With the use of the in-kernel osd library. Implement read/write
of data from/to osd-objects according to information specified
in the objects-layout.

Support for stripping over mirrors with a received stripe_unit.
There are however a few constrains which are not supported:
 1. Stripe Unit must be a multiple of PAGE_SIZE
 2. stripe length (stripe_unit * number_of_stripes) can not be
    bigger then 32bit.

Also support raid-groups and partial-layout. Partial-layout is
when not all the groups are received on the line, addressing
only a partial range of the file.

TODO:
  Only raid0! raid 4/5/6 support will come at later stage

A none supported layout will send IO through the MDS

[Important fallout from the last rebase]
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
[gfp_flags]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Non-rpc layout driver such as for objects and blocks
implement their own I/O path and error handling logic.
Therefore bypass NFS-based error handling for these layout drivers.

[fix lseg ref-count bugs, and null de-refs]
[Fall out from: non-rpc layout drivers]
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
[get rid of PNFS_USE_RPC_CODE]
[get rid of __nfs4_write_done_cb]
[revert useless change in nfs4_write_done_cb]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

allocate and deallocate per-inode private pnfs_layout_hdr
in preparation for I/O implementation.
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

[gfp_flags]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

When a new layout is received in objio_alloc_lseg all device_ids
referenced are retrieved. The device information is queried for from MDS
and then the osd_device is looked-up from the osd-initiator library. The
devices are cached in a per-mount-point list, for later use. At unmount
all devices are "put" back to the library.

objlayout_get_deviceinfo(), objlayout_put_deviceinfo() middleware
API for retrieving device information given a device_id.

TODO: The device cache can get big. Cap its size. Keep an LRU and start
      to return devices which were not used, when list gets to big, or
      when new entries allocation fail.

[pnfs-obj: Bugs in new global-device-cache code]
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
[gfp_flags]
[use global device cache]
[use layout driver in global device cache]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

objlayout_alloc_lseg prepares an xdr_stream and calls the
raid engins objio_alloc_lseg() to allocate a private
pnfs_layout_segment.

objio_osd.c::objio_alloc_lseg() uses passed xdr_stream to
decode and store the layout_segment information in an
objio_segment struct, using the pnfs_osd_xdr.h API for
the actual parsing the layout xdr.

objlayout_free_lseg calls objio_free_lseg() to free the
allocated space.
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
[gfp_flags]
[removed "extern" from function definitions]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

* Add the fs/nfs/objlayout/pnfs_osd_xdr_cli.c file, which will
  include the XDR encode/decode implementations for the pNFS
  client objlayout driver.

[Wrong type in comments]
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

* Define the PNFS_OBJLAYOUT Kconfig option in the nfs
  master Kconfig file.
* Add the objlayout driver to the Kernel's Kbuild system.
* Add the fs/nfs/objlayout/Kbuild file for building the
  objlayoutdriver.ko driver
* Define fs/nfs/objlayout/objio_osd.c, register the driver on module
  initialization and unregister on exit.

[pnfs-obj: remove of CONFIG_PNFS fallout]
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
[added "unsure" clause]
[depend on NFS_V4_1]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

A pNFS client auto-negotiates a lot of features (minorversion level,
pNFS layout type, etc.).  This is convenient, but makes certain kinds of
failures hard for a user to detect.

For example, if the client falls back on 4.0, or falls back to MDS IO
because the user didn't connect to the right iscsi disks before
mounting, the only symptoms may be reduced performance, which may not be
noticed till long after the actual failure, and may be difficult for a
user to diagnose.

However, such "failures" may also be perfectly normal in some cases, so
we don't want to spam the system logs with them.

One approach would be to put some more information into
/proc/self/mountstats.
Signed-off-by: NJ. Bruce Fields <bfields@fieldses.org>
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>
[pnfs: add commit client stats]
[fixup data types for "ret" variables in pnfs_try_to* inline funcs.]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>
[fix definition of show_pnfs for !CONFIG_PNFS]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>
[nfs41: Fix show_sessions in the not CONFIG_NFS_V4_1 case]
    There is a build error when CONFIG_NFS_V4 is set but
    CONFIG_NFS_V4_1 is *not* set. show_sessions() prototype
    was unbalanced between the two cases.
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
[pnfs: super.c remove CONFIG_PNFS]
Signed-off-by: NAndy Adamson <andros@netapp.com>
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Use recalled range to invalidate particular layout segments in the layout cache.
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Add offset and count parameters to pnfs_update_layout and use them to get
the layout in the pageio path.

