* Fix compile warning on 32-bit machines
* Fix locking error in secure VM support
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJeMiC8AAoJEJ2a6ncsY3GfGg8H/03p+jc/aCKcA75ZeQPlzhmu
KWvSBbPATNcQiYOLfIvbB9AMXUPoyIfiblW/On8G6COFypsIhhUTwEfPUjWIBHNX
IwCfzoyf0gDRTi7A7gTDD06ZE+stikxJu59agX2Gc8kTIQ8ge340VR8J95Ol8/n2
/hVA8S/ORrdv8/KaCcvvIwc1V7OV6xBuGsTUOUvywzBTGDKd0CAbNzRwtS8LmWcM
OCkZX4G5DpFIYdsnjSBaSfwEVPAf3G1DzyQ801emwRnbAGYYgfakd1LwqdLDxptt
6CFHuIENEmmweJKMf9FBLWg+fOMl8wsv9l4mBIYt7coq5XPpi07yJ6yqSaJEToQ=
=Hmfo
-----END PGP SIGNATURE-----
Merge tag 'kvm-ppc-next-5.6-2' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc into HEAD
Second KVM PPC update for 5.6
* Fix compile warning on 32-bit machines
* Fix locking error in secure VM support
Fixes: 3a167beac0 ("kvm: powerpc: Add kvmppc_ops callback")
Signed-off-by: David Michael <fedora.dm0@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
When migrate_vma_setup() fails in kvmppc_svm_page_out(),
release kvm->arch.uvmem_lock before returning.
Fixes: ca9f494267 ("KVM: PPC: Book3S HV: Support for running secure guests")
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Reviewed-by: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Use kvm_vcpu_gfn_to_hva() when retrieving the host page size so that the
correct set of memslots is used when handling x86 page faults in SMM.
Fixes: 54bf36aac5 ("KVM: x86: use vcpu-specific functions to read/write/translate GFNs")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Remove kvm_arch_vcpu_init() and kvm_arch_vcpu_uninit() now that all
arch specific implementations are nops.
Acked-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Fold init() into create() now that the two are called back-to-back by
common KVM code (kvm_vcpu_init() calls kvm_arch_vcpu_init() as its last
action, and kvm_vm_ioctl_create_vcpu() calls kvm_arch_vcpu_create()
immediately thereafter). Rinse and repeat for kvm_arch_vcpu_uninit()
and kvm_arch_vcpu_destroy(). This paves the way for removing
kvm_arch_vcpu_{un}init() entirely.
Note, calling kvmppc_mmu_destroy() if kvmppc_core_vcpu_create() fails
may or may not be necessary. Move it along with the more obvious call
to kvmppc_subarch_vcpu_uninit() so as not to inadvertantly introduce a
functional change and/or bug.
No functional change intended.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Remove kvm_arch_vcpu_setup() now that all arch specific implementations
are nops.
Acked-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Fold setup() into create() now that the two are called back-to-back by
common KVM code. This paves the way for removing kvm_arch_vcpu_setup().
Note, BookE directly implements kvm_arch_vcpu_setup() and PPC's common
kvm_arch_vcpu_create() is responsible for its own cleanup, thus the only
cleanup required when directly invoking kvmppc_core_vcpu_setup() is to
call .vcpu_free(), which is the BookE specific portion of PPC's
kvm_arch_vcpu_destroy() by way of kvmppc_core_vcpu_free().
No functional change intended.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Now that all architectures tightly couple vcpu allocation/free with the
mandatory calls to kvm_{un}init_vcpu(), move the sequences verbatim to
common KVM code.
Move both allocation and initialization in a single patch to eliminate
thrash in arch specific code. The bisection benefits of moving the two
pieces in separate patches is marginal at best, whereas the odds of
introducing a transient arch specific bug are non-zero.
Acked-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add kvm_vcpu_destroy() and wire up all architectures to call the common
function instead of their arch specific implementation. The common
destruction function will be used by future patches to move allocation
and initialization of vCPUs to common KVM code, i.e. to free resources
that are allocated by arch agnostic code.
No functional change intended.
