* [PULL] KVM/ARM fixes for v4.4-rc6
@ 2015-12-18 13:00 ` Marc Zyngier
0 siblings, 0 replies; 6+ messages in thread
From: Marc Zyngier @ 2015-12-18 13:00 UTC (permalink / raw)
To: Paolo Bonzini, Gleb Natapov
Cc: Christoffer Dall, kvm, kvmarm, linux-arm-kernel
Hi Paolo,
We have a one line fix for the VGIC this time around, fixing a patch
that went in -rc2. Oh well. Hopefully this is the last one for v4.4.
Please pull!
Thanks,
M.
The following changes since commit 0de58f852875a0f0dcfb120bb8433e4e73c7803b:
ARM/arm64: KVM: correct PTE uncachedness check (2015-12-04 16:30:17 +0000)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm.git tags/kvm-arm-for-v4.4-rc6
for you to fetch changes up to fdec12c12ed4333afb49c9948c29fbd5fb52da97:
KVM: arm/arm64: vgic: Fix kvm_vgic_map_is_active's dist check (2015-12-11 16:33:31 +0000)
----------------------------------------------------------------
KVM/ARM fixes for v4.4-rc6
- Fix for the active interrupt detection code, affecting
the timer interrupt injection.
----------------------------------------------------------------
Christoffer Dall (1):
KVM: arm/arm64: vgic: Fix kvm_vgic_map_is_active's dist check
virt/kvm/arm/vgic.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
^ permalink raw reply [flat|nested] 6+ messages in thread* [PULL] KVM/ARM fixes for v4.4-rc6
@ 2015-12-18 13:00 ` Marc Zyngier
0 siblings, 0 replies; 6+ messages in thread
From: Marc Zyngier @ 2015-12-18 13:00 UTC (permalink / raw)
To: linux-arm-kernel
Hi Paolo,
We have a one line fix for the VGIC this time around, fixing a patch
that went in -rc2. Oh well. Hopefully this is the last one for v4.4.
Please pull!
Thanks,
M.
The following changes since commit 0de58f852875a0f0dcfb120bb8433e4e73c7803b:
ARM/arm64: KVM: correct PTE uncachedness check (2015-12-04 16:30:17 +0000)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm.git tags/kvm-arm-for-v4.4-rc6
for you to fetch changes up to fdec12c12ed4333afb49c9948c29fbd5fb52da97:
KVM: arm/arm64: vgic: Fix kvm_vgic_map_is_active's dist check (2015-12-11 16:33:31 +0000)
----------------------------------------------------------------
KVM/ARM fixes for v4.4-rc6
- Fix for the active interrupt detection code, affecting
the timer interrupt injection.
----------------------------------------------------------------
Christoffer Dall (1):
KVM: arm/arm64: vgic: Fix kvm_vgic_map_is_active's dist check
virt/kvm/arm/vgic.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
^ permalink raw reply [flat|nested] 6+ messages in thread* [PATCH] KVM: arm/arm64: vgic-new: Try to insert multi-source SGIs at once
2015-12-18 13:00 ` Marc Zyngier
@ 2015-12-18 13:00 ` Marc Zyngier
-1 siblings, 0 replies; 6+ messages in thread
From: Marc Zyngier @ 2015-12-18 13:00 UTC (permalink / raw)
To: Paolo Bonzini, Gleb Natapov
Cc: Christoffer Dall, kvm, kvmarm, linux-arm-kernel
Improve handling of multi-source SGIs: instead of only inserting
one source per SGI per run, try to insert them all at once.
Hopefully this is a rare case.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic.c | 30 ++++++++++++++++++++++++++----
1 file changed, 26 insertions(+), 4 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index c6f04a1..052f917 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -371,15 +371,22 @@ static int compute_ap_list_depth(struct kvm_vcpu *vcpu)
struct vgic_irq *irq;
int count = 0;
- list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list)
- count++;
-
+ list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list) {
+ spin_lock(&irq->irq_lock);
+ /* GICv2 SGIs can count for more than one... */
+ if (irq->intid < VGIC_NR_SGIS && irq->source)
+ count += hweight8(irq->source);
+ else
+ count++;
+ spin_unlock(&irq->irq_lock);
+ }
return count;
}
static void vgic_populate_lrs(struct kvm_vcpu *vcpu)
{
struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+ u32 model = vcpu->kvm->arch.vgic.vgic_model;
struct vgic_irq *irq;
int count = 0;
@@ -390,8 +397,23 @@ static void vgic_populate_lrs(struct kvm_vcpu *vcpu)
list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list) {
spin_lock(&irq->irq_lock);
- if (vgic_target_oracle(irq) == vcpu)
+
+ if (unlikely(vgic_target_oracle(irq) != vcpu))
+ goto next;
+
+ /*
+ * If we get an SGI with multiple sources, try to get
+ * them in all at once.
