commit b4d37cdb77 upstream.
When deassigning a KVM_IRQFD, don't clobber the irqfd's copy of the IRQ's
routing entry as doing so breaks kvm_arch_irq_bypass_del_producer() on x86
and arm64, which explicitly look for KVM_IRQ_ROUTING_MSI. Instead, to
handle a concurrent routing update, verify that the irqfd is still active
before consuming the routing information. As evidenced by the x86 and
arm64 bugs, and another bug in kvm_arch_update_irqfd_routing() (see below),
clobbering the entry type without notifying arch code is surprising and
error prone.
As a bonus, checking that the irqfd is active provides a convenient
location for documenting _why_ KVM must not consume the routing entry for
an irqfd that is in the process of being deassigned: once the irqfd is
deleted from the list (which happens *before* the eventfd is detached), it
will no longer receive updates via kvm_irq_routing_update(), and so KVM
could deliver an event using stale routing information (relative to
KVM_SET_GSI_ROUTING returning to userspace).
As an even better bonus, explicitly checking for the irqfd being active
fixes a similar bug to the one the clobbering is trying to prevent: if an
irqfd is deactivated, and then its routing is changed,
kvm_irq_routing_update() won't invoke kvm_arch_update_irqfd_routing()
(because the irqfd isn't in the list). And so if the irqfd is in bypass
mode, IRQs will continue to be posted using the old routing information.
As for kvm_arch_irq_bypass_del_producer(), clobbering the routing type
results in KVM incorrectly keeping the IRQ in bypass mode, which is
especially problematic on AMD as KVM tracks IRQs that are being posted to
a vCPU in a list whose lifetime is tied to the irqfd.
Without the help of KASAN to detect use-after-free, the most common
sympton on AMD is a NULL pointer deref in amd_iommu_update_ga() due to
the memory for irqfd structure being re-allocated and zeroed, resulting
in irqfd->irq_bypass_data being NULL when read by
avic_update_iommu_vcpu_affinity():
BUG: kernel NULL pointer dereference, address: 0000000000000018
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
PGD 40cf2b9067 P4D 40cf2b9067 PUD 408362a067 PMD 0
Oops: Oops: 0000 [#1] SMP
CPU: 6 UID: 0 PID: 40383 Comm: vfio_irq_test
Tainted: G U W O 6.19.0-smp--5dddc257e6b2-irqfd #31 NONE
Tainted: [U]=USER, [W]=WARN, [O]=OOT_MODULE
Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 34.78.2-0 09/05/2025
RIP: 0010:amd_iommu_update_ga+0x19/0xe0
Call Trace:
<TASK>
avic_update_iommu_vcpu_affinity+0x3d/0x90 [kvm_amd]
__avic_vcpu_load+0xf4/0x130 [kvm_amd]
kvm_arch_vcpu_load+0x89/0x210 [kvm]
vcpu_load+0x30/0x40 [kvm]
kvm_arch_vcpu_ioctl_run+0x45/0x620 [kvm]
kvm_vcpu_ioctl+0x571/0x6a0 [kvm]
__se_sys_ioctl+0x6d/0xb0
do_syscall_64+0x6f/0x9d0
entry_SYSCALL_64_after_hwframe+0x4b/0x53
RIP: 0033:0x46893b
</TASK>
---[ end trace 0000000000000000 ]---
If AVIC is inhibited when the irfd is deassigned, the bug will manifest as
list corruption, e.g. on the next irqfd assignment.
list_add corruption. next->prev should be prev (ffff8d474d5cd588),
but was 0000000000000000. (next=ffff8d8658f86530).
------------[ cut here ]------------
kernel BUG at lib/list_debug.c:31!
Oops: invalid opcode: 0000 [#1] SMP
CPU: 128 UID: 0 PID: 80818 Comm: vfio_irq_test
Tainted: G U W O 6.19.0-smp--f19dc4d680ba-irqfd #28 NONE
Tainted: [U]=USER, [W]=WARN, [O]=OOT_MODULE
Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 34.78.2-0 09/05/2025
RIP: 0010:__list_add_valid_or_report+0x97/0xc0
Call Trace:
<TASK>
avic_pi_update_irte+0x28e/0x2b0 [kvm_amd]
kvm_pi_update_irte+0xbf/0x190 [kvm]
kvm_arch_irq_bypass_add_producer+0x72/0x90 [kvm]
irq_bypass_register_consumer+0xcd/0x170 [irqbypass]
kvm_irqfd+0x4c6/0x540 [kvm]
kvm_vm_ioctl+0x118/0x5d0 [kvm]
__se_sys_ioctl+0x6d/0xb0
do_syscall_64+0x6f/0x9d0
entry_SYSCALL_64_after_hwframe+0x4b/0x53
</TASK>
---[ end trace 0000000000000000 ]---
On Intel and arm64, the bug is less noisy, as the end result is that the
device keeps posting IRQs to the vCPU even after it's been deassigned.
Note, the worst of the breakage can be traced back to commit cb21073767
("KVM: Pass new routing entries and irqfd when updating IRTEs"), as before
that commit KVM would pull the routing information from the per-VM routing
table. But as above, similar bugs have existed since support for IRQ
bypass was added. E.g. if a routing change finished before irq_shutdown()
invoked kvm_arch_irq_bypass_del_producer(), VMX and SVM would see stale
routing information and potentially leave the irqfd in bypass mode.
Alternatively, x86 could be fixed by explicitly checking irq_bypass_vcpu
instead of irq_entry.type in kvm_arch_irq_bypass_del_producer(), and arm64
could be modified to utilize irq_bypass_vcpu in a similar manner. But (a)
that wouldn't fix the routing updates bug, and (b) fixing core code doesn't
preclude x86 (or arm64) from adding such code as a sanity check (spoiler
alert).
Fixes: f70c20aaf1 ("KVM: Add an arch specific hooks in 'struct kvm_kernel_irqfd'")
Fixes: cb21073767 ("KVM: Pass new routing entries and irqfd when updating IRTEs")
Fixes: a0d7e2fc61 ("KVM: arm64: vgic-v4: Only attempt vLPI mapping for actual MSIs")
Cc: stable@vger.kernel.org
Cc: Marc Zyngier <maz@kernel.org>
Cc: Oliver Upton <oupton@kernel.org>
Link: https://patch.msgid.link/20260113174606.104978-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9935df5333 upstream.
Reject attempts to disable KVM_MEM_GUEST_MEMFD on a memslot that was
initially created with a guest_memfd binding, as KVM doesn't support
toggling KVM_MEM_GUEST_MEMFD on existing memslots. KVM prevents enabling
KVM_MEM_GUEST_MEMFD, but doesn't prevent clearing the flag.
Failure to reject the new memslot results in a use-after-free due to KVM
not unbinding from the guest_memfd instance. Unbinding on a FLAGS_ONLY
change is easy enough, and can/will be done as a hardening measure (in
anticipation of KVM supporting dirty logging on guest_memfd at some point),
but fixing the use-after-free would only address the immediate symptom.
