[mirror_ubuntu-focal-kernel.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
 * @max_page_shift: high limit for page_shift - 0 means no limit
 * @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,
			unsigned long max_page_shift,
			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, BITS_PER_LONG);
	if (max_page_shift)
		m = min_t(unsigned long, max_page_shift - page_shift, m);
	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, BITS_PER_LONG));
				skip = 1 << m;
				mask = skip - 1;
				base = pfn;
				p = 0;
			} else {
				if (base + p != pfn) {
					tmp = (unsigned long)p;
					m = find_first_bit(&tmp, BITS_PER_LONG);
					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
762f899a	40	* @max_page_shift: high limit for page_shift - 0 means no limit
e126ba97 EC	41	* @count: number of PAGE_SIZE pages covered by umem
	42	* @shift: page shift for the compound pages found in the region
	43	* @ncont: number of compund pages
	44	* @order: log2 of the number of compound pages
	45	*/
762f899a MD	46	void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr,
	47	unsigned long max_page_shift,
	48	int count, int shift,
e126ba97 EC	49	int ncont, int order)
e126ba97 EC	50	{
e126ba97 EC	51	unsigned long tmp;
e126ba97 EC	52	unsigned long m;
eeb8461e	53	int i, k;
e126ba97 EC	54	u64 base = 0;
	55	int p = 0;
	56	int skip;
	57	int mask;
	58	u64 len;
	59	u64 pfn;
eeb8461e YH	60	struct scatterlist *sg;
eeb8461e YH	61	int entry;
f39f8697	62	unsigned long page_shift = ilog2(umem->page_size);
e126ba97	63
cc149f75 HE	64	/* With ODP we must always match OS page size. */
	65	if (umem->odp_data) {
	66	*count = ib_umem_page_count(umem);
	67	*shift = PAGE_SHIFT;
	68	ncont = count;
	69	if (order)
	70	order = ilog2(roundup_pow_of_two(count));
	71
	72	return;
	73	}
	74
f39f8697	75	addr = addr >> page_shift;
e126ba97	76	tmp = (unsigned long)addr;
fffd6873	77	m = find_first_bit(&tmp, BITS_PER_LONG);
762f899a MD	78	if (max_page_shift)
762f899a MD	79	m = min_t(unsigned long, max_page_shift - page_shift, m);
e126ba97 EC	80	skip = 1 << m;
	81	mask = skip - 1;
	82	i = 0;
eeb8461e	83	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
f39f8697 YH	84	len = sg_dma_len(sg) >> page_shift;
f39f8697 YH	85	pfn = sg_dma_address(sg) >> page_shift;
eeb8461e YH	86	for (k = 0; k < len; k++) {
	87	if (!(i & mask)) {
	88	tmp = (unsigned long)pfn;
fffd6873	89	m = min_t(unsigned long, m, find_first_bit(&tmp, BITS_PER_LONG));
eeb8461e YH	90	skip = 1 << m;
	91	mask = skip - 1;
	92	base = pfn;
	93	p = 0;
	94	} else {
	95	if (base + p != pfn) {
	96	tmp = (unsigned long)p;
fffd6873	97	m = find_first_bit(&tmp, BITS_PER_LONG);
e126ba97 EC	98	skip = 1 << m;
	99	mask = skip - 1;
	100	base = pfn;
	101	p = 0;
e126ba97	102	}
e126ba97	103	}
eeb8461e YH	104	p++;
eeb8461e YH	105	i++;
e126ba97	106	}
eeb8461e	107	}
e126ba97 EC	108
	109	if (i) {
	110	m = min_t(unsigned long, ilog2(roundup_pow_of_two(i)), m);
	111
	112	if (order)
	113	*order = ilog2(roundup_pow_of_two(i) >> m);
	114
	115	*ncont = DIV_ROUND_UP(i, (1 << m));
	116	} else {
	117	m = 0;
	118
	119	if (order)
	120	*order = 0;
	121
	122	*ncont = 0;
	123	}
f39f8697	124	*shift = page_shift + m;
e126ba97 EC	125	*count = i;
	126	}
	127
cc149f75 HE	128	#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
	129	static u64 umem_dma_to_mtt(dma_addr_t umem_dma)
	130	{
	131	u64 mtt_entry = umem_dma & ODP_DMA_ADDR_MASK;
	132
	133	if (umem_dma & ODP_READ_ALLOWED_BIT)
	134	mtt_entry \|= MLX5_IB_MTT_READ;
	135	if (umem_dma & ODP_WRITE_ALLOWED_BIT)
	136	mtt_entry \|= MLX5_IB_MTT_WRITE;
	137
	138	return mtt_entry;
	139	}
	140	#endif
	141
	142	/*
	143	* Populate the given array with bus addresses from the umem.
