uipc_mbuf.c revision 1.100.2.16 1 1.100.2.16 yamt /* $NetBSD: uipc_mbuf.c,v 1.100.2.16 2007/11/15 11:44:56 yamt Exp $ */
2 1.42 thorpej
3 1.42 thorpej /*-
4 1.53 thorpej * Copyright (c) 1999, 2001 The NetBSD Foundation, Inc.
5 1.42 thorpej * All rights reserved.
6 1.42 thorpej *
7 1.42 thorpej * This code is derived from software contributed to The NetBSD Foundation
8 1.42 thorpej * by Jason R. Thorpe of the Numerical Aerospace Simulation Facility,
9 1.42 thorpej * NASA Ames Research Center.
10 1.42 thorpej *
11 1.42 thorpej * Redistribution and use in source and binary forms, with or without
12 1.42 thorpej * modification, are permitted provided that the following conditions
13 1.42 thorpej * are met:
14 1.42 thorpej * 1. Redistributions of source code must retain the above copyright
15 1.42 thorpej * notice, this list of conditions and the following disclaimer.
16 1.42 thorpej * 2. Redistributions in binary form must reproduce the above copyright
17 1.42 thorpej * notice, this list of conditions and the following disclaimer in the
18 1.42 thorpej * documentation and/or other materials provided with the distribution.
19 1.42 thorpej * 3. All advertising materials mentioning features or use of this software
20 1.42 thorpej * must display the following acknowledgement:
21 1.42 thorpej * This product includes software developed by the NetBSD
22 1.42 thorpej * Foundation, Inc. and its contributors.
23 1.42 thorpej * 4. Neither the name of The NetBSD Foundation nor the names of its
24 1.42 thorpej * contributors may be used to endorse or promote products derived
25 1.42 thorpej * from this software without specific prior written permission.
26 1.42 thorpej *
27 1.42 thorpej * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
28 1.42 thorpej * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
29 1.42 thorpej * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
30 1.42 thorpej * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
31 1.42 thorpej * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32 1.42 thorpej * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33 1.42 thorpej * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34 1.42 thorpej * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35 1.42 thorpej * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36 1.42 thorpej * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37 1.42 thorpej * POSSIBILITY OF SUCH DAMAGE.
38 1.42 thorpej */
39 1.10 cgd
40 1.1 cgd /*
41 1.9 mycroft * Copyright (c) 1982, 1986, 1988, 1991, 1993
42 1.9 mycroft * The Regents of the University of California. All rights reserved.
43 1.1 cgd *
44 1.1 cgd * Redistribution and use in source and binary forms, with or without
45 1.1 cgd * modification, are permitted provided that the following conditions
46 1.1 cgd * are met:
47 1.1 cgd * 1. Redistributions of source code must retain the above copyright
48 1.1 cgd * notice, this list of conditions and the following disclaimer.
49 1.1 cgd * 2. Redistributions in binary form must reproduce the above copyright
50 1.1 cgd * notice, this list of conditions and the following disclaimer in the
51 1.1 cgd * documentation and/or other materials provided with the distribution.
52 1.70 agc * 3. Neither the name of the University nor the names of its contributors
53 1.1 cgd * may be used to endorse or promote products derived from this software
54 1.1 cgd * without specific prior written permission.
55 1.1 cgd *
56 1.1 cgd * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
57 1.1 cgd * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
58 1.1 cgd * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
59 1.1 cgd * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
60 1.1 cgd * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
61 1.1 cgd * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
62 1.1 cgd * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
63 1.1 cgd * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
64 1.1 cgd * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
65 1.1 cgd * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
66 1.1 cgd * SUCH DAMAGE.
67 1.1 cgd *
68 1.26 fvdl * @(#)uipc_mbuf.c 8.4 (Berkeley) 2/14/95
69 1.1 cgd */
70 1.56 lukem
71 1.56 lukem #include <sys/cdefs.h>
72 1.100.2.16 yamt __KERNEL_RCSID(0, "$NetBSD: uipc_mbuf.c,v 1.100.2.16 2007/11/15 11:44:56 yamt Exp $");
73 1.69 martin
74 1.69 martin #include "opt_mbuftrace.h"
75 1.100.2.6 yamt #include "opt_ddb.h"
76 1.24 mrg
77 1.6 mycroft #include <sys/param.h>
78 1.6 mycroft #include <sys/systm.h>
79 1.6 mycroft #include <sys/proc.h>
80 1.6 mycroft #include <sys/malloc.h>
81 1.1 cgd #define MBTYPES
82 1.6 mycroft #include <sys/mbuf.h>
83 1.6 mycroft #include <sys/kernel.h>
84 1.6 mycroft #include <sys/syslog.h>
85 1.6 mycroft #include <sys/domain.h>
86 1.6 mycroft #include <sys/protosw.h>
87 1.28 thorpej #include <sys/pool.h>
88 1.27 matt #include <sys/socket.h>
89 1.55 simonb #include <sys/sysctl.h>
90 1.55 simonb
91 1.27 matt #include <net/if.h>
92 1.14 christos
93 1.65 thorpej #include <uvm/uvm.h>
94 1.23 mrg
95 1.100.2.16 yamt pool_cache_t mb_cache; /* mbuf cache */
96 1.100.2.16 yamt pool_cache_t mcl_cache; /* mbuf cluster cache */
97 1.53 thorpej
98 1.18 thorpej struct mbstat mbstat;
99 1.18 thorpej int max_linkhdr;
100 1.18 thorpej int max_protohdr;
101 1.18 thorpej int max_hdr;
102 1.18 thorpej int max_datalen;
103 1.18 thorpej
104 1.65 thorpej static int mb_ctor(void *, void *, int);
105 1.65 thorpej
106 1.100.2.6 yamt static void *mclpool_alloc(struct pool *, int);
107 1.100.2.6 yamt static void mclpool_release(struct pool *, void *);
108 1.58 thorpej
109 1.100.2.6 yamt static struct pool_allocator mclpool_allocator = {
110 1.100.2.11 yamt .pa_alloc = mclpool_alloc,
111 1.100.2.11 yamt .pa_free = mclpool_release,
112 1.58 thorpej };
113 1.58 thorpej
114 1.79 junyoung static struct mbuf *m_copym0(struct mbuf *, int, int, int, int);
115 1.85 yamt static struct mbuf *m_split0(struct mbuf *, int, int, int);
116 1.86 yamt static int m_copyback0(struct mbuf **, int, int, const void *, int, int);
117 1.85 yamt
118 1.85 yamt /* flags for m_copyback0 */
119 1.85 yamt #define M_COPYBACK0_COPYBACK 0x0001 /* copyback from cp */
120 1.85 yamt #define M_COPYBACK0_PRESERVE 0x0002 /* preserve original data */
121 1.85 yamt #define M_COPYBACK0_COW 0x0004 /* do copy-on-write */
122 1.85 yamt #define M_COPYBACK0_EXTEND 0x0008 /* extend chain */
123 1.28 thorpej
124 1.100.2.6 yamt static const char mclpool_warnmsg[] =
125 1.42 thorpej "WARNING: mclpool limit reached; increase NMBCLUSTERS";
126 1.63 thorpej
127 1.63 thorpej MALLOC_DEFINE(M_MBUF, "mbuf", "mbuf");
128 1.42 thorpej
129 1.64 matt #ifdef MBUFTRACE
130 1.64 matt struct mownerhead mowners = LIST_HEAD_INITIALIZER(mowners);
131 1.64 matt struct mowner unknown_mowners[] = {
132 1.100.2.11 yamt MOWNER_INIT("unknown", "free"),
133 1.100.2.11 yamt MOWNER_INIT("unknown", "data"),
134 1.100.2.11 yamt MOWNER_INIT("unknown", "header"),
135 1.100.2.11 yamt MOWNER_INIT("unknown", "soname"),
136 1.100.2.11 yamt MOWNER_INIT("unknown", "soopts"),
137 1.100.2.11 yamt MOWNER_INIT("unknown", "ftable"),
138 1.100.2.11 yamt MOWNER_INIT("unknown", "control"),
139 1.100.2.11 yamt MOWNER_INIT("unknown", "oobdata"),
140 1.64 matt };
141 1.100.2.11 yamt struct mowner revoked_mowner = MOWNER_INIT("revoked", "");
142 1.64 matt #endif
143 1.64 matt
144 1.100.2.15 yamt #define MEXT_LOCK(m) mutex_enter(&(m)->m_ext.ext_lock)
145 1.100.2.15 yamt #define MEXT_UNLOCK(m) mutex_exit(&(m)->m_ext.ext_lock)
146 1.100.2.7 yamt #define MEXT_ISEMBEDDED(m) ((m)->m_ext_ref == (m))
147 1.100.2.7 yamt
148 1.100.2.8 yamt static inline void
149 1.100.2.8 yamt mcl_inc_reference(struct mbuf *m)
150 1.100.2.8 yamt {
151 1.100.2.8 yamt
152 1.100.2.10 yamt MEXT_LOCK(m);
153 1.100.2.10 yamt m->m_ext.ext_refcnt++;
154 1.100.2.10 yamt MEXT_UNLOCK(m);
155 1.100.2.8 yamt }
156 1.100.2.8 yamt
157 1.100.2.12 yamt static inline bool
158 1.100.2.8 yamt mcl_dec_and_test_reference(struct mbuf *m)
159 1.100.2.8 yamt {
160 1.100.2.12 yamt bool gotzero;
161 1.100.2.8 yamt
162 1.100.2.8 yamt MEXT_LOCK(m);
163 1.100.2.10 yamt KASSERT(m->m_ext.ext_refcnt > 0);
164 1.100.2.8 yamt m->m_ext.ext_refcnt--;
165 1.100.2.10 yamt gotzero = (m->m_ext.ext_refcnt == 0);
166 1.100.2.8 yamt MEXT_UNLOCK(m);
167 1.100.2.8 yamt
168 1.100.2.10 yamt return gotzero;
169 1.100.2.8 yamt }
170 1.100.2.7 yamt
171 1.100.2.10 yamt #define MCLADDREFERENCE(o, n) \
172 1.100.2.7 yamt do { \
173 1.100.2.7 yamt KASSERT(((o)->m_flags & M_EXT) != 0); \
174 1.100.2.7 yamt KASSERT(((n)->m_flags & M_EXT) == 0); \
175 1.100.2.7 yamt KASSERT((o)->m_ext.ext_refcnt >= 1); \
176 1.100.2.7 yamt (n)->m_flags |= ((o)->m_flags & M_EXTCOPYFLAGS); \
177 1.100.2.8 yamt mcl_inc_reference((o)); \
178 1.100.2.7 yamt (n)->m_ext_ref = (o)->m_ext_ref; \
179 1.100.2.7 yamt _MOWNERREF((n), (n)->m_flags); \
180 1.100.2.7 yamt MCLREFDEBUGN((n), __FILE__, __LINE__); \
181 1.100.2.7 yamt } while (/* CONSTCOND */ 0)
182 1.100.2.7 yamt
183 1.28 thorpej /*
184 1.68 simonb * Initialize the mbuf allocator.
