From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-pj1-f74.google.com (mail-pj1-f74.google.com [209.85.216.74]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id CD9AE2571B9 for ; Mon, 18 Aug 2025 15:25:36 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.74 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1755530738; cv=none; b=j3iof7vaPfN5mB/t0rySo0+8yLT8EFcOeXSl0JNKnxVAI0oZPuFTuttlYfky9V1S1E96ClWSvmqOt87lOZvI31TUbldjX1WVgZM0mgzM/bOPRDpHoC0NZZO/mmiV+301hqvQnnzgNVyPBoB6fIqp/l/LSqwetMBfDGKcJnjwMF0= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1755530738; c=relaxed/simple; bh=cQQ74RqI0oi17iisfFPsBnqCiIujEhsD/7CJ215Qn5s=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=LdtrVI762OJ2P9Wqyt3u+XM6mzAJzVAr2k4V7V+EpU8GqTUlXY0IhnsYfcwvwh4tfOlfsgldZI5w1IhoVDjQh4vdTPX2vOrgqddUZibJf1uzoFBRE41XcncpMPoCMQx1E1xsY9Wd/ruprrf6NKRQtilxAkYkTR+9oG1i29QaiOA= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--seanjc.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=a+NpuiLE; arc=none smtp.client-ip=209.85.216.74 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--seanjc.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="a+NpuiLE" Received: by mail-pj1-f74.google.com with SMTP id 98e67ed59e1d1-32326e09c5fso4262913a91.2 for ; Mon, 18 Aug 2025 08:25:36 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1755530736; x=1756135536; darn=lists.linux.dev; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=CsoYx/ZJm+dL7FRV6jyG7YgqbQ7bK3pBhdAAvFLAFxA=; b=a+NpuiLEk6Btbskx4rNZBgLp131SSEjNdXxqbYBO/3Ocwn6UcBVhCcogwd+CqLu35t DdP2ZuEEa+kAgmiPo3D50eWdobE/t/vqVQrZp9x4WDyODyd82cqg57jy0p31JbaXHAhY qsIBL/VIjDknwXjjYwhxOjpWK6SJfq2sN9iwjzg+e9t5he+5HhZd8DyagLHHefVzhNnk iUxoHoVjE+Ztf34I15I/WriUotEISMDV4/3PfleUrIZN+MfiIBjxdd44qStoITOMuyPe FXPvyMFBYkwsZn3Xe0mRVT5Ka8YPAQ25aVHby9qdqAy+AaZmu/yM6IuIFErX7V7yBeD9 JF7g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1755530736; x=1756135536; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=CsoYx/ZJm+dL7FRV6jyG7YgqbQ7bK3pBhdAAvFLAFxA=; b=uPtoPxhbgqYFrKLsydJHwF2P2iG4k6iRysHG2he8EENWIt9XSd1iqAsuwXHWs40rc9 P2znm2sqVHFGLVB457k/tOwOSEw5dzV2mtTeT8B7JXlFRXCZaD4GT5namcSkGIk1TMih GvF0dSmXdg5aqoChD264RyPdPQgbT3KRSAkp7AYM3rx8QXp47SQpOynR93dTdZ/PzYcx 6kIt76N0gMeHEgV2SGSSolVAE55s60ruv9uwUXOfC1rnKXx5YEOfFNMtkJjJwC3NttqJ Ew+9uQUNBsYTSsxm13dmav3myKDUMuEDC5Yh7Hcm1PJ+AsojvTascg+DmPnSG1VZUdI9 5CkQ== X-Forwarded-Encrypted: i=1; AJvYcCWotwSNPC3xC7P3ugiHys52HXrVkxiuB+Io5FU+l2fiEfxUWZdQtnYqEOGZY82mvrj0dX6iTSUA/0s=@lists.linux.dev X-Gm-Message-State: AOJu0Yyl9wL2FO5w89H/wT5A+XyyxClRx9lF/qJG9nFS+fLAFABanwLh 8Mkbu/4cndE7JIVw+UhJYeQk6bpLFl2z9bj3873RnBWrxylzN3DOygO6JF87nnHFCmx2r8uFzty 0JEggGQ== X-Google-Smtp-Source: AGHT+IFCmpj8FbJmK1+JWLBztlT6f28DkrWlTFsNjoSajBRjHfMiKFT/6+vAkpnb7u9ULN6Oz/FIAv/3KBU= X-Received: from pjee5.prod.google.com ([2002:a17:90b:5785:b0:311:ef56:7694]) (user=seanjc job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:52c6:b0:31e:f3b7:49c6 with SMTP id 98e67ed59e1d1-323421b64b7mr17300718a91.15.1755530736134; Mon, 18 Aug 2025 08:25:36 -0700 (PDT) Date: Mon, 18 Aug 2025 08:25:34 -0700 In-Reply-To: <20250818143204.GH3289052@noisy.programming.kicks-ass.net> Precedence: bulk X-Mailing-List: loongarch@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250806195706.1650976-1-seanjc@google.com> <20250806195706.1650976-10-seanjc@google.com> <20250815113951.GC4067720@noisy.programming.kicks-ass.net> <20250818143204.GH3289052@noisy.programming.kicks-ass.net> Message-ID: Subject: Re: [PATCH v5 09/44] perf/x86: Switch LVTPC to/from mediated PMI vector on guest load/put context From: Sean Christopherson To: Peter Zijlstra Cc: Marc Zyngier , Oliver Upton , Tianrui Zhao , Bibo Mao , Huacai Chen , Anup Patel , Paul Walmsley , Palmer Dabbelt , Albert Ou , Xin Li , "H. Peter Anvin" , Andy Lutomirski , Ingo Molnar , Arnaldo Carvalho de Melo , Namhyung Kim , Paolo Bonzini , linux-arm-kernel@lists.infradead.org, kvmarm@lists.linux.dev, kvm@vger.kernel.org, loongarch@lists.linux.dev, kvm-riscv@lists.infradead.org, linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org, Kan Liang , Yongwei Ma , Mingwei Zhang , Xiong Zhang , Sandipan Das , Dapeng Mi Content-Type: text/plain; charset="us-ascii" On Mon, Aug 18, 2025, Peter Zijlstra wrote: > On Fri, Aug 15, 2025 at 08:55:25AM -0700, Sean Christopherson wrote: > > On Fri, Aug 15, 2025, Sean Christopherson wrote: > > > On Fri, Aug 15, 2025, Peter Zijlstra wrote: > > > > > diff --git a/kernel/events/core.c b/kernel/events/core.c > > > > > index e1df3c3bfc0d..ad22b182762e 