~hc/RK356X_SDK_RELEASE.git

..	..	@@ -1,3 +1,4 @@
	1	+// SPDX-License-Identifier: GPL-2.0
1	2	/*
2	3	* Copyright (C) 2016 CNEX Labs
3	4	* Initial: Javier Gonzalez <javier@cnexlabs.com>
..	..	@@ -12,9 +13,13 @@
12	13	* General Public License for more details.
13	14	*
14	15	* pblk-recovery.c - pblk's recovery path
	16	+ *
	17	+ * The L2P recovery path is single threaded as the L2P table is updated in order
	18	+ * following the line sequence ID.
15	19	*/
16	20
17	21	#include "pblk.h"
	22	+#include "pblk-trace.h"
18	23
19	24	int pblk_recov_check_emeta(struct pblk pblk, struct line_emeta emeta_buf)
20	25	{
..	..	@@ -85,134 +90,63 @@
85	90	return 0;
86	91	}
87	92
88		-static int pblk_calc_sec_in_line(struct pblk pblk, struct pblk_line line)
	93	+static void pblk_update_line_wp(struct pblk pblk, struct pblk_line line,
	94	+ u64 written_secs)
89	95	{
90		- struct nvm_tgt_dev *dev = pblk->dev;
91		- struct nvm_geo *geo = &dev->geo;
	96	+ struct pblk_line_mgmt *l_mg = &pblk->l_mg;
	97	+ int i;
	98	+
	99	+ for (i = 0; i < written_secs; i += pblk->min_write_pgs)
	100	+ __pblk_alloc_page(pblk, line, pblk->min_write_pgs);
	101	+
	102	+ spin_lock(&l_mg->free_lock);
	103	+ if (written_secs > line->left_msecs) {
	104	+ /*
	105	+ * We have all data sectors written
	106	+ * and some emeta sectors written too.
	107	+ */
	108	+ line->left_msecs = 0;
	109	+ } else {
	110	+ /* We have only some data sectors written. */
	111	+ line->left_msecs -= written_secs;
	112	+ }
	113	+ spin_unlock(&l_mg->free_lock);
	114	+}
	115	+
	116	+static u64 pblk_sec_in_open_line(struct pblk pblk, struct pblk_line line)
	117	+{
92	118	struct pblk_line_meta *lm = &pblk->lm;
93	119	int nr_bb = bitmap_weight(line->blk_bitmap, lm->blk_per_line);
	120	+ u64 written_secs = 0;
	121	+ int valid_chunks = 0;
	122	+ int i;
94	123
95		- return lm->sec_per_line - lm->smeta_sec - lm->emeta_sec[0] -
96		- nr_bb * geo->clba;
	124	+ for (i = 0; i < lm->blk_per_line; i++) {
	125	+ struct nvm_chk_meta *chunk = &line->chks[i];
	126	+
	127	+ if (chunk->state & NVM_CHK_ST_OFFLINE)
	128	+ continue;
	129	+
	130	+ written_secs += chunk->wp;
	131	+ valid_chunks++;
	132	+ }
	133	+
	134	+ if (lm->blk_per_line - nr_bb != valid_chunks)
	135	+ pblk_err(pblk, "recovery line %d is bad\n", line->id);
	136	+
	137	+ pblk_update_line_wp(pblk, line, written_secs - lm->smeta_sec);
	138	+
	139	+ return written_secs;
97	140	}
98	141
99	142	struct pblk_recov_alloc {
100	143	struct ppa_addr *ppa_list;
101		- struct pblk_sec_meta *meta_list;
	144	+ void *meta_list;
102	145	struct nvm_rq *rqd;
103	146	void *data;
104	147	dma_addr_t dma_ppa_list;
105	148	dma_addr_t dma_meta_list;
106	149	};
107		-
108		-static int pblk_recov_read_oob(struct pblk pblk, struct pblk_line line,
109		- struct pblk_recov_alloc p, u64 r_ptr)
110		-{
111		- struct nvm_tgt_dev *dev = pblk->dev;
112		- struct nvm_geo *geo = &dev->geo;
113		- struct ppa_addr *ppa_list;
114		- struct pblk_sec_meta *meta_list;
115		- struct nvm_rq *rqd;
116		- struct bio *bio;
117		- void *data;
118		- dma_addr_t dma_ppa_list, dma_meta_list;
119		- u64 r_ptr_int;
120		- int left_ppas;
121		- int rq_ppas, rq_len;
122		- int i, j;
123		- int ret = 0;
124		-
125		- ppa_list = p.ppa_list;
126		- meta_list = p.meta_list;
127		- rqd = p.rqd;
128		- data = p.data;
129		- dma_ppa_list = p.dma_ppa_list;
130		- dma_meta_list = p.dma_meta_list;
131		-
132		- left_ppas = line->cur_sec - r_ptr;
133		- if (!left_ppas)
134		- return 0;
135		-
136		- r_ptr_int = r_ptr;
137		-
138		-next_read_rq:
139		- memset(rqd, 0, pblk_g_rq_size);
140		-
141		- rq_ppas = pblk_calc_secs(pblk, left_ppas, 0);
142		- if (!rq_ppas)
143		- rq_ppas = pblk->min_write_pgs;
144		- rq_len = rq_ppas * geo->csecs;
145		-
146		- bio = bio_map_kern(dev->q, data, rq_len, GFP_KERNEL);
147		- if (IS_ERR(bio))
148		- return PTR_ERR(bio);
149		-
