David Woodhouse [Fri, 10 Dec 2021 16:36:23 +0000 (16:36 +0000)]
KVM: x86/xen: Add KVM_IRQ_ROUTING_XEN_EVTCHN and event channel delivery
This adds basic support for delivering 2 level event channels to a guest.
Initially, it only supports delivery via the IRQ routing table, triggered
by an eventfd. In order to do so, it has a kvm_xen_set_evtchn_fast()
function which will use the pre-mapped shared_info page if it already
exists and is still valid, while the slow path through the irqfd_inject
workqueue will remap the shared_info page if necessary.
It sets the bits in the shared_info page but not the vcpu_info; that is
deferred to __kvm_xen_has_interrupt() which raises the vector to the
appropriate vCPU.
Add a 'verbose' mode to xen_shinfo_test while adding test cases for this.
David Woodhouse [Fri, 10 Dec 2021 16:36:22 +0000 (16:36 +0000)]
KVM: x86/xen: Maintain valid mapping of Xen shared_info page
Use the newly reinstated gfn_to_pfn_cache to maintain a kernel mapping
of the Xen shared_info page so that it can be accessed in atomic context.
Note that we do not participate in dirty tracking for the shared info
page and we do not explicitly mark it dirty every single tim we deliver
an event channel interrupts. We wouldn't want to do that even if we *did*
have a valid vCPU context with which to do so.
David Woodhouse [Fri, 10 Dec 2021 16:36:21 +0000 (16:36 +0000)]
KVM: Reinstate gfn_to_pfn_cache with invalidation support
This can be used in two modes. There is an atomic mode where the cached
mapping is accessed while holding the rwlock, and a mode where the
physical address is used by a vCPU in guest mode.
For the latter case, an invalidation will wake the vCPU with the new
KVM_REQ_GPC_INVALIDATE, and the architecture will need to refresh any
caches it still needs to access before entering guest mode again.
Only one vCPU can be targeted by the wake requests; it's simple enough
to make it wake all vCPUs or even a mask but I don't see a use case for
that additional complexity right now.
Invalidation happens from the invalidate_range_start MMU notifier, which
needs to be able to sleep in order to wake the vCPU and wait for it.
This means that revalidation potentially needs to "wait" for the MMU
operation to complete and the invalidate_range_end notifier to be
invoked. Like the vCPU when it takes a page fault in that period, we
just spin — fixing that in a future patch by implementing an actual
*wait* may be another part of shaving this particularly hirsute yak.
As noted in the comments in the function itself, the only case where
the invalidate_range_start notifier is expected to be called *without*
being able to sleep is when the OOM reaper is killing the process. In
that case, we expect the vCPU threads already to have exited, and thus
there will be nothing to wake, and no reason to wait. So we clear the
KVM_REQUEST_WAIT bit and send the request anyway, then complain loudly
if there actually *was* anything to wake up.
David Woodhouse [Fri, 10 Dec 2021 16:36:20 +0000 (16:36 +0000)]
KVM: Warn if mark_page_dirty() is called without an active vCPU
The various kvm_write_guest() and mark_page_dirty() functions must only
ever be called in the context of an active vCPU, because if dirty ring
tracking is enabled it may simply oops when kvm_get_running_vcpu()
returns NULL for the vcpu and then kvm_dirty_ring_get() dereferences it.
This oops was reported by "butt3rflyh4ck" <[email protected]> in
https://lore.kernel.org/kvm/CAFcO6XOmoS7EacN_n6v4Txk7xL7iqRa2gABg3F7E3Naf5uG94g@mail.gmail.com/
That actual bug will be fixed under separate cover but this warning
should help to prevent new ones from being added.
David Woodhouse [Thu, 9 Dec 2021 15:09:38 +0000 (15:09 +0000)]
x86/kvm: Silence per-cpu pr_info noise about KVM clocks and steal time
I made the actual CPU bringup go nice and fast... and then Linux spends
half a minute printing stupid nonsense about clocks and steal time for
each of 256 vCPUs. Don't do that. Nobody cares.
Eric Hankland [Tue, 30 Nov 2021 07:42:21 +0000 (15:42 +0800)]
KVM: x86: Update vPMCs when retiring branch instructions
When KVM retires a guest branch instruction through emulation,
increment any vPMCs that are configured to monitor "branch
instructions retired," and update the sample period of those counters
so that they will overflow at the right time.
Signed-off-by: Eric Hankland <[email protected]>
[jmattson:
- Split the code to increment "branch instructions retired" into a
separate commit.
- Moved/consolidated the calls to kvm_pmu_trigger_event() in the
emulation of VMLAUNCH/VMRESUME to accommodate the evolution of
that code.
] Fixes: f5132b01386b ("KVM: Expose a version 2 architectural PMU to a guests") Signed-off-by: Jim Mattson <[email protected]>
Message-Id: <20211130074221[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
Eric Hankland [Tue, 30 Nov 2021 07:42:20 +0000 (15:42 +0800)]
KVM: x86: Update vPMCs when retiring instructions
When KVM retires a guest instruction through emulation, increment any
vPMCs that are configured to monitor "instructions retired," and
update the sample period of those counters so that they will overflow
at the right time.
Signed-off-by: Eric Hankland <[email protected]>
[jmattson:
- Split the code to increment "branch instructions retired" into a
separate commit.
- Added 'static' to kvm_pmu_incr_counter() definition.
- Modified kvm_pmu_incr_counter() to check pmc->perf_event->state ==
PERF_EVENT_STATE_ACTIVE.
] Fixes: f5132b01386b ("KVM: Expose a version 2 architectural PMU to a guests") Signed-off-by: Jim Mattson <[email protected]>
[likexu:
- Drop checks for pmc->perf_event or event state or event type
- Increase a counter once its umask bits and the first 8 select bits are matched
- Rewrite kvm_pmu_incr_counter() with a less invasive approach to the host perf;
- Rename kvm_pmu_record_event to kvm_pmu_trigger_event;
- Add counter enable and CPL check for kvm_pmu_trigger_event();
] Cc: Peter Zijlstra <[email protected]> Signed-off-by: Like Xu <[email protected]>
Message-Id: <20211130074221[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
Like Xu [Tue, 30 Nov 2021 07:42:19 +0000 (15:42 +0800)]
KVM: x86/pmu: Add pmc->intr to refactor kvm_perf_overflow{_intr}()
Depending on whether intr should be triggered or not, KVM registers
two different event overflow callbacks in the perf_event context.
The code skeleton of these two functions is very similar, so
the pmc->intr can be stored into pmc from pmc_reprogram_counter()
which provides smaller instructions footprint against the
u-architecture branch predictor.
The __kvm_perf_overflow() can be called in non-nmi contexts
and a flag is needed to distinguish the caller context and thus
avoid a check on kvm_is_in_guest(), otherwise we might get
warnings from suspicious RCU or check_preemption_disabled().
Like Xu [Tue, 30 Nov 2021 07:42:18 +0000 (15:42 +0800)]
KVM: x86/pmu: Reuse pmc_perf_hw_id() and drop find_fixed_event()
Since we set the same semantic event value for the fixed counter in
pmc->eventsel, returning the perf_hw_id for the fixed counter via
find_fixed_event() can be painlessly replaced by pmc_perf_hw_id()
with the help of pmc_is_fixed() check.
Like Xu [Tue, 30 Nov 2021 07:42:17 +0000 (15:42 +0800)]
KVM: x86/pmu: Refactoring find_arch_event() to pmc_perf_hw_id()
The find_arch_event() returns a "unsigned int" value,
which is used by the pmc_reprogram_counter() to
program a PERF_TYPE_HARDWARE type perf_event.
The returned value is actually the kernel defined generic
perf_hw_id, let's rename it to pmc_perf_hw_id() with simpler
incoming parameters for better self-explanation.
Like Xu [Tue, 30 Nov 2021 07:42:16 +0000 (15:42 +0800)]
KVM: x86/pmu: Setup pmc->eventsel for fixed PMCs
The current pmc->eventsel for fixed counter is underutilised. The
pmc->eventsel can be setup for all known available fixed counters
since we have mapping between fixed pmc index and
the intel_arch_events array.
Either gp or fixed counter, it will simplify the later checks for
consistency between eventsel and perf_hw_id.
Paolo Bonzini [Thu, 9 Dec 2021 19:10:04 +0000 (14:10 -0500)]
KVM: x86: avoid out of bounds indices for fixed performance counters
Because IceLake has 4 fixed performance counters but KVM only
supports 3, it is possible for reprogram_fixed_counters to pass
to reprogram_fixed_counter an index that is out of bounds for the
fixed_pmc_events array.
Ultimately intel_find_fixed_event, which is the only place that uses
fixed_pmc_events, handles this correctly because it checks against the
size of fixed_pmc_events anyway. Every other place operates on the
fixed_counters[] array which is sized according to INTEL_PMC_MAX_FIXED.
However, it is cleaner if the unsupported performance counters are culled
early on in reprogram_fixed_counters.
Lai Jiangshan [Thu, 16 Dec 2021 02:19:38 +0000 (10:19 +0800)]
KVM: VMX: Mark VCPU_EXREG_CR3 dirty when !CR0_PG -> CR0_PG if EPT + !URG
When !CR0_PG -> CR0_PG, vcpu->arch.cr3 becomes active, but GUEST_CR3 is
still vmx->ept_identity_map_addr if EPT + !URG. So VCPU_EXREG_CR3 is
considered to be dirty and GUEST_CR3 needs to be updated in this case.
Lai Jiangshan [Thu, 16 Dec 2021 02:19:37 +0000 (10:19 +0800)]
KVM: x86/mmu: Reconstruct shadow page root if the guest PDPTEs is changed
For shadow paging, the page table needs to be reconstructed before the
coming VMENTER if the guest PDPTEs is changed.
But not all paths that call load_pdptrs() will cause the page tables to be
reconstructed. Normally, kvm_mmu_reset_context() and kvm_mmu_free_roots()
are used to launch later reconstruction.
The commit d81135a57aa6("KVM: x86: do not reset mmu if CR0.CD and
CR0.NW are changed") skips kvm_mmu_reset_context() after load_pdptrs()
when changing CR0.CD and CR0.NW.
The commit 21823fbda552("KVM: x86: Invalidate all PGDs for the current
PCID on MOV CR3 w/ flush") skips kvm_mmu_free_roots() after
load_pdptrs() when rewriting the CR3 with the same value.
The commit a91a7c709600("KVM: X86: Don't reset mmu context when
toggling X86_CR4_PGE") skips kvm_mmu_reset_context() after
load_pdptrs() when changing CR4.PGE.
Guests like linux would keep the PDPTEs unchanged for every instance of
pagetable, so this missing reconstruction has no problem for linux
guests.
Fixes: d81135a57aa6("KVM: x86: do not reset mmu if CR0.CD and CR0.NW are changed") Fixes: 21823fbda552("KVM: x86: Invalidate all PGDs for the current PCID on MOV CR3 w/ flush") Fixes: a91a7c709600("KVM: X86: Don't reset mmu context when toggling X86_CR4_PGE") Suggested-by: Sean Christopherson <[email protected]> Signed-off-by: Lai Jiangshan <[email protected]>
Message-Id: <20211216021938[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
Lai Jiangshan [Thu, 16 Dec 2021 02:19:36 +0000 (10:19 +0800)]
KVM: VMX: Save HOST_CR3 in vmx_set_host_fs_gs()
The host CR3 in the vcpu thread can only be changed when scheduling,
so commit 15ad9762d69f ("KVM: VMX: Save HOST_CR3 in vmx_prepare_switch_to_guest()")
changed vmx.c to only save it in vmx_prepare_switch_to_guest().
However, it also has to be synced in vmx_sync_vmcs_host_state() when switching VMCS.
vmx_set_host_fs_gs() is called in both places, so rename it to
vmx_set_vmcs_host_state() and make it update HOST_CR3.
"Commit 24cd19a28cb7 ('KVM: X86: Update mmu->pdptrs only when it is
changed') breaks nested VMs with EPT in L0 and PAE shadow paging in L2.
Reproducing is trivial, just disable EPT in L1 and run a VM. I haven't
investigating how it breaks things."
TEST_ASSERT in SEV ioctl was allowing errors because it checked return
value was good OR the FW error code was OK. This TEST_ASSERT should
require both (aka. AND) values are OK. Removes the LAUNCH_START from the
mirror VM because this call correctly fails because mirror VMs cannot
call this command. Currently issues with the PSP driver functions mean
the firmware error is not always reset to SEV_RET_SUCCESS when a call is
successful. Mainly sev_platform_init() doesn't correctly set the fw
error if the platform has already been initialized.
Mirrors should not be able to call LAUNCH_START. Remove the call on the
mirror to correct the test before fixing sev_ioctl() to correctly assert
on this failed ioctl.
Paolo Bonzini [Fri, 7 Jan 2022 15:43:02 +0000 (10:43 -0500)]
Merge tag 'kvm-riscv-5.17-1' of https://github.com/kvm-riscv/linux into HEAD
KVM/riscv changes for 5.17, take #1
- Use common KVM implementation of MMU memory caches
- SBI v0.2 support for Guest
- Initial KVM selftests support
- Fix to avoid spurious virtual interrupts after clearing hideleg CSR
- Update email address for Anup and Atish
Vincent Chen [Mon, 27 Dec 2021 03:05:14 +0000 (11:05 +0800)]
KVM: RISC-V: Avoid spurious virtual interrupts after clearing hideleg CSR
When the last VM is terminated, the host kernel will invoke function
hardware_disable_nolock() on each CPU to disable the related virtualization
functions. Here, RISC-V currently only clears hideleg CSR and hedeleg CSR.
This behavior will cause the host kernel to receive spurious interrupts if
hvip CSR has pending interrupts and the corresponding enable bits in vsie
CSR are asserted. To avoid it, hvip CSR and vsie CSR must be cleared
before clearing hideleg CSR.
Anup Patel [Tue, 5 Oct 2021 12:39:56 +0000 (18:09 +0530)]
KVM: selftests: Add initial support for RISC-V 64-bit
We add initial support for RISC-V 64-bit in KVM selftests using
which we can cross-compile and run arch independent tests such as:
demand_paging_test
dirty_log_test
kvm_create_max_vcpus,
kvm_page_table_test
set_memory_region_test
kvm_binary_stats_test
All VM guest modes defined in kvm_util.h require at least 48-bit
guest virtual address so to use KVM RISC-V selftests hardware
need to support at least Sv48 MMU for guest (i.e. VS-mode).
Anup Patel [Fri, 26 Nov 2021 11:35:51 +0000 (17:05 +0530)]
RISC-V: KVM: Add VM capability to allow userspace get GPA bits
The number of GPA bits supported for a RISC-V Guest/VM is based on the
MMU mode used by the G-stage translation. The KVM RISC-V will detect and
use the best possible MMU mode for the G-stage in kvm_arch_init().
We add a generic VM capability KVM_CAP_VM_GPA_BITS which can be used by
the KVM userspace to get the number of GPA (guest physical address) bits
supported for a Guest/VM.
Anup Patel [Fri, 26 Nov 2021 05:18:41 +0000 (10:48 +0530)]
RISC-V: KVM: Forward SBI experimental and vendor extensions
The SBI experimental extension space is for temporary (or experimental)
stuff whereas SBI vendor extension space is for hardware vendor specific
stuff. Both these SBI extension spaces won't be standardized by the SBI
specification so let's blindly forward such SBI calls to the userspace.
Atish Patra [Thu, 18 Nov 2021 08:39:12 +0000 (00:39 -0800)]
RISC-V: KVM: Add SBI HSM extension in KVM
SBI HSM extension allows OS to start/stop harts any time. It also allows
ordered booting of harts instead of random booting.
Implement SBI HSM exntesion and designate the vcpu 0 as the boot vcpu id.
All other non-zero non-booting vcpus should be brought up by the OS
implementing HSM extension. If the guest OS doesn't implement HSM
extension, only single vcpu will be available to OS.
Atish Patra [Thu, 18 Nov 2021 08:39:10 +0000 (00:39 -0800)]
RISC-V: KVM: Add SBI v0.2 base extension
SBI v0.2 base extension defined to allow backward compatibility and
probing of future extensions. This is also the only mandatory SBI
extension that must be implemented by SBI implementors.
Atish Patra [Thu, 18 Nov 2021 08:39:09 +0000 (00:39 -0800)]
RISC-V: KVM: Reorganize SBI code by moving SBI v0.1 to its own file
With SBI v0.2, there may be more SBI extensions in future. It makes more
sense to group related extensions in separate files. Guest kernel will
choose appropriate SBI version dynamically.
Move the existing implementation to a separate file so that it can be
removed in future without much conflict.
Atish Patra [Thu, 18 Nov 2021 08:39:08 +0000 (00:39 -0800)]
RISC-V: KVM: Mark the existing SBI implementation as v0.1
The existing SBI specification impelementation follows v0.1
specification. The latest specification allows more scalability
and performance improvements.
Rename the existing implementation as v0.1 and provide a way
to allow future extensions.
KVM: RISC-V: Use common KVM implementation of MMU memory caches
Use common KVM's implementation of the MMU memory caches, which for all
intents and purposes is semantically identical to RISC-V's version, the
only difference being that the common implementation will fall back to an
atomic allocation if there's a KVM bug that triggers a cache underflow.
RISC-V appears to have based its MMU code on arm64 before the conversion
to the common caches in commit c1a33aebe91d ("KVM: arm64: Use common KVM
implementation of MMU memory caches"), despite having also copy-pasted
the definition of KVM_ARCH_NR_OBJS_PER_MEMORY_CACHE in kvm_types.h.
Opportunistically drop the superfluous wrapper
kvm_riscv_stage2_flush_cache(), whose name is very, very confusing as
"cache flush" in the context of MMU code almost always refers to flushing
hardware caches, not freeing unused software objects.
