| Age | Commit message (Collapse) | Author |
|---|
|
cpu-local slices
It has been observed, that highly-threaded, non-cpu-bound applications
running under cpu.cfs_quota_us constraints can hit a high percentage of
periods throttled while simultaneously not consuming the allocated
amount of quota. This use case is typical of user-interactive non-cpu
bound applications, such as those running in kubernetes or mesos when
run on multiple cpu cores.
This has been root caused to cpu-local run queue being allocated per cpu
bandwidth slices, and then not fully using that slice within the period.
At which point the slice and quota expires. This expiration of unused
slice results in applications not being able to utilize the quota for
which they are allocated.
The non-expiration of per-cpu slices was recently fixed by
'commit 512ac999d275 ("sched/fair: Fix bandwidth timer clock drift
condition")'. Prior to that it appears that this had been broken since
at least 'commit 51f2176d74ac ("sched/fair: Fix unlocked reads of some
cfs_b->quota/period")' which was introduced in v3.16-rc1 in 2014. That
added the following conditional which resulted in slices never being
expired.
if (cfs_rq->runtime_expires != cfs_b->runtime_expires) {
/* extend local deadline, drift is bounded above by 2 ticks */
cfs_rq->runtime_expires += TICK_NSEC;
Because this was broken for nearly 5 years, and has recently been fixed
and is now being noticed by many users running kubernetes
(https://github.com/kubernetes/kubernetes/issues/67577) it is my opinion
that the mechanisms around expiring runtime should be removed
altogether.
This allows quota already allocated to per-cpu run-queues to live longer
than the period boundary. This allows threads on runqueues that do not
use much CPU to continue to use their remaining slice over a longer
period of time than cpu.cfs_period_us. However, this helps prevent the
above condition of hitting throttling while also not fully utilizing
your cpu quota.
This theoretically allows a machine to use slightly more than its
allotted quota in some periods. This overflow would be bounded by the
remaining quota left on each per-cpu runqueueu. This is typically no
more than min_cfs_rq_runtime=1ms per cpu. For CPU bound tasks this will
change nothing, as they should theoretically fully utilize all of their
quota in each period. For user-interactive tasks as described above this
provides a much better user/application experience as their cpu
utilization will more closely match the amount they requested when they
hit throttling. This means that cpu limits no longer strictly apply per
period for non-cpu bound applications, but that they are still accurate
over longer timeframes.
This greatly improves performance of high-thread-count, non-cpu bound
applications with low cfs_quota_us allocation on high-core-count
machines. In the case of an artificial testcase (10ms/100ms of quota on
80 CPU machine), this commit resulted in almost 30x performance
improvement, while still maintaining correct cpu quota restrictions.
That testcase is available at https://github.com/indeedeng/fibtest.
Fixes: 512ac999d275 ("sched/fair: Fix bandwidth timer clock drift condition")
Signed-off-by: Dave Chiluk <chiluk+linux@indeed.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: John Hammond <jhammond@indeed.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kyle Anderson <kwa@yelp.com>
Cc: Gabriel Munos <gmunoz@netflix.com>
Cc: Peter Oskolkov <posk@posk.io>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Brendan Gregg <bgregg@netflix.com>
Link: https://lkml.kernel.org/r/1563900266-19734-2-git-send-email-chiluk+linux@indeed.com
|
|
The current active_mm reference counting is confusing and sub-optimal.
Rewrite the code to explicitly consider the 4 separate cases:
user -> user
When switching between two user tasks, all we need to consider
is switch_mm().
user -> kernel
When switching from a user task to a kernel task (which
doesn't have an associated mm) we retain the last mm in our
active_mm. Increment a reference count on active_mm.
kernel -> kernel
When switching between kernel threads, all we need to do is
pass along the active_mm reference.
kernel -> user
When switching between a kernel and user task, we must switch
from the last active_mm to the next mm, hoping of course that
these are the same. Decrement a reference on the active_mm.
The code keeps a different order, because as you'll note, both 'to
user' cases require switch_mm().
And where the old code would increment/decrement for the 'kernel ->
kernel' case, the new code observes this is a neutral operation and
avoids touching the reference count.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Rik van Riel <riel@surriel.com>
Reviewed-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: luto@kernel.org
|
|
A rather embarrasing mistake had us call sched_setscheduler() before
initializing the parameters passed to it.
Fixes: 1a763fd7c633 ("rcu/tree: Call setschedule() gp ktread to SCHED_FIFO outside of atomic region")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paul E. McKenney <paulmck@linux.ibm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
|
|
The patch moving bits into mutex.c was a little too much; by also
moving struct mutex_waiter a few less common CONFIGs would no longer
build.
Fixes: 5f35d5a66b3e ("locking/mutex: Make __mutex_owner static to mutex.c")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
|
|
Since commit c66d4bd110a1f8 ("genirq/affinity: Add new callback for
(re)calculating interrupt sets"), irq_create_affinity_masks() returns
NULL in case of single vector. This change has caused regression on some
drivers, such as lpfc.
The problem is that single vector requests can happen in some generic cases:
1) kdump kernel
2) irq vectors resource is close to exhaustion.
If in that situation the affinity mask for a single vector is not created,
every caller has to handle the special case.
There is no reason why the mask cannot be created, so remove the check for
a single vector and create the mask.
Fixes: c66d4bd110a1f8 ("genirq/affinity: Add new callback for (re)calculating interrupt sets")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20190805011906.5020-1-ming.lei@redhat.com
|
|
Instead of using its own logic for k-/vmalloc rely on
kvmalloc which is actually doing quite the same.
