KVM: x86: Use __try_cmpxchg_user() to update guest PTE A/D bits
Use the recently introduced __try_cmpxchg_user() to update guest PTE A/D
bits instead of mapping the PTE into kernel address space. The VM_PFNMAP
path is broken as it assumes that vm_pgoff is the base pfn of the mapped
VMA range, which is conceptually wrong as vm_pgoff is the offset relative
to the file and has nothing to do with the pfn. The horrific hack worked
for the original use case (backing guest memory with /dev/mem), but leads
to accessing "random" pfns for pretty much any other VM_PFNMAP case.
Fixes: b89f9ae1559d ("X86/KVM: Handle PFNs outside of kernel reach when touching GPTEs") Debugged-by: Tadeusz Struk <tadeusz.struk@linaro.org> Tested-by: Tadeusz Struk <tadeusz.struk@linaro.org> Reported-by: syzbot+6cde2282daa792c49ab8@syzkaller.appspotmail.com Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220202004945.2540433-4-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Peter Zijlstra [Wed, 2 Feb 2022 00:49:42 +0000 (00:49 +0000)]
x86/uaccess: Implement macros for CMPXCHG on user addresses
Add support for CMPXCHG loops on userspace addresses. Provide both an
"unsafe" version for tight loops that do their own uaccess begin/end, as
well as a "safe" version for use cases where the CMPXCHG is not buried in
a loop, e.g. KVM will resume the guest instead of looping when emulation
of a guest atomic accesses fails the CMPXCHG.
Provide 8-byte versions for 32-bit kernels so that KVM can do CMPXCHG on
guest PAE PTEs, which are accessed via userspace addresses.
Guard the asm_volatile_goto() variation with CC_HAS_ASM_GOTO_TIED_OUTPUT,
the "+m" constraint fails on some compilers that otherwise support
CC_HAS_ASM_GOTO_OUTPUT.
Cc: stable@vger.kernel.org Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Co-developed-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220202004945.2540433-3-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Kconfig: Add option for asm goto w/ tied outputs to workaround clang-13 bug
Add a config option to guard (future) usage of asm_volatile_goto() that
includes "tied outputs", i.e. "+" constraints that specify both an input
and output parameter. clang-13 has a bug[1] that causes compilation of
such inline asm to fail, and KVM wants to use a "+m" constraint to
implement a uaccess form of CMPXCHG[2]. E.g. the test code fails with
<stdin>:1:29: error: invalid operand in inline asm: '.long (${1:l}) - .'
int foo(int *x) { asm goto (".long (%l[bar]) - .\n": "+m"(*x) ::: bar); return *x; bar: return 0; }
^
<stdin>:1:29: error: unknown token in expression
<inline asm>:1:9: note: instantiated into assembly here
.long () - .
^
2 errors generated.
on clang-13, but passes on gcc (with appropriate asm goto support). The
bug is fixed in clang-14, but won't be backported to clang-13 as the
changes are too invasive/risky.
gcc also had a similar bug[3], fixed in gcc-11, where gcc failed to
account for its behavior of assigning two numbers to tied outputs (one
for input, one for output) when evaluating symbolic references.
Suggested-by: Nick Desaulniers <ndesaulniers@google.com> Reviewed-by: Nick Desaulniers <ndesaulniers@google.com> Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220202004945.2540433-2-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Peter Gonda [Thu, 7 Apr 2022 21:02:33 +0000 (14:02 -0700)]
KVM, SEV: Add KVM_EXIT_SHUTDOWN metadata for SEV-ES
If an SEV-ES guest requests termination, exit to userspace with
KVM_EXIT_SYSTEM_EVENT and a dedicated SEV_TERM type instead of -EINVAL
so that userspace can take appropriate action.
See AMD's GHCB spec section '4.1.13 Termination Request' for more details.
Suggested-by: Sean Christopherson <seanjc@google.com> Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Cc: kvm@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Peter Gonda <pgonda@google.com> Reported-by: kernel test robot <lkp@intel.com>
Message-Id: <20220407210233.782250-1-pgonda@google.com>
[Add documentatino. - Paolo] Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: nVMX: Clear IDT vectoring on nested VM-Exit for double/triple fault
Clear the IDT vectoring field in vmcs12 on next VM-Exit due to a double
or triple fault. Per the SDM, a VM-Exit isn't considered to occur during
event delivery if the exit is due to an intercepted double fault or a
triple fault. Opportunistically move the default clearing (no event
"pending") into the helper so that it's more obvious that KVM does indeed
handle this case.
Note, the double fault case is worded rather wierdly in the SDM:
The original event results in a double-fault exception that causes the
VM exit directly.
Temporarily ignoring injected events, double faults can _only_ occur if
an exception occurs while attempting to deliver a different exception,
i.e. there's _always_ an original event. And for injected double fault,
while there's no original event, injected events are never subject to
interception.
Presumably the SDM is calling out that a the vectoring info will be valid
if a different exit occurs after a double fault, e.g. if a #PF occurs and
is intercepted while vectoring #DF, then the vectoring info will show the
double fault. In other words, the clause can simply be read as:
The VM exit is caused by a double-fault exception.
Fixes: 4db0cb0272e4 ("KVM: nVMX: Exiting from L2 to L1") Cc: Chenyi Qiang <chenyi.qiang@intel.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220407002315.78092-4-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: nVMX: Leave most VM-Exit info fields unmodified on failed VM-Entry
Don't modify vmcs12 exit fields except EXIT_REASON and EXIT_QUALIFICATION
when performing a nested VM-Exit due to failed VM-Entry. Per the SDM,
only the two aformentioned fields are filled and "All other VM-exit
information fields are unmodified".
Fixes: 4db0cb0272e4 ("KVM: nVMX: Exiting from L2 to L1") Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220407002315.78092-3-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: x86: Drop WARNs that assert a triple fault never "escapes" from L2
Remove WARNs that sanity check that KVM never lets a triple fault for L2
escape and incorrectly end up in L1. In normal operation, the sanity
check is perfectly valid, but it incorrectly assumes that it's impossible
for userspace to induce KVM_REQ_TRIPLE_FAULT without bouncing through
KVM_RUN (which guarantees kvm_check_nested_state() will see and handle
the triple fault).
The WARN can currently be triggered if userspace injects a machine check
while L2 is active and CR4.MCE=0. And a future fix to allow save/restore
of KVM_REQ_TRIPLE_FAULT, e.g. so that a synthesized triple fault isn't
lost on migration, will make it trivially easy for userspace to trigger
the WARN.
Clearing KVM_REQ_TRIPLE_FAULT when forcibly leaving guest mode is
tempting, but wrong, especially if/when the request is saved/restored,
e.g. if userspace restores events (including a triple fault) and then
restores nested state (which may forcibly leave guest mode). Ignoring
the fact that KVM doesn't currently provide the necessary APIs, it's
userspace's responsibility to manage pending events during save/restore.
