When assigning an interrupt to a vcpu, it is not unlikely that
the level of the hierarchy implementing irq_set_vcpu_affinity
is not the top level (think a generic MSI domain on top of a
virtualization aware interrupt controller).
In such a case, let's iterate over the hierarchy until we find
an irqchip implementing it.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
That commit was part of the changes moving x86 to the generic CPU hotplug
interrupt migration code. The force flag was required on x86 before the
hierarchical irqdomain rework, but invoking set_affinity() with force=true
stayed and had no side effects.
At some point in the past, the force flag got repurposed to support the
exynos timer interrupt affinity setting to a not yet online CPU, so the
interrupt controller callback does not verify the supplied affinity mask
against cpu_online_mask.
Setting the flag in the CPU hotplug code causes the cpu online masking to
be blocked on these irq controllers and results in potentially affining an
interrupt to the CPU which is unplugged, i.e. instead of moving it away,
it's just reassigned to it.
As the force flags is not longer needed on x86, it's safe to revert that
patch so the ARM irqchips which use the force flag work again.
Add comments to that effect, so this won't happen again.
Note: The online mask handling should be done in the generic code and the
force flag and the masking in the irq chips removed all together, but
that's not a change possible for 4.13.
Fixes: 77f85e66aa ("genirq/cpuhotplug: Set force affinity flag on hotplug migration")
Reported-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Will Deacon <will.deacon@arm.com>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: LAK <linux-arm-kernel@lists.infradead.org>
Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1707271217590.3109@nanos
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Interrupts with the IRQF_FORCE_RESUME flag set have also the
IRQF_NO_SUSPEND flag set. They are not disabled in the suspend path, but
must be forcefully resumed. That's used by XEN to keep IPIs enabled beyond
the suspension of device irqs. Force resume works by pretending that the
interrupt was disabled and then calling __irq_enable().
Incrementing the disabled depth counter was enough to do that, but with the
recent changes which use state flags to avoid unnecessary hardware access,
this is not longer sufficient. If the state flags are not set, then the
hardware callbacks are not invoked and the interrupt line stays disabled in
"hardware".
Set the disabled and masked state when pretending that an interrupt got
disabled by suspend.
Fixes: bf22ff45be ("genirq: Avoid unnecessary low level irq function calls")
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: xen-devel@lists.xenproject.org
Cc: boris.ostrovsky@oracle.com
Link: http://lkml.kernel.org/r/20170717174703.4603-2-jgross@suse.com
Pull irq fix from Thomas Gleixner:
"Fix the fallout from reworking the locking and resource management in
request/free_irq()"
* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq: Keep chip buslock across irq_request/release_resources()
Moving the irq_request/release_resources() callbacks out of the spinlocked,
irq disabled and bus locked region, unearthed an interesting abuse of the
irq_bus_lock/irq_bus_sync_unlock() callbacks.
The OMAP GPIO driver does merily power management inside of them. The
irq_request_resources() callback of this GPIO irqchip calls a function
which reads a GPIO register. That read aborts now because the clock of the
GPIO block is not magically enabled via the irq_bus_lock() callback.
Move the callbacks under the bus lock again to prevent this. In the
free_irq() path this requires to drop the bus_lock before calling
synchronize_irq() and reaquiring it before calling the
irq_release_resources() callback.
The bus lock can't be held because:
1) The data which has been changed between bus_lock/un_lock is cached in
the irq chip driver private data and needs to go out to the irq chip
via the slow bus (usually SPI or I2C) before calling
synchronize_irq().
That's the reason why this bus_lock/unlock magic exists in the first
place, as you cannot do SPI/I2C transactions while holding desc->lock
with interrupts disabled.
2) synchronize_irq() will actually deadlock, if there is a handler on
flight. These chips use threaded handlers for obvious reasons, as
they allow to do SPI/I2C communication. When the threaded handler
returns then bus_lock needs to be taken in irq_finalize_oneshot() as
we need to talk to the actual irq chip once more. After that the
threaded handler is marked done, which makes synchronize_irq() return.
