[linux.git] / fs / btrfs / raid-stripe-tree.c

// SPDX-License-Identifier: GPL-2.0
/*
 * Copyright (C) 2023 Western Digital Corporation or its affiliates.
 */

#include <linux/btrfs_tree.h>
#include "ctree.h"
#include "fs.h"
#include "accessors.h"
#include "transaction.h"
#include "disk-io.h"
#include "raid-stripe-tree.h"
#include "volumes.h"
#include "print-tree.h"

int btrfs_delete_raid_extent(struct btrfs_trans_handle *trans, u64 start, u64 length)
{
	struct btrfs_fs_info *fs_info = trans->fs_info;
	struct btrfs_root *stripe_root = fs_info->stripe_root;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct extent_buffer *leaf;
	u64 found_start;
	u64 found_end;
	u64 end = start + length;
	int slot;
	int ret;

	if (!stripe_root)
		return 0;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	while (1) {
		key.objectid = start;
		key.type = BTRFS_RAID_STRIPE_KEY;
		key.offset = length;

		ret = btrfs_search_slot(trans, stripe_root, &key, path, -1, 1);
		if (ret < 0)
			break;
		if (ret > 0) {
			ret = 0;
			if (path->slots[0] == 0)
				break;
			path->slots[0]--;
		}

		leaf = path->nodes[0];
		slot = path->slots[0];
		btrfs_item_key_to_cpu(leaf, &key, slot);
		found_start = key.objectid;
		found_end = found_start + key.offset;

		/* That stripe ends before we start, we're done. */
		if (found_end <= start)
			break;

		trace_btrfs_raid_extent_delete(fs_info, start, end,
					       found_start, found_end);

		ASSERT(found_start >= start && found_end <= end);
		ret = btrfs_del_item(trans, stripe_root, path);
		if (ret)
			break;

		btrfs_release_path(path);
	}

	btrfs_free_path(path);
	return ret;
}

static int btrfs_insert_one_raid_extent(struct btrfs_trans_handle *trans,
					struct btrfs_io_context *bioc)
{
	struct btrfs_fs_info *fs_info = trans->fs_info;
	struct btrfs_key stripe_key;
	struct btrfs_root *stripe_root = fs_info->stripe_root;
	const int num_stripes = btrfs_bg_type_to_factor(bioc->map_type);
	u8 encoding = btrfs_bg_flags_to_raid_index(bioc->map_type);
	struct btrfs_stripe_extent *stripe_extent;
	const size_t item_size = struct_size(stripe_extent, strides, num_stripes);
	int ret;

	stripe_extent = kzalloc(item_size, GFP_NOFS);
	if (!stripe_extent) {
		btrfs_abort_transaction(trans, -ENOMEM);
		btrfs_end_transaction(trans);
		return -ENOMEM;
	}

	trace_btrfs_insert_one_raid_extent(fs_info, bioc->logical, bioc->size,
					   num_stripes);
	btrfs_set_stack_stripe_extent_encoding(stripe_extent, encoding);
	for (int i = 0; i < num_stripes; i++) {
		u64 devid = bioc->stripes[i].dev->devid;
		u64 physical = bioc->stripes[i].physical;
		u64 length = bioc->stripes[i].length;
		struct btrfs_raid_stride *raid_stride = &stripe_extent->strides[i];

		if (length == 0)
			length = bioc->size;

		btrfs_set_stack_raid_stride_devid(raid_stride, devid);
		btrfs_set_stack_raid_stride_physical(raid_stride, physical);
	}

	stripe_key.objectid = bioc->logical;
	stripe_key.type = BTRFS_RAID_STRIPE_KEY;
	stripe_key.offset = bioc->size;

	ret = btrfs_insert_item(trans, stripe_root, &stripe_key, stripe_extent,
				item_size);
	if (ret)
		btrfs_abort_transaction(trans, ret);

	kfree(stripe_extent);

	return ret;
}

int btrfs_insert_raid_extent(struct btrfs_trans_handle *trans,
			     struct btrfs_ordered_extent *ordered_extent)
{
	struct btrfs_io_context *bioc;
	int ret;

	if (!btrfs_fs_incompat(trans->fs_info, RAID_STRIPE_TREE))
		return 0;

	list_for_each_entry(bioc, &ordered_extent->bioc_list, rst_ordered_entry) {
		ret = btrfs_insert_one_raid_extent(trans, bioc);
		if (ret)
			return ret;
	}

	while (!list_empty(&ordered_extent->bioc_list)) {
		bioc = list_first_entry(&ordered_extent->bioc_list,
					typeof(*bioc), rst_ordered_entry);
		list_del(&bioc->rst_ordered_entry);
		btrfs_put_bioc(bioc);
	}

