Due to userspace assumptions, the sched_scan_stop operation must
be synchronous, i.e. once it returns a new scheduled scan must be
able to start immediately. Document this in the API.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

While framing the TDLS Setup Confirmation frame, the driver needs to
know if the TDLS peer is VHT/HT/WMM capable and thus shall construct
the VHT/HT operation / WMM parameter elements accordingly. Supplicant
determines if the TDLS peer is VHT/HT/WMM capable based on the
presence of the respective IEs in the received TDLS Setup Response frame.

The host driver should not need to parse the received TDLS Response
frame and thus, should be able to rely on the supplicant to indicate
the capability of the peer through additional flags while transmitting
the TDLS Setup Confirmation frame through tdls_mgmt operations.
Signed-off-by: NSunil Dutt Undekari <usdutt@qti.qualcomm.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

There's no way the driver can pre-build the radiotap header,
so remove the comment stating that it can.
Reported-by: NJouni Malinen <j@w1.fi>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

NAPI was originally added to mac80211 a long time ago (by John in
commit 4e6cbfd0 in July 2010), but then removed years later
(by Stanislaw in commit 30c97120 in February 2013). No driver
ever used it, so that was fine.

Now I'm adding support for NAPI to our driver, so add some code
to mac80211 again  to support NAPI. John was originally wrapping
some (but not nearly all NAPI-related functions), but that doesn't
scale very well with the number of functions that are there, some
of which are even only inlines. Thus, instead of doing that, let
the drivers manage the NAPI struct, except for napi_add() which is
needed so mac80211 knows how to call napi_gro_receive().

Also remove some no longer needed definitions that were left when
NAPI support was removed.
Reviewed-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Reviewed-by: NEyal Shapira <eyal@wizery.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

There's no driver using this flag and consequently no userspace
application is actually looking at it. As it seems unlikely for
any driver to start using it, remove it and the (very little)
code that used it.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

In case of beacon_loss with IEEE80211_HW_CONNECTION_MONITOR
device, mac80211 probes the ap (and disconnects on timeout)
but ignores the ack.

If we already got an ack, there's no reason to continue
disconnecting. this can help devices that supports
IEEE80211_HW_CONNECTION_MONITOR only partially (e.g. take
care of keep alives, but does not probe the ap.

In case the device wants to disconnect without probing,
it can just call ieee80211_connection_loss.
Signed-off-by: NEliad Peller <eliadx.peller@intel.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

It actually goes back to 2004 ([PATCH] Concurrent O_SYNC write support)
when sync_page_range() had been introduced; generic_file_write{,v}() correctly
synced
	pos_after_write - written .. pos_after_write - 1
but generic_file_aio_write() synced
	pos_before_write .. pos_before_write + written - 1
instead.  Which is not the same thing with O_APPEND, obviously.
A couple of years later correct variant had been killed off when
everything switched to use of generic_file_aio_write().

All users of generic_file_aio_write() are affected, and the same bug
has been copied into other instances of ->aio_write().

The fix is trivial; the only subtle point is that generic_write_sync()
ought to be inlined to avoid calculations useless for the majority of
calls.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This is a patch to improve swap readahead algorithm.  It's from Hugh and
I slightly changed it.

Hugh's original changelog:

swapin readahead does a blind readahead, whether or not the swapin is
sequential.  This may be ok on harddisk, because large reads have
relatively small costs, and if the readahead pages are unneeded they can
be reclaimed easily - though, what if their allocation forced reclaim of
useful pages? But on SSD devices large reads are more expensive than
small ones: if the readahead pages are unneeded, reading them in caused
significant overhead.

This patch adds very simplistic random read detection.  Stealing the
PageReadahead technique from Konstantin Khlebnikov's patch, avoiding the
vma/anon_vma sophistications of Shaohua Li's patch, swapin_nr_pages()
simply looks at readahead's current success rate, and narrows or widens
its readahead window accordingly.  There is little science to its
heuristic: it's about as stupid as can be whilst remaining effective.

The table below shows elapsed times (in centiseconds) when running a
single repetitive swapping load across a 1000MB mapping in 900MB ram
with 1GB swap (the harddisk tests had taken painfully too long when I
used mem=500M, but SSD shows similar results for that).