Acked-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add a pre-allocation arch hook to handle checks that are currently done
by arch specific code prior to allocating the vCPU object. This paves
the way for moving the allocation to common KVM code.
Acked-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Remove the superfluous kvm_arch_vcpu_free() as it is no longer called
from commmon KVM code. Note, kvm_arch_vcpu_destroy() *is* called from
common code, i.e. choosing which function to whack is not completely
arbitrary.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Move the kvm_cpu_{un}init() calls to common PPC code as an intermediate
step towards removing kvm_cpu_{un}init() altogether.
No functional change intended.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Move the initialization of oldpir so that the call to kvm_vcpu_init() is
at the top of kvmppc_core_vcpu_create_e500mc(). oldpir is only use
when loading/putting a vCPU, which currently cannot be done until after
kvm_arch_vcpu_create() completes. Reording the call to kvm_vcpu_init()
paves the way for moving the invocation to common PPC code.
No functional change intended.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Call kvm_vcpu_init() in kvmppc_core_vcpu_create_pr() prior to allocating
the book3s and shadow_vcpu objects in preparation of moving said call to
common PPC code. Although kvm_vcpu_init() has an arch callback, the
callback is empty for Book3S PR, i.e. barring unseen black magic, moving
the allocation has no real functional impact.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Move allocation of all flavors of PPC vCPUs to common PPC code. All
variants either allocate 'struct kvm_vcpu' directly, or require that
the embedded 'struct kvm_vcpu' member be located at offset 0, i.e.
guarantee that the allocation can be directly interpreted as a 'struct
kvm_vcpu' object.
Remove the message from the build-time assertion regarding placement of
the struct, as compatibility with the arch usercopy region is no longer
the sole dependent on 'struct kvm_vcpu' being at offset zero.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
In preparation for moving vcpu allocation to common PPC code, add an
explicit, albeit redundant, build-time assert to ensure the vcpu member
is located at offset 0. The assert is redundant in the sense that
kvmppc_core_vcpu_create_e500() contains a functionally identical assert.
The motiviation for adding the extra assert is to provide visual
confirmation of the correctness of moving vcpu allocation to common
code.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Explicitly free the shared page if kvmppc_mmu_init() fails during
kvmppc_core_vcpu_create(), as the page is freed only in
kvmppc_core_vcpu_free(), which is not reached via kvm_vcpu_uninit().
Fixes: 96bc451a15 ("KVM: PPC: Introduce shared page")
Cc: stable@vger.kernel.org
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Call kvm_vcpu_uninit() if vcore creation fails to avoid leaking any
resources allocated by kvm_vcpu_init(), i.e. the vcpu->run page.
Fixes: 371fefd6f2 ("KVM: PPC: Allow book3s_hv guests to use SMT processor modes")
Cc: stable@vger.kernel.org
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Implement the H_SVM_INIT_ABORT hcall which the Ultravisor can use to
abort an SVM after it has issued the H_SVM_INIT_START and before the
H_SVM_INIT_DONE hcalls. This hcall could be used when Ultravisor
encounters security violations or other errors when starting an SVM.
Note that this hcall is different from UV_SVM_TERMINATE ucall which
is used by HV to terminate/cleanup an VM that has becore secure.
The H_SVM_INIT_ABORT basically undoes operations that were done
since the H_SVM_INIT_START hcall - i.e page-out all the VM pages back
to normal memory, and terminate the SVM.
(If we do not bring the pages back to normal memory, the text/data
of the VM would be stuck in secure memory and since the SVM did not
go secure, its MSR_S bit will be clear and the VM wont be able to
access its pages even to do a clean exit).
Based on patches and discussion with Paul Mackerras, Ram Pai and
Bharata Rao.
Signed-off-by: Ram Pai <linuxram@linux.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.ibm.com>
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Add 'skip_page_out' parameter to kvmppc_uvmem_drop_pages() so the
callers can specify whetheter or not to skip paging out pages. This
will be needed in a follow-on patch that implements H_SVM_INIT_ABORT
hcall.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Given that in kvm_create_vm() there is:
kvm->mm = current->mm;
And that on every kvm_*_ioctl we have:
if (kvm->mm != current->mm)
return -EIO;
I see no reason to keep using current->mm instead of kvm->mm.