150		- bio->bi_iter.bi_sector = 0; /* internal bio */
151		- bio_set_op_attrs(bio, REQ_OP_READ, 0);
152		-
153		- rqd->bio = bio;
154		- rqd->opcode = NVM_OP_PREAD;
155		- rqd->meta_list = meta_list;
156		- rqd->nr_ppas = rq_ppas;
157		- rqd->ppa_list = ppa_list;
158		- rqd->dma_ppa_list = dma_ppa_list;
159		- rqd->dma_meta_list = dma_meta_list;
160		-
161		- if (pblk_io_aligned(pblk, rq_ppas))
162		- rqd->flags = pblk_set_read_mode(pblk, PBLK_READ_SEQUENTIAL);
163		- else
164		- rqd->flags = pblk_set_read_mode(pblk, PBLK_READ_RANDOM);
165		-
166		- for (i = 0; i < rqd->nr_ppas; ) {
167		- struct ppa_addr ppa;
168		- int pos;
169		-
170		- ppa = addr_to_gen_ppa(pblk, r_ptr_int, line->id);
171		- pos = pblk_ppa_to_pos(geo, ppa);
172		-
173		- while (test_bit(pos, line->blk_bitmap)) {
174		- r_ptr_int += pblk->min_write_pgs;
175		- ppa = addr_to_gen_ppa(pblk, r_ptr_int, line->id);
176		- pos = pblk_ppa_to_pos(geo, ppa);
177		- }
178		-
179		- for (j = 0; j < pblk->min_write_pgs; j++, i++, r_ptr_int++)
180		- rqd->ppa_list[i] =
181		- addr_to_gen_ppa(pblk, r_ptr_int, line->id);
182		- }
183		-
184		- /* If read fails, more padding is needed */
185		- ret = pblk_submit_io_sync(pblk, rqd);
186		- if (ret) {
187		- pblk_err(pblk, "I/O submission failed: %d\n", ret);
188		- return ret;
189		- }
190		-
191		- atomic_dec(&pblk->inflight_io);
192		-
193		- /* At this point, the read should not fail. If it does, it is a problem
194		- * we cannot recover from here. Need FTL log.
195		- */
196		- if (rqd->error && rqd->error != NVM_RSP_WARN_HIGHECC) {
197		- pblk_err(pblk, "L2P recovery failed (%d)\n", rqd->error);
198		- return -EINTR;
199		- }
200		-
201		- for (i = 0; i < rqd->nr_ppas; i++) {
202		- u64 lba = le64_to_cpu(meta_list[i].lba);
203		-
204		- if (lba == ADDR_EMPTY \|\| lba > pblk->rl.nr_secs)
205		- continue;
206		-
207		- pblk_update_map(pblk, lba, rqd->ppa_list[i]);
208		- }
209		-
210		- left_ppas -= rq_ppas;
211		- if (left_ppas > 0)
212		- goto next_read_rq;
213		-
214		- return 0;
215		-}
216	150
217	151	static void pblk_recov_complete(struct kref *ref)
218	152	{
..	..	@@ -223,10 +157,11 @@
223	157
224	158	static void pblk_end_io_recov(struct nvm_rq *rqd)
225	159	{
	160	+ struct ppa_addr *ppa_list = nvm_rq_to_ppa_list(rqd);
226	161	struct pblk_pad_rq *pad_rq = rqd->private;
227	162	struct pblk *pblk = pad_rq->pblk;
228	163
229		- pblk_up_page(pblk, rqd->ppa_list, rqd->nr_ppas);
	164	+ pblk_up_chunk(pblk, ppa_list[0]);
230	165
231	166	pblk_free_rqd(pblk, rqd, PBLK_WRITE_INT);
232	167
..	..	@@ -234,21 +169,20 @@
234	169	kref_put(&pad_rq->ref, pblk_recov_complete);
235	170	}
236	171
237		-static int pblk_recov_pad_oob(struct pblk pblk, struct pblk_line line,
238		- int left_ppas)
	172	+/* pad line using line bitmap. */
	173	+static int pblk_recov_pad_line(struct pblk pblk, struct pblk_line line,
	174	+ int left_ppas)
239	175	{
240	176	struct nvm_tgt_dev *dev = pblk->dev;
241	177	struct nvm_geo *geo = &dev->geo;
242		- struct ppa_addr *ppa_list;
243		- struct pblk_sec_meta *meta_list;
	178	+ void *meta_list;
244	179	struct pblk_pad_rq *pad_rq;
245	180	struct nvm_rq *rqd;
246		- struct bio *bio;
	181	+ struct ppa_addr *ppa_list;
247	182	void *data;
248		- dma_addr_t dma_ppa_list, dma_meta_list;
249	183	__le64 *lba_list = emeta_to_lbas(pblk, line->emeta->buf);
250	184	u64 w_ptr = line->cur_sec;
251		- int left_line_ppas, rq_ppas, rq_len;
	185	+ int left_line_ppas, rq_ppas;
252	186	int i, j;
253	187	int ret = 0;
254	188
..	..	@@ -271,45 +205,29 @@
271	205	kref_init(&pad_rq->ref);
272	206
273	207	next_pad_rq:
274		- rq_ppas = pblk_calc_secs(pblk, left_ppas, 0);
	208	+ rq_ppas = pblk_calc_secs(pblk, left_ppas, 0, false);
275	209	if (rq_ppas < pblk->min_write_pgs) {
276	210	pblk_err(pblk, "corrupted pad line %d\n", line->id);
277		- goto fail_free_pad;
	211	+ goto fail_complete;