185 1.28 thorpej */
186 1.4 jtc void
187 1.62 thorpej mbinit(void)
188 1.1 cgd {
189 1.65 thorpej
190 1.65 thorpej KASSERT(sizeof(struct _m_ext) <= MHLEN);
191 1.67 simonb KASSERT(sizeof(struct mbuf) == MSIZE);
192 1.65 thorpej
193 1.100.2.6 yamt mclpool_allocator.pa_backingmap = mb_map;
194 1.53 thorpej
195 1.100.2.16 yamt mb_cache = pool_cache_init(msize, 0, 0, 0, "mbpl",
196 1.100.2.16 yamt NULL, IPL_VM, mb_ctor, NULL, NULL);
197 1.100.2.16 yamt KASSERT(mb_cache != NULL);
198 1.100.2.16 yamt
199 1.100.2.16 yamt mcl_cache = pool_cache_init(mclbytes, 0, 0, 0, "mclpl",
200 1.100.2.16 yamt &mclpool_allocator, IPL_VM, NULL, NULL, NULL);
201 1.100.2.16 yamt KASSERT(mcl_cache != NULL);
202 1.59 thorpej
203 1.100.2.16 yamt pool_cache_set_drain_hook(mb_cache, m_reclaim, NULL);
204 1.100.2.16 yamt pool_cache_set_drain_hook(mcl_cache, m_reclaim, NULL);
205 1.37 thorpej
206 1.37 thorpej /*
207 1.39 thorpej * Set the hard limit on the mclpool to the number of
208 1.39 thorpej * mbuf clusters the kernel is to support. Log the limit
209 1.39 thorpej * reached message max once a minute.
210 1.39 thorpej */
211 1.100.2.16 yamt pool_cache_sethardlimit(mcl_cache, nmbclusters, mclpool_warnmsg, 60);
212 1.42 thorpej
213 1.39 thorpej /*
214 1.42 thorpej * Set a low water mark for both mbufs and clusters. This should
215 1.42 thorpej * help ensure that they can be allocated in a memory starvation
216 1.42 thorpej * situation. This is important for e.g. diskless systems which
217 1.42 thorpej * must allocate mbufs in order for the pagedaemon to clean pages.
218 1.37 thorpej */
219 1.100.2.16 yamt pool_cache_setlowat(mb_cache, mblowat);
220 1.100.2.16 yamt pool_cache_setlowat(mcl_cache, mcllowat);
221 1.64 matt
222 1.64 matt #ifdef MBUFTRACE
223 1.64 matt {
224 1.64 matt /*
225 1.64 matt * Attach the unknown mowners.
226 1.64 matt */
227 1.64 matt int i;
228 1.64 matt MOWNER_ATTACH(&revoked_mowner);
229 1.64 matt for (i = sizeof(unknown_mowners)/sizeof(unknown_mowners[0]);
230 1.64 matt i-- > 0; )
231 1.64 matt MOWNER_ATTACH(&unknown_mowners[i]);
232 1.64 matt }
233 1.64 matt #endif
234 1.42 thorpej }
235 1.42 thorpej
236 1.75 atatat /*
237 1.75 atatat * sysctl helper routine for the kern.mbuf subtree. nmbclusters may
238 1.75 atatat * or may not be writable, and mblowat and mcllowat need range
239 1.75 atatat * checking and pool tweaking after being reset.
240 1.75 atatat */
241 1.75 atatat static int
242 1.75 atatat sysctl_kern_mbuf(SYSCTLFN_ARGS)
243 1.42 thorpej {
244 1.42 thorpej int error, newval;
245 1.75 atatat struct sysctlnode node;
246 1.42 thorpej
247 1.75 atatat node = *rnode;
248 1.75 atatat node.sysctl_data = &newval;
249 1.75 atatat switch (rnode->sysctl_num) {
250 1.42 thorpej case MBUF_NMBCLUSTERS:
251 1.76 atatat if (mb_map != NULL) {
252 1.80 atatat node.sysctl_flags &= ~CTLFLAG_READWRITE;
253 1.80 atatat node.sysctl_flags |= CTLFLAG_READONLY;
254 1.75 atatat }
255 1.75 atatat /* FALLTHROUGH */
256 1.42 thorpej case MBUF_MBLOWAT:
257 1.42 thorpej case MBUF_MCLLOWAT:
258 1.75 atatat newval = *(int*)rnode->sysctl_data;
259 1.75 atatat break;
260 1.75 atatat default:
261 1.75 atatat return (EOPNOTSUPP);
262 1.75 atatat }
263 1.75 atatat
264 1.75 atatat error = sysctl_lookup(SYSCTLFN_CALL(&node));
265 1.75 atatat if (error || newp == NULL)
266 1.42 thorpej return (error);
267 1.75 atatat if (newval < 0)
268 1.75 atatat return (EINVAL);
269 1.75 atatat
270 1.75 atatat switch (node.sysctl_num) {
271 1.75 atatat case MBUF_NMBCLUSTERS:
272 1.75 atatat if (newval < nmbclusters)
273 1.75 atatat return (EINVAL);
274 1.75 atatat nmbclusters = newval;
275 1.100.2.16 yamt pool_cache_sethardlimit(mcl_cache, nmbclusters,
276 1.100.2.16 yamt mclpool_warnmsg, 60);
277 1.75 atatat break;
278 1.75 atatat case MBUF_MBLOWAT:
279 1.75 atatat mblowat = newval;
280 1.100.2.16 yamt pool_cache_setlowat(mb_cache, mblowat);
281 1.75 atatat break;
282 1.75 atatat case MBUF_MCLLOWAT:
283 1.76 atatat mcllowat = newval;
284 1.100.2.16 yamt pool_cache_setlowat(mcl_cache, mcllowat);
285 1.75 atatat break;
286 1.75 atatat }
287 1.75 atatat
288 1.75 atatat return (0);
289 1.75 atatat }
290 1.75 atatat
291 1.64 matt #ifdef MBUFTRACE
292 1.75 atatat static int
293 1.75 atatat sysctl_kern_mbuf_mowners(SYSCTLFN_ARGS)
294 1.75 atatat {
295 1.75 atatat struct mowner *mo;
296 1.75 atatat size_t len = 0;
297 1.75 atatat int error = 0;
298 1.75 atatat
299 1.75 atatat if (namelen != 0)
300 1.75 atatat return (EINVAL);
301 1.75 atatat if (newp != NULL)
302 1.75 atatat return (EPERM);
303 1.75 atatat
304 1.75 atatat LIST_FOREACH(mo, &mowners, mo_link) {
305 1.75 atatat if (oldp != NULL) {
306 1.75 atatat if (*oldlenp - len < sizeof(*mo)) {
307 1.75 atatat error = ENOMEM;
308 1.75 atatat break;
309 1.75 atatat }
310 1.100.2.13 yamt error = copyout(mo, (char *)oldp + len, sizeof(*mo));
311 1.75 atatat if (error)
312 1.75 atatat break;
313 1.64 matt }
314 1.75 atatat len += sizeof(*mo);
315 1.75 atatat }
316 1.75 atatat
317 1.75 atatat if (error == 0)
318 1.64 matt *oldlenp = len;
319 1.75 atatat
320 1.75 atatat return (error);
321 1.75 atatat }
322 1.75 atatat #endif /* MBUFTRACE */
323 1.75 atatat
324 1.75 atatat SYSCTL_SETUP(sysctl_kern_mbuf_setup, "sysctl kern.mbuf subtree setup")
325 1.75 atatat {
326 1.75 atatat
327 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
328 1.80 atatat CTLFLAG_PERMANENT,
329 1.75 atatat CTLTYPE_NODE, "kern", NULL,
330 1.75 atatat NULL, 0, NULL, 0,
331 1.75 atatat CTL_KERN, CTL_EOL);
332 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
333 1.80 atatat CTLFLAG_PERMANENT,
334 1.82 atatat CTLTYPE_NODE, "mbuf",
335 1.82 atatat SYSCTL_DESCR("mbuf control variables"),
336 1.75 atatat NULL, 0, NULL, 0,
337 1.75 atatat CTL_KERN, KERN_MBUF, CTL_EOL);
338 1.75 atatat
339 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
340 1.80 atatat CTLFLAG_PERMANENT|CTLFLAG_IMMEDIATE,
341 1.82 atatat CTLTYPE_INT, "msize",
342 1.82 atatat SYSCTL_DESCR("mbuf base size"),
343 1.75 atatat NULL, msize, NULL, 0,
344 1.75 atatat CTL_KERN, KERN_MBUF, MBUF_MSIZE, CTL_EOL);
345 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
346 1.80 atatat CTLFLAG_PERMANENT|CTLFLAG_IMMEDIATE,
347 1.82 atatat CTLTYPE_INT, "mclbytes",
348 1.82 atatat SYSCTL_DESCR("mbuf cluster size"),
349 1.75 atatat NULL, mclbytes, NULL, 0,
350 1.75 atatat CTL_KERN, KERN_MBUF, MBUF_MCLBYTES, CTL_EOL);
351 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
352 1.80 atatat CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
353 1.82 atatat CTLTYPE_INT, "nmbclusters",
354 1.82 atatat SYSCTL_DESCR("Limit on the number of mbuf clusters"),
355 1.75 atatat sysctl_kern_mbuf, 0, &nmbclusters, 0,
356 1.75 atatat CTL_KERN, KERN_MBUF, MBUF_NMBCLUSTERS, CTL_EOL);
357 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
358 1.80 atatat CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
359 1.82 atatat CTLTYPE_INT, "mblowat",
360 1.82 atatat SYSCTL_DESCR("mbuf low water mark"),
361 1.75 atatat sysctl_kern_mbuf, 0, &mblowat, 0,
362 1.75 atatat CTL_KERN, KERN_MBUF, MBUF_MBLOWAT, CTL_EOL);
363 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
364 1.80 atatat CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
365 1.82 atatat CTLTYPE_INT, "mcllowat",
366 1.82 atatat SYSCTL_DESCR("mbuf cluster low water mark"),
367 1.75 atatat sysctl_kern_mbuf, 0, &mcllowat, 0,
368 1.75 atatat CTL_KERN, KERN_MBUF, MBUF_MCLLOWAT, CTL_EOL);
369 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
370 1.80 atatat CTLFLAG_PERMANENT,
371 1.82 atatat CTLTYPE_STRUCT, "stats",