By doing so, we would reduce the use of 'global' variables on code, relying
more in the contents of kvm struct.
Signed-off-by: Leonardo Bras <leonardo@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Given that in kvm_create_vm() there is:
kvm->mm = current->mm;
And that on every kvm_*_ioctl we have:
if (kvm->mm != current->mm)
return -EIO;
I see no reason to keep using current->mm instead of kvm->mm.
By doing so, we would reduce the use of 'global' variables on code, relying
more in the contents of kvm struct.
Signed-off-by: Leonardo Bras <leonardo@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Fixes gcc '-Wunused-but-set-variable' warning:
arch/powerpc/kvm/emulate_loadstore.c: In function kvmppc_emulate_loadstore:
arch/powerpc/kvm/emulate_loadstore.c:87:6: warning: variable ra set but not used [-Wunused-but-set-variable]
arch/powerpc/kvm/emulate_loadstore.c: In function kvmppc_emulate_loadstore:
arch/powerpc/kvm/emulate_loadstore.c:87:10: warning: variable rs set but not used [-Wunused-but-set-variable]
arch/powerpc/kvm/emulate_loadstore.c: In function kvmppc_emulate_loadstore:
arch/powerpc/kvm/emulate_loadstore.c:87:14: warning: variable rt set but not used [-Wunused-but-set-variable]
They are not used since commit 2b33cb585f ("KVM: PPC: Reimplement
LOAD_FP/STORE_FP instruction mmio emulation with analyse_instr() input")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: zhengbin <zhengbin13@huawei.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
* Fix a bug where we try to do an ultracall on a system without an ultravisor.
KVM:
- Fix uninitialised sysreg accessor
- Fix handling of demand-paged device mappings
- Stop spamming the console on IMPDEF sysregs
- Relax mappings of writable memslots
- Assorted cleanups
MIPS:
- Now orphan, James Hogan is stepping down
x86:
- MAINTAINERS change, so long Radim and thanks for all the fish
- supported CPUID fixes for AMD machines without SPEC_CTRL
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
iQEcBAABAgAGBQJd/1+WAAoJEL/70l94x66DFuYH/A8x/P6BuCpppdGoEw+VGy7X
E8141dHTd7b1Wgi0kDNLRREr4QIfArvavGe0z0W8p4fGtcVjXdyhhfPd0UK6dfKG
9P66phY4AGPjde/8q/qSdFup9yshpcFwSVYdRC0L1w86dBRlXwuqk6K5zsRyCU4b
38v5Q3rPdMnWWB0K88/GMvAyQmPkgMOXJvhoecKeDQ+9IZ3ub6DBBNGM/xTJ9Y3z
vUe2BoYkZ3KKn6sfP66PdprBVI1EOrrAoj/l4BSuo/yUPcQsxTihXMkh5iGl18TF
h7TN9eq2Bn2ryh0TsaSK8opuePcotVvx7oll3ERtSV4e+89z5FDt4vVcY1VyRuc=
=adm7
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull KVM fixes from Paolo Bonzini:
"PPC:
- Fix a bug where we try to do an ultracall on a system without an
ultravisor
KVM:
- Fix uninitialised sysreg accessor
- Fix handling of demand-paged device mappings
- Stop spamming the console on IMPDEF sysregs
- Relax mappings of writable memslots
- Assorted cleanups
MIPS:
- Now orphan, James Hogan is stepping down
x86:
- MAINTAINERS change, so long Radim and thanks for all the fish
- supported CPUID fixes for AMD machines without SPEC_CTRL"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
MAINTAINERS: remove Radim from KVM maintainers
MAINTAINERS: Orphan KVM for MIPS
kvm: x86: Host feature SSBD doesn't imply guest feature AMD_SSBD
kvm: x86: Host feature SSBD doesn't imply guest feature SPEC_CTRL_SSBD
KVM: PPC: Book3S HV: Don't do ultravisor calls on systems without ultravisor
KVM: arm/arm64: Properly handle faulting of device mappings
KVM: arm64: Ensure 'params' is initialised when looking up sys register
KVM: arm/arm64: Remove excessive permission check in kvm_arch_prepare_memory_region
KVM: arm64: Don't log IMP DEF sysreg traps
KVM: arm64: Sanely ratelimit sysreg messages
KVM: arm/arm64: vgic: Use wrapper function to lock/unlock all vcpus in kvm_vgic_create()
KVM: arm/arm64: vgic: Fix potential double free dist->spis in __kvm_vgic_destroy()
KVM: arm/arm64: Get rid of unused arg in cpu_init_hyp_mode()
Commit 22945688ac ("KVM: PPC: Book3S HV: Support reset of secure
guest") added a call to uv_svm_terminate, which is an ultravisor
call, without any check that the guest is a secure guest or even that
the system has an ultravisor. On a system without an ultravisor,
the ultracall will degenerate to a hypercall, but since we are not
in KVM guest context, the hypercall will get treated as a system
call, which could have random effects depending on what happens to
be in r0, and could also corrupt the current task's kernel stack.
