summaryrefslogtreecommitdiffstats
path: root/sys/nfs/nfs_bio.c
blob: a32ca7316d846af24ab3f5344d34075fcf4d2586 (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
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
/*
 * Copyright (c) 1989, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * Rick Macklem at The University of Guelph.
 *
 * 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.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
 *
 *	@(#)nfs_bio.c	8.5 (Berkeley) 1/4/94
 * $Id: nfs_bio.c,v 1.18 1995/10/29 15:32:46 phk Exp $
 */

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/resourcevar.h>
#include <sys/signalvar.h>
#include <sys/proc.h>
#include <sys/buf.h>
#include <sys/vnode.h>
#include <sys/mount.h>
#include <sys/kernel.h>

#include <vm/vm.h>

#include <nfs/rpcv2.h>
#include <nfs/nfsproto.h>
#include <nfs/nfs.h>
#include <nfs/nfsmount.h>
#include <nfs/nqnfs.h>
#include <nfs/nfsnode.h>

extern struct buf *nfs_getcacheblk __P((struct vnode *vp, daddr_t bn, int size,
					struct proc *p));

extern struct proc *nfs_iodwant[NFS_MAXASYNCDAEMON];
extern int nfs_numasync;
extern struct nfsstats nfsstats;

/*
 * Ifdefs for FreeBSD-current's merged VM/buffer cache. It is unfortunate
 * that this isn't done inside getblk() and brelse() so these calls
 * wouldn't need to be here.
 */
#ifdef B_VMIO
#define vnode_pager_uncache(vp)
#else
#define vfs_busy_pages(bp, f)
#define vfs_unbusy_pages(bp)
#define vfs_dirty_pages(bp)
#endif

/*
 * Vnode op for read using bio
 * Any similarity to readip() is purely coincidental
 */
int
nfs_bioread(vp, uio, ioflag, cred)
	register struct vnode *vp;
	register struct uio *uio;
	int ioflag;
	struct ucred *cred;
{
	register struct nfsnode *np = VTONFS(vp);
	register int biosize, diff, i;
	struct buf *bp = 0, *rabp;
	struct vattr vattr;
	struct proc *p;
	struct nfsmount *nmp = VFSTONFS(vp->v_mount);
	daddr_t lbn, rabn;
	int bufsize;
	int nra, error = 0, n = 0, on = 0, not_readin;

#ifdef DIAGNOSTIC
	if (uio->uio_rw != UIO_READ)
		panic("nfs_read mode");
#endif
	if (uio->uio_resid == 0)
		return (0);
	if (uio->uio_offset < 0)
		return (EINVAL);
	p = uio->uio_procp;
	if ((nmp->nm_flag & (NFSMNT_NFSV3 | NFSMNT_GOTFSINFO)) == NFSMNT_NFSV3)
		(void)nfs_fsinfo(nmp, vp, cred, p);
	biosize = vp->v_mount->mnt_stat.f_iosize;
	/*
	 * For nfs, cache consistency can only be maintained approximately.
	 * Although RFC1094 does not specify the criteria, the following is
	 * believed to be compatible with the reference port.
	 * For nqnfs, full cache consistency is maintained within the loop.
	 * For nfs:
	 * If the file's modify time on the server has changed since the
	 * last read rpc or you have written to the file,
	 * you may have lost data cache consistency with the
	 * server, so flush all of the file's data out of the cache.
	 * Then force a getattr rpc to ensure that you have up to date
	 * attributes.
	 * NB: This implies that cache data can be read when up to
	 * NFS_ATTRTIMEO seconds out of date. If you find that you need current
	 * attributes this could be forced by setting n_attrstamp to 0 before
	 * the VOP_GETATTR() call.
	 */
	if ((nmp->nm_flag & NFSMNT_NQNFS) == 0) {
		if (np->n_flag & NMODIFIED) {
			if (vp->v_type != VREG) {
				if (vp->v_type != VDIR)
					panic("nfs: bioread, not dir");
				nfs_invaldir(vp);
				error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
				if (error)
					return (error);
			}
			np->n_attrstamp = 0;
			error = VOP_GETATTR(vp, &vattr, cred, p);
			if (error)
				return (error);
			np->n_mtime = vattr.va_mtime.ts_sec;
		} else {
			error = VOP_GETATTR(vp, &vattr, cred, p);
			if (error)
				return (error);
			if (np->n_mtime != vattr.va_mtime.ts_sec) {
				if (vp->v_type == VDIR)
					nfs_invaldir(vp);
				error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
				if (error)
					return (error);
				np->n_mtime = vattr.va_mtime.ts_sec;
			}
		}
	}
	do {

