| Age | Commit message (Collapse) | Author |
|---|
|
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.
The primary motivation for this series is to simplify dealing with enabling
virtualization for Intel's TDX, which needs to enable virtualization
when kvm-intel.ko is loaded, i.e. long before the first VM is created.
That said, this is a nice cleanup on its own. By registering the callbacks
on-demand, the callbacks themselves don't need to check kvm_usage_count,
because their very existence implies a non-zero count.
Patch 1 (re)adds a dedicated lock for kvm_usage_count. This avoids a
lock ordering issue between cpus_read_lock() and kvm_lock. The lock
ordering issue still exist in very rare cases, and will be fixed for
good by switching vm_list to an (S)RCU-protected list.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Today whenever a memslot is moved or deleted, KVM invalidates the entire
page tables and generates fresh ones based on the new memslot layout.
This behavior traditionally was kept because of a bug which was never
fully investigated and caused VM instability with assigned GeForce
GPUs. It generally does not have a huge overhead, because the old
MMU is able to reuse cached page tables and the new one is more
scalabale and can resolve EPT violations/nested page faults in parallel,
but it has worse performance if the guest frequently deletes and
adds small memslots, and it's entirely not viable for TDX. This is
because TDX requires re-accepting of private pages after page dropping.
For non-TDX VMs, this series therefore introduces the
KVM_X86_QUIRK_SLOT_ZAP_ALL quirk, enabling users to control the behavior
of memslot zapping when a memslot is moved/deleted. The quirk is turned
on by default, leading to the zapping of all SPTEs when a memslot is
moved/deleted; users however have the option to turn off the quirk,
which limits the zapping only to those SPTEs hat lie within the range
of memslot being moved/deleted.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Pull kvm fix from Paolo Bonzini:
"Do not always honor guest PAT on CPUs that support self-snoop.
This triggers an issue in the bochsdrm driver, which used ioremap()
instead of ioremap_wc() to map the video RAM.
The revert lets video RAM use the WB memory type instead of the slower
UC memory type"
* tag 'for-linus-6.11' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
Revert "KVM: VMX: Always honor guest PAT on CPUs that support self-snoop"
|
|
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>
|
|
Remove superfluous RSB filling after a VMEXIT when the CPU already has
flushed the RSB after a VMEXIT when AutoIBRS is enabled.
The initial implementation for adding RETPOLINES added an ALTERNATIVES
implementation for filling the RSB after a VMEXIT in commit 117cc7a908c8
("x86/retpoline: Fill return stack buffer on vmexit").
Later, X86_FEATURE_RSB_VMEXIT was added in commit 9756bba28470
("x86/speculation: Fill RSB on vmexit for IBRS") to handle stuffing the
RSB if RETPOLINE=y *or* KERNEL_IBRS=y, i.e. to also stuff the RSB if the
kernel is configured to do IBRS mitigations on entry/exit.
The AutoIBRS (on AMD) feature implementation added in commit e7862eda309e
("x86/cpu: Support AMD Automatic IBRS") used the already-implemented logic
for EIBRS in spectre_v2_determine_rsb_fill_type_on_vmexit() -- but did not
update the code at VMEXIT to act on the mode selected in that function --
resulting in VMEXITs continuing to clear the RSB when RETPOLINES are
enabled, despite the presence of AutoIBRS.
Signed-off-by: Amit Shah <amit.shah@amd.com>
Link: https://lore.kernel.org/r/20240807123531.69677-1-amit@kernel.org
[sean: massage changeloge, drop comment about AMD not needing RSB_VMEXIT_LITE]
Signed-off-by: Sean Christopherson <seanjc@google.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>
|
|
Use KVM_PAGES_PER_HPAGE() instead of open coding equivalent logic that is
anything but obvious.
