GitHub/LineageOS/G12/android_kernel_amlogic_linux-4.9.git
8 years agostaging: lustre: osc: Automatically increase the max_dirty_mb
Hongchao Zhang [Tue, 16 Aug 2016 20:19:22 +0000 (16:19 -0400)]
staging: lustre: osc: Automatically increase the max_dirty_mb

When RPC size or the max RPCs in flight is increased, the actual
limit might be max_dirty_mb. This patch automatically increases
the max_dirty_mb value at connection time and when the related
values are tuned manually by proc file system.

this patch also changes the unit of "cl_dirty" and "cl_dirty_max"
in client_obd from byte to page.

Signed-off-by: Li Xi <lixi@ddn.com>
Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4933
Reviewed-on: http://review.whamcloud.com/10446
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: build master LMV EA dynamically build via readdir
Fan Yong [Tue, 16 Aug 2016 20:19:21 +0000 (16:19 -0400)]
staging: lustre: lmv: build master LMV EA dynamically build via readdir

When creating a striped directory, the master object saves the slave
objects (or shards) as internal sub-directories. The sub-directory's
name is composed of ${shard_FID}:${shard_idx}. With the name, we can
easily to know what the shard is and where it should be.

On the other hand, we need to store some information related with the
striped directory, such as magic, hash type, shards count, and so on.
That is the LMV EA (header). We do NOT store the FID of each shard in
the LMV EA. Instead, when we need the shards' FIDs (such as readdir()
on client-side), we can build the entrie LMV EA on the MDT (in RAM) by
iterating the sub-directory entries that are contained in the master
object of the striped directroy.

Above mechanism can simplify the striped directory create operation.
For very large striped directory, logging the FIDs array in the LMV
EA will be trouble. It also simplify the LFSCK for verifying striped
directory, because it reduces the inconsistency sources.

Another fixing is about the lmv_master_fid in master LMV EA header,
it is redundant information, and may become one of the inconsistency
sources. So replace it with two __u64 padding fields.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5223
Reviewed-on: http://review.whamcloud.com/10751
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: ptlrpc: request gets stuck in UNREGISTERING phase
Andriy Skulysh [Tue, 16 Aug 2016 20:19:20 +0000 (16:19 -0400)]
staging: lustre: ptlrpc: request gets stuck in UNREGISTERING phase

Exit condition from UNREGISTERING phase is releasing of
both reply and bulk buffers.

Call ptlrpc_unregister_bulk() if ptlrpc_unregister_reply()
wasn't completed in async mode before switching to
UNREGISTERING phase.

Signed-off-by: Andriy Skulysh <Andriy_Skulysh@xyratex.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5259
Xyratex-bug-id: MRP-1960
Reviewed-on: http://review.whamcloud.com/10846
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Ann Koehler <amk@cray.com>
Reviewed-by: Liang Zhen <liang.zhen@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: try all stripes for unknown hash functions
wang di [Tue, 16 Aug 2016 20:19:19 +0000 (16:19 -0400)]
staging: lustre: lmv: try all stripes for unknown hash functions

For unknown hash type, LMV should try all stripes to locate
the name entry. But it will only for lookup and unlink, i.e.
we can only list and unlink entries under striped dir with
unknown hash type.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4921
Reviewed-on: http://review.whamcloud.com/10041
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llog: keep llog ctxt indices constant
Mikhail Pershin [Tue, 16 Aug 2016 20:19:18 +0000 (16:19 -0400)]
staging: lustre: llog: keep llog ctxt indices constant

The llog context id table cannot be shrunk easily because that
will cause index shifting and incompatibility between old client
and new server and vice versa.

Patch moves llog_ctxt_id table to the lustre_idl.h because this is
wire protocol data, these values are added to the wirecheck.

Signed-off-by: Mikhail Pershin <mike.pershin@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5218
Reviewed-on: http://review.whamcloud.com/10758
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: ptlrpc: add OBD_CONNECT_UNLINK_CLOSE flag
Lai Siyao [Tue, 16 Aug 2016 20:19:17 +0000 (16:19 -0400)]
staging: lustre: ptlrpc: add OBD_CONNECT_UNLINK_CLOSE flag

Add OBD_CONNECT_UNLINK_CLOSE flag for interop, once this is supported,
client packs file handle in unlink RPC, and MDT will close file before
unlink.

Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4367
Reviewed-on: http://review.whamcloud.com/10426
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: fid: do open-by-fid by default
Lai Siyao [Tue, 16 Aug 2016 20:19:16 +0000 (16:19 -0400)]
staging: lustre: fid: do open-by-fid by default

Currently client open-by-fid often packs name into the request,
but the name may be invalid, eg. NFS export, and even if it's
valid, it may cause inconsistency because this operation is done
on this fid, which is globally unique, but name not.

Since open-by-fid doesn't pack name, for striped dir we can't know
parent stripe fid on client, so we set parent fid the same as
child fid, and MDT has to find its parent fid from linkea (this is
already supported by MDT).

M_CHECK_STALE becomes obsolete.

Unset MDS_OPEN_FL_INTERNAL from open syscall flags, because these
flags are internally used, and should not be set from user space.

It's not necessary to store parent fid in lli_pfid, because MDT
can get it's parent fid from linkea, and now that DNE stripe
directory stores master inode fid in lli_pfid, stop storing parent
fid to avoid conflict.

Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3544
Reviewed-on: http://review.whamcloud.com/7476
Reviewed-on: http://review.whamcloud.com/10692
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: obd: limit lu_object cache
Brian Behlendorf [Tue, 16 Aug 2016 20:19:15 +0000 (16:19 -0400)]
staging: lustre: obd: limit lu_object cache

As the LU cache grows it can consume large enough chunks of
memory that ends up preventing buffers for other objects,
such as the OIs, from being cached and severely impacting
the performance for FID lookups. Limit the lu_object cache
to a maximum of lu_cache_nr objects.

NOTES:

* In order to be able to quickly determine the number of objects in
  the hash table the CFS_HASH_COUNTER flag is added.  This adds an
  atomic_inc/dec to the hash insert/remove paths but is not expected
  to have any measurable impact of performance.

Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5164
Reviewed-on: http://review.whamcloud.com/10237
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Isaac Huang <he.huang@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: obdclass: compile issues with variable not being initialized
James Simmons [Tue, 16 Aug 2016 20:19:14 +0000 (16:19 -0400)]
staging: lustre: obdclass: compile issues with variable not being initialized

One of the versions of gcc I have refuses to build obd_mount.c due to
index not be initialized in function lmd_make_exclusion before it is
used.

