summaryrefslogtreecommitdiff
path: root/fs/btrfs/zoned.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/btrfs/zoned.c')
-rw-r--r--fs/btrfs/zoned.c126
1 files changed, 65 insertions, 61 deletions
diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
index e00036672f33..359a98e6de85 100644
--- a/fs/btrfs/zoned.c
+++ b/fs/btrfs/zoned.c
@@ -37,8 +37,8 @@
#define BTRFS_SB_LOG_FIRST_OFFSET (512ULL * SZ_1G)
#define BTRFS_SB_LOG_SECOND_OFFSET (4096ULL * SZ_1G)
-#define BTRFS_SB_LOG_FIRST_SHIFT const_ilog2(BTRFS_SB_LOG_FIRST_OFFSET)
-#define BTRFS_SB_LOG_SECOND_SHIFT const_ilog2(BTRFS_SB_LOG_SECOND_OFFSET)
+#define BTRFS_SB_LOG_FIRST_SHIFT ilog2(BTRFS_SB_LOG_FIRST_OFFSET)
+#define BTRFS_SB_LOG_SECOND_SHIFT ilog2(BTRFS_SB_LOG_SECOND_OFFSET)
/* Number of superblock log zones */
#define BTRFS_NR_SB_LOG_ZONES 2
@@ -93,7 +93,8 @@ static int sb_write_pointer(struct block_device *bdev, struct blk_zone *zones,
sector_t sector;
for (int i = 0; i < BTRFS_NR_SB_LOG_ZONES; i++) {
- ASSERT(zones[i].type != BLK_ZONE_TYPE_CONVENTIONAL);
+ ASSERT(zones[i].type != BLK_ZONE_TYPE_CONVENTIONAL,
+ "zones[%d].type=%d", i, zones[i].type);
empty[i] = (zones[i].cond == BLK_ZONE_COND_EMPTY);
full[i] = sb_zone_is_full(&zones[i]);
}
@@ -166,14 +167,14 @@ static inline u32 sb_zone_number(int shift, int mirror)
{
u64 zone = U64_MAX;
- ASSERT(mirror < BTRFS_SUPER_MIRROR_MAX);
+ ASSERT(mirror < BTRFS_SUPER_MIRROR_MAX, "mirror=%d", mirror);
switch (mirror) {
case 0: zone = 0; break;
case 1: zone = 1ULL << (BTRFS_SB_LOG_FIRST_SHIFT - shift); break;
case 2: zone = 1ULL << (BTRFS_SB_LOG_SECOND_SHIFT - shift); break;
}
- ASSERT(zone <= U32_MAX);
+ ASSERT(zone <= U32_MAX, "zone=%llu", zone);
return (u32)zone;
}
@@ -240,7 +241,8 @@ static int btrfs_get_dev_zones(struct btrfs_device *device, u64 pos,
unsigned int i;
u32 zno;
- ASSERT(IS_ALIGNED(pos, zinfo->zone_size));
+ ASSERT(IS_ALIGNED(pos, zinfo->zone_size),
+ "pos=%llu zinfo->zone_size=%llu", pos, zinfo->zone_size);
zno = pos >> zinfo->zone_size_shift;
/*
* We cannot report zones beyond the zone end. So, it is OK to
@@ -264,8 +266,8 @@ static int btrfs_get_dev_zones(struct btrfs_device *device, u64 pos,
}
}
- ret = blkdev_report_zones(device->bdev, pos >> SECTOR_SHIFT, *nr_zones,
- copy_zone_info_cb, zones);
+ ret = blkdev_report_zones_cached(device->bdev, pos >> SECTOR_SHIFT,
+ *nr_zones, copy_zone_info_cb, zones);
if (ret < 0) {
btrfs_err(device->fs_info,
"zoned: failed to read zone %llu on %s (devid %llu)",
@@ -494,6 +496,7 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device, bool populate_cache)
case BLK_ZONE_COND_IMP_OPEN:
case BLK_ZONE_COND_EXP_OPEN:
case BLK_ZONE_COND_CLOSED:
+ case BLK_ZONE_COND_ACTIVE:
__set_bit(nreported, zone_info->active_zones);
nactive++;
break;
@@ -896,9 +899,9 @@ int btrfs_sb_log_location_bdev(struct block_device *bdev, int mirror, int rw,
if (sb_zone + 1 >= nr_zones)
return -ENOENT;
- ret = blkdev_report_zones(bdev, zone_start_sector(sb_zone, bdev),
- BTRFS_NR_SB_LOG_ZONES, copy_zone_info_cb,
- zones);
+ ret = blkdev_report_zones_cached(bdev, zone_start_sector(sb_zone, bdev),
