Merge tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi...
[sfrench/cifs-2.6.git] / fs / nfs / objlayout / objlayout.c
1 /*
2  *  pNFS Objects layout driver high level definitions
3  *
4  *  Copyright (C) 2007 Panasas Inc. [year of first publication]
5  *  All rights reserved.
6  *
7  *  Benny Halevy <bhalevy@panasas.com>
8  *  Boaz Harrosh <bharrosh@panasas.com>
9  *
10  *  This program is free software; you can redistribute it and/or modify
11  *  it under the terms of the GNU General Public License version 2
12  *  See the file COPYING included with this distribution for more details.
13  *
14  *  Redistribution and use in source and binary forms, with or without
15  *  modification, are permitted provided that the following conditions
16  *  are met:
17  *
18  *  1. Redistributions of source code must retain the above copyright
19  *     notice, this list of conditions and the following disclaimer.
20  *  2. Redistributions in binary form must reproduce the above copyright
21  *     notice, this list of conditions and the following disclaimer in the
22  *     documentation and/or other materials provided with the distribution.
23  *  3. Neither the name of the Panasas company nor the names of its
24  *     contributors may be used to endorse or promote products derived
25  *     from this software without specific prior written permission.
26  *
27  *  THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
28  *  WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
29  *  MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
30  *  DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
31  *  FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32  *  CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33  *  SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
34  *  BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
35  *  LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
36  *  NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
37  *  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
38  */
39
40 #include <scsi/osd_initiator.h>
41 #include "objlayout.h"
42
43 #define NFSDBG_FACILITY         NFSDBG_PNFS_LD
44 /*
45  * Create a objlayout layout structure for the given inode and return it.
46  */
47 struct pnfs_layout_hdr *
48 objlayout_alloc_layout_hdr(struct inode *inode, gfp_t gfp_flags)
49 {
50         struct objlayout *objlay;
51
52         objlay = kzalloc(sizeof(struct objlayout), gfp_flags);
53         if (objlay) {
54                 spin_lock_init(&objlay->lock);
55                 INIT_LIST_HEAD(&objlay->err_list);
56         }
57         dprintk("%s: Return %p\n", __func__, objlay);
58         return &objlay->pnfs_layout;
59 }
60
61 /*
62  * Free an objlayout layout structure
63  */
64 void
65 objlayout_free_layout_hdr(struct pnfs_layout_hdr *lo)
66 {
67         struct objlayout *objlay = OBJLAYOUT(lo);
68
69         dprintk("%s: objlay %p\n", __func__, objlay);
70
71         WARN_ON(!list_empty(&objlay->err_list));
72         kfree(objlay);
73 }
74
75 /*
76  * Unmarshall layout and store it in pnfslay.
77  */
78 struct pnfs_layout_segment *
79 objlayout_alloc_lseg(struct pnfs_layout_hdr *pnfslay,
80                      struct nfs4_layoutget_res *lgr,
81                      gfp_t gfp_flags)
82 {
83         int status = -ENOMEM;
84         struct xdr_stream stream;
85         struct xdr_buf buf = {
86                 .pages =  lgr->layoutp->pages,
87                 .page_len =  lgr->layoutp->len,
88                 .buflen =  lgr->layoutp->len,
89                 .len = lgr->layoutp->len,
90         };
91         struct page *scratch;
92         struct pnfs_layout_segment *lseg;
93
94         dprintk("%s: Begin pnfslay %p\n", __func__, pnfslay);
95
96         scratch = alloc_page(gfp_flags);
97         if (!scratch)
98                 goto err_nofree;
99
100         xdr_init_decode(&stream, &buf, NULL);
101         xdr_set_scratch_buffer(&stream, page_address(scratch), PAGE_SIZE);
102
103         status = objio_alloc_lseg(&lseg, pnfslay, &lgr->range, &stream, gfp_flags);
104         if (unlikely(status)) {
105                 dprintk("%s: objio_alloc_lseg Return err %d\n", __func__,
106                         status);
