Commit graph

86 commits

Author SHA1 Message Date
Linus Torvalds
c98d767b34 Arm:
"This is a bit of an odd merge window on the KVM/arm64 front. There is
 absolutely no new feature in the pull request. It is purely fixes,
 because it is simply becoming too hard to review new stuff when so
 many AI-fuelled fixes hit the list".
 
 - Significant cleanup of the vgic-v5 PPI support which was merged in
   7.1. This makes the code more maintainable, and squashes a couple
   of bugs in the meantime.
 
 - Set of fixes for the handling of the MMU in an NV context,
   particularly VNCR-triggered faults. S1POE support is fixed
   as well.
 
 - Large set of pKVM fixes, mostly addressing recurring issues
   around hypervisor tracking of donated pages in obscure cases
   where the donation could fail and leave things in a bizarre
   state.
 
 - Fixes for the so-called "lazy vgic init", which resulted in
   sleeping operations in non-preemptible sections. This turned
   out to be far more invasive than initially expected...
 
 - Reduce the overhead of L1/L2 context switch by not touching
   the FP registers.
 
 - Fix the way non-implemented page sizes are dealt with when
   a guest insist on using them for S2 translation.
 
 - The usual set of low-impact fixes and cleanups all over the map.
 
 Loongarch:
 
 - On a request for lazy FPU load, load all FPU state that the VM
   supports instead of enabling only the part (FPU, LSX or LASX)
   that caused the FPU load request.
 
 - Some enhancements about interrupt injection.
 
 - Some bug fixes and other small changes.
 
 RISC-V:
 
 - Batch G-stage TLB flushes for GPA range based page table updates
 
 - Convert HGEI line management to fully per-HART
 
 - Fix missing CSR dirty marking when FWFT state updated via ONE_REG
 
 - Fix stale FWFT feature exposure to Guest/VM
 
 - Speed up dirty logging write faults using MMU rwlock and atomic
   PTE updates using cmpxchg() for permission-only changes
 
 - Use flexible array for APLIC IRQ state
 
 - Use kvm_slot_dirty_track_enabled() for logging enable check on
   a memslot
 
 - Avoid skipping valid pages in kvm_riscv_gstage_wp_range()
 
 - Avoid skipping valid pages in kvm_riscv_gstage_unmap_range()
 
 - Use endian-specific __lelong for NACL shared memory
 
 S390:
 
 - KVM_PRE_FAULT_MEMORY support
 
 - Support for 2G hugepages
 
 - Support for the ASTFLEIE 2 facility
 
 - Support for fast inject using kvm_arch_set_irq_inatomic
 
 - Fix potential leak of uninitialized bytes
 
 - A few more misc gmap fixes.
 
 x86:
 
 - Generic support for the more granular permissions allowed by EPT,
   namely "read" (which was previously usurping the U bit) and
   separate execution bits for kernel and userspace.
 
 - Do not assume that all page tables start with U=1/W=1/NX=0 at the
   root, as AMD GMET needs to have U=0 at the root.
 
 - Introduce common assembly macros for use within Intel and AMD
   vendor-specific vmentry code.  This touches the SPEC_CTRL handling,
   which is now entirely done in assembly for Intel (by reusing the
   AMD code that already existed), and register save/restore which
   uses some macro magic to compute the offsets in the struct.  Both
   of these are preparatory changes for upcoming APX support.
 
 - Clean up KVM's register tracking and storage, primarily to prepare for
   APX support, which expands the maximum number of GPRs from 16 to 32.
 
 - Keep a single copy of the PDPTRs rather than two, since architecturally
   there is just one.
 
 - Handle EXIT_FASTPATH_EXIT_USERSPACE in vendor code to ensure vendor code
   gets a chance to handle things like reaping the PML buffer.
 
 - Update KVM's view of PV async enabling if and only if the MSR write fully
   succeeds.
 
 - Fix a variety of issues where the emulator doesn't honor guest-debug state,
   and clean up related code along the way.
 
 - Synthesize EPT Violation and #NPF "error code" bits when injecting faults
   into L1 that didn't originate in hardware (in which case the VMCS/VMCB
   doesn't hold relevant information).
 
 - Add support for virtualizing (well, emulating) AMD's flavor of CPL>0 CPUID
   faulting.
 
 - Clean up the GPR APIs so that KVM's use of "raw" is consistent, and fix a
   variety of minor bugs along the way.
 
 - Fix an OOB memory access due to not checking the VP ID when handling a
   Hyper-V PV TLB flush for L2.
 
 - Fix a bug in the mediated PMU's handling of fixed counters that allowed the
   guest to bypass the PMU event filter.
 
 - Allow userspace to return EAGAIN when handling SNP and TDX hypercalls, so
   the KVM can forward a "retry" status code to the guest, and reserve all
   unused error codes for future usage.
 
 - Overhaul the TDP MMU => S-EPT code to move as much S-EPT specific logic as
   possible into the TDX code, and to funnel (almost) all S-EPT updates into
   a single chokepoint.  The motivation is largely to prepare for upcoming
   Dynamic PAMT support, but the cleanups are nice to have on their own.
 
 - Plug a hole in shadow page table handling, where KVM fails to recursively
   zap nested EPT/NPT shadow page tables when the nested hypervisor tears
   down its own EPT/NPT page tables from the bottom up
 
 x86 (Intel):
 
 - Support for nested MBEC (Mode-Based Execute Control), see above in the
   generic section; also run with MBEC enabled even for non-nested mode.
 
 - Use the kernel's "enum pg_level" in the TDX APIs instead of the TDX-Module's
   level definitions (which are 0-based).
 
 - Rework the TDX memory APIs to not require/assume that guest memory is
   backed by "struct page" (in prepartion for guest_memfd hugepage support).
 
 - Fix a largely benign bug where KVM TDX would incorrectly state it could
   emulate several x2APIC MSRs.
 
 - Use the "safe" WRMSR API when proxying LBR MSR writes as the to-be-written
   value is guest controlled and completely unvalidated.
 
 x86 (AMD):
 
 - Support for nested GMET (Guest Mode Execution Trap), see above in the
   generic section; also run with GMET enabled even for non-nested mode.
 
 - Fixes and minor cleanups to GHCB handling, on top of the earlier work
   already merged into 7.1-rc.
 
 - Ensure KVM's copy of CR0 and CR3 are up-to-date prior to invoking
   fastpath handlers.
 
 - Add support for virtualizing gPAT (KVM previously just used L1's PAT when
   running L2).
 
 - Fix goofs where KVM mishandles side effects (e.g. single-step and PMC
   updates) when emulating VMRUN.
 
 - Fix a variety of bugs in AVIC's handling of x2APIC MSR interception, most
   notably where KVM didn't disable interception of IRR, ISR, and TMR regs.
 
 - Add support for virtualizing Host-Only/Guest-Only bits in the mediated PMU.
 
 - Don't advertise support for unusable VM types, and account for VM types
   that are disabled by firmware, e.g. to mitigate security vulnerabilities.
 
 - Rewrite the SEV {en,de}crypt debug ioctls as they were riddle with bugs and
   unnecessarily complicated, and add comprehensive tests.
 
 - Clean up and deduplicate the SEV page pinning code.
 
 - Fix minor goofs related to writing back CPUID information after firmware
   rejects a CPUID page for an SNP vCPU.
 
 Generic:
 
 - Rename invalidate_begin() to invalidate_start() throughout KVM to follow
   the kernel's nomenclature, e.g. for mmu_notifiers.
 
 - Use guard() to cleanup up various KVM+VFIO flows.
 
 - Minor cleanups.
 
 guest_memfd:
 
 - Return -EEXIST instead of -EINVAL if userspace attempts to bind a gmem
   range to multiple memslots, and fix the test that was supposed to ensure
   KVM returns -EEXIST.
 