100644 > > > > > --- a/kernel/events/core.c > > > > > +++ b/kernel/events/core.c > > > > > @@ -6408,6 +6408,8 @@ void perf_load_guest_context(unsigned long data) > > > > > task_ctx_sched_out(cpuctx->task_ctx, NULL, EVENT_GUEST); > > > > > } > > > > > > > > > > + arch_perf_load_guest_context(data); > > > > > > > > So I still don't understand why this ever needs to reach the generic > > > > code. x86 pmu driver and x86 kvm can surely sort this out inside of x86, > > > > no? > > > > > > It's definitely possible to handle this entirely within x86, I just don't love > > > switching the LVTPC without the protection of perf_ctx_lock and perf_ctx_disable(). > > > It's not a sticking point for me if you strongly prefer something like this: > > > > > > diff --git a/arch/x86/kvm/pmu.c b/arch/x86/kvm/pmu.c > > > index 0e5048ae86fa..86b81c217b97 100644 > > > --- a/arch/x86/kvm/pmu.c > > > +++ b/arch/x86/kvm/pmu.c > > > @@ -1319,7 +1319,9 @@ void kvm_mediated_pmu_load(struct kvm_vcpu *vcpu) > > > > > > lockdep_assert_irqs_disabled(); > > > > > > - perf_load_guest_context(kvm_lapic_get_reg(vcpu->arch.apic, APIC_LVTPC)); > > > + perf_load_guest_context(); > > > + > > > + perf_load_guest_lvtpc(kvm_lapic_get_reg(vcpu->arch.apic, APIC_LVTPC)); > > > > Hmm, an argument for providing a dedicated perf_load_guest_lvtpc() APIs is that > > it would allow KVM to handle LVTPC writes in KVM's VM-Exit fastpath, i.e. without > > having to do a full put+reload of the guest context. > > > > So if we're confident that switching the host LVTPC outside of > > perf_{load,put}_guest_context() is functionally safe, I'm a-ok with it. > > Let me see. So the hardware sets Masked when it raises the interrupt. > > The interrupt handler clears it from software -- depending on uarch in 3 > different places: > 1) right at the start of the PMI > 2) in the middle, right before enabling the PMU (writing global control) > 3) at the end of the PMI > > the various changelogs adding that code mention spurious PMIs and > malformed PEBS records. > > So the fun all happens when the guest is doing PMI and gets a VM-exit > while still Masked. > > At that point, we can come in and completely rewrite the PMU state, > reroute the PMI and enable things again. Then later, we 'restore' the > PMU state, re-set LVTPC masked to the guest interrupt and 'resume'. > > What could possibly go wrong :/ Kan, I'm assuming, but not knowing, that > writing all the PMU MSRs is somehow serializing state sufficient to not > cause the above mentioned fails? Specifically, clearing PEBS_ENABLE > should inhibit those malformed PEBS records or something? What if the > host also has PEBS and we don't actually clear the bit? > > The current order ensures we rewrite LVTPC when global control is unset; > I think we want to keep that. Yes, for sure. > While staring at this, I note that perf_load_guest_context() will clear > global ctrl, clear all the counter programming, and re-enable an empty > pmu. Now, an empty PMU should result in global control being zero -- > there is nothing run after all. > > But then kvm_mediated_pmu_load() writes an explicit 0 again. Perhaps > replace this with asserting it is 0 instead? Yeah, I like that idea, a lot. This? perf_load_guest_context(); /* * Sanity check that "loading" guest context disabled all counters, as * modifying the LVTPC while host perf is active will cause explosions, * as will loading event selectors and PMCs with guest values. * * VMX will enable/disable counters at VM-Enter/VM-Exit by atomically * loading PERF_GLOBAL_CONTROL. SVM effectively performs the switch by * configuring all events to be GUEST_ONLY. */ WARN_ON_ONCE(rdmsrq(kvm_pmu_ops.PERF_GLOBAL_CTRL)); perf_load_guest_lvtpc(kvm_lapic_get_reg(vcpu->arch.apic, APIC_LVTPC)); > Anyway, this means that moving the LVTPC writing into > kvm_mediated_pmu_load() as you suggest is identical. > perf_load_guest_context() results in global control being 0, we then > assert it is 0, and write LVTPC while it is still 0. > kvm_pmu_load_guest_pmcs() will then frob the MSRs. > > OK, so *IF* doing the VM-exit during PMI is sound, this is something > that needs a comment somewhere. I'm a bit lost here. Are you essentially asking if it's ok to take a VM-Exit while the guest is handling a PMI? If so, that _has_ to work, because there are myriad things that can/will trigger a VM-Exit at any point while the guest is active. > Then going back again, is the easy part, since on the host side, we can never > transition into KVM during a PMI.