KVM devices were manipulating list data structures without any form of
synchronization, and some implementations of the create operations also
suffered from a lack of synchronization.
Now when we've split the xics create operation into create and init, we
can hold the kvm->lock mutex while calling the create operation and when
manipulating the devices list.
The error path in the generic code gets slightly ugly because we have to
take the mutex again and delete the device from the list, but holding
the mutex during anon_inode_getfd or releasing/locking the mutex in the
common non-error path seemed wrong.
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
As we are about to hold the kvm->lock during the create operation on KVM
devices, we should move the call to xics_debugfs_init into its own
function, since holding a mutex over extended amounts of time might not
be a good idea.
Introduce an init operation on the kvm_device_ops struct which cannot
fail and call this, if configured, after the device has been created.
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
KVM/ARM Changes for v4.8 - Take 2
Includes GSI routing support to go along with the new VGIC and a small fix that
has been cooking in -next for a while.
KVM/ARM changes for Linux 4.8
- GICv3 ITS emulation
- Simpler idmap management that fixes potential TLB conflicts
- Honor the kernel protection in HYP mode
- Removal of the old vgic implementation
Up to now, only irqchip routing entries could be set. This patch
adds the capability to insert MSI routing entries.
For ARM64, let's also increase KVM_MAX_IRQ_ROUTES to 4096: this
include SPI irqchip routes plus MSI routes. In the future this
might be extended.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
kvm_setup_default_irq_routing and kvm_setup_empty_irq_routing are
not used by generic code. So let's move the declarations in x86 irq.h
header instead of kvm_host.h.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Suggested-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Enhance kvm_kernel_irq_routing_entry to transport the device id
field, devid. A new flags field makes possible to indicate the
devid is valid. Those additions are used for ARM GICv3 ITS MSI
injection. The original struct msi_msg msi field is replaced by
a new custom structure that embeds the new fields.
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Suggested-by: Radim Krčmář <rkrcmar@redhat.com>
Acked-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
The kvm_io_bus framework is a nice place of holding information about
various MMIO regions for kernel emulated devices.
Add a call to retrieve the kvm_io_device structure which is associated
with a certain MMIO address. This avoids to duplicate kvm_io_bus'
knowledge of MMIO regions without having to fake MMIO calls if a user
needs the device a certain MMIO address belongs to.
This will be used by the ITS emulation to get the associated ITS device
when someone triggers an MSI via an ioctl from userspace.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <marc.zyngier@arm.com>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Tested-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Make kvm_guest_{enter,exit} and __kvm_guest_{enter,exit} trivial wrappers
around the code in context_tracking.h. Name the context_tracking.h functions
consistently with what those for kernel<->user switch.
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The new created_vcpus field makes it possible to avoid the race between
irqchip and VCPU creation in a much nicer way; just check under kvm->lock
whether a VCPU has already been created.
We can then remove KVM_APIC_ARCHITECTURE too, because at this point the
symbol is only governing the default definition of kvm_vcpu_compatible.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The race between creating the irqchip and the first VCPU is
currently fixed by checking the presence of an irqchip before
updating kvm->online_vcpus, and undoing the whole VCPU creation
if someone created the irqchip in the meanwhile.
Instead, introduce a new field in struct kvm that will count VCPUs
under a mutex, without the atomic access and memory ordering that we
need elsewhere to protect the vcpus array. This also plugs the race
and is more easily applicable in all similar circumstances.
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This patch adds a kvm debugfs subdirectory for each VM, which is named
after its pid and file descriptor. The directories contain the same
kind of files that are already in the kvm debugfs directory, but the
data exported through them is now VM specific.
This makes the debugfs kvm data a convenient alternative to the
tracepoints which already have per VM data. The debugfs data is easy
to read and low overhead.
CC: Dan Carpenter <dan.carpenter@oracle.com> [includes fixes by Dan Carpenter]
Signed-off-by: Janosch Frank <frankja@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Some wakeups should not be considered a sucessful poll. For example on
s390 I/O interrupts are usually floating, which means that _ALL_ CPUs
would be considered runnable - letting all vCPUs poll all the time for
transactional like workload, even if one vCPU would be enough.
