Pull aio changes from Ben LaHaise:
"First off, sorry for this pull request being late in the merge window.
Al had raised a couple of concerns about 2 items in the series below.
I addressed the first issue (the race introduced by Gu's use of
mm_populate()), but he has not provided any further details on how he
wants to rework the anon_inode.c changes (which were sent out months
ago but have yet to be commented on).
The bulk of the changes have been sitting in the -next tree for a few
months, with all the issues raised being addressed"
* git://git.kvack.org/~bcrl/aio-next: (22 commits)
aio: rcu_read_lock protection for new rcu_dereference calls
aio: fix race in ring buffer page lookup introduced by page migration support
aio: fix rcu sparse warnings introduced by ioctx table lookup patch
aio: remove unnecessary debugging from aio_free_ring()
aio: table lookup: verify ctx pointer
staging/lustre: kiocb->ki_left is removed
aio: fix error handling and rcu usage in "convert the ioctx list to table lookup v3"
aio: be defensive to ensure request batching is non-zero instead of BUG_ON()
aio: convert the ioctx list to table lookup v3
aio: double aio_max_nr in calculations
aio: Kill ki_dtor
aio: Kill ki_users
aio: Kill unneeded kiocb members
aio: Kill aio_rw_vect_retry()
aio: Don't use ctx->tail unnecessarily
aio: io_cancel() no longer returns the io_event
aio: percpu ioctx refcount
aio: percpu reqs_available
aio: reqs_active -> reqs_available
aio: fix build when migration is disabled
...
After the last architecture switched to generic hard irqs the config
options HAVE_GENERIC_HARDIRQS & GENERIC_HARDIRQS and the related code
for !CONFIG_GENERIC_HARDIRQS can be removed.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Pull SCSI target updates from Nicholas Bellinger:
"Lots of activity again this round for I/O performance optimizations
(per-cpu IDA pre-allocation for vhost + iscsi/target), and the
addition of new fabric independent features to target-core
(COMPARE_AND_WRITE + EXTENDED_COPY).
The main highlights include:
- Support for iscsi-target login multiplexing across individual
network portals
- Generic Per-cpu IDA logic (kent + akpm + clameter)
- Conversion of vhost to use per-cpu IDA pre-allocation for
descriptors, SGLs and userspace page pointer list
- Conversion of iscsi-target + iser-target to use per-cpu IDA
pre-allocation for descriptors
- Add support for generic COMPARE_AND_WRITE (AtomicTestandSet)
emulation for virtual backend drivers
- Add support for generic EXTENDED_COPY (CopyOffload) emulation for
virtual backend drivers.
- Add support for fast memory registration mode to iser-target (Vu)
The patches to add COMPARE_AND_WRITE and EXTENDED_COPY support are of
particular significance, which make us the first and only open source
target to support the full set of VAAI primitives.
Currently Linux clients are lacking upstream support to actually
utilize these primitives. However, with server side support now in
place for folks like MKP + ZAB working on the client, this logic once
reserved for the highest end of storage arrays, can now be run in VMs
on their laptops"
* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending: (50 commits)
target/iscsi: Bump versions to v4.1.0
target: Update copyright ownership/year information to 2013
iscsi-target: Bump default TCP listen backlog to 256
target: Fix >= v3.9+ regression in PR APTPL + ALUA metadata write-out
iscsi-target; Bump default CmdSN Depth to 64
iscsi-target: Remove unnecessary wait_for_completion in iscsi_get_thread_set
iscsi-target: Add thread_set->ts_activate_sem + use common deallocate
iscsi-target: Fix race with thread_pre_handler flush_signals + ISCSI_THREAD_SET_DIE
target: remove unused including <linux/version.h>
iser-target: introduce fast memory registration mode (FRWR)
iser-target: generalize rdma memory registration and cleanup
iser-target: move rdma wr processing to a shared function
target: Enable global EXTENDED_COPY setup/release
target: Add Third Party Copy (3PC) bit in INQUIRY response
target: Enable EXTENDED_COPY setup in spc_parse_cdb
target: Add support for EXTENDED_COPY copy offload emulation
target: Avoid non-existent tg_pt_gp_mem in target_alua_state_check
target: Add global device list for EXTENDED_COPY
target: Make helpers non static for EXTENDED_COPY command setup
target: Make spc_parse_naa_6h_vendor_specific non static
...
Merge more patches from Andrew Morton:
"The rest of MM. Plus one misc cleanup"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (35 commits)
mm/Kconfig: add MMU dependency for MIGRATION.
kernel: replace strict_strto*() with kstrto*()
mm, thp: count thp_fault_fallback anytime thp fault fails
thp: consolidate code between handle_mm_fault() and do_huge_pmd_anonymous_page()
thp: do_huge_pmd_anonymous_page() cleanup
thp: move maybe_pmd_mkwrite() out of mk_huge_pmd()
mm: cleanup add_to_page_cache_locked()
thp: account anon transparent huge pages into NR_ANON_PAGES
truncate: drop 'oldsize' truncate_pagecache() parameter
mm: make lru_add_drain_all() selective
memcg: document cgroup dirty/writeback memory statistics
memcg: add per cgroup writeback pages accounting
memcg: check for proper lock held in mem_cgroup_update_page_stat
memcg: remove MEMCG_NR_FILE_MAPPED
memcg: reduce function dereference
memcg: avoid overflow caused by PAGE_ALIGN
memcg: rename RESOURCE_MAX to RES_COUNTER_MAX
memcg: correct RESOURCE_MAX to ULLONG_MAX
mm: memcg: do not trap chargers with full callstack on OOM
mm: memcg: rework and document OOM waiting and wakeup
...
truncate_pagecache() doesn't care about old size since commit
cedabed49b ("vfs: Fix vmtruncate() regression"). Let's drop it.
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
make lru_add_drain_all() only selectively interrupt the cpus that have
per-cpu free pages that can be drained.
This is important in nohz mode where calling mlockall(), for example,
otherwise will interrupt every core unnecessarily.
This is important on workloads where nohz cores are handling 10 Gb traffic
in userspace. Those CPUs do not enter the kernel and place pages into LRU
pagevecs and they really, really don't want to be interrupted, or they
drop packets on the floor.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Reviewed-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add memcg routines to count writeback pages, later dirty pages will also
be accounted.
After Kame's commit 89c06bd52f ("memcg: use new logic for page stat
accounting"), we can use 'struct page' flag to test page state instead
of per page_cgroup flag. But memcg has a feature to move a page from a
cgroup to another one and may have race between "move" and "page stat
accounting". So in order to avoid the race we have designed a new lock:
mem_cgroup_begin_update_page_stat()
modify page information -->(a)
mem_cgroup_update_page_stat() -->(b)
mem_cgroup_end_update_page_stat()
It requires both (a) and (b)(writeback pages accounting) to be pretected
in mem_cgroup_{begin/end}_update_page_stat(). It's full no-op for
!CONFIG_MEMCG, almost no-op if memcg is disabled (but compiled in), rcu
read lock in the most cases (no task is moving), and spin_lock_irqsave
on top in the slow path.
There're two writeback interfaces to modify: test_{clear/set}_page_writeback().
And the lock order is:
--> memcg->move_lock
--> mapping->tree_lock
Signed-off-by: Sha Zhengju <handai.szj@taobao.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Greg Thelen <gthelen@google.com>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
While accounting memcg page stat, it's not worth to use
MEMCG_NR_FILE_MAPPED as an extra layer of indirection because of the
complexity and presumed performance overhead. We can use
MEM_CGROUP_STAT_FILE_MAPPED directly.
Signed-off-by: Sha Zhengju <handai.szj@taobao.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Fengguang Wu <fengguang.wu@intel.com>
Reviewed-by: Greg Thelen <gthelen@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
RESOURCE_MAX is far too general name, change it to RES_COUNTER_MAX.
Signed-off-by: Sha Zhengju <handai.szj@taobao.com>
Signed-off-by: Qiang Huang <h.huangqiang@huawei.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Cc: Jeff Liu <jeff.liu@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Current RESOURCE_MAX is ULONG_MAX, but the value we used to set resource
limit is unsigned long long, so we can set bigger value than that which is
strange. The XXX_MAX should be reasonable max value, bigger than that
should be overflow.
Notice that this change will affect user output of default *.limit_in_bytes:
before change:
$ cat /cgroup/memory/memory.limit_in_bytes
9223372036854775807
after change:
$ cat /cgroup/memory/memory.limit_in_bytes
18446744073709551615
But it doesn't alter the API in term of input - we can still use "echo -1
> *.limit_in_bytes" to reset the numbers to "unlimited".
