public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Sean Christopherson <seanjc@google.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: linux-kernel@vger.kernel.org, kvm@vger.kernel.org, dmatlack@google.com
Subject: Re: [PATCH v2 01/25] KVM: x86/mmu: avoid indirect call for get_cr3
Date: Tue, 8 Mar 2022 16:16:31 +0000	[thread overview]
Message-ID: <YieBXzkOkB9SZpyp@google.com> (raw)
In-Reply-To: <20220221162243.683208-2-pbonzini@redhat.com>

On Mon, Feb 21, 2022, Paolo Bonzini wrote:
> Most of the time, calls to get_guest_pgd result in calling kvm_read_cr3
> (the exception is only nested TDP).  Check if that is the case if
> retpolines are enabled, thus avoiding an expensive indirect call.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  arch/x86/kvm/mmu.h             | 10 ++++++++++
>  arch/x86/kvm/mmu/mmu.c         | 15 ++++++++-------
>  arch/x86/kvm/mmu/paging_tmpl.h |  2 +-
>  arch/x86/kvm/x86.c             |  2 +-
>  4 files changed, 20 insertions(+), 9 deletions(-)
> 
> diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h
> index 1d0c1904d69a..6ee4436e46f1 100644
> --- a/arch/x86/kvm/mmu.h
> +++ b/arch/x86/kvm/mmu.h
> @@ -116,6 +116,16 @@ static inline void kvm_mmu_load_pgd(struct kvm_vcpu *vcpu)
>  					  vcpu->arch.mmu->shadow_root_level);
>  }
>  
> +extern unsigned long kvm_get_guest_cr3(struct kvm_vcpu *vcpu);

No extern please, it's superfluous and against KVM style.  Moot point though, see
below.

> +static inline unsigned long kvm_mmu_get_guest_pgd(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu)

Wrap the params, no reason to make this line so long.

> +{
> +#ifdef CONFIG_RETPOLINE
> +	if (mmu->get_guest_pgd == kvm_get_guest_cr3)
> +		return kvm_read_cr3(vcpu);

This is unnecessarily fragile and confusing at first glance.  Compilers are smart
enough to generate a non-inline version of functions if they're used for function
pointers, while still inlining where appropriate.  In other words, just drop
kvm_get_guest_cr3() entirely, a al get_pdptr => kvm_pdptr_read().

---
 arch/x86/kvm/mmu.h     |  6 +++---
 arch/x86/kvm/mmu/mmu.c | 11 +++--------
 2 files changed, 6 insertions(+), 11 deletions(-)

diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h
index 3af66b9df640..50528d39de8d 100644
--- a/arch/x86/kvm/mmu.h
+++ b/arch/x86/kvm/mmu.h
@@ -117,11 +117,11 @@ static inline void kvm_mmu_load_pgd(struct kvm_vcpu *vcpu)
 					  vcpu->arch.mmu->shadow_root_level);
 }

-extern unsigned long kvm_get_guest_cr3(struct kvm_vcpu *vcpu);
-static inline unsigned long kvm_mmu_get_guest_pgd(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu)
+static inline unsigned long kvm_mmu_get_guest_pgd(struct kvm_vcpu *vcpu,
+						  struct kvm_mmu *mmu)
 {
 #ifdef CONFIG_RETPOLINE
-	if (mmu->get_guest_pgd == kvm_get_guest_cr3)
+	if (mmu->get_guest_pgd == kvm_read_cr3)
 		return kvm_read_cr3(vcpu);
 #endif
 	return mmu->get_guest_pgd(vcpu);
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 995c3450c20f..cc2414397e4b 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -4234,11 +4234,6 @@ void kvm_mmu_new_pgd(struct kvm_vcpu *vcpu, gpa_t new_pgd)
 }
 EXPORT_SYMBOL_GPL(kvm_mmu_new_pgd);

-unsigned long kvm_get_guest_cr3(struct kvm_vcpu *vcpu)
-{
-	return kvm_read_cr3(vcpu);
-}
-
 static bool sync_mmio_spte(struct kvm_vcpu *vcpu, u64 *sptep, gfn_t gfn,
 			   unsigned int access)
 {
@@ -4793,7 +4788,7 @@ static void init_kvm_tdp_mmu(struct kvm_vcpu *vcpu)
 	context->invlpg = NULL;
 	context->shadow_root_level = kvm_mmu_get_tdp_level(vcpu);
 	context->direct_map = true;
-	context->get_guest_pgd = kvm_get_guest_cr3;
+	context->get_guest_pgd = kvm_read_cr3;
 	context->get_pdptr = kvm_pdptr_read;
 	context->inject_page_fault = kvm_inject_page_fault;
 	context->root_level = role_regs_to_root_level(&regs);
@@ -4968,7 +4963,7 @@ static void init_kvm_softmmu(struct kvm_vcpu *vcpu)

 	kvm_init_shadow_mmu(vcpu, &regs);