So if we hold bus_lock accross the synchronize_irq() call, the
handler cannot mark itself done because it blocks on the bus
lock. That in turn makes synchronize_irq() wait forever on the
threaded handler to complete....
Add the missing unlock of desc->request_mutex in the error path of
__free_irq() and add a bunch of comments to explain the locking and
protection rules.
Fixes: 46e48e2573 ("genirq: Move irq resource handling out of spinlocked region")
Reported-and-tested-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Reported-and-tested-by: Tony Lindgren <tony@atomide.com>
Reported-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Not-longer-ranted-at-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Grygorii Strashko <grygorii.strashko@ti.com>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Pull irq fixes from Thomas Gleixner:
- A few fixes mopping up the fallout of the big irq overhaul
- Move the interrupt resource management logic out of the spin locked,
irq disabled region to avoid unnecessary restrictions of the resource
callbacks
- Preparation for reworking the per cpu irq request function.
* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqdomain: Allow ACPI device nodes to be used as irqdomain identifiers
genirq/debugfs: Remove redundant NULL pointer check
genirq: Allow to pass the IRQF_TIMER flag with percpu irq request
genirq/timings: Move free timings out of spinlocked region
genirq: Move irq resource handling out of spinlocked region
genirq: Add mutex to irq desc to serialize request/free_irq()
genirq: Move bus locking into __setup_irq()
genirq: Force inlining of __irq_startup_managed to prevent build failure
genirq/debugfs: Fix build for !CONFIG_IRQ_DOMAIN
A number of irqchip implementations are (ab)using the irqdomain allocator
by passing a fwnode that is neither a FWNODE_OF or a FWNODE_IRQCHIP.
This is pretty bad, but it also feels pretty crap to force these drivers to
allocate their own irqchip_fwid when they already have a proper fwnode.
Instead, let's teach the irqdomain allocator about ACPI device nodes, and
add some lovely name generation code... Tested on an arm64 D05 system.
Reported-and-tested-by: John Garry <john.garry@huawei.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Agustin Vega-Frias <agustinv@codeaurora.org>
Cc: Ma Jun <majun258@huawei.com>
Cc: Hanjun Guo <hanjun.guo@linaro.org>
Link: http://lkml.kernel.org/r/20170707083959.10349-1-marc.zyngier@arm.com
debugfs_remove() can be called with a NULL pointer.
Fixes: 087cdfb662 ("genirq/debugfs: Add proper debugfs interface")
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The irq timings infrastructure tracks when interrupts occur in order to
statistically predict te next interrupt event.
There is no point to track timer interrupts and try to predict them because
the next expiration time is already known. This can be avoided via the
IRQF_TIMER flag which is passed by timer drivers in request_irq(). It marks
the interrupt as timer based which alloes to ignore these interrupts in the
timings code.
Per CPU interrupts which are requested via request_percpu_+irq() have no
flag argument, so marking per cpu timer interrupts is not possible and they
get tracked pointlessly.
Add __request_percpu_irq() as a variant of request_percpu_irq() with a
flags argument and make request_percpu_irq() an inline wrapper passing
flags = 0.
The flag parameter is restricted to IRQF_TIMER as all other IRQF_ flags
make no sense for per cpu interrupts.
The next step is to convert all existing users of request_percpu_irq() and
then remove the wrapper and the underscores.
[ tglx: Massaged changelog ]
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: peterz@infradead.org
Cc: nicolas.pitre@linaro.org
Cc: vincent.guittot@linaro.org
Cc: rafael@kernel.org
Link: http://lkml.kernel.org/r/1499344144-3964-1-git-send-email-daniel.lezcano@linaro.org
No point to do memory management from a interrupt disabled spin locked
region.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Heiko Stuebner <heiko@sntech.de>
Cc: Julia Cartwright <julia@ni.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Brian Norris <briannorris@chromium.org>
Cc: Doug Anderson <dianders@chromium.org>
Cc: linux-rockchip@lists.infradead.org
Cc: John Keeping <john@metanate.com>
Cc: linux-gpio@vger.kernel.org
Link: http://lkml.kernel.org/r/20170629214344.196130646@linutronix.de
Aside of being conceptually wrong, there is also an actual (hard to
trigger and mostly theoretical) problem.