Hence this adds a test for the guest being a secure guest before
doing uv_svm_terminate().
Fixes: 22945688ac ("KVM: PPC: Book3S HV: Support reset of secure guest")
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
VCPU_CR is the offset of arch.regs.ccr in kvm_vcpu.
arch/powerpc/include/asm/kvm_host.h defines arch.regs as a struct
pt_regs, and arch/powerpc/include/asm/ptrace.h defines the ccr field
of pt_regs as "unsigned long ccr". Since unsigned long is 64 bits, a
64-bit load needs to be used to load it, unless an endianness specific
correction offset is added to access the desired subpart. In this
case there is no reason to _not_ use a 64 bit load though.
Fixes: 6c85b7bc63 ("powerpc/kvm: Use UV_RETURN ucall to return to ultravisor")
Cc: stable@vger.kernel.org # v5.4+
Signed-off-by: Marcus Comstedt <marcus@mc.pp.se>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20191215094900.46740-1-marcus@mc.pp.se
* small x86 cleanup
* fix for an x86-specific out-of-bounds write on a ioctl (not guest triggerable,
data not attacker-controlled)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (GNU/Linux)
iQEcBAABAgAGBQJd551cAAoJEL/70l94x66D+JkH/R3eEOyvckPmYmzd0lnV8mQ/
7e0n2G/aD+iLZkcCbUnMaImdmSJmoEEJCPjgPk/5nJ3zUi5b/ABWyidEM5uf19Hl
rzKBg0DR7BiQptPnZv2JMwEVKu3JOTchMykqu9xXChQlICocZ0xjdOA6nQ19p0Lv
FulDw5MUaWrXevIzCBskQ38zJejRQA6CpD1lQkHn7LKS9p3p+BsAOd/Ouy87RfWG
b3ktECNbXyO6KStrrhgm+z8pviWY+kqYklyBlDOOwxWif0x8WvNDpQLoVo+ZuhLU
Me8YJ1BN75vFlxzh6ZK5exBUnm9E3fGVKIaaF+dpuds2x+j4HnYl+lZCm89MdqY=
=Q4v7
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull more KVM updates from Paolo Bonzini:
- PPC secure guest support
- small x86 cleanup
- fix for an x86-specific out-of-bounds write on a ioctl (not guest
triggerable, data not attacker-controlled)
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
kvm: vmx: Stop wasting a page for guest_msrs
KVM: x86: fix out-of-bounds write in KVM_GET_EMULATED_CPUID (CVE-2019-19332)
Documentation: kvm: Fix mention to number of ioctls classes
powerpc: Ultravisor: Add PPC_UV config option
KVM: PPC: Book3S HV: Support reset of secure guest
KVM: PPC: Book3S HV: Handle memory plug/unplug to secure VM
KVM: PPC: Book3S HV: Radix changes for secure guest
KVM: PPC: Book3S HV: Shared pages support for secure guests
KVM: PPC: Book3S HV: Support for running secure guests
mm: ksm: Export ksm_madvise()
KVM x86: Move kvm cpuid support out of svm
Add support for reset of secure guest via a new ioctl KVM_PPC_SVM_OFF.
