mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Yan Zhao <yan.y.zhao@intel.com>
To: "Edgecombe, Rick P" <rick.p.edgecombe@intel.com>
Cc: "pbonzini@redhat.com" <pbonzini@redhat.com>,
	"seanjc@google.com" <seanjc@google.com>,
	"Huang, Kai" <kai.huang@intel.com>,
	"sagis@google.com" <sagis@google.com>,
	"isaku.yamahata@gmail.com" <isaku.yamahata@gmail.com>,
	"Aktas, Erdem" <erdemaktas@google.com>,
	"kvm@vger.kernel.org" <kvm@vger.kernel.org>,
	"dmatlack@google.com" <dmatlack@google.com>,
	"Yamahata, Isaku" <isaku.yamahata@intel.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v3 13/17] KVM: x86/tdp_mmu: Support mirror root for TDP MMU
Date: Wed, 26 Jun 2024 13:05:02 +0800	[thread overview]
Message-ID: <ZnuhfnLH+m3cV2/U@yzhao56-desk.sh.intel.com> (raw)
In-Reply-To: <70dab5f4fb69c072493efe4b3198305ae262b545.camel@intel.com>

On Wed, Jun 26, 2024 at 04:33:20AM +0800, Edgecombe, Rick P wrote:
> On Tue, 2024-06-25 at 13:43 +0800, Yan Zhao wrote:
> > > > > I was originally suspicious of the asymmetry of the tear down of mirror
> > > > > and
> > > > > direct roots vs the allocation. Do you see a concrete problem, or just
> > > > > advocating for safety?
> > > IMO it's a concrete problem, though rare up to now. e.g.
> > > 
> > > After repeatedly hot-plugping and hot-unplugping memory, which increases
> > > memslots generation, kvm_mmu_zap_all_fast() will be called to invalidate >
> > > direct
> > > roots when the memslots generation wraps around.
> 
> Hmm, yes. I'm not sure about putting the check there though. It adds even more
> confusion to the lifecycle.
>  - mirror_root_hpa != INVALID_PAGE check in a different placed than
>    root.hpa != INVALID_PAGE check.
>  - they get allocated in the same place
>  - they are torn down in the different places.
> 
> Can you think of clearer fix for it. Maybe we can just move the mirror root
> allocation such that it's not subjected to the reload path? Like something that
> matches the tear down in kvm_mmu_destroy()?
But we still need the reload path to have each vcpu to hold a ref count of the
mirror root.
What about below fix?

diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h
index 026e8edfb0bd..4decd13457ec 100644
--- a/arch/x86/kvm/mmu.h
+++ b/arch/x86/kvm/mmu.h
@@ -127,9 +127,28 @@ void kvm_mmu_sync_prev_roots(struct kvm_vcpu *vcpu);
 void kvm_mmu_track_write(struct kvm_vcpu *vcpu, gpa_t gpa, const u8 *new,
                         int bytes);

+static inline bool kvm_has_mirrored_tdp(const struct kvm *kvm)
+{
+       return kvm->arch.vm_type == KVM_X86_TDX_VM;
+}
+
+static inline bool kvm_mmu_root_hpa_is_invalid(struct kvm_vcpu *vcpu)
+{
+       return vcpu->arch.mmu->root.hpa == INVALID_PAGE;
+}
+
+static inline bool kvm_mmu_mirror_root_hpa_is_invalid(struct kvm_vcpu *vcpu)
+{
+       if (!kvm_has_mirrored_tdp(vcpu->kvm))
+               return false;
+
+       return vcpu->arch.mmu->mirror_root_hpa == INVALID_PAGE;
+}
+
 static inline int kvm_mmu_reload(struct kvm_vcpu *vcpu)
 {
-       if (likely(vcpu->arch.mmu->root.hpa != INVALID_PAGE))
+       if (!kvm_mmu_root_hpa_is_invalid(vcpu) &&
+           !kvm_mmu_mirror_root_hpa_is_invalid(vcpu))
                return 0;

        return kvm_mmu_load(vcpu);
@@ -322,11 +341,6 @@ static inline gpa_t kvm_translate_gpa(struct kvm_vcpu *vcpu,
        return translate_nested_gpa(vcpu, gpa, access, exception);
 }

