[Bug 13112] receive_xattr heap overread with non null terminated name and xattr filter
samba-bugs at samba.org
samba-bugs at samba.org
Tue Oct 31 18:46:33 UTC 2017
https://bugzilla.samba.org/show_bug.cgi?id=13112
--- Comment #1 from jeriko.one at gmx.us ---
This issue can also be reached in 3.1.2, but through a different path as the
earlier version doesn't have an xattr filter. I believe the same fix for 3.1.3
would work for 3.1.2 (using read_sbuf) If I should open a separate bug for this
please let me know.
It requires the daemon to use:
fake super = yes
so that am_root == -1
the sender needs to send at least 2 xattrs, at least one without the
USER_PREFIX, and one that doesn't have a terminating NULL.
same issue of name being read in with read_buf, so it's not guaranteed to be
NULL terminated.
698 read_buf(f, name, name_len);
705 #ifdef HAVE_LINUX_XATTRS
706 /* Non-root can only save the user namespace. */
707 if (am_root <= 0 && !HAS_PREFIX(name, USER_PREFIX)) {
708 if (!am_root) {
709 free(ptr);
710 continue;
711 }
712 name -= RPRE_LEN;
713 name_len += RPRE_LEN;
714 memcpy(name, RSYNC_PREFIX, RPRE_LEN);
715 need_sort = 1;
716 }
After reading the xattrs it'll sort them if need be. need_sort is set in the
previous code snippet.
747 if (need_sort && count > 1)
748 qsort(temp_xattr.items, count, sizeof (rsync_xa),
rsync_xal_compare_names);
116 static int rsync_xal_compare_names(const void *x1, const void *x2)
117 {
118 const rsync_xa *xa1 = x1;
119 const rsync_xa *xa2 = x2;
120 return strcmp(xa1->name, xa2->name);
121 }
rsync_xal_compare_names is expecting the names to be NULL terminated, but that
isn't guaranteed.
ASan output:
=================================================================
==29001==ERROR: AddressSanitizer: heap-buffer-overflow on address
0x60200000eaa0 at pc 0x7ffff6eb3185 bp 0x7fffffff99d0 sp 0x7fffffff9178
READ of size 18 at 0x60200000eaa0 thread T0
#0 0x7ffff6eb3184
(/usr/lib/gcc/x86_64-pc-linux-gnu/5.4.0/libasan.so.2+0x47184)
#1 0x49b7e6 in rsync_xal_compare_names
/home/raj/rsync/rsync-3.1.2/xattrs.c:120
#2 0x7ffff64ecc9b (/lib64/libc.so.6+0x34c9b)
#3 0x7ffff64ecfe9 in qsort_r (/lib64/libc.so.6+0x34fe9)
#4 0x49f443 in receive_xattr /home/raj/rsync/rsync-3.1.2/xattrs.c:748
#5 0x40be5b in recv_file_entry /home/raj/rsync/rsync-3.1.2/flist.c:1126
#6 0x415b95 in recv_file_list /home/raj/rsync/rsync-3.1.2/flist.c:2494
#7 0x45478f in do_server_recv /home/raj/rsync/rsync-3.1.2/main.c:1025
#8 0x454fd9 in start_server /home/raj/rsync/rsync-3.1.2/main.c:1104
#9 0x4ae074 in rsync_module /home/raj/rsync/rsync-3.1.2/clientserver.c:1003
#10 0x4ae68a in start_daemon
/home/raj/rsync/rsync-3.1.2/clientserver.c:1094
#11 0x48e271 in start_accept_loop /home/raj/rsync/rsync-3.1.2/socket.c:618
#12 0x4aee41 in daemon_main /home/raj/rsync/rsync-3.1.2/clientserver.c:1196
#13 0x457c02 in main /home/raj/rsync/rsync-3.1.2/main.c:1621
#14 0x7ffff64d866f in __libc_start_main (/lib64/libc.so.6+0x2066f)
#15 0x404778 in _start (/home/raj/rsync/release_asan/bin/rsync+0x404778)
0x60200000eaa0 is located 0 bytes to the right of 16-byte region
[0x60200000ea90,0x60200000eaa0)
allocated by thread T0 here:
#0 0x7ffff6f04572 in malloc
(/usr/lib/gcc/x86_64-pc-linux-gnu/5.4.0/libasan.so.2+0x98572)
#1 0x44ddad in _new_array /home/raj/rsync/rsync-3.1.2/util2.c:70
#2 0x49f012 in receive_xattr /home/raj/rsync/rsync-3.1.2/xattrs.c:694
#3 0x40be5b in recv_file_entry /home/raj/rsync/rsync-3.1.2/flist.c:1126
#4 0x415b95 in recv_file_list /home/raj/rsync/rsync-3.1.2/flist.c:2494
#5 0x45478f in do_server_recv /home/raj/rsync/rsync-3.1.2/main.c:1025
#6 0x454fd9 in start_server /home/raj/rsync/rsync-3.1.2/main.c:1104
#7 0x4ae074 in rsync_module /home/raj/rsync/rsync-3.1.2/clientserver.c:1003
#8 0x4ae68a in start_daemon /home/raj/rsync/rsync-3.1.2/clientserver.c:1094
#9 0x48e271 in start_accept_loop /home/raj/rsync/rsync-3.1.2/socket.c:618
#10 0x4aee41 in daemon_main /home/raj/rsync/rsync-3.1.2/clientserver.c:1196
#11 0x457c02 in main /home/raj/rsync/rsync-3.1.2/main.c:1621
#12 0x7ffff64d866f in __libc_start_main (/lib64/libc.so.6+0x2066f)
SUMMARY: AddressSanitizer: heap-buffer-overflow ??:0 ??
Shadow bytes around the buggy address:
0x0c047fff9d00: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9d10: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9d20: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9d30: fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa fa
0x0c047fff9d40: fa fa fa fa fa fa fa fa fa fa fa fa fa fa 00 00
=>0x0c047fff9d50: fa fa 00 00[fa]fa 04 fa fa fa 00 07 fa fa fd fd
0x0c047fff9d60: fa fa 02 fa fa fa 02 fa fa fa 00 01 fa fa 00 01
0x0c047fff9d70: fa fa 06 fa fa fa fd fd fa fa fd fa fa fa 02 fa
0x0c047fff9d80: fa fa fd fa fa fa fd fa fa fa fd fa fa fa fd fa
0x0c047fff9d90: fa fa 02 fa fa fa 03 fa fa fa 00 01 fa fa fd fd
0x0c047fff9da0: fa fa 00 04 fa fa fd fd fa fa fd fd fa fa fd fa
Shadow byte legend (one shadow byte represents 8 application bytes):
Addressable: 00
Partially addressable: 01 02 03 04 05 06 07
Heap left redzone: fa
Heap right redzone: fb
Freed heap region: fd
Stack left redzone: f1
Stack mid redzone: f2
Stack right redzone: f3
Stack partial redzone: f4
Stack after return: f5
Stack use after scope: f8
Global redzone: f9
Global init order: f6
Poisoned by user: f7
Container overflow: fc
Array cookie: ac
Intra object redzone: bb
ASan internal: fe
==29001==ABORTING
--
You are receiving this mail because:
You are the QA Contact for the bug.
More information about the rsync
mailing list