Marc Zyngier [Tue, 4 Jan 2022 14:03:43 +0000 (14:03 +0000)]
Merge branch kvm-arm64/selftest/irq-injection into kvmarm-master/next
* kvm-arm64/selftest/irq-injection:
: .
: New tests from Ricardo Koller:
: "This series adds a new test, aarch64/vgic-irq, that validates the injection of
: different types of IRQs from userspace using various methods and configurations"
: .
KVM: selftests: aarch64: Add test for restoring active IRQs
KVM: selftests: aarch64: Add ISPENDR write tests in vgic_irq
KVM: selftests: aarch64: Add tests for IRQFD in vgic_irq
KVM: selftests: Add IRQ GSI routing library functions
KVM: selftests: aarch64: Add test_inject_fail to vgic_irq
KVM: selftests: aarch64: Add tests for LEVEL_INFO in vgic_irq
KVM: selftests: aarch64: Level-sensitive interrupts tests in vgic_irq
KVM: selftests: aarch64: Add preemption tests in vgic_irq
KVM: selftests: aarch64: Cmdline arg to set EOI mode in vgic_irq
KVM: selftests: aarch64: Cmdline arg to set number of IRQs in vgic_irq test
KVM: selftests: aarch64: Abstract the injection functions in vgic_irq
KVM: selftests: aarch64: Add vgic_irq to test userspace IRQ injection
KVM: selftests: aarch64: Add vGIC library functions to deal with vIRQ state
KVM: selftests: Add kvm_irq_line library function
KVM: selftests: aarch64: Add GICv3 register accessor library functions
KVM: selftests: aarch64: Add function for accessing GICv3 dist and redist registers
KVM: selftests: aarch64: Move gic_v3.h to shared headers
Marc Zyngier [Tue, 4 Jan 2022 14:03:26 +0000 (14:03 +0000)]
Merge branch kvm-arm64/selftest/ipa into kvmarm-master/next
* kvm-arm64/selftest/ipa:
: .
: Expand the KVM/arm64 selftest infrastructure to discover
: supported page sizes at runtime, support 16kB pages, and
: find out about the original M1 stupidly small IPA space.
: .
KVM: selftests: arm64: Add support for various modes with 16kB page size
KVM: selftests: arm64: Add support for VM_MODE_P36V48_{4K,64K}
KVM: selftests: arm64: Rework TCR_EL1 configuration
KVM: selftests: arm64: Check for supported page sizes
KVM: selftests: arm64: Introduce a variable default IPA size
KVM: selftests: arm64: Initialise default guest mode at test startup time
Zenghui Yu [Thu, 30 Dec 2021 14:15:35 +0000 (22:15 +0800)]
KVM: arm64: Fix comment typo in kvm_vcpu_finalize_sve()
kvm_arm_init_arch_resources() was renamed to kvm_arm_init_sve() in
commit a3be836df7cb ("KVM: arm/arm64: Demote
kvm_arm_init_arch_resources() to just set up SVE"). Fix the function
name in comment of kvm_vcpu_finalize_sve().
Ricardo Koller [Tue, 9 Nov 2021 02:39:06 +0000 (18:39 -0800)]
KVM: selftests: aarch64: Add test for restoring active IRQs
Add a test that restores multiple IRQs in active state, it does it by
writing into ISACTIVER from the guest and using KVM ioctls. This test
tries to emulate what would happen during a live migration: restore
active IRQs.
Ricardo Koller [Tue, 9 Nov 2021 02:39:00 +0000 (18:39 -0800)]
KVM: selftests: aarch64: Level-sensitive interrupts tests in vgic_irq
Add a cmdline arg for using level-sensitive interrupts (vs the default
edge-triggered). Then move the handler into a generic handler function
that takes the type of interrupt (level vs. edge) as an arg. When
handling line-sensitive interrupts it sets the line to low after
acknowledging the IRQ.
Ricardo Koller [Tue, 9 Nov 2021 02:38:59 +0000 (18:38 -0800)]
KVM: selftests: aarch64: Add preemption tests in vgic_irq
Add tests for IRQ preemption (having more than one activated IRQ at the
same time). This test injects multiple concurrent IRQs and handles them
without handling the actual exceptions. This is done by masking
interrupts for the whole test.
Ricardo Koller [Tue, 9 Nov 2021 02:38:57 +0000 (18:38 -0800)]
KVM: selftests: aarch64: Cmdline arg to set number of IRQs in vgic_irq test
Add the ability to specify the number of vIRQs exposed by KVM (arg
defaults to 64). Then extend the KVM_IRQ_LINE test by injecting all
available SPIs at once (specified by the nr-irqs arg). As a bonus,
inject all SGIs at once as well.
Ricardo Koller [Tue, 9 Nov 2021 02:38:56 +0000 (18:38 -0800)]
KVM: selftests: aarch64: Abstract the injection functions in vgic_irq
Build an abstraction around the injection functions, so the preparation
and checking around the actual injection can be shared between tests.
All functions are stored as pointers in arrays of kvm_inject_desc's
which include the pointer and what kind of interrupts they can inject.
Ricardo Koller [Tue, 9 Nov 2021 02:38:55 +0000 (18:38 -0800)]
KVM: selftests: aarch64: Add vgic_irq to test userspace IRQ injection
Add a new KVM selftest, vgic_irq, for testing userspace IRQ injection. This
particular test injects an SPI using KVM_IRQ_LINE on GICv3 and verifies
that the IRQ is handled in the guest. The next commits will add more
types of IRQs and different modes.
Ricardo Koller [Tue, 9 Nov 2021 02:38:51 +0000 (18:38 -0800)]
KVM: selftests: aarch64: Add function for accessing GICv3 dist and redist registers
Add a generic library function for reading and writing GICv3 distributor
and redistributor registers. Then adapt some functions to use it; more
will come and use it in the next commit.
Marc Zyngier [Mon, 27 Dec 2021 12:48:09 +0000 (12:48 +0000)]
KVM: selftests: arm64: Add support for various modes with 16kB page size
The 16kB page size is not a popular choice, due to only a few CPUs
actually implementing support for it. However, it can lead to some
interesting performance improvements given the right uarch choices.