278	212	}
279		-
280		- rq_len = rq_ppas * geo->csecs;
281		-
282		- meta_list = nvm_dev_dma_alloc(dev->parent, GFP_KERNEL, &dma_meta_list);
283		- if (!meta_list) {
284		- ret = -ENOMEM;
285		- goto fail_free_pad;
286		- }
287		-
288		- ppa_list = (void *)(meta_list) + pblk_dma_meta_size;
289		- dma_ppa_list = dma_meta_list + pblk_dma_meta_size;
290		-
291		- bio = pblk_bio_map_addr(pblk, data, rq_ppas, rq_len,
292		- PBLK_VMALLOC_META, GFP_KERNEL);
293		- if (IS_ERR(bio)) {
294		- ret = PTR_ERR(bio);
295		- goto fail_free_meta;
296		- }
297		-
298		- bio->bi_iter.bi_sector = 0; /* internal bio */
299		- bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
300	213
301	214	rqd = pblk_alloc_rqd(pblk, PBLK_WRITE_INT);
302	215
303		- rqd->bio = bio;
	216	+ ret = pblk_alloc_rqd_meta(pblk, rqd);
	217	+ if (ret) {
	218	+ pblk_free_rqd(pblk, rqd, PBLK_WRITE_INT);
	219	+ goto fail_complete;
	220	+ }
	221	+
	222	+ rqd->bio = NULL;
304	223	rqd->opcode = NVM_OP_PWRITE;
305		- rqd->flags = pblk_set_progr_mode(pblk, PBLK_WRITE);
306		- rqd->meta_list = meta_list;
	224	+ rqd->is_seq = 1;
307	225	rqd->nr_ppas = rq_ppas;
308		- rqd->ppa_list = ppa_list;
309		- rqd->dma_ppa_list = dma_ppa_list;
310		- rqd->dma_meta_list = dma_meta_list;
311	226	rqd->end_io = pblk_end_io_recov;
312	227	rqd->private = pad_rq;
	228	+
	229	+ ppa_list = nvm_rq_to_ppa_list(rqd);
	230	+ meta_list = rqd->meta_list;
313	231
314	232	for (i = 0; i < rqd->nr_ppas; ) {
315	233	struct ppa_addr ppa;
..	..	@@ -327,24 +245,29 @@
327	245
328	246	for (j = 0; j < pblk->min_write_pgs; j++, i++, w_ptr++) {
329	247	struct ppa_addr dev_ppa;
	248	+ struct pblk_sec_meta *meta;
330	249	__le64 addr_empty = cpu_to_le64(ADDR_EMPTY);
331	250
332	251	dev_ppa = addr_to_gen_ppa(pblk, w_ptr, line->id);
333	252
334	253	pblk_map_invalidate(pblk, dev_ppa);
335		- lba_list[w_ptr] = meta_list[i].lba = addr_empty;
336		- rqd->ppa_list[i] = dev_ppa;
	254	+ lba_list[w_ptr] = addr_empty;
	255	+ meta = pblk_get_meta(pblk, meta_list, i);
	256	+ meta->lba = addr_empty;
	257	+ ppa_list[i] = dev_ppa;
337	258	}
338	259	}
339	260
340	261	kref_get(&pad_rq->ref);
341		- pblk_down_page(pblk, rqd->ppa_list, rqd->nr_ppas);
	262	+ pblk_down_chunk(pblk, ppa_list[0]);
342	263
343		- ret = pblk_submit_io(pblk, rqd);
	264	+ ret = pblk_submit_io(pblk, rqd, data);
344	265	if (ret) {
345	266	pblk_err(pblk, "I/O submission failed: %d\n", ret);
346		- pblk_up_page(pblk, rqd->ppa_list, rqd->nr_ppas);
347		- goto fail_free_bio;
	267	+ pblk_up_chunk(pblk, ppa_list[0]);
	268	+ kref_put(&pad_rq->ref, pblk_recov_complete);
	269	+ pblk_free_rqd(pblk, rqd, PBLK_WRITE_INT);
	270	+ goto fail_complete;
348	271	}
349	272
350	273	left_line_ppas -= rq_ppas;
..	..	@@ -352,13 +275,9 @@
352	275	if (left_ppas && left_line_ppas)
353	276	goto next_pad_rq;
354	277
	278	+fail_complete:
355	279	kref_put(&pad_rq->ref, pblk_recov_complete);
356		-
357		- if (!wait_for_completion_io_timeout(&pad_rq->wait,
358		- msecs_to_jiffies(PBLK_COMMAND_TIMEOUT_MS))) {
359		- pblk_err(pblk, "pad write timed out\n");
360		- ret = -ETIME;
361		- }
	280	+ wait_for_completion(&pad_rq->wait);
362	281
363	282	if (!pblk_line_is_full(line))
364	283	pblk_err(pblk, "corrupted padded line: %d\n", line->id);
..	..	@@ -367,172 +286,92 @@
367	286	free_rq:
368	287	kfree(pad_rq);
369	288	return ret;
370		-
371		-fail_free_bio:
372		- bio_put(bio);
373		-fail_free_meta:
374		- nvm_dev_dma_free(dev->parent, meta_list, dma_meta_list);
375		-fail_free_pad:
376		- kfree(pad_rq);
377		- vfree(data);
378		- return ret;
379	289	}
380	290
381		-/* When this function is called, it means that not all upper pages have been
382		- * written in a page that contains valid data. In order to recover this data, we
383		- * first find the write pointer on the device, then we pad all necessary
384		- * sectors, and finally attempt to read the valid data