This ioctl will be issued by QEMU during reset and includes the
the following steps:
- Release all device pages of the secure guest.
- Ask UV to terminate the guest via UV_SVM_TERMINATE ucall
- Unpin the VPA pages so that they can be migrated back to secure
side when guest becomes secure again. This is required because
pinned pages can't be migrated.
- Reinit the partition scoped page tables
After these steps, guest is ready to issue UV_ESM call once again
to switch to secure mode.
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
[Implementation of uv_svm_terminate() and its call from
guest shutdown path]
Signed-off-by: Ram Pai <linuxram@us.ibm.com>
[Unpinning of VPA pages]
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Register the new memslot with UV during plug and unregister
the memslot during unplug. In addition, release all the
device pages during unplug.
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
- After the guest becomes secure, when we handle a page fault of a page
belonging to SVM in HV, send that page to UV via UV_PAGE_IN.
- Whenever a page is unmapped on the HV side, inform UV via UV_PAGE_INVAL.
- Ensure all those routines that walk the secondary page tables of
the guest don't do so in case of secure VM. For secure guest, the
active secondary page tables are in secure memory and the secondary
page tables in HV are freed when guest becomes secure.
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
A secure guest will share some of its pages with hypervisor (Eg. virtio
bounce buffers etc). Support sharing of pages between hypervisor and
ultravisor.
Shared page is reachable via both HV and UV side page tables. Once a
secure page is converted to shared page, the device page that represents
the secure page is unmapped from the HV side page tables.
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
A pseries guest can be run as secure guest on Ultravisor-enabled
POWER platforms. On such platforms, this driver will be used to manage
the movement of guest pages between the normal memory managed by
hypervisor (HV) and secure memory managed by Ultravisor (UV).
HV is informed about the guest's transition to secure mode via hcalls:
H_SVM_INIT_START: Initiate securing a VM
H_SVM_INIT_DONE: Conclude securing a VM
As part of H_SVM_INIT_START, register all existing memslots with
the UV. H_SVM_INIT_DONE call by UV informs HV that transition of
the guest to secure mode is complete.
These two states (transition to secure mode STARTED and transition
to secure mode COMPLETED) are recorded in kvm->arch.secure_guest.
Setting these states will cause the assembly code that enters the
guest to call the UV_RETURN ucall instead of trying to enter the
guest directly.
Migration of pages betwen normal and secure memory of secure
guest is implemented in H_SVM_PAGE_IN and H_SVM_PAGE_OUT hcalls.
H_SVM_PAGE_IN: Move the content of a normal page to secure page
H_SVM_PAGE_OUT: Move the content of a secure page to normal page
Private ZONE_DEVICE memory equal to the amount of secure memory
available in the platform for running secure guests is created.
Whenever a page belonging to the guest becomes secure, a page from
this private device memory is used to represent and track that secure
page on the HV side. The movement of pages between normal and secure
memory is done via migrate_vma_pages() using UV_PAGE_IN and
UV_PAGE_OUT ucalls.
In order to prevent the device private pages (that correspond to pages
of secure guest) from participating in KSM merging, H_SVM_PAGE_IN
calls ksm_madvise() under read version of mmap_sem. However
ksm_madvise() needs to be under write lock. Hence we call
kvmppc_svm_page_in with mmap_sem held for writing, and it then
downgrades to a read lock after calling ksm_madvise.
[paulus@ozlabs.org - roll in patch "KVM: PPC: Book3S HV: Take write
mmap_sem when calling ksm_madvise"]
Signed-off-by: Bharata B Rao <bharata@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
We failed to activate the mitigation for Spectre-RSB (Return Stack
Buffer, aka. ret2spec) on context switch, on CPUs prior to Power9
DD2.3.
That allows a process to poison the RSB (called Link Stack on Power
CPUs) and possibly misdirect speculative execution of another process.
If the victim process can be induced to execute a leak gadget then it
may be possible to extract information from the victim via a side
channel.
The fix is to correctly activate the link stack flush mitigation on
all CPUs that have any mitigation of Spectre v2 in userspace enabled.
There's a second commit which adds a link stack flush in the KVM guest
exit path. A leak via that path has not been demonstrated, but we
believe it's at least theoretically possible.
