All of lore.kernel.org
 help / color / mirror / Atom feed
From: Steve Rutherford <srutherford@google.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: kvm@vger.kernel.org, ahonig@google.com
Subject: Re: [RFC PATCH 2/4] KVM: x86: Add KVM exit for IOAPIC EOIs
Date: Wed, 13 May 2015 15:18:19 -0700	[thread overview]
Message-ID: <20150513221819.GB23867@google.com> (raw)
In-Reply-To: <5552FEC6.9090305@redhat.com>

On Wed, May 13, 2015 at 09:35:34AM +0200, Paolo Bonzini wrote:
> 
> 
> On 13/05/2015 03:47, Steve Rutherford wrote:
> > Adds KVM_EXIT_IOAPIC_EOI which passes the interrupt vector up to
> > userspace.
> > 
> > Uses a per VCPU exit bitmap to decide whether or not the IOAPIC needs
> > to be informed (which is identical to the EOI_EXIT_BITMAP field used
> > by modern x86 processors, but can also be used to elide kvm IOAPIC EOI
> > exits on older processors).
> > 
> > [Note: A prototype using ResampleFDs found that decoupling the EOI
> > from the VCPU's thread made it possible for the VCPU to not see a
> > recent EOI after reentering the guest. This does not match real
> > hardware.]
> 
> Can you explain this better (and perhaps add a testcase to ioapic.c)?

This is for passing EOIs to the IOAPIC for level-triggered IOAPIC interrupts. Note that this patch requires some other patch to set the bitmaps. 

The sequential level-triggered interrupts test in ioapic.c effectively tests one side of this (If an EOI fails to be delivered to the IOAPIC, then any subsequent interrupts on that EOI's vector will be coalesced, and the second interrupt in the test won't fire.). 
The other side, checking that the guest /only/ exits on the specified vectors, is harder to instrument from within the KVM unit tests.

Steve

> 
> Paolo
> 
> > Compile tested for Intel x86.
> > 
> > Signed-off-by: Steve Rutherford <srutherford@google.com>
> > ---
> >  Documentation/virtual/kvm/api.txt | 10 ++++++++++
> >  arch/x86/include/asm/kvm_host.h   |  3 +++
> >  arch/x86/kvm/lapic.c              |  9 +++++++++
> >  arch/x86/kvm/x86.c                | 11 +++++++++++
> >  include/linux/kvm_host.h          |  1 +
> >  include/uapi/linux/kvm.h          |  5 +++++
> >  6 files changed, 39 insertions(+)
> > 
> > diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
> > index 0744b4e..dd92996 100644
> > --- a/Documentation/virtual/kvm/api.txt
> > +++ b/Documentation/virtual/kvm/api.txt
> > @@ -3285,6 +3285,16 @@ Valid values for 'type' are:
> >  	 */
> >  	__u64 kvm_valid_regs;
> >  	__u64 kvm_dirty_regs;
> > +
> > +	/* KVM_EXIT_IOAPIC_EOI */
> > +        struct {
> > +	       __u8 vector;
> > +        } eoi;
> > +
> > +Indicates that an eoi of a level triggered IOAPIC interrupt on vector has
> > +occurred, which should be handled by the userspace IOAPIC. Triggers when
> > +the Irqchip has been split between userspace and the kernel.
> > +
> >  	union {
> >  		struct kvm_sync_regs regs;
> >  		char padding[1024];
> > diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
> > index 3ddc134..b1978f1 100644
> > --- a/arch/x86/include/asm/kvm_host.h
> > +++ b/arch/x86/include/asm/kvm_host.h
> > @@ -539,6 +539,9 @@ struct kvm_vcpu_arch {
> >  	struct {
> >  		bool pv_unhalted;
> >  	} pv;
> > +
> > +	u64 eoi_exit_bitmaps[4];
> > +	int pending_ioapic_eoi;
> >  };
> >  
> >  struct kvm_lpage_info {
> > diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> > index bc392a6..42fada6f 100644
> > --- a/arch/x86/kvm/lapic.c
> > +++ b/arch/x86/kvm/lapic.c
> > @@ -860,6 +860,15 @@ int kvm_apic_compare_prio(struct kvm_vcpu *vcpu1, struct kvm_vcpu *vcpu2)
> >  
> >  static void kvm_ioapic_send_eoi(struct kvm_lapic *apic, int vector)
> >  {
> > +	if (irqchip_split(apic->vcpu->kvm)) {
> > +		if (test_bit(vector,
> > +			     (void *) apic->vcpu->arch.eoi_exit_bitmaps)) {
> > +			apic->vcpu->arch.pending_ioapic_eoi = vector;
> > +			kvm_make_request(KVM_REQ_IOAPIC_EOI_EXIT, apic->vcpu);
> > +		}
> > +		return;
> > +	}
> > +
> >  	if (kvm_ioapic_handles_vector(apic->vcpu->kvm, vector)) {
> >  		int trigger_mode;
> >  		if (apic_test_vector(vector, apic->regs + APIC_TMR))
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index 7505b39..cc27c35 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -6324,6 +6324,17 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
> >  			kvm_handle_pmu_event(vcpu);
> >  		if (kvm_check_request(KVM_REQ_PMI, vcpu))
> >  			kvm_deliver_pmi(vcpu);
> > +		if (kvm_check_request(KVM_REQ_IOAPIC_EOI_EXIT, vcpu)) {
> > +			BUG_ON(vcpu->arch.pending_ioapic_eoi > 255);
> > +			if (test_bit(vcpu->arch.pending_ioapic_eoi,
> > +				     (void *) vcpu->arch.eoi_exit_bitmaps)) {
> > +				vcpu->run->exit_reason = KVM_EXIT_IOAPIC_EOI;
> > +				vcpu->run->eoi.vector =
> > +						vcpu->arch.pending_ioapic_eoi;
> > +				r = 0;
> > +				goto out;
> > +			}
> > +		}
> >  		if (kvm_check_request(KVM_REQ_SCAN_IOAPIC, vcpu))
> >  			vcpu_scan_ioapic(vcpu);
> >  		if (kvm_check_request(KVM_REQ_APIC_PAGE_RELOAD, vcpu))
> > diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
> > index 277b7a1..cef20ad 100644
> > --- a/include/linux/kvm_host.h
> > +++ b/include/linux/kvm_host.h
> > @@ -134,6 +134,7 @@ static inline bool is_error_page(struct page *page)
> >  #define KVM_REQ_ENABLE_IBS        23
> >  #define KVM_REQ_DISABLE_IBS       24
> >  #define KVM_REQ_APIC_PAGE_RELOAD  25
> > +#define KVM_REQ_IOAPIC_EOI_EXIT   26
> >  
> >  #define KVM_USERSPACE_IRQ_SOURCE_ID		0
> >  #define KVM_IRQFD_RESAMPLE_IRQ_SOURCE_ID	1
> > diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
> > index 7d06dc4..2305948 100644
> > --- a/include/uapi/linux/kvm.h
> > +++ b/include/uapi/linux/kvm.h
> > @@ -183,6 +183,7 @@ struct kvm_s390_skeys {
> >  #define KVM_EXIT_EPR              23
> >  #define KVM_EXIT_SYSTEM_EVENT     24
> >  #define KVM_EXIT_S390_STSI        25
> > +#define KVM_EXIT_IOAPIC_EOI       26
> >  
> >  /* For KVM_EXIT_INTERNAL_ERROR */
> >  /* Emulate instruction failed. */
> > @@ -329,6 +330,10 @@ struct kvm_run {
> >  			__u8 sel1;
> >  			__u16 sel2;
> >  		} s390_stsi;
> > +		/* KVM_EXIT_IOAPIC_EOI */
> > +		struct {
> > +			__u8 vector;
> > +		} eoi;
> >  		/* Fix the size of the union. */
> >  		char padding[256];
> >  	};
> > 

