genirq/msi: Switch to new irq spreading infrastructure

Switch MSI over to the new spreading code. If a pci device contains a valid
pointer to a cpumask, then this mask is used for spreading otherwise the
online cpu mask is used. This allows a driver to restrict the spread to a
subset of CPUs, e.g. cpus on a particular node.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Christoph Hellwig <hch@lst.de>
Cc: axboe@fb.com
Cc: keith.busch@intel.com
Cc: agordeev@redhat.com
Cc: linux-block@vger.kernel.org
Link: http://lkml.kernel.org/r/1473862739-15032-4-git-send-email-hch@lst.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

2 files changed, 87 insertions, 72 deletions

diff --git a/drivers/pci/msi.c b/drivers/pci/msi.c
index 0db72ba24003..06100dde0e86 100644
--- a/drivers/pci/msi.c
+++ b/drivers/pci/msi.c
@@ -549,15 +549,23 @@ error_attrs:
         return ret;
 }
 
-static struct msi_desc *msi_setup_entry(struct pci_dev *dev, int nvec)
+static struct msi_desc *
+msi_setup_entry(struct pci_dev *dev, int nvec, bool affinity)
 {
-        u16 control;
+        struct cpumask *masks = NULL;
         struct msi_desc *entry;
+        u16 control;
+
+        if (affinity) {
+                masks = irq_create_affinity_masks(dev->irq_affinity, nvec);
+                if (!masks)
+                        pr_err("Unable to allocate affinity masks, ignoring\n");
+        }
 
         /* MSI Entry Initialization */
-        entry = alloc_msi_entry(&dev->dev, nvec, NULL);
+        entry = alloc_msi_entry(&dev->dev, nvec, masks);
         if (!entry)
-                return NULL;
+                goto out;
 
         pci_read_config_word(dev, dev->msi_cap + PCI_MSI_FLAGS, &control);
 
@@ -568,7 +576,6 @@ static struct msi_desc *msi_setup_entry(struct pci_dev *dev, int nvec)
         entry->msi_attrib.default_irq   = dev->irq;     /* Save IOAPIC IRQ */
         entry->msi_attrib.multi_cap     = (control & PCI_MSI_FLAGS_QMASK) >> 1;
         entry->msi_attrib.multiple      = ilog2(__roundup_pow_of_two(nvec));
-        entry->affinity                 = dev->irq_affinity;
 
         if (control & PCI_MSI_FLAGS_64BIT)
                 entry->mask_pos = dev->msi_cap + PCI_MSI_MASK_64;
@@ -579,6 +586,8 @@ static struct msi_desc *msi_setup_entry(struct pci_dev *dev, int nvec)
         if (entry->msi_attrib.maskbit)
                 pci_read_config_dword(dev, entry->mask_pos, &entry->masked);
 
+out:
+        kfree(masks);
         return entry;
 }
 
@@ -607,7 +616,7 @@ static int msi_verify_entries(struct pci_dev *dev)
  * an error, and a positive return value indicates the number of interrupts
  * which could have been allocated.
  */
-static int msi_capability_init(struct pci_dev *dev, int nvec)
+static int msi_capability_init(struct pci_dev *dev, int nvec, bool affinity)
 {
         struct msi_desc *entry;
         int ret;
@@ -615,7 +624,7 @@ static int msi_capability_init(struct pci_dev *dev, int nvec)
 
         pci_msi_set_enable(dev, 0);     /* Disable MSI during set up */
 
-        entry = msi_setup_entry(dev, nvec);
+        entry = msi_setup_entry(dev, nvec, affinity);
         if (!entry)
                 return -ENOMEM;
 
@@ -678,28 +687,29 @@ static void __iomem *msix_map_region(struct pci_dev *dev, unsigned nr_entries)
 }
 
 static int msix_setup_entries(struct pci_dev *dev, void __iomem *base,
-                              struct msix_entry *entries, int nvec)
+                              struct msix_entry *entries, int nvec,
+                              bool affinity)
 {
-        const struct cpumask *mask = NULL;
+        struct cpumask *curmsk, *masks = NULL;
         struct msi_desc *entry;
-        int cpu = -1, i;
-
-        for (i = 0; i < nvec; i++) {
-                if (dev->irq_affinity) {
-                        cpu = cpumask_next(cpu, dev->irq_affinity);
-                        if (cpu >= nr_cpu_ids)
-                                cpu = cpumask_first(dev->irq_affinity);
-                        mask = cpumask_of(cpu);
-                }
+        int ret, i;
 
