@@ -372,11 +372,10 @@ static void f2fs_write_end_io(struct bio *bio)
#ifdef CONFIG_BLK_DEV_ZONED
static void f2fs_zone_write_end_io(struct bio *bio)
{
- struct f2fs_bio_info *io = (struct f2fs_bio_info *)bio->bi_private;
+ struct f2fs_sb_info *sbi = iostat_get_bio_private(bio);
- bio->bi_private = io->bi_private;
- complete(&io->zone_wait);
f2fs_write_end_io(bio);
+ up(&sbi->available_open_zones);
}
#endif
@@ -532,6 +531,24 @@ static void __submit_merged_bio(struct f2fs_bio_info *io)
if (!io->bio)
return;
+#ifdef CONFIG_BLK_DEV_ZONED
+ if (io->open_zone) {
+ /*
+ * if there is no open zone, it will wait for last IO in
+ * previous zone before submitting new IO.
+ */
+ down(&fio->sbi->available_open_zones);
+ io->open_zone = false;
+ io->zone_opened = true;
+ }
+
+ if (io->close_zone) {
+ io->bio->bi_end_io = f2fs_zone_write_end_io;
+ io->zone_opened = false;
+ io->close_zone = false;
+ }
+#endif
+
if (is_read_io(fio->op)) {
trace_f2fs_prepare_read_bio(io->sbi->sb, fio->type, io->bio);
f2fs_submit_read_bio(io->sbi, io->bio, fio->type);
@@ -605,9 +622,9 @@ int f2fs_init_write_merge_io(struct f2fs_sb_info *sbi)
INIT_LIST_HEAD(&io->bio_list);
init_f2fs_rwsem(&io->bio_list_lock);
#ifdef CONFIG_BLK_DEV_ZONED
- init_completion(&io->zone_wait);
- io->zone_pending_bio = NULL;
- io->bi_private = NULL;
+ io->open_zone = false;
+ io->zone_opened = false;
+ io->close_zone = false;
#endif
}
}
@@ -638,6 +655,31 @@ static void __f2fs_submit_merged_write(struct f2fs_sb_info *sbi,
f2fs_up_write(&io->io_rwsem);
}
+void f2fs_blkzoned_submit_merged_write(struct f2fs_sb_info *sbi, int type)
+{
+#ifdef CONFIG_BLK_DEV_ZONED
+ struct f2fs_bio_info *io;
+
+ if (!f2fs_sb_has_blkzoned(sbi))
+ return;
+
+ io = sbi->write_io[PAGE_TYPE(type)] + f2fs_get_segment_temp(sbi, type);
+
+ f2fs_down_write(&io->io_rwsem);
+ if (io->zone_opened) {
+ if (io->bio) {
+ io->close_zone = true;
+ __submit_merged_bio(io);
+ } else {
+ up(&sbi->available_open_zones);
+ io->zone_opened = false;
+ }
+ }
+ f2fs_up_write(&io->io_rwsem);
+#endif
+
+}
+
static void __submit_merged_write_cond(struct f2fs_sb_info *sbi,
struct inode *inode, struct page *page,
nid_t ino, enum page_type type, bool force)
@@ -922,24 +964,21 @@ int f2fs_merge_page_bio(struct f2fs_io_info *fio)
}
#ifdef CONFIG_BLK_DEV_ZONED
-static bool is_end_zone_blkaddr(struct f2fs_sb_info *sbi, block_t blkaddr)
+bool f2fs_is_blkaddr_zone_boundary(struct f2fs_sb_info *sbi,
+ block_t blkaddr, bool start)
{
- struct block_device *bdev = sbi->sb->s_bdev;
- int devi = 0;
+ if (!f2fs_blkaddr_in_seqzone(sbi, blkaddr))
+ return false;
if (f2fs_is_multi_device(sbi)) {
- devi = f2fs_target_device_index(sbi, blkaddr);
- if (blkaddr < FDEV(devi).start_blk ||
- blkaddr > FDEV(devi).end_blk) {
- f2fs_err(sbi, "Invalid block %x", blkaddr);
- return false;
- }
+ int devi = f2fs_target_device_index(sbi, blkaddr);
+
