1 /* 2 * Copyright (c) 1982, 1986, 1989, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 8 * 1. Redistributions of source code must retain the above copyright 9 * notice, this list of conditions and the following disclaimer. 10 * 2. Redistributions in binary form must reproduce the above copyright 11 * notice, this list of conditions and the following disclaimer in the 12 * documentation and/or other materials provided with the distribution. 13 * 3. All advertising materials mentioning features or use of this software 14 * must display the following acknowledgement: 15 * This product includes software developed by the University of 16 * California, Berkeley and its contributors. 17 * 4. Neither the name of the University nor the names of its contributors 18 * may be used to endorse or promote products derived from this software 19 * without specific prior written permission. 20 * 21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 24 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 31 * SUCH DAMAGE. 32 * 33 * @(#)ffs_vnops.c 8.15 (Berkeley) 5/14/95 34 * $FreeBSD$ 35 */ 36 37 #include <sys/param.h> 38 #include <sys/systm.h> 39 #include <sys/resourcevar.h> 40 #include <sys/signalvar.h> 41 #include <sys/kernel.h> 42 #include <sys/stat.h> 43 #include <sys/buf.h> 44 #include <sys/proc.h> 45 #include <sys/mount.h> 46 #include <sys/vnode.h> 47 #include <sys/conf.h> 48 49 #include <machine/limits.h> 50 51 #include <vm/vm.h> 52 #include <vm/vm_page.h> 53 #include <vm/vm_object.h> 54 #include <vm/vm_extern.h> 55 56 #include <ufs/ufs/quota.h> 57 #include <ufs/ufs/inode.h> 58 #include <ufs/ufs/ufsmount.h> 59 #include <ufs/ufs/ufs_extern.h> 60 61 #include <ufs/ffs/fs.h> 62 #include <ufs/ffs/ffs_extern.h> 63 64 static int ffs_fsync __P((struct vop_fsync_args *)); 65 static int ffs_getpages __P((struct vop_getpages_args *)); 66 static int ffs_putpages __P((struct vop_putpages_args *)); 67 static int ffs_read __P((struct vop_read_args *)); 68 static int ffs_write __P((struct vop_write_args *)); 69 70 /* Global vfs data structures for ufs. */ 71 vop_t **ffs_vnodeop_p; 72 static struct vnodeopv_entry_desc ffs_vnodeop_entries[] = { 73 { &vop_default_desc, (vop_t *) ufs_vnoperate }, 74 { &vop_fsync_desc, (vop_t *) ffs_fsync }, 75 { &vop_getpages_desc, (vop_t *) ffs_getpages }, 76 { &vop_putpages_desc, (vop_t *) ffs_putpages }, 77 { &vop_read_desc, (vop_t *) ffs_read }, 78 { &vop_balloc_desc, (vop_t *) ffs_balloc }, 79 { &vop_reallocblks_desc, (vop_t *) ffs_reallocblks }, 80 { &vop_write_desc, (vop_t *) ffs_write }, 81 { NULL, NULL } 82 }; 83 static struct vnodeopv_desc ffs_vnodeop_opv_desc = 84 { &ffs_vnodeop_p, ffs_vnodeop_entries }; 85 86 vop_t **ffs_specop_p; 87 static struct vnodeopv_entry_desc ffs_specop_entries[] = { 88 { &vop_default_desc, (vop_t *) ufs_vnoperatespec }, 89 { &vop_fsync_desc, (vop_t *) ffs_fsync }, 90 { NULL, NULL } 91 }; 92 static struct vnodeopv_desc ffs_specop_opv_desc = 93 { &ffs_specop_p, ffs_specop_entries }; 94 95 vop_t **ffs_fifoop_p; 96 static struct vnodeopv_entry_desc ffs_fifoop_entries[] = { 97 { &vop_default_desc, (vop_t *) ufs_vnoperatefifo }, 98 { &vop_fsync_desc, (vop_t *) ffs_fsync }, 99 { NULL, NULL } 100 }; 101 static struct vnodeopv_desc ffs_fifoop_opv_desc = 102 { &ffs_fifoop_p, ffs_fifoop_entries }; 103 104 VNODEOP_SET(ffs_vnodeop_opv_desc); 105 VNODEOP_SET(ffs_specop_opv_desc); 106 VNODEOP_SET(ffs_fifoop_opv_desc); 107 108 #include <ufs/ufs/ufs_readwrite.c> 109 110 /* 111 * Synch an open file. 112 */ 113 /* ARGSUSED */ 114 static int 115 ffs_fsync(ap) 116 struct vop_fsync_args /* { 117 struct vnode *a_vp; 118 struct ucred *a_cred; 119 int a_waitfor; 120 struct proc *a_p; 121 } */ *ap; 122 { 123 struct vnode *vp = ap->a_vp; 124 struct buf *bp; 125 struct buf *nbp; 126 int s, error, passes, skipmeta; 127 daddr_t lbn; 128 129 130 if (vp->v_type == VBLK) { 131 lbn = INT_MAX; 132 if (vp->v_specmountpoint != NULL && 133 (vp->v_specmountpoint->mnt_flag & MNT_SOFTDEP)) 134 softdep_fsync_mountdev(vp); 135 } else { 136 struct inode *ip; 137 ip = VTOI(vp); 138 lbn = lblkno(ip->i_fs, (ip->i_size + ip->i_fs->fs_bsize - 1)); 139 } 140 141 /* 142 * Flush all dirty buffers associated with a vnode. 