Re: [Qemu-devel] [RESEND PATCH v16 6/6] hw/vfio/platform: add irqfd support

[Peter Maydell]

On 15 June 2015 at 17:33, Eric Auger <address@hidden> wrote:
> This patch aims at optimizing IRQ handling using irqfd framework.
>
> Instead of handling the eventfds on user-side they are handled on
> kernel side using
> - the KVM irqfd framework,
> - the VFIO driver virqfd framework.
>
> the virtual IRQ completion is trapped at interrupt controller
> This removes the need for fast/slow path swap.
>
> Overall this brings significant performance improvements.
>
> Signed-off-by: Alvise Rigo <address@hidden>
> Signed-off-by: Eric Auger <address@hidden>
> Reviewed-by: Alex Bennée <address@hidden>
> Tested-by: Vikram Sethi <address@hidden>
>
> ---
> v15 -> v16:
> - add Vikram's T-b
>
> v13 -> v14:
> - use connect_irq_notifier
> - remove trace_vfio_platform_start_eventfd
>
> v12 -> v13:
> - setup the new mechanism for starting irqfd, based on
>   LinkPropertySetter override
> - use kvm_irqchip_[add,remove]_irqfd_notifier new functions: no need
>   to bother about gsi (hence virtualID could be removed with small
>   change in trace-events)
>
> v10 -> v11:
> - Add Alex' Reviewed-by
> - introduce kvm_accel in this patch and initialize it
>
> v5 -> v6
> - rely on kvm_irqfds_enabled() and kvm_resamplefds_enabled()
> - guard KVM code with #ifdef CONFIG_KVM
>
> v3 -> v4:
> [Alvise Rigo]
> Use of VFIO Platform driver v6 unmask/virqfd feature and removal
> of resamplefd handler. Physical IRQ unmasking is now done in
> VFIO driver.
>
> v3:
> [Eric Auger]
> initial support with resamplefd handled on QEMU side since the
> unmask was not supported on VFIO platform driver v5.
> ---
>  hw/vfio/platform.c              | 107 
> ++++++++++++++++++++++++++++++++++++++++
>  include/hw/vfio/vfio-platform.h |   2 +
>  trace-events                    |   1 +
>  3 files changed, 110 insertions(+)
>
> diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c
> index 9382bb7..1d44085 100644
> --- a/hw/vfio/platform.c
> +++ b/hw/vfio/platform.c
> @@ -26,6 +26,7 @@
>  #include "hw/sysbus.h"
>  #include "trace.h"
>  #include "hw/platform-bus.h"
> +#include "sysemu/kvm.h"
>
>  /*
>   * Functions used whatever the injection method
> @@ -51,6 +52,7 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev,
>      intp->pin = info.index;
>      intp->flags = info.flags;
>      intp->state = VFIO_IRQ_INACTIVE;
> +    intp->kvm_accel = false;
>
>      sysbus_init_irq(sbdev, &intp->qemuirq);
>
> @@ -61,6 +63,13 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev,
>          error_report("vfio: Error: trigger event_notifier_init failed ");
>          return NULL;
>      }
> +    /* Get an eventfd for resample/unmask */
> +    ret = event_notifier_init(&intp->unmask, 0);
> +    if (ret) {
> +        g_free(intp);
> +        error_report("vfio: Error: resample event_notifier_init failed eoi");

This error message seems rather cryptic; what is it trying to tell the user?

