* [GIT PULL 4/8] KVM: s390: deliver floating interrupts in order of priority
2015-03-31 13:01 [GIT PULL 0/8] KVM: s390: Features and fixes for 4.1 (kvm/next) Christian Borntraeger
` (2 preceding siblings ...)
2015-03-31 13:01 ` [GIT PULL 3/8] KVM: s390: fix get_all_floating_irqs Christian Borntraeger
@ 2015-03-31 13:01 ` Christian Borntraeger
2015-03-31 13:02 ` [GIT PULL 5/8] KVM: s390: cpu timer irq priority Christian Borntraeger
` (4 subsequent siblings)
8 siblings, 0 replies; 13+ messages in thread
From: Christian Borntraeger @ 2015-03-31 13:01 UTC (permalink / raw)
To: KVM
Cc: Paolo Bonzini, Marcelo Tosatti, Alexander Graf, Cornelia Huck,
Jens Freimann, linux-s390, Christian Borntraeger
From: Jens Freimann <jfrei@linux.vnet.ibm.com>
This patch makes interrupt handling compliant to the z/Architecture
Principles of Operation with regard to interrupt priorities.
Add a bitmap for pending floating interrupts. Each bit relates to a
interrupt type and its list. A turned on bit indicates that a list
contains items (interrupts) which need to be delivered. When delivering
interrupts on a cpu we can merge the existing bitmap for cpu-local
interrupts and floating interrupts and have a single mechanism for
delivery.
Currently we have one list for all kinds of floating interrupts and a
corresponding spin lock. This patch adds a separate list per
interrupt type. An exception to this are service signal and machine check
interrupts, as there can be only one pending interrupt at a time.
Signed-off-by: Jens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
---
arch/s390/include/asm/kvm_host.h | 30 +-
arch/s390/kvm/interrupt.c | 834 ++++++++++++++++++++++-----------------
arch/s390/kvm/kvm-s390.c | 4 +-
arch/s390/kvm/kvm-s390.h | 2 +-
arch/s390/kvm/priv.c | 9 +-
5 files changed, 511 insertions(+), 368 deletions(-)
diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index b8d1e97..d01fc58 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -344,6 +344,11 @@ enum irq_types {
IRQ_PEND_COUNT
};
+/* We have 2M for virtio device descriptor pages. Smallest amount of
+ * memory per page is 24 bytes (1 queue), so (2048*1024) / 24 = 87381
+ */
+#define KVM_S390_MAX_VIRTIO_IRQS 87381
+
/*
* Repressible (non-floating) machine check interrupts
* subclass bits in MCIC
@@ -421,13 +426,32 @@ struct kvm_s390_local_interrupt {
unsigned long pending_irqs;
};
+#define FIRQ_LIST_IO_ISC_0 0
+#define FIRQ_LIST_IO_ISC_1 1
+#define FIRQ_LIST_IO_ISC_2 2
+#define FIRQ_LIST_IO_ISC_3 3
+#define FIRQ_LIST_IO_ISC_4 4
+#define FIRQ_LIST_IO_ISC_5 5
+#define FIRQ_LIST_IO_ISC_6 6
+#define FIRQ_LIST_IO_ISC_7 7
+#define FIRQ_LIST_PFAULT 8
+#define FIRQ_LIST_VIRTIO 9
+#define FIRQ_LIST_COUNT 10
+#define FIRQ_CNTR_IO 0
+#define FIRQ_CNTR_SERVICE 1
+#define FIRQ_CNTR_VIRTIO 2
+#define FIRQ_CNTR_PFAULT 3
+#define FIRQ_MAX_COUNT 4
+
struct kvm_s390_float_interrupt {
+ unsigned long pending_irqs;
spinlock_t lock;
- struct list_head list;
- atomic_t active;
+ struct list_head lists[FIRQ_LIST_COUNT];
+ int counters[FIRQ_MAX_COUNT];
+ struct kvm_s390_mchk_info mchk;
+ struct kvm_s390_ext_info srv_signal;
int next_rr_cpu;
unsigned long idle_mask[BITS_TO_LONGS(KVM_MAX_VCPUS)];
- unsigned int irq_count;
};
struct kvm_hw_wp_info_arch {
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 65b4eee..298d63b 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -22,6 +22,7 @@
#include <asm/dis.h>
#include <asm/uaccess.h>
#include <asm/sclp.h>
+#include <asm/isc.h>
#include "kvm-s390.h"
#include "gaccess.h"
#include "trace-s390.h"
@@ -34,11 +35,6 @@
#define PFAULT_DONE 0x0680
#define VIRTIO_PARAM 0x0d00
-static int is_ioint(u64 type)
-{
- return ((type & 0xfffe0000u) != 0xfffe0000u);
-}
-
int psw_extint_disabled(struct kvm_vcpu *vcpu)
{
return !(vcpu->arch.sie_block->gpsw.mask & PSW_MASK_EXT);
@@ -74,70 +70,25 @@ static int ckc_interrupts_enabled(struct kvm_vcpu *vcpu)
return 1;
}
-static u64 int_word_to_isc_bits(u32 int_word)
+static inline int is_ioirq(unsigned long irq_type)
{
- u8 isc = (int_word & 0x38000000) >> 27;
+ return ((irq_type >= IRQ_PEND_IO_ISC_0) &&
+ (irq_type <= IRQ_PEND_IO_ISC_7));
+}
+static uint64_t isc_to_isc_bits(int isc)
+{
return (0x80 >> isc) << 24;
}
-static int __must_check __interrupt_is_deliverable(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
+static inline u8 int_word_to_isc(u32 int_word)
{
- switch (inti->type) {
- case KVM_S390_INT_EXTERNAL_CALL:
- if (psw_extint_disabled(vcpu))
- return 0;
- if (vcpu->arch.sie_block->gcr[0] & 0x2000ul)
- return 1;
- return 0;
- case KVM_S390_INT_EMERGENCY:
- if (psw_extint_disabled(vcpu))
- return 0;
- if (vcpu->arch.sie_block->gcr[0] & 0x4000ul)
- return 1;
- return 0;
- case KVM_S390_INT_CLOCK_COMP:
- return ckc_interrupts_enabled(vcpu);
- case KVM_S390_INT_CPU_TIMER:
- if (psw_extint_disabled(vcpu))
- return 0;
- if (vcpu->arch.sie_block->gcr[0] & 0x400ul)
- return 1;
- return 0;
- case KVM_S390_INT_SERVICE:
- case KVM_S390_INT_PFAULT_INIT:
- case KVM_S390_INT_PFAULT_DONE:
- case KVM_S390_INT_VIRTIO:
- if (psw_extint_disabled(vcpu))
- return 0;
- if (vcpu->arch.sie_block->gcr[0] & 0x200ul)
- return 1;
- return 0;
- case KVM_S390_PROGRAM_INT:
- case KVM_S390_SIGP_STOP:
- case KVM_S390_SIGP_SET_PREFIX:
- case KVM_S390_RESTART:
- return 1;
- case KVM_S390_MCHK:
- if (psw_mchk_disabled(vcpu))
- return 0;
- if (vcpu->arch.sie_block->gcr[14] & inti->mchk.cr14)
- return 1;
- return 0;
