diff mbox series

[PULL,19/21] ioapic: allow buggy guests mishandling level-triggered interrupts to make progress

Message ID 1557953433-19663-20-git-send-email-pbonzini@redhat.com (mailing list archive)
State New, archived
Headers show
Series [PULL,01/21] hw/input: Add a CONFIG_PS2 switch for the ps2.c file | expand

Commit Message

Paolo Bonzini May 15, 2019, 8:50 p.m. UTC
From: Vitaly Kuznetsov <vkuznets@redhat.com>

It was found that Hyper-V 2016 on KVM in some configurations (q35 machine +
piix4-usb-uhci) hangs on boot. Root-cause was that one of Hyper-V
level-triggered interrupt handler performs EOI before fixing the cause of
the interrupt. This results in IOAPIC keep re-raising the level-triggered
interrupt after EOI because irq-line remains asserted.

Gory details: https://www.spinics.net/lists/kvm/msg184484.html
(the whole thread).

Turns out we were dealing with similar issues before; in-kernel IOAPIC
implementation has commit 184564efae4d ("kvm: ioapic: conditionally delay
irq delivery duringeoi broadcast") which describes a very similar issue.

Steal the idea from the above mentioned commit for IOAPIC implementation in
QEMU. SUCCESSIVE_IRQ_MAX_COUNT, delay and the comment are borrowed as well.

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Message-Id: <20190402080215.10747-1-vkuznets@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 hw/intc/ioapic.c                  | 57 +++++++++++++++++++++++++++++++++++----
 hw/intc/trace-events              |  1 +
 include/hw/i386/ioapic_internal.h |  3 +++
 3 files changed, 56 insertions(+), 5 deletions(-)

Comments

Marc-André Lureau July 4, 2019, 12:57 p.m. UTC | #1
Hi

On Thu, May 16, 2019 at 1:04 AM Paolo Bonzini <pbonzini@redhat.com> wrote:
>
> From: Vitaly Kuznetsov <vkuznets@redhat.com>
>
> It was found that Hyper-V 2016 on KVM in some configurations (q35 machine +
> piix4-usb-uhci) hangs on boot. Root-cause was that one of Hyper-V
> level-triggered interrupt handler performs EOI before fixing the cause of
> the interrupt. This results in IOAPIC keep re-raising the level-triggered
> interrupt after EOI because irq-line remains asserted.
>
> Gory details: https://www.spinics.net/lists/kvm/msg184484.html
> (the whole thread).
>
> Turns out we were dealing with similar issues before; in-kernel IOAPIC
> implementation has commit 184564efae4d ("kvm: ioapic: conditionally delay
> irq delivery duringeoi broadcast") which describes a very similar issue.
>
> Steal the idea from the above mentioned commit for IOAPIC implementation in
> QEMU. SUCCESSIVE_IRQ_MAX_COUNT, delay and the comment are borrowed as well.
>
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> Message-Id: <20190402080215.10747-1-vkuznets@redhat.com>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

After this commit, I get the following ASAN error when booting a VM:

/home/elmarco/src/qq/hw/intc/ioapic.c:266:27: runtime error: index 41
out of bounds for type 'int [24]'
=================================================================
==9761==ERROR: AddressSanitizer: heap-buffer-overflow on address
0x61b00000312c at pc 0x55e40b8f9e74 bp 0x7f269c1d32a0 sp
0x7f269c1d3290
WRITE of size 4 at 0x61b00000312c thread T4 (CPU 0/KVM)
    #0 0x55e40b8f9e73 in ioapic_eoi_broadcast
/home/elmarco/src/qq/hw/intc/ioapic.c:266
    #1 0x55e40bd3fa4e in kvm_arch_handle_exit
/home/elmarco/src/qq/target/i386/kvm.c:3644
    #2 0x55e40b7d1fd7 in kvm_cpu_exec
/home/elmarco/src/qq/accel/kvm/kvm-all.c:2083
    #3 0x55e40b6e435f in qemu_kvm_cpu_thread_fn /home/elmarco/src/qq/cpus.c:1282
    #4 0x55e40ce9ba42 in qemu_thread_start
/home/elmarco/src/qq/util/qemu-thread-posix.c:502
    #5 0x7f26ac3435a1 in start_thread (/lib64/libpthread.so.0+0x85a1)
    #6 0x7f26ac270302 in __clone (/lib64/libc.so.6+0xfb302)

Address 0x61b00000312c is a wild pointer.
SUMMARY: AddressSanitizer: heap-buffer-overflow
/home/elmarco/src/qq/hw/intc/ioapic.c:266 in ioapic_eoi_broadcast
Shadow bytes around the buggy address:
  0x0c367fff85d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
  0x0c367fff85e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
  0x0c367fff85f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
  0x0c367fff8600: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
  0x0c367fff8610: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 fa fa
=>0x0c367fff8620: fa fa fa fa fa[fa]fa fa fa fa fa fa fa fa fa fa
  0x0c367fff8630: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
  0x0c367fff8640: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
  0x0c367fff8650: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
  0x0c367fff8660: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
  0x0c367fff8670: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
Shadow byte legend (one shadow byte represents 8 application bytes):
  Addressable:           00
  Partially addressable: 01 02 03 04 05 06 07
  Heap left redzone:       fa
  Freed heap region:       fd
  Stack left redzone:      f1
  Stack mid redzone:       f2
  Stack right redzone:     f3
  Stack after return:      f5
  Stack use after scope:   f8
  Global redzone:          f9
  Global init order:       f6
  Poisoned by user:        f7
  Container overflow:      fc
  Array cookie:            ac
  Intra object redzone:    bb
  ASan internal:           fe
  Left alloca redzone:     ca
  Right alloca redzone:    cb
  Shadow gap:              cc
Thread T4 (CPU 0/KVM) created by T0 here:
    #0 0x7f26af884965 in pthread_create (/lib64/libasan.so.5+0x3a965)
    #1 0x55e40ce9be8b in qemu_thread_create
/home/elmarco/src/qq/util/qemu-thread-posix.c:539
    #2 0x55e40b6ea347 in qemu_kvm_start_vcpu /home/elmarco/src/qq/cpus.c:2018
    #3 0x55e40b6eb03b in qemu_init_vcpu /home/elmarco/src/qq/cpus.c:2084
    #4 0x55e40bb52352 in x86_cpu_realizefn
/home/elmarco/src/qq/target/i386/cpu.c:5382
    #5 0x55e40c017aed in device_set_realized
/home/elmarco/src/qq/hw/core/qdev.c:834
    #6 0x55e40c95e703 in property_set_bool
/home/elmarco/src/qq/qom/object.c:2074
    #7 0x55e40c959332 in object_property_set
/home/elmarco/src/qq/qom/object.c:1266
    #8 0x55e40c961edb in object_property_set_qobject
/home/elmarco/src/qq/qom/qom-qobject.c:27
    #9 0x55e40c959700 in object_property_set_bool
/home/elmarco/src/qq/qom/object.c:1332
    #10 0x55e40ba831a3 in pc_new_cpu /home/elmarco/src/qq/hw/i386/pc.c:1531
    #11 0x55e40ba838a0 in pc_cpus_init /home/elmarco/src/qq/hw/i386/pc.c:1579
    #12 0x55e40ba9f394 in pc_q35_init /home/elmarco/src/qq/hw/i386/pc_q35.c:183
    #13 0x55e40baa19ab in pc_init_v4_0 /home/elmarco/src/qq/hw/i386/pc_q35.c:385
    #14 0x55e40c035aa7 in machine_run_board_init
/home/elmarco/src/qq/hw/core/machine.c:1033
    #15 0x55e40bda59d8 in main /home/elmarco/src/qq/vl.c:4494
    #16 0x7f26ac198f32 in __libc_start_main (/lib64/libc.so.6+0x23f32)

