The oom_score_adj scale ranges from -1000 to 1000 and represents the
proportion of memory available to the process at allocation time.  This
means an oom_score_adj value of 300, for example, will bias a process as
though it was using an extra 30.0% of available memory and a value of
-350 will discount 35.0% of available memory from its usage.

The oom killer badness heuristic also uses this scale to report the oom
score for each eligible process in determining the "best" process to
kill.  Thus, it can only differentiate each process's memory usage by
0.1% of system RAM.

On large systems, this can end up being a large amount of memory: 256MB
on 256GB systems, for example.

This can be fixed by having the badness heuristic to use the actual
memory usage in scoring threads and then normalizing it to the
oom_score_adj scale for userspace.  This results in better comparison
between eligible threads for kill and no change from the userspace
perspective.
Suggested-by: NKOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Tested-by: NDave Jones <davej@redhat.com>
Signed-off-by: NDavid Rientjes <rientjes@google.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Remove vmtruncate_range(), and remove the truncate_range method from
struct inode_operations: only tmpfs ever supported it, and tmpfs has now
converted over to using the fallocate method of file_operations.

Update Documentation accordingly, adding (setlease and) fallocate lines.
And while we're in mm.h, remove duplicate declarations of shmem_lock() and
shmem_file_setup(): everyone is now using the ones in shmem_fs.h.
Based-on-patch-by: NCong Wang <amwang@redhat.com>
Signed-off-by: NHugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Cong Wang <amwang@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

A missing validation of the value returned by find_vma() could cause a
NULL ptr dereference when walking the pagetable.

This is triggerable from usermode by a simple user by trying to read a
page info out of /proc/pid/pagemap which doesn't exist.

Introduced by commit 025c5b24 ("thp: optimize away unnecessary page
table locking").
Signed-off-by: NSasha Levin <levinsasha928@gmail.com>
Reviewed-by: NNaoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: <stable@vger.kernel.org>		[3.4.x]
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

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

If a file OPEN is denied due to a share lock, the resulting
NFS4ERR_SHARE_DENIED is currently mapped to the default EIO.
This patch adds a more appropriate mapping, and brings Linux
into line with what Solaris 10 does.

See https://bugzilla.kernel.org/show_bug.cgi?id=43286Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@vger.kernel.org

We're already invalidating the data cache, and setting the new change
attribute. Since directories don't care about the i_size field, there
is no need to be forcing any extra revalidation of the page cache.

We do keep the NFS_INO_INVALID_ATTR flag, in order to force an
attribute cache revalidation on stat() calls since we do not
update the mtime and ctime fields.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

The results from a call to nfs4_proc_create_session() should always
be fed into nfs4_handle_reclaim_lease_error, so that we can
handle errors such as NFS4ERR_SEQ_MISORDERED correctly.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Let nfs4_schedule_session_recovery() handle the details of choosing
between resetting the session, and other session related recovery.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

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

Ensure that we handle NFS4ERR_DELAY errors separately, and then
let nfs4_recovery_handle_error() handle all other cases.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

In order to avoid races with other RPC calls that end up setting the
NFS4CLNT_BIND_CONN_TO_SESSION flag.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

This changes the interfaces in <asm/word-at-a-time.h> to be a bit more
complicated, but a lot more generic.

In particular, it allows us to really do the operations efficiently on
both little-endian and big-endian machines, pretty much regardless of
machine details.  For example, if you can rely on a fast population
count instruction on your architecture, this will allow you to make your
optimized <asm/word-at-a-time.h> file with that.

NOTE! The "generic" version in include/asm-generic/word-at-a-time.h is
not truly generic, it actually only works on big-endian.  Why? Because
on little-endian the generic algorithms are wasteful, since you can
inevitably do better. The x86 implementation is an example of that.

(The only truly non-generic part of the asm-generic implementation is
the "find_zero()" function, and you could make a little-endian version
of it.  And if the Kbuild infrastructure allowed us to pick a particular
header file, that would be lovely)

The <asm/word-at-a-time.h> functions are as follows:

 - WORD_AT_A_TIME_CONSTANTS: specific constants that the algorithm
   uses.

 - has_zero(): take a word, and determine if it has a zero byte in it.
   It gets the word, the pointer to the constant pool, and a pointer to
   an intermediate "data" field it can set.

   This is the "quick-and-dirty" zero tester: it's what is run inside
   the hot loops.