- case KVM_S390_INT_IO_MIN...KVM_S390_INT_IO_MAX:
- if (psw_ioint_disabled(vcpu))
- return 0;
- if (vcpu->arch.sie_block->gcr[6] &
- int_word_to_isc_bits(inti->io.io_int_word))
- return 1;
- return 0;
- default:
- printk(KERN_WARNING "illegal interrupt type %llx\n",
- inti->type);
- BUG();
- }
- return 0;
+ return (int_word & 0x38000000) >> 27;
+}
+
+static inline unsigned long pending_floating_irqs(struct kvm_vcpu *vcpu)
+{
+ return vcpu->kvm->arch.float_int.pending_irqs;
}
static inline unsigned long pending_local_irqs(struct kvm_vcpu *vcpu)
@@ -145,12 +96,31 @@ static inline unsigned long pending_local_irqs(struct kvm_vcpu *vcpu)
return vcpu->arch.local_int.pending_irqs;
}
-static unsigned long deliverable_local_irqs(struct kvm_vcpu *vcpu)
+static unsigned long disable_iscs(struct kvm_vcpu *vcpu,
+ unsigned long active_mask)
{
- unsigned long active_mask = pending_local_irqs(vcpu);
+ int i;
+
+ for (i = 0; i <= MAX_ISC; i++)
+ if (!(vcpu->arch.sie_block->gcr[6] & isc_to_isc_bits(i)))
+ active_mask &= ~(1UL << (IRQ_PEND_IO_ISC_0 + i));
+
+ return active_mask;
+}
+
+static unsigned long deliverable_irqs(struct kvm_vcpu *vcpu)
+{
+ unsigned long active_mask;
+
+ active_mask = pending_local_irqs(vcpu);
+ active_mask |= pending_floating_irqs(vcpu);
if (psw_extint_disabled(vcpu))
active_mask &= ~IRQ_PEND_EXT_MASK;
+ if (psw_ioint_disabled(vcpu))
+ active_mask &= ~IRQ_PEND_IO_MASK;
+ else
+ active_mask = disable_iscs(vcpu, active_mask);
if (!(vcpu->arch.sie_block->gcr[0] & 0x2000ul))
__clear_bit(IRQ_PEND_EXT_EXTERNAL, &active_mask);
if (!(vcpu->arch.sie_block->gcr[0] & 0x4000ul))
@@ -159,8 +129,13 @@ static unsigned long deliverable_local_irqs(struct kvm_vcpu *vcpu)
__clear_bit(IRQ_PEND_EXT_CLOCK_COMP, &active_mask);
if (!(vcpu->arch.sie_block->gcr[0] & 0x400ul))
__clear_bit(IRQ_PEND_EXT_CPU_TIMER, &active_mask);
+ if (!(vcpu->arch.sie_block->gcr[0] & 0x200ul))
+ __clear_bit(IRQ_PEND_EXT_SERVICE, &active_mask);
if (psw_mchk_disabled(vcpu))
active_mask &= ~IRQ_PEND_MCHK_MASK;
+ if (!(vcpu->arch.sie_block->gcr[14] &
+ vcpu->kvm->arch.float_int.mchk.cr14))
+ __clear_bit(IRQ_PEND_MCHK_REP, &active_mask);
/*
* STOP irqs will never be actively delivered. They are triggered via
@@ -202,6 +177,16 @@ static void __set_cpuflag(struct kvm_vcpu *vcpu, u32 flag)
atomic_set_mask(flag, &vcpu->arch.sie_block->cpuflags);
}
+static void set_intercept_indicators_io(struct kvm_vcpu *vcpu)
+{
+ if (!(pending_floating_irqs(vcpu) & IRQ_PEND_IO_MASK))
+ return;
+ else if (psw_ioint_disabled(vcpu))
+ __set_cpuflag(vcpu, CPUSTAT_IO_INT);
+ else
+ vcpu->arch.sie_block->lctl |= LCTL_CR6;
+}
+
static void set_intercept_indicators_ext(struct kvm_vcpu *vcpu)
{
if (!(pending_local_irqs(vcpu) & IRQ_PEND_EXT_MASK))
@@ -228,43 +213,15 @@ static void set_intercept_indicators_stop(struct kvm_vcpu *vcpu)
__set_cpuflag(vcpu, CPUSTAT_STOP_INT);
}
-/* Set interception request for non-deliverable local interrupts */
-static void set_intercept_indicators_local(struct kvm_vcpu *vcpu)
+/* Set interception request for non-deliverable interrupts */
+static void set_intercept_indicators(struct kvm_vcpu *vcpu)
{
+ set_intercept_indicators_io(vcpu);
set_intercept_indicators_ext(vcpu);
set_intercept_indicators_mchk(vcpu);
set_intercept_indicators_stop(vcpu);
}
-static void __set_intercept_indicator(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
-{
- switch (inti->type) {
- case KVM_S390_INT_SERVICE:
- case KVM_S390_INT_PFAULT_DONE:
- case KVM_S390_INT_VIRTIO:
- if (psw_extint_disabled(vcpu))
- __set_cpuflag(vcpu, CPUSTAT_EXT_INT);
- else
- vcpu->arch.sie_block->lctl |= LCTL_CR0;
- break;
- case KVM_S390_MCHK:
- if (psw_mchk_disabled(vcpu))
- vcpu->arch.sie_block->ictl |= ICTL_LPSW;
- else
- vcpu->arch.sie_block->lctl |= LCTL_CR14;
- break;
- case KVM_S390_INT_IO_MIN...KVM_S390_INT_IO_MAX:
- if (psw_ioint_disabled(vcpu))
- __set_cpuflag(vcpu, CPUSTAT_IO_INT);
- else
- vcpu->arch.sie_block->lctl |= LCTL_CR6;
- break;
- default:
- BUG();
- }
-}
-
static u16 get_ilc(struct kvm_vcpu *vcpu)
{
switch (vcpu->arch.sie_block->icptcode) {
@@ -350,42 +307,72 @@ static int __must_check __deliver_pfault_init(struct kvm_vcpu *vcpu)
static int __must_check __deliver_machine_check(struct kvm_vcpu *vcpu)
{
+ struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int;
- struct kvm_s390_mchk_info mchk;
+ struct kvm_s390_mchk_info mchk = {};
unsigned long adtl_status_addr;
- int rc;
+ int deliver = 0;
+ int rc = 0;
+ spin_lock(&fi->lock);
spin_lock(&li->lock);
- mchk = li->irq.mchk;
+ if (test_bit(IRQ_PEND_MCHK_EX, &li->pending_irqs) ||
+ test_bit(IRQ_PEND_MCHK_REP, &li->pending_irqs)) {
+ /*
+ * If there was an exigent machine check pending, then any
+ * repressible machine checks that might have been pending
+ * are indicated along with it, so always clear bits for
+ * repressible and exigent interrupts
+ */
+ mchk = li->irq.mchk;
+ clear_bit(IRQ_PEND_MCHK_EX, &li->pending_irqs);
+ clear_bit(IRQ_PEND_MCHK_REP, &li->pending_irqs);
+ memset(&li->irq.mchk, 0, sizeof(mchk));
+ deliver = 1;
+ }
/*
- * If there was an exigent machine check pending, then any repressible
- * machine checks that might have been pending are indicated along
- * with it, so always clear both bits
+ * We indicate floating repressible conditions along with
+ * other pending conditions. Channel Report Pending and Channel
+ * Subsystem damage are the only two and and are indicated by
+ * bits in mcic and masked in cr14.
