~hc/RK356X_SDK_RELEASE.git

..	..	@@ -28,6 +28,7 @@
28	28	#include <linux/uaccess.h>
29	29	#include <linux/fiemap.h>
30	30	#include <linux/backing-dev.h>
	31	+#include <linux/iomap.h>
31	32	#include "ext4_jbd2.h"
32	33	#include "ext4_extents.h"
33	34	#include "xattr.h"
..	..	@@ -83,13 +84,6 @@
83	84	et->et_checksum = ext4_extent_block_csum(inode, eh);
84	85	}
85	86
86		-static int ext4_split_extent(handle_t *handle,
87		- struct inode *inode,
88		- struct ext4_ext_path **ppath,
89		- struct ext4_map_blocks *map,
90		- int split_flag,
91		- int flags);
92		-
93	87	static int ext4_split_extent_at(handle_t *handle,
94	88	struct inode *inode,
95	89	struct ext4_ext_path **ppath,
..	..	@@ -97,32 +91,41 @@
97	91	int split_flag,
98	92	int flags);
99	93
100		-static int ext4_find_delayed_extent(struct inode *inode,
101		- struct extent_status *newes);
102		-
103		-static int ext4_ext_truncate_extend_restart(handle_t *handle,
104		- struct inode *inode,
105		- int needed)
	94	+static int ext4_ext_trunc_restart_fn(struct inode inode, int dropped)
106	95	{
107		- int err;
108		-
109		- if (!ext4_handle_valid(handle))
110		- return 0;
111		- if (handle->h_buffer_credits >= needed)
112		- return 0;
113	96	/*
114		- * If we need to extend the journal get a few extra blocks
115		- * while we're at it for efficiency's sake.
	97	+ * Drop i_data_sem to avoid deadlock with ext4_map_blocks. At this
	98	+ * moment, get_block can be called only for blocks inside i_size since
	99	+ * page cache has been already dropped and writes are blocked by
	100	+ * i_mutex. So we can safely drop the i_data_sem here.
116	101	*/
117		- needed += 3;
118		- err = ext4_journal_extend(handle, needed - handle->h_buffer_credits);
119		- if (err <= 0)
120		- return err;
121		- err = ext4_truncate_restart_trans(handle, inode, needed);
122		- if (err == 0)
123		- err = -EAGAIN;
	102	+ BUG_ON(EXT4_JOURNAL(inode) == NULL);
	103	+ ext4_discard_preallocations(inode, 0);
	104	+ up_write(&EXT4_I(inode)->i_data_sem);
	105	+ *dropped = 1;
	106	+ return 0;
	107	+}
124	108
125		- return err;
	109	+/*
	110	+ * Make sure 'handle' has at least 'check_cred' credits. If not, restart
	111	+ * transaction with 'restart_cred' credits. The function drops i_data_sem
	112	+ * when restarting transaction and gets it after transaction is restarted.
	113	+ *
	114	+ * The function returns 0 on success, 1 if transaction had to be restarted,
	115	+ * and < 0 in case of fatal error.
	116	+ */
	117	+int ext4_datasem_ensure_credits(handle_t handle, struct inode inode,
	118	+ int check_cred, int restart_cred,
	119	+ int revoke_cred)
	120	+{
	121	+ int ret;
	122	+ int dropped = 0;
	123	+
	124	+ ret = ext4_journal_ensure_credits_fn(handle, check_cred, restart_cred,
	125	+ revoke_cred, ext4_ext_trunc_restart_fn(inode, &dropped));
	126	+ if (dropped)
	127	+ down_write(&EXT4_I(inode)->i_data_sem);
	128	+ return ret;
126	129	}
127	130
128	131	/*
..	..	@@ -133,14 +136,24 @@
133	136	static int ext4_ext_get_access(handle_t handle, struct inode inode,
134	137	struct ext4_ext_path *path)
135	138	{
	139	+ int err = 0;
	140	+
136	141	if (path->p_bh) {
137	142	/* path points to block */
138	143	BUFFER_TRACE(path->p_bh, "get_write_access");
139		- return ext4_journal_get_write_access(handle, path->p_bh);
	144	+ err = ext4_journal_get_write_access(handle, path->p_bh);
	145	+ /*
	146	+ * The extent buffer's verified bit will be set again in
	147	+ * __ext4_ext_dirty(). We could leave an inconsistent
	148	+ * buffer if the extents updating procudure break off du
	149	+ * to some error happens, force to check it again.