 - "prep_zero_mask()": take the word, the data that has_zero() produced,
   and the constant pool, and generate an *exact* mask of which byte had
   the first zero.  This is run directly *outside* the loop, and allows
   the "has_zero()" function to answer the "is there a zero byte"
   question without necessarily getting exactly *which* byte is the
   first one to contain a zero.

   If you do multiple byte lookups concurrently (eg "hash_name()", which
   looks for both NUL and '/' bytes), after you've done the prep_zero_mask()
   phase, the result of those can be or'ed together to get the "either
   or" case.

 - The result from "prep_zero_mask()" can then be fed into "find_zero()"
   (to find the byte offset of the first byte that was zero) or into
   "zero_bytemask()" (to find the bytemask of the bytes preceding the
   zero byte).

   The existence of zero_bytemask() is optional, and is not necessary
   for the normal string routines.  But dentry name hashing needs it, so
   if you enable DENTRY_WORD_AT_A_TIME you need to expose it.

This changes the generic strncpy_from_user() function and the dentry
hashing functions to use these modified word-at-a-time interfaces.  This
gets us back to the optimized state of the x86 strncpy that we lost in
the previous commit when moving over to the generic version.
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

If the EXCHGID4_FLAG_CONFIRMED_R flag is set, the client is in theory
supposed to already know the correct value of the seqid, in which case
RFC5661 states that it should ignore the value returned.

Also ensure that if the sanity check in nfs4_check_cl_exchange_flags
fails, then we must not change the nfs_client fields.

Finally, clean up the code: we don't need to retest the value of
'status' unless it can change.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Ensure that we destroy our lease on last unmount
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Cc: Weston Andros Adamson <dros@netapp.com>

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

For backward compatibility with nfs-utils.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Cc: Weston Andros Adamson <dros@netapp.com>

Apparently the patch "NFS: Always use the same SETCLIENTID boot verifier"
is tickling a Linux nfs server bug, and causing a regression: the server
can get into a situation where it keeps replying NFS4ERR_SEQ_MISORDERED
to our CREATE_SESSION request even when we are sending the correct
sequence ID.

Fix this by purging the lease and then retrying.
Reported-by: NBryan Schumaker <bjschuma@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Otherwise we can end up not sending a new exchange-id/setclientid
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Try to consolidate the error handling for nfs4_reclaim_lease into
a single function instead of doing a bit here, and a bit there...

Also ensure that NFS4CLNT_PURGE_STATE handles errors correctly.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Exchange ID can be called in a lease reclaim situation, so it
will deadlock if it then tries to write out dirty NFS pages.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

The state manager can handle SEQ4_STATUS_CB_PATH_DOWN* flags with a
BIND_CONN_TO_SESSION instead of destroying the session and creating a new one.
Signed-off-by: NWeston Andros Adamson <dros@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

This patch adds the BIND_CONN_TO_SESSION operation which is needed for
upcoming SP4_MACH_CRED work and useful for recovering from broken connections
without destroying the session.
Signed-off-by: NWeston Andros Adamson <dros@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

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

Keep track of the number of bytes read or written via buffered, direct, and
mem-mapped i/o for use by mdsthreshold size_io hints.
Signed-off-by: NAndy Adamson <andros@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

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

We only support one layout type per file system, so one threshold_item4 per
mdsthreshold4.
Signed-off-by: NAndy Adamson <andros@netapp.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

And make sure that everything using it explicitly includes
that header file.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

UDP stack needs a minimum hash size value for proper operation and also
uses alloc_large_system_hash() for proper NUMA distribution of its hash
tables and automatic sizing depending on available system memory.

On some low memory situations, udp_table_init() must ignore the
alloc_large_system_hash() result and reallocs a bigger memory area.

As we cannot easily free old hash table, we leak it and kmemleak can
issue a warning.

This patch adds a low limit parameter to alloc_large_system_hash() to
solve this problem.

We then specify UDP_HTABLE_SIZE_MIN for UDP/UDPLite hash table
allocation.
Reported-by: NMark Asselstine <mark.asselstine@windriver.com>
Reported-by: NTim Bird <tim.bird@am.sony.com>
Signed-off-by: NEric Dumazet <eric.dumazet@gmail.com>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Ensure that a process that uses the nfs_client->cl_cons_state test
for whether the initialisation process is finished does not read
stale data.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Since the struct nfs_client gets added to the global nfs_client_list
before it is initialised, it is possible that rpc_pipefs_event can
end up trying to create idmapper entries on such a thing.

The solution is to have the mount notification wait for the
initialisation of each nfs_client to complete, and then to
skip any entries for which the it failed.
Reported-by: NStanislav Kinsbursky <skinsbursky@parallels.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Acked-by: NStanislav Kinsbursky <skinsbursky@parallels.com>

Session initialisation is not complete until the lease manager
has run. We need to ensure that both nfs4_init_session and
nfs4_init_ds_session do so, and that they check for any resulting
errors in clp->cl_cons_state.

Only after this is done, can nfs4_ds_connect check the contents
of clp->cl_exchange_flags.
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
Cc: Andy Adamson <andros@netapp.com>

Save the server major and minor ID results from EXCHANGE_ID, as they
are needed for detecting server trunking.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

"noresvport" and "discrtry" can be passed to nfs_create_rpc_client()
by setting flags in the passed-in nfs_client.  This change makes it
easy to add new flags.

Note that these settings are now "sticky" over the lifetime of a
struct nfs_client, and may even be copied when an nfs_client is
cloned.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Clean up: Continue to rationalize the locking in nfs_get_client() by
moving the logic that handles the case where a matching server IP
address is not found.

When we support server trunking detection, client initialization may
return a different nfs_client struct than was passed to it.  Change
the synopsis of the init_client methods to return an nfs_client.

The client initialization logic in nfs_get_client() is not much more
than a wrapper around ->init_client.  It's simpler to keep the little
bits of error handling in the version-specific init_client methods.

No behavior change is expected.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Clean up: Code that takes and releases nfs_client_lock remains in
nfs_get_client().  Logic that handles a pre-existing nfs_client is
moved to a separate function.

No behavior change is expected.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Currently our NFS client assigns a unique SETCLIENTID boot verifier
for each server IP address it knows about.  It's set to CURRENT_TIME
when the struct nfs_client for that server IP is created.

During the SETCLIENTID operation, our client also presents an
nfs_client_id4 string to servers, as an identifier on which the server
can hang all of this client's NFSv4 state.  Our client's
nfs_client_id4 string is unique for each server IP address.

An NFSv4 server is obligated to wipe all NFSv4 state associated with
an nfs_client_id4 string when the client presents the same
nfs_client_id4 string along with a changed SETCLIENTID boot verifier.

When our client unmounts the last of a server's shares, it destroys
that server's struct nfs_client.  The next time the client mounts that
NFS server, it creates a fresh struct nfs_client with a fresh boot
verifier.  On seeing the fresh verifer, the server wipes any previous
NFSv4 state associated with that nfs_client_id4.

However, NFSv4.1 clients are supposed to present the same
nfs_client_id4 string to all servers.  And, to support Transparent
State Migration, the same nfs_client_id4 string should be presented
to all NFSv4.0 servers so they recognize that migrated state for this
client belongs with state a server may already have for this client.
(This is known as the Uniform Client String model).

If the nfs_client_id4 string is the same but the boot verifier changes
for each server IP address, SETCLIENTID and EXCHANGE_ID operations
from such a client could unintentionally result in a server wiping a
client's previously obtained lease.

Thus, if our NFS client is going to use a fixed nfs_client_id4 string,
either for NFSv4.0 or NFSv4.1 mounts, our NFS client should use a
boot verifier that does not change depending on server IP address.
Replace our current per-nfs_client boot verifier with a per-nfs_net
boot verifier.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

nfs4_reset_all_state() refreshes the boot verifier a server sees to
trigger that server to wipe this client's state.  This function is
invoked when an NFSv4.1 server reports that it has revoked some or
all of a client's NFSv4 state.

To facilitate server trunking discovery, we will eventually want to
move the cl_boot_time field to a more global structure.  The Uniform
Client String model (and specifically, server trunking detection)
requires that all servers see the same boot verifier until the client
actually does reboot, and not a fresh verifier every time the client
unmounts and remounts the server.

Without the cl_boot_time field, however, nfs4_reset_all_state() will
have to find some other way to force the server to purge the client's
NFSv4 state.

Because these verifiers are opaque (ie, the server doesn't know or
care that they happen to be timestamps), we can force the server
to wipe NFSv4 state by updating the boot verifier as we do now, then
immediately afterwards establish a fresh client ID using the old boot
verifier again.

Hopefully there are no extra paranoid server implementations that keep
track of the client's boot verifiers and prevent clients from reusing
a previous one.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Clean up:  this structure is unused.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>

Clean up: update to use matching types in "if" expressions.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>