[RFC PATCH 07/16] KVM: PPC: Book3S HV: add a GET_ESB_FD control to the XIVE native device

["Cédric Le Goater" <clg@kaod.org>]

This will let the guest create a memory mapping to expose the ESB MMIO
regions used to control the interrupt sources, to trigger events, to
EOI or to turn off the sources.

Signed-off-by: Cédric Le Goater <clg@kaod.org>
---
 arch/powerpc/include/uapi/asm/kvm.h   |  2 +
 arch/powerpc/kvm/book3s_xive_native.c | 88 +++++++++++++++++++++++++++++++++++
 2 files changed, 90 insertions(+)

diff --git a/arch/powerpc/include/uapi/asm/kvm.h b/arch/powerpc/include/uapi/asm/kvm.h
index 84fde0295939..cf54fad1faf6 100644
--- a/arch/powerpc/include/uapi/asm/kvm.h
+++ b/arch/powerpc/include/uapi/asm/kvm.h
@@ -675,6 +675,8 @@ struct kvm_ppc_cpu_char {
 
 /* POWER9 XIVE Native Interrupt Controller */
 #define KVM_DEV_XIVE_GRP_SOURCES	1	/* 64-bit source attributes */
+#define KVM_DEV_XIVE_GRP_CTRL		2
+#define   KVM_DEV_XIVE_GET_ESB_FD	1
 
 /* Layout of 64-bit XIVE source attribute values */
 #define  KVM_XIVE_LEVEL_SENSITIVE	(1ULL << 0)
diff --git a/arch/powerpc/kvm/book3s_xive_native.c b/arch/powerpc/kvm/book3s_xive_native.c
index 6ebb628462d4..395298c5fca8 100644
--- a/arch/powerpc/kvm/book3s_xive_native.c
+++ b/arch/powerpc/kvm/book3s_xive_native.c
@@ -249,6 +249,78 @@ static int kvmppc_xive_native_set_source(struct kvmppc_xive *xive, long irq,
 	return 0;
 }
 
+static int xive_native_esb_fault(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct kvmppc_xive *xive = vma->vm_file->private_data;
+	struct kvmppc_xive_src_block *sb;
+	struct kvmppc_xive_irq_state *state;
+	struct xive_irq_data *xd;
+	u32 hw_num;
+	u16 src;
+	unsigned long irq, pfn;
+
+	/* There are two pages (trigger and EOI) per IRQ */
+	irq = vmf->pgoff / 2;
+
+	sb = kvmppc_xive_find_source(xive, irq, &src);
+	if (!sb) {
+		pr_err("%s: source %lx not found !\n", __func__, irq);
+		return VM_FAULT_SIGBUS;
+	}
+
+	state = &sb->irq_state[src];
+	kvmppc_xive_select_irq(state, &hw_num, &xd);
+
+	arch_spin_lock(&sb->lock);
+
+	/* First page is the trigger page */
+	if (vmf->pgoff % 2)
+		pfn = xd->eoi_page >> PAGE_SHIFT;
+	else
+		pfn = xd->trig_page >> PAGE_SHIFT;
+
+	arch_spin_unlock(&sb->lock);
+
+	vm_insert_pfn(vma, vmf->address, pfn);
+	return VM_FAULT_NOPAGE;
+}
+
+static const struct vm_operations_struct xive_native_esb_vmops = {
+	.fault = xive_native_esb_fault,
+};
+
+static int xive_native_esb_mmap(struct file *file, struct vm_area_struct *vma)
+{
+	unsigned long len = vma->vm_end - vma->vm_start;
+
+	/* There are two pages (trigger and EOI) per IRQ */
+	if ((len >> PAGE_SHIFT) >= KVMPPC_XIVE_NR_IRQS * 2)
+		return -EINVAL;
+
+	vma->vm_flags |= VM_IO | VM_PFNMAP;
+	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+	vma->vm_ops = &xive_native_esb_vmops;
+	return 0;
+}
+
+static const struct file_operations xive_native_esb_fops = {
+	.mmap = xive_native_esb_mmap,
+};
+
+static int kvmppc_xive_native_get_esb_fd(struct kvmppc_xive *xive, u64 addr)
+{
+	u64 __user *ubufp = (u64 __user *) addr;
+	int ret;
+
+	ret = anon_inode_getfd("esb", &xive_native_esb_fops, xive,
+				O_RDWR | O_CLOEXEC);
+	if (ret < 0)
+		return ret;
+
+	return put_user(ret, ubufp);
+}
+
 static int kvmppc_xive_native_set_attr(struct kvm_device *dev,
 				       struct kvm_device_attr *attr)
 {
@@ -265,6 +337,16 @@ static int kvmppc_xive_native_set_attr(struct kvm_device *dev,
 static int kvmppc_xive_native_get_attr(struct kvm_device *dev,
 				       struct kvm_device_attr *attr)
 {
+	struct kvmppc_xive *xive = dev->private;
+
+	switch (attr->group) {
+	case KVM_DEV_XIVE_GRP_CTRL:
+		switch (attr->attr) {
+		case KVM_DEV_XIVE_GET_ESB_FD:
+			return kvmppc_xive_native_get_esb_fd(xive, attr->addr);
+		}
+		break;
+	}
 	return -ENXIO;
 }
 
@@ -277,6 +359,12 @@ static int kvmppc_xive_native_has_attr(struct kvm_device *dev,
 		    attr->attr < KVMPPC_XIVE_NR_IRQS)
 			return 0;
 		break;
+	case KVM_DEV_XIVE_GRP_CTRL:
+		switch (attr->attr) {
+		case KVM_DEV_XIVE_GET_ESB_FD:
+			return 0;
+		}
+		break;
 	}
 	return -ENXIO;
 }
-- 
2.13.6