This reverts commit 26e7deadaa.
Sonny reported that one of their tests started failing on the latest
kernel on their Chrome OS platform. The root cause is that the above
commit removed the protection line of empty zone, while the parser used in
the test relies on the protection line to mark the end of each zone.
Let's revert it to avoid breaking userspace testing or applications.
Fixes: 26e7deadaa ("mm/vmstat.c: do not show lowmem reserve protection information of empty zone)"
Reported-by: Sonny Rao <[email protected]>
Signed-off-by: Baoquan He <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: David Hildenbrand <[email protected]>
Acked-by: David Rientjes <[email protected]>
Cc: <[email protected]> [5.8.x]
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
For some applications, we need to allocate almost all memory as hugepages.
However, on a running system, higher-order allocations can fail if the
memory is fragmented. Linux kernel currently does on-demand compaction as
we request more hugepages, but this style of compaction incurs very high
latency. Experiments with one-time full memory compaction (followed by
hugepage allocations) show that kernel is able to restore a highly
fragmented memory state to a fairly compacted memory state within <1 sec
for a 32G system. Such data suggests that a more proactive compaction can
help us allocate a large fraction of memory as hugepages keeping
allocation latencies low.
For a more proactive compaction, the approach taken here is to define a
new sysctl called 'vm.compaction_proactiveness' which dictates bounds for
external fragmentation which kcompactd tries to maintain.
The tunable takes a value in range [0, 100], with a default of 20.
Note that a previous version of this patch [1] was found to introduce too
many tunables (per-order extfrag{low, high}), but this one reduces them to
just one sysctl. Also, the new tunable is an opaque value instead of
asking for specific bounds of "external fragmentation", which would have
been difficult to estimate. The internal interpretation of this opaque
value allows for future fine-tuning.
Currently, we use a simple translation from this tunable to [low, high]
"fragmentation score" thresholds (low=100-proactiveness, high=low+10%).
The score for a node is defined as weighted mean of per-zone external
fragmentation. A zone's present_pages determines its weight.
To periodically check per-node score, we reuse per-node kcompactd threads,
which are woken up every 500 milliseconds to check the same. If a node's
score exceeds its high threshold (as derived from user-provided
proactiveness value), proactive compaction is started until its score
reaches its low threshold value. By default, proactiveness is set to 20,
which implies threshold values of low=80 and high=90.
This patch is largely based on ideas from Michal Hocko [2]. See also the
LWN article [3].
Performance data
================
System: x64_64, 1T RAM, 80 CPU threads.
Kernel: 5.6.0-rc3 + this patch
echo madvise | sudo tee /sys/kernel/mm/transparent_hugepage/enabled
echo madvise | sudo tee /sys/kernel/mm/transparent_hugepage/defrag
Before starting the driver, the system was fragmented from a userspace
program that allocates all memory and then for each 2M aligned section,
frees 3/4 of base pages using munmap. The workload is mainly anonymous
userspace pages, which are easy to move around. I intentionally avoided
unmovable pages in this test to see how much latency we incur when
hugepage allocations hit direct compaction.
1. Kernel hugepage allocation latencies
With the system in such a fragmented state, a kernel driver then allocates
as many hugepages as possible and measures allocation latency:
(all latency values are in microseconds)
- With vanilla 5.6.0-rc3
percentile latency
–––––––––– –––––––
5 7894
10 9496
25 12561
30 15295
40 18244
50 21229
60 27556
75 30147
80 31047
90 32859
95 33799
Total 2M hugepages allocated = 383859 (749G worth of hugepages out of 762G
total free => 98% of free memory could be allocated as hugepages)
- With 5.6.0-rc3 + this patch, with proactiveness=20
sysctl -w vm.compaction_proactiveness=20
percentile latency
–––––––––– –––––––
5 2
10 2
25 3
30 3
40 3
50 4
60 4
75 4
80 4
90 5
95 429
Total 2M hugepages allocated = 384105 (750G worth of hugepages out of 762G
total free => 98% of free memory could be allocated as hugepages)
2. JAVA heap allocation
In this test, we first fragment memory using the same method as for (1).
Then, we start a Java process with a heap size set to 700G and request the
heap to be allocated with THP hugepages. We also set THP to madvise to
allow hugepage backing of this heap.
/usr/bin/time
java -Xms700G -Xmx700G -XX:+UseTransparentHugePages -XX:+AlwaysPreTouch
The above command allocates 700G of Java heap using hugepages.
- With vanilla 5.6.0-rc3
17.39user 1666.48system 27:37.89elapsed
- With 5.6.0-rc3 + this patch, with proactiveness=20
8.35user 194.58system 3:19.62elapsed
Elapsed time remains around 3:15, as proactiveness is further increased.
Note that proactive compaction happens throughout the runtime of these
workloads. The situation of one-time compaction, sufficient to supply
hugepages for following allocation stream, can probably happen for more
extreme proactiveness values, like 80 or 90.
In the above Java workload, proactiveness is set to 20. The test starts
with a node's score of 80 or higher, depending on the delay between the
fragmentation step and starting the benchmark, which gives more-or-less
time for the initial round of compaction. As t he benchmark consumes
hugepages, node's score quickly rises above the high threshold (90) and
proactive compaction starts again, which brings down the score to the low
threshold level (80). Repeat.
bpftrace also confirms proactive compaction running 20+ times during the
runtime of this Java benchmark. kcompactd threads consume 100% of one of
the CPUs while it tries to bring a node's score within thresholds.
Backoff behavior
================
Above workloads produce a memory state which is easy to compact. However,
if memory is filled with unmovable pages, proactive compaction should
essentially back off. To test this aspect:
- Created a kernel driver that allocates almost all memory as hugepages
followed by freeing first 3/4 of each hugepage.
- Set proactiveness=40
- Note that proactive_compact_node() is deferred maximum number of times
with HPAGE_FRAG_CHECK_INTERVAL_MSEC of wait between each check
(=> ~30 seconds between retries).
