1 // SPDX-License-Identifier: GPL-2.0-only
2 /*
3 * Copyright (C) 2024, SUSE LLC
4 *
5 * Authors: Enzo Matsumiya <ematsumiya@suse.de>
6 *
7 * This file implements I/O compression support for SMB2 messages (SMB 3.1.1 only).
8 * See compress/ for implementation details of each algorithm.
9 *
10 * References:
11 * MS-SMB2 "3.1.4.4 Compressing the Message"
12 * MS-SMB2 "3.1.5.3 Decompressing the Chained Message"
13 * MS-XCA - for details of the supported algorithms
14 */
15 #include <linux/slab.h>
16 #include <linux/kernel.h>
17 #include <linux/uio.h>
18 #include <linux/sort.h>
19
20 #include "cifsglob.h"
21 #include "../common/smb2pdu.h"
22 #include "cifsproto.h"
23 #include "smb2proto.h"
24
25 #include "compress/lz77.h"
26 #include "compress.h"
27
28 /*
29 * The heuristic_*() functions below try to determine data compressibility.
30 *
31 * Derived from fs/btrfs/compression.c, changing coding style, some parameters, and removing
32 * unused parts.
33 *
34 * Read that file for better and more detailed explanation of the calculations.
35 *
36 * The algorithms are ran in a collected sample of the input (uncompressed) data.
37 * The sample is formed of 2K reads in PAGE_SIZE intervals, with a maximum size of 4M.
38 *
39 * Parsing the sample goes from "low-hanging fruits" (fastest algorithms, likely compressible)
40 * to "need more analysis" (likely uncompressible).
41 */
42
43 struct bucket {
44 unsigned int count;
45 };
46
47 /**
48 * has_low_entropy() - Compute Shannon entropy of the sampled data.
49 * @bkt: Bytes counts of the sample.
50 * @slen: Size of the sample.
51 *
52 * Return: true if the level (percentage of number of bits that would be required to
53 * compress the data) is below the minimum threshold.
54 *
55 * Note:
56 * There _is_ an entropy level here that's > 65 (minimum threshold) that would indicate a
57 * possibility of compression, but compressing, or even further analysing, it would waste so much
58 * resources that it's simply not worth it.
59 *
60 * Also Shannon entropy is the last computed heuristic; if we got this far and ended up
61 * with uncertainty, just stay on the safe side and call it uncompressible.
62 */
63 static bool has_low_entropy(struct bucket *bkt, size_t slen)
64 {
65 const size_t threshold = 65, max_entropy = 8 * ilog2(16);
66 size_t i, p, p2, len, sum = 0;
67
68 #define pow4(n) (n * n * n * n)
69 len = ilog2(pow4(slen));
70
71 for (i = 0; i < 256 && bkt[i].count > 0; i++) {
72 p = bkt[i].count;
73 p2 = ilog2(pow4(p));
74 sum += p * (len - p2);
75 }
76
77 sum /= slen;
78
79 return ((sum * 100 / max_entropy) <= threshold);
80 }
81
82 #define BYTE_DIST_BAD 0
83 #define BYTE_DIST_GOOD 1
84 #define BYTE_DIST_MAYBE 2
85 /**
86 * calc_byte_distribution() - Compute byte distribution on the sampled data.
87 * @bkt: Byte counts of the sample.
88 * @slen: Size of the sample.
89 *
90 * Return:
91 * BYTE_DIST_BAD: A "hard no" for compression -- a computed uniform distribution of
92 * the bytes (e.g. random or encrypted data).
93 * BYTE_DIST_GOOD: High probability (normal (Gaussian) distribution) of the data being
94 * compressible.
95 * BYTE_DIST_MAYBE: When computed byte distribution resulted in "low > n < high"
96 * grounds. has_low_entropy() should be used for a final decision.