143 */ 144 passes = NIADDR + 1; 145 skipmeta = 0; 146 if (ap->a_waitfor == MNT_WAIT) 147 skipmeta = 1; 148 s = splbio(); 149 loop: 150 for (bp = TAILQ_FIRST(&vp->v_dirtyblkhd); bp; 151 bp = TAILQ_NEXT(bp, b_vnbufs)) 152 bp->b_flags &= ~B_SCANNED; 153 for (bp = TAILQ_FIRST(&vp->v_dirtyblkhd); bp; bp = nbp) { 154 nbp = TAILQ_NEXT(bp, b_vnbufs); 155 /* 156 * First time through on a synchronous call, 157 * or if it's already scheduled, skip to the next 158 * buffer 159 */ 160 if ((bp->b_flags & B_SCANNED) || 161 ((skipmeta == 1) && (bp->b_lblkno < 0)) || 162 BUF_LOCK(bp, LK_EXCLUSIVE | LK_NOWAIT)) 163 continue; 164 if ((bp->b_flags & B_DELWRI) == 0) 165 panic("ffs_fsync: not dirty"); 166 /* 167 * If data is outstanding to another vnode, or we were 168 * asked to wait for everything, or it's not a file or BDEV, 169 * start the IO on this buffer immediatly. 170 */ 171 bp->b_flags |= B_SCANNED; 172 if (((bp->b_vp != vp) || (ap->a_waitfor == MNT_WAIT)) || 173 ((vp->v_type != VREG) && (vp->v_type != VBLK))) { 174 175 /* 176 * On our final pass through, do all I/O synchronously 177 * so that we can find out if our flush is failing 178 * because of write errors. 179 */ 180 if (passes > 0 || (ap->a_waitfor != MNT_WAIT)) { 181 if ((bp->b_flags & B_CLUSTEROK) && 182 ap->a_waitfor != MNT_WAIT) { 183 BUF_UNLOCK(bp); 184 (void) vfs_bio_awrite(bp); 185 } else { 186 bremfree(bp); 187 splx(s); 188 (void) bawrite(bp); 189 s = splbio(); 190 } 191 } else { 192 bremfree(bp); 193 splx(s); 194 if ((error = bwrite(bp)) != 0) 195 return (error); 196 s = splbio(); 197 } 198 } else if ((vp->v_type == VREG) && (bp->b_lblkno >= lbn)) { 199 /* 200 * If the buffer is for data that has been truncated 201 * off the file, then throw it away. 202 */ 203 bremfree(bp); 204 bp->b_flags |= B_INVAL | B_NOCACHE; 205 splx(s); 206 brelse(bp); 207 s = splbio(); 208 } else { 209 BUF_UNLOCK(bp); 210 vfs_bio_awrite(bp); 211 } 212 /* 213 * Since we may have slept during the I/O, we need 214 * to start from a known point. 215 */ 216 nbp = TAILQ_FIRST(&vp->v_dirtyblkhd); 217 } 218 /* 219 * If we were asked to do this synchronously, then go back for 220 * another pass, this time doing the metadata. 221 */ 222 if (skipmeta) { 223 skipmeta = 0; 224 goto loop; 225 } 226 227 if (ap->a_waitfor == MNT_WAIT) { 228 while (vp->v_numoutput) { 229 vp->v_flag |= VBWAIT; 230 (void) tsleep((caddr_t)&vp->v_numoutput, 231 PRIBIO + 4, "ffsfsn", 0); 232 } 233 234 /* 235 * Ensure that any filesystem metatdata associated 236 * with the vnode has been written. 237 */ 238 splx(s); 239 if ((error = softdep_sync_metadata(ap)) != 0) 240 return (error); 241 s = splbio(); 242 243 if (!TAILQ_EMPTY(&vp->v_dirtyblkhd)) { 244 /* 245 * Block devices associated with filesystems may 246 * have new I/O requests posted for them even if 247 * the vnode is locked, so no amount of trying will 248 * get them clean. Thus we give block devices a 249 * good effort, then just give up. For all other file 250 * types, go around and try again until it is clean. 251 */ 252 if (passes > 0) { 253 passes -= 1; 254 goto loop; 255 } 256 #ifdef DIAGNOSTIC 257 if (vp->v_type != VBLK) 258 vprint("ffs_fsync: dirty", vp); 259 #endif 260 } 261 } 262 splx(s); 263 return (UFS_UPDATE(vp, ap->a_waitfor == MNT_WAIT)); 264 } 265