KVM/arm64 fixes for 6.14, take #2
- Large set of fixes for vector handling, specially in the interactions
between host and guest state. This fixes a number of bugs affecting
actual deployments, and greatly simplifies the FP/SIMD/SVE handling.
Thanks to Mark Rutland for dealing with this thankless task.
- Fix an ugly race between vcpu and vgic creation/init, resulting in
unexpected behaviours.
- Fix use of kernel VAs at EL2 when emulating timers with nVHE.
- Small set of pKVM improvements and cleanups.
If userspace creates vcpus, then a vgic, we end-up in a situation
where irqchip_in_kernel() will return true, but no private interrupt
has been allocated for these vcpus. This situation will continue
until userspace initialises the vgic, at which point we fix the
early vcpus. Should a vcpu run or be initialised in the interval,
bad things may happen.
An obvious solution is to move this fix-up phase to the point where
the vgic is created. This ensures that from that point onwards,
all vcpus have their private interrupts, as new vcpus will directly
allocate them.
With that, we have the invariant that when irqchip_in_kernel() is
true, all vcpus have their private interrupts.
Reported-by: Alexander Potapenko <glider@google.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250212182558.2865232-3-maz@kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
We currently spit out a warning if making a timer interrupt pending
fails. But not only this is loud and easy to trigger from userspace,
we also fail to do anything useful with that information.
Dropping the warning is the easiest thing to do for now. We can
always add error reporting if we really want in the future.
Reported-by: Alexander Potapenko <glider@google.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250212182558.2865232-2-maz@kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
When allocating guest stage-2 page-table pages at EL2, pKVM can consume
pages from the host-provided kvm_hyp_memcache. As pgtable.c expects
zeroed pages, guest_s2_zalloc_page() actively implements this zeroing
with a PAGE_SIZE memset. Unfortunately, we don't check the page
alignment of the host-provided address before doing so, which could
lead to the memset overrunning the page if the host was malicious.
Fix this by simply force-aligning all kvm_hyp_memcache allocations to
page boundaries.
Fixes: 60dfe093ec ("KVM: arm64: Instantiate guest stage-2 page-tables at EL2")
Reported-by: Ben Simner <ben.simner@cl.cam.ac.uk>
Signed-off-by: Quentin Perret <qperret@google.com>
Link: https://lore.kernel.org/r/20250213153615.3642515-1-qperret@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
Now that EL2 has gained some early timer emulation, it accesses
the offsets pointed to by the timer structure, both of which
live in the KVM structure.
Of course, these are *kernel* pointers, so the dereferencing
of these pointers in non-kernel code must be itself be offset.
Given switch.h its own version of timer_get_offset() and use that
instead.
Fixes: b86fc215dc ("KVM: arm64: Handle counter access early in non-HYP context")
Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Tested-by: Anders Roxell <anders.roxell@linaro.org>
Link: https://lore.kernel.org/r/20250212173454.2864462-1-maz@kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
At the end of kvm_arch_vcpu_load_fp() we check that no bits are set in
SVCR. We only check this for protected mode despite this mattering
equally for non-protected mode, and the comment above this is confusing.
Remove the comment and simplify the check, moving from WARN_ON() to
WARN_ON_ONCE() to avoid spamming the log.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
In non-protected KVM modes, while the guest FPSIMD/SVE/SME state is live on the
CPU, the host's active SVE VL may differ from the guest's maximum SVE VL:
* For VHE hosts, when a VM uses NV, ZCR_EL2 contains a value constrained
by the guest hypervisor, which may be less than or equal to that
guest's maximum VL.
Note: in this case the value of ZCR_EL1 is immaterial due to E2H.
* For nVHE/hVHE hosts, ZCR_EL1 contains a value written by the guest,
which may be less than or greater than the guest's maximum VL.
Note: in this case hyp code traps host SVE usage and lazily restores
ZCR_EL2 to the host's maximum VL, which may be greater than the
guest's maximum VL.
This can be the case between exiting a guest and kvm_arch_vcpu_put_fp().
If a softirq is taken during this period and the softirq handler tries
to use kernel-mode NEON, then the kernel will fail to save the guest's
FPSIMD/SVE state, and will pend a SIGKILL for the current thread.
This happens because kvm_arch_vcpu_ctxsync_fp() binds the guest's live
FPSIMD/SVE state with the guest's maximum SVE VL, and
fpsimd_save_user_state() verifies that the live SVE VL is as expected
before attempting to save the register state:
| if (WARN_ON(sve_get_vl() != vl)) {
| force_signal_inject(SIGKILL, SI_KERNEL, 0, 0);
| return;
| }
Fix this and make this a bit easier to reason about by always eagerly
switching ZCR_EL{1,2} at hyp during guest<->host transitions. With this
happening, there's no need to trap host SVE usage, and the nVHE/nVHE
__deactivate_cptr_traps() logic can be simplified to enable host access
to all present FPSIMD/SVE/SME features.
