Add support for AMD SEV and SEV-ES intra-host migration support. Intra
host migration provides a low-cost mechanism for userspace VMM upgrades.
In the common case for intra host migration, we can rely on the normal
ioctls for passing data from one VMM to the next. SEV, SEV-ES, and other
confidential compute environments make most of this information opaque, and
render KVM ioctls such as "KVM_GET_REGS" irrelevant. As a result, we need
the ability to pass this opaque metadata from one VMM to the next. The
easiest way to do this is to leave this data in the kernel, and transfer
ownership of the metadata from one KVM VM (or vCPU) to the next. In-kernel
hand off makes it possible to move any data that would be
unsafe/impossible for the kernel to hand directly to userspace, and
cannot be reproduced using data that can be handed to userspace.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Pull char/misc driver updates from Greg KH:
"Here is the big set of char and misc and other tiny driver subsystem
updates for 5.16-rc1.
Loads of things in here, all of which have been in linux-next for a
while with no reported problems (except for one called out below.)
Included are:
- habanana labs driver updates, including dma_buf usage, reviewed and
acked by the dma_buf maintainers
- iio driver update (going through this tree not staging as they
really do not belong going through that tree anymore)
- counter driver updates
- hwmon driver updates that the counter drivers needed, acked by the
hwmon maintainer
- xillybus driver updates
- binder driver updates
- extcon driver updates
- dma_buf module namespaces added (will cause a build error in arm64
for allmodconfig, but that change is on its way through the drm
tree)
- lkdtm driver updates
- pvpanic driver updates
- phy driver updates
- virt acrn and nitr_enclaves driver updates
- smaller char and misc driver updates"
* tag 'char-misc-5.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc: (386 commits)
comedi: dt9812: fix DMA buffers on stack
comedi: ni_usb6501: fix NULL-deref in command paths
arm64: errata: Enable TRBE workaround for write to out-of-range address
arm64: errata: Enable workaround for TRBE overwrite in FILL mode
coresight: trbe: Work around write to out of range
coresight: trbe: Make sure we have enough space
coresight: trbe: Add a helper to determine the minimum buffer size
coresight: trbe: Workaround TRBE errata overwrite in FILL mode
coresight: trbe: Add infrastructure for Errata handling
coresight: trbe: Allow driver to choose a different alignment
coresight: trbe: Decouple buffer base from the hardware base
coresight: trbe: Add a helper to pad a given buffer area
coresight: trbe: Add a helper to calculate the trace generated
coresight: trbe: Defer the probe on offline CPUs
coresight: trbe: Fix incorrect access of the sink specific data
coresight: etm4x: Add ETM PID for Kryo-5XX
coresight: trbe: Prohibit trace before disabling TRBE
coresight: trbe: End the AUX handle on truncation
coresight: trbe: Do not truncate buffer on IRQ
coresight: trbe: Fix handling of spurious interrupts
...
Pull documentation updates from Jonathan Corbet:
"This is a relatively unexciting cycle for documentation.
- Some small scripts/kerneldoc fixes
- More Chinese translation work, but at a much reduced rate.
- The tip-tree maintainer's handbook
...plus the usual array of build fixes, typo fixes, etc"
* tag 'docs-5.16' of git://git.lwn.net/linux: (53 commits)
kernel-doc: support DECLARE_PHY_INTERFACE_MASK()
docs/zh_CN: add core-api xarray translation
docs/zh_CN: add core-api assoc_array translation
speakup: Fix typo in documentation "boo" -> "boot"
docs: submitting-patches: make section about the Link: tag more explicit
docs: deprecated.rst: Clarify open-coded arithmetic with literals
scripts: documentation-file-ref-check: fix bpf selftests path
scripts: documentation-file-ref-check: ignore hidden files
coding-style.rst: trivial: fix location of driver model macros
docs: f2fs: fix text alignment
docs/zh_CN add PCI pci.rst translation
docs/zh_CN add PCI index.rst translation
docs: translations: zh_CN: memory-hotplug.rst: fix a typo
docs: translations: zn_CN: irq-affinity.rst: add a missing extension
block: add documentation for inflight
scripts: kernel-doc: Ignore __alloc_size() attribute
docs: pdfdocs: Adjust \headheight for fancyhdr
docs: UML: user_mode_linux_howto_v2 edits
docs: use the lore redirector everywhere
docs: proc.rst: mountinfo: align columns
...
