From: "Xu, Like" <like.xu@intel.com>
To: Sean Christopherson <seanjc@google.com>,
Like Xu <like.xu@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
Paolo Bonzini <pbonzini@redhat.com>,
eranian@google.com, kvm@vger.kernel.org,
Ingo Molnar <mingo@redhat.com>,
Thomas Gleixner <tglx@linutronix.de>,
Vitaly Kuznetsov <vkuznets@redhat.com>,
Wanpeng Li <wanpengli@tencent.com>,
Jim Mattson <jmattson@google.com>, Joerg Roedel <joro@8bytes.org>,
Andi Kleen <andi@firstfloor.org>,
Kan Liang <kan.liang@linux.intel.com>,
wei.w.wang@intel.com, luwei.kang@intel.com,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 06/17] KVM: x86/pmu: Add IA32_PEBS_ENABLE MSR emulation for extended PEBS
Date: Thu, 7 Jan 2021 20:38:40 +0800 [thread overview]
Message-ID: <1d7d366d-dc99-8bd1-d2f7-e9328827fe8f@intel.com> (raw)
In-Reply-To: <X/TV9nZw49XFwDF/@google.com>
Hi Sean,
On 2021/1/6 5:11, Sean Christopherson wrote:
> On Mon, Jan 04, 2021, Like Xu wrote:
>> If IA32_PERF_CAPABILITIES.PEBS_BASELINE [bit 14] is set, the
>> IA32_PEBS_ENABLE MSR exists and all architecturally enumerated fixed
>> and general purpose counters have corresponding bits in IA32_PEBS_ENABLE
>> that enable generation of PEBS records. The general-purpose counter bits
>> start at bit IA32_PEBS_ENABLE[0], and the fixed counter bits start at
>> bit IA32_PEBS_ENABLE[32].
>>
>> When guest PEBS is enabled, the IA32_PEBS_ENABLE MSR will be
>> added to the perf_guest_switch_msr() and switched during the
>> VMX transitions just like CORE_PERF_GLOBAL_CTRL MSR.
>>
>> Originally-by: Andi Kleen <ak@linux.intel.com>
>> Co-developed-by: Kan Liang <kan.liang@linux.intel.com>
>> Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
>> Co-developed-by: Luwei Kang <luwei.kang@intel.com>
>> Signed-off-by: Luwei Kang <luwei.kang@intel.com>
>> Signed-off-by: Like Xu <like.xu@linux.intel.com>
>> ---
>> arch/x86/events/intel/core.c | 20 ++++++++++++++++++++
>> arch/x86/include/asm/kvm_host.h | 1 +
>> arch/x86/include/asm/msr-index.h | 6 ++++++
>> arch/x86/kvm/vmx/pmu_intel.c | 28 ++++++++++++++++++++++++++++
>> 4 files changed, 55 insertions(+)
>>
>> diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
>> index af457f8cb29d..6453b8a6834a 100644
>> --- a/arch/x86/events/intel/core.c
>> +++ b/arch/x86/events/intel/core.c
>> @@ -3715,6 +3715,26 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr)
>> *nr = 2;
>> }
>>
>> + if (cpuc->pebs_enabled & ~cpuc->intel_ctrl_host_mask) {
>> + arr[1].msr = MSR_IA32_PEBS_ENABLE;
>> + arr[1].host = cpuc->pebs_enabled & ~cpuc->intel_ctrl_guest_mask;
>> + arr[1].guest = cpuc->pebs_enabled & ~cpuc->intel_ctrl_host_mask;
>> + /*
>> + * The guest PEBS will be disabled once the host PEBS is enabled
>> + * since the both enabled case may brings a unknown PMI to
>> + * confuse host and the guest PEBS overflow PMI would be missed.
>> + */
>> + if (arr[1].host)
>> + arr[1].guest = 0;
>> + arr[0].guest |= arr[1].guest;
> Can't you modify the code that strips the PEBS counters from the guest's
> value instead of poking into the array entry after the fact?
Ah, nice move.
>
> Also, why is this scenario even allowed? Can't we force exclude_guest for
> events that use PEBS?
The attr.exclude_guest can be configured for each event, and
it's not shared with other perf_events on the same CPU,
and changing it will not take effect when the event is running.
Host perf would still allow to create or run the PEBS perf_event
for vcpu when host is using PEBS counter. One reason is that the
perf scheduler needs to know which event owns which PEBS counter.
