[GIT PULL 0/9] KVM: s390: Fixes and cleanups for 3.17

[GIT PULL 0/9] KVM: s390: Fixes and cleanups for 3.17

[Christian Borntraeger]

Paolo,

this should be the last bunch of s390 patches for 3.17 (on top of the
mp_state changes). Please consider to apply.

The following changes since commit 6352e4d2dd9a349024a41356148eced553e1dce4:

  KVM: s390: implement KVM_(S|G)ET_MP_STATE for user space state control (2014-07-10 14:11:17 +0200)

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/kvms390/linux.git  tags/kvm-s390-20140721

for you to fetch changes up to e59d120f96687a606db0513c427f10e30a427cc4:

  KVM: s390: add ipte to trace event decoding (2014-07-21 13:22:47 +0200)

----------------------------------------------------------------
Bugfixes
--------
- add IPTE to trace event decoder
- document and advertise KVM_CAP_S390_IRQCHIP

Cleanups
--------
- Reuse kvm_vcpu_block for s390
- Get rid of tasklet for wakup processing

----------------------------------------------------------------
Christian Borntraeger (1):
      KVM: s390: add ipte to trace event decoding

Cornelia Huck (3):
      KVM: document target of capability enablement
      KVM: s390: document KVM_CAP_S390_IRQCHIP
      KVM: s390: advertise KVM_CAP_S390_IRQCHIP

David Hildenbrand (5):
      KVM: s390: cleanup handle_wait by reusing kvm_vcpu_block
      KVM: s390: remove _bh locking from local_int.lock
      KVM: s390: remove _bh locking from start_stop_lock
      KVM: s390: move vcpu wakeup code to a central point
      KVM: s390: remove the tasklet used by the hrtimer

 Documentation/virtual/kvm/api.txt |  27 ++++++++--
 arch/s390/include/asm/kvm_host.h  |   2 -
 arch/s390/include/uapi/asm/sie.h  |   1 +
 arch/s390/kvm/interrupt.c         | 100 ++++++++++++--------------------------
 arch/s390/kvm/kvm-s390.c          |  18 ++++---
 arch/s390/kvm/kvm-s390.h          |   2 +-
 arch/s390/kvm/sigp.c              |  36 ++++++--------
 7 files changed, 82 insertions(+), 104 deletions(-)

[GIT PULL 1/9] KVM: s390: cleanup handle_wait by reusing kvm_vcpu_block

[Christian Borntraeger]

From: David Hildenbrand <dahi@linux.vnet.ibm.com>

This patch cleans up the code in handle_wait by reusing the common code
function kvm_vcpu_block.

signal_pending(), kvm_cpu_has_pending_timer() and kvm_arch_vcpu_runnable() are
sufficient for checking if we need to wake-up that VCPU. kvm_vcpu_block
uses these functions, so no checks are lost.

The flag "timer_due" can be removed - kvm_cpu_has_pending_timer() tests whether
the timer is pending, thus the vcpu is correctly woken up.

Signed-off-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/include/asm/kvm_host.h |  1 -
 arch/s390/kvm/interrupt.c        | 41 +++++-----------------------------------
 arch/s390/kvm/kvm-s390.c         |  3 +++
 3 files changed, 8 insertions(+), 37 deletions(-)

diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index c2ba020..b3acf28 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -305,7 +305,6 @@ struct kvm_s390_local_interrupt {
 	struct list_head list;
 	atomic_t active;
 	struct kvm_s390_float_interrupt *float_int;
-	int timer_due; /* event indicator for waitqueue below */
 	wait_queue_head_t *wq;
 	atomic_t *cpuflags;
 	unsigned int action_bits;
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 90c8de2..5fd11ce 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -585,60 +585,32 @@ int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu)
 int kvm_s390_handle_wait(struct kvm_vcpu *vcpu)
 {
 	u64 now, sltime;
-	DECLARE_WAITQUEUE(wait, current);
 
 	vcpu->stat.exit_wait_state++;
-	if (kvm_cpu_has_interrupt(vcpu))
-		return 0;
 
-	__set_cpu_idle(vcpu);
-	spin_lock_bh(&vcpu->arch.local_int.lock);
-	vcpu->arch.local_int.timer_due = 0;
-	spin_unlock_bh(&vcpu->arch.local_int.lock);
+	/* fast path */
+	if (kvm_cpu_has_pending_timer(vcpu) || kvm_arch_vcpu_runnable(vcpu))
+		return 0;
 
 	if (psw_interrupts_disabled(vcpu)) {
 		VCPU_EVENT(vcpu, 3, "%s", "disabled wait");
-		__unset_cpu_idle(vcpu);
 		return -EOPNOTSUPP; /* disabled wait */
 	}
 