No functional change intended, and verified by compiling with the below
assertions:
BUILD_BUG_ON((1UL << KVM_HPAGE_GFN_SHIFT(PG_LEVEL_4K)) !=
KVM_PAGES_PER_HPAGE(PG_LEVEL_4K));
BUILD_BUG_ON((1UL << KVM_HPAGE_GFN_SHIFT(PG_LEVEL_2M)) !=
KVM_PAGES_PER_HPAGE(PG_LEVEL_2M));
BUILD_BUG_ON((1UL << KVM_HPAGE_GFN_SHIFT(PG_LEVEL_1G)) !=
KVM_PAGES_PER_HPAGE(PG_LEVEL_1G));
Link: https://lore.kernel.org/r/20240809194335.1726916-19-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Replace all of the open coded '1' literals used to mark a PTE list as
having many/multiple entries with a proper define. It's hard enough to
read the code with one magic bit, and a future patch to support "locking"
a single rmap will add another.
No functional change intended.
Link: https://lore.kernel.org/r/20240809194335.1726916-17-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Fold mmu_spte_age() into its sole caller now that aging and testing for
young SPTEs is handled in a common location, i.e. doesn't require more
helpers.
Opportunistically remove the use of mmu_spte_get_lockless(), as mmu_lock
is held (for write!), and marking SPTEs for access tracking outside of
mmu_lock is unsafe (at least, as written). I.e. using the lockless
accessor is quite misleading.
No functional change intended.
Link: https://lore.kernel.org/r/20240809194335.1726916-16-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Rework kvm_handle_gfn_range() into an aging-specic helper,
kvm_rmap_age_gfn_range(). In addition to purging a bunch of unnecessary
boilerplate code, this sets the stage for aging rmap SPTEs outside of
mmu_lock.
Note, there's a small functional change, as kvm_test_age_gfn() will now
return immediately if a young SPTE is found, whereas previously KVM would
continue iterating over other levels.
Link: https://lore.kernel.org/r/20240809194335.1726916-15-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Convert kvm_unmap_gfn_range(), which is the helper that zaps rmap SPTEs in
response to an mmu_notifier invalidation, to use __kvm_rmap_zap_gfn_range()
and feed in range->may_block. In other words, honor NEED_RESCHED by way of
cond_resched() when zapping rmaps. This fixes a long-standing issue where
KVM could process an absurd number of rmap entries without ever yielding,
e.g. if an mmu_notifier fired on a PUD (or larger) range.
Opportunistically rename __kvm_zap_rmap() to kvm_zap_rmap(), and drop the
old kvm_zap_rmap(). Ideally, the shuffling would be done in a different
patch, but that just makes the compiler unhappy, e.g.
arch/x86/kvm/mmu/mmu.c:1462:13: error: ‘kvm_zap_rmap’ defined but not used
Reported-by: Peter Xu <peterx@redhat.com>
Link: https://lore.kernel.org/r/20240809194335.1726916-14-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Add a dedicated helper to walk and zap rmaps for a given memslot so that
the code can be shared between KVM-initiated zaps and mmu_notifier
invalidations.
No functional change intended.
Link: https://lore.kernel.org/r/20240809194335.1726916-13-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Add a @can_yield param to __walk_slot_rmaps() to control whether or not
dropping mmu_lock and conditionally rescheduling is allowed. This will
allow using __walk_slot_rmaps() and thus cond_resched() to handle
mmu_notifier invalidations, which usually allow blocking/yielding, but not
when invoked by the OOM killer.
Link: https://lore.kernel.org/r/20240809194335.1726916-12-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move walk_slot_rmaps() and friends up near for_each_slot_rmap_range() so
that the walkers can be used to handle mmu_notifier invalidations, and so
that similar function has some amount of locality in code.
No functional change intended.
Link: https://lore.kernel.org/r/20240809194335.1726916-11-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
WARN if KVM gets an MMIO cache hit on a RET_PF_WRITE_PROTECTED fault, as
KVM should return RET_PF_WRITE_PROTECTED if and only if there is a memslot,
and creating a memslot is supposed to invalidate the MMIO cache by virtue
of changing the memslot generation.
