Skip to contentSkip to conversation

fs: Set file_handle::handle_bytes before referencing file_handle::f_handle

PATCH In review Kees Cook 3 participants 4 messages Apr 4, 2024 – Apr 5, 2024

Subsystems (4)
lore
Share & tools
lore

Kees Cook

author
#0 lore raw

PATCH v2 1/1 fs: Set file_handle::handle_bytes before referencing file_handle::f_handle

Since __counted_by(handle_bytes) was added to struct file_handle, we need to explicitly set it in the one place it wasn't yet happening prior to accessing the flex array "f_handle". For robustness also check for a negative value for handle_bytes, which is possible for an "int", but nothing appears to set.

Fixes 1b43c4629756 ("fs: Annotate struct file_handle with __counted_by() and use struct_size()")

---
Cc: Jan Kara <[email protected]>
Cc: Chuck Lever <[email protected]>
Cc: Gustavo A. R. Silva <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Jeff Layton <[email protected]>
Cc: Amir Goldstein <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
v2: more bounds checking, add comments, dropped reviews since logic changed
v1: https://lore.kernel.org/all/[email protected]/

---
 fs/fhandle.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
fs/fhandle.c +9 -2
index 8a7f86c2139a..854f866eaad2 100644
@@ -40,6 +40,11 @@ static long do_sys_name_to_handle(const struct path *path,
 			 GFP_KERNEL);
 	if (!handle)
 		return -ENOMEM;
+	/*
+	 * Since handle->f_handle is about to be written, make sure the
+	 * associated __counted_by(handle_bytes) variable is correct.
+	 */
+	handle->handle_bytes = f_handle.handle_bytes;
 
 	/* convert handle size to multiple of sizeof(u32) */
 	handle_dwords = f_handle.handle_bytes >> 2;
@@ -51,8 +56,8 @@ static long do_sys_name_to_handle(const struct path *path,
 	handle->handle_type = retval;
 	/* convert handle size to bytes */
 	handle_bytes = handle_dwords * sizeof(u32);
-	handle->handle_bytes = handle_bytes;
-	if ((handle->handle_bytes > f_handle.handle_bytes) ||
+	/* check if handle_bytes would have exceeded the allocation */
+	if ((handle_bytes < 0) || (handle_bytes > f_handle.handle_bytes) ||
 	    (retval == FILEID_INVALID) || (retval < 0)) {
 		/* As per old exportfs_encode_fh documentation
 		 * we could return ENOSPC to indicate overflow
@@ -68,6 +73,8 @@ static long do_sys_name_to_handle(const struct path *path,
 		handle_bytes = 0;
 	} else
 		retval = 0;
+	/* the "valid" number of bytes may fewer than originally allocated */
+	handle->handle_bytes = handle_bytes;
 	/* copy the mount id */
 	if (put_user(real_mount(path->mnt)->mnt_id, mnt_id) ||
 	    copy_to_user(ufh, handle,
Signature · 1 line
2.34.1
#1 lore raw
Quoted 5 lines · Kees Cook

Since __counted_by(handle_bytes) was added to struct file_handle, we need to explicitly set it in the one place it wasn't yet happening prior to accessing the flex array "f_handle". For robustness also check for a negative value for handle_bytes, which is possible for an "int", but nothing appears to set.

Why not change handle_bytes from an int to a u32?

Also, what a grotty function.

        handle_dwords = f_handle.handle_bytes >> 2;
...
        handle_bytes = handle_dwords * sizeof(u32);
Quoted 59 lines
Fixes 1b43c4629756 ("fs: Annotate struct file_handle with __counted_by() and use struct_size()")

---
Cc: Jan Kara <[email protected]>
Cc: Chuck Lever <[email protected]>
Cc: Gustavo A. R. Silva <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Jeff Layton <[email protected]>
Cc: Amir Goldstein <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
v2: more bounds checking, add comments, dropped reviews since logic changed
v1: https://lore.kernel.org/all/[email protected]/

---
 fs/fhandle.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
fs/fhandle.c +9 -2
index 8a7f86c2139a..854f866eaad2 100644
@@ -40,6 +40,11 @@ static long do_sys_name_to_handle(const struct path *path,
 			 GFP_KERNEL);
 	if (!handle)
 		return -ENOMEM;
+	/*
+	 * Since handle->f_handle is about to be written, make sure the
+	 * associated __counted_by(handle_bytes) variable is correct.
+	 */
+	handle->handle_bytes = f_handle.handle_bytes;
 