	    /*
	     * Get a valid lease. If cached data is stale, flush it.
	     */
	    if (nmp->nm_flag & NFSMNT_NQNFS) {
		if (NQNFS_CKINVALID(vp, np, ND_READ)) {
		    do {
			error = nqnfs_getlease(vp, ND_READ, cred, p);
		    } while (error == NQNFS_EXPIRED);
		    if (error)
			return (error);
		    if (np->n_lrev != np->n_brev ||
			(np->n_flag & NQNFSNONCACHE) ||
			((np->n_flag & NMODIFIED) && vp->v_type == VDIR)) {
			if (vp->v_type == VDIR)
			    nfs_invaldir(vp);
			error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
			if (error)
			    return (error);
			np->n_brev = np->n_lrev;
		    }
		} else if (vp->v_type == VDIR && (np->n_flag & NMODIFIED)) {
		    nfs_invaldir(vp);
		    error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
		    if (error)
			return (error);
		}
	    }
	    if (np->n_flag & NQNFSNONCACHE) {
		switch (vp->v_type) {
		case VREG:
			return (nfs_readrpc(vp, uio, cred));
		case VLNK:
			return (nfs_readlinkrpc(vp, uio, cred));
		case VDIR:
			break;
		default:
			printf(" NQNFSNONCACHE: type %x unexpected\n",
				vp->v_type);
		};
	    }
	    switch (vp->v_type) {
	    case VREG:
		nfsstats.biocache_reads++;
		lbn = uio->uio_offset / biosize;
		on = uio->uio_offset & (biosize - 1);
		not_readin = 1;

		/*
		 * Start the read ahead(s), as required.
		 */
		if (nfs_numasync > 0 && nmp->nm_readahead > 0) {
		    for (nra = 0; nra < nmp->nm_readahead &&
			(lbn + 1 + nra) * biosize < np->n_size; nra++) {
			rabn = lbn + 1 + nra;
			if (!incore(vp, rabn)) {
			    rabp = nfs_getcacheblk(vp, rabn, biosize, p);
			    if (!rabp)
				return (EINTR);
			    if ((rabp->b_flags & (B_CACHE|B_DELWRI)) == 0) {
				rabp->b_flags |= (B_READ | B_ASYNC);
				vfs_busy_pages(rabp, 0);
				if (nfs_asyncio(rabp, cred)) {
				    rabp->b_flags |= B_INVAL|B_ERROR;
				    vfs_unbusy_pages(rabp);
				    brelse(rabp);
				}
			    } else {
				brelse(rabp);
			    }
			}
		    }
		}

