| Age | Commit message (Collapse) | Author |
|---|
|
setup_vmcs_config()"
Revert back to clearing VM_{ENTRY,EXIT}_LOAD_IA32_PERF_GLOBAL_CTRL in KVM's
golden VMCS config, as applying the workaround during vCPU creation is
pointless and broken. KVM *unconditionally* clears the controls in the
values returned by vmx_vmentry_ctrl() and vmx_vmexit_ctrl(), as KVM loads
PERF_GLOBAL_CTRL if and only if its necessary to do so. E.g. if KVM wants
to run the guest with the same PERF_GLOBAL_CTRL as the host, then there's
no need to re-load the MSR on entry and exit.
Even worse, the buggy commit failed to apply the erratum where it's
actually needed, add_atomic_switch_msr(). As a result, KVM completely
ignores the erratum for all intents and purposes, i.e. uses the flawed
VMCS controls to load PERF_GLOBAL_CTRL.
To top things off, the patch was intended to be dropped, as the premise
of an L1 VMM being able to pivot on FMS is flawed, and KVM can (and now
does) fully emulate the controls in software. Simply revert the commit,
as all upstream supported kernels that have the buggy commit should also
have commit f4c93d1a0e71 ("KVM: nVMX: Always emulate PERF_GLOBAL_CTRL
VM-Entry/VM-Exit controls"), i.e. the (likely theoretical) live migration
concern is a complete non-issue.
Opportunistically drop the manual "kvm: " scope from the warning about
the erratum, as KVM now uses pr_fmt() to provide the correct scope (v6.1
kernels and earlier don't, but the erratum only applies to CPUs that are
15+ years old; it's not worth a separate patch).
This reverts commit 9d78d6fb186bc4aff41b5d6c4726b76649d3cb53.
Link: https://lore.kernel.org/all/YtnZmCutdd5tpUmz@google.com
Fixes: 9d78d6fb186b ("KVM: VMX: Move LOAD_IA32_PERF_GLOBAL_CTRL errata handling out of setup_vmcs_config()")
Cc: stable@vger.kernel.org
Cc: Vitaly Kuznetsov <vkuznets@redhat.com>
Cc: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Message-ID: <20241119011433.1797921-1-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
- Drop obsolete references to PPC970 KVM, which was removed 10 years ago.
- Fix incorrect references to non-existing ioctls
- List registers supported by KVM_GET/SET_ONE_REG on s390
- Use rST internal links
- Reorganize the introduction to the API document
|
|
KVM x86 misc changes for 6.13
- Clean up and optimize KVM's handling of writes to MSR_IA32_APICBASE.
- Quirk KVM's misguided behavior of initialized certain feature MSRs to
their maximum supported feature set, which can result in KVM creating
invalid vCPU state. E.g. initializing PERF_CAPABILITIES to a non-zero
value results in the vCPU having invalid state if userspace hides PDCM
from the guest, which can lead to save/restore failures.
- Fix KVM's handling of non-canonical checks for vCPUs that support LA57
to better follow the "architecture", in quotes because the actual
behavior is poorly documented. E.g. most MSR writes and descriptor
table loads ignore CR4.LA57 and operate purely on whether the CPU
supports LA57.
- Bypass the register cache when querying CPL from kvm_sched_out(), as
filling the cache from IRQ context is generally unsafe, and harden the
cache accessors to try to prevent similar issues from occuring in the
future.
- Advertise AMD_IBPB_RET to userspace, and fix a related bug where KVM
over-advertises SPEC_CTRL when trying to support cross-vendor VMs.
- Minor cleanups
|
|
Hide KVM's pt_mode module param behind CONFIG_BROKEN, i.e. disable support
for virtualizing Intel PT via guest/host mode unless BROKEN=y. There are
myriad bugs in the implementation, some of which are fatal to the guest,
and others which put the stability and health of the host at risk.
For guest fatalities, the most glaring issue is that KVM fails to ensure
tracing is disabled, and *stays* disabled prior to VM-Enter, which is
necessary as hardware disallows loading (the guest's) RTIT_CTL if tracing
is enabled (enforced via a VMX consistency check). Per the SDM:
If the logical processor is operating with Intel PT enabled (if
IA32_RTIT_CTL.TraceEn = 1) at the time of VM entry, the "load
IA32_RTIT_CTL" VM-entry control must be 0.
On the host side, KVM doesn't validate the guest CPUID configuration
provided by userspace, and even worse, uses the guest configuration to
decide what MSRs to save/load at VM-Enter and VM-Exit. E.g. configuring
guest CPUID to enumerate more address ranges than are supported in hardware
will result in KVM trying to passthrough, save, and load non-existent MSRs,
which generates a variety of WARNs, ToPA ERRORs in the host, a potential
deadlock, etc.
Fixes: f99e3daf94ff ("KVM: x86: Add Intel PT virtualization work mode")
Cc: stable@vger.kernel.org
Cc: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Xiaoyao Li <xiaoyao.li@intel.com>
Tested-by: Adrian Hunter <adrian.hunter@intel.com>
Message-ID: <20241101185031.1799556-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
When getting the current VPID, e.g. to emulate a guest TLB flush, return
vpid01 if L2 is running but with VPID disabled, i.e. if VPID is disabled
in vmcs12. Architecturally, if VPID is disabled, then the guest and host
effectively share VPID=0. KVM emulates this behavior by using vpid01 when
running an L2 with VPID disabled (see prepare_vmcs02_early_rare()), and so
KVM must also treat vpid01 as the current VPID while L2 is active.
Unconditionally treating vpid02 as the current VPID when L2 is active
causes KVM to flush TLB entries for vpid02 instead of vpid01, which
results in TLB entries from L1 being incorrectly preserved across nested
VM-Enter to L2 (L2=>L1 isn't problematic, because the TLB flush after
nested VM-Exit flushes vpid01).
The bug manifests as failures in the vmx_apicv_test KVM-Unit-Test, as KVM
incorrectly retains TLB entries for the APIC-access page across a nested
VM-Enter.
Opportunisticaly add comments at various touchpoints to explain the
architectural requirements, and also why KVM uses vpid01 instead of vpid02.
All credit goes to Chao, who root caused the issue and identified the fix.
Link: https://lore.kernel.org/all/ZwzczkIlYGX+QXJz@intel.com
Fixes: 2b4a5a5d5688 ("KVM: nVMX: Flush current VPID (L1 vs. L2) for KVM_REQ_TLB_FLUSH_GUEST")
Cc: stable@vger.kernel.org
Cc: Like Xu <like.xu.linux@gmail.com>
Debugged-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Tested-by: Chao Gao <chao.gao@intel.com>
Link: https://lore.kernel.org/r/20241031202011.1580522-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Drop the restriction that the PMU version is non-zero when handling writes
to PERF_CAPABILITIES now that KVM unconditionally checks for PDCM support.