Signed-off-by: James Simmons <uja.ornl@gmail.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4629
Reviewed-on: http://review.whamcloud.com/10705
Reviewed-by: John L. Hammond <john.hammond@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>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: ldlm: improve ldlm_lock_create() return value
Emoly Liu [Tue, 16 Aug 2016 20:19:13 +0000 (16:19 -0400)]
staging: lustre: ldlm: improve ldlm_lock_create() return value

ldlm_lock_create() and ldlm_resource_get() always return NULL as
error reporting and "NULL" is interpretted as ENOMEM incorrectly
sometimes. This patch fixes this problem by using ERR_PTR() rather
than NULL.

Signed-off-by: Emoly Liu <emoly.liu@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4524
Reviewed-on: http://review.whamcloud.com/9004
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: fld: add fld description documentation
Patrick Farrell [Tue, 16 Aug 2016 20:19:12 +0000 (16:19 -0400)]
staging: lustre: fld: add fld description documentation

Add subsystem description from Di Wang to header file.

Signed-off-by: Patrick Farrell <paf@cray.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5153
Reviewed-on: http://review.whamcloud.com/10631
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: mdc: always use D_INFO for debug info when mdc_put_rpc_lock fails
wang di [Tue, 16 Aug 2016 20:19:11 +0000 (16:19 -0400)]
staging: lustre: mdc: always use D_INFO for debug info when mdc_put_rpc_lock fails

Also use D_INFO no matter what the error returned from
mdc_put_rpc_lock.

Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4973
Reviewed-on: http://review.whamcloud.com/10150
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: osc: revise unstable pages accounting
Jinshan Xiong [Tue, 16 Aug 2016 20:19:10 +0000 (16:19 -0400)]
staging: lustre: osc: revise unstable pages accounting

A few changes are made in this patch for unstable pages tracking:

1. Remove kernel NFS unstable pages tracking because it killed
   performance
2. Track unstable pages as part of LRU cache. Otherwise Lustre
   can use much more memory than max_cached_mb
3. Remove obd_unstable_pages tracking to avoid using global
   atomic counter
4. Make unstable pages track optional. Tracking unstable pages is
   turned off by default, and can be controlled by
   llite.*.unstable_stats.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4841
Reviewed-on: http://review.whamcloud.com/10003
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: clio: Reduce memory overhead of per-page allocation
Jinshan Xiong [Tue, 16 Aug 2016 20:19:09 +0000 (16:19 -0400)]
staging: lustre: clio: Reduce memory overhead of per-page allocation

A page in clio used to occupy 584 bytes, which will use size-1024
slab cache. This patch reduces the per-page overhead to 512 bytes
so it can use size-512 instead.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4793
Reviewed-on: http://review.whamcloud.com/10070
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: mdt: add mbo_ prefix to members of struct mdt_body
John L. Hammond [Tue, 16 Aug 2016 20:19:08 +0000 (16:19 -0400)]
staging: lustre: mdt: add mbo_ prefix to members of struct mdt_body

Rename each member of struct mdt_body, adding the prefix mbo_.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2675
Reviewed-on: http://review.whamcloud.com/10202
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: set dir LOV xattr length variable
Hongchao Zhang [Tue, 16 Aug 2016 20:19:07 +0000 (16:19 -0400)]
staging: lustre: llite: set dir LOV xattr length variable

the LOV xattr of directory could be either lov_user_md_v1
(size is 32) or lov_user_md_v3 (size is 48), then the actual
size of the LOV xattr should be return.

Signed-off-by: Hongchao Zhang <hongchao.zhang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5100
Reviewed-on: http://review.whamcloud.com/10453
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: jacques-Charles Lafoucriere <jacques-charles.lafoucriere@cea.fr>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: obd: rename lsr_padding to lsr_valid
Niu Yawei [Tue, 16 Aug 2016 20:19:06 +0000 (16:19 -0400)]
staging: lustre: obd: rename lsr_padding to lsr_valid

Simple variable rename.

Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4345
Reviewed-on: http://review.whamcloud.com/10223
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: use the correct mode for striped directory
wang di [Tue, 16 Aug 2016 20:19:05 +0000 (16:19 -0400)]
staging: lustre: llite: use the correct mode for striped directory

Create striped directory with correct mode, which should be
handling same as mkdir.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4929
Reviewed-on: http://review.whamcloud.com/10028
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: Match MDT where the FID locates first
wang di [Tue, 16 Aug 2016 20:19:04 +0000 (16:19 -0400)]
staging: lustre: lmv: Match MDT where the FID locates first

With DNE every object can have two locks in different namespaces:
lookup lock in space of MDT storing direntry and update/open lock
in space of MDT storing inode. In lmv_find_cbdata/lmv_lock_lock,
it should try the MDT that the FID maps to first, since this can
be easily found, and only try others if that fails.

In the error handler of lmv_add_targets, it should check whether
ld_tgt_count is being increased before ld_tgt_count is being -1.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4098
Reviewed-on: http://review.whamcloud.com/8019
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lov: new pattern flag for partially repaired file
Fan Yong [Tue, 16 Aug 2016 20:19:03 +0000 (16:19 -0400)]
staging: lustre: lov: new pattern flag for partially repaired file

When the layout LFSCK repairs orphan OST-object, if the parent
MDT-object was lost, then it will re-create the MDT-object and
regenerate the LOV EA and fill the target LOV EA slot with the
orphan information, and fill other slots with zero (LOV hole);
if related LOV EA slot is invalid or hole, then it will refill
the target LOV EA slot; if the target slot exceeds current LOV
EA tail, then extend the LOV EA, and fill the gaps as zero.

Some of the LOV EA holes may cannot be re-filled finally because
of lost some OST-objects. And even if they can be re-filled, but
there are still some possible race accessings from client before
the re-filling. If the client access the LOV EA with hole(s), it
may cause some strange behaviour, such as trigger LBUG()/LASSERT()
on the client.

So we will make the client to be aware of the LOV EA is incomplete.
We introduce a new LOV EA pattern flag LOV_PATTERN_F_HOLE for that:
any time when the LFSCK repairs the LOV EA with hole(s), the LOV EA
will be marked as LOV_PATTERN_F_HOLE; when all the holes in the LOV
EA are refilled, the LOV_PATTERN_F_HOLE will be dropped.