 - Treat memslot binding offsets and sizes as unsigned values to fix a bug
   where KVM interprets a large "offset + size" as a negative value and allows
   a nonsensical offset.
 
 - Use the inode number instead of the page offset for the NUMA interleaving
   index to fix a bug where the effective index would jump by two for
   consecutive pages (the caller also adds in the page offset).
 
 Selftests:
 
 - Randomize the dirty log test's delay when reaping the bitmap on the first
   pass, as always waiting only 1ms hid a KVM RISC-V bug as the test reaped the
   bitmap before KVM could build up enough state to hit the bug.
 
 - A pile of one-off fixes and cleanups.
 -----BEGIN PGP SIGNATURE-----
 
 iQFIBAABCgAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAmoypI0UHHBib256aW5p
 QHJlZGhhdC5jb20ACgkQv/vSX3jHroNCSAf8CyuvicojnZGFsNcoY4DZ0whBj5l8
 IwVkHaptuyGLSwPAebC/DwtU2NBY4CQdbtLjGmxenf+fSQOtcz1YWKbdY4Nr0hJG
 euAb/wO47RqKp1C4WPi7lL2P4XHgmUgD5XZ1qDnRV9gnGddZi3QLLmKgAtNq9Z4m
 +pAdGD2/RyZeI0Dx61+5ufE/z8YBlH6GrzgBTxfvvRNKqNDFykBenxYHI7seEfHC
 BPz0LxjHDD8BMDUo86qaylZnaeQSmFTGO/eTgvCvabxG6WbM1s0w3K1owmUmP7o3
 p9o8274Xh53iK2JYD/qGOtAJXejIwbYKT5bnvUQX0O1EH2+ouvSFapLQ9Q==
 =sSqM
 -----END PGP SIGNATURE-----

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull kvm updates from Paolo Bonzini:
 "arm64:

     This is a bit of an odd merge window on the KVM/arm64 front. There
     is absolutely no new feature in the pull request. It is purely
     fixes, because it is simply becoming too hard to review new stuff
     when so many AI-fuelled fixes hit the list.

   - Significant cleanup of the vgic-v5 PPI support which was merged in
     7.1. This makes the code more maintainable, and squashes a couple
     of bugs in the meantime

   - Set of fixes for the handling of the MMU in an NV context,
     particularly VNCR-triggered faults. S1POE support is fixed as well

   - Large set of pKVM fixes, mostly addressing recurring issues around
     hypervisor tracking of donated pages in obscure cases where the
     donation could fail and leave things in a bizarre state

   - Fixes for the so-called "lazy vgic init", which resulted in
     sleeping operations in non-preemptible sections. This turned out to
     be far more invasive than initially expected..

   - Reduce the overhead of L1/L2 context switch by not touching the FP
     registers

   - Fix the way non-implemented page sizes are dealt with when a guest
     insist on using them for S2 translation

   - The usual set of low-impact fixes and cleanups all over the map

  Loongarch:

   - On a request for lazy FPU load, load all FPU state that the VM
     supports instead of enabling only the part (FPU, LSX or LASX) that
     caused the FPU load request

   - Some enhancements about interrupt injection

   - Some bug fixes and other small changes

  RISC-V:

   - Batch G-stage TLB flushes for GPA range based page table updates

   - Convert HGEI line management to fully per-HART

   - Fix missing CSR dirty marking when FWFT state updated via ONE_REG

   - Fix stale FWFT feature exposure to Guest/VM

   - Speed up dirty logging write faults using MMU rwlock and atomic PTE
     updates using cmpxchg() for permission-only changes

   - Use flexible array for APLIC IRQ state

   - Use kvm_slot_dirty_track_enabled() for logging enable check on a
     memslot

   - Avoid skipping valid pages in kvm_riscv_gstage_wp_range()

   - Avoid skipping valid pages in kvm_riscv_gstage_unmap_range()

   - Use endian-specific __lelong for NACL shared memory

  S390:

   - KVM_PRE_FAULT_MEMORY support

   - Support for 2G hugepages

   - Support for the ASTFLEIE 2 facility

   - Support for fast inject using kvm_arch_set_irq_inatomic

   - Fix potential leak of uninitialized bytes

   - A few more misc gmap fixes

  x86:

   - Generic support for the more granular permissions allowed by EPT,
     namely "read" (which was previously usurping the U bit) and
     separate execution bits for kernel and userspace

   - Do not assume that all page tables start with U=1/W=1/NX=0 at the
     root, as AMD GMET needs to have U=0 at the root

   - Introduce common assembly macros for use within Intel and AMD
     vendor-specific vmentry code. This touches the SPEC_CTRL handling,
     which is now entirely done in assembly for Intel (by reusing the
     AMD code that already existed), and register save/restore which
     uses some macro magic to compute the offsets in the struct. Both of
     these are preparatory changes for upcoming APX support

   - Clean up KVM's register tracking and storage, primarily to prepare
     for APX support, which expands the maximum number of GPRs from 16
     to 32

   - Keep a single copy of the PDPTRs rather than two, since
     architecturally there is just one

   - Handle EXIT_FASTPATH_EXIT_USERSPACE in vendor code to ensure vendor
     code gets a chance to handle things like reaping the PML buffer

   - Update KVM's view of PV async enabling if and only if the MSR write
     fully succeeds

   - Fix a variety of issues where the emulator doesn't honor
     guest-debug state, and clean up related code along the way

   - Synthesize EPT Violation and #NPF "error code" bits when injecting
     faults into L1 that didn't originate in hardware (in which case the
     VMCS/VMCB doesn't hold relevant information)

   - Add support for virtualizing (well, emulating) AMD's flavor of
     CPL>0 CPUID faulting

   - Clean up the GPR APIs so that KVM's use of "raw" is consistent, and
     fix a variety of minor bugs along the way

   - Fix an OOB memory access due to not checking the VP ID when
     handling a Hyper-V PV TLB flush for L2

   - Fix a bug in the mediated PMU's handling of fixed counters that
     allowed the guest to bypass the PMU event filter

   - Allow userspace to return EAGAIN when handling SNP and TDX
     hypercalls, so the KVM can forward a "retry" status code to the
     guest, and reserve all unused error codes for future usage

   - Overhaul the TDP MMU => S-EPT code to move as much S-EPT specific
     logic as possible into the TDX code, and to funnel (almost) all
     S-EPT updates into a single chokepoint. The motivation is largely
     to prepare for upcoming Dynamic PAMT support, but the cleanups are
     nice to have on their own

   - Plug a hole in shadow page table handling, where KVM fails to
     recursively zap nested EPT/NPT shadow page tables when the nested
     hypervisor tears down its own EPT/NPT page tables from the bottom
     up

  x86 (Intel):

   - Support for nested MBEC (Mode-Based Execute Control), see above in
     the generic section; also run with MBEC enabled even for non-nested
     mode

   - Use the kernel's "enum pg_level" in the TDX APIs instead of the
     TDX-Module's level definitions (which are 0-based)

   - Rework the TDX memory APIs to not require/assume that guest memory
     is backed by "struct page" (in prepartion for guest_memfd hugepage
     support)

   - Fix a largely benign bug where KVM TDX would incorrectly state it
     could emulate several x2APIC MSRs

   - Use the "safe" WRMSR API when proxying LBR MSR writes as the
     to-be-written value is guest controlled and completely unvalidated

  x86 (AMD):

   - Support for nested GMET (Guest Mode Execution Trap), see above in
     the generic section; also run with GMET enabled even for non-nested
     mode

   - Fixes and minor cleanups to GHCB handling, on top of the earlier
     work already merged into 7.1-rc

   - Ensure KVM's copy of CR0 and CR3 are up-to-date prior to invoking
     fastpath handlers

   - Add support for virtualizing gPAT (KVM previously just used L1's
     PAT when running L2)

   - Fix goofs where KVM mishandles side effects (e.g. single-step and
     PMC updates) when emulating VMRUN

