that made no sense
and hid the actual type from plain view.
Replace them with proper uXX or sXX types.
Exception is in lustre_idl.h and lustre_ioctl.h where
they are replaced with __uXX and __sXX to be able to be included
in userspace
First patch in series: modify main includes
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I49f53ea25e28b1af31dd7d4cd36b27cf644464fc
Reviewed-on: http://review.whamcloud.com/13140
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
John L. Hammond [Tue, 23 Dec 2014 22:10:33 +0000 (17:10 -0500)]
LU-2675 obd: remove linux/obd.h
Remove the now unneeded header lustre/include/linux/obd.h.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I70c0b40fded991b1e871bf8624cb91f39e065ea7
Reviewed-on: http://review.whamcloud.com/13123
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jian Yu [Fri, 12 Dec 2014 07:34:15 +0000 (23:34 -0800)]
LU-5905 tests: add version check codes to conf-sanity test 56
This patch adds Lustre version check codes into conf-sanity test
56 to make the test interoperate with servers that do not have
the following changes:
Lustre-commit:
3dc232918c7c9b95398ac6a0dc4c2130adc2851d
Lustre-change: http://review.whamcloud.com/12145
or
Lustre-commit:
e6745116cc372647f785c4063389734620c92e91
Lustre-change: http://review.whamcloud.com/12319
Test-Parameters: alwaysuploadlogs envdefinitions=ONLY=56 \
ossjob=lustre-b2_6 mdsjob=lustre-b2_6 ossbuildno=2 mdsbuildno=2 \
testlist=conf-sanity
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I8c1b41f6912c006d4bb9edd756af3fa345dd55f3
Reviewed-on: http://review.whamcloud.com/13045
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Li Wei <wei.g.li@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Thu, 4 Dec 2014 17:10:44 +0000 (11:10 -0600)]
LU-2675 lnet: remove lnet/include/lnet/linux/
Remove the linux specific headers from lnet/include/lnet/linux/,
moving whatever was worthwhile from them to their parent headers or
elsewhere. Also remove lnet/klnds/socklnd/socklnd_lib-linux.h and
rename lnet/klnds/socklnd/socklnd_lib-linux.c to socklnd_lib.c.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Id10683e95277791f9c99492907369a608b28a1fb
Reviewed-on: http://review.whamcloud.com/12932
Tested-by: Jenkins
Reviewed-by: Isaac Huang <he.huang@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Li Wei [Fri, 28 Nov 2014 09:15:04 +0000 (17:15 +0800)]
LU-6027 mdd: Don't list "trusted.link" for orphans
Commit 09fe679 prevents mdd_xattr_get() from accessing XATTR_NAME_LINK
EAs of orphan objects but leaves mdd_xattr_list() unchanged. This is
problematic for fgetxattr() calls on orphan objects when EA cache
refills are involved. To respond to such a refill, mdt lists all of
the object's EA names and gets the value for each name. In the case
of an orphan, XATTR_NAME_LINK EA will be listed, but corresponding
mdd_xattr_get() call will return -ENOENT, causing the fgetxattr() call
to fail eventually. This patch updates mdd_xattr_list() to filter out
XATTR_NAME_LINK EAs for orphan objects as well, following commit
09fe679.
Change-Id: Ie59e4c9342056bacf57a82cf9bf77cfdedc99f6d
Signed-off-by: Li Wei <wei.g.li@intel.com>
Reviewed-on: http://review.whamcloud.com/12884
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Li Wei [Sat, 29 Nov 2014 03:21:19 +0000 (11:21 +0800)]
LU-6027 osd-zfs: Preserve lu_buf when listing EAs
Current osd_xattr_list() in osd-zfs modifies its lu_buf argument,
leaving lu_buf::lb_buf pointing to the end of the original buffer.
This behavior is unnecessary and counter-intuitive. This patch
preserves the content of the lu_buf argument and updates OSD API to
reduce the chance of future regressions.
Change-Id: I0d3e334848734ca6f44778e87b5ff3b044826784
Signed-off-by: Li Wei <wei.g.li@intel.com>
Reviewed-on: http://review.whamcloud.com/12880
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Johann Lombardi <johann.lombardi@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Tue, 25 Nov 2014 14:50:16 +0000 (08:50 -0600)]
LU-2675 llite: remove some cl wrappers
In llite remove the wrapper functions and macros:
cl_i2info()
cl_i2sbi()
cl_iattr2fd()
cl_inode_info
cl_inode_mode()
cl_inode_{a,m,c}time()
cl_isize_{lock,unlock}()
cl_isize_{read,write,write_nolock}()
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I310c16c842781cbec741a7c43b746b582546badd
Reviewed-on: http://review.whamcloud.com/12850
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Tue, 17 Jun 2014 22:20:54 +0000 (16:20 -0600)]
LU-1399 config: check lustre_cfg_new() return
Check if lustre_cfg_new() fails its allocation and pass that error
on to the caller instead of dereferencing the NULL pointer.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I8a934777ea68a866f5f087a6c0948a6dbc3ebbe5
Reviewed-on: http://review.whamcloud.com/10739
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Patrick Farrell [Mon, 1 Dec 2014 06:56:18 +0000 (00:56 -0600)]
LU-4239 mdt: Optimize fid2path
On directory paths 100 or more deep, fid2path
gives a spurious:
ioctl err -75: Value too large for defined data type
This bug occurs because all of the fids found during
a path lookup are stored in pli_fids and the
array is too small for very deep paths.
The stored fids are not used except for the last one, so
this array is not necessary.