385		- */
386		-static int pblk_recov_scan_all_oob(struct pblk pblk, struct pblk_line line,
387		- struct pblk_recov_alloc p)
	291	+static int pblk_pad_distance(struct pblk pblk, struct pblk_line line)
388	292	{
389	293	struct nvm_tgt_dev *dev = pblk->dev;
390	294	struct nvm_geo *geo = &dev->geo;
391		- struct ppa_addr *ppa_list;
392		- struct pblk_sec_meta *meta_list;
393		- struct nvm_rq *rqd;
394		- struct bio *bio;
395		- void *data;
396		- dma_addr_t dma_ppa_list, dma_meta_list;
397		- u64 w_ptr = 0, r_ptr;
398		- int rq_ppas, rq_len;
399		- int i, j;
400		- int ret = 0;
401		- int rec_round;
402		- int left_ppas = pblk_calc_sec_in_line(pblk, line) - line->cur_sec;
	295	+ int distance = geo->mw_cunits * geo->all_luns * geo->ws_opt;
403	296
404		- ppa_list = p.ppa_list;
405		- meta_list = p.meta_list;
406		- rqd = p.rqd;
407		- data = p.data;
408		- dma_ppa_list = p.dma_ppa_list;
409		- dma_meta_list = p.dma_meta_list;
	297	+ return (distance > line->left_msecs) ? line->left_msecs : distance;
	298	+}
410	299
411		- /* we could recover up until the line write pointer */
412		- r_ptr = line->cur_sec;
413		- rec_round = 0;
	300	+/* Return a chunk belonging to a line by stripe(write order) index */
	301	+static struct nvm_chk_meta pblk_get_stripe_chunk(struct pblk pblk,
	302	+ struct pblk_line *line,
	303	+ int index)
	304	+{
	305	+ struct nvm_tgt_dev *dev = pblk->dev;
	306	+ struct nvm_geo *geo = &dev->geo;
	307	+ struct pblk_lun *rlun;
	308	+ struct ppa_addr ppa;
	309	+ int pos;
414	310
415		-next_rq:
416		- memset(rqd, 0, pblk_g_rq_size);
	311	+ rlun = &pblk->luns[index];
	312	+ ppa = rlun->bppa;
	313	+ pos = pblk_ppa_to_pos(geo, ppa);
417	314
418		- rq_ppas = pblk_calc_secs(pblk, left_ppas, 0);
419		- if (!rq_ppas)
420		- rq_ppas = pblk->min_write_pgs;
421		- rq_len = rq_ppas * geo->csecs;
	315	+ return &line->chks[pos];
	316	+}
422	317
423		- bio = bio_map_kern(dev->q, data, rq_len, GFP_KERNEL);
424		- if (IS_ERR(bio))
425		- return PTR_ERR(bio);
	318	+static int pblk_line_wps_are_unbalanced(struct pblk *pblk,
	319	+ struct pblk_line *line)
	320	+{
	321	+ struct pblk_line_meta *lm = &pblk->lm;
	322	+ int blk_in_line = lm->blk_per_line;
	323	+ struct nvm_chk_meta *chunk;
	324	+ u64 max_wp, min_wp;
	325	+ int i;
426	326
427		- bio->bi_iter.bi_sector = 0; /* internal bio */
428		- bio_set_op_attrs(bio, REQ_OP_READ, 0);
	327	+ i = find_first_zero_bit(line->blk_bitmap, blk_in_line);
429	328
430		- rqd->bio = bio;
431		- rqd->opcode = NVM_OP_PREAD;
432		- rqd->meta_list = meta_list;
433		- rqd->nr_ppas = rq_ppas;
434		- rqd->ppa_list = ppa_list;
435		- rqd->dma_ppa_list = dma_ppa_list;
436		- rqd->dma_meta_list = dma_meta_list;
437		-
438		- if (pblk_io_aligned(pblk, rq_ppas))
439		- rqd->flags = pblk_set_read_mode(pblk, PBLK_READ_SEQUENTIAL);
440		- else
441		- rqd->flags = pblk_set_read_mode(pblk, PBLK_READ_RANDOM);
442		-
443		- for (i = 0; i < rqd->nr_ppas; ) {
444		- struct ppa_addr ppa;
445		- int pos;
446		-
447		- w_ptr = pblk_alloc_page(pblk, line, pblk->min_write_pgs);
448		- ppa = addr_to_gen_ppa(pblk, w_ptr, line->id);
449		- pos = pblk_ppa_to_pos(geo, ppa);
450		-
451		- while (test_bit(pos, line->blk_bitmap)) {
452		- w_ptr += pblk->min_write_pgs;
453		- ppa = addr_to_gen_ppa(pblk, w_ptr, line->id);
454		- pos = pblk_ppa_to_pos(geo, ppa);
455		- }
456		-
457		- for (j = 0; j < pblk->min_write_pgs; j++, i++, w_ptr++)
458		- rqd->ppa_list[i] =
459		- addr_to_gen_ppa(pblk, w_ptr, line->id);
460		- }
461		-
462		- ret = pblk_submit_io_sync(pblk, rqd);
463		- if (ret) {
464		- pblk_err(pblk, "I/O submission failed: %d\n", ret);
465		- return ret;
466		- }
467		-
468		- atomic_dec(&pblk->inflight_io);
469		-
470		- /* This should not happen since the read failed during normal recovery,
471		- * but the media works funny sometimes...