Like Xu [Tue, 29 Mar 2022 23:50:54 +0000 (23:50 +0000)]
KVM: x86: Use static calls to reduce kvm_pmu_ops overhead
Use static calls to improve kvm_pmu_ops performance, following the same
pattern and naming scheme used by kvm-x86-ops.h.
Here are the worst fenced_rdtsc() cycles numbers for the kvm_pmu_ops
functions that is most often called (up to 7 digits of calls) when running
a single perf test case in a guest on an ICX 2.70GHz host (mitigations=on):
Like Xu [Tue, 29 Mar 2022 23:50:53 +0000 (23:50 +0000)]
KVM: x86: Move .pmu_ops to kvm_x86_init_ops and tag as __initdata
The pmu_ops should be moved to kvm_x86_init_ops and tagged as __initdata.
That'll save those precious few bytes, and more importantly make
the original ops unreachable, i.e. make it harder to sneak in post-init
modification bugs.
Suggested-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Like Xu <likexu@tencent.com> Reviewed-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220329235054.3534728-4-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Like Xu [Tue, 29 Mar 2022 23:50:52 +0000 (23:50 +0000)]
KVM: x86: Copy kvm_pmu_ops by value to eliminate layer of indirection
Replace the kvm_pmu_ops pointer in common x86 with an instance of the
struct to save one pointer dereference when invoking functions. Copy the
struct by value to set the ops during kvm_init().
Signed-off-by: Like Xu <likexu@tencent.com>
[sean: Move pmc_is_enabled(), make kvm_pmu_ops static] Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220329235054.3534728-3-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Like Xu [Tue, 29 Mar 2022 23:50:51 +0000 (23:50 +0000)]
KVM: x86: Move kvm_ops_static_call_update() to x86.c
The kvm_ops_static_call_update() is defined in kvm_host.h. That's
completely unnecessary, it should have exactly one caller,
kvm_arch_hardware_setup(). Move the helper to x86.c and have it do the
actual memcpy() of the ops in addition to the static call updates. This
will also allow for cleanly giving kvm_pmu_ops static_call treatment.
Suggested-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Like Xu <likexu@tencent.com>
[sean: Move memcpy() into the helper and rename accordingly] Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220329235054.3534728-2-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Derive the mask of RWX bits reported on EPT violations from the mask of
RWX bits that are shoved into EPT entries; the layout is the same, the
EPT violation bits are simply shifted by three. Use the new shift and a
slight copy-paste of the mask derivation instead of completely open
coding the same to convert between the EPT entry bits and the exit
qualification when synthesizing a nested EPT Violation.
No functional change intended.
Cc: SU Hang <darcy.sh@antgroup.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220329030108.97341-3-darcy.sh@antgroup.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
SU Hang [Tue, 29 Mar 2022 03:01:06 +0000 (11:01 +0800)]
KVM: VMX: replace 0x180 with EPT_VIOLATION_* definition
Using self-expressing macro definition EPT_VIOLATION_GVA_VALIDATION
and EPT_VIOLATION_GVA_TRANSLATED instead of 0x180
in FNAME(walk_addr_generic)().
Signed-off-by: SU Hang <darcy.sh@antgroup.com> Reviewed-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220329030108.97341-2-darcy.sh@antgroup.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Wanpeng Li [Tue, 8 Mar 2022 08:18:49 +0000 (00:18 -0800)]
x86/kvm: Don't waste kvmclock memory if there is nopv parameter
When the "nopv" command line parameter is used, it should not waste
memory for kvmclock.
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
Message-Id: <1646727529-11774-1-git-send-email-wanpengli@tencent.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Like Xu [Wed, 6 Apr 2022 06:37:14 +0000 (14:37 +0800)]
selftests: kvm/x86/xen: Replace a comma in the xen_shinfo_test with semicolon
+WARNING: Possible comma where semicolon could be used
+#397: FILE: tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c:700:
++ tmr.type = KVM_XEN_VCPU_ATTR_TYPE_TIMER,
++ vcpu_ioctl(vm, VCPU_ID, KVM_XEN_VCPU_GET_ATTR, &tmr);
Fixes: a36bd44d7147 ("KVM: x86/xen: Add self tests for KVM_XEN_HVM_CONFIG_EVTCHN_SEND") Signed-off-by: Like Xu <likexu@tencent.com>
Message-Id: <20220406063715.55625-4-likexu@tencent.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: x86: hyper-v: Avoid writing to TSC page without an active vCPU
The following WARN is triggered from kvm_vm_ioctl_set_clock():
WARNING: CPU: 10 PID: 579353 at arch/x86/kvm/../../../virt/kvm/kvm_main.c:3161 mark_page_dirty_in_slot+0x6c/0x80 [kvm]
...
CPU: 10 PID: 579353 Comm: qemu-system-x86 Tainted: G W O 5.16.0.stable #20
Hardware name: LENOVO 20UF001CUS/20UF001CUS, BIOS R1CET65W(1.34 ) 06/17/2021
RIP: 0010:mark_page_dirty_in_slot+0x6c/0x80 [kvm]
...
Call Trace:
<TASK>
? kvm_write_guest+0x114/0x120 [kvm]
kvm_hv_invalidate_tsc_page+0x9e/0xf0 [kvm]
kvm_arch_vm_ioctl+0xa26/0xc50 [kvm]
? schedule+0x4e/0xc0
? __cond_resched+0x1a/0x50
? futex_wait+0x166/0x250
? __send_signal+0x1f1/0x3d0
kvm_vm_ioctl+0x747/0xda0 [kvm]
...
The WARN was introduced by commit 03c0304a86bc ("KVM: Warn if
mark_page_dirty() is called without an active vCPU") but the change seems
to be correct (unlike Hyper-V TSC page update mechanism). In fact, there's
no real need to actually write to guest memory to invalidate TSC page, this
can be done by the first vCPU which goes through kvm_guest_time_update().
vcpu_fp uses the riscv_isa_extension mechanism which gets
defined in hwcap.h but doesn't include that head file.
While it seems to work in most cases, in certain conditions
this can lead to build failures like
../arch/riscv/kvm/vcpu_fp.c: In function ‘kvm_riscv_vcpu_fp_reset’:
../arch/riscv/kvm/vcpu_fp.c:22:13: error: implicit declaration of function ‘riscv_isa_extension_available’ [-Werror=implicit-function-declaration]
22 | if (riscv_isa_extension_available(&isa, f) ||
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../arch/riscv/kvm/vcpu_fp.c:22:49: error: ‘f’ undeclared (first use in this function)
22 | if (riscv_isa_extension_available(&isa, f) ||
Fix this by simply including the necessary header.
Fixes: b561057c8388 ("RISC-V: KVM: Factor-out FP virtualization into separate
sources") Signed-off-by: Heiko Stuebner <heiko@sntech.de> Signed-off-by: Anup Patel <anup@brainfault.org>
KVM: selftests: riscv: Fix alignment of the guest_hang() function
The guest_hang() function is used as the default exception handler
for various KVM selftests applications by setting it's address in
the vstvec CSR. The vstvec CSR requires exception handler base address
to be at least 4-byte aligned so this patch fixes alignment of the
guest_hang() function.