 	/* convert handle size to multiple of sizeof(u32) */
 	handle_dwords = f_handle.handle_bytes >> 2;
@@ -51,8 +56,8 @@ static long do_sys_name_to_handle(const struct path *path,
 	handle->handle_type = retval;
 	/* convert handle size to bytes */
 	handle_bytes = handle_dwords * sizeof(u32);
-	handle->handle_bytes = handle_bytes;
-	if ((handle->handle_bytes > f_handle.handle_bytes) ||
+	/* check if handle_bytes would have exceeded the allocation */
+	if ((handle_bytes < 0) || (handle_bytes > f_handle.handle_bytes) ||
 	    (retval == FILEID_INVALID) || (retval < 0)) {
 		/* As per old exportfs_encode_fh documentation
 		 * we could return ENOSPC to indicate overflow
@@ -68,6 +73,8 @@ static long do_sys_name_to_handle(const struct path *path,
 		handle_bytes = 0;
 	} else
 		retval = 0;
+	/* the "valid" number of bytes may fewer than originally allocated */
+	handle->handle_bytes = handle_bytes;
 	/* copy the mount id */
 	if (put_user(real_mount(path->mnt)->mnt_id, mnt_id) ||
 	    copy_to_user(ufh, handle,
Signature · 1 line
2.34.1
reviewer
#2 lore raw
In reply to Kees Cook
Quoted 42 lines · Kees Cook

Since __counted_by(handle_bytes) was added to struct file_handle, we need to explicitly set it in the one place it wasn't yet happening prior to accessing the flex array "f_handle". For robustness also check for a negative value for handle_bytes, which is possible for an "int", but nothing appears to set.

Fixes 1b43c4629756 ("fs: Annotate struct file_handle with __counted_by() and use struct_size()")

---
Cc: Jan Kara <[email protected]>
Cc: Chuck Lever <[email protected]>
Cc: Gustavo A. R. Silva <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Jeff Layton <[email protected]>
Cc: Amir Goldstein <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
v2: more bounds checking, add comments, dropped reviews since logic changed
v1: https://lore.kernel.org/all/[email protected]/

---
 fs/fhandle.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
fs/fhandle.c +5 -0
index 8a7f86c2139a..854f866eaad2 100644
@@ -40,6 +40,11 @@ static long do_sys_name_to_handle(const struct path *path,
 			 GFP_KERNEL);
 	if (!handle)
 		return -ENOMEM;
+	/*
+	 * Since handle->f_handle is about to be written, make sure the
+	 * associated __counted_by(handle_bytes) variable is correct.
+	 */
+	handle->handle_bytes = f_handle.handle_bytes;
 
 	/* convert handle size to multiple of sizeof(u32) */
 	handle_dwords = f_handle.handle_bytes >> 2;

fs/fhandle.c static long do_sys_name_to_handle(const struct path *path, · line 60

Quoted 7 lines
fs/fhandle.c +2 -2
@@ -51,8 +56,8 @@ static long do_sys_name_to_handle(const struct path *path,
 	handle->handle_type = retval;
 	/* convert handle size to bytes */
 	handle_bytes = handle_dwords * sizeof(u32);
-	handle->handle_bytes = handle_bytes;
-	if ((handle->handle_bytes > f_handle.handle_bytes) ||
+	/* check if handle_bytes would have exceeded the allocation */
+	if ((handle_bytes < 0) || (handle_bytes > f_handle.handle_bytes) ||

This is broken. Let me explain: Userspace passes in struct file_handle (ufh) and says how many bytes it has reserved for the variable length contents in ufh->handle_bytes. We call exportfs_encode_fh() to create the file handle. If it fits into the provided space, the function returns in handle_dwords how many uints it has actually stored. If the handle didn't fit, handle_dword contains number of uints we'd need in the variable length part to be able to fit the handle in.

Now your patch destroys this behavior by storing 0 to handle_bytes in case the handle didn't fit *before* the returned value is actually stored to a struct copied to userspace.

