1 /* $NetBSD: ext2fs_readwrite.c,v 1.31 2003/10/05 17:48:49 bouyer Exp $ */ 2 3 /*- 4 * Copyright (c) 1993 5 * The Regents of the University of California. All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 3. Neither the name of the University nor the names of its contributors 16 * may be used to endorse or promote products derived from this software 17 * without specific prior written permission. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 22 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 29 * SUCH DAMAGE. 30 * 31 * @(#)ufs_readwrite.c 8.8 (Berkeley) 8/4/94 32 * Modified for ext2fs by Manuel Bouyer. 33 */ 34 35 /*- 36 * Copyright (c) 1997 Manuel Bouyer. 37 * 38 * Redistribution and use in source and binary forms, with or without 39 * modification, are permitted provided that the following conditions 40 * are met: 41 * 1. Redistributions of source code must retain the above copyright 42 * notice, this list of conditions and the following disclaimer. 43 * 2. Redistributions in binary form must reproduce the above copyright 44 * notice, this list of conditions and the following disclaimer in the 45 * documentation and/or other materials provided with the distribution. 46 * 3. All advertising materials mentioning features or use of this software 47 * must display the following acknowledgement: 48 * This product includes software developed by Manuel Bouyer. 49 * 4. The name of the author may not be used to endorse or promote products 50 * derived from this software without specific prior written permission. 51 * 52 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 53 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 54 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 55 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 56 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 57 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 58 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 59 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 60 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 61 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 62 * SUCH DAMAGE. 63 * 64 * @(#)ufs_readwrite.c 8.8 (Berkeley) 8/4/94 65 * Modified for ext2fs by Manuel Bouyer. 66 */ 67 68 #include <sys/cdefs.h> 69 __KERNEL_RCSID(0, "$NetBSD: ext2fs_readwrite.c,v 1.31 2003/10/05 17:48:49 bouyer Exp $"); 70 71 #include <sys/param.h> 72 #include <sys/systm.h> 73 #include <sys/resourcevar.h> 74 #include <sys/kernel.h> 75 #include <sys/file.h> 76 #include <sys/stat.h> 77 #include <sys/buf.h> 78 #include <sys/proc.h> 79 #include <sys/mount.h> 80 #include <sys/vnode.h> 81 #include <sys/malloc.h> 82 #include <sys/signalvar.h> 83 84 #include <ufs/ufs/inode.h> 85 #include <ufs/ufs/ufs_extern.h> 86 #include <ufs/ext2fs/ext2fs.h> 87 #include <ufs/ext2fs/ext2fs_extern.h> 88 89 90 #define doclusterread 0 /* XXX underway */ 91 #define doclusterwrite 0 92 93 /* 94 * Vnode op for reading. 95 */ 96 /* ARGSUSED */ 97 int 98 ext2fs_read(v) 99 void *v; 100 { 101 struct vop_read_args /* { 102 struct vnode *a_vp; 103 struct uio *a_uio; 104 int a_ioflag; 105 struct ucred *a_cred; 106 } */ *ap = v; 107 struct vnode *vp; 108 struct inode *ip; 109 struct uio *uio; 110 struct m_ext2fs *fs; 111 struct buf *bp; 112 void *win; 113 vsize_t bytelen; 114 daddr_t lbn, nextlbn; 115 off_t bytesinfile; 116 long size, xfersize, blkoffset; 117 int error; 118 119 vp = ap->a_vp; 120 ip = VTOI(vp); 121 uio = ap->a_uio; 122 123 #ifdef DIAGNOSTIC 124 if (uio->uio_rw != UIO_READ) 125 panic("%s: mode", "ext2fs_read"); 126 127 if (vp->v_type == VLNK) { 128 if ((int)ip->i_e2fs_size < vp->v_mount->mnt_maxsymlinklen || 129 (vp->v_mount->mnt_maxsymlinklen == 0 && 130 ip->i_e2fs_nblock == 0)) 131 panic("%s: short symlink", "ext2fs_read"); 132 } else if (vp->v_type != VREG && vp->v_type != VDIR) 133 panic("%s: type %d", "ext2fs_read", vp->v_type); 134 #endif 135 fs = ip->i_e2fs; 136 if ((u_int64_t)uio->uio_offset > 137 ((u_int64_t)0x80000000 * fs->e2fs_bsize - 1)) 138 return (EFBIG); 139 if (uio->uio_resid == 0) 140 return (0); 141 if (uio->uio_offset >= ip->i_e2fs_size) 142 return (0); 143 144 if (vp->v_type == VREG) { 145 error = 0; 146 while (uio->uio_resid > 0) { 147 bytelen = MIN(ip->i_e2fs_size - uio->uio_offset, 148 uio->uio_resid); 149 150 if (bytelen == 0) { 151 break; 152 } 153 win = ubc_alloc(&vp->v_uobj, uio->uio_offset, 154 &bytelen, UBC_READ); 155 error = uiomove(win, bytelen, uio); 156 ubc_release(win, 0); 157 if (error) { 158 break; 159 } 160 } 161 goto out; 162 } 163 164 for (error = 0, bp = NULL; uio->uio_resid > 0; bp = NULL) { 165 if ((bytesinfile = ip->i_e2fs_size - uio->uio_offset) <= 0) 166 break; 167 lbn = lblkno(fs, uio->uio_offset); 168 nextlbn = lbn + 1; 169 size = fs->e2fs_bsize; 170 blkoffset = blkoff(fs, uio->uio_offset); 171 xfersize = fs->e2fs_bsize - blkoffset; 172 if (uio->uio_resid < xfersize) 173 xfersize = uio->uio_resid; 174 if (bytesinfile < xfersize) 175 xfersize = bytesinfile; 176 177 if (lblktosize(fs, nextlbn) >= ip->i_e2fs_size) 178 error = bread(vp, lbn, size, NOCRED, &bp); 179 else { 180 int nextsize = fs->e2fs_bsize; 181 error = breadn(vp, lbn, 182 size, &nextlbn, &nextsize, 1, NOCRED, &bp); 183 } 184 if (error) 185 break; 186 187 /* 188 * We should only get non-zero b_resid when an I/O error 189 * has occurred, which should cause us to break above. 190 * However, if the short read did not cause an error, 191 * then we want to ensure that we do not uiomove bad 192 * or uninitialized data. 193 */ 194 size -= bp->b_resid; 195 if (size < xfersize) { 196 if (size == 0) 197 break; 198 xfersize = size; 199 } 200 error = uiomove((char *)bp->b_data + blkoffset, xfersize, uio); 201 if (error) 202 break; 203 brelse(bp); 204 } 205 if (bp != NULL) 206 brelse(bp); 207 208 out: 209 if (!(vp->v_mount->mnt_flag & MNT_NOATIME)) { 210 ip->i_flag |= IN_ACCESS; 211 if ((ap->a_ioflag & IO_SYNC) == IO_SYNC) 212 error = VOP_UPDATE(vp, NULL, NULL, UPDATE_WAIT); 213 } 214 return (error); 215 } 216 217 /* 218 * Vnode op for writing. 219 */ 220 int 221 ext2fs_write(v) 222 void *v; 223 { 224 struct vop_write_args /* { 225 struct vnode *a_vp; 226 struct uio *a_uio; 227 int a_ioflag; 228 struct ucred *a_cred; 229 } */ *ap = v; 230 struct vnode *vp; 231 struct uio *uio; 232 struct inode *ip; 233 struct m_ext2fs *fs; 234 struct buf *bp; 235 struct proc *p; 236 daddr_t lbn; 237 off_t osize; 238 int blkoffset, error, flags, ioflag, resid, xfersize; 239 vsize_t bytelen; 240 void *win; 241 off_t oldoff; 242 boolean_t async; 243 int extended=0; 244 245 ioflag = ap->a_ioflag; 246 uio = ap->a_uio; 247 vp = ap->a_vp; 248 ip = VTOI(vp); 249 error = 0; 250 251 #ifdef DIAGNOSTIC 252 if (uio->uio_rw != UIO_WRITE) 253 panic("%s: mode", "ext2fs_write"); 254 #endif 255 256 switch (vp->v_type) { 257 case VREG: 258 if (ioflag & IO_APPEND) 259 uio->uio_offset = ip->i_e2fs_size; 260 if ((ip->i_e2fs_flags & EXT2_APPEND) && 261 uio->uio_offset != ip->i_e2fs_size) 262 return (EPERM); 263 /* FALLTHROUGH */ 264 case VLNK: 265 break; 266 case VDIR: 267 if ((ioflag & IO_SYNC) == 0) 268 panic("%s: nonsync dir write", "ext2fs_write"); 269 break; 270 default: 271 panic("%s: type", "ext2fs_write"); 272 } 273 274 fs = ip->i_e2fs; 275 if (uio->uio_offset < 0 || 276 (u_int64_t)uio->uio_offset + uio->uio_resid > 277 ((u_int64_t)0x80000000 * fs->e2fs_bsize - 1)) 278 return (EFBIG); 279 /* 280 * Maybe this should be above the vnode op call, but so long as 281 * file servers have no limits, I don't think it matters. 