   - Fix a variety of bugs in AVIC's handling of x2APIC MSR
     interception, most notably where KVM didn't disable interception of
     IRR, ISR, and TMR regs

   - Add support for virtualizing Host-Only/Guest-Only bits in the
     mediated PMU

   - Don't advertise support for unusable VM types, and account for VM
     types that are disabled by firmware, e.g. to mitigate security
     vulnerabilities

   - Rewrite the SEV {en,de}crypt debug ioctls as they were riddle with
     bugs and unnecessarily complicated, and add comprehensive tests

   - Clean up and deduplicate the SEV page pinning code

   - Fix minor goofs related to writing back CPUID information after
     firmware rejects a CPUID page for an SNP vCPU

  Generic:

   - Rename invalidate_begin() to invalidate_start() throughout KVM to
     follow the kernel's nomenclature, e.g. for mmu_notifiers

   - Use guard() to cleanup up various KVM+VFIO flows

   - Minor cleanups

  guest_memfd:

   - Return -EEXIST instead of -EINVAL if userspace attempts to bind a
     gmem range to multiple memslots, and fix the test that was supposed
     to ensure KVM returns -EEXIST

   - Treat memslot binding offsets and sizes as unsigned values to fix a
     bug where KVM interprets a large "offset + size" as a negative
     value and allows a nonsensical offset

   - Use the inode number instead of the page offset for the NUMA
     interleaving index to fix a bug where the effective index would
     jump by two for consecutive pages (the caller also adds in the page
     offset)

  Selftests:

   - Randomize the dirty log test's delay when reaping the bitmap on the
     first pass, as always waiting only 1ms hid a KVM RISC-V bug as the
     test reaped the bitmap before KVM could build up enough state to
     hit the bug

   - A pile of one-off fixes and cleanups"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (326 commits)
  KVM: x86/mmu: Ensure hugepage is in by slot before checking max mapping level
  KVM: x86: Fix shadow paging use-after-free due to unexpected role
  KVM: s390: Introducing kvm_arch_set_irq_inatomic fast inject
  KVM: s390: Enable adapter_indicators_set to use mapped pages
  KVM: s390: Add map/unmap ioctl and clean mappings post-guest
  riscv: kvm: Use endian-specific __lelong for NACL shared memory
  KVM: selftests: access_tracking_perf_test: bump number of NUMA nodes to 32
  KVM: s390: vsie: Implement ASTFLEIE facility 2
  KVM: s390: vsie: Refactor handle_stfle
  s390/sclp: Detect ASTFLEIE 2 facility
  KVM: s390: Minor refactor of base/ext facility lists
  KVM: x86/mmu: move pdptrs out of the MMU
  KVM: x86: check that kvm_handle_invpcid is only invoked with shadow paging
  KVM: nSVM: invalidate cached PDPTRs across nested NPT transitions
  KVM: nVMX: remove unnecessary code in prepare_vmcs02_rare
  KVM: x86: remove nested_mmu from mmu_is_nested()
  KVM: arm64: vgic-its: Make ABI commit helpers return void
  KVM: s390: Initialize KVM_S390_GET_CMMA_BITS memory
  LoongArch: KVM: Add missing slots_lock for device register/unregister
  LoongArch: KVM: Validate irqchip index in irqfd routing
  ...
2026-06-19 08:56:49 -07:00
Hyunwoo Kim
8cc8bbbfab KVM: arm64: Bound used_lrs when flushing the pKVM hyp vCPU
flush_hyp_vcpu() copies the host vGIC state into the hyp's private vCPU
on every run. The vGIC list register save and restore use used_lrs as
their loop bound and expect it to stay within the number of implemented
list registers. While this is generally the case, flush_hyp_vcpu()
copies vgic_v3 verbatim and does not enforce this, so a value provided
by the host is used at EL2 to index vgic_lr[] and access ICH_LR<n>_EL2
(host -> EL2).

Fix by clamping used_lrs to the number of implemented list registers
after the copy, as the trusted path already does in
vgic_flush_lr_state(). The number of implemented list registers is
constant after init, so it is replicated once from
kvm_vgic_global_state.nr_lr into hyp_gicv3_nr_lr rather than read on
every entry.

Cc: stable@vger.kernel.org
Fixes: be66e67f17 ("KVM: arm64: Use the pKVM hyp vCPU structure in handle___kvm_vcpu_run()")
Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com>
Reviewed-by: Fuad Tabba <tabba@google.com>
Tested-by: Fuad Tabba <tabba@google.com>
Link: https://patch.msgid.link/20260606175614.83273-3-imv4bel@gmail.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2026-06-07 14:07:06 +01:00
Mark Rutland
2768101b39 arm64: fpsimd: Move SVE save/restore inline
Currently the SVE register save/restore sequences are written in
out-of-line assembly routines. While this works, it's somewhat painful:

* As KVM needs to be able to use the sequences in hyp code, separate
  assembly files are used for the regular kernel and KVM code. While the
  common logic is shared in assembly macros, this still requires some
  duplication, and has lead to some trivial divergence.

* As the SVE LDR/STR instrucitons have limited addressing modes, the
  assembly macros use an awkward pattern requiring negative offsets.
  This could be written more clearly with addresses being generated in C
  code.

* As the FFR does not always exist in streaming mode, some awkward
  conditional branching has been written in assembly which could be
  clearer in C (and would permit the compiler to optimize out
  unnecessary branches in some cases).

* For historical reasons, the assembly macros take some register
  arguments as numerical indices (e.g. "sve_save 0, x1" uses x0 and x1),
  which is simply confusing.

* For historical reasons, the SVE save/restore code and FPSIMD
  save/restore code have a distinct sequences for FPSR and FPCR. Ideally
  this logic would be shared.

* The assembly sequences can't be instrumented, and so it's harder than
  necessary to catch memory safety issues.

To handle the above, move the SVE register save/restore sequences
to inline assembly.

Neither GCC nor LLVM instrument memory arguments to inline assembly, so
explicit instrumentation is added in the same manner as other assembly
routines. This instrumentation is implicitly disabled by Kbuild for nVHE
hyp code.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oupton@kernel.org>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
2026-06-03 16:50:49 +01:00
Mark Rutland
e1b163e405 arm64: fpsimd: Use opaque type for SVE state
As the SVE state size can vary at runtime, we don't have a concrete type
for the in-memory SVE state, and pass this around using a pointer to
void. The functions which save/restore the SVE state have a very unusual
calling convention, expecting a pointer to the FFR *in the middle of*
the in-memory SVE state, which is also passed as a pointer to void.
Passing a pointer to the FFR also requires that callers find the live VL
and perform some arithmetic, which callers implement differently.

Using pointer to void means that it's very easy to introduce errors that
cannot be caught by the compiler (e.g. as 'void **' can be assigned to
'void *'). In general this is unnecessarily confusing and fragile.

Improve this by adding an opaque 'struct arm64_sve_state', and
consistently passing a pointer to this, performing the necessary
offsetting *within* the save/restore functions.

For the moment, the offsetting is performed in a new '_sve_pffr'
assembly macro, using the ADDVL and ADDPL instructions. These add a
multiple of the live vector length and predicate length respectively.
The ADDVL immediate range cannot encode 32, so this is split into two
increments of 16.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oupton@kernel.org>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
2026-06-03 16:50:49 +01:00
Mark Rutland
890712d450 arm64: fpsimd: Move fpsimd save/restore inline
Currently the FPSIMD register save/restore sequences are written in
out-of-line assembly routines. While this works, it's somewhat painful:

* As KVM needs to be able to use the sequences in hyp code, separate
  assembly files are used for the regular kernel and KVM code. While the
  common logic is shared in assembly macros, this still requires some
  duplication, and has lead to some trivial divergence.

* For historical reasons, the assembly macros take some register
  arguments as numerical indices (e.g. "fpsimd_save x0, 8" uses x0 and
  x8), which is simply confusing.