==================================================================
BUG: KASAN: slab-use-after-free in kvm_gmem_release+0x362/0x400 [kvm]
Write of size 8 at addr ffff8881111ae908 by task repro/745
CPU: 7 UID: 1000 PID: 745 Comm: repro Not tainted 6.18.0-rc6-115d5de2eef3-next-kasan #3 NONE
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
Call Trace:
<TASK>
dump_stack_lvl+0x51/0x60
print_report+0xcb/0x5c0
kasan_report+0xb4/0xe0
kvm_gmem_release+0x362/0x400 [kvm]
__fput+0x2fa/0x9d0
task_work_run+0x12c/0x200
do_exit+0x6ae/0x2100
do_group_exit+0xa8/0x230
__x64_sys_exit_group+0x3a/0x50
x64_sys_call+0x737/0x740
do_syscall_64+0x5b/0x900
entry_SYSCALL_64_after_hwframe+0x4b/0x53
RIP: 0033:0x7f581f2eac31
</TASK>
Allocated by task 745 on cpu 6 at 9.746971s:
kasan_save_stack+0x20/0x40
kasan_save_track+0x13/0x50
__kasan_kmalloc+0x77/0x90
kvm_set_memory_region.part.0+0x652/0x1110 [kvm]
kvm_vm_ioctl+0x14b0/0x3290 [kvm]
__x64_sys_ioctl+0x129/0x1a0
do_syscall_64+0x5b/0x900
entry_SYSCALL_64_after_hwframe+0x4b/0x53
Freed by task 745 on cpu 6 at 9.747467s:
kasan_save_stack+0x20/0x40
kasan_save_track+0x13/0x50
__kasan_save_free_info+0x37/0x50
__kasan_slab_free+0x3b/0x60
kfree+0xf5/0x440
kvm_set_memslot+0x3c2/0x1160 [kvm]
kvm_set_memory_region.part.0+0x86a/0x1110 [kvm]
kvm_vm_ioctl+0x14b0/0x3290 [kvm]
__x64_sys_ioctl+0x129/0x1a0
do_syscall_64+0x5b/0x900
entry_SYSCALL_64_after_hwframe+0x4b/0x53
Reported-by: Alexander Potapenko <glider@google.com>
Fixes: a7800aa80e ("KVM: Add KVM_CREATE_GUEST_MEMFD ioctl() for guest-specific backing memory")
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20251202020334.1171351-2-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit ae431059e7 ]
When unbinding a memslot from a guest_memfd instance, remove the bindings
even if the guest_memfd file is dying, i.e. even if its file refcount has
gone to zero. If the memslot is freed before the file is fully released,
nullifying the memslot side of the binding in kvm_gmem_release() will
write to freed memory, as detected by syzbot+KASAN:
==================================================================
BUG: KASAN: slab-use-after-free in kvm_gmem_release+0x176/0x440 virt/kvm/guest_memfd.c:353
Write of size 8 at addr ffff88807befa508 by task syz.0.17/6022
CPU: 0 UID: 0 PID: 6022 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT(full)
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/02/2025
Call Trace:
<TASK>
dump_stack_lvl+0x189/0x250 lib/dump_stack.c:120
print_address_description mm/kasan/report.c:378 [inline]
print_report+0xca/0x240 mm/kasan/report.c:482
kasan_report+0x118/0x150 mm/kasan/report.c:595
kvm_gmem_release+0x176/0x440 virt/kvm/guest_memfd.c:353
__fput+0x44c/0xa70 fs/file_table.c:468
task_work_run+0x1d4/0x260 kernel/task_work.c:227
resume_user_mode_work include/linux/resume_user_mode.h:50 [inline]
exit_to_user_mode_loop+0xe9/0x130 kernel/entry/common.c:43
exit_to_user_mode_prepare include/linux/irq-entry-common.h:225 [inline]
syscall_exit_to_user_mode_work include/linux/entry-common.h:175 [inline]
syscall_exit_to_user_mode include/linux/entry-common.h:210 [inline]
do_syscall_64+0x2bd/0xfa0 arch/x86/entry/syscall_64.c:100
entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7fbeeff8efc9
</TASK>
Allocated by task 6023:
kasan_save_stack mm/kasan/common.c:56 [inline]
kasan_save_track+0x3e/0x80 mm/kasan/common.c:77
poison_kmalloc_redzone mm/kasan/common.c:397 [inline]
__kasan_kmalloc+0x93/0xb0 mm/kasan/common.c:414
kasan_kmalloc include/linux/kasan.h:262 [inline]
__kmalloc_cache_noprof+0x3e2/0x700 mm/slub.c:5758
kmalloc_noprof include/linux/slab.h:957 [inline]
kzalloc_noprof include/linux/slab.h:1094 [inline]
kvm_set_memory_region+0x747/0xb90 virt/kvm/kvm_main.c:2104
kvm_vm_ioctl_set_memory_region+0x6f/0xd0 virt/kvm/kvm_main.c:2154
kvm_vm_ioctl+0x957/0xc60 virt/kvm/kvm_main.c:5201
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:597 [inline]
__se_sys_ioctl+0xfc/0x170 fs/ioctl.c:583
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f
Freed by task 6023:
kasan_save_stack mm/kasan/common.c:56 [inline]
kasan_save_track+0x3e/0x80 mm/kasan/common.c:77
kasan_save_free_info+0x46/0x50 mm/kasan/generic.c:584
poison_slab_object mm/kasan/common.c:252 [inline]
__kasan_slab_free+0x5c/0x80 mm/kasan/common.c:284
kasan_slab_free include/linux/kasan.h:234 [inline]
slab_free_hook mm/slub.c:2533 [inline]
slab_free mm/slub.c:6622 [inline]
kfree+0x19a/0x6d0 mm/slub.c:6829
kvm_set_memory_region+0x9c4/0xb90 virt/kvm/kvm_main.c:2130
kvm_vm_ioctl_set_memory_region+0x6f/0xd0 virt/kvm/kvm_main.c:2154
kvm_vm_ioctl+0x957/0xc60 virt/kvm/kvm_main.c:5201
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:597 [inline]
__se_sys_ioctl+0xfc/0x170 fs/ioctl.c:583
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f
Deliberately don't acquire filemap invalid lock when the file is dying as
the lifecycle of f_mapping is outside the purview of KVM. Dereferencing
the mapping is *probably* fine, but there's no need to invalidate anything
as memslot deletion is responsible for zapping SPTEs, and the only code
that can access the dying file is kvm_gmem_release(), whose core code is
mutually exclusive with unbinding.
Note, the mutual exclusivity is also what makes it safe to access the
bindings on a dying gmem instance. Unbinding either runs with slots_lock
held, or after the last reference to the owning "struct kvm" is put, and
kvm_gmem_release() nullifies the slot pointer under slots_lock, and puts
its reference to the VM after that is done.
Reported-by: syzbot+2479e53d0db9b32ae2aa@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/68fa7a22.a70a0220.3bf6c6.008b.GAE@google.com
Tested-by: syzbot+2479e53d0db9b32ae2aa@syzkaller.appspotmail.com
Fixes: a7800aa80e ("KVM: Add KVM_CREATE_GUEST_MEMFD ioctl() for guest-specific backing memory")
Cc: stable@vger.kernel.org
Cc: Hillf Danton <hdanton@sina.com>
Reviewed-By: Vishal Annapurve <vannapurve@google.com>
Link: https://patch.msgid.link/20251104011205.3853541-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 67b43038ce ]
Remove the RCU-protected attribute from slot->gmem.file. No need to use RCU
primitives rcu_assign_pointer()/synchronize_rcu() to update this pointer.
- slot->gmem.file is updated in 3 places:
kvm_gmem_bind(), kvm_gmem_unbind(), kvm_gmem_release().
All of them are protected by kvm->slots_lock.
- slot->gmem.file is read in 2 paths:
(1) kvm_gmem_populate
kvm_gmem_get_file
__kvm_gmem_get_pfn
(2) kvm_gmem_get_pfn
kvm_gmem_get_file
__kvm_gmem_get_pfn
Path (1) kvm_gmem_populate() requires holding kvm->slots_lock, so
slot->gmem.file is protected by the kvm->slots_lock in this path.