Keep the code around mainly to provide a convenient location to document
why emulated MMIO should be impossible.
Suggested-by: Yuan Yao <yuan.yao@linux.intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-23-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Explicitly query the list of to-be-zapped shadow pages when checking to
see if unprotecting a gfn for retry has succeeded, i.e. if KVM should
retry the faulting instruction.
Add a comment to explain why the list needs to be checked before zapping,
which is the primary motivation for this change.
No functional change intended.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-22-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Fold kvm_mmu_unprotect_page() into kvm_mmu_unprotect_gfn_and_retry() now
that all other direct usage is gone.
No functional change intended.
Link: https://lore.kernel.org/r/20240831001538.336683-21-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Rename reexecute_instruction() to kvm_unprotect_and_retry_on_failure() to
make the intent and purpose of the helper much more obvious.
No functional change intended.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-20-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
When retrying the faulting instruction after emulation failure, refresh
the infinite loop protection fields even if no shadow pages were zapped,
i.e. avoid hitting an infinite loop even when retrying the instruction as
a last-ditch effort to avoid terminating the guest.
Link: https://lore.kernel.org/r/20240831001538.336683-19-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Use kvm_mmu_unprotect_gfn_and_retry() in reexecute_instruction() to pick
up protection against infinite loops, e.g. if KVM somehow manages to
encounter an unsupported instruction and unprotecting the gfn doesn't
allow the vCPU to make forward progress. Other than that, the retry-on-
failure logic is a functionally equivalent, open coded version of
kvm_mmu_unprotect_gfn_and_retry().
Note, the emulation failure path still isn't fully protected, as KVM
won't update the retry protection fields if no shadow pages are zapped
(but this change is still a step forward). That flaw will be addressed
in a future patch.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-18-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Don't bother unprotecting the target gfn if EMULTYPE_WRITE_PF_TO_SP is
set, as KVM will simply report the emulation failure to userspace. This
will allow converting reexecute_instruction() to use
kvm_mmu_unprotect_gfn_instead_retry() instead of kvm_mmu_unprotect_page().
Link: https://lore.kernel.org/r/20240831001538.336683-17-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Drop the manual pfn look when retrying an instruction that KVM failed to
emulation in response to a #PF due to a write-protected gfn. Now that KVM
sets EMULTYPE_ALLOW_RETRY_PF if and only if the page fault hit a write-
protected gfn, i.e. if and only if there's a writable memslot, there's no
need to redo the lookup to avoid retrying an instruction that failed on
emulated MMIO (no slot, or a write to a read-only slot).
I.e. KVM will never attempt to retry an instruction that failed on
emulated MMIO, whereas that was not the case prior to the introduction of
RET_PF_WRITE_PROTECTED.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-16-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move the event re-injection unprotect+retry logic into
kvm_mmu_write_protect_fault(), i.e. unprotect and retry if and only if
the #PF actually hit a write-protected gfn. Note, there is a small
possibility that the gfn was unprotected by a different tasking between
hitting the #PF and acquiring mmu_lock, but in that case, KVM will resume
the guest immediately anyways because KVM will treat the fault as spurious.
As a bonus, unprotecting _after_ handling the page fault also addresses the
case where the installing a SPTE to handle fault encounters a shadowed PTE,
i.e. *creates* a read-only SPTE.
Opportunstically add a comment explaining what on earth the intent of the
code is, as based on the changelog from commit 577bdc496614 ("KVM: Avoid
instruction emulation when event delivery is pending").
Link: https://lore.kernel.org/r/20240831001538.336683-15-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
When getting a gpa from a gva to unprotect the associated gfn when an
event is awating reinjection, walk the guest PTEs for WRITE as there's no
point in unprotecting the gfn if the guest is unable to write the page,
i.e. if write-protection can't trigger emulation.