Vanilla is the 3.6-rc7 kernel on which I started; Shaohua denotes his
Sep 3 patch in mmotm and linux-next; HughOld denotes my Oct 1 patch
which Shaohua showed to be defective; HughNew this Nov 14 patch, with
page_cluster as usual at default of 3 (8-page reads); HughPC4 this same
patch with page_cluster 4 (16-page reads); HughPC0 with page_cluster 0
(1-page reads: no readahead).

HDD for swapping to harddisk, SSD for swapping to VertexII SSD.  Seq for
sequential access to the mapping, cycling five times around; Rand for
the same number of random touches.  Anon for a MAP_PRIVATE anon mapping;
Shmem for a MAP_SHARED anon mapping, equivalent to tmpfs.

One weakness of Shaohua's vma/anon_vma approach was that it did not
optimize Shmem: seen below.  Konstantin's approach was perhaps mistuned,
50% slower on Seq: did not compete and is not shown below.

HDD        Vanilla Shaohua HughOld HughNew HughPC4 HughPC0
Seq Anon     73921   76210   75611   76904   78191  121542
Seq Shmem    73601   73176   73855   72947   74543  118322
Rand Anon   895392  831243  871569  845197  846496  841680
Rand Shmem 1058375 1053486  827935  764955  764376  756489

SSD        Vanilla Shaohua HughOld HughNew HughPC4 HughPC0
Seq Anon     24634   24198   24673   25107   21614   70018
Seq Shmem    24959   24932   25052   25703   22030   69678
Rand Anon    43014   26146   28075   25989   26935   25901
Rand Shmem   45349   45215   28249   24268   24138   24332

These tests are, of course, two extremes of a very simple case: under
heavier mixed loads I've not yet observed any consistent improvement or
degradation, and wider testing would be welcome.

Shaohua Li:

Test shows Vanilla is slightly better in sequential workload than Hugh's
patch.  I observed with Hugh's patch sometimes the readahead size is
shrinked too fast (from 8 to 1 immediately) in sequential workload if
there is no hit.  And in such case, continuing doing readahead is good
actually.

I don't prepare a sophisticated algorithm for the sequential workload
because so far we can't guarantee sequential accessed pages are swap out
sequentially.  So I slightly change Hugh's heuristic - don't shrink
readahead size too fast.

Here is my test result (unit second, 3 runs average):
	Vanilla		Hugh		New
Seq	356		370		360
Random	4525		2447		2444

Attached graph is the swapin/swapout throughput I collected with 'vmstat
2'.  The first part is running a random workload (till around 1200 of
the x-axis) and the second part is running a sequential workload.
swapin and swapout throughput are almost identical in steady state in
both workloads.  These are expected behavior.  while in Vanilla, swapin
is much bigger than swapout especially in random workload (because wrong
readahead).

Original patches by: Shaohua Li and Konstantin Khlebnikov.

[fengguang.wu@intel.com: swapin_nr_pages() can be static]
Signed-off-by: NHugh Dickins <hughd@google.com>
Signed-off-by: NShaohua Li <shli@fusionio.com>
Signed-off-by: NFengguang Wu <fengguang.wu@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This capabilities weren't propagated to the radiotap header.
We don't set here the VHT_KNOWN / MCS_HAVE flag because not
all the low level drivers will know how to properly flag
the frames, hence the low level driver will be in charge
of setting IEEE80211_RADIOTAP_MCS_HAVE_FEC,
IEEE80211_RADIOTAP_MCS_HAVE_STBC and / or
IEEE80211_RADIOTAP_VHT_KNOWN_STBC according to its
capabilities.
Signed-off-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

ieee80211_rx_status.flags is full. Define a new vht_flag
variable to be able to set more VHT related flags and make
room in flags.
Signed-off-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Acked-by: Kalle Valo <kvalo@qca.qualcomm.com> [ath10k]
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

The purpose of this housekeeping is to make some room for
VHT flags. The radiotap vendor fields weren't in use.
Signed-off-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

This changes 'do_execve()' to get the executable name as a 'struct
filename', and to free it when it is done.  This is what the normal
users want, and it simplifies and streamlines their error handling.

The controlled lifetime of the executable name also fixes a
use-after-free problem with the trace_sched_process_exec tracepoint: the
lifetime of the passed-in string for kernel users was not at all
obvious, and the user-mode helper code used UMH_WAIT_EXEC to serialize
the pathname allocation lifetime with the execve() having finished,
which in turn meant that the trace point that happened after
mm_release() of the old process VM ended up using already free'd memory.