107                 goto err;
108         }
109
110         __free_page(scratch);
111
112         dprintk("%s: Return %p\n", __func__, lseg);
113         return lseg;
114
115 err:
116         __free_page(scratch);
117 err_nofree:
118         dprintk("%s: Err Return=>%d\n", __func__, status);
119         return ERR_PTR(status);
120 }
121
122 /*
123  * Free a layout segement
124  */
125 void
126 objlayout_free_lseg(struct pnfs_layout_segment *lseg)
127 {
128         dprintk("%s: freeing layout segment %p\n", __func__, lseg);
129
130         if (unlikely(!lseg))
131                 return;
132
133         objio_free_lseg(lseg);
134 }
135
136 /*
137  * I/O Operations
138  */
139 static inline u64
140 end_offset(u64 start, u64 len)
141 {
142         u64 end;
143
144         end = start + len;
145         return end >= start ? end : NFS4_MAX_UINT64;
146 }
147
148 /* last octet in a range */
149 static inline u64
150 last_byte_offset(u64 start, u64 len)
151 {
152         u64 end;
153
154         BUG_ON(!len);
155         end = start + len;
156         return end > start ? end - 1 : NFS4_MAX_UINT64;
157 }
158
159 void _fix_verify_io_params(struct pnfs_layout_segment *lseg,
160                            struct page ***p_pages, unsigned *p_pgbase,
161                            u64 offset, unsigned long count)
162 {
163         u64 lseg_end_offset;
164
165         BUG_ON(offset < lseg->pls_range.offset);
166         lseg_end_offset = end_offset(lseg->pls_range.offset,
167                                      lseg->pls_range.length);
168         BUG_ON(offset >= lseg_end_offset);
169         WARN_ON(offset + count > lseg_end_offset);
170
171         if (*p_pgbase > PAGE_SIZE) {
172                 dprintk("%s: pgbase(0x%x) > PAGE_SIZE\n", __func__, *p_pgbase);
173                 *p_pages += *p_pgbase >> PAGE_SHIFT;
174                 *p_pgbase &= ~PAGE_MASK;
175         }
176 }
177
178 /*
179  * I/O done common code
180  */
181 static void
182 objlayout_iodone(struct objlayout_io_res *oir)
183 {
184         if (likely(oir->status >= 0)) {
185                 objio_free_result(oir);
186         } else {
187                 struct objlayout *objlay = oir->objlay;
188
189                 spin_lock(&objlay->lock);
190                 objlay->delta_space_valid = OBJ_DSU_INVALID;
191                 list_add(&objlay->err_list, &oir->err_list);
192                 spin_unlock(&objlay->lock);
193         }
194 }
195
196 /*
197  * objlayout_io_set_result - Set an osd_error code on a specific osd comp.
198  *
199  * The @index component IO failed (error returned from target). Register
200  * the error for later reporting at layout-return.
201  */
202 void
203 objlayout_io_set_result(struct objlayout_io_res *oir, unsigned index,
204                         struct pnfs_osd_objid *pooid, int osd_error,
205                         u64 offset, u64 length, bool is_write)
206 {
207         struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[index];
208
209         BUG_ON(index >= oir->num_comps);
210         if (osd_error) {
211                 ioerr->oer_component = *pooid;
212                 ioerr->oer_comp_offset = offset;
213                 ioerr->oer_comp_length = length;
214                 ioerr->oer_iswrite = is_write;
215                 ioerr->oer_errno = osd_error;
216
217                 dprintk("%s: err[%d]: errno=%d is_write=%d dev(%llx:%llx) "
218                         "par=0x%llx obj=0x%llx offset=0x%llx length=0x%llx\n",
219                         __func__, index, ioerr->oer_errno,
220                         ioerr->oer_iswrite,
221                         _DEVID_LO(&ioerr->oer_component.oid_device_id),
222                         _DEVID_HI(&ioerr->oer_component.oid_device_id),
223                         ioerr->oer_component.oid_partition_id,
224                         ioerr->oer_component.oid_object_id,
225                         ioerr->oer_comp_offset,
226                         ioerr->oer_comp_length);
227         } else {
228                 /* User need not call if no error is reported */
229                 ioerr->oer_errno = 0;
230         }
231 }
232
233 /* Function scheduled on rpc workqueue to call ->nfs_readlist_complete().