  reply	other threads:[~2015-05-13 22:18 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-05-13  1:47 [RFC PATCH 1/4] KVM: x86: Split the APIC from the rest of IRQCHIP Steve Rutherford
2015-05-13  1:47 ` [RFC PATCH 2/4] KVM: x86: Add KVM exit for IOAPIC EOIs Steve Rutherford
2015-05-13  7:35   ` Paolo Bonzini
2015-05-13 22:18     ` Steve Rutherford [this message]
2015-05-24 16:46   ` Avi Kivity
2015-05-27  2:06     ` Steve Rutherford
2015-05-27  5:32       ` Avi Kivity
2015-05-28 21:58         ` Steve Rutherford
2015-05-13  1:47 ` [RFC PATCH 3/4] KVM: x86: Add EOI exit bitmap inference Steve Rutherford
2015-05-13  6:12   ` Jan Kiszka
2015-05-13  8:04     ` Paolo Bonzini
2015-05-13  8:10       ` Jan Kiszka
2015-05-13  9:24         ` Paolo Bonzini
2015-05-13 10:25           ` Jan Kiszka
2015-05-13 13:04             ` Paolo Bonzini
2015-05-13 13:19               ` Jan Kiszka
2015-05-13 22:21       ` Steve Rutherford
2015-05-15  2:38       ` Steve Rutherford
2015-05-13  7:51   ` Paolo Bonzini
2015-05-13 22:24     ` Steve Rutherford
2015-05-14  9:20       ` Paolo Bonzini
2015-05-14 15:23         ` Alex Williamson
2015-05-14 15:46           ` Paolo Bonzini
2015-05-14 16:04             ` Alex Williamson
2015-05-14 22:10               ` Steve Rutherford
2015-05-14 22:35                 ` Alex Williamson
2015-05-14 23:21                   ` Steve Rutherford
2015-05-13  1:47 ` [RFC PATCH 4/4] KVM: x86: Add support for local interrupt requests from userspace Steve Rutherford
2015-05-13  6:12   ` Jan Kiszka
2015-05-13 22:41     ` Steve Rutherford
2015-05-15 13:27       ` Jan Kiszka
2015-05-13  7:22   ` Paolo Bonzini
2015-05-13 23:13     ` Steve Rutherford
2015-05-13  7:57 ` [RFC PATCH 1/4] KVM: x86: Split the APIC from the rest of IRQCHIP Paolo Bonzini
2015-05-13 22:10   ` Steve Rutherford
2015-05-14  9:12     ` Wu, Feng
2015-05-14 19:29       ` Andrew Honig
2015-05-15  1:28         ` Wu, Feng
2015-05-15  5:03         ` Wanpeng Li
2015-05-15 18:10           ` Steve Rutherford
2015-05-18  2:11             ` Wanpeng Li

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=20150513221819.GB23867@google.com \
    --to=srutherford@google.com \
    --cc=ahonig@google.com \
    --cc=kvm@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.