1 files changed, 59 insertions, 39 deletions

diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c
index d2a9a1ee5361..208638384cd2 100644
--- a/fs/btrfs/raid56.c
+++ b/fs/btrfs/raid56.c
@@ -31,7 +31,7 @@
 #include <linux/hash.h>
 #include <linux/list_sort.h>
 #include <linux/raid/xor.h>
-#include <linux/vmalloc.h>
+#include <linux/mm.h>
 #include <asm/div64.h>
 #include "ctree.h"
 #include "extent_map.h"
@@ -149,7 +149,7 @@ struct btrfs_raid_bio {
 
 	int generic_bio_cnt;
 
-	atomic_t refs;
+	refcount_t refs;
 
 	atomic_t stripes_pending;
 
@@ -218,12 +218,9 @@ int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info)
 	 * of a failing mount.
 	 */
 	table_size = sizeof(*table) + sizeof(*h) * num_entries;
-	table = kzalloc(table_size, GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT);
-	if (!table) {
-		table = vzalloc(table_size);
-		if (!table)
-			return -ENOMEM;
-	}
+	table = kvzalloc(table_size, GFP_KERNEL);
+	if (!table)
+		return -ENOMEM;
 
 	spin_lock_init(&table->cache_lock);
 	INIT_LIST_HEAD(&table->stripe_cache);
@@ -389,7 +386,7 @@ static void __remove_rbio_from_cache(struct btrfs_raid_bio *rbio)
 		if (bio_list_empty(&rbio->bio_list)) {
 			if (!list_empty(&rbio->hash_list)) {
 				list_del_init(&rbio->hash_list);
-				atomic_dec(&rbio->refs);
+				refcount_dec(&rbio->refs);
 				BUG_ON(!list_empty(&rbio->plug_list));
 			}
 		}
@@ -480,7 +477,7 @@ static void cache_rbio(struct btrfs_raid_bio *rbio)
 
 	/* bump our ref if we were not in the list before */
 	if (!test_and_set_bit(RBIO_CACHE_BIT, &rbio->flags))
-		atomic_inc(&rbio->refs);
+		refcount_inc(&rbio->refs);
 
 	if (!list_empty(&rbio->stripe_cache)){
 		list_move(&rbio->stripe_cache, &table->stripe_cache);
@@ -677,11 +674,9 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio)
 	struct btrfs_raid_bio *freeit = NULL;
 	struct btrfs_raid_bio *cache_drop = NULL;
 	int ret = 0;
-	int walk = 0;
 
 	spin_lock_irqsave(&h->lock, flags);
 	list_for_each_entry(cur, &h->hash_list, hash_list) {
-		walk++;
 		if (cur->bbio->raid_map[0] == rbio->bbio->raid_map[0]) {
 			spin_lock(&cur->bio_list_lock);
 
@@ -691,7 +686,7 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio)
 			    test_bit(RBIO_CACHE_BIT, &cur->flags) &&
 			    !test_bit(RBIO_RMW_LOCKED_BIT, &cur->flags)) {
 				list_del_init(&cur->hash_list);
-				atomic_dec(&cur->refs);
+				refcount_dec(&cur->refs);
 
 				steal_rbio(cur, rbio);
 				cache_drop = cur;
@@ -740,7 +735,7 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio)
 		}
 	}
 lockit:
-	atomic_inc(&rbio->refs);
+	refcount_inc(&rbio->refs);
 	list_add(&rbio->hash_list, &h->hash_list);
 out:
 	spin_unlock_irqrestore(&h->lock, flags);
@@ -786,7 +781,7 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio)
 		}
 
 		list_del_init(&rbio->hash_list);
-		atomic_dec(&rbio->refs);
+		refcount_dec(&rbio->refs);
 
 		/*
 		 * we use the plug list to hold all the rbios
@@ -803,7 +798,7 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio)
 			list_del_init(&rbio->plug_list);
 
 			list_add(&next->hash_list, &h->hash_list);
-			atomic_inc(&next->refs);
+			refcount_inc(&next->refs);
 			spin_unlock(&rbio->bio_list_lock);
 			spin_unlock_irqrestore(&h->lock, flags);
 
@@ -845,8 +840,7 @@ static void __free_raid_bio(struct btrfs_raid_bio *rbio)
 {
 	int i;
 