* For historical reasons, the SVE save/restore code and FPSIMD
  save/restore code have distinct sequences for FPSR and FPCR. Ideally
  this logic would be shared.

* The assembly sequences can't be instrumented, and so it's harder than
  necessary to catch memory safety issues.

To handle the above, move the FPSIMD register save/restore sequences to
inline assembly, and share the FPSR+FPCR save/restore with SVE.

Neither GCC nor LLVM instrument memory arguments to inline assembly, so
explicit instrumentation is added in the same manner as other assembly
routines. This instrumentation is implicitly disabled by Kbuild for nVHE
hyp code.

I've used the SVE sequence for restoring FPCR, which uses an
unconditional write to FPCR, rather than the conditional write used by
the FPSIMD assembly sequence. I believe that in practice, this doesn't
matter to a real workload, and given it's possible for the mis-predicted
branch to cost more than the necessary micro-architectural
synchronization, I strongly suspect any performance impact is within the
noise.

Looking at the history, the FPSIMD assembly sequence was changed to use
a conditional write to FPCR since 2014 in commit:

  5959e25729 ("arm64: fpsimd: avoid restoring fpcr if the contents haven't change")

... as described in the commit message, this was based on an expectation
of implementation style, and was not based on benchmarking.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oupton@kernel.org>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
2026-06-03 16:50:49 +01:00
Mark Rutland
1277531fca arm64: fpsimd: Split FPSR/FPCR from SVE save/restore
Regardless of whether the vector registers are saved in FPSIMD or SVE
format, we store FPSR and FPCR in user_fpsimd_state::{fpsr,fpcr}.

For historical reasons, the functions which save/restore SVE context
take a pointer to user_fpsimd_state::fpsr, and use this to access both
user_fpsimd_state::fpsr and user_fpsimd_state::fpcr. This is
unnecessarily fragile.

Move the save/restore of FPSR and FPCR into separate helper functions
which take a pointer to user_fpsimd_state. I've used read_sysreg_s() and
write_sysreg_s() as contemporary versions of LLVM will refuse to
directly assemble accesses to FPCR or FPSR unless the "fp" arch
extension is enabled.

For the moment, fpsimd_save_state() and fpsimd_load_state() are left
as-is with their own logic to save/restore FPSR and FPCR. This will be
unified in subsequent patches.

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: James Morse <james.morse@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oupton@kernel.org>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
2026-06-03 16:50:48 +01:00
Marc Zyngier
f8078d51ee Merge branch kvm-arm64/vgic-v5-ppi into kvmarm-master/next
* kvm-arm64/vgic-v5-ppi: (40 commits)
  : .
  : Add initial GICv5 support for KVM guests, only adding PPI support
  : for the time being. Patches courtesy of Sascha Bischoff.
  :
  : From the cover letter:
  :
  : "This is v7 of the patch series to add the virtual GICv5 [1] device
  : (vgic_v5). Only PPIs are supported by this initial series, and the
  : vgic_v5 implementation is restricted to the CPU interface,
  : only. Further patch series are to follow in due course, and will add
  : support for SPIs, LPIs, the GICv5 IRS, and the GICv5 ITS."
  : .
  KVM: arm64: selftests: Add no-vgic-v5 selftest
  KVM: arm64: selftests: Introduce a minimal GICv5 PPI selftest
  KVM: arm64: gic-v5: Communicate userspace-driveable PPIs via a UAPI
  Documentation: KVM: Introduce documentation for VGICv5
  KVM: arm64: gic-v5: Probe for GICv5 device
  KVM: arm64: gic-v5: Set ICH_VCTLR_EL2.En on boot
  KVM: arm64: gic-v5: Introduce kvm_arm_vgic_v5_ops and register them
  KVM: arm64: gic-v5: Hide FEAT_GCIE from NV GICv5 guests
  KVM: arm64: gic: Hide GICv5 for protected guests
  KVM: arm64: gic-v5: Mandate architected PPI for PMU emulation on GICv5
  KVM: arm64: gic-v5: Enlighten arch timer for GICv5
  irqchip/gic-v5: Introduce minimal irq_set_type() for PPIs
  KVM: arm64: gic-v5: Initialise ID and priority bits when resetting vcpu
  KVM: arm64: gic-v5: Create and initialise vgic_v5
  KVM: arm64: gic-v5: Support GICv5 interrupts with KVM_IRQ_LINE
  KVM: arm64: gic-v5: Implement direct injection of PPIs
  KVM: arm64: Introduce set_direct_injection irq_op
  KVM: arm64: gic-v5: Trap and mask guest ICC_PPI_ENABLERx_EL1 writes
  KVM: arm64: gic-v5: Check for pending PPIs
  KVM: arm64: gic-v5: Clear TWI if single task running
  ...

Signed-off-by: Marc Zyngier <maz@kernel.org>
2026-04-08 12:22:35 +01:00
Sascha Bischoff
5aefaf11f9 KVM: arm64: gic: Hide GICv5 for protected guests
We don't support running protected guest with GICv5 at the moment.
Therefore, be sure that we don't expose it to the guest at all by
actively hiding it when running a protected guest.

Signed-off-by: Sascha Bischoff <sascha.bischoff@arm.com>
Link: https://patch.msgid.link/20260319154937.3619520-34-sascha.bischoff@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2026-03-19 18:21:29 +00:00
Sascha Bischoff
af325e87af KVM: arm64: gic-v5: Add vgic-v5 save/restore hyp interface
Introduce the following hyp functions to save/restore GICv5 state:

* __vgic_v5_save_apr()
* __vgic_v5_restore_vmcr_apr()
* __vgic_v5_save_ppi_state()	- no hypercall required
* __vgic_v5_restore_ppi_state()	- no hypercall required
* __vgic_v5_save_state()	- no hypercall required
* __vgic_v5_restore_state()	- no hypercall required

Note that the functions tagged as not requiring hypercalls are always
called directly from the same context. They are either called via the
vgic_save_state()/vgic_restore_state() path when running with VHE, or
via __hyp_vgic_save_state()/__hyp_vgic_restore_state() otherwise. This
mimics how vgic_v3_save_state()/vgic_v3_restore_state() are
implemented.

Overall, the state of the following registers is saved/restored:

* ICC_ICSR_EL1
* ICH_APR_EL2
* ICH_PPI_ACTIVERx_EL2
* ICH_PPI_DVIRx_EL2
* ICH_PPI_ENABLERx_EL2
* ICH_PPI_PENDRx_EL2
* ICH_PPI_PRIORITYRx_EL2
* ICH_VMCR_EL2

All of these are saved/restored to/from the KVM vgic_v5 CPUIF shadow
state, with the exception of the PPI active, pending, and enable
state. The pending state is saved and restored from kvm_host_data as
any changes here need to be tracked and propagated back to the
vgic_irq shadow structures (coming in a future commit). Therefore, an
entry and an exit copy is required. The active and enable state is
restored from the vgic_v5 CPUIF, but is saved to kvm_host_data. Again,
this needs to by synced back into the shadow data structures.

The ICSR must be save/restored as this register is shared between host
and guest. Therefore, to avoid leaking host state to the guest, this
must be saved and restored. Moreover, as this can by used by the host
at any time, it must be save/restored eagerly. Note: the host state is
not preserved as the host should only use this register when
preemption is disabled.

As with GICv3, the VMCR is eagerly saved as this is required when
checking if interrupts can be injected or not, and therefore impacts
things such as WFI.

As part of restoring the ICH_VMCR_EL2 and ICH_APR_EL2, GICv3-compat
mode is also disabled by setting the ICH_VCTLR_EL2.V3 bit to 0. The
correspoinding GICv3-compat mode enable is part of the VMCR & APR
restore for a GICv3 guest as it only takes effect when actually
running a guest.

