Currently, KVM for ARM64 maps at stage 2 memory that is considered device
(i.e. it is not RAM) with DEVICE_nGnRE memory attributes; this setting
overrides (as per the ARM architecture [1]) any device MMIO mapping
present at stage 1, resulting in a set-up whereby a guest operating
system cannot determine device MMIO mapping memory attributes on its
own but it is always overridden by the KVM stage 2 default.
This set-up does not allow guest operating systems to select device
memory attributes independently from KVM stage-2 mappings
(refer to [1], "Combining stage 1 and stage 2 memory type attributes"),
which turns out to be an issue in that guest operating systems
(e.g. Linux) may request to map devices MMIO regions with memory
attributes that guarantee better performance (e.g. gathering
attribute - that for some devices can generate larger PCIe memory
writes TLPs) and specific operations (e.g. unaligned transactions)
such as the NormalNC memory type.
The default device stage 2 mapping was chosen in KVM for ARM64 since
it was considered safer (i.e. it would not allow guests to trigger
uncontained failures ultimately crashing the machine) but this
turned out to be asynchronous (SError) defeating the purpose.
Failures containability is a property of the platform and is independent
from the memory type used for MMIO device memory mappings.
Actually, DEVICE_nGnRE memory type is even more problematic than
Normal-NC memory type in terms of faults containability in that e.g.
aborts triggered on DEVICE_nGnRE loads cannot be made, architecturally,
synchronous (i.e. that would imply that the processor should issue at
most 1 load transaction at a time - it cannot pipeline them - otherwise
the synchronous abort semantics would break the no-speculation attribute
attached to DEVICE_XXX memory).
This means that regardless of the combined stage1+stage2 mappings a
platform is safe if and only if device transactions cannot trigger
uncontained failures and that in turn relies on platform capabilities
and the device type being assigned (i.e. PCIe AER/DPC error containment
and RAS architecture[3]); therefore the default KVM device stage 2
memory attributes play no role in making device assignment safer
for a given platform (if the platform design adheres to design
guidelines outlined in [3]) and therefore can be relaxed.
For all these reasons, relax the KVM stage 2 device memory attributes
from DEVICE_nGnRE to Normal-NC.
The NormalNC was chosen over a different Normal memory type default
at stage-2 (e.g. Normal Write-through) to avoid cache allocation/snooping.
Relaxing S2 KVM device MMIO mappings to Normal-NC is not expected to
trigger any issue on guest device reclaim use cases either (i.e. device
MMIO unmap followed by a device reset) at least for PCIe devices, in that
in PCIe a device reset is architected and carried out through PCI config
space transactions that are naturally ordered with respect to MMIO
transactions according to the PCI ordering rules.
Having Normal-NC S2 default puts guests in control (thanks to
stage1+stage2 combined memory attributes rules [1]) of device MMIO
regions memory mappings, according to the rules described in [1]
and summarized here ([(S1) - stage1], [(S2) - stage 2]):
S1 | S2 | Result
NORMAL-WB | NORMAL-NC | NORMAL-NC
NORMAL-WT | NORMAL-NC | NORMAL-NC
NORMAL-NC | NORMAL-NC | NORMAL-NC
DEVICE<attr> | NORMAL-NC | DEVICE<attr>
It is worth noting that currently, to map devices MMIO space to user
space in a device pass-through use case the VFIO framework applies memory
attributes derived from pgprot_noncached() settings applied to VMAs, which
result in device-nGnRnE memory attributes for the stage-1 VMM mappings.
This means that a userspace mapping for device MMIO space carried
out with the current VFIO framework and a guest OS mapping for the same
MMIO space may result in a mismatched alias as described in [2].
Defaulting KVM device stage-2 mappings to Normal-NC attributes does not
change anything in this respect, in that the mismatched aliases would
only affect (refer to [2] for a detailed explanation) ordering between
the userspace and GuestOS mappings resulting stream of transactions
(i.e. it does not cause loss of property for either stream of
transactions on its own), which is harmless given that the userspace
and GuestOS access to the device is carried out through independent
transactions streams.
A Normal-NC flag is not present today. So add a new kvm_pgtable_prot
(KVM_PGTABLE_PROT_NORMAL_NC) flag for it, along with its
corresponding PTE value 0x5 (0b101) determined from [1].
Lastly, adapt the stage2 PTE property setter function
(stage2_set_prot_attr) to handle the NormalNC attribute.
The entire discussion leading to this patch series may be followed through
the following links.