Order cache layout segments in the following order:
* offset (ascending)
* length (descending)
* iomode (RW before READ)

Test byte range against the layout segment in use in pnfs_{read,write}_pg_test
so not to coalesce pages not using the same layout segment.

[fix lseg ordering]
[clean up pnfs_find_lseg lseg arg]
[remove unnecessary FIXME]
[fix ordering in pnfs_insert_layout]
[clean up pnfs_insert_layout]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Initialize xdr_stream and xdr_buf using an array of page pointers
and length of buffer.
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

pnfs deviceids are unique per server, per layout type.
struct nfs_client is currently used to distinguish deviceids from
different nfs servers, yet these may clash between different layout
types on the same server.  Therefore, use the layout driver associated
with each deviceid at insertion time to look it up, unhash, or
delete it.
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Note: This functionlaity is incomplete as all layout segments referring to
the 'to be removed device id' need to be reaped, and all in flight I/O drained.

[use be32 res in nfs4_callback_devicenotify]
[use nfs_client to qualify deviceid for cb_notify_deviceid]
[use global deviceid cache for CB_NOTIFY_DEVICEID]
[refactor device cache _lookup_deviceid]
[refactor device cache _find_get_deviceid]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>
[Bug in new global-device-cache code]
[layout_driver MUST set free_deviceid_node if using dev-cache]
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Use the pnfs_layoutdriver_type both as a qualifier for the deviceid,
distinguishing deviceid from different layout types on the server,
and for freeing the layout-driver allocated structure containing the
nfs4_deviceid_node.

[BUG in _deviceid_purge_client]
[layout_driver MUST set free_deviceid_node if using dev-cache]
[let ver < 4.1 compile]
Signed-off-by: NBoaz Harrosh <bharrosh@panasas.com>
[removed EXPORT_SYMBOL_GPL(nfs4_deviceid_purge_client)]
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Move deviceid cache from the pnfs files layout driver to the
generic layer in preparation for the objects layout driver.
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Some definitions in the header file depend on nfs_fs.h so pnfs.h can't
be included independently.
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

deviceids are unique per server, per layout type.
Therefore, in the global cache in the files layout driver
deviceids from different servers may clash so we need
to qualify them with a struct nfs_client that represents
the nfs server that returned the deviceid.

Introduced in 2.6.39 commit ea8eecdd
"NFSv4.1 move deviceid cache to filelayout driver"
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Signed-off-by: NJim Rees <rees@umich.edu>
Signed-off-by: NBenny Halevy <bhalevy@panasas.com>

Currently, writebacks may end up recursing back into the filesystem due to
GFP_KERNEL direct reclaims in the pnfs subsystem.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Prevents an infinite loop as list was never emptied.
Signed-off-by: NAndy Adamson <andros@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Free the slot and resend the RPC with new session <slot#,seq#>.

For nfs4_async_handle_error, return -EAGAIN and set the task->tk_status to 0
to restart the async rpc in the rpc_restart_call_prepare state which resets
the slot.

For nfs4_handle_exception, retrying a call that uses nfs4_call_sync will
reset the slot via nfs41_call_sync_prepare.

For open/close/lock/locku/delegreturn/layoutcommit/unlink/rename/write
cachethis is true, so these operations will not trigger an
NFS4ERR_RETRY_UNCACHED_REP.
Signed-off-by: NAndy Adamson <andros@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

On a remount, the VFS layer will clear the MS_SYNCHRONOUS bit on the
assumption that the flags on the mount syscall will have it set if the
remounted fs is supposed to keep it.

