nilfs2: fix NULL i_assoc_inode dereference in nilfs_mdt_save_to_shadow_map
PATCH 1/1 nilfs2: fix NULL i_assoc_inode dereference in nilfs_mdt_save_to_shadow_map
The DAT inode's btree node cache (i_assoc_inode) is initialized lazily during btree operations. However, nilfs_mdt_save_to_shadow_map() assumes i_assoc_inode is already initialized when copying dirty pages to the shadow map during GC.
If NILFS_IOCTL_CLEAN_SEGMENTS is called immediately after mount before any btree operation has occurred on the DAT inode, i_assoc_inode is NULL leading to a general protection fault.
Fix this by calling nilfs_attach_btree_node_cache() on the DAT inode in nilfs_dat_read() at mount time, ensuring i_assoc_inode is always initialized before any GC operation can use it.
--- fs/nilfs2/dat.c | 3 +++ 1 file changed, 3 insertions(+)
fs/nilfs2/dat.c +3 -0
@@ -524,6 +524,9 @@ int nilfs_dat_read(struct super_block *sb, size_t entry_size,if (err) goto failed; + err =nilfs_attach_btree_node_cache (dat); +if (err) + goto failed; err =nilfs_read_inode_common (dat, raw_inode);if (err) goto failed;
Signature · 1 line
2.43.0
Ryusuke Konishi
maintainerQuoted 39 lines · Deepanshu Kartikey
The DAT inode's btree node cache (
i_assoc_inode) is initialized lazily during btree operations. However,nilfs_mdt_save_to_shadow_map()assumesi_assoc_inodeis already initialized when copying dirty pages to the shadow map during GC.If
NILFS_IOCTL_CLEAN_SEGMENTSis called immediately after mount before any btree operation has occurred on the DAT inode,i_assoc_inodeis NULL leading to a general protection fault.Fix this by calling
nilfs_attach_btree_node_cache()on the DAT inode innilfs_dat_read()at mount time, ensuringi_assoc_inodeis always initialized before any GC operation can use it.Reported-by [email protected]Tested-by [email protected]Fixes e897be17a441 ("nilfs2: fix lockdep warnings in page operations for btree nodes")Signed-off-by Deepanshu Kartikey <[email protected]>--- fs/nilfs2/dat.c | 3 +++ 1 file changed, 3 insertions(+)fs/nilfs2/dat.c +3
-1@@ -524,6 +524,9 @@ int nilfs_dat_read(struct super_block *sb, size_t entry_size,if (err) goto failed; + err =nilfs_attach_btree_node_cache (dat); +if (err) + goto failed; err =nilfs_read_inode_common (dat, raw_inode);if (err) goto failed; --
Quoted 2 lines
2.43.0
As I've already answered in a discussion about another patch for the same issue, I'd like to adopt Deepanshu's patch as the fix for this issue, after doing a little more review (mainly for error cases) and testing just to be sure.
This fix pre-allocates inodes for the b-tree node cache and is a comprehensive stabilization technique that covers multiple potential cases regarding non-existent b-tree node cache references. Since DAT usually starts using b-tree mapping very quickly, I believe this approach is not overkill and is a practical solution.
Thanks,
Ryusuke Konishi