[deliverable/linux.git] / drivers / infiniband / hw / mlx5 / mem.c

/*
 * Copyright (c) 2013-2015, 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.
 */

#include <linux/module.h>
#include <rdma/ib_umem.h>
#include <rdma/ib_umem_odp.h>
#include "mlx5_ib.h"

/* @umem: umem object to scan
 * @addr: ib virtual address requested by the user
 * @count: number of PAGE_SIZE pages covered by umem
 * @shift: page shift for the compound pages found in the region
 * @ncont: number of compund pages
 * @order: log2 of the number of compound pages
 */
void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift,
			int *ncont, int *order)
{
	unsigned long tmp;
	unsigned long m;
	int i, k;
	u64 base = 0;
	int p = 0;
	int skip;
	int mask;
	u64 len;
	u64 pfn;
	struct scatterlist *sg;
	int entry;
	unsigned long page_shift = ilog2(umem->page_size);

	/* With ODP we must always match OS page size. */
	if (umem->odp_data) {
		*count = ib_umem_page_count(umem);
		*shift = PAGE_SHIFT;
		*ncont = *count;
		if (order)
			*order = ilog2(roundup_pow_of_two(*count));

		return;
	}

	addr = addr >> page_shift;
	tmp = (unsigned long)addr;
	m = find_first_bit(&tmp, sizeof(tmp));
	skip = 1 << m;
	mask = skip - 1;
	i = 0;
	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
		len = sg_dma_len(sg) >> page_shift;
		pfn = sg_dma_address(sg) >> page_shift;
		for (k = 0; k < len; k++) {
			if (!(i & mask)) {
				tmp = (unsigned long)pfn;
				m = min_t(unsigned long, m, find_first_bit(&tmp, sizeof(tmp)));
				skip = 1 << m;
				mask = skip - 1;
				base = pfn;
				p = 0;
			} else {
				if (base + p != pfn) {
					tmp = (unsigned long)p;
					m = find_first_bit(&tmp, sizeof(tmp));
					skip = 1 << m;
					mask = skip - 1;
					base = pfn;
					p = 0;
				}
			}
			p++;
			i++;
		}
	}

	if (i) {
		m = min_t(unsigned long, ilog2(roundup_pow_of_two(i)), m);

		if (order)
			*order = ilog2(roundup_pow_of_two(i) >> m);

		*ncont = DIV_ROUND_UP(i, (1 << m));
	} else {
		m  = 0;

		if (order)
			*order = 0;

		*ncont = 0;
	}
	*shift = page_shift + m;
	*count = i;
}

#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
static u64 umem_dma_to_mtt(dma_addr_t umem_dma)
{
	u64 mtt_entry = umem_dma & ODP_DMA_ADDR_MASK;

	if (umem_dma & ODP_READ_ALLOWED_BIT)
		mtt_entry |= MLX5_IB_MTT_READ;
	if (umem_dma & ODP_WRITE_ALLOWED_BIT)
		mtt_entry |= MLX5_IB_MTT_WRITE;

	return mtt_entry;
}
#endif

/*
 * Populate the given array with bus addresses from the umem.
 *
 * dev - mlx5_ib device
 * umem - umem to use to fill the pages
 * page_shift - determines the page size used in the resulting array
 * offset - offset into the umem to start from,
 *          only implemented for ODP umems
 * num_pages - total number of pages to fill
 * pas - bus addresses array to fill
 * access_flags - access flags to set on all present pages.
		  use enum mlx5_ib_mtt_access_flags for this.
 */
void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
			    int page_shift, size_t offset, size_t num_pages,
			    __be64 *pas, int access_flags)
{
	unsigned long umem_page_shift = ilog2(umem->page_size);
	int shift = page_shift - umem_page_shift;
	int mask = (1 << shift) - 1;
	int i, k;
	u64 cur = 0;
	u64 base;
	int len;
	struct scatterlist *sg;
	int entry;
#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
	const bool odp = umem->odp_data != NULL;

	if (odp) {
		WARN_ON(shift != 0);
		WARN_ON(access_flags != (MLX5_IB_MTT_READ | MLX5_IB_MTT_WRITE));

		for (i = 0; i < num_pages; ++i) {
			dma_addr_t pa = umem->odp_data->dma_list[offset + i];

			pas[i] = cpu_to_be64(umem_dma_to_mtt(pa));
		}
		return;
	}
#endif

	i = 0;
	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
		len = sg_dma_len(sg) >> umem_page_shift;
		base = sg_dma_address(sg);
		for (k = 0; k < len; k++) {
			if (!(i & mask)) {
				cur = base + (k << umem_page_shift);
				cur |= access_flags;

				pas[i >> shift] = cpu_to_be64(cur);
				mlx5_ib_dbg(dev, "pas[%d] 0x%llx\n",
					    i >> shift, be64_to_cpu(pas[i >> shift]));
			}  else
				mlx5_ib_dbg(dev, "=====> 0x%llx\n",
					    base + (k << umem_page_shift));
			i++;
		}
	}
}

void mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
			  int page_shift, __be64 *pas, int access_flags)
{
	return __mlx5_ib_populate_pas(dev, umem, page_shift, 0,
				      ib_umem_num_pages(umem), pas,
				      access_flags);
}
int mlx5_ib_get_buf_offset(u64 addr, int page_shift, u32 *offset)
{
	u64 page_size;
	u64 page_mask;
	u64 off_size;
	u64 off_mask;
	u64 buf_off;

	page_size = (u64)1 << page_shift;
	page_mask = page_size - 1;
	buf_off = addr & page_mask;
	off_size = page_size >> 6;
	off_mask = off_size - 1;

	if (buf_off & off_mask)
		return -EINVAL;