97 */
98 static int calc_byte_distribution(struct bucket *bkt, size_t slen)
99 {
100 const size_t low = 64, high = 200, threshold = slen * 90 / 100;
101 size_t sum = 0;
102 int i;
103
104 for (i = 0; i < low; i++)
105 sum += bkt[i].count;
106
107 if (sum > threshold)
108 return BYTE_DIST_BAD;
109
110 for (; i < high && bkt[i].count > 0; i++) {
111 sum += bkt[i].count;
112 if (sum > threshold)
113 break;
114 }
115
116 if (i <= low)
117 return BYTE_DIST_GOOD;
118
119 if (i >= high)
120 return BYTE_DIST_BAD;
121
122 return BYTE_DIST_MAYBE;
123 }
124
125 static bool is_mostly_ascii(const struct bucket *bkt)
126 {
127 size_t count = 0;
128 int i;
129
130 for (i = 0; i < 256; i++)
131 if (bkt[i].count > 0)
132 /* Too many non-ASCII (0-63) bytes. */
133 if (++count > 64)
134 return false;
135
136 return true;
137 }
138
139 static bool has_repeated_data(const u8 *sample, size_t len)
140 {
141 size_t s = len / 2;
142
143 return (!memcmp(&sample[0], &sample[s], s));
144 }
145
146 static int cmp_bkt(const void *_a, const void *_b)
147 {
148 const struct bucket *a = _a, *b = _b;
149
150 /* Reverse sort. */
151 if (a->count > b->count)
152 return -1;
153
154 return 1;
155 }
156
157 /*
158 * TODO:
159 * Support other iter types, if required.
160 * Only ITER_XARRAY is supported for now.
161 */
162 static int collect_sample(const struct iov_iter *iter, ssize_t max, u8 *sample)
163 {
164 struct folio *folios[16], *folio;
165 unsigned int nr, i, j, npages;
166 loff_t start = iter->xarray_start + iter->iov_offset;
167 pgoff_t last, index = start / PAGE_SIZE;
168 size_t len, off, foff;
169 void *p;
170 int s = 0;
171
172 last = (start + max - 1) / PAGE_SIZE;
173 do {
174 nr = xa_extract(iter->xarray, (void **)folios, index, last, ARRAY_SIZE(folios),
175 XA_PRESENT);
176 if (nr == 0)
177 return -EIO;
178
179 for (i = 0; i < nr; i++) {
180 folio = folios[i];
181 npages = folio_nr_pages(folio);
182 foff = start - folio_pos(folio);
183 off = foff % PAGE_SIZE;
184
185 for (j = foff / PAGE_SIZE; j < npages; j++) {
186 size_t len2;
187
188 len = min_t(size_t, max, PAGE_SIZE - off);
189 len2 = min_t(size_t, len, SZ_2K);
190
191 p = kmap_local_page(folio_page(folio, j));
192 memcpy(&sample[s], p, len2);
193 kunmap_local(p);
194
195 s += len2;
196
197 if (len2 < SZ_2K || s >= max - SZ_2K)
198 return s;
199
200 max -= len;
201 if (max <= 0)
202 return s;
203
204 start += len;
205 off = 0;
206 index++;
207 }
208 }
209 } while (nr == ARRAY_SIZE(folios));
210
211 return s;
212 }
213
214 /**
215 * is_compressible() - Determines if a chunk of data is compressible.
216 * @data: Iterator containing uncompressed data.
217 *
218 * Return: true if @data is compressible, false otherwise.
219 *
220 * Tests shows that this function is quite reliable in predicting data compressibility,
221 * matching close to 1:1 with the behaviour of LZ77 compression success and failures.