I start investigating.

> ---
>  hw/intc/ioapic.c                  | 57 +++++++++++++++++++++++++++++++++++----
>  hw/intc/trace-events              |  1 +
>  include/hw/i386/ioapic_internal.h |  3 +++
>  3 files changed, 56 insertions(+), 5 deletions(-)
>
> diff --git a/hw/intc/ioapic.c b/hw/intc/ioapic.c
> index 9d75f84..7074489 100644
> --- a/hw/intc/ioapic.c
> +++ b/hw/intc/ioapic.c
> @@ -139,6 +139,15 @@ static void ioapic_service(IOAPICCommonState *s)
>      }
>  }
>
> +#define SUCCESSIVE_IRQ_MAX_COUNT 10000
> +
> +static void delayed_ioapic_service_cb(void *opaque)
> +{
> +    IOAPICCommonState *s = opaque;
> +
> +    ioapic_service(s);
> +}
> +
>  static void ioapic_set_irq(void *opaque, int vector, int level)
>  {
>      IOAPICCommonState *s = opaque;
> @@ -222,13 +231,39 @@ void ioapic_eoi_broadcast(int vector)
>          }
>          for (n = 0; n < IOAPIC_NUM_PINS; n++) {
>              entry = s->ioredtbl[n];
> -            if ((entry & IOAPIC_LVT_REMOTE_IRR)
> -                && (entry & IOAPIC_VECTOR_MASK) == vector) {
> -                trace_ioapic_clear_remote_irr(n, vector);
> -                s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
> -                if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) {
> +
> +            if ((entry & IOAPIC_VECTOR_MASK) != vector ||
> +                ((entry >> IOAPIC_LVT_TRIGGER_MODE_SHIFT) & 1) != IOAPIC_TRIGGER_LEVEL) {
> +                continue;
> +            }
> +
> +            if (!(entry & IOAPIC_LVT_REMOTE_IRR)) {
> +                continue;
> +            }
> +
> +            trace_ioapic_clear_remote_irr(n, vector);
> +            s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
> +
> +            if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) {
> +                ++s->irq_eoi[vector];
> +                if (s->irq_eoi[vector] >= SUCCESSIVE_IRQ_MAX_COUNT) {
> +                    /*
> +                     * Real hardware does not deliver the interrupt immediately
> +                     * during eoi broadcast, and this lets a buggy guest make
> +                     * slow progress even if it does not correctly handle a
> +                     * level-triggered interrupt. Emulate this behavior if we
> +                     * detect an interrupt storm.
> +                     */
> +                    s->irq_eoi[vector] = 0;
> +                    timer_mod_anticipate(s->delayed_ioapic_service_timer,
> +                                         qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) +
> +                                         NANOSECONDS_PER_SECOND / 100);
> +                    trace_ioapic_eoi_delayed_reassert(vector);
> +                } else {
>                      ioapic_service(s);
>                  }
> +            } else {
> +                s->irq_eoi[vector] = 0;
>              }
>          }
>      }
> @@ -401,6 +436,9 @@ static void ioapic_realize(DeviceState *dev, Error **errp)
>      memory_region_init_io(&s->io_memory, OBJECT(s), &ioapic_io_ops, s,
>                            "ioapic", 0x1000);
>
> +    s->delayed_ioapic_service_timer =
> +        timer_new_ns(QEMU_CLOCK_VIRTUAL, delayed_ioapic_service_cb, s);
> +
>      qdev_init_gpio_in(dev, ioapic_set_irq, IOAPIC_NUM_PINS);
>
>      ioapics[ioapic_no] = s;
> @@ -408,6 +446,14 @@ static void ioapic_realize(DeviceState *dev, Error **errp)
>      qemu_add_machine_init_done_notifier(&s->machine_done);
>  }
>
> +static void ioapic_unrealize(DeviceState *dev, Error **errp)
> +{
> +    IOAPICCommonState *s = IOAPIC_COMMON(dev);
> +
> +    timer_del(s->delayed_ioapic_service_timer);
> +    timer_free(s->delayed_ioapic_service_timer);
> +}
> +
>  static Property ioapic_properties[] = {
>      DEFINE_PROP_UINT8("version", IOAPICCommonState, version, IOAPIC_VER_DEF),
>      DEFINE_PROP_END_OF_LIST(),
> @@ -419,6 +465,7 @@ static void ioapic_class_init(ObjectClass *klass, void *data)
>      DeviceClass *dc = DEVICE_CLASS(klass);
>
>      k->realize = ioapic_realize;
> +    k->unrealize = ioapic_unrealize;
>      /*
>       * If APIC is in kernel, we need to update the kernel cache after
>       * migration, otherwise first 24 gsi routes will be invalid.
> diff --git a/hw/intc/trace-events b/hw/intc/trace-events
> index a28bdce..90c9d07 100644
> --- a/hw/intc/trace-events
> +++ b/hw/intc/trace-events
> @@ -25,6 +25,7 @@ apic_mem_writel(uint64_t addr, uint32_t val) "0x%"PRIx64" = 0x%08x"
>  ioapic_set_remote_irr(int n) "set remote irr for pin %d"
>  ioapic_clear_remote_irr(int n, int vector) "clear remote irr for pin %d vector %d"
>  ioapic_eoi_broadcast(int vector) "EOI broadcast for vector %d"
> +ioapic_eoi_delayed_reassert(int vector) "delayed reassert on EOI broadcast for vector %d"
>  ioapic_mem_read(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem read addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" retval 0x%"PRIx32
>  ioapic_mem_write(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem write addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" val 0x%"PRIx32
>  ioapic_set_irq(int vector, int level) "vector: %d level: %d"
> diff --git a/include/hw/i386/ioapic_internal.h b/include/hw/i386/ioapic_internal.h
> index 9848f39..07002f9 100644
> --- a/include/hw/i386/ioapic_internal.h
> +++ b/include/hw/i386/ioapic_internal.h
> @@ -96,6 +96,7 @@ typedef struct IOAPICCommonClass {
>      SysBusDeviceClass parent_class;
>
>      DeviceRealize realize;
> +    DeviceUnrealize unrealize;
>      void (*pre_save)(IOAPICCommonState *s);
>      void (*post_load)(IOAPICCommonState *s);
>  } IOAPICCommonClass;
> @@ -111,6 +112,8 @@ struct IOAPICCommonState {
>      uint8_t version;
>      uint64_t irq_count[IOAPIC_NUM_PINS];
>      int irq_level[IOAPIC_NUM_PINS];
> +    int irq_eoi[IOAPIC_NUM_PINS];
> +    QEMUTimer *delayed_ioapic_service_timer;
>  };
>
>  void ioapic_reset_common(DeviceState *dev);
> --
> 1.8.3.1
>
>
>
Li Qiang July 4, 2019, 1 p.m. UTC | #2
I have posted a fix for this several weeks ago:

-->https://www.mail-archive.com/qemu-devel@nongnu.org/msg626186.html


Thanks,
Li Qiang

Marc-André Lureau <marcandre.lureau@gmail.com> 于2019年7月4日周四 下午8:57写道:

> Hi
>
> On Thu, May 16, 2019 at 1:04 AM Paolo Bonzini <pbonzini@redhat.com> wrote:
> >
> > From: Vitaly Kuznetsov <vkuznets@redhat.com>
> >
> > It was found that Hyper-V 2016 on KVM in some configurations (q35
> machine +
> > piix4-usb-uhci) hangs on boot. Root-cause was that one of Hyper-V
> > level-triggered interrupt handler performs EOI before fixing the cause of
> > the interrupt. This results in IOAPIC keep re-raising the level-triggered
> > interrupt after EOI because irq-line remains asserted.
> >
> > Gory details: https://www.spinics.net/lists/kvm/msg184484.html
> > (the whole thread).
> >
> > Turns out we were dealing with similar issues before; in-kernel IOAPIC
> > implementation has commit 184564efae4d ("kvm: ioapic: conditionally delay
> > irq delivery duringeoi broadcast") which describes a very similar issue.
> >
> > Steal the idea from the above mentioned commit for IOAPIC implementation
> in
> > QEMU. SUCCESSIVE_IRQ_MAX_COUNT, delay and the comment are borrowed as
> well.
> >
> > Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> > Message-Id: <20190402080215.10747-1-vkuznets@redhat.com>
> > Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>
> After this commit, I get the following ASAN error when booting a VM:
>
> /home/elmarco/src/qq/hw/intc/ioapic.c:266:27: runtime error: index 41
> out of bounds for type 'int [24]'
> =================================================================
> ==9761==ERROR: AddressSanitizer: heap-buffer-overflow on address
> 0x61b00000312c at pc 0x55e40b8f9e74 bp 0x7f269c1d32a0 sp
> 0x7f269c1d3290
> WRITE of size 4 at 0x61b00000312c thread T4 (CPU 0/KVM)
>     #0 0x55e40b8f9e73 in ioapic_eoi_broadcast
> /home/elmarco/src/qq/hw/intc/ioapic.c:266
>     #1 0x55e40bd3fa4e in kvm_arch_handle_exit
> /home/elmarco/src/qq/target/i386/kvm.c:3644
>     #2 0x55e40b7d1fd7 in kvm_cpu_exec
> /home/elmarco/src/qq/accel/kvm/kvm-all.c:2083
>     #3 0x55e40b6e435f in qemu_kvm_cpu_thread_fn
> /home/elmarco/src/qq/cpus.c:1282
>     #4 0x55e40ce9ba42 in qemu_thread_start
> /home/elmarco/src/qq/util/qemu-thread-posix.c:502
>     #5 0x7f26ac3435a1 in start_thread (/lib64/libpthread.so.0+0x85a1)
>     #6 0x7f26ac270302 in __clone (/lib64/libc.so.6+0xfb302)
>
> Address 0x61b00000312c is a wild pointer.
> SUMMARY: AddressSanitizer: heap-buffer-overflow
> /home/elmarco/src/qq/hw/intc/ioapic.c:266 in ioapic_eoi_broadcast
> Shadow bytes around the buggy address:
>   0x0c367fff85d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>   0x0c367fff85e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>   0x0c367fff85f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>   0x0c367fff8600: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>   0x0c367fff8610: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 fa fa
> =>0x0c367fff8620: fa fa fa fa fa[fa]fa fa fa fa fa fa fa fa fa fa
>   0x0c367fff8630: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>   0x0c367fff8640: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>   0x0c367fff8650: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>   0x0c367fff8660: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>   0x0c367fff8670: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
> Shadow byte legend (one shadow byte represents 8 application bytes):
>   Addressable:           00
>   Partially addressable: 01 02 03 04 05 06 07
>   Heap left redzone:       fa
>   Freed heap region:       fd
>   Stack left redzone:      f1
>   Stack mid redzone:       f2
>   Stack right redzone:     f3
>   Stack after return:      f5
>   Stack use after scope:   f8
>   Global redzone:          f9
>   Global init order:       f6
>   Poisoned by user:        f7
>   Container overflow:      fc
>   Array cookie:            ac
>   Intra object redzone:    bb
>   ASan internal:           fe
>   Left alloca redzone:     ca
>   Right alloca redzone:    cb
>   Shadow gap:              cc
> Thread T4 (CPU 0/KVM) created by T0 here:
>     #0 0x7f26af884965 in pthread_create (/lib64/libasan.so.5+0x3a965)
>     #1 0x55e40ce9be8b in qemu_thread_create
> /home/elmarco/src/qq/util/qemu-thread-posix.c:539
>     #2 0x55e40b6ea347 in qemu_kvm_start_vcpu
> /home/elmarco/src/qq/cpus.c:2018
>     #3 0x55e40b6eb03b in qemu_init_vcpu /home/elmarco/src/qq/cpus.c:2084
>     #4 0x55e40bb52352 in x86_cpu_realizefn
> /home/elmarco/src/qq/target/i386/cpu.c:5382
>     #5 0x55e40c017aed in device_set_realized
> /home/elmarco/src/qq/hw/core/qdev.c:834
>     #6 0x55e40c95e703 in property_set_bool
> /home/elmarco/src/qq/qom/object.c:2074
>     #7 0x55e40c959332 in object_property_set
> /home/elmarco/src/qq/qom/object.c:1266
>     #8 0x55e40c961edb in object_property_set_qobject
> /home/elmarco/src/qq/qom/qom-qobject.c:27
>     #9 0x55e40c959700 in object_property_set_bool
> /home/elmarco/src/qq/qom/object.c:1332
>     #10 0x55e40ba831a3 in pc_new_cpu /home/elmarco/src/qq/hw/i386/pc.c:1531
>     #11 0x55e40ba838a0 in pc_cpus_init
> /home/elmarco/src/qq/hw/i386/pc.c:1579
>     #12 0x55e40ba9f394 in pc_q35_init
> /home/elmarco/src/qq/hw/i386/pc_q35.c:183
>     #13 0x55e40baa19ab in pc_init_v4_0
> /home/elmarco/src/qq/hw/i386/pc_q35.c:385
>     #14 0x55e40c035aa7 in machine_run_board_init
> /home/elmarco/src/qq/hw/core/machine.c:1033
>     #15 0x55e40bda59d8 in main /home/elmarco/src/qq/vl.c:4494
>     #16 0x7f26ac198f32 in __libc_start_main (/lib64/libc.so.6+0x23f32)
>
> I start investigating.
>
> > ---
> >  hw/intc/ioapic.c                  | 57
> +++++++++++++++++++++++++++++++++++----
> >  hw/intc/trace-events              |  1 +
> >  include/hw/i386/ioapic_internal.h |  3 +++
> >  3 files changed, 56 insertions(+), 5 deletions(-)
> >
> > diff --git a/hw/intc/ioapic.c b/hw/intc/ioapic.c
> > index 9d75f84..7074489 100644
> > --- a/hw/intc/ioapic.c