Add support for this page size for various PA/VA combinations.
Marc Zyngier [Mon, 27 Dec 2021 12:48:08 +0000 (12:48 +0000)]
KVM: selftests: arm64: Add support for VM_MODE_P36V48_{4K,64K}
Some of the arm64 systems out there have an IPA space that is
positively tiny. Nonetheless, they make great KVM hosts.
Add support for 36bit IPA support with 4kB pages, which makes
some of the fruity machines happy. Whilst we're at it, add support
for 64kB pages as well, though these boxes have no support for it.
Marc Zyngier [Mon, 27 Dec 2021 12:48:05 +0000 (12:48 +0000)]
KVM: selftests: arm64: Introduce a variable default IPA size
Contrary to popular belief, there is no such thing as a default
IPA size on arm64. Anything goes, and implementations are the
usual Wild West.
The selftest infrastructure default to 40bit IPA, which obviously
doesn't work for some systems out there.
Turn VM_MODE_DEFAULT from a constant into a variable, and let
guest_modes_append_default() populate it, depending on what
the HW can do. In order to preserve the current behaviour, we
still pick 40bits IPA as the default if it is available, and
the largest supported IPA space otherwise.
Marc Zyngier [Mon, 27 Dec 2021 12:48:04 +0000 (12:48 +0000)]
KVM: selftests: arm64: Initialise default guest mode at test startup time
As we are going to add support for a variable default mode on arm64,
let's make sure it is setup first by using a constructor that gets
called before the actual test runs.
Paolo Bonzini [Tue, 21 Dec 2021 17:51:09 +0000 (12:51 -0500)]
Merge remote-tracking branch 'kvm/master' into HEAD
Pick commit fdba608f15e2 ("KVM: VMX: Wake vCPU when delivering posted
IRQ even if vCPU == this vCPU"). In addition to fixing a bug, it
also aligns the non-nested and nested usage of triggering posted
interrupts, allowing for additional cleanups.
KVM: VMX: Wake vCPU when delivering posted IRQ even if vCPU == this vCPU
Drop a check that guards triggering a posted interrupt on the currently
running vCPU, and more importantly guards waking the target vCPU if
triggering a posted interrupt fails because the vCPU isn't IN_GUEST_MODE.
If a vIRQ is delivered from asynchronous context, the target vCPU can be
the currently running vCPU and can also be blocking, in which case
skipping kvm_vcpu_wake_up() is effectively dropping what is supposed to
be a wake event for the vCPU.
The "do nothing" logic when "vcpu == running_vcpu" mostly works only
because the majority of calls to ->deliver_posted_interrupt(), especially
when using posted interrupts, come from synchronous KVM context. But if
a device is exposed to the guest using vfio-pci passthrough, the VFIO IRQ
and vCPU are bound to the same pCPU, and the IRQ is _not_ configured to
use posted interrupts, wake events from the device will be delivered to
KVM from IRQ context, e.g.
Fuad Tabba [Wed, 8 Dec 2021 19:32:57 +0000 (19:32 +0000)]
KVM: arm64: Fix comment on barrier in kvm_psci_vcpu_on()
The barrier is there for power_off rather than power_state.
Probably typo in commit 358b28f09f0ab074 ("arm/arm64: KVM: Allow
a VCPU to fully reset itself").
Fuad Tabba [Wed, 8 Dec 2021 19:32:56 +0000 (19:32 +0000)]
KVM: arm64: Fix comment for kvm_reset_vcpu()
The comment for kvm_reset_vcpu() refers to the sysreg table as
being the table above, probably because of the code extracted at
commit f4672752c321ea36 ("arm64: KVM: virtual CPU reset").
Fix the comment to remove the potentially confusing reference.
KVM: selftests: Add test to verify TRIPLE_FAULT on invalid L2 guest state
Add a selftest to attempt to enter L2 with invalid guests state by
exiting to userspace via I/O from L2, and then using KVM_SET_SREGS to set
invalid guest state (marking TR unusable is arbitrary chosen for its
relative simplicity).
This is a regression test for a bug introduced by commit c8607e4a086f
("KVM: x86: nVMX: don't fail nested VM entry on invalid guest state if
!from_vmentry"), which incorrectly set vmx->fail=true when L2 had invalid
guest state and ultimately triggered a WARN due to nested_vmx_vmexit()
seeing vmx->fail==true while attempting to synthesize a nested VM-Exit.
The is also a functional test to verify that KVM sythesizes TRIPLE_FAULT
for L2, which is somewhat arbitrary behavior, instead of emulating L2.
KVM should never emulate L2 due to invalid guest state, as it's
architecturally impossible for L1 to run an L2 guest with invalid state
as nested VM-Enter should always fail, i.e. L1 needs to do the emulation.
Stuffing state via KVM ioctl() is a non-architctural, out-of-band case,
hence the TRIPLE_FAULT being rather arbitrary.
KVM: VMX: Fix stale docs for kvm-intel.emulate_invalid_guest_state
Update the documentation for kvm-intel's emulate_invalid_guest_state to
rectify the description of KVM's default behavior, and to document that
the behavior and thus parameter only applies to L1.
KVM: nVMX: Synthesize TRIPLE_FAULT for L2 if emulation is required
Synthesize a triple fault if L2 guest state is invalid at the time of
VM-Enter, which can happen if L1 modifies SMRAM or if userspace stuffs
guest state via ioctls(), e.g. KVM_SET_SREGS. KVM should never emulate
invalid guest state, since from L1's perspective, it's architecturally
impossible for L2 to have invalid state while L2 is running in hardware.
E.g. attempts to set CR0 or CR4 to unsupported values will either VM-Exit
or #GP.
Modifying vCPU state via RSM+SMRAM and ioctl() are the only paths that
can trigger this scenario, as nested VM-Enter correctly rejects any
attempt to enter L2 with invalid state.
RSM is a straightforward case as (a) KVM follows AMD's SMRAM layout and
behavior, and (b) Intel's SDM states that loading reserved CR0/CR4 bits
via RSM results in shutdown, i.e. there is precedent for KVM's behavior.
