All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ben Gardon <bgardon@google.com>
To: linux-kernel@vger.kernel.org, kvm@vger.kernel.org
Cc: Paolo Bonzini <pbonzini@redhat.com>, Peter Xu <peterx@redhat.com>,
	Sean Christopherson <seanjc@google.com>,
	Peter Shier <pshier@google.com>,
	David Matlack <dmatlack@google.com>,
	Mingwei Zhang <mizhang@google.com>,
	Yulei Zhang <yulei.kernel@gmail.com>,
	Wanpeng Li <kernellwp@gmail.com>,
	Xiao Guangrong <xiaoguangrong.eric@gmail.com>,
	Kai Huang <kai.huang@intel.com>,
	Keqian Zhu <zhukeqian1@huawei.com>,
	David Hildenbrand <david@redhat.com>,
	Ben Gardon <bgardon@google.com>
Subject: [PATCH 03/15] KVM: x86/mmu: Factor wrprot for nested PML out of make_spte
Date: Mon, 15 Nov 2021 15:45:51 -0800	[thread overview]
Message-ID: <20211115234603.2908381-4-bgardon@google.com> (raw)
In-Reply-To: <20211115234603.2908381-1-bgardon@google.com>

When running a nested VM, KVM write protects SPTEs in the EPT/NPT02
instead of using PML for dirty tracking. This avoids expensive
translation later, when emptying the Page Modification Log. In service
of removing the vCPU pointer from make_spte, factor the check for nested
PML out of the function.


Signed-off-by: Ben Gardon <bgardon@google.com>
---
 arch/x86/kvm/mmu/spte.c | 10 +++++++---
 arch/x86/kvm/mmu/spte.h |  3 ++-
 2 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kvm/mmu/spte.c b/arch/x86/kvm/mmu/spte.c
index 04d26e913941..3cf08a534a16 100644
--- a/arch/x86/kvm/mmu/spte.c
+++ b/arch/x86/kvm/mmu/spte.c
@@ -92,7 +92,8 @@ static bool kvm_is_mmio_pfn(kvm_pfn_t pfn)
 bool make_spte(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
 	       struct kvm_memory_slot *slot, unsigned int pte_access,
 	       gfn_t gfn, kvm_pfn_t pfn, u64 old_spte, bool prefetch,
-	       bool can_unsync, bool host_writable, u64 *new_spte)
+	       bool can_unsync, bool host_writable, bool ad_need_write_protect,
+	       u64 *new_spte)
 {
 	int level = sp->role.level;
 	u64 spte = SPTE_MMU_PRESENT_MASK;
@@ -100,7 +101,7 @@ bool make_spte(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
 
 	if (sp->role.ad_disabled)
 		spte |= SPTE_TDP_AD_DISABLED_MASK;
-	else if (kvm_vcpu_ad_need_write_protect(vcpu))
+	else if (ad_need_write_protect)
 		spte |= SPTE_TDP_AD_WRPROT_ONLY_MASK;
 
 	/*
@@ -195,8 +196,11 @@ bool vcpu_make_spte(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
 		    gfn_t gfn, kvm_pfn_t pfn, u64 old_spte, bool prefetch,
 		    bool can_unsync, bool host_writable, u64 *new_spte)
 {
+	bool ad_need_write_protect = kvm_vcpu_ad_need_write_protect(vcpu);
+
 	return make_spte(vcpu, sp, slot, pte_access, gfn, pfn, old_spte,
-			 prefetch, can_unsync, host_writable, new_spte);
+			 prefetch, can_unsync, host_writable,
+			 ad_need_write_protect, new_spte);
 
 }
 
diff --git a/arch/x86/kvm/mmu/spte.h b/arch/x86/kvm/mmu/spte.h
index 14f18082d505..bcf58602f224 100644
--- a/arch/x86/kvm/mmu/spte.h
+++ b/arch/x86/kvm/mmu/spte.h
@@ -332,7 +332,8 @@ static inline u64 get_mmio_spte_generation(u64 spte)
 bool make_spte(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
 	       struct kvm_memory_slot *slot, unsigned int pte_access,
 	       gfn_t gfn, kvm_pfn_t pfn, u64 old_spte, bool prefetch,
-	       bool can_unsync, bool host_writable, u64 *new_spte);
+	       bool can_unsync, bool host_writable, bool ad_need_write_protect,
+	       u64 *new_spte);
 bool vcpu_make_spte(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
 		    struct kvm_memory_slot *slot,
 		    unsigned int pte_access, gfn_t gfn, kvm_pfn_t pfn,
-- 
2.34.0.rc1.387.gb447b232ab-goog


