projects/linux·blob v5.16 · refs

350 lines · 10.2 KiB

1// SPDX-License-Identifier: GPL-2.0
2
3/*
4 * Important notes about in-place decompression
5 *
6 * At least on x86, the kernel is decompressed in place: the compressed data
7 * is placed to the end of the output buffer, and the decompressor overwrites
8 * most of the compressed data. There must be enough safety margin to
9 * guarantee that the write position is always behind the read position.
10 *
11 * The safety margin for ZSTD with a 128 KB block size is calculated below.
12 * Note that the margin with ZSTD is bigger than with GZIP or XZ!
13 *
14 * The worst case for in-place decompression is that the beginning of
15 * the file is compressed extremely well, and the rest of the file is
16 * uncompressible. Thus, we must look for worst-case expansion when the
17 * compressor is encoding uncompressible data.
18 *
19 * The structure of the .zst file in case of a compressed kernel is as follows.
20 * Maximum sizes (as bytes) of the fields are in parenthesis.
21 *
22 * Frame Header: (18)
23 * Blocks: (N)
24 * Checksum: (4)
25 *
26 * The frame header and checksum overhead is at most 22 bytes.
27 *
28 * ZSTD stores the data in blocks. Each block has a header whose size is
29 * a 3 bytes. After the block header, there is up to 128 KB of payload.
30 * The maximum uncompressed size of the payload is 128 KB. The minimum
31 * uncompressed size of the payload is never less than the payload size
32 * (excluding the block header).
33 *
34 * The assumption, that the uncompressed size of the payload is never
35 * smaller than the payload itself, is valid only when talking about
36 * the payload as a whole. It is possible that the payload has parts where
37 * the decompressor consumes more input than it produces output. Calculating
38 * the worst case for this would be tricky. Instead of trying to do that,
39 * let's simply make sure that the decompressor never overwrites any bytes
40 * of the payload which it is currently reading.
41 *
42 * Now we have enough information to calculate the safety margin. We need
43 * - 22 bytes for the .zst file format headers;
44 * - 3 bytes per every 128 KiB of uncompressed size (one block header per
45 * block); and
46 * - 128 KiB (biggest possible zstd block size) to make sure that the
47 * decompressor never overwrites anything from the block it is currently
48 * reading.
49 *
50 * We get the following formula:
51 *
52 * safety_margin = 22 + uncompressed_size * 3 / 131072 + 131072
53 * <= 22 + (uncompressed_size >> 15) + 131072
54 */
55
56/*
57 * Preboot environments #include "path/to/decompress_unzstd.c".
58 * All of the source files we depend on must be #included.
59 * zstd's only source dependency is xxhash, which has no source
60 * dependencies.
61 *
62 * When UNZSTD_PREBOOT is defined we declare __decompress(), which is
63 * used for kernel decompression, instead of unzstd().
64 *
65 * Define __DISABLE_EXPORTS in preboot environments to prevent symbols
66 * from xxhash and zstd from being exported by the EXPORT_SYMBOL macro.
67 */
68#ifdef STATIC
69# define UNZSTD_PREBOOT
70# include "xxhash.c"
71# include "zstd/decompress_sources.h"
72#endif
73
74#include <linux/decompress/mm.h>
75#include <linux/kernel.h>
76#include <linux/zstd.h>
77
78/* 128MB is the maximum window size supported by zstd. */
79#define ZSTD_WINDOWSIZE_MAX (1 << ZSTD_WINDOWLOG_MAX)
80/*
81 * Size of the input and output buffers in multi-call mode.
82 * Pick a larger size because it isn't used during kernel decompression,
83 * since that is single pass, and we have to allocate a large buffer for
84 * zstd's window anyway. The larger size speeds up initramfs decompression.