Signed-off-by: Marc Koderer <marc@koderer.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
|
|
Booting a large arm64 server (HiSi D05) leads to the following
shouting at boot time:
[ 20.722132] debugfs: File 'irqchip@(____ptrval____)-3' in directory 'domains' already present!
[ 20.730851] debugfs: File 'irqchip@(____ptrval____)-3' in directory 'domains' already present!
[ 20.739560] debugfs: File 'irqchip@(____ptrval____)-3' in directory 'domains' already present!
[ 20.748267] debugfs: File 'irqchip@(____ptrval____)-3' in directory 'domains' already present!
[ 20.756975] debugfs: File 'irqchip@(____ptrval____)-3' in directory 'domains' already present!
[ 20.765683] debugfs: File 'irqchip@(____ptrval____)-3' in directory 'domains' already present!
[ 20.774391] debugfs: File 'irqchip@(____ptrval____)-3' in directory 'domains' already present!
and many more... Evidently, we expect something a bit more informative
than ____ptrval____, and certainly we want all of our domains, not just
the first one.
For that, turn the %p used to generate the fwnode name into something
that won't be repainted (%pa). Given that we've now fixed all users to
pass a pointer to a PA, it will actually do the right thing.
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Marc Zyngier <maz@kernel.org>
|
|
Pull networking fixes from David Miller:
"Yeah I should have sent a pull request last week, so there is a lot
more here than usual:
1) Fix memory leak in ebtables compat code, from Wenwen Wang.
2) Several kTLS bug fixes from Jakub Kicinski (circular close on
disconnect etc.)
3) Force slave speed check on link state recovery in bonding 802.3ad
mode, from Thomas Falcon.
4) Clear RX descriptor bits before assigning buffers to them in
stmmac, from Jose Abreu.
5) Several missing of_node_put() calls, mostly wrt. for_each_*() OF
loops, from Nishka Dasgupta.
6) Double kfree_skb() in peak_usb can driver, from Stephane Grosjean.
7) Need to hold sock across skb->destructor invocation, from Cong
Wang.
8) IP header length needs to be validated in ipip tunnel xmit, from
Haishuang Yan.
9) Use after free in ip6 tunnel driver, also from Haishuang Yan.
10) Do not use MSI interrupts on r8169 chips before RTL8168d, from
Heiner Kallweit.
11) Upon bridge device init failure, we need to delete the local fdb.
From Nikolay Aleksandrov.
12) Handle erros from of_get_mac_address() properly in stmmac, from
Martin Blumenstingl.
13) Handle concurrent rename vs. dump in netfilter ipset, from Jozsef
Kadlecsik.
14) Setting NETIF_F_LLTX on mac80211 causes complete breakage with
some devices, so revert. From Johannes Berg.
15) Fix deadlock in rxrpc, from David Howells.
16) Fix Kconfig deps of enetc driver, we must have PHYLIB. From Yue
Haibing.
17) Fix mvpp2 crash on module removal, from Matteo Croce.
18) Fix race in genphy_update_link, from Heiner Kallweit.
19) bpf_xdp_adjust_head() stopped working with generic XDP when we
fixes generic XDP to support stacked devices properly, fix from
Jesper Dangaard Brouer.
20) Unbalanced RCU locking in rt6_update_exception_stamp_rt(), from
David Ahern.
21) Several memory leaks in new sja1105 driver, from Vladimir Oltean"
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (214 commits)
net: dsa: sja1105: Fix memory leak on meta state machine error path
net: dsa: sja1105: Fix memory leak on meta state machine normal path
net: dsa: sja1105: Really fix panic on unregistering PTP clock
net: dsa: sja1105: Use the LOCKEDS bit for SJA1105 E/T as well
net: dsa: sja1105: Fix broken learning with vlan_filtering disabled
net: dsa: qca8k: Add of_node_put() in qca8k_setup_mdio_bus()
net: sched: sample: allow accessing psample_group with rtnl
net: sched: police: allow accessing police->params with rtnl
net: hisilicon: Fix dma_map_single failed on arm64
net: hisilicon: fix hip04-xmit never return TX_BUSY
net: hisilicon: make hip04_tx_reclaim non-reentrant
tc-testing: updated vlan action tests with batch create/delete
net sched: update vlan action for batched events operations
net: stmmac: tc: Do not return a fragment entry
net: stmmac: Fix issues when number of Queues >= 4
net: stmmac: xgmac: Fix XGMAC selftests
be2net: disable bh with spin_lock in be_process_mcc
net: cxgb3_main: Fix a resource leak in a error path in 'init_one()'
net: ethernet: sun4i-emac: Support phy-handle property for finding PHYs
net: bridge: move default pvid init/deinit to NETDEV_REGISTER/UNREGISTER
...
|
|
It is not desirable to relax the ABI to allow tagged user addresses into
the kernel indiscriminately. This patch introduces a prctl() interface
for enabling or disabling the tagged ABI with a global sysctl control
for preventing applications from enabling the relaxed ABI (meant for
testing user-space prctl() return error checking without reconfiguring
the kernel). The ABI properties are inherited by threads of the same
application and fork()'ed children but cleared on execve(). A Kconfig
option allows the overall disabling of the relaxed ABI.
The PR_SET_TAGGED_ADDR_CTRL will be expanded in the future to handle
MTE-specific settings like imprecise vs precise exceptions.
