- 15 Nov, 2019 4 commits
-
-
Andrea Arcangeli authored
It's enough to check the value and issue the direct call. After this commit is applied, here the most common retpolines executed under a high resolution timer workload in the guest on a VMX host: [..] @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 267 @[]: 2256 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 __kvm_wait_lapic_expire+284 vmx_vcpu_run.part.97+1091 vcpu_enter_guest+377 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 2390 @[]: 33410 @total: 315707 Note the highest hit above is __delay so probably not worth optimizing even if it would be more frequent than 2k hits per sec. Signed-off-by:Andrea Arcangeli <aarcange@redhat.com> Signed-off-by:
Paolo Bonzini <pbonzini@redhat.com>
-
Andrea Arcangeli authored
It's enough to check the exit value and issue a direct call to avoid the retpoline for all the common vmexit reasons. After this commit is applied, here the most common retpolines executed under a high resolution timer workload in the guest on a SVM host: [..] @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 ktime_get_update_offsets_now+70 hrtimer_interrupt+131 smp_apic_timer_interrupt+106 apic_timer_interrupt+15 start_sw_timer+359 restart_apic_timer+85 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 1940 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_r12+33 force_qs_rnp+217 rcu_gp_kthread+1270 kthread+268 ret_from_fork+34 ]: 4644 @[]: 25095 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 lapic_next_event+28 clockevents_program_event+148 hrtimer_start_range_ns+528 start_sw_timer+356 restart_apic_timer+85 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 41474 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 clockevents_program_event+148 hrtimer_start_range_ns+528 start_sw_timer+356 restart_apic_timer+85 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 41474 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 ktime_get+58 clockevents_program_event+84 hrtimer_start_range_ns+528 start_sw_timer+356 restart_apic_timer+85 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 41887 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 lapic_next_event+28 clockevents_program_event+148 hrtimer_try_to_cancel+168 hrtimer_cancel+21 kvm_set_lapic_tscdeadline_msr+43 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 42723 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 clockevents_program_event+148 hrtimer_try_to_cancel+168 hrtimer_cancel+21 kvm_set_lapic_tscdeadline_msr+43 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 42766 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 ktime_get+58 clockevents_program_event+84 hrtimer_try_to_cancel+168 hrtimer_cancel+21 kvm_set_lapic_tscdeadline_msr+43 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 42848 @[ trace_retpoline+1 __trace_retpoline+30 __x86_indirect_thunk_rax+33 ktime_get+58 start_sw_timer+279 restart_apic_timer+85 kvm_set_msr_common+1497 msr_interception+142 vcpu_enter_guest+684 kvm_arch_vcpu_ioctl_run+261 kvm_vcpu_ioctl+559 do_vfs_ioctl+164 ksys_ioctl+96 __x64_sys_ioctl+22 do_syscall_64+89 entry_SYSCALL_64_after_hwframe+68 ]: 499845 @total: 1780243 SVM has no TSC based programmable preemption timer so it is invoking ktime_get() frequently. Signed-off-by: -
Andrea Arcangeli authored
It's enough to check the exit value and issue a direct call to avoid the retpoline for all the common vmexit reasons. Of course CONFIG_RETPOLINE already forbids gcc to use indirect jumps while compiling all switch() statements, however switch() would still allow the compiler to bisect the case value. It's more efficient to prioritize the most frequent vmexits instead. The halt may be slow paths from the point of the guest, but not necessarily so from the point of the host if the host runs at full CPU capacity and no host CPU is ever left idle. Signed-off-by:
-
Andrea Arcangeli authored
Eliminate wasteful call/ret non RETPOLINE case and unnecessary fentry dynamic tracing hooking points. Signed-off-by:
-
- 11 Nov, 2019 1 commit
-
-