	150	+ */
	151	+ if (!err)
	152	+ clear_buffer_verified(path->p_bh);
140	153	}
141	154	/* path points to leaf/index in inode body */
142	155	/* we use in-core data, no need to protect them */
143		- return 0;
	156	+ return err;
144	157	}
145	158
146	159	/*
..	..	@@ -149,8 +162,9 @@
149	162	* - ENOMEM
150	163	* - EIO
151	164	*/
152		-int __ext4_ext_dirty(const char where, unsigned int line, handle_t handle,
153		- struct inode inode, struct ext4_ext_path path)
	165	+static int __ext4_ext_dirty(const char *where, unsigned int line,
	166	+ handle_t handle, struct inode inode,
	167	+ struct ext4_ext_path *path)
154	168	{
155	169	int err;
156	170
..	..	@@ -160,12 +174,18 @@
160	174	/* path points to block */
161	175	err = __ext4_handle_dirty_metadata(where, line, handle,
162	176	inode, path->p_bh);
	177	+ /* Extents updating done, re-set verified flag */
	178	+ if (!err)
	179	+ set_buffer_verified(path->p_bh);
163	180	} else {
164	181	/* path points to leaf/index in inode body */
165	182	err = ext4_mark_inode_dirty(handle, inode);
166	183	}
167	184	return err;
168	185	}
	186	+
	187	+#define ext4_ext_dirty(handle, inode, path) \
	188	+ __ext4_ext_dirty(__func__, __LINE__, (handle), (inode), (path))
169	189
170	190	static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
171	191	struct ext4_ext_path *path,
..	..	@@ -290,58 +310,14 @@
290	310	{
291	311	struct ext4_ext_path path = ppath;
292	312	int unwritten = ext4_ext_is_unwritten(path[path->p_depth].p_ext);
	313	+ int flags = EXT4_EX_NOCACHE \| EXT4_GET_BLOCKS_PRE_IO;
	314	+
	315	+ if (nofail)
	316	+ flags \|= EXT4_GET_BLOCKS_METADATA_NOFAIL \| EXT4_EX_NOFAIL;
293	317
294	318	return ext4_split_extent_at(handle, inode, ppath, lblk, unwritten ?
295	319	EXT4_EXT_MARK_UNWRIT1\|EXT4_EXT_MARK_UNWRIT2 : 0,
296		- EXT4_EX_NOCACHE \| EXT4_GET_BLOCKS_PRE_IO \|
297		- (nofail ? EXT4_GET_BLOCKS_METADATA_NOFAIL:0));
298		-}
299		-
300		-/*
301		- * Calculate the number of metadata blocks needed
302		- * to allocate @blocks
303		- * Worse case is one block per extent
304		- */
305		-int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblock)
306		-{
307		- struct ext4_inode_info *ei = EXT4_I(inode);
308		- int idxs;
309		-
310		- idxs = ((inode->i_sb->s_blocksize - sizeof(struct ext4_extent_header))
311		- / sizeof(struct ext4_extent_idx));
312		-
313		- /*
314		- * If the new delayed allocation block is contiguous with the
315		- * previous da block, it can share index blocks with the
316		- * previous block, so we only need to allocate a new index
317		- * block every idxs leaf blocks. At ldxs**2 blocks, we need
318		- * an additional index block, and at ldxs**3 blocks, yet
319		- * another index blocks.
