1 /*-
2 * Copyright (c) 2003-2023 Tim Kientzle
3 * Copyright (c) 2008 Anselm Strauss
4 * All rights reserved.
5 *
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
8 * are met:
9 * 1. Redistributions of source code must retain the above copyright
10 * notice, this list of conditions and the following disclaimer.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
14 *
15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR
16 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
17 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
18 * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT,
19 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
20 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
21 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
22 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
23 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
24 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
25 */
26
27 #include "test.h"
28
29 /*
30 * Detailed byte-for-byte verification of the format of a zip archive
31 * written in streaming mode with Zip64 extensions enabled.
32 */
33
34 static unsigned long
bitcrc32(unsigned long c,void * _p,size_t s)35 bitcrc32(unsigned long c, void *_p, size_t s)
36 {
37 /* This is a drop-in replacement for crc32() from zlib.
38 * Libarchive should be able to correctly generate
39 * uncompressed zip archives (including correct CRCs) even
40 * when zlib is unavailable, and this function helps us verify
41 * that. Yes, this is very, very slow and unsuitable for
42 * production use, but it's correct, compact, and works well
43 * enough for this particular usage. Libarchive internally
44 * uses a much more efficient implementation. */
45 const unsigned char *p = _p;
46 int bitctr;
47
48 if (p == NULL)
49 return (0);
50
51 for (; s > 0; --s) {
52 c ^= *p++;
53 for (bitctr = 8; bitctr > 0; --bitctr) {
54 if (c & 1) c = (c >> 1);
55 else c = (c >> 1) ^ 0xedb88320;
56 c ^= 0x80000000;
57 }
58 }
59 return (c);
60 }
61
62 /* Quick and dirty: Read 2-byte and 4-byte integers from Zip file. */
i2(const unsigned char * p)63 static unsigned i2(const unsigned char *p) { return ((p[0] & 0xff) | ((p[1] & 0xff) << 8)); }
i4(const unsigned char * p)64 static unsigned i4(const unsigned char *p) { return (i2(p) | (i2(p + 2) << 16)); }
65 /* We're only working with small values here; ignore the 4 high bytes. */
i8(const unsigned char * p)66 static unsigned i8(const unsigned char *p) { return (i4(p)); }
67
DEFINE_TEST(test_write_format_zip64_stream)68 DEFINE_TEST(test_write_format_zip64_stream)
69 {
70 struct archive *a;
71 struct archive_entry *ae;
72 size_t used, buffsize = 1000000;
73 unsigned long crc;
74 unsigned long compressed_size = 0;
75 int file_perm = 00644;
76 int zip_version = 45;
77 int zip_compression = 8;
78 short file_uid = 10, file_gid = 20;
79 unsigned char *buff, *buffend, *p;
80 unsigned char *central_header, *local_header, *eocd, *eocd_record;
81 unsigned char *extension_start, *extension_end;
82 unsigned char *data_start, *data_end;
83 char file_data[] = {'1', '2', '3', '4', '5', '6', '7', '8'};
84 const char *file_name = "file";
85
86 #ifndef HAVE_ZLIB_H
87 zip_compression = 0;
88 #endif
89
90 buff = malloc(buffsize);
91
92 /* Create a new archive in memory. */
93 assert((a = archive_write_new()) != NULL);
94 assertEqualIntA(a, ARCHIVE_OK, archive_write_set_format_zip(a));
95 assertEqualIntA(a, ARCHIVE_OK,
96 archive_write_set_options(a, "zip:zip64"));
97 assertEqualIntA(a, ARCHIVE_OK,
98 archive_write_set_options(a, "zip:experimental"));
99 assertEqualIntA(a, ARCHIVE_OK,
100 archive_write_open_memory(a, buff, buffsize, &used));
101
102 assert((ae = archive_entry_new()) != NULL);
103 archive_entry_copy_pathname(ae, file_name);
104 archive_entry_set_mode(ae, AE_IFREG | file_perm);
105 archive_entry_set_uid(ae, file_uid);
106 archive_entry_set_gid(ae, file_gid);
