| | varnish-cache/bin/varnishd/cache/cache_rfc2616.c |
| 0 |
|
/*- |
| 1 |
|
* Copyright (c) 2006 Verdens Gang AS |
| 2 |
|
* Copyright (c) 2006-2011 Varnish Software AS |
| 3 |
|
* All rights reserved. |
| 4 |
|
* |
| 5 |
|
* Author: Poul-Henning Kamp <phk@phk.freebsd.dk> |
| 6 |
|
* |
| 7 |
|
* SPDX-License-Identifier: BSD-2-Clause |
| 8 |
|
* |
| 9 |
|
* Redistribution and use in source and binary forms, with or without |
| 10 |
|
* modification, are permitted provided that the following conditions |
| 11 |
|
* are met: |
| 12 |
|
* 1. Redistributions of source code must retain the above copyright |
| 13 |
|
* notice, this list of conditions and the following disclaimer. |
| 14 |
|
* 2. Redistributions in binary form must reproduce the above copyright |
| 15 |
|
* notice, this list of conditions and the following disclaimer in the |
| 16 |
|
* documentation and/or other materials provided with the distribution. |
| 17 |
|
* |
| 18 |
|
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND |
| 19 |
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
| 20 |
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE |
| 21 |
|
* ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE |
| 22 |
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL |
| 23 |
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS |
| 24 |
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) |
| 25 |
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT |
| 26 |
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY |
| 27 |
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF |
| 28 |
|
* SUCH DAMAGE. |
| 29 |
|
*/ |
| 30 |
|
|
| 31 |
|
#include "config.h" |
| 32 |
|
|
| 33 |
|
#include <stdlib.h> |
| 34 |
|
|
| 35 |
|
#include "cache_varnishd.h" |
| 36 |
|
|
| 37 |
|
#include "vtim.h" |
| 38 |
|
#include "vct.h" |
| 39 |
|
|
| 40 |
|
/*-------------------------------------------------------------------- |
| 41 |
|
* TTL and Age calculation in Varnish |
| 42 |
|
* |
| 43 |
|
* RFC2616 has a lot to say about how caches should calculate the TTL |
| 44 |
|
* and expiry times of objects, but it sort of misses the case that |
| 45 |
|
* applies to Varnish: the server-side cache. |
| 46 |
|
* |
| 47 |
|
* A normal cache, shared or single-client, has no symbiotic relationship |
| 48 |
|
* with the server, and therefore must take a very defensive attitude |
| 49 |
|
* if the Data/Expiry/Age/max-age data does not make sense. Overall |
| 50 |
|
* the policy described in section 13 of RFC 2616 results in no caching |
| 51 |
|
* happening on the first little sign of trouble. |
| 52 |
|
* |
| 53 |
|
* Varnish on the other hand tries to offload as many transactions from |
| 54 |
|
* the backend as possible, and therefore just passing through everything |
| 55 |
|
* if there is a clock-skew between backend and Varnish is not a workable |
| 56 |
|
* choice. |
| 57 |
|
* |
| 58 |
|
* Varnish implements a policy which is RFC2616 compliant when there |
| 59 |
|
* is no clockskew, and falls as gracefully as possible otherwise. |
| 60 |
|