320		- */
321		- if (ei->i_da_metadata_calc_len &&
322		- ei->i_da_metadata_calc_last_lblock+1 == lblock) {
323		- int num = 0;
324		-
325		- if ((ei->i_da_metadata_calc_len % idxs) == 0)
326		- num++;
327		- if ((ei->i_da_metadata_calc_len % (idxs*idxs)) == 0)
328		- num++;
329		- if ((ei->i_da_metadata_calc_len % (idxsidxsidxs)) == 0) {
330		- num++;
331		- ei->i_da_metadata_calc_len = 0;
332		- } else
333		- ei->i_da_metadata_calc_len++;
334		- ei->i_da_metadata_calc_last_lblock++;
335		- return num;
336		- }
337		-
338		- /*
339		- * In the worst case we need a new set of index blocks at
340		- * every level of the inode's extent tree.
341		- */
342		- ei->i_da_metadata_calc_len = 1;
343		- ei->i_da_metadata_calc_last_lblock = lblock;
344		- return ext_depth(inode) + 1;
	320	+ flags);
345	321	}
346	322
347	323	static int
..	..	@@ -389,12 +365,13 @@
389	365	}
390	366
391	367	static int ext4_valid_extent_entries(struct inode *inode,
392		- struct ext4_extent_header *eh,
393		- ext4_fsblk_t *pblk, int depth)
	368	+ struct ext4_extent_header *eh,
	369	+ ext4_lblk_t lblk, ext4_fsblk_t *pblk,
	370	+ int depth)
394	371	{
395	372	unsigned short entries;
396	373	ext4_lblk_t lblock = 0;
397		- ext4_lblk_t prev = 0;
	374	+ ext4_lblk_t cur = 0;
398	375
399	376	if (eh->eh_entries == 0)
400	377	return 1;
..	..	@@ -404,38 +381,51 @@
404	381	if (depth == 0) {
405	382	/* leaf entries */
406	383	struct ext4_extent *ext = EXT_FIRST_EXTENT(eh);
407		- struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es;
408		- ext4_fsblk_t pblock = 0;
	384	+
	385	+ /*
	386	+ * The logical block in the first entry should equal to
	387	+ * the number in the index block.
	388	+ */
	389	+ if (depth != ext_depth(inode) &&
	390	+ lblk != le32_to_cpu(ext->ee_block))
	391	+ return 0;
409	392	while (entries) {
410	393	if (!ext4_valid_extent(inode, ext))
411	394	return 0;
412	395
413	396	/* Check for overlapping extents */
414	397	lblock = le32_to_cpu(ext->ee_block);
415		- if ((lblock <= prev) && prev) {
416		- pblock = ext4_ext_pblock(ext);
417		- es->s_last_error_block = cpu_to_le64(pblock);
	398	+ if (lblock < cur) {
	399	+ *pblk = ext4_ext_pblock(ext);
418	400	return 0;
419	401	}
420		- prev = lblock + ext4_ext_get_actual_len(ext) - 1;
	402	+ cur = lblock + ext4_ext_get_actual_len(ext);
421	403	ext++;
422	404	entries--;
423	405	}
424	406	} else {
425	407	struct ext4_extent_idx *ext_idx = EXT_FIRST_INDEX(eh);
	408	+
	409	+ /*
	410	+ * The logical block in the first entry should equal to
	411	+ * the number in the parent index block.