372 1.82 atatat SYSCTL_DESCR("mbuf allocation statistics"),
373 1.75 atatat NULL, 0, &mbstat, sizeof(mbstat),
374 1.75 atatat CTL_KERN, KERN_MBUF, MBUF_STATS, CTL_EOL);
375 1.75 atatat #ifdef MBUFTRACE
376 1.80 atatat sysctl_createv(clog, 0, NULL, NULL,
377 1.80 atatat CTLFLAG_PERMANENT,
378 1.82 atatat CTLTYPE_STRUCT, "mowners",
379 1.82 atatat SYSCTL_DESCR("Information about mbuf owners"),
380 1.75 atatat sysctl_kern_mbuf_mowners, 0, NULL, 0,
381 1.75 atatat CTL_KERN, KERN_MBUF, MBUF_MOWNERS, CTL_EOL);
382 1.75 atatat #endif /* MBUFTRACE */
383 1.28 thorpej }
384 1.28 thorpej
385 1.100.2.6 yamt static void *
386 1.62 thorpej mclpool_alloc(struct pool *pp, int flags)
387 1.28 thorpej {
388 1.100.2.12 yamt bool waitok = (flags & PR_WAITOK) ? true : false;
389 1.28 thorpej
390 1.93 yamt return ((void *)uvm_km_alloc_poolpage(mb_map, waitok));
391 1.1 cgd }
392 1.1 cgd
393 1.100.2.6 yamt static void
394 1.62 thorpej mclpool_release(struct pool *pp, void *v)
395 1.1 cgd {
396 1.1 cgd
397 1.93 yamt uvm_km_free_poolpage(mb_map, (vaddr_t)v);
398 1.65 thorpej }
399 1.65 thorpej
400 1.65 thorpej /*ARGSUSED*/
401 1.65 thorpej static int
402 1.65 thorpej mb_ctor(void *arg, void *object, int flags)
403 1.65 thorpej {
404 1.65 thorpej struct mbuf *m = object;
405 1.65 thorpej
406 1.65 thorpej #ifdef POOL_VTOPHYS
407 1.65 thorpej m->m_paddr = POOL_VTOPHYS(m);
408 1.65 thorpej #else
409 1.65 thorpej m->m_paddr = M_PADDR_INVALID;
410 1.65 thorpej #endif
411 1.65 thorpej return (0);
412 1.1 cgd }
413 1.1 cgd
414 1.14 christos void
415 1.59 thorpej m_reclaim(void *arg, int flags)
416 1.1 cgd {
417 1.27 matt struct domain *dp;
418 1.81 matt const struct protosw *pr;
419 1.27 matt struct ifnet *ifp;
420 1.100.2.16 yamt int s;
421 1.1 cgd
422 1.100.2.16 yamt KERNEL_LOCK(1, NULL);
423 1.100.2.16 yamt s = splvm();
424 1.91 matt DOMAIN_FOREACH(dp) {
425 1.33 thorpej for (pr = dp->dom_protosw;
426 1.33 thorpej pr < dp->dom_protoswNPROTOSW; pr++)
427 1.33 thorpej if (pr->pr_drain)
428 1.33 thorpej (*pr->pr_drain)();
429 1.91 matt }
430 1.92 matt IFNET_FOREACH(ifp) {
431 1.27 matt if (ifp->if_drain)
432 1.27 matt (*ifp->if_drain)(ifp);
433 1.92 matt }
434 1.1 cgd splx(s);
435 1.1 cgd mbstat.m_drain++;
436 1.100.2.16 yamt KERNEL_UNLOCK_ONE(NULL);
437 1.1 cgd }
438 1.1 cgd
439 1.1 cgd /*
440 1.1 cgd * Space allocation routines.
441 1.1 cgd * These are also available as macros
442 1.1 cgd * for critical paths.
443 1.1 cgd */
444 1.1 cgd struct mbuf *
445 1.62 thorpej m_get(int nowait, int type)
446 1.1 cgd {
447 1.27 matt struct mbuf *m;
448 1.1 cgd
449 1.5 cgd MGET(m, nowait, type);
450 1.1 cgd return (m);
451 1.1 cgd }
452 1.1 cgd
453 1.1 cgd struct mbuf *
454 1.62 thorpej m_gethdr(int nowait, int type)
455 1.1 cgd {
456 1.27 matt struct mbuf *m;
457 1.1 cgd
458 1.5 cgd MGETHDR(m, nowait, type);
459 1.1 cgd return (m);
460 1.1 cgd }
461 1.1 cgd
462 1.1 cgd struct mbuf *
463 1.62 thorpej m_getclr(int nowait, int type)
464 1.1 cgd {
465 1.27 matt struct mbuf *m;
466 1.1 cgd
467 1.5 cgd MGET(m, nowait, type);
468 1.1 cgd if (m == 0)
469 1.71 simonb return (NULL);
470 1.100.2.13 yamt memset(mtod(m, void *), 0, MLEN);
471 1.1 cgd return (m);
472 1.1 cgd }
473 1.1 cgd
474 1.64 matt void
475 1.64 matt m_clget(struct mbuf *m, int nowait)
476 1.64 matt {
477 1.71 simonb
478 1.64 matt MCLGET(m, nowait);
479 1.64 matt }
480 1.64 matt
481 1.1 cgd struct mbuf *
482 1.62 thorpej m_free(struct mbuf *m)
483 1.1 cgd {
484 1.27 matt struct mbuf *n;
485 1.1 cgd
486 1.1 cgd MFREE(m, n);
487 1.1 cgd return (n);
488 1.1 cgd }
489 1.1 cgd
490 1.9 mycroft void
491 1.62 thorpej m_freem(struct mbuf *m)
492 1.1 cgd {
493 1.27 matt struct mbuf *n;
494 1.1 cgd
495 1.1 cgd if (m == NULL)
496 1.1 cgd return;
497 1.1 cgd do {
498 1.1 cgd MFREE(m, n);
499 1.18 thorpej m = n;
500 1.18 thorpej } while (m);
501 1.1 cgd }
502 1.1 cgd
503 1.64 matt #ifdef MBUFTRACE
504 1.83 jonathan /*
505 1.83 jonathan * Walk a chain of mbufs, claiming ownership of each mbuf in the chain.
506 1.83 jonathan */
507 1.64 matt void
508 1.83 jonathan m_claimm(struct mbuf *m, struct mowner *mo)
509 1.64 matt {
510 1.71 simonb
511 1.64 matt for (; m != NULL; m = m->m_next)
512 1.64 matt MCLAIM(m, mo);
513 1.64 matt }
514 1.64 matt #endif
515 1.64 matt
516 1.1 cgd /*
517 1.1 cgd * Mbuffer utility routines.
518 1.1 cgd */
519 1.1 cgd
520 1.1 cgd /*
521 1.1 cgd * Lesser-used path for M_PREPEND:
522 1.1 cgd * allocate new mbuf to prepend to chain,
523 1.1 cgd * copy junk along.
524 1.1 cgd */
525 1.1 cgd struct mbuf *
526 1.62 thorpej m_prepend(struct mbuf *m, int len, int how)
527 1.1 cgd {
528 1.1 cgd struct mbuf *mn;
529 1.1 cgd
530 1.9 mycroft MGET(mn, how, m->m_type);
531 1.1 cgd if (mn == (struct mbuf *)NULL) {
532 1.1 cgd m_freem(m);
533 1.1 cgd return ((struct mbuf *)NULL);
534 1.1 cgd }
535 1.1 cgd if (m->m_flags & M_PKTHDR) {
536 1.100.2.6 yamt M_MOVE_PKTHDR(mn, m);
537 1.64 matt } else {
538 1.64 matt MCLAIM(mn, m->m_owner);
539 1.1 cgd }
540 1.1 cgd mn->m_next = m;
541 1.1 cgd m = mn;
542 1.1 cgd if (len < MHLEN)
543 1.1 cgd MH_ALIGN(m, len);
544 1.1 cgd m->m_len = len;
545 1.1 cgd return (m);
546 1.1 cgd }
547 1.1 cgd
548 1.1 cgd /*
549 1.1 cgd * Make a copy of an mbuf chain starting "off0" bytes from the beginning,
550 1.1 cgd * continuing for "len" bytes. If len is M_COPYALL, copy to end of mbuf.
551 1.1 cgd * The wait parameter is a choice of M_WAIT/M_DONTWAIT from caller.
552 1.1 cgd */
553 1.1 cgd int MCFail;
554 1.1 cgd
555 1.1 cgd struct mbuf *
556 1.62 thorpej m_copym(struct mbuf *m, int off0, int len, int wait)
557 1.1 cgd {
558 1.71 simonb
559 1.44 itojun return m_copym0(m, off0, len, wait, 0); /* shallow copy on M_EXT */
560 1.44 itojun }
561 1.44 itojun
562 1.44 itojun struct mbuf *
563 1.62 thorpej m_dup(struct mbuf *m, int off0, int len, int wait)
564 1.44 itojun {
565 1.71 simonb
566 1.44 itojun return m_copym0(m, off0, len, wait, 1); /* deep copy */
567 1.44 itojun }
568 1.44 itojun
569 1.44 itojun static struct mbuf *
570 1.62 thorpej m_copym0(struct mbuf *m, int off0, int len, int wait, int deep)
571 1.44 itojun {
572 1.27 matt struct mbuf *n, **np;
573 1.27 matt int off = off0;
574 1.1 cgd struct mbuf *top;
575 1.1 cgd int copyhdr = 0;
576 1.1 cgd
577 1.1 cgd if (off < 0 || len < 0)
578 1.43 thorpej panic("m_copym: off %d, len %d", off, len);
579 1.1 cgd if (off == 0 && m->m_flags & M_PKTHDR)
580 1.1 cgd copyhdr = 1;
581 1.1 cgd while (off > 0) {
582 1.1 cgd if (m == 0)
583 1.90 matt panic("m_copym: m == 0, off %d", off);
584 1.1 cgd if (off < m->m_len)
585 1.1 cgd break;
586 1.1 cgd off -= m->m_len;
587 1.1 cgd m = m->m_next;
588 1.1 cgd }
589 1.1 cgd np = ⊤
590 1.1 cgd top = 0;
591 1.1 cgd while (len > 0) {
592 1.1 cgd if (m == 0) {
593 1.1 cgd if (len != M_COPYALL)
594 1.90 matt panic("m_copym: m == 0, len %d [!COPYALL]",
595 1.90 matt len);
596 1.1 cgd break;
597 1.1 cgd }
598 1.1 cgd MGET(n, wait, m->m_type);
599 1.1 cgd *np = n;
600 1.1 cgd if (n == 0)
601 1.1 cgd goto nospace;
602 1.64 matt MCLAIM(n, m->m_owner);
603 1.1 cgd if (copyhdr) {
604 1.1 cgd M_COPY_PKTHDR(n, m);
605 1.1 cgd if (len == M_COPYALL)
606 1.1 cgd n->m_pkthdr.len -= off0;
607 1.1 cgd else
608 1.1 cgd n->m_pkthdr.len = len;
609 1.1 cgd copyhdr = 0;
610 1.1 cgd }
611 1.9 mycroft n->m_len = min(len, m->m_len - off);
612 1.1 cgd if (m->m_flags & M_EXT) {
613 1.44 itojun if (!deep) {
614 1.44 itojun n->m_data = m->m_data + off;
615 1.44 itojun MCLADDREFERENCE(m, n);
616 1.44 itojun } else {
617 1.48 itojun /*
618 1.50 itojun * we are unsure about the way m was allocated.