KVM: selftests: riscv: Set PTE A and D bits in VS-stage page table
Supporting hardware updates of PTE A and D bits is optional for any
RISC-V implementation so current software strategy is to always set
these bits in both G-stage (hypervisor) and VS-stage (guest kernel).
If PTE A and D bits are not set by software (hypervisor or guest)
then RISC-V implementations not supporting hardware updates of these
bits will cause traps even for perfectly valid PTEs.
Based on above explanation, the VS-stage page table created by various
KVM selftest applications is not correct because PTE A and D bits are
not set. This patch fixes VS-stage page table programming of PTE A and
D bits for KVM selftests.
RISC-V: KVM: Don't clear hgatp CSR in kvm_arch_vcpu_put()
We might have RISC-V systems (such as QEMU) where VMID is not part
of the TLB entry tag so these systems will have to flush all TLB
entries upon any change in hgatp.VMID.
Currently, we zero-out hgatp CSR in kvm_arch_vcpu_put() and we
re-program hgatp CSR in kvm_arch_vcpu_load(). For above described
systems, this will flush all TLB entries whenever VCPU exits to
user-space hence reducing performance.
This patch fixes above described performance issue by not clearing
hgatp CSR in kvm_arch_vcpu_put().
Oliver Upton [Wed, 6 Apr 2022 23:56:15 +0000 (23:56 +0000)]
selftests: KVM: Free the GIC FD when cleaning up in arch_timer
In order to correctly destroy a VM, all references to the VM must be
freed. The arch_timer selftest creates a VGIC for the guest, which
itself holds a reference to the VM.
Oliver Upton [Wed, 6 Apr 2022 23:56:14 +0000 (23:56 +0000)]
selftests: KVM: Don't leak GIC FD across dirty log test iterations
dirty_log_perf_test instantiates a VGICv3 for the guest (if supported by
hardware) to reduce the overhead of guest exits. However, the test does
not actually close the GIC fd when cleaning up the VM between test
iterations, meaning that the VM is never actually destroyed in the
kernel.
While this is generally a bad idea, the bug was detected from the kernel
spewing about duplicate debugfs entries as subsequent VMs happen to
reuse the same FD even though the debugfs directory is still present.
Abstract away the notion of setup/cleanup of the GIC FD from the test
by creating arch-specific helpers for test setup/cleanup. Close the GIC
FD on VM cleanup and do nothing for the other architectures.
Fixes: d34d1c6265a9 ("KVM: selftests: Add vgic initialization for dirty log perf test for ARM") Reviewed-by: Jing Zhang <jingzhangos@google.com> Signed-off-by: Oliver Upton <oupton@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20220406235615.1447180-3-oupton@google.com
Oliver Upton [Wed, 6 Apr 2022 23:56:13 +0000 (23:56 +0000)]
KVM: Don't create VM debugfs files outside of the VM directory
Unfortunately, there is no guarantee that KVM was able to instantiate a
debugfs directory for a particular VM. To that end, KVM shouldn't even
attempt to create new debugfs files in this case. If the specified
parent dentry is NULL, debugfs_create_file() will instantiate files at
the root of debugfs.
For arm64, it is possible to create the vgic-state file outside of a
VM directory, the file is not cleaned up when a VM is destroyed.
Nonetheless, the corresponding struct kvm is freed when the VM is
destroyed.
Nip the problem in the bud for all possible errant debugfs file
creations by initializing kvm->debugfs_dentry to -ENOENT. In so doing,
debugfs_create_file() will fail instead of creating the file in the root
directory.
Cc: stable@kernel.org Fixes: 9a9e2947c36b ("kvm: no need to check return value of debugfs_create functions") Signed-off-by: Oliver Upton <oupton@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20220406235615.1447180-2-oupton@google.com
When testing a kernel with commit fb69b29fcaf0 ("KVM: arm64:
Allow SMCCC_ARCH_WORKAROUND_3 to be discovered and migrated")
get-reg-list output
vregs: Number blessed registers: 234
vregs: Number registers: 238
vregs: There are 1 new registers.
Consider adding them to the blessed reg list with the following lines:
KVM_REG_ARM_FW_REG(3),
vregs: PASS
...
That output inspired two changes: 1) add the new register to the
blessed list and 2) explain why "Number registers" is actually four
larger than "Number blessed registers" (on the system used for
testing), even though only one register is being stated as new.
The reason is that some registers are host dependent and they get
filtered out when comparing with the blessed list. The system
used for the test apparently had three filtered registers.
Paolo Bonzini [Wed, 6 Apr 2022 17:13:42 +0000 (13:13 -0400)]
KVM: avoid NULL pointer dereference in kvm_dirty_ring_push
kvm_vcpu_release() will call kvm_dirty_ring_free(), freeing
ring->dirty_gfns and setting it to NULL. Afterwards, it calls
kvm_arch_vcpu_destroy().
However, if closing the file descriptor races with KVM_RUN in such away
that vcpu->arch.st.preempted == 0, the following call stack leads to a
NULL pointer dereference in kvm_dirty_run_push():
Introduce a test for aarch64 that ensures non-mixed-width vCPUs
(all 64bit vCPUs or all 32bit vcPUs) can be configured, and
mixed-width vCPUs cannot be configured.
Reiji Watanabe [Tue, 29 Mar 2022 03:19:23 +0000 (20:19 -0700)]
KVM: arm64: mixed-width check should be skipped for uninitialized vCPUs
KVM allows userspace to configure either all EL1 32bit or 64bit vCPUs
for a guest. At vCPU reset, vcpu_allowed_register_width() checks
if the vcpu's register width is consistent with all other vCPUs'.
Since the checking is done even against vCPUs that are not initialized
(KVM_ARM_VCPU_INIT has not been done) yet, the uninitialized vCPUs
are erroneously treated as 64bit vCPU, which causes the function to
incorrectly detect a mixed-width VM.
Introduce KVM_ARCH_FLAG_EL1_32BIT and KVM_ARCH_FLAG_REG_WIDTH_CONFIGURED
bits for kvm->arch.flags. A value of the EL1_32BIT bit indicates that
the guest needs to be configured with all 32bit or 64bit vCPUs, and
a value of the REG_WIDTH_CONFIGURED bit indicates if a value of the
EL1_32BIT bit is valid (already set up). Values in those bits are set at
the first KVM_ARM_VCPU_INIT for the guest based on KVM_ARM_VCPU_EL1_32BIT
configuration for the vCPU.
Check vcpu's register width against those new bits at the vcpu's
KVM_ARM_VCPU_INIT (instead of against other vCPUs' register width).