	329	+ /* If there is one or zero good chunks in the line,
	330	+ * the write pointers can't be unbalanced.
472	331	*/
473		- if (!rec_round++ && !rqd->error) {
474		- rec_round = 0;
475		- for (i = 0; i < rqd->nr_ppas; i++, r_ptr++) {
476		- u64 lba = le64_to_cpu(meta_list[i].lba);
	332	+ if (i >= (blk_in_line - 1))
	333	+ return 0;
477	334
478		- if (lba == ADDR_EMPTY \|\| lba > pblk->rl.nr_secs)
479		- continue;
	335	+ chunk = pblk_get_stripe_chunk(pblk, line, i);
	336	+ max_wp = chunk->wp;
	337	+ if (max_wp > pblk->max_write_pgs)
	338	+ min_wp = max_wp - pblk->max_write_pgs;
	339	+ else
	340	+ min_wp = 0;
480	341
481		- pblk_update_map(pblk, lba, rqd->ppa_list[i]);
482		- }
	342	+ i = find_next_zero_bit(line->blk_bitmap, blk_in_line, i + 1);
	343	+ while (i < blk_in_line) {
	344	+ chunk = pblk_get_stripe_chunk(pblk, line, i);
	345	+ if (chunk->wp > max_wp \|\| chunk->wp < min_wp)
	346	+ return 1;
	347	+
	348	+ i = find_next_zero_bit(line->blk_bitmap, blk_in_line, i + 1);
483	349	}
484	350
485		- /* Reached the end of the written line */
486		- if (rqd->error == NVM_RSP_ERR_EMPTYPAGE) {
487		- int pad_secs, nr_error_bits, bit;
488		- int ret;
489		-
490		- bit = find_first_bit((void *)&rqd->ppa_status, rqd->nr_ppas);
491		- nr_error_bits = rqd->nr_ppas - bit;
492		-
493		- /* Roll back failed sectors */
494		- line->cur_sec -= nr_error_bits;
495		- line->left_msecs += nr_error_bits;
496		- bitmap_clear(line->map_bitmap, line->cur_sec, nr_error_bits);
497		-
498		- pad_secs = pblk_pad_distance(pblk);
499		- if (pad_secs > line->left_msecs)
500		- pad_secs = line->left_msecs;
501		-
502		- ret = pblk_recov_pad_oob(pblk, line, pad_secs);
503		- if (ret)
504		- pblk_err(pblk, "OOB padding failed (err:%d)\n", ret);
505		-
506		- ret = pblk_recov_read_oob(pblk, line, p, r_ptr);
507		- if (ret)
508		- pblk_err(pblk, "OOB read failed (err:%d)\n", ret);
509		-
510		- left_ppas = 0;
511		- }
512		-
513		- left_ppas -= rq_ppas;
514		- if (left_ppas > 0)
515		- goto next_rq;
516		-
517		- return ret;
	351	+ return 0;
518	352	}
519	353
520	354	static int pblk_recov_scan_oob(struct pblk pblk, struct pblk_line line,
521		- struct pblk_recov_alloc p, int *done)
	355	+ struct pblk_recov_alloc p)
522	356	{
523	357	struct nvm_tgt_dev *dev = pblk->dev;
	358	+ struct pblk_line_meta *lm = &pblk->lm;
524	359	struct nvm_geo *geo = &dev->geo;
525	360	struct ppa_addr *ppa_list;
526		- struct pblk_sec_meta *meta_list;
	361	+ void *meta_list;
527	362	struct nvm_rq *rqd;
528		- struct bio *bio;
529	363	void *data;
530	364	dma_addr_t dma_ppa_list, dma_meta_list;
531		- u64 paddr;
532		- int rq_ppas, rq_len;
	365	+ __le64 *lba_list;
	366	+ u64 paddr = pblk_line_smeta_start(pblk, line) + lm->smeta_sec;
	367	+ bool padded = false;
	368	+ int rq_ppas;
533	369	int i, j;
534		- int ret = 0;
535		- int left_ppas = pblk_calc_sec_in_line(pblk, line);
	370	+ int ret;
	371	+ u64 left_ppas = pblk_sec_in_open_line(pblk, line) - lm->smeta_sec;
	372	+
	373	+ if (pblk_line_wps_are_unbalanced(pblk, line))
	374	+ pblk_warn(pblk, "recovering unbalanced line (%d)\n", line->id);
536	375
537	376	ppa_list = p.ppa_list;
538	377	meta_list = p.meta_list;
..	..	@@ -541,41 +380,32 @@
541	380	dma_ppa_list = p.dma_ppa_list;
542	381	dma_meta_list = p.dma_meta_list;
543	382
544		- *done = 1;
	383	+ lba_list = emeta_to_lbas(pblk, line->emeta->buf);
545	384
546	385	next_rq:
547	386	memset(rqd, 0, pblk_g_rq_size);
548	387