Following AMD's SMRAM layout is important as AMD's layout saves/restores
the descriptor cache information, including CS.RPL and SS.RPL, and also
defines all the fields relevant to invalid guest state as read-only, i.e.
so long as the vCPU had valid state before the SMI, which is guaranteed
for L2, RSM will generate valid state unless SMRAM was modified. Intel's
layout saves/restores only the selector, which means that scenarios where
the selector and cached RPL don't match, e.g. conforming code segments,
would yield invalid guest state. Intel CPUs fudge around this issued by
stuffing SS.RPL and CS.RPL on RSM. Per Intel's SDM on the "Default
Treatment of RSM", paraphrasing for brevity:
IF internal storage indicates that the [CPU was post-VMXON]
THEN
enter VMX operation (root or non-root);
restore VMX-critical state as defined in Section 34.14.1;
set to their fixed values any bits in CR0 and CR4 whose values must
be fixed in VMX operation [unless coming from an unrestricted guest];
IF RFLAGS.VM = 0 AND (in VMX root operation OR the
“unrestricted guest” VM-execution control is 0)
THEN
CS.RPL := SS.DPL;
SS.RPL := SS.DPL;
FI;
restore current VMCS pointer;
FI;
Note that Intel CPUs also overwrite the fixed CR0/CR4 bits, whereas KVM
will sythesize TRIPLE_FAULT in this scenario. KVM's behavior is allowed
as both Intel and AMD define CR0/CR4 SMRAM fields as read-only, i.e. the
only way for CR0 and/or CR4 to have illegal values is if they were
modified by the L1 SMM handler, and Intel's SDM "SMRAM State Save Map"
section states "modifying these registers will result in unpredictable
behavior".
KVM's ioctl() behavior is less straightforward. Because KVM allows
ioctls() to be executed in any order, rejecting an ioctl() if it would
result in invalid L2 guest state is not an option as KVM cannot know if
a future ioctl() would resolve the invalid state, e.g. KVM_SET_SREGS, or
drop the vCPU out of L2, e.g. KVM_SET_NESTED_STATE. Ideally, KVM would
reject KVM_RUN if L2 contained invalid guest state, but that carries the
risk of a false positive, e.g. if RSM loaded invalid guest state and KVM
exited to userspace. Setting a flag/request to detect such a scenario is
undesirable because (a) it's extremely unlikely to add value to KVM as a
whole, and (b) KVM would need to consider ioctl() interactions with such
a flag, e.g. if userspace migrated the vCPU while the flag were set.
KVM: VMX: Always clear vmx->fail on emulation_required
Revert a relatively recent change that set vmx->fail if the vCPU is in L2
and emulation_required is true, as that behavior is completely bogus.
Setting vmx->fail and synthesizing a VM-Exit is contradictory and wrong:
(a) it's impossible to have both a VM-Fail and VM-Exit
(b) vmcs.EXIT_REASON is not modified on VM-Fail
(c) emulation_required refers to guest state and guest state checks are
always VM-Exits, not VM-Fails.
For KVM specifically, emulation_required is handled before nested exits
in __vmx_handle_exit(), thus setting vmx->fail has no immediate effect,
i.e. KVM calls into handle_invalid_guest_state() and vmx->fail is ignored.
Setting vmx->fail can ultimately result in a WARN in nested_vmx_vmexit()
firing when tearing down the VM as KVM never expects vmx->fail to be set
when L2 is active, KVM always reflects those errors into L1.
Andrew Jones [Tue, 14 Dec 2021 15:18:42 +0000 (16:18 +0100)]
selftests: KVM: Fix non-x86 compiling
Attempting to compile on a non-x86 architecture fails with
include/kvm_util.h: In function ‘vm_compute_max_gfn’:
include/kvm_util.h:79:21: error: dereferencing pointer to incomplete type ‘struct kvm_vm’
return ((1ULL << vm->pa_bits) >> vm->page_shift) - 1;
^~
This is because the declaration of struct kvm_vm is in
lib/kvm_util_internal.h as an effort to make it private to
the test lib code. We can still provide arch specific functions,
though, by making the generic function symbols weak. Do that to
fix the compile error.
Marc Orr [Thu, 9 Dec 2021 15:52:57 +0000 (07:52 -0800)]
KVM: x86: Always set kvm_run->if_flag
The kvm_run struct's if_flag is a part of the userspace/kernel API. The
SEV-ES patches failed to set this flag because it's no longer needed by
QEMU (according to the comment in the source code). However, other
hypervisors may make use of this flag. Therefore, set the flag for
guests with encrypted registers (i.e., with guest_state_protected set).
KVM: x86/mmu: Don't advance iterator after restart due to yielding
After dropping mmu_lock in the TDP MMU, restart the iterator during
tdp_iter_next() and do not advance the iterator. Advancing the iterator
results in skipping the top-level SPTE and all its children, which is
fatal if any of the skipped SPTEs were not visited before yielding.
When zapping all SPTEs, i.e. when min_level == root_level, restarting the
iter and then invoking tdp_iter_next() is always fatal if the current gfn
has as a valid SPTE, as advancing the iterator results in try_step_side()
skipping the current gfn, which wasn't visited before yielding.
Sprinkle WARNs on iter->yielded being true in various helpers that are
often used in conjunction with yielding, and tag the helper with
__must_check to reduce the probabily of improper usage.
Failing to zap a top-level SPTE manifests in one of two ways. If a valid
SPTE is skipped by both kvm_tdp_mmu_zap_all() and kvm_tdp_mmu_put_root(),
the shadow page will be leaked and KVM will WARN accordingly.
If kvm_tdp_mmu_zap_all() skips a gfn/SPTE but that SPTE is then zapped by
kvm_tdp_mmu_put_root(), KVM triggers a use-after-free in the form of
marking a struct page as dirty/accessed after it has been put back on the
free list. This directly triggers a WARN due to encountering a page with
page_count() == 0, but it can also lead to data corruption and additional
errors in the kernel.