Link: https://lore.kernel.org/all/20230907181459.18145-3-ankita@nvidia.com
Link: https://lore.kernel.org/r/20231205033015.10044-1-ankita@nvidia.com
[1] section D8.5.5 - DDI0487J_a_a-profile_architecture_reference_manual.pdf
[2] section B2.8 - DDI0487J_a_a-profile_architecture_reference_manual.pdf
[3] sections 1.7.7.3/1.8.5.2/appendix C - DEN0029H_SBSA_7.1.pdf
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Will Deacon <will@kernel.org>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Ankit Agrawal <ankita@nvidia.com>
Link: https://lore.kernel.org/r/20240224150546.368-2-ankita@nvidia.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
FEAT_LPA2 increases the maximum levels of translation from 4 to 5 for
the 4KB page case, when IA is >48 bits. While we can still use 4 levels
for stage2 translation in this case (due to stage2 allowing concatenated
page tables for first level lookup), the same kvm_pgtable library is
used for the hyp stage1 page tables and stage1 does not support
concatenation.
Therefore, modify the library to support up to 5 levels. Previous
patches already laid the groundwork for this by refactoring code to work
in terms of KVM_PGTABLE_FIRST_LEVEL and KVM_PGTABLE_LAST_LEVEL. So we
just need to change these macros.
The hardware sometimes encodes the new level differently from the
others: One such place is when reading the level from the FSC field in
the ESR_EL2 register. We never expect to see the lowest level (-1) here
since the stage 2 page tables always use concatenated tables for first
level lookup and therefore only use 4 levels of lookup. So we get away
with just adding a comment to explain why we are not being careful about
decoding level -1.
For stage2 VTCR_EL2.SL2 is introduced to encode the new start level.
However, since we always use concatenated page tables for first level
look up at stage2 (and therefore we will never need the new extra level)
we never touch this new field.
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20231127111737.1897081-10-ryan.roberts@arm.com
With the introduction of FEAT_LPA2, the Arm ARM adds a new level of
translation, level -1, so levels can now be in the range [-1;3]. 3 is
always the last level and the first level is determined based on the
number of VA bits in use.
Convert level variables to use a signed type in preparation for
supporting this new level -1.
Since the last level is always anchored at 3, and the first level varies
to suit the number of VA/IPA bits, take the opportunity to replace
KVM_PGTABLE_MAX_LEVELS with the 2 macros KVM_PGTABLE_FIRST_LEVEL and
KVM_PGTABLE_LAST_LEVEL. This removes the assumption from the code that
levels run from 0 to KVM_PGTABLE_MAX_LEVELS - 1, which will soon no
longer be true.
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20231127111737.1897081-9-ryan.roberts@arm.com
Implement a simple policy whereby if the HW supports FEAT_LPA2 for the
page size we are using, always use LPA2-style page-tables for stage 2
and hyp stage 1 (assuming an nvhe hyp), regardless of the VMM-requested
IPA size or HW-implemented PA size. When in use we can now support up to
52-bit IPA and PA sizes.
We use the previously created cpu feature to track whether LPA2 is
supported for deciding whether to use the LPA2 or classic pte format.
Note that FEAT_LPA2 brings support for bigger block mappings (512GB with
4KB, 64GB with 16KB). We explicitly don't enable these in the library
because stage2_apply_range() works on batch sizes of the largest used
block mapping, and increasing the size of the batch would lead to soft
lockups. See commit 5994bc9e05 ("KVM: arm64: Limit
stage2_apply_range() batch size to largest block").
With the addition of LPA2 support in the hypervisor, the PA size
supported by the HW must be capped with a runtime decision, rather than
simply using a compile-time decision based on PA_BITS. For example, on a
system that advertises 52 bit PA but does not support FEAT_LPA2, A 4KB
or 16KB kernel compiled with LPA2 support must still limit the PA size
to 48 bits.
Therefore, move the insertion of the PS field into TCR_EL2 out of
__kvm_hyp_init assembly code and instead do it in cpu_prepare_hyp_mode()
where the rest of TCR_EL2 is prepared. This allows us to figure out PS
with kvm_get_parange(), which has the appropriate logic to ensure the
above requirement. (and the PS field of VTCR_EL2 is already populated
this way).
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20231127111737.1897081-8-ryan.roberts@arm.com
Add stub functions which is initially always return false. These provide
the hooks that we need to update the range-based TLBI routines, whose
operands are encoded differently depending on whether lpa2 is enabled or
not.
The kernel and kvm will enable the use of lpa2 asynchronously in future,
and part of that enablement will involve fleshing out their respective
hook to advertise when it is using lpa2.
Since the kernel's decision to use lpa2 relies on more than just whether
the HW supports the feature, it can't just use the same static key as
kvm. This is another reason to use separate functions. lpa2_is_enabled()
is already implemented as part of Ard's kernel lpa2 series. Since kvm
will make its decision solely based on HW support, kvm_lpa2_is_enabled()
will be defined as system_supports_lpa2() once kvm starts using lpa2.