* Our "clockless cache" model is synthesized from the bits of RFC2616 |
| 61 |
|
* that talks about how a cache should react to a clockless origin server, |
| 62 |
|
* and more or less uses the inverse logic for the opposite relationship. |
| 63 |
|
* |
| 64 |
|
*/ |
| 65 |
|
|
| 66 |
|
static inline unsigned |
| 67 |
2200 |
rfc2616_time(const char *p) |
| 68 |
|
{ |
| 69 |
|
char *ep; |
| 70 |
|
unsigned long val; |
| 71 |
2200 |
if (*p == '-') |
| 72 |
80 |
return (0); |
| 73 |
2120 |
val = strtoul(p, &ep, 10); |
| 74 |
2120 |
if (val > UINT_MAX) |
| 75 |
0 |
return (UINT_MAX); |
| 76 |
2120 |
while (vct_issp(*ep)) |
| 77 |
0 |
ep++; |
| 78 |
|
/* We accept ',' as an end character because we may be parsing a |
| 79 |
|
* multi-element Cache-Control part. We accept '.' to be future |
| 80 |
|
* compatible with fractional seconds. */ |
| 81 |
2120 |
if (*ep == '\0' || *ep == ',' || *ep == '.') |
| 82 |
1960 |
return (val); |
| 83 |
160 |
return (0); |
| 84 |
2200 |
} |
| 85 |
|
|
| 86 |
|
void |
| 87 |
81036 |
RFC2616_Ttl(struct busyobj *bo, vtim_real now, vtim_real *t_origin, |
| 88 |
|
float *ttl, float *grace, float *keep) |
| 89 |
|
{ |
| 90 |
|
unsigned max_age, age; |
| 91 |
|
vtim_real h_date, h_expires; |
| 92 |
|
const char *p; |
| 93 |
|
const struct http *hp; |
| 94 |
|
|
| 95 |
81036 |
CHECK_OBJ_NOTNULL(bo, BUSYOBJ_MAGIC); |
| 96 |
81036 |
CHECK_OBJ_NOTNULL(bo->fetch_objcore, OBJCORE_MAGIC); |
| 97 |
81036 |
assert(now != 0.0 && !isnan(now)); |
| 98 |
81036 |
AN(t_origin); |
| 99 |
81036 |
AN(ttl); |
| 100 |
81036 |
AN(grace); |
| 101 |
81036 |
AN(keep); |
| 102 |
|
|
| 103 |
81036 |
*t_origin = now; |
| 104 |
81036 |
*ttl = cache_param->default_ttl; |
| 105 |
81036 |
*grace = cache_param->default_grace; |
| 106 |
81036 |
*keep = cache_param->default_keep; |
| 107 |
|
|
| 108 |
81036 |
hp = bo->beresp; |
| 109 |
|
|
| 110 |
81036 |
max_age = age = 0; |
| 111 |
81036 |
h_expires = 0; |
| 112 |
81036 |
h_date = 0; |
| 113 |
|
|
| 114 |
|
/* |
| 115 |
|
* Initial cacheability determination per [RFC2616, 13.4] |
| 116 |
|
* We do not support ranges to the backend yet, so 206 is out. |
| 117 |
|
*/ |
| 118 |
|
|
| 119 |
81036 |
if (http_GetHdr(hp, H_Age, &p)) { |
| 120 |
880 |
age = rfc2616_time(p); |
| 121 |
880 |
*t_origin -= age; |
| 122 |
880 |
} |
| 123 |
|
|
| 124 |
81036 |
if (bo->fetch_objcore->flags & OC_F_PRIVATE) { |
| 125 |
|
/* Pass object. Halt the processing here, keeping only the |
| 126 |
|
* parsed value of t_origin, as that will be needed to |
| 127 |
|
* synthesize a correct Age header in delivery. The |
| 128 |
|
* SLT_TTL log tag at the end of this function is |
| 129 |
|
* deliberately skipped to avoid confusion when reading |
| 130 |
|
* the log.*/ |
| 131 |
26716 |
*ttl = -1; |
| 132 |
26716 |
*grace = 0; |
| 133 |
26716 |
*keep = 0; |
| 134 |
26716 |
return; |
| 135 |
|
} |
| 136 |
|
|
| 137 |
54320 |
if (http_GetHdr(hp, H_Expires, &p)) |
| 138 |
320 |
h_expires = VTIM_parse(p); |
| 139 |
|
|
| 140 |
54320 |
if (http_GetHdr(hp, H_Date, &p)) |
| 141 |
54320 |