This can result in huge CPU usage for large guests.
This patch lets architectures provide a way to qualify wakeups if they
should be considered a good/bad wakeups in regard to polls.
For s390 the implementation will fence of halt polling for anything but
known good, single vCPU events. The s390 implementation for floating
interrupts does a wakeup for one vCPU, but the interrupt will be delivered
by whatever CPU checks first for a pending interrupt. We prefer the
woken up CPU by marking the poll of this CPU as "good" poll.
This code will also mark several other wakeup reasons like IPI or
expired timers as "good". This will of course also mark some events as
not sucessful. As KVM on z runs always as a 2nd level hypervisor,
we prefer to not poll, unless we are really sure, though.
This patch successfully limits the CPU usage for cases like uperf 1byte
transactional ping pong workload or wakeup heavy workload like OLTP
while still providing a proper speedup.
This also introduced a new vcpu stat "halt_poll_no_tuning" that marks
wakeups that are considered not good for polling.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Radim Krčmář <rkrcmar@redhat.com> (for an earlier version)
Cc: David Matlack <dmatlack@google.com>
Cc: Wanpeng Li <kernellwp@gmail.com>
[Rename config symbol. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If we don't support a mechanism for bypassing IRQs, don't register as
a consumer. This eliminates meaningless dev_info()s when the connect
fails between producer and consumer, such as on AMD systems where
kvm_x86_ops->update_pi_irte is not implemented
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The KVM_MAX_VCPUS define provides the maximum number of vCPUs per guest, and
also the upper limit for vCPU ids. This is okay for all archs except PowerPC
which can have higher ids, depending on the cpu/core/thread topology. In the
worst case (single threaded guest, host with 8 threads per core), it limits
the maximum number of vCPUS to KVM_MAX_VCPUS / 8.
This patch separates the vCPU numbering from the total number of vCPUs, with
the introduction of KVM_MAX_VCPU_ID, as the maximal valid value for vCPU ids
plus one.
The corresponding KVM_CAP_MAX_VCPU_ID allows userspace to validate vCPU ids
before passing them to KVM_CREATE_VCPU.
This patch only implements KVM_MAX_VCPU_ID with a specific value for PowerPC.
Other archs continue to return KVM_MAX_VCPUS instead.
Suggested-by: Radim Krcmar <rkrcmar@redhat.com>
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Commit c896939f7c ("KVM: use heuristic for fast VCPU lookup by id") added
a return path that prevents vcpu ids to exceed KVM_MAX_VCPUS. This is a
problem for powerpc where vcpu ids can grow up to 8*KVM_MAX_VCPUS.
This patch simply reverses the logic so that we only try fast path if the
vcpu id can be tried as an index in kvm->vcpus[]. The slow path is not
affected by the change.
Reviewed-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
kvm_make_request and kvm_check_request imply a producer-consumer
relationship; add implicit memory barriers to them. There was indeed
already a place that was adding an explicit smp_mb() to order between
kvm_check_request and the processing of the request. That memory
barrier can be removed (as an added benefit, kvm_check_request can use
smp_mb__after_atomic which is free on x86).
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The problem:
On -rt, an emulated LAPIC timer instances has the following path:
1) hard interrupt
2) ksoftirqd is scheduled
3) ksoftirqd wakes up vcpu thread
4) vcpu thread is scheduled
This extra context switch introduces unnecessary latency in the
LAPIC path for a KVM guest.
The solution:
Allow waking up vcpu thread from hardirq context,
thus avoiding the need for ksoftirqd to be scheduled.
Normal waitqueues make use of spinlocks, which on -RT
are sleepable locks. Therefore, waking up a waitqueue
waiter involves locking a sleeping lock, which
is not allowed from hard interrupt context.
cyclictest command line:
This patch reduces the average latency in my tests from 14us to 11us.
