nfs_kq.c revision 1.31
1 /* $NetBSD: nfs_kq.c,v 1.31 2021/10/11 01:49:08 thorpej Exp $ */ 2 3 /*- 4 * Copyright (c) 2002, 2008 The NetBSD Foundation, Inc. 5 * All rights reserved. 6 * 7 * This code is derived from software contributed to The NetBSD Foundation 8 * by Jaromir Dolecek. 9 * 10 * Redistribution and use in source and binary forms, with or without 11 * modification, are permitted provided that the following conditions 12 * are met: 13 * 1. Redistributions of source code must retain the above copyright 14 * notice, this list of conditions and the following disclaimer. 15 * 2. Redistributions in binary form must reproduce the above copyright 16 * notice, this list of conditions and the following disclaimer in the 17 * documentation and/or other materials provided with the distribution. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS 20 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED 21 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 22 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS 23 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 24 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 25 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 26 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 27 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 28 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 29 * POSSIBILITY OF SUCH DAMAGE. 30 */ 31 32 #include <sys/cdefs.h> 33 __KERNEL_RCSID(0, "$NetBSD: nfs_kq.c,v 1.31 2021/10/11 01:49:08 thorpej Exp $"); 34 35 #include <sys/param.h> 36 #include <sys/systm.h> 37 #include <sys/condvar.h> 38 #include <sys/kernel.h> 39 #include <sys/proc.h> 40 #include <sys/kmem.h> 41 #include <sys/mount.h> 42 #include <sys/mutex.h> 43 #include <sys/vnode.h> 44 #include <sys/unistd.h> 45 #include <sys/file.h> 46 #include <sys/kthread.h> 47 48 #include <nfs/rpcv2.h> 49 #include <nfs/nfsproto.h> 50 #include <nfs/nfs.h> 51 #include <nfs/nfsnode.h> 52 #include <nfs/nfs_var.h> 53 54 struct kevq { 55 SLIST_ENTRY(kevq) kev_link; 56 struct vnode *vp; 57 u_int usecount; 58 u_int flags; 59 #define KEVQ_BUSY 0x01 /* currently being processed */ 60 struct timespec omtime; /* old modification time */ 61 struct timespec octime; /* old change time */ 62 nlink_t onlink; /* old number of references to file */ 63 kcondvar_t cv; 64 }; 65 SLIST_HEAD(kevqlist, kevq); 66 67 static kmutex_t nfskq_lock; 68 static struct lwp *nfskq_thread; 69 static kcondvar_t nfskq_cv; 70 static struct kevqlist kevlist = SLIST_HEAD_INITIALIZER(kevlist); 71 static bool nfskq_thread_exit; 72 73 void 74 nfs_kqinit(void) 75 { 76 77 mutex_init(&nfskq_lock, MUTEX_DEFAULT, IPL_NONE); 78 cv_init(&nfskq_cv, "nfskqpw"); 79 } 80 81 void 82 nfs_kqfini(void) 83 { 84 85 if (nfskq_thread != NULL) { 86 mutex_enter(&nfskq_lock); 87 nfskq_thread_exit = true; 88 cv_broadcast(&nfskq_cv); 89 do { 90 cv_wait(&nfskq_cv, &nfskq_lock); 91 } while (nfskq_thread != NULL); 92 mutex_exit(&nfskq_lock); 93 } 94 mutex_destroy(&nfskq_lock); 95 cv_destroy(&nfskq_cv); 96 } 97 98 /* 99 * This quite simplistic routine periodically checks for server changes 100 * of any of the watched files every NFS_MINATTRTIMO/2 seconds. 101 * Only changes in size, modification time, change time and nlinks 102 * are being checked, everything else is ignored. 103 * The routine only calls VOP_GETATTR() when it's likely it would get 104 * some new data, i.e. when the vnode expires from attrcache. This 105 * should give same result as periodically running stat(2) from userland, 106 * while keeping CPU/network usage low, and still provide proper kevent 107 * semantics. 