tar: postpone creation of symlinks with "suspicious" targets. Closes 8411

function                                             old     new   delta
data_extract_all                                     968    1038     +70
tar_main                                             952     986     +34
scan_tree                                            258     262      +4
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 3/0 up/down: 108/0)             Total: 108 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

6 files changed, 130 insertions, 36 deletions

diff --git a/archival/libarchive/data_extract_all.c b/archival/libarchive/data_extract_all.c
index 1830ffb8d..1ce927c2f 100644
--- a/archival/libarchive/data_extract_all.c
+++ b/archival/libarchive/data_extract_all.c
@@ -128,10 +128,11 @@ void FAST_FUNC data_extract_all(archive_handle_t *archive_handle)
                 res = link(hard_link, dst_name);
                 if (res != 0 && !(archive_handle->ah_flags & ARCHIVE_EXTRACT_QUIET)) {
                         /* shared message */
-                        bb_perror_msg("can't create %slink "
-                                        "%s to %s", "hard",
+                        bb_perror_msg("can't create %slink '%s' to '%s'",
+                                        "hard",
                                         dst_name,
-                                        hard_link);
+                                        hard_link
+                        );
                 }
                 /* Hardlinks have no separate mode/ownership, skip chown/chmod */
                 goto ret;
@@ -178,15 +179,44 @@ void FAST_FUNC data_extract_all(archive_handle_t *archive_handle)
         case S_IFLNK:
                 /* Symlink */
 //TODO: what if file_header->link_target == NULL (say, corrupted tarball?)
+
+                /* To avoid a directory traversal attack via symlinks,
+                 * for certain link targets postpone creation of symlinks.
+                 *
+                 * For example, consider a .tar created via:
+                 *  $ tar cvf bug.tar anything.txt
+                 *  $ ln -s /tmp symlink
+                 *  $ tar --append -f bug.tar symlink
+                 *  $ rm symlink
+                 *  $ mkdir symlink
+                 *  $ tar --append -f bug.tar symlink/evil.py
+                 *
+                 * This will result in an archive that contains:
+                 *  $ tar --list -f bug.tar
+                 *  anything.txt
+                 *  symlink [-> /tmp]
+                 *  symlink/evil.py
+                 *
+                 * Untarring bug.tar would otherwise place evil.py in '/tmp'.
+                 */
+                if (file_header->link_target[0] == '/'
+                 || strstr(file_header->link_target, "..")
+                ) {
+                        llist_add_to(&archive_handle->symlink_placeholders,
+                                xasprintf("%s%c%s", file_header->name, '\0', file_header->link_target)
+                        );
+                        break;
+                }
                 res = symlink(file_header->link_target, dst_name);
                 if (res != 0
                  && !(archive_handle->ah_flags & ARCHIVE_EXTRACT_QUIET)
                 ) {
                         /* shared message */
-                        bb_perror_msg("can't create %slink "
-                                "%s to %s", "sym",
+                        bb_perror_msg("can't create %slink '%s' to '%s'",
+                                "sym",
                                 dst_name,
-                                file_header->link_target);
+                                file_header->link_target
+                        );
                 }
                 break;
         case S_IFSOCK:
diff --git a/archival/tar.c b/archival/tar.c
index 0fc574dfd..280ded4e1 100644
--- a/archival/tar.c
+++ b/archival/tar.c
@@ -22,24 +22,6 @@
  *
  * Licensed under GPLv2 or later, see file LICENSE in this source tree.
  */
-/* TODO: security with -C DESTDIR option can be enhanced.
- * Consider tar file created via:
- * $ tar cvf bug.tar anything.txt
- * $ ln -s /tmp symlink
- * $ tar --append -f bug.tar symlink
- * $ rm symlink
- * $ mkdir symlink
- * $ tar --append -f bug.tar symlink/evil.py
- *
- * This will result in an archive which contains:
- * $ tar --list -f bug.tar
- * anything.txt
- * symlink
- * symlink/evil.py
- *
- * Untarring it puts evil.py in '/tmp' even if the -C DESTDIR is given.
- * This doesn't feel right, and IIRC GNU tar doesn't do that.
- */
 