Co-authored-by: Timothy Hayes <timothy.hayes@arm.com>
Signed-off-by: Timothy Hayes <timothy.hayes@arm.com>
Signed-off-by: Sascha Bischoff <sascha.bischoff@arm.com>
Link: https://patch.msgid.link/20260319154937.3619520-17-sascha.bischoff@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2026-03-19 18:21:28 +00:00
Vincent Donnefort
8bbeb4d169 KVM: arm64: Initialise hyp_nr_cpus for nVHE hyp
Knowing the number of CPUs is necessary for determining the boundaries
of per-cpu variables, which will be used for upcoming hypervisor
tracing. hyp_nr_cpus which stores this value, is only initialised for
the pKVM hypervisor. Make it accessible for the nVHE hypervisor as well.

With the kernel now responsible for initialising hyp_nr_cpus, the
nr_cpus parameter is no longer needed in __pkvm_init.

Signed-off-by: Vincent Donnefort <vdonnefort@google.com>
Link: https://patch.msgid.link/20260309162516.2623589-22-vdonnefort@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2026-03-11 08:51:16 +00:00
Marc Zyngier
eb33ffa2bd KVM: arm64: GICv3: nv: Resync LRs/VMCR/HCR early for better MI emulation
The current approach to nested GICv3 support is to not do anything
while L2 is running, wait a transition from L2 to L1 to resync
LRs, VMCR and HCR, and only then evaluate the state to decide
whether to generate a maintenance interrupt.

This doesn't provide a good quality of emulation, and it would be
far preferable to find out early that we need to perform a switch.

Move the LRs/VMCR and HCR resync into vgic_v3_sync_nested(), so
that we have most of the state available. As we turning the vgic
off at this stage to avoid a screaming host MI, add a new helper
vgic_v3_flush_nested() that switches the vgic on again. The MI can
then be directly injected as required.

Tested-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Link: https://msgid.link/20251120172540.2267180-35-maz@kernel.org
Signed-off-by: Oliver Upton <oupton@kernel.org>
2025-11-24 14:29:14 -08:00
Marc Zyngier
cf72ee6371 KVM: arm64: Eagerly save VMCR on exit
We currently save/restore the VMCR register in a pretty lazy way
(on load/put, consistently with what we do with the APRs).

However, we are going to need the group-enable bits that are backed
by VMCR on each entry (so that we can avoid injecting interrupts for
disabled groups).

Move the synchronisation from put to sync, which results in some minor
churn in the nVHE hypercalls to simplify things.

Tested-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Link: https://msgid.link/20251120172540.2267180-21-maz@kernel.org
Signed-off-by: Oliver Upton <oupton@kernel.org>
2025-11-24 14:29:13 -08:00
Marc Zyngier
146a050f2d KVM: arm64: nv: Nested GICv3 emulation
When entering a nested VM, we set up the hypervisor control interface
based on what the guest hypervisor has set. Especially, we investigate
each list register written by the guest hypervisor whether HW bit is
set.  If so, we translate hw irq number from the guest's point of view
to the real hardware irq number if there is a mapping.

Co-developed-by: Jintack Lim <jintack@cs.columbia.edu>
Signed-off-by: Jintack Lim <jintack@cs.columbia.edu>
[Christoffer: Redesigned execution flow around vcpu load/put]
Co-developed-by: Christoffer Dall <christoffer.dall@arm.com>
Signed-off-by: Christoffer Dall <christoffer.dall@arm.com>
[maz: Rewritten to support GICv3 instead of GICv2, NV2 support]
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20250225172930.1850838-9-maz@kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
2025-03-03 14:57:04 -08:00
Pierre-Clément Tosi
ea9d7c83d1 KVM: arm64: Fix __pkvm_init_switch_pgd call ABI
Fix the mismatch between the (incorrect) C signature, C call site, and
asm implementation by aligning all three on an API passing the
parameters (pgd and SP) separately, instead of as a bundled struct.

Remove the now unnecessary memory accesses while the MMU is off from the
asm, which simplifies the C caller (as it does not need to convert a VA
struct pointer to PA) and makes the code slightly more robust by
offsetting the struct fields from C and properly expressing the call to
the C compiler (e.g. type checker and kCFI).

Fixes: f320bc742b ("KVM: arm64: Prepare the creation of s1 mappings at EL2")
Signed-off-by: Pierre-Clément Tosi <ptosi@google.com>
Acked-by: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20240610063244.2828978-3-ptosi@google.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
2024-06-20 17:40:53 +00:00
Fuad Tabba
66d5b53e20 KVM: arm64: Allocate memory mapped at hyp for host sve state in pKVM
Protected mode needs to maintain (save/restore) the host's sve
state, rather than relying on the host kernel to do that. This is
to avoid leaking information to the host about guests and the
type of operations they are performing.

As a first step towards that, allocate memory mapped at hyp, per
cpu, for the host sve state. The following patch will use this
memory to save/restore the host state.

Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Fuad Tabba <tabba@google.com>
Link: https://lore.kernel.org/r/20240603122852.3923848-6-tabba@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2024-06-04 15:06:33 +01:00
Fuad Tabba
45f4ea9bcf KVM: arm64: Fix prototype for __sve_save_state/__sve_restore_state
Since the prototypes for __sve_save_state/__sve_restore_state at
hyp were added, the underlying macro has acquired a third
parameter for saving/restoring ffr.

Fix the prototypes to account for the third parameter, and
restore the ffr for the guest since it is saved.

Suggested-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Fuad Tabba <tabba@google.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20240603122852.3923848-3-tabba@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2024-06-04 15:06:32 +01:00
Fuad Tabba
87bb39ed40 KVM: arm64: Reintroduce __sve_save_state
Now that the hypervisor is handling the host sve state in
protected mode, it needs to be able to save it.

This reverts commit e66425fc9b ("KVM: arm64: Remove unused
__sve_save_state").

Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Fuad Tabba <tabba@google.com>
Link: https://lore.kernel.org/r/20240603122852.3923848-2-tabba@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2024-06-04 15:06:32 +01:00
Marc Zyngier
948e1a53c2 KVM: arm64: Simplify vgic-v3 hypercalls
Consolidate the GICv3 VMCR accessor hypercalls into the APR save/restore
hypercalls so that all of the EL2 GICv3 state is covered by a single pair
of hypercalls.

Signed-off-by: Fuad Tabba <tabba@google.com>
Acked-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20240423150538.2103045-17-tabba@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
2024-05-01 16:48:14 +01:00
Bjorn Helgaas
75841d89f3 KVM: arm64: Fix typos
Fix typos, most reported by "codespell arch/arm64".  Only touches comments,
no code changes.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: James Morse <james.morse@arm.com>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: Zenghui Yu <yuzenghui@huawei.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: kvmarm@lists.linux.dev
Reviewed-by: Zenghui Yu <yuzenghui@huawei.com>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Link: https://lore.kernel.org/r/20240103231605.1801364-6-helgaas@kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
2024-02-24 09:13:33 +00:00
Oliver Upton
27cde4c0fe KVM: arm64: Rename helpers for VHE vCPU load/put
The names for the helpers we expose to the 'generic' KVM code are a bit
imprecise; we switch the EL0 + EL1 sysreg context and setup trap
controls that do not need to change for every guest entry/exit. Rename +
shuffle things around a bit in preparation for loading the stage-2 MMU
context on vcpu_load().

Link: https://lore.kernel.org/r/20231018233212.2888027-5-oliver.upton@linux.dev
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
2023-10-20 17:52:01 +00:00
Jean-Philippe Brucker
373beef00f KVM: arm64: nvhe: Ignore SVE hint in SMCCC function ID
When SVE is enabled, the host may set bit 16 in SMCCC function IDs, a
hint that indicates an unused SVE state. At the moment NVHE doesn't
account for this bit when inspecting the function ID, and rejects most
calls. Clear the hint bit before comparing function IDs.