		/*
		 * If the block is in the cache and has the required data
		 * in a valid region, just copy it out.
		 * Otherwise, get the block and write back/read in,
		 * as required.
		 */
again:
		bufsize = biosize;
		if ((lbn + 1) * biosize > np->n_size) {
			bufsize = np->n_size - lbn * biosize;
			bufsize = (bufsize + DEV_BSIZE - 1) & ~(DEV_BSIZE - 1);
		}
		bp = nfs_getcacheblk(vp, lbn, bufsize, p);
		if (!bp)
			return (EINTR);
		if ((bp->b_flags & B_CACHE) == 0) {
			bp->b_flags |= B_READ;
			not_readin = 0;
			vfs_busy_pages(bp, 0);
			error = nfs_doio(bp, cred, p);
			if (error) {
			    brelse(bp);
			    return (error);
			}
		}
		if (bufsize > on) {
			n = min((unsigned)(bufsize - on), uio->uio_resid);
		} else {
			n = 0;
		}
		diff = np->n_size - uio->uio_offset;
		if (diff < n)
			n = diff;
		if (not_readin && n > 0) {
			if (on < bp->b_validoff || (on + n) > bp->b_validend) {
				bp->b_flags |= B_NOCACHE;
				if (bp->b_dirtyend > 0) {
				    if ((bp->b_flags & B_DELWRI) == 0)
					panic("nfsbioread");
				    if (VOP_BWRITE(bp) == EINTR)
					return (EINTR);
				} else
				    brelse(bp);
				goto again;
			}
		}
		vp->v_lastr = lbn;
		diff = (on >= bp->b_validend) ? 0 : (bp->b_validend - on);
		if (diff < n)
			n = diff;
		break;
	    case VLNK:
		nfsstats.biocache_readlinks++;
		bp = nfs_getcacheblk(vp, (daddr_t)0, NFS_MAXPATHLEN, p);
		if (!bp)
			return (EINTR);
		if ((bp->b_flags & B_CACHE) == 0) {
			bp->b_flags |= B_READ;
			vfs_busy_pages(bp, 0);
			error = nfs_doio(bp, cred, p);
			if (error) {
				bp->b_flags |= B_ERROR;
				brelse(bp);
				return (error);
			}
		}
		n = min(uio->uio_resid, NFS_MAXPATHLEN - bp->b_resid);
		on = 0;
		break;
	    case VDIR:
		nfsstats.biocache_readdirs++;
		lbn = uio->uio_offset / NFS_DIRBLKSIZ;
		on = uio->uio_offset & (NFS_DIRBLKSIZ - 1);
		bp = nfs_getcacheblk(vp, lbn, NFS_DIRBLKSIZ, p);
		if (!bp)
		    return (EINTR);
		if ((bp->b_flags & B_CACHE) == 0) {
		    bp->b_flags |= B_READ;
		    vfs_busy_pages(bp, 0);
		    error = nfs_doio(bp, cred, p);
		    if (error) {
			brelse(bp);
			while (error == NFSERR_BAD_COOKIE) {
			    nfs_invaldir(vp);
			    error = nfs_vinvalbuf(vp, 0, cred, p, 1);
			    /*
			     * Yuck! The directory has been modified on the
			     * server. The only way to get the block is by
			     * reading from the beginning to get all the
			     * offset cookies.
			     */
			    for (i = 0; i <= lbn && !error; i++) {
				bp = nfs_getcacheblk(vp, i, NFS_DIRBLKSIZ, p);
				if (!bp)
				    return (EINTR);
				if ((bp->b_flags & B_DONE) == 0) {
				    bp->b_flags |= B_READ;
				    vfs_busy_pages(bp, 0);
				    error = nfs_doio(bp, cred, p);
				    if (error)
					brelse(bp);
				}
			    }
			}
			if (error)
			    return (error);
		    }
		}

		/*
		 * If not eof and read aheads are enabled, start one.
		 * (You need the current block first, so that you have the
		 *  directory offset cookie of the next block.)
		 */
		if (nfs_numasync > 0 && nmp->nm_readahead > 0 &&
		    (np->n_direofoffset == 0 ||
		    (lbn + 1) * NFS_DIRBLKSIZ < np->n_direofoffset) &&
		    !(np->n_flag & NQNFSNONCACHE) &&
		    !incore(vp, lbn + 1)) {
			rabp = nfs_getcacheblk(vp, lbn + 1, NFS_DIRBLKSIZ, p);
			if (rabp) {
			    if ((rabp->b_flags & (B_CACHE|B_DELWRI)) == 0) {
				rabp->b_flags |= (B_READ | B_ASYNC);
				vfs_busy_pages(rabp, 0);
				if (nfs_asyncio(rabp, cred)) {
				    rabp->b_flags |= B_INVAL|B_ERROR;
				    vfs_unbusy_pages(rabp);
				    brelse(rabp);
				}
			    } else {
				brelse(rabp);
			    }
			}
		}
		n = min(uio->uio_resid, NFS_DIRBLKSIZ - bp->b_resid - on);
		break;
	    default:
		printf(" nfs_bioread: type %x unexpected\n",vp->v_type);
		break;
	    };

	    if (n > 0) {
		error = uiomove(bp->b_data + on, (int)n, uio);
	    }
	    switch (vp->v_type) {
	    case VREG:
		break;
	    case VLNK:
		n = 0;
		break;
	    case VDIR:
		if (np->n_flag & NQNFSNONCACHE)
			bp->b_flags |= B_INVAL;
		break;
	    default:
		printf(" nfs_bioread: type %x unexpected\n",vp->v_type);
	    }
 	    brelse(bp);
	} while (error == 0 && uio->uio_resid > 0 && n > 0);
	return (error);
}

