powerpc/pseries/pci: Add MSI domains

Two IRQ domains are added on top of default machine IRQ domain.

First, the top level "pSeries-PCI-MSI" domain deals with the MSI
specificities. In this domain, the HW IRQ numbers are generated by the
PCI MSI layer, they compose a unique ID for an MSI source with the PCI
device identifier and the MSI vector number.

These numbers can be quite large on a pSeries machine running under
the IBM Hypervisor and /sys/kernel/irq/ and /proc/interrupts will
require small fixes to show them correctly.

Second domain is the in-the-middle "pSeries-MSI" domain which acts as
a proxy between the PCI MSI subsystem and the machine IRQ subsystem.
It usually allocate the MSI vector numbers but, on pSeries machines,
this is done by the RTAS FW and RTAS returns IRQ numbers in the IRQ
number space of the machine. This is why the in-the-middle "pSeries-MSI"
domain has the same HW IRQ numbers as its parent domain.

Only the XIVE (P9/P10) parent domain is supported for now. We still
need to add support for IRQ domain hierarchy under XICS.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210701132750.1475580-6-clg@kaod.org

arch/powerpc/include/asm/pci-bridge.h

@@ -126,6 +126,11 @@ struct pci_controller {
 #endif	/* CONFIG_PPC64 */
 
 	void *private_data;
+
+	/* IRQ domain hierarchy */
+	struct irq_domain	*dev_domain;
+	struct irq_domain	*msi_domain;
+	struct fwnode_handle	*fwnode;
 };
 
 /* These are used for config access before all the PCI probing

arch/powerpc/kernel/pci-common.c

@@ -29,6 +29,7 @@
 #include <linux/slab.h>
 #include <linux/vgaarb.h>
 #include <linux/numa.h>
+#include <linux/msi.h>
 
 #include <asm/processor.h>
 #include <asm/io.h>
@@ -1060,11 +1061,16 @@ void pcibios_bus_add_device(struct pci_dev *dev)
 
 int pcibios_add_device(struct pci_dev *dev)
 {
+	struct irq_domain *d;
+
 #ifdef CONFIG_PCI_IOV
 	if (ppc_md.pcibios_fixup_sriov)
 		ppc_md.pcibios_fixup_sriov(dev);
 #endif /* CONFIG_PCI_IOV */
 
+	d = dev_get_msi_domain(&dev->bus->dev);
+	if (d)
+		dev_set_msi_domain(&dev->dev, d);
 	return 0;
 }
 

arch/powerpc/platforms/pseries/msi.c

@@ -13,6 +13,7 @@
 #include <asm/hw_irq.h>
 #include <asm/ppc-pci.h>
 #include <asm/machdep.h>
+#include <asm/xive.h>
 
 #include "pseries.h"
 
@@ -518,6 +519,190 @@ static int rtas_setup_msi_irqs(struct pci_dev *pdev, int nvec_in, int type)
 	return 0;
 }
 