[1] https://patchwork.kernel.org/patch/11098289/
[2] https://lore.kernel.org/linux-mm/[email protected]/
[3] https://lwn.net/Articles/817905/
Signed-off-by: Nitin Gupta <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Tested-by: Oleksandr Natalenko <[email protected]>
Reviewed-by: Vlastimil Babka <[email protected]>
Reviewed-by: Khalid Aziz <[email protected]>
Reviewed-by: Oleksandr Natalenko <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Cc: Khalid Aziz <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: Mike Kravetz <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Cc: David Rientjes <[email protected]>
Cc: Nitin Gupta <[email protected]>
Cc: Oleksandr Natalenko <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
Currently the kernel stack is being accounted per-zone. There is no need
to do that. In addition due to being per-zone, memcg has to keep a
separate MEMCG_KERNEL_STACK_KB. Make the stat per-node and deprecate
MEMCG_KERNEL_STACK_KB as memcg_stat_item is an extension of
node_stat_item. In addition localize the kernel stack stats updates to
account_kernel_stack().
Signed-off-by: Shakeel Butt <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: Roman Gushchin <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
To implement per-object slab memory accounting, we need to convert slab
vmstat counters to bytes. Actually, out of 4 levels of counters: global,
per-node, per-memcg and per-lruvec only two last levels will require
byte-sized counters. It's because global and per-node counters will be
counting the number of slab pages, and per-memcg and per-lruvec will be
counting the amount of memory taken by charged slab objects.
Converting all vmstat counters to bytes or even all slab counters to bytes
would introduce an additional overhead. So instead let's store global and
per-node counters in pages, and memcg and lruvec counters in bytes.
To make the API clean all access helpers (both on the read and write
sides) are dealing with bytes.
To avoid back-and-forth conversions a new flavor of read-side helpers is
introduced, which always returns values in pages: node_page_state_pages()
and global_node_page_state_pages().
Actually new helpers are just reading raw values. Old helpers are simple
wrappers, which will complain on an attempt to read byte value, because at
the moment no one actually needs bytes.
Thanks to Johannes Weiner for the idea of having the byte-sized API on top
of the page-sized internal storage.
Signed-off-by: Roman Gushchin <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: Vlastimil Babka <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
Cc: Christoph Lameter <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Tejun Heo <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
Merge more updates from Andrew Morton:
"More mm/ work, plenty more to come
Subsystems affected by this patch series: slub, memcg, gup, kasan,
pagealloc, hugetlb, vmscan, tools, mempolicy, memblock, hugetlbfs,
thp, mmap, kconfig"
* akpm: (131 commits)
arm64: mm: use ARCH_HAS_DEBUG_WX instead of arch defined
x86: mm: use ARCH_HAS_DEBUG_WX instead of arch defined
riscv: support DEBUG_WX
mm: add DEBUG_WX support
drivers/base/memory.c: cache memory blocks in xarray to accelerate lookup
mm/thp: rename pmd_mknotpresent() as pmd_mkinvalid()
powerpc/mm: drop platform defined pmd_mknotpresent()
mm: thp: don't need to drain lru cache when splitting and mlocking THP
hugetlbfs: get unmapped area below TASK_UNMAPPED_BASE for hugetlbfs
sparc32: register memory occupied by kernel as memblock.memory
include/linux/memblock.h: fix minor typo and unclear comment
mm, mempolicy: fix up gup usage in lookup_node
tools/vm/page_owner_sort.c: filter out unneeded line
mm: swap: memcg: fix memcg stats for huge pages
mm: swap: fix vmstats for huge pages
mm: vmscan: limit the range of LRU type balancing
mm: vmscan: reclaim writepage is IO cost
mm: vmscan: determine anon/file pressure balance at the reclaim root
mm: balance LRU lists based on relative thrashing
mm: only count actual rotations as LRU reclaim cost
...
Having statistics on pages scanned and pages reclaimed for both anon and
file pages makes it easier to evaluate changes to LRU balancing.
While at it, clean up the stat-keeping mess for isolation, putback,
reclaim stats etc. a bit: first the physical LRU operation (isolation and
putback), followed by vmstats, reclaim_stats, and then vm events.
Signed-off-by: Johannes Weiner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Rik van Riel <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
Because the lowmem reserve protection of a zone can't tell anything if the
zone is empty, except of adding one more line in /proc/zoneinfo.
Let's remove it from that zone's showing.
Signed-off-by: Baoquan He <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
Pull networking updates from David Miller:
1) Allow setting bluetooth L2CAP modes via socket option, from Luiz
Augusto von Dentz.
2) Add GSO partial support to igc, from Sasha Neftin.
3) Several cleanups and improvements to r8169 from Heiner Kallweit.
4) Add IF_OPER_TESTING link state and use it when ethtool triggers a
device self-test. From Andrew Lunn.
5) Start moving away from custom driver versions, use the globally
defined kernel version instead, from Leon Romanovsky.
6) Support GRO vis gro_cells in DSA layer, from Alexander Lobakin.
7) Allow hard IRQ deferral during NAPI, from Eric Dumazet.
8) Add sriov and vf support to hinic, from Luo bin.
9) Support Media Redundancy Protocol (MRP) in the bridging code, from
Horatiu Vultur.
10) Support netmap in the nft_nat code, from Pablo Neira Ayuso.
11) Allow UDPv6 encapsulation of ESP in the ipsec code, from Sabrina
Dubroca. Also add ipv6 support for espintcp.
12) Lots of ReST conversions of the networking documentation, from Mauro
Carvalho Chehab.
13) Support configuration of ethtool rxnfc flows in bcmgenet driver,
from Doug Berger.
14) Allow to dump cgroup id and filter by it in inet_diag code, from
Dmitry Yakunin.
15) Add infrastructure to export netlink attribute policies to
userspace, from Johannes Berg.