CPU0 CPU1
free_irq(X) interrupt X
spin_lock(desc->lock)
wake irq thread()
spin_unlock(desc->lock)
spin_lock(desc->lock)
remove action()
shutdown_irq()
release_resources() thread_handler()
spin_unlock(desc->lock) access released resources.
synchronize_irq()
Move the release resources invocation after synchronize_irq() so it's
guaranteed that the threaded handler has finished.
Move the resource request call out of the desc->lock held region as well,
so the invocation context is the same for both request and release.
This solves the problems with those functions on RT as well.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Heiko Stuebner <heiko@sntech.de>
Cc: Julia Cartwright <julia@ni.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Brian Norris <briannorris@chromium.org>
Cc: Doug Anderson <dianders@chromium.org>
Cc: linux-rockchip@lists.infradead.org
Cc: John Keeping <john@metanate.com>
Cc: linux-gpio@vger.kernel.org
Link: http://lkml.kernel.org/r/20170629214344.117028181@linutronix.de
The irq_request/release_resources() callbacks ar currently invoked under
desc->lock with interrupts disabled. This is a source of problems on RT and
conceptually not required.
Add a seperate mutex to struct irq_desc which allows to serialize
request/free_irq(), which can be used to move the resource functions out of
the desc->lock held region.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Heiko Stuebner <heiko@sntech.de>
Cc: Julia Cartwright <julia@ni.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Brian Norris <briannorris@chromium.org>
Cc: Doug Anderson <dianders@chromium.org>
Cc: linux-rockchip@lists.infradead.org
Cc: John Keeping <john@metanate.com>
Cc: linux-gpio@vger.kernel.org
Link: http://lkml.kernel.org/r/20170629214344.039220922@linutronix.de
There is no point in having the irq_bus_lock() protection around all
callers to __setup_irq().
Move it into __setup_irq(). This is also a preparatory patch for addressing
the issues with the irq resource callbacks.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Cc: Heiko Stuebner <heiko@sntech.de>
Cc: Julia Cartwright <julia@ni.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Brian Norris <briannorris@chromium.org>
Cc: Doug Anderson <dianders@chromium.org>
Cc: linux-rockchip@lists.infradead.org
Cc: John Keeping <john@metanate.com>
Cc: linux-gpio@vger.kernel.org
Link: http://lkml.kernel.org/r/20170629214343.960949031@linutronix.de
If CONFIG_SMP=n, and gcc (e.g. 4.1.2) decides not to inline
__irq_startup_managed(), the build fails with:
kernel/built-in.o: In function `irq_startup':
(.text+0x38ed8): undefined reference to `irq_set_affinity_locked'
Fix this by forcing inlining of __irq_startup_managed().
Fixes: 761ea388e8 ("genirq: Handle managed irqs gracefully in irq_startup()")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Link: http://lkml.kernel.org/r/1499162761-12398-1-git-send-email-geert@linux-m68k.org
Fix this build error:
kernel/irq/internals.h:440:20: error: inlining failed in call to always_inline
'irq_domain_debugfs_init': function body not available
kernel/irq/debugfs.c:202:2: note: called from here
irq_domain_debugfs_init(root_dir);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/alpine.LFD.2.20.1707041124000.1712@schleppi
around. Highlights include:
- Conversion of a bunch of security documentation into RST
- The conversion of the remaining DocBook templates by The Amazing
Mauro Machine. We can now drop the entire DocBook build chain.
- The usual collection of fixes and minor updates.
-----BEGIN PGP SIGNATURE-----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=cVjZ
-----END PGP SIGNATURE-----
Merge tag 'docs-4.13' of git://git.lwn.net/linux
Pull documentation updates from Jonathan Corbet:
"There has been a fair amount of activity in the docs tree this time
around. Highlights include:
- Conversion of a bunch of security documentation into RST
- The conversion of the remaining DocBook templates by The Amazing
Mauro Machine. We can now drop the entire DocBook build chain.