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Will Deacon <will@kernel.org>
|
|
When a process creates a new trigger by writing into /proc/pressure/*
files, permissions to write such a file should be used to determine whether
the process is allowed to do so or not. Current implementation would also
require such a process to have setsched capability. Setting of psi trigger
thread's scheduling policy is an implementation detail and should not be
exposed to the user level. Remove the permission check by using _nocheck
version of the function.
Suggested-by: Nick Kralevich <nnk@google.com>
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: lizefan@huawei.com
Cc: mingo@redhat.com
Cc: akpm@linux-foundation.org
Cc: kernel-team@android.com
Cc: dennisszhou@gmail.com
Cc: dennis@kernel.org
Cc: hannes@cmpxchg.org
Cc: axboe@kernel.dk
Link: https://lkml.kernel.org/r/20190730013310.162367-1-surenb@google.com
|
|
PSI defaults to a FIFO-99 thread, reduce this to FIFO-1.
FIFO-99 is the very highest priority available to SCHED_FIFO and
it not a suitable default; it would indicate the psi work is the
most important work on the machine.
Since Real-Time tasks will have pre-allocated memory and locked it in
place, Real-Time tasks do not care about PSI. All it needs is to be
above OTHER.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Tested-by: Suren Baghdasaryan <surenb@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
|
|
{push,pull}_dl_task() always calls {de,}activate_task() with .flags=0
which sets p->on_rq=TASK_ON_RQ_MIGRATING.
{push,pull}_dl_task()->{de,}activate_task()->{de,en}queue_task()->
{de,en}queue_task_dl() calls {sub,add}_{running,rq}_bw() since
p->on_rq==TASK_ON_RQ_MIGRATING.
So {sub,add}_{running,rq}_bw() in {push,pull}_dl_task() is
double-accounting for that task.
Fix it by removing rq/running bw accounting in [push/pull]_dl_task().
Fixes: 7dd778841164 ("sched/core: Unify p->on_rq updates")
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Luca Abeni <luca.abeni@santannapisa.it>
Cc: Daniel Bristot de Oliveira <bristot@redhat.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Qais Yousef <qais.yousef@arm.com>
Link: https://lkml.kernel.org/r/20190802145945.18702-2-dietmar.eggemann@arm.com
|
|
Use the mutex flag macro instead of hard code value inside
__mutex_owner().
Signed-off-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: mingo@redhat.com
Cc: will@kernel.org
Link: https://lkml.kernel.org/r/1564585504-3543-2-git-send-email-mojha@codeaurora.org
|
|
__mutex_owner() should only be used by the mutex api's.
So, to put this restiction let's move the __mutex_owner()
function definition from linux/mutex.h to mutex.c file.
There exist functions that uses __mutex_owner() like
mutex_is_locked() and mutex_trylock_recursive(), So
to keep legacy thing intact move them as well and
export them.
Move mutex_waiter structure also to keep it private to the
file.
Signed-off-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: mingo@redhat.com
Cc: will@kernel.org
Link: https://lkml.kernel.org/r/1564585504-3543-1-git-send-email-mojha@codeaurora.org
|
|
Currently rwsems is the only locking primitive that lacks this
debug feature. Add it under CONFIG_DEBUG_RWSEMS and do the magic
checking in the locking fastpath (trylock) operation such that
we cover all cases. The unlocking part is pretty straightforward.
Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Waiman Long <longman@redhat.com>
Cc: mingo@kernel.org
Cc: Davidlohr Bueso <dave@stgolabs.net>
Link: https://lkml.kernel.org/r/20190729044735.9632-1-dave@stgolabs.net
|
|
When the handoff bit is set by a writer, no other tasks other than
the setting writer itself is allowed to acquire the lock. If the
to-be-handoff'ed writer goes to sleep, there will be a wakeup latency
period where the lock is free, but no one can acquire it. That is less
than ideal.
To reduce that latency, the handoff writer will now optimistically spin
on the owner if it happens to be a on-cpu writer. It will spin until
it releases the lock and the to-be-handoff'ed writer can then acquire
the lock immediately without any delay. Of course, if the owner is not
a on-cpu writer, the to-be-handoff'ed writer will have to sleep anyway.
The optimistic spinning code is also modified to not stop spinning
when the handoff bit is set. This will prevent an occasional setting of
handoff bit from causing a bunch of optimistic spinners from entering
into the wait queue causing significant reduction in throughput.
On a 1-socket 22-core 44-thread Skylake system, the AIM7 shared_memory
workload was run with 7000 users. The throughput (jobs/min) of the
following kernels were as follows:
1) 5.2-rc6
- 8,092,486
2) 5.2-rc6 + tip's rwsem patches
- 7,567,568
3) 5.2-rc6 + tip's rwsem patches + this patch
- 7,954,545
Using perf-record(1), the %cpu time used by rwsem_down_write_slowpath(),
rwsem_down_write_failed() and their callees for the 3 kernels were 1.70%,
5.46% and 2.08% respectively.
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: x86@kernel.org
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: huang ying <huang.ying.caritas@gmail.com>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Link: https://lkml.kernel.org/r/20190625143913.24154-1-longman@redhat.com
|
|
IMA will use the module_signature format for append signatures, so export
the relevant definitions and factor out the code which verifies that the
appended signature trailer is valid.
Also, create a CONFIG_MODULE_SIG_FORMAT option so that IMA can select it
and be able to use mod_check_sig() without having to depend on either
CONFIG_MODULE_SIG or CONFIG_MODULES.
s390 duplicated the definition of struct module_signature so now they can
use the new <linux/module_signature.h> header instead.