16) Several optimizations to sch_fq scheduler, from Eric Dumazet.
17) Fallback to the default qdisc if qdisc init fails because otherwise
a packet scheduler init failure will make a device inoperative. From
Jesper Dangaard Brouer.
18) Several RISCV bpf jit optimizations, from Luke Nelson.
19) Correct the return type of the ->ndo_start_xmit() method in several
drivers, it's netdev_tx_t but many drivers were using
'int'. From Yunjian Wang.
20) Add an ethtool interface for PHY master/slave config, from Oleksij
Rempel.
21) Add BPF iterators, from Yonghang Song.
22) Add cable test infrastructure, including ethool interfaces, from
Andrew Lunn. Marvell PHY driver is the first to support this
facility.
23) Remove zero-length arrays all over, from Gustavo A. R. Silva.
24) Calculate and maintain an explicit frame size in XDP, from Jesper
Dangaard Brouer.
25) Add CAP_BPF, from Alexei Starovoitov.
26) Support terse dumps in the packet scheduler, from Vlad Buslov.
27) Support XDP_TX bulking in dpaa2 driver, from Ioana Ciornei.
28) Add devm_register_netdev(), from Bartosz Golaszewski.
29) Minimize qdisc resets, from Cong Wang.
30) Get rid of kernel_getsockopt and kernel_setsockopt in order to
eliminate set_fs/get_fs calls. From Christoph Hellwig.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (2517 commits)
selftests: net: ip_defrag: ignore EPERM
net_failover: fixed rollback in net_failover_open()
Revert "tipc: Fix potential tipc_aead refcnt leak in tipc_crypto_rcv"
Revert "tipc: Fix potential tipc_node refcnt leak in tipc_rcv"
vmxnet3: allow rx flow hash ops only when rss is enabled
hinic: add set_channels ethtool_ops support
selftests/bpf: Add a default $(CXX) value
tools/bpf: Don't use $(COMPILE.c)
bpf, selftests: Use bpf_probe_read_kernel
s390/bpf: Use bcr 0,%0 as tail call nop filler
s390/bpf: Maintain 8-byte stack alignment
selftests/bpf: Fix verifier test
selftests/bpf: Fix sample_cnt shared between two threads
bpf, selftests: Adapt cls_redirect to call csum_level helper
bpf: Add csum_level helper for fixing up csum levels
bpf: Fix up bpf_skb_adjust_room helper's skb csum setting
sfc: add missing annotation for efx_ef10_try_update_nic_stats_vf()
crypto/chtls: IPv6 support for inline TLS
Crypto/chcr: Fixes a coccinile check error
Crypto/chcr: Fixes compilations warnings
...
Merge updates from Andrew Morton:
"A few little subsystems and a start of a lot of MM patches.
Subsystems affected by this patch series: squashfs, ocfs2, parisc,
vfs. With mm subsystems: slab-generic, slub, debug, pagecache, gup,
swap, memcg, pagemap, memory-failure, vmalloc, kasan"
* emailed patches from Andrew Morton <[email protected]>: (128 commits)
kasan: move kasan_report() into report.c
mm/mm_init.c: report kasan-tag information stored in page->flags
ubsan: entirely disable alignment checks under UBSAN_TRAP
kasan: fix clang compilation warning due to stack protector
x86/mm: remove vmalloc faulting
mm: remove vmalloc_sync_(un)mappings()
x86/mm/32: implement arch_sync_kernel_mappings()
x86/mm/64: implement arch_sync_kernel_mappings()
mm/ioremap: track which page-table levels were modified
mm/vmalloc: track which page-table levels were modified
mm: add functions to track page directory modifications
s390: use __vmalloc_node in stack_alloc
powerpc: use __vmalloc_node in alloc_vm_stack
arm64: use __vmalloc_node in arch_alloc_vmap_stack
mm: remove vmalloc_user_node_flags
mm: switch the test_vmalloc module to use __vmalloc_node
mm: remove __vmalloc_node_flags_caller
mm: remove both instances of __vmalloc_node_flags
mm: remove the prot argument to __vmalloc_node
mm: remove the pgprot argument to __vmalloc
...
After an NFS page has been written it is considered "unstable" until a
COMMIT request succeeds. If the COMMIT fails, the page will be
re-written.
These "unstable" pages are currently accounted as "reclaimable", either
in WB_RECLAIMABLE, or in NR_UNSTABLE_NFS which is included in a
'reclaimable' count. This might have made sense when sending the COMMIT
required a separate action by the VFS/MM (e.g. releasepage() used to
send a COMMIT). However now that all writes generated by ->writepages()
will automatically be followed by a COMMIT (since commit 919e3bd9a8
("NFS: Ensure we commit after writeback is complete")) it makes more
sense to treat them as writeback pages.
So this patch removes NR_UNSTABLE_NFS and accounts unstable pages in
NR_WRITEBACK and WB_WRITEBACK.
A particular effect of this change is that when
wb_check_background_flush() calls wb_over_bg_threshold(), the latter
will report 'true' a lot less often as the 'unstable' pages are no
longer considered 'dirty' (as there is nothing that writeback can do
about them anyway).
Currently wb_check_background_flush() will trigger writeback to NFS even
when there are relatively few dirty pages (if there are lots of unstable
pages), this can result in small writes going to the server (10s of
Kilobytes rather than a Megabyte) which hurts throughput. With this
patch, there are fewer writes which are each larger on average.
Where the NR_UNSTABLE_NFS count was included in statistics
virtual-files, the entry is retained, but the value is hard-coded as
zero. static trace points and warning printks which mentioned this
counter no longer report it.
[[email protected]: re-layout comment]
[[email protected]: fix printk warning]
Signed-off-by: NeilBrown <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: Jan Kara <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Acked-by: Trond Myklebust <[email protected]>
Acked-by: Michal Hocko <[email protected]> [mm]
Cc: Christoph Hellwig <[email protected]>
Cc: Chuck Lever <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
Instead of having all the sysctl handlers deal with user pointers, which
is rather hairy in terms of the BPF interaction, copy the input to and
from userspace in common code. This also means that the strings are
always NUL-terminated by the common code, making the API a little bit
safer.