blkaddr -= FDEV(devi).start_blk;
- bdev = FDEV(devi).bdev;
}
- return bdev_is_zoned(bdev) &&
- f2fs_blkz_is_seq(sbi, devi, blkaddr) &&
- (blkaddr % sbi->blocks_per_blkz == sbi->blocks_per_blkz - 1);
+
+ if (start)
+ return (blkaddr % sbi->blocks_per_blkz) == 0;
+ return (blkaddr % sbi->blocks_per_blkz == sbi->blocks_per_blkz - 1);
}
#endif
@@ -950,20 +989,14 @@ void f2fs_submit_page_write(struct f2fs_io_info *fio)
struct f2fs_bio_info *io = sbi->write_io[btype] + fio->temp;
struct page *bio_page;
enum count_type type;
+#ifdef CONFIG_BLK_DEV_ZONED
+ bool blkzoned = f2fs_sb_has_blkzoned(sbi) && btype < META;
+#endif
f2fs_bug_on(sbi, is_read_io(fio->op));
f2fs_down_write(&io->io_rwsem);
next:
-#ifdef CONFIG_BLK_DEV_ZONED
- if (f2fs_sb_has_blkzoned(sbi) && btype < META && io->zone_pending_bio) {
- wait_for_completion_io(&io->zone_wait);
- bio_put(io->zone_pending_bio);
- io->zone_pending_bio = NULL;
- io->bi_private = NULL;
- }
-#endif
-
if (fio->in_list) {
spin_lock(&io->io_lock);
if (list_empty(&io->io_list)) {
@@ -991,6 +1024,12 @@ void f2fs_submit_page_write(struct f2fs_io_info *fio)
type = WB_DATA_TYPE(bio_page, fio->compressed_page);
inc_page_count(sbi, type);
+#ifdef CONFIG_BLK_DEV_ZONED
+ if (blkzoned &&
+ f2fs_is_blkaddr_zone_boundary(sbi, fio->new_blkaddr, true))
+ io->open_zone = true;
+#endif
+
if (io->bio &&
(!io_is_mergeable(sbi, io->bio, io, fio, io->last_block_in_bio,
fio->new_blkaddr) ||
@@ -1016,15 +1055,11 @@ void f2fs_submit_page_write(struct f2fs_io_info *fio)
io->last_block_in_bio = fio->new_blkaddr;
trace_f2fs_submit_page_write(fio->page, fio);
+
#ifdef CONFIG_BLK_DEV_ZONED
- if (f2fs_sb_has_blkzoned(sbi) && btype < META &&
- is_end_zone_blkaddr(sbi, fio->new_blkaddr)) {
- bio_get(io->bio);
- reinit_completion(&io->zone_wait);
- io->bi_private = io->bio->bi_private;
- io->bio->bi_private = io;
- io->bio->bi_end_io = f2fs_zone_write_end_io;
- io->zone_pending_bio = io->bio;
+ if (blkzoned &&
+ f2fs_is_blkaddr_zone_boundary(sbi, fio->new_blkaddr, false)) {
+ io->close_zone = true;
__submit_merged_bio(io);
}
#endif
@@ -1241,16 +1241,16 @@ struct f2fs_bio_info {
struct bio *bio; /* bios to merge */
sector_t last_block_in_bio; /* last block number */
struct f2fs_io_info fio; /* store buffered io info. */
-#ifdef CONFIG_BLK_DEV_ZONED
- struct completion zone_wait; /* condition value for the previous open zone to close */
- struct bio *zone_pending_bio; /* pending bio for the previous zone */
- void *bi_private; /* previous bi_private for pending bio */
-#endif
struct f2fs_rwsem io_rwsem; /* blocking op for bio */
spinlock_t io_lock; /* serialize DATA/NODE IOs */
struct list_head io_list; /* track fios */
struct list_head bio_list; /* bio entry list head */
struct f2fs_rwsem bio_list_lock; /* lock to protect bio entry list */
+#ifdef CONFIG_BLK_DEV_ZONED
+ bool open_zone; /* open a zone */