/*
 * Vnode op for write using bio
 */
int
nfs_write(ap)
	struct vop_write_args /* {
		struct vnode *a_vp;
		struct uio *a_uio;
		int  a_ioflag;
		struct ucred *a_cred;
	} */ *ap;
{
	register int biosize;
	register struct uio *uio = ap->a_uio;
	struct proc *p = uio->uio_procp;
	register struct vnode *vp = ap->a_vp;
	struct nfsnode *np = VTONFS(vp);
	register struct ucred *cred = ap->a_cred;
	int ioflag = ap->a_ioflag;
	struct buf *bp;
	struct vattr vattr;
	struct nfsmount *nmp = VFSTONFS(vp->v_mount);
	daddr_t lbn;
	int bufsize;
	int n, on, error = 0, iomode, must_commit;

#ifdef DIAGNOSTIC
	if (uio->uio_rw != UIO_WRITE)
		panic("nfs_write mode");
	if (uio->uio_segflg == UIO_USERSPACE && uio->uio_procp != curproc)
		panic("nfs_write proc");
#endif
	if (vp->v_type != VREG)
		return (EIO);
	if (np->n_flag & NWRITEERR) {
		np->n_flag &= ~NWRITEERR;
		return (np->n_error);
	}
	if ((nmp->nm_flag & (NFSMNT_NFSV3 | NFSMNT_GOTFSINFO)) == NFSMNT_NFSV3)
		(void)nfs_fsinfo(nmp, vp, cred, p);
	if (ioflag & (IO_APPEND | IO_SYNC)) {
		if (np->n_flag & NMODIFIED) {
			np->n_attrstamp = 0;
			error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
			if (error)
				return (error);
		}
		if (ioflag & IO_APPEND) {
			np->n_attrstamp = 0;
			error = VOP_GETATTR(vp, &vattr, cred, p);
			if (error)
				return (error);
			uio->uio_offset = np->n_size;
		}
	}
	if (uio->uio_offset < 0)
		return (EINVAL);
	if (uio->uio_resid == 0)
		return (0);
	/*
	 * Maybe this should be above the vnode op call, but so long as
	 * file servers have no limits, i don't think it matters
	 */
	if (p && uio->uio_offset + uio->uio_resid >
	      p->p_rlimit[RLIMIT_FSIZE].rlim_cur) {
		psignal(p, SIGXFSZ);
		return (EFBIG);
	}
	/*
	 * I use nm_rsize, not nm_wsize so that all buffer cache blocks
	 * will be the same size within a filesystem. nfs_writerpc will
	 * still use nm_wsize when sizing the rpc's.
	 */
	biosize = vp->v_mount->mnt_stat.f_iosize;
	do {

		/*
		 * XXX make sure we aren't cached in the VM page cache
		 */
		/*
		 * Check for a valid write lease.
		 */
		if ((nmp->nm_flag & NFSMNT_NQNFS) &&
		    NQNFS_CKINVALID(vp, np, ND_WRITE)) {
			do {
				error = nqnfs_getlease(vp, ND_WRITE, cred, p);
			} while (error == NQNFS_EXPIRED);
			if (error)
				return (error);
			if (np->n_lrev != np->n_brev ||
			    (np->n_flag & NQNFSNONCACHE)) {
				error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
				if (error)
					return (error);
				np->n_brev = np->n_lrev;
			}
		}
		if ((np->n_flag & NQNFSNONCACHE) && uio->uio_iovcnt == 1) {
		    iomode = NFSV3WRITE_FILESYNC;
		    error = nfs_writerpc(vp, uio, cred, &iomode, &must_commit);
		    if (must_commit)
			nfs_clearcommit(vp->v_mount);
		    return (error);
		}
		nfsstats.biocache_writes++;
		lbn = uio->uio_offset / biosize;
		on = uio->uio_offset & (biosize-1);
		n = min((unsigned)(biosize - on), uio->uio_resid);
again:
		if (uio->uio_offset + n > np->n_size) {
			np->n_size = uio->uio_offset + n;
			vnode_pager_setsize(vp, (u_long)np->n_size);
		}
		bufsize = biosize;
		if ((lbn + 1) * biosize > np->n_size) {
			bufsize = np->n_size - lbn * biosize;
			bufsize = (bufsize + DEV_BSIZE - 1) & ~(DEV_BSIZE - 1);
		}
		bp = nfs_getcacheblk(vp, lbn, bufsize, p);
		if (!bp)
			return (EINTR);
		if (bp->b_wcred == NOCRED) {
			crhold(cred);
			bp->b_wcred = cred;
		}
		np->n_flag |= NMODIFIED;