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20231127111737.1897081-3-ryan.roberts@arm.com
Implement the helper kvm_tlb_flush_vmid_range() that acts
as a wrapper for range-based TLB invalidations. For the
given VMID, use the range-based TLBI instructions to do
the job or fallback to invalidating all the TLB entries.
Signed-off-by: Raghavendra Rao Ananta <rananta@google.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230811045127.3308641-11-rananta@google.com
Userspace is allowed to select any PAGE_SIZE aligned hva to back guest
memory. This is even the case with hugepages, although it is a rather
suboptimal configuration as PTE level mappings are used at stage-2.
The arm64 page aging handlers have an assumption that the specified
range is exactly one page/block of memory, which in the aforementioned
case is not necessarily true. All together this leads to the WARN() in
kvm_age_gfn() firing.
However, the WARN is only part of the issue as the table walkers visit
at most a single leaf PTE. For hugepage-backed memory in a memslot that
isn't hugepage-aligned, page aging entirely misses accesses to the
hugepage beyond the first page in the memslot.
Add a new walker dedicated to handling page aging MMU notifiers capable
of walking a range of PTEs. Convert kvm(_test)_age_gfn() over to the new
walker and drop the WARN that caught the issue in the first place. The
implementation of this walker was inspired by the test_clear_young()
implementation by Yu Zhao [*], but repurposed to address a bug in the
existing aging implementation.
Cc: stable@vger.kernel.org # v5.15
Fixes: 056aad67f8 ("kvm: arm/arm64: Rework gpa callback handlers")
Link: https://lore.kernel.org/kvmarm/20230526234435.662652-6-yuzhao@google.com/
Co-developed-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: Reiji Watanabe <reijiw@google.com>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Link: https://lore.kernel.org/r/20230627235405.4069823-1-oliver.upton@linux.dev
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
- Eager page splitting optimization for dirty logging, optionally
allowing for a VM to avoid the cost of block splitting in the stage-2
fault path.
- Arm FF-A proxy for pKVM, allowing a pKVM host to safely interact with
services that live in the Secure world. pKVM intervenes on FF-A calls
to guarantee the host doesn't misuse memory donated to the hyp or a
pKVM guest.
- Support for running the split hypervisor with VHE enabled, known as
'hVHE' mode. This is extremely useful for testing the split
hypervisor on VHE-only systems, and paves the way for new use cases
that depend on having two TTBRs available at EL2.
- Generalized framework for configurable ID registers from userspace.
KVM/arm64 currently prevents arbitrary CPU feature set configuration
from userspace, but the intent is to relax this limitation and allow
userspace to select a feature set consistent with the CPU.
- Enable the use of Branch Target Identification (FEAT_BTI) in the
hypervisor.
- Use a separate set of pointer authentication keys for the hypervisor
when running in protected mode, as the host is untrusted at runtime.
- Ensure timer IRQs are consistently released in the init failure
paths.
- Avoid trapping CTR_EL0 on systems with Enhanced Virtualization Traps
(FEAT_EVT), as it is a register commonly read from userspace.
- Erratum workaround for the upcoming AmpereOne part, which has broken
hardware A/D state management.
As a consequence of the hVHE series reworking the arm64 software
features framework, the for-next/module-alloc branch from the arm64 tree
comes along for the ride.
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQSNXHjWXuzMZutrKNKivnWIJHzdFgUCZJWrhwAKCRCivnWIJHzd
Fs07AP9xliv5yIoQtRgPZXc0QDPyUm7zg8f5KDgqCVJtyHXcvAEAmmerBr39nbPc
XoMXALKx6NGt836P0C+bhvRcHdFPGwE=
=c/Xh
-----END PGP SIGNATURE-----
Merge tag 'kvmarm-6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD
KVM/arm64 updates for 6.5
- Eager page splitting optimization for dirty logging, optionally
allowing for a VM to avoid the cost of block splitting in the stage-2
fault path.
- Arm FF-A proxy for pKVM, allowing a pKVM host to safely interact with
services that live in the Secure world. pKVM intervenes on FF-A calls
to guarantee the host doesn't misuse memory donated to the hyp or a
pKVM guest.
- Support for running the split hypervisor with VHE enabled, known as
'hVHE' mode. This is extremely useful for testing the split
hypervisor on VHE-only systems, and paves the way for new use cases
that depend on having two TTBRs available at EL2.
- Generalized framework for configurable ID registers from userspace.
KVM/arm64 currently prevents arbitrary CPU feature set configuration
from userspace, but the intent is to relax this limitation and allow
userspace to select a feature set consistent with the CPU.
- Enable the use of Branch Target Identification (FEAT_BTI) in the
hypervisor.
- Use a separate set of pointer authentication keys for the hypervisor
when running in protected mode, as the host is untrusted at runtime.
