public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Vitaly Kuznetsov <vkuznets@redhat.com>
To: Sean Christopherson <seanjc@google.com>
Cc: kvm@vger.kernel.org, Paolo Bonzini <pbonzini@redhat.com>,
	Maxim Levitsky <mlevitsk@redhat.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 10/14] KVM: x86: Make Hyper-V emulation optional
Date: Thu, 30 Nov 2023 16:11:43 +0100	[thread overview]
Message-ID: <87y1efmg28.fsf@redhat.com> (raw)
In-Reply-To: <ZWfl3ahamXPPoIGB@google.com>

Sean Christopherson <seanjc@google.com> writes:

...

>
>>  static bool nested_get_vmcs12_pages(struct kvm_vcpu *vcpu)
>>  {
>> @@ -3552,11 +3563,13 @@ static int nested_vmx_run(struct kvm_vcpu *vcpu, bool launch)
>>  	if (!nested_vmx_check_permission(vcpu))
>>  		return 1;
>>  
>> +#ifdef CONFIG_KVM_HYPERV
>>  	evmptrld_status = nested_vmx_handle_enlightened_vmptrld(vcpu, launch);
>>  	if (evmptrld_status == EVMPTRLD_ERROR) {
>>  		kvm_queue_exception(vcpu, UD_VECTOR);
>>  		return 1;
>>  	}
>> +#endif
>>  
>>  	kvm_pmu_trigger_event(vcpu, PERF_COUNT_HW_BRANCH_INSTRUCTIONS);
>
> This fails to build with CONFIG_KVM_HYPERV=n && CONFIG_KVM_WERROR=y:
>
> arch/x86/kvm/vmx/nested.c:3577:9: error: variable 'evmptrld_status' is uninitialized when used here [-Werror,-Wuninitialized]
>         if (CC(evmptrld_status == EVMPTRLD_VMFAIL))
>                ^~~~~~~~~~~~~~~
>
> Sadly, simply wrapping with an #ifdef also fails because then evmptrld_status
> becomes unused.  I'd really prefer to avoid having to tag it __maybe_unused, and
> adding more #ifdef would also be ugly.  Any ideas?

A couple,

- we can try putting all eVMPTR logic under 'if (1)' just to create a
  block where we can define evmptrld_status. Not sure this is nicer than
  another #ifdef wrapping evmptrld_status and I'm not sure what to do
  with kvm_pmu_trigger_event() -- can it just go above
  nested_vmx_handle_enlightened_vmptrld()?

- we can add a helper, e.g. 'evmptr_is_vmfail()' and make it just return
  'false' when !CONFIG_KVM_HYPERV.

- rewrite this as a switch to avoid the need for having the local
  variable, (untested)

diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c
index c5ec0ef51ff7..b26ce7d596e9 100644
--- a/arch/x86/kvm/vmx/nested.c
+++ b/arch/x86/kvm/vmx/nested.c
@@ -3553,22 +3553,23 @@ static int nested_vmx_run(struct kvm_vcpu *vcpu, bool launch)
        enum nvmx_vmentry_status status;
        struct vcpu_vmx *vmx = to_vmx(vcpu);
        u32 interrupt_shadow = vmx_get_interrupt_shadow(vcpu);
-       enum nested_evmptrld_status evmptrld_status;
 
        if (!nested_vmx_check_permission(vcpu))
                return 1;
 
-       evmptrld_status = nested_vmx_handle_enlightened_vmptrld(vcpu, launch);
-       if (evmptrld_status == EVMPTRLD_ERROR) {
+       switch (nested_vmx_handle_enlightened_vmptrld(vcpu, launch)) {
+       case EVMPTRLD_ERROR:
                kvm_queue_exception(vcpu, UD_VECTOR);
                return 1;
+       case EVMPTRLD_VMFAIL:
+               trace_kvm_nested_vmenter_failed("evmptrld_status", 0);
+               return nested_vmx_failInvalid(vcpu);
+       default:
+               break;
        }
 
        kvm_pmu_trigger_event(vcpu, PERF_COUNT_HW_BRANCH_INSTRUCTIONS);
 