		if ((bp->b_blkno * DEV_BSIZE) + bp->b_dirtyend > np->n_size) {
			bp->b_dirtyend = np->n_size - (bp->b_blkno * DEV_BSIZE);
		}

		/*
		 * If the new write will leave a contiguous dirty
		 * area, just update the b_dirtyoff and b_dirtyend,
		 * otherwise force a write rpc of the old dirty area.
		 */
		if (bp->b_dirtyend > 0 &&
		    (on > bp->b_dirtyend || (on + n) < bp->b_dirtyoff)) {
			bp->b_proc = p;
			if (VOP_BWRITE(bp) == EINTR)
				return (EINTR);
			goto again;
		}

		/*
		 * Check for valid write lease and get one as required.
		 * In case getblk() and/or bwrite() delayed us.
		 */
		if ((nmp->nm_flag & NFSMNT_NQNFS) &&
		    NQNFS_CKINVALID(vp, np, ND_WRITE)) {
			do {
				error = nqnfs_getlease(vp, ND_WRITE, cred, p);
			} while (error == NQNFS_EXPIRED);
			if (error) {
				brelse(bp);
				return (error);
			}
			if (np->n_lrev != np->n_brev ||
			    (np->n_flag & NQNFSNONCACHE)) {
				brelse(bp);
				error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
				if (error)
					return (error);
				np->n_brev = np->n_lrev;
				goto again;
			}
		}
		error = uiomove((char *)bp->b_data + on, n, uio);
		if (error) {
			bp->b_flags |= B_ERROR;
			brelse(bp);
			return (error);
		}
		if (bp->b_dirtyend > 0) {
			bp->b_dirtyoff = min(on, bp->b_dirtyoff);
			bp->b_dirtyend = max((on + n), bp->b_dirtyend);
		} else {
			bp->b_dirtyoff = on;
			bp->b_dirtyend = on + n;
		}
		if (bp->b_validend == 0 || bp->b_validend < bp->b_dirtyoff ||
		    bp->b_validoff > bp->b_dirtyend) {
			bp->b_validoff = bp->b_dirtyoff;
			bp->b_validend = bp->b_dirtyend;
		} else {
			bp->b_validoff = min(bp->b_validoff, bp->b_dirtyoff);
			bp->b_validend = max(bp->b_validend, bp->b_dirtyend);
		}
		/*
		 * If the lease is non-cachable or IO_SYNC do bwrite().
		 */
		if ((np->n_flag & NQNFSNONCACHE) || (ioflag & IO_SYNC)) {
			bp->b_proc = p;
			error = VOP_BWRITE(bp);
			if (error)
				return (error);
			if (np->n_flag & NQNFSNONCACHE) {
				error = nfs_vinvalbuf(vp, V_SAVE, cred, p, 1);
				if (error)
					return (error);
			}
		} else if ((n + on) == biosize &&
			(nmp->nm_flag & NFSMNT_NQNFS) == 0) {
			bp->b_proc = (struct proc *)0;
			bp->b_flags |= B_ASYNC;
			(void)nfs_writebp(bp, 0);
		} else
			bdwrite(bp);
	} while (uio->uio_resid > 0 && n > 0);
	return (0);
}

