* [PULL 1/8] KVM: s390: Provide access to program parameter
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:27 ` [PULL 2/8] KVM: s390: expose gbea register to userspace Christian Borntraeger
` (7 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
commit d208c79d63e06457eef077af770d23dc4cde4d43 (KVM: s390: Enable
the LPP facility for guests) enabled the LPP instruction for guests.
We should expose the program parameter as a pseudo register for
migration/reset etc. Lets also reset this value on initial CPU
reset.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Reviewed-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: Jason J. Herne <jjherne@linux.vnet.ibm.com>
---
arch/s390/include/asm/kvm_host.h | 4 +++-
arch/s390/include/uapi/asm/kvm.h | 1 +
arch/s390/kvm/kvm-s390.c | 9 +++++++++
3 files changed, 13 insertions(+), 1 deletion(-)
diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index 2c69ba2..062b78c 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -107,7 +107,9 @@ struct kvm_s390_sie_block {
__u64 gbea; /* 0x0180 */
__u8 reserved188[24]; /* 0x0188 */
__u32 fac; /* 0x01a0 */
- __u8 reserved1a4[68]; /* 0x01a4 */
+ __u8 reserved1a4[58]; /* 0x01a4 */
+ __u64 pp; /* 0x01de */
+ __u8 reserved1e6[2]; /* 0x01e6 */
__u64 itdba; /* 0x01e8 */
__u8 reserved1f0[16]; /* 0x01f0 */
} __attribute__((packed));
diff --git a/arch/s390/include/uapi/asm/kvm.h b/arch/s390/include/uapi/asm/kvm.h
index cb4c1eb8..7663244 100644
--- a/arch/s390/include/uapi/asm/kvm.h
+++ b/arch/s390/include/uapi/asm/kvm.h
@@ -76,4 +76,5 @@ struct kvm_sync_regs {
#define KVM_REG_S390_PFTOKEN (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x5)
#define KVM_REG_S390_PFCOMPARE (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x6)
#define KVM_REG_S390_PFSELECT (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x7)
+#define KVM_REG_S390_PP (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x8)
#endif
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 18959bb..0262936 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -386,6 +386,7 @@ static void kvm_s390_vcpu_initial_reset(struct kvm_vcpu *vcpu)
vcpu->arch.guest_fpregs.fpc = 0;
asm volatile("lfpc %0" : : "Q" (vcpu->arch.guest_fpregs.fpc));
vcpu->arch.sie_block->gbea = 1;
+ vcpu->arch.sie_block->pp = 0;
vcpu->arch.pfault_token = KVM_S390_PFAULT_TOKEN_INVALID;
kvm_clear_async_pf_completion_queue(vcpu);
atomic_set_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags);
@@ -571,6 +572,10 @@ static int kvm_arch_vcpu_ioctl_get_one_reg(struct kvm_vcpu *vcpu,
r = put_user(vcpu->arch.pfault_select,
(u64 __user *)reg->addr);
break;
+ case KVM_REG_S390_PP:
+ r = put_user(vcpu->arch.sie_block->pp,
+ (u64 __user *)reg->addr);
+ break;
default:
break;
}
@@ -612,6 +617,10 @@ static int kvm_arch_vcpu_ioctl_set_one_reg(struct kvm_vcpu *vcpu,
r = get_user(vcpu->arch.pfault_select,
(u64 __user *)reg->addr);
break;
+ case KVM_REG_S390_PP:
+ r = get_user(vcpu->arch.sie_block->pp,
+ (u64 __user *)reg->addr);
+ break;
default:
break;
}
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* [PULL 2/8] KVM: s390: expose gbea register to userspace
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
2014-03-04 11:27 ` [PULL 1/8] KVM: s390: Provide access to program parameter Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:27 ` [PULL 3/8] KVM: s390: Simplify online vcpus counting for stsi Christian Borntraeger
` (6 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
For migration/reset we want to expose the guest breaking event
address register to userspace. Lets use ONE_REG for that purpose.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Reviewed-by: Jason J. Herne <jjherne@linux.vnet.ibm.com>
---
arch/s390/include/uapi/asm/kvm.h | 1 +
arch/s390/kvm/kvm-s390.c | 8 ++++++++
2 files changed, 9 insertions(+)
diff --git a/arch/s390/include/uapi/asm/kvm.h b/arch/s390/include/uapi/asm/kvm.h
index 7663244..2f0ade2 100644
--- a/arch/s390/include/uapi/asm/kvm.h
+++ b/arch/s390/include/uapi/asm/kvm.h
@@ -77,4 +77,5 @@ struct kvm_sync_regs {
#define KVM_REG_S390_PFCOMPARE (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x6)
#define KVM_REG_S390_PFSELECT (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x7)
#define KVM_REG_S390_PP (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x8)
+#define KVM_REG_S390_GBEA (KVM_REG_S390 | KVM_REG_SIZE_U64 | 0x9)
#endif
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 0262936..a3e4c07 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -576,6 +576,10 @@ static int kvm_arch_vcpu_ioctl_get_one_reg(struct kvm_vcpu *vcpu,
r = put_user(vcpu->arch.sie_block->pp,
(u64 __user *)reg->addr);
break;
+ case KVM_REG_S390_GBEA:
+ r = put_user(vcpu->arch.sie_block->gbea,
+ (u64 __user *)reg->addr);
+ break;
default:
break;
}
@@ -621,6 +625,10 @@ static int kvm_arch_vcpu_ioctl_set_one_reg(struct kvm_vcpu *vcpu,
r = get_user(vcpu->arch.sie_block->pp,
(u64 __user *)reg->addr);
break;
+ case KVM_REG_S390_GBEA:
+ r = get_user(vcpu->arch.sie_block->gbea,
+ (u64 __user *)reg->addr);
+ break;
default:
break;
}
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* [PULL 3/8] KVM: s390: Simplify online vcpus counting for stsi
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
2014-03-04 11:27 ` [PULL 1/8] KVM: s390: Provide access to program parameter Christian Borntraeger
2014-03-04 11:27 ` [PULL 2/8] KVM: s390: expose gbea register to userspace Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:27 ` [PULL 4/8] KVM: s390: Fixed CC of SIGP SET_PREFIX handler Christian Borntraeger
` (5 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
From: Jens Freimann <jfrei@linux.vnet.ibm.com>
We don't need to loop over all cpus to get the number of
vcpus. Let's use the available counter online_vcpus instead.
Signed-off-by: Jens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
arch/s390/kvm/priv.c | 7 +------
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c
index 75beea6..ae9e8ee 100644
--- a/arch/s390/kvm/priv.c
+++ b/arch/s390/kvm/priv.c
@@ -396,15 +396,10 @@ static int handle_stidp(struct kvm_vcpu *vcpu)
static void handle_stsi_3_2_2(struct kvm_vcpu *vcpu, struct sysinfo_3_2_2 *mem)
{
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
int cpus = 0;
int n;
- spin_lock(&fi->lock);
- for (n = 0; n < KVM_MAX_VCPUS; n++)
- if (fi->local_int[n])
- cpus++;
- spin_unlock(&fi->lock);
+ cpus = atomic_read(&vcpu->kvm->online_vcpus);
/* deal with other level 3 hypervisors */
if (stsi(mem, 3, 2, 2))
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* [PULL 4/8] KVM: s390: Fixed CC of SIGP SET_PREFIX handler
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
` (2 preceding siblings ...)