	412	+ */
	413	+ if (depth != ext_depth(inode) &&
	414	+ lblk != le32_to_cpu(ext_idx->ei_block))
	415	+ return 0;
426	416	while (entries) {
427	417	if (!ext4_valid_extent_idx(inode, ext_idx))
428	418	return 0;
429	419
430	420	/* Check for overlapping index extents */
431	421	lblock = le32_to_cpu(ext_idx->ei_block);
432		- if ((lblock <= prev) && prev) {
	422	+ if (lblock < cur) {
433	423	*pblk = ext4_idx_pblock(ext_idx);
434	424	return 0;
435	425	}
436	426	ext_idx++;
437	427	entries--;
438		- prev = lblock;
	428	+ cur = lblock + 1;
439	429	}
440	430	}
441	431	return 1;
..	..	@@ -443,7 +433,7 @@
443	433
444	434	static int __ext4_ext_check(const char *function, unsigned int line,
445	435	struct inode inode, struct ext4_extent_header eh,
446		- int depth, ext4_fsblk_t pblk)
	436	+ int depth, ext4_fsblk_t pblk, ext4_lblk_t lblk)
447	437	{
448	438	const char *error_msg;
449	439	int max = 0, err = -EFSCORRUPTED;
..	..	@@ -469,7 +459,11 @@
469	459	error_msg = "invalid eh_entries";
470	460	goto corrupted;
471	461	}
472		- if (!ext4_valid_extent_entries(inode, eh, &pblk, depth)) {
	462	+ if (unlikely((eh->eh_entries == 0) && (depth > 0))) {
	463	+ error_msg = "eh_entries is 0 but eh_depth is > 0";
	464	+ goto corrupted;
	465	+ }
	466	+ if (!ext4_valid_extent_entries(inode, eh, lblk, &pblk, depth)) {
473	467	error_msg = "invalid extent entries";
474	468	goto corrupted;
475	469	}
..	..	@@ -487,18 +481,19 @@
487	481	return 0;
488	482
489	483	corrupted:
490		- ext4_error_inode(inode, function, line, 0,
491		- "pblk %llu bad header/extent: %s - magic %x, "
492		- "entries %u, max %u(%u), depth %u(%u)",
493		- (unsigned long long) pblk, error_msg,
494		- le16_to_cpu(eh->eh_magic),
495		- le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max),
496		- max, le16_to_cpu(eh->eh_depth), depth);
	484	+ ext4_error_inode_err(inode, function, line, 0, -err,
	485	+ "pblk %llu bad header/extent: %s - magic %x, "
	486	+ "entries %u, max %u(%u), depth %u(%u)",
	487	+ (unsigned long long) pblk, error_msg,
	488	+ le16_to_cpu(eh->eh_magic),
	489	+ le16_to_cpu(eh->eh_entries),
	490	+ le16_to_cpu(eh->eh_max),
	491	+ max, le16_to_cpu(eh->eh_depth), depth);
497	492	return err;
498	493	}
499	494
500	495	#define ext4_ext_check(inode, eh, depth, pblk) \
501		- __ext4_ext_check(__func__, __LINE__, (inode), (eh), (depth), (pblk))
	496	+ __ext4_ext_check(__func__, __LINE__, (inode), (eh), (depth), (pblk), 0)
502	497
503	498	int ext4_ext_check_inode(struct inode *inode)
504	499	{
..	..	@@ -531,26 +526,32 @@
531	526
532	527	static struct buffer_head *
533	528	__read_extent_tree_block(const char *function, unsigned int line,
534		- struct inode *inode, ext4_fsblk_t pblk, int depth,
535		- int flags)
	529	+ struct inode inode, struct ext4_extent_idx idx,
	530	+ int depth, int flags)
536	531	{
537	532	struct buffer_head *bh;
538	533	int err;
	534	+ gfp_t gfp_flags = __GFP_MOVABLE \| GFP_NOFS;
	535	+ ext4_fsblk_t pblk;
539	536
540		- bh = sb_getblk_gfp(inode->i_sb, pblk, __GFP_MOVABLE \| GFP_NOFS);
	537	+ if (flags & EXT4_EX_NOFAIL)
	538	+ gfp_flags \|= __GFP_NOFAIL;
	539	+
	540	+ pblk = ext4_idx_pblock(idx);