+ */
+ if (model == KVM_DEV_TYPE_ARM_VGIC_V2 &&
+ irq->intid < VGIC_NR_SGIS) {
+ while(irq->source && count < vcpu->arch.vgic_cpu.nr_lr)
+ vgic_populate_lr(vcpu, irq, count++);
+ } else {
vgic_populate_lr(vcpu, irq, count++);
+ }
+
+ next:
spin_unlock(&irq->irq_lock);
if (count == vcpu->arch.vgic_cpu.nr_lr)
--
2.1.4
^ permalink raw reply related [flat|nested] 6+ messages in thread* [PATCH] KVM: arm/arm64: vgic-new: Try to insert multi-source SGIs at once
@ 2015-12-18 13:00 ` Marc Zyngier
0 siblings, 0 replies; 6+ messages in thread
From: Marc Zyngier @ 2015-12-18 13:00 UTC (permalink / raw)
To: linux-arm-kernel
Improve handling of multi-source SGIs: instead of only inserting
one source per SGI per run, try to insert them all at once.
Hopefully this is a rare case.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
---
virt/kvm/arm/vgic/vgic.c | 30 ++++++++++++++++++++++++++----
1 file changed, 26 insertions(+), 4 deletions(-)
diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c
index c6f04a1..052f917 100644
--- a/virt/kvm/arm/vgic/vgic.c
+++ b/virt/kvm/arm/vgic/vgic.c
@@ -371,15 +371,22 @@ static int compute_ap_list_depth(struct kvm_vcpu *vcpu)
struct vgic_irq *irq;
int count = 0;
- list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list)
- count++;
-
+ list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list) {
+ spin_lock(&irq->irq_lock);
+ /* GICv2 SGIs can count for more than one... */
+ if (irq->intid < VGIC_NR_SGIS && irq->source)
+ count += hweight8(irq->source);
+ else
+ count++;
+ spin_unlock(&irq->irq_lock);
+ }
return count;
}
static void vgic_populate_lrs(struct kvm_vcpu *vcpu)
{
struct vgic_cpu *vgic_cpu = &vcpu->arch.vgic_cpu;
+ u32 model = vcpu->kvm->arch.vgic.vgic_model;
struct vgic_irq *irq;
int count = 0;
@@ -390,8 +397,23 @@ static void vgic_populate_lrs(struct kvm_vcpu *vcpu)
list_for_each_entry(irq, &vgic_cpu->ap_list_head, ap_list) {
spin_lock(&irq->irq_lock);
- if (vgic_target_oracle(irq) == vcpu)
+
+ if (unlikely(vgic_target_oracle(irq) != vcpu))
+ goto next;
+
+ /*
+ * If we get an SGI with multiple sources, try to get
+ * them in all at once.
+ */
+ if (model == KVM_DEV_TYPE_ARM_VGIC_V2 &&
+ irq->intid < VGIC_NR_SGIS) {
+ while(irq->source && count < vcpu->arch.vgic_cpu.nr_lr)
+ vgic_populate_lr(vcpu, irq, count++);
+ } else {
vgic_populate_lr(vcpu, irq, count++);
+ }
+
+ next:
spin_unlock(&irq->irq_lock);
if (count == vcpu->arch.vgic_cpu.nr_lr)
--
2.1.4
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PULL] KVM/ARM fixes for v4.4-rc6
2015-12-18 13:00 ` Marc Zyngier
@ 2015-12-18 14:01 ` Marc Zyngier
-1 siblings, 0 replies; 6+ messages in thread
From: Marc Zyngier @ 2015-12-18 14:01 UTC (permalink / raw)
To: Paolo Bonzini, Gleb Natapov; +Cc: linux-arm-kernel, kvm, kvmarm
Idiot alert, sending the wrong patch...
On 18/12/15 13:00, Marc Zyngier wrote:
> Hi Paolo,
>
> We have a one line fix for the VGIC this time around, fixing a patch
> that went in -rc2. Oh well. Hopefully this is the last one for v4.4.
>
> Please pull!
Or don't, actually.
Sorry about the noise, the real one is coming in a minute, once I've
collected a couple of brain cells...
Thanks,
M.
--
Jazz is not dead. It just smells funny...
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PULL] KVM/ARM fixes for v4.4-rc6
@ 2015-12-18 14:01 ` Marc Zyngier
0 siblings, 0 replies; 6+ messages in thread
From: Marc Zyngier @ 2015-12-18 14:01 UTC (permalink / raw)
To: linux-arm-kernel
Idiot alert, sending the wrong patch...
On 18/12/15 13:00, Marc Zyngier wrote:
> Hi Paolo,
>
> We have a one line fix for the VGIC this time around, fixing a patch
> that went in -rc2. Oh well. Hopefully this is the last one for v4.4.
>
> Please pull!
Or don't, actually.
Sorry about the noise, the real one is coming in a minute, once I've
collected a couple of brain cells...
Thanks,
M.
--
Jazz is not dead. It just smells funny...
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2015-12-18 14:01 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-12-18 13:00 [PULL] KVM/ARM fixes for v4.4-rc6 Marc Zyngier
2015-12-18 13:00 ` Marc Zyngier
2015-12-18 13:00 ` [PATCH] KVM: arm/arm64: vgic-new: Try to insert multi-source SGIs at once Marc Zyngier
2015-12-18 13:00 ` Marc Zyngier
2015-12-18 14:01 ` [PULL] KVM/ARM fixes for v4.4-rc6 Marc Zyngier
2015-12-18 14:01 ` Marc Zyngier
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.