234  * This is because the osd completion is called with ints-off from
235  * the block layer
236  */
237 static void _rpc_read_complete(struct work_struct *work)
238 {
239         struct rpc_task *task;
240         struct nfs_read_data *rdata;
241
242         dprintk("%s enter\n", __func__);
243         task = container_of(work, struct rpc_task, u.tk_work);
244         rdata = container_of(task, struct nfs_read_data, task);
245
246         pnfs_ld_read_done(rdata);
247 }
248
249 void
250 objlayout_read_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
251 {
252         struct nfs_read_data *rdata = oir->rpcdata;
253
254         oir->status = rdata->task.tk_status = status;
255         if (status >= 0)
256                 rdata->res.count = status;
257         else
258                 rdata->pnfs_error = status;
259         objlayout_iodone(oir);
260         /* must not use oir after this point */
261
262         dprintk("%s: Return status=%zd eof=%d sync=%d\n", __func__,
263                 status, rdata->res.eof, sync);
264
265         if (sync)
266                 pnfs_ld_read_done(rdata);
267         else {
268                 INIT_WORK(&rdata->task.u.tk_work, _rpc_read_complete);
269                 schedule_work(&rdata->task.u.tk_work);
270         }
271 }
272
273 /*
274  * Perform sync or async reads.
275  */
276 enum pnfs_try_status
277 objlayout_read_pagelist(struct nfs_read_data *rdata)
278 {
279         loff_t offset = rdata->args.offset;
280         size_t count = rdata->args.count;
281         int err;
282         loff_t eof;
283
284         eof = i_size_read(rdata->inode);
285         if (unlikely(offset + count > eof)) {
286                 if (offset >= eof) {
287                         err = 0;
288                         rdata->res.count = 0;
289                         rdata->res.eof = 1;
290                         /*FIXME: do we need to call pnfs_ld_read_done() */
291                         goto out;
292                 }
293                 count = eof - offset;
294         }
295
296         rdata->res.eof = (offset + count) >= eof;
297         _fix_verify_io_params(rdata->lseg, &rdata->args.pages,
298                               &rdata->args.pgbase,
299                               rdata->args.offset, rdata->args.count);
300
301         dprintk("%s: inode(%lx) offset 0x%llx count 0x%Zx eof=%d\n",
302                 __func__, rdata->inode->i_ino, offset, count, rdata->res.eof);
303
304         err = objio_read_pagelist(rdata);
305  out:
306         if (unlikely(err)) {
307                 rdata->pnfs_error = err;
308                 dprintk("%s: Returned Error %d\n", __func__, err);
309                 return PNFS_NOT_ATTEMPTED;
310         }
311         return PNFS_ATTEMPTED;
312 }
313
314 /* Function scheduled on rpc workqueue to call ->nfs_writelist_complete().
315  * This is because the osd completion is called with ints-off from
316  * the block layer
317  */
318 static void _rpc_write_complete(struct work_struct *work)
319 {
320         struct rpc_task *task;
321         struct nfs_write_data *wdata;
322
323         dprintk("%s enter\n", __func__);
324         task = container_of(work, struct rpc_task, u.tk_work);
325         wdata = container_of(task, struct nfs_write_data, task);
326
327         pnfs_ld_write_done(wdata);
328 }
329
330 void
331 objlayout_write_done(struct objlayout_io_res *oir, ssize_t status, bool sync)
332 {
333         struct nfs_write_data *wdata = oir->rpcdata;
334
335         oir->status = wdata->task.tk_status = status;
336         if (status >= 0) {
337                 wdata->res.count = status;
338                 wdata->verf.committed = oir->committed;
339         } else {
340                 wdata->pnfs_error = status;
341         }
342         objlayout_iodone(oir);
343         /* must not use oir after this point */
344
345         dprintk("%s: Return status %zd committed %d sync=%d\n", __func__,
346                 status, wdata->verf.committed, sync);
347
348         if (sync)
349                 pnfs_ld_write_done(wdata);
350         else {
351                 INIT_WORK(&wdata->task.u.tk_work, _rpc_write_complete);
352                 schedule_work(&wdata->task.u.tk_work);
353         }
354 }
355
356 /*
357  * Perform sync or async writes.
