/* * IRQ domain support for SH INTC subsystem * * Copyright (C) 2012 Paul Mundt * * This file is subject to the terms and conditions of the GNU General Public * License. See the file "COPYING" in the main directory of this archive * for more details. */ #define pr_fmt(fmt) "intc: " fmt #include <linux/irqdomain.h> #include <linux/sh_intc.h> #include <linux/export.h> #include "internals.h" /** * intc_irq_domain_evt_xlate() - Generic xlate for vectored IRQs. * * This takes care of exception vector to hwirq translation through * by way of evt2irq() translation. * * Note: For platforms that use a flat vector space without INTEVT this * basically just mimics irq_domain_xlate_onecell() by way of a nopped * out evt2irq() implementation. */ static int intc_evt_xlate(struct irq_domain *d, struct device_node *ctrlr, const u32 *intspec, unsigned int intsize, unsigned long *out_hwirq, unsigned int *out_type) { if (WARN_ON(intsize < 1)) return -EINVAL; *out_hwirq = evt2irq(intspec[0]); *out_type = IRQ_TYPE_NONE; return 0; } static const struct irq_domain_ops intc_evt_ops = { .xlate = intc_evt_xlate, }; void __init intc_irq_domain_init(struct intc_desc_int *d, struct intc_hw_desc *hw) { unsigned int irq_base, irq_end; /* * Quick linear revmap check */ irq_base = evt2irq(hw->vectors[0].vect); irq_end = evt2irq(hw->vectors[hw->nr_vectors - 1].vect); /* * Linear domains have a hard-wired assertion that IRQs start at * 0 in order to make some performance optimizations. Lamely * restrict the linear case to these conditions here, taking the * tree penalty for linear cases with non-zero hwirq bases. */ if (irq_base == 0 && irq_end == (irq_base + hw->nr_vectors - 1)) d->domain = irq_domain_add_linear(NULL, hw->nr_vectors, &intc_evt_ops, NULL); else d->domain = irq_domain_add_tree(NULL, &intc_evt_ops, NULL); BUG_ON(!d->domain); } |