h_date = VTIM_parse(p); |
| 142 |
|
|
| 143 |
54320 |
switch (http_GetStatus(hp)) { |
| 144 |
|
case 302: /* Moved Temporarily */ |
| 145 |
|
case 307: /* Temporary Redirect */ |
| 146 |
|
/* |
| 147 |
|
* https://tools.ietf.org/html/rfc7231#section-6.1 |
| 148 |
|
* |
| 149 |
|
* Do not apply the default ttl, only set a ttl if Cache-Control |
| 150 |
|
* or Expires are present. Uncacheable otherwise. |
| 151 |
|
*/ |
| 152 |
40 |
*ttl = -1.; |
| 153 |
|
/* FALLTHROUGH */ |
| 154 |
|
case 200: /* OK */ |
| 155 |
|
case 203: /* Non-Authoritative Information */ |
| 156 |
|
case 204: /* No Content */ |
| 157 |
|
case 300: /* Multiple Choices */ |
| 158 |
|
case 301: /* Moved Permanently */ |
| 159 |
|
case 304: /* Not Modified - handled like 200 */ |
| 160 |
|
case 404: /* Not Found */ |
| 161 |
|
case 410: /* Gone */ |
| 162 |
|
case 414: /* Request-URI Too Large */ |
| 163 |
|
/* |
| 164 |
|
* First find any relative specification from the backend |
| 165 |
|
* These take precedence according to RFC2616, 13.2.4 |
| 166 |
|
*/ |
| 167 |
|
|
| 168 |
106319 |
if ((http_GetHdrField(hp, H_Cache_Control, "s-maxage", &p) || |
| 169 |
53039 |
http_GetHdrField(hp, H_Cache_Control, "max-age", &p)) && |
| 170 |
53280 |
p != NULL) { |
| 171 |
1160 |
max_age = rfc2616_time(p); |
| 172 |
1160 |
*ttl = max_age; |
| 173 |
1160 |
break; |
| 174 |
|
} |
| 175 |
|
|
| 176 |
|
/* No expire header, fall back to default */ |
| 177 |
52120 |
if (h_expires == 0) |
| 178 |
51960 |
break; |
| 179 |
|
|
| 180 |
|
|
| 181 |
|
/* If backend told us it is expired already, don't cache. */ |
| 182 |
160 |
if (h_expires < h_date) { |
| 183 |
120 |
*ttl = 0; |
| 184 |
120 |
break; |
| 185 |
|
} |
| 186 |
|
|
| 187 |
40 |
if (h_date == 0 || |
| 188 |
40 |
fabs(h_date - now) < cache_param->clock_skew) { |
| 189 |
|
/* |
| 190 |
|
* If we have no Date: header or if it is |
| 191 |
|
* sufficiently close to our clock we will |
| 192 |
|
* trust Expires: relative to our own clock. |
| 193 |
|
*/ |
| 194 |
0 |
if (h_expires < now) |
| 195 |
0 |
*ttl = 0; |
| 196 |
|
else |
| 197 |
0 |
*ttl = h_expires - now; |
| 198 |
0 |
break; |
| 199 |
|
} else { |
| 200 |
|
/* |
| 201 |
|
* But even if the clocks are out of whack we can still |
| 202 |
|
* derive a relative time from the two headers. |
| 203 |
|
* (the negative ttl case is caught above) |
| 204 |
|
*/ |
| 205 |
40 |
*ttl = (int)(h_expires - h_date); |
| 206 |
|
} |
| 207 |
40 |
break; |
| 208 |
|
default: |
| 209 |
1040 |
*ttl = -1.; |
| 210 |
1040 |
break; |
| 211 |
|
} |
| 212 |
|
|
| 213 |
|
/* |
| 214 |
|
* RFC5861 outlines a way to control the use of stale responses. |
| 215 |
|
* We use this to initialize the grace period. |
| 216 |
|
*/ |
| 217 |
54320 |
if (*ttl >= 0 && http_GetHdrField(hp, H_Cache_Control, |
| 218 |
53240 |
"stale-while-revalidate", &p) && p != NULL) { |
| 219 |
160 |
*grace = rfc2616_time(p); |
| 220 |
160 |
} |
| 221 |
|
|
| 222 |
108640 |
VSLb(bo->vsl, SLT_TTL, |
| 223 |
|
"RFC %.0f %.0f %.0f %.0f %.0f %.0f %.0f %u %s", |
| 224 |
54320 |
*ttl, *grace, *keep, now, |
| 225 |