108 * The poller thread is created when first vnode is added to watch list, 109 * and exits when the watch list is empty. The overhead of thread creation 110 * isn't really important, neither speed of attach and detach of knote. 111 */ 112 /* ARGSUSED */ 113 static void 114 nfs_kqpoll(void *arg) 115 { 116 struct kevq *ke; 117 struct vattr attr; 118 struct lwp *l = curlwp; 119 u_quad_t osize; 120 121 mutex_enter(&nfskq_lock); 122 while (!nfskq_thread_exit) { 123 SLIST_FOREACH(ke, &kevlist, kev_link) { 124 /* skip if still in attrcache */ 125 if (nfs_getattrcache(ke->vp, &attr) != ENOENT) 126 continue; 127 128 /* 129 * Mark entry busy, release lock and check 130 * for changes. 131 */ 132 ke->flags |= KEVQ_BUSY; 133 mutex_exit(&nfskq_lock); 134 135 /* save v_size, nfs_getattr() updates it */ 136 osize = ke->vp->v_size; 137 138 memset(&attr, 0, sizeof(attr)); 139 vn_lock(ke->vp, LK_SHARED | LK_RETRY); 140 (void) VOP_GETATTR(ke->vp, &attr, l->l_cred); 141 VOP_UNLOCK(ke->vp); 142 143 /* following is a bit fragile, but about best 144 * we can get */ 145 if (attr.va_size != osize) { 146 int extended = (attr.va_size > osize); 147 VN_KNOTE(ke->vp, NOTE_WRITE 148 | (extended ? NOTE_EXTEND : 0)); 149 ke->omtime = attr.va_mtime; 150 } else if (attr.va_mtime.tv_sec != ke->omtime.tv_sec 151 || attr.va_mtime.tv_nsec != ke->omtime.tv_nsec) { 152 VN_KNOTE(ke->vp, NOTE_WRITE); 153 ke->omtime = attr.va_mtime; 154 } 155 156 if (attr.va_ctime.tv_sec != ke->octime.tv_sec 157 || attr.va_ctime.tv_nsec != ke->octime.tv_nsec) { 158 VN_KNOTE(ke->vp, NOTE_ATTRIB); 159 ke->octime = attr.va_ctime; 160 } 161 162 if (attr.va_nlink != ke->onlink) { 163 VN_KNOTE(ke->vp, NOTE_LINK); 164 ke->onlink = attr.va_nlink; 165 } 166 167 mutex_enter(&nfskq_lock); 168 ke->flags &= ~KEVQ_BUSY; 169 cv_signal(&ke->cv); 170 } 171 172 if (SLIST_EMPTY(&kevlist)) { 173 /* Nothing more to watch, exit */ 174 nfskq_thread = NULL; 175 mutex_exit(&nfskq_lock); 176 kthread_exit(0); 177 } 178 179 /* wait a while before checking for changes again */ 180 cv_timedwait(&nfskq_cv, &nfskq_lock, 181 NFS_MINATTRTIMO * hz / 2); 182 } 183 nfskq_thread = NULL; 184 cv_broadcast(&nfskq_cv); 185 mutex_exit(&nfskq_lock); 186 } 187 188 static void 189 filt_nfsdetach(struct knote *kn) 190 { 191 struct vnode *vp = (struct vnode *)kn->kn_hook; 192 struct kevq *ke; 193 194 mutex_enter(vp->v_interlock); 195 SLIST_REMOVE(&vp->v_klist, kn, knote, kn_selnext); 196 mutex_exit(vp->v_interlock); 197 198 /* Remove the vnode from watch list */ 199 mutex_enter(&nfskq_lock); 200 SLIST_FOREACH(ke, &kevlist, kev_link) { 201 if (ke->vp == vp) { 202 while (ke->flags & KEVQ_BUSY) { 203 cv_wait(&ke->cv, &nfskq_lock); 204 } 205 206 if (ke->usecount > 1) { 207 /* keep, other kevents need this */ 208 ke->usecount--; 209 } else { 210 /* last user, g/c */ 211 cv_destroy(&ke->cv); 212 SLIST_REMOVE(&kevlist, ke, kevq, kev_link); 213 kmem_free(ke, sizeof(*ke)); 214 } 215 break; 216 } 217 } 218 mutex_exit(&nfskq_lock); 219 } 220 221 static int 222 filt_nfsread(struct knote *kn, long hint) 223 { 224 struct vnode *vp = (struct vnode *)kn->kn_hook; 225 int rv; 226 227 /* 228 * filesystem is gone, so set the EOF flag and schedule 229 * the knote for deletion. 