+ BTRFS_NR_SB_LOG_ZONES,
+ copy_zone_info_cb, zones);
if (ret < 0)
return ret;
if (unlikely(ret != BTRFS_NR_SB_LOG_ZONES))
@@ -1055,8 +1058,10 @@ u64 btrfs_find_allocatable_zones(struct btrfs_device *device, u64 hole_start,
bool have_sb;
int i;
- ASSERT(IS_ALIGNED(hole_start, zinfo->zone_size));
- ASSERT(IS_ALIGNED(num_bytes, zinfo->zone_size));
+ ASSERT(IS_ALIGNED(hole_start, zinfo->zone_size),
+ "hole_start=%llu zinfo->zone_size=%llu", hole_start, zinfo->zone_size);
+ ASSERT(IS_ALIGNED(num_bytes, zinfo->zone_size),
+ "num_bytes=%llu zinfo->zone_size=%llu", num_bytes, zinfo->zone_size);
while (pos < hole_end) {
begin = pos >> shift;
@@ -1172,8 +1177,10 @@ int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size)
u64 pos;
int ret;
- ASSERT(IS_ALIGNED(start, zinfo->zone_size));
- ASSERT(IS_ALIGNED(size, zinfo->zone_size));
+ ASSERT(IS_ALIGNED(start, zinfo->zone_size),
+ "start=%llu, zinfo->zone_size=%llu", start, zinfo->zone_size);
+ ASSERT(IS_ALIGNED(size, zinfo->zone_size),
+ "size=%llu, zinfo->zone_size=%llu", size, zinfo->zone_size);
if (begin + nbits > zinfo->nr_zones)
return -ERANGE;
@@ -1317,6 +1324,7 @@ static int btrfs_load_zone_info(struct btrfs_fs_info *fs_info, int zone_idx,
if (!btrfs_dev_is_sequential(device, info->physical)) {
up_read(&dev_replace->rwsem);
info->alloc_offset = WP_CONVENTIONAL;
+ info->capacity = device->zone_info->zone_size;
return 0;
}
@@ -1522,6 +1530,8 @@ static int btrfs_load_block_group_raid0(struct btrfs_block_group *bg,
u64 last_alloc)
{
struct btrfs_fs_info *fs_info = bg->fs_info;
+ u64 stripe_nr = 0, stripe_offset = 0;
+ u32 stripe_index = 0;
if ((map->type & BTRFS_BLOCK_GROUP_DATA) && !fs_info->stripe_root) {
btrfs_err(fs_info, "zoned: data %s needs raid-stripe-tree",
@@ -1529,28 +1539,26 @@ static int btrfs_load_block_group_raid0(struct btrfs_block_group *bg,
return -EINVAL;
}
+ if (last_alloc) {
+ u32 factor = map->num_stripes;
+
+ stripe_nr = last_alloc >> BTRFS_STRIPE_LEN_SHIFT;
+ stripe_offset = last_alloc & BTRFS_STRIPE_LEN_MASK;
+ stripe_nr = div_u64_rem(stripe_nr, factor, &stripe_index);
+ }
+
for (int i = 0; i < map->num_stripes; i++) {
if (zone_info[i].alloc_offset == WP_MISSING_DEV)
continue;
if (zone_info[i].alloc_offset == WP_CONVENTIONAL) {
- u64 stripe_nr, full_stripe_nr;
- u64 stripe_offset;
- int stripe_index;
-
- stripe_nr = div64_u64(last_alloc, map->stripe_size);
- stripe_offset = stripe_nr * map->stripe_size;
- full_stripe_nr = div_u64(stripe_nr, map->num_stripes);
- div_u64_rem(stripe_nr, map->num_stripes, &stripe_index);
- zone_info[i].alloc_offset =
- full_stripe_nr * map->stripe_size;
+ zone_info[i].alloc_offset = btrfs_stripe_nr_to_offset(stripe_nr);
if (stripe_index > i)
- zone_info[i].alloc_offset += map->stripe_size;
+ zone_info[i].alloc_offset += BTRFS_STRIPE_LEN;
else if (stripe_index == i)
- zone_info[i].alloc_offset +=
- (last_alloc - stripe_offset);
+ zone_info[i].alloc_offset += stripe_offset;
}
if (test_bit(0, active) != test_bit(i, active)) {
@@ -1574,6 +1582,8 @@ static int btrfs_load_block_group_raid10(struct btrfs_block_group *bg,