Additionally, from Frank Zago:
It is not necessary to copy back and forth the fid, linkno,
and path.
Keep the input getinfo_fid2path, and remove
struct path_lookup_info
which was mostly a copy of getinfo_fid2path. This saves
several copies and a memory allocation.
The fid to be returned back to user space is stored in
fp->gf_fid.
Also added doxygen comments and replaced an 'EXIT' and
'return rc' with 'RETURN(rc)'.
Signed-off-by: Patrick Farrell <paf@cray.com>
Signed-off-by: Frank Zago <fzago@cray.com>
Change-Id: I357b4d91b9c65a836d289d5815c763306ac34565
Reviewed-on: http://review.whamcloud.com/10717
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Li Xi [Wed, 8 Oct 2014 12:27:22 +0000 (20:27 +0800)]
LU-3727 nfs: Fix ll_get_parent() LBUG caused by permission
When ll_get_parent() is trying to get the attributes of parent
directory, EACCES will be returned if NFS daemon user does not
have the execute permission of the child directory. This patch
fixes the problem by skipping permission check for ll_get_parent().
Signed-off-by: Li Xi <lixi@ddn.com>
Change-Id: Id922d170c89519b755f69a2818d6ed53f8cc7ad9
Reviewed-on: http://review.whamcloud.com/7327
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Tested-by: Jenkins
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Fri, 19 Dec 2014 16:44:05 +0000 (10:44 -0600)]
LU-6055 hsm: improve error handling in sanity-hsm
In path2fid() propagate errors from 'lfs path2fid'. In
file_creation_failure() redirect the output of df to stderr so that it
is not captured by command substitution.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I7fabf46af4aa36c6affef45cf8c9b044a7aa29dd
Reviewed-on: http://review.whamcloud.com/13154
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Oleg Drokin [Fri, 26 Sep 2014 02:15:36 +0000 (22:15 -0400)]
LU-5648 ofd: In destroy orphan case always let MDS know last id
It appears that if we think MDS asked us to delete too many
precreate objects (more than 2 MAX_PRECREATE counts)
then we don't do the precreation, but do not let MDS know it needs
to advance the last_id forward by quite a bit.
This lead to reassigning these objects to other files after recovery.
Change-Id: Icbd5b9e3cd81129f56ee5cd3aafd5f418affbecc
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Reviewed-on: http://review.whamcloud.com/12067
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Bob Glossman [Tue, 16 Dec 2014 21:18:19 +0000 (13:18 -0800)]
LU-6035 kernel: kernel update RHEL6.6 [2.6.32-504.3.3.el6]
Update RHEL6.6 kernel to 2.6.32-504.3.3.el6
Test-Parameters: clientdistro=el6.6 mdsdistro=el6.6\
ossdistro=el6.6 mdsfilesystemtype=ldiskfs\
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I07f3a84b776e3bcd8f1b7ba2250e0bbb3d47f010
Reviewed-on: http://review.whamcloud.com/13106
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Alexander.Boyko [Wed, 20 Aug 2014 13:07:35 +0000 (17:07 +0400)]
LU-5528 ptlrpc: fix race between connect vs resend
Buggy code at ptlrpc_connect_interpret()
finish:
rc = ptlrpc_import_recovery_state_machine(imp);
...
Set import connection flags
When import has FULL state ptlrpc_import_recovery_state_machine()
wakeup all waiters on import and all delayed request, which was
resented. And it could happened that request was send without
updated flags and AT is disabled. If such request is in progress
on the server, server drop the new instance, and could do early reply
for it. But this early reply confuse client, cause it wait real
reply(no AT for this request). Client try to touch buffer outside
reply and got EPROTO error.
The same bug existed for initital connect too. Import became FULL
before import connection flags was set.
Signed-off-by: Alexander Boyko <alexander_boyko@xyratex.com>
Xyratex-bug-id: MRP-2034
Change-Id: Ic6959e3194c23bdd88a7da1a8f0ea9b4cd4285f0
Reviewed-on: http://review.whamcloud.com/11723
Reviewed-by: Li Wei <wei.g.li@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Alexander Boyko <alexander.boyko@seagate.com>
Reviewed-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Niu Yawei [Tue, 16 Dec 2014 02:59:55 +0000 (21:59 -0500)]
LU-5773 test: reduce thread count
Reduce thread count for the test_1c of obdfilter-survery test to
prevent OOM on vm.
Test-Parameters: alwaysuploadlogs \
envdefinitions=SLOW=yes,ENABLE_QUOTA=yes \
mdtfilesystemtype=ldiskfs mdsfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs \
mdtcount=1 testlist=obdfilter-survey,obdfilter-survey,obdfilter-survey
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I83926a538827dea6682bd0f84434ba5d351cfd88
Reviewed-on: http://review.whamcloud.com/13078
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Kit Westneat [Sat, 13 Dec 2014 16:11:53 +0000 (11:11 -0500)]
LU-6024 nodemap: fix ACL handling in mdt_finish_open
This patch adds ACL mapping code to mdt_finish_open so that clients
get a correctly mapped list. The code is essentially the same as in
mdt_getattr_internal.
Signed-off-by: Kit Westneat <kit.westneat@gmail.com>
Change-Id: I3cf0ba2128ba8c4027359329f423d2327005f082
Reviewed-on: http://review.whamcloud.com/13062
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Andrew Perepechko <andrew.perepechko@seagate.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Thu, 11 Dec 2014 13:17:59 +0000 (16:17 +0300)]
LU-6023 build: fix libselinux support
Enabling correct libselinux-devel detection by searching for
the same header file in autoconf that is used in code.