+	__set_cpu_idle(vcpu);
 	if (!ckc_interrupts_enabled(vcpu)) {
 		VCPU_EVENT(vcpu, 3, "%s", "enabled wait w/o timer");
 		goto no_timer;
 	}
 
 	now = get_tod_clock_fast() + vcpu->arch.sie_block->epoch;
-	if (vcpu->arch.sie_block->ckc < now) {
-		__unset_cpu_idle(vcpu);
-		return 0;
-	}
-
 	sltime = tod_to_ns(vcpu->arch.sie_block->ckc - now);
-
 	hrtimer_start(&vcpu->arch.ckc_timer, ktime_set (0, sltime) , HRTIMER_MODE_REL);
 	VCPU_EVENT(vcpu, 5, "enabled wait via clock comparator: %llx ns", sltime);
 no_timer:
 	srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx);
-	spin_lock(&vcpu->arch.local_int.float_int->lock);
-	spin_lock_bh(&vcpu->arch.local_int.lock);
-	add_wait_queue(&vcpu->wq, &wait);
-	while (list_empty(&vcpu->arch.local_int.list) &&
-		list_empty(&vcpu->arch.local_int.float_int->list) &&
-		(!vcpu->arch.local_int.timer_due) &&
-		!signal_pending(current) &&
-		!kvm_s390_si_ext_call_pending(vcpu)) {
-		set_current_state(TASK_INTERRUPTIBLE);
-		spin_unlock_bh(&vcpu->arch.local_int.lock);
-		spin_unlock(&vcpu->arch.local_int.float_int->lock);
-		schedule();
-		spin_lock(&vcpu->arch.local_int.float_int->lock);
-		spin_lock_bh(&vcpu->arch.local_int.lock);
-	}
+	kvm_vcpu_block(vcpu);
 	__unset_cpu_idle(vcpu);
-	__set_current_state(TASK_RUNNING);
-	remove_wait_queue(&vcpu->wq, &wait);
-	spin_unlock_bh(&vcpu->arch.local_int.lock);
-	spin_unlock(&vcpu->arch.local_int.float_int->lock);
 	vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu);
 
 	hrtimer_try_to_cancel(&vcpu->arch.ckc_timer);
@@ -649,11 +621,8 @@ void kvm_s390_tasklet(unsigned long parm)
 {
 	struct kvm_vcpu *vcpu = (struct kvm_vcpu *) parm;
 
-	spin_lock(&vcpu->arch.local_int.lock);
-	vcpu->arch.local_int.timer_due = 1;
 	if (waitqueue_active(&vcpu->wq))
 		wake_up_interruptible(&vcpu->wq);
-	spin_unlock(&vcpu->arch.local_int.lock);
 }
 
 /*
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index fdf88f7..ecb1357 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -1068,6 +1068,9 @@ retry:
 		goto retry;
 	}
 
+	/* nothing to do, just clear the request */
+	clear_bit(KVM_REQ_UNHALT, &vcpu->requests);
+
 	return 0;
 }
 
-- 
1.8.4.2

[GIT PULL 2/9] KVM: s390: remove _bh locking from local_int.lock

[Christian Borntraeger]

From: David Hildenbrand <dahi@linux.vnet.ibm.com>

local_int.lock is not used in a bottom-half handler anymore, therefore we can
turn it into an ordinary spin_lock at all occurrences.

Signed-off-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/kvm/interrupt.c | 32 ++++++++++++++++----------------
 arch/s390/kvm/kvm-s390.c  |  4 ++--
 arch/s390/kvm/sigp.c      | 20 ++++++++++----------
 3 files changed, 28 insertions(+), 28 deletions(-)

diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 5fd11ce..86575b4 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -544,13 +544,13 @@ int kvm_cpu_has_interrupt(struct kvm_vcpu *vcpu)
 	int rc = 0;
 
 	if (atomic_read(&li->active)) {
-		spin_lock_bh(&li->lock);
+		spin_lock(&li->lock);
 		list_for_each_entry(inti, &li->list, list)
 			if (__interrupt_is_deliverable(vcpu, inti)) {
 				rc = 1;
 				break;
 			}
-		spin_unlock_bh(&li->lock);
+		spin_unlock(&li->lock);
 	}
 
 	if ((!rc) && atomic_read(&fi->active)) {
@@ -645,13 +645,13 @@ void kvm_s390_clear_local_irqs(struct kvm_vcpu *vcpu)
 	struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int;
 	struct kvm_s390_interrupt_info  *n, *inti = NULL;
 