Signed-off-by: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Acked-by: Jessica Yu <jeyu@kernel.org>
Reviewed-by: Philipp Rudo <prudo@linux.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
|
|
Userspace can get suspend stats from the suspend stats debugfs node.
Since debugfs doesn't have stable ABI, expose suspend stats in
sysfs under /sys/power/suspend_stats.
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
|
memremap.c implements MM functionality for ZONE_DEVICE, so it really
should be in the mm/ directory, not the kernel/ one.
Link: http://lkml.kernel.org/r/20190722094143.18387-1-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Acked-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
The kernel-doc parser doesn't handle expressions with %foo*. Instead,
when an asterisk should be part of a constant, it uses an alternative
notation: `foo*`.
Link: http://lkml.kernel.org/r/7f18c2e0b5e39e6b7eb55ddeb043b8b260b49f2d.1563361575.git.mchehab+samsung@kernel.org
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Cc: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Pull arm swiotlb support from Christoph Hellwig:
"This fixes a cascade of regressions that originally started with the
addition of the ia64 port, but only got fatal once we removed most
uses of block layer bounce buffering in Linux 4.18.
The reason is that while the original i386/PAE code that was the first
architecture that supported > 4GB of memory without an iommu decided
to leave bounce buffering to the subsystems, which in those days just
mean block and networking as no one else consumed arbitrary userspace
memory.
Later with ia64, x86_64 and other ports we assumed that either an
iommu or something that fakes it up ("software IOTLB" in beautiful
Intel speak) is present and that subsystems can rely on that for
dealing with addressing limitations in devices. Except that the ARM
LPAE scheme that added larger physical address to 32-bit ARM did not
follow that scheme and thus only worked by chance and only for block
and networking I/O directly to highmem.
Long story, short fix - add swiotlb support to arm when build for LPAE
platforms, which actuallys turns out to be pretty trivial with the
modern dma-direct / swiotlb code to fix the Linux 4.18-ish regression"
* tag 'arm-swiotlb-5.3' of git://git.infradead.org/users/hch/dma-mapping:
arm: use swiotlb for bounce buffering on LPAE configs
dma-mapping: check pfn validity in dma_common_{mmap,get_sgtable}
|
|
Pull dma-mapping regression fixes from Christoph Hellwig:
"Two related regression fixes for changes from this merge window to fix
alignment issues introduced in the CMA allocation rework (Nicolin
Chen)"
* tag 'dma-mapping-5.3-3' of git://git.infradead.org/users/hch/dma-mapping:
dma-contiguous: page-align the size in dma_free_contiguous()
dma-contiguous: do not overwrite align in dma_alloc_contiguous()
|
|
The more aggressive forward-progress tests can interfere with rcutorture
shutdown, resulting in false-positive diagnostics. This commit therefore
ends any such tests 30 seconds prior to shutdown.
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
It is possible that the rcuperf kernel test runs concurrently with init
starting up. During this time, the system is running all grace periods
as expedited. However, rcuperf can also be run for normal GP tests.
Right now, it depends on a holdoff time before starting the test to
ensure grace periods start later. This works fine with the default
holdoff time however it is not robust in situations where init takes
greater than the holdoff time to finish running. Or, as in my case:
I modified the rcuperf test locally to also run a thread that did
preempt disable/enable in a loop. This had the effect of slowing down
init. The end result was that the "batches:" counter in rcuperf was 0
causing a division by 0 error in the results. This counter was 0 because
only expedited GPs seem to happen, not normal ones which led to the
rcu_state.gp_seq counter remaining constant across grace periods which
unexpectedly happen to be expedited. The system was running expedited
RCU all the time because rcu_unexpedited_gp() would not have run yet
from init. In other words, the test would concurrently with init
booting in expedited GP mode.
To fix this properly, this commit waits until system_state is set to
SYSTEM_RUNNING before starting the test. This change is made just
before kernel_init() invokes rcu_end_inkernel_boot(), and this latter
is what turns off boot-time expediting of RCU grace periods.
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
The functions torture_onoff_cleanup() and torture_shuffle_cleanup()
are declared static and marked EXPORT_SYMBOL_GPL(), which is at best an
odd combination. Because these functions are not used outside of the
kernel/torture.c file they are defined in, this commit removes their
EXPORT_SYMBOL_GPL() marking.
Fixes: cc47ae083026 ("rcutorture: Abstract torture-test cleanup")
Signed-off-by: Denis Efremov <efremov@linux.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
During an actual call_rcu() flood, there would be frequent trips to
userspace (in-kernel call_rcu() floods must be otherwise housebroken).
Userspace execution allows a great many things to interrupt execution,
and rcutorture needs to also allow such interruptions. This commit
therefore causes call_rcu() floods to occasionally invoke schedule(),
thus preventing spurious rcutorture failures due to other parts of the
kernel becoming irate at the call_rcu() flood events.
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
The rcu_bh rcuperf type was removed by commit 620d246065cd("rcuperf:
Remove the "rcu_bh" and "sched" torture types"), but it lives on in the
MODULE_PARM_DESC() of perf_type. This commit therefore changes that
module-parameter description to substitute srcu for rcu_bh.
Signed-off-by: Xiao Yang <ice_yangxiao@163.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
The debug_locks flag can never be true at the end of
rcu_read_lock_sched_held() because it is already checked by the earlier
call todebug_lockdep_rcu_enabled(). This commit therefore removes this
redundant check.