  parent reply	other threads:[~2021-11-16  3:16 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-15 23:45 [PATCH 00/15] Currently disabling dirty logging with the TDP MMU is extremely slow. On a 96 vCPU / 96G VM it takes ~45 seconds to disable dirty logging with the TDP MMU, as opposed to ~3.5 seconds with the legacy MMU. This series optimizes TLB flushes and introduces in-place large page promotion, to bring the disable dirty log time down to ~2 seconds Ben Gardon
2021-11-15 23:45 ` [PATCH 01/15] KVM: x86/mmu: Remove redundant flushes when disabling dirty logging Ben Gardon
2021-11-18  8:26   ` Paolo Bonzini
2021-11-15 23:45 ` [PATCH 02/15] KVM: x86/mmu: Introduce vcpu_make_spte Ben Gardon
2021-11-15 23:45 ` Ben Gardon [this message]
2021-11-15 23:45 ` [PATCH 04/15] KVM: x86/mmu: Factor mt_mask out of make_spte Ben Gardon
2021-11-15 23:45 ` [PATCH 05/15] KVM: x86/mmu: Remove need for a vcpu from kvm_slot_page_track_is_active Ben Gardon
2021-11-18  8:25   ` Paolo Bonzini
2021-11-15 23:45 ` [PATCH 06/15] KVM: x86/mmu: Remove need for a vcpu from mmu_try_to_unsync_pages Ben Gardon
2021-11-18  8:25   ` Paolo Bonzini
2021-11-15 23:45 ` [PATCH 07/15] KVM: x86/mmu: Factor shadow_zero_check out of make_spte Ben Gardon
2021-11-15 23:45 ` [PATCH 08/15] KVM: x86/mmu: Replace vcpu argument with kvm pointer in make_spte Ben Gardon
2021-11-15 23:45 ` [PATCH 09/15] KVM: x86/mmu: Factor out the meat of reset_tdp_shadow_zero_bits_mask Ben Gardon
2021-11-15 23:45 ` [PATCH 10/15] KVM: x86/mmu: Propagate memslot const qualifier Ben Gardon
2021-11-18  8:27   ` Paolo Bonzini
2021-11-15 23:45 ` [PATCH 11/15] KVM: x86/MMU: Refactor vmx_get_mt_mask Ben Gardon
2021-11-18  8:30   ` Paolo Bonzini
2021-11-18 15:30     ` Sean Christopherson
2021-11-19  9:02       ` Paolo Bonzini
2021-11-22 18:11         ` Ben Gardon
2021-11-22 18:46           ` Sean Christopherson
2021-11-15 23:46 ` [PATCH 12/15] KVM: x86/mmu: Factor out part of vmx_get_mt_mask which does not depend on vcpu Ben Gardon
2021-11-15 23:46 ` [PATCH 13/15] KVM: x86/mmu: Add try_get_mt_mask to x86_ops Ben Gardon
2021-11-15 23:46 ` [PATCH 14/15] KVM: x86/mmu: Make kvm_is_mmio_pfn usable outside of spte.c Ben Gardon
2021-11-15 23:46 ` [PATCH 15/15] KVM: x86/mmu: Promote pages in-place when disabling dirty logging Ben Gardon
2021-11-25  4:18   ` Peter Xu
2021-11-29 18:31     ` Ben Gardon
2021-11-30  0:13       ` Sean Christopherson
2021-11-30  7:28       ` Peter Xu
2021-11-30 16:01         ` Sean Christopherson
2021-12-01  1:59           ` Peter Xu
2021-11-15 23:58 ` [PATCH 00/15] Currently disabling dirty logging with the TDP MMU is extremely slow. On a 96 vCPU / 96G VM it takes ~45 seconds to disable dirty logging with the TDP MMU, as opposed to ~3.5 seconds with the legacy MMU. This series optimizes TLB flushes and introduces in-place large page promotion, to bring the disable dirty log time down to ~2 seconds Ben Gardon

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=20211115234603.2908381-4-bgardon@google.com \
    --to=bgardon@google.com \
    --cc=david@redhat.com \
    --cc=dmatlack@google.com \
    --cc=kai.huang@intel.com \
    --cc=kernellwp@gmail.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mizhang@google.com \
    --cc=pbonzini@redhat.com \
    --cc=peterx@redhat.com \
    --cc=pshier@google.com \
    --cc=seanjc@google.com \
    --cc=xiaoguangrong.eric@gmail.com \
    --cc=yulei.kernel@gmail.com \
    --cc=zhukeqian1@huawei.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.