To date, VMM-directed TSC synchronization and migration has been a bit
messy. KVM has some baked-in heuristics around TSC writes to infer if
the VMM is attempting to synchronize. This is problematic, as it depends
on host userspace writing to the guest's TSC within 1 second of the last
write.
A much cleaner approach to configuring the guest's views of the TSC is to
simply migrate the TSC offset for every vCPU. Offsets are idempotent,
and thus not subject to change depending on when the VMM actually
reads/writes values from/to KVM. The VMM can then read the TSC once with
KVM_GET_CLOCK to capture a (realtime, host_tsc) pair at the instant when
the guest is paused.
Cc: David Matlack <dmatlack@google.com>
Cc: Sean Christopherson <seanjc@google.com>
Signed-off-by: Oliver Upton <oupton@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20210916181538.968978-8-oupton@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Handling the migration of TSCs correctly is difficult, in part because
Linux does not provide userspace with the ability to retrieve a (TSC,
realtime) clock pair for a single instant in time. In lieu of a more
convenient facility, KVM can report similar information in the kvm_clock
structure.
Provide userspace with a host TSC & realtime pair iff the realtime clock
is based on the TSC. If userspace provides KVM_SET_CLOCK with a valid
realtime value, advance the KVM clock by the amount of elapsed time. Do
not step the KVM clock backwards, though, as it is a monotonic
oscillator.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Oliver Upton <oupton@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20210916181538.968978-5-oupton@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Initial KVM RISC-V support
Following features are supported by the initial KVM RISC-V support:
1. No RISC-V specific KVM IOCTL
2. Loadable KVM RISC-V module
3. Minimal possible KVM world-switch which touches only GPRs and few CSRs
4. Works on both RV64 and RV32 host
5. Full Guest/VM switch via vcpu_get/vcpu_put infrastructure
6. KVM ONE_REG interface for VCPU register access from KVM user-space
7. Interrupt controller emulation in KVM user-space
8. Timer and IPI emuation in kernel
9. Both Sv39x4 and Sv48x4 supported for RV64 host
10. MMU notifiers supported
11. Generic dirty log supported
12. FP lazy save/restore supported
13. SBI v0.1 emulation for Guest/VM
14. Forward unhandled SBI calls to KVM user-space
15. Hugepage support for Guest/VM
16. IOEVENTFD support for Vhost
Document RISC-V specific parts of the KVM API, such as:
- The interrupt numbers passed to the KVM_INTERRUPT ioctl.
- The states supported by the KVM_{GET,SET}_MP_STATE ioctls.
- The registers supported by the KVM_{GET,SET}_ONE_REG interface
and the encoding of those register ids.
- The exit reason KVM_EXIT_RISCV_SBI for SBI calls forwarded to
userspace tool.
CC: Jonathan Corbet <corbet@lwn.net>
CC: linux-doc@vger.kernel.org
Signed-off-by: Anup Patel <anup.patel@wdc.com>
Acked-by: Palmer Dabbelt <palmerdabbelt@google.com>
KVM_MAX_VCPU_ID is not specifying the highest allowed vcpu-id, but the
number of allowed vcpu-ids. This has already led to confusion, so
rename KVM_MAX_VCPU_ID to KVM_MAX_VCPU_IDS to make its semantics more
clear
Suggested-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20210913135745.13944-3-jgross@suse.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Pull KVM updates from Paolo Bonzini:
"ARM:
- Page ownership tracking between host EL1 and EL2
- Rely on userspace page tables to create large stage-2 mappings
- Fix incompatibility between pKVM and kmemleak
- Fix the PMU reset state, and improve the performance of the virtual
PMU
- Move over to the generic KVM entry code
- Address PSCI reset issues w.r.t. save/restore
- Preliminary rework for the upcoming pKVM fixed feature
- A bunch of MM cleanups
- a vGIC fix for timer spurious interrupts
- Various cleanups
s390:
- enable interpretation of specification exceptions
- fix a vcpu_idx vs vcpu_id mixup
x86:
- fast (lockless) page fault support for the new MMU
- new MMU now the default
- increased maximum allowed VCPU count
- allow inhibit IRQs on KVM_RUN while debugging guests
- let Hyper-V-enabled guests run with virtualized LAPIC as long as