For a new client, it recongizes the pattern flag LOV_PATTERN_F_HOLE,
then it can permit/forbid some opertions on the file with LOV holes:

 1) Normal read/write the file with LOV EA hole is permitted, but the
    application will get EIO error when read data from the dummy slot
    or write data to the dummy slot.
 2) The users can dump the recovered data via some common read tools,
    such as "dd conv=sync,noerror".

 3) Append data to the file which has LOV EA hole will get EIO failure.

 4) Other operations will skip the LOV EA hole(s), and will not get
    failures, such as {s,g}etattr, {s,g}getxattr, stat, chown/chgrp,
    chmod, touch, unlink, and so on.

For an old client, since it will not recognize the new pattern flag
LOV_PATTERN_F_HOLE. So the LOV EA with hole will be dicarded with
failure, but it will not cause the client to be crashed.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4675
Reviewed-on: http://review.whamcloud.com/10042
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: validate lock with correct stripe FID
wang di [Tue, 16 Aug 2016 20:19:02 +0000 (16:19 -0400)]
staging: lustre: lmv: validate lock with correct stripe FID

In ll_lookup_it_finish, we need use the real parent(stripe)
FID to validate the parent UPDATE lock.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4925
Reviewed-on: http://review.whamcloud.com/10026
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: a few fixes about readdir of striped dir.
wang di [Tue, 16 Aug 2016 20:19:01 +0000 (16:19 -0400)]
staging: lustre: llite: a few fixes about readdir of striped dir.

Normally we know the value of op_mea1 when ll_readdir is called.
In the case of '.' or '..' op_mea1 is unknown so for that case
fetch the real parents FID.

Signed-off-by: wang di <di.wang@intel.com>
Signed-off-by: Li Xi <lixi@ddn.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4603
Reviewed-on: http://review.whamcloud.com/9191
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Li Xi <pkuelelixi@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: uapi: reduce scope of lustre_idl.h
John L. Hammond [Tue, 16 Aug 2016 20:19:00 +0000 (16:19 -0400)]
staging: lustre: uapi: reduce scope of lustre_idl.h

Move the definition of OBD_OCD_VERSION() and similar macros from
lustre_idl.h to lustre_ver.h. These macros are primarily used in
comparisons to LUSTRE_VERSION_CODE which is defined in lustre_ver.h
and so should be defined there as well. Move a few definitions
(related to FIDs, quota and striping) from lustre_idl.h to
lustre_user.h.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5065
Reviewed-on: http://review.whamcloud.com/10336
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: Change readdir BRW metrics
Jinshan Xiong [Tue, 16 Aug 2016 20:18:59 +0000 (16:18 -0400)]
staging: lustre: llite: Change readdir BRW metrics

To simplify the code, change the metrics from bytes to pages.

Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5034
Reviewed-on: http://review.whamcloud.com/10275
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: Fix the deadlock in balance_dirty_pages()
Jinshan Xiong [Tue, 16 Aug 2016 20:18:58 +0000 (16:18 -0400)]
staging: lustre: llite: Fix the deadlock in balance_dirty_pages()

If the page is already dirtied in ll_write_end() and kernel tries
to call balance_dirty_pages() to write back dirty pages in the same
thread, this is deadlock case if the page is already held by clio.

This can also fix the issue of LU-4873.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4977
Reviewed-on: http://review.whamcloud.com/10149
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: Remove static declaration in anonymous union
Christopher J. Morrone [Tue, 16 Aug 2016 20:18:57 +0000 (16:18 -0400)]
staging: lustre: Remove static declaration in anonymous union

It is not permitted in C++ to have a static declaration inside
of an anonymous union. The g++ compiler will complaine with an
error like this:

 error: struct ost_id::<anonymous union>::ostid invalid; an
 anonymous union can only have non-static data members [-fpermissive]

This patch changes the code to use an unnamed struct in place of
"struct ostid" inside of the anonymous union. That name declaration
was completely unnecessary anyway, since it was not used anywhere else.

Signed-off-by: Christopher J. Morrone <morrone2@llnl.gov>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4987
Reviewed-on: http://review.whamcloud.com/10176
Reviewed-by: Robert Read <robert.read@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: fix inconsistencies of root squash feature
Gregoire Pichon [Tue, 16 Aug 2016 20:18:56 +0000 (16:18 -0400)]
staging: lustre: llite: fix inconsistencies of root squash feature

Root squash exhibits inconsistent behaviour on a client when
enabled. If a file is not cached on the client, then root will get
a permission denied error when accessing the file. When
the file has recently been accessed by a regular user and is
still in cache, root will be able to access the file without error
because the permission check is only done by the client that
isn't aware of root squash.

While the only real security benefit from root squash is to deny
clients access to files owned by root itself, it also makes sense
to treat file access on the client in a consistent manner
regardless of whether the file is in cache or not.

This patch adds root squash settings to llite so that client is able
to apply root squashing when it is relevant.

Configuration of MDT root squash settings will automatically be
applied to llite config log as well.

Update cfs_str2num_check() routine by removing any modification
of the specified string parameter. Since string can come from ls_str
field of a lstr structure, this avoids inconsistent ls_len field.

Signed-off-by: Gregoire Pichon <gregoire.pichon@bull.net>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-1778
Reviewed-on: http://review.whamcloud.com/5700
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: validate names
John L. Hammond [Tue, 16 Aug 2016 20:18:55 +0000 (16:18 -0400)]
staging: lustre: llite: validate names

In ll_prep_md_op_data() validate names according to the same formula
used in mdd_name_check(). Add mdc_pack_name() to validate the name
actually packed in the request.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4992
Reviewed-on: http://review.whamcloud.com/10198
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: separate master object with master stripe
wang di [Tue, 16 Aug 2016 20:18:54 +0000 (16:18 -0400)]
staging: lustre: lmv: separate master object with master stripe

Separate master stripe with master object, so
1. stripeEA only exists on master object.
2. sub-stripe object will be inserted into master object
as sub-directory, and it can get the master object by "..".

By this, it will remove those specilities for stripe0 in
LMV and LOD. And also simplify LFSCK, i.e. consistency check
would be easier.

When then master object becomes an orphan, we should
mark all of its sub-stripes as dead object as well,
otherwise client might still be able to create files
under these stripes.

A few fixes for striped directory layout lock:

 1. stripe 0 should be locked as EX, same as other stripes.
 2. Acquire the layout for directory, when it is being unliked.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4690
Reviewed-on: http://review.whamcloud.com/9511
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: ptlrpc: Early replies need to honor at_max
Chris Horn [Tue, 16 Aug 2016 20:18:53 +0000 (16:18 -0400)]
staging: lustre: ptlrpc: Early replies need to honor at_max

When determining whether an early reply can be sent the server will
calculate the new deadline based on an offset from the request
arrival time. However, when actually setting the new deadline
the server offsets the current time. This can result in deadlines
being extended more than at_max seconds past the request arrival
time. Instead, the server should offset the arrival time when updating
its request timeout.