Link: https://lore.kernel.org/r/20240802185511.305849-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Add a quirk to control KVM's misguided initialization of select feature
MSRs to KVM's max configuration, as enabling features by default violates
KVM's approach of letting userspace own the vCPU model, and is actively
problematic for MSRs that are conditionally supported, as the vCPU will
end up with an MSR value that userspace can't restore. E.g. if the vCPU
is configured with PDCM=0, userspace will save and attempt to restore a
non-zero PERF_CAPABILITIES, thanks to KVM's meddling.
Link: https://lore.kernel.org/r/20240802185511.305849-4-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
As a result of a recent investigation, it was determined that x86 CPUs
which support 5-level paging, don't always respect CR4.LA57 when doing
canonical checks.
In particular:
1. MSRs which contain a linear address, allow full 57-bitcanonical address
regardless of CR4.LA57 state. For example: MSR_KERNEL_GS_BASE.
2. All hidden segment bases and GDT/IDT bases also behave like MSRs.
This means that full 57-bit canonical address can be loaded to them
regardless of CR4.LA57, both using MSRS (e.g GS_BASE) and instructions
(e.g LGDT).
3. TLB invalidation instructions also allow the user to use full 57-bit
address regardless of the CR4.LA57.
Finally, it must be noted that the CPU doesn't prevent the user from
disabling 5-level paging, even when the full 57-bit canonical address is
present in one of the registers mentioned above (e.g GDT base).
In fact, this can happen without any userspace help, when the CPU enters
SMM mode - some MSRs, for example MSR_KERNEL_GS_BASE are left to contain
a non-canonical address in regard to the new mode.
Since most of the affected MSRs and all segment bases can be read and
written freely by the guest without any KVM intervention, this patch makes
the emulator closely follow hardware behavior, which means that the
emulator doesn't take in the account the guest CPUID support for 5-level
paging, and only takes in the account the host CPU support.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20240906221824.491834-4-mlevitsk@redhat.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
When querying guest CPL to determine if a vCPU was preempted while in
kernel mode, bypass the register cache, i.e. always read SS.AR_BYTES from
the VMCS on Intel CPUs. If the kernel is running with full preemption
enabled, using the register cache in the preemption path can result in
stale and/or uninitialized data being cached in the segment cache.
In particular the following scenario is currently possible:
- vCPU is just created, and the vCPU thread is preempted before
SS.AR_BYTES is written in vmx_vcpu_reset().
- When scheduling out the vCPU task, kvm_arch_vcpu_in_kernel() =>
vmx_get_cpl() reads and caches '0' for SS.AR_BYTES.
- vmx_vcpu_reset() => seg_setup() configures SS.AR_BYTES, but doesn't
invoke vmx_segment_cache_clear() to invalidate the cache.
As a result, KVM retains a stale value in the cache, which can be read,
e.g. via KVM_GET_SREGS. Usually this is not a problem because the VMX
segment cache is reset on each VM-Exit, but if the userspace VMM (e.g KVM
selftests) reads and writes system registers just after the vCPU was
created, _without_ modifying SS.AR_BYTES, userspace will write back the
stale '0' value and ultimately will trigger a VM-Entry failure due to
incorrect SS segment type.
Note, the VM-Enter failure can also be avoided by moving the call to
vmx_segment_cache_clear() until after the vmx_vcpu_reset() initializes all
segments. However, while that change is correct and desirable (and will
come along shortly), it does not address the underlying problem that
accessing KVM's register caches from !task context is generally unsafe.
In addition to fixing the immediate bug, bypassing the cache for this
particular case will allow hardening KVM register caching log to assert
that the caches are accessed only when KVM _knows_ it is safe to do so.
Fixes: de63ad4cf497 ("KVM: X86: implement the logic for spinlock optimization")
Reported-by: Maxim Levitsky <mlevitsk@redhat.com>
Closes: https://lore.kernel.org/all/20240716022014.240960-3-mlevitsk@redhat.com
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20241009175002.1118178-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Remove the unused variable "gpa" in __invept().
The INVEPT instruction only supports two types: VMX_EPT_EXTENT_CONTEXT (1)
and VMX_EPT_EXTENT_GLOBAL (2). Neither of these types requires a third
variable "gpa".
The "gpa" variable for __invept() is always set to 0 and was originally
introduced for the old non-existent type VMX_EPT_EXTENT_INDIVIDUAL_ADDR
(0). This type was removed by commit 2b3c5cbc0d81 ("kvm: don't use bit24
for detecting address-specific invalidation capability") and
commit 63f3ac48133a ("KVM: VMX: clean up declaration of VPID/EPT
invalidation types").
Since this variable is not useful for error handling either, remove it to
avoid confusion.
No functional changes expected.
Cc: Yuan Yao <yuan.yao@intel.com>
Signed-off-by: Yan Zhao <yan.y.zhao@intel.com>
Link: https://lore.kernel.org/r/20241014045931.1061-1-yan.y.zhao@intel.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Use __kvm_faultin_page() get the APIC access page so that KVM can
precisely release the refcounted page, i.e. to remove yet another user
of kvm_pfn_to_refcounted_page(). While the path isn't handling a guest
page fault, the semantics are effectively the same; KVM just happens to
be mapping the pfn into a VMCS field instead of a secondary MMU.
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <20241010182427.1434605-52-seanjc@google.com>
|
|
Hold mmu_lock across kvm_release_pfn_clean() when refreshing the APIC
access page address to ensure that KVM doesn't mark a page/folio as
accessed after it has been unmapped. Practically speaking marking a folio
accesses is benign in this scenario, as KVM does hold a reference (it's
really just marking folios dirty that is problematic), but there's no
reason not to be paranoid (moving the APIC access page isn't a hot path),
and no reason to be different from other mmu_notifier-protected flows in
KVM.
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <20241010182427.1434605-51-seanjc@google.com>
|
|
Reset the segment cache after segment initialization in vmx_vcpu_reset()
to harden KVM against caching stale/uninitialized data. Without the
recent fix to bypass the cache in kvm_arch_vcpu_put(), the following
scenario is possible:
- vCPU is just created, and the vCPU thread is preempted before
SS.AR_BYTES is written in vmx_vcpu_reset().
- When scheduling out the vCPU task, kvm_arch_vcpu_in_kernel() =>
vmx_get_cpl() reads and caches '0' for SS.AR_BYTES.
- vmx_vcpu_reset() => seg_setup() configures SS.AR_BYTES, but doesn't
invoke vmx_segment_cache_clear() to invalidate the cache.
As a result, KVM retains a stale value in the cache, which can be read,
e.g. via KVM_GET_SREGS. Usually this is not a problem because the VMX
segment cache is reset on each VM-Exit, but if the userspace VMM (e.g KVM
selftests) reads and writes system registers just after the vCPU was
created, _without_ modifying SS.AR_BYTES, userspace will write back the
stale '0' value and ultimately will trigger a VM-Entry failure due to
incorrect SS segment type.
