These days, we experienced one guest crash with 8 cores and 3 disks,
with qemu error logs as bellow:
qemu-system-x86_64: /build/qemu-2.0.0/kvm-all.c:984:
kvm_irqchip_commit_routes: Assertion `ret == 0' failed.
And then we found one patch(bdf026317d) in qemu tree, which said
could fix this bug.
Execute the following script will reproduce the BUG quickly:
irq_affinity.sh
========================================================================
vda_irq_num=25
vdb_irq_num=27
while [ 1 ]
do
for irq in {1,2,4,8,10,20,40,80}
do
echo $irq > /proc/irq/$vda_irq_num/smp_affinity
echo $irq > /proc/irq/$vdb_irq_num/smp_affinity
dd if=/dev/vda of=/dev/zero bs=4K count=100 iflag=direct
dd if=/dev/vdb of=/dev/zero bs=4K count=100 iflag=direct
done
done
========================================================================
The following qemu log is added in the qemu code and is displayed when
this bug reproduced:
kvm_irqchip_commit_routes: max gsi: 1008, nr_allocated_irq_routes: 1024,
irq_routes->nr: 1024, gsi_count: 1024.
That's to say when irq_routes->nr == 1024, there are 1024 routing entries,
but in the kernel code when routes->nr >= 1024, will just return -EINVAL;
The nr is the number of the routing entries which is in of
[1 ~ KVM_MAX_IRQ_ROUTES], not the index in [0 ~ KVM_MAX_IRQ_ROUTES - 1].
This patch fix the BUG above.
Cc: stable@vger.kernel.org
Signed-off-by: Xiubo Li <lixiubo@cmss.chinamobile.com>
Signed-off-by: Wei Tang <tangwei@cmss.chinamobile.com>
Signed-off-by: Zhang Zhuoyu <zhangzhuoyu@cmss.chinamobile.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When changing the active bit from an MMIO trap, we decide to
explode if the intid is that of a private interrupt.
This flawed logic comes from the fact that we were assuming that
kvm_vcpu_kick() as called by kvm_arm_halt_vcpu() would not return before
the called vcpu responded, but this is not the case, so we need to
perform this wait even for private interrupts.
Dropping the BUG_ON seems like the right thing to do.
[ Commit message tweaked by Christoffer ]
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
When reading back from the list registers, we need to perform
two actions for level interrupts:
1) clear the soft-pending bit if the interrupt is not pending
anymore *in the list register*
2) resample the line level and propagate it to the pending state
But these two actions shouldn't be linked, and we should *always*
resample the line level, no matter what state is in the list
register. Otherwise, we may end-up injecting spurious interrupts
that have been already retired.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
When reading back from the list registers, we need to perform
two actions for level interrupts:
1) clear the soft-pending bit if the interrupt is not pending
anymore *in the list register*
2) resample the line level and propagate it to the pending state
But these two actions shouldn't be linked, and we should *always*
resample the line level, no matter what state is in the list
register. Otherwise, we may end-up injecting spurious interrupts
that have been already retired.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
When saving the state of the list registers, it is critical to
reset them zero, as we could otherwise leave unexpected EOI
interrupts pending for virtual level interrupts.
Cc: stable@vger.kernel.org # v4.6+
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.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>
KVM/ARM Changes for v4.7 take 2
"The GIC is dead; Long live the GIC"
This set of changes include the new vgic, which is a reimplementation of
our horribly broken legacy vgic implementation. The two implementations
will live side-by-side (with the new being the configured default) for
one kernel release and then we'll remove it.
Also fixes a non-critical issue with virtual abort injection to guests.
When modifying the active state of an interrupt via the MMIO interface,
we should ensure that the write has the intended effect.
If a guest sets an interrupt to active, but that interrupt is already
flushed into a list register on a running VCPU, then that VCPU will
write the active state back into the struct vgic_irq upon returning from
the guest and syncing its state. This is a non-benign race, because the
guest can observe that an interrupt is not active, and it can have a
reasonable expectations that other VCPUs will not ack any IRQs, and then
set the state to active, and expect it to stay that way. Currently we
are not honoring this case.
Thefore, change both the SACTIVE and CACTIVE mmio handlers to stop the
world, change the irq state, potentially queue the irq if we're setting
it to active, and then continue.
