#include "transaction.h"
#include "print-tree.h"
#include "volumes.h"
+#include "math.h"
struct stripe {
struct btrfs_device *dev;
#define is_parity_stripe(x) ( ((x) == BTRFS_RAID5_P_STRIPE) || ((x) == BTRFS_RAID6_Q_STRIPE) )
-#define map_lookup_size(n) (sizeof(struct map_lookup) + \
- (sizeof(struct btrfs_bio_stripe) * (n)))
-
static LIST_HEAD(fs_uuids);
static struct btrfs_device *__find_device(struct list_head *head, u64 devid,
list_for_each(cur, &fs_devices->devices) {
device = list_entry(cur, struct btrfs_device, dev_list);
if (device->fd != -1) {
+ fsync(device->fd);
+ if (posix_fadvise(device->fd, 0, 0, POSIX_FADV_DONTNEED))
+ fprintf(stderr, "Warning, could not drop caches\n");
close(device->fd);
device->fd = -1;
}
list_for_each(cur, head) {
device = list_entry(cur, struct btrfs_device, dev_list);
+ if (!device->name) {
+ printk("no name for device %llu, skip it now\n", device->devid);
+ continue;
+ }
fd = open(device->name, flags);
if (fd < 0) {
return 0;
}
-static u64 div_factor(u64 num, int factor)
-{
- if (factor == 10)
- return num;
- num *= factor;
- return num / 10;
-}
-
static u64 chunk_bytes_by_type(u64 type, u64 calc_size, int num_stripes,
int sub_stripes)
{
{
u64 dev_offset;
struct btrfs_fs_info *info = extent_root->fs_info;
- struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
+ struct btrfs_root *chunk_root = info->chunk_root;
struct btrfs_stripe *stripes;
struct btrfs_device *device = NULL;
struct btrfs_chunk *chunk;
struct list_head private_devs;
- struct list_head *dev_list = &extent_root->fs_info->fs_devices->devices;
+ struct list_head *dev_list = &info->fs_devices->devices;
struct list_head *cur;
struct map_lookup *map;
int min_stripe_size = 1 * 1024 * 1024;
if (!chunk)
return -ENOMEM;
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
+ map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map) {
kfree(chunk);
return -ENOMEM;
map->ce.start = key.offset;
map->ce.size = *num_bytes;
- ret = insert_existing_cache_extent(
- &extent_root->fs_info->mapping_tree.cache_tree,
- &map->ce);
+ ret = insert_cache_extent(&info->mapping_tree.cache_tree, &map->ce);
BUG_ON(ret);
if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
{
u64 dev_offset;
struct btrfs_fs_info *info = extent_root->fs_info;
- struct btrfs_root *chunk_root = extent_root->fs_info->chunk_root;
+ struct btrfs_root *chunk_root = info->chunk_root;
struct btrfs_stripe *stripes;
struct btrfs_device *device = NULL;
struct btrfs_chunk *chunk;
- struct list_head *dev_list = &extent_root->fs_info->fs_devices->devices;
+ struct list_head *dev_list = &info->fs_devices->devices;
struct list_head *cur;
struct map_lookup *map;
u64 calc_size = 8 * 1024 * 1024;
if (!chunk)
return -ENOMEM;
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
+ map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map) {
kfree(chunk);
return -ENOMEM;
map->ce.start = key.offset;
map->ce.size = num_bytes;
- ret = insert_existing_cache_extent(
- &extent_root->fs_info->mapping_tree.cache_tree,
- &map->ce);
+ ret = insert_cache_extent(&info->mapping_tree.cache_tree, &map->ce);
BUG_ON(ret);
kfree(chunk);
struct map_lookup *map;
int ret;
- ce = find_first_cache_extent(&map_tree->cache_tree, logical);
+ ce = search_cache_extent(&map_tree->cache_tree, logical);
BUG_ON(!ce);
BUG_ON(ce->start > logical || ce->start + ce->size < logical);
map = container_of(ce, struct map_lookup, ce);
struct cache_extent *ce;
struct map_lookup *map;
