ext4: fix i_data_sem unlock order in ext4_ind_migrate()
PATCH v3 1/1 ext4: fix i_data_sem unlock order in ext4_ind_migrate()
Fuzzing reports a possible deadlock in jbd2_log_wait_commit.
This issue is triggered when an EXT4_IOC_MIGRATE ioctl is set to require synchronous updates because the file descriptor is opened with O_SYNC. This can lead to the jbd2_journal_stop() function calling jbd2_might_wait_for_commit(), potentially causing a deadlock if the EXT4_IOC_MIGRATE call races with a write(2) system call.
This problem only arises when CONFIG_PROVE_LOCKING is enabled. In this case, the jbd2_might_wait_for_commit macro locks jbd2_handle in the jbd2_journal_stop function while i_data_sem is locked. This triggers lockdep because the jbd2_journal_start function might also lock the same jbd2_handle simultaneously.
Found by Linux Verification Center (linuxtesting.org) with syzkaller.
---
v2: New addresses have been added and Ritesh Harjani has been noted as a reviewer.
v3: Description updated.
fs/ext4/migrate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/ext4/migrate.c +1 -1
@@ -663,8 +663,8 @@ int ext4_ind_migrate(struct inode *inode)if (unlikely (ret2 && !ret)) ret = ret2; errout: -ext4_journal_stop (handle);up_write (&EXT4_I (inode)->i_data_sem); +ext4_journal_stop (handle); out_unlock:ext4_writepages_up_write (inode->i_sb, alloc_ctx);return ret;
Signature · 1 line
2.43.0
Quoted 9 lines · Artem Sadovnikov
Fuzzing reports a possible deadlock in
jbd2_log_wait_commit.This issue is triggered when an
EXT4_IOC_MIGRATEioctl is set to require synchronous updates because the file descriptor is opened withO_SYNC. This can lead to thejbd2_journal_stop()function callingjbd2_might_wait_for_commit(), potentially causing a deadlock if theEXT4_IOC_MIGRATEcall races with a write(2) system call.[...]
Applied, thanks!
[1/1] ext4: fix i_data_sem unlock order in ext4_ind_migrate()
commit: cc749e61c011c255d81b192a822db650c68b313f
Best regards,
Signature · 1 line
Theodore Ts'o <[email protected]>