To solve the kernel string lifetime issue, this simply introduces
"getname_kernel()" that works like the normal user-space getname()
function, except with the source coming from kernel memory.

As Oleg points out, this also means that we could drop the tcomm[] array
from 'struct linux_binprm', since the pathname lifetime now covers
setup_new_exec().  That would be a separate cleanup.
Reported-by: NIgor Zhbanov <i.zhbanov@samsung.com>
Tested-by: NSteven Rostedt <rostedt@goodmis.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The element ID list is currently almost sorted by amendment
or similar topic, but the order is difficult to maintain and
not very transparent. Sort the list by ID instead, and add
a lot of missing IDs.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

In case we will get regulatory request with rule
where max_bandwidth_khz is set to 0 handle this
case as a special one.

If max_bandwidth_khz == 0 we should calculate maximum
available bandwidth base on all frequency contiguous rules.
In case we need auto calculation we just have to set:

country PL: DFS-ETSI
        (2402 - 2482 @ 40), (N/A, 20)
        (5170 - 5250 @ AUTO), (N/A, 20)
        (5250 - 5330 @ AUTO), (N/A, 20), DFS
        (5490 - 5710 @ 80), (N/A, 27), DFS

This mean we will calculate maximum bw for rules where
AUTO (N/A) were set, 160MHz (5330 - 5170) in example above.
So we will get:
        (5170 - 5250 @ 160), (N/A, 20)
        (5250 - 5330 @ 160), (N/A, 20), DFS

In other case:
country FR: DFS-ETSI
        (2402 - 2482 @ 40), (N/A, 20)
        (5170 - 5250 @ AUTO), (N/A, 20)
        (5250 - 5330 @ 80), (N/A, 20), DFS
        (5490 - 5710 @ 80), (N/A, 27), DFS

We will get 80MHz (5250 - 5170):
        (5170 - 5250 @ 80), (N/A, 20)
        (5250 - 5330 @ 80), (N/A, 20), DFS

Base on this calculations we will set correct channel
bandwidth flags (eg. IEEE80211_CHAN_NO_80MHZ).

We don't need any changes in CRDA or internal regulatory.
Signed-off-by: NJanusz Dziedzic <janusz.dziedzic@tieto.com>
[extend nl80211 description a bit, fix typo]
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

It was possible to break interface combinations in
the following way:

 combo 1: iftype = AP, num_ifaces = 2, num_chans = 2,
 combo 2: iftype = AP, num_ifaces = 1, num_chans = 1, radar = HT20

With the above interface combinations it was
possible to:

 step 1. start AP on DFS channel by matching combo 2
 step 2. start AP on non-DFS channel by matching combo 1

This was possible beacuse (step 2) did not consider
if other interfaces require radar detection.

The patch changes how cfg80211 tracks channels -
instead of channel itself now a complete chandef
is stored.
Signed-off-by: NMichal Kazior <michal.kazior@tieto.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

When receiving an IBSS_JOINED event select the BSS object
based on the {bssid, channel} couple rather than the bssid
only.
With the current approach if another cell having the same
BSSID (but using a different channel) exists then cfg80211
picks up the wrong BSS object.
The result is a mismatching channel configuration between
cfg80211 and the driver, that can lead to any sort of
problem.

The issue can be triggered by having an IBSS sitting on
given channel and then asking the driver to create a new
cell using the same BSSID but with a different frequency.
By passing the channel to cfg80211_get_bss() we can solve
this ambiguity and retrieve/create the correct BSS object.
All the users of cfg80211_ibss_joined() have been changed
accordingly.

Moreover WARN when cfg80211_ibss_joined() gets a NULL
channel as argument and remove a bogus call of the same
function in ath6kl (it does not make sense to call
cfg80211_ibss_joined() with a zero BSSID on ibss-leave).

Cc: Kalle Valo <kvalo@qca.qualcomm.com>
Cc: Arend van Spriel <arend@broadcom.com>
Cc: Bing Zhao <bzhao@marvell.com>
Cc: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Cc: libertas-dev@lists.infradead.org
Acked-by: NKalle Valo <kvalo@qca.qualcomm.com>
Signed-off-by: NAntonio Quartulli <antonio@open-mesh.com>
[minor code cleanup in ath6kl]
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

Action, disassoc and deauth frames are bufferable, and as such don't
have the PM bit in the frame control field reserved which means we
need to react to the bit when receiving in such a frame.