619 1.50 itojun * copy into multiple MCLBYTES cluster mbufs.
620 1.48 itojun */
621 1.44 itojun MCLGET(n, wait);
622 1.50 itojun n->m_len = 0;
623 1.50 itojun n->m_len = M_TRAILINGSPACE(n);
624 1.50 itojun n->m_len = min(n->m_len, len);
625 1.50 itojun n->m_len = min(n->m_len, m->m_len - off);
626 1.100.2.13 yamt memcpy(mtod(n, void *), mtod(m, char *) + off,
627 1.44 itojun (unsigned)n->m_len);
628 1.44 itojun }
629 1.1 cgd } else
630 1.100.2.13 yamt memcpy(mtod(n, void *), mtod(m, char *) + off,
631 1.1 cgd (unsigned)n->m_len);
632 1.1 cgd if (len != M_COPYALL)
633 1.1 cgd len -= n->m_len;
634 1.50 itojun off += n->m_len;
635 1.50 itojun #ifdef DIAGNOSTIC
636 1.50 itojun if (off > m->m_len)
637 1.50 itojun panic("m_copym0 overrun");
638 1.50 itojun #endif
639 1.50 itojun if (off == m->m_len) {
640 1.50 itojun m = m->m_next;
641 1.50 itojun off = 0;
642 1.50 itojun }
643 1.1 cgd np = &n->m_next;
644 1.1 cgd }
645 1.1 cgd if (top == 0)
646 1.1 cgd MCFail++;
647 1.1 cgd return (top);
648 1.1 cgd nospace:
649 1.1 cgd m_freem(top);
650 1.1 cgd MCFail++;
651 1.71 simonb return (NULL);
652 1.1 cgd }
653 1.1 cgd
654 1.1 cgd /*
655 1.18 thorpej * Copy an entire packet, including header (which must be present).
656 1.18 thorpej * An optimization of the common case `m_copym(m, 0, M_COPYALL, how)'.
657 1.18 thorpej */
658 1.18 thorpej struct mbuf *
659 1.62 thorpej m_copypacket(struct mbuf *m, int how)
660 1.18 thorpej {
661 1.18 thorpej struct mbuf *top, *n, *o;
662 1.18 thorpej
663 1.18 thorpej MGET(n, how, m->m_type);
664 1.18 thorpej top = n;
665 1.18 thorpej if (!n)
666 1.18 thorpej goto nospace;
667 1.18 thorpej
668 1.64 matt MCLAIM(n, m->m_owner);
669 1.18 thorpej M_COPY_PKTHDR(n, m);
670 1.18 thorpej n->m_len = m->m_len;
671 1.18 thorpej if (m->m_flags & M_EXT) {
672 1.18 thorpej n->m_data = m->m_data;
673 1.18 thorpej MCLADDREFERENCE(m, n);
674 1.18 thorpej } else {
675 1.30 perry memcpy(mtod(n, char *), mtod(m, char *), n->m_len);
676 1.18 thorpej }
677 1.18 thorpej
678 1.18 thorpej m = m->m_next;
679 1.18 thorpej while (m) {
680 1.18 thorpej MGET(o, how, m->m_type);
681 1.18 thorpej if (!o)
682 1.18 thorpej goto nospace;
683 1.18 thorpej
684 1.64 matt MCLAIM(o, m->m_owner);
685 1.18 thorpej n->m_next = o;
686 1.18 thorpej n = n->m_next;
687 1.18 thorpej
688 1.18 thorpej n->m_len = m->m_len;
689 1.18 thorpej if (m->m_flags & M_EXT) {
690 1.18 thorpej n->m_data = m->m_data;
691 1.18 thorpej MCLADDREFERENCE(m, n);
692 1.18 thorpej } else {
693 1.30 perry memcpy(mtod(n, char *), mtod(m, char *), n->m_len);
694 1.18 thorpej }
695 1.18 thorpej
696 1.18 thorpej m = m->m_next;
697 1.18 thorpej }
698 1.18 thorpej return top;
699 1.18 thorpej nospace:
700 1.18 thorpej m_freem(top);
701 1.18 thorpej MCFail++;
702 1.71 simonb return NULL;
703 1.18 thorpej }
704 1.18 thorpej
705 1.18 thorpej /*
706 1.1 cgd * Copy data from an mbuf chain starting "off" bytes from the beginning,
707 1.1 cgd * continuing for "len" bytes, into the indicated buffer.
708 1.1 cgd */
709 1.14 christos void
710 1.100 martin m_copydata(struct mbuf *m, int off, int len, void *vp)
711 1.1 cgd {
712 1.94 tron unsigned count;
713 1.100.2.13 yamt void * cp = vp;
714 1.1 cgd
715 1.1 cgd if (off < 0 || len < 0)
716 1.90 matt panic("m_copydata: off %d, len %d", off, len);
717 1.1 cgd while (off > 0) {
718 1.94 tron if (m == NULL)
719 1.94 tron panic("m_copydata: m == NULL, off %d", off);
720 1.1 cgd if (off < m->m_len)
721 1.1 cgd break;
722 1.1 cgd off -= m->m_len;
723 1.1 cgd m = m->m_next;
724 1.1 cgd }
725 1.1 cgd while (len > 0) {
726 1.94 tron if (m == NULL)
727 1.94 tron panic("m_copydata: m == NULL, len %d", len);
728 1.9 mycroft count = min(m->m_len - off, len);
729 1.100.2.13 yamt memcpy(cp, mtod(m, char *) + off, count);
730 1.1 cgd len -= count;
731 1.100.2.13 yamt cp = (char *)cp + count;
732 1.1 cgd off = 0;
733 1.1 cgd m = m->m_next;
734 1.1 cgd }
735 1.1 cgd }
736 1.1 cgd
737 1.1 cgd /*
738 1.1 cgd * Concatenate mbuf chain n to m.
739 1.72 itojun * n might be copied into m (when n->m_len is small), therefore data portion of
740 1.72 itojun * n could be copied into an mbuf of different mbuf type.
741 1.1 cgd * Any m_pkthdr is not updated.
742 1.1 cgd */
743 1.14 christos void
744 1.62 thorpej m_cat(struct mbuf *m, struct mbuf *n)
745 1.1 cgd {
746 1.73 yamt
747 1.1 cgd while (m->m_next)
748 1.1 cgd m = m->m_next;
749 1.1 cgd while (n) {
750 1.77 itojun if (M_READONLY(m) || n->m_len > M_TRAILINGSPACE(m)) {
751 1.1 cgd /* just join the two chains */
752 1.1 cgd m->m_next = n;
753 1.1 cgd return;
754 1.1 cgd }
755 1.1 cgd /* splat the data from one into the other */
756 1.100.2.13 yamt memcpy(mtod(m, char *) + m->m_len, mtod(n, void *),
757 1.1 cgd (u_int)n->m_len);
758 1.1 cgd m->m_len += n->m_len;
759 1.1 cgd n = m_free(n);
760 1.1 cgd }
761 1.1 cgd }
762 1.1 cgd
763 1.11 mycroft void
764 1.62 thorpej m_adj(struct mbuf *mp, int req_len)
765 1.1 cgd {
766 1.27 matt int len = req_len;
767 1.27 matt struct mbuf *m;
768 1.27 matt int count;
769 1.1 cgd
770 1.1 cgd if ((m = mp) == NULL)
771 1.1 cgd return;
772 1.1 cgd if (len >= 0) {
773 1.1 cgd /*
774 1.1 cgd * Trim from head.
775 1.1 cgd */
776 1.1 cgd while (m != NULL && len > 0) {
777 1.1 cgd if (m->m_len <= len) {
778 1.1 cgd len -= m->m_len;
779 1.1 cgd m->m_len = 0;
780 1.1 cgd m = m->m_next;
781 1.1 cgd } else {
782 1.1 cgd m->m_len -= len;
783 1.1 cgd m->m_data += len;
784 1.1 cgd len = 0;
785 1.1 cgd }
786 1.1 cgd }
787 1.1 cgd m = mp;
788 1.1 cgd if (mp->m_flags & M_PKTHDR)
789 1.1 cgd m->m_pkthdr.len -= (req_len - len);
790 1.1 cgd } else {
791 1.1 cgd /*
792 1.1 cgd * Trim from tail. Scan the mbuf chain,
793 1.1 cgd * calculating its length and finding the last mbuf.
794 1.1 cgd * If the adjustment only affects this mbuf, then just
795 1.1 cgd * adjust and return. Otherwise, rescan and truncate
796 1.1 cgd * after the remaining size.
797 1.1 cgd */
798 1.1 cgd len = -len;
799 1.1 cgd count = 0;
800 1.1 cgd for (;;) {
801 1.1 cgd count += m->m_len;
802 1.1 cgd if (m->m_next == (struct mbuf *)0)
803 1.1 cgd break;
804 1.1 cgd m = m->m_next;
805 1.1 cgd }
806 1.1 cgd if (m->m_len >= len) {
807 1.1 cgd m->m_len -= len;
808 1.8 deraadt if (mp->m_flags & M_PKTHDR)
809 1.8 deraadt mp->m_pkthdr.len -= len;
810 1.1 cgd return;
811 1.1 cgd }
812 1.1 cgd count -= len;
813 1.1 cgd if (count < 0)
814 1.1 cgd count = 0;
815 1.1 cgd /*
816 1.1 cgd * Correct length for chain is "count".
817 1.1 cgd * Find the mbuf with last data, adjust its length,
818 1.1 cgd * and toss data from remaining mbufs on chain.
819 1.1 cgd */
820 1.1 cgd m = mp;
821 1.1 cgd if (m->m_flags & M_PKTHDR)
822 1.1 cgd m->m_pkthdr.len = count;
823 1.1 cgd for (; m; m = m->m_next) {
824 1.1 cgd if (m->m_len >= count) {
825 1.1 cgd m->m_len = count;
826 1.1 cgd break;
827 1.1 cgd }
828 1.1 cgd count -= m->m_len;
829 1.1 cgd }
830 1.100.2.6 yamt if (m)
831 1.100.2.6 yamt while (m->m_next)
832 1.100.2.6 yamt (m = m->m_next)->m_len = 0;
833 1.1 cgd }
834 1.1 cgd }
835 1.1 cgd
836 1.1 cgd /*
837 1.100.2.6 yamt * Rearrange an mbuf chain so that len bytes are contiguous
838 1.1 cgd * and in the data area of an mbuf (so that mtod and dtom
839 1.1 cgd * will work for a structure of size len). Returns the resulting
840 1.1 cgd * mbuf chain on success, frees it and returns null on failure.
