bcache: handle cache set verify_ondisk properly for bucket size > 8MB

In bch_btree_cache_alloc() when CONFIG_BCACHE_DEBUG is configured,
allocate memory for c->verify_ondisk may fail if the bucket size > 8MB,
which will require __get_free_pages() to allocate continuous pages
with order > 11 (the default MAX_ORDER of Linux buddy allocator). Such
over size allocation will fail, and cause 2 problems,
- When CONFIG_BCACHE_DEBUG is configured,  bch_btree_verify() does not
  work, because c->verify_ondisk is NULL and bch_btree_verify() returns
  immediately.
- bch_btree_cache_alloc() will fail due to c->verify_ondisk allocation
  failed, then the whole cache device registration fails. And because of
  this failure, the first problem of bch_btree_verify() has no chance to
  be triggered.

This patch fixes the above problem by two means,
1) If pages allocation of c->verify_ondisk fails, set it to NULL and
   returns bch_btree_cache_alloc() with -ENOMEM.
2) When calling __get_free_pages() to allocate c->verify_ondisk pages,
   use ilog2(meta_bucket_pages(&c->sb)) to make sure ilog2() will always
   generate a pages order <= MAX_ORDER (or CONFIG_FORCE_MAX_ZONEORDER).
   Then the buddy system won't directly reject the allocation request.

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 20:00:33 +08:00 коммит произвёл Jens Axboe
Родитель c954ac8d66
Коммит bf6af17065
1 изменённых файлов: 10 добавлений и 2 удалений

Просмотреть файл

@ -738,7 +738,7 @@ void bch_btree_cache_free(struct cache_set *c)
if (c->verify_data) if (c->verify_data)
list_move(&c->verify_data->list, &c->btree_cache); list_move(&c->verify_data->list, &c->btree_cache);
free_pages((unsigned long) c->verify_ondisk, ilog2(bucket_pages(c))); free_pages((unsigned long) c->verify_ondisk, ilog2(meta_bucket_pages(&c->sb)));
#endif #endif
list_splice(&c->btree_cache_freeable, list_splice(&c->btree_cache_freeable,
@ -785,7 +785,15 @@ int bch_btree_cache_alloc(struct cache_set *c)
mutex_init(&c->verify_lock); mutex_init(&c->verify_lock);
c->verify_ondisk = (void *) c->verify_ondisk = (void *)
__get_free_pages(GFP_KERNEL|__GFP_COMP, ilog2(bucket_pages(c))); __get_free_pages(GFP_KERNEL|__GFP_COMP, ilog2(meta_bucket_pages(&c->sb)));
if (!c->verify_ondisk) {
/*
* Don't worry about the mca_rereserve buckets
* allocated in previous for-loop, they will be
* handled properly in bch_cache_set_unregister().
*/
return -ENOMEM;
}
c->verify_data = mca_bucket_alloc(c, &ZERO_KEY, GFP_KERNEL); c->verify_data = mca_bucket_alloc(c, &ZERO_KEY, GFP_KERNEL);