Messages in this thread | | | From | "usvyatsky, ilya" <> | Subject | General dcache question | Date | Mon, 1 Nov 2004 15:42:58 -0500 |
| |
Hi folks, Sorry if this question does not really belong to this list, but it seems to me that someone here can simply know the answer. I am looking at Linux dcache mechanism trying to retreive a name (actually, full path) of an active inode. It seems that each active inode has a linked list of dentry's (i_dentry) that holds a (partial) list of its names (links in fact), and each dentry has d_name and d_parent. Moshe Bar in his "Linux File Systems" book states that for every active inode i_dentry would always contain at least one active dentry. Moreover, such d_entry would have a valid d_parent field pointing at active dcache entry containing a name of the parent directory. He also says that only leaves with d_count equal to 0 are on the dcache LRU list. Now, I wonder if the following assumption is true: For each active inode, there always be enough dcache entries (linked through d_parent pointer) to restore the full path of the inode from the root. I also would appreciate it very much if someone familiar with 2.6 kernel would confirm that this assumption still holds there. Thanks a lot, Ilya Usvyatsky.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |