linux-toradex.git/kernel/sysctl.c, branch v5.10-rc3

mm: allow a controlled amount of unfairness in the page lock

2020-09-17T17:26:41+00:00

Commit 2a9127fcf229 ("mm: rewrite wait_on_page_bit_common() logic") made
the page locking entirely fair, in that if a waiter came in while the
lock was held, the lock would be transferred to the lockers strictly in
order.

That was intended to finally get rid of the long-reported watchdog
failures that involved the page lock under extreme load, where a process
could end up waiting essentially forever, as other page lockers stole
the lock from under it.

It also improved some benchmarks, but it ended up causing huge
performance regressions on others, simply because fair lock behavior
doesn't end up giving out the lock as aggressively, causing better
worst-case latency, but potentially much worse average latencies and
throughput.

Instead of reverting that change entirely, this introduces a controlled
amount of unfairness, with a sysctl knob to tune it if somebody needs
to.  But the default value should hopefully be good for any normal load,
allowing a few rounds of lock stealing, but enforcing the strict
ordering before the lock has been stolen too many times.

There is also a hint from Matthieu Baerts that the fair page coloring
may end up exposing an ABBA deadlock that is hidden by the usual
optimistic lock stealing, and while the unfairness doesn't fix the
fundamental issue (and I'm still looking at that), it avoids it in
practice.

The amount of unfairness can be modified by writing a new value to the
'sysctl_page_lock_unfairness' variable (default value of 5, exposed
through /proc/sys/vm/page_lock_unfairness), but that is hopefully
something we'd use mainly for debugging rather than being necessary for
any deep system tuning.

This whole issue has exposed just how critical the page lock can be, and
how contended it gets under certain locks.  And the main contention
doesn't really seem to be anything related to IO (which was the origin
of this lock), but for things like just verifying that the page file
mapping is stable while faulting in the page into a page table.

Link: https://lore.kernel.org/linux-fsdevel/ed8442fd-6f54-dd84-cd4a-941e8b7ee603@MichaelLarabel.com/
Link: https://www.phoronix.com/scan.php?page=article&item=linux-50-59&num=1
Link: https://lore.kernel.org/linux-fsdevel/c560a38d-8313-51fb-b1ec-e904bd8836bc@tessares.net/
Reported-and-tested-by: Michael Larabel 
Tested-by: Matthieu Baerts 
Cc: Dave Chinner 
Cc: Matthew Wilcox 
Cc: Chris Mason 
Cc: Jan Kara 
Cc: Amir Goldstein 
Signed-off-by: Linus Torvalds

bpf, sysctl: Let bpf_stats_handler take a kernel pointer buffer

2020-08-25T04:11:40+00:00

Commit 32927393dc1c ("sysctl: pass kernel pointers to ->proc_handler")
changed ctl_table.proc_handler to take a kernel pointer. Adjust the
signature of bpf_stats_handler to match ctl_table.proc_handler which
fixes the following sparse warning:

kernel/sysctl.c:226:49: warning: incorrect type in argument 3 (different address spaces)
kernel/sysctl.c:226:49:    expected void *
kernel/sysctl.c:226:49:    got void [noderef] __user *buffer
kernel/sysctl.c:2640:35: warning: incorrect type in initializer (incompatible argument 3 (different address spaces))
kernel/sysctl.c:2640:35:    expected int ( [usertype] *proc_handler )( ... )
kernel/sysctl.c:2640:35:    got int ( * )( ... )

Fixes: 32927393dc1c ("sysctl: pass kernel pointers to ->proc_handler")
Signed-off-by: Tobias Klauser 
Signed-off-by: Alexei Starovoitov 
Cc: Christoph Hellwig 
Link: https://lore.kernel.org/bpf/20200824142047.22043-1-tklauser@distanz.ch

mm: use unsigned types for fragmentation score

2020-08-12T17:57:56+00:00

Proactive compaction uses per-node/zone "fragmentation score" which is
always in range [0, 100], so use unsigned type of these scores as well as
for related constants.