549		- rq_ppas = pblk_calc_secs(pblk, left_ppas, 0);
	388	+ rq_ppas = pblk_calc_secs(pblk, left_ppas, 0, false);
550	389	if (!rq_ppas)
551	390	rq_ppas = pblk->min_write_pgs;
552		- rq_len = rq_ppas * geo->csecs;
553	391
554		- bio = bio_map_kern(dev->q, data, rq_len, GFP_KERNEL);
555		- if (IS_ERR(bio))
556		- return PTR_ERR(bio);
557		-
558		- bio->bi_iter.bi_sector = 0; /* internal bio */
559		- bio_set_op_attrs(bio, REQ_OP_READ, 0);
560		-
561		- rqd->bio = bio;
	392	+retry_rq:
	393	+ rqd->bio = NULL;
562	394	rqd->opcode = NVM_OP_PREAD;
563	395	rqd->meta_list = meta_list;
564	396	rqd->nr_ppas = rq_ppas;
565	397	rqd->ppa_list = ppa_list;
566	398	rqd->dma_ppa_list = dma_ppa_list;
567	399	rqd->dma_meta_list = dma_meta_list;
	400	+ ppa_list = nvm_rq_to_ppa_list(rqd);
568	401
569	402	if (pblk_io_aligned(pblk, rq_ppas))
570		- rqd->flags = pblk_set_read_mode(pblk, PBLK_READ_SEQUENTIAL);
571		- else
572		- rqd->flags = pblk_set_read_mode(pblk, PBLK_READ_RANDOM);
	403	+ rqd->is_seq = 1;
573	404
574	405	for (i = 0; i < rqd->nr_ppas; ) {
575	406	struct ppa_addr ppa;
576	407	int pos;
577	408
578		- paddr = pblk_alloc_page(pblk, line, pblk->min_write_pgs);
579	409	ppa = addr_to_gen_ppa(pblk, paddr, line->id);
580	410	pos = pblk_ppa_to_pos(geo, ppa);
581	411
..	..	@@ -585,53 +415,62 @@
585	415	pos = pblk_ppa_to_pos(geo, ppa);
586	416	}
587	417
588		- for (j = 0; j < pblk->min_write_pgs; j++, i++, paddr++)
589		- rqd->ppa_list[i] =
590		- addr_to_gen_ppa(pblk, paddr, line->id);
	418	+ for (j = 0; j < pblk->min_write_pgs; j++, i++)
	419	+ ppa_list[i] =
	420	+ addr_to_gen_ppa(pblk, paddr + j, line->id);
591	421	}
592	422
593		- ret = pblk_submit_io_sync(pblk, rqd);
	423	+ ret = pblk_submit_io_sync(pblk, rqd, data);
594	424	if (ret) {
595	425	pblk_err(pblk, "I/O submission failed: %d\n", ret);
596		- bio_put(bio);
597	426	return ret;
598	427	}
599	428
600	429	atomic_dec(&pblk->inflight_io);
601	430
602		- /* Reached the end of the written line */
603		- if (rqd->error) {
604		- int nr_error_bits, bit;
	431	+ /* If a read fails, do a best effort by padding the line and retrying */
	432	+ if (rqd->error && rqd->error != NVM_RSP_WARN_HIGHECC) {
	433	+ int pad_distance, ret;
605	434
606		- bit = find_first_bit((void *)&rqd->ppa_status, rqd->nr_ppas);
607		- nr_error_bits = rqd->nr_ppas - bit;
	435	+ if (padded) {
	436	+ pblk_log_read_err(pblk, rqd);
	437	+ return -EINTR;
	438	+ }
608	439
609		- /* Roll back failed sectors */
610		- line->cur_sec -= nr_error_bits;
611		- line->left_msecs += nr_error_bits;
612		- bitmap_clear(line->map_bitmap, line->cur_sec, nr_error_bits);
	440	+ pad_distance = pblk_pad_distance(pblk, line);
	441	+ ret = pblk_recov_pad_line(pblk, line, pad_distance);
	442	+ if (ret) {
	443	+ return ret;
	444	+ }
613	445
614		- left_ppas = 0;
615		- rqd->nr_ppas = bit;
616		-
617		- if (rqd->error != NVM_RSP_ERR_EMPTYPAGE)
618		- *done = 0;
	446	+ padded = true;
	447	+ goto retry_rq;
619	448	}
620	449
621		- for (i = 0; i < rqd->nr_ppas; i++) {
622		- u64 lba = le64_to_cpu(meta_list[i].lba);
	450	+ pblk_get_packed_meta(pblk, rqd);
623	451
624		- if (lba == ADDR_EMPTY \|\| lba > pblk->rl.nr_secs)
	452	+ for (i = 0; i < rqd->nr_ppas; i++) {
	453	+ struct pblk_sec_meta *meta = pblk_get_meta(pblk, meta_list, i);
	454	+ u64 lba = le64_to_cpu(meta->lba);
	455	+
	456	+ lba_list[paddr++] = cpu_to_le64(lba);
	457	+
	458	+ if (lba == ADDR_EMPTY \|\| lba >= pblk->capacity)
625	459	continue;