	541	+ bh = sb_getblk_gfp(inode->i_sb, pblk, gfp_flags);
541	542	if (unlikely(!bh))
542	543	return ERR_PTR(-ENOMEM);
543	544
544	545	if (!bh_uptodate_or_lock(bh)) {
545	546	trace_ext4_ext_load_extent(inode, pblk, _RET_IP_);
546		- err = bh_submit_read(bh);
	547	+ err = ext4_read_bh(bh, 0, NULL);
547	548	if (err < 0)
548	549	goto errout;
549	550	}
550	551	if (buffer_verified(bh) && !(flags & EXT4_EX_FORCE_CACHE))
551	552	return bh;
552		- err = __ext4_ext_check(function, line, inode,
553		- ext_block_hdr(bh), depth, pblk);
	553	+ err = __ext4_ext_check(function, line, inode, ext_block_hdr(bh),
	554	+ depth, pblk, le32_to_cpu(idx->ei_block));
554	555	if (err)
555	556	goto errout;
556	557	set_buffer_verified(bh);
..	..	@@ -568,8 +569,8 @@
568	569
569	570	}
570	571
571		-#define read_extent_tree_block(inode, pblk, depth, flags) \
572		- __read_extent_tree_block(__func__, __LINE__, (inode), (pblk), \
	572	+#define read_extent_tree_block(inode, idx, depth, flags) \
	573	+ __read_extent_tree_block(__func__, __LINE__, (inode), (idx), \
573	574	(depth), (flags))
574	575
575	576	/*
..	..	@@ -589,6 +590,12 @@
589	590	down_read(&ei->i_data_sem);
590	591	depth = ext_depth(inode);
591	592
	593	+ /* Don't cache anything if there are no external extent blocks */
	594	+ if (!depth) {
	595	+ up_read(&ei->i_data_sem);
	596	+ return ret;
	597	+ }
	598	+
592	599	path = kcalloc(depth + 1, sizeof(struct ext4_ext_path),
593	600	GFP_NOFS);
594	601	if (path == NULL) {
..	..	@@ -596,9 +603,6 @@
596	603	return -ENOMEM;
597	604	}
598	605
599		- /* Don't cache anything if there are no external extent blocks */
600		- if (depth == 0)
601		- goto out;
602	606	path[0].p_hdr = ext_inode_hdr(inode);
603	607	ret = ext4_ext_check(inode, path[0].p_hdr, depth, 0);
604	608	if (ret)
..	..	@@ -616,8 +620,7 @@
616	620	i--;
617	621	continue;
618	622	}
619		- bh = read_extent_tree_block(inode,
620		- ext4_idx_pblock(path[i].p_idx++),
	623	+ bh = read_extent_tree_block(inode, path[i].p_idx++,
621	624	depth - i - 1,
622	625	EXT4_EX_FORCE_CACHE);
623	626	if (IS_ERR(bh)) {
..	..	@@ -642,21 +645,22 @@
642	645	{
643	646	int k, l = path->p_depth;
644	647
645		- ext_debug("path:");
	648	+ ext_debug(inode, "path:");
646	649	for (k = 0; k <= l; k++, path++) {
647	650	if (path->p_idx) {
648		- ext_debug(" %d->%llu", le32_to_cpu(path->p_idx->ei_block),
649		- ext4_idx_pblock(path->p_idx));
	651	+ ext_debug(inode, " %d->%llu",
	652	+ le32_to_cpu(path->p_idx->ei_block),
	653	+ ext4_idx_pblock(path->p_idx));
650	654	} else if (path->p_ext) {
651		- ext_debug(" %d:[%d]%d:%llu ",
	655	+ ext_debug(inode, " %d:[%d]%d:%llu ",
652	656	le32_to_cpu(path->p_ext->ee_block),
653	657	ext4_ext_is_unwritten(path->p_ext),
654	658	ext4_ext_get_actual_len(path->p_ext),
655	659	ext4_ext_pblock(path->p_ext));
656	660	} else
657		- ext_debug(" []");
	661	+ ext_debug(inode, " []");
658	662	}
659		- ext_debug("\n");
	663	+ ext_debug(inode, "\n");
660	664	}
661	665
662	666	static void ext4_ext_show_leaf(struct inode inode, struct ext4_ext_path path)
..	..	@@ -672,14 +676,14 @@
672	676	eh = path[depth].p_hdr;
673	677	ex = EXT_FIRST_EXTENT(eh);
674	678