- The usual collection of fixes and minor updates"
* tag 'docs-4.13' of git://git.lwn.net/linux: (90 commits)
scripts/kernel-doc: handle DECLARE_HASHTABLE
Documentation: atomic_ops.txt is core-api/atomic_ops.rst
Docs: clean up some DocBook loose ends
Make the main documentation title less Geocities
Docs: Use kernel-figure in vidioc-g-selection.rst
Docs: fix table problems in ras.rst
Docs: Fix breakage with Sphinx 1.5 and upper
Docs: Include the Latex "ifthen" package
doc/kokr/howto: Only send regression fixes after -rc1
docs-rst: fix broken links to dynamic-debug-howto in kernel-parameters
doc: Document suitability of IBM Verse for kernel development
Doc: fix a markup error in coding-style.rst
docs: driver-api: i2c: remove some outdated information
Documentation: DMA API: fix a typo in a function name
Docs: Insert missing space to separate link from text
doc/ko_KR/memory-barriers: Update control-dependencies example
Documentation, kbuild: fix typo "minimun" -> "minimum"
docs: Fix some formatting issues in request-key.rst
doc: ReSTify keys-trusted-encrypted.txt
doc: ReSTify keys-request-key.txt
...
An interrupt behaves with a burst of activity with periodic interval of time
followed by one or two peaks of longer interval.
As the time intervals are periodic, statistically speaking they follow a normal
distribution and each interrupts can be tracked individually.
Add a mechanism to compute the statistics on all interrupts, except the
timers which are deterministic from a prediction point of view, as their
expiry time is known.
The goal is to extract the periodicity for each interrupt, with the last
timestamp and sum them, so the next event can be predicted to a certain
extent.
Taking the earliest prediction gives the expected wakeup on the system
(assuming a timer won't expire before).
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: "Rafael J . Wysocki" <rafael@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Link: http://lkml.kernel.org/r/1498227072-5980-2-git-send-email-daniel.lezcano@linaro.org
The interrupt framework gives a lot of information about each interrupt. It
does not keep track of when those interrupts occur though, which is a
prerequisite for estimating the next interrupt arrival for power management
purposes.
Add a mechanism to record the timestamp for each interrupt occurrences in a
per-CPU circular buffer to help with the prediction of the next occurrence
using a statistical model.
Each CPU can store up to IRQ_TIMINGS_SIZE events <irq, timestamp>, the
current value of IRQ_TIMINGS_SIZE is 32.
Each event is encoded into a single u64, where the high 48 bits are used
for the timestamp and the low 16 bits are for the irq number.
A static key is introduced so when the irq prediction is switched off at
runtime, the overhead is near to zero.
It results in most of the code in internals.h for inline reasons and a very
few in the new file timings.c. The latter will contain more in the next patch
which will provide the statistical model for the next event prediction.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: "Rafael J . Wysocki" <rafael@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Link: http://lkml.kernel.org/r/1498227072-5980-1-git-send-email-daniel.lezcano@linaro.org
debugfs_remove() has it's own NULL pointer check. Remove the conditional
and make irq_remove_debugfs_entry() an inline helper
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
It did seem like a good idea at the time, but it never really
caught on, and auto-recursive domains remain unused 3 years after
having been introduced.
Oh well, time for a late spring cleanup.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
We can have irq domains that are identified by the same fwnode
(because they are serviced by the same HW), and yet have different
functionnality (because they serve different busses, for example).
This is what we use the bus_token field.
Since we don't use this field when generating the domain name,
all the aliasing domains will get the same name, and the debugfs
file creation fails. Also, bus_token is updated by individual drivers,
and the core code is unaware of that update.
In order to sort this mess, let's introduce a helper that takes care
of updating bus_token, and regenerate the debugfs file.