Link mount_osd_ldiskfs.so with -lselinux.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I085e91464a263b6d1b1be13971e54dfc80a4e347
Reviewed-on: http://review.whamcloud.com/13030
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Andriy Skylysh [Tue, 16 Dec 2014 03:16:50 +0000 (20:16 -0700)]
LU-4660 osd-ldiskfs: fix osd_bufs_get() error handling
Unlock pages just locked during cleanup if page loading failed.
Clean up code style, add function comment blocks.
Signed-off-by: Andriy Skulysh <Andriy_Skulysh@xyratex.com>
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Ie523b0cfea30bc55ecd58db0849b1b005a393fc7
Reviewed-on: http://review.whamcloud.com/9344
Tested-by: Jenkins
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Fri, 19 Dec 2014 18:09:51 +0000 (12:09 -0600)]
LU-6058 test: log to stderr
In test-framework.sh's log() function print the message to stderr
instead of stdout. This ensures that the message will not be not
captured by command substitution. Similarly print bash stack traces to
stderr from error().
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: If9ca4af1c8d66f9bb3767f4fc3c04fb6966de915
Reviewed-on: http://review.whamcloud.com/13155
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Niu Yawei [Mon, 17 Nov 2014 06:59:44 +0000 (01:59 -0500)]
LU-5916 lod: inherit default pool setting properly
When parent directory has no default striping but has only
default pool setting, the pool setting should be inherited
by subdir properly.
This patch fixed the problem by changing the
LOVEA_DELETE_VALUES() to check pool name as well, and it
removed the unused function lod_store_def_striping().
Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Change-Id: I7ce5251b6c601309aeb91a1b903bde02d0712e25
Reviewed-on: http://review.whamcloud.com/12746
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
John L. Hammond [Wed, 5 Nov 2014 17:40:02 +0000 (11:40 -0600)]
LU-5814 lov: flatten struct lov_stripe_md
Flatten out the lsm_wire struct from the middle of struct
lov_stripe_md and remove the member name macros.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I53455d4f42860d4c7e5541f1f18b0c451a022c28
Reviewed-on: http://review.whamcloud.com/12581
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Tue, 4 Nov 2014 22:03:36 +0000 (16:03 -0600)]
LU-5971 llite: merge lclient.h into llite/vvp_internal.h
Move the definition of struct cl_client_cache to
lustre/include/cl_object.h and move the rest of
lustre/include/lclient.h in to lustre/llite/vvp_internal.h.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Ib3ca8f1693c517a77c5bfd7c7cced9bfc440c747
Reviewed-on: http://review.whamcloud.com/12592
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Andreas Dilger [Thu, 18 Dec 2014 00:42:04 +0000 (17:42 -0700)]
LU-4536 tests: disable sanity test_65ic for ZFS
Disable sanity.sh test_65ic due to frequent test timeouts on
review-zfs testing. This will allow the rest of the tests to be
enforced, and will allow this problem to be debugged separately.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I1ea93b5c2dfb0a3bbe0265a283ab7512044d6cd2
Reviewed-on: http://review.whamcloud.com/13114
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Frank Zago [Thu, 11 Dec 2014 16:49:04 +0000 (11:49 -0500)]
LU-5396 obdclass: make some functions static
Some functions and variables are only used in their C file, so reduce
their scope. This reduces the code size, and fixes sparse warnings
such as:
warning: symbol 'proc_lnet_routes' was not declared.
Should it be static?
warning: symbol 'proc_lnet_routers' was not declared.
Should it be static?
Some prototypes were removed from C files and added to the proper
header.
Signed-off-by: Frank Zago <fzago@cray.com>
Change-Id: I35a7ae4d9efd7098d93d3500f4854ef074fce179
Reviewed-on: http://review.whamcloud.com/12220
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jinshan Xiong [Mon, 15 Dec 2014 20:48:23 +0000 (12:48 -0800)]
LU-6028 ldlm: move LDLM_GID_ANY to lustre_dlm.h
lustre_idl.h only includes wire data; lustre_dlm.h is the
right place for LDLM_GID_ANY.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Change-Id: I832796e924116e80c079494f6928c037e7e4f2aa
Reviewed-on: http://review.whamcloud.com/13074
Tested-by: Jenkins
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Mon, 15 Dec 2014 19:15:43 +0000 (14:15 -0500)]
LU-3963 libcfs: Use kernel's strncasecmp and remove cfs_get_blocked_sigs
Remove libcfs function cfs_strncasecmp() since the kernel
already has its own strncasecmp(). Lastly remove from libcfs.h
cfs_get_blocked_sigs() since this function no longer exist.
Change-Id: Ief5bac8ac23114c8e34d988906b06dda8eb77306
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Reviewed-on: http://review.whamcloud.com/13070
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Reviewed-by: Patrick Farrell <paf@cray.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Thu, 11 Dec 2014 13:27:21 +0000 (16:27 +0300)]
LU-6001 build: fix autoconf DLC detection output
checking for yaml_parser_initialize in -lyaml... no
checking whether to enable dlc... yes
no (libyaml not present)
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I29b968295b508522f9158ac920ee3bb862ae4b4e
Reviewed-on: http://review.whamcloud.com/13031
Tested-by: Jenkins
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Amir Shehata <amir.shehata@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Thu, 11 Dec 2014 15:13:14 +0000 (10:13 -0500)]
LU-5651 test: run replay-single test 93 only when supported.