Invalidating the cache after writing the VMCS doesn't address the general
issue of cache accesses from IRQ context being unsafe, but it does prevent
KVM from clobbering the VMCS, i.e. mitigates the harm done _if_ KVM has a
bug that results in an unsafe cache access.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Fixes: 2fb92db1ec08 ("KVM: VMX: Cache vmcs segment fields")
[sean: rework changelog to account for previous patch]
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20241009175002.1118178-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Pull x86 kvm updates from Paolo Bonzini:
"x86:
- KVM currently invalidates the entirety of the page tables, not just
those for the memslot being touched, when a memslot is moved or
deleted.
This does not traditionally have particularly noticeable overhead,
but Intel's TDX will require the guest to re-accept private pages
if they are dropped from the secure EPT, which is a non starter.
Actually, the only reason why this is not already being done is a
bug which was never fully investigated and caused VM instability
with assigned GeForce GPUs, so allow userspace to opt into the new
behavior.
- Advertise AVX10.1 to userspace (effectively prep work for the
"real" AVX10 functionality that is on the horizon)
- Rework common MSR handling code to suppress errors on userspace
accesses to unsupported-but-advertised MSRs
This will allow removing (almost?) all of KVM's exemptions for
userspace access to MSRs that shouldn't exist based on the vCPU
model (the actual cleanup is non-trivial future work)
- Rework KVM's handling of x2APIC ICR, again, because AMD (x2AVIC)
splits the 64-bit value into the legacy ICR and ICR2 storage,
whereas Intel (APICv) stores the entire 64-bit value at the ICR
offset
- Fix a bug where KVM would fail to exit to userspace if one was
triggered by a fastpath exit handler
- Add fastpath handling of HLT VM-Exit to expedite re-entering the
guest when there's already a pending wake event at the time of the
exit
- Fix a WARN caused by RSM entering a nested guest from SMM with
invalid guest state, by forcing the vCPU out of guest mode prior to
signalling SHUTDOWN (the SHUTDOWN hits the VM altogether, not the
nested guest)
- Overhaul the "unprotect and retry" logic to more precisely identify
cases where retrying is actually helpful, and to harden all retry
paths against putting the guest into an infinite retry loop
- Add support for yielding, e.g. to honor NEED_RESCHED, when zapping
rmaps in the shadow MMU
- Refactor pieces of the shadow MMU related to aging SPTEs in
prepartion for adding multi generation LRU support in KVM
- Don't stuff the RSB after VM-Exit when RETPOLINE=y and AutoIBRS is
enabled, i.e. when the CPU has already flushed the RSB
- Trace the per-CPU host save area as a VMCB pointer to improve
readability and cleanup the retrieval of the SEV-ES host save area
- Remove unnecessary accounting of temporary nested VMCB related
allocations
- Set FINAL/PAGE in the page fault error code for EPT violations if
and only if the GVA is valid. If the GVA is NOT valid, there is no
guest-side page table walk and so stuffing paging related metadata
is nonsensical
- Fix a bug where KVM would incorrectly synthesize a nested VM-Exit
instead of emulating posted interrupt delivery to L2
- Add a lockdep assertion to detect unsafe accesses of vmcs12
structures
- Harden eVMCS loading against an impossible NULL pointer deref
(really truly should be impossible)
- Minor SGX fix and a cleanup
- Misc cleanups
Generic:
- Register KVM's cpuhp and syscore callbacks when enabling
virtualization in hardware, as the sole purpose of said callbacks
is to disable and re-enable virtualization as needed
- Enable virtualization when KVM is loaded, not right before the
first VM is created
Together with the previous change, this simplifies a lot the logic
of the callbacks, because their very existence implies
virtualization is enabled
- Fix a bug that results in KVM prematurely exiting to userspace for
coalesced MMIO/PIO in many cases, clean up the related code, and
add a testcase
- Fix a bug in kvm_clear_guest() where it would trigger a buffer
overflow _if_ the gpa+len crosses a page boundary, which thankfully
is guaranteed to not happen in the current code base. Add WARNs in
more helpers that read/write guest memory to detect similar bugs
Selftests:
- Fix a goof that caused some Hyper-V tests to be skipped when run on
bare metal, i.e. NOT in a VM
- Add a regression test for KVM's handling of SHUTDOWN for an SEV-ES
guest
- Explicitly include one-off assets in .gitignore. Past Sean was
completely wrong about not being able to detect missing .gitignore
entries
- Verify userspace single-stepping works when KVM happens to handle a
VM-Exit in its fastpath
- Misc cleanups"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (127 commits)
Documentation: KVM: fix warning in "make htmldocs"
s390: Enable KVM_S390_UCONTROL config in debug_defconfig
selftests: kvm: s390: Add VM run test case
KVM: SVM: let alternatives handle the cases when RSB filling is required
KVM: VMX: Set PFERR_GUEST_{FINAL,PAGE}_MASK if and only if the GVA is valid
KVM: x86/mmu: Use KVM_PAGES_PER_HPAGE() instead of an open coded equivalent
KVM: x86/mmu: Add KVM_RMAP_MANY to replace open coded '1' and '1ul' literals
KVM: x86/mmu: Fold mmu_spte_age() into kvm_rmap_age_gfn_range()
KVM: x86/mmu: Morph kvm_handle_gfn_range() into an aging specific helper
KVM: x86/mmu: Honor NEED_RESCHED when zapping rmaps and blocking is allowed
KVM: x86/mmu: Add a helper to walk and zap rmaps for a memslot
KVM: x86/mmu: Plumb a @can_yield parameter into __walk_slot_rmaps()
KVM: x86/mmu: Move walk_slot_rmaps() up near for_each_slot_rmap_range()
KVM: x86/mmu: WARN on MMIO cache hit when emulating write-protected gfn
KVM: x86/mmu: Detect if unprotect will do anything based on invalid_list
KVM: x86/mmu: Subsume kvm_mmu_unprotect_page() into the and_retry() version
KVM: x86: Rename reexecute_instruction()=>kvm_unprotect_and_retry_on_failure()
KVM: x86: Update retry protection fields when forcing retry on emulation failure
KVM: x86: Apply retry protection to "unprotect on failure" path
KVM: x86: Check EMULTYPE_WRITE_PF_TO_SP before unprotecting gfn
...
|
|
KVM VMX changes for 6.12:
- Set FINAL/PAGE in the page fault error code for EPT Violations if and only
if the GVA is valid. If the GVA is NOT valid, there is no guest-side page
table walk and so stuffing paging related metadata is nonsensical.
- Fix a bug where KVM would incorrectly synthesize a nested VM-Exit instead of
emulating posted interrupt delivery to L2.
- Add a lockdep assertion to detect unsafe accesses of vmcs12 structures.
- Harden eVMCS loading against an impossible NULL pointer deref (really truly
should be impossible).