As most handler just pass through the data to one of the common handlers
a lot of the changes are mechnical.
Signed-off-by: Christoph Hellwig <[email protected]>
Acked-by: Andrey Ignatov <[email protected]>
Signed-off-by: Al Viro <[email protected]>
The thp_fault_fallback and thp_file_fallback vmstats are incremented if
either the hugepage allocation fails through the page allocator or the
hugepage charge fails through mem cgroup.
This patch leaves this field untouched but adds two new fields,
thp_{fault,file}_fallback_charge, which is incremented only when the mem
cgroup charge fails.
This distinguishes between attempted hugepage allocations that fail due to
fragmentation (or low memory conditions) and those that fail due to mem
cgroup limits. That can be used to determine the impact of fragmentation
on the system by excluding faults that failed due to memcg usage.
Signed-off-by: David Rientjes <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: Yang Shi <[email protected]>
Acked-by: Kirill A. Shutemov <[email protected]>
Cc: Mike Rapoport <[email protected]>
Cc: Jeremy Cline <[email protected]>
Cc: Andrea Arcangeli <[email protected]>
Cc: Mike Kravetz <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
The existing thp_fault_fallback indicates when thp attempts to allocate a
hugepage but fails, or if the hugepage cannot be charged to the mem cgroup
hierarchy.
Extend this to shmem as well. Adds a new thp_file_fallback to complement
thp_file_alloc that gets incremented when a hugepage is attempted to be
allocated but fails, or if it cannot be charged to the mem cgroup
hierarchy.
Additionally, remove the check for CONFIG_TRANSPARENT_HUGE_PAGECACHE from
shmem_alloc_hugepage() since it is only called with this configuration
option.
Signed-off-by: David Rientjes <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: Yang Shi <[email protected]>
Acked-by: Kirill A. Shutemov <[email protected]>
Cc: Mike Rapoport <[email protected]>
Cc: Jeremy Cline <[email protected]>
Cc: Andrea Arcangeli <[email protected]>
Cc: Mike Kravetz <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
Now that pages are "DMA-pinned" via pin_user_page*(), and unpinned via
unpin_user_pages*(), we need some visibility into whether all of this is
working correctly.
Add two new fields to /proc/vmstat:
nr_foll_pin_acquired
nr_foll_pin_released
These are documented in Documentation/core-api/pin_user_pages.rst. They
represent the number of pages (since boot time) that have been pinned
("nr_foll_pin_acquired") and unpinned ("nr_foll_pin_released"), via
pin_user_pages*() and unpin_user_pages*().
In the absence of long-running DMA or RDMA operations that hold pages
pinned, the above two fields will normally be equal to each other.
Also: update Documentation/core-api/pin_user_pages.rst, to remove an
earlier (now confirmed untrue) claim about a performance problem with
/proc/vmstat.
Also: update Documentation/core-api/pin_user_pages.rst to rename the new
/proc/vmstat entries, to the names listed here.
Signed-off-by: John Hubbard <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: Jan Kara <[email protected]>
Acked-by: Kirill A. Shutemov <[email protected]>
Cc: Ira Weiny <[email protected]>
Cc: Jérôme Glisse <[email protected]>
Cc: "Matthew Wilcox (Oracle)" <[email protected]>
Cc: Al Viro <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Dan Williams <[email protected]>
Cc: Dave Chinner <[email protected]>
Cc: Jason Gunthorpe <[email protected]>
Cc: Jonathan Corbet <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Mike Kravetz <[email protected]>
Cc: Shuah Khan <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
Use common names from vmstat array when possible. This gives not much
difference in code size for now, but should help in keeping interfaces
consistent.
add/remove: 0/2 grow/shrink: 2/0 up/down: 70/-72 (-2)
Function old new delta
memory_stat_format 984 1050 +66
memcg_stat_show 957 961 +4
memcg1_event_names 32 - -32
mem_cgroup_lru_names 40 - -40
Total: Before=14485337, After=14485335, chg -0.00%
Link: http://lkml.kernel.org/r/157113012508.453.80391533767219371.stgit@buzz
Signed-off-by: Konstantin Khlebnikov <[email protected]>
Acked-by: Andrew Morton <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Statistics in vmstat is combined from counters with different structure,
but names for them are merged into one array.
This patch adds trivial helpers to get name for each item:
const char *zone_stat_name(enum zone_stat_item item);
const char *numa_stat_name(enum numa_stat_item item);
const char *node_stat_name(enum node_stat_item item);
const char *writeback_stat_name(enum writeback_stat_item item);
const char *vm_event_name(enum vm_event_item item);
Names for enum writeback_stat_item are folded in the middle of
vmstat_text so this patch moves declaration into header to calculate
offset of following items.
Also this patch reuses piece of node stat names for lru list names:
const char *lru_list_name(enum lru_list lru);
This returns common lru list names: "inactive_anon", "active_anon",
"inactive_file", "active_file", "unevictable".
[[email protected]: do not use size of vmstat_text as count of /proc/vmstat items]
Link: http://lkml.kernel.org/r/157152151769.4139.15423465513138349343.stgit@buzz
Link: https://lore.kernel.org/linux-mm/[email protected]/T/#u
Link: http://lkml.kernel.org/r/157113012325.453.562783073839432766.stgit@buzz
Signed-off-by: Konstantin Khlebnikov <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Cc: Randy Dunlap <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Vladimir Davydov <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: YueHaibing <[email protected]>
Cc: Stephen Rothwell <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
pagetypeinfo_showfree_print is called by zone->lock held in irq mode.
This is not really nice because it blocks both any interrupts on that
cpu and the page allocator. On large machines this might even trigger
the hard lockup detector.