When a client receives an early reply it doesn't know the server side
arrival time so we use the original sent time as an approximation.

Signed-off-by: Chris Horn <hornc@cray.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4578
Reviewed-on: http://review.whamcloud.com/9100
Reviewed-by: James Simmons <uja.ornl@gmail.com>
Reviewed-by: Christopher J. Morrone <chris.morrone.llnl@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: add error handler in inode prepare phase
wang di [Tue, 16 Aug 2016 20:18:52 +0000 (16:18 -0400)]
staging: lustre: llite: add error handler in inode prepare phase

Add error handler during inode inialization, so inode will
become bad inode if something bad happens during inode prepare
phase, otherwise the striped directory will not get its layout
and being mis-regarded as normal directory.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4930
Reviewed-on: http://review.whamcloud.com/10170
Reviewed-by: Lai Siyao <lai.siyao@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>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: move ioctls to lustre_ioctl.h
John L. Hammond [Tue, 16 Aug 2016 20:18:51 +0000 (16:18 -0400)]
staging: lustre: move ioctls to lustre_ioctl.h

Move ioctl definitions and related functions from lustre_dlm.h,
lustre_lib.h, obd.h, to lustre_ioctl.h. Replace the definitions of
retired ioctls with comment.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4961
Reviewed-on: http://review.whamcloud.com/10139
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Robert Read <robert.read@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: ldlm: flock completion fixes.
Vitaly Fertman [Tue, 16 Aug 2016 20:18:50 +0000 (16:18 -0400)]
staging: lustre: ldlm: flock completion fixes.

Move checks for FAILED, DESTROYED flags under ldlm spinlock,
destroy flock atomically with the check it is not destroyed yet.
Do not put the granted flock into the resource if this is
UNLOCK, TEST, or DEADLOCK'ed flock.

Later a regression for this patch was reported under LU-7626.
The refcount nonzero (1) after lock cleanup errors was reported.
The reason is that the case LCK_NL was not handled for obdecho.
Patch 17791 resolved this issue which has been combined into
this upstream patch.

Signed-off-by: Vitaly Fertman <vitaly_fertman@xyratex.com>
Signed-off-by: Andriy Skulysh <andriy.skulysh@seagate.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2177
Reviewed-by: Alexey Lyashkov <alexey_lyashkov@xyratex.com>
Reviewed-by: Andriy Skulysh <andriy_skulysh@xyratex.com>
Reviewed-by: Vitaly Fertman <vitaly_fertman@xyratex.com>
Xyratex-bug-id: MRP-1588
Reviewed-on: http://review.whamcloud.com/10005
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7626
Reviewed-by: Mirza Arshad Mirza Hussain <arshad.hussain@seagate.com>
Reviewed-by: Alexey Leonidovich Lyashkov <alexey.lyashkov@seagate.com>
Reviewed-on: http://review.whamcloud.com/17791
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: reorder LOV_MAGIC_* definition
Fan Yong [Tue, 16 Aug 2016 20:18:49 +0000 (16:18 -0400)]
staging: lustre: reorder LOV_MAGIC_* definition

Since all the LOV_MAGIC_* definitions have the same
postfix values break that value out into its own
definition. With this we can check whether the magic's
postfix match the LOV_MAGIC_MAGIC or not: if yes,
then it is quite possible that the clients has
encountered an newer LOV magic. This extra information
can let us handle those cases more gracefully.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4941
Reviewed-on: http://review.whamcloud.com/10045
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: obdclass: unified flow control interfaces
Fan Yong [Tue, 16 Aug 2016 20:18:48 +0000 (16:18 -0400)]
staging: lustre: obdclass: unified flow control interfaces

Unify the flow control interfaces for MDC RPC and FLD RPC.
We allow to adjust the maximum inflight RPCs count via /sys
interface.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4687
Reviewed-on: http://review.whamcloud.com/9562
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: rename request to preq in lmv_getattr_name()
John L. Hammond [Tue, 16 Aug 2016 20:18:47 +0000 (16:18 -0400)]
staging: lustre: lmv: rename request to preq in lmv_getattr_name()

Rename request to preq.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4826
Reviewed-on: http://review.whamcloud.com/9863
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: cleanup req in lmv_getattr_name()
John L. Hammond [Tue, 16 Aug 2016 20:18:46 +0000 (16:18 -0400)]
staging: lustre: lmv: cleanup req in lmv_getattr_name()

In lmv_getattr_name() don't return a freed request in the error path.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4826
Reviewed-on: http://review.whamcloud.com/9863
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lov: handle the case of stripe size is not power 2
Jinshan Xiong [Tue, 16 Aug 2016 20:18:45 +0000 (16:18 -0400)]
staging: lustre: lov: handle the case of stripe size is not power 2

Calculate the end of current stripe correctly when the stripe size
is not power 2.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4860
Reviewed-on: http://review.whamcloud.com/9882
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: avoid a deadlock in page write
Jinshan Xiong [Tue, 16 Aug 2016 20:18:44 +0000 (16:18 -0400)]
staging: lustre: llite: avoid a deadlock in page write

For a partial page write, it will have to issue a READ RPC firstly
to get a full uptodate page. If another page is already locked by
this thread it can easily cause deadlock.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4873
Reviewed-on: http://review.whamcloud.com/9928
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: Ensure lmv_intent_lookup cleans up reqp
Nathaniel Clark [Tue, 16 Aug 2016 20:18:43 +0000 (16:18 -0400)]
staging: lustre: lmv: Ensure lmv_intent_lookup cleans up reqp

Ensure there aren't invalid pointers hanging around after
ptlrpc_req_finished is called.

