]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
btrfs: add might_sleep() annotations
authorChenXiaoSong <chenxiaosong2@huawei.com>
Wed, 16 Nov 2022 14:23:53 +0000 (22:23 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sat, 14 Dec 2024 18:51:22 +0000 (19:51 +0100)
commit6e2ac5f585a0f5b285c6245cd6eda8e4408868d9
tree15a6dd805638e040131a70b7f55549a493ae70a4
parentab6cfcf8ed2c7496f55d020b65b1d8cd55d9a2cb
btrfs: add might_sleep() annotations

[ Upstream commit a4c853af0c511d7e0f7cb306bbc8a4f1dbdb64ca ]

Add annotations to functions that might sleep due to allocations or IO
and could be called from various contexts. In case of btrfs_search_slot
it's not obvious why it would sleep:

    btrfs_search_slot
      setup_nodes_for_search
        reada_for_balance
          btrfs_readahead_node_child
            btrfs_readahead_tree_block
              btrfs_find_create_tree_block
                alloc_extent_buffer
                  kmem_cache_zalloc
                    /* allocate memory non-atomically, might sleep */
                    kmem_cache_alloc(GFP_NOFS|__GFP_NOFAIL|__GFP_ZERO)
              read_extent_buffer_pages
                submit_extent_page
                  /* disk IO, might sleep */
                  submit_one_bio

Other examples where the sleeping could happen is in 3 places might
sleep in update_qgroup_limit_item(), as shown below:

  update_qgroup_limit_item
    btrfs_alloc_path
      /* allocate memory non-atomically, might sleep */
      kmem_cache_zalloc(btrfs_path_cachep, GFP_NOFS)

Signed-off-by: ChenXiaoSong <chenxiaosong2@huawei.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Stable-dep-of: 3ed51857a50f ("btrfs: add a sanity check for btrfs root in btrfs_search_slot()")
Signed-off-by: Sasha Levin <sashal@kernel.org>
fs/btrfs/ctree.c