Miaohe Lin authored
The odd out jump label is really not needed. Get rid of it by return true directly while r < 0 as suggested by Paolo. This further lead to var changed being unused. Remove it too. Signed-off-by:
Miaohe Lin <linmiaohe@huawei.com> Signed-off-by:
-
- 02 Nov, 2019 1 commit
-
-
Marcelo Tosatti authored
Commit 0bc48bea ("KVM: x86: update master clock before computing kvmclock_offset") switches the order of operations to avoid the conversion TSC (without frequency correction) -> system_timestamp (with frequency correction), which might cause a time jump. However, it leaves any other masterclock update unsafe, which includes, at the moment: * HV_X64_MSR_REFERENCE_TSC MSR write. * TSC writes. * Host suspend/resume. Avoid the time jump issue by using frequency uncorrected CLOCK_MONOTONIC_RAW clock. Its the guests time keeping software responsability to track and correct a reference clock such as UTC. This fixes forward time jump (which can result in failure to bring up a vCPU) during vCPU hotplug: Oct 11 14:48:33 storage kernel: CPU2 has been hot-added Oct 11 14:48:34 storage kernel: CPU3 has been hot-added Oct 11 14:49:22 storage kernel: smpboot: Booting Node 0 Processor 2 APIC 0x2 <-- time jump of almost 1 minute Oct 11 14:49:22 storage kernel: smpboot: do_boot_cpu failed(-1) to wakeup CPU#2 Oct 11 14:49:23 storage kernel: smpboot: Booting Node 0 Processor 3 APIC 0x3 Oct 11 14:49:23 storage kernel: kvm-clock: cpu 3, msr 0:7ff640c1, secondary cpu clock Which happens because: /* * Wait 10s total for a response from AP */ boot_error = -1; timeout = jiffies + 10*HZ; while (time_before(jiffies, timeout)) { ... } Analyzed-by:
Igor Mammedov <imammedo@redhat.com> Signed-off-by:
Marcelo Tosatti <mtosatti@redhat.com> Signed-off-by:
-
- 31 Oct, 2019 1 commit
-
-
Paolo Bonzini authored
Merge tag 'kvm-ppc-next-5.5-1' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc into HEAD KVM PPC update for 5.5 * Add capability to tell userspace whether we can single-step the guest. * Improve the allocation of XIVE virtual processor IDs, to reduce the risk of running out of IDs when running many VMs on POWER9. * Rewrite interrupt synthesis code to deliver interrupts in virtual mode when appropriate. * Minor cleanups and improvements.
-
- 25 Oct, 2019 1 commit
-
-
Yi Wang authored
We get two warning when build kernel with W=1: arch/x86/kernel/kvm.c:872:6: warning: no previous prototype for ‘arch_haltpoll_enable’ [-Wmissing-prototypes] arch/x86/kernel/kvm.c:885:6: warning: no previous prototype for ‘arch_haltpoll_disable’ [-Wmissing-prototypes] Including the missing head file can fix this. Signed-off-by:
Yi Wang <wang.yi59@zte.com.cn> Signed-off-by:
-
- 22 Oct, 2019 32 commits
-
-
Sean Christopherson authored
Add a new helper, kvm_put_kvm_no_destroy(), to handle putting a borrowed reference[*] to the VM when installing a new file descriptor fails. KVM expects the refcount to remain valid in this case, as the in-progress ioctl() has an explicit reference to the VM. The primary motiviation for the helper is to document that the 'kvm' pointer is still valid after putting the borrowed reference, e.g. to document that doing mutex(&kvm->lock) immediately after putting a ref to kvm isn't broken. [*] When exposing a new object to userspace via a file descriptor, e.g. a new vcpu, KVM grabs a reference to itself (the VM) prior to making the object visible to userspace to avoid prematurely freeing the VM in the scenario where userspace immediately closes file descriptor. Signed-off-by:Sean Christopherson <sean.j.christopherson@intel.com> Signed-off-by:
-
Aaron Lewis authored
Ensure that IA32_XSS appears in KVM_GET_MSR_INDEX_LIST if it can be set to a non-zero value. Reviewed-by:
Jim Mattson <jmattson@google.com> Signed-off-by:
Aaron Lewis <aaronlewis@google.com> Change-Id: Ia2d644f69e2d6d8c27d7e0a7a45c2bf9c42bf5ff Signed-off-by:
-
Aaron Lewis authored
AMD CPUs now support XSAVES in a limited fashion (they require IA32_XSS to be zero). AMD has no equivalent of Intel's "Enable XSAVES/XRSTORS" VM-execution control. Instead, XSAVES is always available to the guest when supported on the host. Reviewed-by:
-
Aaron Lewis authored
Hoist support for RDMSR/WRMSR of IA32_XSS from vmx into common code so that it can be used for svm as well. Right now, kvm only allows the guest IA32_XSS to be zero, so the guest's usage of XSAVES will be exactly the same as XSAVEC. Reviewed-by:
-
Aaron Lewis authored
Hoist the vendor-specific code related to loading the hardware IA32_XSS MSR with guest/host values on VM-entry/VM-exit to common x86 code. Reviewed-by:
-
Aaron Lewis authored
When the guest can execute the XSAVES/XRSTORS instructions, use wrmsr to set the hardware IA32_XSS MSR to guest/host values on VM-entry/VM-exit, rather than the MSR-load areas. By using the same approach as AMD, we will be able to use a common implementation for both (in the next patch). Reviewed-by:
-
Aaron Lewis authored
When the guest can execute the XSAVES/XRSTORS instructions, set the hardware IA32_XSS MSR to guest/host values on VM-entry/VM-exit. Note that vcpu->arch.ia32_xss is currently guaranteed to be 0 on AMD, since there is no way to change it. Suggested-by:
-
Aaron Lewis authored
The kvm_vcpu variable, guest_xcr0_loaded, is a waste of an 'int' and a conditional branch. VMX and SVM are the only users, and both unconditionally pair kvm_load_guest_xcr0() with kvm_put_guest_xcr0() making this check unnecessary. Without this variable, the predicates in kvm_load_guest_xcr0 and kvm_put_guest_xcr0 should match. Suggested-by:
-
Aaron Lewis authored
Volume 4 of the SDM says that IA32_XSS is supported if CPUID(EAX=0DH,ECX=1):EAX.XSS[bit 3] is set, so only the X86_FEATURE_XSAVES check is necessary (X86_FEATURE_XSAVES is the Linux name for CPUID(EAX=0DH,ECX=1):EAX.XSS[bit 3]). Fixes: 4d763b16 ("KVM: VMX: check CPUID before allowing read/write of IA32_XSS") Reviewed-by:
-
Aaron Lewis authored
Cache whether XSAVES is enabled in the guest by adding xsaves_enabled to vcpu->arch. Reviewed-by:
-
Xiaoyao Li authored
Rename {vmx,nested_vmx}_vcpu_setup() to match what they really do. Signed-off-by:Xiaoyao Li <xiaoyao.li@intel.com> Signed-off-by:
-
Xiaoyao Li authored
Move the initialization of vmx->guest_msrs[] from vmx_vcpu_setup() to vmx_create_vcpu(), and put it right after its allocation. This also is the preperation for next patch. Signed-off-by:
-
Xiaoyao Li authored
... It can be removed here because the same code is called later in vmx_vcpu_reset() as the flow: kvm_arch_vcpu_setup() -> kvm_vcpu_reset() -> vmx_vcpu_reset() Signed-off-by:
-
Xiaoyao Li authored
Move the code that writes vmx->vpid to vmcs from vmx_vcpu_reset() to vmx_vcpu_setup(), because vmx->vpid is allocated when creating vcpu and never changed. So we don't need to update the vmcs.vpid when resetting vcpu. Signed-off-by:
-
Like Xu authored
Replace the explicit declaration of "u64 reprogram_pmi" with the generic macro DECLARE_BITMAP for all possible appropriate number of bits. Suggested-by:
Like Xu <like.xu@linux.intel.com> Signed-off-by:
-
Miaohe Lin authored
If we reach here with r = 0, we will reassign r = 0 unnecesarry, then do the label set_irqchip_out work. If we reach here with r != 0, then we will do the label work directly. So this if statement and r = 0 assignment is redundant. Signed-off-by:
-
Suthikulpanit, Suravee authored
Generally, APICv for all vcpus in the VM are enable/disable in the same manner. So, get_enable_apicv() should represent APICv status of the VM instead of each VCPU. Modify kvm_x86_ops.get_enable_apicv() to take struct kvm as parameter instead of struct kvm_vcpu. Reviewed-by:
Vitaly Kuznetsov <vkuznets@redhat.com> Signed-off-by:
Suravee Suthikulpanit <suravee.suthikulpanit@amd.com> Signed-off-by:
-
Sean Christopherson authored