Note, the entire flow should be guarded on the access being a write, and
even better should be conditioned on actually triggering a write-protect
fault. This will be addressed in a future commit.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-14-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
If getting the gpa for a gva fails, e.g. because the gva isn't mapped in
the guest page tables, don't try to unprotect the invalid gfn. This is
mostly a performance fix (avoids unnecessarily taking mmu_lock), as
for_each_gfn_valid_sp_with_gptes() won't explode on garbage input, it's
simply pointless.
Link: https://lore.kernel.org/r/20240831001538.336683-13-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Now that retry_instruction() is reasonably tiny, fold it into its sole
caller, x86_emulate_instruction(). In addition to getting rid of the
absurdly confusing retry_instruction() name, handling the retry in
x86_emulate_instruction() pairs it back up with the code that resets
last_retry_{eip,address}.
No functional change intended.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-12-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move the sanity checks for EMULTYPE_ALLOW_RETRY_PF to the top of
x86_emulate_instruction(). In addition to deduplicating a small amount
of code, this makes the connection between EMULTYPE_ALLOW_RETRY_PF and
EMULTYPE_PF even more explicit, and will allow dropping retry_instruction()
entirely.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-11-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Try to unprotect shadow pages if and only if indirect_shadow_pages is non-
zero, i.e. iff there is at least one protected such shadow page. Pre-
checking indirect_shadow_pages avoids taking mmu_lock for write when the
gfn is write-protected by a third party, i.e. not for KVM shadow paging,
and in the *extremely* unlikely case that a different task has already
unprotected the last shadow page.
Link: https://lore.kernel.org/r/20240831001538.336683-10-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move the anti-infinite-loop protection provided by last_retry_{eip,addr}
into kvm_mmu_write_protect_fault() so that it guards unprotect+retry that
never hits the emulator, as well as reexecute_instruction(), which is the
last ditch "might as well try it" logic that kicks in when emulation fails
on an instruction that faulted on a write-protected gfn.
Add a new helper, kvm_mmu_unprotect_gfn_and_retry(), to set the retry
fields and deduplicate other code (with more to come).
Link: https://lore.kernel.org/r/20240831001538.336683-9-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Store the gpa used to unprotect the faulting gfn for retry as a gpa_t, not
an unsigned long. This fixes a bug where 32-bit KVM would unprotect and
retry the wrong gfn if the gpa had bits 63:32!=0. In practice, this bug
is functionally benign, as unprotecting the wrong gfn is purely a
performance issue (thanks to the anti-infinite-loop logic). And of course,
almost no one runs 32-bit KVM these days.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-8-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Read RIP from vCPU state instead of pulling it from the emulation context
when filling last_retry_eip, which is part of the anti-infinite-loop
protection used when unprotecting and retrying instructions that hit a
write-protected gfn.
This will allow reusing the anti-infinite-loop protection in flows that
never make it into the emulator.
No functional change intended, as ctxt->eip is set to kvm_rip_read() in
init_emulate_ctxt(), and EMULTYPE_PF emulation is mutually exclusive with
EMULTYPE_NO_DECODE and EMULTYPE_SKIP, i.e. always goes through
x86_decode_emulated_instruction() and hasn't advanced ctxt->eip (yet).
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-7-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Resume the guest and thus skip emulation of a non-PTE-writing instruction
if and only if unprotecting the gfn actually zapped at least one shadow
page. If the gfn is write-protected for some reason other than shadow
paging, attempting to unprotect the gfn will effectively fail, and thus
retrying the instruction is all but guaranteed to be pointless. This bug
has existed for a long time, but was effectively fudged around by the
retry RIP+address anti-loop detection.
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
When doing "fast unprotection" of nested TDP page tables, skip emulation
if and only if at least one gfn was unprotected, i.e. continue with
emulation if simply resuming is likely to hit the same fault and risk
putting the vCPU into an infinite loop.