Signed-off-by: Sha Zhengju <handai.szj@taobao.com>
Signed-off-by: Qiang Huang <h.huangqiang@huawei.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Cc: Jeff Liu <jeff.liu@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The memcg OOM handling is incredibly fragile and can deadlock. When a
task fails to charge memory, it invokes the OOM killer and loops right
there in the charge code until it succeeds. Comparably, any other task
that enters the charge path at this point will go to a waitqueue right
then and there and sleep until the OOM situation is resolved. The problem
is that these tasks may hold filesystem locks and the mmap_sem; locks that
the selected OOM victim may need to exit.
For example, in one reported case, the task invoking the OOM killer was
about to charge a page cache page during a write(), which holds the
i_mutex. The OOM killer selected a task that was just entering truncate()
and trying to acquire the i_mutex:
OOM invoking task:
mem_cgroup_handle_oom+0x241/0x3b0
mem_cgroup_cache_charge+0xbe/0xe0
add_to_page_cache_locked+0x4c/0x140
add_to_page_cache_lru+0x22/0x50
grab_cache_page_write_begin+0x8b/0xe0
ext3_write_begin+0x88/0x270
generic_file_buffered_write+0x116/0x290
__generic_file_aio_write+0x27c/0x480
generic_file_aio_write+0x76/0xf0 # takes ->i_mutex
do_sync_write+0xea/0x130
vfs_write+0xf3/0x1f0
sys_write+0x51/0x90
system_call_fastpath+0x18/0x1d
OOM kill victim:
do_truncate+0x58/0xa0 # takes i_mutex
do_last+0x250/0xa30
path_openat+0xd7/0x440
do_filp_open+0x49/0xa0
do_sys_open+0x106/0x240
sys_open+0x20/0x30
system_call_fastpath+0x18/0x1d
The OOM handling task will retry the charge indefinitely while the OOM
killed task is not releasing any resources.
A similar scenario can happen when the kernel OOM killer for a memcg is
disabled and a userspace task is in charge of resolving OOM situations.
In this case, ALL tasks that enter the OOM path will be made to sleep on
the OOM waitqueue and wait for userspace to free resources or increase
the group's limit. But a userspace OOM handler is prone to deadlock
itself on the locks held by the waiting tasks. For example one of the
sleeping tasks may be stuck in a brk() call with the mmap_sem held for
writing but the userspace handler, in order to pick an optimal victim,
may need to read files from /proc/<pid>, which tries to acquire the same
mmap_sem for reading and deadlocks.
This patch changes the way tasks behave after detecting a memcg OOM and
makes sure nobody loops or sleeps with locks held:
1. When OOMing in a user fault, invoke the OOM killer and restart the
fault instead of looping on the charge attempt. This way, the OOM
victim can not get stuck on locks the looping task may hold.
2. When OOMing in a user fault but somebody else is handling it
(either the kernel OOM killer or a userspace handler), don't go to
sleep in the charge context. Instead, remember the OOMing memcg in
the task struct and then fully unwind the page fault stack with
-ENOMEM. pagefault_out_of_memory() will then call back into the
memcg code to check if the -ENOMEM came from the memcg, and then
either put the task to sleep on the memcg's OOM waitqueue or just
restart the fault. The OOM victim can no longer get stuck on any
lock a sleeping task may hold.
Debugged by Michal Hocko.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reported-by: azurIt <azurit@pobox.sk>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
System calls and kernel faults (uaccess, gup) can handle an out of memory
situation gracefully and just return -ENOMEM.
Enable the memcg OOM killer only for user faults, where it's really the
only option available.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: azurIt <azurit@pobox.sk>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Unlike global OOM handling, memory cgroup code will invoke the OOM killer
in any OOM situation because it has no way of telling faults occuring in
kernel context - which could be handled more gracefully - from
user-triggered faults.
Pass a flag that identifies faults originating in user space from the
architecture-specific fault handlers to generic code so that memcg OOM
handling can be improved.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: azurIt <azurit@pobox.sk>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The caller of the iterator might know that some nodes or even subtrees
should be skipped but there is no way to tell iterators about that so the
only choice left is to let iterators to visit each node and do the
selection outside of the iterating code. This, however, doesn't scale
well with hierarchies with many groups where only few groups are
interesting.
This patch adds mem_cgroup_iter_cond variant of the iterator with a
callback which gets called for every visited node. There are three
possible ways how the callback can influence the walk. Either the node is
visited, it is skipped but the tree walk continues down the tree or the
whole subtree of the current group is skipped.
[hughd@google.com: fix memcg-less page reclaim]
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Cc: Balbir Singh <bsingharora@gmail.com>
Cc: Glauber Costa <glommer@openvz.org>
Cc: Greg Thelen <gthelen@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Michel Lespinasse <walken@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Ying Han <yinghan@google.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Soft reclaim has been done only for the global reclaim (both background
and direct). Since "memcg: integrate soft reclaim tighter with zone
shrinking code" there is no reason for this limitation anymore as the soft
limit reclaim doesn't use any special code paths and it is a part of the
zone shrinking code which is used by both global and targeted reclaims.
From the semantic point of view it is natural to consider soft limit
before touching all groups in the hierarchy tree which is touching the
hard limit because soft limit tells us where to push back when there is a
memory pressure. It is not important whether the pressure comes from the
limit or imbalanced zones.
This patch simply enables soft reclaim unconditionally in
mem_cgroup_should_soft_reclaim so it is enabled for both global and
targeted reclaim paths. mem_cgroup_soft_reclaim_eligible needs to learn
about the root of the reclaim to know where to stop checking soft limit
state of parents up the hierarchy. Say we have
A (over soft limit)
\
B (below s.l., hit the hard limit)
/ \
C D (below s.l.)
B is the source of the outside memory pressure now for D but we shouldn't
soft reclaim it because it is behaving well under B subtree and we can
still reclaim from C (pressumably it is over the limit).
mem_cgroup_soft_reclaim_eligible should therefore stop climbing up the
hierarchy at B (root of the memory pressure).
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Glauber Costa <glommer@openvz.org>
Reviewed-by: Tejun Heo <tj@kernel.org>
Cc: Balbir Singh <bsingharora@gmail.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Michel Lespinasse <walken@google.com>
Cc: Ying Han <yinghan@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patchset is sitting out of tree for quite some time without any
objections. I would be really happy if it made it into 3.12. I do not
want to push it too hard but I think this work is basically ready and
waiting more doesn't help.
The basic idea is quite simple. Pull soft reclaim into shrink_zone in the
first step and get rid of the previous soft reclaim infrastructure.
shrink_zone is done in two passes now. First it tries to do the soft
limit reclaim and it falls back to reclaim-all mode if no group is over
the limit or no pages have been scanned. The second pass happens at the
same priority so the only time we waste is the memcg tree walk which has
been updated in the third step to have only negligible overhead.
As a bonus we will get rid of a _lot_ of code by this and soft reclaim
will not stand out like before when it wasn't integrated into the zone
shrinking code and it reclaimed at priority 0 (the testing results show
that some workloads suffers from such an aggressive reclaim). The clean
up is in a separate patch because I felt it would be easier to review that
way.
The second step is soft limit reclaim integration into targeted reclaim.
It should be rather straight forward. Soft limit has been used only for
the global reclaim so far but it makes sense for any kind of pressure
coming from up-the-hierarchy, including targeted reclaim.
The third step (patches 4-8) addresses the tree walk overhead by enhancing
memcg iterators to enable skipping whole subtrees and tracking number of
over soft limit children at each level of the hierarchy. This information
is updated same way the old soft limit tree was updated (from
memcg_check_events) so we shouldn't see an additional overhead. In fact
mem_cgroup_update_soft_limit is much simpler than tree manipulation done
previously.
__shrink_zone uses mem_cgroup_soft_reclaim_eligible as a predicate for
mem_cgroup_iter so the decision whether a particular group should be
visited is done at the iterator level which allows us to decide to skip
the whole subtree as well (if there is no child in excess). This reduces
the tree walk overhead considerably.
* TEST 1
========
My primary test case was a parallel kernel build with 2 groups (make is
running with -j8 with a distribution .config in a separate cgroup without
any hard limit) on a 32 CPU machine booted with 1GB memory and both builds
run taskset to Node 0 cpus.