Handle caching CR3 (from VMX's VMCS) into struct kvm_vcpu via the common cache_reg() callback and drop the dedicated decache_cr3(). The name decache_cr3() is somewhat confusing as the caching behavior of CR3 follows that of GPRs, RFLAGS and PDPTRs, (handled via cache_reg()), and has nothing in common with the caching behavior of CR0/CR4 (whose decache_cr{0,4}_guest_bits() likely provided the 'decache' verbiage). This would effectivel adds a BUG() if KVM attempts to cache CR3 on SVM. Change it to a WARN_ON_ONCE() -- if the cache never requires filling, the value is already in the right place -- and opportunistically add one in VMX to provide an equivalent check. Signed-off-by: -
Sean Christopherson authored
Add helpers to prettify code that tests and/or marks whether or not a register is available and/or dirty. Suggested-by:
-
Sean Christopherson authored
Now that indexing into arch.regs is either protected by WARN_ON_ONCE or done with hardcoded enums, combine all definitions for registers that are tracked by regs_avail and regs_dirty into 'enum kvm_reg'. Having a single enum type will simplify additional cleanup related to regs_avail and regs_dirty. Signed-off-by:
-
Sean Christopherson authored
Add WARN_ON_ONCE() checks in kvm_register_{read,write}() to detect reg values that would cause KVM to overflow vcpu->arch.regs. Change the reg param to an 'int' to make it clear that the reg index is unverified. Regarding the overhead of WARN_ON_ONCE(), now that all fixed GPR reads and writes use dedicated accessors, e.g. kvm_rax_read(), the overhead is limited to flows where the reg index is generated at runtime. And there is at least one historical bug where KVM has generated an out-of- bounds access to arch.regs (see commit b68f3cc7, "KVM: x86: Always use 32-bit SMRAM save state for 32-bit kernels"). Adding the WARN_ON_ONCE() protection paves the way for additional cleanup related to kvm_reg and kvm_reg_ex. Signed-off-by: -
Sean Christopherson authored
Rework vmx_set_rflags() to avoid the extra code need to handle emulation of real mode and invalid state when unrestricted guest is disabled. The primary reason for doing so is to avoid the call to vmx_get_rflags(), which will incur a VMREAD when RFLAGS is not already available. When running nested VMs, the majority of calls to vmx_set_rflags() will occur without an associated vmx_get_rflags(), i.e. when stuffing GUEST_RFLAGS during transitions between vmcs01 and vmcs02. Note, vmx_get_rflags() guarantees RFLAGS is marked available. Signed-off-by:
-
Sean Christopherson authored
Capture struct vcpu_vmx in a local variable to improve the readability of vmx_{g,s}et_rflags(). No functional change intended. Signed-off-by: -
Sean Christopherson authored
Skip the VMWRITE to update GUEST_CR3 if CR3 is not available, i.e. has not been read from the VMCS since the last VM-Enter. If vcpu->arch.cr3 is stale, kvm_read_cr3(vcpu) will refresh vcpu->arch.cr3 from the VMCS, meaning KVM will do a VMREAD and then VMWRITE the value it just pulled from the VMCS. Note, this is a purely theoretical change, no instances of skipping the VMREAD+VMWRITE have been observed with this change. Tested-by:
Reto Buerki <reet@codelabs.ch> Tested-by:
-
Tom Lendacky authored
Performing a WBINVD and DF_FLUSH are expensive operations. Currently, a WBINVD/DF_FLUSH is performed every time an SEV guest terminates. However, the WBINVD/DF_FLUSH is only required when an ASID is being re-allocated to a new SEV guest. Also, a single WBINVD/DF_FLUSH can enable all ASIDs that have been disassociated from guests through DEACTIVATE. To reduce the number of WBINVD/DF_FLUSH invocations, introduce a new ASID bitmap to track ASIDs that need to be reclaimed. When an SEV guest is terminated, add its ASID to the reclaim bitmap instead of clearing the bitmap in the existing SEV ASID bitmap. This delays the need to perform a WBINVD/DF_FLUSH invocation when an SEV guest terminates until all of the available SEV ASIDs have been used. At that point, the WBINVD/DF_FLUSH invocation can be performed and all ASIDs in the reclaim bitmap moved to the available ASIDs bitmap. The semaphore around DEACTIVATE can be changed to a read semaphore with the semaphore taken in write mode before performing the WBINVD/DF_FLUSH. Tested-by:
David Rientjes <rientjes@google.com> Signed-off-by:
Tom Lendacky <thomas.lendacky@amd.com> Signed-off-by:
-
Tom Lendacky authored