841 1.1 cgd * If there is room, it will add up to max_protohdr-len extra bytes to the
842 1.1 cgd * contiguous region in an attempt to avoid being called next time.
843 1.1 cgd */
844 1.1 cgd int MPFail;
845 1.1 cgd
846 1.1 cgd struct mbuf *
847 1.62 thorpej m_pullup(struct mbuf *n, int len)
848 1.1 cgd {
849 1.27 matt struct mbuf *m;
850 1.27 matt int count;
851 1.1 cgd int space;
852 1.1 cgd
853 1.1 cgd /*
854 1.1 cgd * If first mbuf has no cluster, and has room for len bytes
855 1.1 cgd * without shifting current data, pullup into it,
856 1.1 cgd * otherwise allocate a new mbuf to prepend to the chain.
857 1.1 cgd */
858 1.1 cgd if ((n->m_flags & M_EXT) == 0 &&
859 1.1 cgd n->m_data + len < &n->m_dat[MLEN] && n->m_next) {
860 1.1 cgd if (n->m_len >= len)
861 1.1 cgd return (n);
862 1.1 cgd m = n;
863 1.1 cgd n = n->m_next;
864 1.1 cgd len -= m->m_len;
865 1.1 cgd } else {
866 1.1 cgd if (len > MHLEN)
867 1.1 cgd goto bad;
868 1.1 cgd MGET(m, M_DONTWAIT, n->m_type);
869 1.1 cgd if (m == 0)
870 1.1 cgd goto bad;
871 1.64 matt MCLAIM(m, n->m_owner);
872 1.1 cgd m->m_len = 0;
873 1.1 cgd if (n->m_flags & M_PKTHDR) {
874 1.100.2.6 yamt M_MOVE_PKTHDR(m, n);
875 1.1 cgd }
876 1.1 cgd }
877 1.1 cgd space = &m->m_dat[MLEN] - (m->m_data + m->m_len);
878 1.1 cgd do {
879 1.1 cgd count = min(min(max(len, max_protohdr), space), n->m_len);
880 1.100.2.13 yamt memcpy(mtod(m, char *) + m->m_len, mtod(n, void *),
881 1.1 cgd (unsigned)count);
882 1.1 cgd len -= count;
883 1.1 cgd m->m_len += count;
884 1.1 cgd n->m_len -= count;
885 1.1 cgd space -= count;
886 1.1 cgd if (n->m_len)
887 1.1 cgd n->m_data += count;
888 1.1 cgd else
889 1.1 cgd n = m_free(n);
890 1.1 cgd } while (len > 0 && n);
891 1.1 cgd if (len > 0) {
892 1.1 cgd (void) m_free(m);
893 1.1 cgd goto bad;
894 1.1 cgd }
895 1.1 cgd m->m_next = n;
896 1.1 cgd return (m);
897 1.1 cgd bad:
898 1.1 cgd m_freem(n);
899 1.1 cgd MPFail++;
900 1.71 simonb return (NULL);
901 1.60 thorpej }
902 1.60 thorpej
903 1.60 thorpej /*
904 1.60 thorpej * Like m_pullup(), except a new mbuf is always allocated, and we allow
905 1.60 thorpej * the amount of empty space before the data in the new mbuf to be specified
906 1.60 thorpej * (in the event that the caller expects to prepend later).
907 1.60 thorpej */
908 1.60 thorpej int MSFail;
909 1.60 thorpej
910 1.60 thorpej struct mbuf *
911 1.60 thorpej m_copyup(struct mbuf *n, int len, int dstoff)
912 1.60 thorpej {
913 1.60 thorpej struct mbuf *m;
914 1.60 thorpej int count, space;
915 1.60 thorpej
916 1.60 thorpej if (len > (MHLEN - dstoff))
917 1.60 thorpej goto bad;
918 1.60 thorpej MGET(m, M_DONTWAIT, n->m_type);
919 1.60 thorpej if (m == NULL)
920 1.60 thorpej goto bad;
921 1.64 matt MCLAIM(m, n->m_owner);
922 1.60 thorpej m->m_len = 0;
923 1.60 thorpej if (n->m_flags & M_PKTHDR) {
924 1.100.2.6 yamt M_MOVE_PKTHDR(m, n);
925 1.60 thorpej }
926 1.60 thorpej m->m_data += dstoff;
927 1.60 thorpej space = &m->m_dat[MLEN] - (m->m_data + m->m_len);
928 1.60 thorpej do {
929 1.60 thorpej count = min(min(max(len, max_protohdr), space), n->m_len);
930 1.100.2.13 yamt memcpy(mtod(m, char *) + m->m_len, mtod(n, void *),
931 1.60 thorpej (unsigned)count);
932 1.60 thorpej len -= count;
933 1.60 thorpej m->m_len += count;
934 1.60 thorpej n->m_len -= count;
935 1.60 thorpej space -= count;
936 1.60 thorpej if (n->m_len)
937 1.60 thorpej n->m_data += count;
938 1.60 thorpej else
939 1.60 thorpej n = m_free(n);
940 1.60 thorpej } while (len > 0 && n);
941 1.60 thorpej if (len > 0) {
942 1.60 thorpej (void) m_free(m);
943 1.60 thorpej goto bad;
944 1.60 thorpej }
945 1.60 thorpej m->m_next = n;
946 1.60 thorpej return (m);
947 1.60 thorpej bad:
948 1.60 thorpej m_freem(n);
949 1.60 thorpej MSFail++;
950 1.60 thorpej return (NULL);
951 1.9 mycroft }
952 1.9 mycroft
953 1.9 mycroft /*
954 1.9 mycroft * Partition an mbuf chain in two pieces, returning the tail --
955 1.9 mycroft * all but the first len0 bytes. In case of failure, it returns NULL and
956 1.9 mycroft * attempts to restore the chain to its original state.
957 1.9 mycroft */
958 1.9 mycroft struct mbuf *
959 1.62 thorpej m_split(struct mbuf *m0, int len0, int wait)
960 1.9 mycroft {
961 1.85 yamt
962 1.87 yamt return m_split0(m0, len0, wait, 1);
963 1.85 yamt }
964 1.85 yamt
965 1.85 yamt static struct mbuf *
966 1.85 yamt m_split0(struct mbuf *m0, int len0, int wait, int copyhdr)
967 1.85 yamt {
968 1.27 matt struct mbuf *m, *n;
969 1.22 thorpej unsigned len = len0, remain, len_save;
970 1.9 mycroft
971 1.9 mycroft for (m = m0; m && len > m->m_len; m = m->m_next)
972 1.9 mycroft len -= m->m_len;
973 1.9 mycroft if (m == 0)
974 1.71 simonb return (NULL);
975 1.9 mycroft remain = m->m_len - len;
976 1.85 yamt if (copyhdr && (m0->m_flags & M_PKTHDR)) {
977 1.9 mycroft MGETHDR(n, wait, m0->m_type);
978 1.9 mycroft if (n == 0)
979 1.71 simonb return (NULL);
980 1.100.2.11 yamt MCLAIM(n, m0->m_owner);
981 1.9 mycroft n->m_pkthdr.rcvif = m0->m_pkthdr.rcvif;
982 1.9 mycroft n->m_pkthdr.len = m0->m_pkthdr.len - len0;
983 1.22 thorpej len_save = m0->m_pkthdr.len;
984 1.9 mycroft m0->m_pkthdr.len = len0;
985 1.9 mycroft if (m->m_flags & M_EXT)
986 1.9 mycroft goto extpacket;
987 1.9 mycroft if (remain > MHLEN) {
988 1.9 mycroft /* m can't be the lead packet */
989 1.9 mycroft MH_ALIGN(n, 0);
990 1.9 mycroft n->m_next = m_split(m, len, wait);
991 1.9 mycroft if (n->m_next == 0) {
992 1.9 mycroft (void) m_free(n);
993 1.22 thorpej m0->m_pkthdr.len = len_save;
994 1.71 simonb return (NULL);
995 1.9 mycroft } else
996 1.9 mycroft return (n);
997 1.9 mycroft } else
998 1.9 mycroft MH_ALIGN(n, remain);
999 1.9 mycroft } else if (remain == 0) {
1000 1.9 mycroft n = m->m_next;
1001 1.9 mycroft m->m_next = 0;
1002 1.9 mycroft return (n);
1003 1.9 mycroft } else {
1004 1.9 mycroft MGET(n, wait, m->m_type);
1005 1.9 mycroft if (n == 0)
1006 1.71 simonb return (NULL);
1007 1.64 matt MCLAIM(n, m->m_owner);
1008 1.9 mycroft M_ALIGN(n, remain);
1009 1.9 mycroft }
1010 1.9 mycroft extpacket:
1011 1.9 mycroft if (m->m_flags & M_EXT) {
1012 1.9 mycroft n->m_data = m->m_data + len;
1013 1.100.2.1 yamt MCLADDREFERENCE(m, n);
1014 1.9 mycroft } else {
1015 1.100.2.13 yamt memcpy(mtod(n, void *), mtod(m, char *) + len, remain);
1016 1.9 mycroft }
1017 1.9 mycroft n->m_len = remain;
1018 1.9 mycroft m->m_len = len;
1019 1.9 mycroft n->m_next = m->m_next;
1020 1.9 mycroft m->m_next = 0;
1021 1.9 mycroft return (n);
1022 1.9 mycroft }
1023 1.9 mycroft /*
1024 1.9 mycroft * Routine to copy from device local memory into mbufs.
1025 1.9 mycroft */
1026 1.9 mycroft struct mbuf *
1027 1.62 thorpej m_devget(char *buf, int totlen, int off0, struct ifnet *ifp,
1028 1.62 thorpej void (*copy)(const void *from, void *to, size_t len))
1029 1.9 mycroft {
1030 1.27 matt struct mbuf *m;
1031 1.9 mycroft struct mbuf *top = 0, **mp = ⊤
1032 1.27 matt int off = off0, len;
1033 1.27 matt char *cp;
1034 1.9 mycroft char *epkt;
1035 1.9 mycroft
1036 1.9 mycroft cp = buf;
1037 1.9 mycroft epkt = cp + totlen;
1038 1.9 mycroft if (off) {
1039 1.13 cgd /*
1040 1.13 cgd * If 'off' is non-zero, packet is trailer-encapsulated,
1041 1.13 cgd * so we have to skip the type and length fields.