Fixes: 59ea50f5602f ("KVM: arm64: Prevent mixed-width VM creation") Signed-off-by: Reiji Watanabe <reijiw@google.com> Reviewed-by: Oliver Upton <oupton@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20220329031924.619453-2-reijiw@google.com
Oliver Upton [Fri, 1 Apr 2022 19:46:52 +0000 (19:46 +0000)]
KVM: arm64: Don't split hugepages outside of MMU write lock
It is possible to take a stage-2 permission fault on a page larger than
PAGE_SIZE. For example, when running a guest backed by 2M HugeTLB, KVM
eagerly maps at the largest possible block size. When dirty logging is
enabled on a memslot, KVM does *not* eagerly split these 2M stage-2
mappings and instead clears the write bit on the pte.
Since dirty logging is always performed at PAGE_SIZE granularity, KVM
lazily splits these 2M block mappings down to PAGE_SIZE in the stage-2
fault handler. This operation must be done under the write lock. Since
commit 20f96d750fea ("KVM: arm64: Add fast path to handle permission
relaxation during dirty logging"), the stage-2 fault handler
conditionally takes the read lock on permission faults with dirty
logging enabled. To that end, it is possible to split a 2M block mapping
while only holding the read lock.
The problem is demonstrated by running kvm_page_table_test with 2M
anonymous HugeTLB, which splats like so:
WARNING: CPU: 5 PID: 15276 at arch/arm64/kvm/hyp/pgtable.c:153 stage2_map_walk_leaf+0x124/0x158
Fix the issue by only acquiring the read lock if the guest faulted on a
PAGE_SIZE granule w/ dirty logging enabled. Add a WARN to catch locking
bugs in future changes.
Fixes: 20f96d750fea ("KVM: arm64: Add fast path to handle permission relaxation during dirty logging") Cc: Jing Zhang <jingzhangos@google.com> Signed-off-by: Oliver Upton <oupton@google.com> Reviewed-by: Reiji Watanabe <reijiw@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20220401194652.950240-1-oupton@google.com
Oliver Upton [Tue, 22 Mar 2022 18:35:38 +0000 (18:35 +0000)]
KVM: arm64: Drop unneeded minor version check from PSCI v1.x handler
We already sanitize the guest's PSCI version when it is being written by
userspace, rejecting unsupported version numbers. Additionally, the
'minor' parameter to kvm_psci_1_x_call() is a constant known at compile
time for all callsites.
Though it is benign, the additional check against the
PSCI kvm_psci_1_x_call() is unnecessary and likely to be missed the next
time KVM raises its maximum PSCI version. Drop the check altogether and
rely on sanitization when the PSCI version is set by userspace.
Oliver Upton [Tue, 22 Mar 2022 18:35:37 +0000 (18:35 +0000)]
KVM: arm64: Actually prevent SMC64 SYSTEM_RESET2 from AArch32
The SMCCC does not allow the SMC64 calling convention to be used from
AArch32. While KVM checks to see if the calling convention is allowed in
PSCI_1_0_FN_PSCI_FEATURES, it does not actually prevent calls to
unadvertised PSCI v1.0+ functions.
Hoist the check to see if the requested function is allowed into
kvm_psci_call(), thereby preventing SMC64 calls from AArch32 for all
PSCI versions.
Fixes: ed47efd863d9 ("KVM: arm64: Expose PSCI SYSTEM_RESET2 call to the guest") Acked-by: Will Deacon <will@kernel.org> Reviewed-by: Reiji Watanabe <reijiw@google.com> Signed-off-by: Oliver Upton <oupton@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20220322183538.2757758-3-oupton@google.com
Oliver Upton [Tue, 22 Mar 2022 18:35:36 +0000 (18:35 +0000)]
KVM: arm64: Generally disallow SMC64 for AArch32 guests
The only valid calling SMC calling convention from an AArch32 state is
SMC32. Disallow any PSCI function that sets the SMC64 function ID bit
when called from AArch32 rather than comparing against known SMC64 PSCI
functions.
Note that without this change KVM advertises the SMC64 flavor of
SYSTEM_RESET2 to AArch32 guests.
Fixes: ed47efd863d9 ("KVM: arm64: Expose PSCI SYSTEM_RESET2 call to the guest") Acked-by: Will Deacon <will@kernel.org> Reviewed-by: Reiji Watanabe <reijiw@google.com> Reviewed-by: Andrew Jones <drjones@redhat.com> Signed-off-by: Oliver Upton <oupton@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20220322183538.2757758-2-oupton@google.com
KVM: x86/mmu: Resolve nx_huge_pages when kvm.ko is loaded
Resolve nx_huge_pages to true/false when kvm.ko is loaded, leaving it as
-1 is technically undefined behavior when its value is read out by
param_get_bool(), as boolean values are supposed to be '0' or '1'.
Alternatively, KVM could define a custom getter for the param, but the
auto value doesn't depend on the vendor module in any way, and printing
"auto" would be unnecessarily unfriendly to the user.
In addition to fixing the undefined behavior, resolving the auto value
also fixes the scenario where the auto value resolves to N and no vendor
module is loaded. Previously, -1 would result in Y being printed even
though KVM would ultimately disable the mitigation.
Rename the existing MMU module init/exit helpers to clarify that they're
invoked with respect to the vendor module, and add comments to document
why KVM has two separate "module init" flows.
=========================================================================
UBSAN: invalid-load in kernel/params.c:320:33
load of value 255 is not a valid value for type '_Bool'
CPU: 6 PID: 892 Comm: tail Not tainted 5.17.0-rc3+ #799
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
Call Trace:
<TASK>
dump_stack_lvl+0x34/0x44
ubsan_epilogue+0x5/0x40
__ubsan_handle_load_invalid_value.cold+0x43/0x48
param_get_bool.cold+0xf/0x14
param_attr_show+0x55/0x80
module_attr_show+0x1c/0x30
sysfs_kf_seq_show+0x93/0xc0
seq_read_iter+0x11c/0x450
new_sync_read+0x11b/0x1a0
vfs_read+0xf0/0x190
ksys_read+0x5f/0xe0
do_syscall_64+0x3b/0xc0
entry_SYSCALL_64_after_hwframe+0x44/0xae
</TASK>
=========================================================================
Fixes: 78311cba7f22 ("kvm: mmu: ITLB_MULTIHIT mitigation") Cc: stable@vger.kernel.org Reported-by: Bruno Goncalves <bgoncalv@redhat.com> Reported-by: Jan Stancek <jstancek@redhat.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220331221359.3912754-1-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Merge tag 'trace-v5.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull more tracing updates from Steven Rostedt:
- Rename the staging files to give them some meaning. Just
stage1,stag2,etc, does not show what they are for
- Check for NULL from allocation in bootconfig
- Hold event mutex for dyn_event call in user events
- Mark user events to broken (to work on the API)
- Remove eBPF updates from user events
- Remove user events from uapi header to keep it from being installed.
- Move ftrace_graph_is_dead() into inline as it is called from hot
paths and also convert it into a static branch.