*/
- clear_bit(IRQ_PEND_MCHK_EX, &li->pending_irqs);
- clear_bit(IRQ_PEND_MCHK_REP, &li->pending_irqs);
- memset(&li->irq.mchk, 0, sizeof(mchk));
+ if (test_and_clear_bit(IRQ_PEND_MCHK_REP, &fi->pending_irqs)) {
+ mchk.mcic |= fi->mchk.mcic;
+ mchk.cr14 |= fi->mchk.cr14;
+ memset(&fi->mchk, 0, sizeof(mchk));
+ deliver = 1;
+ }
spin_unlock(&li->lock);
+ spin_unlock(&fi->lock);
- VCPU_EVENT(vcpu, 4, "interrupt: machine check mcic=%llx",
- mchk.mcic);
- trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, KVM_S390_MCHK,
- mchk.cr14, mchk.mcic);
-
- rc = kvm_s390_vcpu_store_status(vcpu, KVM_S390_STORE_STATUS_PREFIXED);
- rc |= read_guest_lc(vcpu, __LC_VX_SAVE_AREA_ADDR,
- &adtl_status_addr, sizeof(unsigned long));
- rc |= kvm_s390_vcpu_store_adtl_status(vcpu, adtl_status_addr);
- rc |= put_guest_lc(vcpu, mchk.mcic,
- (u64 __user *) __LC_MCCK_CODE);
- rc |= put_guest_lc(vcpu, mchk.failing_storage_address,
- (u64 __user *) __LC_MCCK_FAIL_STOR_ADDR);
- rc |= write_guest_lc(vcpu, __LC_PSW_SAVE_AREA,
- &mchk.fixed_logout, sizeof(mchk.fixed_logout));
- rc |= write_guest_lc(vcpu, __LC_MCK_OLD_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= read_guest_lc(vcpu, __LC_MCK_NEW_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
+ if (deliver) {
+ VCPU_EVENT(vcpu, 4, "interrupt: machine check mcic=%llx",
+ mchk.mcic);
+ trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id,
+ KVM_S390_MCHK,
+ mchk.cr14, mchk.mcic);
+
+ rc = kvm_s390_vcpu_store_status(vcpu,
+ KVM_S390_STORE_STATUS_PREFIXED);
+ rc |= read_guest_lc(vcpu, __LC_VX_SAVE_AREA_ADDR,
+ &adtl_status_addr,
+ sizeof(unsigned long));
+ rc |= kvm_s390_vcpu_store_adtl_status(vcpu,
+ adtl_status_addr);
+ rc |= put_guest_lc(vcpu, mchk.mcic,
+ (u64 __user *) __LC_MCCK_CODE);
+ rc |= put_guest_lc(vcpu, mchk.failing_storage_address,
+ (u64 __user *) __LC_MCCK_FAIL_STOR_ADDR);
+ rc |= write_guest_lc(vcpu, __LC_PSW_SAVE_AREA,
+ &mchk.fixed_logout,
+ sizeof(mchk.fixed_logout));
+ rc |= write_guest_lc(vcpu, __LC_MCK_OLD_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ rc |= read_guest_lc(vcpu, __LC_MCK_NEW_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ }
return rc ? -EFAULT : 0;
}
@@ -597,16 +584,27 @@ static int __must_check __deliver_prog(struct kvm_vcpu *vcpu)
return rc ? -EFAULT : 0;
}
-static int __must_check __deliver_service(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
+static int __must_check __deliver_service(struct kvm_vcpu *vcpu)
{
- int rc;
+ struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
+ struct kvm_s390_ext_info ext;
+ int rc = 0;
+
+ spin_lock(&fi->lock);
+ if (!(test_bit(IRQ_PEND_EXT_SERVICE, &fi->pending_irqs))) {
+ spin_unlock(&fi->lock);
+ return 0;
+ }
+ ext = fi->srv_signal;
+ memset(&fi->srv_signal, 0, sizeof(ext));
+ clear_bit(IRQ_PEND_EXT_SERVICE, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
VCPU_EVENT(vcpu, 4, "interrupt: sclp parm:%x",
- inti->ext.ext_params);
+ ext.ext_params);
vcpu->stat.deliver_service_signal++;
- trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, inti->type,
- inti->ext.ext_params, 0);
+ trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, KVM_S390_INT_SERVICE,
+ ext.ext_params, 0);
rc = put_guest_lc(vcpu, EXT_IRQ_SERVICE_SIG, (u16 *)__LC_EXT_INT_CODE);
rc |= put_guest_lc(vcpu, 0, (u16 *)__LC_EXT_CPU_ADDR);
@@ -614,106 +612,146 @@ static int __must_check __deliver_service(struct kvm_vcpu *vcpu,
&vcpu->arch.sie_block->gpsw, sizeof(psw_t));
rc |= read_guest_lc(vcpu, __LC_EXT_NEW_PSW,
&vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= put_guest_lc(vcpu, inti->ext.ext_params,
+ rc |= put_guest_lc(vcpu, ext.ext_params,
(u32 *)__LC_EXT_PARAMS);
+
return rc ? -EFAULT : 0;
}
-static int __must_check __deliver_pfault_done(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
+static int __must_check __deliver_pfault_done(struct kvm_vcpu *vcpu)
{
- int rc;
+ struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
+ struct kvm_s390_interrupt_info *inti;
+ int rc = 0;
- trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id,
- KVM_S390_INT_PFAULT_DONE, 0,
- inti->ext.ext_params2);
+ spin_lock(&fi->lock);
+ inti = list_first_entry_or_null(&fi->lists[FIRQ_LIST_PFAULT],
+ struct kvm_s390_interrupt_info,
+ list);
+ if (inti) {
+ trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id,
+ KVM_S390_INT_PFAULT_DONE, 0,
+ inti->ext.ext_params2);
+ list_del(&inti->list);
+ fi->counters[FIRQ_CNTR_PFAULT] -= 1;
+ }
+ if (list_empty(&fi->lists[FIRQ_LIST_PFAULT]))
+ clear_bit(IRQ_PEND_PFAULT_DONE, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
- rc = put_guest_lc(vcpu, EXT_IRQ_CP_SERVICE, (u16 *)__LC_EXT_INT_CODE);
- rc |= put_guest_lc(vcpu, PFAULT_DONE, (u16 *)__LC_EXT_CPU_ADDR);
- rc |= write_guest_lc(vcpu, __LC_EXT_OLD_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= read_guest_lc(vcpu, __LC_EXT_NEW_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= put_guest_lc(vcpu, inti->ext.ext_params2,
- (u64 *)__LC_EXT_PARAMS2);
+ if (inti) {
+ rc = put_guest_lc(vcpu, EXT_IRQ_CP_SERVICE,
+ (u16 *)__LC_EXT_INT_CODE);
+ rc |= put_guest_lc(vcpu, PFAULT_DONE,
+ (u16 *)__LC_EXT_CPU_ADDR);
+ rc |= write_guest_lc(vcpu, __LC_EXT_OLD_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ rc |= read_guest_lc(vcpu, __LC_EXT_NEW_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ rc |= put_guest_lc(vcpu, inti->ext.ext_params2,
+ (u64 *)__LC_EXT_PARAMS2);
+ kfree(inti);
+ }
return rc ? -EFAULT : 0;
}
-static int __must_check __deliver_virtio(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
+static int __must_check __deliver_virtio(struct kvm_vcpu *vcpu)
{
- int rc;
+ struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int;
+ struct kvm_s390_interrupt_info *inti;
+ int rc = 0;
- VCPU_EVENT(vcpu, 4, "interrupt: virtio parm:%x,parm64:%llx",
- inti->ext.ext_params, inti->ext.ext_params2);
- vcpu->stat.deliver_virtio_interrupt++;
- trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, inti->type,
- inti->ext.ext_params,
- inti->ext.ext_params2);
+ spin_lock(&fi->lock);
+ inti = list_first_entry_or_null(&fi->lists[FIRQ_LIST_VIRTIO],
+ struct kvm_s390_interrupt_info,
+ list);
+ if (inti) {
+ VCPU_EVENT(vcpu, 4,
+ "interrupt: virtio parm:%x,parm64:%llx",
+ inti->ext.ext_params, inti->ext.ext_params2);
+ vcpu->stat.deliver_virtio_interrupt++;
+ trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id,
+ inti->type,
+ inti->ext.ext_params,
+ inti->ext.ext_params2);
+ list_del(&inti->list);
+ fi->counters[FIRQ_CNTR_VIRTIO] -= 1;
+ }
+ if (list_empty(&fi->lists[FIRQ_LIST_VIRTIO]))
+ clear_bit(IRQ_PEND_VIRTIO, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