The new test for replay-single will only work for lustre
versions that contain commit
f61cec84. This patch does
lustre version checking to over failing on platforms that
are not fixed.
Change-Id: I0feffb4b3c131db68c115a111bc87a229bda23e3
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Reviewed-on: http://review.whamcloud.com/12942
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 1 Dec 2014 19:57:06 +0000 (13:57 -0600)]
LU-2675 mdc: remove mdc_sendpage()
Remove the unused function mdc_sendpage() and supporting crud.
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: I09792c5b25f10351eead9e67d42de5821f4d7bc1
Reviewed-on: http://review.whamcloud.com/12893
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 27 Oct 2014 19:40:29 +0000 (14:40 -0500)]
LU-2675 llapi: find_param cleanup part II
Add fp_ prefixes to the remaining members of struct find_param. Also
add underscores between words in several member names:
obduuid -> fp_obd_uuid
stripesize_units -> fp_stripe_size_units
...
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Iaa62a6b44e2929f3925525a97fec0434243dec1b
Reviewed-on: http://review.whamcloud.com/12851
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Dmitry Eremin [Wed, 29 Oct 2014 14:17:03 +0000 (17:17 +0300)]
LU-5577 ptlrpc: fix comparison between signed and unsigned
Change return type and size argiments of lustre_msg_hdr_size(),
lustre_msg_buf{len,count}() and req_capsule_*_size() to __u32.
Change type of req_format->rf_idx and req_format->rf_fields.nr
to size_t. Also return zero for incorrect message magic instead
of -EINVAL. This will be more robust because of few of them after
LASSERTF(0, "...") and will not be returned. In the rest places
it return zero size instead of huge number after implicit
unsigned conversion.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: Iae8b83b4285a1ea2fa010f9de2792426c1772b2c
Reviewed-on: http://review.whamcloud.com/12475
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bobi Jam [Tue, 28 Oct 2014 03:30:38 +0000 (11:30 +0800)]
LU-5823 clio: add coo_getstripe interface
Use cl_object_operations::coo_getstripe() to handle
LL_IOC_LOV_GETSTRIPE ops.
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Change-Id: Iee0d007f773ead35434eb88e7ca2f31c0bc42c11
Reviewed-on: http://review.whamcloud.com/12452
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Mon, 13 Oct 2014 17:22:00 +0000 (21:22 +0400)]
LU-5577 obdclass: do_body_ops->dbo_{read,write} return ssize_t
Fix warnings when use return value from do_body_ops->dbo_{read,write}.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I6f3324e0799eff699b44e7b2b09ee4b3d2298c0d
Reviewed-on: http://review.whamcloud.com/12388
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Fri, 10 Oct 2014 16:55:00 +0000 (20:55 +0400)]
LU-5577 obdclass: (*dt_index_page_build_t) arg nob to size_t
Change the type accordant usage and check for correct input.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: Ic80025e95eddbc9fb51efd83c54573e5f310bb0b
Reviewed-on: http://review.whamcloud.com/12382
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Fri, 10 Oct 2014 15:41:29 +0000 (19:41 +0400)]
LU-5577 obdclass: change cl_fault_io->ft_nob to size_t
Change the type accordant usage.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I866b9e0fc24c1d2fd281b40d03e8b4d73f159e31
Reviewed-on: http://review.whamcloud.com/12380
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Fri, 10 Oct 2014 14:53:52 +0000 (18:53 +0400)]
LU-5577 libcfs: fix warnings in libcfs/curproc.h
Fix compiler warnings that annoying from searching other issues.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I17cdc1d94b58d2c2f3284665090ad7eef991d5ac
Reviewed-on: http://review.whamcloud.com/12379
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Wed, 5 Nov 2014 01:24:49 +0000 (20:24 -0500)]
LU-5577 obd: change brw_page->count to unsigned
Pages count is unsigned. So, change the type accordant usage.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I71a9c9e31f90a7e6aa456801c6f47dc9ee9c0df2
Reviewed-on: http://review.whamcloud.com/12378
Tested-by: Jenkins
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Dmitry Eremin [Tue, 16 Dec 2014 15:05:48 +0000 (18:05 +0300)]
LU-6029 test: add wrap up duration in conf-sanity test_84
The time of execution the function target_finish_recovery() can be
very long but "obd->obd_recovery_end = cfs_time_current_sec();" set
at the end. Therefore the recovery time always will be more than
hard limit and depends how long target_finish_recovery() function
was executed.
To avoid instability in test_84 just add 5 seconds as wrap up time.
Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Change-Id: I53263f9a248429f59741f701fcc7b5b6d693325e
Reviewed-on: http://review.whamcloud.com/13087
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Sergey Cheremencev <sergey_cheremencev@xyratex.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Lai Siyao [Thu, 14 Aug 2014 13:22:50 +0000 (21:22 +0800)]
LU-5461 mdc: don't add to page cache upon failure
Reading directory pages may fail on MDS, in this case client should
not cache a non-up-to-date directory page, because it will cause
a later read on the same page fail.
Add sanity 242 for this.
Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Change-Id: Icbb55c0ecc4b88913b08ac8998c2673ec15b18ea
Reviewed-on: http://review.whamcloud.com/11450
Tested-by: Jenkins
Reviewed-by: Fan Yong <fan.yong@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
JC Lafoucriere [Mon, 3 Nov 2014 02:17:45 +0000 (21:17 -0500)]
LU-4178 tests: add messages to sanity-hsm
Add messages in case of file creation failure so we
can diagnose the failure.