Signed-off-by: Nitin Gupta 
Signed-off-by: Andrew Morton 
Reviewed-by: Baoquan He 
Cc: Luis Chamberlain 
Cc: Kees Cook 
Cc: Iurii Zaikin 
Cc: Vlastimil Babka 
Cc: Joonsoo Kim 
Link: http://lkml.kernel.org/r/20200618010319.13159-1-nigupta@nvidia.com
Signed-off-by: Linus Torvalds

mm: proactive compaction

2020-08-12T17:57:56+00:00

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/20161230131412.GI13301@dhcp22.suse.cz/
[3] https://lwn.net/Articles/817905/

Signed-off-by: Nitin Gupta 
Signed-off-by: Andrew Morton 
Tested-by: Oleksandr Natalenko 
Reviewed-by: Vlastimil Babka 
Reviewed-by: Khalid Aziz 
Reviewed-by: Oleksandr Natalenko 
Cc: Vlastimil Babka 
Cc: Khalid Aziz 
Cc: Michal Hocko 
Cc: Mel Gorman 
Cc: Matthew Wilcox 
Cc: Mike Kravetz 
Cc: Joonsoo Kim 
Cc: David Rientjes 
Cc: Nitin Gupta 
Cc: Oleksandr Natalenko 
Link: http://lkml.kernel.org/r/20200616204527.19185-1-nigupta@nvidia.com
Signed-off-by: Linus Torvalds

mm: adjust vm_committed_as_batch according to vm overcommit policy

2020-08-07T18:33:26+00:00

When checking a performance change for will-it-scale scalability mmap test
[1], we found very high lock contention for spinlock of percpu counter
'vm_committed_as':

    94.14%     0.35%  [kernel.kallsyms]         [k] _raw_spin_lock_irqsave
    48.21% _raw_spin_lock_irqsave;percpu_counter_add_batch;__vm_enough_memory;mmap_region;do_mmap;
    45.91% _raw_spin_lock_irqsave;percpu_counter_add_batch;__do_munmap;

Actually this heavy lock contention is not always necessary.  The
'vm_committed_as' needs to be very precise when the strict
OVERCOMMIT_NEVER policy is set, which requires a rather small batch number
for the percpu counter.

So keep 'batch' number unchanged for strict OVERCOMMIT_NEVER policy, and
lift it to 64X for OVERCOMMIT_ALWAYS and OVERCOMMIT_GUESS policies.  Also
add a sysctl handler to adjust it when the policy is reconfigured.

Benchmark with the same testcase in [1] shows 53% improvement on a 8C/16T
desktop, and 2097%(20X) on a 4S/72C/144T server.  We tested with test
platforms in 0day (server, desktop and laptop), and 80%+ platforms shows
improvements with that test.  And whether it shows improvements depends on
if the test mmap size is bigger than the batch number computed.

And if the lift is 16X, 1/3 of the platforms will show improvements,
though it should help the mmap/unmap usage generally, as Michal Hocko
mentioned:

: I believe that there are non-synthetic worklaods which would benefit from
: a larger batch.  E.g.  large in memory databases which do large mmaps
: during startups from multiple threads.

[1] https://lore.kernel.org/lkml/20200305062138.GI5972@shao2-debian/

Signed-off-by: Feng Tang 
Signed-off-by: Andrew Morton 
Acked-by: Michal Hocko 
Cc: Matthew Wilcox (Oracle) 
Cc: Johannes Weiner 
Cc: Mel Gorman 
Cc: Qian Cai 
Cc: Kees Cook 
Cc: Andi Kleen 
Cc: Tim Chen 
Cc: Dave Hansen 
Cc: Huang Ying 
Cc: Christoph Lameter 
Cc: Dennis Zhou 
Cc: Haiyang Zhang 
Cc: kernel test robot 
Cc: "K. Y. Srinivasan" 
Cc: Tejun Heo 
Link: http://lkml.kernel.org/r/1589611660-89854-4-git-send-email-feng.tang@intel.com
Link: http://lkml.kernel.org/r/1592725000-73486-4-git-send-email-feng.tang@intel.com
Link: http://lkml.kernel.org/r/1594389708-60781-5-git-send-email-feng.tang@intel.com
Signed-off-by: Linus Torvalds