85 */
86#define ZSTD_IOBUF_SIZE (1 << 17)
87
88static int INIT handle_zstd_error(size_t ret, void (*error)(char *x))
89{
90 const zstd_error_code err = zstd_get_error_code(ret);
91
92 if (!zstd_is_error(ret))
93 return 0;
94
95 /*
96 * zstd_get_error_name() cannot be used because error takes a char *
97 * not a const char *
98 */
99 switch (err) {
100 case ZSTD_error_memory_allocation:
101 error("ZSTD decompressor ran out of memory");
102 break;
103 case ZSTD_error_prefix_unknown:
104 error("Input is not in the ZSTD format (wrong magic bytes)");
105 break;
106 case ZSTD_error_dstSize_tooSmall:
107 case ZSTD_error_corruption_detected:
108 case ZSTD_error_checksum_wrong:
109 error("ZSTD-compressed data is corrupt");
110 break;
111 default:
112 error("ZSTD-compressed data is probably corrupt");
113 break;
114 }
115 return -1;
116}
117
118/*
119 * Handle the case where we have the entire input and output in one segment.
120 * We can allocate less memory (no circular buffer for the sliding window),
121 * and avoid some memcpy() calls.
122 */
123static int INIT decompress_single(const u8 *in_buf, long in_len, u8 *out_buf,
124 long out_len, long *in_pos,
125 void (*error)(char *x))
126{
127 const size_t wksp_size = zstd_dctx_workspace_bound();
128 void *wksp = large_malloc(wksp_size);
129 zstd_dctx *dctx = zstd_init_dctx(wksp, wksp_size);
130 int err;
131 size_t ret;
132
133 if (dctx == NULL) {
134 error("Out of memory while allocating zstd_dctx");
135 err = -1;
136 goto out;
137 }
138 /*
139 * Find out how large the frame actually is, there may be junk at
140 * the end of the frame that zstd_decompress_dctx() can't handle.
141 */
142 ret = zstd_find_frame_compressed_size(in_buf, in_len);
143 err = handle_zstd_error(ret, error);
144 if (err)
145 goto out;
146 in_len = (long)ret;
147
148 ret = zstd_decompress_dctx(dctx, out_buf, out_len, in_buf, in_len);
149 err = handle_zstd_error(ret, error);
150 if (err)
151 goto out;
152
153 if (in_pos != NULL)
154 *in_pos = in_len;
155
156 err = 0;
157out:
158 if (wksp != NULL)
159 large_free(wksp);
160 return err;
161}
162
163static int INIT __unzstd(unsigned char *in_buf, long in_len,
164 long (*fill)(void*, unsigned long),
165 long (*flush)(void*, unsigned long),
166 unsigned char *out_buf, long out_len,
167 long *in_pos,
168 void (*error)(char *x))
169{
170 zstd_in_buffer in;
171 zstd_out_buffer out;
172 zstd_frame_header header;
173 void *in_allocated = NULL;
174 void *out_allocated = NULL;
175 void *wksp = NULL;
176 size_t wksp_size;
177 zstd_dstream *dstream;
178 int err;
179 size_t ret;
180
181 /*
182 * ZSTD decompression code won't be happy if the buffer size is so big
183 * that its end address overflows. When the size is not provided, make
184 * it as big as possible without having the end address overflow.
185 */
186 if (out_len == 0)
187 out_len = UINTPTR_MAX - (uintptr_t)out_buf;
188
189 if (fill == NULL && flush == NULL)
190 /*
191 * We can decompress faster and with less memory when we have a
192 * single chunk.
193 */
194 return decompress_single(in_buf, in_len, out_buf, out_len,
195 in_pos, error);
196
197 /*
198 * If in_buf is not provided, we must be using fill(), so allocate
199 * a large enough buffer. If it is provided, it must be at least
200 * ZSTD_IOBUF_SIZE large.
201 */
202 if (in_buf == NULL) {
203 in_allocated = large_malloc(ZSTD_IOBUF_SIZE);
204 if (in_allocated == NULL) {
205 error("Out of memory while allocating input buffer");
206 err = -1;
207 goto out;
208 }
209 in_buf = in_allocated;
210 in_len = 0;
211 }
212 /* Read the first chunk, since we need to decode the frame header. */
213 if (fill != NULL)
214 in_len = fill(in_buf, ZSTD_IOBUF_SIZE);
215 if (in_len < 0) {
216 error("ZSTD-compressed data is truncated");
217 err = -1;
218 goto out;
219 }
220 /* Set the first non-empty input buffer. */
221 in.src = in_buf;
222 in.pos = 0;
223 in.size = in_len;
224 /* Allocate the output buffer if we are using flush(). */
225 if (flush != NULL) {
226 out_allocated = large_malloc(ZSTD_IOBUF_SIZE);
227 if (out_allocated == NULL) {
228 error("Out of memory while allocating output buffer");
229 err = -1;
230 goto out;
231 }
232 out_buf = out_allocated;
233 out_len = ZSTD_IOBUF_SIZE;
234 }
235 /* Set the output buffer. */
236 out.dst = out_buf;
237 out.pos = 0;
238 out.size = out_len;
239
240 /*
241 * We need to know the window size to allocate the zstd_dstream.