We take this chance to slightly optimize these functions by not stopping
the world when touching private interrupts where there is inherently no
possible race.
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Now that the new VGIC implementation has reached feature parity with
the old one, add the new files to the build system and add a Kconfig
option to switch between the two versions.
We set the default to the new version to get maximum test coverage,
in case people experience problems they can switch back to the old
behaviour if needed.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
We now store the mapped hardware IRQ number in our struct, so we
don't need the irq_phys_map for the new VGIC.
Implement the hardware IRQ mapping on top of the reworked arch
timer interface.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
Connect to the new VGIC to the irqfd framework, so that we can
inject IRQs.
GSI routing and MSI routing is not yet implemented.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
map_resources is the last initialization step. It is executed on
first VCPU run. At that stage the code checks that userspace has provided
the base addresses for the relevant VGIC regions, which depend on the
type of VGIC that is exposed to the guest. Also we check if the two
regions overlap.
If the checks succeeded, we register the respective register frames with
the kvm_io_bus framework.
If we emulate a GICv2, the function also forces vgic_init execution if
it has not been executed yet. Also we map the virtual GIC CPU interface
onto the guest's CPU interface.
Signed-off-by: Eric Auger <eric.auger@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
This patch allocates and initializes the data structures used
to model the vgic distributor and virtual cpu interfaces. At that
stage the number of IRQs and number of virtual CPUs is frozen.
Signed-off-by: Eric Auger <eric.auger@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
This patch implements the vgic_creation function which is
called on CREATE_IRQCHIP VM IOCTL (v2 only) or KVM_CREATE_DEVICE
Signed-off-by: Eric Auger <eric.auger@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
Implements kvm_vgic_hyp_init and vgic_probe function.
This uses the new firmware independent VGIC probing to support both ACPI
and DT based systems (code from Marc Zyngier).
The vgic_global struct is enriched with new fields populated
by those functions.
Signed-off-by: Eric Auger <eric.auger@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
Using the VMCR accessors we provide access to GIC CPU interface state
to userland by wiring it up to the existing userland interface.
[Marc: move and make VMCR accessors static, streamline MMIO handlers]
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
Since the GIC CPU interface is always virtualized by the hardware,
we don't have CPU interface state information readily available in our
emulation if userland wants to save or restore it.
Fortunately the GIC hypervisor interface provides the VMCR register to
access the required virtual CPU interface bits.
Provide wrappers for GICv2 and GICv3 hosts to have access to this
register.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
Userland may want to save and restore the state of the in-kernel VGIC,
so we provide the code which takes a userland request and translate
that into calls to our MMIO framework.
From Christoffer:
When accessing the VGIC state from userspace we really don't want a VCPU
to be messing with the state at the same time, and the API specifies
that we should return -EBUSY if any VCPUs are running.
Check and prevent VCPUs from running by grabbing their mutexes, one by
one, and error out if we fail.
(Note: This could potentially be simplified to just do a simple check
and see if any VCPUs are running, and return -EBUSY then, without
enforcing the locking throughout the duration of the uaccess, if we
think that taking/releasing all these mutexes for every single GIC
register access is too heavyweight.)
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Userland can access the emulated GIC to save and restore its state
for initialization or migration purposes.
The kvm_io_bus API requires an absolute gpa, which does not fit the
KVM_DEV_ARM_VGIC_GRP_DIST_REGS user API, that only provides relative
offsets. So we provide a wrapper to plug into our MMIO framework and
find the respective register handler.
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
This patch implements the switches for KVM_DEV_ARM_VGIC_GRP_DIST_REGS
and KVM_DEV_ARM_VGIC_GRP_CPU_REGS API which allows the userspace to
access VGIC registers.
Signed-off-by: Eric Auger <eric.auger@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
This patch implements the KVM_DEV_ARM_VGIC_GRP_ADDR group which
enables to set the base address of GIC regions as seen by the guest.
Signed-off-by: Eric Auger <eric.auger@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
kvm_vgic_addr is used by the userspace to set the base address of
the following register regions, as seen by the guest:
- distributor(v2 and v3),
- re-distributors (v3),
- CPU interface (v2).
Signed-off-by: Eric Auger <eric.auger@linaro.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Christoffer Dall <christoffer.dall@linaro.org>
