summaryrefslogtreecommitdiffstats
path: root/sys/ufs/ffs/ffs_rawread.c
blob: 3afa6c3bd1193b1656c52c02cf363f92aa54fc8f (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
/*-
 * Copyright (c) 2000-2003 Tor Egge
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/fcntl.h>
#include <sys/file.h>
#include <sys/stat.h>
#include <sys/proc.h>
#include <sys/limits.h>
#include <sys/mount.h>
#include <sys/namei.h>
#include <sys/vnode.h>
#include <sys/conf.h>
#include <sys/filio.h>
#include <sys/ttycom.h>
#include <sys/bio.h>
#include <sys/buf.h>
#include <ufs/ufs/extattr.h>
#include <ufs/ufs/quota.h>
#include <ufs/ufs/inode.h>
#include <ufs/ufs/ufsmount.h>
#include <ufs/ufs/ufs_extern.h>
#include <ufs/ffs/fs.h>
#include <ufs/ffs/ffs_extern.h>

#include <vm/vm.h>
#include <vm/vm_extern.h>
#include <vm/vm_object.h>
#include <sys/kernel.h>
#include <sys/sysctl.h>

static int ffs_rawread_readahead(struct vnode *vp,
				 caddr_t udata,
				 off_t offset,
				 size_t len,
				 struct thread *td,
				 struct buf *bp,
				 caddr_t sa);
static int ffs_rawread_main(struct vnode *vp,
			    struct uio *uio);

static int ffs_rawread_sync(struct vnode *vp, struct thread *td);

int ffs_rawread(struct vnode *vp, struct uio *uio, int *workdone);

void ffs_rawread_setup(void);

SYSCTL_DECL(_vfs_ffs);

static int ffsrawbufcnt = 4;
SYSCTL_INT(_vfs_ffs, OID_AUTO, ffsrawbufcnt, CTLFLAG_RD, &ffsrawbufcnt, 0,
	   "Buffers available for raw reads");

static int allowrawread = 1;
SYSCTL_INT(_vfs_ffs, OID_AUTO, allowrawread, CTLFLAG_RW, &allowrawread, 0,
	   "Flag to enable raw reads");

static int rawreadahead = 1;
SYSCTL_INT(_vfs_ffs, OID_AUTO, rawreadahead, CTLFLAG_RW, &rawreadahead, 0,
	   "Flag to enable readahead for long raw reads");


void
ffs_rawread_setup(void)
{
	ffsrawbufcnt = (nswbuf > 100 ) ? (nswbuf - (nswbuf >> 4)) : nswbuf - 8;
}


static int
ffs_rawread_sync(struct vnode *vp, struct thread *td)
{
	int spl;
	int error;
	int upgraded;
	struct bufobj *bo;

	/* Check for dirty mmap, pending writes and dirty buffers */
	spl = splbio();
	VI_LOCK(vp);
	bo = &vp->v_bufobj;
	if (bo->bo_numoutput > 0 ||
	    bo->bo_dirty.bv_cnt > 0 ||
	    (vp->v_iflag & VI_OBJDIRTY) != 0) {
		splx(spl);
		VI_UNLOCK(vp);
		
		if (VOP_ISLOCKED(vp, td) != LK_EXCLUSIVE) {
			upgraded = 1;
			/* Upgrade to exclusive lock, this might block */
			VOP_LOCK(vp, LK_UPGRADE, td);
		} else
			upgraded = 0;
			
		
		/* Attempt to msync mmap() regions to clean dirty mmap */ 
		VI_LOCK(vp);
		if ((vp->v_iflag & VI_OBJDIRTY) != 0) {
			VI_UNLOCK(vp);
			if (vp->v_object != NULL) {
				VM_OBJECT_LOCK(vp->v_object);
				vm_object_page_clean(vp->v_object, 0, 0, OBJPC_SYNC);
				VM_OBJECT_UNLOCK(vp->v_object);
			}
			VI_LOCK(vp);
		}

		/* Wait for pending writes to complete */
		spl = splbio();
		error = bufobj_wwait(&vp->v_bufobj, 0, 0);
		if (error != 0) {
			/* XXX: can't happen with a zero timeout ??? */
			splx(spl);
			VI_UNLOCK(vp);
			if (upgraded != 0)
				VOP_LOCK(vp, LK_DOWNGRADE, td);
			return (error);
		}
		/* Flush dirty buffers */
		if (bo->bo_dirty.bv_cnt > 0) {
			splx(spl);
			VI_UNLOCK(vp);
			if ((error = ffs_syncvnode(vp, MNT_WAIT)) != 0) {
				if (upgraded != 0)
					VOP_LOCK(vp, LK_DOWNGRADE, td);
				return (error);
			}
			VI_LOCK(vp);
			spl = splbio();
			if (bo->bo_numoutput > 0 || bo->bo_dirty.bv_cnt > 0)
				panic("ffs_rawread_sync: dirty bufs");
		}
		splx(spl);
		VI_UNLOCK(vp);
		if (upgraded != 0)
			VOP_LOCK(vp, LK_DOWNGRADE, td);
	} else {
		splx(spl);
		VI_UNLOCK(vp);
	}
	return 0;
}