Note, it's entirely possible to get a false negative, e.g. if a different
vCPU faults on the same gfn and unprotects the gfn first, but that's a
relatively rare edge case, and emulating is still functionally ok, i.e.
saving a few cycles by avoiding emulation isn't worth the risk of putting
the vCPU into an infinite loop.
Opportunistically rewrite the relevant comment to document in gory detail
exactly what scenario the "fast unprotect" logic is handling.
Fixes: 147277540bbc ("kvm: svm: Add support for additional SVM NPF error codes")
Cc: Yuan Yao <yuan.yao@intel.com>
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-5-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Trigger KVM's various "unprotect gfn" paths if and only if the page fault
was a write to a write-protected gfn. To do so, add a new page fault
return code, RET_PF_WRITE_PROTECTED, to explicitly and precisely track
such page faults.
If a page fault requires emulation for any MMIO (or any reason besides
write-protection), trying to unprotect the gfn is pointless and risks
putting the vCPU into an infinite loop. E.g. KVM will put the vCPU into
an infinite loop if the vCPU manages to trigger MMIO on a page table walk.
Fixes: 147277540bbc ("kvm: svm: Add support for additional SVM NPF error codes")
Reviewed-by: Yuan Yao <yuan.yao@intel.com>
Link: https://lore.kernel.org/r/20240831001538.336683-4-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Drop the globally visible PFERR_NESTED_GUEST_PAGE and replace it with a
more appropriately named is_write_to_guest_page_table(). The macro name
is misleading, because while all nNPT walks match PAGE|WRITE|PRESENT, the
reverse is not true.
No functional change intended.
Link: https://lore.kernel.org/r/20240831001538.336683-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Add lockdep assertions in get_vmcs12() and get_shadow_vmcs12() to verify
the vCPU's mutex is held, as the returned VMCS objects are dynamically
allocated/freed when nested VMX is turned on/off, i.e. accessing vmcs12
structures without holding vcpu->mutex is susceptible to use-after-free.
Waive the assertion if the VM is being destroyed, as KVM currently forces
a nested VM-Exit when freeing the vCPU. If/when that wart is fixed, the
assertion can/should be converted to an unqualified lockdep assertion.
See also https://lore.kernel.org/all/Zsd0TqCeY3B5Sb5b@google.com.
Link: https://lore.kernel.org/r/20240906043413.1049633-8-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>
|
|
Fold kvm_get_apic_interrupt() into kvm_cpu_get_interrupt() now that nVMX
essentially open codes kvm_get_apic_interrupt() in order to correctly
emulate nested posted interrupts.
Opportunistically stop exporting kvm_cpu_get_interrupt(), as the
aforementioned nVMX flow was the only user in vendor code.
Link: https://lore.kernel.org/r/20240906043413.1049633-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
When synthensizing a nested VM-Exit due to an external interrupt, pend a
nested posted interrupt if the external interrupt vector matches L2's PI
notification vector, i.e. if the interrupt is a PI notification for L2.
This fixes a bug where KVM will incorrectly inject VM-Exit instead of
processing nested posted interrupt when IPI virtualization is enabled.
Per the SDM, detection of the notification vector doesn't occur until the
interrupt is acknowledge and deliver to the CPU core.
If the external-interrupt exiting VM-execution control is 1, any unmasked
external interrupt causes a VM exit (see Section 26.2). If the "process
posted interrupts" VM-execution control is also 1, this behavior is
changed and the processor handles an external interrupt as follows:
1. The local APIC is acknowledged; this provides the processor core
with an interrupt vector, called here the physical vector.
2. If the physical vector equals the posted-interrupt notification
vector, the logical processor continues to the next step. Otherwise,
a VM exit occurs as it would normally due to an external interrupt;
the vector is saved in the VM-exit interruption-information field.
For the most part, KVM has avoided problems because a PI NV for L2 that
arrives will L2 is active will be processed by hardware, and KVM checks
for a pending notification vector during nested VM-Enter. Thus, to hit
the bug, the PI NV interrupt needs to sneak its way into L1's vIRR while
L2 is active.
