KVM module handling different per Architecture - ppc64el

[Christian Ehrhardt ]

Hi,
the Intel KVM module is build as a modules - the core being KVM and then the chip vendor specific kvm-intel / kvm-svm.
On ppc64el the situation is similar there is kvm.ko and then kvm-pr and kvm-hv for problem state and hardware virtualization.

Currently using kvm users can get:
modprobe: FATAL: Module kvm-hv not found ...

But when evaluating if I should add a new recommends to qemu-kvm I realized this module to kernel package mismatch.

This is caused by one difference, on intel all is in one package:

dpkg -S /lib/modules/4.9.0-11-generic/kernel/arch/x86/kvm/kvm-amd.ko /lib/modules/4.9.0-11-generic/kernel/arch/x86/kvm/kvm-intel.ko /lib/modules/4.9.0-11-generic/kernel/arch/x86/kvm/kvm.ko
linux-image-4.9.0-11-generic: /lib/modules/4.9.0-11-generic/kernel/arch/x86/kvm/kvm-amd.ko
linux-image-4.9.0-11-generic: /lib/modules/4.9.0-11-generic/kernel/arch/x86/kvm/kvm-intel.ko
linux-image-4.9.0-11-generic: /lib/modules/4.9.0-11-generic/kernel/arch/x86/kvm/kvm.ko

But on ppc64el it is not:

$ dpkg -S /lib/modules/4.9.0-11-generic/kernel/arch/powerpc/kvm/kvm.ko /lib/modules/4.9.0-11-generic/kernel/arch/powerpc/kvm/kvm-hv.ko /lib/modules/4.9.0-11-generic/kernel/arch/powerpc/kvm/kvm-pr.ko
linux-image-4.9.0-11-generic: /lib/modules/4.9.0-11-generic/kernel/arch/powerpc/kvm/kvm.ko
linux-image-extra-4.9.0-11-generic: /lib/modules/4.9.0-11-generic/kernel/arch/powerpc/kvm/kvm-hv.ko
linux-image-extra-4.9.0-11-generic: /lib/modules/4.9.0-11-generic/kernel/arch/powerpc/kvm/kvm-pr.ko

Those should be in just one of these packages.
At a discussion on #stable-kernel it was suggested that it might just have been missed so far:

<infinity> smb: I doubt it was a conscious move, but rather that kvm.ko is in the module include list (as inherited from x86 builds), while the ppc-specific modules are not.

I don't know, but I wanted to open the bug to get it resolved - and if possible all in the base image.

[Brad Figg (brad-figg)]

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1657734

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

[Christian Ehrhardt  (paelzer)]

Hi Brad(bot) o/,
this won't need an apport-collect so setting confirmed now.

[Tim Gardner (timg-tpi)]

UBUNTU: [Config] powerpc: Add kvm-hv and kvm-pr to the generic inclusion list

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 4.9.0-15.16

---------------
linux (4.9.0-15.16) zesty; urgency=low

  [ Tim Gardner ]

  * Release Tracking Bug
    - LP: #1658101

  * Zesty update to v4.9.5 stable release (LP: #1658099)
    - Input: xpad - use correct product id for x360w controllers
    - Input: i8042 - add Pegatron touchpad to noloop table
    - pinctrl: imx: fix imx_pinctrl_desc initialization
    - pinctrl: sh-pfc: r8a7795: Use lookup function for bias data
    - pinctrl: sh-pfc: Add helper to handle bias lookup table
    - regulator: tps65086: Fix 25mV ranges for BUCK regulators
    - regulator: axp20x: Fix axp809 ldo_io registration error on cold boot
    - drm/tegra: dpaux: Fix error handling
    - drm/vc4: Fix a couple error codes in vc4_cl_lookup_bos()
    - drm/savage: dereferencing an error pointer
    - selftests: do not require bash to run netsocktests testcase
    - selftests: do not require bash for the generated test
    - zram: revalidate disk under init_lock
    - zram: support BDI_CAP_STABLE_WRITES
    - dax: fix deadlock with DAX 4k holes
    - mm: pmd dirty emulation in page fault handler
    - mm: fix devm_memremap_pages crash, use mem_hotplug_{begin, done}
    - ocfs2: fix crash caused by stale lvb with fsdlm plugin
    - mm, memcg: fix the active list aging for lowmem requests when memcg is enabled
    - mm: support anonymous stable page
    - mm/slab.c: fix SLAB freelist randomization duplicate entries
    - mm/hugetlb.c: fix reservation race when freeing surplus pages
    - KVM: x86: fix emulation of "MOV SS, null selector"
    - KVM: eventfd: fix NULL deref irqbypass consumer
    - jump_labels: API for flushing deferred jump label updates
    - KVM: x86: flush pending lapic jump label updates on module unload
    - KVM: x86: fix NULL deref in vcpu_scan_ioapic
    - KVM: x86: add Align16 instruction flag
    - KVM: x86: add asm_safe wrapper
    - KVM: x86: emulate FXSAVE and FXRSTOR
    - KVM: x86: Introduce segmented_write_std
    - efi/libstub/arm*: Pass latest memory map to the kernel
    - efi/x86: Prune invalid memory map entries and fix boot regression
    - x86/efi: Don't allocate memmap through memblock after mm_init()
    - nl80211: fix sched scan netlink socket owner destruction
    - gpio: Move freeing of GPIO hogs before numbing of the device
    - xfs: Timely free truncated dirty pages
    - bridge: netfilter: Fix dropping packets that moving through bridge interface
    - x86/cpu/AMD: Clean up cpu_llc_id assignment per topology feature
    - x86/bugs: Separate AMD E400 erratum and C1E bug
    - x86/CPU/AMD: Fix Bulldozer topology
    - wusbcore: Fix one more crypto-on-the-stack bug
    - usb: musb: fix runtime PM in debugfs
    - USB: serial: kl5kusb105: fix line-state error handling
    - USB: serial: ch341: fix initial modem-control state
    - USB: serial: ch341: fix resume after reset
    - USB: serial: ch341: fix open error handling
    - USB: serial: ch341: fix control-message error handling
    - USB: serial: ch341: fix open and resume after B0
    - Input: elants_i2c - avoid divide by 0 errors on bad touchscreen data
    - i2c: print correct device invalid address
    - i2c: fix kern...