2014-03-04 11:27 ` [PULL 3/8] KVM: s390: Simplify online vcpus counting for stsi Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:27 ` [PULL 5/8] KVM: s390: get rid of local_int array Christian Borntraeger
` (4 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
From: Thomas Huth <thuth@linux.vnet.ibm.com>
When SIGP SET_PREFIX is called with an illegal CPU id, it must return
the condition code 3 ("not operational") instead of 1. Also fixed the
order in which the checks are done - CC3 has a higher priority than CC1.
And while we're at it, this patch also get rid of the floating interrupt
lock here by using kvm_get_vcpu() to get the local_int struct of the
destination CPU.
Signed-off-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
arch/s390/kvm/sigp.c | 24 ++++++++----------------
1 file changed, 8 insertions(+), 16 deletions(-)
diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c
index fe9442d..466eefa 100644
--- a/arch/s390/kvm/sigp.c
+++ b/arch/s390/kvm/sigp.c
@@ -249,12 +249,18 @@ static int __sigp_set_arch(struct kvm_vcpu *vcpu, u32 parameter)
static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address,
u64 *reg)
{
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
- struct kvm_s390_local_interrupt *li = NULL;
+ struct kvm_s390_local_interrupt *li;
+ struct kvm_vcpu *dst_vcpu = NULL;
struct kvm_s390_interrupt_info *inti;
int rc;
u8 tmp;
+ if (cpu_addr < KVM_MAX_VCPUS)
+ dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
+ if (!dst_vcpu)
+ return SIGP_CC_NOT_OPERATIONAL;
+ li = &dst_vcpu->arch.local_int;
+
/* make sure that the new value is valid memory */
address = address & 0x7fffe000u;
if (copy_from_guest_absolute(vcpu, &tmp, address, 1) ||
@@ -268,18 +274,6 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address,
if (!inti)
return SIGP_CC_BUSY;
- spin_lock(&fi->lock);
- if (cpu_addr < KVM_MAX_VCPUS)
- li = fi->local_int[cpu_addr];
-
- if (li == NULL) {
- *reg &= 0xffffffff00000000UL;
- *reg |= SIGP_STATUS_INCORRECT_STATE;
- rc = SIGP_CC_STATUS_STORED;
- kfree(inti);
- goto out_fi;
- }
-
spin_lock_bh(&li->lock);
/* cpu must be in stopped state */
if (!(atomic_read(li->cpuflags) & CPUSTAT_STOPPED)) {
@@ -302,8 +296,6 @@ 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);
-out_fi:
- spin_unlock(&fi->lock);
return rc;
}
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* [PULL 5/8] KVM: s390: get rid of local_int array
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
` (3 preceding siblings ...)
2014-03-04 11:27 ` [PULL 4/8] KVM: s390: Fixed CC of SIGP SET_PREFIX handler Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:27 ` [PULL 6/8] virtio_ccw: fix vcdev pointer handling issues Christian Borntraeger
` (3 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
From: Jens Freimann <jfrei@linux.vnet.ibm.com>
We can use kvm_get_vcpu() now and don't need the
local_int array in the floating_int struct anymore.
This also means we don't have to hold the float_int.lock
in some places.
Signed-off-by: Jens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
arch/s390/include/asm/kvm_host.h | 1 -
arch/s390/kvm/interrupt.c | 6 +-
arch/s390/kvm/kvm-s390.c | 5 +-
arch/s390/kvm/sigp.c | 124 ++++++++++++++++-----------------------
4 files changed, 56 insertions(+), 80 deletions(-)
diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index 062b78c..734d302 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -215,7 +215,6 @@ struct kvm_s390_float_interrupt {
int next_rr_cpu;
unsigned long idle_mask[(KVM_MAX_VCPUS + sizeof(long) - 1)
/ sizeof(long)];
- struct kvm_s390_local_interrupt *local_int[KVM_MAX_VCPUS];
unsigned int irq_count;
};
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index fff070b..1d0f9d5 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -692,6 +692,7 @@ static int __inject_vm(struct kvm *kvm, struct kvm_s390_interrupt_info *inti)
struct kvm_s390_local_interrupt *li;
struct kvm_s390_float_interrupt *fi;
struct kvm_s390_interrupt_info *iter;
+ struct kvm_vcpu *dst_vcpu = NULL;
int sigcpu;
int rc = 0;
@@ -726,9 +727,10 @@ static int __inject_vm(struct kvm *kvm, struct kvm_s390_interrupt_info *inti)
sigcpu = fi->next_rr_cpu++;
if (sigcpu == KVM_MAX_VCPUS)
sigcpu = fi->next_rr_cpu = 0;
- } while (fi->local_int[sigcpu] == NULL);
+ } while (kvm_get_vcpu(kvm, sigcpu) == NULL);
}
- li = fi->local_int[sigcpu];
+ dst_vcpu = kvm_get_vcpu(kvm, sigcpu);
+ li = &dst_vcpu->arch.local_int;
spin_lock_bh(&li->lock);
atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags);
if (waitqueue_active(li->wq))
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index a3e4c07..9136f8d 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -460,11 +460,8 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
spin_lock_init(&vcpu->arch.local_int.lock);
INIT_LIST_HEAD(&vcpu->arch.local_int.list);
vcpu->arch.local_int.float_int = &kvm->arch.float_int;
- spin_lock(&kvm->arch.float_int.lock);
- kvm->arch.float_int.local_int[id] = &vcpu->arch.local_int;
vcpu->arch.local_int.wq = &vcpu->wq;
vcpu->arch.local_int.cpuflags = &vcpu->arch.sie_block->cpuflags;
- spin_unlock(&kvm->arch.float_int.lock);
rc = kvm_vcpu_init(vcpu, kvm, id);
if (rc)
@@ -952,7 +949,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
atomic_clear_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags);
- BUG_ON(vcpu->kvm->arch.float_int.local_int[vcpu->vcpu_id] == NULL);
+ BUG_ON(kvm_get_vcpu(vcpu->kvm, vcpu->vcpu_id) == NULL);
switch (kvm_run->exit_reason) {
case KVM_EXIT_S390_SIEIC:
diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c
index 466eefa..3fe44c4 100644
--- a/arch/s390/kvm/sigp.c
+++ b/arch/s390/kvm/sigp.c
@@ -23,29 +23,30 @@
static int __sigp_sense(struct kvm_vcpu *vcpu, u16 cpu_addr,
u64 *reg)
{
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
+ struct kvm_s390_local_interrupt *li;
+ struct kvm_vcpu *dst_vcpu = NULL;
+ int cpuflags;
int rc;
if (cpu_addr >= KVM_MAX_VCPUS)
return SIGP_CC_NOT_OPERATIONAL;
- spin_lock(&fi->lock);
- if (fi->local_int[cpu_addr] == NULL)
- rc = SIGP_CC_NOT_OPERATIONAL;
- else if (!(atomic_read(fi->local_int[cpu_addr]->cpuflags)
- & (CPUSTAT_ECALL_PEND | CPUSTAT_STOPPED)))
+ dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
+ if (!dst_vcpu)
+ return SIGP_CC_NOT_OPERATIONAL;
+ li = &dst_vcpu->arch.local_int;
+
+ cpuflags = atomic_read(li->cpuflags);
+ if (!(cpuflags & (CPUSTAT_ECALL_PEND | CPUSTAT_STOPPED)))
rc = SIGP_CC_ORDER_CODE_ACCEPTED;
else {
*reg &= 0xffffffff00000000UL;
- if (atomic_read(fi->local_int[cpu_addr]->cpuflags)
- & CPUSTAT_ECALL_PEND)
+ if (cpuflags & CPUSTAT_ECALL_PEND)
*reg |= SIGP_STATUS_EXT_CALL_PENDING;
- if (atomic_read(fi->local_int[cpu_addr]->cpuflags)
- & CPUSTAT_STOPPED)
+ if (cpuflags & CPUSTAT_STOPPED)
*reg |= SIGP_STATUS_STOPPED;
rc = SIGP_CC_STATUS_STORED;
}
- spin_unlock(&fi->lock);
VCPU_EVENT(vcpu, 4, "sensed status of cpu %x rc %x", cpu_addr, rc);
return rc;
@@ -53,10 +54,9 @@ static int __sigp_sense(struct kvm_vcpu *vcpu, u16 cpu_addr,
static int __sigp_emergency(struct kvm_vcpu *vcpu, u16 cpu_addr)
{
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
struct kvm_s390_local_interrupt *li;
struct kvm_s390_interrupt_info *inti;
- int rc;
+ struct kvm_vcpu *dst_vcpu = NULL;
if (cpu_addr >= KVM_MAX_VCPUS)
return SIGP_CC_NOT_OPERATIONAL;
@@ -68,13 +68,10 @@ static int __sigp_emergency(struct kvm_vcpu *vcpu, u16 cpu_addr)
inti->type = KVM_S390_INT_EMERGENCY;
inti->emerg.code = vcpu->vcpu_id;
- spin_lock(&fi->lock);
- li = fi->local_int[cpu_addr];
- if (li == NULL) {
- rc = SIGP_CC_NOT_OPERATIONAL;
- kfree(inti);
- goto unlock;
- }
+ dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
+ if (!dst_vcpu)
+ return SIGP_CC_NOT_OPERATIONAL;
+ li = &dst_vcpu->arch.local_int;
spin_lock_bh(&li->lock);
list_add_tail(&inti->list, &li->list);
atomic_set(&li->active, 1);
@@ -82,11 +79,9 @@ static int __sigp_emergency(struct kvm_vcpu *vcpu, u16 cpu_addr)
if (waitqueue_active(li->wq))
wake_up_interruptible(li->wq);
spin_unlock_bh(&li->lock);
- rc = SIGP_CC_ORDER_CODE_ACCEPTED;
VCPU_EVENT(vcpu, 4, "sent sigp emerg to cpu %x", cpu_addr);
-unlock:
- spin_unlock(&fi->lock);
- return rc;
+
+ return SIGP_CC_ORDER_CODE_ACCEPTED;
}
static int __sigp_conditional_emergency(struct kvm_vcpu *vcpu, u16 cpu_addr,
@@ -122,10 +117,9 @@ static int __sigp_conditional_emergency(struct kvm_vcpu *vcpu, u16 cpu_addr,
static int __sigp_external_call(struct kvm_vcpu *vcpu, u16 cpu_addr)
{
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
struct kvm_s390_local_interrupt *li;
struct kvm_s390_interrupt_info *inti;
- int rc;
+ struct kvm_vcpu *dst_vcpu = NULL;
if (cpu_addr >= KVM_MAX_VCPUS)
return SIGP_CC_NOT_OPERATIONAL;
@@ -137,13 +131,10 @@ static int __sigp_external_call(struct kvm_vcpu *vcpu, u16 cpu_addr)
inti->type = KVM_S390_INT_EXTERNAL_CALL;
inti->extcall.code = vcpu->vcpu_id;
- spin_lock(&fi->lock);
- li = fi->local_int[cpu_addr];
- if (li == NULL) {
- rc = SIGP_CC_NOT_OPERATIONAL;
- kfree(inti);
- goto unlock;
- }
+ dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
+ if (!dst_vcpu)
+ return SIGP_CC_NOT_OPERATIONAL;
+ li = &dst_vcpu->arch.local_int;
spin_lock_bh(&li->lock);
list_add_tail(&inti->list, &li->list);
atomic_set(&li->active, 1);
@@ -151,11 +142,9 @@ static int __sigp_external_call(struct kvm_vcpu *vcpu, u16 cpu_addr)
if (waitqueue_active(li->wq))
wake_up_interruptible(li->wq);
spin_unlock_bh(&li->lock);
- rc = SIGP_CC_ORDER_CODE_ACCEPTED;
VCPU_EVENT(vcpu, 4, "sent sigp ext call to cpu %x", cpu_addr);
-unlock:
- spin_unlock(&fi->lock);
- return rc;
+
+ return SIGP_CC_ORDER_CODE_ACCEPTED;
}
static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action)
@@ -189,31 +178,26 @@ out:
static int __sigp_stop(struct kvm_vcpu *vcpu, u16 cpu_addr, int action)
{
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
struct kvm_s390_local_interrupt *li;
+ struct kvm_vcpu *dst_vcpu = NULL;
int rc;
if (cpu_addr >= KVM_MAX_VCPUS)
return SIGP_CC_NOT_OPERATIONAL;
- spin_lock(&fi->lock);
- li = fi->local_int[cpu_addr];
- if (li == NULL) {
- rc = SIGP_CC_NOT_OPERATIONAL;
- goto unlock;
- }
+ 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);
-unlock:
- spin_unlock(&fi->lock);
VCPU_EVENT(vcpu, 4, "sent sigp stop to cpu %x", cpu_addr);
if ((action & ACTION_STORE_ON_STOP) != 0 && rc == -ESHUTDOWN) {
/* If the CPU has already been stopped, we still have
* to save the status when doing stop-and-store. This
* has to be done after unlocking all spinlocks. */
- struct kvm_vcpu *dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
rc = kvm_s390_store_status_unloaded(dst_vcpu,
KVM_S390_STORE_STATUS_NOADDR);
}
@@ -333,28 +317,26 @@ static int __sigp_store_status_at_addr(struct kvm_vcpu *vcpu, u16 cpu_id,
static int __sigp_sense_running(struct kvm_vcpu *vcpu, u16 cpu_addr,
u64 *reg)
{
+ struct kvm_s390_local_interrupt *li;
+ struct kvm_vcpu *dst_vcpu = NULL;
int rc;
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
if (cpu_addr >= KVM_MAX_VCPUS)
return SIGP_CC_NOT_OPERATIONAL;
- spin_lock(&fi->lock);