Signed-off-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4826
Reviewed-on: http://review.whamcloud.com/9841
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: lookup remote migrating object in LMV
wang di [Tue, 16 Aug 2016 20:18:42 +0000 (16:18 -0400)]
staging: lustre: lmv: lookup remote migrating object in LMV

If remote object is being found in a migrating directory,
it should continue to lookup the object in remote MDT,
instead of return.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4805
Reviewed-on: http://review.whamcloud.com/9806
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: access lum_stripe_offset as little endian
John L. Hammond [Tue, 16 Aug 2016 20:18:41 +0000 (16:18 -0400)]
staging: lustre: lmv: access lum_stripe_offset as little endian

By the time that a struct lmv_user_md reaches lmv_placement_policy()
it has already been converted to little endian. Therefore use the
appropriate macros around accesses to this this field. This issue was
found by rewriting the definition of struct lmv_user_md to use the
__leXX typedefs and running sparse.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4738
Reviewed-on: http://review.whamcloud.com/9671
Reviewed-by: Dmitry Eremin <dmitry.eremin@intel.com>
Reviewed-by: Swapnil Pimpale <spimpale@ddn.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: simplify inline functions in lustre_fid.h
John L. Hammond [Tue, 16 Aug 2016 20:18:40 +0000 (16:18 -0400)]
staging: lustre: simplify inline functions in lustre_fid.h

Several inline functions return a structure that was passed in.
Their is no need for this so just make these function void.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2675
Reviewed-on: http://review.whamcloud.com/8641
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: use bool for several function in lustre_idl.h/lustre_fid.h
John L. Hammond [Tue, 16 Aug 2016 20:18:39 +0000 (16:18 -0400)]
staging: lustre: use bool for several function in lustre_idl.h/lustre_fid.h

Change the return type of several predicate functions from int to bool.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2675
Reviewed-on: http://review.whamcloud.com/8641
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: const correct FID/OSTID/... helpers
John L. Hammond [Tue, 16 Aug 2016 20:18:38 +0000 (16:18 -0400)]
staging: lustre: const correct FID/OSTID/... helpers

Add a const qualifier wherever possible to the pointer parameters of
the inline helper functions in lustre_idl.h and lustre_fid.h.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2675
Reviewed-on: http://review.whamcloud.com/8641
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: don't need to const __u64 parameters for lustre_idl.h
John L. Hammond [Tue, 16 Aug 2016 20:18:37 +0000 (16:18 -0400)]
staging: lustre: don't need to const __u64 parameters for lustre_idl.h

Remove the const for the __u64 parameters for inline functions
in lustre_idl.h.

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2675
Reviewed-on: http://review.whamcloud.com/8641
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: mdc: fixup MDS_SWAP_LAYOUTS ELC handling
John L. Hammond [Tue, 16 Aug 2016 20:18:36 +0000 (16:18 -0400)]
staging: lustre: mdc: fixup MDS_SWAP_LAYOUTS ELC handling

In mdc_ioc_swap_layouts() cancel *any* unused locks with LAYOUT or
XATTR IBITS set on the two files. (This matches the locks acquired in
mdt_swap_layouts(). Previously only locks that conflicted with a CR
LAYOUT lock were cancelled.)

Signed-off-by: John L. Hammond <john.hammond@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4320
Reviewed-on: http://review.whamcloud.com/9329
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: a few fixes for migration.
wang di [Tue, 16 Aug 2016 20:18:35 +0000 (16:18 -0400)]
staging: lustre: llite: a few fixes for migration.

1. Clear the client dentry cache before migrating file/directory
   to the remote MDT.

2. Do not return stripe information to client, if it did not get
   the layout lock.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4682
Reviewed-on: http://review.whamcloud.com/9522
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: osc: allow to call brw_commit() multiple times
Jinshan Xiong [Tue, 16 Aug 2016 20:18:34 +0000 (16:18 -0400)]
staging: lustre: osc: allow to call brw_commit() multiple times

Sometimes the rq_commit_cb of BRW RPC can be called twice if that RPC
has already committed at reply time. This will cause inaccuracy of
unstable pages accounting and then assertion.

Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3274
Reviewed-on: http://review.whamcloud.com/8215
Reviewed-by: Prakash Surya <surya1@llnl.gov>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: enable clients to inject error for lfsck
Fan Yong [Tue, 16 Aug 2016 20:18:33 +0000 (16:18 -0400)]
staging: lustre: llite: enable clients to inject error for lfsck

This enables the client to inject an error by altering
the parent FID in order to test if the server side file
system checker behaves properly.

Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3951
Reviewed-on: http://review.whamcloud.com/7667
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: libcfs: Only dump log once per sec. to avoid EEXIST
Ryan Haasken [Tue, 16 Aug 2016 20:18:32 +0000 (16:18 -0400)]
staging: lustre: libcfs: Only dump log once per sec. to avoid EEXIST

Since the log file name contains the current time in seconds, dumping
the logs more than once per second causes EEXIST errors to be emitted.
Add a static variable to libcfs_debug_dumplog_internal that records
the time of the last Lustre log dump.  If the current time in seconds
is equal to the last time, do not dump logs again.

Note that this is not thread-safe.  However, in the rare case that two
threads try to access last_dump_time simultaneously, the worst thing
that could happen is that one of the threads will get an EEXIST error
when trying to write the log file.  This is no worse than the current
situation, and it is not likely to happen.

Signed-off-by: Ryan Haasken <haasken@cray.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4129
Reviewed-on: http://review.whamcloud.com/8964
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: fix issue found by Klocwork Insight tool
Dmitry Eremin [Tue, 16 Aug 2016 20:18:31 +0000 (16:18 -0400)]
staging: lustre: lmv: fix issue found by Klocwork Insight tool

'plock.cookie' might be used uninitialized in this function.

sscanf format specification '%d' expects type 'int *' for 'd',
but parameter 3 has a different type '__u32*'

Signed-off-by: Dmitry Eremin <dmitry.eremin@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4629
Reviewed-on: http://review.whamcloud.com/9390
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>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: add ability to migrate inodes.
wang di [Tue, 16 Aug 2016 20:18:30 +0000 (16:18 -0400)]
staging: lustre: add ability to migrate inodes.

Add client support to migrate the individual inodes
from one MDT to another MDT, and this functionality
will only migrate inode layout on MDT but not touch
data object on OST.

The directory will be migrated from top to the bottom,
i.e. migrating parent first, then migrating the child.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-2430
Reviewed-on: http://review.whamcloud.com/6662
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: obdclass: bug fixes for lu_device_type handling
Fan Yong [Tue, 16 Aug 2016 20:18:29 +0000 (16:18 -0400)]
staging: lustre: obdclass: bug fixes for lu_device_type handling

There was no protection when inc/dec lu_device_type::ldt_device_nr,
which may caused the ldt_device_nr to be wrong and trigger assert.
This patch redefine lu_device_type::ldt_device_nr as atomic type.

There was no protection when add/del lu_device_type::ldt_linkage
into/from the global lu_device_types list, which may caused bad
address accessing. This patch uses the existing obd_types_lock
to protect related operations.