* tag 'trace-v5.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing: Move user_events.h temporarily out of include/uapi
ftrace: Make ftrace_graph_is_dead() a static branch
tracing: Set user_events to BROKEN
tracing/user_events: Remove eBPF interfaces
tracing/user_events: Hold event_mutex during dyn_event_add
proc: bootconfig: Add null pointer check
tracing: Rename the staging files for trace_events
Merge tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk fix from Stephen Boyd:
"A single revert to fix a boot regression seen when clk_put() started
dropping rate range requests. It's best to keep various systems
booting so we'll kick this out and try again next time"
* tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
Revert "clk: Drop the rate range on clk_put()"
Merge tag 'x86-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Thomas Gleixner:
"A set of x86 fixes and updates:
- Make the prctl() for enabling dynamic XSTATE components correct so
it adds the newly requested feature to the permission bitmap
instead of overwriting it. Add a selftest which validates that.
- Unroll string MMIO for encrypted SEV guests as the hypervisor
cannot emulate it.
- Handle supervisor states correctly in the FPU/XSTATE code so it
takes the feature set of the fpstate buffer into account. The
feature sets can differ between host and guest buffers. Guest
buffers do not contain supervisor states. So far this was not an
issue, but with enabling PASID it needs to be handled in the buffer
offset calculation and in the permission bitmaps.
- Avoid a gazillion of repeated CPUID invocations in by caching the
values early in the FPU/XSTATE code.
- Enable CONFIG_WERROR in x86 defconfig.
- Make the X86 defconfigs more useful by adapting them to Y2022
reality"
* tag 'x86-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/fpu/xstate: Consolidate size calculations
x86/fpu/xstate: Handle supervisor states in XSTATE permissions
x86/fpu/xsave: Handle compacted offsets correctly with supervisor states
x86/fpu: Cache xfeature flags from CPUID
x86/fpu/xsave: Initialize offset/size cache early
x86/fpu: Remove unused supervisor only offsets
x86/fpu: Remove redundant XCOMP_BV initialization
x86/sev: Unroll string mmio with CC_ATTR_GUEST_UNROLL_STRING_IO
x86/config: Make the x86 defconfigs a bit more usable
x86/defconfig: Enable WERROR
selftests/x86/amx: Update the ARCH_REQ_XCOMP_PERM test
x86/fpu/xstate: Fix the ARCH_REQ_XCOMP_PERM implementation
Merge tag 'core-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull RT signal fix from Thomas Gleixner:
"Revert the RT related signal changes. They need to be reworked and
generalized"
* tag 'core-urgent-2022-04-03' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
Revert "signal, x86: Delay calling signals in atomic on RT enabled kernels"
Merge tag 'dma-mapping-5.18-1' of git://git.infradead.org/users/hch/dma-mapping
Pull more dma-mapping updates from Christoph Hellwig:
- fix a regression in dma remap handling vs AMD memory encryption (me)
- finally kill off the legacy PCI DMA API (Christophe JAILLET)
* tag 'dma-mapping-5.18-1' of git://git.infradead.org/users/hch/dma-mapping:
dma-mapping: move pgprot_decrypted out of dma_pgprot
PCI/doc: cleanup references to the legacy PCI DMA API
PCI: Remove the deprecated "pci-dma-compat.h" API
Stephen Boyd [Sun, 3 Apr 2022 02:28:18 +0000 (19:28 -0700)]
Revert "clk: Drop the rate range on clk_put()"
This reverts commit 0bc6a8728d95e710c7f6cb09aa25958dfa6f63db. There are
multiple reports that this breaks boot on various systems. The common
theme is that orphan clks are having rates set on them when that isn't
expected. Let's revert it out for now so that -rc1 boots.
Merge tag 'perf-tools-for-v5.18-2022-04-02' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux
Pull more perf tools updates from Arnaldo Carvalho de Melo:
- Avoid SEGV if core.cpus isn't set in 'perf stat'.
- Stop depending on .git files for building PERF-VERSION-FILE, used in
'perf --version', fixing some perf tools build scenarios.
- Convert tracepoint.py example to python3.
- Update UAPI header copies from the kernel sources: socket,
mman-common, msr-index, KVM, i915 and cpufeatures.
- Update copy of libbpf's hashmap.c.
- Directly return instead of using local ret variable in
evlist__create_syswide_maps(), found by coccinelle.
* tag 'perf-tools-for-v5.18-2022-04-02' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
perf python: Convert tracepoint.py example to python3
perf evlist: Directly return instead of using local ret variable
perf cpumap: More cpu map reuse by merge.
perf cpumap: Add is_subset function
perf evlist: Rename cpus to user_requested_cpus
perf tools: Stop depending on .git files for building PERF-VERSION-FILE
tools headers cpufeatures: Sync with the kernel sources
tools headers UAPI: Sync drm/i915_drm.h with the kernel sources
tools headers UAPI: Sync linux/kvm.h with the kernel sources
tools kvm headers arm64: Update KVM headers from the kernel sources
tools arch x86: Sync the msr-index.h copy with the kernel sources
tools headers UAPI: Sync asm-generic/mman-common.h with the kernel
perf beauty: Update copy of linux/socket.h with the kernel sources
perf tools: Update copy of libbpf's hashmap.c
perf stat: Avoid SEGV if core.cpus isn't set
Merge tag 'kbuild-fixes-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- Fix empty $(PYTHON) expansion.
- Fix UML, which got broken by the attempt to suppress Clang warnings.
- Fix warning message in modpost.
* tag 'kbuild-fixes-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
modpost: restore the warning message for missing symbol versions
Revert "um: clang: Strip out -mno-global-merge from USER_CFLAGS"
kbuild: Remove '-mno-global-merge'
kbuild: fix empty ${PYTHON} in scripts/link-vmlinux.sh
kconfig: remove stale comment about removed kconfig_print_symbol()
Merge tag 'mips_5.18_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux
Pull MIPS fixes from Thomas Bogendoerfer:
- build fix for gpio
- fix crc32 build problems
- check for failed memory allocations
* tag 'mips_5.18_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
MIPS: crypto: Fix CRC32 code
MIPS: rb532: move GPIOD definition into C-files
MIPS: lantiq: check the return value of kzalloc()
mips: sgi-ip22: add a check for the return of kzalloc()
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm fixes from Paolo Bonzini:
- Only do MSR filtering for MSRs accessed by rdmsr/wrmsr
- Documentation improvements
- Prevent module exit until all VMs are freed
- PMU Virtualization fixes
- Fix for kvm_irq_delivery_to_apic_fast() NULL-pointer dereferences
- Other miscellaneous bugfixes
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (42 commits)
KVM: x86: fix sending PV IPI
KVM: x86/mmu: do compare-and-exchange of gPTE via the user address
KVM: x86: Remove redundant vm_entry_controls_clearbit() call
KVM: x86: cleanup enter_rmode()
KVM: x86: SVM: fix tsc scaling when the host doesn't support it
kvm: x86: SVM: remove unused defines
KVM: x86: SVM: move tsc ratio definitions to svm.h
KVM: x86: SVM: fix avic spec based definitions again
KVM: MIPS: remove reference to trap&emulate virtualization
KVM: x86: document limitations of MSR filtering
KVM: x86: Only do MSR filtering when access MSR by rdmsr/wrmsr
KVM: x86/emulator: Emulate RDPID only if it is enabled in guest
KVM: x86/pmu: Fix and isolate TSX-specific performance event logic
KVM: x86: mmu: trace kvm_mmu_set_spte after the new SPTE was set
KVM: x86/svm: Clear reserved bits written to PerfEvtSeln MSRs
KVM: x86: Trace all APICv inhibit changes and capture overall status
KVM: x86: Add wrappers for setting/clearing APICv inhibits
KVM: x86: Make APICv inhibit reasons an enum and cleanup naming
KVM: X86: Handle implicit supervisor access with SMAP
KVM: X86: Rename variable smap to not_smap in permission_fault()
...
