V1->V2:
	In patch 2, when imode is not set to S_IFREG, the inode also needs
	to be initialized. Otherwise, the check can be bypassed, causing
	the BUG_ON. (found in the review by yangerkun)
V2->V3:
	a. add EXT4_IGET_BAD flag to prevent unexpected bad inode.
	b. check bad quota inode in vfs_setup_quota_inode() instead of in
	   ext4_quota_enable() for more generic approach to this problem.
	c. add helper to check quota inums.

Baokun Li (4):
  ext4: fix bug_on in __es_tree_search caused by bad quota inode
  ext4: add helper to check quota inums
  ext4: add EXT4_IGET_BAD flag to prevent unexpected bad inode
  ext4: fix bug_on in __es_tree_search caused by bad boot loader inode

 fs/ext4/ext4.h   |  3 ++-
 fs/ext4/inode.c  |  8 +++++++-
 fs/ext4/ioctl.c  |  5 +++--
 fs/ext4/super.c  | 28 +++++++++++++++++++++++++---
 fs/quota/dquot.c |  2 ++
 5 files changed, 39 insertions(+), 7 deletions(-)

-- 
2.31.1

On Wed, 26 Oct 2022 12:23:06 +0800, Baokun Li wrote:
> V1->V2:
> 	In patch 2, when imode is not set to S_IFREG, the inode also needs
> 	to be initialized. Otherwise, the check can be bypassed, causing
> 	the BUG_ON. (found in the review by yangerkun)
> V2->V3:
> 	a. add EXT4_IGET_BAD flag to prevent unexpected bad inode.
> 	b. check bad quota inode in vfs_setup_quota_inode() instead of in
> 	   ext4_quota_enable() for more generic approach to this problem.
> 	c. add helper to check quota inums.
> 
> [...]

Applied, thanks!

[1/4] ext4: fix bug_on in __es_tree_search caused by bad quota inode
      commit: c7e9666f28ba9bdeeb99fb0c60a27dbb88f452f4
[2/4] ext4: add helper to check quota inums
      commit: 9c4883f1b41181f2096e2ee4e98111008b77165c
[3/4] ext4: add EXT4_IGET_BAD flag to prevent unexpected bad inode
      commit: 10f525fda2faff81f6cfce2a6bc4b50a5254d9ea
[4/4] ext4: fix bug_on in __es_tree_search caused by bad boot loader inode
      commit: db14233edaf579153d8c92bf3a0ba27ceb87eabc

Best regards,
-- 
Theodore Ts'o <tytso@mit.edu>