+ bool zone_opened; /* zone has been opened */
+ bool close_zone; /* close a zone */
+#endif
};
#define FDEV(i) (sbi->devs[i])
@@ -1572,6 +1572,7 @@ struct f2fs_sb_info {
unsigned int max_open_zones; /* max open zone resources of the zoned device */
/* For adjust the priority writing position of data in zone UFS */
unsigned int blkzone_alloc_policy;
+ struct semaphore available_open_zones; /* available open zones */
#endif
/* for node-related operations */
@@ -3861,6 +3862,7 @@ void f2fs_destroy_bio_entry_cache(void);
void f2fs_submit_read_bio(struct f2fs_sb_info *sbi, struct bio *bio,
enum page_type type);
int f2fs_init_write_merge_io(struct f2fs_sb_info *sbi);
+void f2fs_blkzoned_submit_merged_write(struct f2fs_sb_info *sbi, int type);
void f2fs_submit_merged_write(struct f2fs_sb_info *sbi, enum page_type type);
void f2fs_submit_merged_write_cond(struct f2fs_sb_info *sbi,
struct inode *inode, struct page *page,
@@ -3870,6 +3872,10 @@ void f2fs_submit_merged_ipu_write(struct f2fs_sb_info *sbi,
void f2fs_flush_merged_writes(struct f2fs_sb_info *sbi);
int f2fs_submit_page_bio(struct f2fs_io_info *fio);
int f2fs_merge_page_bio(struct f2fs_io_info *fio);
+#ifdef CONFIG_BLK_DEV_ZONED
+bool f2fs_is_blkaddr_zone_boundary(struct f2fs_sb_info *sbi,
+ block_t blkaddr, bool start);
+#endif
void f2fs_submit_page_write(struct f2fs_io_info *fio);
struct block_device *f2fs_target_device(struct f2fs_sb_info *sbi,
block_t blk_addr, sector_t *sector);
@@ -4541,6 +4547,27 @@ static inline bool f2fs_blkz_is_seq(struct f2fs_sb_info *sbi, int devi,
return test_bit(zno, FDEV(devi).blkz_seq);
}
+
+static inline bool f2fs_blkaddr_in_seqzone(struct f2fs_sb_info *sbi,
+ block_t blkaddr)
+{
+ struct block_device *bdev = sbi->sb->s_bdev;
+ int devi = 0;
+
+ if (f2fs_is_multi_device(sbi)) {
+ devi = f2fs_target_device_index(sbi, blkaddr);
+ if (blkaddr < FDEV(devi).start_blk ||
+ blkaddr > FDEV(devi).end_blk) {
+ f2fs_err(sbi, "Invalid block %x", blkaddr);
+ return false;
+ }
+ blkaddr -= FDEV(devi).start_blk;
+ bdev = FDEV(devi).bdev;
+ }
+
+ return bdev_is_zoned(bdev) &&
+ f2fs_blkz_is_seq(sbi, devi, blkaddr);
+}
#endif
static inline int f2fs_bdev_index(struct f2fs_sb_info *sbi,
@@ -237,6 +237,13 @@ static inline void __update_iostat_latency(struct bio_iostat_ctx *iostat_ctx,
spin_unlock_irqrestore(&sbi->iostat_lat_lock, flags);
}
+void *iostat_get_bio_private(struct bio *bio)
+{
+ struct bio_iostat_ctx *iostat_ctx = bio->bi_private;
+
+ return iostat_ctx->sbi;
+}
+
void iostat_update_and_unbind_ctx(struct bio *bio)
{
struct bio_iostat_ctx *iostat_ctx = bio->bi_private;
@@ -58,6 +58,7 @@ static inline struct bio_post_read_ctx *get_post_read_ctx(struct bio *bio)
return iostat_ctx->post_read_ctx;
}
+extern void *iostat_get_bio_private(struct bio *bio);
extern void iostat_update_and_unbind_ctx(struct bio *bio);
extern void iostat_alloc_and_bind_ctx(struct f2fs_sb_info *sbi,