Without IPI virtualization, the scenario is practically impossible to hit,
modulo L1 doing weird things (see below), as the ordering between
vmx_deliver_posted_interrupt() and nested VM-Enter effectively guarantees
that either the sender will see the vCPU as being in_guest_mode(), or the
receiver will see the interrupt in its vIRR.
With IPI virtualization, introduced by commit d588bb9be1da ("KVM: VMX:
enable IPI virtualization"), the sending CPU effectively implements a rough
equivalent of vmx_deliver_posted_interrupt(), sans the nested PI NV check.
If the target vCPU has a valid PID, the CPU will send a PI NV interrupt
based on _L1's_ PID, as the sender's because IPIv table points at L1 PIDs.
PIR := 32 bytes at PID_ADDR;
// under lock
PIR[V] := 1;
store PIR at PID_ADDR;
// release lock
NotifyInfo := 8 bytes at PID_ADDR + 32;
// under lock
IF NotifyInfo.ON = 0 AND NotifyInfo.SN = 0; THEN
NotifyInfo.ON := 1;
SendNotify := 1;
ELSE
SendNotify := 0;
FI;
store NotifyInfo at PID_ADDR + 32;
// release lock
IF SendNotify = 1; THEN
send an IPI specified by NotifyInfo.NDST and NotifyInfo.NV;
FI;
As a result, the target vCPU ends up receiving an interrupt on KVM's
POSTED_INTR_VECTOR while L2 is running, with an interrupt in L1's PIR for
L2's nested PI NV. The POSTED_INTR_VECTOR interrupt triggers a VM-Exit
from L2 to L0, KVM moves the interrupt from L1's PIR to vIRR, triggers a
KVM_REQ_EVENT prior to re-entry to L2, and calls vmx_check_nested_events(),
effectively bypassing all of KVM's "early" checks on nested PI NV.
Without IPI virtualization, the bug can likely be hit only if L1 programs
an assigned device to _post_ an interrupt to L2's notification vector, by
way of L1's PID.PIR. Doing so would allow the interrupt to get into L1's
vIRR without KVM checking vmcs12's NV. Which is architecturally allowed,
but unlikely behavior for a hypervisor.
Cc: Zeng Guang <guang.zeng@intel.com>
Reviewed-by: Chao Gao <chao.gao@intel.com>
Link: https://lore.kernel.org/r/20240906043413.1049633-5-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
In the should-be-impossible scenario that kvm_cpu_get_interrupt() doesn't
return a valid vector after checking kvm_cpu_has_interrupt(), skip VM-Exit
injection to reduce the probability of crashing/confusing L1. Now that
KVM gets the IRQ _before_ calling nested_vmx_vmexit(), squashing the
VM-Exit injection is trivial since there are no actions that need to be
undone.
Reviewed-by: Chao Gao <chao.gao@intel.com>
Link: https://lore.kernel.org/r/20240906043413.1049633-4-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Move the logic to get the to-be-acknowledge IRQ for a nested VM-Exit from
nested_vmx_vmexit() to vmx_check_nested_events(), which is subtly the one
and only path where KVM invokes nested_vmx_vmexit() with
EXIT_REASON_EXTERNAL_INTERRUPT. A future fix will perform a last-minute
check on L2's nested posted interrupt notification vector, just before
injecting a nested VM-Exit. To handle that scenario correctly, KVM needs
to get the interrupt _before_ injecting VM-Exit, as simply querying the
highest priority interrupt, via kvm_cpu_has_interrupt(), would result in
TOCTOU bug, as a new, higher priority interrupt could arrive between
kvm_cpu_has_interrupt() and kvm_cpu_get_interrupt().
Unfortunately, simply moving the call to kvm_cpu_get_interrupt() doesn't
suffice, as a VMWRITE to GUEST_INTERRUPT_STATUS.SVI is hiding in
kvm_get_apic_interrupt(), and acknowledging the interrupt before nested
VM-Exit would cause the VMWRITE to hit vmcs02 instead of vmcs01.
