bcache: handle cache prio_buckets and disk_buckets properly for bucket size > 8MB

Similar to c->uuids, struct cache's prio_buckets and disk_buckets also
have the potential memory allocation failure during cache registration
if the bucket size > 8MB.

ca->prio_buckets can be stored on cache device in multiple buckets, its
in-memory space is allocated by kzalloc() interface but normally
allocated by alloc_pages() because the size > KMALLOC_MAX_CACHE_SIZE.

So allocation of ca->prio_buckets has the MAX_ORDER restriction too. If
the bucket size > 8MB, by default the page allocator will fail because
the page order > 11 (default MAX_ORDER value). ca->prio_buckets should
also use meta_bucket_bytes(), meta_bucket_pages() to decide its memory
size and use alloc_meta_bucket_pages() to allocate pages, to avoid the
allocation failure during cache set registration when bucket size > 8MB.

ca->disk_buckets is a single bucket size memory buffer, it is used to
iterate each bucket of ca->prio_buckets, and compose the bio based on
memory of ca->disk_buckets, then write ca->disk_buckets memory to cache
disk one-by-one for each bucket of ca->prio_buckets. ca->disk_buckets
should have in-memory size exact to the meta_bucket_pages(), this is the
size that ca->prio_buckets will be stored into each on-disk bucket.

This patch fixes the above issues and handle cache's prio_buckets and
disk_buckets properly for bucket size larger than 8MB.

Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This commit is contained in:
Coly Li 2020-07-25 20:00:32 +08:00 committed by Jens Axboe
parent 21e478ddb2
commit c954ac8d66
2 changed files with 10 additions and 9 deletions

View file

@ -782,11 +782,12 @@ static inline unsigned int meta_bucket_bytes(struct cache_sb *sb)
return meta_bucket_pages(sb) << PAGE_SHIFT;
}
#define prios_per_bucket(c) \
((bucket_bytes(c) - sizeof(struct prio_set)) / \
#define prios_per_bucket(ca) \
((meta_bucket_bytes(&(ca)->sb) - sizeof(struct prio_set)) / \
sizeof(struct bucket_disk))
#define prio_buckets(c) \
DIV_ROUND_UP((size_t) (c)->sb.nbuckets, prios_per_bucket(c))
#define prio_buckets(ca) \
DIV_ROUND_UP((size_t) (ca)->sb.nbuckets, prios_per_bucket(ca))
static inline size_t sector_to_bucket(struct cache_set *c, sector_t s)
{

View file

@ -563,7 +563,7 @@ static void prio_io(struct cache *ca, uint64_t bucket, int op,
bio->bi_iter.bi_sector = bucket * ca->sb.bucket_size;
bio_set_dev(bio, ca->bdev);
bio->bi_iter.bi_size = bucket_bytes(ca);
bio->bi_iter.bi_size = meta_bucket_bytes(&ca->sb);
bio->bi_end_io = prio_endio;
bio->bi_private = ca;
@ -621,7 +621,7 @@ int bch_prio_write(struct cache *ca, bool wait)
p->next_bucket = ca->prio_buckets[i + 1];
p->magic = pset_magic(&ca->sb);
p->csum = bch_crc64(&p->magic, bucket_bytes(ca) - 8);
p->csum = bch_crc64(&p->magic, meta_bucket_bytes(&ca->sb) - 8);
bucket = bch_bucket_alloc(ca, RESERVE_PRIO, wait);
BUG_ON(bucket == -1);
@ -674,7 +674,7 @@ static int prio_read(struct cache *ca, uint64_t bucket)
prio_io(ca, bucket, REQ_OP_READ, 0);
if (p->csum !=
bch_crc64(&p->magic, bucket_bytes(ca) - 8)) {
bch_crc64(&p->magic, meta_bucket_bytes(&ca->sb) - 8)) {
pr_warn("bad csum reading priorities\n");
goto out;
}
@ -2230,7 +2230,7 @@ void bch_cache_release(struct kobject *kobj)
ca->set->cache[ca->sb.nr_this_dev] = NULL;
}
free_pages((unsigned long) ca->disk_buckets, ilog2(bucket_pages(ca)));
free_pages((unsigned long) ca->disk_buckets, ilog2(meta_bucket_pages(&ca->sb)));
kfree(ca->prio_buckets);
vfree(ca->buckets);
@ -2327,7 +2327,7 @@ static int cache_alloc(struct cache *ca)
goto err_prio_buckets_alloc;
}
ca->disk_buckets = alloc_bucket_pages(GFP_KERNEL, ca);
ca->disk_buckets = alloc_meta_bucket_pages(GFP_KERNEL, &ca->sb);
if (!ca->disk_buckets) {
err = "ca->disk_buckets alloc failed";
goto err_disk_buckets_alloc;