[v2,06/25] iommu, x86: No need to migrating irq for VT-d Posted-Interrupts

Commit Message

Wu, Feng Dec. 3, 2014, 7:39 a.m. UTC

We don't need to migrate the irqs for VT-d Posted-Interrupts here.
When 'pst' is set in IRTE, the associated irq will be posted to
guests instead of interrupt remapping. The destination of the
interrupt is set in Posted-Interrupts Descriptor, and the migration
happens during vCPU scheduling.

Signed-off-by: Feng Wu <[e-mail]>
Reviewed-by: Jiang Liu <[e-mail]>
---
 drivers/iommu/intel_irq_remapping.c |   10 ++++++++++
 1 files changed, 10 insertions(+), 0 deletions(-)

Patch

diff --git a/drivers/iommu/intel_irq_remapping.c b/drivers/iommu/intel_irq_remapping.c
index 749cb93..01786a8 100644
--- a/drivers/iommu/intel_irq_remapping.c
+++ b/drivers/iommu/intel_irq_remapping.c
@@ -976,10 +976,20 @@  intel_ir_set_affinity(struct irq_data *data, const struct cpumask *mask,
 {
 	struct intel_ir_data *ir_data = data->chip_data;
 	struct irte *irte = &ir_data->irte_entry;
+	struct irte_pi *irte_pi = (struct irte_pi *)irte;
 	struct irq_cfg *cfg = irqd_cfg(data);
 	struct irq_data *parent = data->parent_data;
 	int ret;
 
+	/*
+	 * If the interrupt is for posting, it is used by guests,
+	 * we cannot set irq affinity here.
+	 */
+	if (irte_pi->pst == 1) {
+		pr_warn("cannot set irq affinity for posted-interrupts\n");
+		return -EBUSY;
+	}
+
 	ret = parent->chip->irq_set_affinity(parent, mask, force);
 	if (ret < 0 || ret == IRQ_SET_MASK_OK_DONE)
 		return ret;