-	WARN_ON(atomic_read(&rbio->refs) < 0);
-	if (!atomic_dec_and_test(&rbio->refs))
+	if (!refcount_dec_and_test(&rbio->refs))
 		return;
 
 	WARN_ON(!list_empty(&rbio->stripe_cache));
@@ -874,7 +868,7 @@ static void free_raid_bio(struct btrfs_raid_bio *rbio)
  * this frees the rbio and runs through all the bios in the
  * bio_list and calls end_io on them
  */
-static void rbio_orig_end_io(struct btrfs_raid_bio *rbio, int err)
+static void rbio_orig_end_io(struct btrfs_raid_bio *rbio, blk_status_t err)
 {
 	struct bio *cur = bio_list_get(&rbio->bio_list);
 	struct bio *next;
@@ -887,7 +881,7 @@ static void rbio_orig_end_io(struct btrfs_raid_bio *rbio, int err)
 	while (cur) {
 		next = cur->bi_next;
 		cur->bi_next = NULL;
-		cur->bi_error = err;
+		cur->bi_status = err;
 		bio_endio(cur);
 		cur = next;
 	}
@@ -900,7 +894,7 @@ static void rbio_orig_end_io(struct btrfs_raid_bio *rbio, int err)
 static void raid_write_end_io(struct bio *bio)
 {
 	struct btrfs_raid_bio *rbio = bio->bi_private;
-	int err = bio->bi_error;
+	blk_status_t err = bio->bi_status;
 	int max_errors;
 
 	if (err)
@@ -917,7 +911,7 @@ static void raid_write_end_io(struct bio *bio)
 	max_errors = (rbio->operation == BTRFS_RBIO_PARITY_SCRUB) ?
 		     0 : rbio->bbio->max_errors;
 	if (atomic_read(&rbio->error) > max_errors)
-		err = -EIO;
+		err = BLK_STS_IOERR;
 
 	rbio_orig_end_io(rbio, err);
 }
@@ -999,7 +993,7 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_fs_info *fs_info,
 	rbio->stripe_npages = stripe_npages;
 	rbio->faila = -1;
 	rbio->failb = -1;
-	atomic_set(&rbio->refs, 1);
+	refcount_set(&rbio->refs, 1);
 	atomic_set(&rbio->error, 0);
 	atomic_set(&rbio->stripes_pending, 0);
 
@@ -1095,7 +1089,7 @@ static int rbio_add_io_page(struct btrfs_raid_bio *rbio,
 		 * devices or if they are not contiguous
 		 */
 		if (last_end == disk_start && stripe->dev->bdev &&
-		    !last->bi_error &&
+		    !last->bi_status &&
 		    last->bi_bdev == stripe->dev->bdev) {
 			ret = bio_add_page(last, page, PAGE_SIZE, 0);
 			if (ret == PAGE_SIZE)
@@ -1104,10 +1098,7 @@ static int rbio_add_io_page(struct btrfs_raid_bio *rbio,
 	}
 
 	/* put a new bio on the list */
-	bio = btrfs_io_bio_alloc(GFP_NOFS, bio_max_len >> PAGE_SHIFT?:1);
-	if (!bio)
-		return -ENOMEM;
-
+	bio = btrfs_io_bio_alloc(bio_max_len >> PAGE_SHIFT ?: 1);
 	bio->bi_iter.bi_size = 0;
 	bio->bi_bdev = stripe->dev->bdev;
 	bio->bi_iter.bi_sector = disk_start >> 9;
@@ -1145,20 +1136,27 @@ static void validate_rbio_for_rmw(struct btrfs_raid_bio *rbio)
 static void index_rbio_pages(struct btrfs_raid_bio *rbio)
 {
 	struct bio *bio;
-	struct bio_vec *bvec;
 	u64 start;
 	unsigned long stripe_offset;
 	unsigned long page_index;
-	int i;
 
 	spin_lock_irq(&rbio->bio_list_lock);
 	bio_list_for_each(bio, &rbio->bio_list) {
+		struct bio_vec bvec;
+		struct bvec_iter iter;
+		int i = 0;
+
 		start = (u64)bio->bi_iter.bi_sector << 9;
 		stripe_offset = start - rbio->bbio->raid_map[0];
 		page_index = stripe_offset >> PAGE_SHIFT;
 