> > +++ b/hw/intc/ioapic.c
> > @@ -139,6 +139,15 @@ static void ioapic_service(IOAPICCommonState *s)
> >      }
> >  }
> >
> > +#define SUCCESSIVE_IRQ_MAX_COUNT 10000
> > +
> > +static void delayed_ioapic_service_cb(void *opaque)
> > +{
> > +    IOAPICCommonState *s = opaque;
> > +
> > +    ioapic_service(s);
> > +}
> > +
> >  static void ioapic_set_irq(void *opaque, int vector, int level)
> >  {
> >      IOAPICCommonState *s = opaque;
> > @@ -222,13 +231,39 @@ void ioapic_eoi_broadcast(int vector)
> >          }
> >          for (n = 0; n < IOAPIC_NUM_PINS; n++) {
> >              entry = s->ioredtbl[n];
> > -            if ((entry & IOAPIC_LVT_REMOTE_IRR)
> > -                && (entry & IOAPIC_VECTOR_MASK) == vector) {
> > -                trace_ioapic_clear_remote_irr(n, vector);
> > -                s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
> > -                if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 <<
> n))) {
> > +
> > +            if ((entry & IOAPIC_VECTOR_MASK) != vector ||
> > +                ((entry >> IOAPIC_LVT_TRIGGER_MODE_SHIFT) & 1) !=
> IOAPIC_TRIGGER_LEVEL) {
> > +                continue;
> > +            }
> > +
> > +            if (!(entry & IOAPIC_LVT_REMOTE_IRR)) {
> > +                continue;
> > +            }
> > +
> > +            trace_ioapic_clear_remote_irr(n, vector);
> > +            s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
> > +
> > +            if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) {
> > +                ++s->irq_eoi[vector];
> > +                if (s->irq_eoi[vector] >= SUCCESSIVE_IRQ_MAX_COUNT) {
> > +                    /*
> > +                     * Real hardware does not deliver the interrupt
> immediately
> > +                     * during eoi broadcast, and this lets a buggy
> guest make
> > +                     * slow progress even if it does not correctly
> handle a
> > +                     * level-triggered interrupt. Emulate this behavior
> if we
> > +                     * detect an interrupt storm.
> > +                     */
> > +                    s->irq_eoi[vector] = 0;
> > +
> timer_mod_anticipate(s->delayed_ioapic_service_timer,
> > +
>  qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) +
> > +                                         NANOSECONDS_PER_SECOND / 100);
> > +                    trace_ioapic_eoi_delayed_reassert(vector);
> > +                } else {
> >                      ioapic_service(s);
> >                  }
> > +            } else {
> > +                s->irq_eoi[vector] = 0;
> >              }
> >          }
> >      }
> > @@ -401,6 +436,9 @@ static void ioapic_realize(DeviceState *dev, Error
> **errp)
> >      memory_region_init_io(&s->io_memory, OBJECT(s), &ioapic_io_ops, s,
> >                            "ioapic", 0x1000);
> >
> > +    s->delayed_ioapic_service_timer =
> > +        timer_new_ns(QEMU_CLOCK_VIRTUAL, delayed_ioapic_service_cb, s);
> > +
> >      qdev_init_gpio_in(dev, ioapic_set_irq, IOAPIC_NUM_PINS);
> >
> >      ioapics[ioapic_no] = s;
> > @@ -408,6 +446,14 @@ static void ioapic_realize(DeviceState *dev, Error
> **errp)
> >      qemu_add_machine_init_done_notifier(&s->machine_done);
> >  }
> >
> > +static void ioapic_unrealize(DeviceState *dev, Error **errp)
> > +{
> > +    IOAPICCommonState *s = IOAPIC_COMMON(dev);
> > +
> > +    timer_del(s->delayed_ioapic_service_timer);
> > +    timer_free(s->delayed_ioapic_service_timer);
> > +}
> > +
> >  static Property ioapic_properties[] = {
> >      DEFINE_PROP_UINT8("version", IOAPICCommonState, version,
> IOAPIC_VER_DEF),
> >      DEFINE_PROP_END_OF_LIST(),
> > @@ -419,6 +465,7 @@ static void ioapic_class_init(ObjectClass *klass,
> void *data)
> >      DeviceClass *dc = DEVICE_CLASS(klass);
> >
> >      k->realize = ioapic_realize;
> > +    k->unrealize = ioapic_unrealize;
> >      /*
> >       * If APIC is in kernel, we need to update the kernel cache after
> >       * migration, otherwise first 24 gsi routes will be invalid.
> > diff --git a/hw/intc/trace-events b/hw/intc/trace-events
> > index a28bdce..90c9d07 100644
> > --- a/hw/intc/trace-events
> > +++ b/hw/intc/trace-events
> > @@ -25,6 +25,7 @@ apic_mem_writel(uint64_t addr, uint32_t val)
> "0x%"PRIx64" = 0x%08x"
> >  ioapic_set_remote_irr(int n) "set remote irr for pin %d"
> >  ioapic_clear_remote_irr(int n, int vector) "clear remote irr for pin %d
> vector %d"
> >  ioapic_eoi_broadcast(int vector) "EOI broadcast for vector %d"
> > +ioapic_eoi_delayed_reassert(int vector) "delayed reassert on EOI
> broadcast for vector %d"
> >  ioapic_mem_read(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t
> val) "ioapic mem read addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8"
> retval 0x%"PRIx32
> >  ioapic_mem_write(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t
> val) "ioapic mem write addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8"
> val 0x%"PRIx32
> >  ioapic_set_irq(int vector, int level) "vector: %d level: %d"
> > diff --git a/include/hw/i386/ioapic_internal.h
> b/include/hw/i386/ioapic_internal.h
> > index 9848f39..07002f9 100644
> > --- a/include/hw/i386/ioapic_internal.h
> > +++ b/include/hw/i386/ioapic_internal.h
> > @@ -96,6 +96,7 @@ typedef struct IOAPICCommonClass {
> >      SysBusDeviceClass parent_class;
> >
> >      DeviceRealize realize;
> > +    DeviceUnrealize unrealize;
> >      void (*pre_save)(IOAPICCommonState *s);
> >      void (*post_load)(IOAPICCommonState *s);
> >  } IOAPICCommonClass;
> > @@ -111,6 +112,8 @@ struct IOAPICCommonState {
> >      uint8_t version;
> >      uint64_t irq_count[IOAPIC_NUM_PINS];
> >      int irq_level[IOAPIC_NUM_PINS];
> > +    int irq_eoi[IOAPIC_NUM_PINS];
> > +    QEMUTimer *delayed_ioapic_service_timer;
> >  };
> >
> >  void ioapic_reset_common(DeviceState *dev);
> > --
> > 1.8.3.1
> >
> >
> >
>
>
> --
> Marc-André Lureau
>
>
Marc-André Lureau July 4, 2019, 1:05 p.m. UTC | #3
Hi