We do NOT need lu_types_stop() any longer. Such function scans
the global lu_device_types list, and for each type item on it
which has zerod lu_device_type::ldt_device_nr, call its stop()
method. In fact, the lu_device_type::ldt_device_nr only will be
zero when the last lu_device_fini() is called, and at that time,
inside the lu_device_fini(), its stop() method will be called.
So it is unnecessary to call the stop() again via lu_types_stop().

Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4604
Reviewed-on: http://review.whamcloud.com/8694
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: obdclass: fix lmd_parse() to handle comma-separated NIDs
Jian Yu [Tue, 16 Aug 2016 20:18:28 +0000 (16:18 -0400)]
staging: lustre: obdclass: fix lmd_parse() to handle comma-separated NIDs

This patch handles the  upgrade situation that old mountdata already
contains comma-separated NIDs. The correct way to fix the original
issue is to parse comma-separated NIDs in lmd_parse().

Signed-off-by: Jian Yu <jian.yu@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4460
Reviewed-on: http://review.whamcloud.com/8918
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Bobi Jam <bobijam@gmail.com>
Reviewed-by: Sebastien Buisson <sebastien.buisson@bull.net>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: delete striped directory
wang di [Tue, 16 Aug 2016 20:18:27 +0000 (16:18 -0400)]
staging: lustre: delete striped directory

Add delete striped directory, it includes

1. enable sync log between MDTs, so slave objects will
   be delete by unlink log, which is similar as deleting ost
   object.

2. retrieve layout information of striped directory on MDT,
   then lock all of the slave objects before unlink.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-on: http://review.whamcloud.com/7445
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: fix "getdirstripe" to show stripe info
wang di [Tue, 16 Aug 2016 20:18:26 +0000 (16:18 -0400)]
staging: lustre: llite: fix "getdirstripe" to show stripe info

Fix "lfs getdirstripe", so it can show layout information
of striped directory

[root@testnode tests]# ../utils/lfs getdirstripe /mnt/lustre/test1
/mnt/lustre/test1
lmv_stripe_count: 2
lmv_stripe_offset: 0
mdtidx               FID[seq:oid:ver]
     0               [0x280000400:0x1:0x0]
     1               [0x2c0000400:0x1:0x0]

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-on: http://review.whamcloud.com/7228
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: create striped directory
wang di [Tue, 16 Aug 2016 20:18:25 +0000 (16:18 -0400)]
staging: lustre: create striped directory

1. client send create request to the master MDT, which
  will allocate FIDs and create slaves. for all of slaves.

2. Client needs to revalidate slaves during intent getattr
   and open request.

3. lmv_stripe_md will include attributes(size, nlink etc)
   from all of stripe, which will be protected by UPDATE lock.
   client needs to merge these attributes when update inode.

4. send create request to the MDT where the file is located,
   which can help creating master stripe of striped directory.

Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3529
Reviewed-on: http://review.whamcloud.com/7196
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: replace obd_free_memmd with lmv_free_memmd
wang di [Tue, 16 Aug 2016 20:18:24 +0000 (16:18 -0400)]
staging: lustre: lmv: replace obd_free_memmd with lmv_free_memmd

Use lmv_free_memmd for proper cleanup instead of
the generic obd_free_memmd.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: remove lmv_get_easize
wang di [Tue, 16 Aug 2016 20:18:23 +0000 (16:18 -0400)]
staging: lustre: lmv: remove lmv_get_easize

Completely replace lmv_get_easize with lmv_mds_md_size.
With this change we can delete lmv_get_easize.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: change handling of lmv striping information
wang di [Tue, 16 Aug 2016 20:18:22 +0000 (16:18 -0400)]
staging: lustre: lmv: change handling of lmv striping information

The lmv_[un]pack_md function are used to calculate the
size of the data used to represent the LMV striping data.
The original code was straight forward in its calculate
with lmv_get_easize since only one type of data format
could exist. We want to be able to support different
version of this data in the future so this patch moves
to generating the size of the data using the stripe count
and which LMV_MAGIC_* version.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: remove duplicate MAX_HASH_*
wang di [Tue, 16 Aug 2016 20:18:21 +0000 (16:18 -0400)]
staging: lustre: lmv: remove duplicate MAX_HASH_*

The MAX_HASH_* macros already exist in obd.h. Remove
the duplicated defines in lustre_idl.h.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: lmv: remove unused lmv_get_mea function
wang di [Tue, 16 Aug 2016 20:18:20 +0000 (16:18 -0400)]
staging: lustre: lmv: remove unused lmv_get_mea function

The function lmv_get_mea() is not used so remove it.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: mdc: don't take rpc lock for readdir case
wang di [Tue, 16 Aug 2016 20:18:19 +0000 (16:18 -0400)]
staging: lustre: mdc: don't take rpc lock for readdir case

If the operation is IT_READDIR don't need to handle the
mdc RPC lock.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/10761
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-4906
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: change remove parameter to bool
wang di [Tue, 16 Aug 2016 20:18:18 +0000 (16:18 -0400)]
staging: lustre: llite: change remove parameter to bool

Change the 3rd parameter remove to a bool for ll_release_page
function.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: pass inode to ll_release_page
wang di [Tue, 16 Aug 2016 20:18:17 +0000 (16:18 -0400)]
staging: lustre: llite: pass inode to ll_release_page

Add a inode parameter to ll_release_page. This will be
used in the future.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: style cleanup for llite_internal.h
wang di [Tue, 16 Aug 2016 20:18:16 +0000 (16:18 -0400)]
staging: lustre: llite: style cleanup for llite_internal.h

Group function prototypes together related to dir.c. Move
ll_release_page to be with function declarations.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: remove comment from ll_dir_read
wang di [Tue, 16 Aug 2016 20:18:15 +0000 (16:18 -0400)]
staging: lustre: llite: remove comment from ll_dir_read

Remove comment about fixing swabbing that is not needed.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: lustre: llite: add md_op_data parameter to ll_get_dir_page
wang di [Tue, 16 Aug 2016 20:18:14 +0000 (16:18 -0400)]
staging: lustre: llite: add md_op_data parameter to ll_get_dir_page

Pass in struct md_op_data for ll_get_dir_page function.

Signed-off-by: wang di <di.wang@intel.com>
Reviewed-on: http://review.whamcloud.com/7043
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3531
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: most: aim-network: fix interrupt unsafe spinlocks
Christian Gromm [Thu, 18 Aug 2016 13:28:27 +0000 (15:28 +0200)]
staging: most: aim-network: fix interrupt unsafe spinlocks

The networking AIM does not use the *_irqsave and *_irqrestore flavored
spinlock functions. The rx_completion callback, however, can be called
from an interrupt context.