Considering the pagetypeinfo is a debugging tool we do not really need
exact numbers here. The primary reason to look at the outuput is to see
how pageblocks are spread among different migratetypes and low number of
pages is much more interesting therefore putting a bound on the number
of pages on the free_list sounds like a reasonable tradeoff.
The new output will simply tell
[...]
Node 6, zone Normal, type Movable >100000 >100000 >100000 >100000 41019 31560 23996 10054 3229 983 648
instead of
Node 6, zone Normal, type Movable 399568 294127 221558 102119 41019 31560 23996 10054 3229 983 648
The limit has been chosen arbitrary and it is a subject of a future
change should there be a need for that.
While we are at it, also drop the zone lock after each free_list
iteration which will help with the IRQ and page allocator responsiveness
even further as the IRQ lock held time is always bound to those 100k
pages.
[[email protected]: tweak comment text, per David Hildenbrand]
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Michal Hocko <[email protected]>
Suggested-by: Andrew Morton <[email protected]>
Reviewed-by: Waiman Long <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Acked-by: David Hildenbrand <[email protected]>
Acked-by: Rafael Aquini <[email protected]>
Acked-by: David Rientjes <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Cc: Greg Kroah-Hartman <[email protected]>
Cc: Jann Horn <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Konstantin Khlebnikov <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Roman Gushchin <[email protected]>
Cc: Song Liu <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
/proc/pagetypeinfo is a debugging tool to examine internal page
allocator state wrt to fragmentation. It is not very useful for any
other use so normal users really do not need to read this file.
Waiman Long has noticed that reading this file can have negative side
effects because zone->lock is necessary for gathering data and that a)
interferes with the page allocator and its users and b) can lead to hard
lockups on large machines which have very long free_list.
Reduce both issues by simply not exporting the file to regular users.
Link: http://lkml.kernel.org/r/[email protected]
Fixes: 467c996c1e ("Print out statistics in relation to fragmentation avoidance to /proc/pagetypeinfo")
Signed-off-by: Michal Hocko <[email protected]>
Reported-by: Waiman Long <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Acked-by: Waiman Long <[email protected]>
Acked-by: Rafael Aquini <[email protected]>
Acked-by: David Rientjes <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Cc: David Hildenbrand <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Roman Gushchin <[email protected]>
Cc: Konstantin Khlebnikov <[email protected]>
Cc: Jann Horn <[email protected]>
Cc: Song Liu <[email protected]>
Cc: Greg Kroah-Hartman <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Add SPDX license identifiers to all files which:
- Have no license information of any form
- Have EXPORT_.*_SYMBOL_GPL inside which was used in the
initial scan/conversion to ignore the file
These files fall under the project license, GPL v2 only. The resulting SPDX
license identifier is:
GPL-2.0-only
Signed-off-by: Thomas Gleixner <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
totalram_pages, zone->managed_pages and totalhigh_pages updates are
protected by managed_page_count_lock, but readers never care about it.
Convert these variables to atomic to avoid readers potentially seeing a
store tear.
This patch converts zone->managed_pages. Subsequent patches will convert
totalram_panges, totalhigh_pages and eventually managed_page_count_lock
will be removed.
Main motivation was that managed_page_count_lock handling was complicating
things. It was discussed in length here,
https://lore.kernel.org/patchwork/patch/995739/#1181785 So it seemes
better to remove the lock and convert variables to atomic, with preventing
poteintial store-to-read tearing as a bonus.
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Arun KS <[email protected]>
Suggested-by: Michal Hocko <[email protected]>
Suggested-by: Vlastimil Babka <[email protected]>
Reviewed-by: Konstantin Khlebnikov <[email protected]>
Reviewed-by: David Hildenbrand <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Reviewed-by: Pavel Tatashin <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Scan through the whole array to see if an update is needed. While we're
at it, use sizeof() to be safe against any possible type changes in the
future.
The bug here is that we wouldn't sync per-cpu counters into global ones
if there was an update of numa_stats for higher cpus. Highly
theoretical one though because it is much more probable that zone_stats
are updated so we would refresh anyway. So I wouldn't bother to mark
this for stable, yet something nice to fix.
[[email protected]: changelog enhancement]
Link: http://lkml.kernel.org/r/[email protected]
Fixes: 1d90ca897c ("mm: update NUMA counter threshold size")
Signed-off-by: Janne Huttunen <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Refaults happen during transitions between workingsets as well as in-place
thrashing. Knowing the difference between the two has a range of
applications, including measuring the impact of memory shortage on the
system performance, as well as the ability to smarter balance pressure
between the filesystem cache and the swap-backed workingset.
During workingset transitions, inactive cache refaults and pushes out
established active cache. When that active cache isn't stale, however,
and also ends up refaulting, that's bonafide thrashing.
Introduce a new page flag that tells on eviction whether the page has been
active or not in its lifetime. This bit is then stored in the shadow
entry, to classify refaults as transitioning or thrashing.
How many page->flags does this leave us with on 32-bit?
20 bits are always page flags
21 if you have an MMU
23 with the zone bits for DMA, Normal, HighMem, Movable
29 with the sparsemem section bits
30 if PAE is enabled
31 with this patch.
So on 32-bit PAE, that leaves 1 bit for distinguishing two NUMA nodes. If
that's not enough, the system can switch to discontigmem and re-gain the 6
or 7 sparsemem section bits.
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Johannes Weiner <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]>
Tested-by: Daniel Drake <[email protected]>
Tested-by: Suren Baghdasaryan <[email protected]>
Cc: Christopher Lameter <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Mike Galbraith <[email protected]>
Cc: Peter Enderborg <[email protected]>
Cc: Randy Dunlap <[email protected]>
Cc: Shakeel Butt <[email protected]>
Cc: Tejun Heo <[email protected]>
Cc: Vinayak Menon <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
The vmstat counter NR_INDIRECTLY_RECLAIMABLE_BYTES was introduced by
commit eb59254608 ("mm: introduce NR_INDIRECTLY_RECLAIMABLE_BYTES") with
the goal of accounting objects that can be reclaimed, but cannot be
allocated via a SLAB_RECLAIM_ACCOUNT cache. This is now possible via
kmalloc() with __GFP_RECLAIMABLE flag, and the dcache external names user
is converted.