they do not enable the Hyper-V "AutoEOI" feature
- fixes and optimizations for the toggling of AMD AVIC (virtualized
LAPIC)
- tuning for the case when two-dimensional paging (EPT/NPT) is
disabled
- bugfixes and cleanups, especially with respect to vCPU reset and
choosing a paging mode based on CR0/CR4/EFER
- support for 5-level page table on AMD processors
Generic:
- MMU notifier invalidation callbacks do not take mmu_lock unless
necessary
- improved caching of LRU kvm_memory_slot
- support for histogram statistics
- add statistics for halt polling and remote TLB flush requests"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (210 commits)
KVM: Drop unused kvm_dirty_gfn_invalid()
KVM: x86: Update vCPU's hv_clock before back to guest when tsc_offset is adjusted
KVM: MMU: mark role_regs and role accessors as maybe unused
KVM: MIPS: Remove a "set but not used" variable
x86/kvm: Don't enable IRQ when IRQ enabled in kvm_wait
KVM: stats: Add VM stat for remote tlb flush requests
KVM: Remove unnecessary export of kvm_{inc,dec}_notifier_count()
KVM: x86/mmu: Move lpage_disallowed_link further "down" in kvm_mmu_page
KVM: x86/mmu: Relocate kvm_mmu_page.tdp_mmu_page for better cache locality
Revert "KVM: x86: mmu: Add guest physical address check in translate_gpa()"
KVM: x86/mmu: Remove unused field mmio_cached in struct kvm_mmu_page
kvm: x86: Increase KVM_SOFT_MAX_VCPUS to 710
kvm: x86: Increase MAX_VCPUS to 1024
kvm: x86: Set KVM_MAX_VCPU_ID to 4*KVM_MAX_VCPUS
KVM: VMX: avoid running vmx_handle_exit_irqoff in case of emulation
KVM: x86/mmu: Don't freak out if pml5_root is NULL on 4-level host
KVM: s390: index kvm->arch.idle_mask by vcpu_idx
KVM: s390: Enable specification exception interpretation
KVM: arm64: Trim guest debug exception handling
KVM: SVM: Add 5-level page table support for SVM
...
Pull documentation updates from Jonathan Corbet:
"Yet another set of documentation changes:
- A reworking of PDF generation to yield better results for documents
using CJK fonts in particular.
- A new set of translations into traditional Chinese, a dialect for
which I am assured there is a community of interested readers.
- A lot more regular Chinese translation work as well.
... plus the usual assortment of updates, fixes, typo tweaks, etc"
* tag 'docs-5.15' of git://git.lwn.net/linux: (55 commits)
docs: sphinx-requirements: Move sphinx_rtd_theme to top
docs: pdfdocs: Enable language-specific font choice of zh_TW translations
docs: pdfdocs: Teach xeCJK about character classes of quotation marks
docs: pdfdocs: Permit AutoFakeSlant for CJK fonts
docs: pdfdocs: One-half spacing for CJK translations
docs: pdfdocs: Add conf.py local to translations for ascii-art alignment
docs: pdfdocs: Preserve inter-phrase space in Korean translations
docs: pdfdocs: Choose Serif font as CJK mainfont if possible
docs: pdfdocs: Add CJK-language-specific font settings
docs: pdfdocs: Refactor config for CJK document
scripts/kernel-doc: Override -Werror from KCFLAGS with KDOC_WERROR
docs/zh_CN: Add zh_CN/accounting/psi.rst
doc: align Italian translation
Documentation/features/vm: riscv supports THP now
docs/zh_CN: add infiniband user_verbs translation
docs/zh_CN: add infiniband user_mad translation
docs/zh_CN: add infiniband tag_matching translation
docs/zh_CN: add infiniband sysfs translation
docs/zh_CN: add infiniband opa_vnic translation
docs/zh_CN: add infiniband ipoib translation
...
KVM_GUESTDBG_BLOCKIRQ will allow KVM to block all interrupts
while running.
This change is mostly intended for more robust single stepping
of the guest and it has the following benefits when enabled:
* Resuming from a breakpoint is much more reliable.
When resuming execution from a breakpoint, with interrupts enabled,
more often than not, KVM would inject an interrupt and make the CPU
jump immediately to the interrupt handler and eventually return to
the breakpoint, to trigger it again.
From the user point of view it looks like the CPU never executed a
single instruction and in some cases that can even prevent forward
progress, for example, when the breakpoint is placed by an automated
script (e.g lx-symbols), which does something in response to the
breakpoint and then continues the guest automatically.