>
>> + *nr = 2;
>> + } else if (*nr == 1) {
>> + /* Remove MSR_IA32_PEBS_ENABLE from MSR switch list in KVM */
>> + arr[1].msr = MSR_IA32_PEBS_ENABLE;
>> + arr[1].host = arr[1].guest = 0;
>> + *nr = 2;
> Similar to above, rather then check "*nr == 1", this should properly integrate
> with the "x86_pmu.pebs && x86_pmu.pebs_no_isolation" logic instead of poking
> into the array after the fact.
Thanks, it makes sense to me and I'll figure it out w/ your clearly code.
>
> By incorporating both suggestions, the logic can be streamlined significantly,
> and IMO makes the overall flow much more understandable. Untested...
>
> diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
> index d4569bfa83e3..c5cc7e558c8e 100644
> --- a/arch/x86/events/intel/core.c
> +++ b/arch/x86/events/intel/core.c
> @@ -3708,24 +3708,39 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr)
> arr[0].msr = MSR_CORE_PERF_GLOBAL_CTRL;
> arr[0].host = x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_guest_mask;
> arr[0].guest = x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_host_mask;
> - if (x86_pmu.flags & PMU_FL_PEBS_ALL)
> - arr[0].guest &= ~cpuc->pebs_enabled;
> - else
> - arr[0].guest &= ~(cpuc->pebs_enabled & PEBS_COUNTER_MASK);
> +
> + /*
> + * Disable PEBS in the guest if PEBS is used by the host; enabling PEBS
> + * in both will lead to unexpected PMIs in the host and/or missed PMIs
> + * in the guest.
> + */
> + if (cpuc->pebs_enabled & ~cpuc->intel_ctrl_guest_mask) {
> + if (x86_pmu.flags & PMU_FL_PEBS_ALL)
> + arr[0].guest &= ~cpuc->pebs_enabled;
> + else
> + arr[0].guest &= ~(cpuc->pebs_enabled & PEBS_COUNTER_MASK);
> + }
> *nr = 1;
>
> - if (x86_pmu.pebs && x86_pmu.pebs_no_isolation) {
> - /*
> - * If PMU counter has PEBS enabled it is not enough to
> - * disable counter on a guest entry since PEBS memory
> - * write can overshoot guest entry and corrupt guest
> - * memory. Disabling PEBS solves the problem.
> - *
> - * Don't do this if the CPU already enforces it.
> - */
> + if (x86_pmu.pebs) {
> arr[1].msr = MSR_IA32_PEBS_ENABLE;
> - arr[1].host = cpuc->pebs_enabled;
> - arr[1].guest = 0;
> + arr[1].host = cpuc->pebs_enabled & ~cpuc->intel_ctrl_guest_mask;
> +
> + /*
> + * Host and guest PEBS are mutually exclusive. Load the guest
> + * value iff PEBS is disabled in the host. If PEBS is enabled
> + * in the host and the CPU supports PEBS isolation, disabling
> + * the counters is sufficient (see above); skip the MSR loads
s/above/9b545c04abd4/
> + * by stuffing guest=host (KVM will remove the entry). Without
> + * isolation, PEBS must be explicitly disabled prior to
> + * VM-Enter to prevent PEBS writes from overshooting VM-Enter.
> + */
> + if (!arr[1].host)
> + arr[1].guest = cpuc->pebs_enabled & ~cpuc->intel_ctrl_host_mask;
miss "arr[0].guest |= arr[1].guest;" here to make it enabled in the
global_ctrl msr.
Sean, if you have more comments on other patches, just let me know.