- if (fi->local_int[cpu_addr] == NULL)
- rc = SIGP_CC_NOT_OPERATIONAL;
- else {
- if (atomic_read(fi->local_int[cpu_addr]->cpuflags)
- & CPUSTAT_RUNNING) {
- /* running */
- rc = SIGP_CC_ORDER_CODE_ACCEPTED;
- } else {
- /* not running */
- *reg &= 0xffffffff00000000UL;
- *reg |= SIGP_STATUS_NOT_RUNNING;
- rc = SIGP_CC_STATUS_STORED;
- }
+ dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
+ if (!dst_vcpu)
+ return SIGP_CC_NOT_OPERATIONAL;
+ li = &dst_vcpu->arch.local_int;
+ if (atomic_read(li->cpuflags) & CPUSTAT_RUNNING) {
+ /* running */
+ rc = SIGP_CC_ORDER_CODE_ACCEPTED;
+ } else {
+ /* not running */
+ *reg &= 0xffffffff00000000UL;
+ *reg |= SIGP_STATUS_NOT_RUNNING;
+ rc = SIGP_CC_STATUS_STORED;
}
- spin_unlock(&fi->lock);
VCPU_EVENT(vcpu, 4, "sensed running status of cpu %x rc %x", cpu_addr,
rc);
@@ -365,26 +347,22 @@ static int __sigp_sense_running(struct kvm_vcpu *vcpu, u16 cpu_addr,
/* Test whether the destination CPU is available and not busy */
static int sigp_check_callable(struct kvm_vcpu *vcpu, u16 cpu_addr)
{
- struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
struct kvm_s390_local_interrupt *li;
int rc = SIGP_CC_ORDER_CODE_ACCEPTED;
+ struct kvm_vcpu *dst_vcpu = NULL;
if (cpu_addr >= KVM_MAX_VCPUS)
return SIGP_CC_NOT_OPERATIONAL;
- spin_lock(&fi->lock);
- li = fi->local_int[cpu_addr];
- if (li == NULL) {
- rc = SIGP_CC_NOT_OPERATIONAL;
- goto out;
- }
-
+ dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr);
+ if (!dst_vcpu)
+ return SIGP_CC_NOT_OPERATIONAL;
+ li = &dst_vcpu->arch.local_int;
spin_lock_bh(&li->lock);
if (li->action_bits & ACTION_STOP_ON_STOP)
rc = SIGP_CC_BUSY;
spin_unlock_bh(&li->lock);
-out:
- spin_unlock(&fi->lock);
+
return rc;
}
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* [PULL 6/8] virtio_ccw: fix vcdev pointer handling issues
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
` (4 preceding siblings ...)
2014-03-04 11:27 ` [PULL 5/8] KVM: s390: get rid of local_int array Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:27 ` [PULL 7/8] s390/airq: add support for irq ranges Christian Borntraeger
` (2 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
From: Heinz Graalfs <graalfs@linux.vnet.ibm.com>
The interrupt handler virtio_ccw_int_handler() using the vcdev pointer
is protected by the ccw_device lock. Resetting the pointer within the
ccw_device structure should be done when holding this lock.
Also resetting the vcdev pointer (under the ccw_device lock) prior to
freeing the vcdev pointer memory removes a critical path.
Signed-off-by: Heinz Graalfs <graalfs@linux.vnet.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
drivers/s390/kvm/virtio_ccw.c | 35 ++++++++++++++++++++++++++++-------
1 file changed, 28 insertions(+), 7 deletions(-)
diff --git a/drivers/s390/kvm/virtio_ccw.c b/drivers/s390/kvm/virtio_ccw.c
index 0fc5848..413c630 100644
--- a/drivers/s390/kvm/virtio_ccw.c
+++ b/drivers/s390/kvm/virtio_ccw.c
@@ -636,6 +636,8 @@ static void virtio_ccw_int_handler(struct ccw_device *cdev,
struct virtqueue *vq;
struct virtio_driver *drv;
+ if (!vcdev)
+ return;
/* Check if it's a notification from the host. */
if ((intparm == 0) &&
(scsw_stctl(&irb->scsw) ==
@@ -734,23 +736,37 @@ static int virtio_ccw_probe(struct ccw_device *cdev)
return 0;
}
+static struct virtio_ccw_device *virtio_grab_drvdata(struct ccw_device *cdev)
+{
+ unsigned long flags;
+ struct virtio_ccw_device *vcdev;
+
+ spin_lock_irqsave(get_ccwdev_lock(cdev), flags);
+ vcdev = dev_get_drvdata(&cdev->dev);
+ if (!vcdev) {
+ spin_unlock_irqrestore(get_ccwdev_lock(cdev), flags);
+ return NULL;
+ }
+ dev_set_drvdata(&cdev->dev, NULL);
+ spin_unlock_irqrestore(get_ccwdev_lock(cdev), flags);
+ return vcdev;
+}
+
static void virtio_ccw_remove(struct ccw_device *cdev)
{
- struct virtio_ccw_device *vcdev = dev_get_drvdata(&cdev->dev);
+ struct virtio_ccw_device *vcdev = virtio_grab_drvdata(cdev);
- if (cdev->online) {
+ if (vcdev && cdev->online)
unregister_virtio_device(&vcdev->vdev);
- dev_set_drvdata(&cdev->dev, NULL);
- }
cdev->handler = NULL;
}
static int virtio_ccw_offline(struct ccw_device *cdev)
{
- struct virtio_ccw_device *vcdev = dev_get_drvdata(&cdev->dev);
+ struct virtio_ccw_device *vcdev = virtio_grab_drvdata(cdev);
- unregister_virtio_device(&vcdev->vdev);
- dev_set_drvdata(&cdev->dev, NULL);
+ if (vcdev)
+ unregister_virtio_device(&vcdev->vdev);
return 0;
}
@@ -759,6 +775,7 @@ static int virtio_ccw_online(struct ccw_device *cdev)
{
int ret;
struct virtio_ccw_device *vcdev;
+ unsigned long flags;
vcdev = kzalloc(sizeof(*vcdev), GFP_KERNEL);
if (!vcdev) {
@@ -786,7 +803,9 @@ static int virtio_ccw_online(struct ccw_device *cdev)
INIT_LIST_HEAD(&vcdev->virtqueues);
spin_lock_init(&vcdev->lock);
+ spin_lock_irqsave(get_ccwdev_lock(cdev), flags);
dev_set_drvdata(&cdev->dev, vcdev);
+ spin_unlock_irqrestore(get_ccwdev_lock(cdev), flags);
vcdev->vdev.id.vendor = cdev->id.cu_type;
vcdev->vdev.id.device = cdev->id.cu_model;
ret = register_virtio_device(&vcdev->vdev);
@@ -797,7 +816,9 @@ static int virtio_ccw_online(struct ccw_device *cdev)
}
return 0;
out_put:
+ spin_lock_irqsave(get_ccwdev_lock(cdev), flags);
dev_set_drvdata(&cdev->dev, NULL);
+ spin_unlock_irqrestore(get_ccwdev_lock(cdev), flags);
put_device(&vcdev->vdev.dev);
return ret;
out_free:
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* [PULL 7/8] s390/airq: add support for irq ranges
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
` (5 preceding siblings ...)