- Ensure timer IRQs are consistently released in the init failure
paths.
- Avoid trapping CTR_EL0 on systems with Enhanced Virtualization Traps
(FEAT_EVT), as it is a register commonly read from userspace.
- Erratum workaround for the upcoming AmpereOne part, which has broken
hardware A/D state management.
As a consequence of the hVHE series reworking the arm64 software
features framework, the for-next/module-alloc branch from the arm64 tree
comes along for the ride.
* kvm-arm64/eager-page-splitting:
: Eager Page Splitting, courtesy of Ricardo Koller.
:
: Dirty logging performance is dominated by the cost of splitting
: hugepages to PTE granularity. On systems that mere mortals can get their
: hands on, each fault incurs the cost of a full break-before-make
: pattern, wherein the broadcast invalidation and ensuing serialization
: significantly increases fault latency.
:
: The goal of eager page splitting is to move the cost of hugepage
: splitting out of the stage-2 fault path and instead into the ioctls
: responsible for managing the dirty log:
:
: - If manual protection is enabled for the VM, hugepage splitting
: happens in the KVM_CLEAR_DIRTY_LOG ioctl. This is desirable as it
: provides userspace granular control over hugepage splitting.
:
: - Otherwise, if userspace relies on the legacy dirty log behavior
: (clear on collection), hugepage splitting is done at the moment dirty
: logging is enabled for a particular memslot.
:
: Support for eager page splitting requires explicit opt-in from
: userspace, which is realized through the
: KVM_CAP_ARM_EAGER_SPLIT_CHUNK_SIZE capability.
arm64: kvm: avoid overflow in integer division
KVM: arm64: Use local TLBI on permission relaxation
KVM: arm64: Split huge pages during KVM_CLEAR_DIRTY_LOG
KVM: arm64: Open-code kvm_mmu_write_protect_pt_masked()
KVM: arm64: Split huge pages when dirty logging is enabled
KVM: arm64: Add kvm_uninit_stage2_mmu()
KVM: arm64: Refactor kvm_arch_commit_memory_region()
KVM: arm64: Add kvm_pgtable_stage2_split()
KVM: arm64: Add KVM_CAP_ARM_EAGER_SPLIT_CHUNK_SIZE
KVM: arm64: Export kvm_are_all_memslots_empty()
KVM: arm64: Add helper for creating unlinked stage2 subtrees
KVM: arm64: Add KVM_PGTABLE_WALK flags for skipping CMOs and BBM TLBIs
KVM: arm64: Rename free_removed to free_unlinked
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
The preorder callback on the kvm_pgtable_stage2_map() path can replace
a table with a block, then recursively free the detached table. The
higher-level walking logic stashes the old page table entry and
then walks the freed table, invoking the leaf callback and
potentially freeing pgtable pages prematurely.
In normal operation, the call to tear down the detached stage-2
is indirected and uses an RCU callback to trigger the freeing.
RCU is not available to pKVM, which is where this bug is
triggered.
Change the behavior of the walker to reload the page table entry
after invoking the walker callback on preorder traversal, as it
does for leaf entries.
Tested on Pixel 6.
Fixes: 5c359cca1f ("KVM: arm64: Tear down unlinked stage-2 subtree after break-before-make")
Suggested-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230522103258.402272-1-tabba@google.com
Add a new stage2 function, kvm_pgtable_stage2_split(), for splitting a
range of huge pages. This will be used for eager-splitting huge pages
into PAGE_SIZE pages. The goal is to avoid having to split huge pages
on write-protection faults, and instead use this function to do it
ahead of time for large ranges (e.g., all guest memory in 1G chunks at
a time).
Signed-off-by: Ricardo Koller <ricarkol@google.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Link: https://lore.kernel.org/r/20230426172330.1439644-7-ricarkol@google.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Add a capability for userspace to specify the eager split chunk size.
The chunk size specifies how many pages to break at a time, using a
single allocation. Bigger the chunk size, more pages need to be
allocated ahead of time.
Suggested-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Ricardo Koller <ricarkol@google.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Link: https://lore.kernel.org/r/20230426172330.1439644-6-ricarkol@google.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Add a stage2 helper, kvm_pgtable_stage2_create_unlinked(), for
creating unlinked tables (which is the opposite of
kvm_pgtable_stage2_free_unlinked()). Creating an unlinked table is
useful for splitting level 1 and 2 entries into subtrees of PAGE_SIZE
PTEs. For example, a level 1 entry can be split into PAGE_SIZE PTEs
by first creating a fully populated tree, and then use it to replace
the level 1 entry in a single step. This will be used in a subsequent
commit for eager huge-page splitting (a dirty-logging optimization).
Signed-off-by: Ricardo Koller <ricarkol@google.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Link: https://lore.kernel.org/r/20230426172330.1439644-4-ricarkol@google.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Add two flags to kvm_pgtable_visit_ctx, KVM_PGTABLE_WALK_SKIP_BBM_TLBI
and KVM_PGTABLE_WALK_SKIP_CMO, to indicate that the walk should not
perform TLB invalidations (TLBIs) in break-before-make (BBM) nor cache
maintenance operations (CMO). This will be used by a future commit to
create unlinked tables not accessible to the HW page-table walker.
Signed-off-by: Ricardo Koller <ricarkol@google.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Link: https://lore.kernel.org/r/20230426172330.1439644-3-ricarkol@google.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Normalize on referring to tables outside of an active paging structure
as 'unlinked'.
A subsequent change to KVM will add support for building page tables
that are not part of an active paging structure. The existing
'removed_table' terminology is quite clunky when applied in this
context.
Signed-off-by: Ricardo Koller <ricarkol@google.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Link: https://lore.kernel.org/r/20230426172330.1439644-2-ricarkol@google.com
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Until now, the page table walker counted increments to the PA and IPA
of a walk in two separate places. While the PA is incremented as soon as
a leaf PTE is installed in stage2_map_walker_try_leaf(), the IPA is
actually bumped in the generic table walker context. Critically,
__kvm_pgtable_visit() rereads the PTE after the LEAF callback returns
to work out if a table or leaf was installed, and only bumps the IPA for
a leaf PTE.
This arrangement worked fine when we handled faults behind the write lock,
as the walker had exclusive access to the stage-2 page tables. However,
commit 1577cb5823 ("KVM: arm64: Handle stage-2 faults in parallel")
started handling all stage-2 faults behind the read lock, opening up a
race where a walker could increment the PA but not the IPA of a walk.
Nothing good ensues, as the walker starts mapping with the incorrect
IPA -> PA relationship.
For example, assume that two vCPUs took a data abort on the same IPA.
One observes that dirty logging is disabled, and the other observed that
it is enabled:
vCPU attempting PMD mapping vCPU attempting PTE mapping
====================================== =====================================
/* install PMD */
stage2_make_pte(ctx, leaf);
data->phys += granule;
/* replace PMD with a table */
stage2_try_break_pte(ctx, data->mmu);
stage2_make_pte(ctx, table);
/* table is observed */
ctx.old = READ_ONCE(*ptep);
table = kvm_pte_table(ctx.old, level);
/*
* map walk continues w/o incrementing
* IPA.
*/
__kvm_pgtable_walk(..., level + 1);
Bring an end to the whole mess by using the IPA as the single source of
truth for how far along a walk has gotten. Work out the correct PA to
map by calculating the IPA offset from the beginning of the walk and add
that to the starting physical address.
Cc: stable@vger.kernel.org
Fixes: 1577cb5823 ("KVM: arm64: Handle stage-2 faults in parallel")
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230421071606.1603916-2-oliver.upton@linux.dev
The page table walkers are invoked outside fault handling paths, such as
write protecting a range of memory. EAGAIN is generally used by the
walkers to retry execution due to races on a particular PTE, like taking
an access fault on a PTE being invalidated from another thread.
This early return behavior is undesirable for walkers that operate
outside a fault handler. Suppress EAGAIN and continue the walk if
operating outside a fault handler.
Link: https://lore.kernel.org/r/20221202185156.696189-3-oliver.upton@linux.dev
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
* kvm-arm64/pkvm-vcpu-state: (25 commits)
: .
: Large drop of pKVM patches from Will Deacon and co, adding
: a private vm/vcpu state at EL2, managed independently from
: the EL1 state. From the cover letter:
:
: "This is version six of the pKVM EL2 state series, extending the pKVM
: hypervisor code so that it can dynamically instantiate and manage VM
: data structures without the host being able to access them directly.
: These structures consist of a hyp VM, a set of hyp vCPUs and the stage-2
: page-table for the MMU. The pages used to hold the hypervisor structures
: are returned to the host when the VM is destroyed."
: .