Path (2) kvm_gmem_get_pfn() does not require holding kvm->slots_lock.
However, it's also not guarded by rcu_read_lock() and rcu_read_unlock().
So synchronize_rcu() in kvm_gmem_unbind()/kvm_gmem_release() actually
will not wait for the readers in kvm_gmem_get_pfn() due to lack of RCU
read-side critical section.
The path (2) kvm_gmem_get_pfn() is safe without RCU protection because:
a) kvm_gmem_bind() is called on a new memslot, before the memslot is
visible to kvm_gmem_get_pfn().
b) kvm->srcu ensures that kvm_gmem_unbind() and freeing of a memslot
occur after the memslot is no longer visible to kvm_gmem_get_pfn().
c) get_file_active() ensures that kvm_gmem_get_pfn() will not access the
stale file if kvm_gmem_release() sets it to NULL. This is because if
kvm_gmem_release() occurs before kvm_gmem_get_pfn(), get_file_active()
will return NULL; if get_file_active() does not return NULL,
kvm_gmem_release() should not occur until after kvm_gmem_get_pfn()
releases the file reference.
Signed-off-by: Yan Zhao <yan.y.zhao@intel.com>
Message-ID: <20241104084303.29909-1-yan.y.zhao@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Stable-dep-of: ae431059e7 ("KVM: guest_memfd: Remove bindings on memslot deletion when gmem is dying")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 4af18dc6a9 ]
Refactor guest_memfd usage of __kvm_gmem_get_pfn() to pass the index into
the guest_memfd file instead of the gfn, i.e. resolve the index based on
the slot+gfn in the caller instead of in __kvm_gmem_get_pfn(). This will
allow kvm_gmem_get_pfn() to retrieve and return the specific "struct page",
which requires the index into the folio, without a redoing the index
calculation multiple times (which isn't costly, just hard to follow).
Opportunistically add a kvm_gmem_get_index() helper to make the copy+pasted
code easier to understand.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <20241010182427.1434605-46-seanjc@google.com>
Stable-dep-of: ae431059e7 ("KVM: guest_memfd: Remove bindings on memslot deletion when gmem is dying")
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit dca6c88532 ]
Add new members to strut kvm_gfn_range to indicate which mapping
(private-vs-shared) to operate on: enum kvm_gfn_range_filter
attr_filter. Update the core zapping operations to set them appropriately.
TDX utilizes two GPA aliases for the same memslots, one for memory that is
for private memory and one that is for shared. For private memory, KVM
cannot always perform the same operations it does on memory for default
VMs, such as zapping pages and having them be faulted back in, as this
requires guest coordination. However, some operations such as guest driven
conversion of memory between private and shared should zap private memory.
Internally to the MMU, private and shared mappings are tracked on separate
roots. Mapping and zapping operations will operate on the respective GFN
alias for each root (private or shared). So zapping operations will by
default zap both aliases. Add fields in struct kvm_gfn_range to allow
callers to specify which aliases so they can only target the aliases
appropriate for their specific operation.
There was feedback that target aliases should be specified such that the
default value (0) is to operate on both aliases. Several options were
considered. Several variations of having separate bools defined such
that the default behavior was to process both aliases. They either allowed
nonsensical configurations, or were confusing for the caller. A simple
enum was also explored and was close, but was hard to process in the
caller. Instead, use an enum with the default value (0) reserved as a
disallowed value. Catch ranges that didn't have the target aliases
specified by looking for that specific value.
Set target alias with enum appropriately for these MMU operations:
- For KVM's mmu notifier callbacks, zap shared pages only because private
pages won't have a userspace mapping
- For setting memory attributes, kvm_arch_pre_set_memory_attributes()
chooses the aliases based on the attribute.
- For guest_memfd invalidations, zap private only.
Link: https://lore.kernel.org/kvm/ZivIF9vjKcuGie3s@google.com/
Signed-off-by: Isaku Yamahata <isaku.yamahata@intel.com>
Co-developed-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Message-ID: <20240718211230.1492011-3-rick.p.edgecombe@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Stable-dep-of: 9129633d56 ("KVM: x86/mmu: Prevent installing hugepages when mem attributes are changing")
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 459a35111b upstream.
Convert HAVE_KVM_IRQ_BYPASS into a tristate so that selecting
IRQ_BYPASS_MANAGER follows KVM={m,y}, i.e. doesn't force irqbypass.ko to
be built-in.
Note, PPC allows building KVM as a module, but selects HAVE_KVM_IRQ_BYPASS
from a boolean Kconfig, i.e. KVM PPC unnecessarily forces irqbpass.ko to
be built-in. But that flaw is a longstanding PPC specific issue.
Fixes: 61df71ee99 ("kvm: move "select IRQ_BYPASS_MANAGER" to common code")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20250315024623.2363994-1-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d96c77bd4e upstream.
kvm_vm_create_worker_thread() is meant to be used for kthreads that
can consume significant amounts of CPU time on behalf of a VM or in
response to how the VM behaves (for example how it accesses its memory).
Therefore it wants to charge the CPU time consumed by that work to
the VM's container.
However, because of these threads, cgroups which have kvm instances
inside never complete freezing. This can be trivially reproduced:
root@test ~# mkdir /sys/fs/cgroup/test
root@test ~# echo $$ > /sys/fs/cgroup/test/cgroup.procs
root@test ~# qemu-system-x86_64 -nographic -enable-kvm
and in another terminal:
root@test ~# echo 1 > /sys/fs/cgroup/test/cgroup.freeze
root@test ~# cat /sys/fs/cgroup/test/cgroup.events
populated 1
frozen 0
The cgroup freezing happens in the signal delivery path but
kvm_nx_huge_page_recovery_worker, while joining non-root cgroups, never
calls into the signal delivery path and thus never gets frozen. Because
the cgroup freezer determines whether a given cgroup is frozen by
comparing the number of frozen threads to the total number of threads
in the cgroup, the cgroup never becomes frozen and users waiting for
the state transition may hang indefinitely.
Since the worker kthread is tied to a user process, it's better if
it behaves similarly to user tasks as much as possible, including
being able to send SIGSTOP and SIGCONT. In fact, vhost_task is all
that kvm_vm_create_worker_thread() wanted to be and more: not only it
inherits the userspace process's cgroups, it has other niceties like
being parented properly in the process tree. Use it instead of the
homegrown alternative.
Incidentally, the new code is also better behaved when you flip recovery
back and forth to disabled and back to enabled. If your recovery period
is 1 minute, it will run the next recovery after 1 minute independent
of how many times you flipped the parameter.
(Commit message based on emails from Tejun).
Reported-by: Tejun Heo <tj@kernel.org>
Reported-by: Luca Boccassi <bluca@debian.org>
Acked-by: Tejun Heo <tj@kernel.org>
Tested-by: Luca Boccassi <bluca@debian.org>
Cc: stable@vger.kernel.org
Reviewed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Pull kvm fixes from Paolo Bonzini:
"ARM64:
- Fix the guest view of the ID registers, making the relevant fields
writable from userspace (affecting ID_AA64DFR0_EL1 and
ID_AA64PFR1_EL1)
- Correcly expose S1PIE to guests, fixing a regression introduced in
6.12-rc1 with the S1POE support
- Fix the recycling of stage-2 shadow MMUs by tracking the context
(are we allowed to block or not) as well as the recycling state
- Address a couple of issues with the vgic when userspace
misconfigures the emulation, resulting in various splats. Headaches
courtesy of our Syzkaller friends
- Stop wasting space in the HYP idmap, as we are dangerously close to
the 4kB limit, and this has already exploded in -next
- Fix another race in vgic_init()
- Fix a UBSAN error when faking the cache topology with MTE enabled
RISCV:
- RISCV: KVM: use raw_spinlock for critical section in imsic
x86:
- A bandaid for lack of XCR0 setup in selftests, which causes trouble
if the compiler is configured to have x86-64-v3 (with AVX) as the
default ISA. Proper XCR0 setup will come in the next merge window.