2014-03-04 11:27 ` [PULL 6/8] virtio_ccw: fix vcdev pointer handling issues Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:27 ` [PULL 8/8] virtio-ccw: virtio-ccw adapter interrupt support Christian Borntraeger
2014-03-04 11:34 ` [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Paolo Bonzini
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add airq_iv_alloc and airq_iv_free to allocate and free consecutive
ranges of irqs from the interrupt vector.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
arch/s390/include/asm/airq.h | 14 ++++++++--
drivers/s390/cio/airq.c | 66 +++++++++++++++++++++++++++-----------------
2 files changed, 53 insertions(+), 27 deletions(-)
diff --git a/arch/s390/include/asm/airq.h b/arch/s390/include/asm/airq.h
index 4bbb595..bd93ff6 100644
--- a/arch/s390/include/asm/airq.h
+++ b/arch/s390/include/asm/airq.h
@@ -44,11 +44,21 @@ struct airq_iv {
struct airq_iv *airq_iv_create(unsigned long bits, unsigned long flags);
void airq_iv_release(struct airq_iv *iv);
-unsigned long airq_iv_alloc_bit(struct airq_iv *iv);
-void airq_iv_free_bit(struct airq_iv *iv, unsigned long bit);
+unsigned long airq_iv_alloc(struct airq_iv *iv, unsigned long num);
+void airq_iv_free(struct airq_iv *iv, unsigned long bit, unsigned long num);
unsigned long airq_iv_scan(struct airq_iv *iv, unsigned long start,
unsigned long end);
+static inline unsigned long airq_iv_alloc_bit(struct airq_iv *iv)
+{
+ return airq_iv_alloc(iv, 1);
+}
+
+static inline void airq_iv_free_bit(struct airq_iv *iv, unsigned long bit)
+{
+ airq_iv_free(iv, bit, 1);
+}
+
static inline unsigned long airq_iv_end(struct airq_iv *iv)
{
return iv->end;
diff --git a/drivers/s390/cio/airq.c b/drivers/s390/cio/airq.c
index f055df0..445564c 100644
--- a/drivers/s390/cio/airq.c
+++ b/drivers/s390/cio/airq.c
@@ -186,55 +186,71 @@ void airq_iv_release(struct airq_iv *iv)
EXPORT_SYMBOL(airq_iv_release);
/**
- * airq_iv_alloc_bit - allocate an irq bit from an interrupt vector
+ * airq_iv_alloc - allocate irq bits from an interrupt vector
* @iv: pointer to an interrupt vector structure
+ * @num: number of consecutive irq bits to allocate
*
- * Returns the bit number of the allocated irq, or -1UL if no bit
- * is available or the AIRQ_IV_ALLOC flag has not been specified
+ * Returns the bit number of the first irq in the allocated block of irqs,
+ * or -1UL if no bit is available or the AIRQ_IV_ALLOC flag has not been
+ * specified
*/
-unsigned long airq_iv_alloc_bit(struct airq_iv *iv)
+unsigned long airq_iv_alloc(struct airq_iv *iv, unsigned long num)
{
- unsigned long bit;
+ unsigned long bit, i;
- if (!iv->avail)
+ if (!iv->avail || num == 0)
return -1UL;
spin_lock(&iv->lock);
bit = find_first_bit_inv(iv->avail, iv->bits);
- if (bit < iv->bits) {
- clear_bit_inv(bit, iv->avail);
- if (bit >= iv->end)
- iv->end = bit + 1;
- } else
+ while (bit + num <= iv->bits) {
+ for (i = 1; i < num; i++)
+ if (!test_bit_inv(bit + i, iv->avail))
+ break;
+ if (i >= num) {
+ /* Found a suitable block of irqs */
+ for (i = 0; i < num; i++)
+ clear_bit_inv(bit + i, iv->avail);
+ if (bit + num >= iv->end)
+ iv->end = bit + num + 1;
+ break;
+ }
+ bit = find_next_bit_inv(iv->avail, iv->bits, bit + i + 1);
+ }
+ if (bit + num > iv->bits)
bit = -1UL;
spin_unlock(&iv->lock);
return bit;
}
-EXPORT_SYMBOL(airq_iv_alloc_bit);
+EXPORT_SYMBOL(airq_iv_alloc);
/**
- * airq_iv_free_bit - free an irq bit of an interrupt vector
+ * airq_iv_free - free irq bits of an interrupt vector
* @iv: pointer to interrupt vector structure
- * @bit: number of the irq bit to free
+ * @bit: number of the first irq bit to free
+ * @num: number of consecutive irq bits to free
*/
-void airq_iv_free_bit(struct airq_iv *iv, unsigned long bit)
+void airq_iv_free(struct airq_iv *iv, unsigned long bit, unsigned long num)
{
- if (!iv->avail)
+ unsigned long i;
+
+ if (!iv->avail || num == 0)
return;
spin_lock(&iv->lock);
- /* Clear (possibly left over) interrupt bit */
- clear_bit_inv(bit, iv->vector);
- /* Make the bit position available again */
- set_bit_inv(bit, iv->avail);
- if (bit == iv->end - 1) {
+ for (i = 0; i < num; i++) {
+ /* Clear (possibly left over) interrupt bit */
+ clear_bit_inv(bit + i, iv->vector);
+ /* Make the bit positions available again */
+ set_bit_inv(bit + i, iv->avail);
+ }
+ if (bit + num >= iv->end) {
/* Find new end of bit-field */
- while (--iv->end > 0)
- if (!test_bit_inv(iv->end - 1, iv->avail))
- break;
+ while (iv->end > 0 && !test_bit_inv(iv->end - 1, iv->avail))
+ iv->end--;
}
spin_unlock(&iv->lock);
}
-EXPORT_SYMBOL(airq_iv_free_bit);
+EXPORT_SYMBOL(airq_iv_free);
/**
* airq_iv_scan - scan interrupt vector for non-zero bits
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* [PULL 8/8] virtio-ccw: virtio-ccw adapter interrupt support.
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
` (6 preceding siblings ...)
2014-03-04 11:27 ` [PULL 7/8] s390/airq: add support for irq ranges Christian Borntraeger
@ 2014-03-04 11:27 ` Christian Borntraeger
2014-03-04 11:34 ` [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Paolo Bonzini
8 siblings, 0 replies; 10+ messages in thread
From: Christian Borntraeger @ 2014-03-04 11:27 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann, Christian Borntraeger
From: Cornelia Huck <cornelia.huck@de.ibm.com>
Implement the new CCW_CMD_SET_IND_ADAPTER command and try to enable
adapter interrupts for every device on the first startup. If the host
does not support adapter interrupts, fall back to normal I/O interrupts.
virtio-ccw adapter interrupts use the same isc as normal I/O subchannels
and share a summary indicator for all devices sharing the same indicator
area.