	return 0;
}

int btrfs_get_raid_extent_offset(struct btrfs_fs_info *fs_info,
				 u64 logical, u64 *length, u64 map_type,
				 u32 stripe_index, struct btrfs_io_stripe *stripe)
{
	struct btrfs_root *stripe_root = fs_info->stripe_root;
	struct btrfs_stripe_extent *stripe_extent;
	struct btrfs_key stripe_key;
	struct btrfs_key found_key;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	const u64 end = logical + *length;
	int num_stripes;
	u8 encoding;
	u64 offset;
	u64 found_logical;
	u64 found_length;
	u64 found_end;
	int slot;
	int ret;

	stripe_key.objectid = logical;
	stripe_key.type = BTRFS_RAID_STRIPE_KEY;
	stripe_key.offset = 0;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	if (stripe->is_scrub) {
		path->skip_locking = 1;
		path->search_commit_root = 1;
	}

	ret = btrfs_search_slot(NULL, stripe_root, &stripe_key, path, 0, 0);
	if (ret < 0)
		goto free_path;
	if (ret) {
		if (path->slots[0] != 0)
			path->slots[0]--;
	}

	while (1) {
		leaf = path->nodes[0];
		slot = path->slots[0];

		btrfs_item_key_to_cpu(leaf, &found_key, slot);
		found_logical = found_key.objectid;
		found_length = found_key.offset;
		found_end = found_logical + found_length;

		if (found_logical > end) {
			ret = -ENOENT;
			goto out;
		}

		if (in_range(logical, found_logical, found_length))
			break;

		ret = btrfs_next_item(stripe_root, path);
		if (ret)
			goto out;
	}

	offset = logical - found_logical;

	/*
	 * If we have a logically contiguous, but physically non-continuous
	 * range, we need to split the bio. Record the length after which we
	 * must split the bio.
	 */
	if (end > found_end)
		*length -= end - found_end;

	num_stripes = btrfs_num_raid_stripes(btrfs_item_size(leaf, slot));
	stripe_extent = btrfs_item_ptr(leaf, slot, struct btrfs_stripe_extent);
	encoding = btrfs_stripe_extent_encoding(leaf, stripe_extent);

	if (encoding != btrfs_bg_flags_to_raid_index(map_type)) {
		ret = -EUCLEAN;
		btrfs_handle_fs_error(fs_info, ret,
				      "on-disk stripe encoding %d doesn't match RAID index %d",
				      encoding,
				      btrfs_bg_flags_to_raid_index(map_type));
		goto out;
	}

	for (int i = 0; i < num_stripes; i++) {
		struct btrfs_raid_stride *stride = &stripe_extent->strides[i];
		u64 devid = btrfs_raid_stride_devid(leaf, stride);
		u64 physical = btrfs_raid_stride_physical(leaf, stride);

		if (devid != stripe->dev->devid)
			continue;

		if ((map_type & BTRFS_BLOCK_GROUP_DUP) && stripe_index != i)
			continue;

		stripe->physical = physical + offset;

		trace_btrfs_get_raid_extent_offset(fs_info, logical, *length,
						   stripe->physical, devid);

		ret = 0;
		goto free_path;
	}

	/* If we're here, we haven't found the requested devid in the stripe. */
	ret = -ENOENT;
out:
	if (ret > 0)
		ret = -ENOENT;
	if (ret && ret != -EIO && !stripe->is_scrub) {
		if (IS_ENABLED(CONFIG_BTRFS_DEBUG))
			btrfs_print_tree(leaf, 1);
		btrfs_err(fs_info,
		"cannot find raid-stripe for logical [%llu, %llu] devid %llu, profile %s",
			  logical, logical + *length, stripe->dev->devid,
			  btrfs_bg_type_to_raid_name(map_type));
	}
free_path:
	btrfs_free_path(path);