- Fix an issue where KVM would not ignore low bits of the nested CR3
and potentially leak up to 31 bytes out of the guest memory's
bounds
- Fix case in which an out-of-date cached value for the segments
could by returned by KVM_GET_SREGS.
- More cleanups for KVM_X86_QUIRK_SLOT_ZAP_ALL
- Override MTRR state for KVM confidential guests, making it WB by
default as is already the case for Hyper-V guests.
Generic:
- Remove a couple of unused functions"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (27 commits)
RISCV: KVM: use raw_spinlock for critical section in imsic
KVM: selftests: Fix out-of-bounds reads in CPUID test's array lookups
KVM: selftests: x86: Avoid using SSE/AVX instructions
KVM: nSVM: Ignore nCR3[4:0] when loading PDPTEs from memory
KVM: VMX: reset the segment cache after segment init in vmx_vcpu_reset()
KVM: x86: Clean up documentation for KVM_X86_QUIRK_SLOT_ZAP_ALL
KVM: x86/mmu: Add lockdep assert to enforce safe usage of kvm_unmap_gfn_range()
KVM: x86/mmu: Zap only SPs that shadow gPTEs when deleting memslot
x86/kvm: Override default caching mode for SEV-SNP and TDX
KVM: Remove unused kvm_vcpu_gfn_to_pfn_atomic
KVM: Remove unused kvm_vcpu_gfn_to_pfn
KVM: arm64: Ensure vgic_ready() is ordered against MMIO registration
KVM: arm64: vgic: Don't check for vgic_ready() when setting NR_IRQS
KVM: arm64: Fix shift-out-of-bounds bug
KVM: arm64: Shave a few bytes from the EL2 idmap code
KVM: arm64: Don't eagerly teardown the vgic on init error
KVM: arm64: Expose S1PIE to guests
KVM: arm64: nv: Clarify safety of allowing TLBI unmaps to reschedule
KVM: arm64: nv: Punt stage-2 recycling to a vCPU request
KVM: arm64: nv: Do not block when unmapping stage-2 if disallowed
...
Sean noted that ever since commit 152e11f6df ("sched/fair: Implement
delayed dequeue") KVM's preemption notifiers have started
mis-classifying preemption vs blocking.
Notably p->on_rq is no longer sufficient to determine if a task is
runnable or blocked -- the aforementioned commit introduces tasks that
remain on the runqueue even through they will not run again, and
should be considered blocked for many cases.
Add the task_is_runnable() helper to classify things and audit all
external users of the p->on_rq state. Also add a few comments.
Fixes: 152e11f6df ("sched/fair: Implement delayed dequeue")
Reported-by: Sean Christopherson <seanjc@google.com>
Tested-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20241010091843.GK33184@noisy.programming.kicks-ass.net
Pull x86 kvm updates from Paolo Bonzini:
"x86:
- KVM currently invalidates the entirety of the page tables, not just
those for the memslot being touched, when a memslot is moved or
deleted.
This does not traditionally have particularly noticeable overhead,
but Intel's TDX will require the guest to re-accept private pages
if they are dropped from the secure EPT, which is a non starter.
Actually, the only reason why this is not already being done is a
bug which was never fully investigated and caused VM instability
with assigned GeForce GPUs, so allow userspace to opt into the new
behavior.
- Advertise AVX10.1 to userspace (effectively prep work for the
"real" AVX10 functionality that is on the horizon)
- Rework common MSR handling code to suppress errors on userspace
accesses to unsupported-but-advertised MSRs
This will allow removing (almost?) all of KVM's exemptions for
userspace access to MSRs that shouldn't exist based on the vCPU
model (the actual cleanup is non-trivial future work)
- Rework KVM's handling of x2APIC ICR, again, because AMD (x2AVIC)
splits the 64-bit value into the legacy ICR and ICR2 storage,
whereas Intel (APICv) stores the entire 64-bit value at the ICR
offset
- Fix a bug where KVM would fail to exit to userspace if one was
triggered by a fastpath exit handler
- Add fastpath handling of HLT VM-Exit to expedite re-entering the
guest when there's already a pending wake event at the time of the
exit
- Fix a WARN caused by RSM entering a nested guest from SMM with
invalid guest state, by forcing the vCPU out of guest mode prior to
signalling SHUTDOWN (the SHUTDOWN hits the VM altogether, not the
nested guest)
- Overhaul the "unprotect and retry" logic to more precisely identify
cases where retrying is actually helpful, and to harden all retry
paths against putting the guest into an infinite retry loop
- Add support for yielding, e.g. to honor NEED_RESCHED, when zapping
rmaps in the shadow MMU
- Refactor pieces of the shadow MMU related to aging SPTEs in
prepartion for adding multi generation LRU support in KVM
- Don't stuff the RSB after VM-Exit when RETPOLINE=y and AutoIBRS is
enabled, i.e. when the CPU has already flushed the RSB
- Trace the per-CPU host save area as a VMCB pointer to improve
readability and cleanup the retrieval of the SEV-ES host save area
- Remove unnecessary accounting of temporary nested VMCB related
allocations
- Set FINAL/PAGE in the page fault error code for EPT violations if
and only if the GVA is valid. If the GVA is NOT valid, there is no
guest-side page table walk and so stuffing paging related metadata
is nonsensical
- Fix a bug where KVM would incorrectly synthesize a nested VM-Exit
instead of emulating posted interrupt delivery to L2
- Add a lockdep assertion to detect unsafe accesses of vmcs12
structures
- Harden eVMCS loading against an impossible NULL pointer deref
(really truly should be impossible)
- Minor SGX fix and a cleanup
- Misc cleanups
Generic:
- Register KVM's cpuhp and syscore callbacks when enabling
virtualization in hardware, as the sole purpose of said callbacks
is to disable and re-enable virtualization as needed
- Enable virtualization when KVM is loaded, not right before the
first VM is created
Together with the previous change, this simplifies a lot the logic
of the callbacks, because their very existence implies
virtualization is enabled
- Fix a bug that results in KVM prematurely exiting to userspace for
coalesced MMIO/PIO in many cases, clean up the related code, and
add a testcase
- Fix a bug in kvm_clear_guest() where it would trigger a buffer
overflow _if_ the gpa+len crosses a page boundary, which thankfully
is guaranteed to not happen in the current code base. Add WARNs in
more helpers that read/write guest memory to detect similar bugs
Selftests:
- Fix a goof that caused some Hyper-V tests to be skipped when run on
bare metal, i.e. NOT in a VM
- Add a regression test for KVM's handling of SHUTDOWN for an SEV-ES
guest
- Explicitly include one-off assets in .gitignore. Past Sean was
completely wrong about not being able to detect missing .gitignore
entries
- Verify userspace single-stepping works when KVM happens to handle a
VM-Exit in its fastpath
- Misc cleanups"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (127 commits)
Documentation: KVM: fix warning in "make htmldocs"
s390: Enable KVM_S390_UCONTROL config in debug_defconfig
selftests: kvm: s390: Add VM run test case
KVM: SVM: let alternatives handle the cases when RSB filling is required
KVM: VMX: Set PFERR_GUEST_{FINAL,PAGE}_MASK if and only if the GVA is valid
KVM: x86/mmu: Use KVM_PAGES_PER_HPAGE() instead of an open coded equivalent
KVM: x86/mmu: Add KVM_RMAP_MANY to replace open coded '1' and '1ul' literals
KVM: x86/mmu: Fold mmu_spte_age() into kvm_rmap_age_gfn_range()
KVM: x86/mmu: Morph kvm_handle_gfn_range() into an aging specific helper
KVM: x86/mmu: Honor NEED_RESCHED when zapping rmaps and blocking is allowed
KVM: x86/mmu: Add a helper to walk and zap rmaps for a memslot
KVM: x86/mmu: Plumb a @can_yield parameter into __walk_slot_rmaps()
KVM: x86/mmu: Move walk_slot_rmaps() up near for_each_slot_rmap_range()
KVM: x86/mmu: WARN on MMIO cache hit when emulating write-protected gfn
KVM: x86/mmu: Detect if unprotect will do anything based on invalid_list
KVM: x86/mmu: Subsume kvm_mmu_unprotect_page() into the and_retry() version
KVM: x86: Rename reexecute_instruction()=>kvm_unprotect_and_retry_on_failure()
KVM: x86: Update retry protection fields when forcing retry on emulation failure
KVM: x86: Apply retry protection to "unprotect on failure" path
KVM: x86: Check EMULTYPE_WRITE_PF_TO_SP before unprotecting gfn
...