	*offset = buf_off >> ilog2(off_size);
	return 0;
}
Commit	Line	Data
e126ba97	1	/*
6cf0a15f	2	* Copyright (c) 2013-2015, Mellanox Technologies. All rights reserved.
e126ba97 EC	3	*
	4	* This software is available to you under a choice of one of two
	5	* licenses. You may choose to be licensed under the terms of the GNU
	6	* General Public License (GPL) Version 2, available from the file
	7	* COPYING in the main directory of this source tree, or the
	8	* OpenIB.org BSD license below:
	9	*
	10	* Redistribution and use in source and binary forms, with or
	11	* without modification, are permitted provided that the following
	12	* conditions are met:
	13	*
	14	* - Redistributions of source code must retain the above
	15	* copyright notice, this list of conditions and the following
	16	* disclaimer.
	17	*
	18	* - Redistributions in binary form must reproduce the above
	19	* copyright notice, this list of conditions and the following
	20	* disclaimer in the documentation and/or other materials
	21	* provided with the distribution.
	22	*
	23	* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
	24	* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
	25	* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
	26	* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
	27	* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
	28	* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
	29	* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
	30	* SOFTWARE.
	31	*/
	32
	33	#include <linux/module.h>
	34	#include <rdma/ib_umem.h>
cc149f75	35	#include <rdma/ib_umem_odp.h>
e126ba97 EC	36	#include "mlx5_ib.h"
	37
	38	/* @umem: umem object to scan
	39	* @addr: ib virtual address requested by the user
	40	* @count: number of PAGE_SIZE pages covered by umem
	41	* @shift: page shift for the compound pages found in the region
	42	* @ncont: number of compund pages
	43	* @order: log2 of the number of compound pages
	44	*/
	45	void mlx5_ib_cont_pages(struct ib_umem umem, u64 addr, int count, int *shift,
	46	int ncont, int order)
	47	{
e126ba97 EC	48	unsigned long tmp;
e126ba97 EC	49	unsigned long m;
eeb8461e	50	int i, k;
e126ba97 EC	51	u64 base = 0;
	52	int p = 0;
	53	int skip;
	54	int mask;
	55	u64 len;
	56	u64 pfn;
eeb8461e YH	57	struct scatterlist *sg;
eeb8461e YH	58	int entry;
f39f8697	59	unsigned long page_shift = ilog2(umem->page_size);
e126ba97	60
cc149f75 HE	61	/* With ODP we must always match OS page size. */
	62	if (umem->odp_data) {
	63	*count = ib_umem_page_count(umem);
	64	*shift = PAGE_SHIFT;
	65	ncont = count;
	66	if (order)
	67	order = ilog2(roundup_pow_of_two(count));
	68
	69	return;
	70	}
	71
f39f8697	72	addr = addr >> page_shift;
e126ba97 EC	73	tmp = (unsigned long)addr;
	74	m = find_first_bit(&tmp, sizeof(tmp));
	75	skip = 1 << m;
	76	mask = skip - 1;
	77	i = 0;
eeb8461e	78	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
f39f8697 YH	79	len = sg_dma_len(sg) >> page_shift;
f39f8697 YH	80	pfn = sg_dma_address(sg) >> page_shift;
eeb8461e YH	81	for (k = 0; k < len; k++) {
	82	if (!(i & mask)) {
	83	tmp = (unsigned long)pfn;
7835bfb5	84	m = min_t(unsigned long, m, find_first_bit(&tmp, sizeof(tmp)));
eeb8461e YH	85	skip = 1 << m;
	86	mask = skip - 1;
	87	base = pfn;
	88	p = 0;
	89	} else {
	90	if (base + p != pfn) {
	91	tmp = (unsigned long)p;
	92	m = find_first_bit(&tmp, sizeof(tmp));
e126ba97 EC	93	skip = 1 << m;
	94	mask = skip - 1;
	95	base = pfn;
	96	p = 0;
e126ba97	97	}
e126ba97	98	}
eeb8461e YH	99	p++;
eeb8461e YH	100	i++;
e126ba97	101	}
eeb8461e	102	}
e126ba97 EC	103
	104	if (i) {
	105	m = min_t(unsigned long, ilog2(roundup_pow_of_two(i)), m);
	106
	107	if (order)
