Hide Forgot
Description of problem: vIOMMU support (VT-d emulation) is available in QEMU 2.6, but does not support the necessary notifier calling points to support an external IOMMU cache, such as necessary for vfio-pci. We don't want users enabling configurations which we know to be broken and currently unsupported. Prevent them. Version-Release number of selected component (if applicable): How reproducible: 100% Steps to Reproduce: 1. Enable vfio-pci device with -M q35,iommu=on 2. 3. Actual results: VM starts, but vfio-pci device does not work Expected results: vfio-pci device should either work or the configuration should generate an error. Here we're trying to enable the latter. Additional info: Patches posted upstream: https://lists.nongnu.org/archive/html/qemu-devel/2016-06/msg04253.html
Fix included in qemu-kvm-rhev-2.6.0-12.el7
Reproduced: Versions: qemu-kvm-rhev-2.6.0-11.el7.x86_64 Steps: 1. Boot guest with iommu=on and vfio-pci device # lspci | grep Eth 01:00.0 Ethernet controller: Intel Corporation Ethernet Controller 10-Gigabit X540-AT2 (rev 01) ... # /usr/libexec/qemu-kvm \ -M q35,iommu=on \ -device vfio-pci,host=01:00.0 (qemu) qemu-kvm: -device vfio-pci,host=01:00.0,bus=root.1,id=pf: iommu map to non memory area 0 qemu-kvm: -device vfio-pci,host=01:00.0,bus=root.1,id=pf: iommu map to non memory area 1000 qemu-kvm: -device vfio-pci,host=01:00.0,bus=root.1,id=pf: iommu map to non memory area 2000 ... qemu-kvm: -device vfio-pci,host=01:00.0: iommu map to non memory area f09c000 qemu-kvm: -device vfio-pci,host=01:00.0: iommu map to non memory area f09d000 ... Hi Alex, In your description, the guest should work. But it doesn't, qemu will keep printing "qemu-kvm: -device vfio-pci,host=01:00.0,bus=root.1,id=pf: iommu map to non memory area xxx",it looks like in a indefinite loop. So is this bug reproduced? Thanks. Verified: Versions: qemu-img-rhev-2.6.0-13.el7.x86_64 Steps: 1. Boot guest with iommu=on and vfio-pci device # /usr/libexec/qemu-kvm \ -M q35,iommu=on \ -device vfio-pci,host=01:00.0 (qemu) qemu: hardware error: Device at bus root.1 addr 00.0 requires iommu notifier which is currently not supported by intel-iommu emulation CPU #0: EAX=00000000 EBX=00000000 ECX=00000000 EDX=000206a1 ESI=00000000 EDI=00000000 EBP=00000000 ESP=00000000 EIP=0000fff0 EFL=00000002 [-------] CPL=0 II=0 A20=1 SMM=0 HLT=0 ES =0000 00000000 0000ffff 00009300 CS =f000 ffff0000 0000ffff 00009b00 SS =0000 00000000 0000ffff 00009300 DS =0000 00000000 0000ffff 00009300 FS =0000 00000000 0000ffff 00009300 GS =0000 00000000 0000ffff 00009300 LDT=0000 00000000 0000ffff 00008200 TR =0000 00000000 0000ffff 00008b00 GDT= 00000000 0000ffff IDT= 00000000 0000ffff CR0=60000010 CR2=00000000 CR3=00000000 CR4=00000000 DR0=0000000000000000 DR1=0000000000000000 DR2=0000000000000000 DR3=0000000000000000 DR6=00000000ffff0ff0 DR7=0000000000000400 EFER=0000000000000000 FCW=037f FSW=0000 [ST=0] FTW=00 MXCSR=00001f80 FPR0=0000000000000000 0000 FPR1=0000000000000000 0000 FPR2=0000000000000000 0000 FPR3=0000000000000000 0000 FPR4=0000000000000000 0000 FPR5=0000000000000000 0000 FPR6=0000000000000000 0000 FPR7=0000000000000000 0000 XMM00=00000000000000000000000000000000 XMM01=00000000000000000000000000000000 XMM02=00000000000000000000000000000000 XMM03=00000000000000000000000000000000 XMM04=00000000000000000000000000000000 XMM05=00000000000000000000000000000000 XMM06=00000000000000000000000000000000 XMM07=00000000000000000000000000000000 CPU #1: EAX=00000000 EBX=00000000 ECX=00000000 EDX=000206a1 ESI=00000000 EDI=00000000 EBP=00000000 ESP=00000000 EIP=0000fff0 EFL=00000002 [-------] CPL=0 II=0 A20=1 SMM=0 HLT=0 ES =0000 00000000 0000ffff 00009300 CS =f000 ffff0000 0000ffff 00009b00 SS =0000 00000000 0000ffff 00009300 DS =0000 00000000 0000ffff 00009300 FS =0000 00000000 0000ffff 00009300 GS =0000 00000000 0000ffff 00009300 LDT=0000 00000000 0000ffff 00008200 TR =0000 00000000 0000ffff 00008b00 GDT= 00000000 0000ffff IDT= 00000000 0000ffff CR0=60000010 CR2=00000000 CR3=00000000 CR4=00000000 DR0=0000000000000000 DR1=0000000000000000 DR2=0000000000000000 DR3=0000000000000000 DR6=00000000ffff0ff0 DR7=0000000000000400 EFER=0000000000000000 FCW=037f FSW=0000 [ST=0] FTW=00 MXCSR=00001f80 FPR0=0000000000000000 0000 FPR1=0000000000000000 0000 FPR2=0000000000000000 0000 FPR3=0000000000000000 0000 FPR4=0000000000000000 0000 FPR5=0000000000000000 0000 FPR6=0000000000000000 0000 FPR7=0000000000000000 0000 XMM00=00000000000000000000000000000000 XMM01=00000000000000000000000000000000 XMM02=00000000000000000000000000000000 XMM03=00000000000000000000000000000000 XMM04=00000000000000000000000000000000 XMM05=00000000000000000000000000000000 XMM06=00000000000000000000000000000000 XMM07=00000000000000000000000000000000 Aborted (core dumped) Qemu quit with prompting 'hardware error' which seems to be expected. But qemu core dump. So I file a new bug[1] to track this new issue. [1]Bug 1356445 - qemu core dump when booting with 'iommu=on' and vfio-pci
This is the expected behavior.
According to Comment 4 and Comment 5, set this bug to VERIFIED.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHBA-2016-2673.html