626	460
627		- pblk_update_map(pblk, lba, rqd->ppa_list[i]);
	461	+ line->nr_valid_lbas++;
	462	+ pblk_update_map(pblk, lba, ppa_list[i]);
628	463	}
629	464
630	465	left_ppas -= rq_ppas;
631	466	if (left_ppas > 0)
632	467	goto next_rq;
633	468
634		- return ret;
	469	+#ifdef CONFIG_NVM_PBLK_DEBUG
	470	+ WARN_ON(padded && !pblk_line_is_full(line));
	471	+#endif
	472	+
	473	+ return 0;
635	474	}
636	475
637	476	/* Scan line for lbas on out of bound area */
..	..	@@ -641,18 +480,18 @@
641	480	struct nvm_geo *geo = &dev->geo;
642	481	struct nvm_rq *rqd;
643	482	struct ppa_addr *ppa_list;
644		- struct pblk_sec_meta *meta_list;
	483	+ void *meta_list;
645	484	struct pblk_recov_alloc p;
646	485	void *data;
647	486	dma_addr_t dma_ppa_list, dma_meta_list;
648		- int done, ret = 0;
	487	+ int ret = 0;
649	488
650	489	meta_list = nvm_dev_dma_alloc(dev->parent, GFP_KERNEL, &dma_meta_list);
651	490	if (!meta_list)
652	491	return -ENOMEM;
653	492
654		- ppa_list = (void *)(meta_list) + pblk_dma_meta_size;
655		- dma_ppa_list = dma_meta_list + pblk_dma_meta_size;
	493	+ ppa_list = (void *)(meta_list) + pblk_dma_meta_size(pblk);
	494	+ dma_ppa_list = dma_meta_list + pblk_dma_meta_size(pblk);
656	495
657	496	data = kcalloc(pblk->max_write_pgs, geo->csecs, GFP_KERNEL);
658	497	if (!data) {
..	..	@@ -660,7 +499,8 @@
660	499	goto free_meta_list;
661	500	}
662	501
663		- rqd = pblk_alloc_rqd(pblk, PBLK_READ);
	502	+ rqd = mempool_alloc(&pblk->r_rq_pool, GFP_KERNEL);
	503	+ memset(rqd, 0, pblk_g_rq_size);
664	504
665	505	p.ppa_list = ppa_list;
666	506	p.meta_list = meta_list;
..	..	@@ -669,24 +509,17 @@
669	509	p.dma_ppa_list = dma_ppa_list;
670	510	p.dma_meta_list = dma_meta_list;
671	511
672		- ret = pblk_recov_scan_oob(pblk, line, p, &done);
	512	+ ret = pblk_recov_scan_oob(pblk, line, p);
673	513	if (ret) {
674		- pblk_err(pblk, "could not recover L2P from OOB\n");
	514	+ pblk_err(pblk, "could not recover L2P form OOB\n");
675	515	goto out;
676		- }
677		-
678		- if (!done) {
679		- ret = pblk_recov_scan_all_oob(pblk, line, p);
680		- if (ret) {
681		- pblk_err(pblk, "could not recover L2P from OOB\n");
682		- goto out;
683		- }
684	516	}
685	517
686	518	if (pblk_line_is_full(line))
687	519	pblk_line_recov_close(pblk, line);
688	520
689	521	out:
	522	+ mempool_free(rqd, &pblk->r_rq_pool);
690	523	kfree(data);
691	524	free_meta_list:
692	525	nvm_dev_dma_free(dev->parent, meta_list, dma_meta_list);
..	..	@@ -775,7 +608,7 @@
775	608	}
776	609
777	610	static int pblk_line_was_written(struct pblk_line *line,
778		- struct pblk *pblk)
	611	+ struct pblk *pblk)
779	612	{
780	613
781	614	struct pblk_line_meta *lm = &pblk->lm;
..	..	@@ -795,10 +628,24 @@
795	628	bppa = pblk->luns[smeta_blk].bppa;
796	629	chunk = &line->chks[pblk_ppa_to_pos(geo, bppa)];
797	630
798		- if (chunk->state & NVM_CHK_ST_FREE)
799		- return 0;
	631	+ if (chunk->state & NVM_CHK_ST_CLOSED \|\|
	632	+ (chunk->state & NVM_CHK_ST_OPEN
	633	+ && chunk->wp >= lm->smeta_sec))
	634	+ return 1;
800	635
801		- return 1;
	636	+ return 0;
	637	+}
	638	+
	639	+static bool pblk_line_is_open(struct pblk pblk, struct pblk_line line)
	640	+{
	641	+ struct pblk_line_meta *lm = &pblk->lm;
	642	+ int i;
	643	+
	644	+ for (i = 0; i < lm->blk_per_line; i++)
	645	+ if (line->chks[i].state & NVM_CHK_ST_OPEN)
	646	+ return true;
	647	+
	648	+ return false;
802	649	}
803	650
804	651	struct pblk_line pblk_recov_l2p(struct pblk pblk)