Performing a WBINVD and DF_FLUSH are expensive operations. The SEV support currently performs this WBINVD/DF_FLUSH combination when an SEV guest is terminated, so there is no need for it to be done before LAUNCH. However, when the SEV firmware transitions the platform from UNINIT state to INIT state, all ASIDs will be marked invalid across all threads. Therefore, as part of transitioning the platform to INIT state, perform a WBINVD/DF_FLUSH after a successful INIT in the PSP/SEV device driver. Since the PSP/SEV device driver is x86 only, it can reference and use the WBINVD related functions directly. Cc: Gary Hook <gary.hook@amd.com> Cc: Herbert Xu <herbert@gondor.apana.org.au> Cc: "David S. Miller" <davem@davemloft.net> Tested-by:
-
Sean Christopherson authored
Write the desired L2 CR3 into vmcs02.GUEST_CR3 during nested VM-Enter instead of deferring the VMWRITE until vmx_set_cr3(). If the VMWRITE is deferred, then KVM can consume a stale vmcs02.GUEST_CR3 when it refreshes vmcs12->guest_cr3 during nested_vmx_vmexit() if the emulated VM-Exit occurs without actually entering L2, e.g. if the nested run is squashed because nested VM-Enter (from L1) is putting L2 into HLT. Note, the above scenario can occur regardless of whether L1 is intercepting HLT, e.g. L1 can intercept HLT and then re-enter L2 with vmcs.GUEST_ACTIVITY_STATE=HALTED. But practically speaking, a VMM will likely put a guest into HALTED if and only if it's not intercepting HLT. In an ideal world where EPT *requires* unrestricted guest (and vice versa), VMX could handle CR3 similar to how it handles RSP and RIP, e.g. mark CR3 dirty and conditionally load it at vmx_vcpu_run(). But the unrestricted guest silliness complicates the dirty tracking logic to the point that explicitly handling vmcs02.GUEST_CR3 during nested VM-Enter is a simpler overall implementation. Cc: stable@vger.kernel.org Reported-and-tested-by:
Liran Alon <liran.alon@oracle.com> Signed-off-by:
-
Tom Lendacky authored
The SEV firmware DEACTIVATE command disassociates an SEV guest from an ASID, clears the WBINVD indicator on all threads and indicates that the SEV firmware DF_FLUSH command must be issued before the ASID can be re-used. The SEV firmware DF_FLUSH command will return an error if a WBINVD has not been performed on every thread before it has been invoked. A window exists between the WBINVD and the invocation of the DF_FLUSH command where an SEV firmware DEACTIVATE command could be invoked on another thread, clearing the WBINVD indicator. This will cause the subsequent SEV firmware DF_FLUSH command to fail which, in turn, results in the SEV firmware ACTIVATE command failing for the reclaimed ASID. This results in the SEV guest failing to start. Use a mutex to close the WBINVD/DF_FLUSH window by obtaining the mutex before the DEACTIVATE and releasing it after the DF_FLUSH. This ensures that any DEACTIVATE cannot run before a DF_FLUSH has completed. Fixes: 59414c98 ("KVM: SVM: Add support for KVM_SEV_LAUNCH_START command") Tested-by:
-
Tom Lendacky authored
The SEV ASID bitmap currently is not protected against parallel SEV guest startups. This can result in an SEV guest failing to start because another SEV guest could have been assigned the same ASID value. Use a mutex to serialize access to the SEV ASID bitmap. Fixes: 1654efcb ("KVM: SVM: Add KVM_SEV_INIT command") Tested-by:
-
Paolo Bonzini authored
Merge tag 'kvm-ppc-fixes-5.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc into HEAD PPC KVM fix for 5.4 - Fix a bug in the XIVE code which can cause a host crash.
-
Paolo Bonzini authored
Merge tag 'kvmarm-fixes-5.4-2' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD KVM/arm fixes for 5.4, take #2 Special PMU edition: - Fix cycle counter truncation - Fix cycle counter overflow limit on pure 64bit system - Allow chained events to be actually functional - Correct sample period after overflow
-
Paolo Bonzini authored
After resetting the vCPU, the kvmclock MSR keeps the previous value but it is not enabled. This can be confusing, so fix it. Signed-off-by:
-