struct bio *bio, struct bio_post_read_ctx *ctx);
@@ -68,6 +69,7 @@ extern void f2fs_destroy_iostat(struct f2fs_sb_info *sbi);
#else
static inline void f2fs_update_iostat(struct f2fs_sb_info *sbi, struct inode *inode,
enum iostat_type type, unsigned long long io_bytes) {}
+static inline void *iostat_get_bio_private(struct bio *bio) { return bio->bi_private; }
static inline void iostat_update_and_unbind_ctx(struct bio *bio) {}
static inline void iostat_alloc_and_bind_ctx(struct f2fs_sb_info *sbi,
struct bio *bio, struct bio_post_read_ctx *ctx) {}
@@ -3230,6 +3230,10 @@ static int __allocate_new_segment(struct f2fs_sb_info *sbi, int type,
return err;
stat_inc_seg_type(sbi, curseg);
locate_dirty_segment(sbi, old_segno);
+
+ if (new_sec)
+ f2fs_blkzoned_submit_merged_write(sbi, type);
+
return 0;
}
@@ -4316,6 +4320,30 @@ static int restore_curseg_summaries(struct f2fs_sb_info *sbi)
return -EINVAL;
}
+#ifdef CONFIG_BLK_DEV_ZONED
+ if (f2fs_sb_has_blkzoned(sbi)) {
+ for (type = 0; type < NR_PERSISTENT_LOG; type++) {
+ struct curseg_info *curseg = CURSEG_I(sbi, type);
+ enum page_type ptype;
+ enum temp_type temp;
+
+ /* current segment locates in non-seqzone */
+ if (!f2fs_blkaddr_in_seqzone(sbi,
+ START_BLOCK(sbi, curseg->segno)))
+ continue;
+
+ /* write pointer of zone is zero */
+ if (f2fs_is_blkaddr_zone_boundary(sbi,
+ NEXT_FREE_BLKADDR(sbi, curseg), true))
+ continue;
+
+ ptype = PAGE_TYPE(type);
+ temp = f2fs_get_segment_temp(sbi, type);
+ down(&sbi->available_open_zones);
+ sbi->write_io[ptype][temp].zone_opened = true;
+ }
+ }
+#endif
return 0;
}
@@ -5645,11 +5673,23 @@ static void destroy_curseg(struct f2fs_sb_info *sbi)
if (!array)
return;
- SM_I(sbi)->curseg_array = NULL;
for (i = 0; i < NR_CURSEG_TYPE; i++) {
kfree(array[i].sum_blk);
kfree(array[i].journal);
+
+#ifdef CONFIG_BLK_DEV_ZONED
+ if (f2fs_sb_has_blkzoned(sbi) && i < NR_PERSISTENT_LOG) {
+ enum page_type ptype = PAGE_TYPE(i);
+ enum temp_type temp = f2fs_get_segment_temp(sbi, i);
+
+ if (sbi->write_io[ptype][temp].zone_opened) {
+ up(&sbi->available_open_zones);
+ sbi->write_io[ptype][temp].zone_opened = false;
+ }
+ }
+#endif
}
+ SM_I(sbi)->curseg_array = NULL;
kfree(array);
}
@@ -26,7 +26,8 @@
#define IS_DATASEG(t) ((t) <= CURSEG_COLD_DATA)
#define IS_NODESEG(t) ((t) >= CURSEG_HOT_NODE && (t) <= CURSEG_COLD_NODE)
-#define SE_PAGETYPE(se) ((IS_NODESEG((se)->type) ? NODE : DATA))
+#define PAGE_TYPE(t) (IS_NODESEG(t) ? NODE : DATA)
+#define SE_PAGETYPE(se) (PAGE_TYPE((se)->type))
static inline void sanity_check_seg_type(struct f2fs_sb_info *sbi,
unsigned short seg_type)
@@ -3923,6 +3923,8 @@ static int init_blkz_info(struct f2fs_sb_info *sbi, int devi)
sbi->max_open_zones, F2FS_OPTION(sbi).active_logs);
return -EINVAL;
}
+
+ sema_init(&sbi->available_open_zones, sbi->max_open_zones);
}
zone_sectors = bdev_zone_sectors(bdev);