KVM: arm64: Use the pKVM hyp vCPU structure in handle___kvm_vcpu_run()
KVM: arm64: Don't unnecessarily map host kernel sections at EL2
KVM: arm64: Explicitly map 'kvm_vgic_global_state' at EL2
KVM: arm64: Maintain a copy of 'kvm_arm_vmid_bits' at EL2
KVM: arm64: Unmap 'kvm_arm_hyp_percpu_base' from the host
KVM: arm64: Return guest memory from EL2 via dedicated teardown memcache
KVM: arm64: Instantiate guest stage-2 page-tables at EL2
KVM: arm64: Consolidate stage-2 initialisation into a single function
KVM: arm64: Add generic hyp_memcache helpers
KVM: arm64: Provide I-cache invalidation by virtual address at EL2
KVM: arm64: Initialise hypervisor copies of host symbols unconditionally
KVM: arm64: Add per-cpu fixmap infrastructure at EL2
KVM: arm64: Instantiate pKVM hypervisor VM and vCPU structures from EL1
KVM: arm64: Add infrastructure to create and track pKVM instances at EL2
KVM: arm64: Rename 'host_kvm' to 'host_mmu'
KVM: arm64: Add hyp_spinlock_t static initializer
KVM: arm64: Include asm/kvm_mmu.h in nvhe/mem_protect.h
KVM: arm64: Add helpers to pin memory shared with the hypervisor at EL2
KVM: arm64: Prevent the donation of no-map pages
KVM: arm64: Implement do_donate() helper for donating memory
...
Signed-off-by: Marc Zyngier <maz@kernel.org>
Exclusive table walks are the only supported table walk in the hyp, as
there is no construct like RCU available in the hypervisor code. Reject
any attempt to do a shared table walk by returning an error and allowing
the caller to clean up the mess.
Suggested-by: Will Deacon <will@kernel.org>
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221118182222.3932898-4-oliver.upton@linux.dev
Rather than passing through the state of the KVM_PGTABLE_WALK_SHARED
flag, just take a pointer to the whole walker structure instead. Move
around struct kvm_pgtable and the RCU indirection such that the
associated ifdeffery remains in one place while ensuring the walker +
flags definitions precede their use.
No functional change intended.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221118182222.3932898-2-oliver.upton@linux.dev
Mapping pages in a guest page-table from within the pKVM hypervisor at
EL2 may require cache maintenance to ensure that the initialised page
contents is visible even to non-cacheable (e.g. MMU-off) accesses from
the guest.
In preparation for performing this maintenance at EL2, introduce a
per-vCPU fixmap which allows the pKVM hypervisor to map guest pages
temporarily into its stage-1 page-table for the purposes of cache
maintenance and, in future, poisoning on the reclaim path. The use of a
fixmap avoids the need for memory allocation or locking on the map()
path.
Tested-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Quentin Perret <qperret@google.com>
Co-developed-by: Will Deacon <will@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221110190259.26861-15-will@kernel.org
Introduce a global table (and lock) to track pKVM instances at EL2, and
provide hypercalls that can be used by the untrusted host to create and
destroy pKVM VMs and their vCPUs. pKVM VM/vCPU state is directly
accessible only by the trusted hypervisor (EL2).
Each pKVM VM is directly associated with an untrusted host KVM instance,
and is referenced by the host using an opaque handle. Future patches
will provide hypercalls to allow the host to initialize/set/get pKVM
VM/vCPU state using the opaque handle.
Tested-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Co-developed-by: Will Deacon <will@kernel.org>
Signed-off-by: Will Deacon <will@kernel.org>
[maz: silence warning on unmap_donated_memory_noclear()]
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221110190259.26861-13-will@kernel.org
The stage-2 map walker has been made parallel-aware, and as such can be
called while only holding the read side of the MMU lock. Rip out the
conditional locking in user_mem_abort() and instead grab the read lock.
Continue to take the write lock from other callsites to
kvm_pgtable_stage2_map().
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107220033.1895655-1-oliver.upton@linux.dev
Use RCU to safely walk the stage-2 page tables in parallel. Acquire and
release the RCU read lock when traversing the page tables. Defer the
freeing of table memory to an RCU callback. Indirect the calls into RCU
and provide stubs for hypervisor code, as RCU is not available in such a
context.
The RCU protection doesn't amount to much at the moment, as readers are
already protected by the read-write lock (all walkers that free table
memory take the write lock). Nonetheless, a subsequent change will
futher relax the locking requirements around the stage-2 MMU, thereby
depending on RCU.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107215644.1895162-9-oliver.upton@linux.dev
The break-before-make sequence is a bit annoying as it opens a window
wherein memory is unmapped from the guest. KVM should replace the PTE
as quickly as possible and avoid unnecessary work in between.
Presently, the stage-2 map walker tears down a removed table before
installing a block mapping when coalescing a table into a block. As the
removed table is no longer visible to hardware walkers after the
DSB+TLBI, it is possible to move the remaining cleanup to happen after
installing the new PTE.
Reshuffle the stage-2 map walker to install the new block entry in
the pre-order callback. Unwire all of the teardown logic and replace
it with a call to kvm_pgtable_stage2_free_removed() after fixing
the PTE. The post-order visitor is now completely unnecessary, so drop
it. Finally, touch up the comments to better represent the now
simplified map walker.
Note that the call to tear down the unlinked stage-2 is indirected
as a subsequent change will use an RCU callback to trigger tear down.