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
The rcu_dereference_raw_notrace() API name is confusing. It is equivalent
to rcu_dereference_raw() except that it also does sparse pointer checking.
There are only a few users of rcu_dereference_raw_notrace(). This patches
renames all of them to be rcu_dereference_raw_check() with the "_check()"
indicating sparse checking.
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
[ paulmck: Fix checkpatch warnings about parentheses. ]
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
The return value of rcu_spawn_one_boost_kthread() is not used any longer.
This commit therefore changes its return type from int to void, and
removes the cast to void from its callers.
Signed-off-by: Byungchul Park <byungchul.park@lge.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
Because pointer output is now obfuscated, and because what you really
want to know is whether or not the callback lists are empty, this commit
replaces the srcu_data structure's head callback pointer printout with
a single character that is "." is the callback list is empty or "C"
otherwise.
This is the only remaining user of rcu_segcblist_head(), so this
commit also removes this function's definition. It also turns out that
rcu_segcblist_tail() no longer has any callers, so this commit removes
that function's definition while in the area. They were both marked
"Interim", and their end has come.
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
The synchronize_rcu_expedited() function has an INIT_WORK_ONSTACK(),
but lacks the corresponding destroy_work_on_stack(). This commit
therefore adds destroy_work_on_stack().
Reported-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
Acked-by: Andrea Arcangeli <aarcange@redhat.com>
|
|
This commit adds a rcu_cpu_stall_ftrace_dump kernel boot parameter, that,
when set, causes the trace buffer to be dumped after an RCU CPU stall
warning is printed. This kernel boot parameter is disabled by default,
maintaining compatibility with previous behavior.
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
Commit bb73c52bad36 ("rcu: Don't disable preemption for Tiny and Tree
RCU readers") removed the barrier() calls from rcu_read_lock() and
rcu_write_lock() in CONFIG_PREEMPT=n&&CONFIG_PREEMPT_COUNT=n kernels.
Within RCU, this commit was OK, but it failed to account for things like
get_user() that can pagefault and that can be reordered by the compiler.
Lack of the barrier() calls in rcu_read_lock() and rcu_read_unlock()
can cause these page faults to migrate into RCU read-side critical
sections, which in CONFIG_PREEMPT=n kernels could result in too-short
grace periods and arbitrary misbehavior. Please see commit 386afc91144b
("spinlocks and preemption points need to be at least compiler barriers")
and Linus's commit 66be4e66a7f4 ("rcu: locking and unlocking need to
always be at least barriers"), this last of which restores the barrier()
call to both rcu_read_lock() and rcu_read_unlock().
This commit removes barrier() calls that are no longer needed given that
the addition of them in Linus's commit noted above. The combination of
this commit and Linus's commit effectively reverts commit bb73c52bad36
("rcu: Don't disable preemption for Tiny and Tree RCU readers").
Reported-by: Herbert Xu <herbert@gondor.apana.org.au>
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
[ paulmck: Fix embarrassing typo located by Alan Stern. ]
|
|
The TASKS03 and TREE04 rcutorture scenarios produce the following
lockdep complaint:
------------------------------------------------------------------------
================================
WARNING: inconsistent lock state
5.2.0-rc1+ #513 Not tainted
--------------------------------
inconsistent {IN-HARDIRQ-W} -> {HARDIRQ-ON-W} usage.
migration/1/14 [HC0[0]:SC0[0]:HE1:SE1] takes:
(____ptrval____) (tick_broadcast_lock){?...}, at: tick_broadcast_offline+0xf/0x70
{IN-HARDIRQ-W} state was registered at:
lock_acquire+0xb0/0x1c0
_raw_spin_lock_irqsave+0x3c/0x50
tick_broadcast_switch_to_oneshot+0xd/0x40
tick_switch_to_oneshot+0x4f/0xd0
hrtimer_run_queues+0xf3/0x130
run_local_timers+0x1c/0x50
update_process_times+0x1c/0x50
tick_periodic+0x26/0xc0
tick_handle_periodic+0x1a/0x60
smp_apic_timer_interrupt+0x80/0x2a0
apic_timer_interrupt+0xf/0x20
_raw_spin_unlock_irqrestore+0x4e/0x60
rcu_nocb_gp_kthread+0x15d/0x590
kthread+0xf3/0x130
ret_from_fork+0x3a/0x50
irq event stamp: 171
hardirqs last enabled at (171): [<ffffffff8a201a37>] trace_hardirqs_on_thunk+0x1a/0x1c
hardirqs last disabled at (170): [<ffffffff8a201a53>] trace_hardirqs_off_thunk+0x1a/0x1c
softirqs last enabled at (0): [<ffffffff8a264ee0>] copy_process.part.56+0x650/0x1cb0
softirqs last disabled at (0): [<0000000000000000>] 0x0
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(tick_broadcast_lock);
<Interrupt>
lock(tick_broadcast_lock);
*** DEADLOCK ***
1 lock held by migration/1/14:
#0: (____ptrval____) (clockevents_lock){+.+.}, at: tick_offline_cpu+0xf/0x30
stack backtrace:
CPU: 1 PID: 14 Comm: migration/1 Not tainted 5.2.0-rc1+ #513
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS Bochs 01/01/2011
Call Trace:
dump_stack+0x5e/0x8b
print_usage_bug+0x1fc/0x216
? print_shortest_lock_dependencies+0x1b0/0x1b0
mark_lock+0x1f2/0x280
__lock_acquire+0x1e0/0x18f0
? __lock_acquire+0x21b/0x18f0
? _raw_spin_unlock_irqrestore+0x4e/0x60
lock_acquire+0xb0/0x1c0
? tick_broadcast_offline+0xf/0x70
_raw_spin_lock+0x33/0x40
? tick_broadcast_offline+0xf/0x70
tick_broadcast_offline+0xf/0x70
tick_offline_cpu+0x16/0x30
take_cpu_down+0x7d/0xa0
multi_cpu_stop+0xa2/0xe0
? cpu_stop_queue_work+0xc0/0xc0
cpu_stopper_thread+0x6d/0x100
smpboot_thread_fn+0x169/0x240
kthread+0xf3/0x130
? sort_range+0x20/0x20
? kthread_cancel_delayed_work_sync+0x10/0x10
ret_from_fork+0x3a/0x50
------------------------------------------------------------------------
To reproduce, run the following rcutorture test:
tools/testing/selftests/rcutorture/bin/kvm.sh --duration 5 --kconfig "CONFIG_DEBUG_LOCK_ALLOC=y CONFIG_PROVE_LOCKING=y" --configs "TASKS03 TREE04"
It turns out that tick_broadcast_offline() was an innocent bystander.
