Michael S. Tsirkin 1a577b7247 KVM: fix race with level interrupts ... When more than 1 source id is in use for the same GSI, we have the following race related to handling irq_states race: CPU 0 clears bit 0. CPU 0 read irq_state as 0. CPU 1 sets level to 1. CPU 1 calls kvm_ioapic_set_irq(1). CPU 0 calls kvm_ioapic_set_irq(0). Now ioapic thinks the level is 0 but irq_state is not 0. Fix by performing all irq_states bitmap handling under pic/ioapic lock. This also removes the need for atomics with irq_states handling. Reported-by: Gleb Natapov <gleb@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>		2012-07-20 16:12:00 -03:00
..
assigned-dev.c	KVM: Convert intx_mask_lock to spin lock	2012-03-20 12:41:24 +02:00
async_pf.c	KVM: fix the race while wakeup all pv guest	2011-01-12 11:29:03 +02:00
async_pf.h	KVM: Halt vcpu if page it tries to access is swapped out	2011-01-12 11:21:39 +02:00
coalesced_mmio.c	KVM: make checks stricter in coalesced_mmio_in_range()	2011-12-27 11:17:07 +02:00
coalesced_mmio.h	KVM: Make coalesced mmio use a device per zone	2011-09-25 19:17:57 +03:00
eventfd.c	KVM: Intelligent device lookup on I/O bus	2011-09-25 19:17:59 +03:00
ioapic.c	KVM: fix race with level interrupts	2012-07-20 16:12:00 -03:00
ioapic.h	KVM: fix race with level interrupts	2012-07-20 16:12:00 -03:00
iodev.h	KVM: remove in_range from io devices	2009-09-10 08:33:05 +03:00
iommu.c	KVM: lock slots_lock around device assignment	2012-04-19 00:04:18 -03:00
irq_comm.c	KVM: fix race with level interrupts	2012-07-20 16:12:00 -03:00
Kconfig	KVM: Introduce direct MSI message injection for in-kernel irqchips	2012-04-24 15:59:47 +03:00
kvm_main.c	KVM: handle last_boosted_vcpu = 0 case	2012-07-06 14:11:18 -03:00