Indicator bits for the individual virtqueues may be contained in the same
indicator area for different devices.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
---
arch/s390/include/asm/irq.h | 1 +
arch/s390/kernel/irq.c | 1 +
drivers/s390/kvm/virtio_ccw.c | 278 ++++++++++++++++++++++++++++++++++++++++--
3 files changed, 270 insertions(+), 10 deletions(-)
diff --git a/arch/s390/include/asm/irq.h b/arch/s390/include/asm/irq.h
index 5f8bcc5..35f0faa 100644
--- a/arch/s390/include/asm/irq.h
+++ b/arch/s390/include/asm/irq.h
@@ -53,6 +53,7 @@ enum interruption_class {
IRQIO_PCI,
IRQIO_MSI,
IRQIO_VIR,
+ IRQIO_VAI,
NMI_NMI,
CPU_RST,
NR_ARCH_IRQS
diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c
index bb27a26..c288ef7 100644
--- a/arch/s390/kernel/irq.c
+++ b/arch/s390/kernel/irq.c
@@ -84,6 +84,7 @@ static const struct irq_class irqclass_sub_desc[NR_ARCH_IRQS] = {
[IRQIO_PCI] = {.name = "PCI", .desc = "[I/O] PCI Interrupt" },
[IRQIO_MSI] = {.name = "MSI", .desc = "[I/O] MSI Interrupt" },
[IRQIO_VIR] = {.name = "VIR", .desc = "[I/O] Virtual I/O Devices"},
+ [IRQIO_VAI] = {.name = "VAI", .desc = "[I/O] Virtual I/O Devices AI"},
[NMI_NMI] = {.name = "NMI", .desc = "[NMI] Machine Check"},
[CPU_RST] = {.name = "RST", .desc = "[CPU] CPU Restart"},
};
diff --git a/drivers/s390/kvm/virtio_ccw.c b/drivers/s390/kvm/virtio_ccw.c
index 413c630..6a2b5fd 100644
--- a/drivers/s390/kvm/virtio_ccw.c
+++ b/drivers/s390/kvm/virtio_ccw.c
@@ -1,7 +1,7 @@
/*
* ccw based virtio transport
*
- * Copyright IBM Corp. 2012
+ * Copyright IBM Corp. 2012, 2014
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License (version 2 only)
@@ -32,6 +32,8 @@
#include <asm/cio.h>
#include <asm/ccwdev.h>
#include <asm/virtio-ccw.h>
+#include <asm/isc.h>
+#include <asm/airq.h>
/*
* virtio related functions
@@ -58,6 +60,8 @@ struct virtio_ccw_device {
unsigned long indicators;
unsigned long indicators2;
struct vq_config_block *config_block;
+ bool is_thinint;
+ void *airq_info;
};
struct vq_info_block {
@@ -72,15 +76,38 @@ struct virtio_feature_desc {
__u8 index;
} __packed;
+struct virtio_thinint_area {
+ unsigned long summary_indicator;
+ unsigned long indicator;
+ u64 bit_nr;
+ u8 isc;
+} __packed;
+
struct virtio_ccw_vq_info {
struct virtqueue *vq;
int num;
void *queue;
struct vq_info_block *info_block;
+ int bit_nr;
struct list_head node;
long cookie;
};
+#define VIRTIO_AIRQ_ISC IO_SCH_ISC /* inherit from subchannel */
+
+#define VIRTIO_IV_BITS (L1_CACHE_BYTES * 8)
+#define MAX_AIRQ_AREAS 20
+
+static int virtio_ccw_use_airq = 1;
+
+struct airq_info {
+ rwlock_t lock;
+ u8 summary_indicator;
+ struct airq_struct airq;
+ struct airq_iv *aiv;
+};
+static struct airq_info *airq_areas[MAX_AIRQ_AREAS];
+
#define CCW_CMD_SET_VQ 0x13
#define CCW_CMD_VDEV_RESET 0x33
#define CCW_CMD_SET_IND 0x43
@@ -91,6 +118,7 @@ struct virtio_ccw_vq_info {
#define CCW_CMD_WRITE_CONF 0x21
#define CCW_CMD_WRITE_STATUS 0x31
#define CCW_CMD_READ_VQ_CONF 0x32
+#define CCW_CMD_SET_IND_ADAPTER 0x73
#define VIRTIO_CCW_DOING_SET_VQ 0x00010000
#define VIRTIO_CCW_DOING_RESET 0x00040000
@@ -102,6 +130,7 @@ struct virtio_ccw_vq_info {
#define VIRTIO_CCW_DOING_SET_IND 0x01000000
#define VIRTIO_CCW_DOING_READ_VQ_CONF 0x02000000
#define VIRTIO_CCW_DOING_SET_CONF_IND 0x04000000
+#define VIRTIO_CCW_DOING_SET_IND_ADAPTER 0x08000000
#define VIRTIO_CCW_INTPARM_MASK 0xffff0000
static struct virtio_ccw_device *to_vc_device(struct virtio_device *vdev)
@@ -109,6 +138,125 @@ static struct virtio_ccw_device *to_vc_device(struct virtio_device *vdev)
return container_of(vdev, struct virtio_ccw_device, vdev);
}
+static void drop_airq_indicator(struct virtqueue *vq, struct airq_info *info)
+{
+ unsigned long i, flags;
+
+ write_lock_irqsave(&info->lock, flags);
+ for (i = 0; i < airq_iv_end(info->aiv); i++) {
+ if (vq == (void *)airq_iv_get_ptr(info->aiv, i)) {
+ airq_iv_free_bit(info->aiv, i);
+ airq_iv_set_ptr(info->aiv, i, 0);
+ break;
+ }
+ }
+ write_unlock_irqrestore(&info->lock, flags);
+}
+
+static void virtio_airq_handler(struct airq_struct *airq)
+{
+ struct airq_info *info = container_of(airq, struct airq_info, airq);
+ unsigned long ai;
+
+ inc_irq_stat(IRQIO_VAI);
+ read_lock(&info->lock);
+ /* Walk through indicators field, summary indicator active. */
+ for (ai = 0;;) {
+ ai = airq_iv_scan(info->aiv, ai, airq_iv_end(info->aiv));
+ if (ai == -1UL)
+ break;
+ vring_interrupt(0, (void *)airq_iv_get_ptr(info->aiv, ai));
+ }
+ info->summary_indicator = 0;
+ smp_wmb();
+ /* Walk through indicators field, summary indicator not active. */
+ for (ai = 0;;) {
+ ai = airq_iv_scan(info->aiv, ai, airq_iv_end(info->aiv));
+ if (ai == -1UL)
+ break;
+ vring_interrupt(0, (void *)airq_iv_get_ptr(info->aiv, ai));
+ }
+ read_unlock(&info->lock);
+}
+
+static struct airq_info *new_airq_info(void)
+{
+ struct airq_info *info;
+ int rc;
+
+ info = kzalloc(sizeof(*info), GFP_KERNEL);
+ if (!info)
+ return NULL;
+ rwlock_init(&info->lock);
+ info->aiv = airq_iv_create(VIRTIO_IV_BITS, AIRQ_IV_ALLOC | AIRQ_IV_PTR);
+ if (!info->aiv) {
+ kfree(info);
+ return NULL;
+ }
+ info->airq.handler = virtio_airq_handler;
+ info->airq.lsi_ptr = &info->summary_indicator;
+ info->airq.lsi_mask = 0xff;
+ info->airq.isc = VIRTIO_AIRQ_ISC;
+ rc = register_adapter_interrupt(&info->airq);
+ if (rc) {
+ airq_iv_release(info->aiv);
+ kfree(info);
+ return NULL;
+ }
+ return info;
+}
+
+static void destroy_airq_info(struct airq_info *info)
+{
+ if (!info)
+ return;
+
+ unregister_adapter_interrupt(&info->airq);
+ airq_iv_release(info->aiv);
+ kfree(info);
+}
+
+static unsigned long get_airq_indicator(struct virtqueue *vqs[], int nvqs,
+ u64 *first, void **airq_info)
+{
+ int i, j;
+ struct airq_info *info;
+ unsigned long indicator_addr = 0;
+ unsigned long bit, flags;
+
+ for (i = 0; i < MAX_AIRQ_AREAS && !indicator_addr; i++) {
+ if (!airq_areas[i])
+ airq_areas[i] = new_airq_info();
+ info = airq_areas[i];
+ if (!info)
+ return 0;
+ write_lock_irqsave(&info->lock, flags);
+ bit = airq_iv_alloc(info->aiv, nvqs);
+ if (bit == -1UL) {
+ /* Not enough vacancies. */
+ write_unlock_irqrestore(&info->lock, flags);
+ continue;
+ }
+ *first = bit;
+ *airq_info = info;
+ indicator_addr = (unsigned long)info->aiv->vector;
+ for (j = 0; j < nvqs; j++) {
+ airq_iv_set_ptr(info->aiv, bit + j,
+ (unsigned long)vqs[j]);
+ }
+ write_unlock_irqrestore(&info->lock, flags);
+ }
+ return indicator_addr;
+}
+
+static void virtio_ccw_drop_indicators(struct virtio_ccw_device *vcdev)
+{
+ struct virtio_ccw_vq_info *info;
+
+ list_for_each_entry(info, &vcdev->virtqueues, node)
+ drop_airq_indicator(info->vq, vcdev->airq_info);
+}
+
static int doing_io(struct virtio_ccw_device *vcdev, __u32 flag)
{
unsigned long flags;
@@ -145,6 +293,51 @@ static int ccw_io_helper(struct virtio_ccw_device *vcdev,
return ret ? ret : vcdev->err;
}
+static void virtio_ccw_drop_indicator(struct virtio_ccw_device *vcdev,
+ struct ccw1 *ccw)
+{
+ int ret;
+ unsigned long *indicatorp = NULL;
+ struct virtio_thinint_area *thinint_area = NULL;
+ struct airq_info *airq_info = vcdev->airq_info;
+
+ if (vcdev->is_thinint) {
+ thinint_area = kzalloc(sizeof(*thinint_area),
+ GFP_DMA | GFP_KERNEL);
+ if (!thinint_area)
+ return;
+ thinint_area->summary_indicator =
+ (unsigned long) &airq_info->summary_indicator;
+ thinint_area->isc = VIRTIO_AIRQ_ISC;
+ ccw->cmd_code = CCW_CMD_SET_IND_ADAPTER;
+ ccw->count = sizeof(*thinint_area);
+ ccw->cda = (__u32)(unsigned long) thinint_area;
+ } else {
+ indicatorp = kmalloc(sizeof(&vcdev->indicators),
+ GFP_DMA | GFP_KERNEL);
+ if (!indicatorp)
+ return;
+ *indicatorp = 0;
+ ccw->cmd_code = CCW_CMD_SET_IND;
+ ccw->count = sizeof(vcdev->indicators);
+ ccw->cda = (__u32)(unsigned long) indicatorp;
+ }
+ /* Deregister indicators from host. */
+ vcdev->indicators = 0;
+ ccw->flags = 0;
+ ret = ccw_io_helper(vcdev, ccw,
+ vcdev->is_thinint ?
+ VIRTIO_CCW_DOING_SET_IND_ADAPTER :
+ VIRTIO_CCW_DOING_SET_IND);
+ if (ret && (ret != -ENODEV))
+ dev_info(&vcdev->cdev->dev,
+ "Failed to deregister indicators (%d)\n", ret);
+ else if (vcdev->is_thinint)
+ virtio_ccw_drop_indicators(vcdev);
+ kfree(indicatorp);
+ kfree(thinint_area);
+}
+
static inline long do_kvm_notify(struct subchannel_id schid,
unsigned long queue_index,
long cookie)
@@ -232,11 +425,13 @@ static void virtio_ccw_del_vqs(struct virtio_device *vdev)
{
struct virtqueue *vq, *n;
struct ccw1 *ccw;
+ struct virtio_ccw_device *vcdev = to_vc_device(vdev);
ccw = kzalloc(sizeof(*ccw), GFP_DMA | GFP_KERNEL);
if (!ccw)
return;
+ virtio_ccw_drop_indicator(vcdev, ccw);
list_for_each_entry_safe(vq, n, &vdev->vqs, list)
virtio_ccw_del_vq(vq, ccw);
@@ -326,6 +521,54 @@ out_err:
return ERR_PTR(err);
}
+static int virtio_ccw_register_adapter_ind(struct virtio_ccw_device *vcdev,
+ struct virtqueue *vqs[], int nvqs,
+ struct ccw1 *ccw)
+{
+ int ret;
+ struct virtio_thinint_area *thinint_area = NULL;
+ struct airq_info *info;
+
+ thinint_area = kzalloc(sizeof(*thinint_area), GFP_DMA | GFP_KERNEL);
+ if (!thinint_area) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ /* Try to get an indicator. */
+ thinint_area->indicator = get_airq_indicator(vqs, nvqs,
+ &thinint_area->bit_nr,
+ &vcdev->airq_info);
+ if (!thinint_area->indicator) {
+ ret = -ENOSPC;
+ goto out;
+ }
+ info = vcdev->airq_info;
+ thinint_area->summary_indicator =
+ (unsigned long) &info->summary_indicator;
+ thinint_area->isc = VIRTIO_AIRQ_ISC;
+ ccw->cmd_code = CCW_CMD_SET_IND_ADAPTER;
+ ccw->flags = CCW_FLAG_SLI;
+ ccw->count = sizeof(*thinint_area);
+ ccw->cda = (__u32)(unsigned long)thinint_area;
+ ret = ccw_io_helper(vcdev, ccw, VIRTIO_CCW_DOING_SET_IND_ADAPTER);
+ if (ret) {
+ if (ret == -EOPNOTSUPP) {
+ /*
+ * The host does not support adapter interrupts
+ * for virtio-ccw, stop trying.