About version compatibility: the host's PSCI driver initially probes the
firmware for a SMCCC version number. If the firmware implements a
protocol recent enough (1.3), subsequent SMCCC calls have the hint bit
set. Since the hint bit was reserved in earlier versions of the
protocol, clearing it is fine regardless of the version in use.

When a new hint is added to the protocol in the future, it will be added
to ARM_SMCCC_CALL_HINTS and NVHE will handle it straight away. This
patch only clears known hints and leaves reserved bits as is, because
future SMCCC versions could use reserved bits as modifiers for the
function ID, rather than hints.

Fixes: cfa7ff959a ("arm64: smccc: Support SMCCC v1.3 SVE register saving hint")
Reported-by: Ben Horgan <ben.horgan@arm.com>
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230911145254.934414-4-jean-philippe@linaro.org
2023-09-12 13:07:37 +01:00
Marc Zyngier
6f617d3aa6 KVM: arm64: Key use of VHE instructions in nVHE code off ARM64_KVM_HVHE
We can now start with the fun stuff: if we enable VHE *only* for
the hypervisor, we need to generate the VHE instructions when
accessing the system registers.

For this, reporpose the alternative sequence to be keyed off
ARM64_KVM_HVHE in the nVHE hypervisor code, and only there.

Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230609162200.2024064-10-maz@kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
2023-06-12 23:17:23 +00:00
Marc Zyngier
57e784b407 KVM: arm64: Remove alternatives from sysreg accessors in VHE hypervisor context
In the VHE hypervisor code, we should be using the remapped VHE
accessors, no ifs, no buts. No need to generate any alternative.

Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230609162200.2024064-9-maz@kernel.org
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
2023-06-12 23:17:23 +00:00
Quentin Perret
8669651ce0 KVM: arm64: Provide sanitized SYS_ID_AA64SMFR0_EL1 to nVHE
We will need a sanitized copy of SYS_ID_AA64SMFR0_EL1 from the nVHE EL2
code shortly, so make sure to provide it with a copy.

Signed-off-by: Quentin Perret <qperret@google.com>
Acked-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20230201103755.1398086-2-qperret@google.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
2023-02-02 21:46:43 +00:00
Will Deacon
73f38ef2ae KVM: arm64: Maintain a copy of 'kvm_arm_vmid_bits' at EL2
Sharing 'kvm_arm_vmid_bits' between EL1 and EL2 allows the host to
modify the variable arbitrarily, potentially leading to all sorts of
shenanians as this is used to configure the VTTBR register for the
guest stage-2.

In preparation for unmapping host sections entirely from EL2, maintain
a copy of 'kvm_arm_vmid_bits' in the pKVM hypervisor and initialise it
from the host value while it is still trusted.

Tested-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221110190259.26861-23-will@kernel.org
2022-11-11 17:19:35 +00:00
Will Deacon
13e248aab7 KVM: arm64: Provide I-cache invalidation by virtual address at EL2
In preparation for handling cache maintenance of guest pages from within
the pKVM hypervisor at EL2, introduce an EL2 copy of icache_inval_pou()
which will later be plumbed into the stage-2 page-table cache
maintenance callbacks, ensuring that the initial contents of pages
mapped as executable into the guest stage-2 page-table is visible to the
instruction fetcher.

Tested-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221110190259.26861-17-will@kernel.org
2022-11-11 17:16:25 +00:00
Vladimir Murzin
def8c222f0 arm64: Add support of PAuth QARMA3 architected algorithm
QARMA3 is relaxed version of the QARMA5 algorithm which expected to
reduce the latency of calculation while still delivering a suitable
level of security.

Support for QARMA3 can be discovered via ID_AA64ISAR2_EL1

    APA3, bits [15:12] Indicates whether the QARMA3 algorithm is
                       implemented in the PE for address
                       authentication in AArch64 state.

    GPA3, bits [11:8]  Indicates whether the QARMA3 algorithm is
                       implemented in the PE for generic code
                       authentication in AArch64 state.

Signed-off-by: Vladimir Murzin <vladimir.murzin@arm.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20220224124952.119612-4-vladimir.murzin@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
2022-02-25 13:38:52 +00:00
Marc Zyngier
e66425fc9b KVM: arm64: Remove unused __sve_save_state
Now that we don't have any users left for __sve_save_state, remove
it altogether. Should we ever need to save the SVE state from the
hypervisor again, we can always re-introduce it.

Suggested-by: Zenghui Yu <yuzenghui@huawei.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
2021-11-22 16:01:39 +00:00
Fuad Tabba
6c30bfb18d KVM: arm64: Add handlers for protected VM System Registers
Add system register handlers for protected VMs. These cover Sys64
registers (including feature id registers), and debug.

No functional change intended as these are not hooked in yet to
the guest exit handlers introduced earlier. So when trapping is
triggered, the exit handlers let the host handle it, as before.

Reviewed-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20211010145636.1950948-8-tabba@google.com
2021-10-11 14:57:29 +01:00
Fuad Tabba
1460b4b25f KVM: arm64: Restore mdcr_el2 from vcpu
On deactivating traps, restore the value of mdcr_el2 from the
newly created and preserved host value vcpu context, rather than
directly reading the hardware register.

Up until and including this patch the two values are the same,
i.e., the hardware register and the vcpu one. A future patch will
be changing the value of mdcr_el2 on activating traps, and this
ensures that its value will be restored.

No functional change intended.

Signed-off-by: Fuad Tabba <tabba@google.com>
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210817081134.2918285-7-tabba@google.com
2021-08-20 11:12:17 +01:00
Marc Zyngier
5c92a7643b Merge branch 'kvm-arm64/nvhe-panic-info' into kvmarm-master/next
Signed-off-by: Marc Zyngier <maz@kernel.org>
2021-04-13 15:38:03 +01:00
Andrew Scull
f79e616f27 KVM: arm64: Use BUG and BUG_ON in nVHE hyp
hyp_panic() reports the address of the panic by using ELR_EL2, but this
isn't a useful address when hyp_panic() is called directly. Replace such
direct calls with BUG() and BUG_ON() which use BRK to trigger an
exception that then goes to hyp_panic() with the correct address. Also
remove the hyp_panic() declaration from the header file to avoid
accidental misuse.

Signed-off-by: Andrew Scull <ascull@google.com>
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210318143311.839894-5-ascull@google.com
2021-04-01 09:54:37 +01:00
Marc Zyngier
7c4199375a KVM: arm64: Drop the CPU_FTR_REG_HYP_COPY infrastructure
Now that the read_ctr macro has been specialised for nVHE,
the whole CPU_FTR_REG_HYP_COPY infrastrcture looks completely
overengineered.

Simplify it by populating the two u64 quantities (MMFR0 and 1)
that the hypervisor need.

Reviewed-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
2021-03-25 11:01:03 +00:00
Quentin Perret
f320bc742b KVM: arm64: Prepare the creation of s1 mappings at EL2
When memory protection is enabled, the EL2 code needs the ability to
create and manage its own page-table. To do so, introduce a new set of
hypercalls to bootstrap a memory management system at EL2.

This leads to the following boot flow in nVHE Protected mode:

 1. the host allocates memory for the hypervisor very early on, using
    the memblock API;

 2. the host creates a set of stage 1 page-table for EL2, installs the
    EL2 vectors, and issues the __pkvm_init hypercall;

 3. during __pkvm_init, the hypervisor re-creates its stage 1 page-table
    and stores it in the memory pool provided by the host;

 4. the hypervisor then extends its stage 1 mappings to include a
    vmemmap in the EL2 VA space, hence allowing to use the buddy
    allocator introduced in a previous patch;

 5. the hypervisor jumps back in the idmap page, switches from the
    host-provided page-table to the new one, and wraps up its
    initialization by enabling the new allocator, before returning to
    the host.