	108	*order = ilog2(roundup_pow_of_two(i) >> m);
	109
	110	*ncont = DIV_ROUND_UP(i, (1 << m));
	111	} else {
	112	m = 0;
	113
	114	if (order)
	115	*order = 0;
	116
	117	*ncont = 0;
	118	}
f39f8697	119	*shift = page_shift + m;
e126ba97 EC	120	*count = i;
	121	}
	122
cc149f75 HE	123	#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
	124	static u64 umem_dma_to_mtt(dma_addr_t umem_dma)
	125	{
	126	u64 mtt_entry = umem_dma & ODP_DMA_ADDR_MASK;
	127
	128	if (umem_dma & ODP_READ_ALLOWED_BIT)
	129	mtt_entry \|= MLX5_IB_MTT_READ;
	130	if (umem_dma & ODP_WRITE_ALLOWED_BIT)
	131	mtt_entry \|= MLX5_IB_MTT_WRITE;
	132
	133	return mtt_entry;
	134	}
	135	#endif
	136
	137	/*
	138	* Populate the given array with bus addresses from the umem.
	139	*
	140	* dev - mlx5_ib device
	141	* umem - umem to use to fill the pages
	142	* page_shift - determines the page size used in the resulting array
832a6b06 HE	143	* offset - offset into the umem to start from,
	144	* only implemented for ODP umems
	145	* num_pages - total number of pages to fill
cc149f75 HE	146	* pas - bus addresses array to fill
	147	* access_flags - access flags to set on all present pages.
	148	use enum mlx5_ib_mtt_access_flags for this.
	149	*/
832a6b06 HE	150	void __mlx5_ib_populate_pas(struct mlx5_ib_dev dev, struct ib_umem umem,
	151	int page_shift, size_t offset, size_t num_pages,
	152	__be64 *pas, int access_flags)
e126ba97	153	{
f39f8697 YH	154	unsigned long umem_page_shift = ilog2(umem->page_size);
f39f8697 YH	155	int shift = page_shift - umem_page_shift;
e126ba97	156	int mask = (1 << shift) - 1;
eeb8461e	157	int i, k;
e126ba97 EC	158	u64 cur = 0;
	159	u64 base;
	160	int len;
eeb8461e YH	161	struct scatterlist *sg;
eeb8461e YH	162	int entry;
cc149f75 HE	163	#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
	164	const bool odp = umem->odp_data != NULL;
	165
	166	if (odp) {
cc149f75 HE	167	WARN_ON(shift != 0);
	168	WARN_ON(access_flags != (MLX5_IB_MTT_READ \| MLX5_IB_MTT_WRITE));
	169
	170	for (i = 0; i < num_pages; ++i) {
832a6b06	171	dma_addr_t pa = umem->odp_data->dma_list[offset + i];
cc149f75 HE	172
	173	pas[i] = cpu_to_be64(umem_dma_to_mtt(pa));
	174	}
	175	return;
	176	}
	177	#endif
e126ba97 EC	178
e126ba97 EC	179	i = 0;
eeb8461e	180	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
f39f8697	181	len = sg_dma_len(sg) >> umem_page_shift;
eeb8461e YH	182	base = sg_dma_address(sg);
	183	for (k = 0; k < len; k++) {
	184	if (!(i & mask)) {
f39f8697	185	cur = base + (k << umem_page_shift);
cc149f75	186	cur \|= access_flags;
e126ba97	187
eeb8461e YH	188	pas[i >> shift] = cpu_to_be64(cur);
	189	mlx5_ib_dbg(dev, "pas[%d] 0x%llx\n",
	190	i >> shift, be64_to_cpu(pas[i >> shift]));
	191	} else
	192	mlx5_ib_dbg(dev, "=====> 0x%llx\n",
f39f8697	193	base + (k << umem_page_shift));
eeb8461e	194	i++;
e126ba97	195	}
eeb8461e	196	}
e126ba97 EC	197	}
e126ba97 EC	198
832a6b06 HE	199	void mlx5_ib_populate_pas(struct mlx5_ib_dev dev, struct ib_umem umem,
	200	int page_shift, __be64 *pas, int access_flags)
	201	{
	202	return __mlx5_ib_populate_pas(dev, umem, page_shift, 0,
	203	ib_umem_num_pages(umem), pas,
	204	access_flags);
	205	}
e126ba97 EC	206	int mlx5_ib_get_buf_offset(u64 addr, int page_shift, u32 *offset)
	207	{
	208	u64 page_size;
	209	u64 page_mask;
	210	u64 off_size;
	211	u64 off_mask;
	212	u64 buf_off;
	213
f241e749	214	page_size = (u64)1 << page_shift;
e126ba97 EC	215	page_mask = page_size - 1;
	216	buf_off = addr & page_mask;
	217	off_size = page_size >> 6;
	218	off_mask = off_size - 1;
	219
	220	if (buf_off & off_mask)
	221	return -EINVAL;
	222
	223	*offset = buf_off >> ilog2(off_size);
	224	return 0;
	225	}