In protected nVHE/hVHE modes, the host's state is always saved/restored
by hyp, and the guest's state is saved prior to exit to the host, so
from the host's PoV the guest never has live FPSIMD/SVE/SME state, and
the host's ZCR_EL1 is never clobbered by hyp.
Fixes: 8c8010d69c ("KVM: arm64: Save/restore SVE state for nVHE")
Fixes: 2e3cf82063 ("KVM: arm64: nv: Ensure correct VL is loaded before saving SVE state")
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Cc: Will Deacon <will@kernel.org>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-9-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
The shared hyp switch header has a number of static functions which
might not be used by all files that include the header, and when unused
they will provoke compiler warnings, e.g.
| In file included from arch/arm64/kvm/hyp/nvhe/hyp-main.c:8:
| ./arch/arm64/kvm/hyp/include/hyp/switch.h:703:13: warning: 'kvm_hyp_handle_dabt_low' defined but not used [-Wunused-function]
| 703 | static bool kvm_hyp_handle_dabt_low(struct kvm_vcpu *vcpu, u64 *exit_code)
| | ^~~~~~~~~~~~~~~~~~~~~~~
| ./arch/arm64/kvm/hyp/include/hyp/switch.h:682:13: warning: 'kvm_hyp_handle_cp15_32' defined but not used [-Wunused-function]
| 682 | static bool kvm_hyp_handle_cp15_32(struct kvm_vcpu *vcpu, u64 *exit_code)
| | ^~~~~~~~~~~~~~~~~~~~~~
| ./arch/arm64/kvm/hyp/include/hyp/switch.h:662:13: warning: 'kvm_hyp_handle_sysreg' defined but not used [-Wunused-function]
| 662 | static bool kvm_hyp_handle_sysreg(struct kvm_vcpu *vcpu, u64 *exit_code)
| | ^~~~~~~~~~~~~~~~~~~~~
| ./arch/arm64/kvm/hyp/include/hyp/switch.h:458:13: warning: 'kvm_hyp_handle_fpsimd' defined but not used [-Wunused-function]
| 458 | static bool kvm_hyp_handle_fpsimd(struct kvm_vcpu *vcpu, u64 *exit_code)
| | ^~~~~~~~~~~~~~~~~~~~~
| ./arch/arm64/kvm/hyp/include/hyp/switch.h:329:13: warning: 'kvm_hyp_handle_mops' defined but not used [-Wunused-function]
| 329 | static bool kvm_hyp_handle_mops(struct kvm_vcpu *vcpu, u64 *exit_code)
| | ^~~~~~~~~~~~~~~~~~~
Mark these functions as 'inline' to suppress this warning. This
shouldn't result in any functional change.
At the same time, avoid the use of __alias() in the header and alias
kvm_hyp_handle_iabt_low() and kvm_hyp_handle_watchpt_low() to
kvm_hyp_handle_memory_fault() using CPP, matching the style in the rest
of the kernel. For consistency, kvm_hyp_handle_memory_fault() is also
marked as 'inline'.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Acked-by: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-8-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
The hyp exit handling logic is largely shared between VHE and nVHE/hVHE,
with common logic in arch/arm64/kvm/hyp/include/hyp/switch.h. The code
in the header depends on function definitions provided by
arch/arm64/kvm/hyp/vhe/switch.c and arch/arm64/kvm/hyp/nvhe/switch.c
when they include the header.
This is an unusual header dependency, and prevents the use of
arch/arm64/kvm/hyp/include/hyp/switch.h in other files as this would
result in compiler warnings regarding missing definitions, e.g.
| In file included from arch/arm64/kvm/hyp/nvhe/hyp-main.c:8:
| ./arch/arm64/kvm/hyp/include/hyp/switch.h:733:31: warning: 'kvm_get_exit_handler_array' used but never defined
| 733 | static const exit_handler_fn *kvm_get_exit_handler_array(struct kvm_vcpu *vcpu);
| | ^~~~~~~~~~~~~~~~~~~~~~~~~~
| ./arch/arm64/kvm/hyp/include/hyp/switch.h:735:13: warning: 'early_exit_filter' used but never defined
| 735 | static void early_exit_filter(struct kvm_vcpu *vcpu, u64 *exit_code);
| | ^~~~~~~~~~~~~~~~~
Refactor the logic such that the header doesn't depend on anything from
the C files. There should be no functional change as a result of this
patch.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Acked-by: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-7-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
For historical reasons, the VHE and nVHE/hVHE implementations of
__activate_cptr_traps() pair with a common implementation of
__kvm_reset_cptr_el2(), which ideally would be named
__deactivate_cptr_traps().
Rename __kvm_reset_cptr_el2() to __deactivate_cptr_traps(), and split it
into separate VHE and nVHE/hVHE variants so that each can be paired with
its corresponding implementation of __activate_cptr_traps().