107 archive_entry_set_mtime(ae, 0, 0);
108 assertEqualInt(0, archive_write_header(a, ae));
109 archive_entry_free(ae);
110 assertEqualInt(8, archive_write_data(a, file_data, sizeof(file_data)));
111 assertEqualIntA(a, ARCHIVE_OK, archive_write_close(a));
112 assertEqualInt(ARCHIVE_OK, archive_write_free(a));
113 buffend = buff + used;
114 dumpfile("constructed.zip", buff, used);
115
116 /* Verify "End of Central Directory" record. */
117 /* Get address of end-of-central-directory record. */
118 eocd_record = p = buffend - 22; /* Assumes there is no zip comment field. */
119 failure("End-of-central-directory begins with PK\\005\\006 signature");
120 assertEqualMem(p, "PK\005\006", 4);
121 failure("This must be disk 0");
122 assertEqualInt(i2(p + 4), 0);
123 failure("Central dir must start on disk 0");
124 assertEqualInt(i2(p + 6), 0);
125 failure("All central dir entries are on this disk");
126 assertEqualInt(i2(p + 8), i2(p + 10));
127 eocd = buff + i4(p + 12) + i4(p + 16);
128 failure("no zip comment");
129 assertEqualInt(i2(p + 20), 0);
130
131 /* Get address of first entry in central directory. */
132 central_header = p = buff + i4(buffend - 6);
133 failure("Central file record at offset %d should begin with"
134 " PK\\001\\002 signature",
135 i4(buffend - 10));
136
137 /* Verify file entry in central directory. */
138 assertEqualMem(p, "PK\001\002", 4); /* Signature */
139 assertEqualInt(i2(p + 4), 3 * 256 + zip_version); /* Version made by */
140 assertEqualInt(i2(p + 6), zip_version); /* Version needed to extract */
141 assertEqualInt(i2(p + 8), 8); /* Flags */
142 assertEqualInt(i2(p + 10), zip_compression); /* Compression method */
143 assertEqualInt(i2(p + 12), 0); /* File time */
144 assertEqualInt(i2(p + 14), 33); /* File date */
145 crc = bitcrc32(0, file_data, sizeof(file_data));
146 assertEqualInt(i4(p + 16), crc); /* CRC-32 */
147 compressed_size = i4(p + 20); /* Compressed size */
148 assertEqualInt(i4(p + 24), sizeof(file_data)); /* Uncompressed size */
149 assertEqualInt(i2(p + 28), strlen(file_name)); /* Pathname length */
150 /* assertEqualInt(i2(p + 30), 28); */ /* Extra field length: See below */
151 assertEqualInt(i2(p + 32), 0); /* File comment length */
152 assertEqualInt(i2(p + 34), 0); /* Disk number start */
153 assertEqualInt(i2(p + 36), 0); /* Internal file attrs */
154 assertEqualInt(i4(p + 38) >> 16 & 01777, file_perm); /* External file attrs */
155 assertEqualInt(i4(p + 42), 0); /* Offset of local header */
156 assertEqualMem(p + 46, file_name, strlen(file_name)); /* Pathname */
157 p = extension_start = central_header + 46 + strlen(file_name);
158 extension_end = extension_start + i2(central_header + 30);
159
160 assertEqualInt(i2(p), 0x7875); /* 'ux' extension header */
161 assertEqualInt(i2(p + 2), 11); /* 'ux' size */
162 /* TODO: verify 'ux' contents */
163 p += 4 + i2(p + 2);
164
165 assertEqualInt(i2(p), 0x5455); /* 'UT' extension header */
166 assertEqualInt(i2(p + 2), 5); /* 'UT' size */
167 assertEqualInt(p[4], 1); /* 'UT' flags */
168 assertEqualInt(i4(p + 5), 0); /* 'UT' mtime */
169 p += 4 + i2(p + 2);
170
171 /* Note: We don't expect to see zip64 extension in the central
172 * directory, since the writer knows the actual full size by
173 * the time it is ready to write the central directory and has
174 * no reason to insert it then. Info-Zip seems to do the same
175 * thing. */
176
177 /* Just in case: Report any extra extensions. */
178 while (p < extension_end) {
179 failure("Unexpected extension 0x%04X", i2(p));
180 assert(0);
181 p += 4 + i2(p + 2);
182 }
183
184 /* Should have run exactly to end of extra data. */
185 assertEqualAddress(p, extension_end);
186
187 assertEqualAddress(p, eocd);
188
189 /* After Central dir, we find Zip64 eocd and Zip64 eocd locator. */
190 assertEqualMem(p, "PK\006\006", 4); /* Zip64 eocd */
191 assertEqualInt(i8(p + 4), 44); /* We're using v1 Zip64 eocd */