After all, interrupts are supposed to be disabled throughout
take_cpu_down(), and therefore should have been disabled upon entry to
tick_offline_cpu() and thus to tick_broadcast_offline(). This suggests
that one of the CPU-hotplug notifiers was incorrectly enabling interrupts,
and leaving them enabled on return.
Some debugging code showed that the culprit was sched_cpu_dying().
It had irqs enabled after return from sched_tick_stop(). Which in turn
had irqs enabled after return from cancel_delayed_work_sync(). Which is a
wrapper around __cancel_work_timer(). Which can sleep in the case where
something else is concurrently trying to cancel the same delayed work,
and as Thomas Gleixner pointed out on IRC, sleeping is a decidedly bad
idea when you are invoked from take_cpu_down(), regardless of the state
you leave interrupts in upon return.
Code inspection located no reason why the delayed work absolutely
needed to be canceled from sched_tick_stop(): The work is not
bound to the outgoing CPU by design, given that the whole point is
to collect statistics without disturbing the outgoing CPU.
This commit therefore simply drops the cancel_delayed_work_sync() from
sched_tick_stop(). Instead, a new ->state field is added to the tick_work
structure so that the delayed-work handler function sched_tick_remote()
can avoid reposting itself. A cpu_is_offline() check is also added to
sched_tick_remote() to avoid mucking with the state of an offlined CPU
(though it does appear safe to do so). The sched_tick_start() and
sched_tick_stop() functions also update ->state, and sched_tick_start()
also schedules the delayed work if ->state indicates that it is not
already in flight.
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
[ paulmck: Apply Peter Zijlstra and Frederic Weisbecker atomics feedback. ]
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
|
|
Security is a wonderful thing, but so is the ability to debug based on
lockdep warnings. This commit therefore makes lockdep lock addresses
visible in the clear.
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
Because __rcu_read_unlock() can be preempted just before the call to
rcu_read_unlock_special(), it is possible for a task to be preempted just
before it would have fully exited its RCU read-side critical section.
This would result in a needless extension of that critical section until
that task was resumed, which might in turn result in a needlessly
long grace period, needless RCU priority boosting, and needless
force-quiescent-state actions. Therefore, rcu_note_context_switch()
invokes __rcu_read_unlock() followed by rcu_preempt_deferred_qs() when
it detects this situation. This action by rcu_note_context_switch()
ends the RCU read-side critical section immediately.
Of course, once the task resumes, it will invoke rcu_read_unlock_special()
redundantly. This is harmless because the fact that a preemption
happened means that interrupts, preemption, and softirqs cannot
have been disabled, so there would be no deferred quiescent state.
While ->rcu_read_lock_nesting remains less than zero, none of the
->rcu_read_unlock_special.b bits can be set, and they were all zeroed by
the call to rcu_note_context_switch() at task-preemption time. Therefore,
setting ->rcu_read_unlock_special.b.exp_hint to false has no effect.
Therefore, the extra call to rcu_preempt_deferred_qs_irqrestore()
would return immediately. With one possible exception, which is
if an expedited grace period started just as the task was being
resumed, which could leave ->exp_deferred_qs set. This will cause
rcu_preempt_deferred_qs_irqrestore() to invoke rcu_report_exp_rdp(),
reporting the quiescent state, just as it should. (Such an expedited
grace period won't affect the preemption code path due to interrupts
having already been disabled.)
But when rcu_note_context_switch() invokes __rcu_read_unlock(), it
is doing so with preemption disabled, hence __rcu_read_unlock() will
unconditionally defer the quiescent state, only to immediately invoke
rcu_preempt_deferred_qs(), thus immediately reporting the deferred
quiescent state. It turns out to be safe (and faster) to instead
just invoke rcu_preempt_deferred_qs() without the __rcu_read_unlock()
middleman.
Because this is the invocation during the preemption (as opposed to
the invocation just after the resume), at least one of the bits in
->rcu_read_unlock_special.b must be set and ->rcu_read_lock_nesting
must be negative. This means that rcu_preempt_need_deferred_qs() must
return true, avoiding the early exit from rcu_preempt_deferred_qs().
Thus, rcu_preempt_deferred_qs_irqrestore() will be invoked immediately,
as required.