static int
ffs_rawread_readahead(struct vnode *vp,
		      caddr_t udata,
		      off_t offset,
		      size_t len,
		      struct thread *td,
		      struct buf *bp,
		      caddr_t sa)
{
	int error;
	u_int iolen;
	off_t blockno;
	int blockoff;
	int bsize;
	struct vnode *dp;
	int bforwards;
	struct inode *ip;
	ufs2_daddr_t blkno;
	
	bsize = vp->v_mount->mnt_stat.f_iosize;
	
	ip = VTOI(vp);
	dp = ip->i_devvp;

	iolen = ((vm_offset_t) udata) & PAGE_MASK;
	bp->b_bcount = len;
	if (bp->b_bcount + iolen > bp->b_kvasize) {
		bp->b_bcount = bp->b_kvasize;
		if (iolen != 0)
			bp->b_bcount -= PAGE_SIZE;
	}
	bp->b_flags = 0;	/* XXX necessary ? */
	bp->b_iocmd = BIO_READ;
	bp->b_iodone = bdone;
	bp->b_data = udata;
	bp->b_saveaddr = sa;
	blockno = offset / bsize;
	blockoff = (offset % bsize) / DEV_BSIZE;
	if ((daddr_t) blockno != blockno) {
		return EINVAL; /* blockno overflow */
	}
	
	bp->b_lblkno = bp->b_blkno = blockno;
	
	error = ufs_bmaparray(vp, bp->b_lblkno, &blkno, NULL, &bforwards, NULL);
	if (error != 0)
		return error;
	if (blkno == -1) {

		/* Fill holes with NULs to preserve semantics */
		
		if (bp->b_bcount + blockoff * DEV_BSIZE > bsize)
			bp->b_bcount = bsize - blockoff * DEV_BSIZE;
		bp->b_bufsize = bp->b_bcount;
		
		if (vmapbuf(bp) < 0)
			return EFAULT;
		
		if (ticks - PCPU_GET(switchticks) >= hogticks)
			uio_yield();
		bzero(bp->b_data, bp->b_bufsize);

		/* Mark operation completed (similar to bufdone()) */

		bp->b_resid = 0;
		bp->b_flags |= B_DONE;
		return 0;
	}
	bp->b_blkno = blkno + blockoff;
	bp->b_offset = bp->b_iooffset = (blkno + blockoff) * DEV_BSIZE;
	
	if (bp->b_bcount + blockoff * DEV_BSIZE > bsize * (1 + bforwards))
		bp->b_bcount = bsize * (1 + bforwards) - blockoff * DEV_BSIZE;
	bp->b_bufsize = bp->b_bcount;
	
	if (vmapbuf(bp) < 0)
		return EFAULT;
	
	BO_STRATEGY(&dp->v_bufobj, bp);
	return 0;
}


static int
ffs_rawread_main(struct vnode *vp,
		 struct uio *uio)
{
	int error, nerror;
	struct buf *bp, *nbp, *tbp;
	caddr_t sa, nsa, tsa;
	u_int iolen;
	int spl;
	caddr_t udata;
	long resid;
	off_t offset;
	struct thread *td;
	
	td = uio->uio_td ? uio->uio_td : curthread;
	udata = uio->uio_iov->iov_base;
	resid = uio->uio_resid;
	offset = uio->uio_offset;

	/*
	 * keep the process from being swapped
	 */
	PHOLD(td->td_proc);
	
	error = 0;
	nerror = 0;
	
	bp = NULL;
	nbp = NULL;
	sa = NULL;
	nsa = NULL;
	
	while (resid > 0) {
		
		if (bp == NULL) { /* Setup first read */
			/* XXX: Leave some bufs for swap */
			bp = getpbuf(&ffsrawbufcnt);
			sa = bp->b_data;
			bp->b_vp = vp; 
			error = ffs_rawread_readahead(vp, udata, offset,
						     resid, td, bp, sa);
			if (error != 0)
				break;
			
			if (resid > bp->b_bufsize) { /* Setup fist readahead */
				/* XXX: Leave bufs for swap */
				if (rawreadahead != 0) 
					nbp = trypbuf(&ffsrawbufcnt);
				else
					nbp = NULL;
				if (nbp != NULL) {
					nsa = nbp->b_data;
					nbp->b_vp = vp;
					
					nerror = ffs_rawread_readahead(vp, 
								       udata +
								       bp->b_bufsize,
								       offset +
								       bp->b_bufsize,
								       resid -
								       bp->b_bufsize,
								       td,
								       nbp,
								       nsa);
					if (nerror) {
						relpbuf(nbp, &ffsrawbufcnt);
						nbp = NULL;
					}
				}
			}
		}
		
