linux-kselftest.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Yang Zhong <yang.zhong@intel.com>
To: x86@kernel.org, kvm@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
	linux-kselftest@vger.kernel.org, tglx@linutronix.de,
	mingo@redhat.com, bp@alien8.de, dave.hansen@linux.intel.com,
	pbonzini@redhat.com, corbet@lwn.net, shuah@kernel.org
Cc: seanjc@google.com, jun.nakajima@intel.com, kevin.tian@intel.com,
	jing2.liu@linux.intel.com, jing2.liu@intel.com,
	guang.zeng@intel.com, wei.w.wang@intel.com, yang.zhong@intel.com
Subject: [PATCH v4 21/21] kvm: x86: Disable interception for IA32_XFD on demand
Date: Wed, 29 Dec 2021 05:13:28 -0800	[thread overview]
Message-ID: <20211229131328.12283-22-yang.zhong@intel.com> (raw)
In-Reply-To: <20211229131328.12283-1-yang.zhong@intel.com>

From: Kevin Tian <kevin.tian@intel.com>

Always intercepting IA32_XFD causes non-negligible overhead when this
register is updated frequently in the guest.

Disable r/w emulation after intercepting the first WRMSR(IA32_XFD)
with a non-zero value.

Disable WRMSR emulation implies that IA32_XFD becomes out-of-sync
with the software states in fpstate and the per-cpu xfd cache. Call
fpu_sync_guest_vmexit_xfd_state() to bring them back in-sync, before
preemption is enabled.

Always trap #NM once write interception is disabled for IA32_XFD.
The #NM exception is rare if the guest doesn't use dynamic features.
Otherwise, there is at most one exception per guest task given a
dynamic feature.

p.s. We have confirmed that SDM is being revised to say that
when setting IA32_XFD[18] the AMX register state is not guaranteed
to be preserved. This clarification avoids adding mess for a creative
guest which sets IA32_XFD[18]=1 before saving active AMX state to
its own storage.

Signed-off-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jing Liu <jing2.liu@intel.com>
Signed-off-by: Yang Zhong <yang.zhong@intel.com>
---
 arch/x86/include/asm/kvm_host.h |  1 +
 arch/x86/kvm/vmx/vmx.c          | 23 ++++++++++++++++++-----
 arch/x86/kvm/vmx/vmx.h          |  2 +-
 arch/x86/kvm/x86.c              |  3 +++
 4 files changed, 23 insertions(+), 6 deletions(-)

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 555f4de47ef2..a372dfe6f407 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -640,6 +640,7 @@ struct kvm_vcpu_arch {
 	u64 smi_count;
 	bool tpr_access_reporting;
 	bool xsaves_enabled;
+	bool xfd_no_write_intercept;
 	u64 ia32_xss;
 	u64 microcode_version;
 	u64 arch_capabilities;
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c
index 638665b3e241..21c4d7409433 100644
--- a/arch/x86/kvm/vmx/vmx.c
+++ b/arch/x86/kvm/vmx/vmx.c
@@ -162,6 +162,7 @@ static u32 vmx_possible_passthrough_msrs[MAX_POSSIBLE_PASSTHROUGH_MSRS] = {
 	MSR_FS_BASE,
 	MSR_GS_BASE,
 	MSR_KERNEL_GS_BASE,
+	MSR_IA32_XFD,
 	MSR_IA32_XFD_ERR,
 #endif
 	MSR_IA32_SYSENTER_CS,
@@ -766,10 +767,11 @@ void vmx_update_exception_bitmap(struct kvm_vcpu *vcpu)
 	}
 
 	/*
-	 * Trap #NM if guest xfd contains a non-zero value so guest XFD_ERR
-	 * can be saved timely.
+	 * Disabling xfd interception indicates that dynamically-enabled
+	 * features might be used in the guest. Always trap #NM in this case
+	 * for proper virtualization of guest xfd_err.
 	 */
-	if (vcpu->arch.guest_fpu.fpstate->xfd)
+	if (vcpu->arch.xfd_no_write_intercept)
 		eb |= (1u << NM_VECTOR);
 
 	vmcs_write32(EXCEPTION_BITMAP, eb);
@@ -1971,9 +1973,20 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
 		break;
 	case MSR_IA32_XFD:
 		ret = kvm_set_msr_common(vcpu, msr_info);
-		/* Update #NM interception according to guest xfd */
-		if (!ret)
+		/*
+		 * Always intercepting WRMSR could incur non-negligible
+		 * overhead given xfd might be touched in guest context
+		 * switch. Disable write interception upon the first write
+		 * with a non-zero value (indicate potential guest usage
+		 * on dynamic xfeatures). Also update exception bitmap
+		 * to trap #NM for proper virtualization of guest xfd_err.
+		 */
+		if (!ret && data) {
+			vmx_disable_intercept_for_msr(vcpu, MSR_IA32_XFD,
+						      MSR_TYPE_RW);
+			vcpu->arch.xfd_no_write_intercept = true;
 			vmx_update_exception_bitmap(vcpu);
+		}
 		break;
 #endif
 	case MSR_IA32_SYSENTER_CS:
diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h
index bf9d3051cd6c..0a00242a91e7 100644
--- a/arch/x86/kvm/vmx/vmx.h
+++ b/arch/x86/kvm/vmx/vmx.h
@@ -340,7 +340,7 @@ struct vcpu_vmx {
 	struct lbr_desc lbr_desc;
 