At the same time, fold kvm_write_cptr_el2() into its callers. This
makes it clear in-context whether a write is made to the CPACR_EL1
encoding or the CPTR_EL2 encoding, and removes the possibility of
confusion as to whether kvm_write_cptr_el2() reformats the sysreg fields
as cpacr_clear_set() does.
In the nVHE/hVHE implementation of __activate_cptr_traps(), placing the
sysreg writes within the if-else blocks requires that the call to
__activate_traps_fpsimd32() is moved earlier, but as this was always
called before writing to CPTR_EL2/CPACR_EL1, this should not result in a
functional change.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Acked-by: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-6-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
When KVM is in VHE mode, the host kernel tries to save and restore the
configuration of CPACR_EL1.SMEN (i.e. CPTR_EL2.SMEN when HCR_EL2.E2H=1)
across kvm_arch_vcpu_load_fp() and kvm_arch_vcpu_put_fp(), since the
configuration may be clobbered by hyp when running a vCPU. This logic
has historically been broken, and is currently redundant.
This logic was originally introduced in commit:
861262ab86 ("KVM: arm64: Handle SME host state when running guests")
At the time, the VHE hyp code would reset CPTR_EL2.SMEN to 0b00 when
returning to the host, trapping host access to SME state. Unfortunately,
this was unsafe as the host could take a softirq before calling
kvm_arch_vcpu_put_fp(), and if a softirq handler were to use kernel mode
NEON the resulting attempt to save the live FPSIMD/SVE/SME state would
result in a fatal trap.
That issue was limited to VHE mode. For nVHE/hVHE modes, KVM always
saved/restored the host kernel's CPACR_EL1 value, and configured
CPTR_EL2.TSM to 0b0, ensuring that host usage of SME would not be
trapped.
The issue above was incidentally fixed by commit:
375110ab51 ("KVM: arm64: Fix resetting SME trap values on reset for (h)VHE")
That commit changed the VHE hyp code to configure CPTR_EL2.SMEN to 0b01
when returning to the host, permitting host kernel usage of SME,
avoiding the issue described above. At the time, this was not identified
as a fix for commit 861262ab86.
Now that the host eagerly saves and unbinds its own FPSIMD/SVE/SME
state, there's no need to save/restore the state of the EL0 SME trap.
The kernel can safely save/restore state without trapping, as described
above, and will restore userspace state (including trap controls) before
returning to userspace.
Remove the redundant logic.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Acked-by: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-5-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
When KVM is in VHE mode, the host kernel tries to save and restore the
configuration of CPACR_EL1.ZEN (i.e. CPTR_EL2.ZEN when HCR_EL2.E2H=1)
across kvm_arch_vcpu_load_fp() and kvm_arch_vcpu_put_fp(), since the
configuration may be clobbered by hyp when running a vCPU. This logic is
currently redundant.
The VHE hyp code unconditionally configures CPTR_EL2.ZEN to 0b01 when
returning to the host, permitting host kernel usage of SVE.
Now that the host eagerly saves and unbinds its own FPSIMD/SVE/SME
state, there's no need to save/restore the state of the EL0 SVE trap.
The kernel can safely save/restore state without trapping, as described
above, and will restore userspace state (including trap controls) before
returning to userspace.
Remove the redundant logic.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Acked-by: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-4-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
Now that the host eagerly saves its own FPSIMD/SVE/SME state,
non-protected KVM never needs to save the host FPSIMD/SVE/SME state,
and the code to do this is never used. Protected KVM still needs to
save/restore the host FPSIMD/SVE state to avoid leaking guest state to
the host (and to avoid revealing to the host whether the guest used
FPSIMD/SVE/SME), and that code needs to be retained.
Remove the unused code and data structures.
To avoid the need for a stub copy of kvm_hyp_save_fpsimd_host() in the
VHE hyp code, the nVHE/hVHE version is moved into the shared switch
header, where it is only invoked when KVM is in protected mode.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Acked-by: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-3-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
There are several problems with the way hyp code lazily saves the host's
FPSIMD/SVE state, including:
* Host SVE being discarded unexpectedly due to inconsistent
configuration of TIF_SVE and CPACR_ELx.ZEN. This has been seen to
result in QEMU crashes where SVE is used by memmove(), as reported by
Eric Auger:
https://issues.redhat.com/browse/RHEL-68997
* Host SVE state is discarded *after* modification by ptrace, which was an
unintentional ptrace ABI change introduced with lazy discarding of SVE state.
* The host FPMR value can be discarded when running a non-protected VM,
where FPMR support is not exposed to a VM, and that VM uses
FPSIMD/SVE. In these cases the hyp code does not save the host's FPMR
before unbinding the host's FPSIMD/SVE/SME state, leaving a stale
value in memory.
Avoid these by eagerly saving and "flushing" the host's FPSIMD/SVE/SME
state when loading a vCPU such that KVM does not need to save any of the
host's FPSIMD/SVE/SME state. For clarity, fpsimd_kvm_prepare() is
removed and the necessary call to fpsimd_save_and_flush_cpu_state() is
placed in kvm_arch_vcpu_load_fp(). As 'fpsimd_state' and 'fpmr_ptr'
should not be used, they are set to NULL; all uses of these will be
removed in subsequent patches.