Open code a rough equivalent to kvm_cpu_get_interrupt() so that the IRQ
is acknowledged after emulating VM-Exit, taking care to avoid the TOCTOU
issue described above.
Opportunistically convert the WARN_ON() to a WARN_ON_ONCE(). If KVM has
a bug that results in a false positive from kvm_cpu_has_interrupt(),
spamming dmesg won't help the situation.
Note, nested_vmx_reflect_vmexit() can never reflect external interrupts as
they are always "wanted" by L0.
Link: https://lore.kernel.org/r/20240906043413.1049633-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Split the "ack" phase, i.e. the movement of an interrupt from IRR=>ISR,
out of kvm_get_apic_interrupt() and into a separate API so that nested
VMX can acknowledge a specific interrupt _after_ emulating a VM-Exit from
L2 to L1.
To correctly emulate nested posted interrupts while APICv is active, KVM
must:
1. find the highest pending interrupt.
2. check if that IRQ is L2's notification vector
3. emulate VM-Exit if the IRQ is NOT the notification vector
4. ACK the IRQ in L1 _after_ VM-Exit
When APICv is active, the process of moving the IRQ from the IRR to the
ISR also requires a VMWRITE to update vmcs01.GUEST_INTERRUPT_STATUS.SVI,
and so acknowledging the interrupt before switching to vmcs01 would result
in marking the IRQ as in-service in the wrong VMCS.
KVM currently fudges around this issue by doing kvm_get_apic_interrupt()
smack dab in the middle of emulating VM-Exit, but that hack doesn't play
nice with nested posted interrupts, as notification vector IRQs don't
trigger a VM-Exit in the first place.
Cc: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20240906043413.1049633-2-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>
|
|
Commit 238adc77051a ("KVM: Cleanup LAPIC interface") removed
kvm_lapic_get_base() but leave declaration.
And other two declarations were never implenmented since introduction.
Signed-off-by: Yue Haibing <yuehaibing@huawei.com>
Link: https://lore.kernel.org/r/20240830022537.2403873-1-yuehaibing@huawei.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
Leave nested mode before synthesizing shutdown (a.k.a. TRIPLE_FAULT) if
RSM fails when resuming L2 (a.k.a. guest mode). Architecturally, shutdown
on RSM occurs _before_ the transition back to guest mode on both Intel and
AMD.
On Intel, per the SDM pseudocode, SMRAM state is loaded before critical
VMX state:
restore state normally from SMRAM;
...
CR4.VMXE := value stored internally;
IF internal storage indicates that the logical processor had been in
VMX operation (root or non-root)
THEN
enter VMX operation (root or non-root);
restore VMX-critical state as defined in Section 32.14.1;
...
restore current VMCS pointer;
FI;
AMD's APM is both less clearcut and more explicit. Because AMD CPUs save
VMCB and guest state in SMRAM itself, given the lack of anything in the
APM to indicate a shutdown in guest mode is possible, a straightforward
reading of the clause on invalid state is that _what_ state is invalid is
irrelevant, i.e. all roads lead to shutdown.
An RSM causes a processor shutdown if an invalid-state condition is
found in the SMRAM state-save area.
This fixes a bug found by syzkaller where synthesizing shutdown for L2
led to a nested VM-Exit (if L1 is intercepting shutdown), which in turn
caused KVM to complain about trying to cancel a nested VM-Enter (see
commit 759cbd59674a ("KVM: x86: nSVM/nVMX: set nested_run_pending on VM
entry which is a result of RSM").
Note, Paolo pointed out that KVM shouldn't set nested_run_pending until
after loading SMRAM state. But as above, that's only half the story, KVM
shouldn't transition to guest mode either. Unfortunately, fixing that
mess requires rewriting the nVMX and nSVM RSM flows to not piggyback
their nested VM-Enter flows, as executing the nested VM-Enter flows after
loading state from SMRAM would clobber much of said state.