Signed-off-by: JC Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Change-Id: I080c4cbb6bfe5a65d964f32a36d7c1dba35d0c27
Reviewed-on: http://review.whamcloud.com/8113
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Wang Shilong [Mon, 20 Oct 2014 12:41:03 +0000 (20:41 +0800)]
LU-5771 ldiskfs: cleanup orphan inode in error path
Commit from upstream:
4538821993f4486c76090dfb377c60c0a0e71ba3
There were some error paths in ext4_delete_inode() which was not
dropping the inode from the orphan list. This could lead to a BUG_ON
on umount when the orphan list is discovered to be non-empty.
Signed-off-by: Wang Shilong <wshilong@ddn.com>
Change-Id: Ica2c87ae378cc23eed70c1bfeb96db3852d218dc
Reviewed-on: http://review.whamcloud.com/12349
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Bruno Faccini [Tue, 28 Oct 2014 13:50:00 +0000 (14:50 +0100)]
LU-3852 tests: skip tests with large file when no room
Strengthen sanity-hsm sub-tests using large files by skipping
then if not enough space instead to fail.
Also re-enable test_251, and test_[200,221,223b] (for LU-4178),
who have all been wrongly disabled because of their multiple
failures on an auto-test node with very low disk space for the
Servers VMs!!
Also fix a bug in sanity/test_56w that should not be run if
$OSTCOUNT < 2, that was triggered due to my very restictive
test environment osscount/ostcount=1 to limit the Lustre FS size
as much as possible.
Test-Parameters: ostcount=1 ostsizegb=1 envdefinitions=SLOW=yes
Test-Parameters: osscount=1 mdscount=1 testlist=sanity-hsm
Signed-off-by: Bruno Faccini <bruno.faccini@intel.com>
Change-Id: I26bc36e4304a916f29149e6ad19306ed59c9a616
Reviewed-on: http://review.whamcloud.com/12456
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Nunez <james.a.nunez@intel.com>
Andreas Dilger [Tue, 16 Dec 2014 19:55:15 +0000 (12:55 -0700)]
LU-5242 tests: except sanity test_132 for ZFS
Sanity test_132() is failing regularly on ZFS, but SOM is not a
feature we support. Disable this test until the problem can be
resolved. The SOM functionality itself will continue to be tested
with ldiskfs, and should not be OSD-specific.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: I8fa38d993be0ab5afd9f293d9878cb862fc50c46
Reviewed-on: http://review.whamcloud.com/13093
Tested-by: Jenkins
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Bob Glossman [Wed, 10 Dec 2014 15:13:05 +0000 (07:13 -0800)]
LU-6011 kernel: kernel update [RHEL7 3.10.0-123.13.1.el7]
update RHEL7 kernel to 3.10.0-123.13.1.el7
Test-Parameters: clientdistro=el7 mdsfilesystemtype=ldiskfs\
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I6f9234f982a97646fa38c350d4740672cfc331e7
Reviewed-on: http://review.whamcloud.com/13022
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Tue, 4 Nov 2014 02:26:33 +0000 (21:26 -0500)]
LU-4536 ldlm: Recalculate interval in ldlm_pool_recalc()
Instead of rechecking a static value, recalculate to see if pool stats
need to be updated.
Add newline so message will print instead of warning about missing
newline.
Test-Parameters: mdsfilesystemtype=zfs mdtfilesystemtype=zfs ostfilesystemtype=zfs testlist=sanity,sanity,sanity,sanity,sanity,sanity,sanity,sanity,sanity,sanity
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: Ic31cc0c1d09a85a9bd5ee04ac34c388263190df1
Reviewed-on: http://review.whamcloud.com/12547
Tested-by: Jenkins
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Nathaniel Clark [Wed, 10 Dec 2014 01:13:04 +0000 (20:13 -0500)]
LU-5986 test: Ensure correct start for conf-sanity/84
For review-zfs:
1) test 79 fails to reformat nodes
2) test 80 fails silently
3) test 84 chokes
Fix test 79 to reformat at end and fix 80 die when it should.
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Change-Id: I42c635fffd89eda9ccd6f3f9d739ff8ac75afcf6
Reviewed-on: http://review.whamcloud.com/13016
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Sat, 27 Sep 2014 23:10:34 +0000 (07:10 +0800)]
LU-6012 scrub: NOT miss to auto detect inconsistent OI mapping
When full scrub is triggered automatically, its flags should
be set as SF_INCONSISTENT.
For lookup case, we should check whether current OI mapping is
consistent or not, even if the current OI scrub flags is NOT
SF_INCONSISTENT.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I99ea077ae79fcdfedd7bb16c2a664714e0ea5ea3
Reviewed-on: http://review.whamcloud.com/13020
Tested-by: Jenkins
Tested-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Andreas Dilger [Fri, 14 Nov 2014 21:06:34 +0000 (14:06 -0700)]
LU-1445 lod: clean up lod_fld_lookup() return codes
Don't return "rc" when it is known that this will always be "0".
This confuses the reader into thinking that this is an error path
when it is in fact a no-op shortcut that returns success.
Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Ie4571226b3e90c866b958cf6ab65f6077abcab07
Reviewed-on: http://review.whamcloud.com/12727
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Fan Yong [Wed, 24 Sep 2014 09:30:56 +0000 (17:30 +0800)]
LU-5885 lfsck: deadlock when remove striped dir
There is potential deadlock race condition between removing
striped directory and namespace LFSCK. Consider the following
scenario:
1) The LFSCK thread obtained the master object firstly, at
that time, the master object has not been destroyed yet.
