mthca_memfree.c

linux内核源码

/*
 * Copyright (c) 2004, 2005 Topspin Communications.  All rights reserved.
 * Copyright (c) 2005 Cisco Systems.  All rights reserved.
 * Copyright (c) 2005 Mellanox Technologies. All rights reserved.
 *
 * This software is available to you under a choice of one of two
 * licenses.  You may choose to be licensed under the terms of the GNU
 * General Public License (GPL) Version 2, available from the file
 * COPYING in the main directory of this source tree, or the
 * OpenIB.org BSD license below:
 *
 *     Redistribution and use in source and binary forms, with or
 *     without modification, are permitted provided that the following
 *     conditions are met:
 *
 *      - Redistributions of source code must retain the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer.
 *
 *      - Redistributions in binary form must reproduce the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer in the documentation and/or other materials
 *        provided with the distribution.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 *
 * $Id$
 */

#include <linux/mm.h>
#include <linux/scatterlist.h>
#include <linux/sched.h>

#include <asm/page.h>

#include "mthca_memfree.h"
#include "mthca_dev.h"
#include "mthca_cmd.h"

/*
 * We allocate in as big chunks as we can, up to a maximum of 256 KB
 * per chunk.
 */
enum {
	MTHCA_ICM_ALLOC_SIZE   = 1 << 18,
	MTHCA_TABLE_CHUNK_SIZE = 1 << 18
};

struct mthca_user_db_table {
	struct mutex mutex;
	struct {
		u64                uvirt;
		struct scatterlist mem;
		int                refcount;
	}                page[0];
};

static void mthca_free_icm_pages(struct mthca_dev *dev, struct mthca_icm_chunk *chunk)
{
	int i;

	if (chunk->nsg > 0)
		pci_unmap_sg(dev->pdev, chunk->mem, chunk->npages,
			     PCI_DMA_BIDIRECTIONAL);

	for (i = 0; i < chunk->npages; ++i)
		__free_pages(sg_page(&chunk->mem[i]),
			     get_order(chunk->mem[i].length));
}

static void mthca_free_icm_coherent(struct mthca_dev *dev, struct mthca_icm_chunk *chunk)
{
	int i;

	for (i = 0; i < chunk->npages; ++i) {
		dma_free_coherent(&dev->pdev->dev, chunk->mem[i].length,
				  lowmem_page_address(sg_page(&chunk->mem[i])),
				  sg_dma_address(&chunk->mem[i]));
	}
}

void mthca_free_icm(struct mthca_dev *dev, struct mthca_icm *icm, int coherent)
{
	struct mthca_icm_chunk *chunk, *tmp;

	if (!icm)
		return;

	list_for_each_entry_safe(chunk, tmp, &icm->chunk_list, list) {
		if (coherent)
			mthca_free_icm_coherent(dev, chunk);
		else
			mthca_free_icm_pages(dev, chunk);

		kfree(chunk);
	}

	kfree(icm);
}

static int mthca_alloc_icm_pages(struct scatterlist *mem, int order, gfp_t gfp_mask)
{
	struct page *page;

	page = alloc_pages(gfp_mask, order);
	if (!page)
		return -ENOMEM;

	sg_set_page(mem, page, PAGE_SIZE << order, 0);
	return 0;
}

static int mthca_alloc_icm_coherent(struct device *dev, struct scatterlist *mem,
				    int order, gfp_t gfp_mask)
{
	void *buf = dma_alloc_coherent(dev, PAGE_SIZE << order, &sg_dma_address(mem),
				       gfp_mask);
	if (!buf)
		return -ENOMEM;

	sg_set_buf(mem, buf, PAGE_SIZE << order);
	BUG_ON(mem->offset);
	sg_dma_len(mem) = PAGE_SIZE << order;
	return 0;
}

struct mthca_icm *mthca_alloc_icm(struct mthca_dev *dev, int npages,
				  gfp_t gfp_mask, int coherent)
{
	struct mthca_icm *icm;
	struct mthca_icm_chunk *chunk = NULL;
	int cur_order;
	int ret;

