ext4: fix indirect punch hole corruption

Commit 4f579ae7de56 (ext4: fix punch hole on files with indirect
mapping) rewrote FALLOC_FL_PUNCH_HOLE for ext4 files with indirect
mapping. However, there are bugs in several corner cases. This fixes 5
distinct bugs:

1. When there is at least one entire level of indirection between the
start and end of the punch range and the end of the punch range is the
first block of its level, we can't return early; we have to free the
intervening levels.

2. When the end is at a higher level of indirection than the start and
ext4_find_shared returns a top branch for the end, we still need to free
the rest of the shared branch it returns; we can't decrement partial2.

3. When a punch happens within one level of indirection, we need to
converge on an indirect block that contains the start and end. However,
because the branches returned from ext4_find_shared do not necessarily
start at the same level (e.g., the partial2 chain will be shallower if
the last block occurs at the beginning of an indirect group), the walk
of the two chains can end up "missing" each other and freeing a bunch of
extra blocks in the process. This mismatch can be handled by first
making sure that the chains are at the same level, then walking them
together until they converge.

4. When the punch happens within one level of indirection and
ext4_find_shared returns a top branch for the start, we must free it,
but only if the end does not occur within that branch.

5. When the punch happens within one level of indirection and
ext4_find_shared returns a top branch for the end, then we shouldn't
free the block referenced by the end of the returned chain (this mirrors
the different levels case).

Signed-off-by: Omar Sandoval <[e-mail]>

fs/ext4/indirect.c

···
 				 * to free. Everything was covered by the start
 				 * of the range.
 				 */
-				return 0;
-			} else {
-				/* Shared branch grows from an indirect block */
-				partial2--;
+				goto do_indirects;
 			}
 		} else {
 			/*
···
 	/* Punch happened within the same level (n == n2) */
 	partial = ext4_find_shared(inode, n, offsets, chain, &nr);
 	partial2 = ext4_find_shared(inode, n2, offsets2, chain2, &nr2);
-	/*
-	 * ext4_find_shared returns Indirect structure which
-	 * points to the last element which should not be
-	 * removed by truncate. But this is end of the range
-	 * in punch_hole so we need to point to the next element
-	 */
-	partial2->p++;
-	while ((partial > chain) || (partial2 > chain2)) {
-		/* We're at the same block, so we're almost finished */
-		if ((partial->bh && partial2->bh) &&
-		    (partial->bh->b_blocknr == partial2->bh->b_blocknr)) {
-			if ((partial > chain) && (partial2 > chain2)) {
-				ext4_free_branches(handle, inode, partial->bh,
-						   partial->p + 1,
-						   partial2->p,
-						   (chain+n-1) - partial);
-				BUFFER_TRACE(partial->bh, "call brelse");
-				brelse(partial->bh);
-				BUFFER_TRACE(partial2->bh, "call brelse");
-				brelse(partial2->bh);
+
+	/* Free top, but only if partial2 isn't its subtree. */
+	if (nr) {
+		int level = min(partial - chain, partial2 - chain2);
+		int i;
+		int subtree = 1;
+
+		for (i = 0; i <= level; i++) {
+			if (offsets[i] != offsets2[i]) {
+				subtree = 0;
+				break;
 			}
+		}
+
+		if (!subtree) {
+			if (partial == chain) {
+				/* Shared branch grows from the inode */
+				ext4_free_branches(handle, inode, NULL,
+						   &nr, &nr+1,
+						   (chain+n-1) - partial);
+				*partial->p = 0;
+			} else {
+				/* Shared branch grows from an indirect block */
+				BUFFER_TRACE(partial->bh, "get_write_access");
+				ext4_free_branches(handle, inode, partial->bh,
+						   partial->p,
+						   partial->p+1,
+						   (chain+n-1) - partial);
+			}
+		}
+	}
+
+	if (!nr2) {
+		/*
+		 * ext4_find_shared returns Indirect structure which
+		 * points to the last element which should not be
+		 * removed by truncate. But this is end of the range
+		 * in punch_hole so we need to point to the next element
+		 */
+		partial2->p++;
+	}
+
+	while (partial > chain || partial2 > chain2) {
+		int depth = (chain+n-1) - partial;
+		int depth2 = (chain2+n2-1) - partial2;
+
+		if (partial > chain && partial2 > chain2 &&
+		    partial->bh->b_blocknr == partial2->bh->b_blocknr) {
+			/*
+			 * We've converged on the same block. Clear the range,
+			 * then we're done.
+			 */
+			ext4_free_branches(handle, inode, partial->bh,
+					   partial->p + 1,
+					   partial2->p,
+					   (chain+n-1) - partial);
+			BUFFER_TRACE(partial->bh, "call brelse");
+			brelse(partial->bh);
+			BUFFER_TRACE(partial2->bh, "call brelse");
+			brelse(partial2->bh);
 			return 0;
 		}
+
 		/*
-		 * Clear the ends of indirect blocks on the shared branch
-		 * at the start of the range
+		 * The start and end partial branches may not be at the same
+		 * level even though the punch happened within one level. So, we
+		 * give them a chance to arrive at the same level, then walk
+		 * them in step with each other until we converge on the same
+		 * block.
 		 */
-		if (partial > chain) {
+		if (partial > chain && depth <= depth2) {
 			ext4_free_branches(handle, inode, partial->bh,
-				   partial->p + 1,
-				   (__le32 *)partial->bh->b_data+addr_per_block,
-				   (chain+n-1) - partial);
+					   partial->p + 1,
+					   (__le32 *)partial->bh->b_data+addr_per_block,
+					   (chain+n-1) - partial);
 			BUFFER_TRACE(partial->bh, "call brelse");
 			brelse(partial->bh);
 			partial--;
 		}
-		/*
-		 * Clear the ends of indirect blocks on the shared branch
-		 * at the end of the range
-		 */
-		if (partial2 > chain2) {
+		if (partial2 > chain2 && depth2 <= depth) {
 			ext4_free_branches(handle, inode, partial2->bh,
 					   (__le32 *)partial2->bh->b_data,
 					   partial2->p,
-					   (chain2+n-1) - partial2);
+					   (chain2+n2-1) - partial2);
 			BUFFER_TRACE(partial2->bh, "call brelse");
 			brelse(partial2->bh);
 			partial2--;
 		}
 	}
+	return 0;
 
 do_indirects:
 	/* Kill the remaining (whole) subtrees */