- Minor SGX fix and a cleanup.
|
|
into HEAD
KVM VMX and x86 PAT MSR macro cleanup for 6.12:
- Add common defines for the x86 architectural memory types, i.e. the types
that are shared across PAT, MTRRs, VMCSes, and EPTPs.
- Clean up the various VMX MSR macros to make the code self-documenting
(inasmuch as possible), and to make it less painful to add new macros.
|
|
KVM x86 misc changes for 6.12
- Advertise AVX10.1 to userspace (effectively prep work for the "real" AVX10
functionality that is on the horizon).
- Rework common MSR handling code to suppress errors on userspace accesses to
unsupported-but-advertised MSRs. This will allow removing (almost?) all of
KVM's exemptions for userspace access to MSRs that shouldn't exist based on
the vCPU model (the actual cleanup is non-trivial future work).
- Rework KVM's handling of x2APIC ICR, again, because AMD (x2AVIC) splits the
64-bit value into the legacy ICR and ICR2 storage, whereas Intel (APICv)
stores the entire 64-bit value a the ICR offset.
- Fix a bug where KVM would fail to exit to userspace if one was triggered by
a fastpath exit handler.
- Add fastpath handling of HLT VM-Exit to expedite re-entering the guest when
there's already a pending wake event at the time of the exit.
- Finally fix the RSM vs. nested VM-Enter WARN by forcing the vCPU out of
guest mode prior to signalling SHUTDOWN (architecturally, the SHUTDOWN is
supposed to hit L1, not L2).
|
|
This reverts commit 377b2f359d1f71c75f8cc352b5c81f2210312d83.
This caused a regression with the bochsdrm driver, which used ioremap()
instead of ioremap_wc() to map the video RAM. After the commit, the
WB memory type is used without the IGNORE_PAT, resulting in the slower
UC memory type. In fact, UC is slow enough to basically cause guests
to not boot... but only on new processors such as Sapphire Rapids and
Cascade Lake. Coffee Lake for example works properly, though that might
also be an effect of being on a larger, more NUMA system.
The driver has been fixed but that does not help older guests. Until we
figure out whether Cascade Lake and newer processors are working as
intended, revert the commit. Long term we might add a quirk, but the
details depend on whether the processors are working as intended: for
example if they are, the quirk might reference bochs-compatible devices,
e.g. in the name and documentation, so that userspace can disable the
quirk by default and only leave it enabled if such a device is being
exposed to the guest.
If instead this is actually a bug in CLX+, then the actions we need to
take are different and depend on the actual cause of the bug.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Set PFERR_GUEST_{FINAL,PAGE}_MASK based on EPT_VIOLATION_GVA_TRANSLATED if
and only if EPT_VIOLATION_GVA_IS_VALID is also set in exit qualification.
Per the SDM, bit 8 (EPT_VIOLATION_GVA_TRANSLATED) is valid if and only if
bit 7 (EPT_VIOLATION_GVA_IS_VALID) is set, and is '0' if bit 7 is '0'.
Bit 7 (a.k.a. EPT_VIOLATION_GVA_IS_VALID)
Set if the guest linear-address field is valid. The guest linear-address
field is valid for all EPT violations except those resulting from an
attempt to load the guest PDPTEs as part of the execution of the MOV CR
instruction and those due to trace-address pre-translation
Bit 8 (a.k.a. EPT_VIOLATION_GVA_TRANSLATED)
If bit 7 is 1:
• Set if the access causing the EPT violation is to a guest-physical
address that is the translation of a linear address.
• Clear if the access causing the EPT violation is to a paging-structure
entry as part of a page walk or the update of an accessed or dirty bit.
Reserved if bit 7 is 0 (cleared to 0).
Failure to guard the logic on GVA_IS_VALID results in KVM marking the page
fault as PFERR_GUEST_PAGE_MASK when there is no known GVA, which can put
the vCPU into an infinite loop due to kvm_mmu_page_fault() getting false
positive on its PFERR_NESTED_GUEST_PAGE logic (though only because that
logic is also buggy/flawed).
In practice, this is largely a non-issue because so GVA_IS_VALID is almost
always set. However, when TDX comes along, GVA_IS_VALID will *never* be
set, as the TDX Module deliberately clears bits 12:7 in exit qualification,
e.g. so that the faulting virtual address and other metadata that aren't
practically useful for the hypervisor aren't leaked to the untrusted host.
When exit is due to EPT violation, bits 12-7 of the exit qualification
are cleared to 0.
Fixes: eebed2438923 ("kvm: nVMX: Add support for fast unprotection of nested guest page tables")
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Explicitly invalidate posted_intr_nv when emulating nested VM-Enter and
posted interrupts are disabled to make it clear that posted_intr_nv is
valid if and only if nested posted interrupts are enabled, and as a cheap
way to harden against KVM bugs.
KVM initializes posted_intr_nv to -1 at vCPU creation and resets it to -1
when unloading vmcs12 and/or leaving nested mode, i.e. this is not a bug
fix (or at least, it's not intended to be a bug fix).
Note, tracking nested.posted_intr_nv as a u16 subtly adds a measure of
safety, as it prevents unintentionally matching KVM's informal "no IRQ"
vector of -1, stored as a signed int. Because a u16 can be always be
represented as a signed int, the effective "invalid" value of
posted_intr_nv, 65535, will be preserved as-is when comparing against an
int, i.e. will be zero-extended, not sign-extended, and thus won't get a
false positive if KVM is buggy and compares posted_intr_nv against -1.
Opportunistically add a comment in vmx_deliver_nested_posted_interrupt()
to call out that it must check vmx->nested.posted_intr_nv, not the vector
in vmcs12, which is presumably the _entire_ reason nested.posted_intr_nv
exists. E.g. vmcs12 is a KVM-controlled snapshot, so there are no TOCTOU
races to worry about, the only potential badness is if the vCPU leaves
nested and frees vmcs12 between the sender checking is_guest_mode() and
dereferencing the vmcs12 pointer.
Link: https://lore.kernel.org/r/20240906043413.1049633-7-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
When SGX EDECCSSA support was added to KVM in commit 16a7fe3728a8
("KVM/VMX: Allow exposing EDECCSSA user leaf function to KVM guest"), it
forgot to clear the X86_FEATURE_SGX_EDECCSSA bit in KVM CPU caps when
KVM SGX is disabled. Fix it.
Fixes: 16a7fe3728a8 ("KVM/VMX: Allow exposing EDECCSSA user leaf function to KVM guest")
Signed-off-by: Kai Huang <kai.huang@intel.com>
Link: https://lore.kernel.org/r/20240905120837.579102-1-kai.huang@intel.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Register the "disable virtualization in an emergency" callback just
before KVM enables virtualization in hardware, as there is no functional
need to keep the callbacks registered while KVM happens to be loaded, but
is inactive, i.e. if KVM hasn't enabled virtualization.