Merge tag 'for-5.18/drivers-2022-04-02' of git://git.kernel.dk/linux-block
Pull block driver fix from Jens Axboe:
"Got two reports on nbd spewing warnings on load now, which is a
regression from a commit that went into your tree yesterday.
Revert the problematic change for now"
* tag 'for-5.18/drivers-2022-04-02' of git://git.kernel.dk/linux-block:
Revert "nbd: fix possible overflow on 'first_minor' in nbd_dev_add()"
Merge tag 'tag-chrome-platform-for-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/chrome-platform/linux
Pull chrome platform updates from Benson Leung:
"cros_ec_typec:
- Check for EC device - Fix a crash when using the cros_ec_typec
driver on older hardware not capable of typec commands
- Make try power role optional
- Mux configuration reorganization series from Prashant
cros_ec_debugfs:
- Fix use after free. Thanks Tzung-bi
sensorhub:
- cros_ec_sensorhub fixup - Split trace include file
misc:
- Add new mailing list for chrome-platform development:
chrome-platform@lists.linux.dev
Now with patchwork!"
* tag 'tag-chrome-platform-for-v5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/chrome-platform/linux:
platform/chrome: cros_ec_debugfs: detach log reader wq from devm
platform: chrome: Split trace include file
platform/chrome: cros_ec_typec: Update mux flags during partner removal
platform/chrome: cros_ec_typec: Configure muxes at start of port update
platform/chrome: cros_ec_typec: Get mux state inside configure_mux
platform/chrome: cros_ec_typec: Move mux flag checks
platform/chrome: cros_ec_typec: Check for EC device
platform/chrome: cros_ec_typec: Make try power role optional
MAINTAINERS: platform-chrome: Add new chrome-platform@lists.linux.dev list
After being merged, user_events become more visible to a wider audience
that have concerns with the current API.
It is too late to fix this for this release, but instead of a full
revert, just mark it as BROKEN (which prevents it from being selected in
make config). Then we can work finding a better API. If that fails,
then it will need to be completely reverted.
To not have the code silently bitrot, still allow building it with
COMPILE_TEST.
And to prevent the uapi header from being installed, then later changed,
and then have an old distro user space see the old version, move the
header file out of the uapi directory.
Surround the include with CONFIG_COMPILE_TEST to the current location,
but when the BROKEN tag is taken off, it will use the uapi directory,
and fail to compile. This is a good way to remind us to move the header
back.
tracing: Move user_events.h temporarily out of include/uapi
While user_events API is under development and has been marked for broken
to not let the API become fixed, move the header file out of the uapi
directory. This is to prevent it from being installed, then later changed,
and then have an old distro user space update with a new kernel, where
applications see the user_events being available, but the old header is in
place, and then they get compiled incorrectly.
Also, surround the include with CONFIG_COMPILE_TEST to the current
location, but when the BROKEN tag is taken off, it will use the uapi
directory, and fail to compile. This is a good way to remind us to move
the header back.
Christophe Leroy [Wed, 30 Mar 2022 07:00:19 +0000 (09:00 +0200)]
ftrace: Make ftrace_graph_is_dead() a static branch
ftrace_graph_is_dead() is used on hot paths, it just reads a variable
in memory and is not worth suffering function call constraints.
For instance, at entry of prepare_ftrace_return(), inlining it avoids
saving prepare_ftrace_return() parameters to stack and restoring them
after calling ftrace_graph_is_dead().
While at it using a static branch is even more performant and is
rather well adapted considering that the returned value will almost
never change.
Inline ftrace_graph_is_dead() and replace 'kill_ftrace_graph' bool
by a static branch.
After being merged, user_events become more visible to a wider audience
that have concerns with the current API. It is too late to fix this for
this release, but instead of a full revert, just mark it as BROKEN (which
prevents it from being selected in make config). Then we can work finding
a better API. If that fails, then it will need to be completely reverted.
tracing: Rename the staging files for trace_events
When looking for implementation of different phases of the creation of the
TRACE_EVENT() macro, it is pretty useless when all helper macro
redefinitions are in files labeled "stageX_defines.h". Rename them to
state which phase the files are for. For instance, when looking for the
defines that are used to create the event fields, seeing
"stage4_event_fields.h" gives the developer a good idea that the defines
are in that file.
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Hou Wenlong [Tue, 15 Mar 2022 09:35:13 +0000 (17:35 +0800)]
KVM: x86/mmu: Don't rebuild page when the page is synced and no tlb flushing is required
Before Commit 32f7c9f979226 ("KVM: X86: Change kvm_sync_page()
to return true when remote flush is needed"), the return value
of kvm_sync_page() indicates whether the page is synced, and
kvm_mmu_get_page() would rebuild page when the sync fails.
But now, kvm_sync_page() returns false when the page is
synced and no tlb flushing is required, which leads to
rebuild page in kvm_mmu_get_page(). So return the return
value of mmu->sync_page() directly and check it in
kvm_mmu_get_page(). If the sync fails, the page will be
zapped and the invalid_list is not empty, so set flush as
true is accepted in mmu_sync_children().
Cc: stable@vger.kernel.org Fixes: 32f7c9f979226 ("KVM: X86: Change kvm_sync_page() to return true when remote flush is needed") Signed-off-by: Hou Wenlong <houwenlong.hwl@antgroup.com> Acked-by: Lai Jiangshan <jiangshanlai@gmail.com>
Message-Id: <0dabeeb789f57b0d793f85d073893063e692032d.1647336064.git.houwenlong.hwl@antgroup.com>
[mmu_sync_children should not flush if the page is zapped. - Paolo] Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Jon Kohler [Thu, 24 Mar 2022 00:44:39 +0000 (20:44 -0400)]
KVM: x86: optimize PKU branching in kvm_load_{guest|host}_xsave_state
kvm_load_{guest|host}_xsave_state handles xsave on vm entry and exit,
part of which is managing memory protection key state. The latest
arch.pkru is updated with a rdpkru, and if that doesn't match the base
host_pkru (which about 70% of the time), we issue a __write_pkru.