1042 1.13 cgd */
1043 1.100.2.6 yamt cp += off + 2 * sizeof(uint16_t);
1044 1.100.2.6 yamt totlen -= 2 * sizeof(uint16_t);
1045 1.9 mycroft }
1046 1.9 mycroft MGETHDR(m, M_DONTWAIT, MT_DATA);
1047 1.9 mycroft if (m == 0)
1048 1.71 simonb return (NULL);
1049 1.9 mycroft m->m_pkthdr.rcvif = ifp;
1050 1.9 mycroft m->m_pkthdr.len = totlen;
1051 1.9 mycroft m->m_len = MHLEN;
1052 1.9 mycroft
1053 1.9 mycroft while (totlen > 0) {
1054 1.9 mycroft if (top) {
1055 1.9 mycroft MGET(m, M_DONTWAIT, MT_DATA);
1056 1.9 mycroft if (m == 0) {
1057 1.9 mycroft m_freem(top);
1058 1.71 simonb return (NULL);
1059 1.9 mycroft }
1060 1.9 mycroft m->m_len = MLEN;
1061 1.9 mycroft }
1062 1.9 mycroft len = min(totlen, epkt - cp);
1063 1.9 mycroft if (len >= MINCLSIZE) {
1064 1.9 mycroft MCLGET(m, M_DONTWAIT);
1065 1.19 mycroft if ((m->m_flags & M_EXT) == 0) {
1066 1.20 mycroft m_free(m);
1067 1.19 mycroft m_freem(top);
1068 1.71 simonb return (NULL);
1069 1.19 mycroft }
1070 1.19 mycroft m->m_len = len = min(len, MCLBYTES);
1071 1.9 mycroft } else {
1072 1.9 mycroft /*
1073 1.9 mycroft * Place initial small packet/header at end of mbuf.
1074 1.9 mycroft */
1075 1.9 mycroft if (len < m->m_len) {
1076 1.9 mycroft if (top == 0 && len + max_linkhdr <= m->m_len)
1077 1.9 mycroft m->m_data += max_linkhdr;
1078 1.9 mycroft m->m_len = len;
1079 1.9 mycroft } else
1080 1.9 mycroft len = m->m_len;
1081 1.9 mycroft }
1082 1.9 mycroft if (copy)
1083 1.100.2.13 yamt copy(cp, mtod(m, void *), (size_t)len);
1084 1.9 mycroft else
1085 1.100.2.13 yamt memcpy(mtod(m, void *), cp, (size_t)len);
1086 1.9 mycroft cp += len;
1087 1.9 mycroft *mp = m;
1088 1.9 mycroft mp = &m->m_next;
1089 1.9 mycroft totlen -= len;
1090 1.9 mycroft if (cp == epkt)
1091 1.9 mycroft cp = buf;
1092 1.9 mycroft }
1093 1.9 mycroft return (top);
1094 1.18 thorpej }
1095 1.18 thorpej
1096 1.18 thorpej /*
1097 1.18 thorpej * Copy data from a buffer back into the indicated mbuf chain,
1098 1.18 thorpej * starting "off" bytes from the beginning, extending the mbuf
1099 1.18 thorpej * chain if necessary.
1100 1.18 thorpej */
1101 1.18 thorpej void
1102 1.86 yamt m_copyback(struct mbuf *m0, int off, int len, const void *cp)
1103 1.18 thorpej {
1104 1.85 yamt #if defined(DEBUG)
1105 1.85 yamt struct mbuf *origm = m0;
1106 1.85 yamt int error;
1107 1.85 yamt #endif /* defined(DEBUG) */
1108 1.85 yamt
1109 1.85 yamt if (m0 == NULL)
1110 1.85 yamt return;
1111 1.85 yamt
1112 1.85 yamt #if defined(DEBUG)
1113 1.85 yamt error =
1114 1.85 yamt #endif /* defined(DEBUG) */
1115 1.85 yamt m_copyback0(&m0, off, len, cp,
1116 1.85 yamt M_COPYBACK0_COPYBACK|M_COPYBACK0_EXTEND, M_DONTWAIT);
1117 1.85 yamt
1118 1.85 yamt #if defined(DEBUG)
1119 1.85 yamt if (error != 0 || (m0 != NULL && origm != m0))
1120 1.85 yamt panic("m_copyback");
1121 1.85 yamt #endif /* defined(DEBUG) */
1122 1.85 yamt }
1123 1.85 yamt
1124 1.85 yamt struct mbuf *
1125 1.86 yamt m_copyback_cow(struct mbuf *m0, int off, int len, const void *cp, int how)
1126 1.85 yamt {
1127 1.85 yamt int error;
1128 1.85 yamt
1129 1.85 yamt /* don't support chain expansion */
1130 1.85 yamt KDASSERT(off + len <= m_length(m0));
1131 1.85 yamt
1132 1.85 yamt error = m_copyback0(&m0, off, len, cp,
1133 1.85 yamt M_COPYBACK0_COPYBACK|M_COPYBACK0_COW, how);
1134 1.85 yamt if (error) {
1135 1.85 yamt /*
1136 1.85 yamt * no way to recover from partial success.
1137 1.85 yamt * just free the chain.
1138 1.85 yamt */
1139 1.85 yamt m_freem(m0);
1140 1.85 yamt return NULL;
1141 1.85 yamt }
1142 1.85 yamt return m0;
1143 1.85 yamt }
1144 1.85 yamt
1145 1.85 yamt /*
1146 1.85 yamt * m_makewritable: ensure the specified range writable.
1147 1.85 yamt */
1148 1.85 yamt int
1149 1.85 yamt m_makewritable(struct mbuf **mp, int off, int len, int how)
1150 1.85 yamt {
1151 1.85 yamt int error;
1152 1.85 yamt #if defined(DEBUG)
1153 1.85 yamt struct mbuf *n;
1154 1.85 yamt int origlen, reslen;
1155 1.85 yamt
1156 1.85 yamt origlen = m_length(*mp);
1157 1.85 yamt #endif /* defined(DEBUG) */
1158 1.85 yamt
1159 1.85 yamt #if 0 /* M_COPYALL is large enough */
1160 1.85 yamt if (len == M_COPYALL)
1161 1.85 yamt len = m_length(*mp) - off; /* XXX */
1162 1.85 yamt #endif
1163 1.85 yamt
1164 1.85 yamt error = m_copyback0(mp, off, len, NULL,
1165 1.85 yamt M_COPYBACK0_PRESERVE|M_COPYBACK0_COW, how);
1166 1.85 yamt
1167 1.85 yamt #if defined(DEBUG)
1168 1.85 yamt reslen = 0;
1169 1.85 yamt for (n = *mp; n; n = n->m_next)
1170 1.85 yamt reslen += n->m_len;
1171 1.85 yamt if (origlen != reslen)
1172 1.85 yamt panic("m_makewritable: length changed");
1173 1.85 yamt if (((*mp)->m_flags & M_PKTHDR) != 0 && reslen != (*mp)->m_pkthdr.len)
1174 1.85 yamt panic("m_makewritable: inconsist");
1175 1.85 yamt #endif /* defined(DEBUG) */
1176 1.85 yamt
1177 1.85 yamt return error;
1178 1.85 yamt }
1179 1.85 yamt
1180 1.85 yamt int
1181 1.86 yamt m_copyback0(struct mbuf **mp0, int off, int len, const void *vp, int flags,
1182 1.86 yamt int how)
1183 1.85 yamt {
1184 1.27 matt int mlen;
1185 1.85 yamt struct mbuf *m, *n;
1186 1.85 yamt struct mbuf **mp;
1187 1.18 thorpej int totlen = 0;
1188 1.86 yamt const char *cp = vp;
1189 1.18 thorpej
1190 1.85 yamt KASSERT(mp0 != NULL);
1191 1.85 yamt KASSERT(*mp0 != NULL);
1192 1.85 yamt KASSERT((flags & M_COPYBACK0_PRESERVE) == 0 || cp == NULL);
1193 1.85 yamt KASSERT((flags & M_COPYBACK0_COPYBACK) == 0 || cp != NULL);
1194 1.85 yamt
1195 1.100.2.6 yamt /*
1196 1.100.2.6 yamt * we don't bother to update "totlen" in the case of M_COPYBACK0_COW,
1197 1.100.2.6 yamt * assuming that M_COPYBACK0_EXTEND and M_COPYBACK0_COW are exclusive.
1198 1.100.2.6 yamt */
1199 1.100.2.6 yamt
1200 1.100.2.6 yamt KASSERT((~flags & (M_COPYBACK0_EXTEND|M_COPYBACK0_COW)) != 0);
1201 1.100.2.6 yamt
1202 1.85 yamt mp = mp0;
1203 1.85 yamt m = *mp;
1204 1.18 thorpej while (off > (mlen = m->m_len)) {
1205 1.18 thorpej off -= mlen;
1206 1.18 thorpej totlen += mlen;
1207 1.100.2.6 yamt if (m->m_next == NULL) {
1208 1.100.2.6 yamt int tspace;
1209 1.100.2.6 yamt extend:
1210 1.85 yamt if ((flags & M_COPYBACK0_EXTEND) == 0)
1211 1.85 yamt goto out;
1212 1.100.2.6 yamt
1213 1.100.2.6 yamt /*
1214 1.100.2.6 yamt * try to make some space at the end of "m".
1215 1.100.2.6 yamt */
1216 1.100.2.6 yamt
1217 1.100.2.6 yamt mlen = m->m_len;
1218 1.100.2.6 yamt if (off + len >= MINCLSIZE &&
1219 1.100.2.6 yamt (m->m_flags & M_EXT) == 0 && m->m_len == 0) {
1220 1.100.2.6 yamt MCLGET(m, how);
1221 1.100.2.6 yamt }
1222 1.100.2.6 yamt tspace = M_TRAILINGSPACE(m);
1223 1.100.2.6 yamt if (tspace > 0) {
1224 1.100.2.6 yamt tspace = min(tspace, off + len);
1225 1.100.2.6 yamt KASSERT(tspace > 0);
1226 1.100.2.6 yamt memset(mtod(m, char *) + m->m_len, 0,
1227 1.100.2.6 yamt min(off, tspace));
1228 1.100.2.6 yamt m->m_len += tspace;
1229 1.100.2.6 yamt off += mlen;
1230 1.100.2.6 yamt totlen -= mlen;
1231 1.100.2.6 yamt continue;
1232 1.100.2.6 yamt }
1233 1.100.2.6 yamt
1234 1.100.2.6 yamt /*
1235 1.100.2.6 yamt * need to allocate an mbuf.
1236 1.100.2.6 yamt */
1237 1.100.2.6 yamt
1238 1.100.2.6 yamt if (off + len >= MINCLSIZE) {
1239 1.100.2.6 yamt n = m_getcl(how, m->m_type, 0);
1240 1.100.2.6 yamt } else {
1241 1.100.2.6 yamt n = m_get(how, m->m_type);
1242 1.100.2.6 yamt }
1243 1.100.2.6 yamt if (n == NULL) {
1244 1.18 thorpej goto out;
1245 1.100.2.6 yamt }
1246 1.100.2.6 yamt n->m_len = 0;
1247 1.100.2.6 yamt n->m_len = min(M_TRAILINGSPACE(n), off + len);
1248 1.100.2.6 yamt memset(mtod(n, char *), 0, min(n->m_len, off));
1249 1.18 thorpej m->m_next = n;
1250 1.18 thorpej }
1251 1.85 yamt mp = &m->m_next;
1252 1.18 thorpej m = m->m_next;
1253 1.18 thorpej }
1254 1.18 thorpej while (len > 0) {
1255 1.85 yamt mlen = m->m_len - off;
1256 1.85 yamt if (mlen != 0 && M_READONLY(m)) {
1257 1.85 yamt char *datap;
1258 1.85 yamt int eatlen;
1259 1.85 yamt
1260 1.85 yamt /*
1261 1.85 yamt * this mbuf is read-only.