I was mostly interested in 2 setups. Default - no soft limit set and -
and 0 soft limit set to both groups. The first one should tell us whether
the rework regresses the default behavior while the second one should show
us improvements in an extreme case where both workloads are always over
the soft limit.
/usr/bin/time -v has been used to collect the statistics and each
configuration had 3 runs after fresh boot without any other load on the
system.
base is mmotm-2013-07-18-16-40
rework all 8 patches applied on top of base
* No-limit
User
no-limit/base: min: 651.92 max: 672.65 avg: 664.33 std: 8.01 runs: 6
no-limit/rework: min: 657.34 [100.8%] max: 668.39 [99.4%] avg: 663.13 [99.8%] std: 3.61 runs: 6
System
no-limit/base: min: 69.33 max: 71.39 avg: 70.32 std: 0.79 runs: 6
no-limit/rework: min: 69.12 [99.7%] max: 71.05 [99.5%] avg: 70.04 [99.6%] std: 0.59 runs: 6
Elapsed
no-limit/base: min: 398.27 max: 422.36 avg: 408.85 std: 7.74 runs: 6
no-limit/rework: min: 386.36 [97.0%] max: 438.40 [103.8%] avg: 416.34 [101.8%] std: 18.85 runs: 6
The results are within noise. Elapsed time has a bigger variance but the
average looks good.
* 0-limit
User
0-limit/base: min: 573.76 max: 605.63 avg: 585.73 std: 12.21 runs: 6
0-limit/rework: min: 645.77 [112.6%] max: 666.25 [110.0%] avg: 656.97 [112.2%] std: 7.77 runs: 6
System
0-limit/base: min: 69.57 max: 71.13 avg: 70.29 std: 0.54 runs: 6
0-limit/rework: min: 68.68 [98.7%] max: 71.40 [100.4%] avg: 69.91 [99.5%] std: 0.87 runs: 6
Elapsed
0-limit/base: min: 1306.14 max: 1550.17 avg: 1430.35 std: 90.86 runs: 6
0-limit/rework: min: 404.06 [30.9%] max: 465.94 [30.1%] avg: 434.81 [30.4%] std: 22.68 runs: 6
The improvement is really huge here (even bigger than with my previous
testing and I suspect that this highly depends on the storage). Page
fault statistics tell us at least part of the story:
Minor
0-limit/base: min: 37180461.00 max: 37319986.00 avg: 37247470.00 std: 54772.71 runs: 6
0-limit/rework: min: 36751685.00 [98.8%] max: 36805379.00 [98.6%] avg: 36774506.33 [98.7%] std: 17109.03 runs: 6
Major
0-limit/base: min: 170604.00 max: 221141.00 avg: 196081.83 std: 18217.01 runs: 6
0-limit/rework: min: 2864.00 [1.7%] max: 10029.00 [4.5%] avg: 5627.33 [2.9%] std: 2252.71 runs: 6
Same as with my previous testing Minor faults are more or less within
noise but Major fault count is way bellow the base kernel.
While this looks as a nice win it is fair to say that 0-limit
configuration is quite artificial. So I was playing with 0-no-limit
loads as well.
* TEST 2
========
The following results are from 2 groups configuration on a 16GB machine
(single NUMA node).
- A running stream IO (dd if=/dev/zero of=local.file bs=1024) with
2*TotalMem with 0 soft limit.
- B running a mem_eater which consumes TotalMem-1G without any limit. The
mem_eater consumes the memory in 100 chunks with 1s nap after each
mmap+poppulate so that both loads have chance to fight for the memory.
The expected result is that B shouldn't be reclaimed and A shouldn't see
a big dropdown in elapsed time.
User
base: min: 2.68 max: 2.89 avg: 2.76 std: 0.09 runs: 3
rework: min: 3.27 [122.0%] max: 3.74 [129.4%] avg: 3.44 [124.6%] std: 0.21 runs: 3
System
base: min: 86.26 max: 88.29 avg: 87.28 std: 0.83 runs: 3
rework: min: 81.05 [94.0%] max: 84.96 [96.2%] avg: 83.14 [95.3%] std: 1.61 runs: 3
Elapsed
base: min: 317.28 max: 332.39 avg: 325.84 std: 6.33 runs: 3
rework: min: 281.53 [88.7%] max: 298.16 [89.7%] avg: 290.99 [89.3%] std: 6.98 runs: 3
System time improved slightly as well as Elapsed. My previous testing
has shown worse numbers but this again seem to depend on the storage
speed.
My theory is that the writeback doesn't catch up and prio-0 soft reclaim
falls into wait on writeback page too often in the base kernel. The
patched kernel doesn't do that because the soft reclaim is done from the
kswapd/direct reclaim context. This can be seen on the following graph
nicely. The A's group usage_in_bytes regurarly drops really low very often.
All 3 runs
http://labs.suse.cz/mhocko/soft_limit_rework/stream_io-vs-mem_eater/stream.png
resp. a detail of the single run
http://labs.suse.cz/mhocko/soft_limit_rework/stream_io-vs-mem_eater/stream-one-run.png
mem_eater seems to be doing better as well. It gets to the full
allocation size faster as can be seen on the following graph:
http://labs.suse.cz/mhocko/soft_limit_rework/stream_io-vs-mem_eater/mem_eater-one-run.png
/proc/meminfo collected during the test also shows that rework kernel
hasn't swapped that much (well almost not at all):
base: max: 123900 K avg: 56388.29 K
rework: max: 300 K avg: 128.68 K
kswapd and direct reclaim statistics are of no use unfortunatelly because
soft reclaim is not accounted properly as the counters are hidden by
global_reclaim() checks in the base kernel.
* TEST 3
========
Another test was the same configuration as TEST2 except the stream IO was
replaced by a single kbuild (16 parallel jobs bound to Node0 cpus same as
in TEST1) and mem_eater allocated TotalMem-200M so kbuild had only 200MB
left.
Kbuild did better with the rework kernel here as well:
User
base: min: 860.28 max: 872.86 avg: 868.03 std: 5.54 runs: 3
rework: min: 880.81 [102.4%] max: 887.45 [101.7%] avg: 883.56 [101.8%] std: 2.83 runs: 3
System
base: min: 84.35 max: 85.06 avg: 84.79 std: 0.31 runs: 3
rework: min: 85.62 [101.5%] max: 86.09 [101.2%] avg: 85.79 [101.2%] std: 0.21 runs: 3
Elapsed
base: min: 135.36 max: 243.30 avg: 182.47 std: 45.12 runs: 3
rework: min: 110.46 [81.6%] max: 116.20 [47.8%] avg: 114.15 [62.6%] std: 2.61 runs: 3
Minor
base: min: 36635476.00 max: 36673365.00 avg: 36654812.00 std: 15478.03 runs: 3
rework: min: 36639301.00 [100.0%] max: 36695541.00 [100.1%] avg: 36665511.00 [100.0%] std: 23118.23 runs: 3
Major
base: min: 14708.00 max: 53328.00 avg: 31379.00 std: 16202.24 runs: 3
rework: min: 302.00 [2.1%] max: 414.00 [0.8%] avg: 366.33 [1.2%] std: 47.22 runs: 3
Again we can see a significant improvement in Elapsed (it also seems to
be more stable), there is a huge dropdown for the Major page faults and
much more swapping:
base: max: 583736 K avg: 112547.43 K
rework: max: 4012 K avg: 124.36 K
Graphs from all three runs show the variability of the kbuild quite
nicely. It even seems that it took longer after every run with the base
kernel which would be quite surprising as the source tree for the build is
removed and caches are dropped after each run so the build operates on a
freshly extracted sources everytime.
http://labs.suse.cz/mhocko/soft_limit_rework/stream_io-vs-mem_eater/kbuild-mem_eater.png
My other testing shows that this is just a matter of timing and other runs
behave differently the std for Elapsed time is similar ~50. Example of
other three runs:
http://labs.suse.cz/mhocko/soft_limit_rework/stream_io-vs-mem_eater/kbuild-mem_eater2.png
So to wrap this up. The series is still doing good and improves the soft
limit.
The testing results for bunch of cgroups with both stream IO and kbuild
loads can be found in "memcg: track children in soft limit excess to
improve soft limit".
This patch:
Memcg soft reclaim has been traditionally triggered from the global
reclaim paths before calling shrink_zone. mem_cgroup_soft_limit_reclaim
then picked up a group which exceeds the soft limit the most and reclaimed
it with 0 priority to reclaim at least SWAP_CLUSTER_MAX pages.