-                entry = alloc_msi_entry(&dev->dev, 1, NULL);
+        if (affinity) {
+                masks = irq_create_affinity_masks(dev->irq_affinity, nvec);
+                if (!masks)
+                        pr_err("Unable to allocate affinity masks, ignoring\n");
+        }
+
+        for (i = 0, curmsk = masks; i < nvec; i++) {
+                entry = alloc_msi_entry(&dev->dev, 1, curmsk);
                 if (!entry) {
                         if (!i)
                                 iounmap(base);
                         else
                                 free_msi_irqs(dev);
                         /* No enough memory. Don't try again */
-                        return -ENOMEM;
+                        ret = -ENOMEM;
+                        goto out;
                 }
 
                 entry->msi_attrib.is_msix       = 1;
@@ -710,11 +720,14 @@ static int msix_setup_entries(struct pci_dev *dev, void __iomem *base,
                         entry->msi_attrib.entry_nr = i;
                 entry->msi_attrib.default_irq   = dev->irq;
                 entry->mask_base                = base;
-                entry->affinity                 = mask;
 
                 list_add_tail(&entry->list, dev_to_msi_list(&dev->dev));
+                if (masks)
+                        curmsk++;
         }
-
+        ret = 0;
+out:
+        kfree(masks);
         return 0;
 }
 
@@ -743,8 +756,8 @@ static void msix_program_entries(struct pci_dev *dev,
  * single MSI-X irq. A return of zero indicates the successful setup of
  * requested MSI-X entries with allocated irqs or non-zero for otherwise.
  **/
-static int msix_capability_init(struct pci_dev *dev,
-                                struct msix_entry *entries, int nvec)
+static int msix_capability_init(struct pci_dev *dev, struct msix_entry *entries,
+                                int nvec, bool affinity)
 {
         int ret;
         u16 control;
@@ -759,7 +772,7 @@ static int msix_capability_init(struct pci_dev *dev,
         if (!base)
                 return -ENOMEM;
 
-        ret = msix_setup_entries(dev, base, entries, nvec);
+        ret = msix_setup_entries(dev, base, entries, nvec, affinity);
         if (ret)
                 return ret;
 
@@ -939,22 +952,8 @@ int pci_msix_vec_count(struct pci_dev *dev)
 }
 EXPORT_SYMBOL(pci_msix_vec_count);
 
-/**
- * pci_enable_msix - configure device's MSI-X capability structure
- * @dev: pointer to the pci_dev data structure of MSI-X device function
- * @entries: pointer to an array of MSI-X entries (optional)
- * @nvec: number of MSI-X irqs requested for allocation by device driver
- *
- * Setup the MSI-X capability structure of device function with the number
- * of requested irqs upon its software driver call to request for
- * MSI-X mode enabled on its hardware device function. A return of zero
- * indicates the successful configuration of MSI-X capability structure
- * with new allocated MSI-X irqs. A return of < 0 indicates a failure.
- * Or a return of > 0 indicates that driver request is exceeding the number
- * of irqs or MSI-X vectors available. Driver should use the returned value to
- * re-send its request.
- **/
-int pci_enable_msix(struct pci_dev *dev, struct msix_entry *entries, int nvec)
+static int __pci_enable_msix(struct pci_dev *dev, struct msix_entry *entries,
+                             int nvec, bool affinity)
 {
         int nr_entries;
         int i, j;
@@ -986,7 +985,27 @@ int pci_enable_msix(struct pci_dev *dev, struct msix_entry *entries, int nvec)
                 dev_info(&dev->dev, "can't enable MSI-X (MSI IRQ already assigned)\n");
                 return -EINVAL;
         }
-        return msix_capability_init(dev, entries, nvec);
+        return msix_capability_init(dev, entries, nvec, affinity);
+}
+
+/**
+ * pci_enable_msix - configure device's MSI-X capability structure
+ * @dev: pointer to the pci_dev data structure of MSI-X device function
+ * @entries: pointer to an array of MSI-X entries (optional)
+ * @nvec: number of MSI-X irqs requested for allocation by device driver
+ *
+ * Setup the MSI-X capability structure of device function with the number
+ * of requested irqs upon its software driver call to request for
+ * MSI-X mode enabled on its hardware device function. A return of zero
+ * indicates the successful configuration of MSI-X capability structure
+ * with new allocated MSI-X irqs. A return of < 0 indicates a failure.
+ * Or a return of > 0 indicates that driver request is exceeding the number
+ * of irqs or MSI-X vectors available. Driver should use the returned value to
+ * re-send its request.
+ **/
+int pci_enable_msix(struct pci_dev *dev, struct msix_entry *entries, int nvec)
+{
+        return __pci_enable_msix(dev, entries, nvec, false);
 }
 EXPORT_SYMBOL(pci_enable_msix);
 