The counter is however still useful for accounting direct page allocations
(i.e. not slab) with a shrinker, such as the ION page pool. So keep it,
and:
- change granularity to pages to be more like other counters; sub-page
allocations should be able to use kmalloc
- rename the counter to NR_KERNEL_MISC_RECLAIMABLE
- expose the counter again in vmstat as "nr_kernel_misc_reclaimable"; we can
again remove the check for not printing "hidden" counters
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Vlastimil Babka <[email protected]>
Acked-by: Christoph Lameter <[email protected]>
Acked-by: Roman Gushchin <[email protected]>
Cc: Vijayanand Jitta <[email protected]>
Cc: Laura Abbott <[email protected]>
Cc: Sumit Semwal <[email protected]>
Cc: David Rientjes <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Michal Hocko <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
5dd0b16cda ("mm/vmstat: Make NR_TLB_REMOTE_FLUSH_RECEIVED available even
on UP") made the availability of the NR_TLB_REMOTE_FLUSH* counters inside
the kernel unconditional to reduce #ifdef soup, but (either to avoid
showing dummy zero counters to userspace, or because that code was missed)
didn't update the vmstat_array, meaning that all following counters would
be shown with incorrect values.
This only affects kernel builds with
CONFIG_VM_EVENT_COUNTERS=y && CONFIG_DEBUG_TLBFLUSH=y && CONFIG_SMP=n.
Link: http://lkml.kernel.org/r/[email protected]
Fixes: 5dd0b16cda ("mm/vmstat: Make NR_TLB_REMOTE_FLUSH_RECEIVED available even on UP")
Signed-off-by: Jann Horn <[email protected]>
Reviewed-by: Kees Cook <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Acked-by: Roman Gushchin <[email protected]>
Cc: Davidlohr Bueso <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Cc: Christoph Lameter <[email protected]>
Cc: Kemi Wang <[email protected]>
Cc: Andy Lutomirski <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Revert commit c7f26ccfb2 ("mm/vmstat.c: fix vmstat_update() preemption
BUG"). Steven saw a "using smp_processor_id() in preemptible" message
and added a preempt_disable() section around it to keep it quiet. This
is not the right thing to do it does not fix the real problem.
vmstat_update() is invoked by a kworker on a specific CPU. This worker
it bound to this CPU. The name of the worker was "kworker/1:1" so it
should have been a worker which was bound to CPU1. A worker which can
run on any CPU would have a `u' before the first digit.
smp_processor_id() can be used in a preempt-enabled region as long as
the task is bound to a single CPU which is the case here. If it could
run on an arbitrary CPU then this is the problem we have an should seek
to resolve.
Not only this smp_processor_id() must not be migrated to another CPU but
also refresh_cpu_vm_stats() which might access wrong per-CPU variables.
Not to mention that other code relies on the fact that such a worker
runs on one specific CPU only.
Therefore revert that commit and we should look instead what broke the
affinity mask of the kworker.
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Sebastian Andrzej Siewior <[email protected]>
Cc: Steven J. Hill <[email protected]>
Cc: Tejun Heo <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Variants of proc_create{,_data} that directly take a struct seq_operations
argument and drastically reduces the boilerplate code in the callers.
All trivial callers converted over.
Signed-off-by: Christoph Hellwig <[email protected]>
Patch series "indirectly reclaimable memory", v2.
This patchset introduces the concept of indirectly reclaimable memory
and applies it to fix the issue of when a big number of dentries with
external names can significantly affect the MemAvailable value.
This patch (of 3):
Introduce a concept of indirectly reclaimable memory and adds the
corresponding memory counter and /proc/vmstat item.
Indirectly reclaimable memory is any sort of memory, used by the kernel
(except of reclaimable slabs), which is actually reclaimable, i.e. will
be released under memory pressure.
The counter is in bytes, as it's not always possible to count such
objects in pages. The name contains BYTES by analogy to
NR_KERNEL_STACK_KB.
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Roman Gushchin <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Mel Gorman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Attempting to hotplug CPUs with CONFIG_VM_EVENT_COUNTERS enabled can
cause vmstat_update() to report a BUG due to preemption not being
disabled around smp_processor_id().
Discovered on Ubiquiti EdgeRouter Pro with Cavium Octeon II processor.
BUG: using smp_processor_id() in preemptible [00000000] code:
kworker/1:1/269
caller is vmstat_update+0x50/0xa0
CPU: 0 PID: 269 Comm: kworker/1:1 Not tainted
4.16.0-rc4-Cavium-Octeon-00009-gf83bbd5-dirty #1
Workqueue: mm_percpu_wq vmstat_update
Call Trace:
show_stack+0x94/0x128
dump_stack+0xa4/0xe0
check_preemption_disabled+0x118/0x120
vmstat_update+0x50/0xa0
process_one_work+0x144/0x348
worker_thread+0x150/0x4b8
kthread+0x110/0x140
ret_from_kernel_thread+0x14/0x1c
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Steven J. Hill <[email protected]>
Reviewed-by: Andrew Morton <[email protected]>
Cc: Tejun Heo <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
This is the second step which introduces a tunable interface that allow
numa stats configurable for optimizing zone_statistics(), as suggested
by Dave Hansen and Ying Huang.
=========================================================================
When page allocation performance becomes a bottleneck and you can
tolerate some possible tool breakage and decreased numa counter
precision, you can do:
echo 0 > /proc/sys/vm/numa_stat
In this case, numa counter update is ignored. We can see about
*4.8%*(185->176) drop of cpu cycles per single page allocation and
reclaim on Jesper's page_bench01 (single thread) and *8.1%*(343->315)
drop of cpu cycles per single page allocation and reclaim on Jesper's
page_bench03 (88 threads) running on a 2-Socket Broadwell-based server
(88 threads, 126G memory).