-	context->get_guest_pgd	   = kvm_get_guest_cr3;
+	context->get_guest_pgd	   = kvm_read_cr3;
 	context->get_pdptr         = kvm_pdptr_read;
 	context->inject_page_fault = kvm_inject_page_fault;
 }
@@ -5000,7 +4995,7 @@ static void init_kvm_nested_mmu(struct kvm_vcpu *vcpu)
 		return;

 	g_context->mmu_role.as_u64 = new_role.as_u64;
-	g_context->get_guest_pgd     = kvm_get_guest_cr3;
+	g_context->get_guest_pgd     = kvm_read_cr3;
 	g_context->get_pdptr         = kvm_pdptr_read;
 	g_context->inject_page_fault = kvm_inject_page_fault;
 	g_context->root_level        = new_role.base.level;

base-commit: c31df3e63672c14d8b52e34606c823e2166024b8
--

  reply	other threads:[~2022-03-08 16:16 UTC|newest]

Thread overview: 66+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-21 16:22 [PATCH v2 00/25] KVM MMU refactoring part 2: role changes Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 01/25] KVM: x86/mmu: avoid indirect call for get_cr3 Paolo Bonzini
2022-03-08 16:16   ` Sean Christopherson [this message]
2022-03-08 16:21     ` Paolo Bonzini
2022-03-08 16:32       ` Sean Christopherson
2022-03-08 16:43         ` Paolo Bonzini
2022-03-08 16:53           ` Sean Christopherson
2022-03-08 17:14             ` Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 02/25] KVM: x86/mmu: nested EPT cannot be used in SMM Paolo Bonzini
2022-03-08 16:18   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 03/25] KVM: x86/mmu: constify uses of struct kvm_mmu_role_regs Paolo Bonzini
2022-03-08 16:22   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 04/25] KVM: x86/mmu: pull computation of kvm_mmu_role_regs to kvm_init_mmu Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 05/25] KVM: x86/mmu: rephrase unclear comment Paolo Bonzini
2022-03-08 16:39   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 06/25] KVM: nVMX/nSVM: do not monkey-patch inject_page_fault callback Paolo Bonzini
2022-03-08 17:13   ` Sean Christopherson
2022-03-08 20:34     ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 07/25] KVM: x86/mmu: remove "bool base_only" arguments Paolo Bonzini
2022-03-08 17:15   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 08/25] KVM: x86/mmu: split cpu_mode from mmu_role Paolo Bonzini
2022-03-08 17:36   ` Sean Christopherson
2022-03-08 17:49     ` Paolo Bonzini
2022-03-08 18:55   ` Sean Christopherson
2022-03-09  9:58     ` Paolo Bonzini
2022-03-09 15:38       ` Sean Christopherson
2022-03-09 15:40         ` Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 09/25] KVM: x86/mmu: do not recompute root level from kvm_mmu_role_regs Paolo Bonzini
2022-03-08 17:41   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 10/25] KVM: x86/mmu: remove ept_ad field Paolo Bonzini
2022-03-08 17:42   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 11/25] KVM: x86/mmu: remove kvm_calc_shadow_root_page_role_common Paolo Bonzini
2022-03-08 17:48   ` Sean Christopherson
2022-03-08 17:50     ` Paolo Bonzini
2022-03-08 18:17       ` Sean Christopherson
2022-03-08 18:18         ` Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 12/25] KVM: x86/mmu: cleanup computation of MMU roles for two-dimensional paging Paolo Bonzini
2022-03-08 18:11   ` Sean Christopherson
2022-03-08 18:24     ` Paolo Bonzini
2022-03-08 18:44       ` Sean Christopherson
2022-03-08 18:38     ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 13/25] KVM: x86/mmu: cleanup computation of MMU roles for shadow paging Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 14/25] KVM: x86/mmu: store shadow EFER.NX in the MMU role Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 15/25] KVM: x86/mmu: remove extended bits from mmu_role, rename field Paolo Bonzini
2022-03-08 19:02   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 16/25] KVM: x86/mmu: rename kvm_mmu_role union Paolo Bonzini
2022-03-08 19:15   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 17/25] KVM: x86/mmu: remove redundant bits from extended role Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 18/25] KVM: x86/mmu: remove valid " Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 19/25] KVM: x86/mmu: simplify and/or inline computation of shadow MMU roles Paolo Bonzini
2022-03-08 19:35   ` Sean Christopherson
2022-03-08 19:41     ` Sean Christopherson
2022-03-09 10:33     ` Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 20/25] KVM: x86/mmu: pull CPU mode computation to kvm_init_mmu Paolo Bonzini
2022-03-08 19:45   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 21/25] KVM: x86/mmu: replace shadow_root_level with root_role.level Paolo Bonzini
2022-03-08 19:48   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 22/25] KVM: x86/mmu: replace root_level with cpu_mode.base.level Paolo Bonzini
2022-03-08 19:49   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 23/25] KVM: x86/mmu: replace direct_map with root_role.direct Paolo Bonzini
2022-03-08 19:52   ` Sean Christopherson
2022-02-21 16:22 ` [PATCH v2 24/25] KVM: x86/mmu: initialize constant-value fields just once Paolo Bonzini
2022-03-08 20:58   ` Sean Christopherson
2022-03-09 10:34     ` Paolo Bonzini
2022-02-21 16:22 ` [PATCH v2 25/25] KVM: x86/mmu: extract initialization of the page walking data Paolo Bonzini
2022-03-08 20:02   ` Sean Christopherson

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=YieBXzkOkB9SZpyp@google.com \
    --to=seanjc@google.com \
    --cc=dmatlack@google.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pbonzini@redhat.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