The infrastructure requires per-node-zone trees which hold over-limit
groups and keep them up-to-date (via memcg_check_events) which is not cost
free. Although this overhead hasn't turned out to be a bottle neck the
implementation is suboptimal because mem_cgroup_update_tree has no idea
which zones consumed memory over the limit so we could easily end up
having a group on a node-zone tree having only few pages from that
node-zone.
This patch doesn't try to fix node-zone trees management because it seems
that integrating soft reclaim into zone shrinking sounds much easier and
more appropriate for several reasons. First of all 0 priority reclaim was
a crude hack which might lead to big stalls if the group's LRUs are big
and hard to reclaim (e.g. a lot of dirty/writeback pages). Soft reclaim
should be applicable also to the targeted reclaim which is awkward right
now without additional hacks. Last but not least the whole infrastructure
eats quite some code.
After this patch shrink_zone is done in 2 passes. First it tries to do
the soft reclaim if appropriate (only for global reclaim for now to keep
compatible with the original state) and fall back to ignoring soft limit
if no group is eligible to soft reclaim or nothing has been scanned during
the first pass. Only groups which are over their soft limit or any of
their parents up the hierarchy is over the limit are considered eligible
during the first pass.
Soft limit tree which is not necessary anymore will be removed in the
follow up patch to make this patch smaller and easier to review.
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Glauber Costa <glommer@openvz.org>
Reviewed-by: Tejun Heo <tj@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Ying Han <yinghan@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Michel Lespinasse <walken@google.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Balbir Singh <bsingharora@gmail.com>
Cc: Glauber Costa <glommer@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull vfs pile 4 from Al Viro:
"list_lru pile, mostly"
This came out of Andrew's pile, Al ended up doing the merge work so that
Andrew didn't have to.
Additionally, a few fixes.
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (42 commits)
super: fix for destroy lrus
list_lru: dynamically adjust node arrays
shrinker: Kill old ->shrink API.
shrinker: convert remaining shrinkers to count/scan API
staging/lustre/libcfs: cleanup linux-mem.h
staging/lustre/ptlrpc: convert to new shrinker API
staging/lustre/obdclass: convert lu_object shrinker to count/scan API
staging/lustre/ldlm: convert to shrinkers to count/scan API
hugepage: convert huge zero page shrinker to new shrinker API
i915: bail out earlier when shrinker cannot acquire mutex
drivers: convert shrinkers to new count/scan API
fs: convert fs shrinkers to new scan/count API
xfs: fix dquot isolation hang
xfs-convert-dquot-cache-lru-to-list_lru-fix
xfs: convert dquot cache lru to list_lru
xfs: rework buffer dispose list tracking
xfs-convert-buftarg-lru-to-generic-code-fix
xfs: convert buftarg LRU to generic code
fs: convert inode and dentry shrinking to be node aware
vmscan: per-node deferred work
...
Pull led updates from Bryan Wu:
"Sorry for the late pull request, since I'm just back from vacation.
LED subsystem updates for 3.12:
- pca9633 driver DT supporting and pca9634 chip supporting
- restore legacy device attributes for lp5521
- other fixing and updates"
* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/linux-leds: (28 commits)
leds: wm831x-status: Request a REG resource
leds: trigger: ledtrig-backlight: Fix invalid memory access in fb_event notification callback
leds-pca963x: Fix device tree parsing
leds-pca9633: Rename to leds-pca963x
leds-pca9633: Add mutex to the ledout register
leds-pca9633: Unique naming of the LEDs
leds-pca9633: Add support for PCA9634
leds: lp5562: use LP55xx common macros for device attributes
Documentation: leds-lp5521,lp5523: update device attribute information
leds: lp5523: remove unnecessary writing commands
leds: lp5523: restore legacy device attributes
leds: lp5523: LED MUX configuration on initializing
leds: lp5523: make separate API for loading engine
leds: lp5521: remove unnecessary writing commands
leds: lp5521: restore legacy device attributes
leds: lp55xx: add common macros for device attributes
leds: lp55xx: add common data structure for program
Documentation: leds: Fix a typo
leds: ss4200: Fix incorrect placement of __initdata
leds: clevo-mail: Fix incorrect placement of __initdata
...
This round the updates contain:
* A new driver for the Freescale PAMU IOMMU from Varun Sethi.
This driver has cooked for a while and required changes to the
IOMMU-API and infrastructure that were already merged before.
* Updates for the ARM-SMMU driver from Will Deacon
* Various fixes, the most important one is probably a fix from
Alex Williamson for a memory leak in the VT-d page-table
freeing code
In summary not all that much. The biggest part in the diffstat is the
new PAMU driver.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)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=UZoD
-----END PGP SIGNATURE-----
Merge tag 'iommu-updates-v3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu
Pull IOMMU Updates from Joerg Roedel:
"This round the updates contain:
- A new driver for the Freescale PAMU IOMMU from Varun Sethi.
This driver has cooked for a while and required changes to the
IOMMU-API and infrastructure that were already merged before.
- Updates for the ARM-SMMU driver from Will Deacon
- Various fixes, the most important one is probably a fix from Alex
Williamson for a memory leak in the VT-d page-table freeing code
In summary not all that much. The biggest part in the diffstat is the
new PAMU driver"
* tag 'iommu-updates-v3.12' of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:
intel-iommu: Fix leaks in pagetable freeing
iommu/amd: Fix resource leak in iommu_init_device()
iommu/amd: Clean up unnecessary MSI/MSI-X capability find
iommu/arm-smmu: Simplify VMID and ASID allocation
iommu/arm-smmu: Don't use VMIDs for stage-1 translations
iommu/arm-smmu: Tighten up global fault reporting
iommu/arm-smmu: Remove broken big-endian check
iommu/fsl: Remove unnecessary 'fsl-pamu' prefixes
iommu/fsl: Fix whitespace problems noticed by git-am
iommu/fsl: Freescale PAMU driver and iommu implementation.
iommu/fsl: Add additional iommu attributes required by the PAMU driver.
powerpc: Add iommu domain pointer to device archdata
iommu/exynos: Remove dead code (set_prefbuf)
1) ACPI-based PCI hotplug (ACPIPHP) fixes related to spurious events
After the recent ACPIPHP changes we've seen some interesting breakage
on a system that triggers device check notifications during boot for
non-existing devices. Although those notifications are really
spurious, we should be able to deal with them nevertheless and that
shouldn't introduce too much overhead. Four commits to make that
work properly.
2) Memory hotplug and hibernation mutual exclusion rework
This was maent to be a cleanup, but it happens to fix a classical
ABBA deadlock between system suspend/hibernation and ACPI memory
hotplug which is possible if they are started roughly at the same
time. Three commits rework memory hotplug so that it doesn't
acquire pm_mutex and make hibernation use device_hotplug_lock
which prevents it from racing with memory hotplug.
3) ACPI Intel LPSS (Low-Power Subsystem) driver crash fix
The ACPI LPSS driver crashes during boot on Apple Macbook Air with
Haswell that has slightly unusual BIOS configuration in which one
of the LPSS device's _CRS method doesn't return all of the information
expected by the driver. Fix from Mika Westerberg, for stable.
4) ACPICA fix related to Store->ArgX operation
AML interpreter fix for obscure breakage that causes AML to be
executed incorrectly on some machines (observed in practice). From
Bob Moore.
5) ACPI core fix for PCI ACPI device objects lookup
There still are cases in which there is more than one ACPI device
object matching a given PCI device and we don't choose the one that
the BIOS expects us to choose, so this makes the lookup take more
criteria into account in those cases.
6) Fix to prevent cpuidle from crashing in some rare cases
If the result of cpuidle_get_driver() is NULL, which can happen on
some systems, cpuidle_driver_ref() will crash trying to use that
pointer and the Daniel Fu's fix prevents that from happening.
7) cpufreq fixes related to CPU hotplug
Stephen Boyd reported a number of concurrency problems with cpufreq
related to CPU hotplug which are addressed by a series of fixes
from Srivatsa S Bhat and Viresh Kumar.
8) cpufreq fix for time conversion in time_in_state attribute
Time conversion carried out by cpufreq when user space attempts to
read /sys/devices/system/cpu/cpu*/cpufreq/stats/time_in_state won't
work correcty if cputime_t doesn't map directly to jiffies. Fix
from Andreas Schwab.
9) Revert of a troublesome cpufreq commit
Commit 7c30ed5 (cpufreq: make sure frequency transitions are
serialized) was intended to address some known concurrency problems
in cpufreq related to the ordering of transitions, but unfortunately
it introduced several problems of its own, so I decided to revert it
now and address the original problems later in a more robust way.