358  */
359 enum pnfs_try_status
360 objlayout_write_pagelist(struct nfs_write_data *wdata,
361                          int how)
362 {
363         int err;
364
365         _fix_verify_io_params(wdata->lseg, &wdata->args.pages,
366                               &wdata->args.pgbase,
367                               wdata->args.offset, wdata->args.count);
368
369         err = objio_write_pagelist(wdata, how);
370         if (unlikely(err)) {
371                 wdata->pnfs_error = err;
372                 dprintk("%s: Returned Error %d\n", __func__, err);
373                 return PNFS_NOT_ATTEMPTED;
374         }
375         return PNFS_ATTEMPTED;
376 }
377
378 void
379 objlayout_encode_layoutcommit(struct pnfs_layout_hdr *pnfslay,
380                               struct xdr_stream *xdr,
381                               const struct nfs4_layoutcommit_args *args)
382 {
383         struct objlayout *objlay = OBJLAYOUT(pnfslay);
384         struct pnfs_osd_layoutupdate lou;
385         __be32 *start;
386
387         dprintk("%s: Begin\n", __func__);
388
389         spin_lock(&objlay->lock);
390         lou.dsu_valid = (objlay->delta_space_valid == OBJ_DSU_VALID);
391         lou.dsu_delta = objlay->delta_space_used;
392         objlay->delta_space_used = 0;
393         objlay->delta_space_valid = OBJ_DSU_INIT;
394         lou.olu_ioerr_flag = !list_empty(&objlay->err_list);
395         spin_unlock(&objlay->lock);
396
397         start = xdr_reserve_space(xdr, 4);
398
399         BUG_ON(pnfs_osd_xdr_encode_layoutupdate(xdr, &lou));
400
401         *start = cpu_to_be32((xdr->p - start - 1) * 4);
402
403         dprintk("%s: Return delta_space_used %lld err %d\n", __func__,
404                 lou.dsu_delta, lou.olu_ioerr_flag);
405 }
406
407 static int
408 err_prio(u32 oer_errno)
409 {
410         switch (oer_errno) {
411         case 0:
412                 return 0;
413
414         case PNFS_OSD_ERR_RESOURCE:
415                 return OSD_ERR_PRI_RESOURCE;
416         case PNFS_OSD_ERR_BAD_CRED:
417                 return OSD_ERR_PRI_BAD_CRED;
418         case PNFS_OSD_ERR_NO_ACCESS:
419                 return OSD_ERR_PRI_NO_ACCESS;
420         case PNFS_OSD_ERR_UNREACHABLE:
421                 return OSD_ERR_PRI_UNREACHABLE;
422         case PNFS_OSD_ERR_NOT_FOUND:
423                 return OSD_ERR_PRI_NOT_FOUND;
424         case PNFS_OSD_ERR_NO_SPACE:
425                 return OSD_ERR_PRI_NO_SPACE;
426         default:
427                 WARN_ON(1);
428                 /* fallthrough */
429         case PNFS_OSD_ERR_EIO:
430                 return OSD_ERR_PRI_EIO;
431         }
432 }
433
434 static void
435 merge_ioerr(struct pnfs_osd_ioerr *dest_err,
436             const struct pnfs_osd_ioerr *src_err)
437 {
438         u64 dest_end, src_end;
439
440         if (!dest_err->oer_errno) {
441                 *dest_err = *src_err;
442                 /* accumulated device must be blank */
443                 memset(&dest_err->oer_component.oid_device_id, 0,
444                         sizeof(dest_err->oer_component.oid_device_id));
445
446                 return;
447         }
448
449         if (dest_err->oer_component.oid_partition_id !=
450                                 src_err->oer_component.oid_partition_id)
451                 dest_err->oer_component.oid_partition_id = 0;
452
453         if (dest_err->oer_component.oid_object_id !=
454                                 src_err->oer_component.oid_object_id)
455                 dest_err->oer_component.oid_object_id = 0;
456
457         if (dest_err->oer_comp_offset > src_err->oer_comp_offset)
458                 dest_err->oer_comp_offset = src_err->oer_comp_offset;
459
460         dest_end = end_offset(dest_err->oer_comp_offset,
461                               dest_err->oer_comp_length);
462         src_end =  end_offset(src_err->oer_comp_offset,
463                               src_err->oer_comp_length);
464         if (dest_end < src_end)
465                 dest_end = src_end;
466
467         dest_err->oer_comp_length = dest_end - dest_err->oer_comp_offset;
468
469         if ((src_err->oer_iswrite == dest_err->oer_iswrite) &&
470             (err_prio(src_err->oer_errno) > err_prio(dest_err->oer_errno))) {
471                         dest_err->oer_errno = src_err->oer_errno;
472         } else if (src_err->oer_iswrite) {
473                 dest_err->oer_iswrite = true;