		spl = splbio();
		bwait(bp, PRIBIO, "rawrd");
		splx(spl);
		
		vunmapbuf(bp);
		
		iolen = bp->b_bcount - bp->b_resid;
		if (iolen == 0 && (bp->b_ioflags & BIO_ERROR) == 0) {
			nerror = 0;	/* Ignore possible beyond EOF error */
			break; /* EOF */
		}
		
		if ((bp->b_ioflags & BIO_ERROR) != 0) {
			error = bp->b_error;
			break;
		}
		resid -= iolen;
		udata += iolen;
		offset += iolen;
		if (iolen < bp->b_bufsize) {
			/* Incomplete read.  Try to read remaining part */
			error = ffs_rawread_readahead(vp,
						      udata,
						      offset,
						      bp->b_bufsize - iolen,
						      td,
						      bp,
						      sa);
			if (error != 0)
				break;
		} else if (nbp != NULL) { /* Complete read with readahead */
			
			tbp = bp;
			bp = nbp;
			nbp = tbp;
			
			tsa = sa;
			sa = nsa;
			nsa = tsa;
			
			if (resid <= bp->b_bufsize) { /* No more readaheads */
				relpbuf(nbp, &ffsrawbufcnt);
				nbp = NULL;
			} else { /* Setup next readahead */
				nerror = ffs_rawread_readahead(vp,
							       udata +
							       bp->b_bufsize,
							       offset +
							       bp->b_bufsize,
							       resid -
							       bp->b_bufsize,
							       td,
							       nbp,
							       nsa);
				if (nerror != 0) {
					relpbuf(nbp, &ffsrawbufcnt);
					nbp = NULL;
				}
			}
		} else if (nerror != 0) {/* Deferred Readahead error */
			break;		
		}  else if (resid > 0) { /* More to read, no readahead */
			error = ffs_rawread_readahead(vp, udata, offset,
						      resid, td, bp, sa);
			if (error != 0)
				break;
		}
	}
	
	if (bp != NULL)
		relpbuf(bp, &ffsrawbufcnt);
	if (nbp != NULL) {			/* Run down readahead buffer */
		spl = splbio();
		bwait(nbp, PRIBIO, "rawrd");
		splx(spl);
		vunmapbuf(nbp);
		relpbuf(nbp, &ffsrawbufcnt);
	}
	
	if (error == 0)
		error = nerror;
	PRELE(td->td_proc);
	uio->uio_iov->iov_base = udata;
	uio->uio_resid = resid;
	uio->uio_offset = offset;
	return error;
}


int
ffs_rawread(struct vnode *vp,
	    struct uio *uio,
	    int *workdone)
{
	if (allowrawread != 0 &&
	    uio->uio_iovcnt == 1 && 
	    uio->uio_segflg == UIO_USERSPACE &&
	    uio->uio_resid == uio->uio_iov->iov_len &&
	    (((uio->uio_td != NULL) ? uio->uio_td : curthread)->td_pflags &
	     TDP_DEADLKTREAT) == 0) {
		int secsize;		/* Media sector size */
		off_t filebytes;	/* Bytes left of file */
		int blockbytes;		/* Bytes left of file in full blocks */
		int partialbytes;	/* Bytes in last partial block */
		int skipbytes;		/* Bytes not to read in ffs_rawread */
		struct inode *ip;
		int error;
		

		/* Only handle sector aligned reads */
		ip = VTOI(vp);
		secsize = ip->i_devvp->v_bufobj.bo_bsize;
		if ((uio->uio_offset & (secsize - 1)) == 0 &&
		    (uio->uio_resid & (secsize - 1)) == 0) {
			
			/* Sync dirty pages and buffers if needed */
			error = ffs_rawread_sync(vp,
						 (uio->uio_td != NULL) ?
						 uio->uio_td : curthread);
			if (error != 0)
				return error;
			
			/* Check for end of file */
			if (ip->i_size > uio->uio_offset) {
				filebytes = ip->i_size - uio->uio_offset;

				/* No special eof handling needed ? */
				if (uio->uio_resid <= filebytes) {
					*workdone = 1;
					return ffs_rawread_main(vp, uio);
				}
				
				partialbytes = ((unsigned int) ip->i_size) %
					ip->i_fs->fs_bsize;
				blockbytes = (int) filebytes - partialbytes;
				if (blockbytes > 0) {
					skipbytes = uio->uio_resid -
						blockbytes;
					uio->uio_resid = blockbytes;
					error = ffs_rawread_main(vp, uio);
					uio->uio_resid += skipbytes;
					if (error != 0)
						return error;
					/* Read remaining part using buffer */
				}
			}
		}
	}
	*workdone = 0;
	return 0;
}
OpenPOWER on IntegriCloud