Daniel writes:
Paolo asked for numbers from kvm-unit-tests/tscdeadline_latency
benchmark on mainline. The test was run 1000 times on
tip/sched/core 4.4.0-rc8-01134-g0905f04:
./x86-run x86/tscdeadline_latency.flat -cpu host
with idle=poll.
The test seems not to deliver really stable numbers though most of
them are smaller. Paolo write:
"Anything above ~10000 cycles means that the host went to C1 or
lower---the number means more or less nothing in that case.
The mean shows an improvement indeed."
Before:
min max mean std
count 1000.000000 1000.000000 1000.000000 1000.000000
mean 5162.596000 2019270.084000 5824.491541 20681.645558
std 75.431231 622607.723969 89.575700 6492.272062
min 4466.000000 23928.000000 5537.926500 585.864966
25% 5163.000000 1613252.750000 5790.132275 16683.745433
50% 5175.000000 2281919.000000 5834.654000 23151.990026
75% 5190.000000 2382865.750000 5861.412950 24148.206168
max 5228.000000 4175158.000000 6254.827300 46481.048691
After
min max mean std
count 1000.000000 1000.00000 1000.000000 1000.000000
mean 5143.511000 2076886.10300 5813.312474 21207.357565
std 77.668322 610413.09583 86.541500 6331.915127
min 4427.000000 25103.00000 5529.756600 559.187707
25% 5148.000000 1691272.75000 5784.889825 17473.518244
50% 5160.000000 2308328.50000 5832.025000 23464.837068
75% 5172.000000 2393037.75000 5853.177675 24223.969976
max 5222.000000 3922458.00000 6186.720500 42520.379830
[Patch was originaly based on the swait implementation found in the -rt
tree. Daniel ported it to mainline's version and gathered the
benchmark numbers for tscdeadline_latency test.]
Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-rt-users@vger.kernel.org
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/1455871601-27484-4-git-send-email-wagi@monom.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
To date, we have implemented two I/O usage models for persistent memory,
PMEM (a persistent "ram disk") and DAX (mmap persistent memory into
userspace). This series adds a third, DAX-GUP, that allows DAX mappings
to be the target of direct-i/o. It allows userspace to coordinate
DMA/RDMA from/to persistent memory.
The implementation leverages the ZONE_DEVICE mm-zone that went into
4.3-rc1 (also discussed at kernel summit) to flag pages that are owned
and dynamically mapped by a device driver. The pmem driver, after
mapping a persistent memory range into the system memmap via
devm_memremap_pages(), arranges for DAX to distinguish pfn-only versus
page-backed pmem-pfns via flags in the new pfn_t type.
The DAX code, upon seeing a PFN_DEV+PFN_MAP flagged pfn, flags the
resulting pte(s) inserted into the process page tables with a new
_PAGE_DEVMAP flag. Later, when get_user_pages() is walking ptes it keys
off _PAGE_DEVMAP to pin the device hosting the page range active.
Finally, get_page() and put_page() are modified to take references
against the device driver established page mapping.
Finally, this need for "struct page" for persistent memory requires
memory capacity to store the memmap array. Given the memmap array for a
large pool of persistent may exhaust available DRAM introduce a
mechanism to allocate the memmap from persistent memory. The new
"struct vmem_altmap *" parameter to devm_memremap_pages() enables
arch_add_memory() to use reserved pmem capacity rather than the page
allocator.
This patch (of 18):
The core has developed a need for a "pfn_t" type [1]. Move the existing
pfn_t in KVM to kvm_pfn_t [2].
[1]: https://lists.01.org/pipermail/linux-nvdimm/2015-September/002199.html
[2]: https://lists.01.org/pipermail/linux-nvdimm/2015-September/002218.html
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since the numbers now overlap, it makes sense to enumerate
them in asm/kvm_host.h rather than linux/kvm_host.h. Functions
that refer to architecture-specific requests are also moved
to arch/.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