On Thu, Jul 4, 2019 at 5:01 PM Li Qiang <liq3ea@gmail.com> wrote:
>
> I have posted a fix for this several weeks ago:
>
> -->https://www.mail-archive.com/qemu-devel@nongnu.org/msg626186.html

Your patch looks reasonable, but I am not really able to review it.

I hope Paolo and Vitaly will take care of it.

Thanks

>
>
> Thanks,
> Li Qiang
>
> Marc-André Lureau <marcandre.lureau@gmail.com> 于2019年7月4日周四 下午8:57写道:
>>
>> Hi
>>
>> On Thu, May 16, 2019 at 1:04 AM Paolo Bonzini <pbonzini@redhat.com> wrote:
>> >
>> > From: Vitaly Kuznetsov <vkuznets@redhat.com>
>> >
>> > It was found that Hyper-V 2016 on KVM in some configurations (q35 machine +
>> > piix4-usb-uhci) hangs on boot. Root-cause was that one of Hyper-V
>> > level-triggered interrupt handler performs EOI before fixing the cause of
>> > the interrupt. This results in IOAPIC keep re-raising the level-triggered
>> > interrupt after EOI because irq-line remains asserted.
>> >
>> > Gory details: https://www.spinics.net/lists/kvm/msg184484.html
>> > (the whole thread).
>> >
>> > Turns out we were dealing with similar issues before; in-kernel IOAPIC
>> > implementation has commit 184564efae4d ("kvm: ioapic: conditionally delay
>> > irq delivery duringeoi broadcast") which describes a very similar issue.
>> >
>> > Steal the idea from the above mentioned commit for IOAPIC implementation in
>> > QEMU. SUCCESSIVE_IRQ_MAX_COUNT, delay and the comment are borrowed as well.
>> >
>> > Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
>> > Message-Id: <20190402080215.10747-1-vkuznets@redhat.com>
>> > Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>>
>> After this commit, I get the following ASAN error when booting a VM:
>>
>> /home/elmarco/src/qq/hw/intc/ioapic.c:266:27: runtime error: index 41
>> out of bounds for type 'int [24]'
>> =================================================================
>> ==9761==ERROR: AddressSanitizer: heap-buffer-overflow on address
>> 0x61b00000312c at pc 0x55e40b8f9e74 bp 0x7f269c1d32a0 sp
>> 0x7f269c1d3290
>> WRITE of size 4 at 0x61b00000312c thread T4 (CPU 0/KVM)
>>     #0 0x55e40b8f9e73 in ioapic_eoi_broadcast
>> /home/elmarco/src/qq/hw/intc/ioapic.c:266
>>     #1 0x55e40bd3fa4e in kvm_arch_handle_exit
>> /home/elmarco/src/qq/target/i386/kvm.c:3644
>>     #2 0x55e40b7d1fd7 in kvm_cpu_exec
>> /home/elmarco/src/qq/accel/kvm/kvm-all.c:2083
>>     #3 0x55e40b6e435f in qemu_kvm_cpu_thread_fn /home/elmarco/src/qq/cpus.c:1282
>>     #4 0x55e40ce9ba42 in qemu_thread_start
>> /home/elmarco/src/qq/util/qemu-thread-posix.c:502
>>     #5 0x7f26ac3435a1 in start_thread (/lib64/libpthread.so.0+0x85a1)
>>     #6 0x7f26ac270302 in __clone (/lib64/libc.so.6+0xfb302)
>>
>> Address 0x61b00000312c is a wild pointer.
>> SUMMARY: AddressSanitizer: heap-buffer-overflow
>> /home/elmarco/src/qq/hw/intc/ioapic.c:266 in ioapic_eoi_broadcast
>> Shadow bytes around the buggy address:
>>   0x0c367fff85d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>>   0x0c367fff85e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>>   0x0c367fff85f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>>   0x0c367fff8600: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>>   0x0c367fff8610: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 fa fa
>> =>0x0c367fff8620: fa fa fa fa fa[fa]fa fa fa fa fa fa fa fa fa fa
>>   0x0c367fff8630: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>>   0x0c367fff8640: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>>   0x0c367fff8650: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>>   0x0c367fff8660: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>>   0x0c367fff8670: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
>> Shadow byte legend (one shadow byte represents 8 application bytes):
>>   Addressable:           00
>>   Partially addressable: 01 02 03 04 05 06 07
>>   Heap left redzone:       fa
>>   Freed heap region:       fd
>>   Stack left redzone:      f1
>>   Stack mid redzone:       f2
>>   Stack right redzone:     f3
>>   Stack after return:      f5
>>   Stack use after scope:   f8
>>   Global redzone:          f9
>>   Global init order:       f6
>>   Poisoned by user:        f7
>>   Container overflow:      fc
>>   Array cookie:            ac