RCU is not available to pKVM, so there is a need to use different
implementations on pKVM and non-pKVM VMs.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Ben Gardon <bgardon@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107215644.1895162-8-oliver.upton@linux.dev
Use an opaque type for pteps and require visitors explicitly dereference
the pointer before using. Protecting page table memory with RCU requires
that KVM dereferences RCU-annotated pointers before using. However, RCU
is not available for use in the nVHE hypervisor and the opaque type can
be conditionally annotated with RCU for the stage-2 MMU.
Call the type a 'pteref' to avoid a naming collision with raw pteps. No
functional change intended.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107215644.1895162-7-oliver.upton@linux.dev
A subsequent change to KVM will move the tear down of an unlinked
stage-2 subtree out of the critical path of the break-before-make
sequence.
Introduce a new helper for tearing down unlinked stage-2 subtrees.
Leverage the existing stage-2 free walkers to do so, with a deep call
into __kvm_pgtable_walk() as the subtree is no longer reachable from the
root.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107215644.1895162-6-oliver.upton@linux.dev
As a prerequisite for getting visitors off of struct kvm_pgtable, pass
mm_ops through the visitor context.
No functional change intended.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Ben Gardon <bgardon@google.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107215644.1895162-4-oliver.upton@linux.dev
Rather than reading the ptep all over the shop, read the ptep once from
__kvm_pgtable_visit() and stick it in the visitor context. Reread the
ptep after visiting a leaf in case the callback installed a new table
underneath.
No functional change intended.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Ben Gardon <bgardon@google.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107215644.1895162-3-oliver.upton@linux.dev
Passing new arguments by value to the visitor callbacks is extremely
inflexible for stuffing new parameters used by only some of the
visitors. Use a context structure instead and pass the pointer through
to the visitor callback.
While at it, redefine the 'flags' parameter to the visitor to contain
the bit indicating the phase of the walk. Pass the entire set of flags
through the context structure such that the walker can communicate
additional state to the visitor callback.
No functional change intended.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Ben Gardon <bgardon@google.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221107215644.1895162-2-oliver.upton@linux.dev
Work out the minimum page table level where KVM supports block mappings
at compile time. While at it, rewrite the comment around supported block
mappings to directly describe what KVM supports instead of phrasing in
terms of what it does not.
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221007234151.461779-2-oliver.upton@linux.dev
Normally we include the full register name in the defines for fields within
registers but this has not been followed for ID registers. In preparation
for automatic generation of defines add the _EL1s into the defines for
ID_AA64MMFR0_EL1 to follow the convention. No functional changes.
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Kristina Martsenko <kristina.martsenko@arm.com>
Link: https://lore.kernel.org/r/20220905225425.1871461-5-broonie@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Ganapatrao reported that the kvm_pgtable->mmu pointer is more or
less hardcoded to the main S2 mmu structure, while the nested
code needs it to point to other instances (as we have one instance
per nested context).
Rework the initialisation of the kvm_pgtable structure so that
this assumtion doesn't hold true anymore. This requires some
minor changes to the order in which things are initialised
(the mmu->arch pointer being the critical one).
Reported-by: Ganapatrao Kulkarni <gankulkarni@os.amperecomputing.com>
Reviewed-by: Ganapatrao Kulkarni <gankulkarni@os.amperecomputing.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20211129200150.351436-5-maz@kernel.org
Implement kvm_pgtable_hyp_unmap() which can be used to remove hypervisor
stage-1 mappings at EL2.
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20211215161232.1480836-6-qperret@google.com
Introduce helper functions in the KVM stage-2 and stage-1 page-table
manipulation library allowing to retrieve the enum kvm_pgtable_prot of a
PTE. This will be useful to implement custom walkers outside of
pgtable.c.
Signed-off-by: Quentin Perret <qperret@google.com>
Reviewed-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210809152448.1810400-17-qperret@google.com
Introduce infrastructure allowing to manipulate software bits in stage-1
and stage-2 page-tables using additional entries in the kvm_pgtable_prot
enum.
This is heavily inspired by Marc's implementation of a similar feature
in the NV patch series, but adapted to allow stage-1 changes as well:
https://lore.kernel.org/kvmarm/20210510165920.1913477-56-maz@kernel.org/
Suggested-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Reviewed-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210809152448.1810400-12-qperret@google.com
Much of the stage-2 manipulation logic relies on being able to destroy
block mappings if e.g. installing a smaller mapping in the range. The
rationale for this behaviour is that stage-2 mappings can always be
re-created lazily. However, this gets more complicated when the stage-2
page-table is used to store metadata about the underlying pages. In such
cases, destroying a block mapping may lead to losing part of the state,
and confuse the user of those metadata (such as the hypervisor in nVHE
protected mode).