-		bio_for_each_segment_all(bvec, bio, i)
-			rbio->bio_pages[page_index + i] = bvec->bv_page;
+		if (bio_flagged(bio, BIO_CLONED))
+			bio->bi_iter = btrfs_io_bio(bio)->iter;
+
+		bio_for_each_segment(bvec, bio, iter) {
+			rbio->bio_pages[page_index + i] = bvec.bv_page;
+			i++;
+		}
 	}
 	spin_unlock_irq(&rbio->bio_list_lock);
 }
@@ -1432,11 +1430,14 @@ static int fail_bio_stripe(struct btrfs_raid_bio *rbio,
  */
 static void set_bio_pages_uptodate(struct bio *bio)
 {
-	struct bio_vec *bvec;
-	int i;
+	struct bio_vec bvec;
+	struct bvec_iter iter;
+
+	if (bio_flagged(bio, BIO_CLONED))
+		bio->bi_iter = btrfs_io_bio(bio)->iter;
 
-	bio_for_each_segment_all(bvec, bio, i)
-		SetPageUptodate(bvec->bv_page);
+	bio_for_each_segment(bvec, bio, iter)
+		SetPageUptodate(bvec.bv_page);
 }
 
 /*
@@ -1451,7 +1452,7 @@ static void raid_rmw_end_io(struct bio *bio)
 {
 	struct btrfs_raid_bio *rbio = bio->bi_private;
 
-	if (bio->bi_error)
+	if (bio->bi_status)
 		fail_bio_stripe(rbio, bio);
 	else
 		set_bio_pages_uptodate(bio);
@@ -1994,7 +1995,7 @@ static void raid_recover_end_io(struct bio *bio)
 	 * we only read stripe pages off the disk, set them
 	 * up to date if there were no errors
 	 */
-	if (bio->bi_error)
+	if (bio->bi_status)
 		fail_bio_stripe(rbio, bio);
 	else
 		set_bio_pages_uptodate(bio);
@@ -2120,6 +2121,11 @@ int raid56_parity_recover(struct btrfs_fs_info *fs_info, struct bio *bio,
 	struct btrfs_raid_bio *rbio;
 	int ret;
 
+	if (generic_io) {
+		ASSERT(bbio->mirror_num == mirror_num);
+		btrfs_io_bio(bio)->mirror_num = mirror_num;
+	}
+
 	rbio = alloc_rbio(fs_info, bbio, stripe_len);
 	if (IS_ERR(rbio)) {
 		if (generic_io)
@@ -2196,6 +2202,8 @@ static void read_rebuild_work(struct btrfs_work *work)
 /*
  * The following code is used to scrub/replace the parity stripe
  *
+ * Caller must have already increased bio_counter for getting @bbio.
+ *
  * Note: We need make sure all the pages that add into the scrub/replace
  * raid bio are correct and not be changed during the scrub/replace. That
  * is those pages just hold metadata or file data with checksum.
@@ -2233,6 +2241,12 @@ raid56_parity_alloc_scrub_rbio(struct btrfs_fs_info *fs_info, struct bio *bio,
 	ASSERT(rbio->stripe_npages == stripe_nsectors);
 	bitmap_copy(rbio->dbitmap, dbitmap, stripe_nsectors);
 
+	/*
+	 * We have already increased bio_counter when getting bbio, record it
+	 * so we can free it at rbio_orig_end_io().
+	 */
+	rbio->generic_bio_cnt = 1;
+
 	return rbio;
 }
 
@@ -2520,7 +2534,7 @@ static void raid56_parity_scrub_end_io(struct bio *bio)
 {
 	struct btrfs_raid_bio *rbio = bio->bi_private;
 
-	if (bio->bi_error)
+	if (bio->bi_status)
 		fail_bio_stripe(rbio, bio);
 	else
 		set_bio_pages_uptodate(bio);
@@ -2675,6 +2689,12 @@ raid56_alloc_missing_rbio(struct btrfs_fs_info *fs_info, struct bio *bio,
 		return NULL;
 	}
 
+	/*
+	 * When we get bbio, we have already increased bio_counter, record it
+	 * so we can free it at rbio_orig_end_io()
+	 */
+	rbio->generic_bio_cnt = 1;
+
 	return rbio;
 }