474                 dest_err->oer_errno = src_err->oer_errno;
475         }
476 }
477
478 static void
479 encode_accumulated_error(struct objlayout *objlay, __be32 *p)
480 {
481         struct objlayout_io_res *oir, *tmp;
482         struct pnfs_osd_ioerr accumulated_err = {.oer_errno = 0};
483
484         list_for_each_entry_safe(oir, tmp, &objlay->err_list, err_list) {
485                 unsigned i;
486
487                 for (i = 0; i < oir->num_comps; i++) {
488                         struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[i];
489
490                         if (!ioerr->oer_errno)
491                                 continue;
492
493                         printk(KERN_ERR "%s: err[%d]: errno=%d is_write=%d "
494                                 "dev(%llx:%llx) par=0x%llx obj=0x%llx "
495                                 "offset=0x%llx length=0x%llx\n",
496                                 __func__, i, ioerr->oer_errno,
497                                 ioerr->oer_iswrite,
498                                 _DEVID_LO(&ioerr->oer_component.oid_device_id),
499                                 _DEVID_HI(&ioerr->oer_component.oid_device_id),
500                                 ioerr->oer_component.oid_partition_id,
501                                 ioerr->oer_component.oid_object_id,
502                                 ioerr->oer_comp_offset,
503                                 ioerr->oer_comp_length);
504
505                         merge_ioerr(&accumulated_err, ioerr);
506                 }
507                 list_del(&oir->err_list);
508                 objio_free_result(oir);
509         }
510
511         pnfs_osd_xdr_encode_ioerr(p, &accumulated_err);
512 }
513
514 void
515 objlayout_encode_layoutreturn(struct pnfs_layout_hdr *pnfslay,
516                               struct xdr_stream *xdr,
517                               const struct nfs4_layoutreturn_args *args)
518 {
519         struct objlayout *objlay = OBJLAYOUT(pnfslay);
520         struct objlayout_io_res *oir, *tmp;
521         __be32 *start;
522
523         dprintk("%s: Begin\n", __func__);
524         start = xdr_reserve_space(xdr, 4);
525         BUG_ON(!start);
526
527         spin_lock(&objlay->lock);
528
529         list_for_each_entry_safe(oir, tmp, &objlay->err_list, err_list) {
530                 __be32 *last_xdr = NULL, *p;
531                 unsigned i;
532                 int res = 0;
533
534                 for (i = 0; i < oir->num_comps; i++) {
535                         struct pnfs_osd_ioerr *ioerr = &oir->ioerrs[i];
536
537                         if (!ioerr->oer_errno)
538                                 continue;
539
540                         dprintk("%s: err[%d]: errno=%d is_write=%d "
541                                 "dev(%llx:%llx) par=0x%llx obj=0x%llx "
542                                 "offset=0x%llx length=0x%llx\n",
543                                 __func__, i, ioerr->oer_errno,
544                                 ioerr->oer_iswrite,
545                                 _DEVID_LO(&ioerr->oer_component.oid_device_id),
546                                 _DEVID_HI(&ioerr->oer_component.oid_device_id),
547                                 ioerr->oer_component.oid_partition_id,
548                                 ioerr->oer_component.oid_object_id,
549                                 ioerr->oer_comp_offset,
550                                 ioerr->oer_comp_length);
551
552                         p = pnfs_osd_xdr_ioerr_reserve_space(xdr);
553                         if (unlikely(!p)) {
554                                 res = -E2BIG;
555                                 break; /* accumulated_error */
556                         }
557
558                         last_xdr = p;
559                         pnfs_osd_xdr_encode_ioerr(p, &oir->ioerrs[i]);
560                 }
561
562                 /* TODO: use xdr_write_pages */
563                 if (unlikely(res)) {
564                         /* no space for even one error descriptor */
565                         BUG_ON(!last_xdr);
566
567                         /* we've encountered a situation with lots and lots of
568                          * errors and no space to encode them all. Use the last
569                          * available slot to report the union of all the
570                          * remaining errors.