 //config:config TAR
 //config:       bool "tar (40 kb)"
@@ -296,6 +278,23 @@ static void chksum_and_xwrite(int fd, struct tar_header_t* hp)
         xwrite(fd, hp, sizeof(*hp));
 }
 
+static void replace_symlink_placeholders(llist_t *list)
+{
+        while (list) {
+                char *target;
+
+                target = list->data + strlen(list->data) + 1;
+                if (symlink(target, list->data)) {
+                        /* shared message */
+                        bb_error_msg_and_die("can't create %slink '%s' to '%s'",
+                                "sym",
+                                list->data, target
+                        );
+                }
+                list = list->link;
+        }
+}
+
 #if ENABLE_FEATURE_TAR_GNU_EXTENSIONS
 static void writeLongname(int fd, int type, const char *name, int dir)
 {
@@ -1252,6 +1251,8 @@ int tar_main(int argc UNUSED_PARAM, char **argv)
         while (get_header_tar(tar_handle) == EXIT_SUCCESS)
                 bb_got_signal = EXIT_SUCCESS; /* saw at least one header, good */
 
+        replace_symlink_placeholders(tar_handle->symlink_placeholders);
+
         /* Check that every file that should have been extracted was */
         while (tar_handle->accept) {
                 if (!find_list_entry(tar_handle->reject, tar_handle->accept->data)
diff --git a/archival/tar_symlink_attack b/archival/tar_symlink_attack
new file mode 100755
index 000000000..35455f200
--- /dev/null
+++ b/archival/tar_symlink_attack
@@ -0,0 +1,16 @@
+#!/bin/sh
+# Makes "symlink attack" tarball (needs GNU tar for --append)
+
+true >anything.txt
+tar cvf tar_symlink_attack.tar anything.txt
+rm anything.txt
+
+ln -s /tmp symlink
+tar --append -f tar_symlink_attack.tar symlink
+rm symlink
+
+mkdir symlink
+echo BUG >symlink/bb_test_evilfile
+tar --append -f tar_symlink_attack.tar symlink/bb_test_evilfile
+rm symlink/bb_test_evilfile
+rmdir symlink
diff --git a/coreutils/link.c b/coreutils/link.c
index 56832fdf6..6e20dafe3 100644
--- a/coreutils/link.c
+++ b/coreutils/link.c
@@ -33,7 +33,7 @@ int link_main(int argc UNUSED_PARAM, char **argv)
         if (link(argv[0], argv[1]) != 0) {
                 /* shared message */
                 bb_perror_msg_and_die("can't create %slink "
-                                        "%s to %s", "hard",
+                                        "'%s' to '%s'", "hard",
                                         argv[1], argv[0]
                 );
         }
diff --git a/include/bb_archive.h b/include/bb_archive.h
index 2b9c5f04c..d3762415f 100644
--- a/include/bb_archive.h
+++ b/include/bb_archive.h
@@ -64,6 +64,9 @@ typedef struct archive_handle_t {
         /* Currently processed file's header */
         file_header_t *file_header;
 
+        /* List of symlink placeholders */
+        llist_t *symlink_placeholders;
+
         /* Process the header component, e.g. tar -t */
         void FAST_FUNC (*action_header)(const file_header_t *);
 