282 */ 283 p = uio->uio_procp; 284 if (vp->v_type == VREG && p && 285 uio->uio_offset + uio->uio_resid > 286 p->p_rlimit[RLIMIT_FSIZE].rlim_cur) { 287 psignal(p, SIGXFSZ); 288 return (EFBIG); 289 } 290 if (uio->uio_resid == 0) 291 return (0); 292 293 async = vp->v_mount->mnt_flag & MNT_ASYNC; 294 resid = uio->uio_resid; 295 osize = ip->i_e2fs_size; 296 297 if (vp->v_type == VREG) { 298 while (uio->uio_resid > 0) { 299 oldoff = uio->uio_offset; 300 blkoffset = blkoff(fs, uio->uio_offset); 301 bytelen = MIN(fs->e2fs_bsize - blkoffset, 302 uio->uio_resid); 303 304 error = ufs_balloc_range(vp, uio->uio_offset, 305 bytelen, ap->a_cred, 0); 306 if (error) { 307 break; 308 } 309 win = ubc_alloc(&vp->v_uobj, uio->uio_offset, 310 &bytelen, UBC_WRITE); 311 error = uiomove(win, bytelen, uio); 312 ubc_release(win, 0); 313 if (error) { 314 break; 315 } 316 317 /* 318 * update UVM's notion of the size now that we've 319 * copied the data into the vnode's pages. 320 */ 321 322 if (vp->v_size < uio->uio_offset) { 323 uvm_vnp_setsize(vp, uio->uio_offset); 324 extended = 1; 325 } 326 327 /* 328 * flush what we just wrote if necessary. 329 * XXXUBC simplistic async flushing. 330 */ 331 332 if (!async && oldoff >> 16 != uio->uio_offset >> 16) { 333 simple_lock(&vp->v_interlock); 334 error = VOP_PUTPAGES(vp, (oldoff >> 16) << 16, 335 (uio->uio_offset >> 16) << 16, PGO_CLEANIT); 336 } 337 } 338 if (error == 0 && ioflag & IO_SYNC) { 339 simple_lock(&vp->v_interlock); 340 error = VOP_PUTPAGES(vp, trunc_page(oldoff), 341 round_page(blkroundup(fs, uio->uio_offset)), 342 PGO_CLEANIT | PGO_SYNCIO); 343 } 344 345 goto out; 346 } 347 348 flags = ioflag & IO_SYNC ? B_SYNC : 0; 349 for (error = 0; uio->uio_resid > 0;) { 350 lbn = lblkno(fs, uio->uio_offset); 351 blkoffset = blkoff(fs, uio->uio_offset); 352 xfersize = MIN(fs->e2fs_bsize - blkoffset, uio->uio_resid); 353 if (xfersize < fs->e2fs_bsize) 354 flags |= B_CLRBUF; 355 else 356 flags &= ~B_CLRBUF; 357 error = ext2fs_balloc(ip, 358 lbn, blkoffset + xfersize, ap->a_cred, &bp, flags); 359 if (error) 360 break; 361 if (ip->i_e2fs_size < uio->uio_offset + xfersize) { 362 ip->i_e2fs_size = uio->uio_offset + xfersize; 363 } 364 error = uiomove((char *)bp->b_data + blkoffset, xfersize, uio); 365 366 /* 367 * update UVM's notion of the size now that we've 368 * copied the data into the vnode's pages. 369 */ 370 371 if (vp->v_size < uio->uio_offset) { 372 uvm_vnp_setsize(vp, uio->uio_offset); 373 extended = 1; 374 } 375 376 if (ioflag & IO_SYNC) 377 (void)bwrite(bp); 378 else if (xfersize + blkoffset == fs->e2fs_bsize) 379 bawrite(bp); 380 else 381 bdwrite(bp); 382 if (error || xfersize == 0) 383 break; 384 } 385 386 /* 387 * If we successfully wrote any data, and we are not the superuser 388 * we clear the setuid and setgid bits as a precaution against 389 * tampering. 390 */ 391 392 out: 393 ip->i_flag |= IN_CHANGE | IN_UPDATE; 394 if (resid > uio->uio_resid && ap->a_cred && ap->a_cred->cr_uid != 0) 395 ip->i_e2fs_mode &= ~(ISUID | ISGID); 396 if (resid > uio->uio_resid) 397 VN_KNOTE(vp, NOTE_WRITE | (extended ? NOTE_EXTEND : 0)); 398 if (error) { 399 (void) VOP_TRUNCATE(vp, osize, ioflag & IO_SYNC, ap->a_cred, 400 uio->uio_procp); 401 uio->uio_offset -= resid - uio->uio_resid; 402 uio->uio_resid = resid; 403 } else if (resid > uio->uio_resid && (ioflag & IO_SYNC) == IO_SYNC) 404 error = VOP_UPDATE(vp, NULL, NULL, UPDATE_WAIT); 405 KASSERT(vp->v_size == ip->i_e2fs_size); 406 return (error); 407 } 408