This commit therefore simplifies the CONFIG_PREEMPT=y version of
rcu_note_context_switch() by removing the "else if" branch of its
"if" statement. This change means that all callers that would have
invoked rcu_read_unlock_special() followed by rcu_preempt_deferred_qs()
will now simply invoke rcu_preempt_deferred_qs(), thus avoiding the
rcu_read_unlock_special() middleman when __rcu_read_unlock() is preempted.
Cc: rcu@vger.kernel.org
Cc: kernel-team@android.com
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
Threaded interrupts provide additional interesting interactions between
RCU and raise_softirq() that can result in self-deadlocks in v5.0-2 of
the Linux kernel. These self-deadlocks can be provoked in susceptible
kernels within a few minutes using the following rcutorture command on
an 8-CPU system:
tools/testing/selftests/rcutorture/bin/kvm.sh --duration 5 --configs "TREE03" --bootargs "threadirqs"
Although post-v5.2 RCU commits have at least greatly reduced the
probability of these self-deadlocks, this was entirely by accident.
Although this sort of accident should be rowdily celebrated on those
rare occasions when it does occur, such celebrations should be quickly
followed by a principled patch, which is what this patch purports to be.
The key point behind this patch is that when in_interrupt() returns
true, __raise_softirq_irqoff() will never attempt a wakeup. Therefore,
if in_interrupt(), calls to raise_softirq*() are both safe and
extremely cheap.
This commit therefore replaces the in_irq() calls in the "if" statement
in rcu_read_unlock_special() with in_interrupt() and simplifies the
"if" condition to the following:
if (irqs_were_disabled && use_softirq &&
(in_interrupt() ||
(exp && !t->rcu_read_unlock_special.b.deferred_qs))) {
raise_softirq_irqoff(RCU_SOFTIRQ);
} else {
/* Appeal to the scheduler. */
}
The rationale behind the "if" condition is as follows:
1. irqs_were_disabled: If interrupts are enabled, we should
instead appeal to the scheduler so as to let the upcoming
irq_enable()/local_bh_enable() do the rescheduling for us.
2. use_softirq: If this kernel isn't using softirq, then
raise_softirq_irqoff() will be unhelpful.
3. a. in_interrupt(): If this returns true, the subsequent
call to raise_softirq_irqoff() is guaranteed not to
do a wakeup, so that call will be both very cheap and
quite safe.
b. Otherwise, if !in_interrupt() the raise_softirq_irqoff()
might do a wakeup, which is expensive and, in some
contexts, unsafe.
i. The "exp" (an expedited RCU grace period is being
blocked) says that the wakeup is worthwhile, and:
ii. The !.deferred_qs says that scheduler locks
cannot be held, so the wakeup will be safe.
Backporting this requires considerable care, so no auto-backport, please!
Fixes: 05f415715ce45 ("rcu: Speed up expedited GPs when interrupting RCU reader")
Reported-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
In !use_softirq runs, we clearly cannot rely on raise_softirq() and
its lightweight bit setting, so we must instead do some form of wakeup.
In the absence of a self-IPI when interrupts are disabled, these wakeups
can be delayed until the next interrupt occurs. This means that calling
invoke_rcu_core() doesn't actually do any expediting.
In this case, it is better to take the "else" clause, which sets the
current CPU's resched bits and, if there is an expedited grace period
in flight, uses IRQ-work to force the needed self-IPI. This commit
therefore removes the "else if" clause that calls invoke_rcu_core().
Reported-by: Scott Wood <swood@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
|
|
Since we always allocate memory, allocate just a little bit more
for the BPF program in case it need to override user input with
bigger value. The canonical example is TCP_CONGESTION where
input string might be too small to override (nv -> bbr or cubic).
16 bytes are chosen to match the size of TCP_CA_NAME_MAX and can
be extended in the future if needed.
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
This adds the P_PIDFD type to waitid().
One of the last remaining bits for the pidfd api is to make it possible
to wait on pidfds. With P_PIDFD added to waitid() the parts of userspace
that want to use the pidfd api to exclusively manage processes can do so
now.
One of the things this will unblock in the future is the ability to make
it possible to retrieve the exit status via waitid(P_PIDFD) for
non-parent processes if handed a _suitable_ pidfd that has this feature
set. This is similar to what you can do on FreeBSD with kqueue(). It
might even end up being possible to wait on a process as a non-parent if
an appropriate property is enabled on the pidfd.
With P_PIDFD no scoping of the process identified by the pidfd is
possible, i.e. it explicitly blocks things such as wait4(-1), wait4(0),
waitid(P_ALL), waitid(P_PGID) etc. It only allows for semantics
equivalent to wait4(pid), waitid(P_PID). Users that need scoping should
rely on pid-based wait*() syscalls for now.
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: David Howells <dhowells@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: Andy Lutomirsky <luto@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Aleksa Sarai <cyphar@cyphar.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Link: https://lore.kernel.org/r/20190727222229.6516-2-christian@brauner.io
|
|
Timer deletion on PREEMPT_RT is prone to priority inversion and live
locks. The hrtimer code has a synchronization mechanism for this. Posix CPU
timers will grow one.
But that mechanism cannot be invoked while holding the k_itimer lock
because that can deadlock against the running timer callback. So the lock
must be dropped which allows the timer to be freed.
The timer free can be prevented by taking RCU readlock before dropping the
lock, but because the rcu_head is part of the 'it' union a concurrent free
will overwrite the hrtimer on which the task is trying to synchronize.
Move the rcu_head out of the union to prevent this.