Benchmark link provided by Jesper D Brouer (increase loop times to
10000000):
https://github.com/netoptimizer/prototype-kernel/tree/master/kernel/mm/bench
=========================================================================
When page allocation performance is not a bottleneck and you want all
tooling to work, you can do:
echo 1 > /proc/sys/vm/numa_stat
This is system default setting.
Many thanks to Michal Hocko, Dave Hansen, Ying Huang and Vlastimil Babka
for comments to help improve the original patch.
[[email protected]: make sure mutex is a global static]
Link: http://lkml.kernel.org/r/20171107213809.GA4314@beast
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Kemi Wang <[email protected]>
Signed-off-by: Kees Cook <[email protected]>
Reported-by: Jesper Dangaard Brouer <[email protected]>
Suggested-by: Dave Hansen <[email protected]>
Suggested-by: Ying Huang <[email protected]>
Acked-by: Vlastimil Babka <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Cc: "Luis R . Rodriguez" <[email protected]>
Cc: Kees Cook <[email protected]>
Cc: Jonathan Corbet <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Christopher Lameter <[email protected]>
Cc: Sebastian Andrzej Siewior <[email protected]>
Cc: Andrey Ryabinin <[email protected]>
Cc: Tim Chen <[email protected]>
Cc: Andi Kleen <[email protected]>
Cc: Aaron Lu <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
There is significant overhead in cache bouncing caused by zone counters
(NUMA associated counters) update in parallel in multi-threaded page
allocation (suggested by Dave Hansen).
This patch updates NUMA counter threshold to a fixed size of MAX_U16 - 2,
as a small threshold greatly increases the update frequency of the global
counter from local per cpu counter(suggested by Ying Huang).
The rationality is that these statistics counters don't affect the
kernel's decision, unlike other VM counters, so it's not a problem to use
a large threshold.
With this patchset, we see 31.3% drop of CPU cycles(537-->369) for per
single page allocation and reclaim on Jesper's page_bench03 benchmark.
Benchmark provided by Jesper D Brouer(increase loop times to 10000000):
https://github.com/netoptimizer/prototype-kernel/tree/master/kernel/mm/
bench
Threshold CPU cycles Throughput(88 threads)
32 799 241760478
64 640 301628829
125 537 358906028 <==> system by default (base)
256 468 412397590
512 428 450550704
4096 399 482520943
20000 394 489009617
30000 395 488017817
65533 369(-31.3%) 521661345(+45.3%) <==> with this patchset
N/A 342(-36.3%) 562900157(+56.8%) <==> disable zone_statistics
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Kemi Wang <[email protected]>
Reported-by: Jesper Dangaard Brouer <[email protected]>
Suggested-by: Dave Hansen <[email protected]>
Suggested-by: Ying Huang <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Cc: Aaron Lu <[email protected]>
Cc: Andi Kleen <[email protected]>
Cc: Christopher Lameter <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Tim Chen <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Patch series "Separate NUMA statistics from zone statistics", v2.
Each page allocation updates a set of per-zone statistics with a call to
zone_statistics(). As discussed in 2017 MM summit, these are a
substantial source of overhead in the page allocator and are very rarely
consumed. This significant overhead in cache bouncing caused by zone
counters (NUMA associated counters) update in parallel in multi-threaded
page allocation (pointed out by Dave Hansen).
A link to the MM summit slides:
http://people.netfilter.org/hawk/presentations/MM-summit2017/MM-summit2017-JesperBrouer.pdf
To mitigate this overhead, this patchset separates NUMA statistics from
zone statistics framework, and update NUMA counter threshold to a fixed
size of MAX_U16 - 2, as a small threshold greatly increases the update
frequency of the global counter from local per cpu counter (suggested by
Ying Huang). The rationality is that these statistics counters don't
need to be read often, unlike other VM counters, so it's not a problem
to use a large threshold and make readers more expensive.
With this patchset, we see 31.3% drop of CPU cycles(537-->369, see
below) for per single page allocation and reclaim on Jesper's
page_bench03 benchmark. Meanwhile, this patchset keeps the same style
of virtual memory statistics with little end-user-visible effects (only
move the numa stats to show behind zone page stats, see the first patch
for details).
I did an experiment of single page allocation and reclaim concurrently
using Jesper's page_bench03 benchmark on a 2-Socket Broadwell-based
server (88 processors with 126G memory) with different size of threshold
of pcp counter.
Benchmark provided by Jesper D Brouer(increase loop times to 10000000):
https://github.com/netoptimizer/prototype-kernel/tree/master/kernel/mm/bench
Threshold CPU cycles Throughput(88 threads)
32 799 241760478
64 640 301628829
125 537 358906028 <==> system by default
256 468 412397590
512 428 450550704
4096 399 482520943
20000 394 489009617
30000 395 488017817
65533 369(-31.3%) 521661345(+45.3%) <==> with this patchset
N/A 342(-36.3%) 562900157(+56.8%) <==> disable zone_statistics
This patch (of 3):
In this patch, NUMA statistics is separated from zone statistics
framework, all the call sites of NUMA stats are changed to use
numa-stats-specific functions, it does not have any functionality change
except that the number of NUMA stats is shown behind zone page stats
when users *read* the zone info.