 	/* Save desired MSR intercept (read: pass-through) state */
-#define MAX_POSSIBLE_PASSTHROUGH_MSRS	14
+#define MAX_POSSIBLE_PASSTHROUGH_MSRS	15
 	struct {
 		DECLARE_BITMAP(read, MAX_POSSIBLE_PASSTHROUGH_MSRS);
 		DECLARE_BITMAP(write, MAX_POSSIBLE_PASSTHROUGH_MSRS);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 76e1941db223..4990d6de5359 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -10027,6 +10027,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 	if (vcpu->arch.guest_fpu.xfd_err)
 		wrmsrl(MSR_IA32_XFD_ERR, 0);
 
+	if (vcpu->arch.xfd_no_write_intercept)
+		fpu_sync_guest_vmexit_xfd_state();
+
 	/*
 	 * Consume any pending interrupts, including the possible source of
 	 * VM-Exit on SVM and any ticks that occur between VM-Exit and now.

  parent reply	other threads:[~2021-12-29 13:15 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-29 13:13 [PATCH v4 00/21] AMX Support in KVM Yang Zhong
2021-12-29 13:13 ` [PATCH v4 01/21] x86/fpu: Extend fpu_xstate_prctl() with guest permissions Yang Zhong
2021-12-29 13:13 ` [PATCH v4 02/21] x86/fpu: Prepare guest FPU for dynamically enabled FPU features Yang Zhong
2021-12-29 13:13 ` [PATCH v4 03/21] kvm: x86: Fix xstate_required_size() to follow XSTATE alignment rule Yang Zhong
2022-01-04 19:54   ` Sean Christopherson
2021-12-29 13:13 ` [PATCH v4 04/21] kvm: x86: Exclude unpermitted xfeatures at KVM_GET_SUPPORTED_CPUID Yang Zhong
2021-12-29 13:13 ` [PATCH v4 05/21] x86/fpu: Make XFD initialization in __fpstate_reset() a function argument Yang Zhong
2021-12-29 13:13 ` [PATCH v4 06/21] x86/fpu: Add guest support to xfd_enable_feature() Yang Zhong
2021-12-29 13:13 ` [PATCH v4 07/21] x86/fpu: Provide fpu_update_guest_perm_features() for guest Yang Zhong
2021-12-29 13:13 ` [PATCH v4 08/21] kvm: x86: Check and enable permitted dynamic xfeatures at KVM_SET_CPUID2 Yang Zhong
2021-12-29 16:55   ` Sean Christopherson
2021-12-30  2:28     ` Tian, Kevin
2021-12-29 13:13 ` [PATCH v4 09/21] x86/fpu: Provide fpu_update_guest_xfd() for IA32_XFD emulation Yang Zhong
2021-12-29 13:13 ` [PATCH v4 10/21] kvm: x86: Add emulation for IA32_XFD Yang Zhong
2022-01-04 19:32   ` Sean Christopherson
2022-01-05  0:22     ` Tian, Kevin
2021-12-29 13:13 ` [PATCH v4 11/21] x86/fpu: Prepare xfd_err in struct fpu_guest Yang Zhong
2021-12-29 13:13 ` [PATCH v4 12/21] kvm: x86: Intercept #NM for saving IA32_XFD_ERR Yang Zhong
2022-01-04 20:01   ` Sean Christopherson
2022-01-05  0:27     ` Tian, Kevin
2021-12-29 13:13 ` [PATCH v4 13/21] kvm: x86: Emulate IA32_XFD_ERR for guest Yang Zhong
2021-12-29 13:13 ` [PATCH v4 14/21] kvm: x86: Disable RDMSR interception of IA32_XFD_ERR Yang Zhong
2022-01-04 19:34   ` Sean Christopherson
2022-01-05  0:23     ` Tian, Kevin
2021-12-29 13:13 ` [PATCH v4 15/21] kvm: x86: Add XCR0 support for Intel AMX Yang Zhong
2021-12-29 13:13 ` [PATCH v4 16/21] kvm: x86: Add CPUID " Yang Zhong
2021-12-29 13:13 ` [PATCH v4 17/21] x86/fpu: Add uabi_size to guest_fpu Yang Zhong
2021-12-29 13:13 ` [PATCH v4 18/21] kvm: x86: Add support for getting/setting expanded xstate buffer Yang Zhong
2022-01-04 19:46   ` Sean Christopherson
2022-01-04 20:45     ` Paolo Bonzini
2022-01-05  4:03       ` Tian, Kevin
2021-12-29 13:13 ` [PATCH v4 19/21] kvm: selftests: Add support for KVM_CAP_XSAVE2 Yang Zhong
2021-12-29 13:13 ` [PATCH v4 20/21] x86/fpu: Provide fpu_sync_guest_vmexit_xfd_state() Yang Zhong
2021-12-29 13:13 ` Yang Zhong [this message]
2021-12-30  8:46 ` [PATCH v4 00/21] AMX Support in KVM Tian, Kevin
2022-01-04 18:36 ` Paolo Bonzini
2022-01-04 18:54   ` Sean Christopherson
2022-01-05  0:10     ` Tian, Kevin

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=20211229131328.12283-22-yang.zhong@intel.com \
    --to=yang.zhong@intel.com \
    --cc=bp@alien8.de \
    --cc=corbet@lwn.net \
    --cc=dave.hansen@linux.intel.com \
    --cc=guang.zeng@intel.com \
    --cc=jing2.liu@intel.com \
    --cc=jing2.liu@linux.intel.com \
    --cc=jun.nakajima@intel.com \
    --cc=kevin.tian@intel.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=seanjc@google.com \
    --cc=shuah@kernel.org \
    --cc=tglx@linutronix.de \
    --cc=wei.w.wang@intel.com \
    --cc=x86@kernel.org \
    /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;
as well as URLs for NNTP newsgroup(s).