Historical problems go back at least as far as v5.17, e.g. erroneous
assumptions about TIF_SVE being clear in commit:
8383741ab2 ("KVM: arm64: Get rid of host SVE tracking/saving")
... and so this eager save+flush probably needs to be backported to ALL
stable trees.
Fixes: 93ae6b01ba ("KVM: arm64: Discard any SVE state when entering KVM guests")
Fixes: 8c845e2731 ("arm64/sve: Leave SVE enabled on syscall if we don't context switch")
Fixes: ef3be86021 ("KVM: arm64: Add save/restore support for FPMR")
Reported-by: Eric Auger <eauger@redhat.com>
Reported-by: Wilco Dijkstra <wilco.dijkstra@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Tested-by: Mark Brown <broonie@kernel.org>
Tested-by: Eric Auger <eric.auger@redhat.com>
Acked-by: Will Deacon <will@kernel.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Florian Weimer <fweimer@redhat.com>
Cc: Fuad Tabba <tabba@google.com>
Cc: Jeremy Linton <jeremy.linton@arm.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oliver.upton@linux.dev>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250210195226.1215254-2-mark.rutland@arm.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
When the handling of a guest stage-2 permission fault races with an MMU
notifier, the faulting page might be gone from the guest's stage-2 by
the point we attempt to call (p)kvm_pgtable_stage2_relax_perms(). In the
normal KVM case, this leads to returning -EAGAIN which user_mem_abort()
handles correctly by simply re-entering the guest. However, the pKVM
hypercall implementation has additional logic to check the page state
using __check_host_shared_guest() which gets confused with absence of a
page mapped at the requested IPA and returns -ENOENT, hence breaking
user_mem_abort() and hilarity ensues.
Luckily, several of the hypercalls for managing the stage-2 page-table
of NP guests have no effect on the pKVM ownership tracking (wrprotect,
test_clear_young, mkyoung, and crucially relax_perms), so the extra
state checking logic is in fact not strictly necessary. So, to fix the
discrepancy between standard KVM and pKVM, let's just drop the
superfluous __check_host_shared_guest() logic from those hypercalls and
make the extra state checking a debug assertion dependent on
CONFIG_NVHE_EL2_DEBUG as we already do for other transitions.
Signed-off-by: Quentin Perret <qperret@google.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250207145438.1333475-3-qperret@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
The check_host_shared_guest() path expects to find a last-level valid
PTE in the guest's stage-2 page-table. However, it checks the PTE's
level before its validity, which makes it hard for callers to figure out
what went wrong.
To make error handling simpler, check the PTE's validity first.
Signed-off-by: Quentin Perret <qperret@google.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250207145438.1333475-2-qperret@google.com
Signed-off-by: Marc Zyngier <maz@kernel.org>
When waking a VM's NX huge page recovery thread, ensure the thread is
actually alive before trying to wake it. Now that the thread is spawned
on-demand during KVM_RUN, a VM without a recovery thread is reachable via
the related module params.
BUG: kernel NULL pointer dereference, address: 0000000000000040
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
RIP: 0010:vhost_task_wake+0x5/0x10
Call Trace:
<TASK>
set_nx_huge_pages+0xcc/0x1e0 [kvm]
param_attr_store+0x8a/0xd0
module_attr_store+0x1a/0x30
kernfs_fop_write_iter+0x12f/0x1e0
vfs_write+0x233/0x3e0
ksys_write+0x60/0xd0
do_syscall_64+0x5b/0x160
entry_SYSCALL_64_after_hwframe+0x4b/0x53
RIP: 0033:0x7f3b52710104
</TASK>
Modules linked in: kvm_intel kvm
CR2: 0000000000000040
Fixes: 931656b9e2 ("kvm: defer huge page recovery vhost task to later")
Cc: stable@vger.kernel.org
Cc: Keith Busch <kbusch@kernel.org>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20250124234623.3609069-1-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The only statement in a kvm_arch_post_init_vm implementation
can be moved into the x86 kvm_arch_init_vm. Do so and remove all
traces from architecture-independent code.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM/arm64 fixes for 6.14, take #1
- Correctly clean the BSS to the PoC before allowing EL2 to access it
on nVHE/hVHE/protected configurations
- Propagate ownership of debug registers in protected mode after
the rework that landed in 6.14-rc1
- Stop pretending that we can run the protected mode without a GICv3
being present on the host
- Fix a use-after-free situation that can occur if a vcpu fails to
initialise the NV shadow S2 MMU contexts
- Always evaluate the need to arm a background timer for fully emulated
guest timers
- Fix the emulation of EL1 timers in the absence of FEAT_ECV
- Correctly handle the EL2 virtual timer, specially when HCR_EL2.E2H==0
- some selftest fixes
- move some kvm-related functions from mm into kvm
- remove all usage of page->index and page->lru from kvm
- fixes and cleanups for vsie
SYNTHESIZED_F() generally is used together with setup_force_cpu_cap(),
i.e. when it makes sense to present the feature even if cpuid does not
have it *and* the VM is not able to see the difference. For example,
it can be used when mitigations on the host automatically protect
the guest as well.