 6. the host can free the now unused page-table created for EL2, and
    will now need to issue hypercalls to make changes to the EL2 stage 1
    mappings instead of modifying them directly.

Note that for the sake of simplifying the review, this patch focuses on
the hypervisor side of things. In other words, this only implements the
new hypercalls, but does not make use of them from the host yet. The
host-side changes will follow in a subsequent patch.

Credits to Will for __pkvm_init_switch_pgd.

Acked-by: Will Deacon <will@kernel.org>
Co-authored-by: Will Deacon <will@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210319100146.1149909-18-qperret@google.com
2021-03-19 12:01:21 +00:00
Marc Zyngier
a1baa01f76 Linux 5.12-rc3
-----BEGIN PGP SIGNATURE-----
 
 iQFSBAABCAA8FiEEq68RxlopcLEwq+PEeb4+QwBBGIYFAmBOgu4eHHRvcnZhbGRz
 QGxpbnV4LWZvdW5kYXRpb24ub3JnAAoJEHm+PkMAQRiGUd0H/3Ey8aWjVAig9Pe+
 VQVZKwG+LXWH6UmUx5qyaTxophhmGnWLvkigJMn63qIg4eQtfp2gNFHK+T4OJNIP
 ybnkjFZ337x4J9zD6m8mt4Wmelq9iW2wNOS+3YZAyYiGlXfMGM7SlYRCQRQznTED
 2O/JCMsOoP+Z8tr5ah/bzs0dANsXmTZ3QqRP2uzb6irKTgFR3/weOhj+Ht1oJ4Aq
 V+bgdcwhtk20hJhlvVeqws+o74LR789tTDCknlz/YNMv9e6VPfyIQ5vJAcFmZATE
 Ezj9yzkZ4IU+Ux6ikAyaFyBU8d1a4Wqye3eHCZBsEo6tcSAhbTZ90eoU86vh6ajS
 LZjwkNw=
 =6y1u
 -----END PGP SIGNATURE-----

Merge tag 'v5.12-rc3' into kvm-arm64/host-stage2

Linux 5.12-rc3

Signed-off-by: Marc Zyngier <maz@kernel.org>

# gpg: Signature made Sun 14 Mar 2021 21:41:02 GMT
# gpg:                using RSA key ABAF11C65A2970B130ABE3C479BE3E4300411886
# gpg:                issuer "torvalds@linux-foundation.org"
# gpg: Can't check signature: No public key
2021-03-19 12:01:04 +00:00
Marc Zyngier
52029198c1 KVM: arm64: Rework SVE host-save/guest-restore
In order to keep the code readable, move the host-save/guest-restore
sequences in their own functions, with the following changes:
- the hypervisor ZCR is now set from C code
- ZCR_EL2 is always used as the EL2 accessor

This results in some minor assembler macro rework.
No functional change intended.

Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
2021-03-18 13:57:37 +00:00
Marc Zyngier
297b8603e3 KVM: arm64: Provide KVM's own save/restore SVE primitives
as we are about to change the way KVM deals with SVE, provide
KVM with its own save/restore SVE primitives.

No functional change intended.

Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
2021-03-18 11:23:14 +00:00
Andrew Scull
c4b000c392 KVM: arm64: Fix nVHE hyp panic host context restore
When panicking from the nVHE hyp and restoring the host context, x29 is
expected to hold a pointer to the host context. This wasn't being done
so fix it to make sure there's a valid pointer the host context being
used.

Rather than passing a boolean indicating whether or not the host context
should be restored, instead pass the pointer to the host context. NULL
is passed to indicate that no context should be restored.

Fixes: a2e102e20f ("KVM: arm64: nVHE: Handle hyp panics")
Cc: stable@vger.kernel.org
Signed-off-by: Andrew Scull <ascull@google.com>
[maz: partial rewrite to fit 5.12-rc1]
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210219122406.1337626-1-ascull@google.com
Message-Id: <20210305185254.3730990-4-maz@kernel.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2021-03-06 04:18:40 -05:00
Suzuki K Poulose
b96b0c5de6 KVM: arm64: nvhe: Save the SPE context early
The nVHE KVM hyp drains and disables the SPE buffer, before
entering the guest, as the EL1&0 translation regime
is going to be loaded with that of the guest.

But this operation is performed way too late, because :
  - The owning translation regime of the SPE buffer
    is transferred to EL2. (MDCR_EL2_E2PB == 0)
  - The guest Stage1 is loaded.

Thus the flush could use the host EL1 virtual address,
but use the EL2 translations instead of host EL1, for writing
out any cached data.

Fix this by moving the SPE buffer handling early enough.
The restore path is doing the right thing.

Fixes: 014c4c77aa ("KVM: arm64: Improve debug register save/restore flow")
Cc: stable@vger.kernel.org
Cc: Christoffer Dall <christoffer.dall@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Alexandru Elisei <alexandru.elisei@arm.com>
Reviewed-by: Alexandru Elisei <alexandru.elisei@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210302120345.3102874-1-suzuki.poulose@arm.com
Message-Id: <20210305185254.3730990-2-maz@kernel.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2021-03-06 04:18:39 -05:00
David Brazdil
eeeee7193d KVM: arm64: Bootstrap PSCI SMC handler in nVHE EL2
Add a handler of PSCI SMCs in nVHE hyp code. The handler is initialized
with the version used by the host's PSCI driver and the function IDs it
was configured with. If the SMC function ID matches one of the
configured PSCI calls (for v0.1) or falls into the PSCI function ID
range (for v0.2+), the SMC is handled by the PSCI handler. For now, all
SMCs return PSCI_RET_NOT_SUPPORTED.

Signed-off-by: David Brazdil <dbrazdil@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20201202184122.26046-17-dbrazdil@google.com
2020-12-04 10:08:34 +00:00
David Brazdil
63fec24351 KVM: arm64: Move hyp-init params to a per-CPU struct
Once we start initializing KVM on newly booted cores before the rest of
the kernel, parameters to __do_hyp_init will need to be provided by EL2
rather than EL1. At that point it will not be possible to pass its three
arguments directly because PSCI_CPU_ON only supports one context
argument.

Refactor __do_hyp_init to accept its parameters in a struct. This
prepares the code for KVM booting cores as well as removes any limits on
the number of __do_hyp_init arguments.

Signed-off-by: David Brazdil <dbrazdil@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20201202184122.26046-11-dbrazdil@google.com
2020-12-04 10:08:32 +00:00
Andrew Scull
a2e102e20f KVM: arm64: nVHE: Handle hyp panics
Restore the host context when panicking from hyp to give the best chance
of the panic being clean.

The host requires that registers be preserved such as x18 for the shadow
callstack. If the panic is caused by an exception from EL1, the host
context is still valid so the panic can return straight back to the
host. If the panic comes from EL2 then it's most likely that the hyp
context is active and the host context needs to be restored.

There are windows before and after the host context is saved and
restored that restoration is attempted incorrectly and the panic won't
be clean.

Signed-off-by: Andrew Scull <ascull@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200915104643.2543892-14-ascull@google.com
2020-09-15 18:39:03 +01:00
Andrew Scull
b619d9aa8b KVM: arm64: Introduce hyp context
During __guest_enter, save and restore from a new hyp context rather
than the host context. This is preparation for separation of the hyp and
host context in nVHE.

Signed-off-by: Andrew Scull <ascull@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200915104643.2543892-9-ascull@google.com
2020-09-15 18:39:02 +01:00
Andrew Scull
a0e479523e KVM: arm64: Save chosen hyp vector to a percpu variable
Introduce a percpu variable to hold the address of the selected hyp
vector that will be used with guests. This avoids the selection process
each time a guest is being entered and can be used by nVHE when a
separate vector is introduced for the host.

Signed-off-by: Andrew Scull <ascull@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200915104643.2543892-6-ascull@google.com
2020-09-15 18:39:02 +01:00
Andrew Scull
6a0259ed29 KVM: arm64: Remove hyp_panic arguments
hyp_panic is able to find all the context it needs from within itself so
remove the argument. The __hyp_panic wrapper becomes redundant so is
also removed.