54320 |
*t_origin, h_date, h_expires, max_age, |
| 226 |
54320 |
bo->uncacheable ? "uncacheable" : "cacheable"); |
| 227 |
81036 |
} |
| 228 |
|
|
| 229 |
|
/*-------------------------------------------------------------------- |
| 230 |
|
* Find out if the request can receive a gzipped response |
| 231 |
|
*/ |
| 232 |
|
|
| 233 |
|
unsigned |
| 234 |
142567 |
RFC2616_Req_Gzip(const struct http *hp) |
| 235 |
|
{ |
| 236 |
|
|
| 237 |
|
|
| 238 |
|
/* |
| 239 |
|
* "x-gzip" is for http/1.0 backwards compat, final note in 14.3 |
| 240 |
|
* p104 says to not do q values for x-gzip, so we just test |
| 241 |
|
* for its existence. |
| 242 |
|
*/ |
| 243 |
142567 |
if (http_GetHdrToken(hp, H_Accept_Encoding, "x-gzip", NULL, NULL)) |
| 244 |
80 |
return (1); |
| 245 |
|
|
| 246 |
|
/* |
| 247 |
|
* "gzip" is the real thing, but the 'q' value must be nonzero. |
| 248 |
|
* We do not care a hoot if the client prefers some other |
| 249 |
|
* compression more than gzip: Varnish only does gzip. |
| 250 |
|
*/ |
| 251 |
142487 |
if (http_GetHdrQ(hp, H_Accept_Encoding, "gzip") > 0.) |
| 252 |
24120 |
return (1); |
| 253 |
|
|
| 254 |
|
/* Bad client, no gzip. */ |
| 255 |
118367 |
return (0); |
| 256 |
142567 |
} |
| 257 |
|
|
| 258 |
|
/*--------------------------------------------------------------------*/ |
| 259 |
|
|
| 260 |
|
// rfc7232,l,547,548 |
| 261 |
|
static inline int |
| 262 |
160 |
rfc2616_strong_compare(const char *p, const char *e) |
| 263 |
|
{ |
| 264 |
200 |
if ((p[0] == 'W' && p[1] == '/') || |
| 265 |
80 |
(e[0] == 'W' && e[1] == '/')) |
| 266 |
120 |
return (0); |
| 267 |
|
/* XXX: should we also have http_etag_cmp() ? */ |
| 268 |
40 |
return (strcmp(p, e) == 0); |
| 269 |
160 |
} |
| 270 |
|
|
| 271 |
|
// rfc7232,l,550,552 |
| 272 |
|
static inline int |
| 273 |
840 |
rfc2616_weak_compare(const char *p, const char *e) |
| 274 |
|
{ |
| 275 |
840 |
if (p[0] == 'W' && p[1] == '/') |
| 276 |
160 |
p += 2; |
| 277 |
840 |
if (e[0] == 'W' && e[1] == '/') |
| 278 |
160 |
e += 2; |
| 279 |
|
/* XXX: should we also have http_etag_cmp() ? */ |
| 280 |
840 |
return (strcmp(p, e) == 0); |
| 281 |
|
} |
| 282 |
|
|
| 283 |
|
int |
| 284 |
112397 |
RFC2616_Do_Cond(const struct req *req) |
| 285 |
|
{ |
| 286 |
|
const char *p, *e; |
| 287 |
|
vtim_real ims, lm; |
| 288 |
|
|
| 289 |
112397 |
if (!http_IsStatus(req->resp, 200)) |
| 290 |
13878 |
return (0); |
| 291 |
|
|
| 292 |
|
/* |
| 293 |
|
* rfc7232,l,861,866 |
| 294 |
|
* We MUST ignore If-Modified-Since if we have an If-None-Match header |
| 295 |
|
*/ |
| 296 |
98519 |
if (http_GetHdr(req->http, H_If_None_Match, &p)) { |
| 297 |
1040 |
if (!http_GetHdr(req->resp, H_ETag, &e)) |
| 298 |
40 |
return (0); |
| 299 |
1000 |
if (http_GetHdr(req->http, H_Range, NULL)) |
| 300 |
160 |
return (rfc2616_strong_compare(p, e)); |
| 301 |
|
else |
| 302 |
840 |
return (rfc2616_weak_compare(p, e)); |
| 303 |
|
} |
| 304 |
|
|
| 305 |
97479 |
if (http_GetHdr(req->http, H_If_Modified_Since, &p)) { |
| 306 |
1040 |
ims = VTIM_parse(p); |
| 307 |
1040 |
