I understand from a discussion with aviro on IRC that
quota deadlocks are known to exist.  For the record,
here's one which is caused by a lock ranking error with
ext3.  ext2 will have the same problem.

I am told that write_dquot() is not supposed to need to
instantiate new blocks.  Well, it is doing so here.


Thread 1:

    sys_sync
        fsync_dev
            sync_dquots
                write_dquot                                            lock_dquot()
                    ext3_file_write
                        __generic_file_write
                            ext3_prepare_write
                                block_prepare_write
                                    __block_prepare_write
                                        ext3_get_block
                                            ext3_get_block_handle
                                                ext3_alloc_branch
                                                    ext3_alloc_block   lock_super()
                                                        __down_failed

Thread 2:

    ext3_file_write
        __generic_file_write
            ext3_prepare_write
                block_prepare_write
                    __block_prepare_write
                        ext3_get_block
                            ext3_get_block_handle
                                ext3_alloc_branch
                                    ext3_alloc_block
                                        ext3_new_block            lock_super()
                                            dquot_alloc_block
                                                dqduplicate       wait_on_dquot()
                                                    __wait_on_dquot
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [EMAIL PROTECTED]

Reply via email to