Note, unregistering the callback every time the last VM is destroyed could
have measurable latency due to the synchronize_rcu() needed to ensure all
references to the callback are dropped before KVM is unloaded. But the
latency should be a small fraction of the total latency of disabling
virtualization across all CPUs, and userspace can set enable_virt_at_load
to completely eliminate the runtime overhead.
Add a pointer in kvm_x86_ops to allow vendor code to provide its callback.
There is no reason to force vendor code to do the registration, and either
way KVM would need a new kvm_x86_ops hook.
Suggested-by: Kai Huang <kai.huang@intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Acked-by: Kai Huang <kai.huang@intel.com>
Tested-by: Farrah Chen <farrah.chen@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20240830043600.127750-11-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Rename x86's the per-CPU vendor hooks used to enable virtualization in
hardware to align with the recently renamed arch hooks.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Message-ID: <20240830043600.127750-7-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Add a fastpath for HLT VM-Exits by immediately re-entering the guest if
it has a pending wake event. When virtual interrupt delivery is enabled,
i.e. when KVM doesn't need to manually inject interrupts, this allows KVM
to stay in the fastpath run loop when a vIRQ arrives between the guest
doing CLI and STI;HLT. Without AMD's Idle HLT-intercept support, the CPU
generates a HLT VM-Exit even though KVM will immediately resume the guest.
Note, on bare metal, it's relatively uncommon for a modern guest kernel to
actually trigger this scenario, as the window between the guest checking
for a wake event and committing to HLT is quite small. But in a nested
environment, the timings change significantly, e.g. rudimentary testing
showed that ~50% of HLT exits where HLT-polling was successful would be
serviced by this fastpath, i.e. ~50% of the time that a nested vCPU gets
a wake event before KVM schedules out the vCPU, the wake event was pending
even before the VM-Exit.
Link: https://lore.kernel.org/all/20240528041926.3989-3-manali.shukla@amd.com
Link: https://lore.kernel.org/r/20240802195120.325560-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Rename all APIs related to feature MSRs from get_msr_feature() to
get_feature_msr(). The APIs get "feature MSRs", not "MSR features".
And unlike kvm_{g,s}et_msr_common(), the "feature" adjective doesn't
describe the helper itself.
No functional change intended.
Link: https://lore.kernel.org/r/20240802181935.292540-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Refactor get_msr_feature() to take the index and data pointer as distinct
parameters in anticipation of eliminating "struct kvm_msr_entry" usage
further up the primary callchain.
No functional change intended.
Link: https://lore.kernel.org/r/20240802181935.292540-5-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Rename the "INVALID" internal MSR error return code to "UNSUPPORTED" to
try and make it more clear that access was denied because the MSR itself
is unsupported/unknown. "INVALID" is too ambiguous, as it could just as
easily mean the value for WRMSR as invalid.
Avoid UNKNOWN and UNIMPLEMENTED, as the error code is used for MSRs that
_are_ actually implemented by KVM, e.g. if the MSR is unsupported because
an associated feature flag is not present in guest CPUID.
Opportunistically beef up the comments for the internal MSR error codes.
Link: https://lore.kernel.org/r/20240802181935.292540-4-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
In prepare_vmcs02_rare(), call vmx_segment_cache_clear() instead of
setting segment_cache.bitmask directly. Using the helper minimizes the
chances of prepare_vmcs02_rare() doing the wrong thing in the future, e.g.
if KVM ends up doing more than just zero the bitmask when purging the
cache.
No functional change intended.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20240725175232.337266-2-mlevitsk@redhat.com
[sean: massage changelog]
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Use vmx_misc_preemption_timer_rate() to get the rate in hardware_setup(),
and open code the rate's bitmask in vmx_misc_preemption_timer_rate() so
that the function looks like all the helpers that grab values from
VMX_BASIC and VMX_MISC MSR values.
No functional change intended.
Cc: Shan Kang <shan.kang@intel.com>
Cc: Kai Huang <kai.huang@intel.com>
Signed-off-by: Xin Li <xin3.li@intel.com>
[sean: split to separate patch, write changelog]
Reviewed-by: Kai Huang <kai.huang@intel.com>
Reviewed-by: Xiaoyao Li <xiaoyao.li@intel.com>
Link: https://lore.kernel.org/r/20240605231918.2915961-10-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Track the "basic" capabilities VMX MSR as a single u64 in vmcs_config
instead of splitting it across three fields, that obviously don't combine
into a single 64-bit value, so that KVM can use the macros that define MSR
bits using their absolute position. Replace all open coded shifts and
masks, many of which are relative to the "high" half, with the appropriate
macro.
Opportunistically use VMX_BASIC_32BIT_PHYS_ADDR_ONLY instead of an open
coded equivalent, and clean up the related comment to not reference a
specific SDM section (to the surprise of no one, the comment is stale).
No functional change intended (though obviously the code generation will
be quite different).
Cc: Shan Kang <shan.kang@intel.com>
Cc: Kai Huang <kai.huang@intel.com>
Signed-off-by: Xin Li <xin3.li@intel.com>
[sean: split to separate patch, write changelog]
Reviewed-by: Xiaoyao Li <xiaoyao.li@intel.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/20240605231918.2915961-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
etc.)
Add defines for the architectural memory types that can be shoved into
various MSRs and registers, e.g. MTRRs, PAT, VMX capabilities MSRs, EPTPs,
etc. While most MSRs/registers support only a subset of all memory types,
the values themselves are architectural and identical across all users.
Leave the goofy MTRR_TYPE_* definitions as-is since they are in a uapi
header, but add compile-time assertions to connect the dots (and sanity
check that the msr-index.h values didn't get fat-fingered).
Keep the VMX_EPTP_MT_* defines so that it's slightly more obvious that the
EPTP holds a single memory type in 3 of its 64 bits; those bits just
happen to be 2:0, i.e. don't need to be shifted.
Opportunistically use X86_MEMTYPE_WB instead of an open coded '6' in
setup_vmcs_config().
No functional change intended.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Kai Huang <kai.huang@intel.com>
Reviewed-by: Xiaoyao Li <xiaoyao.li@intel.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Link: https://lore.kernel.org/r/20240605231918.2915961-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
KVM VMX changes for 6.11
- Remove an unnecessary EPT TLB flush when enabling hardware.
- Fix a series of bugs that cause KVM to fail to detect nested pending posted
interrupts as valid wake eents for a vCPU executing HLT in L2 (with
HLT-exiting disable by L1).
- Misc cleanups
|
|
KVM x86/pmu changes for 6.11
- Don't advertise IA32_PERF_GLOBAL_OVF_CTRL as an MSR-to-be-saved, as it reads
'0' and writes from userspace are ignored.
- Update to the newfangled Intel CPU FMS infrastructure.