To improve performance, implement the following optimizations:
1. Reorder if conditions prior to wrpkru in both
kvm_load_{guest|host}_xsave_state.
Flip the ordering of the || condition so that XFEATURE_MASK_PKRU is
checked first, which when instrumented in our environment appeared
to be always true and less overall work than kvm_read_cr4_bits.
For kvm_load_guest_xsave_state, hoist arch.pkru != host_pkru ahead
one position. When instrumented, I saw this be true roughly ~70% of
the time vs the other conditions which were almost always true.
With this change, we will avoid 3rd condition check ~30% of the time.
2. Wrap PKU sections with CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS,
as if the user compiles out this feature, we should not have
these branches at all.
Signed-off-by: Jon Kohler <jon@nutanix.com>
Message-Id: <20220324004439.6709-1-jon@nutanix.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Maxim Levitsky [Tue, 22 Mar 2022 17:40:50 +0000 (19:40 +0200)]
KVM: x86: SVM: allow AVIC to co-exist with a nested guest running
Inhibit the AVIC of the vCPU that is running nested for the duration of the
nested run, so that all interrupts arriving from both its vCPU siblings
and from KVM are delivered using normal IPIs and cause that vCPU to vmexit.
Note that unlike normal AVIC inhibition, there is no need to
update the AVIC mmio memslot, because the nested guest uses its
own set of paging tables.
That also means that AVIC doesn't need to be inhibited VM wide.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220322174050.241850-7-mlevitsk@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Maxim Levitsky [Tue, 22 Mar 2022 17:40:48 +0000 (19:40 +0200)]
KVM: x86: nSVM: implement nested vGIF
In case L1 enables vGIF for L2, the L2 cannot affect L1's GIF, regardless
of STGI/CLGI intercepts, and since VM entry enables GIF, this means
that L1's GIF is always 1 while L2 is running.
Thus in this case leave L1's vGIF in vmcb01, while letting L2
control the vGIF thus implementing nested vGIF.
Also allow KVM to toggle L1's GIF during nested entry/exit
by always using vmcb01.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220322174050.241850-5-mlevitsk@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Maxim Levitsky [Tue, 22 Mar 2022 17:40:47 +0000 (19:40 +0200)]
KVM: x86: nSVM: support PAUSE filtering when L0 doesn't intercept PAUSE
Expose the pause filtering and threshold in the guest CPUID
and support PAUSE filtering when possible:
- If the L0 doesn't intercept PAUSE (cpu_pm=on), then allow L1 to
have full control over PAUSE filtering.
- if the L1 doesn't intercept PAUSE, use host values and update
the adaptive count/threshold even when running nested.
- Otherwise always exit to L1; it is not really possible to merge
the fields correctly. It is expected that in this case, userspace
will not enable this feature in the guest CPUID, to avoid having the
guest update both fields pointlessly.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220322174050.241850-4-mlevitsk@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This was tested with kvm-unit-test that was developed
for this purpose.
Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
Message-Id: <20220322174050.241850-3-mlevitsk@redhat.com>
[Copy all of DEBUGCTL except for reserved bits. - Paolo] Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Li RongQing [Wed, 9 Mar 2022 08:46:50 +0000 (16:46 +0800)]
KVM: x86: Support the vCPU preemption check with nopvspin and realtime hint
If guest kernel is configured with nopvspin, or CONFIG_PARAVIRT_SPINLOCK
is disabled, or guest find its has dedicated pCPUs from realtime hint
feature, the pvspinlock will be disabled, and vCPU preemption check
is disabled too.
Hoever, KVM still can emulating HLT for vCPU for both cases. Checking if a vCPU
is preempted or not can still boost performance in IPI-heavy scenarios such as
unixbench file copy and pipe-based context switching tests: Here the vCPU is
running with a dedicated pCPU, so the guest kernel has nopvspin but is
emulating HLT for the vCPU:
Testcase Base with patch
System Benchmarks Index Values INDEX INDEX
Dhrystone 2 using register variables 3278.4 3277.7
Double-Precision Whetstone 822.8 825.8
Execl Throughput 1296.5 941.1
File Copy 1024 bufsize 2000 maxblocks 2124.2 2142.7
File Copy 256 bufsize 500 maxblocks 1335.9 1353.6
File Copy 4096 bufsize 8000 maxblocks 4256.3 4760.3
Pipe Throughput 1050.1 1054.0
Pipe-based Context Switching 243.3 352.0
Process Creation 820.1 814.4
Shell Scripts (1 concurrent) 2169.0 2086.0
Shell Scripts (8 concurrent) 7710.3 7576.3
System Call Overhead 672.4 673.9
======== =======
System Benchmarks Index Score 1467.2 1483.0
Move the setting of pv_ops.lock.vcpu_is_preempted to kvm_guest_init, so
that it does not depend on pvspinlock.
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Message-Id: <1646815610-43315-1-git-send-email-lirongqing@baidu.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Fri, 25 Feb 2022 14:53:04 +0000 (14:53 +0000)]
KVM: x86: Test case for TSC scaling and offset sync
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Message-Id: <20220225145304.36166-4-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: x86: Don't snapshot "max" TSC if host TSC is constant
Don't snapshot tsc_khz into max_tsc_khz during KVM initialization if the
host TSC is constant, in which case the actual TSC frequency will never
change and thus capturing the "max" TSC during initialization is
unnecessary, KVM can simply use tsc_khz during VM creation.
On CPUs with constant TSC, but not a hardware-specified TSC frequency,
snapshotting max_tsc_khz and using that to set a VM's default TSC
frequency can lead to KVM thinking it needs to manually scale the guest's
TSC if refining the TSC completes after KVM snapshots tsc_khz. The
actual frequency never changes, only the kernel's calculation of what
that frequency is changes. On systems without hardware TSC scaling, this
either puts KVM into "always catchup" mode (extremely inefficient), or
prevents creating VMs altogether.
Ideally, KVM would not be able to race with TSC refinement, or would have
a hook into tsc_refine_calibration_work() to get an alert when refinement
is complete. Avoiding the race altogether isn't practical as refinement
takes a relative eternity; it's deliberately put on a work queue outside
of the normal boot sequence to avoid unnecessarily delaying boot.
Adding a hook is doable, but somewhat gross due to KVM's ability to be
built as a module. And if the TSC is constant, which is likely the case
for every VMX/SVM-capable CPU produced in the last decade, the race can
be hit if and only if userspace is able to create a VM before TSC
refinement completes; refinement is slow, but not that slow.
For now, punt on a proper fix, as not taking a snapshot can help some
uses cases and not taking a snapshot is arguably correct irrespective of
the race with refinement.
[ dwmw2: Rebase on top of KVM-wide default_tsc_khz to ensure that all
vCPUs get the same frequency even if we hit the race. ]
Cc: Suleiman Souhlal <suleiman@google.com> Cc: Anton Romanov <romanton@google.com> Signed-off-by: Sean Christopherson <seanjc@google.com> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Message-Id: <20220225145304.36166-3-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Fri, 25 Feb 2022 14:53:02 +0000 (14:53 +0000)]
KVM: x86: Accept KVM_[GS]ET_TSC_KHZ as a VM ioctl.