- rc = put_guest_lc(vcpu, EXT_IRQ_CP_SERVICE, (u16 *)__LC_EXT_INT_CODE);
- rc |= put_guest_lc(vcpu, VIRTIO_PARAM, (u16 *)__LC_EXT_CPU_ADDR);
- rc |= write_guest_lc(vcpu, __LC_EXT_OLD_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= read_guest_lc(vcpu, __LC_EXT_NEW_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= put_guest_lc(vcpu, inti->ext.ext_params,
- (u32 *)__LC_EXT_PARAMS);
- rc |= put_guest_lc(vcpu, inti->ext.ext_params2,
- (u64 *)__LC_EXT_PARAMS2);
+ if (inti) {
+ rc = put_guest_lc(vcpu, EXT_IRQ_CP_SERVICE,
+ (u16 *)__LC_EXT_INT_CODE);
+ rc |= put_guest_lc(vcpu, VIRTIO_PARAM,
+ (u16 *)__LC_EXT_CPU_ADDR);
+ rc |= write_guest_lc(vcpu, __LC_EXT_OLD_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ rc |= read_guest_lc(vcpu, __LC_EXT_NEW_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ rc |= put_guest_lc(vcpu, inti->ext.ext_params,
+ (u32 *)__LC_EXT_PARAMS);
+ rc |= put_guest_lc(vcpu, inti->ext.ext_params2,
+ (u64 *)__LC_EXT_PARAMS2);
+ kfree(inti);
+ }
return rc ? -EFAULT : 0;
}
static int __must_check __deliver_io(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
+ unsigned long irq_type)
{
- int rc;
+ struct list_head *isc_list;
+ struct kvm_s390_float_interrupt *fi;
+ struct kvm_s390_interrupt_info *inti = NULL;
+ int rc = 0;
- VCPU_EVENT(vcpu, 4, "interrupt: I/O %llx", inti->type);
- vcpu->stat.deliver_io_int++;
- trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, inti->type,
- ((__u32)inti->io.subchannel_id << 16) |
- inti->io.subchannel_nr,
- ((__u64)inti->io.io_int_parm << 32) |
- inti->io.io_int_word);
-
- rc = put_guest_lc(vcpu, inti->io.subchannel_id,
- (u16 *)__LC_SUBCHANNEL_ID);
- rc |= put_guest_lc(vcpu, inti->io.subchannel_nr,
- (u16 *)__LC_SUBCHANNEL_NR);
- rc |= put_guest_lc(vcpu, inti->io.io_int_parm,
- (u32 *)__LC_IO_INT_PARM);
- rc |= put_guest_lc(vcpu, inti->io.io_int_word,
- (u32 *)__LC_IO_INT_WORD);
- rc |= write_guest_lc(vcpu, __LC_IO_OLD_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= read_guest_lc(vcpu, __LC_IO_NEW_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- return rc ? -EFAULT : 0;
-}
+ fi = &vcpu->kvm->arch.float_int;
-static int __must_check __deliver_mchk_floating(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
-{
- struct kvm_s390_mchk_info *mchk = &inti->mchk;
- int rc;
+ spin_lock(&fi->lock);
+ isc_list = &fi->lists[irq_type - IRQ_PEND_IO_ISC_0];
+ inti = list_first_entry_or_null(isc_list,
+ struct kvm_s390_interrupt_info,
+ list);
+ if (inti) {
+ VCPU_EVENT(vcpu, 4, "interrupt: I/O %llx", inti->type);
+ vcpu->stat.deliver_io_int++;
+ trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id,
+ inti->type,
+ ((__u32)inti->io.subchannel_id << 16) |
+ inti->io.subchannel_nr,
+ ((__u64)inti->io.io_int_parm << 32) |
+ inti->io.io_int_word);
+ list_del(&inti->list);
+ fi->counters[FIRQ_CNTR_IO] -= 1;
+ }
+ if (list_empty(isc_list))
+ clear_bit(irq_type, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
+
+ if (inti) {
+ rc = put_guest_lc(vcpu, inti->io.subchannel_id,
+ (u16 *)__LC_SUBCHANNEL_ID);
+ rc |= put_guest_lc(vcpu, inti->io.subchannel_nr,
+ (u16 *)__LC_SUBCHANNEL_NR);
+ rc |= put_guest_lc(vcpu, inti->io.io_int_parm,
+ (u32 *)__LC_IO_INT_PARM);
+ rc |= put_guest_lc(vcpu, inti->io.io_int_word,
+ (u32 *)__LC_IO_INT_WORD);
+ rc |= write_guest_lc(vcpu, __LC_IO_OLD_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ rc |= read_guest_lc(vcpu, __LC_IO_NEW_PSW,
+ &vcpu->arch.sie_block->gpsw,
+ sizeof(psw_t));
+ kfree(inti);
+ }
- VCPU_EVENT(vcpu, 4, "interrupt: machine check mcic=%llx",
- mchk->mcic);
- trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, KVM_S390_MCHK,
- mchk->cr14, mchk->mcic);
-
- rc = kvm_s390_vcpu_store_status(vcpu, KVM_S390_STORE_STATUS_PREFIXED);
- rc |= put_guest_lc(vcpu, mchk->mcic,
- (u64 __user *) __LC_MCCK_CODE);
- rc |= put_guest_lc(vcpu, mchk->failing_storage_address,
- (u64 __user *) __LC_MCCK_FAIL_STOR_ADDR);
- rc |= write_guest_lc(vcpu, __LC_PSW_SAVE_AREA,
- &mchk->fixed_logout, sizeof(mchk->fixed_logout));
- rc |= write_guest_lc(vcpu, __LC_MCK_OLD_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
- rc |= read_guest_lc(vcpu, __LC_MCK_NEW_PSW,
- &vcpu->arch.sie_block->gpsw, sizeof(psw_t));
return rc ? -EFAULT : 0;
}
@@ -721,6 +759,7 @@ typedef int (*deliver_irq_t)(struct kvm_vcpu *vcpu);
static const deliver_irq_t deliver_irq_funcs[] = {
[IRQ_PEND_MCHK_EX] = __deliver_machine_check,
+ [IRQ_PEND_MCHK_REP] = __deliver_machine_check,
[IRQ_PEND_PROG] = __deliver_prog,
[IRQ_PEND_EXT_EMERGENCY] = __deliver_emergency_signal,
[IRQ_PEND_EXT_EXTERNAL] = __deliver_external_call,
@@ -729,36 +768,11 @@ static const deliver_irq_t deliver_irq_funcs[] = {
[IRQ_PEND_RESTART] = __deliver_restart,
[IRQ_PEND_SET_PREFIX] = __deliver_set_prefix,
[IRQ_PEND_PFAULT_INIT] = __deliver_pfault_init,
+ [IRQ_PEND_EXT_SERVICE] = __deliver_service,
+ [IRQ_PEND_PFAULT_DONE] = __deliver_pfault_done,
+ [IRQ_PEND_VIRTIO] = __deliver_virtio,
};
-static int __must_check __deliver_floating_interrupt(struct kvm_vcpu *vcpu,
- struct kvm_s390_interrupt_info *inti)
-{
- int rc;
-
- switch (inti->type) {
- case KVM_S390_INT_SERVICE:
- rc = __deliver_service(vcpu, inti);
- break;
- case KVM_S390_INT_PFAULT_DONE:
- rc = __deliver_pfault_done(vcpu, inti);
- break;
- case KVM_S390_INT_VIRTIO:
- rc = __deliver_virtio(vcpu, inti);
- break;
- case KVM_S390_MCHK:
- rc = __deliver_mchk_floating(vcpu, inti);
- break;
- case KVM_S390_INT_IO_MIN...KVM_S390_INT_IO_MAX:
- rc = __deliver_io(vcpu, inti);
- break;
- default:
- BUG();
- }
-
- return rc;
-}
-
/* Check whether an external call is pending (deliverable or not) */
int kvm_s390_ext_call_pending(struct kvm_vcpu *vcpu)
{
@@ -774,21 +788,9 @@ int kvm_s390_ext_call_pending(struct kvm_vcpu *vcpu)
int kvm_s390_vcpu_has_irq(struct kvm_vcpu *vcpu, int exclude_stop)
{
- struct kvm_s390_float_interrupt *fi = vcpu->arch.local_int.float_int;
- struct kvm_s390_interrupt_info *inti;
int rc;
- rc = !!deliverable_local_irqs(vcpu);
-
- if ((!rc) && atomic_read(&fi->active)) {
- spin_lock(&fi->lock);
- list_for_each_entry(inti, &fi->list, list)
- if (__interrupt_is_deliverable(vcpu, inti)) {
- rc = 1;
- break;
- }
- spin_unlock(&fi->lock);
- }
+ rc = !!deliverable_irqs(vcpu);
if (!rc && kvm_cpu_has_pending_timer(vcpu))
rc = 1;
@@ -907,13 +909,10 @@ void kvm_s390_clear_local_irqs(struct kvm_vcpu *vcpu)
int __must_check kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu)
{
struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int;
- struct kvm_s390_float_interrupt *fi = vcpu->arch.local_int.float_int;