Note, the underlying bug existed even before commit 1af4a96025b3 ("KVM:
x86/mmu: Yield in TDU MMU iter even if no SPTES changed") moved calls to
tdp_mmu_iter_cond_resched() to the beginning of loops, as KVM could still
incorrectly advance past a top-level entry when yielding on a lower-level
entry. But with respect to leaking shadow pages, the bug was introduced
by yielding before processing the current gfn.
Alternatively, tdp_mmu_iter_cond_resched() could simply fall through, or
callers could jump to their "retry" label. The downside of that approach
is that tdp_mmu_iter_cond_resched() _must_ be called before anything else
in the loop, and there's no easy way to enfornce that requirement.
Ideally, KVM would handling the cond_resched() fully within the iterator
macro (the code is actually quite clean) and avoid this entire class of
bugs, but that is extremely difficult do while also supporting yielding
after tdp_mmu_set_spte_atomic() fails. Yielding after failing to set a
SPTE is very desirable as the "owner" of the REMOVED_SPTE isn't strictly
bounded, e.g. if it's zapping a high-level shadow page, the REMOVED_SPTE
may block operations on the SPTE for a significant amount of time.
Wei Wang [Fri, 17 Dec 2021 12:49:34 +0000 (07:49 -0500)]
KVM: x86: remove PMU FIXED_CTR3 from msrs_to_save_all
The fixed counter 3 is used for the Topdown metrics, which hasn't been
enabled for KVM guests. Userspace accessing to it will fail as it's not
included in get_fixed_pmc(). This breaks KVM selftests on ICX+ machines,
which have this counter.
To reproduce it on ICX+ machines, ./state_test reports:
==== Test Assertion Failure ====
lib/x86_64/processor.c:1078: r == nmsrs
pid=4564 tid=4564 - Argument list too long
1 0x000000000040b1b9: vcpu_save_state at processor.c:1077
2 0x0000000000402478: main at state_test.c:209 (discriminator 6)
3 0x00007fbe21ed5f92: ?? ??:0
4 0x000000000040264d: _start at ??:?
Unexpected result from KVM_GET_MSRS, r: 17 (failed MSR was 0x30c)
KVM: x86: Retry page fault if MMU reload is pending and root has no sp
Play nice with a NULL shadow page when checking for an obsolete root in
the page fault handler by flagging the page fault as stale if there's no
shadow page associated with the root and KVM_REQ_MMU_RELOAD is pending.
Invalidating memslots, which is the only case where _all_ roots need to
be reloaded, requests all vCPUs to reload their MMUs while holding
mmu_lock for lock.
The "special" roots, e.g. pae_root when KVM uses PAE paging, are not
backed by a shadow page. Running with TDP disabled or with nested NPT
explodes spectaculary due to dereferencing a NULL shadow page pointer.
Skip the KVM_REQ_MMU_RELOAD check if there is a valid shadow page for the
root. Zapping shadow pages in response to guest activity, e.g. when the
guest frees a PGD, can trigger KVM_REQ_MMU_RELOAD even if the current
vCPU isn't using the affected root. I.e. KVM_REQ_MMU_RELOAD can be seen
with a completely valid root shadow page. This is a bit of a moot point
as KVM currently unloads all roots on KVM_REQ_MMU_RELOAD, but that will
be cleaned up in the future.
Vitaly Kuznetsov [Thu, 16 Dec 2021 16:52:12 +0000 (17:52 +0100)]
KVM: selftests: vmx_pmu_msrs_test: Drop tests mangling guest visible CPUIDs
Host initiated writes to MSR_IA32_PERF_CAPABILITIES should not depend
on guest visible CPUIDs and (incorrect) KVM logic implementing it is
about to change. Also, KVM_SET_CPUID{,2} after KVM_RUN is now forbidden
and causes test to fail.
Vitaly Kuznetsov [Thu, 16 Dec 2021 16:52:13 +0000 (17:52 +0100)]
KVM: x86: Drop guest CPUID check for host initiated writes to MSR_IA32_PERF_CAPABILITIES
The ability to write to MSR_IA32_PERF_CAPABILITIES from the host should
not depend on guest visible CPUID entries, even if just to allow
creating/restoring guest MSRs and CPUIDs in any sequence.
Paolo Bonzini [Sun, 19 Dec 2021 14:27:21 +0000 (15:27 +0100)]
Merge branch 'topic/ppc-kvm' of https://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux into HEAD
Fix conflicts between memslot overhaul and commit 511d25d6b789f ("KVM:
PPC: Book3S: Suppress warnings when allocating too big memory slots")
from the powerpc tree.
Eric Farman [Mon, 13 Dec 2021 21:05:50 +0000 (22:05 +0100)]
KVM: s390: Clarify SIGP orders versus STOP/RESTART
With KVM_CAP_S390_USER_SIGP, there are only five Signal Processor
orders (CONDITIONAL EMERGENCY SIGNAL, EMERGENCY SIGNAL, EXTERNAL CALL,
SENSE, and SENSE RUNNING STATUS) which are intended for frequent use
and thus are processed in-kernel. The remainder are sent to userspace
with the KVM_CAP_S390_USER_SIGP capability. Of those, three orders
(RESTART, STOP, and STOP AND STORE STATUS) have the potential to
inject work back into the kernel, and thus are asynchronous.
Let's look for those pending IRQs when processing one of the in-kernel
SIGP orders, and return BUSY (CC2) if one is in process. This is in
agreement with the Principles of Operation, which states that only one
order can be "active" on a CPU at a time.
Janosch Frank [Mon, 7 Jun 2021 08:07:13 +0000 (08:07 +0000)]
s390: uv: Add offset comments to UV query struct and fix naming
Changes to the struct are easier to manage with offset comments so
let's add some. And now that we know that the last struct member has
the wrong name let's also fix this.
KVM: s390: gaccess: Refactor access address range check
Do not round down the first address to the page boundary, just translate
it normally, which gives the value we care about in the first place.
Given this, translating a single address is just the special case of
translating a range spanning a single page.
Make the output optional, so the function can be used to just check a
range.
Marc Zyngier [Mon, 29 Nov 2021 20:00:45 +0000 (20:00 +0000)]
KVM: arm64: Rework kvm_pgtable initialisation
Ganapatrao reported that the kvm_pgtable->mmu pointer is more or
less hardcoded to the main S2 mmu structure, while the nested
code needs it to point to other instances (as we have one instance
per nested context).