This is the fix for CVE-2019-18660.
-----BEGIN PGP SIGNATURE-----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=NQFt
-----END PGP SIGNATURE-----
Merge tag 'powerpc-spectre-rsb' of powerpc-CVE-2019-18660.bundle
Pull powerpc Spectre-RSB fixes from Michael Ellerman:
"We failed to activate the mitigation for Spectre-RSB (Return Stack
Buffer, aka. ret2spec) on context switch, on CPUs prior to Power9
DD2.3.
That allows a process to poison the RSB (called Link Stack on Power
CPUs) and possibly misdirect speculative execution of another process.
If the victim process can be induced to execute a leak gadget then it
may be possible to extract information from the victim via a side
channel.
The fix is to correctly activate the link stack flush mitigation on
all CPUs that have any mitigation of Spectre v2 in userspace enabled.
There's a second commit which adds a link stack flush in the KVM guest
exit path. A leak via that path has not been demonstrated, but we
believe it's at least theoretically possible.
This is the fix for CVE-2019-18660"
* tag 'powerpc-spectre-rsb' of /home/torvalds/Downloads/powerpc-CVE-2019-18660.bundle:
KVM: PPC: Book3S HV: Flush link stack on guest exit to host kernel
powerpc/book3s64: Fix link stack flush on context switch
- Two fixes from Greg Kurz to fix memory leak bugs in the XIVE code.
-----BEGIN PGP SIGNATURE-----
iQEzBAABCgAdFiEEv0VLfXa2m9eKuaRpnZrqdyxjcZ8FAl3bJKwACgkQnZrqdyxj
cZ92xQgAhgnARWJwh+uazayNrwB12TJA7G25RO8CUEwWaAY/io5QeO7nQCmNZ3cf
TflQpI1dL5qFpzU7uNunHqdqyhlaD0wwkHfrN71molr5sA1uRlIyxwwkE6coZQEC
n/LiGayoxqt2Ra06H4L4SGSjb7fcCl8eYjC3xjTx9Zdf/iXVcwYprBch5kcrToLV
s0NvRvDgwcaqsxQyybTO0wRvME/qz9JFtNUgl6H4PNSt3l/yv+rM+BgjyNR3tyKu
B1G4937GqBIAV4jYmK0a/LDnNfxs9EmOjuJLKCHmVxlfbsg8wasNk3kj+mdrh2O3
ZjCdh782GyGwp/ysddOHmIhXFyQMhQ==
=9kV2
-----END PGP SIGNATURE-----
Merge tag 'kvm-ppc-next-5.5-2' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc into HEAD
Second KVM PPC update for 5.5
- Two fixes from Greg Kurz to fix memory leak bugs in the XIVE code.
We need to check the host page size is big enough to accomodate the
EQ. Let's do this before taking a reference on the EQ page to avoid
a potential leak if the check fails.
Cc: stable@vger.kernel.org # v5.2
Fixes: 13ce3297c5 ("KVM: PPC: Book3S HV: XIVE: Add controls for the EQ configuration")
Signed-off-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The EQ page is allocated by the guest and then passed to the hypervisor
with the H_INT_SET_QUEUE_CONFIG hcall. A reference is taken on the page
before handing it over to the HW. This reference is dropped either when
the guest issues the H_INT_RESET hcall or when the KVM device is released.
But, the guest can legitimately call H_INT_SET_QUEUE_CONFIG several times,
either to reset the EQ (vCPU hot unplug) or to set a new EQ (guest reboot).
In both cases the existing EQ page reference is leaked because we simply
overwrite it in the XIVE queue structure without calling put_page().
This is especially visible when the guest memory is backed with huge pages:
start a VM up to the guest userspace, either reboot it or unplug a vCPU,
quit QEMU. The leak is observed by comparing the value of HugePages_Free in
/proc/meminfo before and after the VM is run.
Ideally we'd want the XIVE code to handle the EQ page de-allocation at the
platform level. This isn't the case right now because the various XIVE
drivers have different allocation needs. It could maybe worth introducing
hooks for this purpose instead of exposing XIVE internals to the drivers,
but this is certainly a huge work to be done later.