> +        return NULL;
> +    }
>
>      QLIST_INSERT_HEAD(&vdev->intp_list, intp, next);
>      return intp;
> @@ -315,6 +324,95 @@ static int vfio_start_eventfd_injection(VFIOINTp *intp)
>      return ret;
>  }
>
> +/*
> + * Functions used for irqfd
> + */
> +
> +#ifdef CONFIG_KVM
> +
> +/**
> + * vfio_set_resample_eventfd - sets the resamplefd for an IRQ
> + * @intp: the IRQ struct handle
> + * programs the VFIO driver to unmask this IRQ when the
> + * intp->unmask eventfd is triggered
> + */
> +static int vfio_set_resample_eventfd(VFIOINTp *intp)
> +{
> +    VFIODevice *vbasedev = &intp->vdev->vbasedev;
> +    struct vfio_irq_set *irq_set;
> +    int argsz, ret;
> +    int32_t *pfd;
> +
> +    argsz = sizeof(*irq_set) + sizeof(*pfd);
> +    irq_set = g_malloc0(argsz);
> +    irq_set->argsz = argsz;
> +    irq_set->flags = VFIO_IRQ_SET_DATA_EVENTFD | VFIO_IRQ_SET_ACTION_UNMASK;
> +    irq_set->index = intp->pin;
> +    irq_set->start = 0;
> +    irq_set->count = 1;
> +    pfd = (int32_t *)&irq_set->data;
> +    *pfd = event_notifier_get_fd(&intp->unmask);
> +    qemu_set_fd_handler(*pfd, NULL, NULL, NULL);
> +    ret = ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, irq_set);
> +    g_free(irq_set);
> +    if (ret < 0) {
> +        error_report("vfio: Failed to set resample eventfd: %m");
> +    }
> +    return ret;
> +}
> +
> +static void vfio_start_irqfd_injection(SysBusDevice *sbdev, qemu_irq irq)
> +{
> +    VFIOPlatformDevice *vdev = VFIO_PLATFORM_DEVICE(sbdev);
> +    VFIOINTp *intp;
> +    bool found = false;
> +
> +    QLIST_FOREACH(intp, &vdev->intp_list, next) {
> +        if (intp->qemuirq == irq) {
> +            found  = true;

Stray double-space.

> +            break;
> +        }
> +    }
> +    assert(found);
> +
> +    /* Get to a known interrupt state */
> +    qemu_set_fd_handler(event_notifier_get_fd(&intp->interrupt),
> +                        NULL, NULL, vdev);
> +
> +    vfio_mask_single_irqindex(&vdev->vbasedev, intp->pin);
> +    qemu_set_irq(intp->qemuirq, 0);
> +
> +    if (kvm_irqchip_add_irqfd_notifier(kvm_state, &intp->interrupt,
> +                                   &intp->unmask, irq) < 0) {
> +        goto fail_irqfd;
> +    }
> +
> +    if (vfio_set_trigger_eventfd(intp, NULL) < 0) {
> +        goto fail_vfio;
> +    }
> +    if (vfio_set_resample_eventfd(intp) < 0) {
> +        goto fail_vfio;
> +    }
> +
> +    /* Let'em rip */

This comment could probably be phrased more informatively :-)

> +    vfio_unmask_single_irqindex(&vdev->vbasedev, intp->pin);
> +
> +    intp->kvm_accel = true;
> +
> +    trace_vfio_platform_start_irqfd_injection(intp->pin,
> +                                     event_notifier_get_fd(&intp->interrupt),
> +                                     event_notifier_get_fd(&intp->unmask));
> +    return;
> +fail_vfio:
> +    kvm_irqchip_remove_irqfd_notifier(kvm_state, &intp->interrupt, irq);
> +fail_irqfd:
> +    vfio_start_eventfd_injection(intp);
> +    vfio_unmask_single_irqindex(&vdev->vbasedev, intp->pin);
> +    return;
> +}
> +
> +#endif /* CONFIG_KVM */
> +
>  /* VFIO skeleton */
>
>  static void vfio_platform_compute_needs_reset(VFIODevice *vbasedev)
> @@ -548,6 +646,7 @@ static void vfio_platform_realize(DeviceState *dev, Error 
> **errp)
>  {
>      VFIOPlatformDevice *vdev = VFIO_PLATFORM_DEVICE(dev);
>      SysBusDevice *sbdev = SYS_BUS_DEVICE(dev);
> +    SysBusDeviceClass *sbc = SYS_BUS_DEVICE_GET_CLASS(dev);
>      VFIODevice *vbasedev = &vdev->vbasedev;
>      VFIOINTp *intp;
>      int i, ret;
> @@ -555,6 +654,13 @@ static void vfio_platform_realize(DeviceState *dev, 
> Error **errp)
>      vbasedev->type = VFIO_DEVICE_TYPE_PLATFORM;
>      vbasedev->ops = &vfio_platform_ops;
>
> +#ifdef CONFIG_KVM
> +    if (kvm_irqfds_enabled() && kvm_resamplefds_enabled() &&
> +        vdev->irqfd_allowed) {
> +        sbc->connect_irq_notifier = vfio_start_irqfd_injection;
> +    }
> +#endif