no_llseek had been defined to NULL two years ago, in commit 868941b144
("fs: remove no_llseek")
To quote that commit,
At -rc1 we'll need do a mechanical removal of no_llseek -
git grep -l -w no_llseek | grep -v porting.rst | while read i; do
sed -i '/\<no_llseek\>/d' $i
done
would do it.
Unfortunately, that hadn't been done. Linus, could you do that now, so
that we could finally put that thing to rest? All instances are of the
form
.llseek = no_llseek,
so it's obviously safe.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull 'struct fd' updates from Al Viro:
"Just the 'struct fd' layout change, with conversion to accessor
helpers"
* tag 'pull-stable-struct_fd' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
add struct fd constructors, get rid of __to_fd()
struct fd: representation change
introduce fd_file(), convert all accessors to it.
KVK generic changes for 6.12:
- Fix a bug that results in KVM prematurely exiting to userspace for coalesced
MMIO/PIO in many cases, clean up the related code, and add a testcase.
- Fix a bug in kvm_clear_guest() where it would trigger a buffer overflow _if_
the gpa+len crosses a page boundary, which thankfully is guaranteed to not
happen in the current code base. Add WARNs in more helpers that read/write
guest memory to detect similar bugs.
When reading or writing a guest page, WARN and bail if offset+len would
result in a read to a different page so that KVM bugs are more likely to
be detected, and so that any such bugs are less likely to escalate to an
out-of-bounds access. E.g. if userspace isn't using guard pages and the
target page is at the end of a memslot.
Note, KVM already hardens itself in similar APIs, e.g. in the "cached"
variants, it's just the vanilla APIs that are playing with fire.
Link: https://lore.kernel.org/r/20240829191413.900740-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Add arch hooks that are invoked when KVM enables/disable virtualization.
x86 will use the hooks to register an "emergency disable" callback, which
is essentially an x86-specific shutdown notifier that is used when the
kernel is doing an emergency reboot/shutdown/kexec.
Add comments for the declarations to help arch code understand exactly
when the callbacks are invoked. Alternatively, the APIs themselves could
communicate most of the same info, but kvm_arch_pre_enable_virtualization()
and kvm_arch_post_disable_virtualization() are a bit cumbersome, and make
it a bit less obvious that they are intended to be implemented as a pair.
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Acked-by: Kai Huang <kai.huang@intel.com>
Tested-by: Farrah Chen <farrah.chen@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20240830043600.127750-9-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add an on-by-default module param, enable_virt_at_load, to let userspace
force virtualization to be enabled in hardware when KVM is initialized,
i.e. just before /dev/kvm is exposed to userspace. Enabling virtualization
during KVM initialization allows userspace to avoid the additional latency
when creating/destroying the first/last VM (or more specifically, on the
0=>1 and 1=>0 edges of creation/destruction).
Now that KVM uses the cpuhp framework to do per-CPU enabling, the latency
could be non-trivial as the cpuhup bringup/teardown is serialized across
CPUs, e.g. the latency could be problematic for use case that need to spin
up VMs quickly.
Prior to commit 10474ae894 ("KVM: Activate Virtualization On Demand"),
KVM _unconditionally_ enabled virtualization during load, i.e. there's no
fundamental reason KVM needs to dynamically toggle virtualization. These
days, the only known argument for not enabling virtualization is to allow
KVM to be autoloaded without blocking other out-of-tree hypervisors, and
such use cases can simply change the module param, e.g. via command line.
Note, the aforementioned commit also mentioned that enabling SVM (AMD's
virtualization extensions) can result in "using invalid TLB entries".
It's not clear whether the changelog was referring to a KVM bug, a CPU
bug, or something else entirely. Regardless, leaving virtualization off
by default is not a robust "fix", as any protection provided is lost the
instant userspace creates the first VM.
Reviewed-by: Chao Gao <chao.gao@intel.com>
Acked-by: Kai Huang <kai.huang@intel.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Tested-by: Farrah Chen <farrah.chen@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20240830043600.127750-8-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Rename the per-CPU hooks used to enable virtualization in hardware to
align with the KVM-wide helpers in kvm_main.c, and to better capture that
the callbacks are invoked on every online CPU.
No functional change intended.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Message-ID: <20240830043600.127750-5-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Rename the various functions (and a variable) that enable virtualization
to prepare for upcoming changes, and to clean up artifacts of KVM's
previous behavior, which required manually juggling locks around
kvm_usage_count.
Drop the "nolock" qualifier from per-CPU functions now that there are no
"nolock" implementations of the "all" variants, i.e. now that calling a
non-nolock function from a nolock function isn't confusing (unlike this
sentence).
Drop "all" from the outer helpers as they no longer manually iterate
over all CPUs, and because it might not be obvious what "all" refers to.
In lieu of the above qualifiers, append "_cpu" to the end of the functions
that are per-CPU helpers for the outer APIs.
Opportunistically prepend "kvm" to all functions to help make it clear
that they are KVM helpers, but mostly because there's no reason not to.
Lastly, use "virtualization" instead of "hardware", because while the
functions do enable virtualization in hardware, there are a _lot_ of
things that KVM enables in hardware.
Defer renaming the arch hooks to future patches, purely to reduce the
amount of churn in a single commit.
Reviewed-by: Chao Gao <chao.gao@intel.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Acked-by: Kai Huang <kai.huang@intel.com>
Tested-by: Farrah Chen <farrah.chen@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20240830043600.127750-4-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Register KVM's cpuhp and syscore callback when enabling virtualization
in hardware instead of registering the callbacks during initialization,
and let the CPU up/down framework invoke the inner enable/disable
functions. Registering the callbacks during initialization makes things
more complex than they need to be, as KVM needs to be very careful about
handling races between enabling CPUs being onlined/offlined and hardware
being enabled/disabled.
Intel TDX support will require KVM to enable virtualization during KVM
initialization, i.e. will add another wrinkle to things, at which point
sorting out the potential races with kvm_usage_count would become even
more complex.