If the script execution takes enough time for another interrupt to
arrive, the guest will be stuck on the same breakpoint RIP forever.
* Normal single stepping is much more predictable, since it won't
land the debugger into an interrupt handler.
* RFLAGS.TF has less chance to be leaked to the guest:
We set that flag behind the guest's back to do single stepping
but if single step lands us into an interrupt/exception handler
it will be leaked to the guest in the form of being pushed
to the stack.
This doesn't completely eliminate this problem as exceptions
can still happen, but at least this reduces the chances
of this happening.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20210811122927.900604-6-mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add yet another spinlock for the TDP MMU and take it when marking indirect
shadow pages unsync. When using the TDP MMU and L1 is running L2(s) with
nested TDP, KVM may encounter shadow pages for the TDP entries managed by
L1 (controlling L2) when handling a TDP MMU page fault. The unsync logic
is not thread safe, e.g. the kvm_mmu_page fields are not atomic, and
misbehaves when a shadow page is marked unsync via a TDP MMU page fault,
which runs with mmu_lock held for read, not write.
Lack of a critical section manifests most visibly as an underflow of
unsync_children in clear_unsync_child_bit() due to unsync_children being
corrupted when multiple CPUs write it without a critical section and
without atomic operations. But underflow is the best case scenario. The
worst case scenario is that unsync_children prematurely hits '0' and
leads to guest memory corruption due to KVM neglecting to properly sync
shadow pages.
Use an entirely new spinlock even though piggybacking tdp_mmu_pages_lock
would functionally be ok. Usurping the lock could degrade performance when
building upper level page tables on different vCPUs, especially since the
unsync flow could hold the lock for a comparatively long time depending on
the number of indirect shadow pages and the depth of the paging tree.
For simplicity, take the lock for all MMUs, even though KVM could fairly
easily know that mmu_lock is held for write. If mmu_lock is held for
write, there cannot be contention for the inner spinlock, and marking
shadow pages unsync across multiple vCPUs will be slow enough that
bouncing the kvm_arch cacheline should be in the noise.
Note, even though L2 could theoretically be given access to its own EPT
entries, a nested MMU must hold mmu_lock for write and thus cannot race
against a TDP MMU page fault. I.e. the additional spinlock only _needs_ to
be taken by the TDP MMU, as opposed to being taken by any MMU for a VM
that is running with the TDP MMU enabled. Holding mmu_lock for read also
prevents the indirect shadow page from being freed. But as above, keep
it simple and always take the lock.
Alternative #1, the TDP MMU could simply pass "false" for can_unsync and
effectively disable unsync behavior for nested TDP. Write protecting leaf
shadow pages is unlikely to noticeably impact traditional L1 VMMs, as such
VMMs typically don't modify TDP entries, but the same may not hold true for
non-standard use cases and/or VMMs that are migrating physical pages (from
L1's perspective).
Alternative #2, the unsync logic could be made thread safe. In theory,
simply converting all relevant kvm_mmu_page fields to atomics and using
atomic bitops for the bitmap would suffice. However, (a) an in-depth audit
would be required, (b) the code churn would be substantial, and (c) legacy
shadow paging would incur additional atomic operations in performance
sensitive paths for no benefit (to legacy shadow paging).
Fixes: a2855afc7e ("KVM: x86/mmu: Allow parallel page faults for the TDP MMU")
Cc: stable@vger.kernel.org
Cc: Ben Gardon <bgardon@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20210812181815.3378104-1-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We would like to avoid taking mmu_lock for .invalidate_range_{start,end}()
notifications that are unrelated to KVM. Because mmu_notifier_count
must be modified while holding mmu_lock for write, and must always
be paired across start->end to stay balanced, lock elision must
happen in both or none. Therefore, in preparation for this change,
this patch prevents memslot updates across range_start() and range_end().
Note, technically flag-only memslot updates could be allowed in parallel,
but stalling a memslot update for a relatively short amount of time is
not a scalability issue, and this is all more than complex enough.
A long note on the locking: a previous version of the patch used an rwsem
to block the memslot update while the MMU notifier run, but this resulted
in the following deadlock involving the pseudo-lock tagged as
"mmu_notifier_invalidate_range_start".