- ce = find_first_cache_extent(&map_tree->cache_tree, *logical);
+ ce = search_cache_extent(&map_tree->cache_tree, *logical);
while (ce) {
ce = next_cache_extent(ce);
u64 rmap_len;
int i, j, nr = 0;
- ce = find_first_cache_extent(&map_tree->cache_tree, chunk_start);
+ ce = search_cache_extent(&map_tree->cache_tree, chunk_start);
BUG_ON(!ce);
map = container_of(ce, struct map_lookup, ce);
stripes_allocated = 1;
}
again:
- ce = find_first_cache_extent(&map_tree->cache_tree, logical);
+ ce = search_cache_extent(&map_tree->cache_tree, logical);
if (!ce) {
if (multi)
kfree(multi);
return NULL;
}
-struct btrfs_device *btrfs_find_device_by_devid(struct btrfs_root *root,
- u64 devid, int instance)
+struct btrfs_device *
+btrfs_find_device_by_devid(struct btrfs_fs_devices *fs_devices,
+ u64 devid, int instance)
{
- struct list_head *head = &root->fs_info->fs_devices->devices;
+ struct list_head *head = &fs_devices->devices;
struct btrfs_device *dev;
- struct list_head *cur;
int num_found = 0;
- list_for_each(cur, head) {
- dev = list_entry(cur, struct btrfs_device, dev_list);
+ list_for_each_entry(dev, head, dev_list) {
if (dev->devid == devid && num_found++ == instance)
return dev;
}
list_for_each(cur, &fs_devices->devices) {
num_stripes++;
}
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
+ map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map)
return -ENOMEM;
map->stripes[i].dev = device;
i++;
}
- ret = insert_existing_cache_extent(&map_tree->cache_tree, &map->ce);
+ ret = insert_cache_extent(&map_tree->cache_tree, &map->ce);
if (ret == -EEXIST) {
struct cache_extent *old;
struct map_lookup *old_map;
- old = find_cache_extent(&map_tree->cache_tree, logical, length);
+ old = lookup_cache_extent(&map_tree->cache_tree,
+ logical, length);
old_map = container_of(old, struct map_lookup, ce);
remove_cache_extent(&map_tree->cache_tree, old);
kfree(old_map);
- ret = insert_existing_cache_extent(&map_tree->cache_tree,
+ ret = insert_cache_extent(&map_tree->cache_tree,
&map->ce);
}
BUG_ON(ret);
int readonly = 0;
int i;
- ce = find_first_cache_extent(&map_tree->cache_tree, chunk_offset);
+ ce = search_cache_extent(&map_tree->cache_tree, chunk_offset);
BUG_ON(!ce);
map = container_of(ce, struct map_lookup, ce);
logical = key->offset;
length = btrfs_chunk_length(leaf, chunk);
- ce = find_first_cache_extent(&map_tree->cache_tree, logical);
+ ce = search_cache_extent(&map_tree->cache_tree, logical);
/* already mapped? */
if (ce && ce->start <= logical && ce->start + ce->size > logical) {
}
num_stripes = btrfs_chunk_num_stripes(leaf, chunk);
- map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS);
+ map = kmalloc(btrfs_map_lookup_size(num_stripes), GFP_NOFS);
if (!map)
return -ENOMEM;
}
}
- ret = insert_existing_cache_extent(&map_tree->cache_tree, &map->ce);
+ ret = insert_cache_extent(&map_tree->cache_tree, &map->ce);
BUG_ON(ret);
return 0;
if (!device) {
printk("warning devid %llu not found already\n",
(unsigned long long)devid);
- device = kmalloc(sizeof(*device), GFP_NOFS);
+ device = kzalloc(sizeof(*device), GFP_NOFS);
if (!device)
return -ENOMEM;
- device->total_ios = 0;
+ device->fd = -1;
list_add(&device->dev_list,
&root->fs_info->fs_devices->devices);
}
{
return &fs_uuids;
}
+
+static int rmw_eb(struct btrfs_fs_info *info,
+ struct extent_buffer *eb, struct extent_buffer *orig_eb)
+{
+ int ret;
+ unsigned long orig_off = 0;
+ unsigned long dest_off = 0;
+ unsigned long copy_len = eb->len;