[ tglx: Fixed up kernel-doc. Rewrote changelog ]
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190730223828.965541887@linutronix.de
|
|
As a preparatory step for adding the PREEMPT RT specific synchronization
mechanism to wait for a running timer callback, rework the timer cancel
retry loops so they call a common function. This allows trivial
substitution in one place.
Originally-by: Anna-Maria Gleixner <anna-maria@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190730223828.874901027@linutronix.de
|
|
do_timer_settime() has a 'flags' argument and uses 'flag' for the interrupt
flags, which is confusing at best.
Rename the argument so 'flags' can be used for interrupt flags as usual.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190730223828.782664411@linutronix.de
|
|
Use the hrtimer_cancel_wait_running() synchronization mechanism to prevent
priority inversion and live locks on PREEMPT_RT.
As a benefit the retry loop gains the missing cpu_relax() on !RT.
[ tglx: Split out of combo patch ]
Signed-off-by: Anna-Maria Gleixner <anna-maria@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190730223828.690771827@linutronix.de
|
|
Use the hrtimer_cancel_wait_running() synchronization mechanism to prevent
priority inversion and live locks on PREEMPT_RT.
[ tglx: Split out of combo patch ]
Signed-off-by: Anna-Maria Gleixner <anna-maria@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190730223828.508744705@linutronix.de
|
|
SCHED_DEADLINE inactive timer needs to run in hardirq context (as
dl_task_timer already does) on PREEMPT_RT
Change the mode to HRTIMER_MODE_REL_HARD.
[ tglx: Fixed up the start site, so mode debugging works ]
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20190731103715.4047-1-juri.lelli@redhat.com
|
|
When PREEMPT_RT is enabled, the soft interrupt thread can be preempted. If
the soft interrupt thread is preempted in the middle of a timer callback,
then calling del_timer_sync() can lead to two issues:
- If the caller is on a remote CPU then it has to spin wait for the timer
handler to complete. This can result in unbound priority inversion.
- If the caller originates from the task which preempted the timer
handler on the same CPU, then spin waiting for the timer handler to
complete is never going to end.
To avoid these issues, add a new lock to the timer base which is held
around the execution of the timer callbacks. If del_timer_sync() detects
that the timer callback is currently running, it blocks on the expiry
lock. When the callback is finished, the expiry lock is dropped by the
softirq thread which wakes up the waiter and the system makes progress.
This addresses both the priority inversion and the life lock issues.
This mechanism is not used for timers which are marked IRQSAFE as for those
preemption is disabled accross the callback and therefore this situation
cannot happen. The callbacks for such timers need to be individually
audited for RT compliance.
The same issue can happen in virtual machines when the vCPU which runs a
timer callback is scheduled out. If a second vCPU of the same guest calls
del_timer_sync() it will spin wait for the other vCPU to be scheduled back
in. The expiry lock mechanism would avoid that. It'd be trivial to enable
this when paravirt spinlocks are enabled in a guest, but it's not clear
whether this is an actual problem in the wild, so for now it's an RT only
mechanism.
As the softirq thread can be preempted with PREEMPT_RT=y, the SMP variant
of del_timer_sync() needs to be used on UP as well.
[ tglx: Refactored it for mainline ]
Signed-off-by: Anna-Maria Gleixner <anna-maria@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190726185753.832418500@linutronix.de
|
|
When PREEMPT_RT is enabled, the soft interrupt thread can be preempted. If
the soft interrupt thread is preempted in the middle of a timer callback,
then calling hrtimer_cancel() can lead to two issues:
- If the caller is on a remote CPU then it has to spin wait for the timer
handler to complete. This can result in unbound priority inversion.
- If the caller originates from the task which preempted the timer
handler on the same CPU, then spin waiting for the timer handler to
complete is never going to end.
To avoid these issues, add a new lock to the timer base which is held
around the execution of the timer callbacks. If hrtimer_cancel() detects
that the timer callback is currently running, it blocks on the expiry
lock. When the callback is finished, the expiry lock is dropped by the
softirq thread which wakes up the waiter and the system makes progress.
This addresses both the priority inversion and the life lock issues.
The same issue can happen in virtual machines when the vCPU which runs a
timer callback is scheduled out. If a second vCPU of the same guest calls
hrtimer_cancel() it will spin wait for the other vCPU to be scheduled back
in. The expiry lock mechanism would avoid that. It'd be trivial to enable
this when paravirt spinlocks are enabled in a guest, but it's not clear
whether this is an actual problem in the wild, so for now it's an RT only
mechanism.
[ tglx: Refactored it for mainline ]
Signed-off-by: Anna-Maria Gleixner <anna-maria@linutronix.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190726185753.737767218@linutronix.de
|
|
On PREEMPT_RT enabled kernels hrtimers which are not explicitely marked for
hard interrupt expiry mode are moved into soft interrupt context either for
latency reasons or because the hrtimer callback takes regular spinlocks or
invokes other functions which are not suitable for hard interrupt context
on PREEMPT_RT.
The hrtimer_sleeper callback is RT compatible in hard interrupt context,
but there is a latency concern: Untrusted userspace can spawn many threads
which arm timers for the same expiry time on the same CPU. On expiry that
causes a latency spike due to the wakeup of a gazillion threads.
OTOH, priviledged real-time user space applications rely on the low latency
of hard interrupt wakeups. These syscall related wakeups are all based on
hrtimer sleepers.
If the current task is in a real-time scheduling class, mark the mode for
hard interrupt expiry.
[ tglx: Split out of a larger combo patch. Added changelog ]
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190726185753.645792403@linutronix.de
|