>>   Intra object redzone:    bb
>>   ASan internal:           fe
>>   Left alloca redzone:     ca
>>   Right alloca redzone:    cb
>>   Shadow gap:              cc
>> Thread T4 (CPU 0/KVM) created by T0 here:
>>     #0 0x7f26af884965 in pthread_create (/lib64/libasan.so.5+0x3a965)
>>     #1 0x55e40ce9be8b in qemu_thread_create
>> /home/elmarco/src/qq/util/qemu-thread-posix.c:539
>>     #2 0x55e40b6ea347 in qemu_kvm_start_vcpu /home/elmarco/src/qq/cpus.c:2018
>>     #3 0x55e40b6eb03b in qemu_init_vcpu /home/elmarco/src/qq/cpus.c:2084
>>     #4 0x55e40bb52352 in x86_cpu_realizefn
>> /home/elmarco/src/qq/target/i386/cpu.c:5382
>>     #5 0x55e40c017aed in device_set_realized
>> /home/elmarco/src/qq/hw/core/qdev.c:834
>>     #6 0x55e40c95e703 in property_set_bool
>> /home/elmarco/src/qq/qom/object.c:2074
>>     #7 0x55e40c959332 in object_property_set
>> /home/elmarco/src/qq/qom/object.c:1266
>>     #8 0x55e40c961edb in object_property_set_qobject
>> /home/elmarco/src/qq/qom/qom-qobject.c:27
>>     #9 0x55e40c959700 in object_property_set_bool
>> /home/elmarco/src/qq/qom/object.c:1332
>>     #10 0x55e40ba831a3 in pc_new_cpu /home/elmarco/src/qq/hw/i386/pc.c:1531
>>     #11 0x55e40ba838a0 in pc_cpus_init /home/elmarco/src/qq/hw/i386/pc.c:1579
>>     #12 0x55e40ba9f394 in pc_q35_init /home/elmarco/src/qq/hw/i386/pc_q35.c:183
>>     #13 0x55e40baa19ab in pc_init_v4_0 /home/elmarco/src/qq/hw/i386/pc_q35.c:385
>>     #14 0x55e40c035aa7 in machine_run_board_init
>> /home/elmarco/src/qq/hw/core/machine.c:1033
>>     #15 0x55e40bda59d8 in main /home/elmarco/src/qq/vl.c:4494
>>     #16 0x7f26ac198f32 in __libc_start_main (/lib64/libc.so.6+0x23f32)
>>
>> I start investigating.
>>
>> > ---
>> >  hw/intc/ioapic.c                  | 57 +++++++++++++++++++++++++++++++++++----
>> >  hw/intc/trace-events              |  1 +
>> >  include/hw/i386/ioapic_internal.h |  3 +++
>> >  3 files changed, 56 insertions(+), 5 deletions(-)
>> >
>> > diff --git a/hw/intc/ioapic.c b/hw/intc/ioapic.c
>> > index 9d75f84..7074489 100644
>> > --- a/hw/intc/ioapic.c
>> > +++ b/hw/intc/ioapic.c
>> > @@ -139,6 +139,15 @@ static void ioapic_service(IOAPICCommonState *s)
>> >      }
>> >  }
>> >
>> > +#define SUCCESSIVE_IRQ_MAX_COUNT 10000
>> > +
>> > +static void delayed_ioapic_service_cb(void *opaque)
>> > +{
>> > +    IOAPICCommonState *s = opaque;
>> > +
>> > +    ioapic_service(s);
>> > +}
>> > +
>> >  static void ioapic_set_irq(void *opaque, int vector, int level)
>> >  {
>> >      IOAPICCommonState *s = opaque;
>> > @@ -222,13 +231,39 @@ void ioapic_eoi_broadcast(int vector)
>> >          }
>> >          for (n = 0; n < IOAPIC_NUM_PINS; n++) {
>> >              entry = s->ioredtbl[n];
>> > -            if ((entry & IOAPIC_LVT_REMOTE_IRR)
>> > -                && (entry & IOAPIC_VECTOR_MASK) == vector) {
>> > -                trace_ioapic_clear_remote_irr(n, vector);
>> > -                s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
>> > -                if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) {
>> > +
>> > +            if ((entry & IOAPIC_VECTOR_MASK) != vector ||
>> > +                ((entry >> IOAPIC_LVT_TRIGGER_MODE_SHIFT) & 1) != IOAPIC_TRIGGER_LEVEL) {
>> > +                continue;
>> > +            }
>> > +
>> > +            if (!(entry & IOAPIC_LVT_REMOTE_IRR)) {
>> > +                continue;
>> > +            }
>> > +
>> > +            trace_ioapic_clear_remote_irr(n, vector);
>> > +            s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
>> > +
>> > +            if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) {
>> > +                ++s->irq_eoi[vector];
>> > +                if (s->irq_eoi[vector] >= SUCCESSIVE_IRQ_MAX_COUNT) {
>> > +                    /*
>> > +                     * Real hardware does not deliver the interrupt immediately
>> > +                     * during eoi broadcast, and this lets a buggy guest make
>> > +                     * slow progress even if it does not correctly handle a
>> > +                     * level-triggered interrupt. Emulate this behavior if we
>> > +                     * detect an interrupt storm.
>> > +                     */
>> > +                    s->irq_eoi[vector] = 0;