/*
 * Get an nfs cache block.
 * Allocate a new one if the block isn't currently in the cache
 * and return the block marked busy. If the calling process is
 * interrupted by a signal for an interruptible mount point, return
 * NULL.
 */
struct buf *
nfs_getcacheblk(vp, bn, size, p)
	struct vnode *vp;
	daddr_t bn;
	int size;
	struct proc *p;
{
	register struct buf *bp;
	struct nfsmount *nmp = VFSTONFS(vp->v_mount);
	int biosize = vp->v_mount->mnt_stat.f_iosize;

	if (nmp->nm_flag & NFSMNT_INT) {
		bp = getblk(vp, bn, size, PCATCH, 0);
		while (bp == (struct buf *)0) {
			if (nfs_sigintr(nmp, (struct nfsreq *)0, p))
				return ((struct buf *)0);
			bp = getblk(vp, bn, size, 0, 2 * hz);
		}
	} else
		bp = getblk(vp, bn, size, 0, 0);

	if( vp->v_type == VREG)
		bp->b_blkno = (bn * biosize) / DEV_BSIZE;

	return (bp);
}

/*
 * Flush and invalidate all dirty buffers. If another process is already
 * doing the flush, just wait for completion.
 */
int
nfs_vinvalbuf(vp, flags, cred, p, intrflg)
	struct vnode *vp;
	int flags;
	struct ucred *cred;
	struct proc *p;
	int intrflg;
{
	register struct nfsnode *np = VTONFS(vp);
	struct nfsmount *nmp = VFSTONFS(vp->v_mount);
	int error = 0, slpflag, slptimeo;

	if ((nmp->nm_flag & NFSMNT_INT) == 0)
		intrflg = 0;
	if (intrflg) {
		slpflag = PCATCH;
		slptimeo = 2 * hz;
	} else {
		slpflag = 0;
		slptimeo = 0;
	}
	/*
	 * First wait for any other process doing a flush to complete.
	 */
	while (np->n_flag & NFLUSHINPROG) {
		np->n_flag |= NFLUSHWANT;
		error = tsleep((caddr_t)&np->n_flag, PRIBIO + 2, "nfsvinval",
			slptimeo);
		if (error && intrflg && nfs_sigintr(nmp, (struct nfsreq *)0, p))
			return (EINTR);
	}

	/*
	 * Now, flush as required.
	 */
	np->n_flag |= NFLUSHINPROG;
	error = vinvalbuf(vp, flags, cred, p, slpflag, 0);
	while (error) {
		if (intrflg && nfs_sigintr(nmp, (struct nfsreq *)0, p)) {
			np->n_flag &= ~NFLUSHINPROG;
			if (np->n_flag & NFLUSHWANT) {
				np->n_flag &= ~NFLUSHWANT;
				wakeup((caddr_t)&np->n_flag);
			}
			return (EINTR);
		}
		error = vinvalbuf(vp, flags, cred, p, 0, slptimeo);
	}
	np->n_flag &= ~(NMODIFIED | NFLUSHINPROG);
	if (np->n_flag & NFLUSHWANT) {
		np->n_flag &= ~NFLUSHWANT;
		wakeup((caddr_t)&np->n_flag);
	}
	return (0);
}

/*
 * Initiate asynchronous I/O. Return an error if no nfsiods are available.
 * This is mainly to avoid queueing async I/O requests when the nfsiods
 * are all hung on a dead server.
 */
int
nfs_asyncio(bp, cred)
	register struct buf *bp;
	struct ucred *cred;
{
	register int i;

	if (nfs_numasync == 0)
		return (EIO);
	for (i = 0; i < NFS_MAXASYNCDAEMON; i++)
	    if (nfs_iodwant[i]) {
		if (bp->b_flags & B_READ) {
			if (bp->b_rcred == NOCRED && cred != NOCRED) {
				crhold(cred);
				bp->b_rcred = cred;
			}
		} else {
			bp->b_flags |= B_WRITEINPROG;
			if (bp->b_wcred == NOCRED && cred != NOCRED) {
				crhold(cred);
				bp->b_wcred = cred;
			}
		}

		TAILQ_INSERT_TAIL(&nfs_bufq, bp, b_freelist);
		nfs_iodwant[i] = (struct proc *)0;
		wakeup((caddr_t)&nfs_iodwant[i]);
		return (0);
	    }