10) Intel Haswell CPU models for intel_pstate from Nell Hardcastle.
11) cpufreq fixes related to system suspend/resume
The recent cpufreq changes that made it preserve CPU sysfs attributes
over suspend/resume cycles introduced a possible NULL pointer
dereference that caused it to crash during the second attempt to
suspend. Three commits from Srivatsa S Bhat fix that problem and a
couple of related issues.
12) cpufreq locking fix
cpufreq_policy_restore() should acquire the lock for reading, but
it acquires it for writing. Fix from Lan Tianyu.
/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)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=U/Jw
-----END PGP SIGNATURE-----
Merge tag 'pm+acpi-fixes-3.12-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull ACPI and power management fixes from Rafael Wysocki:
"All of these commits are fixes that have emerged recently and some of
them fix bugs introduced during this merge window.
Specifics:
1) ACPI-based PCI hotplug (ACPIPHP) fixes related to spurious events
After the recent ACPIPHP changes we've seen some interesting
breakage on a system that triggers device check notifications
during boot for non-existing devices. Although those
notifications are really spurious, we should be able to deal with
them nevertheless and that shouldn't introduce too much overhead.
Four commits to make that work properly.
2) Memory hotplug and hibernation mutual exclusion rework
This was maent to be a cleanup, but it happens to fix a classical
ABBA deadlock between system suspend/hibernation and ACPI memory
hotplug which is possible if they are started roughly at the same
time. Three commits rework memory hotplug so that it doesn't
acquire pm_mutex and make hibernation use device_hotplug_lock
which prevents it from racing with memory hotplug.
3) ACPI Intel LPSS (Low-Power Subsystem) driver crash fix
The ACPI LPSS driver crashes during boot on Apple Macbook Air with
Haswell that has slightly unusual BIOS configuration in which one
of the LPSS device's _CRS method doesn't return all of the
information expected by the driver. Fix from Mika Westerberg, for
stable.
4) ACPICA fix related to Store->ArgX operation
AML interpreter fix for obscure breakage that causes AML to be
executed incorrectly on some machines (observed in practice).
From Bob Moore.
5) ACPI core fix for PCI ACPI device objects lookup
There still are cases in which there is more than one ACPI device
object matching a given PCI device and we don't choose the one
that the BIOS expects us to choose, so this makes the lookup take
more criteria into account in those cases.
6) Fix to prevent cpuidle from crashing in some rare cases
If the result of cpuidle_get_driver() is NULL, which can happen on
some systems, cpuidle_driver_ref() will crash trying to use that
pointer and the Daniel Fu's fix prevents that from happening.
7) cpufreq fixes related to CPU hotplug
Stephen Boyd reported a number of concurrency problems with
cpufreq related to CPU hotplug which are addressed by a series of
fixes from Srivatsa S Bhat and Viresh Kumar.
8) cpufreq fix for time conversion in time_in_state attribute
Time conversion carried out by cpufreq when user space attempts to
read /sys/devices/system/cpu/cpu*/cpufreq/stats/time_in_state
won't work correcty if cputime_t doesn't map directly to jiffies.
Fix from Andreas Schwab.
9) Revert of a troublesome cpufreq commit
Commit 7c30ed5 (cpufreq: make sure frequency transitions are
serialized) was intended to address some known concurrency
problems in cpufreq related to the ordering of transitions, but
unfortunately it introduced several problems of its own, so I
decided to revert it now and address the original problems later
in a more robust way.
10) Intel Haswell CPU models for intel_pstate from Nell Hardcastle.
11) cpufreq fixes related to system suspend/resume
The recent cpufreq changes that made it preserve CPU sysfs
attributes over suspend/resume cycles introduced a possible NULL
pointer dereference that caused it to crash during the second
attempt to suspend. Three commits from Srivatsa S Bhat fix that
problem and a couple of related issues.
12) cpufreq locking fix
cpufreq_policy_restore() should acquire the lock for reading, but
it acquires it for writing. Fix from Lan Tianyu"
* tag 'pm+acpi-fixes-3.12-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (25 commits)
cpufreq: Acquire the lock in cpufreq_policy_restore() for reading
cpufreq: Prevent problems in update_policy_cpu() if last_cpu == new_cpu
cpufreq: Restructure if/else block to avoid unintended behavior
cpufreq: Fix crash in cpufreq-stats during suspend/resume
intel_pstate: Add Haswell CPU models
Revert "cpufreq: make sure frequency transitions are serialized"
cpufreq: Use signed type for 'ret' variable, to store negative error values
cpufreq: Remove temporary fix for race between CPU hotplug and sysfs-writes
cpufreq: Synchronize the cpufreq store_*() routines with CPU hotplug
cpufreq: Invoke __cpufreq_remove_dev_finish() after releasing cpu_hotplug.lock
cpufreq: Split __cpufreq_remove_dev() into two parts
cpufreq: Fix wrong time unit conversion
cpufreq: serialize calls to __cpufreq_governor()
cpufreq: don't allow governor limits to be changed when it is disabled
ACPI / bind: Prefer device objects with _STA to those without it
ACPI / hotplug / PCI: Avoid parent bus rescans on spurious device checks
ACPI / hotplug / PCI: Use _OST to notify firmware about notify status
ACPI / hotplug / PCI: Avoid doing too much for spurious notifies
ACPICA: Fix for a Store->ArgX when ArgX contains a reference to a field.
ACPI / hotplug / PCI: Don't trim devices before scanning the namespace
...
Let's not pollute the include files with inline functions that are only
used in a single place. Especially not if we decide we might want to
change the semantics of said function to make it more efficient..
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull btrfs updates from Chris Mason:
"This is against 3.11-rc7, but was pulled and tested against your tree
as of yesterday. We do have two small incrementals queued up, but I
wanted to get this bunch out the door before I hop on an airplane.
This is a fairly large batch of fixes, performance improvements, and
cleanups from the usual Btrfs suspects.
We've included Stefan Behren's work to index subvolume UUIDs, which is
targeted at speeding up send/receive with many subvolumes or snapshots
in place. It closes a long standing performance issue that was built
in to the disk format.
Mark Fasheh's offline dedup work is also here. In this case offline
means the FS is mounted and active, but the dedup work is not done
inline during file IO. This is a building block where utilities are
able to ask the FS to dedup a series of extents. The kernel takes
care of verifying the data involved really is the same. Today this
involves reading both extents, but we'll continue to evolve the
patches"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (118 commits)
Btrfs: optimize key searches in btrfs_search_slot
Btrfs: don't use an async starter for most of our workers
Btrfs: only update disk_i_size as we remove extents
Btrfs: fix deadlock in uuid scan kthread
Btrfs: stop refusing the relocation of chunk 0
Btrfs: fix memory leak of uuid_root in free_fs_info
btrfs: reuse kbasename helper
btrfs: return btrfs error code for dev excl ops err
Btrfs: allow partial ordered extent completion
Btrfs: convert all bug_ons in free-space-cache.c
Btrfs: add support for asserts
Btrfs: adjust the fs_devices->missing count on unmount
Btrf: cleanup: don't check for root_refs == 0 twice
Btrfs: fix for patch "cleanup: don't check the same thing twice"
Btrfs: get rid of one BUG() in write_all_supers()
Btrfs: allocate prelim_ref with a slab allocater
Btrfs: pass gfp_t to __add_prelim_ref() to avoid always using GFP_ATOMIC
Btrfs: fix race conditions in BTRFS_IOC_FS_INFO ioctl
Btrfs: fix race between removing a dev and writing sbs
Btrfs: remove ourselves from the cluster list under lock
...
The sequence lock (seqlock) was originally designed for the cases where
the readers do not need to block the writers by making the readers retry
the read operation when the data change.
Since then, the use cases have been expanded to include situations where
a thread does not need to change the data (effectively a reader) at all
but have to take the writer lock because it can't tolerate changes to
the protected structure. Some examples are the d_path() function and
the getcwd() syscall in fs/dcache.c where the functions take the writer
lock on rename_lock even though they don't need to change anything in
the protected data structure at all. This is inefficient as a reader is
now blocking other sequence number reading readers from moving forward
by pretending to be a writer.
This patch tries to eliminate this inefficiency by introducing a new
type of locking reader to the seqlock locking mechanism. This new
locking reader will try to take an exclusive lock preventing other
writers and locking readers from going forward. However, it won't
affect the progress of the other sequence number reading readers as the
sequence number won't be changed.