This sets the default TSC frequency for subsequently created vCPUs.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Message-Id: <20220225145304.36166-2-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Zeng Guang [Tue, 15 Mar 2022 14:58:36 +0000 (22:58 +0800)]
KVM: VMX: Prepare VMCS setting for posted interrupt enabling when APICv is available
Currently KVM setup posted interrupt VMCS only depending on
per-vcpu APICv activation status at the vCPU creation time.
However, this status can be toggled dynamically under some
circumstance. So potentially, later posted interrupt enabling
may be problematic without VMCS readiness.
To fix this, always settle the VMCS setting for posted interrupt
as long as APICv is available and lapic locates in kernel.
Signed-off-by: Zeng Guang <guang.zeng@intel.com>
Message-Id: <20220315145836.9910-1-guang.zeng@intel.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Wed, 9 Mar 2022 14:38:35 +0000 (14:38 +0000)]
KVM: x86/xen: Update self test for Xen PV timers
Add test cases for timers in the past, and reading the status of a timer
which has already fired.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Message-Id: <20220309143835.253911-3-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Thu, 3 Mar 2022 15:41:27 +0000 (15:41 +0000)]
KVM: x86/xen: Add self tests for KVM_XEN_HVM_CONFIG_EVTCHN_SEND
Test a combination of event channel send, poll and timer operations.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
Message-Id: <20220303154127.202856-18-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Boris Ostrovsky [Thu, 3 Mar 2022 15:41:26 +0000 (15:41 +0000)]
KVM: x86/xen: handle PV spinlocks slowpath
Add support for SCHEDOP_poll hypercall.
This implementation is optimized for polling for a single channel, which
is what Linux does. Polling for multiple channels is not especially
efficient (and has not been tested).
PV spinlocks slow path uses this hypercall, and explicitly crash if it's
not supported.
[ dwmw2: Rework to use kvm_vcpu_halt(), not supported for 32-bit guests ]
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-17-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Thu, 3 Mar 2022 15:41:25 +0000 (15:41 +0000)]
KVM: x86/xen: Advertise and document KVM_XEN_HVM_CONFIG_EVTCHN_SEND
At the end of the patch series adding this batch of event channel
acceleration features, finally add the feature bit which advertises
them and document it all.
For SCHEDOP_poll we need to wake a polling vCPU when a given port
is triggered, even when it's masked — and we want to implement that
in the kernel, for efficiency. So we want the kernel to know that it
has sole ownership of event channel delivery. Thus, we allow
userspace to make the 'promise' by setting the corresponding feature
bit in its KVM_XEN_HVM_CONFIG call. As we implement SCHEDOP_poll
bypass later, we will do so only if that promise has been made by
userspace.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-16-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Thu, 3 Mar 2022 15:41:24 +0000 (15:41 +0000)]
KVM: x86/xen: Support per-vCPU event channel upcall via local APIC
Windows uses a per-vCPU vector, and it's delivered via the local APIC
basically like an MSI (with associated EOI) unlike the traditional
guest-wide vector which is just magically asserted by Xen (and in the
KVM case by kvm_xen_has_interrupt() / kvm_cpu_get_extint()).
Now that the kernel is able to raise event channel events for itself,
being able to do so for Windows guests is also going to be useful.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-15-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Thu, 3 Mar 2022 15:41:23 +0000 (15:41 +0000)]
KVM: x86/xen: Kernel acceleration for XENVER_version
Turns out this is a fast path for PV guests because they use it to
trigger the event channel upcall. So letting it bounce all the way up
to userspace is not great.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-14-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The event channel corresponding to the timer virq is then used to inject
events once timer deadlines are met. For now we back the PV timer with
hrtimer.
David Woodhouse [Thu, 3 Mar 2022 15:41:21 +0000 (15:41 +0000)]
KVM: x86/xen: Add KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID
In order to intercept hypercalls such as VCPUOP_set_singleshot_timer, we
need to be aware of the Xen CPU numbering.
This looks a lot like the Hyper-V handling of vpidx, for obvious reasons.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-12-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Joao Martins [Thu, 3 Mar 2022 15:41:20 +0000 (15:41 +0000)]
KVM: x86/xen: handle PV IPI vcpu yield
Cooperative Linux guests after an IPI-many may yield vcpu if
any of the IPI'd vcpus were preempted (i.e. runstate is 'runnable'.)
Support SCHEDOP_yield for handling yield.
Signed-off-by: Joao Martins <joao.m.martins@oracle.com> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-11-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Joao Martins [Thu, 3 Mar 2022 15:41:19 +0000 (15:41 +0000)]
KVM: x86/xen: intercept EVTCHNOP_send from guests
Userspace registers a sending @port to either deliver to an @eventfd
or directly back to a local event channel port.
After binding events the guest or host may wish to bind those
events to a particular vcpu. This is usually done for unbound
and and interdomain events. Update requests are handled via the
KVM_XEN_EVTCHN_UPDATE flag.
Unregistered ports are handled by the emulator.
Co-developed-by: Ankur Arora <ankur.a.arora@oracle.com> Co-developed-By: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Joao Martins <joao.m.martins@oracle.com> Signed-off-by: Ankur Arora <ankur.a.arora@oracle.com> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-10-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Thu, 3 Mar 2022 15:41:18 +0000 (15:41 +0000)]
KVM: x86/xen: Support direct injection of event channel events
This adds a KVM_XEN_HVM_EVTCHN_SEND ioctl which allows direct injection
of events given an explicit { vcpu, port, priority } in precisely the
same form that those fields are given in the IRQ routing table.
Userspace is currently able to inject 2-level events purely by setting
the bits in the shared_info and vcpu_info, but FIFO event channels are
harder to deal with; we will need the kernel to take sole ownership of
delivery when we support those.
A patch advertising this feature with a new bit in the KVM_CAP_XEN_HVM
ioctl will be added in a subsequent patch.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-9-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
David Woodhouse [Thu, 3 Mar 2022 15:41:17 +0000 (15:41 +0000)]
KVM: x86/xen: Make kvm_xen_set_evtchn() reusable from other places
Clean it up to return -errno on error consistently, while still being
compatible with the return conventions for kvm_arch_set_irq_inatomic()
and the kvm_set_irq() callback.
We use -ENOTCONN to indicate when the port is masked. No existing users
care, except that it's negative.
Also allow it to optimise the vCPU lookup. Unless we abuse the lapic
map, there is no quick lookup from APIC ID to a vCPU; the logic in
kvm_get_vcpu_by_id() will just iterate over all vCPUs till it finds
the one it wants. So do that just once and stash the result in the
struct kvm_xen_evtchn for next time.
Signed-off-by: David Woodhouse <dwmw@amazon.co.uk> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220303154127.202856-8-dwmw2@infradead.org> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
projects / kernel.git / log