1262 1.85 yamt * allocate a new writable mbuf and try again.
1263 1.85 yamt */
1264 1.85 yamt
1265 1.85 yamt #if defined(DIAGNOSTIC)
1266 1.85 yamt if ((flags & M_COPYBACK0_COW) == 0)
1267 1.85 yamt panic("m_copyback0: read-only");
1268 1.85 yamt #endif /* defined(DIAGNOSTIC) */
1269 1.85 yamt
1270 1.85 yamt /*
1271 1.85 yamt * if we're going to write into the middle of
1272 1.85 yamt * a mbuf, split it first.
1273 1.85 yamt */
1274 1.85 yamt if (off > 0 && len < mlen) {
1275 1.85 yamt n = m_split0(m, off, how, 0);
1276 1.85 yamt if (n == NULL)
1277 1.85 yamt goto enobufs;
1278 1.85 yamt m->m_next = n;
1279 1.85 yamt mp = &m->m_next;
1280 1.85 yamt m = n;
1281 1.85 yamt off = 0;
1282 1.85 yamt continue;
1283 1.85 yamt }
1284 1.85 yamt
1285 1.85 yamt /*
1286 1.85 yamt * XXX TODO coalesce into the trailingspace of
1287 1.85 yamt * the previous mbuf when possible.
1288 1.85 yamt */
1289 1.85 yamt
1290 1.85 yamt /*
1291 1.85 yamt * allocate a new mbuf. copy packet header if needed.
1292 1.85 yamt */
1293 1.85 yamt MGET(n, how, m->m_type);
1294 1.85 yamt if (n == NULL)
1295 1.85 yamt goto enobufs;
1296 1.85 yamt MCLAIM(n, m->m_owner);
1297 1.85 yamt if (off == 0 && (m->m_flags & M_PKTHDR) != 0) {
1298 1.100.2.6 yamt M_MOVE_PKTHDR(n, m);
1299 1.85 yamt n->m_len = MHLEN;
1300 1.85 yamt } else {
1301 1.85 yamt if (len >= MINCLSIZE)
1302 1.85 yamt MCLGET(n, M_DONTWAIT);
1303 1.85 yamt n->m_len =
1304 1.85 yamt (n->m_flags & M_EXT) ? MCLBYTES : MLEN;
1305 1.85 yamt }
1306 1.85 yamt if (n->m_len > len)
1307 1.85 yamt n->m_len = len;
1308 1.85 yamt
1309 1.85 yamt /*
1310 1.85 yamt * free the region which has been overwritten.
1311 1.85 yamt * copying data from old mbufs if requested.
1312 1.85 yamt */
1313 1.85 yamt if (flags & M_COPYBACK0_PRESERVE)
1314 1.85 yamt datap = mtod(n, char *);
1315 1.85 yamt else
1316 1.85 yamt datap = NULL;
1317 1.85 yamt eatlen = n->m_len;
1318 1.85 yamt KDASSERT(off == 0 || eatlen >= mlen);
1319 1.85 yamt if (off > 0) {
1320 1.85 yamt KDASSERT(len >= mlen);
1321 1.85 yamt m->m_len = off;
1322 1.85 yamt m->m_next = n;
1323 1.85 yamt if (datap) {
1324 1.85 yamt m_copydata(m, off, mlen, datap);
1325 1.85 yamt datap += mlen;
1326 1.85 yamt }
1327 1.85 yamt eatlen -= mlen;
1328 1.85 yamt mp = &m->m_next;
1329 1.85 yamt m = m->m_next;
1330 1.85 yamt }
1331 1.85 yamt while (m != NULL && M_READONLY(m) &&
1332 1.85 yamt n->m_type == m->m_type && eatlen > 0) {
1333 1.85 yamt mlen = min(eatlen, m->m_len);
1334 1.85 yamt if (datap) {
1335 1.85 yamt m_copydata(m, 0, mlen, datap);
1336 1.85 yamt datap += mlen;
1337 1.85 yamt }
1338 1.85 yamt m->m_data += mlen;
1339 1.85 yamt m->m_len -= mlen;
1340 1.85 yamt eatlen -= mlen;
1341 1.85 yamt if (m->m_len == 0)
1342 1.85 yamt *mp = m = m_free(m);
1343 1.85 yamt }
1344 1.85 yamt if (eatlen > 0)
1345 1.85 yamt n->m_len -= eatlen;
1346 1.85 yamt n->m_next = m;
1347 1.85 yamt *mp = m = n;
1348 1.85 yamt continue;
1349 1.85 yamt }
1350 1.85 yamt mlen = min(mlen, len);
1351 1.85 yamt if (flags & M_COPYBACK0_COPYBACK) {
1352 1.100.2.13 yamt memcpy(mtod(m, char *) + off, cp, (unsigned)mlen);
1353 1.85 yamt cp += mlen;
1354 1.85 yamt }
1355 1.18 thorpej len -= mlen;
1356 1.18 thorpej mlen += off;
1357 1.18 thorpej off = 0;
1358 1.18 thorpej totlen += mlen;
1359 1.18 thorpej if (len == 0)
1360 1.18 thorpej break;
1361 1.100.2.6 yamt if (m->m_next == NULL) {
1362 1.100.2.6 yamt goto extend;
1363 1.18 thorpej }
1364 1.85 yamt mp = &m->m_next;
1365 1.18 thorpej m = m->m_next;
1366 1.18 thorpej }
1367 1.100.2.6 yamt out: if (((m = *mp0)->m_flags & M_PKTHDR) && (m->m_pkthdr.len < totlen)) {
1368 1.100.2.6 yamt KASSERT((flags & M_COPYBACK0_EXTEND) != 0);
1369 1.18 thorpej m->m_pkthdr.len = totlen;
1370 1.100.2.6 yamt }
1371 1.85 yamt
1372 1.85 yamt return 0;
1373 1.85 yamt
1374 1.85 yamt enobufs:
1375 1.85 yamt return ENOBUFS;
1376 1.66 thorpej }
1377 1.66 thorpej
1378 1.100.2.6 yamt void
1379 1.100.2.6 yamt m_move_pkthdr(struct mbuf *to, struct mbuf *from)
1380 1.100.2.6 yamt {
1381 1.100.2.6 yamt
1382 1.100.2.6 yamt KASSERT((to->m_flags & M_EXT) == 0);
1383 1.100.2.6 yamt KASSERT((to->m_flags & M_PKTHDR) == 0 || m_tag_first(to) == NULL);
1384 1.100.2.6 yamt KASSERT((from->m_flags & M_PKTHDR) != 0);
1385 1.100.2.6 yamt
1386 1.100.2.6 yamt to->m_pkthdr = from->m_pkthdr;
1387 1.100.2.6 yamt to->m_flags = from->m_flags & M_COPYFLAGS;
1388 1.100.2.6 yamt to->m_data = to->m_pktdat;
1389 1.100.2.6 yamt
1390 1.100.2.6 yamt from->m_flags &= ~M_PKTHDR;
1391 1.100.2.6 yamt }
1392 1.100.2.6 yamt
1393 1.66 thorpej /*
1394 1.66 thorpej * Apply function f to the data in an mbuf chain starting "off" bytes from the
1395 1.66 thorpej * beginning, continuing for "len" bytes.
1396 1.66 thorpej */
1397 1.66 thorpej int
1398 1.66 thorpej m_apply(struct mbuf *m, int off, int len,
1399 1.100.2.13 yamt int (*f)(void *, void *, unsigned int), void *arg)
1400 1.66 thorpej {
1401 1.66 thorpej unsigned int count;
1402 1.66 thorpej int rval;
1403 1.66 thorpej
1404 1.66 thorpej KASSERT(len >= 0);
1405 1.66 thorpej KASSERT(off >= 0);
1406 1.66 thorpej
1407 1.66 thorpej while (off > 0) {
1408 1.66 thorpej KASSERT(m != NULL);
1409 1.66 thorpej if (off < m->m_len)
1410 1.66 thorpej break;
1411 1.66 thorpej off -= m->m_len;
1412 1.66 thorpej m = m->m_next;
1413 1.66 thorpej }
1414 1.66 thorpej while (len > 0) {
1415 1.66 thorpej KASSERT(m != NULL);
1416 1.66 thorpej count = min(m->m_len - off, len);
1417 1.66 thorpej
1418 1.100.2.13 yamt rval = (*f)(arg, mtod(m, char *) + off, count);
1419 1.66 thorpej if (rval)
1420 1.66 thorpej return (rval);
1421 1.66 thorpej
1422 1.66 thorpej len -= count;
1423 1.66 thorpej off = 0;
1424 1.66 thorpej m = m->m_next;
1425 1.66 thorpej }
1426 1.66 thorpej
1427 1.66 thorpej return (0);
1428 1.66 thorpej }
1429 1.66 thorpej
1430 1.66 thorpej /*
1431 1.66 thorpej * Return a pointer to mbuf/offset of location in mbuf chain.
1432 1.66 thorpej */
1433 1.66 thorpej struct mbuf *
1434 1.66 thorpej m_getptr(struct mbuf *m, int loc, int *off)
1435 1.66 thorpej {
1436 1.66 thorpej
1437 1.66 thorpej while (loc >= 0) {
1438 1.66 thorpej /* Normal end of search */
1439 1.66 thorpej if (m->m_len > loc) {
1440 1.66 thorpej *off = loc;
1441 1.66 thorpej return (m);
1442 1.66 thorpej } else {
1443 1.66 thorpej loc -= m->m_len;
1444 1.66 thorpej
1445 1.66 thorpej if (m->m_next == NULL) {
1446 1.66 thorpej if (loc == 0) {
1447 1.66 thorpej /* Point at the end of valid data */
1448 1.66 thorpej *off = m->m_len;
1449 1.66 thorpej return (m);
1450 1.66 thorpej } else
1451 1.66 thorpej return (NULL);
1452 1.66 thorpej } else
1453 1.66 thorpej m = m->m_next;
1454 1.66 thorpej }
1455 1.66 thorpej }
1456 1.66 thorpej
1457 1.66 thorpej return (NULL);
1458 1.1 cgd }
1459 1.100.2.2 yamt
1460 1.100.2.2 yamt /*
1461 1.100.2.2 yamt * m_ext_free: release a reference to the mbuf external storage.
1462 1.100.2.2 yamt *
1463 1.100.2.2 yamt * => free the mbuf m itsself as well.