675		- ext_debug("Displaying leaf extents for inode %lu\n", inode->i_ino);
	679	+ ext_debug(inode, "Displaying leaf extents\n");
676	680
677	681	for (i = 0; i < le16_to_cpu(eh->eh_entries); i++, ex++) {
678		- ext_debug("%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
	682	+ ext_debug(inode, "%d:[%d]%d:%llu ", le32_to_cpu(ex->ee_block),
679	683	ext4_ext_is_unwritten(ex),
680	684	ext4_ext_get_actual_len(ex), ext4_ext_pblock(ex));
681	685	}
682		- ext_debug("\n");
	686	+ ext_debug(inode, "\n");
683	687	}
684	688
685	689	static void ext4_ext_show_move(struct inode inode, struct ext4_ext_path path,
..	..	@@ -692,10 +696,9 @@
692	696	struct ext4_extent_idx *idx;
693	697	idx = path[level].p_idx;
694	698	while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) {
695		- ext_debug("%d: move %d:%llu in new index %llu\n", level,
696		- le32_to_cpu(idx->ei_block),
697		- ext4_idx_pblock(idx),
698		- newblock);
	699	+ ext_debug(inode, "%d: move %d:%llu in new index %llu\n",
	700	+ level, le32_to_cpu(idx->ei_block),
	701	+ ext4_idx_pblock(idx), newblock);
699	702	idx++;
700	703	}
701	704
..	..	@@ -704,7 +707,7 @@
704	707
705	708	ex = path[depth].p_ext;
706	709	while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) {
707		- ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n",
	710	+ ext_debug(inode, "move %d:%llu:[%d]%d in new leaf %llu\n",
708	711	le32_to_cpu(ex->ee_block),
709	712	ext4_ext_pblock(ex),
710	713	ext4_ext_is_unwritten(ex),
..	..	@@ -727,11 +730,10 @@
727	730	if (!path)
728	731	return;
729	732	depth = path->p_depth;
730		- for (i = 0; i <= depth; i++, path++)
731		- if (path->p_bh) {
732		- brelse(path->p_bh);
733		- path->p_bh = NULL;
734		- }
	733	+ for (i = 0; i <= depth; i++, path++) {
	734	+ brelse(path->p_bh);
	735	+ path->p_bh = NULL;
	736	+ }
735	737	}
736	738
737	739	/*
..	..	@@ -747,7 +749,7 @@
747	749	struct ext4_extent_idx r, l, *m;
748	750
749	751
750		- ext_debug("binsearch for %u(idx): ", block);
	752	+ ext_debug(inode, "binsearch for %u(idx): ", block);
751	753
752	754	l = EXT_FIRST_INDEX(eh) + 1;
753	755	r = EXT_LAST_INDEX(eh);
..	..	@@ -757,13 +759,13 @@
757	759	r = m - 1;
758	760	else
759	761	l = m + 1;
760		- ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ei_block),
761		- m, le32_to_cpu(m->ei_block),
762		- r, le32_to_cpu(r->ei_block));
	762	+ ext_debug(inode, "%p(%u):%p(%u):%p(%u) ", l,
	763	+ le32_to_cpu(l->ei_block), m, le32_to_cpu(m->ei_block),
	764	+ r, le32_to_cpu(r->ei_block));
763	765	}
764	766
765	767	path->p_idx = l - 1;
766		- ext_debug(" -> %u->%lld ", le32_to_cpu(path->p_idx->ei_block),
	768	+ ext_debug(inode, " -> %u->%lld ", le32_to_cpu(path->p_idx->ei_block),
767	769	ext4_idx_pblock(path->p_idx));
768	770
769	771	#ifdef CHECK_BINSEARCH
..	..	@@ -773,8 +775,8 @@
773	775
774	776	chix = ix = EXT_FIRST_INDEX(eh);
775	777	for (k = 0; k < le16_to_cpu(eh->eh_entries); k++, ix++) {
776		- if (k != 0 &&
777		- le32_to_cpu(ix->ei_block) <= le32_to_cpu(ix[-1].ei_block)) {
	778	+ if (k != 0 && le32_to_cpu(ix->ei_block) <=
	779	+ le32_to_cpu(ix[-1].ei_block)) {
778	780	printk(KERN_DEBUG "k=%d, ix=0x%p, "
779	781	"first=0x%p\n", k,
780	782	ix, EXT_FIRST_INDEX(eh));
