Re: [PATCH v3 7/9] s390: ap: implement PAPQ AQIC interception in kernel

From: Pierre Morel
Date: Tue Feb 19 2019 - 14:16:45 EST

On 16/02/2019 00:11, Tony Krowiak wrote:
On 2/14/19 8:51 AM, Pierre Morel wrote:
We register the AP PQAP instruction hook during the open
of the mediated device. And unregister it on release.

In the AP PQAP instruction hook, if we receive a demand to
enable IRQs,
- we retrieve the vfio_ap_queue based on the APQN we receive
ÂÂ in REG1,
- we retrieve the page of the guest address, (NIB), from
ÂÂ register REG2
- we the mediated device to use the VFIO pinning infratrsucture
ÂÂ to pin the page of the guest address,
- we retrieve the pointer to KVM to register the guest ISC
ÂÂ and retrieve the host ISC
- finaly we activate GISA

If we receive a demand to disable IRQs,
- we deactivate GISA
- unregister from the GIB
- unping the NIB

Signed-off-by: Pierre Morel <pmorel@xxxxxxxxxxxxx>
---
 drivers/s390/crypto/ap_bus.h | 1 +
 drivers/s390/crypto/vfio_ap_ops.c | 191 +++++++++++++++++++++++++++++++++-
 drivers/s390/crypto/vfio_ap_private.h | 2 +
 3 files changed, 191 insertions(+), 3 deletions(-)