Signed-off-by: Waiman Long <Waiman.Long@hp.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull thermal management updates from Zhang Rui:
"We have a lot of SOC changes and a few thermal core fixes this time.
The biggest change is about exynos thermal driver restructure. The
patch set adds TMU (Thermal management Unit) driver support for
exynos5440 platform. There are 3 instances of the TMU controllers so
necessary cleanup/re-structure is done to handle multiple thermal
zone.
The next biggest change is the introduction of the imx thermal driver.
It adds the imx thermal support using Temperature Monitor (TEMPMON)
block found on some Freescale i.MX SoCs. The driver uses syscon
regmap interface to access TEMPMON control registers and calibration
data, and supports cpufreq as the cooling device.
Highlights:
- restructure exynos thermal driver.
- introduce new imx thermal driver.
- fix a bug in thermal core, which powers on the fans unexpectedly
after resume from suspend"
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/rzhang/linux: (46 commits)
drivers: thermal: add check when unregistering cpu cooling
thermal: thermal_core: allow binding with limits on bind_params
drivers: thermal: make usage of CONFIG_THERMAL_HWMON optional
drivers: thermal: parent virtual hwmon with thermal zone
thermal: hwmon: move hwmon support to single file
thermal: exynos: Clean up non-DT remnants
thermal: exynos: Fix potential NULL pointer dereference
thermal: exynos: Fix typos in Kconfig
thermal: ti-soc-thermal: Ensure to compute thermal trend
thermal: ti-soc-thermal: Set the bandgap mask counter delay value
thermal: ti-soc-thermal: Initialize counter_delay field for TI DRA752 sensors
thermal: step_wise: return instance->target by default
thermal: step_wise: cdev only needs update on a new target state
Thermal/cpu_cooling: Return directly for the cpu out of allowed_cpus in the cpufreq_thermal_notifier()
thermal: exynos_tmu: fix wrong error check for mapped memory
thermal: imx: implement thermal alarm interrupt handling
thermal: imx: dynamic passive and SoC specific critical trip points
Documentation: thermal: Explain the exynos thermal driver model
ARM: dts: thermal: exynos: Add documentation for Exynos SoC thermal bindings
thermal: exynos: Support for TMU regulator defined at device tree
...
Pull CIFS fixes from Steve French:
"CIFS update including case insensitive file name matching improvements
for UTF-8 to Unicode, various small cifs fixes, SMB2/SMB3 leasing
improvements, support for following SMB2 symlinks, SMB3 packet signing
improvements"
* 'for-next' of git://git.samba.org/sfrench/cifs-2.6: (25 commits)
CIFS: Respect epoch value from create lease context v2
CIFS: Add create lease v2 context for SMB3
CIFS: Move parsing lease buffer to ops struct
CIFS: Move creating lease buffer to ops struct
CIFS: Store lease state itself rather than a mapped oplock value
CIFS: Replace clientCanCache* bools with an integer
[CIFS] quiet sparse compile warning
cifs: Start using per session key for smb2/3 for signature generation
cifs: Add a variable specific to NTLMSSP for key exchange.
cifs: Process post session setup code in respective dialect functions.
CIFS: convert to use le32_add_cpu()
CIFS: Fix missing lease break
CIFS: Fix a memory leak when a lease break comes
cifs: add winucase_convert.pl to Documentation/ directory
cifs: convert case-insensitive dentry ops to use new case conversion routines
cifs: add new case-insensitive conversion routines that are based on wchar_t's
[CIFS] Add Scott to list of cifs contributors
cifs: Move and expand MAX_SERVER_SIZE definition
cifs: Expand max share name length to 256
cifs: Move string length definitions to uapi
...
Merge first patch-bomb from Andrew Morton:
- Some pidns/fork/exec tweaks
- OCFS2 updates
- Most of MM - there remain quite a few memcg parts which depend on
pending core cgroups changes. Which might have been already merged -
I'll check tomorrow...
- Various misc stuff all over the place
- A few block bits which I never got around to sending to Jens -
relatively minor things.
- MAINTAINERS maintenance
- A small number of lib/ updates
- checkpatch updates
- epoll
- firmware/dmi-scan
- Some kprobes work for S390
- drivers/rtc updates
- hfsplus feature work
- vmcore feature work
- rbtree upgrades
- AOE updates
- pktcdvd cleanups
- PPS
- memstick
- w1
- New "inittmpfs" feature, which does the obvious
- More IPC work from Davidlohr.
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (303 commits)
lz4: fix compression/decompression signedness mismatch
ipc: drop ipc_lock_check
ipc, shm: drop shm_lock_check
ipc: drop ipc_lock_by_ptr
ipc, shm: guard against non-existant vma in shmdt(2)
ipc: document general ipc locking scheme
ipc,msg: drop msg_unlock
ipc: rename ids->rw_mutex
ipc,shm: shorten critical region for shmat
ipc,shm: cleanup do_shmat pasta
ipc,shm: shorten critical region for shmctl
ipc,shm: make shmctl_nolock lockless
ipc,shm: introduce shmctl_nolock
ipc: drop ipcctl_pre_down
ipc,shm: shorten critical region in shmctl_down
ipc,shm: introduce lockless functions to obtain the ipc object
initmpfs: use initramfs if rootfstype= or root= specified
initmpfs: make rootfs use tmpfs when CONFIG_TMPFS enabled
initmpfs: move rootfs code from fs/ramfs/ to init/
initmpfs: move bdi setup from init_rootfs to init_ramfs
...
LZ4 compression and decompression functions require different in
signedness input/output parameters: unsigned char for compression and
signed char for decompression.
Change decompression API to require "(const) unsigned char *".
Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Kyungsik Lee <kyungsik.lee@lge.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Yann Collet <yann.collet.73@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since in some situations the lock can be shared for readers, we shouldn't
be calling it a mutex, rename it to rwsem.
Signed-off-by: Davidlohr Bueso <davidlohr.bueso@hp.com>
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Manfred Spraul <manfred@colorfullife.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When the rootfs code was a wrapper around ramfs, having them in the same
file made sense. Now that it can wrap another filesystem type, move it in
with the init code instead.
This also allows a subsequent patch to access rootfstype= command line
arg.
Signed-off-by: Rob Landley <rob@landley.net>
Cc: Jeff Layton <jlayton@redhat.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Stephen Warren <swarren@nvidia.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jim Cromie <jim.cromie@gmail.com>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
With users of radix_tree_preload() run from interrupt (block/blk-ioc.c is
one such possible user), the following race can happen:
radix_tree_preload()
...
radix_tree_insert()
radix_tree_node_alloc()
if (rtp->nr) {
ret = rtp->nodes[rtp->nr - 1];
<interrupt>
...
radix_tree_preload()
...
radix_tree_insert()
radix_tree_node_alloc()
if (rtp->nr) {
ret = rtp->nodes[rtp->nr - 1];
And we give out one radix tree node twice. That clearly results in radix
tree corruption with different results (usually OOPS) depending on which
two users of radix tree race.
We fix the problem by making radix_tree_node_alloc() always allocate fresh
radix tree nodes when in interrupt. Using preloading when in interrupt
doesn't make sense since all the allocations have to be atomic anyway and
we cannot steal nodes from process-context users because some users rely
on radix_tree_insert() succeeding after radix_tree_preload().
in_interrupt() check is somewhat ugly but we cannot simply key off passed
gfp_mask as that is acquired from root_gfp_mask() and thus the same for
all preload users.
Another part of the fix is to avoid node preallocation in
radix_tree_preload() when passed gfp_mask doesn't allow waiting. Again,
preallocation in such case doesn't make sense and when preallocation would
happen in interrupt we could possibly leak some allocated nodes. However,
some users of radix_tree_preload() require following radix_tree_insert()
to succeed. To avoid unexpected effects for these users,
radix_tree_preload() only warns if passed gfp mask doesn't allow waiting
and we provide a new function radix_tree_maybe_preload() for those users
which get different gfp mask from different call sites and which are
prepared to handle radix_tree_insert() failure.
Signed-off-by: Jan Kara <jack@suse.cz>
Cc: Jens Axboe <jaxboe@fusionio.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Because deletion (of the entire tree) is a relatively common use of the
rbtree_postorder iteration, and because doing it safely means fiddling
with temporary storage, provide a helper to simplify postorder rbtree
iteration.