You shouldn't need this CONFIG_KVM ifdef -- we deliberately provide
"always false" versions of functions like kvm_irqfds_enabled() in
sysemu/kvm.h so that we can avoid littering code with ifdefs.

It looks like your commit f41389ae3c54b failed to provide the
non-KVM version of kvm_resamplefds_enabled(), but the correct
fix is to provide it...

This implies removing the ifdefs earlier around vfio_start_irqfd_injection,
but that should be OK -- we provide stub versions of functions like
kvm_irqchip_add_irqfd_notifier() so the non-KVM code can still compile.

> +
>      trace_vfio_platform_realize(vbasedev->name, vdev->compat);
>
>      ret = vfio_base_device_init(vbasedev);
> @@ -584,6 +690,7 @@ static Property vfio_platform_dev_properties[] = {
>      DEFINE_PROP_BOOL("x-mmap", VFIOPlatformDevice, vbasedev.allow_mmap, 
> true),
>      DEFINE_PROP_UINT32("mmap-timeout-ms", VFIOPlatformDevice,
>                         mmap_timeout, 1100),
> +    DEFINE_PROP_BOOL("x-irqfd", VFIOPlatformDevice, irqfd_allowed, true),
>      DEFINE_PROP_END_OF_LIST(),
>  };
>
> diff --git a/include/hw/vfio/vfio-platform.h b/include/hw/vfio/vfio-platform.h
> index 26b2ad6..c5cf1d7 100644
> --- a/include/hw/vfio/vfio-platform.h
> +++ b/include/hw/vfio/vfio-platform.h
> @@ -41,6 +41,7 @@ typedef struct VFIOINTp {
>      int state; /* inactive, pending, active */
>      uint8_t pin; /* index */
>      uint32_t flags; /* IRQ info flags */
> +    bool kvm_accel; /* set when QEMU bypass through KVM enabled */
>  } VFIOINTp;
>
>  /* function type for user side eventfd handler */
> @@ -57,6 +58,7 @@ typedef struct VFIOPlatformDevice {
>      uint32_t mmap_timeout; /* delay to re-enable mmaps after interrupt */
>      QEMUTimer *mmap_timer; /* allows fast-path resume after IRQ hit */
>      QemuMutex intp_mutex; /* protect the intp_list IRQ state */
> +    bool irqfd_allowed; /* debug option to force irqfd on/off */
>  } VFIOPlatformDevice;
>
>  typedef struct VFIOPlatformDeviceClass {
> diff --git a/trace-events b/trace-events
> index 6060d36..4390381 100644
> --- a/trace-events
> +++ b/trace-events
> @@ -1594,6 +1594,7 @@ vfio_platform_intp_interrupt(int pin, int fd) "Inject 
> IRQ #%d (fd = %d)"
>  vfio_platform_intp_inject_pending_lockheld(int pin, int fd) "Inject pending 
> IRQ #%d (fd = %d)"
>  vfio_platform_populate_interrupts(int pin, int count, int flags) "- IRQ 
> index %d: count %d, flags=0x%x"
>  vfio_intp_interrupt_set_pending(int index) "irq %d is set PENDING"
> +vfio_platform_start_irqfd_injection(int index, int fd, int resamplefd) "IRQ 
> index=%d, fd = %d, resamplefd = %d"
>
>  #hw/acpi/memory_hotplug.c
>  mhp_acpi_invalid_slot_selected(uint32_t slot) "0x%"PRIx32
> --
> 1.8.3.2
>

thanks
-- PMM