-       if (CC(evmptrld_status == EVMPTRLD_VMFAIL))
-               return nested_vmx_failInvalid(vcpu);
-
        if (CC(!evmptr_is_valid(vmx->nested.hv_evmcs_vmptr) &&
               vmx->nested.current_vmptr == INVALID_GPA))
                return nested_vmx_failInvalid(vcpu);

Unfortunately, I had to open code CC() ;-(

Or maybe just another "#ifdef" is not so ugly after all? :-)

-- 
Vitaly


  reply	other threads:[~2023-11-30 15:11 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-25 15:23 [PATCH 00/14] KVM: x86: Make Hyper-V emulation optional Vitaly Kuznetsov
2023-10-25 15:23 ` [PATCH 01/14] KVM: x86: xen: Remove unneeded xen context from struct kvm_arch when !CONFIG_KVM_XEN Vitaly Kuznetsov
2023-10-25 15:23 ` [PATCH 02/14] KVM: x86: hyper-v: Move Hyper-V partition assist page out of Hyper-V emulation context Vitaly Kuznetsov
2023-11-07 18:21   ` Maxim Levitsky
2023-10-25 15:23 ` [PATCH 03/14] KVM: VMX: Split off vmx_onhyperv.{ch} from hyperv.{ch} Vitaly Kuznetsov
2023-10-25 15:23 ` [PATCH 04/14] KVM: x86: hyper-v: Introduce kvm_hv_synic_auto_eoi_set() Vitaly Kuznetsov
2023-10-25 15:23 ` [PATCH 05/14] KVM: x86: hyper-v: Introduce kvm_hv_synic_has_vector() Vitaly Kuznetsov
2023-10-25 15:23 ` [PATCH 06/14] KVM: VMX: Split off hyperv_evmcs.{ch} Vitaly Kuznetsov
2023-10-25 15:23 ` [PATCH 07/14] KVM: x86: hyper-v: Introduce kvm_hv_nested_transtion_tlb_flush() helper Vitaly Kuznetsov
2023-11-07 18:21   ` Maxim Levitsky
2023-10-25 15:24 ` [PATCH 08/14] KVM: selftests: Make all Hyper-V tests explicitly dependent on Hyper-V emulation support in KVM Vitaly Kuznetsov
2023-11-07 18:23   ` Maxim Levitsky
2023-10-25 15:24 ` [PATCH 09/14] KVM: selftests: Fix vmxon_pa == vmcs12_pa == -1ull vmx_set_nested_state_test for !eVMCS case Vitaly Kuznetsov
2023-11-07 18:23   ` Maxim Levitsky
2023-10-25 15:24 ` [PATCH 10/14] KVM: x86: Make Hyper-V emulation optional Vitaly Kuznetsov
2023-11-07 18:25   ` Maxim Levitsky
2023-11-13 13:51     ` Vitaly Kuznetsov
2023-11-30  1:31   ` Sean Christopherson
2023-11-30 15:11     ` Vitaly Kuznetsov [this message]
2023-11-30 18:28       ` Sean Christopherson
2023-11-30  1:38   ` Sean Christopherson
2023-10-25 15:24 ` [PATCH 11/14] KVM: nVMX: hyper-v: Introduce nested_vmx_evmptr12() and nested_vmx_is_evmptr12_valid() helpers Vitaly Kuznetsov
2023-11-07 18:26   ` Maxim Levitsky
2023-11-30  1:24   ` Sean Christopherson
2023-11-30 19:13   ` Sean Christopherson
2023-10-25 15:24 ` [PATCH 12/14] KVM: nVMX: hyper-v: Introduce nested_vmx_evmcs() accessor Vitaly Kuznetsov
2023-10-25 15:24 ` [PATCH 13/14] KVM: nVMX: hyper-v: Hide more stuff under CONFIG_KVM_HYPERV Vitaly Kuznetsov
2023-10-25 15:24 ` [PATCH 14/14] KVM: nSVM: hyper-v: Hide more stuff under CONFIG_KVM_HYPERV/CONFIG_HYPERV Vitaly Kuznetsov
2023-11-30  1:42 ` [PATCH 00/14] KVM: x86: Make Hyper-V emulation optional 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=87y1efmg28.fsf@redhat.com \
    --to=vkuznets@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mlevitsk@redhat.com \
    --cc=pbonzini@redhat.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