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	list_for_each_entry_safe(inti, n, &li->list, list) {
 		list_del(&inti->list);
 		kfree(inti);
 	}
 	atomic_set(&li->active, 0);
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 
 	/* clear pending external calls set by sigp interpretation facility */
 	atomic_clear_mask(CPUSTAT_ECALL_PEND, &vcpu->arch.sie_block->cpuflags);
@@ -670,7 +670,7 @@ void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu)
 	if (atomic_read(&li->active)) {
 		do {
 			deliver = 0;
-			spin_lock_bh(&li->lock);
+			spin_lock(&li->lock);
 			list_for_each_entry_safe(inti, n, &li->list, list) {
 				if (__interrupt_is_deliverable(vcpu, inti)) {
 					list_del(&inti->list);
@@ -681,7 +681,7 @@ void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu)
 			}
 			if (list_empty(&li->list))
 				atomic_set(&li->active, 0);
-			spin_unlock_bh(&li->lock);
+			spin_unlock(&li->lock);
 			if (deliver) {
 				__do_deliver_interrupt(vcpu, inti);
 				kfree(inti);
@@ -727,7 +727,7 @@ void kvm_s390_deliver_pending_machine_checks(struct kvm_vcpu *vcpu)
 	if (atomic_read(&li->active)) {
 		do {
 			deliver = 0;
-			spin_lock_bh(&li->lock);
+			spin_lock(&li->lock);
 			list_for_each_entry_safe(inti, n, &li->list, list) {
 				if ((inti->type == KVM_S390_MCHK) &&
 				    __interrupt_is_deliverable(vcpu, inti)) {
@@ -739,7 +739,7 @@ void kvm_s390_deliver_pending_machine_checks(struct kvm_vcpu *vcpu)
 			}
 			if (list_empty(&li->list))
 				atomic_set(&li->active, 0);
-			spin_unlock_bh(&li->lock);
+			spin_unlock(&li->lock);
 			if (deliver) {
 				__do_deliver_interrupt(vcpu, inti);
 				kfree(inti);
@@ -786,11 +786,11 @@ int kvm_s390_inject_program_int(struct kvm_vcpu *vcpu, u16 code)
 
 	VCPU_EVENT(vcpu, 3, "inject: program check %d (from kernel)", code);
 	trace_kvm_s390_inject_vcpu(vcpu->vcpu_id, inti->type, code, 0, 1);
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	list_add(&inti->list, &li->list);
 	atomic_set(&li->active, 1);
 	BUG_ON(waitqueue_active(li->wq));
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 	return 0;
 }
 
@@ -811,11 +811,11 @@ int kvm_s390_inject_prog_irq(struct kvm_vcpu *vcpu,
 
 	inti->type = KVM_S390_PROGRAM_INT;
 	memcpy(&inti->pgm, pgm_info, sizeof(inti->pgm));
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	list_add(&inti->list, &li->list);
 	atomic_set(&li->active, 1);
 	BUG_ON(waitqueue_active(li->wq));
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 	return 0;
 }
 
@@ -903,12 +903,12 @@ static int __inject_vm(struct kvm *kvm, struct kvm_s390_interrupt_info *inti)
 	}
 	dst_vcpu = kvm_get_vcpu(kvm, sigcpu);
 	li = &dst_vcpu->arch.local_int;
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags);
 	if (waitqueue_active(li->wq))
 		wake_up_interruptible(li->wq);
 	kvm_get_vcpu(kvm, sigcpu)->preempted = true;
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 unlock_fi:
 	spin_unlock(&fi->lock);
 	mutex_unlock(&kvm->lock);
@@ -1050,7 +1050,7 @@ int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu,
 
 	mutex_lock(&vcpu->kvm->lock);
 	li = &vcpu->arch.local_int;
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	if (inti->type == KVM_S390_PROGRAM_INT)
 		list_add(&inti->list, &li->list);
 	else
@@ -1062,7 +1062,7 @@ int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu,
 	if (waitqueue_active(&vcpu->wq))
 		wake_up_interruptible(&vcpu->wq);
 	vcpu->preempted = true;
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 	mutex_unlock(&vcpu->kvm->lock);
 	return 0;
 }
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index ecb1357..a7bda18 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -1522,13 +1522,13 @@ void kvm_s390_vcpu_stop(struct kvm_vcpu *vcpu)
 	online_vcpus = atomic_read(&vcpu->kvm->online_vcpus);
 
 	/* Need to lock access to action_bits to avoid a SIGP race condition */
-	spin_lock_bh(&vcpu->arch.local_int.lock);
+	spin_lock(&vcpu->arch.local_int.lock);
 	atomic_set_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags);
 
 	/* SIGP STOP and SIGP STOP AND STORE STATUS has been fully processed */
 	vcpu->arch.local_int.action_bits &=
 				 ~(ACTION_STOP_ON_STOP | ACTION_STORE_ON_STOP);
-	spin_unlock_bh(&vcpu->arch.local_int.lock);
+	spin_unlock(&vcpu->arch.local_int.lock);
 
 	__disable_ibs_on_vcpu(vcpu);
 
diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c
index fd7fb5c..946992f 100644
--- a/arch/s390/kvm/sigp.c
+++ b/arch/s390/kvm/sigp.c
@@ -135,7 +135,7 @@ static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action)
 		return -ENOMEM;
 	inti->type = KVM_S390_SIGP_STOP;
 
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	if (li->action_bits & ACTION_STOP_ON_STOP) {
 		/* another SIGP STOP is pending */
 		rc = SIGP_CC_BUSY;
@@ -154,7 +154,7 @@ static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action)
 	if (waitqueue_active(li->wq))
 		wake_up_interruptible(li->wq);
 out:
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 
 	return rc;
 }
@@ -243,7 +243,7 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address,
 	if (!inti)
 		return SIGP_CC_BUSY;
 
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	/* cpu must be in stopped state */
 	if (!(atomic_read(li->cpuflags) & CPUSTAT_STOPPED)) {
 		*reg &= 0xffffffff00000000UL;
@@ -264,7 +264,7 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address,
 
 	VCPU_EVENT(vcpu, 4, "set prefix of cpu %02x to %x", cpu_addr, address);
 out_li:
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 	return rc;
 }
 
@@ -280,9 +280,9 @@ static int __sigp_store_status_at_addr(struct kvm_vcpu *vcpu, u16 cpu_id,
 	if (!dst_vcpu)
 		return SIGP_CC_NOT_OPERATIONAL;
 
-	spin_lock_bh(&dst_vcpu->arch.local_int.lock);
+	spin_lock(&dst_vcpu->arch.local_int.lock);
 	flags = atomic_read(dst_vcpu->arch.local_int.cpuflags);
-	spin_unlock_bh(&dst_vcpu->arch.local_int.lock);
+	spin_unlock(&dst_vcpu->arch.local_int.lock);
 	if (!(flags & CPUSTAT_STOPPED)) {
 		*reg &= 0xffffffff00000000UL;
 		*reg |= SIGP_STATUS_INCORRECT_STATE;
@@ -343,10 +343,10 @@ static int sigp_check_callable(struct kvm_vcpu *vcpu, u16 cpu_addr)
 	if (!dst_vcpu)
 		return SIGP_CC_NOT_OPERATIONAL;
 	li = &dst_vcpu->arch.local_int;
-	spin_lock_bh(&li->lock);
+	spin_lock(&li->lock);
 	if (li->action_bits & ACTION_STOP_ON_STOP)
 		rc = SIGP_CC_BUSY;
-	spin_unlock_bh(&li->lock);
+	spin_unlock(&li->lock);
 
 	return rc;
 }
@@ -466,11 +466,11 @@ int kvm_s390_handle_sigp_pei(struct kvm_vcpu *vcpu)
 		dest_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
 		BUG_ON(dest_vcpu == NULL);
 
-		spin_lock_bh(&dest_vcpu->arch.local_int.lock);
+		spin_lock(&dest_vcpu->arch.local_int.lock);
 		if (waitqueue_active(&dest_vcpu->wq))
 			wake_up_interruptible(&dest_vcpu->wq);
 		dest_vcpu->preempted = true;
-		spin_unlock_bh(&dest_vcpu->arch.local_int.lock);
+		spin_unlock(&dest_vcpu->arch.local_int.lock);
 
 		kvm_s390_set_psw_cc(vcpu, SIGP_CC_ORDER_CODE_ACCEPTED);
 		return 0;
-- 
1.8.4.2

[GIT PULL 3/9] KVM: s390: remove _bh locking from start_stop_lock

[Christian Borntraeger]

From: David Hildenbrand <dahi@linux.vnet.ibm.com>

The start_stop_lock is no longer acquired when in atomic context, therefore we
can convert it into an ordinary spin_lock.

Signed-off-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/kvm/kvm-s390.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index a7bda18..b29a031 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -1478,7 +1478,7 @@ void kvm_s390_vcpu_start(struct kvm_vcpu *vcpu)
 
 	trace_kvm_s390_vcpu_start_stop(vcpu->vcpu_id, 1);
 	/* Only one cpu at a time may enter/leave the STOPPED state. */
-	spin_lock_bh(&vcpu->kvm->arch.start_stop_lock);
+	spin_lock(&vcpu->kvm->arch.start_stop_lock);
 	online_vcpus = atomic_read(&vcpu->kvm->online_vcpus);
 
 	for (i = 0; i < online_vcpus; i++) {
@@ -1504,7 +1504,7 @@ void kvm_s390_vcpu_start(struct kvm_vcpu *vcpu)
 	 * Let's play safe and flush the VCPU at startup.
 	 */
 	vcpu->arch.sie_block->ihcpu  = 0xffff;
-	spin_unlock_bh(&vcpu->kvm->arch.start_stop_lock);
+	spin_unlock(&vcpu->kvm->arch.start_stop_lock);
 	return;
 }
 