Note, using the cpuhp framework has a subtle behavioral change: enabling
will be done serially across all CPUs, whereas KVM currently sends an IPI
to all CPUs in parallel. While serializing virtualization enabling could
create undesirable latency, the issue is limited to the 0=>1 transition of
VM creation. And even that can be mitigated, e.g. by letting userspace
force virtualization to be enabled when KVM is initialized.
Cc: Chao Gao <chao.gao@intel.com>
Reviewed-by: Kai Huang <kai.huang@intel.com>
Acked-by: Kai Huang <kai.huang@intel.com>
Tested-by: Farrah Chen <farrah.chen@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20240830043600.127750-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use a dedicated mutex to guard kvm_usage_count to fix a potential deadlock
on x86 due to a chain of locks and SRCU synchronizations. Translating the
below lockdep splat, CPU1 #6 will wait on CPU0 #1, CPU0 #8 will wait on
CPU2 #3, and CPU2 #7 will wait on CPU1 #4 (if there's a writer, due to the
fairness of r/w semaphores).
CPU0 CPU1 CPU2
1 lock(&kvm->slots_lock);
2 lock(&vcpu->mutex);
3 lock(&kvm->srcu);
4 lock(cpu_hotplug_lock);
5 lock(kvm_lock);
6 lock(&kvm->slots_lock);
7 lock(cpu_hotplug_lock);
8 sync(&kvm->srcu);
Note, there are likely more potential deadlocks in KVM x86, e.g. the same
pattern of taking cpu_hotplug_lock outside of kvm_lock likely exists with
__kvmclock_cpufreq_notifier():
cpuhp_cpufreq_online()
|
-> cpufreq_online()
|
-> cpufreq_gov_performance_limits()
|
-> __cpufreq_driver_target()
|
-> __target_index()
|
-> cpufreq_freq_transition_begin()
|
-> cpufreq_notify_transition()
|
-> ... __kvmclock_cpufreq_notifier()
But, actually triggering such deadlocks is beyond rare due to the
combination of dependencies and timings involved. E.g. the cpufreq
notifier is only used on older CPUs without a constant TSC, mucking with
the NX hugepage mitigation while VMs are running is very uncommon, and
doing so while also onlining/offlining a CPU (necessary to generate
contention on cpu_hotplug_lock) would be even more unusual.
The most robust solution to the general cpu_hotplug_lock issue is likely
to switch vm_list to be an RCU-protected list, e.g. so that x86's cpufreq
notifier doesn't to take kvm_lock. For now, settle for fixing the most
blatant deadlock, as switching to an RCU-protected list is a much more
involved change, but add a comment in locking.rst to call out that care
needs to be taken when walking holding kvm_lock and walking vm_list.
======================================================
WARNING: possible circular locking dependency detected
6.10.0-smp--c257535a0c9d-pip #330 Tainted: G S O
------------------------------------------------------
tee/35048 is trying to acquire lock:
ff6a80eced71e0a8 (&kvm->slots_lock){+.+.}-{3:3}, at: set_nx_huge_pages+0x179/0x1e0 [kvm]
but task is already holding lock:
ffffffffc07abb08 (kvm_lock){+.+.}-{3:3}, at: set_nx_huge_pages+0x14a/0x1e0 [kvm]
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #3 (kvm_lock){+.+.}-{3:3}:
__mutex_lock+0x6a/0xb40
mutex_lock_nested+0x1f/0x30
kvm_dev_ioctl+0x4fb/0xe50 [kvm]
__se_sys_ioctl+0x7b/0xd0
__x64_sys_ioctl+0x21/0x30
x64_sys_call+0x15d0/0x2e60
do_syscall_64+0x83/0x160
entry_SYSCALL_64_after_hwframe+0x76/0x7e
-> #2 (cpu_hotplug_lock){++++}-{0:0}:
cpus_read_lock+0x2e/0xb0
static_key_slow_inc+0x16/0x30
kvm_lapic_set_base+0x6a/0x1c0 [kvm]
kvm_set_apic_base+0x8f/0xe0 [kvm]
kvm_set_msr_common+0x9ae/0xf80 [kvm]
vmx_set_msr+0xa54/0xbe0 [kvm_intel]
__kvm_set_msr+0xb6/0x1a0 [kvm]
kvm_arch_vcpu_ioctl+0xeca/0x10c0 [kvm]
kvm_vcpu_ioctl+0x485/0x5b0 [kvm]
__se_sys_ioctl+0x7b/0xd0
__x64_sys_ioctl+0x21/0x30
x64_sys_call+0x15d0/0x2e60
do_syscall_64+0x83/0x160
entry_SYSCALL_64_after_hwframe+0x76/0x7e
-> #1 (&kvm->srcu){.+.+}-{0:0}:
__synchronize_srcu+0x44/0x1a0
synchronize_srcu_expedited+0x21/0x30
kvm_swap_active_memslots+0x110/0x1c0 [kvm]
kvm_set_memslot+0x360/0x620 [kvm]
__kvm_set_memory_region+0x27b/0x300 [kvm]
kvm_vm_ioctl_set_memory_region+0x43/0x60 [kvm]
kvm_vm_ioctl+0x295/0x650 [kvm]
__se_sys_ioctl+0x7b/0xd0
__x64_sys_ioctl+0x21/0x30
x64_sys_call+0x15d0/0x2e60
do_syscall_64+0x83/0x160
entry_SYSCALL_64_after_hwframe+0x76/0x7e
-> #0 (&kvm->slots_lock){+.+.}-{3:3}:
__lock_acquire+0x15ef/0x2e30
lock_acquire+0xe0/0x260
__mutex_lock+0x6a/0xb40
mutex_lock_nested+0x1f/0x30
set_nx_huge_pages+0x179/0x1e0 [kvm]
param_attr_store+0x93/0x100
module_attr_store+0x22/0x40
sysfs_kf_write+0x81/0xb0
kernfs_fop_write_iter+0x133/0x1d0
vfs_write+0x28d/0x380
ksys_write+0x70/0xe0
__x64_sys_write+0x1f/0x30
x64_sys_call+0x281b/0x2e60
do_syscall_64+0x83/0x160
entry_SYSCALL_64_after_hwframe+0x76/0x7e
Cc: Chao Gao <chao.gao@intel.com>
Fixes: 0bf50497f0 ("KVM: Drop kvm_count_lock and instead protect kvm_usage_count with kvm_lock")
Cc: stable@vger.kernel.org
Reviewed-by: Kai Huang <kai.huang@intel.com>
Acked-by: Kai Huang <kai.huang@intel.com>
Tested-by: Farrah Chen <farrah.chen@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20240830043600.127750-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Fold coalesced_mmio_has_room() into its sole caller, coalesced_mmio_write(),
as it's really just a single line of code, has a goofy return value, and
is unnecessarily brittle.
E.g. if coalesced_mmio_has_room() were to check ring->last directly, or
the caller failed to use READ_ONCE(), KVM would be susceptible to TOCTOU
attacks from userspace.
Opportunistically add a comment explaining why on earth KVM leaves one
entry free, which may not be obvious to readers that aren't familiar with
ring buffers.
No functional change intended.
Reviewed-by: Ilias Stamatis <ilstam@amazon.com>
Cc: Paul Durrant <paul@xen.org>
Link: https://lore.kernel.org/r/20240828181446.652474-3-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
The following calculation used in coalesced_mmio_has_room() to check
whether the ring buffer is full is wrong and results in premature exits if
the start of the valid entries is in the first half of the ring buffer.
avail = (ring->first - last - 1) % KVM_COALESCED_MMIO_MAX;
if (avail == 0)
/* full */
Because negative values are handled using two's complement, and KVM
computes the result as an unsigned value, the above will get a false
positive if "first < last" and the ring is half-full.