	144	*
	145	* dev - mlx5_ib device
	146	* umem - umem to use to fill the pages
	147	* page_shift - determines the page size used in the resulting array
832a6b06 HE	148	* offset - offset into the umem to start from,
	149	* only implemented for ODP umems
	150	* num_pages - total number of pages to fill
cc149f75 HE	151	* pas - bus addresses array to fill
	152	* access_flags - access flags to set on all present pages.
	153	use enum mlx5_ib_mtt_access_flags for this.
	154	*/
832a6b06 HE	155	void __mlx5_ib_populate_pas(struct mlx5_ib_dev dev, struct ib_umem umem,
	156	int page_shift, size_t offset, size_t num_pages,
	157	__be64 *pas, int access_flags)
e126ba97	158	{
f39f8697 YH	159	unsigned long umem_page_shift = ilog2(umem->page_size);
f39f8697 YH	160	int shift = page_shift - umem_page_shift;
e126ba97	161	int mask = (1 << shift) - 1;
eeb8461e	162	int i, k;
e126ba97 EC	163	u64 cur = 0;
	164	u64 base;
	165	int len;
eeb8461e YH	166	struct scatterlist *sg;
eeb8461e YH	167	int entry;
cc149f75 HE	168	#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
	169	const bool odp = umem->odp_data != NULL;
	170
	171	if (odp) {
cc149f75 HE	172	WARN_ON(shift != 0);
	173	WARN_ON(access_flags != (MLX5_IB_MTT_READ \| MLX5_IB_MTT_WRITE));
	174
	175	for (i = 0; i < num_pages; ++i) {
832a6b06	176	dma_addr_t pa = umem->odp_data->dma_list[offset + i];
cc149f75 HE	177
	178	pas[i] = cpu_to_be64(umem_dma_to_mtt(pa));
	179	}
	180	return;
	181	}
	182	#endif
e126ba97 EC	183
e126ba97 EC	184	i = 0;
eeb8461e	185	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
f39f8697	186	len = sg_dma_len(sg) >> umem_page_shift;
eeb8461e YH	187	base = sg_dma_address(sg);
	188	for (k = 0; k < len; k++) {
	189	if (!(i & mask)) {
f39f8697	190	cur = base + (k << umem_page_shift);
cc149f75	191	cur \|= access_flags;
e126ba97	192
eeb8461e YH	193	pas[i >> shift] = cpu_to_be64(cur);
	194	mlx5_ib_dbg(dev, "pas[%d] 0x%llx\n",
	195	i >> shift, be64_to_cpu(pas[i >> shift]));
	196	} else
	197	mlx5_ib_dbg(dev, "=====> 0x%llx\n",
f39f8697	198	base + (k << umem_page_shift));
eeb8461e	199	i++;
e126ba97	200	}
eeb8461e	201	}
e126ba97 EC	202	}
e126ba97 EC	203
832a6b06 HE	204	void mlx5_ib_populate_pas(struct mlx5_ib_dev dev, struct ib_umem umem,
	205	int page_shift, __be64 *pas, int access_flags)
	206	{
	207	return __mlx5_ib_populate_pas(dev, umem, page_shift, 0,
	208	ib_umem_num_pages(umem), pas,
	209	access_flags);
	210	}
e126ba97 EC	211	int mlx5_ib_get_buf_offset(u64 addr, int page_shift, u32 *offset)
	212	{
	213	u64 page_size;
	214	u64 page_mask;
	215	u64 off_size;
	216	u64 off_mask;
	217	u64 buf_off;
	218
f241e749	219	page_size = (u64)1 << page_shift;
e126ba97 EC	220	page_mask = page_size - 1;
	221	buf_off = addr & page_mask;
	222	off_size = page_size >> 6;
	223	off_mask = off_size - 1;
	224
	225	if (buf_off & off_mask)
	226	return -EINVAL;
	227
	228	*offset = buf_off >> ilog2(off_size);
	229	return 0;
	230	}