- Use macros instead of open-coded literals to clean up KVM's manipulation of
FIXED_CTR_CTRL MSRs.
|
|
KVM x86 MTRR virtualization removal
Remove support for virtualizing MTRRs on Intel CPUs, along with a nasty CR0.CD
hack, and instead always honor guest PAT on CPUs that support self-snoop.
|
|
KVM x86 misc changes for 6.11
- Add a global struct to consolidate tracking of host values, e.g. EFER, and
move "shadow_phys_bits" into the structure as "maxphyaddr".
- Add KVM_CAP_X86_APIC_BUS_CYCLES_NS to allow configuring the effective APIC
bus frequency, because TDX.
- Print the name of the APICv/AVIC inhibits in the relevant tracepoint.
- Clean up KVM's handling of vendor specific emulation to consistently act on
"compatible with Intel/AMD", versus checking for a specific vendor.
- Misc cleanups
|
|
KVM generic changes for 6.11
- Enable halt poll shrinking by default, as Intel found it to be a clear win.
- Setup empty IRQ routing when creating a VM to avoid having to synchronize
SRCU when creating a split IRQCHIP on x86.
- Rework the sched_in/out() paths to replace kvm_arch_sched_in() with a flag
that arch code can use for hooking both sched_in() and sched_out().
- Take the vCPU @id as an "unsigned long" instead of "u32" to avoid
truncating a bogus value from userspace, e.g. to help userspace detect bugs.
- Mark a vCPU as preempted if and only if it's scheduled out while in the
KVM_RUN loop, e.g. to avoid marking it preempted and thus writing guest
memory when retrieving guest state during live migration blackout.
- A few minor cleanups
|
|
Check for a Requested Virtual Interrupt, i.e. a virtual interrupt that is
pending delivery, in vmx_has_nested_events() and drop the one-off
kvm_x86_ops.guest_apic_has_interrupt() hook.
In addition to dropping a superfluous hook, this fixes a bug where KVM
would incorrectly treat virtual interrupts _for L2_ as always enabled due
to kvm_arch_interrupt_allowed(), by way of vmx_interrupt_blocked(),
treating IRQs as enabled if L2 is active and vmcs12 is configured to exit
on IRQs, i.e. KVM would treat a virtual interrupt for L2 as a valid wake
event based on L1's IRQ blocking status.
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20240607172609.3205077-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move the non-VMX chunk of the "interrupt blocked" checks to a separate
helper so that KVM can reuse the code to detect if interrupts are blocked
for L2, e.g. to determine if a virtual interrupt _for L2_ is a valid wake
event. If L1 disables HLT-exiting for L2, nested APICv is enabled, and L2
HLTs, then L2 virtual interrupts are valid wake events, but if and only if
interrupts are unblocked for L2.
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20240607172609.3205077-4-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
In the vmx_init() error handling path, the __vmx_exit() is done before
kvm_x86_vendor_exit(). They should follow the same order in vmx_exit().
But currently __vmx_exit() is done after kvm_x86_vendor_exit() in
vmx_exit(). Switch the order of them to fix.
Fixes: e32b120071ea ("KVM: VMX: Do _all_ initialization before exposing /dev/kvm to userspace")
Signed-off-by: Kai Huang <kai.huang@intel.com>
Link: https://lore.kernel.org/r/20240627010524.3732488-1-kai.huang@intel.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Remove the completely pointess global INVEPT, i.e. EPT TLB flush, from
KVM's VMX enablement path. KVM always does a targeted TLB flush when
using a "new" EPT root, in quotes because "new" simply means a root that
isn't currently being used by the vCPU.
KVM also _deliberately_ runs with stale TLB entries for defunct roots,
i.e. doesn't do a TLB flush when vCPUs stop using roots, precisely because
KVM does the flush on first use. As called out by the comment in
kvm_mmu_load(), the reason KVM flushes on first use is because KVM can't
guarantee the correctness of past hypervisors.
Jumping back to the global INVEPT, when the painfully terse commit
1439442c7b25 ("KVM: VMX: Enable EPT feature for KVM") was added, the
effective TLB flush being performed was:
static void vmx_flush_tlb(struct kvm_vcpu *vcpu)
{
vpid_sync_vcpu_all(to_vmx(vcpu));
}
I.e. KVM was not flushing EPT TLB entries when allocating a "new" root,
which very strongly suggests that the global INVEPT during hardware
enabling was a misguided hack that addressed the most obvious symptom,
but failed to fix the underlying bug.
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20240608001003.3296640-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Add module descriptions for the vendor modules to fix allmodconfig
'make W=1' warnings:
WARNING: modpost: missing MODULE_DESCRIPTION() in arch/x86/kvm/kvm-intel.o
WARNING: modpost: missing MODULE_DESCRIPTION() in arch/x86/kvm/kvm-amd.o
Signed-off-by: Jeff Johnson <quic_jjohnson@quicinc.com>
Link: https://lore.kernel.org/r/20240622-md-kvm-v2-1-29a60f7c48b1@quicinc.com
[sean: split kvm.ko change to separate commit]
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Now that KVM unconditionally sets l1tf_flush_l1d in kvm_arch_vcpu_load(),
drop the redundant store from vcpu_run(). The flag is cleared only when
VM-Enter is imminent, deep below vcpu_run(), i.e. barring a KVM bug, it's
impossible for l1tf_flush_l1d to be cleared between loading the vCPU and
calling vcpu_run().
Acked-by: Kai Huang <kai.huang@intel.com>
Link: https://lore.kernel.org/r/20240522014013.1672962-7-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Fold the guts of kvm_arch_sched_in() into kvm_arch_vcpu_load(), keying
off the recently added kvm_vcpu.scheduled_out as appropriate.
Note, there is a very slight functional change, as PLE shrink updates will
now happen after blasting WBINVD, but that is quite uninteresting as the
two operations do not interact in any way.
Acked-by: Kai Huang <kai.huang@intel.com>
Link: https://lore.kernel.org/r/20240522014013.1672962-4-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move VMX's {grow,shrink}_ple_window() above vmx_vcpu_load() in preparation
of moving the sched_in logic, which handles shrinking the PLE window, into
vmx_vcpu_load().
No functional change intended.
Acked-by: Kai Huang <kai.huang@intel.com>
Link: https://lore.kernel.org/r/20240522014013.1672962-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Unconditionally honor guest PAT on CPUs that support self-snoop, as
Intel has confirmed that CPUs that support self-snoop always snoop caches
and store buffers. I.e. CPUs with self-snoop maintain cache coherency
even in the presence of aliased memtypes, thus there is no need to trust
the guest behaves and only honor PAT as a last resort, as KVM does today.
Honoring guest PAT is desirable for use cases where the guest has access
to non-coherent DMA _without_ bouncing through VFIO, e.g. when a virtual
(mediated, for all intents and purposes) GPU is exposed to the guest, along
with buffers that are consumed directly by the physical GPU, i.e. which
can't be proxied by the host to ensure writes from the guest are performed
with the correct memory type for the GPU.