1464 1.100.2.2 yamt * => called at splvm.
1465 1.100.2.2 yamt */
1466 1.100.2.2 yamt
1467 1.100.2.2 yamt void
1468 1.100.2.2 yamt m_ext_free(struct mbuf *m)
1469 1.100.2.2 yamt {
1470 1.100.2.12 yamt bool embedded = MEXT_ISEMBEDDED(m);
1471 1.100.2.12 yamt bool dofree = true;
1472 1.100.2.2 yamt
1473 1.100.2.2 yamt KASSERT((m->m_flags & M_EXT) != 0);
1474 1.100.2.2 yamt KASSERT(MEXT_ISEMBEDDED(m->m_ext_ref));
1475 1.100.2.2 yamt KASSERT((m->m_ext_ref->m_flags & M_EXT) != 0);
1476 1.100.2.2 yamt KASSERT((m->m_flags & M_EXT_CLUSTER) ==
1477 1.100.2.2 yamt (m->m_ext_ref->m_flags & M_EXT_CLUSTER));
1478 1.100.2.2 yamt
1479 1.100.2.10 yamt if (!mcl_dec_and_test_reference(m)) {
1480 1.100.2.2 yamt if (embedded) {
1481 1.100.2.15 yamt /*
1482 1.100.2.15 yamt * other mbuf's m_ext_ref still points to us.
1483 1.100.2.15 yamt */
1484 1.100.2.12 yamt dofree = false;
1485 1.100.2.2 yamt } else {
1486 1.100.2.2 yamt m->m_ext_ref = m;
1487 1.100.2.2 yamt }
1488 1.100.2.2 yamt } else {
1489 1.100.2.2 yamt /*
1490 1.100.2.2 yamt * dropping the last reference
1491 1.100.2.2 yamt */
1492 1.100.2.2 yamt if (!embedded) {
1493 1.100.2.10 yamt mcl_inc_reference(m); /* XXX */
1494 1.100.2.2 yamt m_ext_free(m->m_ext_ref);
1495 1.100.2.2 yamt m->m_ext_ref = m;
1496 1.100.2.2 yamt } else if ((m->m_flags & M_EXT_CLUSTER) != 0) {
1497 1.100.2.6 yamt pool_cache_put_paddr((struct pool_cache *)
1498 1.100.2.6 yamt m->m_ext.ext_arg,
1499 1.100.2.2 yamt m->m_ext.ext_buf, m->m_ext.ext_paddr);
1500 1.100.2.2 yamt } else if (m->m_ext.ext_free) {
1501 1.100.2.15 yamt mutex_destroy(&m->m_ext.ext_lock);
1502 1.100.2.3 yamt (*m->m_ext.ext_free)(m,
1503 1.100.2.2 yamt m->m_ext.ext_buf, m->m_ext.ext_size,
1504 1.100.2.2 yamt m->m_ext.ext_arg);
1505 1.100.2.2 yamt /*
1506 1.100.2.2 yamt * 'm' is already freed by the ext_free callback.
1507 1.100.2.2 yamt */
1508 1.100.2.12 yamt dofree = false;
1509 1.100.2.2 yamt } else {
1510 1.100.2.2 yamt free(m->m_ext.ext_buf, m->m_ext.ext_type);
1511 1.100.2.2 yamt }
1512 1.100.2.2 yamt }
1513 1.100.2.2 yamt if (dofree) {
1514 1.100.2.15 yamt if (embedded) {
1515 1.100.2.15 yamt mutex_destroy(&m->m_ext.ext_lock);
1516 1.100.2.15 yamt }
1517 1.100.2.2 yamt pool_cache_put(&mbpool_cache, m);
1518 1.100.2.2 yamt }
1519 1.100.2.2 yamt }
1520 1.100.2.3 yamt
1521 1.100.2.4 yamt #if defined(__HAVE_LAZY_MBUF) || defined(DEBUG)
1522 1.100.2.14 yamt char *
1523 1.100.2.3 yamt m_mapin(struct mbuf *m)
1524 1.100.2.3 yamt {
1525 1.100.2.5 yamt
1526 1.100.2.15 yamt #if defined(__HAVE_LAZY_MBUF)
1527 1.100.2.3 yamt KASSERT((~m->m_flags & (M_EXT|M_EXT_PAGES|M_EXT_LAZY)) == 0);
1528 1.100.2.3 yamt
1529 1.100.2.3 yamt MEXT_LOCK(m);
1530 1.100.2.3 yamt if (m->m_ext.ext_flags & M_EXT_LAZY) {
1531 1.100.2.3 yamt vaddr_t buf = (vaddr_t)m->m_ext.ext_buf;
1532 1.100.2.3 yamt vsize_t size = (vsize_t)m->m_ext.ext_size;
1533 1.100.2.3 yamt vaddr_t va, sva, eva;
1534 1.100.2.3 yamt int i;
1535 1.100.2.3 yamt
1536 1.100.2.3 yamt sva = trunc_page(buf);
1537 1.100.2.3 yamt eva = round_page(buf + size);
1538 1.100.2.3 yamt
1539 1.100.2.3 yamt for (i = 0, va = sva; va < eva; i++, va += PAGE_SIZE) {
1540 1.100.2.3 yamt pmap_kenter_pa(va, VM_PAGE_TO_PHYS(m->m_ext.ext_pgs[i]),
1541 1.100.2.3 yamt VM_PROT_READ);
1542 1.100.2.3 yamt }
1543 1.100.2.3 yamt pmap_update(pmap_kernel());
1544 1.100.2.3 yamt m->m_ext.ext_flags &= ~M_EXT_LAZY;
1545 1.100.2.3 yamt }
1546 1.100.2.3 yamt MEXT_UNLOCK(m);
1547 1.100.2.3 yamt
1548 1.100.2.3 yamt m->m_flags &= ~M_EXT_LAZY;
1549 1.100.2.3 yamt return m->m_data;
1550 1.100.2.4 yamt #else /* defined(__HAVE_LAZY_MBUF) */
1551 1.100.2.4 yamt panic("m_mapin");
1552 1.100.2.4 yamt #endif /* defined(__HAVE_LAZY_MBUF) */
1553 1.100.2.3 yamt }
1554 1.100.2.4 yamt #endif /* defined(__HAVE_LAZY_MBUF) || defined(DEBUG) */
1555 1.100.2.6 yamt
1556 1.100.2.6 yamt #if defined(DDB)
1557 1.100.2.6 yamt void
1558 1.100.2.6 yamt m_print(const struct mbuf *m, const char *modif, void (*pr)(const char *, ...))
1559 1.100.2.6 yamt {
1560 1.100.2.6 yamt char ch;
1561 1.100.2.12 yamt bool opt_c = false;
1562 1.100.2.6 yamt char buf[512];
1563 1.100.2.6 yamt
1564 1.100.2.6 yamt while ((ch = *(modif++)) != '\0') {
1565 1.100.2.6 yamt switch (ch) {
1566 1.100.2.6 yamt case 'c':
1567 1.100.2.12 yamt opt_c = true;
1568 1.100.2.6 yamt break;
1569 1.100.2.6 yamt }
1570 1.100.2.6 yamt }
1571 1.100.2.6 yamt
1572 1.100.2.6 yamt nextchain:
1573 1.100.2.6 yamt (*pr)("MBUF %p\n", m);
1574 1.100.2.16 yamt bitmask_snprintf((u_int)m->m_flags, M_FLAGS_BITS, buf, sizeof(buf));
1575 1.100.2.6 yamt (*pr)(" data=%p, len=%d, type=%d, flags=0x%s\n",
1576 1.100.2.6 yamt m->m_data, m->m_len, m->m_type, buf);
1577 1.100.2.6 yamt (*pr)(" owner=%p, next=%p, nextpkt=%p\n", m->m_owner, m->m_next,
1578 1.100.2.6 yamt m->m_nextpkt);
1579 1.100.2.6 yamt (*pr)(" leadingspace=%u, trailingspace=%u, readonly=%u\n",
1580 1.100.2.6 yamt (int)M_LEADINGSPACE(m), (int)M_TRAILINGSPACE(m),
1581 1.100.2.6 yamt (int)M_READONLY(m));
1582 1.100.2.6 yamt if ((m->m_flags & M_PKTHDR) != 0) {
1583 1.100.2.6 yamt bitmask_snprintf(m->m_pkthdr.csum_flags, M_CSUM_BITS, buf,
1584 1.100.2.6 yamt sizeof(buf));
1585 1.100.2.6 yamt (*pr)(" pktlen=%d, rcvif=%p, csum_flags=0x%s, csum_data=0x%"
1586 1.100.2.6 yamt PRIx32 ", segsz=%u\n",
1587 1.100.2.6 yamt m->m_pkthdr.len, m->m_pkthdr.rcvif,
1588 1.100.2.6 yamt buf, m->m_pkthdr.csum_data, m->m_pkthdr.segsz);
1589 1.100.2.6 yamt }
1590 1.100.2.6 yamt if ((m->m_flags & M_EXT)) {
1591 1.100.2.9 yamt (*pr)(" ext_refcnt=%u, ext_buf=%p, ext_size=%zd, "
1592 1.100.2.6 yamt "ext_free=%p, ext_arg=%p\n",
1593 1.100.2.9 yamt m->m_ext.ext_refcnt,
1594 1.100.2.6 yamt m->m_ext.ext_buf, m->m_ext.ext_size,
1595 1.100.2.6 yamt m->m_ext.ext_free, m->m_ext.ext_arg);
1596 1.100.2.6 yamt }
1597 1.100.2.6 yamt if ((~m->m_flags & (M_EXT|M_EXT_PAGES)) == 0) {
1598 1.100.2.6 yamt vaddr_t sva = (vaddr_t)m->m_ext.ext_buf;
1599 1.100.2.6 yamt vaddr_t eva = sva + m->m_ext.ext_size;
1600 1.100.2.6 yamt int n = (round_page(eva) - trunc_page(sva)) >> PAGE_SHIFT;
1601 1.100.2.6 yamt int i;
1602 1.100.2.6 yamt
1603 1.100.2.6 yamt (*pr)(" pages:");
1604 1.100.2.6 yamt for (i = 0; i < n; i ++) {
1605 1.100.2.6 yamt (*pr)(" %p", m->m_ext.ext_pgs[i]);
1606 1.100.2.6 yamt }
1607 1.100.2.6 yamt (*pr)("\n");
1608 1.100.2.6 yamt }
1609 1.100.2.6 yamt
1610 1.100.2.6 yamt if (opt_c) {
1611 1.100.2.6 yamt m = m->m_next;
1612 1.100.2.6 yamt if (m != NULL) {
1613 1.100.2.6 yamt goto nextchain;
1614 1.100.2.6 yamt }
1615 1.100.2.6 yamt }
1616 1.100.2.6 yamt }
1617 1.100.2.6 yamt #endif /* defined(DDB) */
1618