The "SYNTHESIZED_F(SRSO_USER_KERNEL_NO)" line came in as a conflict
resolution between the CPUID overhaul from the KVM tree and support
for the feature in the x86 tree. Using it right now does not hurt,
or make a difference for that matter, because there is no
setup_force_cpu_cap(X86_FEATURE_SRSO_USER_KERNEL_NO). However, it
is a little less future proof in case such a setup_force_cpu_cap()
appears later, for a case where the kernel somehow is not vulnerable
but the guest would have to apply the mitigation.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The way we deal with the EL2 virtual timer is a bit odd.
We try to cope with E2H being flipped, and adjust which offset
applies to that timer depending on the current E2H value. But that's
a complexity we shouldn't have to worry about.
What we have to deal with is either E2H being RES1, in which case
there is no offset, or E2H being RES0, and the virtual timer simply
does not exist.
Drop the adjusting of the timer offset, which makes things a bit
simpler. At the same time, make sure that accessing the HV timer
when E2H is RES0 results in an UNDEF in the guest.
Suggested-by: Oliver Upton <oliver.upton@linux.dev>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250204110050.150560-4-maz@kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
Both Wei-Lin Chang and Volodymyr Babchuk report that the way we
handle the emulation of EL1 timers with NV is completely wrong,
specially in the case of HCR_EL2.E2H==0.
There are three problems in about as many lines of code:
- With E2H==0, the EL1 timers are overwritten with the EL1 state,
while they should actually contain the EL2 state (as per the timer
map)
- With E2H==1, we run the full EL1 timer emulation even when ECV
is present, hiding a bug in timer_emulate() (see previous patch)
- The comments are actively misleading, and say all the wrong things.
This is only attributable to the code having been initially written
for FEAT_NV, hacked up to handle FEAT_NV2 *in parallel*, and vaguely
hacked again to be FEAT_NV2 only. Oh, and yours truly being a gold
plated idiot.
The fix is obvious: just delete most of the E2H==0 code, have a unified
handling of the timers (because they really are E2H agnostic), and
make sure we don't execute any of that when FEAT_ECV is present.
Fixes: 4bad3068cf ("KVM: arm64: nv: Sync nested timer state with FEAT_NV2")
Reported-by: Wei-Lin Chang <r09922117@csie.ntu.edu.tw>
Reported-by: Volodymyr Babchuk <Volodymyr_Babchuk@epam.com>
Link: https://lore.kernel.org/r/fqiqfjzwpgbzdtouu2pwqlu7llhnf5lmy4hzv5vo6ph4v3vyls@jdcfy3fjjc5k
Link: https://lore.kernel.org/r/87frl51tse.fsf@epam.com
Tested-by: Dmytro Terletskyi <dmytro_terletskyi@epam.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250204110050.150560-3-maz@kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
When updating the interrupt state for an emulated timer, we return
early and skip the setup of a soft timer that runs in parallel
with the guest.
While this is OK if we have set the interrupt pending, it is pretty
wrong if the guest moved CVAL into the future. In that case,
no timer is armed and the guest can wait for a very long time
(it will take a full put/load cycle for the situation to resolve).
This is specially visible with EDK2 running at EL2, but still
using the EL1 virtual timer, which in that case is fully emulated.
Any key-press takes ages to be captured, as there is no UART
interrupt and EDK2 relies on polling from a timer...
The fix is simply to drop the early return. If the timer interrupt
is pending, we will still return early, and otherwise arm the soft
timer.
Fixes: 4d74ecfa64 ("KVM: arm64: Don't arm a hrtimer for an already pending timer")
Cc: stable@vger.kernel.org
Tested-by: Dmytro Terletskyi <dmytro_terletskyi@epam.com>
Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Link: https://lore.kernel.org/r/20250204110050.150560-2-maz@kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
For each vcpu that userspace creates, we allocate a number of
s2_mmu structures that will eventually contain our shadow S2
page tables.
Since this is a dynamically allocated array, we reallocate
the array and initialise the newly allocated elements. Once
everything is correctly initialised, we adjust pointer and size
in the kvm structure, and move on.
But should that initialisation fail *and* the reallocation triggered
a copy to another location, we end-up returning early, with the
kvm structure still containing the (now stale) old pointer. Weeee!
Cure it by assigning the pointer early, and use this to perform
the initialisation. If everything succeeds, we adjust the size.
Otherwise, we just leave the size as it was, no harm done, and the
new memory is as good as the ol' one (we hope...).
