xref: /freebsd/lib/libufs/cgroup.c (revision 5ca8e32633c4ffbbcd6762e5888b6a4ba0708c6c)
1 /*-
2  * SPDX-License-Identifier: BSD-2-Clause
3  *
4  * Copyright (c) 2003 Juli Mallett.  All rights reserved.
5  *
6  * This software was written by Juli Mallett <jmallett@FreeBSD.org> for the
7  * FreeBSD project.  Redistribution and use in source and binary forms, with
8  * or without modification, are permitted provided that the following
9  * conditions are met:
10  *
11  * 1. Redistribution of source code must retain the above copyright notice,
12  *    this list of conditions and the following disclaimer.
13  * 2. Redistribution 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  *
17  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18  * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
19  * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
20  * DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT,
21  * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
22  * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
23  * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
25  * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
26  * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
27  * POSSIBILITY OF SUCH DAMAGE.
28  */
29 
30 #include <sys/param.h>
31 #include <sys/mount.h>
32 #include <sys/disklabel.h>
33 #include <sys/stat.h>
34 
35 #include <ufs/ufs/extattr.h>
36 #include <ufs/ufs/quota.h>
37 #include <ufs/ufs/ufsmount.h>
38 #include <ufs/ufs/dinode.h>
39 #include <ufs/ffs/fs.h>
40 
41 #include <errno.h>
42 #include <fcntl.h>
43 #include <stdio.h>
44 #include <stdlib.h>
45 #include <string.h>
46 #include <unistd.h>
47 
48 #include <libufs.h>
49 
50 ufs2_daddr_t
51 cgballoc(struct uufsd *disk)
52 {
53 	u_int8_t *blksfree;
54 	struct cg *cgp;
55 	struct fs *fs;
56 	long bno;
57 
58 	fs = &disk->d_fs;
59 	cgp = &disk->d_cg;
60 	blksfree = cg_blksfree(cgp);
61 	for (bno = 0; bno < fs->fs_fpg / fs->fs_frag; bno++)
62 		if (ffs_isblock(fs, blksfree, bno))
63 			goto gotit;
64 	return (0);
65 gotit:
66 	fs->fs_cs(fs, cgp->cg_cgx).cs_nbfree--;
67 	ffs_clrblock(fs, blksfree, (long)bno);
68 	ffs_clusteracct(fs, cgp, bno, -1);
69 	cgp->cg_cs.cs_nbfree--;
70 	fs->fs_cstotal.cs_nbfree--;
71 	fs->fs_fmod = 1;
72 	return (cgbase(fs, cgp->cg_cgx) + blkstofrags(fs, bno));
73 }
74 
75 int
76 cgbfree(struct uufsd *disk, ufs2_daddr_t bno, long size)
77 {
78 	u_int8_t *blksfree;
79 	struct fs *fs;
80 	struct cg *cgp;
81 	ufs1_daddr_t fragno, cgbno;
82 	int i, cg, blk, frags, bbase;
83 
84 	fs = &disk->d_fs;
85 	cg = dtog(fs, bno);
86 	if (cgread1(disk, cg) != 1)
87 		return (-1);
88 	cgp = &disk->d_cg;
89 	cgbno = dtogd(fs, bno);
90 	blksfree = cg_blksfree(cgp);
91 	if (size == fs->fs_bsize) {
92 		fragno = fragstoblks(fs, cgbno);
93 		ffs_setblock(fs, blksfree, fragno);
94 		ffs_clusteracct(fs, cgp, fragno, 1);
95 		cgp->cg_cs.cs_nbfree++;
96 		fs->fs_cstotal.cs_nbfree++;
97 		fs->fs_cs(fs, cg).cs_nbfree++;
98 	} else {
99 		bbase = cgbno - fragnum(fs, cgbno);
100 		/*
101 		 * decrement the counts associated with the old frags
102 		 */
103 		blk = blkmap(fs, blksfree, bbase);
104 		ffs_fragacct(fs, blk, cgp->cg_frsum, -1);
105 		/*
106 		 * deallocate the fragment
107 		 */
108 		frags = numfrags(fs, size);
109 		for (i = 0; i < frags; i++)
110 			setbit(blksfree, cgbno + i);
111 		cgp->cg_cs.cs_nffree += i;
112 		fs->fs_cstotal.cs_nffree += i;
113 		fs->fs_cs(fs, cg).cs_nffree += i;
114 		/*
115 		 * add back in counts associated with the new frags
116 		 */
117 		blk = blkmap(fs, blksfree, bbase);
118 		ffs_fragacct(fs, blk, cgp->cg_frsum, 1);
119 		/*
120 		 * if a complete block has been reassembled, account for it
121 		 */
122 		fragno = fragstoblks(fs, bbase);
123 		if (ffs_isblock(fs, blksfree, fragno)) {
124 			cgp->cg_cs.cs_nffree -= fs->fs_frag;
125 			fs->fs_cstotal.cs_nffree -= fs->fs_frag;
126 			fs->fs_cs(fs, cg).cs_nffree -= fs->fs_frag;
127 			ffs_clusteracct(fs, cgp, fragno, 1);
128 			cgp->cg_cs.cs_nbfree++;
129 			fs->fs_cstotal.cs_nbfree++;
130 			fs->fs_cs(fs, cg).cs_nbfree++;
131 		}
132 	}
133 	return cgwrite(disk);
134 }
135 
136 ino_t
137 cgialloc(struct uufsd *disk)
138 {
139 	struct ufs2_dinode *dp2;
140 	u_int8_t *inosused;
141 	struct cg *cgp;
142 	struct fs *fs;
143 	ino_t ino;
144 	int i;
145 
146 	fs = &disk->d_fs;
147 	cgp = &disk->d_cg;
148 	inosused = cg_inosused(cgp);
149 	for (ino = 0; ino < fs->fs_ipg; ino++)
150 		if (isclr(inosused, ino))
151 			goto gotit;
152 	return (0);
153 gotit:
154 	if (fs->fs_magic == FS_UFS2_MAGIC &&
155 	    ino + INOPB(fs) > cgp->cg_initediblk &&
156 	    cgp->cg_initediblk < cgp->cg_niblk) {
157 		char block[MAXBSIZE];
158 		bzero(block, (int)fs->fs_bsize);
159 		dp2 = (struct ufs2_dinode *)&block;
160 		for (i = 0; i < INOPB(fs); i++) {
161 			dp2->di_gen = arc4random();
162 			dp2++;
163 		}
164 		if (bwrite(disk, ino_to_fsba(fs,
165 		    cgp->cg_cgx * fs->fs_ipg + cgp->cg_initediblk),
166 		    block, fs->fs_bsize))
167 			return (0);
168 		cgp->cg_initediblk += INOPB(fs);
169 	}
170 
171 	setbit(inosused, ino);
172 	cgp->cg_irotor = ino;
173 	cgp->cg_cs.cs_nifree--;
174 	fs->fs_cstotal.cs_nifree--;
175 	fs->fs_cs(fs, cgp->cg_cgx).cs_nifree--;
176 	fs->fs_fmod = 1;
177 
178 	return (ino + (cgp->cg_cgx * fs->fs_ipg));
179 }
180 
181 int
182 cgread(struct uufsd *disk)
183 {
184 
185 	if (disk->d_ccg >= disk->d_fs.fs_ncg)
186 		return (0);
187 	return (cgread1(disk, disk->d_ccg++));
188 }
189 
190 /* Short read/write error messages from cgget()/cgput() */
191 static const char *failmsg;
192 
193 int
194 cgread1(struct uufsd *disk, int c)
195 {
196 
197 	if (cgget(disk->d_fd, &disk->d_fs, c, &disk->d_cg) == 0) {
198 		disk->d_lcg = c;
199 		return (1);
200 	}
201 	ERROR(disk, NULL);
202 	if (failmsg != NULL) {
203 		ERROR(disk, failmsg);
204 		return (-1);
205 	}
206 	switch (errno) {
207 	case EINTEGRITY:
208 		ERROR(disk, "cylinder group checks failed");
209 		break;
210 	case EIO:
211 		ERROR(disk, "read error from block device");
212 		break;
213 	default:
214 		ERROR(disk, strerror(errno));
215 		break;
216 	}
217 	return (-1);
218 }
219 
220 int
221 cgget(int devfd, struct fs *fs, int cg, struct cg *cgp)
222 {
223 	uint32_t cghash, calchash;
224 	size_t cnt;
225 
226 	failmsg = NULL;
227 	if ((cnt = pread(devfd, cgp, fs->fs_cgsize,
228 	    fsbtodb(fs, cgtod(fs, cg)) * (fs->fs_fsize / fsbtodb(fs,1)))) < 0)
229 		return (-1);
230 	if (cnt == 0) {
231 		failmsg = "end of file from block device";
232 		errno = EIO;
233 		return (-1);
234 	}
235 	if (cnt != fs->fs_cgsize) {
236 		failmsg = "short read from block device";
237 		errno = EIO;
238 		return (-1);
239 	}
240 	calchash = cgp->cg_ckhash;
241 	if ((fs->fs_metackhash & CK_CYLGRP) != 0) {
242 		cghash = cgp->cg_ckhash;
243 		cgp->cg_ckhash = 0;
244 		calchash = calculate_crc32c(~0L, (void *)cgp, fs->fs_cgsize);
245 		cgp->cg_ckhash = cghash;
246 	}
247 	if (cgp->cg_ckhash != calchash || !cg_chkmagic(cgp) ||
248 	    cgp->cg_cgx != cg) {
249 		errno = EINTEGRITY;
250 		return (-1);
251 	}
252 	return (0);
253 }
254 
255 int
256 cgwrite(struct uufsd *disk)
257 {
258 
259 	return (cgwrite1(disk, disk->d_cg.cg_cgx));
260 }
261 
262 int
263 cgwrite1(struct uufsd *disk, int cg)
264 {
265 	static char errmsg[BUFSIZ];
266 
267 	if (cg == disk->d_cg.cg_cgx) {
268 		if (ufs_disk_write(disk) == -1) {
269 			ERROR(disk, "failed to open disk for writing");
270 			return (-1);
271 		}
272 		if (cgput(disk->d_fd, &disk->d_fs, &disk->d_cg) == 0)
273 			return (0);
274 		ERROR(disk, NULL);
275 		if (failmsg != NULL) {
276 			ERROR(disk, failmsg);
277 			return (-1);
278 		}
279 		switch (errno) {
280 		case EIO:
281 			ERROR(disk, "unable to write cylinder group");
282 			break;
283 		default:
284 			ERROR(disk, strerror(errno));
285 			break;
286 		}
287 		return (-1);
288 	}
289 	snprintf(errmsg, BUFSIZ, "Cylinder group %d in buffer does not match "
290 	    "the cylinder group %d that cgwrite1 requested",
291 	    disk->d_cg.cg_cgx, cg);
292 	ERROR(disk, errmsg);
293 	errno = EDOOFUS;
294 	return (-1);
295 }
296 
297 int
298 cgput(int devfd, struct fs *fs, struct cg *cgp)
299 {
300 	size_t cnt;
301 
302 	if ((fs->fs_metackhash & CK_CYLGRP) != 0) {
303 		cgp->cg_ckhash = 0;
304 		cgp->cg_ckhash =
305 		    calculate_crc32c(~0L, (void *)cgp, fs->fs_cgsize);
306 	}
307 	failmsg = NULL;
308 	if ((cnt = pwrite(devfd, cgp, fs->fs_cgsize,
309 	    fsbtodb(fs, cgtod(fs, cgp->cg_cgx)) *
310 	    (fs->fs_fsize / fsbtodb(fs,1)))) < 0)
311 		return (-1);
312 	if (cnt != fs->fs_cgsize) {
313 		failmsg = "short write to block device";
314 		return (-1);
315 	}
316 	return (0);
317 }
318