@@ -1518,7 +1518,7 @@ void kvm_s390_vcpu_stop(struct kvm_vcpu *vcpu)
 
 	trace_kvm_s390_vcpu_start_stop(vcpu->vcpu_id, 0);
 	/* Only one cpu at a time may enter/leave the STOPPED state. */
-	spin_lock_bh(&vcpu->kvm->arch.start_stop_lock);
+	spin_lock(&vcpu->kvm->arch.start_stop_lock);
 	online_vcpus = atomic_read(&vcpu->kvm->online_vcpus);
 
 	/* Need to lock access to action_bits to avoid a SIGP race condition */
@@ -1547,7 +1547,7 @@ void kvm_s390_vcpu_stop(struct kvm_vcpu *vcpu)
 		__enable_ibs_on_vcpu(started_vcpu);
 	}
 
-	spin_unlock_bh(&vcpu->kvm->arch.start_stop_lock);
+	spin_unlock(&vcpu->kvm->arch.start_stop_lock);
 	return;
 }
 
-- 
1.8.4.2

[GIT PULL 4/9] KVM: s390: move vcpu wakeup code to a central point

[Christian Borntraeger]

From: David Hildenbrand <dahi@linux.vnet.ibm.com>

Let's move the vcpu wakeup code to a central point.

We should set the vcpu->preempted flag only if the target is actually sleeping
and before the real wakeup happens. Otherwise the preempted flag might be set,
when not necessary. This may result in immediate reschedules after schedule()
in some scenarios.

The wakeup code doesn't require the local_int.lock to be held.

Signed-off-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/kvm/interrupt.c | 24 +++++++++++++++---------
 arch/s390/kvm/kvm-s390.h  |  1 +
 arch/s390/kvm/sigp.c      | 20 ++++++--------------
 3 files changed, 22 insertions(+), 23 deletions(-)

diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 86575b4..65396e1 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -617,12 +617,22 @@ no_timer:
 	return 0;
 }
 
+void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu)
+{
+	if (waitqueue_active(&vcpu->wq)) {
+		/*
+		 * The vcpu gave up the cpu voluntarily, mark it as a good
+		 * yield-candidate.
+		 */
+		vcpu->preempted = true;
+		wake_up_interruptible(&vcpu->wq);
+	}
+}
+
 void kvm_s390_tasklet(unsigned long parm)
 {
 	struct kvm_vcpu *vcpu = (struct kvm_vcpu *) parm;
-
-	if (waitqueue_active(&vcpu->wq))
-		wake_up_interruptible(&vcpu->wq);
+	kvm_s390_vcpu_wakeup(vcpu);
 }
 
 /*
@@ -905,10 +915,8 @@ static int __inject_vm(struct kvm *kvm, struct kvm_s390_interrupt_info *inti)
 	li = &dst_vcpu->arch.local_int;
 	spin_lock(&li->lock);
 	atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags);
-	if (waitqueue_active(li->wq))
-		wake_up_interruptible(li->wq);
-	kvm_get_vcpu(kvm, sigcpu)->preempted = true;
 	spin_unlock(&li->lock);
+	kvm_s390_vcpu_wakeup(kvm_get_vcpu(kvm, sigcpu));
 unlock_fi:
 	spin_unlock(&fi->lock);
 	mutex_unlock(&kvm->lock);
@@ -1059,11 +1067,9 @@ int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu,
 	if (inti->type == KVM_S390_SIGP_STOP)
 		li->action_bits |= ACTION_STOP_ON_STOP;
 	atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags);
-	if (waitqueue_active(&vcpu->wq))
-		wake_up_interruptible(&vcpu->wq);
-	vcpu->preempted = true;
 	spin_unlock(&li->lock);
 	mutex_unlock(&vcpu->kvm->lock);
+	kvm_s390_vcpu_wakeup(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h
index 33a0e4b..665eacc 100644
--- a/arch/s390/kvm/kvm-s390.h
+++ b/arch/s390/kvm/kvm-s390.h
@@ -136,6 +136,7 @@ static inline int kvm_s390_user_cpu_state_ctrl(struct kvm *kvm)
 }
 
 int kvm_s390_handle_wait(struct kvm_vcpu *vcpu);
+void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu);
 enum hrtimer_restart kvm_s390_idle_wakeup(struct hrtimer *timer);
 void kvm_s390_tasklet(unsigned long parm);
 void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu);
diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c
index 946992f..c6f1c2b 100644
--- a/arch/s390/kvm/sigp.c
+++ b/arch/s390/kvm/sigp.c
@@ -125,8 +125,9 @@ static int __sigp_external_call(struct kvm_vcpu *vcpu, u16 cpu_addr)
 	return rc ? rc : SIGP_CC_ORDER_CODE_ACCEPTED;
 }
 
-static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action)
+static int __inject_sigp_stop(struct kvm_vcpu *dst_vcpu, int action)
 {
+	struct kvm_s390_local_interrupt *li = &dst_vcpu->arch.local_int;
 	struct kvm_s390_interrupt_info *inti;
 	int rc = SIGP_CC_ORDER_CODE_ACCEPTED;
 