192 assertEqualInt(i2(p + 12), 45); /* Written by Version 4.5 */
193 assertEqualInt(i2(p + 14), 45); /* Needs version 4.5 to extract */
194 assertEqualInt(i4(p + 16), 0); /* This is disk #0 */
195 assertEqualInt(i4(p + 20), 0); /* Dir starts on disk #0 */
196 assertEqualInt(i8(p + 24), 1); /* 1 entry on this disk */
197 assertEqualInt(i8(p + 32), 1); /* 1 entry total */
198 assertEqualInt(i8(p + 40), eocd - central_header); /* size of cd */
199 assertEqualInt(i8(p + 48), central_header - buff); /* start of cd */
200 p += 12 + i8(p + 4);
201
202 assertEqualMem(p, "PK\006\007", 4); /* Zip64 eocd locator */
203 assertEqualInt(i4(p + 4), 0); /* Zip64 eocd is on disk #0 */
204 assertEqualInt(i8(p + 8), eocd - buff); /* Offset of Zip64 eocd */
205 assertEqualInt(i4(p + 16), 1); /* 1 disk */
206 p += 20;
207
208 /* Regular EOCD immediately follows Zip64 records. */
209 assertEqualAddress(p, eocd_record);
210
211 /* Verify local header of file entry. */
212 p = local_header = buff;
213 assertEqualMem(p, "PK\003\004", 4); /* Signature */
214 assertEqualInt(i2(p + 4), zip_version); /* Version needed to extract */
215 assertEqualInt(i2(p + 6), 8); /* Flags: bit 3 = length-at-end */
216 assertEqualInt(i2(p + 8), zip_compression); /* Compression method */
217 assertEqualInt(i2(p + 10), 0); /* File time */
218 assertEqualInt(i2(p + 12), 33); /* File date */
219 assertEqualInt(i4(p + 14), 0); /* CRC-32 */
220 assertEqualInt(i4(p + 18), 0); /* Compressed size must be zero for length-at-end */
221 assertEqualInt(i4(p + 22), 0); /* Uncompressed size must be zero for length-at-end */
222 assertEqualInt(i2(p + 26), strlen(file_name)); /* Pathname length */
223 assertEqualInt(i2(p + 28), 37); /* Extra field length */
224 assertEqualMem(p + 30, file_name, strlen(file_name)); /* Pathname */
225 p = extension_start = local_header + 30 + strlen(file_name);
226 extension_end = extension_start + i2(local_header + 28);
227
228 assertEqualInt(i2(p), 0x7875); /* 'ux' extension header */
229 assertEqualInt(i2(p + 2), 11); /* 'ux' size */
230 assertEqualInt(p[4], 1); /* 'ux' version */
231 assertEqualInt(p[5], 4); /* 'ux' uid size */
232 assertEqualInt(i4(p + 6), file_uid); /* 'Ux' UID */
233 assertEqualInt(p[10], 4); /* 'ux' gid size */
234 assertEqualInt(i4(p + 11), file_gid); /* 'Ux' GID */
235 p += 4 + i2(p + 2);
236
237 assertEqualInt(i2(p), 0x5455); /* 'UT' extension header */
238 assertEqualInt(i2(p + 2), 5); /* 'UT' size */
239 assertEqualInt(p[4], 1); /* 'UT' flags */
240 assertEqualInt(i4(p + 5), 0); /* 'UT' mtime */
241 p += 4 + i2(p + 2);
242
243 assertEqualInt(i2(p), 0x6c78); /* 'xl' experimental extension header */
244 assertEqualInt(i2(p + 2), 9); /* size */
245 assertEqualInt(p[4], 7); /* bitmap of included fields */
246 assertEqualInt(i2(p + 5) >> 8, 3); /* system & version made by */
247 assertEqualInt(i2(p + 7), 0); /* internal file attributes */
248 assertEqualInt(i4(p + 9) >> 16 & 01777, file_perm); /* external file attributes */
249 p += 4 + i2(p + 2);
250
251 /* Just in case: Report any extra extensions. */
252 while (p < extension_end) {
253 failure("Unexpected extension 0x%04X", i2(p));
254 assert(0);
255 p += 4 + i2(p + 2);
256 }
257
258 /* Should have run exactly to end of extra data. */
259 assertEqualAddress(p, extension_end);
260 data_start = p;
261
262 /* Data descriptor should follow compressed data. */
263 while (p < central_header && memcmp(p, "PK\007\010", 4) != 0)
264 ++p;
265 data_end = p;
266 assertEqualInt(data_end - data_start, compressed_size);
267 assertEqualMem(p, "PK\007\010", 4);
268 assertEqualInt(i4(p + 4), crc); /* CRC-32 */
269 assertEqualInt(i8(p + 8), compressed_size); /* compressed size */
270 assertEqualInt(i8(p + 16), sizeof(file_data)); /* uncompressed size */
271
272 /* Central directory should immediately follow the only entry. */
273 assertEqualAddress(p + 24, central_header);
274
275 free(buff);
276 }
277