In the case of "noac" though, MS_SYNCHRONOUS is implied. A remount of
such a mount will lose the MS_SYNCHRONOUS flag since "sync" isn't part
of the mount options.
Reported-by: NMax Matveev <makc@redhat.com>
Signed-off-by: NJeff Layton <jlayton@redhat.com>
Cc: stable@kernel.org
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

When compiling, I was getting this warning:
fs/nfs/nfs4xdr.c: In function ‘decode_secinfo’:
fs/nfs/nfs4xdr.c:4839:6: warning: variable ‘status’ set but not used
[-Wunused-but-set-variable]

We were unconditionally returning 0 as long as there wasn't an error
coming out of xdr_inline_decode().  We probably want to check the error
status coming out of decode_op_hdr() and decode_secinfo_gss(), rather
than assuming that everything is OK all the time.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

When readdir() returns a directory entry for the root of a mounted
filesystem, Linux follows the old convention of returning the inode
number of the covered directory (despite newer versions of POSIX declaring
that this is a bug).
To ensure this continues to work, the NFSv4 readdir implementation requests
the 'mounted-on-fileid' from the server.

However, readdirplus also needs to instantiate an inode for this entry, and
for that, we also need to request the real fileid as per this patch.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

The following patch ensures that we do not get permanently trapped in
the RPC layer when trying to establish a new client id or session.
This again ensures that the state manager can finish in a timely
fashion when the last filesystem to reference the nfs_client exits.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

If a server for some reason keeps sending NFS4ERR_DELAY errors, we can end
up looping forever inside nfs4_proc_create_session, and so the usual
mechanisms for detecting if the nfs_client is dead don't work.

Fix this by ensuring that we loop inside the nfs4_state_manager thread
instead.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

I only want to try other secflavors during an initial mount if
NFS4ERR_WRONGSEC is returned.  nfs4_handle_exception() could
potentially map other errors to EPERM, so we should handle this
error specially for correctness.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

If we fail to contact the gss upcall program, then no message will
be sent to the server.  The client still updated the sequence number,
however, and this lead to NFS4ERR_SEQ_MISMATCH for the next several
RPC calls.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Currently, the state manager may continue to try recovering state forever
even after the last filesystem to reference that nfs_client has umounted.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@kernel.org

While decoding a secinfo reply, I store the list of supported sec
flavors on a page accessible through res->flavors.  Before reading
each new flavor, I do some math to determine if there is enough
space left on this page, and I break out of my read look if there
isn't.  In order to perform this check correctly, I need to use the
address of res->flavors, rather than the address of res.

When this loop was broken early I lied to the caller and told them
that the entire list had been decoded.  This could lead to problems
if the caller tries to use any the garbage data claiming to be a
valid sec flavor.  I fixed this by using res->flavors->num_flavors
as a counter, incrementing it every time a sec flavor is
successfully decoded.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

We were always attempting sec flavor negotiation, even if the user
told us a specific sec flavor to use.  If that sec flavor fails,
we should return an error rather than continuing with sec flavor
negotiation.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

nfs4_lookup_root() is already configured to use either RPC_AUTH_UNIX
or a user specified flavor (through -o sec=<whatever>).  We should
use this flavor first, and only attempt negotiation if it fails
with -EPERM.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

NFS_MAX_SECFLAVORS should already take into account RPC_AUTH_UNIX
and RPC_AUTH_NULL, so we don't need to set aside extra slots
for them.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

There can be an infinite loop if gss_create_upcall() is called without
the userspace program running.  To prevent this, we return -EACCES if
we notice that pipe_version hasn't changed (indicating that the pipe
has not been opened).
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

mark_inode_dirty_sync() grabs the same inode lock!

race conditions between holding the lock in pnfs_set_layoutcommit() and in
mark_inode_dirty_sync() can result in a second call to pnfs_layoutcommit_inode(), but
this will be a noop as NFS_INO_LAYOUTCOMMIT won't be set in the second call
Signed-off-by: NWeston Andros Adamson <dros@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

The inode was used in an earlier version of the code, but it isn't
used anymore.
Signed-off-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

We only need to call nfs_mark_request_dirty() once in nfs_writepage_setup().
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>