Fixes: 4f128f8e1a ("KVM: arm64: nv: Support multiple nested Stage-2 mmu structures")
Reported-by: Alexander Potapenko <glider@google.com>
Tested-by: Alexander Potapenko <glider@google.com>
Link: https://lore.kernel.org/r/20250204145554.774427-1-maz@kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
Pull turbostat updates from Len Brown:
- Fix regression that affinitized forked child in one-shot mode.
- Harden one-shot mode against hotplug online/offline
- Enable RAPL SysWatt column by default
- Add initial PTL, CWF platform support
- Harden initial PMT code in response to early use
- Enable first built-in PMT counter: CWF c1e residency
- Refuse to run on unsupported platforms without --force, to encourage
updating to a version that supports the system, and to avoid
no-so-useful measurement results
* tag 'turbostat-2025.02.02' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux: (25 commits)
tools/power turbostat: version 2025.02.02
tools/power turbostat: Add CPU%c1e BIC for CWF
tools/power turbostat: Harden one-shot mode against cpu offline
tools/power turbostat: Fix forked child affinity regression
tools/power turbostat: Add tcore clock PMT type
tools/power turbostat: version 2025.01.14
tools/power turbostat: Allow adding PMT counters directly by sysfs path
tools/power turbostat: Allow mapping multiple PMT files with the same GUID
tools/power turbostat: Add PMT directory iterator helper
tools/power turbostat: Extend PMT identification with a sequence number
tools/power turbostat: Return default value for unmapped PMT domains
tools/power turbostat: Check for non-zero value when MSR probing
tools/power turbostat: Enhance turbostat self-performance visibility
tools/power turbostat: Add fixed RAPL PSYS divisor for SPR
tools/power turbostat: Fix PMT mmaped file size rounding
tools/power turbostat: Remove SysWatt from DISABLED_BY_DEFAULT
tools/power turbostat: Add an NMI column
tools/power turbostat: add Busy% to "show idle"
tools/power turbostat: Introduce --force parameter
tools/power turbostat: Improve --help output
...
Pull sh updates from John Paul Adrian Glaubitz:
"Fixes and improvements for sh:
- replace seq_printf() with the more efficient
seq_put_decimal_ull_width() to increase performance when stress
reading /proc/interrupts (David Wang)
- migrate sh to the generic rule for built-in DTB to help avoid race
conditions during parallel builds which can occur because Kbuild
decends into arch/*/boot/dts twice (Masahiro Yamada)
- replace select with imply in the board Kconfig for enabling
hardware with complex dependencies. This addresses warnings which
were reported by the kernel test robot (Geert Uytterhoeven)"
* tag 'sh-for-v6.14-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux:
sh: boards: Use imply to enable hardware with complex dependencies
sh: Migrate to the generic rule for built-in DTB
sh: irq: Use seq_put_decimal_ull_width() for decimal values
Summary of Changes since 2024.11.30:
Fix regression in 2023.11.07 that affinitized forked child
in one-shot mode.
Harden one-shot mode against hotplug online/offline
Enable RAPL SysWatt column by default.
Add initial PTL, CWF platform support.
Harden initial PMT code in response to early use.
Enable first built-in PMT counter: CWF c1e residency
Refuse to run on unsupported platforms without --force,
to encourage updating to a version that supports the system,
and to avoid no-so-useful measurement results.
Signed-off-by: Len Brown <len.brown@intel.com>
Pull misc vfs cleanups from Al Viro:
"Two unrelated patches - one is a removal of long-obsolete include in
overlayfs (it used to need fs/internal.h, but the extern it wanted has
been moved back to include/linux/namei.h) and another introduces
convenience helper constructing struct qstr by a NUL-terminated
string"
* tag 'pull-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
add a string-to-qstr constructor
fs/overlayfs/namei.c: get rid of include ../internal.h
Pull MIPS fix from Thomas Bogendoerfer:
"Revert commit breaking sysv ipc for o32 ABI"
* tag 'mips_6.14_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
Revert "mips: fix shmctl/semctl/msgctl syscall for o32"
Pull more smb client updates from Steve French:
- various updates for special file handling: symlink handling,
support for creating sockets, cleanups, new mount options (e.g. to
allow disabling using reparse points for them, and to allow
overriding the way symlinks are saved), and fixes to error paths
- fix for kerberos mounts (allow IAKerb)
- SMB1 fix for stat and for setting SACL (auditing)
- fix an incorrect error code mapping
- cleanups"
* tag 'v6.14-rc-smb3-client-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6: (21 commits)
cifs: Fix parsing native symlinks directory/file type
cifs: update internal version number
cifs: Add support for creating WSL-style symlinks
smb3: add support for IAKerb
cifs: Fix struct FILE_ALL_INFO
cifs: Add support for creating NFS-style symlinks
cifs: Add support for creating native Windows sockets
cifs: Add mount option -o reparse=none
cifs: Add mount option -o symlink= for choosing symlink create type
cifs: Fix creating and resolving absolute NT-style symlinks
cifs: Simplify reparse point check in cifs_query_path_info() function
cifs: Remove symlink member from cifs_open_info_data union
cifs: Update description about ACL permissions
cifs: Rename struct reparse_posix_data to reparse_nfs_data_buffer and move to common/smb2pdu.h
cifs: Remove struct reparse_posix_data from struct cifs_open_info_data
cifs: Remove unicode parameter from parse_reparse_point() function
cifs: Fix getting and setting SACLs over SMB1
cifs: Remove intermediate object of failed create SFU call
cifs: Validate EAs for WSL reparse points
cifs: Change translation of STATUS_PRIVILEGE_NOT_HELD to -EPERM
...