sched/uclamp: Add a new sysctl to control RT default boost value

2020-07-29T11:51:47+00:00

RT tasks by default run at the highest capacity/performance level. When
uclamp is selected this default behavior is retained by enforcing the
requested uclamp.min (p->uclamp_req[UCLAMP_MIN]) of the RT tasks to be
uclamp_none(UCLAMP_MAX), which is SCHED_CAPACITY_SCALE; the maximum
value.

This is also referred to as 'the default boost value of RT tasks'.

See commit 1a00d999971c ("sched/uclamp: Set default clamps for RT tasks").

On battery powered devices, it is desired to control this default
(currently hardcoded) behavior at runtime to reduce energy consumed by
RT tasks.

For example, a mobile device manufacturer where big.LITTLE architecture
is dominant, the performance of the little cores varies across SoCs, and
on high end ones the big cores could be too power hungry.

Given the diversity of SoCs, the new knob allows manufactures to tune
the best performance/power for RT tasks for the particular hardware they
run on.

They could opt to further tune the value when the user selects
a different power saving mode or when the device is actively charging.

The runtime aspect of it further helps in creating a single kernel image
that can be run on multiple devices that require different tuning.

Keep in mind that a lot of RT tasks in the system are created by the
kernel. On Android for instance I can see over 50 RT tasks, only
a handful of which created by the Android framework.

To control the default behavior globally by system admins and device
integrator, introduce the new sysctl_sched_uclamp_util_min_rt_default
to change the default boost value of the RT tasks.

I anticipate this to be mostly in the form of modifying the init script
of a particular device.

To avoid polluting the fast path with unnecessary code, the approach
taken is to synchronously do the update by traversing all the existing
tasks in the system. This could race with a concurrent fork(), which is
dealt with by introducing sched_post_fork() function which will ensure
the racy fork will get the right update applied.

Tested on Juno-r2 in combination with the RT capacity awareness [1].
By default an RT task will go to the highest capacity CPU and run at the
maximum frequency, which is particularly energy inefficient on high end
mobile devices because the biggest core[s] are 'huge' and power hungry.

With this patch the RT task can be controlled to run anywhere by
default, and doesn't cause the frequency to be maximum all the time.
Yet any task that really needs to be boosted can easily escape this
default behavior by modifying its requested uclamp.min value
(p->uclamp_req[UCLAMP_MIN]) via sched_setattr() syscall.

[1] 804d402fb6f6: ("sched/rt: Make RT capacity-aware")

Signed-off-by: Qais Yousef 
Signed-off-by: Peter Zijlstra (Intel) 
Link: https://lkml.kernel.org/r/20200716110347.19553-2-qais.yousef@arm.com

sched/deadline: Impose global limits on sched_attr::sched_period

2020-06-15T12:10:04+00:00

Signed-off-by: Peter Zijlstra (Intel) 
Link: https://lkml.kernel.org/r/20190726161357.397880775@infradead.org

kernel/sysctl.c: ignore out-of-range taint bits introduced via kernel.tainted

2020-06-08T18:05:56+00:00

Users with SYS_ADMIN capability can add arbitrary taint flags to the
running kernel by writing to /proc/sys/kernel/tainted or issuing the
command 'sysctl -w kernel.tainted=...'.  This interface, however, is
open for any integer value and this might cause an invalid set of flags
being committed to the tainted_mask bitset.

This patch introduces a simple way for proc_taint() to ignore any
eventual invalid bit coming from the user input before committing those
bits to the kernel tainted_mask.