-static inline bool kvm_has_mirrored_tdp(const struct kvm *kvm)
-{
-       return kvm->arch.vm_type == KVM_X86_TDX_VM;
-}
-
 static inline gfn_t kvm_gfn_direct_bits(const struct kvm *kvm)
 {
        return kvm->arch.gfn_direct_bits;
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index e1299eb03e63..5e7d92074f70 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -3702,9 +3702,11 @@ static int mmu_alloc_direct_roots(struct kvm_vcpu *vcpu)
        int r;

        if (tdp_mmu_enabled) {
-               if (kvm_has_mirrored_tdp(vcpu->kvm))
+               if (kvm_mmu_mirror_root_hpa_is_invalid(vcpu))
                        kvm_tdp_mmu_alloc_root(vcpu, true);
-               kvm_tdp_mmu_alloc_root(vcpu, false);
+
+               if (kvm_mmu_root_hpa_is_invalid(vcpu))
+                       kvm_tdp_mmu_alloc_root(vcpu, false);
                return 0;
        }



  reply	other threads:[~2024-06-26  5:06 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-06-19 22:35 [PATCH v3 00/17] TDX MMU prep series part 1 Rick Edgecombe
2024-06-19 22:35 ` [PATCH v3 01/17] KVM: x86/tdp_mmu: Rename REMOVED_SPTE to FROZEN_SPTE Rick Edgecombe
2024-06-19 22:35 ` [PATCH v3 02/17] KVM: Add member to struct kvm_gfn_range for target alias Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 03/17] KVM: x86: Add a VM type define for TDX Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 04/17] KVM: x86/mmu: Add an external pointer to struct kvm_mmu_page Rick Edgecombe
2024-07-03  7:03   ` Yan Zhao
2024-06-19 22:36 ` [PATCH v3 05/17] KVM: x86/mmu: Add an is_mirror member for union kvm_mmu_page_role Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 06/17] KVM: x86/mmu: Make kvm_tdp_mmu_alloc_root() return void Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 07/17] KVM: x86/tdp_mmu: Take struct kvm in iter loops Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 08/17] KVM: x86/tdp_mmu: Take a GFN in kvm_tdp_mmu_fast_pf_get_last_sptep() Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 09/17] KVM: x86/mmu: Support GFN direct bits Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 10/17] KVM: x86/tdp_mmu: Extract root invalid check from tdx_mmu_next_root() Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 11/17] KVM: x86/tdp_mmu: Introduce KVM MMU root types to specify page table type Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 12/17] KVM: x86/tdp_mmu: Take root in tdp_mmu_for_each_pte() Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 13/17] KVM: x86/tdp_mmu: Support mirror root for TDP MMU Rick Edgecombe
2024-06-24  8:30   ` Yan Zhao
2024-06-25  0:51     ` Edgecombe, Rick P
2024-06-25  5:43       ` Yan Zhao
2024-06-25 20:33         ` Edgecombe, Rick P
2024-06-26  5:05           ` Yan Zhao [this message]
2024-07-03 19:40             ` Edgecombe, Rick P
2024-07-04  8:09               ` Yan Zhao
2024-07-09 22:36                 ` Edgecombe, Rick P
2024-07-04  8:51   ` Yan Zhao
2024-07-09 22:38     ` Edgecombe, Rick P
2024-07-11 23:54       ` Edgecombe, Rick P
2024-07-12  1:42         ` Yan Zhao
2024-06-19 22:36 ` [PATCH v3 14/17] KVM: x86/tdp_mmu: Propagate attr_filter to MMU notifier callbacks Rick Edgecombe
2024-06-19 22:36 ` [PATCH v3 15/17] KVM: x86/tdp_mmu: Propagate building mirror page tables Rick Edgecombe
2024-06-20  5:15   ` Binbin Wu
2024-06-24 23:52     ` Edgecombe, Rick P
2024-06-19 22:36 ` [PATCH v3 16/17] KVM: x86/tdp_mmu: Propagate tearing down " Rick Edgecombe
2024-06-20  8:44   ` Binbin Wu
2024-06-24 23:55     ` Edgecombe, Rick P
2024-06-19 22:36 ` [PATCH v3 17/17] KVM: x86/tdp_mmu: Take root types for kvm_tdp_mmu_invalidate_all_roots() Rick Edgecombe
2024-06-21  7:10   ` Yan Zhao
2024-06-21 19:08     ` Edgecombe, Rick P
2024-06-24  8:29       ` Yan Zhao
2024-06-24 23:15         ` Edgecombe, Rick P
2024-06-25  6:14           ` Yan Zhao
2024-06-25 20:56             ` Edgecombe, Rick P
2024-06-26  2:25               ` Yan Zhao
2024-07-03 20:00                 ` Edgecombe, Rick P
2024-07-05  1:16                   ` Yan Zhao
2024-07-09 22:52                     ` Edgecombe, Rick P
2024-07-18 15:28       ` Isaku Yamahata
2024-07-18 15:55         ` Edgecombe, Rick P

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=ZnuhfnLH+m3cV2/U@yzhao56-desk.sh.intel.com \
    --to=yan.y.zhao@intel.com \
    --cc=dmatlack@google.com \
    --cc=erdemaktas@google.com \
    --cc=isaku.yamahata@gmail.com \
    --cc=isaku.yamahata@intel.com \
    --cc=kai.huang@intel.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pbonzini@redhat.com \
    --cc=rick.p.edgecombe@intel.com \
    --cc=sagis@google.com \
    --cc=seanjc@google.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

all inboxes | Powered by JetHome®