kernel/virt, branch linux-3.6.y

kvm: Fix nonsense handling of compat ioctl

2012-08-26T18:11:48Z

KVM_SET_SIGNAL_MASK passed a NULL argument leaves the on stack signal sets uninitialized. It then passes them through to kvm_vcpu_ioctl_set_sigmask. We should be passing a NULL in this case not translated garbage. Signed-off-by: Alan Cox Signed-off-by: Marcelo Tosatti

Merge tag 'kvm-3.6-1' of git://git.kernel.org/pub/scm/virt/kvm/kvm

2012-07-24T19:01:20Z

Pull KVM updates from Avi Kivity: "Highlights include - full big real mode emulation on pre-Westmere Intel hosts (can be disabled with emulate_invalid_guest_state=0) - relatively small ppc and s390 updates - PCID/INVPCID support in guests - EOI avoidance; 3.6 guests should perform better on 3.6 hosts on interrupt intensive workloads) - Lockless write faults during live migration - EPT accessed/dirty bits support for new Intel processors" Fix up conflicts in: - Documentation/virtual/kvm/api.txt: Stupid subchapter numbering, added next to each other. - arch/powerpc/kvm/booke_interrupts.S: PPC asm changes clashing with the KVM fixes - arch/s390/include/asm/sigp.h, arch/s390/kvm/sigp.c: Duplicated commits through the kvm tree and the s390 tree, with subsequent edits in the KVM tree. * tag 'kvm-3.6-1' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (93 commits) KVM: fix race with level interrupts x86, hyper: fix build with !CONFIG_KVM_GUEST Revert "apic: fix kvm build on UP without IOAPIC" KVM guest: switch to apic_set_eoi_write, apic_write apic: add apic_set_eoi_write for PV use KVM: VMX: Implement PCID/INVPCID for guests with EPT KVM: Add x86_hyper_kvm to complete detect_hypervisor_platform check KVM: PPC: Critical interrupt emulation support KVM: PPC: e500mc: Fix tlbilx emulation for 64-bit guests KVM: PPC64: booke: Set interrupt computation mode for 64-bit host KVM: PPC: bookehv: Add ESR flag to Data Storage Interrupt KVM: PPC: bookehv64: Add support for std/ld emulation. booke: Added crit/mc exception handler for e500v2 booke/bookehv: Add host crit-watchdog exception support KVM: MMU: document mmu-lock and fast page fault KVM: MMU: fix kvm_mmu_pagetable_walk tracepoint KVM: MMU: trace fast page fault KVM: MMU: fast path of handling guest page fault KVM: MMU: introduce SPTE_MMU_WRITEABLE bit KVM: MMU: fold tlb flush judgement into mmu_spte_update ...

KVM: fix race with level interrupts

2012-07-20T19:12:00Z

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 Signed-off-by: Michael S. Tsirkin Signed-off-by: Marcelo Tosatti

KVM: Fix device assignment threaded irq handler

2012-07-11T10:01:48Z

The kernel no longer allows us to pass NULL for the hard handler without also specifying IRQF_ONESHOT. IRQF_ONESHOT imposes latency in the exit path that we don't need for MSI interrupts. Long term we'd like to inject these interrupts from the hard handler when possible. In the short term, we can create dummy hard handlers that return us to the previous behavior. Credit to Michael for original patch. Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=43328 Signed-off-by: Michael S. Tsirkin Signed-off-by: Alex Williamson Signed-off-by: Avi Kivity

KVM: handle last_boosted_vcpu = 0 case

2012-07-06T17:11:18Z

If last_boosted_vcpu == 0, then we fall through all test cases and may end up with all VCPUs pouncing on vcpu 0. With a large enough guest, this can result in enormous runqueue lock contention, which can prevent vcpu0 from running, leading to a livelock. Changing < to <= makes sure we properly handle that case. Signed-off-by: Rik van Riel Signed-off-by: Marcelo Tosatti

KVM: fix fault page leak

2012-07-03T20:31:50Z

fault_page is forgot to be freed Signed-off-by: Xiao Guangrong Signed-off-by: Marcelo Tosatti

KVM: Sanitize KVM_IRQFD flags

2012-07-03T00:10:30Z

We only know of one so far. Signed-off-by: Alex Williamson Signed-off-by: Marcelo Tosatti

KVM: Pass kvm_irqfd to functions

2012-07-03T00:10:30Z

Prune this down to just the struct kvm_irqfd so we can avoid changing function definition for every flag or field we use. Signed-off-by: Alex Williamson Acked-by: Cornelia Huck Signed-off-by: Marcelo Tosatti

KVM: use KVM_CAP_IRQ_ROUTING to protect the routing related code

2012-06-18T13:06:35Z

The KVM code sometimes uses CONFIG_HAVE_KVM_IRQCHIP to protect code that is related to IRQ routing, which not all in-kernel irqchips may support. Use KVM_CAP_IRQ_ROUTING instead. Signed-off-by: Marc Zyngier Signed-off-by: Christoffer Dall Signed-off-by: Avi Kivity

KVM: Fix PCI header check on device assignment

2012-06-16T02:22:12Z

The masking was wrong (must have been 0x7f), and there is no need to re-read the value as pci_setup_device already does this for us. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=43339 Signed-off-by: Jan Kiszka Acked-by: Alex Williamson Signed-off-by: Marcelo Tosatti