u64 last_alloc)
{
struct btrfs_fs_info *fs_info = bg->fs_info;
+ u64 stripe_nr = 0, stripe_offset = 0;
+ u32 stripe_index = 0;
if ((map->type & BTRFS_BLOCK_GROUP_DATA) && !fs_info->stripe_root) {
btrfs_err(fs_info, "zoned: data %s needs raid-stripe-tree",
@@ -1581,6 +1591,14 @@ static int btrfs_load_block_group_raid10(struct btrfs_block_group *bg,
return -EINVAL;
}
+ if (last_alloc) {
+ u32 factor = map->num_stripes / map->sub_stripes;
+
+ stripe_nr = last_alloc >> BTRFS_STRIPE_LEN_SHIFT;
+ stripe_offset = last_alloc & BTRFS_STRIPE_LEN_MASK;
+ stripe_nr = div_u64_rem(stripe_nr, factor, &stripe_index);
+ }
+
for (int i = 0; i < map->num_stripes; i++) {
if (zone_info[i].alloc_offset == WP_MISSING_DEV)
continue;
@@ -1594,26 +1612,12 @@ static int btrfs_load_block_group_raid10(struct btrfs_block_group *bg,
}
if (zone_info[i].alloc_offset == WP_CONVENTIONAL) {
- u64 stripe_nr, full_stripe_nr;
- u64 stripe_offset;
- int stripe_index;
-
- stripe_nr = div64_u64(last_alloc, map->stripe_size);
- stripe_offset = stripe_nr * map->stripe_size;
- full_stripe_nr = div_u64(stripe_nr,
- map->num_stripes / map->sub_stripes);
- div_u64_rem(stripe_nr,
- (map->num_stripes / map->sub_stripes),
- &stripe_index);
-
- zone_info[i].alloc_offset =
- full_stripe_nr * map->stripe_size;
+ zone_info[i].alloc_offset = btrfs_stripe_nr_to_offset(stripe_nr);
if (stripe_index > (i / map->sub_stripes))
- zone_info[i].alloc_offset += map->stripe_size;
+ zone_info[i].alloc_offset += BTRFS_STRIPE_LEN;
else if (stripe_index == (i / map->sub_stripes))
- zone_info[i].alloc_offset +=
- (last_alloc - stripe_offset);
+ zone_info[i].alloc_offset += stripe_offset;
}
if ((i % map->sub_stripes) == 0) {
@@ -1631,7 +1635,7 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
struct btrfs_chunk_map *map;
u64 logical = cache->start;
u64 length = cache->length;
- struct zone_info *zone_info = NULL;
+ struct zone_info AUTO_KFREE(zone_info);
int ret;
int i;
unsigned long *active = NULL;
@@ -1683,8 +1687,6 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
set_bit(BLOCK_GROUP_FLAG_SEQUENTIAL_ZONE, &cache->runtime_flags);
if (num_conventional > 0) {
- /* Zone capacity is always zone size in emulation */
- cache->zone_capacity = cache->length;
ret = calculate_alloc_pointer(cache, &last_alloc, new);
if (ret) {
btrfs_err(fs_info,
@@ -1693,6 +1695,7 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache, bool new)
goto out;
} else if (map->num_stripes == num_conventional) {
cache->alloc_offset = last_alloc;
+ cache->zone_capacity = cache->length;
set_bit(BLOCK_GROUP_FLAG_ZONE_IS_ACTIVE, &cache->runtime_flags);
goto out;
}
@@ -1753,7 +1756,7 @@ out:
!fs_info->stripe_root) {
btrfs_err(fs_info, "zoned: data %s needs raid-stripe-tree",
btrfs_bg_type_to_raid_name(map->type));
- return -EINVAL;
+ ret = -EINVAL;
}
if (unlikely(cache->alloc_offset > cache->zone_capacity)) {
@@ -1786,7 +1789,6 @@ out:
cache->physical_map = NULL;
}
bitmap_free(active);
- kfree(zone_info);
return ret;
}
@@ -1813,14 +1815,14 @@ bool btrfs_use_zone_append(struct btrfs_bio *bbio)