A separate patch will update all the individual users.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Currently the irq vector spread algorithm is restricted to online CPUs,
which ties the IRQ mapping to the currently online devices and doesn't deal
nicely with the fact that CPUs could come and go rapidly due to e.g. power
management.
Instead assign vectors to all present CPUs to avoid this churn.
Build a map of all possible CPUs for a given node, as the architectures
only provide a map of all onlines CPUs. Do this dynamically on each call
for the vector assingments, which is a bit suboptimal and could be
optimized in the future by provinding a mapping from the arch code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: linux-block@vger.kernel.org
Cc: Sagi Grimberg <sagi@grimberg.me>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: linux-nvme@lists.infradead.org
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20170603140403.27379-5-hch@lst.de
Avoid trying to add a newly online CPU to the effective affinity mask of an
started up interrupt. That interrupt will either stay on the already online
CPU or move around for no value.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235447.431321047@linutronix.de
Many interrupt chips allow only a single CPU as interrupt target. The core
code has no knowledge about that. That's unfortunate as it could avoid
trying to readd a newly online CPU to the effective affinity mask.
Add the status flag and the necessary accessors.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235447.352343969@linutronix.de
If a CPU goes offline, interrupts affine to the CPU are moved away. If the
outgoing CPU is the last CPU in the affinity mask the migration code breaks
the affinity and sets it it all online cpus.
This is a problem for affinity managed interrupts as CPU hotplug is often
used for power management purposes. If the affinity is broken, the
interrupt is not longer affine to the CPUs to which it was allocated.
The affinity spreading allows to lay out multi queue devices in a way that
they are assigned to a single CPU or a group of CPUs. If the last CPU goes
offline, then the queue is not longer used, so the interrupt can be
shutdown gracefully and parked until one of the assigned CPUs comes online
again.
Add a graceful shutdown mechanism into the irq affinity breaking code path,
mark the irq as MANAGED_SHUTDOWN and leave the affinity mask unmodified.
In the online path, scan the active interrupts for managed interrupts and
if the interrupt is functional and the newly online CPU is part of the
affinity mask, restart the interrupt if it is marked MANAGED_SHUTDOWN or if
the interrupts is started up, try to add the CPU back to the effective
affinity mask.
Originally-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20170619235447.273417334@linutronix.de
Affinity managed interrupts should keep their assigned affinity accross CPU
hotplug. To avoid magic hackery in device drivers, the core code shall
manage them transparently and set these interrupts into a managed shutdown
state when the last CPU of the assigned affinity mask goes offline. The
interrupt will be restarted when one of the CPUs in the assigned affinity
mask comes back online.
Add the necessary logic to irq_startup(). If an interrupt is requested and
started up, the code checks whether it is affinity managed and if so, it
checks whether a CPU in the interrupts affinity mask is online. If not, it
puts the interrupt into managed shutdown state.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235447.189851170@linutronix.de
In order to handle managed interrupts gracefully on irq_startup() so they
won't lose their assigned affinity, it's necessary to allow startups which
keep the interrupts in managed shutdown state, if none of the assigend CPUs
is online. This allows drivers to request interrupts w/o the CPUs being
online, which avoid online/offline churn in drivers.
Add a force argument which can override that decision and let only
request_irq() and enable_irq() allow the managed shutdown
handling. enable_irq() is required, because the interrupt might be
requested with IRQF_NOAUTOEN and enable_irq() invokes irq_startup() which
would then wreckage the assignment again. All other callers force startup
and potentially break the assigned affinity.
No functional change as this only adds the function argument.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235447.112094565@linutronix.de
Split out the inner workings of irq_startup() so it can be reused to handle
managed interrupts gracefully.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235447.033235144@linutronix.de
Affinity managed interrupts should keep their assigned affinity accross CPU
hotplug. To avoid magic hackery in device drivers, the core code shall
manage them transparently. This will set these interrupts into a managed
shutdown state when the last CPU of the assigned affinity mask goes
offline. The interrupt will be restarted when one of the CPUs in the
assigned affinity mask comes back online.