	return ret;
}
Commit	Line	Data
02c372e1 JT	1	// SPDX-License-Identifier: GPL-2.0
	2	/*
	3	* Copyright (C) 2023 Western Digital Corporation or its affiliates.
	4	*/
	5
	6	#include <linux/btrfs_tree.h>
	7	#include "ctree.h"
	8	#include "fs.h"
	9	#include "accessors.h"
	10	#include "transaction.h"
	11	#include "disk-io.h"
	12	#include "raid-stripe-tree.h"
	13	#include "volumes.h"
02c372e1 JT	14	#include "print-tree.h"
02c372e1 JT	15
ca41504e JT	16	int btrfs_delete_raid_extent(struct btrfs_trans_handle *trans, u64 start, u64 length)
	17	{
	18	struct btrfs_fs_info *fs_info = trans->fs_info;
	19	struct btrfs_root *stripe_root = fs_info->stripe_root;
	20	struct btrfs_path *path;
	21	struct btrfs_key key;
	22	struct extent_buffer *leaf;
	23	u64 found_start;
	24	u64 found_end;
	25	u64 end = start + length;
	26	int slot;
	27	int ret;
	28
	29	if (!stripe_root)
	30	return 0;
	31
	32	path = btrfs_alloc_path();
	33	if (!path)
	34	return -ENOMEM;
	35
	36	while (1) {
	37	key.objectid = start;
	38	key.type = BTRFS_RAID_STRIPE_KEY;
	39	key.offset = length;
	40
	41	ret = btrfs_search_slot(trans, stripe_root, &key, path, -1, 1);
	42	if (ret < 0)
	43	break;
	44	if (ret > 0) {
	45	ret = 0;
	46	if (path->slots[0] == 0)
	47	break;
	48	path->slots[0]--;
	49	}
	50
	51	leaf = path->nodes[0];
	52	slot = path->slots[0];
	53	btrfs_item_key_to_cpu(leaf, &key, slot);
	54	found_start = key.objectid;
	55	found_end = found_start + key.offset;
	56
	57	/* That stripe ends before we start, we're done. */
	58	if (found_end <= start)
	59	break;
	60
b5e2c2ff JT	61	trace_btrfs_raid_extent_delete(fs_info, start, end,
	62	found_start, found_end);
	63
ca41504e JT	64	ASSERT(found_start >= start && found_end <= end);
	65	ret = btrfs_del_item(trans, stripe_root, path);
	66	if (ret)
	67	break;
	68
	69	btrfs_release_path(path);
	70	}
	71
	72	btrfs_free_path(path);
	73	return ret;
	74	}
	75
02c372e1 JT	76	static int btrfs_insert_one_raid_extent(struct btrfs_trans_handle *trans,
	77	struct btrfs_io_context *bioc)
	78	{
	79	struct btrfs_fs_info *fs_info = trans->fs_info;
	80	struct btrfs_key stripe_key;
	81	struct btrfs_root *stripe_root = fs_info->stripe_root;
	82	const int num_stripes = btrfs_bg_type_to_factor(bioc->map_type);
	83	u8 encoding = btrfs_bg_flags_to_raid_index(bioc->map_type);
	84	struct btrfs_stripe_extent *stripe_extent;
	85	const size_t item_size = struct_size(stripe_extent, strides, num_stripes);
	86	int ret;
	87
	88	stripe_extent = kzalloc(item_size, GFP_NOFS);
	89	if (!stripe_extent) {
	90	btrfs_abort_transaction(trans, -ENOMEM);
	91	btrfs_end_transaction(trans);
	92	return -ENOMEM;
	93	}
	94
b5e2c2ff JT	95	trace_btrfs_insert_one_raid_extent(fs_info, bioc->logical, bioc->size,
b5e2c2ff JT	96	num_stripes);
02c372e1 JT	97	btrfs_set_stack_stripe_extent_encoding(stripe_extent, encoding);
	98	for (int i = 0; i < num_stripes; i++) {
	99	u64 devid = bioc->stripes[i].dev->devid;
	100	u64 physical = bioc->stripes[i].physical;
	101	u64 length = bioc->stripes[i].length;
	102	struct btrfs_raid_stride *raid_stride = &stripe_extent->strides[i];
	103
	104	if (length == 0)
	105	length = bioc->size;
	106
	107	btrfs_set_stack_raid_stride_devid(raid_stride, devid);
	108	btrfs_set_stack_raid_stride_physical(raid_stride, physical);
	109	}
	110
	111	stripe_key.objectid = bioc->logical;
	112	stripe_key.type = BTRFS_RAID_STRIPE_KEY;
	113	stripe_key.offset = bioc->size;
	114
	115	ret = btrfs_insert_item(trans, stripe_root, &stripe_key, stripe_extent,
	116	item_size);
	117	if (ret)
	118	btrfs_abort_transaction(trans, ret);
	119
	120	kfree(stripe_extent);
	121
	122	return ret;
	123	}
	124
	125	int btrfs_insert_raid_extent(struct btrfs_trans_handle *trans,
	126	struct btrfs_ordered_extent *ordered_extent)
	127	{
	128	struct btrfs_io_context *bioc;
	129	int ret;
	130
	131	if (!btrfs_fs_incompat(trans->fs_info, RAID_STRIPE_TREE))
	132	return 0;
	133
	134	list_for_each_entry(bioc, &ordered_extent->bioc_list, rst_ordered_entry) {
