diff options
| author |   Sean Christopherson <seanjc@google.com> | 2021-11-11 02:07:23 +0000 |
|---|---|---|
| committer |   Peter Zijlstra <peterz@infradead.org> | 2021-11-17 14:49:06 +0100 |
| commit | 5c7df80e2ce4c954c80eb4ecf5fa002a5ff5d2d6 (patch) | |
| tree | 4e36d1cb20acedb1e19537a908fa45e76386224e | |
| parent | perf: Protect perf_guest_cbs with RCU (diff) | |
| download | linux-dev-5c7df80e2ce4c954c80eb4ecf5fa002a5ff5d2d6.tar.xz linux-dev-5c7df80e2ce4c954c80eb4ecf5fa002a5ff5d2d6.zip | |
KVM: x86: Register perf callbacks after calling vendor's hardware_setup()
Wait to register perf callbacks until after doing vendor hardaware setup.
VMX's hardware_setup() configures Intel Processor Trace (PT) mode, and a
future fix to register the Intel PT guest interrupt hook if and only if
Intel PT is exposed to the guest will consume the configured PT mode.
Delaying registration to hardware setup is effectively a nop as KVM's perf
hooks all pivot on the per-CPU current_vcpu, which is non-NULL only when
KVM is handling an IRQ/NMI in a VM-Exit path. I.e. current_vcpu will be
NULL throughout both kvm_arch_init() and kvm_arch_hardware_setup().
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20211111020738.2512932-3-seanjc@google.com
| -rw-r--r-- | arch/x86/kvm/x86.c | 7 |
1 files changed, 4 insertions, 3 deletions
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index dc7eb5fddfd3..50f0cd16f2d4 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -8626,8 +8626,6 @@ int kvm_arch_init(void *opaque) kvm_timer_init(); - perf_register_guest_info_callbacks(&kvm_guest_cbs); - if (boot_cpu_has(X86_FEATURE_XSAVE)) { host_xcr0 = xgetbv(XCR_XFEATURE_ENABLED_MASK); supported_xcr0 = host_xcr0 & KVM_SUPPORTED_XCR0; @@ -8659,7 +8657,6 @@ void kvm_arch_exit(void) clear_hv_tscchange_cb(); #endif kvm_lapic_exit(); - perf_unregister_guest_info_callbacks(&kvm_guest_cbs); if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) cpufreq_unregister_notifier(&kvmclock_cpufreq_notifier_block, @@ -11225,6 +11222,8 @@ int kvm_arch_hardware_setup(void *opaque) memcpy(&kvm_x86_ops, ops->runtime_ops, sizeof(kvm_x86_ops)); kvm_ops_static_call_update(); + perf_register_guest_info_callbacks(&kvm_guest_cbs); + if (!kvm_cpu_cap_has(X86_FEATURE_XSAVES)) supported_xss = 0; @@ -11252,6 +11251,8 @@ int kvm_arch_hardware_setup(void *opaque) void kvm_arch_hardware_unsetup(void) { + perf_unregister_guest_info_callbacks(&kvm_guest_cbs); + static_call(kvm_x86_hardware_unsetup)(); } |