linux/include/kvm/arm_arch_timer.h

189 lines
4.9 KiB

KVM: arm/arm64: Support arch timers with a userspace gic If you're running with a userspace gic or other interrupt controller (that is no vgic in the kernel), then you have so far not been able to use the architected timers, because the output of the architected timers, which are driven inside the kernel, was a kernel-only construct between the arch timer code and the vgic. This patch implements the new KVM_CAP_ARM_USER_IRQ feature, where we use a side channel on the kvm_run structure, run->s.regs.device_irq_level, to always notify userspace of the timer output levels when using a userspace irqchip. This works by ensuring that before we enter the guest, if the timer output level has changed compared to what we last told userspace, we don't enter the guest, but instead return to userspace to notify it of the new level. If we are exiting, because of an MMIO for example, and the level changed at the same time, the value is also updated and userspace can sample the line as it needs. This is nicely achieved simply always updating the timer_irq_level field after the main run loop. Note that the kvm_timer_update_irq trace event is changed to show the host IRQ number for the timer instead of the guest IRQ number, because the kernel no longer know which IRQ userspace wires up the timer signal to. Also note that this patch implements all required functionality but does not yet advertise the capability. Reviewed-by: Alexander Graf <agraf@suse.de> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Alexander Graf <agraf@suse.de> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
2016-09-27 21:08:06 +02:00
KVM: arm/arm64: Avoid timer save/restore in vcpu entry/exit We don't need to save and restore the hardware timer state and examine if it generates interrupts on on every entry/exit to the guest. The timer hardware is perfectly capable of telling us when it has expired by signaling interrupts. When taking a vtimer interrupt in the host, we don't want to mess with the timer configuration, we just want to forward the physical interrupt to the guest as a virtual interrupt. We can use the split priority drop and deactivate feature of the GIC to do this, which leaves an EOI'ed interrupt active on the physical distributor, making sure we don't keep taking timer interrupts which would prevent the guest from running. We can then forward the physical interrupt to the VM using the HW bit in the LR of the GIC, like we do already, which lets the guest directly deactivate both the physical and virtual timer simultaneously, allowing the timer hardware to exit the VM and generate a new physical interrupt when the timer output is again asserted later on. We do need to capture this state when migrating VCPUs between physical CPUs, however, which we use the vcpu put/load functions for, which are called through preempt notifiers whenever the thread is scheduled away from the CPU or called directly if we return from the ioctl to userspace. One caveat is that we have to save and restore the timer state in both kvm_timer_vcpu_[put/load] and kvm_timer_[schedule/unschedule], because we can have the following flows: 1. kvm_vcpu_block 2. kvm_timer_schedule 3. schedule 4. kvm_timer_vcpu_put (preempt notifier) 5. schedule (vcpu thread gets scheduled back) 6. kvm_timer_vcpu_load (preempt notifier) 7. kvm_timer_unschedule And a version where we don't actually call schedule: 1. kvm_vcpu_block 2. kvm_timer_schedule 7. kvm_timer_unschedule Since kvm_timer_[schedule/unschedule] may not be followed by put/load, but put/load also may be called independently, we call the timer save/restore functions from both paths. Since they rely on the loaded flag to never save/restore when unnecessary, this doesn't cause any harm, and we ensure that all invokations of either set of functions work as intended. An added benefit beyond not having to read and write the timer sysregs on every entry and exit is that we no longer have to actively write the active state to the physical distributor, because we configured the irq for the vtimer to only get a priority drop when handling the interrupt in the GIC driver (we called irq_set_vcpu_affinity()), and the interrupt stays active after firing on the host. Reviewed-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Christoffer Dall <cdall@linaro.org>
2016-10-16 20:30:38 +02:00
KVM: arm/arm64: Avoid timer save/restore in vcpu entry/exit We don't need to save and restore the hardware timer state and examine if it generates interrupts on on every entry/exit to the guest. The timer hardware is perfectly capable of telling us when it has expired by signaling interrupts. When taking a vtimer interrupt in the host, we don't want to mess with the timer configuration, we just want to forward the physical interrupt to the guest as a virtual interrupt. We can use the split priority drop and deactivate feature of the GIC to do this, which leaves an EOI'ed interrupt active on the physical distributor, making sure we don't keep taking timer interrupts which would prevent the guest from running. We can then forward the physical interrupt to the VM using the HW bit in the LR of the GIC, like we do already, which lets the guest directly deactivate both the physical and virtual timer simultaneously, allowing the timer hardware to exit the VM and generate a new physical interrupt when the timer output is again asserted later on. We do need to capture this state when migrating VCPUs between physical CPUs, however, which we use the vcpu put/load functions for, which are called through preempt notifiers whenever the thread is scheduled away from the CPU or called directly if we return from the ioctl to userspace. One caveat is that we have to save and restore the timer state in both kvm_timer_vcpu_[put/load] and kvm_timer_[schedule/unschedule], because we can have the following flows: 1. kvm_vcpu_block 2. kvm_timer_schedule 3. schedule 4. kvm_timer_vcpu_put (preempt notifier) 5. schedule (vcpu thread gets scheduled back) 6. kvm_timer_vcpu_load (preempt notifier) 7. kvm_timer_unschedule And a version where we don't actually call schedule: 1. kvm_vcpu_block 2. kvm_timer_schedule 7. kvm_timer_unschedule Since kvm_timer_[schedule/unschedule] may not be followed by put/load, but put/load also may be called independently, we call the timer save/restore functions from both paths. Since they rely on the loaded flag to never save/restore when unnecessary, this doesn't cause any harm, and we ensure that all invokations of either set of functions work as intended. An added benefit beyond not having to read and write the timer sysregs on every entry and exit is that we no longer have to actively write the active state to the physical distributor, because we configured the irq for the vtimer to only get a priority drop when handling the interrupt in the GIC driver (we called irq_set_vcpu_affinity()), and the interrupt stays active after firing on the host. Reviewed-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Christoffer Dall <cdall@linaro.org>
2016-10-16 20:30:38 +02:00