Signed-off-by: Rafael Aquini 
Signed-off-by: Andrew Morton 
Reviewed-by: Luis Chamberlain 
Cc: Kees Cook 
Cc: Iurii Zaikin 
Cc: "Theodore Ts'o" 
Link: http://lkml.kernel.org/r/20200512223946.888020-1-aquini@redhat.com
Signed-off-by: Linus Torvalds

panic: add sysctl to dump all CPUs backtraces on oops event

2020-06-08T18:05:56+00:00

Usually when the kernel reaches an oops condition, it's a point of no
return; in case not enough debug information is available in the kernel
splat, one of the last resorts would be to collect a kernel crash dump
and analyze it.  The problem with this approach is that in order to
collect the dump, a panic is required (to kexec-load the crash kernel).
When in an environment of multiple virtual machines, users may prefer to
try living with the oops, at least until being able to properly shutdown
their VMs / finish their important tasks.

This patch implements a way to collect a bit more debug details when an
oops event is reached, by printing all the CPUs backtraces through the
usage of NMIs (on architectures that support that).  The sysctl added
(and documented) here was called "oops_all_cpu_backtrace", and when set
will (as the name suggests) dump all CPUs backtraces.

Far from ideal, this may be the last option though for users that for
some reason cannot panic on oops.  Most of times oopses are clear enough
to indicate the kernel portion that must be investigated, but in virtual
environments it's possible to observe hypervisor/KVM issues that could
lead to oopses shown in other guests CPUs (like virtual APIC crashes).
This patch hence aims to help debug such complex issues without
resorting to kdump.

Signed-off-by: Guilherme G. Piccoli 
Signed-off-by: Andrew Morton 
Reviewed-by: Kees Cook 
Cc: Luis Chamberlain 
Cc: Iurii Zaikin 
Cc: Thomas Gleixner 
Cc: Vlastimil Babka 
Cc: Randy Dunlap 
Cc: Matthew Wilcox 
Link: http://lkml.kernel.org/r/20200327224116.21030-1-gpiccoli@canonical.com
Signed-off-by: Linus Torvalds

kernel/hung_task.c: introduce sysctl to print all traces when a hung task is detected

2020-06-08T18:05:56+00:00

Commit 401c636a0eeb ("kernel/hung_task.c: show all hung tasks before
panic") introduced a change in that we started to show all CPUs
backtraces when a hung task is detected _and_ the sysctl/kernel
parameter "hung_task_panic" is set.  The idea is good, because usually
when observing deadlocks (that may lead to hung tasks), the culprit is
another task holding a lock and not necessarily the task detected as
hung.

The problem with this approach is that dumping backtraces is a slightly
expensive task, specially printing that on console (and specially in
many CPU machines, as servers commonly found nowadays).  So, users that
plan to collect a kdump to investigate the hung tasks and narrow down
the deadlock definitely don't need the CPUs backtrace on dmesg/console,
which will delay the panic and pollute the log (crash tool would easily
grab all CPUs traces with 'bt -a' command).

Also, there's the reciprocal scenario: some users may be interested in
seeing the CPUs backtraces but not have the system panic when a hung
task is detected.  The current approach hence is almost as embedding a
policy in the kernel, by forcing the CPUs backtraces' dump (only) on
hung_task_panic.

This patch decouples the panic event on hung task from the CPUs
backtraces dump, by creating (and documenting) a new sysctl called
"hung_task_all_cpu_backtrace", analog to the approach taken on soft/hard
lockups, that have both a panic and an "all_cpu_backtrace" sysctl to
allow individual control.  The new mechanism for dumping the CPUs
backtraces on hung task detection respects "hung_task_warnings" by not
dumping the traces in case there's no warnings left.

Signed-off-by: Guilherme G. Piccoli 
Signed-off-by: Andrew Morton 
Reviewed-by: Kees Cook 
Cc: Tetsuo Handa 
Link: http://lkml.kernel.org/r/20200327223646.20779-1-gpiccoli@canonical.com
Signed-off-by: Linus Torvalds