@@ -188,6 +191,7 @@ char get_header_ar(archive_handle_t *archive_handle) FAST_FUNC;
 char get_header_cpio(archive_handle_t *archive_handle) FAST_FUNC;
 char get_header_tar(archive_handle_t *archive_handle) FAST_FUNC;
 char get_header_tar_gz(archive_handle_t *archive_handle) FAST_FUNC;
+char get_header_tar_xz(archive_handle_t *archive_handle) FAST_FUNC;
 char get_header_tar_bz2(archive_handle_t *archive_handle) FAST_FUNC;
 char get_header_tar_lzma(archive_handle_t *archive_handle) FAST_FUNC;
 char get_header_tar_xz(archive_handle_t *archive_handle) FAST_FUNC;
diff --git a/testsuite/tar.tests b/testsuite/tar.tests
index 9f7ce1587..1675b07b1 100755
--- a/testsuite/tar.tests
+++ b/testsuite/tar.tests
@@ -10,9 +10,6 @@ unset LC_COLLATE
 unset LC_ALL
 umask 022
 
-rm -rf tar.tempdir 2>/dev/null
-mkdir tar.tempdir && cd tar.tempdir || exit 1
-
 # testing "test name" "script" "expected result" "file input" "stdin"
 
 testing "Empty file is not a tarball" '\
@@ -53,6 +50,7 @@ dd if=/dev/zero bs=512 count=20 2>/dev/null | tar xvf - 2>&1; echo $?
 "" ""
 SKIP=
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 # "tar cf test.tar input input_dir/ input_hard1 input_hard2 input_hard1 input_dir/ input":
 # GNU tar 1.26 records as hardlinks:
 #  input_hard2 -> input_hard1
@@ -64,7 +62,6 @@ SKIP=
 # We also don't use "hrw-r--r--" notation for hardlinks in "tar tv" listing.
 optional FEATURE_TAR_CREATE FEATURE_LS_SORTFILES
 testing "tar hardlinks and repeated files" '\
-rm -rf input_* test.tar 2>/dev/null
 >input_hard1
 ln input_hard1 input_hard2
 mkdir input_dir
@@ -95,10 +92,11 @@ drwxr-xr-x input_dir
 " \
 "" ""
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 optional FEATURE_TAR_CREATE FEATURE_LS_SORTFILES
 testing "tar hardlinks mode" '\
-rm -rf input_* test.tar 2>/dev/null
 >input_hard1
 chmod 741 input_hard1
 ln input_hard1 input_hard2
@@ -128,10 +126,11 @@ Ok: 0
 " \
 "" ""
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 optional FEATURE_TAR_CREATE FEATURE_LS_SORTFILES
 testing "tar symlinks mode" '\
-rm -rf input_* test.tar 2>/dev/null
 >input_file
 chmod 741 input_file
 ln -s input_file input_soft
@@ -159,10 +158,11 @@ lrwxrwxrwx input_file
 " \
 "" ""
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 optional FEATURE_TAR_CREATE FEATURE_TAR_LONG_OPTIONS
 testing "tar --overwrite" "\
-rm -rf input_* test.tar 2>/dev/null
 ln input input_hard
 tar cf test.tar input_hard
 echo WRONG >input
@@ -174,12 +174,13 @@ Ok
 " \
 "Ok\n" ""
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 test x"$SKIP_KNOWN_BUGS" = x"" && {
 # Needs to be run under non-root for meaningful test
 optional FEATURE_TAR_CREATE
 testing "tar writing into read-only dir" '\
-rm -rf input_* test.tar 2>/dev/null
 mkdir input_dir
 >input_dir/input_file
 chmod 550 input_dir
@@ -201,7 +202,9 @@ dr-xr-x--- input_dir
 "" ""
 SKIP=
 }
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 # Had a bug where on extract autodetect first "switched off" -z
 # and then failed to recognize .tgz extension
 optional FEATURE_TAR_CREATE FEATURE_SEAMLESS_GZ GUNZIP
@@ -217,7 +220,9 @@ Ok
 " \
 "" ""