The above might have worked as expected in python for example:
>>> (-86) % 170
84
However it doesn't work the same way in C.
printf("avail: %d\n", (-86) % 170);
printf("avail: %u\n", (-86) % 170);
printf("avail: %u\n", (-86u) % 170u);
Using gcc-11 these print:
avail: -86
avail: 4294967210
avail: 0
For illustration purposes, given a 4-bit integer and a ring size of 0xA
(unsigned), 0xA == 0x1010 == -6, and thus (-6u % 0xA) == 0.
Fix the calculation and allow all but one entries in the buffer to be
used as originally intended.
Note, KVM's behavior is self-healing to some extent, as KVM will allow the
entire buffer to be used if ring->first is beyond the halfway point. In
other words, in the unlikely scenario that a use case benefits from being
able to coalesce more than 86 entries at once, KVM will still provide such
behavior, sometimes.
Note #2, the % operator in C is not the modulo operator but the remainder
operator. Modulo and remainder operators differ with respect to negative
values. But, the relevant values in KVM are all unsigned, so it's a moot
point in this case anyway.
Note #3, this is almost a pure revert of the buggy commit, plus a
READ_ONCE() to provide additional safety. Thue buggy commit justified the
change with "it paves the way for making this function lockless", but it's
not at all clear what was intended, nor is there any evidence that the
buggy code was somehow safer. (a) the fields in question were already
accessed locklessly, from the perspective that they could be modified by
userspace at any time, and (b) the lock guarding the ring itself was
changed, but never dropped, i.e. whatever lockless scheme (SRCU?) was
planned never landed.
Fixes: 105f8d40a7 ("KVM: Calculate available entries in coalesced mmio ring")
Signed-off-by: Ilias Stamatis <ilstam@amazon.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Link: https://lore.kernel.org/r/20240718193543.624039-2-ilstam@amazon.com
[sean: rework changelog to clarify behavior, call out weirdness of buggy commit]
Signed-off-by: Sean Christopherson <seanjc@google.com>
Disallow read-only memslots for SEV-{ES,SNP} VM types, as KVM can't
directly emulate instructions for ES/SNP, and instead the guest must
explicitly request emulation. Unless the guest explicitly requests
emulation without accessing memory, ES/SNP relies on KVM creating an MMIO
SPTE, with the subsequent #NPF being reflected into the guest as a #VC.
But for read-only memslots, KVM deliberately doesn't create MMIO SPTEs,
because except for ES/SNP, doing so requires setting reserved bits in the
SPTE, i.e. the SPTE can't be readable while also generating a #VC on
writes. Because KVM never creates MMIO SPTEs and jumps directly to
emulation, the guest never gets a #VC. And since KVM simply resumes the
guest if ES/SNP guests trigger emulation, KVM effectively puts the vCPU
into an infinite #NPF loop if the vCPU attempts to write read-only memory.
Disallow read-only memory for all VMs with protected state, i.e. for
upcoming TDX VMs as well as ES/SNP VMs. For TDX, it's actually possible
to support read-only memory, as TDX uses EPT Violation #VE to reflect the
fault into the guest, e.g. KVM could configure read-only SPTEs with RX
protections and SUPPRESS_VE=0. But there is no strong use case for
supporting read-only memslots on TDX, e.g. the main historical usage is
to emulate option ROMs, but TDX disallows executing from shared memory.
And if someone comes along with a legitimate, strong use case, the
restriction can always be lifted for TDX.
Don't bother trying to retroactively apply the restriction to SEV-ES
VMs that are created as type KVM_X86_DEFAULT_VM. Read-only memslots can't
possibly work for SEV-ES, i.e. disallowing such memslots is really just
means reporting an error to userspace instead of silently hanging vCPUs.
Trying to deal with the ordering between KVM_SEV_INIT and memslot creation
isn't worth the marginal benefit it would provide userspace.
Fixes: 26c44aa9e0 ("KVM: SEV: define VM types for SEV and SEV-ES")
Fixes: 1dfe571c12 ("KVM: SEV: Add initial SEV-SNP support")
Cc: Peter Gonda <pgonda@google.com>
Cc: Michael Roth <michael.roth@amd.com>
Cc: Vishal Annapurve <vannapurve@google.com>
Cc: Ackerly Tng <ackerleytng@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20240809190319.1710470-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When hot-unplug a device which has many queues, and guest CPU will has
huge jitter, and unplugging is very slow.
It turns out synchronize_srcu() in irqfd_shutdown() caused the guest
jitter and unplugging latency, so replace synchronize_srcu() with
synchronize_srcu_expedited(), to accelerate the unplugging, and reduce
the guest OS jitter, this accelerates the VM reboot too.
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Message-ID: <20240711121130.38917-1-lirongqing@baidu.com>
[Call it just once in irqfd_resampler_shutdown. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
For any changes of struct fd representation we need to
turn existing accesses to fields into calls of wrappers.
Accesses to struct fd::flags are very few (3 in linux/file.h,
1 in net/socket.c, 3 in fs/overlayfs/file.c and 3 more in
explicit initializers).
Those can be dealt with in the commit converting to
new layout; accesses to struct fd::file are too many for that.
This commit converts (almost) all of f.file to
fd_file(f). It's not entirely mechanical ('file' is used as
a member name more than just in struct fd) and it does not
even attempt to distinguish the uses in pointer context from
those in boolean context; the latter will be eventually turned
into a separate helper (fd_empty()).
NOTE: mass conversion to fd_empty(), tempting as it
might be, is a bad idea; better do that piecewise in commit
that convert from fdget...() to CLASS(...).
[conflicts in fs/fhandle.c, kernel/bpf/syscall.c, mm/memcontrol.c
caught by git; fs/stat.c one got caught by git grep]
[fs/xattr.c conflict]
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Right now, large folios are not supported in guest_memfd, and therefore the order
used by kvm_gmem_populate() is always 0. In this scenario, using the up-to-date
bit to track prepared-ness is nice and easy because we have one bit available
per page.
In the future, however, we might have large pages that are partially populated;
for example, in the case of SEV-SNP, if a large page has both shared and private
areas inside, it is necessary to populate it at a granularity that is smaller
than that of the guest_memfd's backing store. In that case we will have
to track preparedness at a 4K level, probably as a bitmap.
In preparation for that, do not use explicitly folio_test_uptodate() and
folio_mark_uptodate(). Return the state of the page directly from
__kvm_gmem_get_pfn(), so that it is expected to apply to 2^N pages
with N=*max_order. The function to mark a range as prepared for now
takes just a folio, but is expected to take also an index and order
(or something like that) when large pages are introduced.
Thanks to Michael Roth for pointing out the issue with large pages.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This check is currently performed by sev_gmem_post_populate(), but it
applies to all callers of kvm_gmem_populate(): the point of the function
is that the memory is being encrypted and some work has to be done
on all the gfns in order to encrypt them.
Therefore, check the KVM_MEMORY_ATTRIBUTE_PRIVATE attribute prior
to invoking the callback, and stop the operation if a shared page
is encountered. Because CONFIG_KVM_PRIVATE_MEM in principle does
not require attributes, this makes kvm_gmem_populate() depend on
CONFIG_KVM_GENERIC_PRIVATE_MEM (which does require them).
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
While currently there is no other attribute than KVM_MEMORY_ATTRIBUTE_PRIVATE,
KVM code such as kvm_mem_is_private() is written to expect their existence.