@@ -151,8 +152,7 @@ static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action)
 	atomic_set(&li->active, 1);
 	li->action_bits |= action;
 	atomic_set_mask(CPUSTAT_STOP_INT, li->cpuflags);
-	if (waitqueue_active(li->wq))
-		wake_up_interruptible(li->wq);
+	kvm_s390_vcpu_wakeup(dst_vcpu);
 out:
 	spin_unlock(&li->lock);
 
@@ -161,7 +161,6 @@ out:
 
 static int __sigp_stop(struct kvm_vcpu *vcpu, u16 cpu_addr, int action)
 {
-	struct kvm_s390_local_interrupt *li;
 	struct kvm_vcpu *dst_vcpu = NULL;
 	int rc;
 
@@ -171,9 +170,8 @@ static int __sigp_stop(struct kvm_vcpu *vcpu, u16 cpu_addr, int action)
 	dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
 	if (!dst_vcpu)
 		return SIGP_CC_NOT_OPERATIONAL;
-	li = &dst_vcpu->arch.local_int;
 
-	rc = __inject_sigp_stop(li, action);
+	rc = __inject_sigp_stop(dst_vcpu, action);
 
 	VCPU_EVENT(vcpu, 4, "sent sigp stop to cpu %x", cpu_addr);
 
@@ -258,8 +256,7 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address,
 
 	list_add_tail(&inti->list, &li->list);
 	atomic_set(&li->active, 1);
-	if (waitqueue_active(li->wq))
-		wake_up_interruptible(li->wq);
+	kvm_s390_vcpu_wakeup(dst_vcpu);
 	rc = SIGP_CC_ORDER_CODE_ACCEPTED;
 
 	VCPU_EVENT(vcpu, 4, "set prefix of cpu %02x to %x", cpu_addr, address);
@@ -466,12 +463,7 @@ int kvm_s390_handle_sigp_pei(struct kvm_vcpu *vcpu)
 		dest_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
 		BUG_ON(dest_vcpu == NULL);
 
-		spin_lock(&dest_vcpu->arch.local_int.lock);
-		if (waitqueue_active(&dest_vcpu->wq))
-			wake_up_interruptible(&dest_vcpu->wq);
-		dest_vcpu->preempted = true;
-		spin_unlock(&dest_vcpu->arch.local_int.lock);
-
+		kvm_s390_vcpu_wakeup(dest_vcpu);
 		kvm_s390_set_psw_cc(vcpu, SIGP_CC_ORDER_CODE_ACCEPTED);
 		return 0;
 	}
-- 
1.8.4.2

[GIT PULL 5/9] KVM: s390: remove the tasklet used by the hrtimer

[Christian Borntraeger]

From: David Hildenbrand <dahi@linux.vnet.ibm.com>

We can get rid of the tasklet used for waking up a VCPU in the hrtimer
code but wakeup the VCPU directly.

Signed-off-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/include/asm/kvm_host.h |  1 -
 arch/s390/kvm/interrupt.c        | 13 +------------
 arch/s390/kvm/kvm-s390.c         |  2 --
 arch/s390/kvm/kvm-s390.h         |  1 -
 4 files changed, 1 insertion(+), 16 deletions(-)

diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index b3acf28..773bef7 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -366,7 +366,6 @@ struct kvm_vcpu_arch {
 	s390_fp_regs      guest_fpregs;
 	struct kvm_s390_local_interrupt local_int;
 	struct hrtimer    ckc_timer;
-	struct tasklet_struct tasklet;
 	struct kvm_s390_pgm_info pgm;
 	union  {
 		struct cpuid	cpu_id;
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 65396e1..1be3d8d 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -629,23 +629,12 @@ void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu)
 	}
 }
 
-void kvm_s390_tasklet(unsigned long parm)
-{
-	struct kvm_vcpu *vcpu = (struct kvm_vcpu *) parm;
-	kvm_s390_vcpu_wakeup(vcpu);
-}
-
-/*
- * low level hrtimer wake routine. Because this runs in hardirq context
- * we schedule a tasklet to do the real work.
- */
 enum hrtimer_restart kvm_s390_idle_wakeup(struct hrtimer *timer)
 {
 	struct kvm_vcpu *vcpu;
 
 	vcpu = container_of(timer, struct kvm_vcpu, arch.ckc_timer);
-	vcpu->preempted = true;
-	tasklet_schedule(&vcpu->arch.tasklet);
+	kvm_s390_vcpu_wakeup(vcpu);
 