Fix this by introducing a new helper ieee80211_is_bufferable_mmpdu()
and using it for the RX path that currently ignores the PM bit in
any non-data frames for doze->wake transitions, but listens to it in
all frames for wake->doze transitions, both of which are wrong.

Also use the new helper in the TX path to clean up the code.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

The scheduled scan matchsets were intended to be a list of filters,
with the found BSS having to pass at least one of them to be passed
to the host. When the RSSI attribute was added, however, this was
broken and currently wpa_supplicant adds that attribute in its own
matchset; however, it doesn't intend that to mean that anything
that passes the RSSI filter should be passed to the host, instead
it wants it to mean that everything needs to also have higher RSSI.

This is semantically problematic because we have a list of filters
like [ SSID1, SSID2, SSID3, RSSI ] with no real indication which
one should be OR'ed and which one AND'ed.

To fix this, move the RSSI filter attribute into each matchset. As
we need to stay backward compatible, treat a matchset with only the
RSSI attribute as a "default RSSI filter" for all other matchsets,
but only if there are other matchsets (an RSSI-only matchset by
itself is still desirable.)

To make driver implementation easier, keep a global min_rssi_thold
for the entire request as well. The only affected driver is ath6kl.

I found this when I looked into the code after Raja Mani submitted
a patch fixing the n_match_sets calculation to disregard the RSSI,
but that patch didn't address the semantic issue.
Reported-by: NRaja Mani <rmani@qti.qualcomm.com>
Acked-by: NLuciano Coelho <luciano.coelho@intel.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

A few places weren't checking that the frame passed to the
function actually has enough data even though the function
clearly documents it must have a payload byte. Make this
safer by changing the function to take an skb and checking
the length inside. The old version is preserved for now as
the rtl* drivers use it and don't have a correct skb.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

There's not a single rate control algorithm actually in
a separate module where the module refcount would be
required. Similarly, there's no specific rate control
module.

Therefore, all the module handling code in rate control
is really just dead code, so remove it.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

Change the code to allow making all the rate control ops
const, nothing ever needs to change them. Also change all
drivers to make use of this and mark the ops const.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

Allow to force SGI, LGI.
Mainly for test purpose.
Signed-off-by: NJanusz Dziedzic <janusz.dziedzic@tieto.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

This required liberally sprinkling 'const' over brcmfmac
and mwifiex but seems like a useful thing to do since the
pointer can't really be written.
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

Addition of the frequency hints showed up couple of places in cfg80211
where pointers could be marked const and a shared function could be used
to fetch a valid channel.
Signed-off-by: NJouni Malinen <j@w1.fi>
[fix mwifiex]
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

This allows drivers to advertise the maximum number of associated
stations they support in AP mode (including P2P GO). User space
applications can use this for cleaner way of handling the limit (e.g.,
hostapd rejecting IEEE 802.11 authentication without manual
configuration of the limit) or to figure out what type of use cases can
be executed with multiple devices before trying and failing.
Signed-off-by: NJouni Malinen <j@w1.fi>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

This clarifies the expected driver behavior on the older
NL80211_ATTR_MAC and NL80211_ATTR_WIPHY_FREQ attributes and adds a new
set of similar attributes with _HINT postfix to enable use of a
recommendation of the initial BSS to choose. This can be helpful for
some drivers that can avoid an additional full scan on connection
request if the information is provided to them (user space tools like
wpa_supplicant already has that information available based on earlier
scans).

In addition, this can be used to get more expected behavior for cases
where a specific BSS should be picked first based on operations like
Interworking network selection or WPS. These cases were already easily
addressed with drivers that leave BSS selection to user space, but there
was no convenient way to do this with drivers that take care of BSS
selection internally without using the NL80211_ATTR_MAC which is not
really desired since it is needed for other purposes to force the
association to remain with the same BSS.
Signed-off-by: NJouni Malinen <j@w1.fi>
[add const, fix policy]
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

A beacon should never have a Channel Switch Announcement information
element with a count of 0, because a count of 1 means switch just
before the next beacon.  So, if a count of 0 was valid in a beacon, it
would have been transmitted in the next channel already, which is
useless.  A CSA count equal to zero is only meaningful in action
frames or probe_responses.

Fix the ieee80211_csa_is_complete() and ieee80211_update_csa()
functions accordingly.

With a CSA count of 0, we won't transmit any CSA beacons, because the
switch will happen before the next TBTT.  To avoid extra work and
potential confusion in the drivers, complete the CSA immediately,
instead of waiting for the driver to call ieee80211_csa_finish().