For now, add a FIXME to call out that transitioning to guest mode before
loading state from SMRAM is wrong.
Link: https://lore.kernel.org/all/CABgObfYaUHXyRmsmg8UjRomnpQ0Jnaog9-L2gMjsjkqChjDYUQ@mail.gmail.com
Reported-by: syzbot+988d9efcdf137bc05f66@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/0000000000007a9acb06151e1670@google.com
Reported-by: Zheyu Ma <zheyuma97@gmail.com>
Closes: https://lore.kernel.org/all/CAMhUBjmXMYsEoVYw_M8hSZjBMHh24i88QYm-RY6HDta5YZ7Wgw@mail.gmail.com
Analyzed-by: Michal Wilczynski <michal.wilczynski@intel.com>
Cc: Kishen Maloor <kishen.maloor@intel.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20240906161337.1118412-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
|
|
clang warns on this because it has an unannotated fall-through between
cases:
arch/x86/kvm/x86.c:4819:2: error: unannotated fall-through between switch labels [-Werror,-Wimplicit-fallthrough]
and while we could annotate it as a fallthrough, the proper fix is to
just add the break for this case, instead of falling through to the
default case and the break there.
gcc also has that warning, but it looks like gcc only warns for the
cases where they fall through to "real code", rather than to just a
break. Odd.
Fixes: d30d9ee94cc0 ("KVM: x86: Only advertise KVM_CAP_READONLY_MEM when supported by VM")
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Tom Dohrmann <erbse.13@gmx.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Commit 92b5265d38f6a ("KVM: Depend on HIGH_RES_TIMERS") added a dependency
to high resolution timers with the comment:
KVM lapic timer and tsc deadline timer based on hrtimer,
setting a leftmost node to rb tree and then do hrtimer reprogram.
If hrtimer not configured as high resolution, hrtimer_enqueue_reprogram
do nothing and then make kvm lapic timer and tsc deadline timer fail.
That was back in 2012, where hrtimer_start_range_ns() would do the
reprogramming with hrtimer_enqueue_reprogram(). But as that was a nop with
high resolution timers disabled, this did not work. But a lot has changed
in the last 12 years.
For example, commit 49a2a07514a3a ("hrtimer: Kick lowres dynticks targets on
timer enqueue") modifies __hrtimer_start_range_ns() to work with low res
timers. There's been lots of other changes that make low res work.
ChromeOS has tested this before as well, and it hasn't seen any issues
with running KVM with high res timers disabled. There could be problems,
especially at low HZ, for guests that do not support kvmclock and rely
on precise delivery of periodic timers to keep their clock running.
This can be the APIC timer (provided by the kernel), the RTC (provided
by userspace), or the i8254 (choice of kernel/userspace). These guests
are few and far between these days, and in the case of the APIC timer +
Intel hosts we can use the preemption timer (which is TSC-based and has
better latency _and_ accuracy).
In KVM, only x86 is requiring CONFIG_HIGH_RES_TIMERS; perhaps a "depends
on HIGH_RES_TIMERS || EXPERT" could be added to virt/kvm, or a pr_warn
could be added to kvm_init if HIGH_RES_TIMERS are not enabled. But in
general, it seems that there must be other code in the kernel (maybe
sound/?) that is relying on having high-enough HZ or hrtimers but that's
not documented anywhere. Whenever you disable it you probably need to
know what you're doing and what your workload is; so the dependency is
not particularly interesting, and we can just remove it.
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Message-ID: <20240821095127.45d17b19@gandalf.local.home>
[Added the last two paragraphs to the commit message. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.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>
|
|
Rename the per-CPU hooks used to enable virtualization in hardware to
align with the KVM-wide helpers in kvm_main.c, and to better capture that
the callbacks are invoked on every online CPU.
No functional change intended.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Message-ID: <20240830043600.127750-5-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