Rework the initialisation of the kvm_pgtable structure so that
this assumtion doesn't hold true anymore. This requires some
minor changes to the order in which things are initialised
(the mmu->arch pointer being the critical one).
Marc Zyngier [Thu, 16 Dec 2021 13:06:09 +0000 (13:06 +0000)]
Merge branch kvm-arm64/pkvm-hyp-sharing into kvmarm-master/next
* kvm-arm64/pkvm-hyp-sharing:
: .
: Series from Quentin Perret, implementing HYP page share/unshare:
:
: This series implements an unshare hypercall at EL2 in nVHE
: protected mode, and makes use of it to unmmap guest-specific
: data-structures from EL2 stage-1 during guest tear-down.
: Crucially, the implementation of the share and unshare
: routines use page refcounts in the host kernel to avoid
: accidentally unmapping data-structures that overlap a common
: page.
: [...]
: .
KVM: arm64: pkvm: Unshare guest structs during teardown
KVM: arm64: Expose unshare hypercall to the host
KVM: arm64: Implement do_unshare() helper for unsharing memory
KVM: arm64: Implement __pkvm_host_share_hyp() using do_share()
KVM: arm64: Implement do_share() helper for sharing memory
KVM: arm64: Introduce wrappers for host and hyp spin lock accessors
KVM: arm64: Extend pkvm_page_state enumeration to handle absent pages
KVM: arm64: pkvm: Refcount the pages shared with EL2
KVM: arm64: Introduce kvm_share_hyp()
KVM: arm64: Implement kvm_pgtable_hyp_unmap() at EL2
KVM: arm64: Hook up ->page_count() for hypervisor stage-1 page-table
KVM: arm64: Fixup hyp stage-1 refcount
KVM: arm64: Refcount hyp stage-1 pgtable pages
KVM: arm64: Provide {get,put}_page() stubs for early hyp allocator
Will Deacon [Wed, 15 Dec 2021 16:12:30 +0000 (16:12 +0000)]
KVM: arm64: Expose unshare hypercall to the host
Introduce an unshare hypercall which can be used to unmap memory from
the hypervisor stage-1 in nVHE protected mode. This will be useful to
update the EL2 ownership state of pages during guest teardown, and
avoids keeping dangling mappings to unreferenced portions of memory.
Will Deacon [Wed, 15 Dec 2021 16:12:29 +0000 (16:12 +0000)]
KVM: arm64: Implement do_unshare() helper for unsharing memory
Tearing down a previously shared memory region results in the borrower
losing access to the underlying pages and returning them to the "owned"
state in the owner.
Implement a do_unshare() helper, along the same lines as do_share(), to
provide this functionality for the host-to-hyp case.
Will Deacon [Wed, 15 Dec 2021 16:12:28 +0000 (16:12 +0000)]
KVM: arm64: Implement __pkvm_host_share_hyp() using do_share()
__pkvm_host_share_hyp() shares memory between the host and the
hypervisor so implement it as an invocation of the new do_share()
mechanism.
Note that double-sharing is no longer permitted (as this allows us to
reduce the number of page-table walks significantly), but is thankfully
no longer relied upon by the host.
Will Deacon [Wed, 15 Dec 2021 16:12:27 +0000 (16:12 +0000)]
KVM: arm64: Implement do_share() helper for sharing memory
By default, protected KVM isolates memory pages so that they are
accessible only to their owner: be it the host kernel, the hypervisor
at EL2 or (in future) the guest. Establishing shared-memory regions
between these components therefore involves a transition for each page
so that the owner can share memory with a borrower under a certain set
of permissions.
Introduce a do_share() helper for safely sharing a memory region between
two components. Currently, only host-to-hyp sharing is implemented, but
the code is easily extended to handle other combinations and the
permission checks for each component are reusable.
Will Deacon [Wed, 15 Dec 2021 16:12:26 +0000 (16:12 +0000)]
KVM: arm64: Introduce wrappers for host and hyp spin lock accessors
In preparation for adding additional locked sections for manipulating
page-tables at EL2, introduce some simple wrappers around the host and
hypervisor locks so that it's a bit easier to read and bit more difficult
to take the wrong lock (or even take them in the wrong order).
Will Deacon [Wed, 15 Dec 2021 16:12:25 +0000 (16:12 +0000)]
KVM: arm64: Extend pkvm_page_state enumeration to handle absent pages
Explicitly name the combination of SW0 | SW1 as reserved in the pte and
introduce a new PKVM_NOPAGE meta-state which, although not directly
stored in the software bits of the pte, can be used to represent an
entry for which there is no underlying page. This is distinct from an
invalid pte, as stage-2 identity mappings for the host are created
lazily and so an invalid pte there is the same as a valid mapping for
the purposes of ownership information.
This state will be used for permission checking during page transitions
in later patches.
Quentin Perret [Wed, 15 Dec 2021 16:12:24 +0000 (16:12 +0000)]
KVM: arm64: pkvm: Refcount the pages shared with EL2
In order to simplify the page tracking infrastructure at EL2 in nVHE
protected mode, move the responsibility of refcounting pages that are
shared multiple times on the host. In order to do so, let's create a
red-black tree tracking all the PFNs that have been shared, along with
a refcount.
Quentin Perret [Wed, 15 Dec 2021 16:12:23 +0000 (16:12 +0000)]
KVM: arm64: Introduce kvm_share_hyp()
The create_hyp_mappings() function can currently be called at any point
in time. However, its behaviour in protected mode changes widely
depending on when it is being called. Prior to KVM init, it is used to
create the temporary page-table used to bring-up the hypervisor, and
later on it is transparently turned into a 'share' hypercall when the
kernel has lost control over the hypervisor stage-1. In order to prepare
the ground for also unsharing pages with the hypervisor during guest
teardown, introduce a kvm_share_hyp() function to make it clear in which
places a share hypercall should be expected, as we will soon need a
matching unshare hypercall in all those places.
projects / linux.git / log