From: Sean Christopherson <seanjc@google.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
maciej.szmigiero@oracle.com
Subject: Re: [PATCH v2 4/8] KVM: x86: do not go through ctxt->ops when emulating rsm
Date: Fri, 14 Oct 2022 21:04:18 +0000 [thread overview]
Message-ID: <Y0nO0quQnVFQruPM@google.com> (raw)
In-Reply-To: <20220929172016.319443-5-pbonzini@redhat.com>
On Thu, Sep 29, 2022, Paolo Bonzini wrote:
> @@ -520,14 +505,14 @@ int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
> u64 smbase;
> int ret;
>
> - smbase = ctxt->ops->get_smbase(ctxt);
> + smbase = vcpu->arch.smbase;
>
> - ret = ctxt->ops->read_phys(ctxt, smbase + 0xfe00, buf, sizeof(buf));
> - if (ret != X86EMUL_CONTINUE)
> + ret = kvm_vcpu_read_guest(vcpu, smbase + 0xfe00, buf, sizeof(buf));
> + if (ret < 0)
> return X86EMUL_UNHANDLEABLE;
>
> - if ((ctxt->ops->get_hflags(ctxt) & X86EMUL_SMM_INSIDE_NMI_MASK) == 0)
> - ctxt->ops->set_nmi_mask(ctxt, false);
> + if ((vcpu->arch.hflags & HF_SMM_INSIDE_NMI_MASK) == 0)
> + static_call(kvm_x86_set_nmi_mask)(vcpu, false);
>
> kvm_smm_changed(vcpu, false);
>
> @@ -535,41 +520,41 @@ int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
> * Get back to real mode, to prepare a safe state in which to load
> * CR0/CR3/CR4/EFER. It's all a bit more complicated if the vCPU
> * supports long mode.
> - *
> - * The ctxt->ops callbacks will handle all side effects when writing
> - * writing MSRs and CRs, e.g. MMU context resets, CPUID
> - * runtime updates, etc.
> */
> - if (emulator_has_longmode(ctxt)) {
> - struct desc_struct cs_desc;
> +#ifdef CONFIG_X86_64
> + if (guest_cpuid_has(vcpu, X86_FEATURE_LM)) {
To fix the "unused 'efer'" issue and avoid multiple guest_cpuid_has() calls, this
as fixup? It's not like we care about the code footprint for 32-bit KVM if the
compiler isn't clever enough to optimize away the dead code.
---
arch/x86/kvm/smm.c | 12 +++++-------
1 file changed, 5 insertions(+), 7 deletions(-)
diff --git a/arch/x86/kvm/smm.c b/arch/x86/kvm/smm.c
index 41ca128478fc..740fca1cf3a3 100644
--- a/arch/x86/kvm/smm.c
+++ b/arch/x86/kvm/smm.c
@@ -500,6 +500,8 @@ static int rsm_load_state_64(struct x86_emulate_ctxt *ctxt,
int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
{
struct kvm_vcpu *vcpu = ctxt->vcpu;
+ bool is_64bit_vcpu = IS_ENABLED(CONFIG_X86_64) &&
+ guest_cpuid_has(vcpu, X86_FEATURE_LM);
unsigned long cr0, cr4, efer;
char buf[512];
u64 smbase;
@@ -521,8 +523,7 @@ int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
* CR0/CR3/CR4/EFER. It's all a bit more complicated if the vCPU
* supports long mode.
*/
-#ifdef CONFIG_X86_64
- if (guest_cpuid_has(vcpu, X86_FEATURE_LM)) {
+ if (is_64bit_vcpu) {
struct kvm_segment cs_desc;
/* Zero CR4.PCIDE before CR0.PG. */
@@ -536,15 +537,13 @@ int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
cs_desc.s = cs_desc.g = cs_desc.present = 1;
kvm_set_segment(vcpu, &cs_desc, VCPU_SREG_CS);
}
-#endif
/* For the 64-bit case, this will clear EFER.LMA. */
cr0 = kvm_read_cr0(vcpu);
if (cr0 & X86_CR0_PE)
kvm_set_cr0(vcpu, cr0 & ~(X86_CR0_PG | X86_CR0_PE));
-#ifdef CONFIG_X86_64
- if (guest_cpuid_has(vcpu, X86_FEATURE_LM)) {
+ if (is_64bit_vcpu) {
/* Clear CR4.PAE before clearing EFER.LME. */
cr4 = kvm_read_cr4(vcpu);
if (cr4 & X86_CR4_PAE)
@@ -554,7 +553,6 @@ int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
efer = 0;
kvm_set_msr(vcpu, MSR_EFER, efer);
}
-#endif
/*
* Give leave_smm() a chance to make ISA-specific changes to the vCPU
@@ -565,7 +563,7 @@ int emulator_leave_smm(struct x86_emulate_ctxt *ctxt)
return X86EMUL_UNHANDLEABLE;
#ifdef CONFIG_X86_64
- if (guest_cpuid_has(vcpu, X86_FEATURE_LM))
+ if (is_64bit_vcpu)
return rsm_load_state_64(ctxt, buf);
else
#endif
base-commit: 8b86d27cc60a150252b04989de818ad4ec85f899
--
next prev parent reply other threads:[~2022-10-14 21:04 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-29 17:20 [PATCH v2 0/8] KVM: x86: allow compiling out SMM support Paolo Bonzini
2022-09-29 17:20 ` [PATCH v2 1/8] KVM: x86: start moving SMM-related functions to new files Paolo Bonzini
2022-10-14 19:38 ` Sean Christopherson
2022-10-24 12:31 ` Maxim Levitsky
2022-09-29 17:20 ` [PATCH v2 2/8] KVM: x86: move SMM entry to a new file Paolo Bonzini
2022-10-24 12:32 ` Maxim Levitsky
2022-09-29 17:20 ` [PATCH v2 3/8] KVM: x86: move SMM exit " Paolo Bonzini
2022-10-24 12:32 ` Maxim Levitsky
2022-09-29 17:20 ` [PATCH v2 4/8] KVM: x86: do not go through ctxt->ops when emulating rsm Paolo Bonzini
2022-10-01 13:10 ` kernel test robot
2022-10-01 13:41 ` kernel test robot
2022-10-02 2:42 ` kernel test robot
2022-10-14 21:04 ` Sean Christopherson [this message]
2022-10-24 12:32 ` Maxim Levitsky
2022-09-29 17:20 ` [PATCH v2 5/8] KVM: allow compiling out SMM support Paolo Bonzini
2022-10-24 12:32 ` Maxim Levitsky
2022-09-29 17:20 ` [PATCH v2 6/8] KVM: x86: compile out vendor-specific code if SMM is disabled Paolo Bonzini
2022-10-24 12:32 ` Maxim Levitsky
2022-09-29 17:20 ` [PATCH v2 7/8] KVM: x86: remove SMRAM address space if SMM is not supported Paolo Bonzini
2022-10-24 12:33 ` Maxim Levitsky
2022-09-29 17:20 ` [PATCH v2 8/8] KVM: x86: do not define KVM_REQ_SMI if SMM disabled Paolo Bonzini
2022-10-14 21:06 ` Sean Christopherson
2022-10-24 12:33 ` Maxim Levitsky
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=Y0nO0quQnVFQruPM@google.com \
--to=seanjc@google.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=maciej.szmigiero@oracle.com \
--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;
as well as URLs for NNTP newsgroup(s).