[PATCH v2] mdrestore: fix extent length overflow in v2 restore path

[Manognya Singuru <msinguru@redhat.com>]

Aisle Research reported that a crafted metadump v2 extent
length can overflow the signed int used for length,
leading to incorrect size calculations and a
potential heap buffer over-read in restore_meta_extent().

Change the length type to uint64_t and ensure all size
arithmetic in 64-bit before converting to size_t for I/O
calls.

Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>
Signed-off-by: Manognya Singuru <msinguru@redhat.com>
---
Changes since v1:
- Fixed indentation

 mdrestore/xfs_mdrestore.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/mdrestore/xfs_mdrestore.c b/mdrestore/xfs_mdrestore.c
index f10c4bef..3368a005 100644
--- a/mdrestore/xfs_mdrestore.c
+++ b/mdrestore/xfs_mdrestore.c
@@ -395,11 +395,9 @@ restore_meta_extent(
 	char		*device,
 	void		*buf,
 	uint64_t	offset,
-	int		len)
+	uint64_t	len)
 {
-	int		io_size;
-
-	io_size = min(len, MDR_IO_BUF_SIZE);
+	size_t io_size = min(len, MDR_IO_BUF_SIZE);
 
 	do {
 		if (fread(buf, io_size, 1, md_fp) != 1)
@@ -428,7 +426,7 @@ restore_v2(
 	int64_t			mb_read = 0;
 	int64_t			bytes_read;
 	uint64_t		offset;
-	int			len;
+	uint64_t		len;
 
 	block_buffer = malloc(MDR_IO_BUF_SIZE);
 	if (block_buffer == NULL)
@@ -442,7 +440,7 @@ restore_v2(
 			XME_ADDR_DATA_DEVICE)
 		fatal("Invalid superblock disk address/length\n");
 
-	len = BBTOB(be32_to_cpu(xme.xme_len));
+	len = BBTOB((uint64_t)be32_to_cpu(xme.xme_len));
 
 	if (fread(block_buffer, len, 1, md_fp) != 1)
 		fatal("error reading from metadump file\n");
@@ -503,7 +501,7 @@ restore_v2(
 			break;
 		}
 
-		len = BBTOB(be32_to_cpu(xme.xme_len));
+		len = BBTOB((uint64_t)be32_to_cpu(xme.xme_len));
 
 		restore_meta_extent(md_fp, fd, device, block_buffer, offset,
 				len);
-- 
2.54.0