+static int pseries_msi_ops_prepare(struct irq_domain *domain, struct device *dev,
+				   int nvec, msi_alloc_info_t *arg)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct msi_desc *desc = first_pci_msi_entry(pdev);
+	int type = desc->msi_attrib.is_msix ? PCI_CAP_ID_MSIX : PCI_CAP_ID_MSI;
+
+	return rtas_prepare_msi_irqs(pdev, nvec, type, arg);
+}
+
+static struct msi_domain_ops pseries_pci_msi_domain_ops = {
+	.msi_prepare	= pseries_msi_ops_prepare,
+};
+
+static void pseries_msi_shutdown(struct irq_data *d)
+{
+	d = d->parent_data;
+	if (d->chip->irq_shutdown)
+		d->chip->irq_shutdown(d);
+}
+
+static void pseries_msi_mask(struct irq_data *d)
+{
+	pci_msi_mask_irq(d);
+	irq_chip_mask_parent(d);
+}
+
+static void pseries_msi_unmask(struct irq_data *d)
+{
+	pci_msi_unmask_irq(d);
+	irq_chip_unmask_parent(d);
+}
+
+static struct irq_chip pseries_pci_msi_irq_chip = {
+	.name		= "pSeries-PCI-MSI",
+	.irq_shutdown	= pseries_msi_shutdown,
+	.irq_mask	= pseries_msi_mask,
+	.irq_unmask	= pseries_msi_unmask,
+	.irq_eoi	= irq_chip_eoi_parent,
+};
+
+static struct msi_domain_info pseries_msi_domain_info = {
+	.flags = (MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
+		  MSI_FLAG_MULTI_PCI_MSI  | MSI_FLAG_PCI_MSIX),
+	.ops   = &pseries_pci_msi_domain_ops,
+	.chip  = &pseries_pci_msi_irq_chip,
+};
+
+static void pseries_msi_compose_msg(struct irq_data *data, struct msi_msg *msg)
+{
+	__pci_read_msi_msg(irq_data_get_msi_desc(data), msg);
+}
+
+static struct irq_chip pseries_msi_irq_chip = {
+	.name			= "pSeries-MSI",
+	.irq_shutdown		= pseries_msi_shutdown,
+	.irq_mask		= irq_chip_mask_parent,
+	.irq_unmask		= irq_chip_unmask_parent,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_set_affinity	= irq_chip_set_affinity_parent,
+	.irq_compose_msi_msg	= pseries_msi_compose_msg,
+};
+
+static int pseries_irq_parent_domain_alloc(struct irq_domain *domain, unsigned int virq,
+					   irq_hw_number_t hwirq)
+{
+	struct irq_fwspec parent_fwspec;
+	int ret;
+
+	parent_fwspec.fwnode = domain->parent->fwnode;
+	parent_fwspec.param_count = 2;
+	parent_fwspec.param[0] = hwirq;
+	parent_fwspec.param[1] = IRQ_TYPE_EDGE_RISING;
+
+	ret = irq_domain_alloc_irqs_parent(domain, virq, 1, &parent_fwspec);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static int pseries_irq_domain_alloc(struct irq_domain *domain, unsigned int virq,
+				    unsigned int nr_irqs, void *arg)
+{
+	struct pci_controller *phb = domain->host_data;
+	msi_alloc_info_t *info = arg;
+	struct msi_desc *desc = info->desc;
+	struct pci_dev *pdev = msi_desc_to_pci_dev(desc);
+	int hwirq;
+	int i, ret;
+
+	hwirq = rtas_query_irq_number(pci_get_pdn(pdev), desc->msi_attrib.entry_nr);
+	if (hwirq < 0) {
+		dev_err(&pdev->dev, "Failed to query HW IRQ: %d\n", hwirq);
+		return hwirq;
+	}
+
+	dev_dbg(&pdev->dev, "%s bridge %pOF %d/%x #%d\n", __func__,
+		phb->dn, virq, hwirq, nr_irqs);
+
+	for (i = 0; i < nr_irqs; i++) {
+		ret = pseries_irq_parent_domain_alloc(domain, virq + i, hwirq + i);
+		if (ret)
+			goto out;
+
+		irq_domain_set_hwirq_and_chip(domain, virq + i, hwirq + i,
+					      &pseries_msi_irq_chip, domain->host_data);
+	}
+
+	return 0;
+
+out:
+	/* TODO: handle RTAS cleanup in ->msi_finish() ? */
+	irq_domain_free_irqs_parent(domain, virq, i - 1);
+	return ret;
+}
+
+static void pseries_irq_domain_free(struct irq_domain *domain, unsigned int virq,
+				    unsigned int nr_irqs)
+{
+	struct irq_data *d = irq_domain_get_irq_data(domain, virq);
+	struct pci_controller *phb = irq_data_get_irq_chip_data(d);
+
+	pr_debug("%s bridge %pOF %d #%d\n", __func__, phb->dn, virq, nr_irqs);
+
+	irq_domain_free_irqs_parent(domain, virq, nr_irqs);
+}
+
+static const struct irq_domain_ops pseries_irq_domain_ops = {
+	.alloc  = pseries_irq_domain_alloc,
+	.free   = pseries_irq_domain_free,
+};
+
+static int __pseries_msi_allocate_domains(struct pci_controller *phb,
+					  unsigned int count)
+{
+	struct irq_domain *parent = irq_get_default_host();
+
+	phb->fwnode = irq_domain_alloc_named_id_fwnode("pSeries-MSI",
+						       phb->global_number);
+	if (!phb->fwnode)
+		return -ENOMEM;
+
+	phb->dev_domain = irq_domain_create_hierarchy(parent, 0, count,
+						      phb->fwnode,
+						      &pseries_irq_domain_ops, phb);
+	if (!phb->dev_domain) {
+		pr_err("PCI: failed to create IRQ domain bridge %pOF (domain %d)\n",
+		       phb->dn, phb->global_number);
+		irq_domain_free_fwnode(phb->fwnode);
+		return -ENOMEM;
+	}
+
+	phb->msi_domain = pci_msi_create_irq_domain(of_node_to_fwnode(phb->dn),
+						    &pseries_msi_domain_info,
+						    phb->dev_domain);
+	if (!phb->msi_domain) {
+		pr_err("PCI: failed to create MSI IRQ domain bridge %pOF (domain %d)\n",
+		       phb->dn, phb->global_number);
+		irq_domain_free_fwnode(phb->fwnode);
+		irq_domain_remove(phb->dev_domain);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+int pseries_msi_allocate_domains(struct pci_controller *phb)
+{
+	int count;
+
+	/* Only supported by the XIVE driver */
+	if (!xive_enabled())
+		return -ENODEV;
+
+	if (!__find_pe_total_msi(phb->dn, &count)) {
+		pr_err("PCI: failed to find MSIs for bridge %pOF (domain %d)\n",
+		       phb->dn, phb->global_number);
+		return -ENOSPC;
+	}
+
+	return __pseries_msi_allocate_domains(phb, count);
+}
+
 static void rtas_msi_pci_irq_fixup(struct pci_dev *pdev)
 {
 	/* No LSI -> leave MSIs (if any) configured */

arch/powerpc/platforms/pseries/pseries.h

@@ -85,6 +85,7 @@ struct pci_host_bridge;
 int pseries_root_bridge_prepare(struct pci_host_bridge *bridge);
 
 extern struct pci_controller_ops pseries_pci_controller_ops;
+int pseries_msi_allocate_domains(struct pci_controller *phb);
 
 unsigned long pseries_memory_block_size(void);
 

arch/powerpc/platforms/pseries/setup.c

@@ -486,6 +486,8 @@ static void __init pSeries_discover_phbs(void)
 
 		/* create pci_dn's for DT nodes under this PHB */
 		pci_devs_phb_init_dynamic(phb);
+
+		pseries_msi_allocate_domains(phb);
 	}
 
 	of_node_put(root);