 	return HRTIMER_NORESTART;
 }
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index b29a031..dd902e6 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -649,8 +649,6 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 			return rc;
 	}
 	hrtimer_init(&vcpu->arch.ckc_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS);
-	tasklet_init(&vcpu->arch.tasklet, kvm_s390_tasklet,
-		     (unsigned long) vcpu);
 	vcpu->arch.ckc_timer.function = kvm_s390_idle_wakeup;
 	get_cpu_id(&vcpu->arch.cpu_id);
 	vcpu->arch.cpu_id.version = 0xff;
diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h
index 665eacc..3862fa2 100644
--- a/arch/s390/kvm/kvm-s390.h
+++ b/arch/s390/kvm/kvm-s390.h
@@ -138,7 +138,6 @@ static inline int kvm_s390_user_cpu_state_ctrl(struct kvm *kvm)
 int kvm_s390_handle_wait(struct kvm_vcpu *vcpu);
 void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu);
 enum hrtimer_restart kvm_s390_idle_wakeup(struct hrtimer *timer);
-void kvm_s390_tasklet(unsigned long parm);
 void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu);
 void kvm_s390_deliver_pending_machine_checks(struct kvm_vcpu *vcpu);
 void kvm_s390_clear_local_irqs(struct kvm_vcpu *vcpu);
-- 
1.8.4.2

[GIT PULL 6/9] KVM: document target of capability enablement

[Christian Borntraeger]

From: Cornelia Huck <cornelia.huck@de.ibm.com>

Capabilities can be enabled on a vcpu or (since recently) on a vm. Document
this and note for the existing capabilites whether they are per-vcpu or
per-vm.

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 Documentation/virtual/kvm/api.txt | 18 +++++++++++++++---
 1 file changed, 15 insertions(+), 3 deletions(-)

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index a41465b..7ab41e9 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -2875,15 +2875,18 @@ The fields in each entry are defined as follows:
 6. Capabilities that can be enabled
 -----------------------------------
 
-There are certain capabilities that change the behavior of the virtual CPU when
-enabled. To enable them, please see section 4.37. Below you can find a list of
-capabilities and what their effect on the vCPU is when enabling them.
+There are certain capabilities that change the behavior of the virtual CPU or
+the virtual machine when enabled. To enable them, please see section 4.37.
+Below you can find a list of capabilities and what their effect on the vCPU or
+the virtual machine is when enabling them.
 
 The following information is provided along with the description:
 
   Architectures: which instruction set architectures provide this ioctl.
       x86 includes both i386 and x86_64.
 
+  Target: whether this is a per-vcpu or per-vm capability.
+
   Parameters: what parameters are accepted by the capability.
 
   Returns: the return value.  General error numbers (EBADF, ENOMEM, EINVAL)
@@ -2893,6 +2896,7 @@ The following information is provided along with the description:
 6.1 KVM_CAP_PPC_OSI
 
 Architectures: ppc
+Target: vcpu
 Parameters: none
 Returns: 0 on success; -1 on error
 
@@ -2907,6 +2911,7 @@ When this capability is enabled, KVM_EXIT_OSI can occur.
 6.2 KVM_CAP_PPC_PAPR
 
 Architectures: ppc
+Target: vcpu
 Parameters: none
 Returns: 0 on success; -1 on error
 
@@ -2926,6 +2931,7 @@ When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur.
 6.3 KVM_CAP_SW_TLB
 
 Architectures: ppc
+Target: vcpu
 Parameters: args[0] is the address of a struct kvm_config_tlb
 Returns: 0 on success; -1 on error
 
@@ -2968,6 +2974,7 @@ For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV:
 6.4 KVM_CAP_S390_CSS_SUPPORT
 
 Architectures: s390
+Target: vcpu
 Parameters: none
 Returns: 0 on success; -1 on error
 
@@ -2979,9 +2986,13 @@ handled in-kernel, while the other I/O instructions are passed to userspace.
 When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST
 SUBCHANNEL intercepts.
 
+Note that even though this capability is enabled per-vcpu, the complete
+virtual machine is affected.
+
 6.5 KVM_CAP_PPC_EPR
 
 Architectures: ppc
+Target: vcpu
 Parameters: args[0] defines whether the proxy facility is active
 Returns: 0 on success; -1 on error
 
@@ -3007,6 +3018,7 @@ This capability connects the vcpu to an in-kernel MPIC device.
 6.7 KVM_CAP_IRQ_XICS
 
 Architectures: ppc
+Target: vcpu
 Parameters: args[0] is the XICS device fd
             args[1] is the XICS CPU number (server ID) for this vcpu
 
-- 
1.8.4.2

[GIT PULL 7/9] KVM: s390: document KVM_CAP_S390_IRQCHIP

[Christian Borntraeger]

From: Cornelia Huck <cornelia.huck@de.ibm.com>

Let's document that this is a capability that may be enabled per-vm.