+
+ ret = read_whole_eb(info, eb, 0);
+ if (ret)
+ return ret;
+
+ if (eb->start + eb->len <= orig_eb->start ||
+ eb->start >= orig_eb->start + orig_eb->len)
+ return 0;
+ /*
+ * | ----- orig_eb ------- |
+ * | ----- stripe ------- |
+ * | ----- orig_eb ------- |
+ * | ----- orig_eb ------- |
+ */
+ if (eb->start > orig_eb->start)
+ orig_off = eb->start - orig_eb->start;
+ if (orig_eb->start > eb->start)
+ dest_off = orig_eb->start - eb->start;
+
+ if (copy_len > orig_eb->len - orig_off)
+ copy_len = orig_eb->len - orig_off;
+ if (copy_len > eb->len - dest_off)
+ copy_len = eb->len - dest_off;
+
+ memcpy(eb->data + dest_off, orig_eb->data + orig_off, copy_len);
+ return 0;
+}
+
+static void split_eb_for_raid56(struct btrfs_fs_info *info,
+ struct extent_buffer *orig_eb,
+ struct extent_buffer **ebs,
+ u64 stripe_len, u64 *raid_map,
+ int num_stripes)
+{
+ struct extent_buffer *eb;
+ u64 start = orig_eb->start;
+ u64 this_eb_start;
+ int i;
+ int ret;
+
+ for (i = 0; i < num_stripes; i++) {
+ if (raid_map[i] >= BTRFS_RAID5_P_STRIPE)
+ break;
+
+ eb = malloc(sizeof(struct extent_buffer) + stripe_len);
+ if (!eb)
+ BUG();
+ memset(eb, 0, sizeof(struct extent_buffer) + stripe_len);
+
+ eb->start = raid_map[i];
+ eb->len = stripe_len;
+ eb->refs = 1;
+ eb->flags = 0;
+ eb->fd = -1;
+ eb->dev_bytenr = (u64)-1;
+
+ this_eb_start = raid_map[i];
+
+ if (start > this_eb_start ||
+ start + orig_eb->len < this_eb_start + stripe_len) {
+ ret = rmw_eb(info, eb, orig_eb);
+ BUG_ON(ret);
+ } else {
+ memcpy(eb->data, orig_eb->data + eb->start - start, stripe_len);
+ }
+ ebs[i] = eb;
+ }
+}
+
+int write_raid56_with_parity(struct btrfs_fs_info *info,
+ struct extent_buffer *eb,
+ struct btrfs_multi_bio *multi,
+ u64 stripe_len, u64 *raid_map)
+{
+ struct extent_buffer *ebs[multi->num_stripes], *p_eb = NULL, *q_eb = NULL;
+ int i;
+ int j;
+ int ret;
+ int alloc_size = eb->len;
+
+ if (stripe_len > alloc_size)
+ alloc_size = stripe_len;
+
+ split_eb_for_raid56(info, eb, ebs, stripe_len, raid_map,
+ multi->num_stripes);
+
+ for (i = 0; i < multi->num_stripes; i++) {
+ struct extent_buffer *new_eb;
+ if (raid_map[i] < BTRFS_RAID5_P_STRIPE) {
+ ebs[i]->dev_bytenr = multi->stripes[i].physical;
+ ebs[i]->fd = multi->stripes[i].dev->fd;
+ multi->stripes[i].dev->total_ios++;
+ BUG_ON(ebs[i]->start != raid_map[i]);
+ continue;
+ }
+ new_eb = kmalloc(sizeof(*eb) + alloc_size, GFP_NOFS);
+ BUG_ON(!new_eb);
+ new_eb->dev_bytenr = multi->stripes[i].physical;
+ new_eb->fd = multi->stripes[i].dev->fd;
+ multi->stripes[i].dev->total_ios++;
+ new_eb->len = stripe_len;
+
+ if (raid_map[i] == BTRFS_RAID5_P_STRIPE)
+ p_eb = new_eb;
+ else if (raid_map[i] == BTRFS_RAID6_Q_STRIPE)
+ q_eb = new_eb;
+ }
+ if (q_eb) {
+ void *pointers[multi->num_stripes];
+ ebs[multi->num_stripes - 2] = p_eb;
+ ebs[multi->num_stripes - 1] = q_eb;
+
+ for (i = 0; i < multi->num_stripes; i++)
+ pointers[i] = ebs[i]->data;
+
+ raid6_gen_syndrome(multi->num_stripes, stripe_len, pointers);
+ } else {
+ ebs[multi->num_stripes - 1] = p_eb;
+ memcpy(p_eb->data, ebs[0]->data, stripe_len);
+ for (j = 1; j < multi->num_stripes - 1; j++) {
+ for (i = 0; i < stripe_len; i += sizeof(unsigned long)) {
+ *(unsigned long *)(p_eb->data + i) ^=
+ *(unsigned long *)(ebs[j]->data + i);
+ }
+ }
+ }
+
+ for (i = 0; i < multi->num_stripes; i++) {
+ ret = write_extent_to_disk(ebs[i]);
+ BUG_ON(ret);
+ if (ebs[i] != eb)
+ kfree(ebs[i]);
+ }
+ return 0;
+}