---
thx,likexu
> + else if (x86_pmu.pebs_no_isolation)
> + arr[1].guest = 0;
> + else
> + arr[1].guest = arr[1].host;
> *nr = 2;
> }
next prev parent reply other threads:[~2021-01-07 12:39 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-04 13:15 [PATCH v3 00/17] KVM: x86/pmu: Add support to enable Guest PEBS via DS Like Xu
2021-01-04 13:15 ` [PATCH v3 01/17] KVM: x86/pmu: Set MSR_IA32_MISC_ENABLE_EMON bit when vPMU is enabled Like Xu
2021-01-04 13:15 ` [PATCH v3 02/17] KVM: x86/pmu: Use IA32_PERF_CAPABILITIES to adjust features visibility Like Xu
2021-01-04 13:15 ` [PATCH v3 03/17] KVM: x86/pmu: Introduce the ctrl_mask value for fixed counter Like Xu
2021-01-13 18:06 ` Peter Zijlstra
2021-01-14 1:58 ` Xu, Like
2021-01-04 13:15 ` [PATCH v3 04/17] perf: x86/ds: Handle guest PEBS overflow PMI and inject it to guest Like Xu
2021-01-13 18:22 ` Peter Zijlstra
2021-01-13 18:27 ` Peter Zijlstra
2021-01-14 3:39 ` Xu, Like
2021-01-15 12:01 ` Peter Zijlstra
2021-01-15 14:30 ` Xu, Like
2021-01-15 14:44 ` Peter Zijlstra
2021-01-15 15:12 ` Xu, Like
2021-01-25 8:26 ` Like Xu
2021-01-25 11:47 ` Peter Zijlstra
2021-02-02 6:31 ` Xu, Like
2021-01-14 18:55 ` Sean Christopherson
2021-01-15 2:49 ` Xu, Like
2021-01-15 17:42 ` Sean Christopherson
2021-01-22 5:30 ` Like Xu
2021-01-04 13:15 ` [PATCH v3 05/17] KVM: x86/pmu: Reprogram guest PEBS event to emulate guest PEBS counter Like Xu
2021-01-15 11:33 ` Peter Zijlstra
2021-01-15 13:53 ` Xu, Like
2021-01-04 13:15 ` [PATCH v3 06/17] KVM: x86/pmu: Add IA32_PEBS_ENABLE MSR emulation for extended PEBS Like Xu
2021-01-05 21:11 ` Sean Christopherson
2021-01-07 12:38 ` Xu, Like [this message]
2021-01-15 14:46 ` Peter Zijlstra
2021-01-15 15:29 ` Xu, Like
2021-01-04 13:15 ` [PATCH v3 07/17] KVM: x86/pmu: Add IA32_DS_AREA MSR emulation to manage guest DS buffer Like Xu
2021-01-05 21:16 ` Sean Christopherson
2021-01-08 3:05 ` Xu, Like
2021-01-04 13:15 ` [PATCH v3 08/17] KVM: x86/pmu: Add PEBS_DATA_CFG MSR emulation to support adaptive PEBS Like Xu
2021-01-04 13:15 ` [PATCH v3 09/17] KVM: x86: Set PEBS_UNAVAIL in IA32_MISC_ENABLE when PEBS is enabled Like Xu
2021-01-04 13:15 ` [PATCH v3 10/17] KVM: x86/pmu: Expose CPUIDs feature bits PDCM, DS, DTES64 Like Xu
2021-01-04 13:15 ` [PATCH v3 11/17] KVM: x86/pmu: Adjust precise_ip to emulate Ice Lake guest PDIR counter Like Xu
2021-01-04 13:15 ` [PATCH v3 12/17] KVM: x86/pmu: Disable guest PEBS when counters are cross-mapped Like Xu
2021-01-04 13:15 ` [PATCH v3 13/17] KVM: x86/pmu: Add hook to emulate pebs for cross-mapped counters Like Xu
2021-01-04 13:15 ` [PATCH v3 14/17] KVM: vmx/pmu: Limit pebs_interrupt_threshold in the guest DS area Like Xu
2021-01-04 13:15 ` [PATCH v3 15/17] KVM: vmx/pmu: Rewrite applicable_counters field in guest PEBS records Like Xu
2021-01-04 13:15 ` [PATCH v3 16/17] KVM: x86/pmu: Save guest pebs reset values when pebs is configured Like Xu
2021-01-04 13:15 ` [PATCH v3 17/17] KVM: x86/pmu: Adjust guest pebs reset values for crpss-mapped counters Like Xu
2021-01-14 19:10 ` [PATCH v3 00/17] KVM: x86/pmu: Add support to enable Guest PEBS via DS Sean Christopherson
2021-01-15 2:02 ` Xu, Like
2021-01-15 17:57 ` Sean Christopherson
2021-01-15 18:27 ` Andi Kleen
2021-01-15 18:51 ` Sean Christopherson
2021-01-15 19:11 ` Andi Kleen
2021-01-22 9:56 ` Peter Zijlstra
2021-01-25 8:08 ` Like Xu
2021-01-25 11:13 ` Peter Zijlstra
2021-01-25 12:07 ` Xu, Like
2021-01-25 12:18 ` Peter Zijlstra
2021-01-25 12:53 ` Xu, Like
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1d7d366d-dc99-8bd1-d2f7-e9328827fe8f@intel.com \
--to=like.xu@intel.com \
--cc=andi@firstfloor.org \
--cc=eranian@google.com \
--cc=jmattson@google.com \
--cc=joro@8bytes.org \
--cc=kan.liang@linux.intel.com \
--cc=kvm@vger.kernel.org \
--cc=like.xu@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=luwei.kang@intel.com \
--cc=mingo@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peterz@infradead.org \
--cc=seanjc@google.com \
--cc=tglx@linutronix.de \
--cc=vkuznets@redhat.com \
--cc=wanpengli@tencent.com \
--cc=wei.w.wang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).