if (!ims || ims > req->t_req) // rfc7232,l,868,869 |
| 308 |
0 |
return (0); |
| 309 |
1040 |
if (http_GetHdr(req->resp, H_Last_Modified, &p)) { |
| 310 |
960 |
lm = VTIM_parse(p); |
| 311 |
960 |
if (!lm || lm > ims) |
| 312 |
280 |
return (0); |
| 313 |
680 |
return (1); |
| 314 |
|
} |
| 315 |
80 |
AZ(ObjGetDouble(req->wrk, req->objcore, OA_LASTMODIFIED, &lm)); |
| 316 |
80 |
if (lm > ims) |
| 317 |
40 |
return (0); |
| 318 |
40 |
return (1); |
| 319 |
|
} |
| 320 |
|
|
| 321 |
96439 |
return (0); |
| 322 |
112397 |
} |
| 323 |
|
|
| 324 |
|
/*--------------------------------------------------------------------*/ |
| 325 |
|
|
| 326 |
|
void |
| 327 |
21080 |
RFC2616_Weaken_Etag(struct http *hp) |
| 328 |
|
{ |
| 329 |
|
const char *p; |
| 330 |
|
|
| 331 |
21080 |
CHECK_OBJ_NOTNULL(hp, HTTP_MAGIC); |
| 332 |
|
|
| 333 |
21080 |
if (!http_GetHdr(hp, H_ETag, &p)) |
| 334 |
20680 |
return; |
| 335 |
400 |
AN(p); |
| 336 |
400 |
if (p[0] == 'W' && p[1] == '/') |
| 337 |
80 |
return; |
| 338 |
320 |
http_Unset(hp, H_ETag); |
| 339 |
320 |
http_PrintfHeader(hp, "ETag: W/%s", p); |
| 340 |
21080 |
} |
| 341 |
|
|
| 342 |
|
/*--------------------------------------------------------------------*/ |
| 343 |
|
|
| 344 |
|
void |
| 345 |
4920 |
RFC2616_Vary_AE(struct http *hp) |
| 346 |
|
{ |
| 347 |
|
const char *vary; |
| 348 |
|
|
| 349 |
4920 |
if (http_GetHdrToken(hp, H_Vary, "Accept-Encoding", NULL, NULL)) |
| 350 |
280 |
return; |
| 351 |
4640 |
if (http_GetHdr(hp, H_Vary, &vary)) { |
| 352 |
80 |
http_Unset(hp, H_Vary); |
| 353 |
80 |
http_PrintfHeader(hp, "Vary: %s, Accept-Encoding", vary); |
| 354 |
80 |
} else { |
| 355 |
4560 |
http_SetHeader(hp, "Vary: Accept-Encoding"); |
| 356 |
|
} |
| 357 |
4920 |
} |
| 358 |
|
|
| 359 |
|
/*--------------------------------------------------------------------*/ |
| 360 |
|
|
| 361 |
|
const char * |
| 362 |
48320 |
RFC2616_Strong_LM(const struct http *hp, struct worker *wrk, |
| 363 |
|
struct objcore *oc) |
| 364 |
|
{ |
| 365 |
48320 |
const char *p = NULL, *e = NULL; |
| 366 |
|
vtim_real lm, d; |
| 367 |
|
|
| 368 |
48320 |
CHECK_OBJ_ORNULL(wrk, WORKER_MAGIC); |
| 369 |
48320 |
CHECK_OBJ_ORNULL(oc, OBJCORE_MAGIC); |
| 370 |
48320 |
CHECK_OBJ_ORNULL(hp, HTTP_MAGIC); |
| 371 |
|
|
| 372 |
48320 |
if (hp != NULL) { |
| 373 |
46760 |
(void)http_GetHdr(hp, H_Last_Modified, &p); |
| 374 |
46760 |
(void)http_GetHdr(hp, H_Date, &e); |
| 375 |
48320 |
} else if (wrk != NULL && oc != NULL) { |
| 376 |
1560 |
p = HTTP_GetHdrPack(wrk, oc, H_Last_Modified); |
| 377 |
1560 |
e = HTTP_GetHdrPack(wrk, oc, H_Date); |
| 378 |
1560 |
} |
| 379 |
|
|
| 380 |
48320 |
if (p == NULL || e == NULL) |
| 381 |
46760 |
return (NULL); |
| 382 |
|
|
| 383 |
1560 |
lm = VTIM_parse(p); |
| 384 |
1560 |
d = VTIM_parse(e); |
| 385 |
|
|
| 386 |
|
/* The cache entry includes a Date value which is at least one second |
| 387 |
|
* after the Last-Modified value. |
| 388 |
|
* [RFC9110 8.8.2.2-6.2] |
| 389 |
|
*/ |
| 390 |
1560 |
return ((lm && d && lm + 1 <= d) ? p : NULL); |
| 391 |
48320 |
} |