iommu.c

linux 内核源代码

/*
 * Copyright (C) 2001 Mike Corrigan & Dave Engebretsen, IBM Corporation
 * 
 * Rewrite, cleanup, new allocation schemes, virtual merging: 
 * Copyright (C) 2004 Olof Johansson, IBM Corporation
 *               and  Ben. Herrenschmidt, IBM Corporation
 *
 * Dynamic DMA mapping support, bus-independent parts.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 * 
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 * 
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
 */


#include <linux/init.h>
#include <linux/types.h>
#include <linux/slab.h>
#include <linux/mm.h>
#include <linux/spinlock.h>
#include <linux/string.h>
#include <linux/dma-mapping.h>
#include <linux/bitops.h>
#include <asm/io.h>
#include <asm/prom.h>
#include <asm/iommu.h>
#include <asm/pci-bridge.h>
#include <asm/machdep.h>
#include <asm/kdump.h>

#define DBG(...)

#ifdef CONFIG_IOMMU_VMERGE
static int novmerge = 0;
#else
static int novmerge = 1;
#endif

static int protect4gb = 1;

static inline unsigned long iommu_num_pages(unsigned long vaddr,
					    unsigned long slen)
{
	unsigned long npages;

	npages = IOMMU_PAGE_ALIGN(vaddr + slen) - (vaddr & IOMMU_PAGE_MASK);
	npages >>= IOMMU_PAGE_SHIFT;

	return npages;
}

static int __init setup_protect4gb(char *str)
{
	if (strcmp(str, "on") == 0)
		protect4gb = 1;
	else if (strcmp(str, "off") == 0)
		protect4gb = 0;

	return 1;
}

static int __init setup_iommu(char *str)
{
	if (!strcmp(str, "novmerge"))
		novmerge = 1;
	else if (!strcmp(str, "vmerge"))
		novmerge = 0;
	return 1;
}

__setup("protect4gb=", setup_protect4gb);
__setup("iommu=", setup_iommu);

static unsigned long iommu_range_alloc(struct iommu_table *tbl,
                                       unsigned long npages,
                                       unsigned long *handle,
                                       unsigned long mask,
                                       unsigned int align_order)
{ 
	unsigned long n, end, i, start;
	unsigned long limit;
	int largealloc = npages > 15;
	int pass = 0;
	unsigned long align_mask;

	align_mask = 0xffffffffffffffffl >> (64 - align_order);

	/* This allocator was derived from x86_64's bit string search */

	/* Sanity check */
	if (unlikely(npages == 0)) {
		if (printk_ratelimit())
			WARN_ON(1);
		return DMA_ERROR_CODE;
	}

	if (handle && *handle)
		start = *handle;
	else
		start = largealloc ? tbl->it_largehint : tbl->it_hint;

	/* Use only half of the table for small allocs (15 pages or less) */
	limit = largealloc ? tbl->it_size : tbl->it_halfpoint;

	if (largealloc && start < tbl->it_halfpoint)
		start = tbl->it_halfpoint;

	/* The case below can happen if we have a small segment appended
	 * to a large, or when the previous alloc was at the very end of
	 * the available space. If so, go back to the initial start.
	 */
	if (start >= limit)
		start = largealloc ? tbl->it_largehint : tbl->it_hint;

 again:

	if (limit + tbl->it_offset > mask) {
		limit = mask - tbl->it_offset + 1;
		/* If we're constrained on address range, first try
		 * at the masked hint to avoid O(n) search complexity,
		 * but on second pass, start at 0.
		 */
		if ((start & mask) >= limit || pass > 0)
			start = 0;
		else
			start &= mask;
	}

	n = find_next_zero_bit(tbl->it_map, limit, start);

	/* Align allocation */
	n = (n + align_mask) & ~align_mask;

	end = n + npages;

	if (unlikely(end >= limit)) {
		if (likely(pass < 2)) {
			/* First failure, just rescan the half of the table.
			 * Second failure, rescan the other half of the table.
			 */
			start = (largealloc ^ pass) ? tbl->it_halfpoint : 0;
			limit = pass ? tbl->it_size : limit;
			pass++;
			goto again;
		} else {
			/* Third failure, give up */
			return DMA_ERROR_CODE;
		}
	}

	for (i = n; i < end; i++)
		if (test_bit(i, tbl->it_map)) {
			start = i+1;
			goto again;
		}

	for (i = n; i < end; i++)
		__set_bit(i, tbl->it_map);

	/* Bump the hint to a new block for small allocs. */
	if (largealloc) {
		/* Don't bump to new block to avoid fragmentation */
		tbl->it_largehint = end;
	} else {
		/* Overflow will be taken care of at the next allocation */
		tbl->it_hint = (end + tbl->it_blocksize - 1) &
		                ~(tbl->it_blocksize - 1);
	}

	/* Update handle for SG allocations */
	if (handle)
		*handle = end;

	return n;
}

static dma_addr_t iommu_alloc(struct iommu_table *tbl, void *page,
		       unsigned int npages, enum dma_data_direction direction,
		       unsigned long mask, unsigned int align_order)
{
	unsigned long entry, flags;
	dma_addr_t ret = DMA_ERROR_CODE;

	spin_lock_irqsave(&(tbl->it_lock), flags);

	entry = iommu_range_alloc(tbl, npages, NULL, mask, align_order);

	if (unlikely(entry == DMA_ERROR_CODE)) {
		spin_unlock_irqrestore(&(tbl->it_lock), flags);
		return DMA_ERROR_CODE;
	}

	entry += tbl->it_offset;	/* Offset into real TCE table */
	ret = entry << IOMMU_PAGE_SHIFT;	/* Set the return dma address */