Introduce the necessary state flag and the accessor functions.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235446.954523476@linutronix.de
If the architecture supports the effective affinity mask, migrating
interrupts away which are not targeted by the effective mask is
pointless.
They can stay in the user or system supplied affinity mask, but won't be
targetted at any given point as the affinity setter functions need to
validate against the online cpu mask anyway.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235446.328488490@linutronix.de
There is currently no way to evaluate the effective affinity mask of a
given interrupt. Many irq chips allow only a single target CPU or a subset
of CPUs in the affinity mask.
Updating the mask at the time of setting the affinity to the subset would
be counterproductive because information for cpu hotplug about assigned
interrupt affinities gets lost. On CPU hotplug it's also pointless to force
migrate an interrupt, which is not targeted at the CPU effectively. But
currently the information is not available.
Provide a seperate mask to be updated by the irq_chip->irq_set_affinity()
implementations. Implement the read only proc files so the user can see the
effective mask as well w/o trying to deduce it from /proc/interrupts.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235446.247834245@linutronix.de
The proc file setup repeats the same ugly type cast for the irq number over
and over. Do it once and hand in the local void pointer.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235446.160866358@linutronix.de
All callers hand in GPF_KERNEL. No point to have an extra argument for
that.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235446.082544752@linutronix.de
The third argument of the internal helper function is unused. Remove it.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235446.004958600@linutronix.de
Now that x86 uses the generic code, the function declaration and inline
stub can move to the core internal header.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.928156166@linutronix.de
Set the force migration flag when migrating interrupts away from an
outgoing CPU.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.681874648@linutronix.de
Interrupts which cannot be migrated in process context, need to be masked
before the affinity is changed forcefully.
Add support for that. Will be compiled out for architectures which do not
have this x86 specific issue.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.604565591@linutronix.de
In order to move x86 to the generic hotplug migration code, add support for
cleaning up move in progress bits.
On architectures which have this x86 specific (mis)feature not enabled,
this is optimized out by the compiler.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.525817311@linutronix.de
Interrupts, which are shut down are tried to be migrated as well. That's
pointless because the interrupt cannot fire and the next startup will move
it to the proper place anyway.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.447550992@linutronix.de
Move the checks for a valid irq chip and the irq_set_affinity() callback
right in front of the whole migration logic. No point in doing a gazillion
of other things when the interrupt cannot be migrated at all.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.354181630@linutronix.de
In case the affinity of an interrupt was broken, a printk is emitted.
But if the affinity cannot be set at all due to a missing
irq_set_affinity() callback or due to a failing callback, the message is
still printed preceeded by a warning/error.
That makes no sense whatsoever.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.274852976@linutronix.de
This is called from stop_machine() with interrupts disabled. No point in
disabling them some more.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.198042748@linutronix.de
So that the affinity code can reuse them.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20170619235445.109426284@linutronix.de
The startup vs. setaffinity ordering of interrupts depends on the
IRQF_NOAUTOEN flag. Chained interrupts are not getting any affinity
assignment at all.
A regular interrupt is started up and then the affinity is set. A
IRQF_NOAUTOEN marked interrupt is not started up, but the affinity is set
nevertheless.
Move the affinity setup to startup_irq() so the ordering is always the same
and chained interrupts get the proper default affinity assigned as well.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235445.020534783@linutronix.de
Rename it with a proper irq_ prefix and make it available for other files
in the core code. Preparatory patch for moving the irq affinity setup
around.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235444.928501004@linutronix.de
No point to have this alloc/free dance of cpumasks. Provide a static mask
for setup_affinity() and protect it proper.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235444.851571573@linutronix.de
If an CPU goes offline, the interrupts are migrated away, but a eventually
pending interrupt move, which has not yet been made effective is kept
pending even if the outgoing CPU is the sole target of the pending affinity
mask. What's worse is, that the pending affinity mask is discarded even if
it would contain a valid subset of the online CPUs.
Implement a helper function which allows to avoid these issues.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Link: http://lkml.kernel.org/r/20170619235444.691345468@linutronix.de
