Hi, this is a rework of the new VGIC series. Many thanks to all reviewers, especially Christoffer. Major changes include a dependency on a new, separate arch_timer rework series [1], a total rework of the MMIO dispatching, which now uses much less kvm_io_bus devices, the proper handling of enabling and disabling the distributor and the much more elaborate handling of activating an IRQ via the MMIO interface. Also the patches have been rearranged a bit in the world switch area and for the GICv3 MMIO handling to be more logical and avoid stub functions that get populated only later. Minor changes are: - removal of vgic_cpu->nr_lr in favour of kvm_vgic_global_state - refinements in the virtual IRQ queueing function - introduction of a separate vgic_clear_lr() function - the removal of the vgic_v[23]_irq_change_affinity functions - caring about SGI injection on SGIPENDR writes - removing locks for atomic reads in the MMIO handlers - removal of ITS stub functions - introducing vgic_irq_is_sgi() - many comment fixes The simple arch_timer rework in the RFC series has been replaced with a separate rework series [1]. Also there is new preparatory patch from Christoffer which gets rid of reading the number of LRs from the vgic_cpu structure. The MMIO data write back fix has also been replaced with a more sophisticated version. I refrained from merging vgic_v[23]_populate_lr() and vgic_v[23]_process_maintenance(), respectively. Having taken a closer look it does not appear worthwhile to me to unify the implementations. Topics which I couldn't manage to address due to timing restraints are: - proper PHYS_CPUID masking when populating GICv2 LRs - proper handling of priorites on injection, also filtering(?) - proper endianess handling in the MMIO framework (not in the VGIC) - handling corner cases when clearing the active state via MMIO - cleanup of register region description with new MMIO framework This series is now based on v4.6-rc3 plus the arch_timer rework[1]. A git tree containing this series and the prerequisites can be found on linux-arm.org: git://linux-arm.org/linux-ap.git branch: vgic-new/v1 http://www.linux-arm.org/git?p=linux-ap.git;a=log;h=refs/heads/vgic-new/v1 Cheers, Andre. =========================== This series is a joint effort to re-implement KVM's GIC emulation. While the current implementation is centered around providing efficient MMIO emulation, the hot path for most guests is actually the guest entry and exit, which currently is rather costly. Also the existing emulation has a global distributor lock, which quickly becomes a bottleneck once the number of VCPUs increases. Additionally the emulation was originally designed for GICv2, adding GICv3 ITS emulation support to this proved to be rather painful. Last, but not least the existing code became less and less maintainable, with many special cases handled explicitly. The new implementation is build around a struct vgic_irq data data structure, which holds all information about a virtual interrupt. Interruts which should be injected are hold in a per-VCPU list, this make the entry/exit path much more efficient. Also the new structure allows to have more fine grained locking - per IRQ and per VCPU - getting rid of the global distributor lock. As a result of the new design ITS emulation fits in more nicely, the respective code will be provided as a follow-up series. This series implements the same feature set as the existing emulation, as a goodie we now implement priorities correctly. To allow an easy transition with good test coverage, but still maintain stability, both implementations live side by side, selectable via a Kconfig option. The default is the new implementation. If this code proves to be reliable, we will later remove the current implementation with an extra patch set. Please have a look at the series, review it and give the code some serious testing (and possibly debugging). All feedback is appreciated. Cheers, Andre. [1] http://lists.infradead.org/pipermail/linux-arm-kernel/2016-April/422580.html Andre Przywara (24): KVM: arm/arm64: pmu: abstract access to number of SPIs KVM: arm/arm64: vgic-new: Add MMIO handling framework KVM: arm/arm64: vgic-new: Export register access interface KVM: arm/arm64: vgic-new: Add CTLR, TYPER and IIDR handlers KVM: arm/arm64: vgic-new: Add ENABLE registers handlers KVM: arm/arm64: vgic-new: Add PENDING registers handlers KVM: arm/arm64: vgic-new: Add PRIORITY registers handlers KVM: arm/arm64: vgic-new: Add ACTIVE registers handlers KVM: arm/arm64: vgic-new: Add CONFIG registers handlers KVM: arm/arm64: vgic-new: Add TARGET registers handlers KVM: arm/arm64: vgic-new: Add SGIR register handler KVM: arm/arm64: vgic-new: Add SGIPENDR register handlers KVM: arm/arm64: vgic-new: Add GICv3 MMIO handling framework KVM: arm/arm64: vgic-new: Add GICv3 CTLR, IIDR, TYPER handlers KVM: arm/arm64: vgic-new: Add GICv3 redistributor IIDR, TYPER handlers KVM: arm/arm64: vgic-new: Add GICv3 IDREGS register handler KVM: arm/arm64: vgic-new: Add GICv3 IROUTER register handlers KVM: arm/arm64: vgic-new: Add GICv3 SGI system register trap handler KVM: arm/arm64: vgic-new: Add userland access to VGIC dist registers KVM: arm/arm64: vgic-new: Add GICH_VMCR accessors KVM: arm/arm64: vgic-new: Add userland GIC CPU interface access KVM: arm/arm64: vgic-new: Wire up irqfd injection KVM: arm/arm64: vgic-new: implement mapped IRQ handling KVM: arm/arm64: vgic-new: enable build Christoffer Dall (6): KVM: arm/arm64: Get rid of vgic_cpu->nr_lr KVM: arm/arm64: Fix MMIO emulation data handling KVM: arm/arm64: vgic-new: Add data structure definitions KVM: arm/arm64: vgic-new: Add acccessor to new struct vgic_irq instance KVM: arm/arm64: vgic-new: Implement virtual IRQ injection KVM: arm/arm64: vgic-new: Add IRQ sorting Eric Auger (12): KVM: arm/arm64: vgic-new: Implement kvm_vgic_vcpu_pending_irq KVM: arm/arm64: vgic-new: vgic_kvm_device: KVM device ops registration KVM: arm/arm64: vgic-new: vgic_kvm_device: KVM_DEV_ARM_VGIC_GRP_NR_IRQS KVM: arm/arm64: vgic-new: vgic_kvm_device: KVM_DEV_ARM_VGIC_GRP_CTRL KVM: arm/arm64: vgic-new: vgic_kvm_device: KVM_DEV_ARM_VGIC_GRP_ADDR KVM: arm/arm64: vgic-new: vgic_kvm_device: access to VGIC registers KVM: arm/arm64: vgic-new: vgic_kvm_device: implement kvm_vgic_addr KVM: arm/arm64: vgic-new: vgic_init: implement kvm_vgic_hyp_init KVM: arm/arm64: vgic-new: vgic_init: implement vgic_create KVM: arm/arm64: vgic-new: vgic_init: implement vgic_init KVM: arm/arm64: vgic-new: vgic_init: implement map_resources KVM: arm/arm64: vgic-new: Add vgic_v2/v3_enable Marc Zyngier (3): KVM: arm/arm64: vgic-new: Add IRQ sync/flush framework KVM: arm/arm64: vgic-new: Add GICv2 world switch backend KVM: arm/arm64: vgic-new: Add GICv3 world switch backend arch/arm/kvm/Kconfig | 7 + arch/arm/kvm/Makefile | 10 + arch/arm/kvm/mmio.c | 14 +- arch/arm64/kvm/Kconfig | 7 + arch/arm64/kvm/Makefile | 10 + include/kvm/arm_vgic.h | 10 +- include/kvm/vgic/vgic.h | 250 +++++++ virt/kvm/arm/hyp/vgic-v2-sr.c | 17 +- virt/kvm/arm/pmu.c | 2 +- virt/kvm/arm/vgic-v2.c | 4 +- virt/kvm/arm/vgic.c | 19 +- virt/kvm/arm/vgic/vgic-v2.c | 355 +++++++++ virt/kvm/arm/vgic/vgic-v3.c | 333 +++++++++ virt/kvm/arm/vgic/vgic.c | 617 +++++++++++++++ virt/kvm/arm/vgic/vgic.h | 138 ++++ virt/kvm/arm/vgic/vgic_init.c | 446 +++++++++++ virt/kvm/arm/vgic/vgic_irqfd.c | 51 ++ virt/kvm/arm/vgic/vgic_kvm_device.c | 522 +++++++++++++ virt/kvm/arm/vgic/vgic_mmio.c | 1411 +++++++++++++++++++++++++++++++++++ virt/kvm/arm/vgic/vgic_mmio.h | 53 ++ 20 files changed, 4242 insertions(+), 34 deletions(-) create mode 100644 include/kvm/vgic/vgic.h create mode 100644 virt/kvm/arm/vgic/vgic-v2.c create mode 100644 virt/kvm/arm/vgic/vgic-v3.c create mode 100644 virt/kvm/arm/vgic/vgic.c create mode 100644 virt/kvm/arm/vgic/vgic.h create mode 100644 virt/kvm/arm/vgic/vgic_init.c create mode 100644 virt/kvm/arm/vgic/vgic_irqfd.c create mode 100644 virt/kvm/arm/vgic/vgic_kvm_device.c create mode 100644 virt/kvm/arm/vgic/vgic_mmio.c create mode 100644 virt/kvm/arm/vgic/vgic_mmio.h -- 2.7.3 -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html