======================================================
WARNING: possible circular locking dependency detected
5.12.0-rc3+ #6 Tainted: G OE
------------------------------------------------------
qemu-system-x86/3069 is trying to acquire lock:
ffffffff9c775ca0 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}, at: __mmu_notifier_invalidate_range_end+0x5/0x190
but task is already holding lock:
ffffaff7410a9160 (&kvm->mmu_notifier_slots_lock){.+.+}-{3:3}, at: kvm_mmu_notifier_invalidate_range_start+0x36d/0x4f0 [kvm]
which lock already depends on the new lock.
This corresponds to the following MMU notifier logic:
invalidate_range_start
take pseudo lock
down_read() (*)
release pseudo lock
invalidate_range_end
take pseudo lock (**)
up_read()
release pseudo lock
At point (*) we take the mmu_notifiers_slots_lock inside the pseudo lock;
at point (**) we take the pseudo lock inside the mmu_notifiers_slots_lock.
This could cause a deadlock (ignoring for a second that the pseudo lock
is not a lock):
- invalidate_range_start waits on down_read(), because the rwsem is
held by install_new_memslots
- install_new_memslots waits on down_write(), because the rwsem is
held till (another) invalidate_range_end finishes
- invalidate_range_end sits waits on the pseudo lock, held by
invalidate_range_start.
Removing the fairness of the rwsem breaks the cycle (in lockdep terms,
it would change the *shared* rwsem readers into *shared recursive*
readers), so open-code the wait using a readers count and a
spinlock. This also allows handling blockable and non-blockable
critical section in the same way.
Losing the rwsem fairness does theoretically allow MMU notifiers to
block install_new_memslots forever. Note that mm/mmu_notifier.c's own
retry scheme in mmu_interval_read_begin also uses wait/wake_up
and is likewise not fair.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The conversion tools used during DocBook/LaTeX/html/Markdown->ReST
conversion and some cut-and-pasted text contain some characters that
aren't easily reachable on standard keyboards and/or could cause
troubles when parsed by the documentation build system.
Replace the occurences of the following characters:
- U+00a0 (' '): NO-BREAK SPACE
as it can cause lines being truncated on PDF output
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Message-Id: <ff70cb42d63f3a1da66af1b21b8d038418ed5189.1626947264.git.mchehab+huawei@kernel.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The conversion tools used during DocBook/LaTeX/html/Markdown->ReST
conversion and some cut-and-pasted text contain some characters that
aren't easily reachable on standard keyboards and/or could cause
troubles when parsed by the documentation build system.
Replace the occurences of the following characters:
- U+00a0 (' '): NO-BREAK SPACE
as it can cause lines being truncated on PDF output
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Link: https://lore.kernel.org/r/ff70cb42d63f3a1da66af1b21b8d038418ed5189.1626947264.git.mchehab+huawei@kernel.org
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Pull documentation updates from Jonathan Corbet:
"This was a reasonably active cycle for documentation; this includes:
- Some kernel-doc cleanups. That script is still regex onslaught from
hell, but it has gotten a little better.
- Improvements to the checkpatch docs, which are also used by the
tool itself.
- A major update to the pathname lookup documentation.
- Elimination of :doc: markup, since our automarkup magic can create
references from filenames without all the extra noise.
- The flurry of Chinese translation activity continues.
Plus, of course, the usual collection of updates, typo fixes, and
warning fixes"
* tag 'docs-5.14' of git://git.lwn.net/linux: (115 commits)
docs: path-lookup: use bare function() rather than literals
docs: path-lookup: update symlink description
docs: path-lookup: update get_link() ->follow_link description
docs: path-lookup: update WALK_GET, WALK_PUT desc
docs: path-lookup: no get_link()
docs: path-lookup: update i_op->put_link and cookie description
docs: path-lookup: i_op->follow_link replaced with i_op->get_link
docs: path-lookup: Add macro name to symlink limit description
docs: path-lookup: remove filename_mountpoint
docs: path-lookup: update do_last() part
docs: path-lookup: update path_mountpoint() part
docs: path-lookup: update path_to_nameidata() part
docs: path-lookup: update follow_managed() part
docs: Makefile: Use CONFIG_SHELL not SHELL
docs: Take a little noise out of the build process
docs: x86: avoid using ReST :doc:`foo` markup
docs: virt: kvm: s390-pv-boot.rst: avoid using ReST :doc:`foo` markup
docs: userspace-api: landlock.rst: avoid using ReST :doc:`foo` markup
docs: trace: ftrace.rst: avoid using ReST :doc:`foo` markup
docs: trace: coresight: coresight.rst: avoid using ReST :doc:`foo` markup
...
