From 0b827cc9788e3e0b66d2ea663bf6fab347ebc44b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jean-Pierre=20Andr=C3=A9?= <jpandre@users.sourceforge.net>
Date: Tue, 30 Aug 2011 15:56:36 +0200
Subject: [PATCH] Avoid endless recursion when MFT extents are described in
 themselves

When getting extents of MFT, we must be sure they are in the MFT part which
has already been mapped, otherwise we fall into an endless recursion.

Situations have been met where extents locations are described in themselves,
as a consequence of a bug, probably unrelated to ntfs-3g.
This is a severe error which chkdsk cannot fix.
---
 libntfs-3g/inode.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

diff --git a/libntfs-3g/inode.c b/libntfs-3g/inode.c
index c68e231c..a4a01348 100644
--- a/libntfs-3g/inode.c
+++ b/libntfs-3g/inode.c
@@ -573,6 +573,9 @@ int ntfs_inode_close(ntfs_inode *ni)
 ntfs_inode *ntfs_extent_inode_open(ntfs_inode *base_ni, const MFT_REF mref)
 {
 	u64 mft_no = MREF_LE(mref);
+	VCN extent_vcn;
+	runlist_element *rl;
+	ntfs_volume *vol;
 	ntfs_inode *ni = NULL;
 	ntfs_inode **extent_nis;
 	int i;
@@ -587,6 +590,37 @@ ntfs_inode *ntfs_extent_inode_open(ntfs_inode *base_ni, const MFT_REF mref)
 			(unsigned long long)mft_no,
 			(unsigned long long)base_ni->mft_no);
 	
+	if (!base_ni->mft_no) {
+			/*
+			 * When getting extents of MFT, we must be sure
+			 * they are in the MFT part which has already
+			 * been mapped, otherwise we fall into an endless
+			 * recursion.
+			 * Situations have been met where extents locations
+			 * are described in themselves.
+			 * This is a severe error which chkdsk cannot fix.
+			 */
+		vol = base_ni->vol;
+		extent_vcn = mft_no << vol->mft_record_size_bits
+				>> vol->cluster_size_bits;
+		rl = vol->mft_na->rl;
+		if (rl) {
+			while (rl->length
+			    && ((rl->vcn + rl->length) <= extent_vcn))
+				rl++;
+		}
+		if (!rl || (rl->lcn < 0)) {
+			ntfs_log_error("MFT is corrupt, cannot read"
+				" its unmapped extent record %lld\n",
+					(long long)mft_no);
+			ntfs_log_error("Note : chkdsk cannot fix this,"
+				" try ntfsfix\n");
+			errno = EIO;
+			ni = (ntfs_inode*)NULL;
+			goto out;
+		}
+	}
+
 	/* Is the extent inode already open and attached to the base inode? */
 	if (base_ni->nr_extents > 0) {
 		extent_nis = base_ni->extent_nis;