- struct kvm_s390_interrupt_info *n, *inti = NULL;
deliver_irq_t func;
- int deliver;
int rc = 0;
unsigned long irq_type;
- unsigned long deliverable_irqs;
+ unsigned long irqs;
__reset_intercept_indicators(vcpu);
@@ -923,44 +922,27 @@ int __must_check kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu)
set_bit(IRQ_PEND_EXT_CLOCK_COMP, &li->pending_irqs);
do {
- deliverable_irqs = deliverable_local_irqs(vcpu);
+ irqs = deliverable_irqs(vcpu);
/* bits are in the order of interrupt priority */
- irq_type = find_first_bit(&deliverable_irqs, IRQ_PEND_COUNT);
+ irq_type = find_first_bit(&irqs, IRQ_PEND_COUNT);
if (irq_type == IRQ_PEND_COUNT)
break;
- func = deliver_irq_funcs[irq_type];
- if (!func) {
- WARN_ON_ONCE(func == NULL);
- clear_bit(irq_type, &li->pending_irqs);
- continue;
+ if (is_ioirq(irq_type)) {
+ rc = __deliver_io(vcpu, irq_type);
+ } else {
+ func = deliver_irq_funcs[irq_type];
+ if (!func) {
+ WARN_ON_ONCE(func == NULL);
+ clear_bit(irq_type, &li->pending_irqs);
+ continue;
+ }
+ rc = func(vcpu);
}
- rc = func(vcpu);
- } while (!rc && irq_type != IRQ_PEND_COUNT);
-
- set_intercept_indicators_local(vcpu);
+ if (rc)
+ break;
+ } while (!rc);
- if (!rc && atomic_read(&fi->active)) {
- do {
- deliver = 0;
- spin_lock(&fi->lock);
- list_for_each_entry_safe(inti, n, &fi->list, list) {
- if (__interrupt_is_deliverable(vcpu, inti)) {
- list_del(&inti->list);
- fi->irq_count--;
- deliver = 1;
- break;
- }
- __set_intercept_indicator(vcpu, inti);
- }
- if (list_empty(&fi->list))
- atomic_set(&fi->active, 0);
- spin_unlock(&fi->lock);
- if (deliver) {
- rc = __deliver_floating_interrupt(vcpu, inti);
- kfree(inti);
- }
- } while (!rc && deliver);
- }
+ set_intercept_indicators(vcpu);
return rc;
}
@@ -1195,80 +1177,182 @@ static int __inject_cpu_timer(struct kvm_vcpu *vcpu)
return 0;
}
+static struct kvm_s390_interrupt_info *get_io_int(struct kvm *kvm,
+ int isc, u32 schid)
+{
+ struct kvm_s390_float_interrupt *fi = &kvm->arch.float_int;
+ struct list_head *isc_list = &fi->lists[FIRQ_LIST_IO_ISC_0 + isc];
+ struct kvm_s390_interrupt_info *iter;
+ u16 id = (schid & 0xffff0000U) >> 16;
+ u16 nr = schid & 0x0000ffffU;
+ spin_lock(&fi->lock);
+ list_for_each_entry(iter, isc_list, list) {
+ if (schid && (id != iter->io.subchannel_id ||
+ nr != iter->io.subchannel_nr))
+ continue;
+ /* found an appropriate entry */
+ list_del_init(&iter->list);
+ fi->counters[FIRQ_CNTR_IO] -= 1;
+ if (list_empty(isc_list))
+ clear_bit(IRQ_PEND_IO_ISC_0 + isc, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
+ return iter;
+ }
+ spin_unlock(&fi->lock);
+ return NULL;
+}
+
+/*
+ * Dequeue and return an I/O interrupt matching any of the interruption
+ * subclasses as designated by the isc mask in cr6 and the schid (if != 0).
+ */
struct kvm_s390_interrupt_info *kvm_s390_get_io_int(struct kvm *kvm,
- u64 cr6, u64 schid)
+ u64 isc_mask, u32 schid)
+{
+ struct kvm_s390_interrupt_info *inti = NULL;
+ int isc;
+
+ for (isc = 0; isc <= MAX_ISC && !inti; isc++) {
+ if (isc_mask & isc_to_isc_bits(isc))
+ inti = get_io_int(kvm, isc, schid);
+ }
+ return inti;
+}
+
+#define SCCB_MASK 0xFFFFFFF8
+#define SCCB_EVENT_PENDING 0x3
+
+static int __inject_service(struct kvm *kvm,
+ struct kvm_s390_interrupt_info *inti)
+{
+ struct kvm_s390_float_interrupt *fi = &kvm->arch.float_int;
+
+ spin_lock(&fi->lock);
+ fi->srv_signal.ext_params |= inti->ext.ext_params & SCCB_EVENT_PENDING;
+ /*
+ * Early versions of the QEMU s390 bios will inject several
+ * service interrupts after another without handling a
+ * condition code indicating busy.
+ * We will silently ignore those superfluous sccb values.
+ * A future version of QEMU will take care of serialization
+ * of servc requests
+ */
+ if (fi->srv_signal.ext_params & SCCB_MASK)
+ goto out;
+ fi->srv_signal.ext_params |= inti->ext.ext_params & SCCB_MASK;
+ set_bit(IRQ_PEND_EXT_SERVICE, &fi->pending_irqs);
+out:
+ spin_unlock(&fi->lock);
+ kfree(inti);
+ return 0;
+}
+
+static int __inject_virtio(struct kvm *kvm,
+ struct kvm_s390_interrupt_info *inti)
+{
+ struct kvm_s390_float_interrupt *fi = &kvm->arch.float_int;
+
+ spin_lock(&fi->lock);
+ if (fi->counters[FIRQ_CNTR_VIRTIO] >= KVM_S390_MAX_VIRTIO_IRQS) {
+ spin_unlock(&fi->lock);
+ return -EBUSY;
+ }
+ fi->counters[FIRQ_CNTR_VIRTIO] += 1;
+ list_add_tail(&inti->list, &fi->lists[FIRQ_LIST_VIRTIO]);
+ set_bit(IRQ_PEND_VIRTIO, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
+ return 0;
+}
+
+static int __inject_pfault_done(struct kvm *kvm,
+ struct kvm_s390_interrupt_info *inti)
+{
+ struct kvm_s390_float_interrupt *fi = &kvm->arch.float_int;
+
+ spin_lock(&fi->lock);
+ if (fi->counters[FIRQ_CNTR_PFAULT] >=
+ (ASYNC_PF_PER_VCPU * KVM_MAX_VCPUS)) {
+ spin_unlock(&fi->lock);
+ return -EBUSY;
+ }
+ fi->counters[FIRQ_CNTR_PFAULT] += 1;
+ list_add_tail(&inti->list, &fi->lists[FIRQ_LIST_PFAULT]);
+ set_bit(IRQ_PEND_PFAULT_DONE, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
+ return 0;
+}
+
+#define CR_PENDING_SUBCLASS 28
+static int __inject_float_mchk(struct kvm *kvm,
+ struct kvm_s390_interrupt_info *inti)
+{
+ struct kvm_s390_float_interrupt *fi = &kvm->arch.float_int;
+
+ spin_lock(&fi->lock);
+ fi->mchk.cr14 |= inti->mchk.cr14 & (1UL << CR_PENDING_SUBCLASS);
+ fi->mchk.mcic |= inti->mchk.mcic;
+ set_bit(IRQ_PEND_MCHK_REP, &fi->pending_irqs);
+ spin_unlock(&fi->lock);
+ kfree(inti);
+ return 0;
+}
+
+static int __inject_io(struct kvm *kvm, struct kvm_s390_interrupt_info *inti)
{
struct kvm_s390_float_interrupt *fi;
- struct kvm_s390_interrupt_info *inti, *iter;
+ struct list_head *list;
+ int isc;
- if ((!schid && !cr6) || (schid && cr6))
- return NULL;
fi = &kvm->arch.float_int;
spin_lock(&fi->lock);
- inti = NULL;
- list_for_each_entry(iter, &fi->list, list) {
- if (!is_ioint(iter->type))
- continue;
- if (cr6 &&
- ((cr6 & int_word_to_isc_bits(iter->io.io_int_word)) == 0))
- continue;
- if (schid) {
- if (((schid & 0x00000000ffff0000) >> 16) !=
- iter->io.subchannel_id)
- continue;
- if ((schid & 0x000000000000ffff) !=
- iter->io.subchannel_nr)
- continue;
- }
- inti = iter;
- break;
- }
- if (inti) {
- list_del_init(&inti->list);
- fi->irq_count--;
+ if (fi->counters[FIRQ_CNTR_IO] >= KVM_S390_MAX_FLOAT_IRQS) {
+ spin_unlock(&fi->lock);
+ return -EBUSY;
}
- if (list_empty(&fi->list))
- atomic_set(&fi->active, 0);
+ fi->counters[FIRQ_CNTR_IO] += 1;
+
+ isc = int_word_to_isc(inti->io.io_int_word);
+ list = &fi->lists[FIRQ_LIST_IO_ISC_0 + isc];
+ list_add_tail(&inti->list, list);
+ set_bit(IRQ_PEND_IO_ISC_0 + isc, &fi->pending_irqs);
spin_unlock(&fi->lock);
- return inti;
+ return 0;
