linux-kernel - [PATCH v4 00/10] Guest LBR Enabling

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <1545816338-1171-1-git-send-email-wei.w.wang@intel.com>
Date:   Wed, 26 Dec 2018 17:25:28 +0800
From:   Wei Wang <wei.w.wang@...el.com>
To:     linux-kernel@...r.kernel.org, kvm@...r.kernel.org,
        pbonzini@...hat.com, ak@...ux.intel.com, peterz@...radead.org
Cc:     kan.liang@...el.com, mingo@...hat.com, rkrcmar@...hat.com,
        like.xu@...el.com, wei.w.wang@...el.com, jannh@...gle.com,
        arei.gonglei@...wei.com
Subject: [PATCH v4 00/10] Guest LBR Enabling

Last Branch Recording (LBR) is a performance monitor unit (PMU) feature
on Intel CPUs that captures branch related info. This patch series enables
this feature to KVM guests.

Here is a conclusion of the fundamental methods that we use:
1) the LBR feature is enabled per guest via QEMU setting of
   KVM_CAP_X86_GUEST_LBR;
2) the LBR stack is passed through to the guest for direct accesses after
   the guest's first access to any of the lbr related MSRs;
3) When the guest uses the LBR feature with the user callstack mode, the
   host will help save/resotre the LBR stack when the vCPU is scheduled
   out/in.

ChangeLog:
v3->v4:
    - perf/x86:
        - Patch 1: change the lbr msr variable type to "unsigned int";
        - Patch 6: add a "no_counter" boolean attr to let callers
          explicitly tell the perf core to not allocate counters for the
          event;
        - Patch 7: change "cpuc->vcpu_lbr" from "boolean" to "u8", and add
          the PF_VCPU and is_kernel_event checks befoer setting it;
        - Patch 8: add the LBR_SELECT save/restore on vCPU switching;
	- Patch 9: move lbr_select_user_callstack to .h, instead of
          exporting it;

    - KVM/x86:
        - Patch 3: use "cap->args[0]" to enable/disable the lbr feature
          from userspace;
        - Patch 4: forbid the enabling the guest lbr when the host and
          guest see different lbr stack entries;
        - Patch 10: in guest_access_lbr_msr, pass through the lbr stack
          only when it has been passed through (i.e. add the check
          !vcpu->arch.lbr_used).
previous:
https://lkml.org/lkml/2018/9/20/507

Like Xu (1):
  KVM/x86/vPMU: Add APIs to support host save/restore the guest lbr
    stack

Wei Wang (9):
  perf/x86: fix the variable type of the LBR MSRs
  perf/x86: add a function to get the lbr stack
  KVM/x86: KVM_CAP_X86_GUEST_LBR
  KVM/x86: intel_pmu_lbr_enable
  KVM/x86: expose MSR_IA32_PERF_CAPABILITIES to the guest
  perf/x86: no counter allocation support
  perf/x86: save/restore LBR_SELECT on vCPU switching
  perf/x86: function to check lbr user callstack mode
  KVM/x86/lbr: lazy save the guest lbr stack

 arch/x86/events/core.c            |  12 +++
 arch/x86/events/intel/lbr.c       |  82 ++++++++---------
 arch/x86/events/perf_event.h      |   6 +-
 arch/x86/include/asm/kvm_host.h   |   7 ++
 arch/x86/include/asm/perf_event.h |  64 +++++++++++++
 arch/x86/kvm/cpuid.c              |   2 +-
 arch/x86/kvm/cpuid.h              |   8 ++
 arch/x86/kvm/pmu.c                |   8 ++
 arch/x86/kvm/pmu.h                |  10 +++
 arch/x86/kvm/pmu_intel.c          | 183 ++++++++++++++++++++++++++++++++++++++
 arch/x86/kvm/vmx.c                | 147 ++++++++++++++++++++++++++++++
 arch/x86/kvm/x86.c                |  14 +++
 include/linux/perf_event.h        |  12 +++
 include/uapi/linux/kvm.h          |   1 +
 include/uapi/linux/perf_event.h   |   3 +-
 kernel/events/core.c              |   7 --
 16 files changed, 514 insertions(+), 52 deletions(-)

-- 
2.7.4