222 */
223 static bool is_compressible(const struct iov_iter *data)
224 {
225 const size_t read_size = SZ_2K, bkt_size = 256, max = SZ_4M;
226 struct bucket *bkt = NULL;
227 size_t len;
228 u8 *sample;
229 bool ret = false;
230 int i;
231
232 /* Preventive double check -- already checked in should_compress(). */
233 len = iov_iter_count(data);
234 if (unlikely(len < read_size))
235 return ret;
236
237 if (len - read_size > max)
238 len = max;
239
240 sample = kvzalloc(len, GFP_KERNEL);
241 if (!sample) {
242 WARN_ON_ONCE(1);
243
244 return ret;
245 }
246
247 /* Sample 2K bytes per page of the uncompressed data. */
248 i = collect_sample(data, len, sample);
249 if (i <= 0) {
250 WARN_ON_ONCE(1);
251
252 goto out;
253 }
254
255 len = i;
256 ret = true;
257
258 if (has_repeated_data(sample, len))
259 goto out;
260
261 bkt = kcalloc(bkt_size, sizeof(*bkt), GFP_KERNEL);
262 if (!bkt) {
263 WARN_ON_ONCE(1);
264 ret = false;
265
266 goto out;
267 }
268
269 for (i = 0; i < len; i++)
270 bkt[sample[i]].count++;
271
272 if (is_mostly_ascii(bkt))
273 goto out;
274
275 /* Sort in descending order */
276 sort(bkt, bkt_size, sizeof(*bkt), cmp_bkt, NULL);
277
278 i = calc_byte_distribution(bkt, len);
279 if (i != BYTE_DIST_MAYBE) {
280 ret = !!i;
281
282 goto out;
283 }
284
285 ret = has_low_entropy(bkt, len);
286 out:
287 kvfree(sample);
288 kfree(bkt);
289
290 return ret;
291 }
292
293 bool should_compress(const struct cifs_tcon *tcon, const struct smb_rqst *rq)
294 {
295 const struct smb2_hdr *shdr = rq->rq_iov->iov_base;
296
297 if (unlikely(!tcon || !tcon->ses || !tcon->ses->server))
298 return false;
299
300 if (!tcon->ses->server->compression.enabled)
301 return false;
302
303 if (!(tcon->share_flags & SMB2_SHAREFLAG_COMPRESS_DATA))
304 return false;
305
306 if (shdr->Command == SMB2_WRITE) {
307 const struct smb2_write_req *wreq = rq->rq_iov->iov_base;
308
309 if (le32_to_cpu(wreq->Length) < SMB_COMPRESS_MIN_LEN)
310 return false;
311
312 return is_compressible(&rq->rq_iter);
313 }
314
315 return (shdr->Command == SMB2_READ);
316 }
317
318 int smb_compress(struct TCP_Server_Info *server, struct smb_rqst *rq, compress_send_fn send_fn)
319 {
320 struct iov_iter iter;
321 u32 slen, dlen;
322 void *src, *dst = NULL;
323 int ret;
324
325 if (!server || !rq || !rq->rq_iov || !rq->rq_iov->iov_base)
326 return -EINVAL;
327
328 if (rq->rq_iov->iov_len != sizeof(struct smb2_write_req))
329 return -EINVAL;
330
331 slen = iov_iter_count(&rq->rq_iter);
332 src = kvzalloc(slen, GFP_KERNEL);
333 if (!src) {
334 ret = -ENOMEM;
335 goto err_free;
336 }
337
338 /* Keep the original iter intact. */
339 iter = rq->rq_iter;
340
341 if (!copy_from_iter_full(src, slen, &iter)) {
342 ret = -EIO;
343 goto err_free;
344 }
345
346 /*
347 * This is just overprovisioning, as the algorithm will error out if @dst reaches 7/8
348 * of @slen.
349 */
350 dlen = slen;
351 dst = kvzalloc(dlen, GFP_KERNEL);
352 if (!dst) {
353 ret = -ENOMEM;
354 goto err_free;
355 }
356
357 ret = lz77_compress(src, slen, dst, &dlen);
358 if (!ret) {
359 struct smb2_compression_hdr hdr = { 0 };
360 struct smb_rqst comp_rq = { .rq_nvec = 3, };
361 struct kvec iov[3];
362
363 hdr.ProtocolId = SMB2_COMPRESSION_TRANSFORM_ID;
364 hdr.OriginalCompressedSegmentSize = cpu_to_le32(slen);
365 hdr.CompressionAlgorithm = SMB3_COMPRESS_LZ77;
366 hdr.Flags = SMB2_COMPRESSION_FLAG_NONE;
367 hdr.Offset = cpu_to_le32(rq->rq_iov[0].iov_len);
368
369 iov[0].iov_base = &hdr;
370 iov[0].iov_len = sizeof(hdr);
371 iov[1] = rq->rq_iov[0];
372 iov[2].iov_base = dst;
373 iov[2].iov_len = dlen;
374
375 comp_rq.rq_iov = iov;
376
377 ret = send_fn(server, 1, &comp_rq);
378 } else if (ret == -EMSGSIZE || dlen >= slen) {
379 ret = send_fn(server, 1, rq);
380 }
381 err_free:
382 kvfree(dst);
383 kvfree(src);
384
385 return ret;
386 }
387
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.