@@ -1039,6 +1058,7 @@ EXPORT_SYMBOL(pci_msi_enabled);
 static int __pci_enable_msi_range(struct pci_dev *dev, int minvec, int maxvec,
                 unsigned int flags)
 {
+        bool affinity = flags & PCI_IRQ_AFFINITY;
         int nvec;
         int rc;
 
@@ -1067,19 +1087,17 @@ static int __pci_enable_msi_range(struct pci_dev *dev, int minvec, int maxvec,
                 nvec = maxvec;
 
         for (;;) {
-                if (flags & PCI_IRQ_AFFINITY) {
-                        dev->irq_affinity = irq_create_affinity_mask(&nvec);
+                if (affinity) {
+                        nvec = irq_calc_affinity_vectors(dev->irq_affinity,
+                                        nvec);
                         if (nvec < minvec)
                                 return -ENOSPC;
                 }
 
-                rc = msi_capability_init(dev, nvec);
+                rc = msi_capability_init(dev, nvec, affinity);
                 if (rc == 0)
                         return nvec;
 
-                kfree(dev->irq_affinity);
-                dev->irq_affinity = NULL;
-
                 if (rc < 0)
                         return rc;
                 if (rc < minvec)
@@ -1111,26 +1129,24 @@ static int __pci_enable_msix_range(struct pci_dev *dev,
                 struct msix_entry *entries, int minvec, int maxvec,
                 unsigned int flags)
 {
-        int nvec = maxvec;
-        int rc;
+        bool affinity = flags & PCI_IRQ_AFFINITY;
+        int rc, nvec = maxvec;
 
         if (maxvec < minvec)
                 return -ERANGE;
 
         for (;;) {
-                if (flags & PCI_IRQ_AFFINITY) {
-                        dev->irq_affinity = irq_create_affinity_mask(&nvec);
+                if (affinity) {
+                        nvec = irq_calc_affinity_vectors(dev->irq_affinity,
+                                        nvec);
                         if (nvec < minvec)
                                 return -ENOSPC;
                 }
 
-                rc = pci_enable_msix(dev, entries, nvec);
+                rc = __pci_enable_msix(dev, entries, nvec, affinity);
                 if (rc == 0)
                         return nvec;
 
-                kfree(dev->irq_affinity);
-                dev->irq_affinity = NULL;
-
                 if (rc < 0)
                         return rc;
                 if (rc < minvec)
diff --git a/kernel/irq/irqdesc.c b/kernel/irq/irqdesc.c
index a623b44f2d4b..5a5a685aba33 100644
--- a/kernel/irq/irqdesc.c
+++ b/kernel/irq/irqdesc.c
@@ -236,25 +236,24 @@ static int alloc_descs(unsigned int start, unsigned int cnt, int node,
         const struct cpumask *mask = NULL;
         struct irq_desc *desc;
         unsigned int flags;
-        int i, cpu = -1;
+        int i;
 
-        if (affinity && cpumask_empty(affinity))
-                return -EINVAL;
+        /* Validate affinity mask(s) */
+        if (affinity) {
+                for (i = 0, mask = affinity; i < cnt; i++, mask++) {
+                        if (cpumask_empty(mask))
+                                return -EINVAL;
+                }
+        }
 
         flags = affinity ? IRQD_AFFINITY_MANAGED : 0;
+        mask = NULL;
 
         for (i = 0; i < cnt; i++) {
                 if (affinity) {
-                        cpu = cpumask_next(cpu, affinity);
-                        if (cpu >= nr_cpu_ids)
-                                cpu = cpumask_first(affinity);
-                        node = cpu_to_node(cpu);
-
-                        /*
-                         * For single allocations we use the caller provided
-                         * mask otherwise we use the mask of the target cpu
-                         */
-                        mask = cnt == 1 ? affinity : cpumask_of(cpu);
+                        node = cpu_to_node(cpumask_first(affinity));
+                        mask = affinity;
+                        affinity++;
                 }
                 desc = alloc_desc(start + i, node, flags, mask, owner);
                 if (!desc)
@@ -481,9 +480,9 @@ EXPORT_SYMBOL_GPL(irq_free_descs);
  * @cnt:        Number of consecutive irqs to allocate.
  * @node:       Preferred node on which the irq descriptor should be allocated
  * @owner:      Owning module (can be NULL)
- * @affinity:   Optional pointer to an affinity mask which hints where the
- *              irq descriptors should be allocated and which default
- *              affinities to use
+ * @affinity:   Optional pointer to an affinity mask array of size @cnt which
+ *              hints where the irq descriptors should be allocated and which
+ *              default affinities to use
  *
  * Returns the first irq number or error code
  */