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Reviewed-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Cc: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Michel Lespinasse <walken@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Postorder iteration yields all of a node's children prior to yielding the
node itself, and this particular implementation also avoids examining the
leaf links in a node after that node has been yielded.
In what I expect will be its most common usage, postorder iteration allows
the deletion of every node in an rbtree without modifying the rbtree nodes
(no _requirement_ that they be nulled) while avoiding referencing child
nodes after they have been "deleted" (most commonly, freed).
I have only updated zswap to use this functionality at this point, but
numerous bits of code (most notably in the filesystem drivers) use a hand
rolled postorder iteration that NULLs child links as it traverses the
tree. Each of those instances could be replaced with this common
implementation.
1 & 2 add rbtree postorder iteration functions.
3 adds testing of the iteration to the rbtree runtime tests
4 allows building the rbtree runtime tests as builtins
5 updates zswap.
This patch:
Add postorder iteration functions for rbtree. These are useful for safely
freeing an entire rbtree without modifying the tree at all.
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Reviewed-by: Seth Jennings <sjenning@linux.vnet.ibm.com>
Cc: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Michel Lespinasse <walken@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For zfcpdump we can't map the HSA storage because it is only available via
a read interface. Therefore, for the new vmcore mmap feature we have
introduce a new mechanism to create mappings on demand.
This patch introduces a new architecture function remap_oldmem_pfn_range()
that should be used to create mappings with remap_pfn_range() for oldmem
areas that can be directly mapped. For zfcpdump this is everything
besides of the HSA memory. For the areas that are not mapped by
remap_oldmem_pfn_range() a generic vmcore a new generic vmcore fault
handler mmap_vmcore_fault() is called.
This handler works as follows:
* Get already available or new page from page cache (find_or_create_page)
* Check if /proc/vmcore page is filled with data (PageUptodate)
* If yes:
Return that page
* If no:
Fill page using __vmcore_read(), set PageUptodate, and return page
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com>
Cc: Jan Willeke <willeke@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For s390 we want to use /proc/vmcore for our SCSI stand-alone dump
(zfcpdump). We have support where the first HSA_SIZE bytes are saved into
a hypervisor owned memory area (HSA) before the kdump kernel is booted.
When the kdump kernel starts, it is restricted to use only HSA_SIZE bytes.
The advantages of this mechanism are:
* No crashkernel memory has to be defined in the old kernel.
* Early boot problems (before kexec_load has been done) can be dumped
* Non-Linux systems can be dumped.
We modify the s390 copy_oldmem_page() function to read from the HSA memory
if memory below HSA_SIZE bytes is requested.
Since we cannot use the kexec tool to load the kernel in this scenario,
we have to build the ELF header in the 2nd (kdump/new) kernel.
So with the following patch set we would like to introduce the new
function that the ELF header for /proc/vmcore can be created in the 2nd
kernel memory.
The following steps are done during zfcpdump execution:
1. Production system crashes
2. User boots a SCSI disk that has been prepared with the zfcpdump tool
3. Hypervisor saves CPU state of boot CPU and HSA_SIZE bytes of memory into HSA
4. Boot loader loads kernel into low memory area
5. Kernel boots and uses only HSA_SIZE bytes of memory
6. Kernel saves registers of non-boot CPUs
7. Kernel does memory detection for dump memory map
8. Kernel creates ELF header for /proc/vmcore
9. /proc/vmcore uses this header for initialization
10. The zfcpdump user space reads /proc/vmcore to write dump to SCSI disk
- copy_oldmem_page() copies from HSA for memory below HSA_SIZE
- copy_oldmem_page() copies from real memory for memory above HSA_SIZE
Currently for s390 we create the ELF core header in the 2nd kernel with a
small trick. We relocate the addresses in the ELF header in a way that
for the /proc/vmcore code it seems to be in the 1st kernel (old) memory
and the read_from_oldmem() returns the correct data. This allows the
/proc/vmcore code to use the ELF header in the 2nd kernel.
This patch:
Exchange the old mechanism with the new and much cleaner function call
override feature that now offcially allows to create the ELF core header
in the 2nd kernel.
To use the new feature the following function have to be defined
by the architecture backend code to read from new memory:
* elfcorehdr_alloc: Allocate ELF header
* elfcorehdr_free: Free the memory of the ELF header
* elfcorehdr_read: Read from ELF header
* elfcorehdr_read_notes: Read from ELF notes
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com>
Cc: Jan Willeke <willeke@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Nobody except search_binary_handler() should touch ->recursion_depth, "int
depth" buys nothing but complicates the code, kill it.
Probably we should also kill "fn" and the !NULL check, ->load_binary
should be always defined. And it can not go away after read_unlock() or
this code is buggy anyway.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Kees Cook <keescook@chromium.org>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Evgeniy Polyakov <zbr@ioremap.net>
Cc: Zach Levis <zml@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The current two insn slot caches both use module_alloc/module_free to
allocate and free insn slot cache pages.
For s390 this is not sufficient since there is the need to allocate insn
slots that are either within the vmalloc module area or within dma memory.
Therefore add a mechanism which allows to specify an own allocator for an
own insn slot cache.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The current kpropes insn caches allocate memory areas for insn slots
with module_alloc(). The assumption is that the kernel image and module
area are both within the same +/- 2GB memory area.
This however is not true for s390 where the kernel image resides within
the first 2GB (DMA memory area), but the module area is far away in the
vmalloc area, usually somewhere close below the 4TB area.
For new pc relative instructions s390 needs insn slots that are within
+/- 2GB of each area. That way we can patch displacements of
pc-relative instructions within the insn slots just like x86 and
powerpc.
The module area works already with the normal insn slot allocator,
however there is currently no way to get insn slots that are within the
first 2GB on s390 (aka DMA area).
Therefore this patch set modifies the kprobes insn slot cache code in
order to allow to specify a custom allocator for the insn slot cache
pages. In addition architecure can now have private insn slot caches
withhout the need to modify common code.
Patch 1 unifies and simplifies the current insn and optinsn caches
implementation. This is a preparation which allows to add more
insn caches in a simple way.
Patch 2 adds the possibility to specify a custom allocator.
Patch 3 makes s390 use the new insn slot mechanisms and adds support for
pc-relative instructions with long displacements.
This patch (of 3):
The two insn caches (insn, and optinsn) each have an own mutex and
alloc/free functions (get_[opt]insn_slot() / free_[opt]insn_slot()).
Since there is the need for yet another insn cache which satifies dma
allocations on s390, unify and simplify the current implementation:
- Move the per insn cache mutex into struct kprobe_insn_cache.
- Move the alloc/free functions to kprobe.h so they are simply
wrappers for the generic __get_insn_slot/__free_insn_slot functions.
The implementation is done with a DEFINE_INSN_CACHE_OPS() macro
which provides the alloc/free functions for each cache if needed.
- move the struct kprobe_insn_cache to kprobe.h which allows to generate
architecture specific insn slot caches outside of the core kprobes
code.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Unclutter -Wmissing-prototypes warning types (enabled at make W=1)
linux/include/linux/syscalls.h:190:18: warning: no previous prototype for 'SyS_semctl' [-Wmissing-prototypes]
asmlinkage long SyS##name(__MAP(x,__SC_LONG,__VA_ARGS__)) \
^
linux/include/linux/syscalls.h:183:2: note: in expansion of macro '__SYSCALL_DEFINEx'
__SYSCALL_DEFINEx(x, sname, __VA_ARGS__)
^
by adding forward declarations right before definitions.
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
All of the other non-trivial !SMP versions of functions in smp.h are
out-of-line in up.c. Move on_each_cpu() there as well.
This allows us to get rid of the #include <linux/irqflags.h>. The
drawback is that this makes both the x86_64 and i386 defconfig !SMP
kernels about 200 bytes larger each.
Signed-off-by: David Daney <david.daney@cavium.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
As in commit f21afc25f9 ("smp.h: Use local_irq_{save,restore}() in
!SMP version of on_each_cpu()"), we don't want to enable irqs if they
are not already enabled. There are currently no known problematical
callers of these functions, but since it is a known failure pattern, we
preemptively fix them.
Since they are not trivial functions, make them non-inline by moving
them to up.c. This also makes it so we don't have to fix #include
dependancies for preempt_{disable,enable}.
Signed-off-by: David Daney <david.daney@cavium.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
compound lock is introduced by commit e9da73d67("thp: compound_lock."), it
is used to serialize put_page against __split_huge_page_refcount(). In
addition, transparent hugepages will be splitted in hwpoison handler and
just one subpage will be poisoned. There is unnecessary to hold compound
lock for hugetlbfs page. This patch replace compound_trans_order by
compond_order in the place where the page is hugetlbfs page.
Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Reviewed-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Tony Luck <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The feature prevents mistrusted filesystems (ie: FUSE mounts created by
unprivileged users) to grow a large number of dirty pages before
throttling. For such filesystems balance_dirty_pages always check bdi
counters against bdi limits. I.e. even if global "nr_dirty" is under
"freerun", it's not allowed to skip bdi checks. The only use case for now
is fuse: it sets bdi max_ratio to 1% by default and system administrators
are supposed to expect that this limit won't be exceeded.
The feature is on if a BDI is marked by BDI_CAP_STRICTLIMIT flag. A
filesystem may set the flag when it initializes its BDI.
The problematic scenario comes from the fact that nobody pays attention to
the NR_WRITEBACK_TEMP counter (i.e. number of pages under fuse
writeback). The implementation of fuse writeback releases original page
(by calling end_page_writeback) almost immediately. A fuse request queued
for real processing bears a copy of original page. Hence, if userspace
fuse daemon doesn't finalize write requests in timely manner, an
aggressive mmap writer can pollute virtually all memory by those temporary
fuse page copies. They are carefully accounted in NR_WRITEBACK_TEMP, but
nobody cares.
To make further explanations shorter, let me use "NR_WRITEBACK_TEMP
problem" as a shortcut for "a possibility of uncontrolled grow of amount
of RAM consumed by temporary pages allocated by kernel fuse to process
writeback".
The problem was very easy to reproduce. There is a trivial example
filesystem implementation in fuse userspace distribution: fusexmp_fh.c. I
added "sleep(1);" to the write methods, then recompiled and mounted it.
Then created a huge file on the mount point and run a simple program which
mmap-ed the file to a memory region, then wrote a data to the region. An
hour later I observed almost all RAM consumed by fuse writeback. Since
then some unrelated changes in kernel fuse made it more difficult to
reproduce, but it is still possible now.
Putting this theoretical happens-in-the-lab thing aside, there is another
thing that really hurts real world (FUSE) users. This is write-through
page cache policy FUSE currently uses. I.e. handling write(2), kernel
fuse populates page cache and flushes user data to the server
synchronously. This is excessively suboptimal. Pavel Emelyanov's patches
("writeback cache policy") solve the problem, but they also make resolving
NR_WRITEBACK_TEMP problem absolutely necessary. Otherwise, simply copying
a huge file to a fuse mount would result in memory starvation. Miklos,
the maintainer of FUSE, believes strictlimit feature the way to go.
And eventually putting FUSE topics aside, there is one more use-case for
strictlimit feature. Using a slow USB stick (mass storage) in a machine
with huge amount of RAM installed is a well-known pain. Let's make simple
computations. Assuming 64GB of RAM installed, existing implementation of
balance_dirty_pages will start throttling only after 9.6GB of RAM becomes
dirty (freerun == 15% of total RAM). So, the command "cp 9GB_file
/media/my-usb-storage/" may return in a few seconds, but subsequent
"umount /media/my-usb-storage/" will take more than two hours if effective
throughput of the storage is, to say, 1MB/sec.
After inclusion of strictlimit feature, it will be trivial to add a knob
(e.g. /sys/devices/virtual/bdi/x:y/strictlimit) to enable it on demand.
Manually or via udev rule. May be I'm wrong, but it seems to be quite a
natural desire to limit the amount of dirty memory for some devices we are
not fully trust (in the sense of sustainable throughput).
[akpm@linux-foundation.org: fix warning in page-writeback.c]
Signed-off-by: Maxim Patlasov <MPatlasov@parallels.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Miklos Szeredi <miklos@szeredi.hu>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Pavel Emelyanov <xemul@parallels.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It's not used globally and could be static.
Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: David Rientjes <rientjes@google.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch is based on KOSAKI's work and I add a little more description,
please refer https://lkml.org/lkml/2012/6/14/74.
Currently, I found system can enter a state that there are lots of free
pages in a zone but only order-0 and order-1 pages which means the zone is
heavily fragmented, then high order allocation could make direct reclaim
path's long stall(ex, 60 seconds) especially in no swap and no compaciton
enviroment. This problem happened on v3.4, but it seems issue still lives
in current tree, the reason is do_try_to_free_pages enter live lock:
kswapd will go to sleep if the zones have been fully scanned and are still
not balanced. As kswapd thinks there's little point trying all over again
to avoid infinite loop. Instead it changes order from high-order to
0-order because kswapd think order-0 is the most important. Look at
73ce02e9 in detail. If watermarks are ok, kswapd will go back to sleep
and may leave zone->all_unreclaimable =3D 0. It assume high-order users
can still perform direct reclaim if they wish.
Direct reclaim continue to reclaim for a high order which is not a
COSTLY_ORDER without oom-killer until kswapd turn on
zone->all_unreclaimble= . This is because to avoid too early oom-kill.
So it means direct_reclaim depends on kswapd to break this loop.
In worst case, direct-reclaim may continue to page reclaim forever when
kswapd sleeps forever until someone like watchdog detect and finally kill
the process. As described in:
http://thread.gmane.org/gmane.linux.kernel.mm/103737
We can't turn on zone->all_unreclaimable from direct reclaim path because
direct reclaim path don't take any lock and this way is racy. Thus this
patch removes zone->all_unreclaimable field completely and recalculates
zone reclaimable state every time.
Note: we can't take the idea that direct-reclaim see zone->pages_scanned
directly and kswapd continue to use zone->all_unreclaimable. Because, it
is racy. commit 929bea7c71 (vmscan: all_unreclaimable() use
zone->all_unreclaimable as a name) describes the detail.
[akpm@linux-foundation.org: uninline zone_reclaimable_pages() and zone_reclaimable()]
Cc: Aaditya Kumar <aaditya.kumar.30@gmail.com>
Cc: Ying Han <yinghan@google.com>
Cc: Nick Piggin <npiggin@gmail.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Bob Liu <lliubbo@gmail.com>
Cc: Neil Zhang <zhangwm@marvell.com>
Cc: Russell King - ARM Linux <linux@arm.linux.org.uk>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Minchan Kim <minchan@kernel.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Lisa Du <cldu@marvell.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently munlock_vma_pages_range() calls follow_page_mask() to obtain
each individual struct page. This entails repeated full page table
translations and page table lock taken for each page separately.
This patch avoids the costly follow_page_mask() where possible, by
iterating over ptes within single pmd under single page table lock. The
first pte is obtained by get_locked_pte() for non-THP page acquired by the
initial follow_page_mask(). The rest of the on-stack pagevec for munlock
is filled up using pte_walk as long as pte_present() and vm_normal_page()
are sufficient to obtain the struct page.
After this patch, a 14% speedup was measured for munlocking a 56GB large
memory area with THP disabled.
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Jörn Engel <joern@logfs.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michel Lespinasse <walken@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pavel reported that in case if vma area get unmapped and then mapped (or
expanded) in-place, the soft dirty tracker won't be able to recognize this
situation since it works on pte level and ptes are get zapped on unmap,
loosing soft dirty bit of course.
So to resolve this situation we need to track actions on vma level, there
VM_SOFTDIRTY flag comes in. When new vma area created (or old expanded)
we set this bit, and keep it here until application calls for clearing
soft dirty bit.
Thus when user space application track memory changes now it can detect if
vma area is renewed.
Reported-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Matt Mackall <mpm@selenic.com>
Cc: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@gmail.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Cc: Rob Landley <rob@landley.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Current early_pfn_to_nid() on arch that support memblock go over
memblock.memory one by one, so will take too many try near the end.
We can use existing memblock_search to find the node id for given pfn,
that could save some time on bigger system that have many entries
memblock.memory array.
Here are the timing differences for several machines. In each case with
the patch less time was spent in __early_pfn_to_nid().
3.11-rc5 with patch difference (%)
-------- ---------- --------------
UV1: 256 nodes 9TB: 411.66 402.47 -9.19 (2.23%)
UV2: 255 nodes 16TB: 1141.02 1138.12 -2.90 (0.25%)
UV2: 64 nodes 2TB: 128.15 126.53 -1.62 (1.26%)
UV2: 32 nodes 2TB: 121.87 121.07 -0.80 (0.66%)
Time in seconds.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Cc: Tejun Heo <tj@kernel.org>
Acked-by: Russ Anderson <rja@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>