2) One RPC service thread destroyed the master and all its
slave objects (shards). Because the LFSCK is referencing
the master object, then the master object will be marked
as dying in RAM. On the other hand, the master object is
referencing all its slave objects, then all slave objects
will be marked as dying in RAM also.
3) The LFSCK thread tries to find some slave object with the
master object referenced. Then it will find that the slave
object is dying. According to the object visibility rules:
the object with dying flag cannot be returned to others.
So the LFSCK thread has to wait until the dying object has
been purged from RAM, then it can allocate a new object (with
the same FID) in RAM. Unfortunately, the LFSCK thread itself
is referencing the master object, and cause the master object
cannot be purged, then cause the slave object cannot be purged
also. So the LFSCK thread will fall into deadlock.
To resolve such trouble, the LFSCK should use non-blocked version
lu_object_find() to locate the slave object of the striped dir,
and return failure immediately (instead of wait) when it finds
dying (slave) object.
This patch also contorls the async pipeline depth between the
LFSCK main engine and the namespace assistant thread to avoid
potential RAM pressure.
Some other code adjustment to avoid potential data overflow
that may cause weird LFSCK statistics information.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Change-Id: I00c601eca8ade5d2e4260c729463f7ecdba0ed53
Reviewed-on: http://review.whamcloud.com/12741
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Amir Shehata [Thu, 8 May 2014 17:47:56 +0000 (10:47 -0700)]
LU-3353 ptlrpc: Suppress error message when imp_sec is freed
There is a race condition on client reconnect when the import
is being destroyed. Some outstanding client bound requests
are being processed when the imp_sec has alread been freed.
Ensure to suppress the error message in import_sec_validate_get()
in that case
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: I44bc27c804259d4e4b6564460318732113b251a9
Reviewed-on: http://review.whamcloud.com/10200
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Nathaniel Clark [Wed, 5 Nov 2014 18:05:22 +0000 (13:05 -0500)]
LU-3573 osd-zfs: Only advance zap cursor as needed
Only advance the zap cursor when ozi_pos is not advanced, otherwise
occasionally the a file could get "lost" because the zap_cursor would
advance over it before the retrieve happened. Handle '..' like '.'
when retrieving ZAP values.
Test-Parameters: mdtfilesystemtype=zfs mdsfilesystemtype=zfs ostfilesystemtype=zfs testlist=conf-sanity,conf-sanity,conf-sanity,conf-sanity
Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Isaac Huang <he.huang@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Change-Id: I916573c70c8828bed6905b5eda9344b4a49b7f11
Reviewed-on: http://review.whamcloud.com/12904
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Kit Westneat [Tue, 8 Jul 2014 18:57:40 +0000 (14:57 -0400)]
LU-4647 nodemap: add tests to sanity-sec for nodemap mapping
Added tests to sanity-sec.sh, as outlined in the original nodemap
spec. The tests currently only work with a single OSS node, but this
will be fixed in a future update. These tests test basic permissions
and quota handling of the nodemapper, as well as ACL mapping.
Signed-off-by: Kit Westneat <kit.westneat@gmail.com>
Change-Id: Ieb29091e5b3110593973a5eb03680e86a769b449
Reviewed-on: http://review.whamcloud.com/10406
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Mon, 8 Dec 2014 19:48:22 +0000 (11:48 -0800)]
LU-5373 test: ignore command return value in sanity test_33b
Since the test is only looking for a panic and the command
used has different returns depending on kernel version,
ignore the command return value in all cases.
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I63e9d0589b1c11736c9afbb8bee8ea4e11b30a4f
Reviewed-on: http://review.whamcloud.com/12992
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Bob Glossman [Tue, 11 Nov 2014 23:16:03 +0000 (15:16 -0800)]
LU-5909 kernel: kernel update RHEL6.6 [2.6.32-504.1.3.el6]
Update RHEL6.6 kernel to 2.6.32-504.1.3.el6
Test-Parameters: clientdistro=el6.6 mdsdistro=el6.6\
ossdistro=el6.6 mdsfilesystemtype=ldiskfs\
mdtfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I6a320f2a2806b12ee7c07645bed212792965da99
Reviewed-on: http://review.whamcloud.com/12815
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Minh Diep <minh.diep@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Henri Doreau [Fri, 31 Oct 2014 23:04:19 +0000 (00:04 +0100)]
LU-5837 llite: ll_getparent cleanup
Avoid unneeded allocation. Get read-only attributes from the user
getparent structure and write the modified attributes only, instead
of populating a whole structure in kernel and copying it back.
Signed-off-by: Henri Doreau <henri.doreau@cea.fr>
Change-Id: Ifc0632870f80733194384d02d1b4962cdcd75658
Reviewed-on: http://review.whamcloud.com/12527
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: frank zago <fzago@cray.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Liang Zhen [Tue, 28 Oct 2014 10:04:51 +0000 (18:04 +0800)]
LU-5485 lnet: peer aliveness status and NI status
A couple of changes to improve aliveness detection:
- When LNet received a message, it can determine peer of this message
is alive
- When LNet recieved a message from remote network, it can determine
router is alive and NI status on router is UP.
Signed-off-by: Liang Zhen <liang.zhen@intel.com>
Change-Id: I7133987c5c8728248cce7bc0a95048b26bc6611a
Reviewed-on: http://review.whamcloud.com/12453
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Isaac Huang <he.huang@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jian Yu [Wed, 19 Nov 2014 21:10:10 +0000 (13:10 -0800)]
LU-5932 tests: load sunrpc module before insmod ptlrpc_gss
Lustre ptlrpc_gss module depends on Linux kernel sunrpc module.