..	..	@@ -841,7 +688,7 @@
841	688	continue;
842	689
843	690	/* Lines that cannot be read are assumed as not written here */
844		- if (pblk_line_read_smeta(pblk, line))
	691	+ if (pblk_line_smeta_read(pblk, line))
845	692	continue;
846	693
847	694	crc = pblk_calc_smeta_crc(pblk, smeta_buf);
..	..	@@ -859,11 +706,13 @@
859	706
860	707	/* The first valid instance uuid is used for initialization */
861	708	if (!valid_uuid) {
862		- memcpy(pblk->instance_uuid, smeta_buf->header.uuid, 16);
	709	+ guid_copy(&pblk->instance_uuid,
	710	+ (guid_t *)&smeta_buf->header.uuid);
863	711	valid_uuid = 1;
864	712	}
865	713
866		- if (memcmp(pblk->instance_uuid, smeta_buf->header.uuid, 16)) {
	714	+ if (!guid_equal(&pblk->instance_uuid,
	715	+ (guid_t *)&smeta_buf->header.uuid)) {
867	716	pblk_debug(pblk, "ignore line %u due to uuid mismatch\n",
868	717	i);
869	718	continue;
..	..	@@ -893,7 +742,7 @@
893	742	}
894	743
895	744	if (!found_lines) {
896		- pblk_setup_uuid(pblk);
	745	+ guid_gen(&pblk->instance_uuid);
897	746
898	747	spin_lock(&l_mg->free_lock);
899	748	WARN_ON_ONCE(!test_and_clear_bit(meta_line,
..	..	@@ -911,7 +760,12 @@
911	760	line->emeta = emeta;
912	761	memset(line->emeta->buf, 0, lm->emeta_len[0]);
913	762
914		- if (pblk_line_read_emeta(pblk, line, line->emeta->buf)) {
	763	+ if (pblk_line_is_open(pblk, line)) {
	764	+ pblk_recov_l2p_from_oob(pblk, line);
	765	+ goto next;
	766	+ }
	767	+
	768	+ if (pblk_line_emeta_read(pblk, line, line->emeta->buf)) {
915	769	pblk_recov_l2p_from_oob(pblk, line);
916	770	goto next;
917	771	}
..	..	@@ -935,6 +789,8 @@
935	789
936	790	spin_lock(&line->lock);
937	791	line->state = PBLK_LINESTATE_CLOSED;
	792	+ trace_pblk_line_state(pblk_disk_name(pblk), line->id,
	793	+ line->state);
938	794	move_list = pblk_line_gc_list(pblk, line);
939	795	spin_unlock(&line->lock);
940	796
..	..	@@ -942,17 +798,25 @@
942	798	list_move_tail(&line->list, move_list);
943	799	spin_unlock(&l_mg->gc_lock);
944	800
945		- kfree(line->map_bitmap);
	801	+ mempool_free(line->map_bitmap, l_mg->bitmap_pool);
946	802	line->map_bitmap = NULL;
947	803	line->smeta = NULL;
948	804	line->emeta = NULL;
949	805	} else {
950		- if (open_lines > 1)
951		- pblk_err(pblk, "failed to recover L2P\n");
	806	+ spin_lock(&line->lock);
	807	+ line->state = PBLK_LINESTATE_OPEN;
	808	+ spin_unlock(&line->lock);
	809	+
	810	+ line->emeta->mem = 0;
	811	+ atomic_set(&line->emeta->sync, 0);
	812	+
	813	+ trace_pblk_line_state(pblk_disk_name(pblk), line->id,
	814	+ line->state);
	815	+
	816	+ data_line = line;
	817	+ line->meta_line = meta_line;
952	818
953	819	open_lines++;
954		- line->meta_line = meta_line;
955		- data_line = line;
956	820	}
957	821	}
958	822
..	..	@@ -961,9 +825,9 @@
961	825	WARN_ON_ONCE(!test_and_clear_bit(meta_line,
962	826	&l_mg->meta_bitmap));
963	827	spin_unlock(&l_mg->free_lock);
964		- pblk_line_replace_data(pblk);
965	828	} else {
966	829	spin_lock(&l_mg->free_lock);
	830	+ l_mg->data_line = data_line;
967	831	/* Allocate next line for preparation */
968	832	l_mg->data_next = pblk_line_get(pblk);
969	833	if (l_mg->data_next) {
..	..	@@ -1000,7 +864,7 @@
1000	864	left_msecs = line->left_msecs;
1001	865	spin_unlock(&l_mg->free_lock);
1002	866
1003		- ret = pblk_recov_pad_oob(pblk, line, left_msecs);
	867	+ ret = pblk_recov_pad_line(pblk, line, left_msecs);
1004	868	if (ret) {
1005	869	pblk_err(pblk, "tear down padding failed (%d)\n", ret);
1006	870	return ret;