>> > +                    timer_mod_anticipate(s->delayed_ioapic_service_timer,
>> > +                                         qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) +
>> > +                                         NANOSECONDS_PER_SECOND / 100);
>> > +                    trace_ioapic_eoi_delayed_reassert(vector);
>> > +                } else {
>> >                      ioapic_service(s);
>> >                  }
>> > +            } else {
>> > +                s->irq_eoi[vector] = 0;
>> >              }
>> >          }
>> >      }
>> > @@ -401,6 +436,9 @@ static void ioapic_realize(DeviceState *dev, Error **errp)
>> >      memory_region_init_io(&s->io_memory, OBJECT(s), &ioapic_io_ops, s,
>> >                            "ioapic", 0x1000);
>> >
>> > +    s->delayed_ioapic_service_timer =
>> > +        timer_new_ns(QEMU_CLOCK_VIRTUAL, delayed_ioapic_service_cb, s);
>> > +
>> >      qdev_init_gpio_in(dev, ioapic_set_irq, IOAPIC_NUM_PINS);
>> >
>> >      ioapics[ioapic_no] = s;
>> > @@ -408,6 +446,14 @@ static void ioapic_realize(DeviceState *dev, Error **errp)
>> >      qemu_add_machine_init_done_notifier(&s->machine_done);
>> >  }
>> >
>> > +static void ioapic_unrealize(DeviceState *dev, Error **errp)
>> > +{
>> > +    IOAPICCommonState *s = IOAPIC_COMMON(dev);
>> > +
>> > +    timer_del(s->delayed_ioapic_service_timer);
>> > +    timer_free(s->delayed_ioapic_service_timer);
>> > +}
>> > +
>> >  static Property ioapic_properties[] = {
>> >      DEFINE_PROP_UINT8("version", IOAPICCommonState, version, IOAPIC_VER_DEF),
>> >      DEFINE_PROP_END_OF_LIST(),
>> > @@ -419,6 +465,7 @@ static void ioapic_class_init(ObjectClass *klass, void *data)
>> >      DeviceClass *dc = DEVICE_CLASS(klass);
>> >
>> >      k->realize = ioapic_realize;
>> > +    k->unrealize = ioapic_unrealize;
>> >      /*
>> >       * If APIC is in kernel, we need to update the kernel cache after
>> >       * migration, otherwise first 24 gsi routes will be invalid.
>> > diff --git a/hw/intc/trace-events b/hw/intc/trace-events
>> > index a28bdce..90c9d07 100644
>> > --- a/hw/intc/trace-events
>> > +++ b/hw/intc/trace-events
>> > @@ -25,6 +25,7 @@ apic_mem_writel(uint64_t addr, uint32_t val) "0x%"PRIx64" = 0x%08x"
>> >  ioapic_set_remote_irr(int n) "set remote irr for pin %d"
>> >  ioapic_clear_remote_irr(int n, int vector) "clear remote irr for pin %d vector %d"
>> >  ioapic_eoi_broadcast(int vector) "EOI broadcast for vector %d"
>> > +ioapic_eoi_delayed_reassert(int vector) "delayed reassert on EOI broadcast for vector %d"
>> >  ioapic_mem_read(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem read addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" retval 0x%"PRIx32
>> >  ioapic_mem_write(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem write addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" val 0x%"PRIx32
>> >  ioapic_set_irq(int vector, int level) "vector: %d level: %d"
>> > diff --git a/include/hw/i386/ioapic_internal.h b/include/hw/i386/ioapic_internal.h
>> > index 9848f39..07002f9 100644
>> > --- a/include/hw/i386/ioapic_internal.h
>> > +++ b/include/hw/i386/ioapic_internal.h
>> > @@ -96,6 +96,7 @@ typedef struct IOAPICCommonClass {
>> >      SysBusDeviceClass parent_class;
>> >
>> >      DeviceRealize realize;
>> > +    DeviceUnrealize unrealize;
>> >      void (*pre_save)(IOAPICCommonState *s);
>> >      void (*post_load)(IOAPICCommonState *s);
>> >  } IOAPICCommonClass;
>> > @@ -111,6 +112,8 @@ struct IOAPICCommonState {
>> >      uint8_t version;
>> >      uint64_t irq_count[IOAPIC_NUM_PINS];
>> >      int irq_level[IOAPIC_NUM_PINS];
>> > +    int irq_eoi[IOAPIC_NUM_PINS];
>> > +    QEMUTimer *delayed_ioapic_service_timer;
>> >  };
>> >
>> >  void ioapic_reset_common(DeviceState *dev);
>> > --
>> > 1.8.3.1
>> >
>> >
>> >
>>
>>
>> --
>> Marc-André Lureau
>>
Paolo Bonzini July 4, 2019, 1:13 p.m. UTC | #4
On 04/07/19 15:05, Marc-André Lureau wrote:
> Hi
> 
> On Thu, Jul 4, 2019 at 5:01 PM Li Qiang <liq3ea@gmail.com> wrote:
>>
>> I have posted a fix for this several weeks ago:
>>
>> -->https://www.mail-archive.com/qemu-devel@nongnu.org/msg626186.html
> 
> Your patch looks reasonable, but I am not really able to review it.
> 
> I hope Paolo and Vitaly will take care of it.