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 # Do we detect XZ-compressed data (even w/o .tar.xz or txz extension)?
 # (the uuencoded hello_world.txz contains one empty file named "hello_world")
 optional UUDECODE FEATURE_TAR_AUTODETECT FEATURE_SEAMLESS_XZ
@@ -236,7 +241,9 @@ AAAEWVo=
 ====
 "
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 # On extract, everything up to and including last ".." component is stripped
 optional FEATURE_TAR_CREATE
 testing "tar strips /../ on extract" "\
@@ -255,7 +262,9 @@ Ok
 " \
 "" ""
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 # attack.tar.bz2 has symlink pointing to a system file
 # followed by a regular file with the same name
 # containing "root::0:0::/root:/bin/sh":
@@ -270,6 +279,7 @@ optional UUDECODE FEATURE_TAR_AUTODETECT FEATURE_SEAMLESS_BZ2
 testing "tar does not extract into symlinks" "\
 >>/tmp/passwd && uudecode -o input && tar xf input 2>&1 && rm passwd; cat /tmp/passwd; echo \$?
 " "\
+tar: can't create symlink 'passwd' to '/tmp/passwd'
 0
 " \
 "" "\
@@ -281,12 +291,15 @@ l4/V8LDoe90yiWJhOJvIypgEfxdyRThQkBVn/bI=
 ====
 "
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
+
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 # And same with -k
 optional UUDECODE FEATURE_TAR_AUTODETECT FEATURE_SEAMLESS_BZ2
 testing "tar -k does not extract into symlinks" "\
 >>/tmp/passwd && uudecode -o input && tar xf input -k 2>&1 && rm passwd; cat /tmp/passwd; echo \$?
 " "\
-tar: can't open 'passwd': File exists
+tar: can't create symlink 'passwd' to '/tmp/passwd'
 0
 " \
 "" "\
@@ -298,7 +311,9 @@ l4/V8LDoe90yiWJhOJvIypgEfxdyRThQkBVn/bI=
 ====
 "
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
+mkdir tar.tempdir && cd tar.tempdir || exit 1
 optional UNICODE_SUPPORT FEATURE_TAR_GNU_EXTENSIONS FEATURE_SEAMLESS_BZ2 FEATURE_TAR_AUTODETECT
 testing "Pax-encoded UTF8 names and symlinks" '\
 tar xvf ../tar.utf8.tar.bz2 2>&1; echo $?
@@ -318,8 +333,36 @@ etc/ssl/certs/f80cc7f6.0 -> EBG_Elektronik_Sertifika_Hizmet_Sağlayıcısı.pem
 " \
 "" ""
 SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
-
-cd .. && rm -rf tar.tempdir || exit 1
+mkdir tar.tempdir && cd tar.tempdir || exit 1
+optional FEATURE_SEAMLESS_BZ2 FEATURE_TAR_AUTODETECT
+testing "Symlink attack: create symlink and then write through it" '\
+exec 2>&1
+uudecode -o input && tar xvf input; echo $?
+ls /tmp/bb_test_evilfile
+ls bb_test_evilfile
+ls symlink/bb_test_evilfile
+' "\
+anything.txt
+symlink
+symlink/bb_test_evilfile
+tar: can't create symlink 'symlink' to '/tmp'
+1
+ls: /tmp/bb_test_evilfile: No such file or directory
+ls: bb_test_evilfile: No such file or directory
+symlink/bb_test_evilfile
+" \
+"" "\
+begin-base64 644 tar_symlink_attack.tar.bz2
+QlpoOTFBWSZTWZgs7bQAALT/hMmQAFBAAf+AEMAGJPPv32AAAIAIMAC5thlR
+omAjAmCMADQT1BqNE0AEwAAjAEwElTKeo9NTR6h6gaeoA0DQNLVdwZZ5iNTk
+AQwCAV6S00QFJYhrlfFkVCEDEGtgNVqYrI0uK3ggnt30gqk4e1TTQm5QIAKa
+SJqzRGSFLMmOloHSAcvLiFxxRiQtQZF+qPxbo173ZDISOAoNoPN4PQPhBhKS
+n8fYaKlioCTzL2oXYczyUUIP4u5IpwoSEwWdtoA=
+====
+"
+SKIP=
+cd .. || exit 1; rm -rf tar.tempdir 2>/dev/null
 
 exit $FAILCOUNT