To keep things simpler, we also switch immediately when the CSA count
is 1, while in theory we should delay the switch until just before the
next TBTT.

Additionally, move the ieee80211_csa_finish() function to cfg.c,
where it makes more sense.
Tested-by: NSimon Wunderlich <sw@simonwunderlich.de>
Acked-by: NSimon Wunderlich <sw@simonwunderlich.de>
Signed-off-by: NLuciano Coelho <luciano.coelho@intel.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>

This reverts commit 08ece5bb.

As it breaks ARM builds and needs more attention
on the ARM side.
Acked-by: NDavid Vrabel <david.vrabel@citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

nfs41_wake_and_assign_slot() relies on the task->tk_msg.rpc_argp and
task->tk_msg.rpc_resp always pointing to the session sequence arguments.

nfs4_proc_open_confirm tries to pull a fast one by reusing the open
sequence structure, thus causing corruption of the NFSv4 slot table.

Cc: stable@vger.kernel.org # 3.12+
Signed-off-by: NTrond Myklebust <trond.myklebust@primarydata.com>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

The grant mapping API does m2p_override unnecessarily: only gntdev needs it,
for blkback and future netback patches it just cause a lock contention, as
those pages never go to userspace. Therefore this series does the following:
- the original functions were renamed to __gnttab_[un]map_refs, with a new
  parameter m2p_override
- based on m2p_override either they follow the original behaviour, or just set
  the private flag and call set_phys_to_machine
- gnttab_[un]map_refs are now a wrapper to call __gnttab_[un]map_refs with
  m2p_override false
- a new function gnttab_[un]map_refs_userspace provides the old behaviour

It also removes a stray space from page.h and change ret to 0 if
XENFEAT_auto_translated_physmap, as that is the only possible return value
there.

v2:
- move the storing of the old mfn in page->index to gnttab_map_refs
- move the function header update to a separate patch

v3:
- a new approach to retain old behaviour where it needed
- squash the patches into one

v4:
- move out the common bits from m2p* functions, and pass pfn/mfn as parameter
- clear page->private before doing anything with the page, so m2p_find_override
  won't race with this

v5:
- change return value handling in __gnttab_[un]map_refs
- remove a stray space in page.h
- add detail why ret = 0 now at some places

v6:
- don't pass pfn to m2p* functions, just get it locally
Signed-off-by: NZoltan Kiss <zoltan.kiss@citrix.com>
Suggested-by: NDavid Vrabel <david.vrabel@citrix.com>
Acked-by: NDavid Vrabel <david.vrabel@citrix.com>
Acked-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>

On x86, SLUB creates and handles <=8192-byte allocations internally.
It passes larger ones up to the allocator.  Saying "up to order 2" is,
at best, ambiguous.  Is that order-1?  Or (order-2 bytes)?  Make
it more clear.

SLOB commits a similar sin.  It *handles* page-size requests, but the
comment says that it passes up "all page size and larger requests".

SLOB also swaps around the order of the very-similarly-named
KMALLOC_SHIFT_HIGH and KMALLOC_SHIFT_MAX #defines.  Make it
consistent with the order of the other two allocators.

Cc: Matt Mackall <mpm@selenic.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Acked-by: NChristoph Lameter <cl@linux-foundation.org>
Acked-by: NDavid Rientjes <rientjes@google.com>
Signed-off-by: NDave Hansen <dave.hansen@linux.intel.com>
Signed-off-by: NPekka Enberg <penberg@kernel.org>

Add my copyright to the zsmalloc source code which I maintain.
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This patch moves zsmalloc under mm directory.

Before that, description will explain why we have needed custom
allocator.

Zsmalloc is a new slab-based memory allocator for storing compressed
pages.  It is designed for low fragmentation and high allocation success
rate on large object, but <= PAGE_SIZE allocations.

zsmalloc differs from the kernel slab allocator in two primary ways to
achieve these design goals.

zsmalloc never requires high order page allocations to back slabs, or
"size classes" in zsmalloc terms.  Instead it allows multiple
single-order pages to be stitched together into a "zspage" which backs
the slab.  This allows for higher allocation success rate under memory
pressure.

Also, zsmalloc allows objects to span page boundaries within the zspage.
This allows for lower fragmentation than could be had with the kernel
slab allocator for objects between PAGE_SIZE/2 and PAGE_SIZE.  With the
kernel slab allocator, if a page compresses to 60% of it original size,
the memory savings gained through compression is lost in fragmentation
because another object of the same size can't be stored in the leftover
space.

This ability to span pages results in zsmalloc allocations not being
directly addressable by the user.  The user is given an
non-dereferencable handle in response to an allocation request.  That
handle must be mapped, using zs_map_object(), which returns a pointer to
the mapped region that can be used.  The mapping is necessary since the
object data may reside in two different noncontigious pages.

The zsmalloc fulfills the allocation needs for zram perfectly

[sjenning@linux.vnet.ibm.com: borrow Seth's quote]
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Acked-by: NNitin Gupta <ngupta@vflare.org>
Reviewed-by: NKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Bob Liu <bob.liu@oracle.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Seth Jennings <sjenning@linux.vnet.ibm.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Make smp_call_function_single and friends more efficient by using a
lockless list.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Now we have memblock_virt_alloc_low to replace original bootmem api in
swiotlb.

But we should not use BOOTMEM_LOW_LIMIT for arch that does not support
CONFIG_NOBOOTMEM, as old api take 0.

| #define alloc_bootmem_low(x) \
|        __alloc_bootmem_low(x, SMP_CACHE_BYTES, 0)
|#define alloc_bootmem_low_pages_nopanic(x) \
|        __alloc_bootmem_low_nopanic(x, PAGE_SIZE, 0)

and we have
 #define BOOTMEM_LOW_LIMIT __pa(MAX_DMA_ADDRESS)
for CONFIG_NOBOOTMEM.

Restore goal to 0 to fix ia64 crash, that Tony found.
Signed-off-by: NYinghai Lu <yinghai@kernel.org>
Reported-by: NTony Luck <tony.luck@gmail.com>
Tested-by: NTony Luck <tony.luck@intel.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

On ARM, address size can be 32 bits or 64 bits (if CONFIG_ARCH_PHYS_ADDR_T_64BIT
is enabled).
We can't assume that the grant frame base address will always fits in an
unsigned long. Use phys_addr_t instead of unsigned long as argument for
gnttab_setup_auto_xlat_frames.
Signed-off-by: NJulien Grall <julien.grall@linaro.org>
Signed-off-by: NStefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>
Reviewed-by: NDavid Vrabel <david.vrabel@citrix.com>

This patch fixes a percpu_ref_put race for se_lun->lun_ref in
transport_lun_remove_cmd() where ->lun_ref could end up being
put more than once per command via different target completion
and fabric release contexts.

It adds a cmpxchg() for se_cmd->lun_ref_active to ensure that
percpu_ref_put() is only ever called once per se_cmd.

This bug was manifesting itself as a LUN shutdown regression
bug in >= v3.13 code, where percpu_ref_kill() would end up
hanging indefinately due to the incorrect percpu_ref count.

(Change se_cmd->lun_ref_active from bool -> int to force at
 least a 4-byte cmpxchg with MIPS ll/sc ins. - Fengguang)
Reported-by: NTommy Apel <tommyapeldk@gmail.com>
Cc: Tommy Apel <tommyapeldk@gmail.com>
Cc: <stable@vger.kernel.org> #3.13+
Signed-off-by: NNicholas Bellinger <nab@linux-iscsi.org>

Commit d5dc77bf ("consolidate compat lookup_dcookie()") coverted all
architectures to the new compat_sys_lookup_dcookie() syscall.

The "len" paramater of the new compat syscall must have the type
compat_size_t in order to enforce zero extension for architectures where
the ABI requires that the caller of a function performed zero and/or
sign extension to 64 bit of all parameters.
Signed-off-by: NHeiko Carstens <heiko.carstens@de.ibm.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: <stable@vger.kernel.org>	[v3.10+]
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

We got a report that the pwritev syscall does not work correctly in
compat mode on s390.

It turned out that with commit 72ec3516 ("switch compat readv/writev
variants to COMPAT_SYSCALL_DEFINE") we lost the zero extension of a
couple of syscall parameters because the some parameter types haven't
been converted from unsigned long to compat_ulong_t.

This is needed for architectures where the ABI requires that the caller
of a function performed zero and/or sign extension to 64 bit of all
parameters.
Signed-off-by: NHeiko Carstens <heiko.carstens@de.ibm.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: <stable@vger.kernel.org>	[v3.10+]
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>