Pull debugfs fix from Greg KH:
"Here is a single debugfs fix from Al to resolve a reported regression
in the driver-core tree. It has been reported to fix the issue"
* tag 'driver-core-6.14-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
debugfs: Fix the missing initializations in __debugfs_file_get()
Pull misc fixes from Andrew Morton:
"21 hotfixes. 8 are cc:stable and the remainder address post-6.13
issues. 13 are for MM and 8 are for non-MM.
All are singletons, please see the changelogs for details"
* tag 'mm-hotfixes-stable-2025-02-01-03-56' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (21 commits)
MAINTAINERS: include linux-mm for xarray maintenance
revert "xarray: port tests to kunit"
MAINTAINERS: add lib/test_xarray.c
mailmap, MAINTAINERS, docs: update Carlos's email address
mm/hugetlb: fix hugepage allocation for interleaved memory nodes
mm: gup: fix infinite loop within __get_longterm_locked
mm, swap: fix reclaim offset calculation error during allocation
.mailmap: update email address for Christopher Obbard
kfence: skip __GFP_THISNODE allocations on NUMA systems
nilfs2: fix possible int overflows in nilfs_fiemap()
mm: compaction: use the proper flag to determine watermarks
kernel: be more careful about dup_mmap() failures and uprobe registering
mm/fake-numa: handle cases with no SRAT info
mm: kmemleak: fix upper boundary check for physical address objects
mailmap: add an entry for Hamza Mahfooz
MAINTAINERS: mailmap: update Yosry Ahmed's email address
scripts/gdb: fix aarch64 userspace detection in get_current_task
mm/vmscan: accumulate nr_demoted for accurate demotion statistics
ocfs2: fix incorrect CPU endianness conversion causing mount failure
mm/zsmalloc: add __maybe_unused attribute for is_first_zpdesc()
...
Pull media fix from Mauro Carvalho Chehab:
"A revert for a regression in the uvcvideo driver"
* tag 'media/v6.14-2' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
Revert "media: uvcvideo: Require entities to have a non-zero unique ID"
gather_bootmem_prealloc() assumes the start nid as 0 and size as
num_node_state(N_MEMORY). That means in case if memory attached numa
nodes are interleaved, then gather_bootmem_prealloc_parallel() will fail
to scan few of these nodes.
Since memory attached numa nodes can be interleaved in any fashion, hence
ensure that the current code checks for all numa node ids
(.size = nr_node_ids). Let's still keep max_threads as N_MEMORY, so that
it can distributes all nr_node_ids among the these many no. threads.
e.g. qemu cmdline
========================
numa_cmd="-numa node,nodeid=1,memdev=mem1,cpus=2-3 -numa node,nodeid=0,cpus=0-1 -numa dist,src=0,dst=1,val=20"
mem_cmd="-object memory-backend-ram,id=mem1,size=16G"
w/o this patch for cmdline (default_hugepagesz=1GB hugepagesz=1GB hugepages=2):
==========================
~ # cat /proc/meminfo |grep -i huge
AnonHugePages: 0 kB
ShmemHugePages: 0 kB
FileHugePages: 0 kB
HugePages_Total: 0
HugePages_Free: 0
HugePages_Rsvd: 0
HugePages_Surp: 0
Hugepagesize: 1048576 kB
Hugetlb: 0 kB
with this patch for cmdline (default_hugepagesz=1GB hugepagesz=1GB hugepages=2):
===========================
~ # cat /proc/meminfo |grep -i huge
AnonHugePages: 0 kB
ShmemHugePages: 0 kB
FileHugePages: 0 kB
HugePages_Total: 2
HugePages_Free: 2
HugePages_Rsvd: 0
HugePages_Surp: 0
Hugepagesize: 1048576 kB
Hugetlb: 2097152 kB
Link: https://lkml.kernel.org/r/f8d8dad3a5471d284f54185f65d575a6aaab692b.1736592534.git.ritesh.list@gmail.com
Fixes: b78b27d029 ("hugetlb: parallelize 1G hugetlb initialization")
Signed-off-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Reported-by: Pavithra Prakash <pavrampu@linux.ibm.com>
Suggested-by: Muchun Song <muchun.song@linux.dev>
Tested-by: Sourabh Jain <sourabhjain@linux.ibm.com>
Reviewed-by: Luiz Capitulino <luizcap@redhat.com>
Acked-by: David Rientjes <rientjes@google.com>
Cc: Donet Tom <donettom@linux.ibm.com>
Cc: Gang Li <gang.li@linux.dev>
Cc: Daniel Jordan <daniel.m.jordan@oracle.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Since nilfs_bmap_lookup_contig() in nilfs_fiemap() calculates its result
by being prepared to go through potentially maxblocks == INT_MAX blocks,
the value in n may experience an overflow caused by left shift of blkbits.