In the meantime, for easier backport, fix both vCPU unplug and guest reboot
leaks by introducing a wrapper around xive_native_configure_queue() that
does the necessary cleanup.
Reported-by: Satheesh Rajendran <sathnaga@linux.vnet.ibm.com>
Cc: stable@vger.kernel.org # v5.2
Fixes: 13ce3297c5 ("KVM: PPC: Book3S HV: XIVE: Add controls for the EQ configuration")
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Greg Kurz <groug@kaod.org>
Tested-by: Lijun Pan <ljp@linux.ibm.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
On some systems that are vulnerable to Spectre v2, it is up to
software to flush the link stack (return address stack), in order to
protect against Spectre-RSB.
When exiting from a guest we do some house keeping and then
potentially exit to C code which is several stack frames deep in the
host kernel. We will then execute a series of returns without
preceeding calls, opening up the possiblity that the guest could have
poisoned the link stack, and direct speculative execution of the host
to a gadget of some sort.
To prevent this we add a flush of the link stack on exit from a guest.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
* Add capability to tell userspace whether we can single-step the guest.
* Improve the allocation of XIVE virtual processor IDs, to reduce the
risk of running out of IDs when running many VMs on POWER9.
* Rewrite interrupt synthesis code to deliver interrupts in virtual
mode when appropriate.
* Minor cleanups and improvements.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJdur0ZAAoJEJ2a6ncsY3Gf/xoH/j4wIOKcSjXFxPBAPvvR01Ld
Yt3n+ly/388uMuB4egsM/H+50CK8mpsMA02mQ40nwD4XoTFbOwhKS5wbgd4rQCoX
KtYr1Ylz+D4egw5W0c8Bu7Qdipt8TvKtSFGqDbARWg9oNiN0ZNd0zbuuzA9VpFkL
e58iwUHj1umWqPzHloqtHTyP1jakd9MMLoY5k+BpRKWSwj9ljUNi6JTGv/j8h2f/
JgKEXQ5Ug7Q3eqkMA+jx5fR5OL39rgDwhczd8WxSPz75UD5D3ijuEcmfXsJcMNHL
APggspJI6CHkjYNFAsGoPX4/MQwo0EOJMlWIgGxIoKAiHZbzCxJkYFb8Ibg59GU=
=LodM
-----END PGP SIGNATURE-----
Merge tag 'kvm-ppc-next-5.5-1' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc into HEAD
KVM PPC update for 5.5
* Add capability to tell userspace whether we can single-step the guest.
* Improve the allocation of XIVE virtual processor IDs, to reduce the
risk of running out of IDs when running many VMs on POWER9.
* Rewrite interrupt synthesis code to deliver interrupts in virtual
mode when appropriate.
* Minor cleanups and improvements.
Add a new helper, kvm_put_kvm_no_destroy(), to handle putting a borrowed
reference[*] to the VM when installing a new file descriptor fails. KVM
expects the refcount to remain valid in this case, as the in-progress
ioctl() has an explicit reference to the VM. The primary motiviation
for the helper is to document that the 'kvm' pointer is still valid
after putting the borrowed reference, e.g. to document that doing
mutex(&kvm->lock) immediately after putting a ref to kvm isn't broken.
[*] When exposing a new object to userspace via a file descriptor, e.g.
a new vcpu, KVM grabs a reference to itself (the VM) prior to making
the object visible to userspace to avoid prematurely freeing the VM
in the scenario where userspace immediately closes file descriptor.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
AIL=2 mode has no known users, so is not well tested or supported.
Disallow guests from selecting this mode because it may become
deprecated in future versions of the architecture.
This policy decision is not left to QEMU because KVM support is
required for AIL=2 (when injecting interrupts).
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
kvmppc_inject_interrupt does not implement LPCR[AIL]!=0 modes, which
can result in the guest receiving interrupts as if LPCR[AIL]=0
contrary to the ISA.