Allow using kvm_range_has_memory_attributes() as a multi-page version of
kvm_mem_is_private(), without it breaking later when more attributes are
introduced.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Use a guard to simplify early returns, and add two more easy
shortcuts. If the requested attributes are invalid, the attributes
xarray will never show them as set. And if testing a single page,
kvm_get_memory_attributes() is more efficient.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Do not allow populating the same page twice with startup data. In the
case of SEV-SNP, for example, the firmware does not allow it anyway,
since the launch-update operation is only possible on pages that are
still shared in the RMP.
Even if it worked, kvm_gmem_populate()'s callback is meant to have side
effects such as updating launch measurements, and updating the same
page twice is unlikely to have the desired results.
Races between calls to the ioctl are not possible because
kvm_gmem_populate() holds slots_lock and the VM should not be running.
But again, even if this worked on other confidential computing technology,
it doesn't matter to guest_memfd.c whether this is something fishy
such as missing synchronization in userspace, or rather something
intentional. One of the racers wins, and the page is initialized by
either kvm_gmem_prepare_folio() or kvm_gmem_populate().
Anyway, out of paranoia, adjust sev_gmem_post_populate() anyway to use
the same errno that kvm_gmem_populate() is using.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It is enough to return 0 if a guest need not do any preparation.
This is in fact how sev_gmem_prepare() works for non-SNP guests,
and it extends naturally to Intel hosts: the x86 callback for
gmem_prepare is optional and returns 0 if not defined.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This is now possible because preparation is done by kvm_gmem_get_pfn()
instead of fallocate(). In practice this is not a limitation, because
even though guest_memfd can be bound to multiple struct kvm, for
hardware implementations of confidential computing only one guest
(identified by an ASID on SEV-SNP, or an HKID on TDX) will be able
to access it.
In the case of intra-host migration (not implemented yet for SEV-SNP,
but we can use SEV-ES as an idea of how it will work), the new struct
kvm inherits the same ASID and preparation need not be repeated.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Initializing the contents of the folio on fallocate() is unnecessarily
restrictive. It means that the page is registered with the firmware and
then it cannot be touched anymore. In particular, this loses the
possibility of using fallocate() to pre-allocate the page for SEV-SNP
guests, because kvm_arch_gmem_prepare() then fails.
It's only when the guest actually accesses the page (and therefore
kvm_gmem_get_pfn() is called) that the page must be cleared from any
stale host data and registered with the firmware. The up-to-date flag
is clear if this has to be done (i.e. it is the first access and
kvm_gmem_populate() has not been called).
All in all, there are enough differences between kvm_gmem_get_pfn() and
kvm_gmem_populate(), that it's better to separate the two flows completely.
Extract the bulk of kvm_gmem_get_folio(), which take a folio and end up
setting its up-to-date flag, to a new function kvm_gmem_prepare_folio();
these are now done only by the non-__-prefixed kvm_gmem_get_pfn().
As a bonus, __kvm_gmem_get_pfn() loses its ugly "bool prepare" argument.
One difference is that fallocate(PUNCH_HOLE) can now race with a
page fault. Potentially this causes a page to be prepared and into the
filemap even after fallocate(PUNCH_HOLE). This is harmless, as it can be
fixed by another hole punching operation, and can be avoided by clearing
the private-page attribute prior to invoking fallocate(PUNCH_HOLE).
This way, the page fault will cause an exit to user space.
The previous semantics, where fallocate() could be used to prepare
the pages in advance of running the guest, can be accessed with
KVM_PRE_FAULT_MEMORY.
For now, accessing a page in one VM will attempt to call
kvm_arch_gmem_prepare() in all of those that have bound the guest_memfd.
Cleaning this up is left to a separate patch.
Suggested-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Allow testing the up-to-date flag in the caller without taking the
lock again.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add "ARCH" to the symbols; shortly, the "prepare" phase will include both
the arch-independent step to clear out contents left in the page by the
host, and the arch-dependent step enabled by CONFIG_HAVE_KVM_GMEM_PREPARE.
For consistency do the same for CONFIG_HAVE_KVM_GMEM_INVALIDATE as well.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We have a perfectly usable folio, use it to retrieve the pfn and order.
All that's needed is a version of folio_file_page that returns a pfn.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The up-to-date flag as is now is not too useful; it tells guest_memfd not
to overwrite the contents of a folio, but it doesn't say that the page
is ready to be mapped into the guest. For encrypted guests, mapping
a private page requires that the "preparation" phase has succeeded,
and at the same time the same page cannot be prepared twice.
So, ensure that folio_mark_uptodate() is only called on a prepared page. If
kvm_gmem_prepare_folio() or the post_populate callback fail, the folio
will not be marked up-to-date; it's not a problem to call clear_highpage()
again on such a page prior to the next preparation attempt.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Right now this is simply more consistent and avoids use of pfn_to_page()
and put_page(). It will be put to more use in upcoming patches, to
ensure that the up-to-date flag is set at the very end of both the
kvm_gmem_get_pfn() and kvm_gmem_populate() flows.
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM generic changes for 6.11
- Enable halt poll shrinking by default, as Intel found it to be a clear win.
- Setup empty IRQ routing when creating a VM to avoid having to synchronize
SRCU when creating a split IRQCHIP on x86.
- Rework the sched_in/out() paths to replace kvm_arch_sched_in() with a flag
that arch code can use for hooking both sched_in() and sched_out().
- Take the vCPU @id as an "unsigned long" instead of "u32" to avoid
truncating a bogus value from userspace, e.g. to help userspace detect bugs.
- Mark a vCPU as preempted if and only if it's scheduled out while in the
KVM_RUN loop, e.g. to avoid marking it preempted and thus writing guest
memory when retrieving guest state during live migration blackout.
- A few minor cleanups
KVM Xen:
Fix a bug where KVM fails to check the validity of an incoming userspace
virtual address and tries to activate a gfn_to_pfn_cache with a kernel address.
Pre-population has been requested several times to mitigate KVM page faults
during guest boot or after live migration. It is also required by TDX
before filling in the initial guest memory with measured contents.
Introduce it as a generic API.
The flags AS_UNMOVABLE and AS_INACCESSIBLE were both added just for guest_memfd;
AS_UNMOVABLE is already in existing versions of Linux, while AS_INACCESSIBLE was
acked for inclusion in 6.11.
But really, they are the same thing: only guest_memfd uses them, at least for
now, and guest_memfd pages are unmovable because they should not be
accessed by the CPU.
So merge them into one; use the AS_INACCESSIBLE name which is more comprehensive.
At the same time, this fixes an embarrassing bug where AS_INACCESSIBLE was used
as a bit mask, despite it being just a bit index.
The bug was mostly benign, because AS_INACCESSIBLE's bit representation (1010)
corresponded to setting AS_UNEVICTABLE (which is already set) and AS_ENOSPC
(except no async writes can happen on the guest_memfd). So the AS_INACCESSIBLE
flag simply had no effect.
Fixes: 1d23040caa ("KVM: guest_memfd: Use AS_INACCESSIBLE when creating guest_memfd inode")
Fixes: c72ceafbd1 ("mm: Introduce AS_INACCESSIBLE for encrypted/confidential memory")
Cc: linux-mm@kvack.org
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: David Hildenbrand <david@redhat.com>
Tested-by: Michael Roth <michael.roth@amd.com>
Reviewed-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Sean Christopherson
Yan Zhao
Liam Merwick
Isaku Yamahata
Paolo Bonzini
Linus Torvalds
Dr. David Alan Gilbert
Peter Zijlstra
Al Viro
Peter Xu
Ilias Stamatis
Li RongQing