{
u64 start = (bbio->bio.bi_iter.bi_sector << SECTOR_SHIFT);
struct btrfs_inode *inode = bbio->inode;
- struct btrfs_fs_info *fs_info = bbio->fs_info;
+ struct btrfs_fs_info *fs_info = inode->root->fs_info;
struct btrfs_block_group *cache;
bool ret = false;
if (!btrfs_is_zoned(fs_info))
return false;
- if (!inode || !is_data_inode(inode))
+ if (!is_data_inode(inode))
return false;
if (btrfs_op(&bbio->bio) != BTRFS_MAP_WRITE)
@@ -1871,7 +1873,7 @@ static void btrfs_rewrite_logical_zoned(struct btrfs_ordered_extent *ordered,
em = btrfs_search_extent_mapping(em_tree, ordered->file_offset,
ordered->num_bytes);
/* The em should be a new COW extent, thus it should not have an offset. */
- ASSERT(em->offset == 0);
+ ASSERT(em->offset == 0, "em->offset=%llu", em->offset);
em->disk_bytenr = logical;
btrfs_free_extent_map(em);
write_unlock(&em_tree->lock);
@@ -2582,7 +2584,8 @@ again:
struct btrfs_space_info *reloc_sinfo = data_sinfo->sub_group[0];
int factor;
- ASSERT(reloc_sinfo->subgroup_id == BTRFS_SUB_GROUP_DATA_RELOC);
+ ASSERT(reloc_sinfo->subgroup_id == BTRFS_SUB_GROUP_DATA_RELOC,
+ "reloc_sinfo->subgroup_id=%d", reloc_sinfo->subgroup_id);
factor = btrfs_bg_type_to_factor(bg->flags);
down_write(&space_info->groups_sem);
@@ -2596,9 +2599,9 @@ again:
space_info->disk_total -= bg->length * factor;
space_info->disk_total -= bg->zone_unusable;
/* There is no allocation ever happened. */
- ASSERT(bg->used == 0);
+ ASSERT(bg->used == 0, "bg->used=%llu", bg->used);
/* No super block in a block group on the zoned setup. */
- ASSERT(bg->bytes_super == 0);
+ ASSERT(bg->bytes_super == 0, "bg->bytes_super=%llu", bg->bytes_super);
spin_unlock(&space_info->lock);
bg->space_info = reloc_sinfo;
@@ -2624,7 +2627,8 @@ again:
/* Allocate new BG in the data relocation space_info. */
space_info = data_sinfo->sub_group[0];
- ASSERT(space_info->subgroup_id == BTRFS_SUB_GROUP_DATA_RELOC);
+ ASSERT(space_info->subgroup_id == BTRFS_SUB_GROUP_DATA_RELOC,
+ "space_info->subgroup_id=%d", space_info->subgroup_id);
ret = btrfs_chunk_alloc(trans, space_info, alloc_flags, CHUNK_ALLOC_FORCE);
btrfs_end_transaction(trans);
if (ret == 1) {
@@ -2754,10 +2758,9 @@ int btrfs_zone_finish_one_bg(struct btrfs_fs_info *fs_info)
return ret < 0 ? ret : 1;
}
-int btrfs_zoned_activate_one_bg(struct btrfs_fs_info *fs_info,
- struct btrfs_space_info *space_info,
- bool do_finish)
+int btrfs_zoned_activate_one_bg(struct btrfs_space_info *space_info, bool do_finish)
{
+ struct btrfs_fs_info *fs_info = space_info->fs_info;
struct btrfs_block_group *bg;
int index;
@@ -2966,7 +2969,8 @@ int btrfs_reset_unused_block_groups(struct btrfs_space_info *space_info, u64 num
* This holds because we currently reset fully used then freed
* block group.
*/
- ASSERT(reclaimed == bg->zone_capacity);
+ ASSERT(reclaimed == bg->zone_capacity,
+ "reclaimed=%llu bg->zone_capacity=%llu", reclaimed, bg->zone_capacity);
bg->free_space_ctl->free_space += reclaimed;
space_info->bytes_zone_unusable -= reclaimed;
spin_unlock(&bg->lock);
generated by cgit (git 2.34.1) at 2025-12-04 11:12:43 +0000