}
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;
+ u64 type = READ_ONCE(inti->type);
+ int rc;
fi = &kvm->arch.float_int;
- spin_lock(&fi->lock);
- if (fi->irq_count >= KVM_S390_MAX_FLOAT_IRQS) {
+
+ switch (type) {
+ case KVM_S390_MCHK:
+ rc = __inject_float_mchk(kvm, inti);
+ break;
+ case KVM_S390_INT_VIRTIO:
+ rc = __inject_virtio(kvm, inti);
+ break;
+ case KVM_S390_INT_SERVICE:
+ rc = __inject_service(kvm, inti);
+ break;
+ case KVM_S390_INT_PFAULT_DONE:
+ rc = __inject_pfault_done(kvm, inti);
+ break;
+ case KVM_S390_INT_IO_MIN...KVM_S390_INT_IO_MAX:
+ rc = __inject_io(kvm, inti);
+ break;
+ default:
rc = -EINVAL;
- goto unlock_fi;
}
- fi->irq_count++;
- if (!is_ioint(inti->type)) {
- list_add_tail(&inti->list, &fi->list);
- } else {
- u64 isc_bits = int_word_to_isc_bits(inti->io.io_int_word);
+ if (rc)
+ return rc;
- /* Keep I/O interrupts sorted in isc order. */
- list_for_each_entry(iter, &fi->list, list) {
- if (!is_ioint(iter->type))
- continue;
- if (int_word_to_isc_bits(iter->io.io_int_word)
- <= isc_bits)
- continue;
- break;
- }
- list_add_tail(&inti->list, &iter->list);
- }
- atomic_set(&fi->active, 1);
- if (atomic_read(&kvm->online_vcpus) == 0)
- goto unlock_fi;
sigcpu = find_first_bit(fi->idle_mask, KVM_MAX_VCPUS);
if (sigcpu == KVM_MAX_VCPUS) {
do {
@@ -1280,7 +1364,7 @@ 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(&li->lock);
- switch (inti->type) {
+ switch (type) {
case KVM_S390_MCHK:
atomic_set_mask(CPUSTAT_STOP_INT, li->cpuflags);
break;
@@ -1293,9 +1377,8 @@ static int __inject_vm(struct kvm *kvm, struct kvm_s390_interrupt_info *inti)
}
spin_unlock(&li->lock);
kvm_s390_vcpu_wakeup(kvm_get_vcpu(kvm, sigcpu));
-unlock_fi:
- spin_unlock(&fi->lock);
- return rc;
+ return 0;
+
}
int kvm_s390_inject_vm(struct kvm *kvm,
@@ -1462,20 +1545,14 @@ int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu, struct kvm_s390_irq *irq)
return rc;
}
-void kvm_s390_clear_float_irqs(struct kvm *kvm)
+static inline void clear_irq_list(struct list_head *_list)
{
- struct kvm_s390_float_interrupt *fi;
- struct kvm_s390_interrupt_info *n, *inti = NULL;
+ struct kvm_s390_interrupt_info *inti, *n;
- fi = &kvm->arch.float_int;
- spin_lock(&fi->lock);
- list_for_each_entry_safe(inti, n, &fi->list, list) {
+ list_for_each_entry_safe(inti, n, _list, list) {
list_del(&inti->list);
kfree(inti);
}
- fi->irq_count = 0;
- atomic_set(&fi->active, 0);
- spin_unlock(&fi->lock);
}
static void inti_to_irq(struct kvm_s390_interrupt_info *inti,
@@ -1486,26 +1563,37 @@ static void inti_to_irq(struct kvm_s390_interrupt_info *inti,
case KVM_S390_INT_PFAULT_INIT:
case KVM_S390_INT_PFAULT_DONE:
case KVM_S390_INT_VIRTIO:
- case KVM_S390_INT_SERVICE:
irq->u.ext = inti->ext;
break;
case KVM_S390_INT_IO_MIN...KVM_S390_INT_IO_MAX:
irq->u.io = inti->io;
break;
- case KVM_S390_MCHK:
- irq->u.mchk = inti->mchk;
- break;
}
}
-static int get_all_floating_irqs(struct kvm *kvm, __user u8 *usrbuf, u64 len)
+void kvm_s390_clear_float_irqs(struct kvm *kvm)
+{
+ struct kvm_s390_float_interrupt *fi = &kvm->arch.float_int;
+ int i;
+
+ spin_lock(&fi->lock);
+ for (i = 0; i < FIRQ_LIST_COUNT; i++)
+ clear_irq_list(&fi->lists[i]);
+ for (i = 0; i < FIRQ_MAX_COUNT; i++)
+ fi->counters[i] = 0;
+ spin_unlock(&fi->lock);
+};
+
+static int get_all_floating_irqs(struct kvm *kvm, __user __u8 *usrbuf, __u64 len)
{
struct kvm_s390_interrupt_info *inti;
struct kvm_s390_float_interrupt *fi;
struct kvm_s390_irq *buf;
+ struct kvm_s390_irq *irq;
int max_irqs;
int ret = 0;
int n = 0;
+ int i;
if (len > KVM_S390_FLIC_MAX_BUFFER || len == 0)
return -EINVAL;
@@ -1523,15 +1611,41 @@ static int get_all_floating_irqs(struct kvm *kvm, __user u8 *usrbuf, u64 len)
fi = &kvm->arch.float_int;
spin_lock(&fi->lock);
- list_for_each_entry(inti, &fi->list, list) {
+ for (i = 0; i < FIRQ_LIST_COUNT; i++) {
+ list_for_each_entry(inti, &fi->lists[i], list) {
+ if (n == max_irqs) {
+ /* signal userspace to try again */
+ ret = -ENOMEM;
+ goto out;
+ }
+ inti_to_irq(inti, &buf[n]);
+ n++;
+ }
+ }
+ if (test_bit(IRQ_PEND_EXT_SERVICE, &fi->pending_irqs)) {
if (n == max_irqs) {
/* signal userspace to try again */
ret = -ENOMEM;
- break;
+ goto out;
}
- inti_to_irq(inti, &buf[n]);
+ irq = (struct kvm_s390_irq *) &buf[n];
+ irq->type = KVM_S390_INT_SERVICE;
+ irq->u.ext = fi->srv_signal;
n++;
}
+ if (test_bit(IRQ_PEND_MCHK_REP, &fi->pending_irqs)) {
+ if (n == max_irqs) {
+ /* signal userspace to try again */
+ ret = -ENOMEM;
+ goto out;
+ }
+ irq = (struct kvm_s390_irq *) &buf[n];
+ irq->type = KVM_S390_MCHK;
+ irq->u.mchk = fi->mchk;
+ n++;
+}
+
+out:
spin_unlock(&fi->lock);
if (!ret && n > 0) {
if (copy_to_user((void __user *) usrbuf,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index a130885..dbc9ca3 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -31,6 +31,7 @@
#include <asm/pgtable.h>
#include <asm/nmi.h>
#include <asm/switch_to.h>
+#include <asm/isc.h>
#include <asm/sclp.h>
#include "kvm-s390.h"
#include "gaccess.h"
@@ -1069,7 +1070,8 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
goto out_err;
spin_lock_init(&kvm->arch.float_int.lock);
- INIT_LIST_HEAD(&kvm->arch.float_int.list);
+ for (i = 0; i < FIRQ_LIST_COUNT; i++)
+ INIT_LIST_HEAD(&kvm->arch.float_int.lists[i]);
init_waitqueue_head(&kvm->arch.ipte_wq);
mutex_init(&kvm->arch.ipte_mutex);
diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h
index c5aefef..343644a 100644
--- a/arch/s390/kvm/kvm-s390.h
+++ b/arch/s390/kvm/kvm-s390.h
@@ -178,7 +178,7 @@ int __must_check kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu,
struct kvm_s390_irq *irq);
int __must_check kvm_s390_inject_program_int(struct kvm_vcpu *vcpu, u16 code);
struct kvm_s390_interrupt_info *kvm_s390_get_io_int(struct kvm *kvm,
- u64 cr6, u64 schid);
+ u64 isc_mask, u32 schid);
int kvm_s390_reinject_io_int(struct kvm *kvm,
struct kvm_s390_interrupt_info *inti);
int kvm_s390_mask_adapter(struct kvm *kvm, unsigned int id, bool masked);
diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c
index 5e4658d..d22d8ee 100644
--- a/arch/s390/kvm/priv.c
+++ b/arch/s390/kvm/priv.c
@@ -294,10 +294,13 @@ reinject_interrupt:
static int handle_tsch(struct kvm_vcpu *vcpu)
{
- struct kvm_s390_interrupt_info *inti;
+ struct kvm_s390_interrupt_info *inti = NULL;
+ const u64 isc_mask = 0xffUL << 24; /* all iscs set */
- inti = kvm_s390_get_io_int(vcpu->kvm, 0,
- vcpu->run->s.regs.gprs[1]);
+ /* a valid schid has at least one bit set */
+ if (vcpu->run->s.regs.gprs[1])
+ inti = kvm_s390_get_io_int(vcpu->kvm, isc_mask,
+ vcpu->run->s.regs.gprs[1]);
/*
* Prepare exit to userspace.