	135	ret = btrfs_insert_one_raid_extent(trans, bioc);
	136	if (ret)
	137	return ret;
	138	}
	139
	140	while (!list_empty(&ordered_extent->bioc_list)) {
	141	bioc = list_first_entry(&ordered_extent->bioc_list,
	142	typeof(*bioc), rst_ordered_entry);
	143	list_del(&bioc->rst_ordered_entry);
	144	btrfs_put_bioc(bioc);
	145	}
	146
b8212814	147	return 0;
02c372e1	148	}
10e27980 JT	149
	150	int btrfs_get_raid_extent_offset(struct btrfs_fs_info *fs_info,
	151	u64 logical, u64 *length, u64 map_type,
	152	u32 stripe_index, struct btrfs_io_stripe *stripe)
	153	{
	154	struct btrfs_root *stripe_root = fs_info->stripe_root;
	155	struct btrfs_stripe_extent *stripe_extent;
	156	struct btrfs_key stripe_key;
	157	struct btrfs_key found_key;
	158	struct btrfs_path *path;
	159	struct extent_buffer *leaf;
	160	const u64 end = logical + *length;
	161	int num_stripes;
	162	u8 encoding;
	163	u64 offset;
	164	u64 found_logical;
	165	u64 found_length;
	166	u64 found_end;
	167	int slot;
	168	int ret;
	169
	170	stripe_key.objectid = logical;
	171	stripe_key.type = BTRFS_RAID_STRIPE_KEY;
	172	stripe_key.offset = 0;
	173
	174	path = btrfs_alloc_path();
	175	if (!path)
	176	return -ENOMEM;
	177
9acaa641 JT	178	if (stripe->is_scrub) {
	179	path->skip_locking = 1;
	180	path->search_commit_root = 1;
	181	}
	182
10e27980 JT	183	ret = btrfs_search_slot(NULL, stripe_root, &stripe_key, path, 0, 0);
	184	if (ret < 0)
	185	goto free_path;
	186	if (ret) {
	187	if (path->slots[0] != 0)
	188	path->slots[0]--;
	189	}
	190
	191	while (1) {
	192	leaf = path->nodes[0];
	193	slot = path->slots[0];
	194
	195	btrfs_item_key_to_cpu(leaf, &found_key, slot);
	196	found_logical = found_key.objectid;
	197	found_length = found_key.offset;
	198	found_end = found_logical + found_length;
	199
	200	if (found_logical > end) {
	201	ret = -ENOENT;
	202	goto out;
	203	}
	204
	205	if (in_range(logical, found_logical, found_length))
	206	break;
	207
	208	ret = btrfs_next_item(stripe_root, path);
	209	if (ret)
	210	goto out;
	211	}
	212
	213	offset = logical - found_logical;
	214
	215	/*
	216	* If we have a logically contiguous, but physically non-continuous
	217	* range, we need to split the bio. Record the length after which we
	218	* must split the bio.
	219	*/
	220	if (end > found_end)
	221	*length -= end - found_end;
	222
	223	num_stripes = btrfs_num_raid_stripes(btrfs_item_size(leaf, slot));
	224	stripe_extent = btrfs_item_ptr(leaf, slot, struct btrfs_stripe_extent);
	225	encoding = btrfs_stripe_extent_encoding(leaf, stripe_extent);
	226
	227	if (encoding != btrfs_bg_flags_to_raid_index(map_type)) {
	228	ret = -EUCLEAN;
	229	btrfs_handle_fs_error(fs_info, ret,
	230	"on-disk stripe encoding %d doesn't match RAID index %d",
	231	encoding,
	232	btrfs_bg_flags_to_raid_index(map_type));
	233	goto out;
	234	}
	235
	236	for (int i = 0; i < num_stripes; i++) {
	237	struct btrfs_raid_stride *stride = &stripe_extent->strides[i];
	238	u64 devid = btrfs_raid_stride_devid(leaf, stride);
	239	u64 physical = btrfs_raid_stride_physical(leaf, stride);
	240
	241	if (devid != stripe->dev->devid)
	242	continue;
	243
	244	if ((map_type & BTRFS_BLOCK_GROUP_DUP) && stripe_index != i)
	245	continue;
	246
247	stripe->physical = physical + offset;
248
b5e2c2ff JT	249	trace_btrfs_get_raid_extent_offset(fs_info, logical, *length,
	250	stripe->physical, devid);
	251
10e27980 JT	252	ret = 0;
	253	goto free_path;
	254	}
	255
	256	/* If we're here, we haven't found the requested devid in the stripe. */
	257	ret = -ENOENT;
	258	out:
	259	if (ret > 0)
	260	ret = -ENOENT;
9acaa641	261	if (ret && ret != -EIO && !stripe->is_scrub) {
10e27980 JT	262	if (IS_ENABLED(CONFIG_BTRFS_DEBUG))
	263	btrfs_print_tree(leaf, 1);
	264	btrfs_err(fs_info,
	265	"cannot find raid-stripe for logical [%llu, %llu] devid %llu, profile %s",
	266	logical, logical + *length, stripe->dev->devid,
	267	btrfs_bg_type_to_raid_name(map_type));
	268	}
	269	free_path:
	270	btrfs_free_path(path);
	271
	272	return ret;
	273	}