	/* We use sg_set_buf for coherent allocs, which assumes low memory */
	BUG_ON(coherent && (gfp_mask & __GFP_HIGHMEM));

	icm = kmalloc(sizeof *icm, gfp_mask & ~(__GFP_HIGHMEM | __GFP_NOWARN));
	if (!icm)
		return icm;

	icm->refcount = 0;
	INIT_LIST_HEAD(&icm->chunk_list);

	cur_order = get_order(MTHCA_ICM_ALLOC_SIZE);

	while (npages > 0) {
		if (!chunk) {
			chunk = kmalloc(sizeof *chunk,
					gfp_mask & ~(__GFP_HIGHMEM | __GFP_NOWARN));
			if (!chunk)
				goto fail;

			sg_init_table(chunk->mem, MTHCA_ICM_CHUNK_LEN);
			chunk->npages = 0;
			chunk->nsg    = 0;
			list_add_tail(&chunk->list, &icm->chunk_list);
		}

		while (1 << cur_order > npages)
			--cur_order;

		if (coherent)
			ret = mthca_alloc_icm_coherent(&dev->pdev->dev,
						       &chunk->mem[chunk->npages],
						       cur_order, gfp_mask);
		else
			ret = mthca_alloc_icm_pages(&chunk->mem[chunk->npages],
						    cur_order, gfp_mask);

		if (!ret) {
			++chunk->npages;

			if (coherent)
				++chunk->nsg;
			else if (chunk->npages == MTHCA_ICM_CHUNK_LEN) {
				chunk->nsg = pci_map_sg(dev->pdev, chunk->mem,
							chunk->npages,
							PCI_DMA_BIDIRECTIONAL);

				if (chunk->nsg <= 0)
					goto fail;
			}

			if (chunk->npages == MTHCA_ICM_CHUNK_LEN)
				chunk = NULL;

			npages -= 1 << cur_order;
		} else {
			--cur_order;
			if (cur_order < 0)
				goto fail;
		}
	}

	if (!coherent && chunk) {
		chunk->nsg = pci_map_sg(dev->pdev, chunk->mem,
					chunk->npages,
					PCI_DMA_BIDIRECTIONAL);

		if (chunk->nsg <= 0)
			goto fail;
	}

	return icm;

fail:
	mthca_free_icm(dev, icm, coherent);
	return NULL;
}

int mthca_table_get(struct mthca_dev *dev, struct mthca_icm_table *table, int obj)
{
	int i = (obj & (table->num_obj - 1)) * table->obj_size / MTHCA_TABLE_CHUNK_SIZE;
	int ret = 0;
	u8 status;

	mutex_lock(&table->mutex);

	if (table->icm[i]) {
		++table->icm[i]->refcount;
		goto out;
	}

	table->icm[i] = mthca_alloc_icm(dev, MTHCA_TABLE_CHUNK_SIZE >> PAGE_SHIFT,
					(table->lowmem ? GFP_KERNEL : GFP_HIGHUSER) |
					__GFP_NOWARN, table->coherent);
	if (!table->icm[i]) {
		ret = -ENOMEM;
		goto out;
	}

	if (mthca_MAP_ICM(dev, table->icm[i], table->virt + i * MTHCA_TABLE_CHUNK_SIZE,
			  &status) || status) {
		mthca_free_icm(dev, table->icm[i], table->coherent);
		table->icm[i] = NULL;
		ret = -ENOMEM;
		goto out;
	}