Cc: Yiwei Zhang <zzyiwei@google.com>
Suggested-by: Yan Zhao <yan.y.zhao@intel.com>
Suggested-by: Kevin Tian <kevin.tian@intel.com>
Tested-by: Xiangfei Ma <xiangfeix.ma@intel.com>
Tested-by: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lore.kernel.org/r/20240309010929.1403984-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Drop KVM's emulation of CR0.CD=1 on Intel CPUs now that KVM no longer
honors guest MTRR memtypes, as forcing UC memory for VMs with
non-coherent DMA only makes sense if the guest is using something other
than PAT to configure the memtype for the DMA region.
Furthermore, KVM has forced WB memory for CR0.CD=1 since commit
fb279950ba02 ("KVM: vmx: obey KVM_QUIRK_CD_NW_CLEARED"), and no known
VMM in existence disables KVM_X86_QUIRK_CD_NW_CLEARED, let alone does
so with non-coherent DMA.
Lastly, commit fb279950ba02 ("KVM: vmx: obey KVM_QUIRK_CD_NW_CLEARED") was
from the same author as commit b18d5431acc7 ("KVM: x86: fix CR0.CD
virtualization"), and followed by a mere month. I.e. forcing UC memory
was likely the result of code inspection or perhaps misdiagnosed failures,
and not the necessitate by a concrete use case.
Update KVM's documentation to note that KVM_X86_QUIRK_CD_NW_CLEARED is now
AMD-only, and to take an erratum for lack of CR0.CD virtualization on
Intel.
Tested-by: Xiangfei Ma <xiangfeix.ma@intel.com>
Tested-by: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lore.kernel.org/r/20240309010929.1403984-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Remove KVM's support for virtualizing guest MTRR memtypes, as full MTRR
adds no value, negatively impacts guest performance, and is a maintenance
burden due to it's complexity and oddities.
KVM's approach to virtualizating MTRRs make no sense, at all. KVM *only*
honors guest MTRR memtypes if EPT is enabled *and* the guest has a device
that may perform non-coherent DMA access. From a hardware virtualization
perspective of guest MTRRs, there is _nothing_ special about EPT. Legacy
shadowing paging doesn't magically account for guest MTRRs, nor does NPT.
Unwinding and deciphering KVM's murky history, the MTRR virtualization
code appears to be the result of misdiagnosed issues when EPT + VT-d with
passthrough devices was enabled years and years ago. And importantly, the
underlying bugs that were fudged around by honoring guest MTRR memtypes
have since been fixed (though rather poorly in some cases).
The zapping GFNs logic in the MTRR virtualization code came from:
commit efdfe536d8c643391e19d5726b072f82964bfbdb
Author: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Date: Wed May 13 14:42:27 2015 +0800
KVM: MMU: fix MTRR update
Currently, whenever guest MTRR registers are changed
kvm_mmu_reset_context is called to switch to the new root shadow page
table, however, it's useless since:
1) the cache type is not cached into shadow page's attribute so that
the original root shadow page will be reused
2) the cache type is set on the last spte, that means we should sync
the last sptes when MTRR is changed
This patch fixs this issue by drop all the spte in the gfn range which
is being updated by MTRR
which was a fix for:
commit 0bed3b568b68e5835ef5da888a372b9beabf7544
Author: Sheng Yang <sheng@linux.intel.com>
AuthorDate: Thu Oct 9 16:01:54 2008 +0800
Commit: Avi Kivity <avi@redhat.com>
CommitDate: Wed Dec 31 16:51:44 2008 +0200
KVM: Improve MTRR structure
As well as reset mmu context when set MTRR.
which was part of a "MTRR/PAT support for EPT" series that also added:
+ if (mt_mask) {
+ mt_mask = get_memory_type(vcpu, gfn) <<
+ kvm_x86_ops->get_mt_mask_shift();
+ spte |= mt_mask;
+ }
where get_memory_type() was a truly gnarly helper to retrieve the guest
MTRR memtype for a given memtype. And *very* subtly, at the time of that
change, KVM *always* set VMX_EPT_IGMT_BIT,
kvm_mmu_set_base_ptes(VMX_EPT_READABLE_MASK |
VMX_EPT_WRITABLE_MASK |
VMX_EPT_DEFAULT_MT << VMX_EPT_MT_EPTE_SHIFT |
VMX_EPT_IGMT_BIT);
which came in via:
commit 928d4bf747e9c290b690ff515d8f81e8ee226d97
Author: Sheng Yang <sheng@linux.intel.com>
AuthorDate: Thu Nov 6 14:55:45 2008 +0800
Commit: Avi Kivity <avi@redhat.com>
CommitDate: Tue Nov 11 21:00:37 2008 +0200
KVM: VMX: Set IGMT bit in EPT entry
There is a potential issue that, when guest using pagetable without vmexit when
EPT enabled, guest would use PAT/PCD/PWT bits to index PAT msr for it's memory,
which would be inconsistent with host side and would cause host MCE due to
inconsistent cache attribute.
The patch set IGMT bit in EPT entry to ignore guest PAT and use WB as default
memory type to protect host (notice that all memory mapped by KVM should be WB).
Note the CommitDates! The AuthorDates strongly suggests Sheng Yang added
the whole "ignoreIGMT things as a bug fix for issues that were detected
during EPT + VT-d + passthrough enabling, but it was applied earlier
because it was a generic fix.
Jumping back to 0bed3b568b68 ("KVM: Improve MTRR structure"), the other
relevant code, or rather lack thereof, is the handling of *host* MMIO.
That fix came in a bit later, but given the author and timing, it's safe
to say it was all part of the same EPT+VT-d enabling mess.
commit 2aaf69dcee864f4fb6402638dd2f263324ac839f
Author: Sheng Yang <sheng@linux.intel.com>
AuthorDate: Wed Jan 21 16:52:16 2009 +0800
Commit: Avi Kivity <avi@redhat.com>
CommitDate: Sun Feb 15 02:47:37 2009 +0200
KVM: MMU: Map device MMIO as UC in EPT
Software are not allow to access device MMIO using cacheable memory type, the
patch limit MMIO region with UC and WC(guest can select WC using PAT and
PCD/PWT).
In addition to the host MMIO and IGMT issues, KVM's MTRR virtualization
was obviously never tested on NPT until much later, which lends further
credence to the theory/argument that this was all the result of
misdiagnosed issues.
Discussion from the EPT+MTRR enabling thread[*] more or less confirms that
Sheng Yang was trying to resolve issues with passthrough MMIO.
* Sheng Yang
: Do you mean host(qemu) would access this memory and if we set it to guest
: MTRR, host access would be broken? We would cover this in our shadow MTRR
: patch, for we encountered this in video ram when doing some experiment with
: VGA assignment.
And in the same thread, there's also what appears to be confirmation of
Intel running into issues with Windows XP related to a guest device driver
mapping DMA with WC in the PAT.