This patch fixes load_module() in test-framework.sh to load the
sunrpc module before loading ptlrpc_gss module by using insmod.
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: Idf22fa9023a2fada9038e16fbc3e8a61530266bc
Reviewed-on: http://review.whamcloud.com/12790
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Wei Liu <wei3.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
John L. Hammond [Mon, 24 Nov 2014 22:08:27 +0000 (16:08 -0600)]
LU-2675 llite: remove lli_lvb
In struct ll_inode_info remove the struct ost_lvb lli_lvb member and
replace it with obd_time lli_{a,m,c}time. Rename ll_merge_lvb() to
ll_merge_attr(). Remove cl_merge_lvb() and replace calls to it with
calls to ll_merge_attr().
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Change-Id: Idaf8a89d2e4243e62a23cab949c3c129001bb9f3
Reviewed-on: http://review.whamcloud.com/12849
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Amir Shehata [Wed, 5 Nov 2014 22:32:59 +0000 (14:32 -0800)]
LU-5828 lnet: showing buffers problem with mulitple CPTs
Overloading an iterator variable in lustre_lnet_show_routing()
caused only the first CPT information to be displayed.
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: Ic75dccc0c3537b8272d1c2687a759fbcc23052e8
Reviewed-on: http://review.whamcloud.com/12593
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Doug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Bob Glossman [Tue, 4 Nov 2014 19:14:35 +0000 (11:14 -0800)]
LU-5858 obdclass: eliminate NULL error return
Always return an ERR_PTR() on errors, never return a NULL,
in lu_object_find_slice(). Also clean up callers who
no longer need special case handling of NULL returns.
Signed-off-by: Bob Glossman <bob.glossman@intel.com>
Change-Id: I57ddb38abaec7caf57bb63a75dbd76e181ba72b2
Reviewed-on: http://review.whamcloud.com/12554
Tested-by: Jenkins
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Frank Zago [Tue, 30 Sep 2014 03:10:50 +0000 (22:10 -0500)]
LU-5396 mdc: (and lmv, mgc, osc) make some functions static
Some functions and variables are only used in their C file, so reduce
their scope. This reduces the code size, and fixes sparse warnings
such as:
warning: symbol 'proc_lnet_routes' was not declared.
Should it be static?
warning: symbol 'proc_lnet_routers' was not declared.
Should it be static?
Some prototypes were removed from C files and added to the proper
header.
Signed-off-by: Frank Zago <fzago@cray.com>
Change-Id: I8dcc5224c1da75cfb5ef7afb1fdb0f72422a3ac0
Reviewed-on: http://review.whamcloud.com/12222
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Frank Zago [Tue, 30 Sep 2014 02:39:31 +0000 (21:39 -0500)]
LU-5396 lov: (and ldlm) make some functions static
Some functions and variables are only used in their C file, so reduce
their scope. This reduces the code size, and fixes sparse warnings
such as:
warning: symbol 'proc_lnet_routes' was not declared.
Should it be static?
warning: symbol 'proc_lnet_routers' was not declared.
Should it be static?
Some prototypes were removed from C files and added to the proper
header.
Signed-off-by: Frank Zago <fzago@cray.com>
Change-Id: I86b7ada5c768f4b875fce55745f7492faabd4617
Reviewed-on: http://review.whamcloud.com/12221
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Frank Zago [Thu, 2 Oct 2014 02:05:25 +0000 (21:05 -0500)]
LU-5396 ptlrpc: make some functions static
Some functions and variables are only used in their C file, so reduce
their scope. This reduces the code size, and fixes sparse warnings
such as:
warning: symbol 'proc_lnet_routes' was not declared.
Should it be static?
warning: symbol 'proc_lnet_routers' was not declared.
Should it be static?
Some prototypes were removed from C files and added to the proper
header.
Signed-off-by: Frank Zago <fzago@cray.com>
Change-Id: Ic30c9c00be7fd161e0eb3aa2505c6d731c3d7a87
Reviewed-on: http://review.whamcloud.com/12219
Tested-by: Jenkins
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Frank Zago [Mon, 8 Sep 2014 00:24:35 +0000 (19:24 -0500)]
LU-5396 llite: make some functions static
Some functions and variables are only used in their C file, so reduce
their scope. This reduces the code size, and fixes sparse warnings
such as:
warning: symbol 'proc_lnet_routes' was not declared.
Should it be static?
warning: symbol 'proc_lnet_routers' was not declared.
Should it be static?
Some prototypes were removed from C files and added to the proper
header.
Signed-off-by: Frank Zago <fzago@cray.com>
Change-Id: Id6b13d2b5ceb30de02b60ed6be24d4a496454b70
Reviewed-on: http://review.whamcloud.com/12211
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Frank Zago [Sun, 7 Sep 2014 18:00:28 +0000 (13:00 -0500)]
LU-5396 libcfs: make some functions static
Some functions and variables are only used in their C file, so reduce
their scope. This reduces the code size, and fixes sparse warnings
such as:
warning: symbol 'proc_lnet_routes' was not declared.
Should it be static?
warning: symbol 'proc_lnet_routers' was not declared.
Should it be static?
Some prototypes were removed from C files and added to the proper
header.