This patch is needed to fix this problem.

Signed-off-by: Andrey Shvetsov <andrey.shvetsov@k2l.de>
Signed-off-by: Christian Gromm <christian.gromm@microchip.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoStaging: android: ion: ion_test.c: fix parenthesis alignment
Ben LeMasurier [Tue, 16 Aug 2016 17:01:44 +0000 (11:01 -0600)]
Staging: android: ion: ion_test.c: fix parenthesis alignment

This fixes the checkpatch.pl "Alignment should match open parenthesis"
issues in ion_test.c.

Signed-off-by: Ben LeMasurier <ben@crypt.ly>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agoStaging: android: ion: ion_heap.c: fix parenthesis alignment
Ben LeMasurier [Wed, 17 Aug 2016 03:02:00 +0000 (21:02 -0600)]
Staging: android: ion: ion_heap.c: fix parenthesis alignment

This fixes the checkpatch.pl "Alignment should match open parenthesis"
issues in ion_heap.c.

Signed-off-by: Ben LeMasurier <ben@crypt.ly>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: most: dim2-hdm: introduce function dim2_transfer_madr
Christian Gromm [Thu, 18 Aug 2016 14:58:47 +0000 (16:58 +0200)]
staging: most: dim2-hdm: introduce function dim2_transfer_madr

This patch removes duplicated code by putting it into the new function
dim2_transfer_madr.

Signed-off-by: Andrey Shvetsov <andrey.shvetsov@k2l.de>
Signed-off-by: Christian Gromm <christian.gromm@microchip.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: most: dim2-hdm: initialize DBR memory of synchronous channels
Christian Gromm [Thu, 18 Aug 2016 14:58:46 +0000 (16:58 +0200)]
staging: most: dim2-hdm: initialize DBR memory of synchronous channels

In the time after a synchronous DIM2 channel is enabled and before the
first buffer travels this channel, the DIM2 macro is babbling random
random data to the network.

This patch initializes the internal DBR memory of the DIM2 that is used
to transfer the synchronous data.

Signed-off-by: Andrey Shvetsov <andrey.shvetsov@k2l.de>
Signed-off-by: Christian Gromm <christian.gromm@microchip.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorinput: make lock_visor_dev a mutex
Tim Sell [Sat, 11 Jun 2016 01:48:26 +0000 (21:48 -0400)]
staging: unisys: visorinput: make lock_visor_dev a mutex

Since lock_visor_dev is a binary semaphore it makes more sense to
use a mutex lock.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorinput: ensure proper locking wrt creation & ints
Tim Sell [Sat, 11 Jun 2016 01:48:25 +0000 (21:48 -0400)]
staging: unisys: visorinput: ensure proper locking wrt creation & ints

Ensure we properly lock between visorinput_channel_interrupt(),
visorinput_open(), and devdata_create().  We now guarantee that:

* interrupts will be disabled and remain disabled during device creation,
  by setting 'paused = true' across device creation

* we canNOT get into visorinput_open() until the device structure is
  totally initialized, by delaying the input_register_device() until the
  end of device initialization

We also now ensure that lock_visor_dev is held across updates of devdata
state, to ensure state consistency.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: Move vbushelper.h to visorbus directory
David Kershner [Sat, 11 Jun 2016 01:48:24 +0000 (21:48 -0400)]
staging: unisys: Move vbushelper.h to visorbus directory

Only visorbus needs this header file so move it to visorbus
directory.

Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: fix visorbus_private.h comments
Tim Sell [Sat, 11 Jun 2016 01:48:23 +0000 (21:48 -0400)]
staging: unisys: visorbus: fix visorbus_private.h comments

This patch ONLY touches comment lines, i.e., NO executable code is
affected.

Comments were fixed in visorbus_private.h:
* Minor typos were corrected.
* Useless comments were removed.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: Rename function to follow existing convention
David Binder [Sat, 11 Jun 2016 01:48:22 +0000 (21:48 -0400)]
staging: unisys: visorbus: Rename function to follow existing convention

Renames visorchipset_device_pause_response to device_pause_response,
thereby following the convention that other responder functions follow.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: Remove notifier-related code from visorbus
David Binder [Sat, 11 Jun 2016 01:48:21 +0000 (21:48 -0400)]
staging: unisys: visorbus: Remove notifier-related code from visorbus

When this functionality was first implemented, visorchipset and visorbus
were separate drivers, which necessitated a registration mechanism for
them to communicate.  More-recently, visorchipset and visorbus were
combined into a single driver, and now exist as separate source files
within the same driver, known as 'visorbus'.  This eliminated the need
for a registration mechanism, but it has remained nevertheless until now.
For the sake of simplification, this registration mechanism is now being
removed.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <timothy.sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: rectify kerneldoc comment for struct
David Binder [Sat, 11 Jun 2016 01:48:20 +0000 (21:48 -0400)]
staging: unisys: visorbus: rectify kerneldoc comment for struct

Fixes the kerneldoc comment for struct visor_device - the struct members
were not listed with the appropriate @ prefix.

Signed-off-by: David Binder <david.binder@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: Move visorbus-unique functions to private header
David Binder [Sat, 11 Jun 2016 01:48:19 +0000 (21:48 -0400)]
staging: unisys: visorbus: Move visorbus-unique functions to private header

Moves function prototypes that are unique to visorbus from
include/visorbus.h to visorbus/visorbus_private.h.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: Rectify commenting in visorchipset.c
David Binder [Sat, 11 Jun 2016 01:48:18 +0000 (21:48 -0400)]
staging: unisys: visorbus: Rectify commenting in visorchipset.c

Adds kerneldoc formatting to appropriate functions. Other multi-line
comments now use proper formatting.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: fix visorchannel.c comments
David Binder [Sat, 11 Jun 2016 01:48:17 +0000 (21:48 -0400)]
staging: unisys: visorbus: fix visorchannel.c comments

This patch ONLY touches comment lines, i.e., NO executable code is
affected.

Comments were fixed in visorchannel.c:
* All functions worthy of documenting now use standard kerneldoc
  formatting.
* Multi-line comments were tweaked so as to use appropriate conventions.
* Minor typos were corrected.
* Useless comments were removed.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Signed-off-by: David Binder <david.binder@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: fix commenting in visorbus_main.c
David Binder [Sat, 11 Jun 2016 01:48:16 +0000 (21:48 -0400)]
staging: unisys: visorbus: fix commenting in visorbus_main.c

This patch ONLY touches comment lines, i.e., NO executable code is
affected.