	/* Put the TCEs in the HW table */
	ppc_md.tce_build(tbl, entry, npages, (unsigned long)page & IOMMU_PAGE_MASK,
			 direction);


	/* Flush/invalidate TLB caches if necessary */
	if (ppc_md.tce_flush)
		ppc_md.tce_flush(tbl);

	spin_unlock_irqrestore(&(tbl->it_lock), flags);

	/* Make sure updates are seen by hardware */
	mb();

	return ret;
}

static void __iommu_free(struct iommu_table *tbl, dma_addr_t dma_addr, 
			 unsigned int npages)
{
	unsigned long entry, free_entry;
	unsigned long i;

	entry = dma_addr >> IOMMU_PAGE_SHIFT;
	free_entry = entry - tbl->it_offset;

	if (((free_entry + npages) > tbl->it_size) ||
	    (entry < tbl->it_offset)) {
		if (printk_ratelimit()) {
			printk(KERN_INFO "iommu_free: invalid entry\n");
			printk(KERN_INFO "\tentry     = 0x%lx\n", entry); 
			printk(KERN_INFO "\tdma_addr  = 0x%lx\n", (u64)dma_addr);
			printk(KERN_INFO "\tTable     = 0x%lx\n", (u64)tbl);
			printk(KERN_INFO "\tbus#      = 0x%lx\n", (u64)tbl->it_busno);
			printk(KERN_INFO "\tsize      = 0x%lx\n", (u64)tbl->it_size);
			printk(KERN_INFO "\tstartOff  = 0x%lx\n", (u64)tbl->it_offset);
			printk(KERN_INFO "\tindex     = 0x%lx\n", (u64)tbl->it_index);
			WARN_ON(1);
		}
		return;
	}

	ppc_md.tce_free(tbl, entry, npages);
	
	for (i = 0; i < npages; i++)
		__clear_bit(free_entry+i, tbl->it_map);
}

static void iommu_free(struct iommu_table *tbl, dma_addr_t dma_addr,
		unsigned int npages)
{
	unsigned long flags;

	spin_lock_irqsave(&(tbl->it_lock), flags);

	__iommu_free(tbl, dma_addr, npages);

	/* Make sure TLB cache is flushed if the HW needs it. We do
	 * not do an mb() here on purpose, it is not needed on any of
	 * the current platforms.
	 */
	if (ppc_md.tce_flush)
		ppc_md.tce_flush(tbl);

	spin_unlock_irqrestore(&(tbl->it_lock), flags);
}

int iommu_map_sg(struct iommu_table *tbl, struct scatterlist *sglist,
		 int nelems, unsigned long mask,
		 enum dma_data_direction direction)
{
	dma_addr_t dma_next = 0, dma_addr;
	unsigned long flags;
	struct scatterlist *s, *outs, *segstart;
	int outcount, incount, i;
	unsigned int align;
	unsigned long handle;

	BUG_ON(direction == DMA_NONE);

	if ((nelems == 0) || !tbl)
		return 0;

	outs = s = segstart = &sglist[0];
	outcount = 1;
	incount = nelems;
	handle = 0;

	/* Init first segment length for backout at failure */
	outs->dma_length = 0;

	DBG("sg mapping %d elements:\n", nelems);

	spin_lock_irqsave(&(tbl->it_lock), flags);

	for_each_sg(sglist, s, nelems, i) {
		unsigned long vaddr, npages, entry, slen;

		slen = s->length;
		/* Sanity check */
		if (slen == 0) {
			dma_next = 0;
			continue;
		}
		/* Allocate iommu entries for that segment */
		vaddr = (unsigned long) sg_virt(s);
		npages = iommu_num_pages(vaddr, slen);
		align = 0;
		if (IOMMU_PAGE_SHIFT < PAGE_SHIFT && slen >= PAGE_SIZE &&
		    (vaddr & ~PAGE_MASK) == 0)
			align = PAGE_SHIFT - IOMMU_PAGE_SHIFT;
		entry = iommu_range_alloc(tbl, npages, &handle,
					  mask >> IOMMU_PAGE_SHIFT, align);

		DBG("  - vaddr: %lx, size: %lx\n", vaddr, slen);

		/* Handle failure */
		if (unlikely(entry == DMA_ERROR_CODE)) {
			if (printk_ratelimit())
				printk(KERN_INFO "iommu_alloc failed, tbl %p vaddr %lx"
				       " npages %lx\n", tbl, vaddr, npages);
			goto failure;
		}

		/* Convert entry to a dma_addr_t */
		entry += tbl->it_offset;
		dma_addr = entry << IOMMU_PAGE_SHIFT;
		dma_addr |= (s->offset & ~IOMMU_PAGE_MASK);

		DBG("  - %lu pages, entry: %lx, dma_addr: %lx\n",
			    npages, entry, dma_addr);

		/* Insert into HW table */
		ppc_md.tce_build(tbl, entry, npages, vaddr & IOMMU_PAGE_MASK, direction);

		/* If we are in an open segment, try merging */
		if (segstart != s) {