* Avi Kavity
: Sheng Yang wrote:
: > Yes... But it's easy to do with assigned devices' mmio, but what if guest
: > specific some non-mmio memory's memory type? E.g. we have met one issue in
: > Xen, that a assigned-device's XP driver specific one memory region as buffer,
: > and modify the memory type then do DMA.
: >
: > Only map MMIO space can be first step, but I guess we can modify assigned
: > memory region memory type follow guest's?
: >
:
: With ept/npt, we can't, since the memory type is in the guest's
: pagetable entries, and these are not accessible.
[*] https://lore.kernel.org/all/1223539317-32379-1-git-send-email-sheng@linux.intel.com
So, for the most part, what likely happened is that 15 years ago, a few
engineers (a) fixed a #MC problem by ignoring guest PAT and (b) initially
"fixed" passthrough device MMIO by emulating *guest* MTRRs. Except for
the below case, everything since then has been a result of those two
intertwined changes.
The one exception, which is actually yet more confirmation of all of the
above, is the revert of Paolo's attempt at "full" virtualization of guest
MTRRs:
commit 606decd67049217684e3cb5a54104d51ddd4ef35
Author: Paolo Bonzini <pbonzini@redhat.com>
Date: Thu Oct 1 13:12:47 2015 +0200
Revert "KVM: x86: apply guest MTRR virtualization on host reserved pages"
This reverts commit fd717f11015f673487ffc826e59b2bad69d20fe5.
It was reported to cause Machine Check Exceptions (bug 104091).
...
commit fd717f11015f673487ffc826e59b2bad69d20fe5
Author: Paolo Bonzini <pbonzini@redhat.com>
Date: Tue Jul 7 14:38:13 2015 +0200
KVM: x86: apply guest MTRR virtualization on host reserved pages
Currently guest MTRR is avoided if kvm_is_reserved_pfn returns true.
However, the guest could prefer a different page type than UC for
such pages. A good example is that pass-throughed VGA frame buffer is
not always UC as host expected.
This patch enables full use of virtual guest MTRRs.
I.e. Paolo tried to add back KVM's behavior before "Map device MMIO as UC
in EPT" and got the same result: machine checks, likely due to the guest
MTRRs not being trustworthy/sane at all times.
Note, Paolo also tried to enable MTRR virtualization on SVM+NPT, but that
too got reverted. Unfortunately, it doesn't appear that anyone ever found
a smoking gun, i.e. exactly why emulating guest MTRRs via NPT PAT caused
extremely slow boot times doesn't appear to have a definitive root cause.
commit fc07e76ac7ffa3afd621a1c3858a503386a14281
Author: Paolo Bonzini <pbonzini@redhat.com>
Date: Thu Oct 1 13:20:22 2015 +0200
Revert "KVM: SVM: use NPT page attributes"
This reverts commit 3c2e7f7de3240216042b61073803b61b9b3cfb22.
Initializing the mapping from MTRR to PAT values was reported to
fail nondeterministically, and it also caused extremely slow boot
(due to caching getting disabled---bug 103321) with assigned devices.
...
commit 3c2e7f7de3240216042b61073803b61b9b3cfb22
Author: Paolo Bonzini <pbonzini@redhat.com>
Date: Tue Jul 7 14:32:17 2015 +0200
KVM: SVM: use NPT page attributes
Right now, NPT page attributes are not used, and the final page
attribute depends solely on gPAT (which however is not synced
correctly), the guest MTRRs and the guest page attributes.
However, we can do better by mimicking what is done for VMX.
In the absence of PCI passthrough, the guest PAT can be ignored
and the page attributes can be just WB. If passthrough is being
used, instead, keep respecting the guest PAT, and emulate the guest
MTRRs through the PAT field of the nested page tables.
The only snag is that WP memory cannot be emulated correctly,
because Linux's default PAT setting only includes the other types.
In short, honoring guest MTRRs for VMX was initially a workaround of
sorts for KVM ignoring guest PAT *and* for KVM not forcing UC for host
MMIO. And while there *are* known cases where honoring guest MTRRs is
desirable, e.g. passthrough VGA frame buffers, the desired behavior in
that case is to get WC instead of UC, i.e. at this point it's for
performance, not correctness.
Furthermore, the complete absence of MTRR virtualization on NPT and
shadow paging proves that, while KVM theoretically can do better, it's
by no means necessary for correctnesss.
Lastly, since kernels mostly rely on firmware to do MTRR setup, and the
host typically provides guest firmware, honoring guest MTRRs is effectively
honoring *host* userspace memtypes, which is also backwards. I.e. it
would be far better for host userspace to communicate its desired memtype
directly to KVM (or perhaps indirectly via VMAs in the host kernel), not
through guest MTRRs.
Tested-by: Xiangfei Ma <xiangfeix.ma@intel.com>
Tested-by: Yongwei Ma <yongwei.ma@intel.com>
Link: https://lore.kernel.org/r/20240309010929.1403984-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
* Fixes and debugging help for the #VE sanity check. Also disable
it by default, even for CONFIG_DEBUG_KERNEL, because it was found
to trigger spuriously (most likely a processor erratum as the
exact symptoms vary by generation).
* Avoid WARN() when two NMIs arrive simultaneously during an NMI-disabled
situation (GIF=0 or interrupt shadow) when the processor supports
virtual NMI. While generally KVM will not request an NMI window
when virtual NMIs are supported, in this case it *does* have to
single-step over the interrupt shadow or enable the STGI intercept,
in order to deliver the latched second NMI.
* Drop support for hand tuning APIC timer advancement from userspace.
Since we have adaptive tuning, and it has proved to work well,
drop the module parameter for manual configuration and with it a
few stupid bugs that it had.
|
|
Use x86_vfm (vendor, family, module) to detect CPUs that are affected by
PERF_GLOBAL_CTRL bugs instead of manually checking the family and model.
The new VFM infrastructure encodes all information in one handy location.
No functional change intended.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Link: https://lore.kernel.org/r/20240520224620.9480-10-tony.luck@intel.com
[sean: massage changelog]
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move shadow_phys_bits into "struct kvm_host_values", i.e. into KVM's
global "kvm_host" variable, so that it is automatically exported for use
in vendor modules. Rename the variable/field to maxphyaddr to more
clearly capture what value it holds, now that it's used outside of the
MMU (and because the "shadow" part is more than a bit misleading as the
variable is not at all unique to shadow paging).
Recomputing the raw/true host.MAXPHYADDR on every use can be subtly
expensive, e.g. it will incur a VM-Exit on the CPUID if KVM is running as
a nested hypervisor. Vendor code already has access to the information,
e.g. by directly doing CPUID or by invoking kvm_get_shadow_phys_bits(), so
there's no tangible benefit to making it MMU-only.
Link: https://lore.kernel.org/r/20240423221521.2923759-5-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