Also the handle_bytes < 0 check is IMHO pointless and confusing. If some filesystem is returning bogus size of the file handle, we have a big problem anyway so I don't think checking like this brings much. If you want to add some paranoia what would make some sense is: Make handle_bytes uint and do:

	if (WARN_ON_ONCE(handle_bytes > MAX_HANDLE_SZ)) {
		handle_bytes = 0;
		retval = -EINVAL;
	}

But as a separate patch please, because it is unrelated to this __counted_by fixup.

Honza

Quoted 15 lines
fs/fhandle.c +2 -0
 	    (retval == FILEID_INVALID) || (retval < 0)) {
 		/* As per old exportfs_encode_fh documentation
 		 * we could return ENOSPC to indicate overflow
@@ -68,6 +73,8 @@ static long do_sys_name_to_handle(const struct path *path,
 		handle_bytes = 0;
 	} else
 		retval = 0;
+	/* the "valid" number of bytes may fewer than originally allocated */
+	handle->handle_bytes = handle_bytes;
 	/* copy the mount id */
 	if (put_user(real_mount(path->mnt)->mnt_id, mnt_id) ||
 	    copy_to_user(ufh, handle,
Signature · 1 line
2.34.1
Signature · 2 lines
Jan Kara <[email protected]>
SUSE Labs, CR
author
#3 lore raw
Quoted 59 lines · Jan Kara
Quoted 42 lines · Kees Cook

Since __counted_by(handle_bytes) was added to struct file_handle, we need to explicitly set it in the one place it wasn't yet happening prior to accessing the flex array "f_handle". For robustness also check for a negative value for handle_bytes, which is possible for an "int", but nothing appears to set.

Fixes 1b43c4629756 ("fs: Annotate struct file_handle with __counted_by() and use struct_size()")

---
Cc: Jan Kara <[email protected]>
Cc: Chuck Lever <[email protected]>
Cc: Gustavo A. R. Silva <[email protected]>
Cc: Christian Brauner <[email protected]>
Cc: Alexander Viro <[email protected]>
Cc: Jeff Layton <[email protected]>
Cc: Amir Goldstein <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
v2: more bounds checking, add comments, dropped reviews since logic changed
v1: https://lore.kernel.org/all/[email protected]/

---
 fs/fhandle.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
fs/fhandle.c +5 -0
index 8a7f86c2139a..854f866eaad2 100644
@@ -40,6 +40,11 @@ static long do_sys_name_to_handle(const struct path *path,
 			 GFP_KERNEL);
 	if (!handle)
 		return -ENOMEM;
+	/*
+	 * Since handle->f_handle is about to be written, make sure the
+	 * associated __counted_by(handle_bytes) variable is correct.
+	 */
+	handle->handle_bytes = f_handle.handle_bytes;
 
 	/* convert handle size to multiple of sizeof(u32) */
 	handle_dwords = f_handle.handle_bytes >> 2;

fs/fhandle.c static long do_sys_name_to_handle(const struct path *path, · line 60

Quoted 7 lines
fs/fhandle.c +2 -2
@@ -51,8 +56,8 @@ static long do_sys_name_to_handle(const struct path *path,
 	handle->handle_type = retval;
 	/* convert handle size to bytes */
 	handle_bytes = handle_dwords * sizeof(u32);
-	handle->handle_bytes = handle_bytes;
-	if ((handle->handle_bytes > f_handle.handle_bytes) ||
+	/* check if handle_bytes would have exceeded the allocation */
+	if ((handle_bytes < 0) || (handle_bytes > f_handle.handle_bytes) ||

This is broken. Let me explain: Userspace passes in struct file_handle (ufh) and says how many bytes it has reserved for the variable length contents in ufh->handle_bytes. We call exportfs_encode_fh() to create the file handle. If it fits into the provided space, the function returns in handle_dwords how many uints it has actually stored. If the handle didn't fit, handle_dword contains number of uints we'd need in the variable length part to be able to fit the handle in.

Quoted 3 lines

Now your patch destroys this behavior by storing 0 to handle_bytes in case the handle didn't fit *before* the returned value is actually stored to a struct copied to userspace.

Ah yes, I see now how they're used for separate things. I will respin again and possibly fix up the "int" to "u32" as suggested by willy.

Signature · 1 line
Kees Cook