	/*
	 * If it is a read or a write already marked B_WRITEINPROG or B_NOCACHE
	 * return EIO so the process will call nfs_doio() and do it
	 * synchronously.
	 */
	if (bp->b_flags & (B_READ | B_WRITEINPROG | B_NOCACHE))
		return (EIO);

	/*
	 * Just turn the async write into a delayed write, instead of
	 * doing in synchronously. Hopefully, at least one of the nfsiods
	 * is currently doing a write for this file and will pick up the
	 * delayed writes before going back to sleep.
	 */
	bp->b_flags |= B_DELWRI;
	reassignbuf(bp, bp->b_vp);
	biodone(bp);
	return (0);
}

/*
 * Do an I/O operation to/from a cache block. This may be called
 * synchronously or from an nfsiod.
 */
int
nfs_doio(bp, cr, p)
	register struct buf *bp;
	struct ucred *cr;
	struct proc *p;
{
	register struct uio *uiop;
	register struct vnode *vp;
	struct nfsnode *np;
	struct nfsmount *nmp;
	int error = 0, diff, len, iomode, must_commit = 0;
	struct uio uio;
	struct iovec io;

	vp = bp->b_vp;
	np = VTONFS(vp);
	nmp = VFSTONFS(vp->v_mount);
	uiop = &uio;
	uiop->uio_iov = &io;
	uiop->uio_iovcnt = 1;
	uiop->uio_segflg = UIO_SYSSPACE;
	uiop->uio_procp = p;

	/*
	 * Historically, paging was done with physio, but no more.
	 */
	if (bp->b_flags & B_PHYS) {
	    /*
	     * ...though reading /dev/drum still gets us here.
	     */
	    io.iov_len = uiop->uio_resid = bp->b_bcount;
	    /* mapping was done by vmapbuf() */
	    io.iov_base = bp->b_data;
	    uiop->uio_offset = ((off_t)bp->b_blkno) * DEV_BSIZE;
	    if (bp->b_flags & B_READ) {
		uiop->uio_rw = UIO_READ;
		nfsstats.read_physios++;
		error = nfs_readrpc(vp, uiop, cr);
	    } else {
		int com;

		iomode = NFSV3WRITE_DATASYNC;
		uiop->uio_rw = UIO_WRITE;
		nfsstats.write_physios++;
		error = nfs_writerpc(vp, uiop, cr, &iomode, &com);
	    }
	    if (error) {
		bp->b_flags |= B_ERROR;
		bp->b_error = error;
	    }
	} else if (bp->b_flags & B_READ) {
	    io.iov_len = uiop->uio_resid = bp->b_bcount;
	    io.iov_base = bp->b_data;
	    uiop->uio_rw = UIO_READ;
	    switch (vp->v_type) {
	    case VREG:
		uiop->uio_offset = ((off_t)bp->b_blkno) * DEV_BSIZE;
		nfsstats.read_bios++;
		error = nfs_readrpc(vp, uiop, cr);
		if (!error) {
		    bp->b_validoff = 0;
		    if (uiop->uio_resid) {
			/*
			 * If len > 0, there is a hole in the file and
			 * no writes after the hole have been pushed to
			 * the server yet.
			 * Just zero fill the rest of the valid area.
			 */
			diff = bp->b_bcount - uiop->uio_resid;
			len = np->n_size - (((u_quad_t)bp->b_blkno) * DEV_BSIZE
				+ diff);
			if (len > 0) {
			    len = min(len, uiop->uio_resid);
			    bzero((char *)bp->b_data + diff, len);
			    bp->b_validend = diff + len;
			} else
			    bp->b_validend = diff;
		    } else
			bp->b_validend = bp->b_bcount;
		}
		if (p && (vp->v_flag & VTEXT) &&
			(((nmp->nm_flag & NFSMNT_NQNFS) &&
			  NQNFS_CKINVALID(vp, np, ND_READ) &&
			  np->n_lrev != np->n_brev) ||
			 (!(nmp->nm_flag & NFSMNT_NQNFS) &&
			  np->n_mtime != np->n_vattr.va_mtime.ts_sec))) {
			uprintf("Process killed due to text file modification\n");
			psignal(p, SIGKILL);
#ifdef __NetBSD__
			p->p_holdcnt++;
#else
			p->p_flag |= P_NOSWAP;
#endif
		}
		break;
	    case VLNK:
		uiop->uio_offset = (off_t)0;
		nfsstats.readlink_bios++;
		error = nfs_readlinkrpc(vp, uiop, cr);
		break;
	    case VDIR:
		nfsstats.readdir_bios++;
		uiop->uio_offset = ((u_quad_t)bp->b_lblkno) * NFS_DIRBLKSIZ;
		if (nmp->nm_flag & NFSMNT_RDIRPLUS) {
			error = nfs_readdirplusrpc(vp, uiop, cr);
			if (error == NFSERR_NOTSUPP)
				nmp->nm_flag &= ~NFSMNT_RDIRPLUS;
		}
		if ((nmp->nm_flag & NFSMNT_RDIRPLUS) == 0)
			error = nfs_readdirrpc(vp, uiop, cr);
		break;
	    default:
		printf("nfs_doio:  type %x unexpected\n",vp->v_type);
		break;
	    };
	    if (error) {
		bp->b_flags |= B_ERROR;
		bp->b_error = error;
	    }
	} else {
	    if (((bp->b_blkno * DEV_BSIZE) + bp->b_dirtyend) > np->n_size)
		bp->b_dirtyend = np->n_size - (bp->b_blkno * DEV_BSIZE);