Signed-off-by: Andrew Scull <ascull@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200915104643.2543892-3-ascull@google.com
2020-09-15 18:39:01 +01:00
David Brazdil
c50cb04303 KVM: arm64: Remove __hyp_text macro, use build rules instead
With nVHE code now fully separated from the rest of the kernel, the effects of
the __hyp_text macro (which had to be applied on all nVHE code) can be
achieved with build rules instead. The macro used to:
  (a) move code to .hyp.text ELF section, now done by renaming .text using
      `objcopy`, and
  (b) `notrace` and `__noscs` would negate effects of CC_FLAGS_FTRACE and
      CC_FLAGS_SCS, respectivelly, now those flags are  erased from
      KBUILD_CFLAGS (same way as in EFI stub).

Note that by removing __hyp_text from code shared with VHE, all VHE code is now
compiled into .text and without `notrace` and `__noscs`.

Use of '.pushsection .hyp.text' removed from assembly files as this is now also
covered by the build rules.

For MAINTAINERS: if needed to re-run, uses of macro were removed with the
following command. Formatting was fixed up manually.

  find arch/arm64/kvm/hyp -type f -name '*.c' -o -name '*.h' \
       -exec sed -i 's/ __hyp_text//g' {} +

Signed-off-by: David Brazdil <dbrazdil@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200625131420.71444-15-dbrazdil@google.com
2020-07-05 18:38:45 +01:00
David Brazdil
9aebdea494 KVM: arm64: Duplicate hyp/timer-sr.c for VHE/nVHE
timer-sr.c contains a HVC handler for setting CNTVOFF_EL2 and two helper
functions for controlling access to physical counter. The former is used by
both VHE/nVHE and is duplicated, the latter are used only by nVHE and moved
to nvhe/timer-sr.c.

Signed-off-by: David Brazdil <dbrazdil@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200625131420.71444-13-dbrazdil@google.com
2020-07-05 18:38:38 +01:00
David Brazdil
13aeb9b400 KVM: arm64: Split hyp/sysreg-sr.c to VHE/nVHE
sysreg-sr.c contains KVM's code for saving/restoring system registers, with
some code shared between VHE/nVHE. These common routines are moved to
a header file, VHE-specific code is moved to vhe/sysreg-sr.c and nVHE-specific
code to nvhe/sysreg-sr.c.

Signed-off-by: David Brazdil <dbrazdil@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200625131420.71444-12-dbrazdil@google.com
2020-07-05 18:38:29 +01:00
David Brazdil
09cf57eba3 KVM: arm64: Split hyp/switch.c to VHE/nVHE
switch.c implements context-switching for KVM, with large parts shared between
VHE/nVHE. These common routines are moved to a header file, VHE-specific code
is moved to vhe/switch.c and nVHE-specific code is moved to nvhe/switch.c.

Previously __kvm_vcpu_run needed a different symbol name for VHE/nVHE. This
is cleaned up and the caller in arm.c simplified.

Signed-off-by: David Brazdil <dbrazdil@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200625131420.71444-10-dbrazdil@google.com
2020-07-05 18:38:21 +01:00
Linus Torvalds
039aeb9deb ARM:
- Move the arch-specific code into arch/arm64/kvm
 - Start the post-32bit cleanup
 - Cherry-pick a few non-invasive pre-NV patches
 
 x86:
 - Rework of TLB flushing
 - Rework of event injection, especially with respect to nested virtualization
 - Nested AMD event injection facelift, building on the rework of generic code
 and fixing a lot of corner cases
 - Nested AMD live migration support
 - Optimization for TSC deadline MSR writes and IPIs
 - Various cleanups
 - Asynchronous page fault cleanups (from tglx, common topic branch with tip tree)
 - Interrupt-based delivery of asynchronous "page ready" events (host side)
 - Hyper-V MSRs and hypercalls for guest debugging
 - VMX preemption timer fixes
 
 s390:
 - Cleanups
 
 Generic:
 - switch vCPU thread wakeup from swait to rcuwait
 
 The other architectures, and the guest side of the asynchronous page fault
 work, will come next week.
 -----BEGIN PGP SIGNATURE-----
 
 iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAl7VJcYUHHBib256aW5p
 QHJlZGhhdC5jb20ACgkQv/vSX3jHroPf6QgAq4wU5wdd1lTGz/i3DIhNVJNJgJlp
 ozLzRdMaJbdbn5RpAK6PEBd9+pt3+UlojpFB3gpJh2Nazv2OzV4yLQgXXXyyMEx1
 5Hg7b4UCJYDrbkCiegNRv7f/4FWDkQ9dx++RZITIbxeskBBCEI+I7GnmZhGWzuC4
 7kj4ytuKAySF2OEJu0VQF6u0CvrNYfYbQIRKBXjtOwuRK4Q6L63FGMJpYo159MBQ
 asg3B1jB5TcuGZ9zrjL5LkuzaP4qZZHIRs+4kZsH9I6MODHGUxKonrkablfKxyKy
 CFK+iaHCuEXXty5K0VmWM3nrTfvpEjVjbMc7e1QGBQ5oXsDM0pqn84syRg==
 =v7Wn
 -----END PGP SIGNATURE-----

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull kvm updates from Paolo Bonzini:
 "ARM:
   - Move the arch-specific code into arch/arm64/kvm

   - Start the post-32bit cleanup

   - Cherry-pick a few non-invasive pre-NV patches

  x86:
   - Rework of TLB flushing

   - Rework of event injection, especially with respect to nested
     virtualization

   - Nested AMD event injection facelift, building on the rework of
     generic code and fixing a lot of corner cases

   - Nested AMD live migration support

   - Optimization for TSC deadline MSR writes and IPIs

   - Various cleanups

   - Asynchronous page fault cleanups (from tglx, common topic branch
     with tip tree)

   - Interrupt-based delivery of asynchronous "page ready" events (host
     side)

   - Hyper-V MSRs and hypercalls for guest debugging

   - VMX preemption timer fixes

  s390:
   - Cleanups

  Generic:
   - switch vCPU thread wakeup from swait to rcuwait

  The other architectures, and the guest side of the asynchronous page
  fault work, will come next week"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (256 commits)
  KVM: selftests: fix rdtsc() for vmx_tsc_adjust_test
  KVM: check userspace_addr for all memslots
  KVM: selftests: update hyperv_cpuid with SynDBG tests
  x86/kvm/hyper-v: Add support for synthetic debugger via hypercalls
  x86/kvm/hyper-v: enable hypercalls regardless of hypercall page
  x86/kvm/hyper-v: Add support for synthetic debugger interface
  x86/hyper-v: Add synthetic debugger definitions
  KVM: selftests: VMX preemption timer migration test
  KVM: nVMX: Fix VMX preemption timer migration
  x86/kvm/hyper-v: Explicitly align hcall param for kvm_hyperv_exit
  KVM: x86/pmu: Support full width counting
  KVM: x86/pmu: Tweak kvm_pmu_get_msr to pass 'struct msr_data' in
  KVM: x86: announce KVM_FEATURE_ASYNC_PF_INT
  KVM: x86: acknowledgment mechanism for async pf page ready notifications
  KVM: x86: interrupt based APF 'page ready' event delivery
  KVM: introduce kvm_read_guest_offset_cached()
  KVM: rename kvm_arch_can_inject_async_page_present() to kvm_arch_can_dequeue_async_page_present()
  KVM: x86: extend struct kvm_vcpu_pv_apf_data with token info
  Revert "KVM: async_pf: Fix #DF due to inject "Page not Present" and "Page Ready" exceptions simultaneously"
  KVM: VMX: Replace zero-length array with flexible-array
  ...
2020-06-03 15:13:47 -07:00