To avoid this, introduce a callback function in the pgtable struct which
is called during all map operations to determine whether the mappings
can use blocks, or should be forced to page granularity. This is used by
the hypervisor when creating the host stage-2 to force page-level
mappings when using non-default protection attributes.
Signed-off-by: Quentin Perret <qperret@google.com>
Reviewed-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210809152448.1810400-11-qperret@google.com
The kvm_pgtable_stage2_find_range() function is used in the host memory
abort path to try and look for the largest block mapping that can be
used to map the faulting address. In order to do so, the function
currently walks the stage-2 page-table and looks for existing
incompatible mappings within the range of the largest possible block.
If incompatible mappings are found, it tries the same procedure again,
but using a smaller block range, and repeats until a matching range is
found (potentially up to page granularity). While this approach has
benefits (mostly in the fact that it proactively coalesces host stage-2
mappings), it can be slow if the ranges are fragmented, and it isn't
optimized to deal with CPUs faulting on the same IPA as all of them will
do all the work every time.
To avoid these issues, remove kvm_pgtable_stage2_find_range(), and walk
the page-table only once in the host_mem_abort() path to find the
closest leaf to the input address. With this, use the corresponding
range if it is invalid and not owned by another entity. If a valid leaf
is found, return -EAGAIN similar to what is done in the
kvm_pgtable_stage2_map() path to optimize concurrent faults.
Reviewed-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210809152448.1810400-7-qperret@google.com
The KVM pgtable API exposes the kvm_pgtable_walk() function to allow
the definition of walkers outside of pgtable.c. However, it is not easy
to implement any of those walkers without some of the low-level helpers.
Move some of them to the header file to allow re-use from other places.
Signed-off-by: Quentin Perret <qperret@google.com>
Reviewed-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210809152448.1810400-6-qperret@google.com
It is becoming a common need to fetch the PTE for a given address
together with its level. Add such a helper.
Signed-off-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Quentin Perret <qperret@google.com>
Reviewed-by: Alexandru Elisei <alexandru.elisei@arm.com>
Link: https://lore.kernel.org/r/20210726153552.1535838-2-maz@kernel.org
To prepare for performing CMOs for guest stage-2 in the fault handlers
in pgtable.c, here introduce two cache maintenance callbacks in struct
kvm_pgtable_mm_ops. We also adjust the comment alignment for the
existing part but make no real content change at all.
Reviewed-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
[maz: fixed up comments and renamed callbacks]
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210617105824.31752-2-wangyanan55@huawei.com
Introduce a new stage 2 configuration flag to specify that all mappings
in a given page-table will be identity-mapped, as will be the case for
the host. This allows to introduce sanity checks in the map path and to
avoid programming errors.
Suggested-by: Will Deacon <will@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210319100146.1149909-34-qperret@google.com
In order to further configure stage 2 page-tables, pass flags to the
init function using a new enum.
The first of these flags allows to disable FWB even if the hardware
supports it as we will need to do so for the host stage 2.
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210319100146.1149909-33-qperret@google.com
Since the host stage 2 will be identity mapped, and since it will own
most of memory, it would preferable for performance to try and use large
block mappings whenever that is possible. To ease this, introduce a new
helper in the KVM page-table code which allows to search for large
ranges of available IPA space. This will be used in the host memory
abort path to greedily idmap large portion of the PA space.
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210319100146.1149909-32-qperret@google.com
As the host stage 2 will be identity mapped, all the .hyp memory regions
and/or memory pages donated to protected guestis will have to marked
invalid in the host stage 2 page-table. At the same time, the hypervisor
will need a way to track the ownership of each physical page to ensure
memory sharing or donation between entities (host, guests, hypervisor) is
legal.
In order to enable this tracking at EL2, let's use the host stage 2
page-table itself. The idea is to use the top bits of invalid mappings
to store the unique identifier of the page owner. The page-table owner
(the host) gets identifier 0 such that, at boot time, it owns the entire
IPA space as the pgd starts zeroed.
Provide kvm_pgtable_stage2_set_owner() which allows to modify the
ownership of pages in the host stage 2. It re-uses most of the map()
logic, but ends up creating invalid mappings instead. This impacts
how we do refcount as we now need to count invalid mappings when they
are used for ownership tracking.
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210319100146.1149909-30-qperret@google.com
The current stage2 page-table allocator uses a memcache to get
pre-allocated pages when it needs any. To allow re-using this code at
EL2 which uses a concept of memory pools, make the memcache argument of
kvm_pgtable_stage2_map() anonymous, and let the mm_ops zalloc_page()
callbacks use it the way they need to.
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210319100146.1149909-26-qperret@google.com
In order to re-use some of the stage 2 setup code at EL2, factor parts
of kvm_arm_setup_stage2() out into separate functions.
No functional change intended.
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210319100146.1149909-23-qperret@google.com