230 */ 231 switch (hint) { 232 case NOTE_REVOKE: 233 KASSERT(mutex_owned(vp->v_interlock)); 234 knote_set_eof(kn, EV_ONESHOT); 235 return (1); 236 case 0: 237 mutex_enter(vp->v_interlock); 238 kn->kn_data = vp->v_size - ((file_t *)kn->kn_obj)->f_offset; 239 rv = (kn->kn_data != 0); 240 mutex_exit(vp->v_interlock); 241 return rv; 242 default: 243 KASSERT(mutex_owned(vp->v_interlock)); 244 kn->kn_data = vp->v_size - ((file_t *)kn->kn_obj)->f_offset; 245 return (kn->kn_data != 0); 246 } 247 } 248 249 static int 250 filt_nfsvnode(struct knote *kn, long hint) 251 { 252 struct vnode *vp = (struct vnode *)kn->kn_hook; 253 int fflags; 254 255 switch (hint) { 256 case NOTE_REVOKE: 257 KASSERT(mutex_owned(vp->v_interlock)); 258 knote_set_eof(kn, 0); 259 if ((kn->kn_sfflags & hint) != 0) 260 kn->kn_fflags |= hint; 261 return (1); 262 case 0: 263 mutex_enter(vp->v_interlock); 264 fflags = kn->kn_fflags; 265 mutex_exit(vp->v_interlock); 266 break; 267 default: 268 KASSERT(mutex_owned(vp->v_interlock)); 269 if ((kn->kn_sfflags & hint) != 0) 270 kn->kn_fflags |= hint; 271 fflags = kn->kn_fflags; 272 break; 273 } 274 275 return (fflags != 0); 276 } 277 278 279 static const struct filterops nfsread_filtops = { 280 .f_flags = FILTEROP_ISFD | FILTEROP_MPSAFE, 281 .f_attach = NULL, 282 .f_detach = filt_nfsdetach, 283 .f_event = filt_nfsread, 284 }; 285 286 static const struct filterops nfsvnode_filtops = { 287 .f_flags = FILTEROP_ISFD | FILTEROP_MPSAFE, 288 .f_attach = NULL, 289 .f_detach = filt_nfsdetach, 290 .f_event = filt_nfsvnode, 291 }; 292 293 int 294 nfs_kqfilter(void *v) 295 { 296 struct vop_kqfilter_args /* { 297 struct vnode *a_vp; 298 struct knote *a_kn; 299 } */ *ap = v; 300 struct vnode *vp; 301 struct knote *kn; 302 struct kevq *ke; 303 int error = 0; 304 struct vattr attr; 305 struct lwp *l = curlwp; 306 307 vp = ap->a_vp; 308 kn = ap->a_kn; 309 switch (kn->kn_filter) { 310 case EVFILT_READ: 311 kn->kn_fop = &nfsread_filtops; 312 break; 313 case EVFILT_VNODE: 314 kn->kn_fop = &nfsvnode_filtops; 315 break; 316 default: 317 return (EINVAL); 318 } 319 320 /* 321 * Put the vnode to watched list. 322 */ 323 324 /* 325 * Fetch current attributes. It's only needed when the vnode 326 * is not watched yet, but we need to do this without lock 327 * held. This is likely cheap due to attrcache, so do it now. 328 */ 329 memset(&attr, 0, sizeof(attr)); 330 vn_lock(vp, LK_SHARED | LK_RETRY); 331 (void) VOP_GETATTR(vp, &attr, l->l_cred); 332 VOP_UNLOCK(vp); 333 334 mutex_enter(&nfskq_lock); 335 336 /* ensure the poller is running */ 337 if (!nfskq_thread) { 338 error = kthread_create(PRI_NONE, 0, NULL, nfs_kqpoll, 339 NULL, &nfskq_thread, "nfskqpoll"); 340 if (error) { 341 mutex_exit(&nfskq_lock); 342 return error; 343 } 344 } 345 346 SLIST_FOREACH(ke, &kevlist, kev_link) { 347 if (ke->vp == vp) 348 break; 349 } 350 351 if (ke) { 352 /* already watched, so just bump usecount */ 353 ke->usecount++; 354 } else { 355 /* need a new one */ 356 ke = kmem_alloc(sizeof(*ke), KM_SLEEP); 357 ke->vp = vp; 358 ke->usecount = 1; 359 ke->flags = 0; 360 ke->omtime = attr.va_mtime; 361 ke->octime = attr.va_ctime; 362 ke->onlink = attr.va_nlink; 363 cv_init(&ke->cv, "nfskqdet"); 364 SLIST_INSERT_HEAD(&kevlist, ke, kev_link); 365 } 366 367 mutex_enter(vp->v_interlock); 368 SLIST_INSERT_HEAD(&vp->v_klist, kn, kn_selnext); 369 kn->kn_hook = vp; 370 mutex_exit(vp->v_interlock); 371 372 /* kick the poller */ 373 cv_signal(&nfskq_cv); 374 mutex_exit(&nfskq_lock); 375 376 return (error); 377 } 378
Indexes created Fri Sep 26 08:10:20 GMT 2025