+ */
+ virtio_ccw_use_airq = 0;
+ pr_info("Adapter interrupts unsupported on host\n");
+ } else
+ dev_warn(&vcdev->cdev->dev,
+ "enabling adapter interrupts = %d\n", ret);
+ virtio_ccw_drop_indicators(vcdev);
+ }
+out:
+ kfree(thinint_area);
+ return ret;
+}
+
static int virtio_ccw_find_vqs(struct virtio_device *vdev, unsigned nvqs,
struct virtqueue *vqs[],
vq_callback_t *callbacks[],
@@ -355,15 +598,23 @@ static int virtio_ccw_find_vqs(struct virtio_device *vdev, unsigned nvqs,
if (!indicatorp)
goto out;
*indicatorp = (unsigned long) &vcdev->indicators;
- /* Register queue indicators with host. */
- vcdev->indicators = 0;
- ccw->cmd_code = CCW_CMD_SET_IND;
- ccw->flags = 0;
- ccw->count = sizeof(vcdev->indicators);
- ccw->cda = (__u32)(unsigned long) indicatorp;
- ret = ccw_io_helper(vcdev, ccw, VIRTIO_CCW_DOING_SET_IND);
- if (ret)
- goto out;
+ if (vcdev->is_thinint) {
+ ret = virtio_ccw_register_adapter_ind(vcdev, vqs, nvqs, ccw);
+ if (ret)
+ /* no error, just fall back to legacy interrupts */
+ vcdev->is_thinint = 0;
+ }
+ if (!vcdev->is_thinint) {
+ /* Register queue indicators with host. */
+ vcdev->indicators = 0;
+ ccw->cmd_code = CCW_CMD_SET_IND;
+ ccw->flags = 0;
+ ccw->count = sizeof(vcdev->indicators);
+ ccw->cda = (__u32)(unsigned long) indicatorp;
+ ret = ccw_io_helper(vcdev, ccw, VIRTIO_CCW_DOING_SET_IND);
+ if (ret)
+ goto out;
+ }
/* Register indicators2 with host for config changes */
*indicatorp = (unsigned long) &vcdev->indicators2;
vcdev->indicators2 = 0;
@@ -665,6 +916,7 @@ static void virtio_ccw_int_handler(struct ccw_device *cdev,
case VIRTIO_CCW_DOING_SET_CONF_IND:
case VIRTIO_CCW_DOING_RESET:
case VIRTIO_CCW_DOING_READ_VQ_CONF:
+ case VIRTIO_CCW_DOING_SET_IND_ADAPTER:
vcdev->curr_io &= ~activity;
wake_up(&vcdev->wait_q);
break;
@@ -795,6 +1047,8 @@ static int virtio_ccw_online(struct ccw_device *cdev)
goto out_free;
}
+ vcdev->is_thinint = virtio_ccw_use_airq; /* at least try */
+
vcdev->vdev.dev.parent = &cdev->dev;
vcdev->vdev.dev.release = virtio_ccw_release_dev;
vcdev->vdev.config = &virtio_ccw_config_ops;
@@ -956,6 +1210,10 @@ module_init(virtio_ccw_init);
static void __exit virtio_ccw_exit(void)
{
+ int i;
+
ccw_driver_unregister(&virtio_ccw_driver);
+ for (i = 0; i < MAX_AIRQ_AREAS; i++)
+ destroy_airq_info(airq_areas[i]);
}
module_exit(virtio_ccw_exit);
--
1.8.4.2
^ permalink raw reply related [flat|nested] 10+ messages in thread* Re: [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes
2014-03-04 11:27 [PULL 0/8] KVM: s390,virtio-ccw: Features, cleanups and fixes Christian Borntraeger
` (7 preceding siblings ...)
2014-03-04 11:27 ` [PULL 8/8] virtio-ccw: virtio-ccw adapter interrupt support Christian Borntraeger
@ 2014-03-04 11:34 ` Paolo Bonzini
8 siblings, 0 replies; 10+ messages in thread
From: Paolo Bonzini @ 2014-03-04 11:34 UTC (permalink / raw)
To: Christian Borntraeger
Cc: Gleb Natapov, KVM, linux-s390, Cornelia Huck, Martin Schwidefsky,
Heinz Graalfs, Thomas Huth, Jens Freimann
Il 04/03/2014 12:27, Christian Borntraeger ha scritto:
> Paolo,
>
> The following changes since commit d3714010c307d26df251c45be9cd12ab6d41f0c4:
>
> KVM: x86: emulator_cmpxchg_emulated should mark_page_dirty (2014-02-26 10:11:08 +0100)
>
> are available in the git repository at:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/kvms390/linux.git tags/kvm-s390-20140304
>
> for you to fetch changes up to 96b14536d935848cffd904f583f67c66169002d8:
>
> virtio-ccw: virtio-ccw adapter interrupt support. (2014-03-04 10:41:04 +0100)
>
> The pull target is kvm/queue.
>
> ----------------------------------------------------------------
> this is the next round of patches for KVM: s390:
> - It contains the guest kernel part for adapter interrupt support
> for virtio-ccw, no irqfd yet
> - cleanups, getting rid of some interrupt data structure
> - ONEREG access to GBEA and PP
> - Fixes for virtio-ccw, sigp
>
> ----------------------------------------------------------------
> Christian Borntraeger (2):
> KVM: s390: Provide access to program parameter
> KVM: s390: expose gbea register to userspace
>
> Cornelia Huck (1):
> virtio-ccw: virtio-ccw adapter interrupt support.
>
> Heinz Graalfs (1):
> virtio_ccw: fix vcdev pointer handling issues
>
> Jens Freimann (2):
> KVM: s390: Simplify online vcpus counting for stsi
> KVM: s390: get rid of local_int array
>
> Martin Schwidefsky (1):
> s390/airq: add support for irq ranges
>
> Michael Mueller (3):
> KVM: s390: implementation of kvm_arch_vcpu_runnable()
> KVM: add kvm_arch_vcpu_runnable() test to kvm_vcpu_on_spin() loop
> KVM/s390: Set preempted flag during vcpu wakeup and interrupt delivery
>
> Takuya Yoshikawa (1):
> KVM: x86: Break kvm_for_each_vcpu loop after finding the VP_INDEX
>
> Thomas Huth (1):
> KVM: s390: Fixed CC of SIGP SET_PREFIX handler
>
> arch/s390/include/asm/airq.h | 14 +-
> arch/s390/include/asm/irq.h | 1 +
> arch/s390/include/asm/kvm_host.h | 5 +-
> arch/s390/include/uapi/asm/kvm.h | 2 +
> arch/s390/kernel/irq.c | 1 +
> arch/s390/kvm/interrupt.c | 9 +-
> arch/s390/kvm/kvm-s390.c | 26 +++-
> arch/s390/kvm/priv.c | 7 +-
> arch/s390/kvm/sigp.c | 148 ++++++++----------
> arch/x86/kvm/x86.c | 7 +-
> drivers/s390/cio/airq.c | 66 +++++----
> drivers/s390/kvm/virtio_ccw.c | 313 ++++++++++++++++++++++++++++++++++++---
> virt/kvm/kvm_main.c | 2 +-
> 13 files changed, 448 insertions(+), 153 deletions(-)
>
Pulled to kvm/queue, thanks.
Paolo
^ permalink raw reply [flat|nested] 10+ messages in thread