In practice, Linux guests cope with this deviation, but it should be
fixed.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
This consolidates the HV interrupt delivery logic into one place.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
reset_msr sets the MSR for interrupt injection, but it's cleaner and
more flexible to provide a single op to set both MSR and PC for the
interrupt.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Add a new attribute to both XIVE and XICS-on-XIVE KVM devices so that
userspace can tell how many interrupt servers it needs. If a VM needs
less than the current default of KVM_MAX_VCPUS (2048), we can allocate
less VPs in OPAL. Combined with a core stride (VSMT) that matches the
number of guest threads per core, this may substantially increases the
number of VMs that can run concurrently with an in-kernel XIVE device.
Since the legacy XIVE KVM device is exposed to userspace through the
XICS KVM API, a new attribute group is added to it for this purpose.
While here, fix the syntax of the existing KVM_DEV_XICS_GRP_SOURCES
in the XICS documentation.
Signed-off-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
The XIVE VP is an internal structure which allow the XIVE interrupt
controller to maintain the interrupt context state of vCPUs non
dispatched on HW threads.
When a guest is started, the XIVE KVM device allocates a block of
XIVE VPs in OPAL, enough to accommodate the highest possible vCPU
id KVM_MAX_VCPU_ID (16384) packed down to KVM_MAX_VCPUS (2048).
With a guest's core stride of 8 and a threading mode of 1 (QEMU's
default), a VM must run at least 256 vCPUs to actually need such a
range of VPs.
A POWER9 system has a limited XIVE VP space : 512k and KVM is
currently wasting this HW resource with large VP allocations,
especially since a typical VM likely runs with a lot less vCPUs.
Make the size of the VP block configurable. Add an nr_servers
field to the XIVE structure and a function to set it for this
purpose.
Split VP allocation out of the device create function. Since the
VP block isn't used before the first vCPU connects to the XIVE KVM
device, allocation is now performed by kvmppc_xive_connect_vcpu().
This gives the opportunity to set nr_servers in between:
kvmppc_xive_create() / kvmppc_xive_native_create()
.
.
kvmppc_xive_set_nr_servers()
.
.
kvmppc_xive_connect_vcpu() / kvmppc_xive_native_connect_vcpu()
The connect_vcpu() functions check that the vCPU id is below nr_servers
and if it is the first vCPU they allocate the VP block. This is protected
against a concurrent update of nr_servers by kvmppc_xive_set_nr_servers()
with the xive->lock mutex.
Also, the block is allocated once for the device lifetime: nr_servers
should stay constant otherwise connect_vcpu() could generate a boggus
VP id and likely crash OPAL. It is thus forbidden to update nr_servers
once the block is allocated.
If the VP allocation fail, return ENOSPC which seems more appropriate to
report the depletion of system wide HW resource than ENOMEM or ENXIO.
A VM using a stride of 8 and 1 thread per core with 32 vCPUs would hence
only need 256 VPs instead of 2048. If the stride is set to match the number
of threads per core, this goes further down to 32.
This will be exposed to userspace by a subsequent patch.
Signed-off-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Reduce code duplication by consolidating the checking of vCPU ids and VP
ids to a common helper used by both legacy and native XIVE KVM devices.
And explain the magic with a comment.
Signed-off-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Print out the VP id of each connected vCPU, this allow to see:
- the VP block base in which OPAL encodes information that may be
useful when debugging
- the packed vCPU id which may differ from the raw vCPU id if the
latter is >= KVM_MAX_VCPUS (2048)
Signed-off-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
If we cannot allocate the XIVE VPs in OPAL, the creation of a XIVE or
XICS-on-XIVE device is aborted as expected, but we leave kvm->arch.xive
set forever since the release method isn't called in this case. Any
subsequent tentative to create a XIVE or XICS-on-XIVE for this VM will
thus always fail (DoS). This is a problem for QEMU since it destroys
and re-creates these devices when the VM is reset: the VM would be
restricted to using the much slower emulated XIVE or XICS forever.
As an alternative to adding rollback, do not assign kvm->arch.xive before
making sure the XIVE VPs are allocated in OPAL.
Cc: stable@vger.kernel.org # v5.2
Fixes: 5422e95103 ("KVM: PPC: Book3S HV: XIVE: Replace the 'destroy' method by a 'release' method")
Signed-off-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
