Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220117085307.93030-1-likexu@tencent.com> <CALMp9eS=1U7T39L-vL_cTXTNN2Li8epjtAPoP_+Hwefe9d+teQ@mail.gmail.com>
 <67a731dd-53ba-0eb8-377f-9707e5c9be1b@intel.com> <CABOYuvbPL0DeEgV4gsC+v786xfBAo3T6+7XQr7cVVzbaoFoEAg@mail.gmail.com>
 <7b5012d8-6ae1-7cde-a381-e82685dfed4f@linux.intel.com> <CALMp9eTOaWxQPfdwMSAn-OYAHKPLcuCyse7BpsSOM35vg5d0Jg@mail.gmail.com>
 <e06db1a5-1b67-28ac-ee4c-34ece5857b1f@linux.intel.com> <CALMp9eSjDro169JjTXyCZn=Rf3PT0uHhdNXEifiXGYQK-Zn8LA@mail.gmail.com>
 <d86ba87b-d98a-53a0-b2cd-5bf77b97b592@linux.intel.com> <CABOYuvZ9SZAWeRkrhhhpMM4XwzMzXv9A1WDpc6z8SUBquf0SFQ@mail.gmail.com>
 <6afcec02-fb44-7b72-e527-6517a94855d4@linux.intel.com> <CALMp9eQ79Cnh1aqNGLR8n5MQuHTwuf=DMjJ2cTb9uXu94GGhEA@mail.gmail.com>
 <2180ea93-5f05-b1c1-7253-e3707da29f8c@linux.intel.com> <CALMp9eQiaXtF3S0QZ=2_SWavFnv6zFHqf_zFXBrxXb9pVYh0nQ@mail.gmail.com>
 <8d9149b5-e56f-b397-1527-9f21a26ad95b@linux.intel.com> <CALMp9eTqNyG-ke1Aq72hn0CVXER63SgVPamzXria76PmqiZvJQ@mail.gmail.com>
 <31cd6e81-fd74-daa1-8518-8a8dfc6174d0@gmail.com>
In-Reply-To: <31cd6e81-fd74-daa1-8518-8a8dfc6174d0@gmail.com>
From:   Jim Mattson <jmattson@google.com>
Date:   Wed, 16 Feb 2022 10:10:05 -0800
Message-ID: <CALMp9eTrKmtP4BHxBNvyG9+bhOAd1jofx0rQz0rF=MtaoShb=w@mail.gmail.com>
Subject: Re: [PATCH kvm/queue v2 2/3] perf: x86/core: Add interface to query
 perfmon_event_map[] directly
To:     Like Xu <like.xu.linux@gmail.com>
Cc:     "Liang, Kan" <kan.liang@linux.intel.com>,
        David Dunn <daviddunn@google.com>,
        Dave Hansen <dave.hansen@intel.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Paolo Bonzini <pbonzini@redhat.com>,
        Sean Christopherson <seanjc@google.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Joerg Roedel <joro@8bytes.org>, kvm@vger.kernel.org,
        linux-kernel@vger.kernel.org, Like Xu <likexu@tencent.com>,
        Stephane Eranian <eranian@google.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Tue, Feb 15, 2022 at 11:36 PM Like Xu <like.xu.linux@gmail.com> wrote:

> Hardware resources will always be limited (the good news is that
> the number of counters will increase in the future), and when we have
> a common need for the same hardware, we should prioritize overall
> fairness over a GUEST-first strategy.

The bad news about the new counters on Intel is that they are less
capable than the existing counters. SInce there is no bitmask for GP
counters in CPUID.0AH, if you give 4 counters to the guest, then the
host is stuck using the less capable counters.

Overall fairness is definitely not what we want. See the other thread,
which has a proposal for a more configurable perf subsystem that
should meet your needs as well as ours.

Can we move all of this discussion to the other thread, BTW?