I have queued the patch, thanks Marc-André and Li Qiang.

Paolo
diff mbox series

Patch

diff --git a/hw/intc/ioapic.c b/hw/intc/ioapic.c
index 9d75f84..7074489 100644
--- a/hw/intc/ioapic.c
+++ b/hw/intc/ioapic.c
@@ -139,6 +139,15 @@  static void ioapic_service(IOAPICCommonState *s)
     }
 }
 
+#define SUCCESSIVE_IRQ_MAX_COUNT 10000
+
+static void delayed_ioapic_service_cb(void *opaque)
+{
+    IOAPICCommonState *s = opaque;
+
+    ioapic_service(s);
+}
+
 static void ioapic_set_irq(void *opaque, int vector, int level)
 {
     IOAPICCommonState *s = opaque;
@@ -222,13 +231,39 @@  void ioapic_eoi_broadcast(int vector)
         }
         for (n = 0; n < IOAPIC_NUM_PINS; n++) {
             entry = s->ioredtbl[n];
-            if ((entry & IOAPIC_LVT_REMOTE_IRR)
-                && (entry & IOAPIC_VECTOR_MASK) == vector) {
-                trace_ioapic_clear_remote_irr(n, vector);
-                s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
-                if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) {
+
+            if ((entry & IOAPIC_VECTOR_MASK) != vector ||
+                ((entry >> IOAPIC_LVT_TRIGGER_MODE_SHIFT) & 1) != IOAPIC_TRIGGER_LEVEL) {
+                continue;
+            }
+
+            if (!(entry & IOAPIC_LVT_REMOTE_IRR)) {
+                continue;
+            }
+
+            trace_ioapic_clear_remote_irr(n, vector);
+            s->ioredtbl[n] = entry & ~IOAPIC_LVT_REMOTE_IRR;
+
+            if (!(entry & IOAPIC_LVT_MASKED) && (s->irr & (1 << n))) {
+                ++s->irq_eoi[vector];
+                if (s->irq_eoi[vector] >= SUCCESSIVE_IRQ_MAX_COUNT) {
+                    /*
+                     * Real hardware does not deliver the interrupt immediately
+                     * during eoi broadcast, and this lets a buggy guest make
+                     * slow progress even if it does not correctly handle a
+                     * level-triggered interrupt. Emulate this behavior if we
+                     * detect an interrupt storm.
+                     */
+                    s->irq_eoi[vector] = 0;
+                    timer_mod_anticipate(s->delayed_ioapic_service_timer,
+                                         qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL) +
+                                         NANOSECONDS_PER_SECOND / 100);
+                    trace_ioapic_eoi_delayed_reassert(vector);
+                } else {
                     ioapic_service(s);
                 }
+            } else {
+                s->irq_eoi[vector] = 0;
             }
         }
     }
@@ -401,6 +436,9 @@  static void ioapic_realize(DeviceState *dev, Error **errp)
     memory_region_init_io(&s->io_memory, OBJECT(s), &ioapic_io_ops, s,
                           "ioapic", 0x1000);
 
+    s->delayed_ioapic_service_timer =
+        timer_new_ns(QEMU_CLOCK_VIRTUAL, delayed_ioapic_service_cb, s);
+
     qdev_init_gpio_in(dev, ioapic_set_irq, IOAPIC_NUM_PINS);
 
     ioapics[ioapic_no] = s;
@@ -408,6 +446,14 @@  static void ioapic_realize(DeviceState *dev, Error **errp)
     qemu_add_machine_init_done_notifier(&s->machine_done);
 }
 
+static void ioapic_unrealize(DeviceState *dev, Error **errp)
+{
+    IOAPICCommonState *s = IOAPIC_COMMON(dev);
+
+    timer_del(s->delayed_ioapic_service_timer);
+    timer_free(s->delayed_ioapic_service_timer);
+}
+
 static Property ioapic_properties[] = {
     DEFINE_PROP_UINT8("version", IOAPICCommonState, version, IOAPIC_VER_DEF),
     DEFINE_PROP_END_OF_LIST(),
@@ -419,6 +465,7 @@  static void ioapic_class_init(ObjectClass *klass, void *data)
     DeviceClass *dc = DEVICE_CLASS(klass);
 
     k->realize = ioapic_realize;
+    k->unrealize = ioapic_unrealize;
     /*
      * If APIC is in kernel, we need to update the kernel cache after
      * migration, otherwise first 24 gsi routes will be invalid.
diff --git a/hw/intc/trace-events b/hw/intc/trace-events
index a28bdce..90c9d07 100644
--- a/hw/intc/trace-events
+++ b/hw/intc/trace-events
@@ -25,6 +25,7 @@  apic_mem_writel(uint64_t addr, uint32_t val) "0x%"PRIx64" = 0x%08x"
 ioapic_set_remote_irr(int n) "set remote irr for pin %d"
 ioapic_clear_remote_irr(int n, int vector) "clear remote irr for pin %d vector %d"
 ioapic_eoi_broadcast(int vector) "EOI broadcast for vector %d"
+ioapic_eoi_delayed_reassert(int vector) "delayed reassert on EOI broadcast for vector %d"
 ioapic_mem_read(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem read addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" retval 0x%"PRIx32
 ioapic_mem_write(uint8_t addr, uint8_t regsel, uint8_t size, uint32_t val) "ioapic mem write addr 0x%"PRIx8" regsel: 0x%"PRIx8" size 0x%"PRIx8" val 0x%"PRIx32
 ioapic_set_irq(int vector, int level) "vector: %d level: %d"
diff --git a/include/hw/i386/ioapic_internal.h b/include/hw/i386/ioapic_internal.h
index 9848f39..07002f9 100644
--- a/include/hw/i386/ioapic_internal.h
+++ b/include/hw/i386/ioapic_internal.h
@@ -96,6 +96,7 @@  typedef struct IOAPICCommonClass {
     SysBusDeviceClass parent_class;
 
     DeviceRealize realize;
+    DeviceUnrealize unrealize;
     void (*pre_save)(IOAPICCommonState *s);
     void (*post_load)(IOAPICCommonState *s);
 } IOAPICCommonClass;
@@ -111,6 +112,8 @@  struct IOAPICCommonState {
     uint8_t version;
     uint64_t irq_count[IOAPIC_NUM_PINS];
     int irq_level[IOAPIC_NUM_PINS];
+    int irq_eoi[IOAPIC_NUM_PINS];
+    QEMUTimer *delayed_ioapic_service_timer;
 };
 
 void ioapic_reset_common(DeviceState *dev);