--
2.3.0
^ permalink raw reply related [flat|nested] 13+ messages in thread* [GIT PULL 8/8] KVM: s390: migrate vcpu interrupt state
2015-03-31 13:01 [GIT PULL 0/8] KVM: s390: Features and fixes for 4.1 (kvm/next) Christian Borntraeger
` (6 preceding siblings ...)
2015-03-31 13:02 ` [GIT PULL 7/8] KVM: s390: refactor vcpu injection function Christian Borntraeger
@ 2015-03-31 13:02 ` Christian Borntraeger
2015-03-31 16:11 ` [GIT PULL 0/8] KVM: s390: Features and fixes for 4.1 (kvm/next) Paolo Bonzini
8 siblings, 0 replies; 13+ messages in thread
From: Christian Borntraeger @ 2015-03-31 13:02 UTC (permalink / raw)
To: KVM
Cc: Paolo Bonzini, Marcelo Tosatti, Alexander Graf, Cornelia Huck,
Jens Freimann, linux-s390, Christian Borntraeger
From: Jens Freimann <jfrei@linux.vnet.ibm.com>
This patch adds support to migrate vcpu interrupts. Two new vcpu ioctls
are added which get/set the complete status of pending interrupts in one
go. The ioctls are marked as available with the new capability
KVM_CAP_S390_IRQ_STATE.
We can not use a ONEREG, as the number of pending local interrupts is not
constant and depends on the number of CPUs.
To retrieve the interrupt state we add an ioctl KVM_S390_GET_IRQ_STATE.
Its input parameter is a pointer to a struct kvm_s390_irq_state which
has a buffer and length. For all currently pending interrupts, we copy
a struct kvm_s390_irq into the buffer and pass it to userspace.
To store interrupt state into a buffer provided by userspace, we add an
ioctl KVM_S390_SET_IRQ_STATE. It passes a struct kvm_s390_irq_state into
the kernel and injects all interrupts contained in the buffer.
Signed-off-by: Jens Freimann <jfrei@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>
---
Documentation/virtual/kvm/api.txt | 61 +++++++++++++++++
arch/s390/kvm/interrupt.c | 140 ++++++++++++++++++++++++++++++++++++++
arch/s390/kvm/kvm-s390.c | 36 ++++++++++
arch/s390/kvm/kvm-s390.h | 4 ++
include/uapi/linux/kvm.h | 11 +++
5 files changed, 252 insertions(+)
diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
index a7c651d..18fb763 100644
--- a/Documentation/virtual/kvm/api.txt
+++ b/Documentation/virtual/kvm/api.txt
@@ -2875,6 +2875,67 @@ KVM_S390_MCHK - machine check interrupt; parameters in .mchk
Note that the vcpu ioctl is asynchronous to vcpu execution.
+4.94 KVM_S390_GET_IRQ_STATE
+
+Capability: KVM_CAP_S390_IRQ_STATE
+Architectures: s390
+Type: vcpu ioctl
+Parameters: struct kvm_s390_irq_state (out)
+Returns: >= number of bytes copied into buffer,
+ -EINVAL if buffer size is 0,
+ -ENOBUFS if buffer size is too small to fit all pending interrupts,
+ -EFAULT if the buffer address was invalid
+
+This ioctl allows userspace to retrieve the complete state of all currently
+pending interrupts in a single buffer. Use cases include migration
+and introspection. The parameter structure contains the address of a
+userspace buffer and its length:
+
+struct kvm_s390_irq_state {
+ __u64 buf;
+ __u32 flags;
+ __u32 len;
+ __u32 reserved[4];
+};
+
+Userspace passes in the above struct and for each pending interrupt a
+struct kvm_s390_irq is copied to the provided buffer.
+
+If -ENOBUFS is returned the buffer provided was too small and userspace
+may retry with a bigger buffer.
+
+4.95 KVM_S390_SET_IRQ_STATE
+
+Capability: KVM_CAP_S390_IRQ_STATE
+Architectures: s390
+Type: vcpu ioctl
+Parameters: struct kvm_s390_irq_state (in)
+Returns: 0 on success,
+ -EFAULT if the buffer address was invalid,
+ -EINVAL for an invalid buffer length (see below),
+ -EBUSY if there were already interrupts pending,
+ errors occurring when actually injecting the
+ interrupt. See KVM_S390_IRQ.
+
+This ioctl allows userspace to set the complete state of all cpu-local
+interrupts currently pending for the vcpu. It is intended for restoring
+interrupt state after a migration. The input parameter is a userspace buffer
+containing a struct kvm_s390_irq_state:
+
+struct kvm_s390_irq_state {
+ __u64 buf;
+ __u32 len;
+ __u32 pad;
+};
+
+The userspace memory referenced by buf contains a struct kvm_s390_irq
+for each interrupt to be injected into the guest.
+If one of the interrupts could not be injected for some reason the
+ioctl aborts.
+
+len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0
+and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq),
+which is the maximum number of possibly pending cpu-local interrupts.