diff --git a/drivers/s390/crypto/ap_bus.h b/drivers/s390/crypto/ap_bus.h
index bfc66e4..323f2aa 100644
--- a/drivers/s390/crypto/ap_bus.h
+++ b/drivers/s390/crypto/ap_bus.h
@@ -43,6 +43,7 @@ static inline int ap_test_bit(unsigned int *ptr, unsigned int nr)
 #define AP_RESPONSE_BUSY 0x05
 #define AP_RESPONSE_INVALID_ADDRESS 0x06
 #define AP_RESPONSE_OTHERWISE_CHANGED 0x07
+#define AP_RESPONSE_INVALID_GISAÂÂÂ 0x08
 #define AP_RESPONSE_Q_FULL 0x10
 #define AP_RESPONSE_NO_PENDING_REPLY 0x10
 #define AP_RESPONSE_INDEX_TOO_BIG 0x11
diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c
index 6eddc2c..5664cf3 100644
--- a/drivers/s390/crypto/vfio_ap_ops.c
+++ b/drivers/s390/crypto/vfio_ap_ops.c
@@ -77,6 +77,28 @@ static void vfio_ap_put_queue(struct vfio_ap_queue *q)
ÂÂÂÂÂ q->dev = NULL;
 }
+/**
+ * vfio_ap_free_irq:
+ * @q: The vfio_ap_queue
+ *
+ * Unpin the guest NIB
+ * Unregister the ISC from the GIB alert
+ * Clear the vfio_ap_queue intern fields
+ */
+static void vfio_ap_free_irq(struct vfio_ap_queue *q)
+{
+ÂÂÂ unsigned long pfn = q->nib >> PAGE_SHIFT;
+
+ÂÂÂ if (!q)
+ÂÂÂÂÂÂÂ return;
+ÂÂÂ if (q->nib)
+ÂÂÂÂÂÂÂ vfio_unpin_pages(mdev_dev(q->matrix->mdev), &pfn, 1);
+ÂÂÂ if (q->isc)
+ÂÂÂÂÂÂÂ kvm_s390_gisc_unregister(q->matrix->kvm, q->isc);
+ÂÂÂ q->nib = 0;
+ÂÂÂ q->isc = 0;
+}
+
 static void vfio_ap_matrix_init(struct ap_config_info *info,
ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ struct ap_matrix *matrix)
 {
@@ -98,6 +120,7 @@ static int vfio_ap_mdev_create(struct kobject *kobj, struct mdev_device *mdev)
ÂÂÂÂÂÂÂÂÂ return -ENOMEM;
ÂÂÂÂÂ }
+ÂÂÂ matrix_mdev->mdev = mdev;
ÂÂÂÂÂ vfio_ap_matrix_init(&matrix_dev->info, &matrix_mdev->matrix);
ÂÂÂÂÂ mdev_set_drvdata(mdev, matrix_mdev);
ÂÂÂÂÂ mutex_lock(&matrix_dev->lock);
@@ -781,6 +804,156 @@ static const struct attribute_group *vfio_ap_mdev_attr_groups[] = {
 };
 /**
+ * vfio_ap_clrirq: Disable Interruption for a APQN
+ *
+ * @dev: the device associated with the ap_queue
+ * @q:ÂÂ the vfio_ap_queue holding AQIC parameters
+ *
+ * Issue the host side PQAP/AQIC
+ * On success: unpin the NIB saved in *q and unregister from GIB
+ * interface
+ *
+ * Return the ap_queue_status returned by the ap_aqic()
+ */
+static struct ap_queue_status vfio_ap_clrirq(struct vfio_ap_queue *q)
+{
+ÂÂÂ struct ap_qirq_ctrl aqic_gisa = {};
+ÂÂÂ struct ap_queue_status status;
+
+ÂÂÂ status = ap_aqic(q->apqn, aqic_gisa, NULL);
+ÂÂÂ if (!status.response_code)
+ÂÂÂÂÂÂÂ vfio_ap_free_irq(q);
+
+ÂÂÂ return status;
+}
+
+/**
+ * vfio_ap_setirq: Enable Interruption for a APQN
+ *
+ * @dev: the device associated with the ap_queue
+ * @q:ÂÂ the vfio_ap_queue holding AQIC parameters
+ *
+ * Pin the NIB saved in *q
+ * Register the guest ISC to GIB interface and retrieve the
+ * host ISC to issue the host side PQAP/AQIC
+ *
+ * Response.status may be set to following Response Code in case of error:
+ * - AP_RESPONSE_INVALID_ADDRESS: vfio_pin_pages failed
+ * - AP_RESPONSE_OTHERWISE_CHANGED: Hypervizor GISA internal error
+ *
+ * Otherwise return the ap_queue_status returned by the ap_aqic()
+ */
+static struct ap_queue_status vfio_ap_setirq(struct vfio_ap_queue *q)
+{
+ÂÂÂ struct ap_qirq_ctrl aqic_gisa = {};
+ÂÂÂ struct ap_queue_status status = {};
+ÂÂÂ struct kvm_s390_gisa *gisa;
+ÂÂÂ struct kvm *kvm;
+ÂÂÂ unsigned long g_pfn, h_nib, h_pfn;
+ÂÂÂ int ret;
+
+ÂÂÂ kvm = q->matrix->kvm;
+ÂÂÂ gisa = kvm->arch.gisa_int.origin;
+
+ÂÂÂ g_pfn = q->nib >> PAGE_SHIFT;
+ÂÂÂ ret = vfio_pin_pages(mdev_dev(q->matrix->mdev), &g_pfn, 1,
+ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ IOMMU_READ | IOMMU_WRITE, &h_pfn);
+ÂÂÂ switch (ret) {
+ÂÂÂ case 1:
+ÂÂÂÂÂÂÂ break;
+ÂÂÂ case -EINVAL:
+ÂÂÂ case -E2BIG:
+ÂÂÂÂÂÂÂ status.response_code = AP_RESPONSE_INVALID_ADDRESS;
+ÂÂÂÂÂÂÂ /* Fallthrough */
+ÂÂÂ default:
+ÂÂÂÂÂÂÂ return status;
+ÂÂÂ }
+
+ÂÂÂ h_nib = (h_pfn << PAGE_SHIFT) | (q->nib & ~PAGE_MASK);
+ÂÂÂ aqic_gisa.gisc = q->isc;
+ÂÂÂ aqic_gisa.isc = kvm_s390_gisc_register(kvm, q->isc);
+ÂÂÂ aqic_gisa.ir = 1;
+ÂÂÂ aqic_gisa.gisa = gisa->next_alert >> 4;
+
+ÂÂÂ status = ap_aqic(q->apqn, aqic_gisa, (void *)h_nib);
+ÂÂÂ if (status.response_code == AP_RESPONSE_INVALID_GISA) {
+ÂÂÂÂÂÂÂ status.response_code = AP_RESPONSE_OTHERWISE_CHANGED;
+ÂÂÂÂÂÂÂ pr_warn("vfio_ap: apqn %02x.%04x: AP_RESPONSE_INVALID_GISA\n",
+ÂÂÂÂÂÂÂÂÂÂÂ (q->apqn >> 8) & 0xff, q->apqn & 0xff);
+ÂÂÂ }
+
+ÂÂÂ if (status.response_code)
+ÂÂÂÂÂÂÂ vfio_ap_free_irq(q);
+
+ÂÂÂ return status;
+}
+
+/**
+ * handle_pqap: PQAP instruction callback
+ *
+ * @vcpu: The vcpu on which we received the PQAP instruction
+ *
+ * Get the general register contents to initialize internal variables.
+ * REG[0]: APQN
+ * REG[1]: IR and ISC
+ * REG[2]: NIB
+ *
+ * Response.status may be set to following Response Code:
+ * - AP_RESPONSE_Q_NOT_AVAIL: if the queue is not available
+ * - AP_RESPONSE_DECONFIGURED: if the queue is not configured
+ * - AP_RESPONSE_NORMAL (0) : in case of successs
+ *ÂÂ Check vfio_ap_setirq() and vfio_ap_clrirq() for other possible RC.
+ *
+ * Return 0 if we could handle the request inside KVM.
+ * otherwise, returns -EOPNOTSUPP to let QEMU handle the fault.
+ */

This function be nothing more than a switch statement for the
function code sent with the PQAP instruction. Each case should
be a call to a appropriate PQAP function handler. This will make
it much easier to add additional handlers for the 6 other
PQAP functions if necessary at some time down the road.

AFAIK there are only two PQAP functions we can intercept.

I will give attention to this function.
I will wait to this on the answers from KVM maintainers to know which KVM functions I can use here.



+static int handle_pqap(struct kvm_vcpu *vcpu)
+{ÂÂÂ int ret.
ÂÂÂÂuint8_t fc;

ÂÂÂÂfc = vcpu->run->s.regs.gprs[0] >> 24;
ÂÂÂÂswitch(fc) {
ÂÂÂÂcase 0x03:
ÂÂÂÂÂÂÂ ret = handle_pqap_aqic(vcpu);
ÂÂÂÂdefault:
ÂÂÂÂÂÂÂ ret = -EOPNOTSUPP;
ÂÂÂÂ}

ÂÂÂÂreturn ret;
}

static int handle_pqap_aqic(struct kvm_vcpu *vcpu) {
+ÂÂÂ uint64_t status;
+ÂÂÂ uint16_t apqn;
ÂÂÂÂstruct device *qdev;
+ÂÂÂ struct vfio_ap_queue *q;
+ÂÂÂ struct ap_queue_status qstatus = {};
+ÂÂÂ struct ap_matrix_mdev *matrix_mdev;
+
+ÂÂÂ /* If we do not use the AIV facility just go to userland */
+ÂÂÂ if (!(vcpu->arch.sie_block->eca & ECA_AIV))
+ÂÂÂÂÂÂÂ return -EOPNOTSUPP;
+
+ÂÂÂ apqn = vcpu->run->s.regs.gprs[0] & 0xffff;
+ÂÂÂ q = vfio_ap_get_queue(apqn);

Replace with:
ÂÂÂÂqdev = vfio_ap_get_queue_dev(apqn);

You asked to revisit the life cycle of the vfio_ap_queues in another thread.
I will do it.

Thanks
Pierre

--
Pierre Morel
Linux/KVM/QEMU in BÃblingen - Germany