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 Documentation/virtual/kvm/api.txt | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index 7ab41e9..f1979c7 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -3023,3 +3023,12 @@ Parameters: args[0] is the XICS device fd
             args[1] is the XICS CPU number (server ID) for this vcpu
 
 This capability connects the vcpu to an in-kernel XICS device.
+
+6.8 KVM_CAP_S390_IRQCHIP
+
+Architectures: s390
+Target: vm
+Parameters: none
+
+This capability enables the in-kernel irqchip for s390. Please refer to
+"4.24 KVM_CREATE_IRQCHIP" for details.
-- 
1.8.4.2

[GIT PULL 8/9] KVM: s390: advertise KVM_CAP_S390_IRQCHIP

[Christian Borntraeger]

From: Cornelia Huck <cornelia.huck@de.ibm.com>

We should advertise all capabilities, including those that can
be enabled.

Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
 arch/s390/kvm/kvm-s390.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index dd902e6..339b34a 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -166,6 +166,7 @@ int kvm_dev_ioctl_check_extension(long ext)
 	case KVM_CAP_IOEVENTFD:
 	case KVM_CAP_DEVICE_CTRL:
 	case KVM_CAP_ENABLE_CAP_VM:
+	case KVM_CAP_S390_IRQCHIP:
 	case KVM_CAP_VM_ATTRIBUTES:
 	case KVM_CAP_MP_STATE:
 		r = 1;
-- 
1.8.4.2

[GIT PULL 9/9] KVM: s390: add ipte to trace event decoding

[Christian Borntraeger]

IPTE intercept can happen, let's decode that.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
---
 arch/s390/include/uapi/asm/sie.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/s390/include/uapi/asm/sie.h b/arch/s390/include/uapi/asm/sie.h
index 5d9cc19..d4096fd 100644
--- a/arch/s390/include/uapi/asm/sie.h
+++ b/arch/s390/include/uapi/asm/sie.h
@@ -108,6 +108,7 @@
 	exit_code_ipa0(0xB2, 0x17, "STETR"),	\
 	exit_code_ipa0(0xB2, 0x18, "PC"),	\
 	exit_code_ipa0(0xB2, 0x20, "SERVC"),	\
+	exit_code_ipa0(0xB2, 0x21, "IPTE"),	\
 	exit_code_ipa0(0xB2, 0x28, "PT"),	\
 	exit_code_ipa0(0xB2, 0x29, "ISKE"),	\
 	exit_code_ipa0(0xB2, 0x2a, "RRBE"),	\
-- 
1.8.4.2

Re: [GIT PULL 8/9] KVM: s390: advertise KVM_CAP_S390_IRQCHIP

[Paolo Bonzini]

Il 21/07/2014 21:14, Christian Borntraeger ha scritto:
> From: Cornelia Huck <cornelia.huck@de.ibm.com>
> 
> We should advertise all capabilities, including those that can
> be enabled.
> 
> Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
> Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
> ---
>  arch/s390/kvm/kvm-s390.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index dd902e6..339b34a 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -166,6 +166,7 @@ int kvm_dev_ioctl_check_extension(long ext)
>  	case KVM_CAP_IOEVENTFD:
>  	case KVM_CAP_DEVICE_CTRL:
>  	case KVM_CAP_ENABLE_CAP_VM:
> +	case KVM_CAP_S390_IRQCHIP:
>  	case KVM_CAP_VM_ATTRIBUTES:
>  	case KVM_CAP_MP_STATE:
>  		r = 1;
> 

Should I apply this to 3.16 as well (or Cc: stable)?

Re: [GIT PULL 8/9] KVM: s390: advertise KVM_CAP_S390_IRQCHIP

[Christian Borntraeger]

On 22/07/14 10:32, Paolo Bonzini wrote:
> Il 21/07/2014 21:14, Christian Borntraeger ha scritto:
>> From: Cornelia Huck <cornelia.huck@de.ibm.com>
>>
>> We should advertise all capabilities, including those that can
>> be enabled.
>>
>> Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
>> Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
>> Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
>> ---
>>  arch/s390/kvm/kvm-s390.c | 1 +
>>  1 file changed, 1 insertion(+)
>>
>> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
>> index dd902e6..339b34a 100644
>> --- a/arch/s390/kvm/kvm-s390.c
>> +++ b/arch/s390/kvm/kvm-s390.c
>> @@ -166,6 +166,7 @@ int kvm_dev_ioctl_check_extension(long ext)
>>  	case KVM_CAP_IOEVENTFD:
>>  	case KVM_CAP_DEVICE_CTRL:
>>  	case KVM_CAP_ENABLE_CAP_VM:
>> +	case KVM_CAP_S390_IRQCHIP:
>>  	case KVM_CAP_VM_ATTRIBUTES:
>>  	case KVM_CAP_MP_STATE:
>>  		r = 1;
>>
> 
> Should I apply this to 3.16 as well (or Cc: stable)?
> 
Hmm, userspace currently does not care. But Cc: stable 3.15 and 3.16 might be "the right thing to do".

Thanks for noticing.

Christian