5. The kvm_run structure
------------------------
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c
index 2f78f1a..ee2dfdc 100644
--- a/arch/s390/kvm/interrupt.c
+++ b/arch/s390/kvm/interrupt.c
@@ -2125,3 +2125,143 @@ int kvm_set_msi(struct kvm_kernel_irq_routing_entry *e, struct kvm *kvm,
{
return -EINVAL;
}
+
+int kvm_s390_set_irq_state(struct kvm_vcpu *vcpu, void __user *irqstate, int len)
+{
+ struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int;
+ struct kvm_s390_irq *buf;
+ int r = 0;
+ int n;
+
+ buf = vmalloc(len);
+ if (!buf)
+ return -ENOMEM;
+
+ if (copy_from_user((void *) buf, irqstate, len)) {
+ r = -EFAULT;
+ goto out_free;
+ }
+
+ /*
+ * Don't allow setting the interrupt state
+ * when there are already interrupts pending
+ */
+ spin_lock(&li->lock);
+ if (li->pending_irqs) {
+ r = -EBUSY;
+ goto out_unlock;
+ }
+
+ for (n = 0; n < len / sizeof(*buf); n++) {
+ r = do_inject_vcpu(vcpu, &buf[n]);
+ if (r)
+ break;
+ }
+
+out_unlock:
+ spin_unlock(&li->lock);
+out_free:
+ vfree(buf);
+
+ return r;
+}
+
+static void store_local_irq(struct kvm_s390_local_interrupt *li,
+ struct kvm_s390_irq *irq,
+ unsigned long irq_type)
+{
+ switch (irq_type) {
+ case IRQ_PEND_MCHK_EX:
+ case IRQ_PEND_MCHK_REP:
+ irq->type = KVM_S390_MCHK;
+ irq->u.mchk = li->irq.mchk;
+ break;
+ case IRQ_PEND_PROG:
+ irq->type = KVM_S390_PROGRAM_INT;
+ irq->u.pgm = li->irq.pgm;
+ break;
+ case IRQ_PEND_PFAULT_INIT:
+ irq->type = KVM_S390_INT_PFAULT_INIT;
+ irq->u.ext = li->irq.ext;
+ break;
+ case IRQ_PEND_EXT_EXTERNAL:
+ irq->type = KVM_S390_INT_EXTERNAL_CALL;
+ irq->u.extcall = li->irq.extcall;
+ break;
+ case IRQ_PEND_EXT_CLOCK_COMP:
+ irq->type = KVM_S390_INT_CLOCK_COMP;
+ break;
+ case IRQ_PEND_EXT_CPU_TIMER:
+ irq->type = KVM_S390_INT_CPU_TIMER;
+ break;
+ case IRQ_PEND_SIGP_STOP:
+ irq->type = KVM_S390_SIGP_STOP;
+ irq->u.stop = li->irq.stop;
+ break;
+ case IRQ_PEND_RESTART:
+ irq->type = KVM_S390_RESTART;
+ break;
+ case IRQ_PEND_SET_PREFIX:
+ irq->type = KVM_S390_SIGP_SET_PREFIX;
+ irq->u.prefix = li->irq.prefix;
+ break;
+ }
+}
+
+int kvm_s390_get_irq_state(struct kvm_vcpu *vcpu, __u8 __user *buf, int len)
+{
+ uint8_t sigp_ctrl = vcpu->kvm->arch.sca->cpu[vcpu->vcpu_id].sigp_ctrl;
+ unsigned long sigp_emerg_pending[BITS_TO_LONGS(KVM_MAX_VCPUS)];
+ struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int;
+ unsigned long pending_irqs;
+ struct kvm_s390_irq irq;
+ unsigned long irq_type;
+ int cpuaddr;
+ int n = 0;
+
+ spin_lock(&li->lock);
+ pending_irqs = li->pending_irqs;
+ memcpy(&sigp_emerg_pending, &li->sigp_emerg_pending,
+ sizeof(sigp_emerg_pending));
+ spin_unlock(&li->lock);
+
+ for_each_set_bit(irq_type, &pending_irqs, IRQ_PEND_COUNT) {
+ memset(&irq, 0, sizeof(irq));
+ if (irq_type == IRQ_PEND_EXT_EMERGENCY)
+ continue;
+ if (n + sizeof(irq) > len)
+ return -ENOBUFS;
+ store_local_irq(&vcpu->arch.local_int, &irq, irq_type);
+ if (copy_to_user(&buf[n], &irq, sizeof(irq)))
+ return -EFAULT;
+ n += sizeof(irq);
+ }
+
+ if (test_bit(IRQ_PEND_EXT_EMERGENCY, &pending_irqs)) {
+ for_each_set_bit(cpuaddr, sigp_emerg_pending, KVM_MAX_VCPUS) {
+ memset(&irq, 0, sizeof(irq));
+ if (n + sizeof(irq) > len)
+ return -ENOBUFS;
+ irq.type = KVM_S390_INT_EMERGENCY;
+ irq.u.emerg.code = cpuaddr;
+ if (copy_to_user(&buf[n], &irq, sizeof(irq)))
+ return -EFAULT;
+ n += sizeof(irq);
+ }
+ }
+
+ if ((sigp_ctrl & SIGP_CTRL_C) &&
+ (atomic_read(&vcpu->arch.sie_block->cpuflags) &
+ CPUSTAT_ECALL_PEND)) {
+ if (n + sizeof(irq) > len)
+ return -ENOBUFS;
+ memset(&irq, 0, sizeof(irq));
+ irq.type = KVM_S390_INT_EXTERNAL_CALL;
+ irq.u.extcall.code = sigp_ctrl & SIGP_CTRL_SCN_MASK;
+ if (copy_to_user(&buf[n], &irq, sizeof(irq)))
+ return -EFAULT;
+ n += sizeof(irq);
+ }
+
+ return n;
+}
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 8bc25d4..3040b14 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -41,6 +41,9 @@
#include "trace-s390.h"
#define MEM_OP_MAX_SIZE 65536 /* Maximum transfer size for KVM_S390_MEM_OP */
+#define LOCAL_IRQS 32
+#define VCPU_IRQS_MAX_BUF (sizeof(struct kvm_s390_irq) * \
+ (KVM_MAX_VCPUS + LOCAL_IRQS))
#define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU
@@ -181,6 +184,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
case KVM_CAP_S390_USER_SIGP:
case KVM_CAP_S390_USER_STSI:
case KVM_CAP_S390_SKEYS:
+ case KVM_CAP_S390_IRQ_STATE:
r = 1;
break;
case KVM_CAP_S390_MEM_OP:
@@ -2500,6 +2504,38 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
r = -EFAULT;
break;
}
+ case KVM_S390_SET_IRQ_STATE: {
+ struct kvm_s390_irq_state irq_state;
+
+ r = -EFAULT;
+ if (copy_from_user(&irq_state, argp, sizeof(irq_state)))
+ break;
+ if (irq_state.len > VCPU_IRQS_MAX_BUF ||
+ irq_state.len == 0 ||
+ irq_state.len % sizeof(struct kvm_s390_irq) > 0) {
+ r = -EINVAL;
+ break;
+ }
+ r = kvm_s390_set_irq_state(vcpu,
+ (void __user *) irq_state.buf,
+ irq_state.len);
+ break;
+ }
+ case KVM_S390_GET_IRQ_STATE: {
+ struct kvm_s390_irq_state irq_state;
+
+ r = -EFAULT;
+ if (copy_from_user(&irq_state, argp, sizeof(irq_state)))
+ break;
+ if (irq_state.len == 0) {
+ r = -EINVAL;
+ break;
+ }
+ r = kvm_s390_get_irq_state(vcpu,
+ (__u8 __user *) irq_state.buf,
+ irq_state.len);
+ break;
+ }
default:
r = -ENOTTY;
}
diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h
index 343644a..ca108b9 100644
--- a/arch/s390/kvm/kvm-s390.h
+++ b/arch/s390/kvm/kvm-s390.h
@@ -272,6 +272,10 @@ int kvm_s390_ext_call_pending(struct kvm_vcpu *vcpu);
extern struct kvm_device_ops kvm_flic_ops;
int kvm_s390_is_stop_irq_pending(struct kvm_vcpu *vcpu);
void kvm_s390_clear_stop_irq(struct kvm_vcpu *vcpu);
+int kvm_s390_set_irq_state(struct kvm_vcpu *vcpu,
+ void __user *buf, int len);
+int kvm_s390_get_irq_state(struct kvm_vcpu *vcpu,
+ __u8 __user *buf, int len);
/* implemented in guestdbg.c */
void kvm_s390_backup_guest_per_regs(struct kvm_vcpu *vcpu);
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index e61f49a..7437ce9 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -558,6 +558,13 @@ struct kvm_s390_irq {
} u;
};
+struct kvm_s390_irq_state {
+ __u64 buf;
+ __u32 flags;
+ __u32 len;
+ __u32 reserved[4];
+};
+
/* for KVM_SET_GUEST_DEBUG */
#define KVM_GUESTDBG_ENABLE 0x00000001
@@ -803,6 +810,7 @@ struct kvm_ppc_smmu_info {
#define KVM_CAP_S390_USER_STSI 109
#define KVM_CAP_S390_SKEYS 110
#define KVM_CAP_S390_INJECT_IRQ 111
+#define KVM_CAP_S390_IRQ_STATE 112
#ifdef KVM_CAP_IRQ_ROUTING
@@ -1185,6 +1193,9 @@ struct kvm_s390_ucas_mapping {
#define KVM_S390_SET_SKEYS _IOW(KVMIO, 0xb3, struct kvm_s390_skeys)
/* Available with KVM_CAP_S390_INJECT_IRQ */
#define KVM_S390_IRQ _IOW(KVMIO, 0xb4, struct kvm_s390_irq)
+/* Available with KVM_CAP_S390_IRQ_STATE */
+#define KVM_S390_SET_IRQ_STATE _IOW(KVMIO, 0xb5, struct kvm_s390_irq_state)
+#define KVM_S390_GET_IRQ_STATE _IOW(KVMIO, 0xb6, struct kvm_s390_irq_state)
#define KVM_DEV_ASSIGN_ENABLE_IOMMU (1 << 0)
#define KVM_DEV_ASSIGN_PCI_2_3 (1 << 1)
--
2.3.0
^ permalink raw reply related [flat|nested] 13+ messages in thread