* All functions worthy of documenting now use standard kerneldoc
  formatting.
* Improper uses of kerneldoc formatting were converted to standard
  multi-line comments.
* Multi-line comments were tweaked so as to use appropriate conventions.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: fix commenting in vbusdevinfo.h
David Binder [Sat, 11 Jun 2016 01:48:15 +0000 (21:48 -0400)]
staging: unisys: visorbus: fix commenting in vbusdevinfo.h

This patch ONLY touches comment lines, i.e., NO executable code is
affected.

* All functions worthy of documenting now use standard kerneldoc
  formatting.
* Multi-line comments were tweaked so as to use appropriate conventions.
* Minor typos were corrected.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: remove unused parameter from function
Tim Sell [Sat, 11 Jun 2016 01:48:14 +0000 (21:48 -0400)]
staging: unisys: visorbus: remove unused parameter from function

The off parameter to visorchannel_create_guts() was never used, so it was
removed.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: Remove reference to unused STANDALONE_CLIENT
Bryan Thompson [Sat, 11 Jun 2016 01:48:13 +0000 (21:48 -0400)]
staging: unisys: Remove reference to unused STANDALONE_CLIENT

The STANDALONE_CLIENT define is no longer used by Unisys driver code.

Signed-off-by: Bryan Thompson <bryan.thompson@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: Remove unused functions
Bryan Thompson [Sat, 11 Jun 2016 01:48:12 +0000 (21:48 -0400)]
staging: unisys: visorbus: Remove unused functions

Remove visorbus_clear_channel, visorchannel_signalqueue_slots_avail,
visorchannel_signalqueue_max_slots, visorchannel_clear, and
visorchannel_debug which are no longer called by any driver.

Signed-off-by: Bryan Thompson <bryan.thompson@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: Remove unnecessary EXPORT_SYMBOL statements
Bryan Thompson [Sat, 11 Jun 2016 01:48:11 +0000 (21:48 -0400)]
staging: unisys: visorbus: Remove unnecessary EXPORT_SYMBOL statements

The driver that is now visorbus started out as multiple separate drivers,
and when they were merged the EXPORT_SYMBOL statements that were required
for separate drivers were left in the code. This patch removes those now
unnecessary exports.

Signed-off-by: Bryan Thompson <bryan.thompson@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: Make visordriver_callback_lock a mutex
Bryan Thompson [Sat, 11 Jun 2016 01:48:10 +0000 (21:48 -0400)]
staging: unisys: visorbus: Make visordriver_callback_lock a mutex

visordriver_callback_lock is just a binary semaphore that logically
makes more sense as a mutex.

Signed-off-by: Bryan Thompson <bryan.thompson@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: remove periodic_work.h/.c
Tim Sell [Sat, 11 Jun 2016 01:48:09 +0000 (21:48 -0400)]
staging: unisys: visorbus: remove periodic_work.h/.c

These files were made no-longer-necessary by recent commits.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: use kernel timer instead of workqueue
Tim Sell [Sat, 11 Jun 2016 01:48:08 +0000 (21:48 -0400)]
staging: unisys: visorbus: use kernel timer instead of workqueue

A kernel timer is now used as the vehicle to periodically call the
channel_interrupt function of registered visor drivers, instead of a
workqueue.

This simplifies a lot of things by making periodic_work.c and
periodic_work.h no longer necessary.  This change also means that the
channel_interrupt() callbacks registered by visor drivers (via
visorbus_register_visor_driver()) will now be called in atomic context
(i.e., canNOT sleep) rather than kernel thread context (CAN sleep).
Fortunately this did NOT necessitate any change to the existing
channel_interrupt() callbacks, because none of them ever perform any
operations that would be invalid in atomic context.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorinput: remove unnecessary locking
Tim Sell [Sat, 11 Jun 2016 01:48:07 +0000 (21:48 -0400)]
staging: unisys: visorinput: remove unnecessary locking

Locking in the _interrupt() function is NOT necessary so long as we ensure
that interrupts have been stopped whenever we need to pause or resume the
device, which we now do.

While a device is paused, we ensure that interrupts stay disabled, i.e.
that the _interrupt() function will NOT be called, yet remember the desired
state in devdata->interrupts_enabled if open() or close() are called are
called while the device is paused.  Then when the device is resumed, we
restore the actual state of interrupts (i.e., whether _interrupt() is going
to be called or not) to the desired state in devdata->interrupts_enabled.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: removed unused periodic_test_workqueue
Tim Sell [Sat, 11 Jun 2016 01:48:06 +0000 (21:48 -0400)]
staging: unisys: visorbus: removed unused periodic_test_workqueue

periodic_test_workqueue was an unused relic from the past, and was removed.

Signed-off-by: Tim Sell <Timothy.Sell@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: include: Remove thread-related enum members
David Binder [Sat, 11 Jun 2016 01:48:05 +0000 (21:48 -0400)]
staging: unisys: include: Remove thread-related enum members

Code relating to ktheads was previously removed from s-Par driver code.
This patch cleans up lingering remnants of kthreads by removing thread-
related enum types.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visornic: Correct comment spelling mistake
David Binder [Sat, 11 Jun 2016 01:48:04 +0000 (21:48 -0400)]
staging: unisys: visornic: Correct comment spelling mistake

Fixes a comment spelling mistake in visornic.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: modify format string to match argument
David Binder [Sat, 11 Jun 2016 01:48:03 +0000 (21:48 -0400)]
staging: unisys: visorbus: modify format string to match argument

Modifies the format string of snprintf to expect an unsigned int
instead of a signed one, per the supplied argument.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: remove unused struct
David Binder [Sat, 11 Jun 2016 01:48:02 +0000 (21:48 -0400)]
staging: unisys: visorbus: remove unused struct

Removes unused struct definition, channel_size_info, in response to
findings by SonarQube.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
8 years agostaging: unisys: visorbus: remove unused module parameters
David Binder [Sat, 11 Jun 2016 01:48:01 +0000 (21:48 -0400)]
staging: unisys: visorbus: remove unused module parameters

Removes unused module parameters from visorbus_main.c, in response to
findings by SonarQube.

Signed-off-by: David Binder <david.binder@unisys.com>
Signed-off-by: David Kershner <david.kershner@unisys.com>
Reviewed-by: Tim Sell <Timothy.Sell@unisys.com>
Acked-By: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>