While it is extremely unlikely to occur, play it safe and cast right hand
expression to wider type to mitigate the issue.
Found by Linux Verification Center (linuxtesting.org) with static analysis
tool SVACE.
Link: https://lkml.kernel.org/r/20250124222133.5323-1-konishi.ryusuke@gmail.com
Fixes: 622daaff0a ("nilfs2: fiemap support")
Signed-off-by: Nikita Zhandarovich <n.zhandarovich@fintech.ru>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
There are 4 NUMA nodes on my machine, and each NUMA node has 32GB of
memory. I have configured 16GB of CMA memory on each NUMA node, and
starting a 32GB virtual machine with device passthrough is extremely slow,
taking almost an hour.
Long term GUP cannot allocate memory from CMA area, so a maximum of 16 GB
of no-CMA memory on a NUMA node can be used as virtual machine memory.
There is 16GB of free CMA memory on a NUMA node, which is sufficient to
pass the order-0 watermark check, causing the __compaction_suitable()
function to consistently return true.
For costly allocations, if the __compaction_suitable() function always
returns true, it causes the __alloc_pages_slowpath() function to fail to
exit at the appropriate point. This prevents timely fallback to
allocating memory on other nodes, ultimately resulting in excessively long
virtual machine startup times.
Call trace:
__alloc_pages_slowpath
if (compact_result == COMPACT_SKIPPED ||
compact_result == COMPACT_DEFERRED)
goto nopage; // should exit __alloc_pages_slowpath() from here
We could use the real unmovable allocation context to have
__zone_watermark_unusable_free() subtract CMA pages, and thus we won't
pass the order-0 check anymore once the non-CMA part is exhausted. There
is some risk that in some different scenario the compaction could in fact
migrate pages from the exhausted non-CMA part of the zone to the CMA part
and succeed, and we'll skip it instead. But only __GFP_NORETRY
allocations should be affected in the immediate "goto nopage" when
compaction is skipped, others will attempt with DEF_COMPACT_PRIORITY
anyway and won't fail without trying to compact-migrate the non-CMA
pageblocks into CMA pageblocks first, so it should be fine.
After this fix, it only takes a few tens of seconds to start a 32GB
virtual machine with device passthrough functionality.
Link: https://lore.kernel.org/lkml/1736335854-548-1-git-send-email-yangge1116@126.com/
Link: https://lkml.kernel.org/r/1737788037-8439-1-git-send-email-yangge1116@126.com
Signed-off-by: yangge <yangge1116@126.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Barry Song <21cnbao@gmail.com>
Cc: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
If a memory allocation fails during dup_mmap(), the maple tree can be left
in an unsafe state for other iterators besides the exit path. All the
locks are dropped before the exit_mmap() call (in mm/mmap.c), but the
incomplete mm_struct can be reached through (at least) the rmap finding
the vmas which have a pointer back to the mm_struct.
Up to this point, there have been no issues with being able to find an
mm_struct that was only partially initialised. Syzbot was able to make
the incomplete mm_struct fail with recent forking changes, so it has been
proven unsafe to use the mm_struct that hasn't been initialised, as
referenced in the link below.
Although 8ac662f5da ("fork: avoid inappropriate uprobe access to
invalid mm") fixed the uprobe access, it does not completely remove the
race.
This patch sets the MMF_OOM_SKIP to avoid the iteration of the vmas on the
oom side (even though this is extremely unlikely to be selected as an oom
victim in the race window), and sets MMF_UNSTABLE to avoid other potential
users from using a partially initialised mm_struct.
When registering vmas for uprobe, skip the vmas in an mm that is marked
unstable. Modifying a vma in an unstable mm may cause issues if the mm
isn't fully initialised.
Link: https://lore.kernel.org/all/6756d273.050a0220.2477f.003d.GAE@google.com/
Link: https://lkml.kernel.org/r/20250127170221.1761366-1-Liam.Howlett@oracle.com
Fixes: d240629148 ("fork: use __mt_dup() to duplicate maple tree in dup_mmap()")
Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Peng Zhang <zhangpeng.00@bytedance.com>
Cc: Matthew Wilcox <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Paolo Bonzini
Marc Zyngier
Quentin Perret
Mark Rutland
Sean Christopherson
Colin Ian King
Oliver Upton
Linus Torvalds
Len Brown
Andrew Morton
Tamir Duberstein
Carlos Bilbao
Ritesh Harjani (IBM)
Zhaoyang Huang
Kairui Song
Christopher Obbard
Marco Elver
Nikita Zhandarovich
yangge
Liam R. Howlett