	++table->icm[i]->refcount;

out:
	mutex_unlock(&table->mutex);
	return ret;
}

void mthca_table_put(struct mthca_dev *dev, struct mthca_icm_table *table, int obj)
{
	int i;
	u8 status;

	if (!mthca_is_memfree(dev))
		return;

	i = (obj & (table->num_obj - 1)) * table->obj_size / MTHCA_TABLE_CHUNK_SIZE;

	mutex_lock(&table->mutex);

	if (--table->icm[i]->refcount == 0) {
		mthca_UNMAP_ICM(dev, table->virt + i * MTHCA_TABLE_CHUNK_SIZE,
				MTHCA_TABLE_CHUNK_SIZE / MTHCA_ICM_PAGE_SIZE,
				&status);
		mthca_free_icm(dev, table->icm[i], table->coherent);
		table->icm[i] = NULL;
	}

	mutex_unlock(&table->mutex);
}

void *mthca_table_find(struct mthca_icm_table *table, int obj, dma_addr_t *dma_handle)
{
	int idx, offset, dma_offset, i;
	struct mthca_icm_chunk *chunk;
	struct mthca_icm *icm;
	struct page *page = NULL;

	if (!table->lowmem)
		return NULL;

	mutex_lock(&table->mutex);

	idx = (obj & (table->num_obj - 1)) * table->obj_size;
	icm = table->icm[idx / MTHCA_TABLE_CHUNK_SIZE];
	dma_offset = offset = idx % MTHCA_TABLE_CHUNK_SIZE;

	if (!icm)
		goto out;

	list_for_each_entry(chunk, &icm->chunk_list, list) {
		for (i = 0; i < chunk->npages; ++i) {
			if (dma_handle && dma_offset >= 0) {
				if (sg_dma_len(&chunk->mem[i]) > dma_offset)
					*dma_handle = sg_dma_address(&chunk->mem[i]) +
						dma_offset;
				dma_offset -= sg_dma_len(&chunk->mem[i]);
			}
			/* DMA mapping can merge pages but not split them,
			 * so if we found the page, dma_handle has already
			 * been assigned to. */
			if (chunk->mem[i].length > offset) {
				page = sg_page(&chunk->mem[i]);
				goto out;
			}
			offset -= chunk->mem[i].length;
		}
	}

out:
	mutex_unlock(&table->mutex);
	return page ? lowmem_page_address(page) + offset : NULL;
}

int mthca_table_get_range(struct mthca_dev *dev, struct mthca_icm_table *table,
			  int start, int end)
{
	int inc = MTHCA_TABLE_CHUNK_SIZE / table->obj_size;
	int i, err;

	for (i = start; i <= end; i += inc) {
		err = mthca_table_get(dev, table, i);
		if (err)
			goto fail;
	}

	return 0;

fail:
	while (i > start) {
		i -= inc;
		mthca_table_put(dev, table, i);
	}

	return err;
}

void mthca_table_put_range(struct mthca_dev *dev, struct mthca_icm_table *table,
			   int start, int end)
{
	int i;

	if (!mthca_is_memfree(dev))
		return;

	for (i = start; i <= end; i += MTHCA_TABLE_CHUNK_SIZE / table->obj_size)
		mthca_table_put(dev, table, i);
}

struct mthca_icm_table *mthca_alloc_icm_table(struct mthca_dev *dev,
					      u64 virt, int obj_size,
					      int nobj, int reserved,
					      int use_lowmem, int use_coherent)
{
	struct mthca_icm_table *table;
	int num_icm;
	unsigned chunk_size;
	int i;
	u8 status;

	num_icm = (obj_size * nobj + MTHCA_TABLE_CHUNK_SIZE - 1) / MTHCA_TABLE_CHUNK_SIZE;

	table = kmalloc(sizeof *table + num_icm * sizeof *table->icm, GFP_KERNEL);
	if (!table)
		return NULL;

	table->virt     = virt;
	table->num_icm  = num_icm;
	table->num_obj  = nobj;
	table->obj_size = obj_size;
	table->lowmem   = use_lowmem;
	table->coherent = use_coherent;
	mutex_init(&table->mutex);

	for (i = 0; i < num_icm; ++i)
		table->icm[i] = NULL;

	for (i = 0; i * MTHCA_TABLE_CHUNK_SIZE < reserved * obj_size; ++i) {
		chunk_size = MTHCA_TABLE_CHUNK_SIZE;