ext2fs_readwrite.c revision 1.33 1 /* $NetBSD: ext2fs_readwrite.c,v 1.33 2004/08/15 07:19:56 mycroft 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 AUTHOR ``AS IS'' AND ANY EXPRESS OR
53 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
54 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
55 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
56 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
57 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
58 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
59 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
60 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
61 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
62 *
63 * @(#)ufs_readwrite.c 8.8 (Berkeley) 8/4/94
64 * Modified for ext2fs by Manuel Bouyer.
65 */
66
67 #include <sys/cdefs.h>
68 __KERNEL_RCSID(0, "$NetBSD: ext2fs_readwrite.c,v 1.33 2004/08/15 07:19:56 mycroft Exp $");
69
70 #include <sys/param.h>
71 #include <sys/systm.h>
72 #include <sys/resourcevar.h>
73 #include <sys/kernel.h>
74 #include <sys/file.h>
75 #include <sys/stat.h>
76 #include <sys/buf.h>
77 #include <sys/proc.h>
78 #include <sys/mount.h>
79 #include <sys/vnode.h>
80 #include <sys/malloc.h>
81 #include <sys/signalvar.h>
82
83 #include <ufs/ufs/inode.h>
84 #include <ufs/ufs/ufsmount.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 struct ufsmount *ump;
119
120 vp = ap->a_vp;
121 ip = VTOI(vp);
122 ump = ip->i_ump;
123 uio = ap->a_uio;
124 error = 0;
125
126 #ifdef DIAGNOSTIC
127 if (uio->uio_rw != UIO_READ)
128 panic("%s: mode", "ext2fs_read");
129
130 if (vp->v_type == VLNK) {
131 if (ip->i_e2fs_size < ump->um_maxsymlinklen ||
132 (ump->um_maxsymlinklen == 0 && ip->i_e2fs_nblock == 0))
133 panic("%s: short symlink", "ext2fs_read");
134 } else if (vp->v_type != VREG && vp->v_type != VDIR)
135 panic("%s: type %d", "ext2fs_read", vp->v_type);
136 #endif
137 fs = ip->i_e2fs;
138 if ((u_int64_t)uio->uio_offset > ump->um_maxfilesize)
139 return (EFBIG);
140 if (uio->uio_resid == 0)
141 return (0);
142 if (uio->uio_offset >= ip->i_e2fs_size)
143 goto out;
144
145 if (vp->v_type == VREG) {
146 while (uio->uio_resid > 0) {
147 bytelen = MIN(ip->i_e2fs_size - uio->uio_offset,
148 uio->uio_resid);
149 if (bytelen == 0)
150 break;
151
152 win = ubc_alloc(&vp->v_uobj, uio->uio_offset,
153 &bytelen, UBC_READ);
154 error = uiomove(win, bytelen, uio);
155 ubc_release(win, 0);
156 if (error)
157 break;
158 }
159 goto out;
160 }
161
162 for (error = 0, bp = NULL; uio->uio_resid > 0; bp = NULL) {
163 bytesinfile = ip->i_e2fs_size - uio->uio_offset;
164 if (bytesinfile <= 0)
165 break;
166 lbn = lblkno(fs, uio->uio_offset);
167 nextlbn = lbn + 1;
168 size = fs->e2fs_bsize;
169 blkoffset = blkoff(fs, uio->uio_offset);
170 xfersize = fs->e2fs_bsize - blkoffset;
171 if (uio->uio_resid < xfersize)
172 xfersize = uio->uio_resid;
173 if (bytesinfile < xfersize)
174 xfersize = bytesinfile;
175
176 if (lblktosize(fs, nextlbn) >= ip->i_e2fs_size)
177 error = bread(vp, lbn, size, NOCRED, &bp);
178 else {
179 int nextsize = fs->e2fs_bsize;
180 error = breadn(vp, lbn,
181 size, &nextlbn, &nextsize, 1, NOCRED, &bp);
182 }
183 if (error)
184 break;
185
186 /*
187 * We should only get non-zero b_resid when an I/O error
188 * has occurred, which should cause us to break above.
189 * However, if the short read did not cause an error,
190 * then we want to ensure that we do not uiomove bad
191 * or uninitialized data.
192 */
193 size -= bp->b_resid;
194 if (size < xfersize) {
195 if (size == 0)
196 break;
197 xfersize = size;
198 }
199 error = uiomove((char *)bp->b_data + blkoffset, xfersize, uio);
200 if (error)
201 break;
202 brelse(bp);
203 }
204 if (bp != NULL)
205 brelse(bp);
206
207 out:
208 if (!(vp->v_mount->mnt_flag & MNT_NOATIME)) {
209 ip->i_flag |= IN_ACCESS;
210 if ((ap->a_ioflag & IO_SYNC) == IO_SYNC)
211 error = VOP_UPDATE(vp, NULL, NULL, UPDATE_WAIT);
212 }
213 return (error);
214 }
215
216 /*
217 * Vnode op for writing.
218 */
219 int
220 ext2fs_write(v)
221 void *v;
222 {
223 struct vop_write_args /* {
224 struct vnode *a_vp;
225 struct uio *a_uio;
226 int a_ioflag;
227 struct ucred *a_cred;
228 } */ *ap = v;
229 struct vnode *vp;
230 struct uio *uio;
231 struct inode *ip;
232 struct m_ext2fs *fs;
233 struct buf *bp;
234 struct proc *p;
235 daddr_t lbn;
236 off_t osize;
237 int blkoffset, error, flags, ioflag, resid, xfersize;
238 vsize_t bytelen;
239 void *win;
240 off_t oldoff;
241 boolean_t async;
242 int extended = 0;
243 struct ufsmount *ump;
244
245 ioflag = ap->a_ioflag;
246 uio = ap->a_uio;
247 vp = ap->a_vp;
248 ip = VTOI(vp);
249 ump = ip->i_ump;
250 error = 0;
251
252 KASSERT(vp->v_size == ip->i_size);
253 #ifdef DIAGNOSTIC
254 if (uio->uio_rw != UIO_WRITE)
255 panic("%s: mode", "ext2fs_write");
256 #endif
257
258 switch (vp->v_type) {
259 case VREG:
260 if (ioflag & IO_APPEND)
261 uio->uio_offset = ip->i_e2fs_size;
262 if ((ip->i_e2fs_flags & EXT2_APPEND) &&
263 uio->uio_offset != ip->i_e2fs_size)
264 return (EPERM);
265 /* FALLTHROUGH */
266 case VLNK:
267 break;
268 case VDIR:
269 if ((ioflag & IO_SYNC) == 0)
270 panic("%s: nonsync dir write", "ext2fs_write");
271 break;
272 default:
273 panic("%s: type", "ext2fs_write");
274 }
275
276 fs = ip->i_e2fs;
277 if (uio->uio_offset < 0 ||
278 (u_int64_t)uio->uio_offset + uio->uio_resid > ump->um_maxfilesize)
279 return (EFBIG);
280 /*
281 * Maybe this should be above the vnode op call, but so long as
282 * file servers have no limits, I don't think it matters.
283 */
284 p = uio->uio_procp;
285 if (vp->v_type == VREG && p &&
286 uio->uio_offset + uio->uio_resid >
287 p->p_rlimit[RLIMIT_FSIZE].rlim_cur) {
288 psignal(p, SIGXFSZ);
289 return (EFBIG);
290 }
291 if (uio->uio_resid == 0)
292 return (0);
293
294 async = vp->v_mount->mnt_flag & MNT_ASYNC;
295 resid = uio->uio_resid;
296 osize = ip->i_e2fs_size;
297
298 if (vp->v_type == VREG) {
299 while (uio->uio_resid > 0) {
300 oldoff = uio->uio_offset;
301 blkoffset = blkoff(fs, uio->uio_offset);
302 bytelen = MIN(fs->e2fs_bsize - blkoffset,
303 uio->uio_resid);
304
305 error = ufs_balloc_range(vp, uio->uio_offset,
306 bytelen, ap->a_cred, 0);
307 if (error)
308 break;
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 * update UVM's notion of the size now that we've
318 * copied the data into the vnode's pages.
319 */
320
321 if (vp->v_size < uio->uio_offset) {
322 uvm_vnp_setsize(vp, uio->uio_offset);
323 extended = 1;
324 }
325
326 /*
327 * flush what we just wrote if necessary.
328 * XXXUBC simplistic async flushing.
329 */
330
331 if (!async && oldoff >> 16 != uio->uio_offset >> 16) {
332 simple_lock(&vp->v_interlock);
333 error = VOP_PUTPAGES(vp, (oldoff >> 16) << 16,
334 (uio->uio_offset >> 16) << 16, PGO_CLEANIT);
335 }
336 }
337 if (error == 0 && ioflag & IO_SYNC) {
338 simple_lock(&vp->v_interlock);
339 error = VOP_PUTPAGES(vp, trunc_page(oldoff),
340 round_page(blkroundup(fs, uio->uio_offset)),
341 PGO_CLEANIT | PGO_SYNCIO);
342 }
343
344 goto out;
345 }
346
347 flags = ioflag & IO_SYNC ? B_SYNC : 0;
348 for (error = 0; uio->uio_resid > 0;) {
349 lbn = lblkno(fs, uio->uio_offset);
350 blkoffset = blkoff(fs, uio->uio_offset);
351 xfersize = MIN(fs->e2fs_bsize - blkoffset, uio->uio_resid);
352 if (xfersize < fs->e2fs_bsize)
353 flags |= B_CLRBUF;
354 else
355 flags &= ~B_CLRBUF;
356 error = ext2fs_balloc(ip,
357 lbn, blkoffset + xfersize, ap->a_cred, &bp, flags);
358 if (error)
359 break;
360 if (ip->i_e2fs_size < uio->uio_offset + xfersize) {
361 ip->i_e2fs_size = uio->uio_offset + xfersize;
362 }
363 error = uiomove((char *)bp->b_data + blkoffset, xfersize, uio);
364
365 /*
366 * update UVM's notion of the size now that we've
367 * copied the data into the vnode's pages.
368 */
369
370 if (vp->v_size < uio->uio_offset) {
371 uvm_vnp_setsize(vp, uio->uio_offset);
372 extended = 1;
373 }
374
375 if (ioflag & IO_SYNC)
376 (void)bwrite(bp);
377 else if (xfersize + blkoffset == fs->e2fs_bsize)
378 bawrite(bp);
379 else
380 bdwrite(bp);
381 if (error || xfersize == 0)
382 break;
383 }
384
385 /*
386 * If we successfully wrote any data, and we are not the superuser
387 * we clear the setuid and setgid bits as a precaution against
388 * tampering.
389 */
390
391 out:
392 ip->i_flag |= IN_CHANGE | IN_UPDATE;
393 if (resid > uio->uio_resid && ap->a_cred && ap->a_cred->cr_uid != 0)
394 ip->i_e2fs_mode &= ~(ISUID | ISGID);
395 if (resid > uio->uio_resid)
396 VN_KNOTE(vp, NOTE_WRITE | (extended ? NOTE_EXTEND : 0));
397 if (error) {
398 (void) VOP_TRUNCATE(vp, osize, ioflag & IO_SYNC, ap->a_cred,
399 uio->uio_procp);
400 uio->uio_offset -= resid - uio->uio_resid;
401 uio->uio_resid = resid;
402 } else if (resid > uio->uio_resid && (ioflag & IO_SYNC) == IO_SYNC)
403 error = VOP_UPDATE(vp, NULL, NULL, UPDATE_WAIT);
404 KASSERT(vp->v_size == ip->i_e2fs_size);
405 return (error);
406 }
407