	    if (bp->b_dirtyend > bp->b_dirtyoff) {
		io.iov_len = uiop->uio_resid = bp->b_dirtyend
		    - bp->b_dirtyoff;
		uiop->uio_offset = ((off_t)bp->b_blkno) * DEV_BSIZE
		    + bp->b_dirtyoff;
		io.iov_base = (char *)bp->b_data + bp->b_dirtyoff;
		uiop->uio_rw = UIO_WRITE;
		nfsstats.write_bios++;
		if ((bp->b_flags & (B_ASYNC | B_NEEDCOMMIT | B_NOCACHE)) == B_ASYNC)
		    iomode = NFSV3WRITE_UNSTABLE;
		else
		    iomode = NFSV3WRITE_FILESYNC;
		bp->b_flags |= B_WRITEINPROG;
		error = nfs_writerpc(vp, uiop, cr, &iomode, &must_commit);
		if (!error && iomode == NFSV3WRITE_UNSTABLE)
		    bp->b_flags |= B_NEEDCOMMIT;
		else
		    bp->b_flags &= ~B_NEEDCOMMIT;
		bp->b_flags &= ~B_WRITEINPROG;

		/*
		 * For an interrupted write, the buffer is still valid
		 * and the write hasn't been pushed to the server yet,
		 * so we can't set B_ERROR and report the interruption
		 * by setting B_EINTR. For the B_ASYNC case, B_EINTR
		 * is not relevant, so the rpc attempt is essentially
		 * a noop.  For the case of a V3 write rpc not being
		 * committed to stable storage, the block is still
		 * dirty and requires either a commit rpc or another
		 * write rpc with iomode == NFSV3WRITE_FILESYNC before
		 * the block is reused. This is indicated by setting
		 * the B_DELWRI and B_NEEDCOMMIT flags.
		 */
    		if (error == EINTR
		    || (!error && (bp->b_flags & B_NEEDCOMMIT))) {
			bp->b_flags &= ~(B_INVAL|B_NOCACHE);
			bp->b_flags |= B_DELWRI;

		/*
		 * Since for the B_ASYNC case, nfs_bwrite() has reassigned the
		 * buffer to the clean list, we have to reassign it back to the
		 * dirty one. Ugh.
		 */
			if (bp->b_flags & B_ASYNC)
				reassignbuf(bp, vp);
			else
				bp->b_flags |= B_EINTR;
	    	} else {
			if (error) {
				bp->b_flags |= B_ERROR;
				bp->b_error = np->n_error = error;
				np->n_flag |= NWRITEERR;
			}
			bp->b_dirtyoff = bp->b_dirtyend = 0;
		}
	    } else {
		bp->b_resid = 0;
		biodone(bp);
		return (0);
	    }
	}
	bp->b_resid = uiop->uio_resid;
	if (must_commit)
		nfs_clearcommit(vp->v_mount);
	biodone(bp);
	return (error);
}
OpenPOWER on IntegriCloud