1 /* $NetBSD: ext2fs_readwrite.c,v 1.12 2000/05/13 23:43:13 perseant Exp $ */ 2 3 /*- 4 * Copyright (c) 1997 Manuel Bouyer. 5 * Copyright (c) 1993 6 * The Regents of the University of California. All rights reserved. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 3. All advertising materials mentioning features or use of this software 17 * must display the following acknowledgement: 18 * This product includes software developed by the University of 19 * California, Berkeley and its contributors. 20 * 4. Neither the name of the University nor the names of its contributors 21 * may be used to endorse or promote products derived from this software 22 * without specific prior written permission. 23 * 24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 27 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 34 * SUCH DAMAGE. 35 * 36 * @(#)ufs_readwrite.c 8.8 (Berkeley) 8/4/94 37 * Modified for ext2fs by Manuel Bouyer. 38 */ 39 40 #include <sys/param.h> 41 #include <sys/systm.h> 42 #include <sys/resourcevar.h> 43 #include <sys/kernel.h> 44 #include <sys/file.h> 45 #include <sys/stat.h> 46 #include <sys/buf.h> 47 #include <sys/proc.h> 48 #include <sys/conf.h> 49 #include <sys/mount.h> 50 #include <sys/vnode.h> 51 #include <sys/malloc.h> 52 #include <sys/signalvar.h> 53 54 #include <vm/vm.h> 55 56 #include <uvm/uvm_extern.h> 57 58 #include <ufs/ufs/quota.h> 59 #include <ufs/ufs/inode.h> 60 #include <ufs/ext2fs/ext2fs.h> 61 #include <ufs/ext2fs/ext2fs_extern.h> 62 63 64 #define doclusterread 0 /* XXX underway */ 65 #define doclusterwrite 0 66 67 /* 68 * Vnode op for reading. 69 */ 70 /* ARGSUSED */ 71 int 72 ext2fs_read(v) 73 void *v; 74 { 75 struct vop_read_args /* { 76 struct vnode *a_vp; 77 struct uio *a_uio; 78 int a_ioflag; 79 struct ucred *a_cred; 80 } */ *ap = v; 81 struct vnode *vp; 82 struct inode *ip; 83 struct uio *uio; 84 struct m_ext2fs *fs; 85 struct buf *bp; 86 ufs_daddr_t lbn, nextlbn; 87 off_t bytesinfile; 88 long size, xfersize, blkoffset; 89 int error; 90 91 vp = ap->a_vp; 92 ip = VTOI(vp); 93 uio = ap->a_uio; 94 95 #ifdef DIAGNOSTIC 96 if (uio->uio_rw != UIO_READ) 97 panic("%s: mode", "ext2fs_read"); 98 99 if (vp->v_type == VLNK) { 100 if ((int)ip->i_e2fs_size < vp->v_mount->mnt_maxsymlinklen || 101 (vp->v_mount->mnt_maxsymlinklen == 0 && 102 ip->i_e2fs_nblock == 0)) 103 panic("%s: short symlink", "ext2fs_read"); 104 } else if (vp->v_type != VREG && vp->v_type != VDIR) 105 panic("%s: type %d", "ext2fs_read", vp->v_type); 106 #endif 107 fs = ip->i_e2fs; 108 if ((u_int64_t)uio->uio_offset > 109 ((u_int64_t)0x80000000 * fs->e2fs_bsize - 1)) 110 return (EFBIG); 111 if (uio->uio_resid == 0) 112 return (0); 113 114 for (error = 0, bp = NULL; uio->uio_resid > 0; bp = NULL) { 115 if ((bytesinfile = ip->i_e2fs_size - uio->uio_offset) <= 0) 116 break; 117 lbn = lblkno(fs, uio->uio_offset); 118 nextlbn = lbn + 1; 119 size = fs->e2fs_bsize; 120 blkoffset = blkoff(fs, uio->uio_offset); 121 xfersize = fs->e2fs_bsize - blkoffset; 122 if (uio->uio_resid < xfersize) 123 xfersize = uio->uio_resid; 124 if (bytesinfile < xfersize) 125 xfersize = bytesinfile; 126 127 if (lblktosize(fs, nextlbn) >= ip->i_e2fs_size) 128 error = bread(vp, lbn, size, NOCRED, &bp); 129 else if (doclusterread) 130 error = cluster_read(vp, 131 ip->i_e2fs_size, lbn, size, NOCRED, &bp); 132 else if (lbn - 1 == vp->v_lastr) { 133 int nextsize = fs->e2fs_bsize; 134 error = breadn(vp, lbn, 135 size, &nextlbn, &nextsize, 1, NOCRED, &bp); 136 } else 137 error = bread(vp, lbn, size, NOCRED, &bp); 138 if (error) 139 break; 140 vp->v_lastr = lbn; 141 142 /* 143 * We should only get non-zero b_resid when an I/O error 144 * has occurred, which should cause us to break above. 