242 * Since we are streaming, we need to allocate a buffer for the sliding
243 * window. The window size varies from 1 KB to ZSTD_WINDOWSIZE_MAX
244 * (8 MB), so it is important to use the actual value so as not to
245 * waste memory when it is smaller.
246 */
247 ret = zstd_get_frame_header(&header, in.src, in.size);
248 err = handle_zstd_error(ret, error);
249 if (err)
250 goto out;
251 if (ret != 0) {
252 error("ZSTD-compressed data has an incomplete frame header");
253 err = -1;
254 goto out;
255 }
256 if (header.windowSize > ZSTD_WINDOWSIZE_MAX) {
257 error("ZSTD-compressed data has too large a window size");
258 err = -1;
259 goto out;
260 }
261
262 /*
263 * Allocate the zstd_dstream now that we know how much memory is
264 * required.
265 */
266 wksp_size = zstd_dstream_workspace_bound(header.windowSize);
267 wksp = large_malloc(wksp_size);
268 dstream = zstd_init_dstream(header.windowSize, wksp, wksp_size);
269 if (dstream == NULL) {
270 error("Out of memory while allocating ZSTD_DStream");
271 err = -1;
272 goto out;
273 }
274
275 /*
276 * Decompression loop:
277 * Read more data if necessary (error if no more data can be read).
278 * Call the decompression function, which returns 0 when finished.
279 * Flush any data produced if using flush().
280 */
281 if (in_pos != NULL)
282 *in_pos = 0;
283 do {
284 /*
285 * If we need to reload data, either we have fill() and can
286 * try to get more data, or we don't and the input is truncated.
287 */
288 if (in.pos == in.size) {
289 if (in_pos != NULL)
290 *in_pos += in.pos;
291 in_len = fill ? fill(in_buf, ZSTD_IOBUF_SIZE) : -1;
292 if (in_len < 0) {
293 error("ZSTD-compressed data is truncated");
294 err = -1;
295 goto out;
296 }
297 in.pos = 0;
298 in.size = in_len;
299 }
300 /* Returns zero when the frame is complete. */
301 ret = zstd_decompress_stream(dstream, &out, &in);
302 err = handle_zstd_error(ret, error);
303 if (err)
304 goto out;
305 /* Flush all of the data produced if using flush(). */
306 if (flush != NULL && out.pos > 0) {
307 if (out.pos != flush(out.dst, out.pos)) {
308 error("Failed to flush()");
309 err = -1;
310 goto out;
311 }
312 out.pos = 0;
313 }
314 } while (ret != 0);
315
316 if (in_pos != NULL)
317 *in_pos += in.pos;
318
319 err = 0;
320out:
321 if (in_allocated != NULL)
322 large_free(in_allocated);
323 if (out_allocated != NULL)
324 large_free(out_allocated);
325 if (wksp != NULL)
326 large_free(wksp);
327 return err;
328}
329
330#ifndef UNZSTD_PREBOOT
331STATIC int INIT unzstd(unsigned char *buf, long len,
332 long (*fill)(void*, unsigned long),
333 long (*flush)(void*, unsigned long),
334 unsigned char *out_buf,
335 long *pos,
336 void (*error)(char *x))
337{
338 return __unzstd(buf, len, fill, flush, out_buf, 0, pos, error);
339}
340#else
341STATIC int INIT __decompress(unsigned char *buf, long len,
342 long (*fill)(void*, unsigned long),
343 long (*flush)(void*, unsigned long),
344 unsigned char *out_buf, long out_len,
345 long *pos,
346 void (*error)(char *x))
347{
348 return __unzstd(buf, len, fill, flush, out_buf, out_len, pos, error);
349}
350#endif