571                          */
572                         encode_accumulated_error(objlay, last_xdr);
573                         goto loop_done;
574                 }
575                 list_del(&oir->err_list);
576                 objio_free_result(oir);
577         }
578 loop_done:
579         spin_unlock(&objlay->lock);
580
581         *start = cpu_to_be32((xdr->p - start - 1) * 4);
582         dprintk("%s: Return\n", __func__);
583 }
584
585
586 /*
587  * Get Device Info API for io engines
588  */
589 struct objlayout_deviceinfo {
590         struct page *page;
591         struct pnfs_osd_deviceaddr da; /* This must be last */
592 };
593
594 /* Initialize and call nfs_getdeviceinfo, then decode and return a
595  * "struct pnfs_osd_deviceaddr *" Eventually objlayout_put_deviceinfo()
596  * should be called.
597  */
598 int objlayout_get_deviceinfo(struct pnfs_layout_hdr *pnfslay,
599         struct nfs4_deviceid *d_id, struct pnfs_osd_deviceaddr **deviceaddr,
600         gfp_t gfp_flags)
601 {
602         struct objlayout_deviceinfo *odi;
603         struct pnfs_device pd;
604         struct super_block *sb;
605         struct page *page, **pages;
606         u32 *p;
607         int err;
608
609         page = alloc_page(gfp_flags);
610         if (!page)
611                 return -ENOMEM;
612
613         pages = &page;
614         pd.pages = pages;
615
616         memcpy(&pd.dev_id, d_id, sizeof(*d_id));
617         pd.layout_type = LAYOUT_OSD2_OBJECTS;
618         pd.pages = &page;
619         pd.pgbase = 0;
620         pd.pglen = PAGE_SIZE;
621         pd.mincount = 0;
622
623         sb = pnfslay->plh_inode->i_sb;
624         err = nfs4_proc_getdeviceinfo(NFS_SERVER(pnfslay->plh_inode), &pd);
625         dprintk("%s nfs_getdeviceinfo returned %d\n", __func__, err);
626         if (err)
627                 goto err_out;
628
629         p = page_address(page);
630         odi = kzalloc(sizeof(*odi), gfp_flags);
631         if (!odi) {
632                 err = -ENOMEM;
633                 goto err_out;
634         }
635         pnfs_osd_xdr_decode_deviceaddr(&odi->da, p);
636         odi->page = page;
637         *deviceaddr = &odi->da;
638         return 0;
639
640 err_out:
641         __free_page(page);
642         return err;
643 }
644
645 void objlayout_put_deviceinfo(struct pnfs_osd_deviceaddr *deviceaddr)
646 {
647         struct objlayout_deviceinfo *odi = container_of(deviceaddr,
648                                                 struct objlayout_deviceinfo,
649                                                 da);
650
651         __free_page(odi->page);
652         kfree(odi);
653 }