145 * However, if the short read did not cause an error, 146 * then we want to ensure that we do not uiomove bad 147 * or uninitialized data. 148 */ 149 size -= bp->b_resid; 150 if (size < xfersize) { 151 if (size == 0) 152 break; 153 xfersize = size; 154 } 155 error = uiomove((char *)bp->b_data + blkoffset, (int)xfersize, 156 uio); 157 if (error) 158 break; 159 brelse(bp); 160 } 161 if (bp != NULL) 162 brelse(bp); 163 if (!(vp->v_mount->mnt_flag & MNT_NOATIME)) { 164 ip->i_flag |= IN_ACCESS; 165 if ((ap->a_ioflag & IO_SYNC) == IO_SYNC) 166 error = VOP_UPDATE(vp, NULL, NULL, UPDATE_WAIT); 167 } 168 return (error); 169 } 170 171 /* 172 * Vnode op for writing. 173 */ 174 int 175 ext2fs_write(v) 176 void *v; 177 { 178 struct vop_write_args /* { 179 struct vnode *a_vp; 180 struct uio *a_uio; 181 int a_ioflag; 182 struct ucred *a_cred; 183 } */ *ap = v; 184 struct vnode *vp; 185 struct uio *uio; 186 struct inode *ip; 187 struct m_ext2fs *fs; 188 struct buf *bp; 189 struct proc *p; 190 ufs_daddr_t lbn; 191 off_t osize; 192 int blkoffset, error, flags, ioflag, resid, size, xfersize; 193 194 ioflag = ap->a_ioflag; 195 uio = ap->a_uio; 196 vp = ap->a_vp; 197 ip = VTOI(vp); 198 199 #ifdef DIAGNOSTIC 200 if (uio->uio_rw != UIO_WRITE) 201 panic("%s: mode", "ext2fs_write"); 202 #endif 203 204 switch (vp->v_type) { 205 case VREG: 206 if (ioflag & IO_APPEND) 207 uio->uio_offset = ip->i_e2fs_size; 208 if ((ip->i_e2fs_flags & EXT2_APPEND) && 209 uio->uio_offset != ip->i_e2fs_size) 210 return (EPERM); 211 /* FALLTHROUGH */ 212 case VLNK: 213 break; 214 case VDIR: 215 if ((ioflag & IO_SYNC) == 0) 216 panic("%s: nonsync dir write", "ext2fs_write"); 217 break; 218 default: 219 panic("%s: type", "ext2fs_write"); 220 } 221 222 fs = ip->i_e2fs; 223 if (uio->uio_offset < 0 || 224 (u_int64_t)uio->uio_offset + uio->uio_resid > 225 ((u_int64_t)0x80000000 * fs->e2fs_bsize - 1)) 226 return (EFBIG); 227 /* 228 * Maybe this should be above the vnode op call, but so long as 229 * file servers have no limits, I don't think it matters. 230 */ 231 p = uio->uio_procp; 232 if (vp->v_type == VREG && p && 233 uio->uio_offset + uio->uio_resid > 234 p->p_rlimit[RLIMIT_FSIZE].rlim_cur) { 235 psignal(p, SIGXFSZ); 236 return (EFBIG); 237 } 238 239 resid = uio->uio_resid; 240 osize = ip->i_e2fs_size; 241 flags = ioflag & IO_SYNC ? B_SYNC : 0; 242 243 for (error = 0; uio->uio_resid > 0;) { 244 lbn = lblkno(fs, uio->uio_offset); 245 blkoffset = blkoff(fs, uio->uio_offset); 246 xfersize = fs->e2fs_bsize - blkoffset; 247 if (uio->uio_resid < xfersize) 248 xfersize = uio->uio_resid; 249 if (fs->e2fs_bsize > xfersize) 250 flags |= B_CLRBUF; 251 else 252 flags &= ~B_CLRBUF; 253 254 error = ext2fs_balloc(ip, 255 lbn, blkoffset + xfersize, ap->a_cred, &bp, flags); 256 if (error) 257 break; 258 if (uio->uio_offset + xfersize > ip->i_e2fs_size) { 259 ip->i_e2fs_size = uio->uio_offset + xfersize; 260 uvm_vnp_setsize(vp, ip->i_e2fs_size); 261 } 262 (void)uvm_vnp_uncache(vp); 263 264 size = fs->e2fs_bsize - bp->b_resid; 265 if (size < xfersize) 266 xfersize = size; 267 268 error = 269 uiomove((char *)bp->b_data + blkoffset, (int)xfersize, uio); 270 if (ioflag & IO_SYNC) 271 (void)bwrite(bp); 272 else if (xfersize + blkoffset == fs->e2fs_bsize) 273 if (doclusterwrite) 274 cluster_write(bp, ip->i_e2fs_size); 275 else 276 bawrite(bp); 277 else 278 bdwrite(bp); 279 if (error || xfersize == 0) 280 break; 281 ip->i_flag |= IN_CHANGE | IN_UPDATE; 282 } 283 /* 284 * If we successfully wrote any data, and we are not the superuser 285 * we clear the setuid and setgid bits as a precaution against 286 * tampering. 287 */ 288 if (resid > uio->uio_resid && ap->a_cred && ap->a_cred->cr_uid != 0) 289 ip->i_e2fs_mode &= ~(ISUID | ISGID); 290 if (error) { 291 if (ioflag & IO_UNIT) { 292 (void)VOP_TRUNCATE(vp, osize, 293 ioflag & IO_SYNC, ap->a_cred, uio->uio_procp); 294 uio->uio_offset -= resid - uio->uio_resid; 295 uio->uio_resid = resid; 296 } 297 } else if (resid > uio->uio_resid && (ioflag & IO_SYNC) == IO_SYNC) 298 error = VOP_UPDATE(vp, NULL, NULL, UPDATE_WAIT); 299 return (error); 300 } 301