Signed-off-by: Frank Zago <fzago@cray.com>
Change-Id: I5bdf94633fb94e435d32691d521ad7c1234018aa
Reviewed-on: http://review.whamcloud.com/12207
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Isaac Huang [Mon, 20 Oct 2014 19:32:20 +0000 (13:32 -0600)]
LU-5772 osd-zfs: irrelevant comment over __osd_xattr_load
Moved comment over __osd_xattr_load() to __osd_xattr_get(),
where it really belongs, and converted it to Doxygen format.
Added a few other minor cleanups as well.
Signed-off-by: Isaac Huang <he.huang@intel.com>
Change-Id: If73048d046419eaa4e23cbc5acde32c09b588996
Reviewed-on: http://review.whamcloud.com/12350
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Ned Bass <bass6@llnl.gov>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jian Yu [Tue, 18 Nov 2014 02:32:20 +0000 (18:32 -0800)]
LU-5443 libcfs: replace direct HZ access with kernel APIs
On some customer's systems, kernel was compiled with HZ defined to
100, instead of 1000. This improves performance for HPC applications.
However, to use these systems with Lustre, customers have to re-build
Lustre for the kernel because Lustre directly uses the defined
constant HZ.
Since kernel 2.6.21, some non-HZ dependent timing APIs become non-
inline functions, which can be used in Lustre codes to replace the
direct HZ access.
These kernel APIs include:
jiffies_to_msecs()
jiffies_to_usecs()
jiffies_to_timespec()
msecs_to_jiffies()
usecs_to_jiffies()
timespec_to_jiffies()
And here are some samples of the replacement:
HZ -> msecs_to_jiffies(MSEC_PER_SEC)
n * HZ -> msecs_to_jiffies(n * MSEC_PER_SEC)
HZ / n -> msecs_to_jiffies(MSEC_PER_SEC / n)
n / HZ -> jiffies_to_msecs(n) / MSEC_PER_SEC
n / HZ * 1000 -> jiffies_to_msecs(n)
This patch replaces the direct HZ access in libcfs module.
The patch also replaces ONE_BILLION with NSEC_PER_SEC,
and ONE_MILLION with USEC_PER_SEC in linux-time.h.
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I33846f378eb876cd8958ff0c397ffb56a552f256
Reviewed-on: http://review.whamcloud.com/11993
Tested-by: Jenkins
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Jian Yu [Wed, 19 Nov 2014 02:36:32 +0000 (18:36 -0800)]
LU-5443 ldiskfs: replace direct HZ access with kernel APIs
On some customer's systems, kernel was compiled with HZ defined to
100, instead of 1000. This improves performance for HPC applications.
However, to use these systems with Lustre, customers have to re-build
Lustre for the kernel because Lustre directly uses the defined
constant HZ.
Since kernel 2.6.21, some non-HZ dependent timing APIs become non-
inline functions, which can be used in Lustre codes to replace the
direct HZ access.
These kernel APIs include:
jiffies_to_msecs()
jiffies_to_usecs()
jiffies_to_timespec()
msecs_to_jiffies()
usecs_to_jiffies()
timespec_to_jiffies()
And here are some samples of the replacement:
HZ -> msecs_to_jiffies(MSEC_PER_SEC)
n * HZ -> msecs_to_jiffies(n * MSEC_PER_SEC)
HZ / n -> msecs_to_jiffies(MSEC_PER_SEC / n)
n / HZ -> jiffies_to_msecs(n) / MSEC_PER_SEC
n / HZ * 1000 -> jiffies_to_msecs(n)
This patch replaces the direct HZ access in ldiskfs module.
Test-Parameters: alwaysuploadlogs envdefinitions=SLOW=yes,MMP_EXCEPT=5 \
mdtfilesystemtype=ldiskfs mdsfilesystemtype=ldiskfs ostfilesystemtype=ldiskfs \
clientcount=4 osscount=2 mdscount=2 austeroptions=-R failover=true iscsi=1 \
testlist=mmp
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: Ic111360083bd6d8973e47767cb1b291915613727
Reviewed-on: http://review.whamcloud.com/12779
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
James Simmons [Mon, 8 Dec 2014 15:59:59 +0000 (10:59 -0500)]
LU-5986 test: fix conflicting conf-sanity 83 test.
Patches for both LU-4119 and LU-5729 introduced test
83 to conf-sanity. The simple fix is to renumber the
test from LU-4119 to test 84.
Change-Id: Idca7c97daface6768a08f7ef7cbd00b601921a1e
Signed-off-by: James Simmons <uja.ornl@gmail.com>
Reviewed-on: http://review.whamcloud.com/12984
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Jenkins
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Jian Yu [Sat, 6 Dec 2014 07:16:37 +0000 (23:16 -0800)]
LU-5996 tests: check spaces and tabs in .sh files
This patch fixes checkpatch.pl script to check the following
coding style rules in .sh files:
- code indent should use tabs where possible
- no space before tabs
- no spaces at the start of a line
Signed-off-by: Jian Yu <jian.yu@intel.com>
Change-Id: I71821d2d15ca218528fcd8fb37119d0e0798027a
Reviewed-on: http://review.whamcloud.com/12976
Tested-by: Jenkins
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Andreas Dilger [Mon, 8 Dec 2014 17:37:02 +0000 (17:37 +0000)]
Revert "LU-4820 osd: drop memcpy in zfs osd"
This caused review-zfs sanity test_44 to fail in all test cases.
This reverts commit
1249edcd71e6a44f92aba1482201b30696e85d0d.
Change-Id: I972c4c68ee67443c999ce74fda6f6960b0e4b30d
Reviewed-on: http://review.whamcloud.com/12990
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Tested-by: Oleg Drokin <oleg.drokin@intel.com>