E.g. cat /proc/zoneinfo
***Base*** ***With this patch***
nr_free_pages 3976 nr_free_pages 3976
nr_zone_inactive_anon 0 nr_zone_inactive_anon 0
nr_zone_active_anon 0 nr_zone_active_anon 0
nr_zone_inactive_file 0 nr_zone_inactive_file 0
nr_zone_active_file 0 nr_zone_active_file 0
nr_zone_unevictable 0 nr_zone_unevictable 0
nr_zone_write_pending 0 nr_zone_write_pending 0
nr_mlock 0 nr_mlock 0
nr_page_table_pages 0 nr_page_table_pages 0
nr_kernel_stack 0 nr_kernel_stack 0
nr_bounce 0 nr_bounce 0
nr_zspages 0 nr_zspages 0
numa_hit 0 *nr_free_cma 0*
numa_miss 0 numa_hit 0
numa_foreign 0 numa_miss 0
numa_interleave 0 numa_foreign 0
numa_local 0 numa_interleave 0
numa_other 0 numa_local 0
*nr_free_cma 0* numa_other 0
... ...
vm stats threshold: 10 vm stats threshold: 10
... ...
The next patch updates the numa stats counter size and threshold.
[[email protected]: coding-style fixes]
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Kemi Wang <[email protected]>
Reported-by: Jesper Dangaard Brouer <[email protected]>
Acked-by: Mel Gorman <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Christopher Lameter <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: Andi Kleen <[email protected]>
Cc: Ying Huang <[email protected]>
Cc: Aaron Lu <[email protected]>
Cc: Tim Chen <[email protected]>
Cc: Dave Hansen <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Patch series "mm, swap: VMA based swap readahead", v4.
The swap readahead is an important mechanism to reduce the swap in
latency. Although pure sequential memory access pattern isn't very
popular for anonymous memory, the space locality is still considered
valid.
In the original swap readahead implementation, the consecutive blocks in
swap device are readahead based on the global space locality estimation.
But the consecutive blocks in swap device just reflect the order of page
reclaiming, don't necessarily reflect the access pattern in virtual
memory space. And the different tasks in the system may have different
access patterns, which makes the global space locality estimation
incorrect.
In this patchset, when page fault occurs, the virtual pages near the
fault address will be readahead instead of the swap slots near the fault
swap slot in swap device. This avoid to readahead the unrelated swap
slots. At the same time, the swap readahead is changed to work on
per-VMA from globally. So that the different access patterns of the
different VMAs could be distinguished, and the different readahead
policy could be applied accordingly. The original core readahead
detection and scaling algorithm is reused, because it is an effect
algorithm to detect the space locality.
In addition to the swap readahead changes, some new sysfs interface is
added to show the efficiency of the readahead algorithm and some other
swap statistics.
This new implementation will incur more small random read, on SSD, the
improved correctness of estimation and readahead target should beat the
potential increased overhead, this is also illustrated in the test
results below. But on HDD, the overhead may beat the benefit, so the
original implementation will be used by default.
The test and result is as follow,
Common test condition
=====================
Test Machine: Xeon E5 v3 (2 sockets, 72 threads, 32G RAM)
Swap device: NVMe disk
Micro-benchmark with combined access pattern
============================================
vm-scalability, sequential swap test case, 4 processes to eat 50G
virtual memory space, repeat the sequential memory writing until 300
seconds. The first round writing will trigger swap out, the following
rounds will trigger sequential swap in and out.
At the same time, run vm-scalability random swap test case in
background, 8 processes to eat 30G virtual memory space, repeat the
random memory write until 300 seconds. This will trigger random swap-in
in the background.
This is a combined workload with sequential and random memory accessing
at the same time. The result (for sequential workload) is as follow,
Base Optimized
---- ---------
throughput 345413 KB/s 414029 KB/s (+19.9%)
latency.average 97.14 us 61.06 us (-37.1%)
latency.50th 2 us 1 us
latency.60th 2 us 1 us
latency.70th 98 us 2 us
latency.80th 160 us 2 us
latency.90th 260 us 217 us
latency.95th 346 us 369 us
latency.99th 1.34 ms 1.09 ms
ra_hit% 52.69% 99.98%
The original swap readahead algorithm is confused by the background
random access workload, so readahead hit rate is lower. The VMA-base
readahead algorithm works much better.
Linpack
=======
The test memory size is bigger than RAM to trigger swapping.
Base Optimized
---- ---------
elapsed_time 393.49 s 329.88 s (-16.2%)
ra_hit% 86.21% 98.82%
The score of base and optimized kernel hasn't visible changes. But the
elapsed time reduced and readahead hit rate improved, so the optimized
kernel runs better for startup and tear down stages. And the absolute
value of readahead hit rate is high, shows that the space locality is
still valid in some practical workloads.
This patch (of 5):
The statistics for total readahead pages and total readahead hits are
recorded and exported via the following sysfs interface.
/sys/kernel/mm/swap/ra_hits
/sys/kernel/mm/swap/ra_total
With them, the efficiency of the swap readahead could be measured, so
that the swap readahead algorithm and parameters could be tuned
accordingly.
[[email protected]: don't display swap stats if CONFIG_SWAP=n]
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: "Huang, Ying" <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Shaohua Li <[email protected]>
Cc: Hugh Dickins <[email protected]>
Cc: Fengguang Wu <[email protected]>
Cc: Tim Chen <[email protected]>
Cc: Dave Hansen <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
global_page_state is error prone as a recent bug report pointed out [1].
It only returns proper values for zone based counters as the enum it
gets suggests. We already have global_node_page_state so let's rename
global_page_state to global_zone_page_state to be more explicit here.
All existing users seems to be correct:
$ git grep "global_page_state(NR_" | sed 's@.*(\(NR_[A-Z_]*\)).*@\1@' | sort | uniq -c
2 NR_BOUNCE
2 NR_FREE_CMA_PAGES
11 NR_FREE_PAGES
1 NR_KERNEL_STACK_KB
1 NR_MLOCK
2 NR_PAGETABLE
This patch shouldn't introduce any functional change.
[1] http://lkml.kernel.org/r/[email protected]
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Michal Hocko <[email protected]>
Signed-off-by: Johannes Weiner <[email protected]>
Cc: Tetsuo Handa <[email protected]>
Cc: Josef Bacik <[email protected]>
Cc: Vladimir Davydov <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Baoquan He