..	..	@@ -814,7 +816,7 @@
814	816	return;
815	817	}
816	818
817		- ext_debug("binsearch for %u: ", block);
	819	+ ext_debug(inode, "binsearch for %u: ", block);
818	820
819	821	l = EXT_FIRST_EXTENT(eh) + 1;
820	822	r = EXT_LAST_EXTENT(eh);
..	..	@@ -825,13 +827,13 @@
825	827	r = m - 1;
826	828	else
827	829	l = m + 1;
828		- ext_debug("%p(%u):%p(%u):%p(%u) ", l, le32_to_cpu(l->ee_block),
829		- m, le32_to_cpu(m->ee_block),
830		- r, le32_to_cpu(r->ee_block));
	830	+ ext_debug(inode, "%p(%u):%p(%u):%p(%u) ", l,
	831	+ le32_to_cpu(l->ee_block), m, le32_to_cpu(m->ee_block),
	832	+ r, le32_to_cpu(r->ee_block));
831	833	}
832	834
833	835	path->p_ext = l - 1;
834		- ext_debug(" -> %d:%llu:[%d]%d ",
	836	+ ext_debug(inode, " -> %d:%llu:[%d]%d ",
835	837	le32_to_cpu(path->p_ext->ee_block),
836	838	ext4_ext_pblock(path->p_ext),
837	839	ext4_ext_is_unwritten(path->p_ext),
..	..	@@ -856,7 +858,7 @@
856	858
857	859	}
858	860
859		-int ext4_ext_tree_init(handle_t handle, struct inode inode)
	861	+void ext4_ext_tree_init(handle_t handle, struct inode inode)
860	862	{
861	863	struct ext4_extent_header *eh;
862	864
..	..	@@ -867,7 +869,6 @@
867	869	eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0));
868	870	eh->eh_generation = 0;
869	871	ext4_mark_inode_dirty(handle, inode);
870		- return 0;
871	872	}
872	873
873	874	struct ext4_ext_path *
..	..	@@ -879,6 +880,10 @@
879	880	struct ext4_ext_path path = orig_path ? orig_path : NULL;
880	881	short int depth, i, ppos = 0;
881	882	int ret;
	883	+ gfp_t gfp_flags = GFP_NOFS;
	884	+
	885	+ if (flags & EXT4_EX_NOFAIL)
	886	+ gfp_flags \|= __GFP_NOFAIL;
882	887
883	888	eh = ext_inode_hdr(inode);
884	889	depth = ext_depth(inode);
..	..	@@ -899,7 +904,7 @@
899	904	if (!path) {
900	905	/* account possible depth increase */
901	906	path = kcalloc(depth + 2, sizeof(struct ext4_ext_path),
902		- GFP_NOFS);
	907	+ gfp_flags);
903	908	if (unlikely(!path))
904	909	return ERR_PTR(-ENOMEM);
905	910	path[0].p_maxdepth = depth + 1;
..	..	@@ -912,7 +917,7 @@
912	917	ext4_cache_extents(inode, eh);
913	918	/* walk through the tree */
914	919	while (i) {
915		- ext_debug("depth %d: num %d, max %d\n",
	920	+ ext_debug(inode, "depth %d: num %d, max %d\n",
916	921	ppos, le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
917	922
918	923	ext4_ext_binsearch_idx(inode, path + ppos, block);
..	..	@@ -920,8 +925,7 @@
920	925	path[ppos].p_depth = i;
921	926	path[ppos].p_ext = NULL;
922	927
923		- bh = read_extent_tree_block(inode, path[ppos].p_block, --i,
924		- flags);
	928	+ bh = read_extent_tree_block(inode, path[ppos].p_idx, --i, flags);
925	929	if (IS_ERR(bh)) {
926	930	ret = PTR_ERR(bh);
927	931	goto err;
..	..	@@ -989,18 +993,20 @@
989	993
990	994	if (logical > le32_to_cpu(curp->p_idx->ei_block)) {
991	995	/* insert after */
992		- ext_debug("insert new index %d after: %llu\n", logical, ptr);
	996	+ ext_debug(inode, "insert new index %d after: %llu\n",
	997	+ logical, ptr);
993	998	ix = curp->p_idx + 1;
994	999	} else {
995	1000	/* insert before */
996		- ext_debug("insert new index %d before: %llu\n", logical, ptr);
	1001	+ ext_debug(inode, "insert new index %d before: %llu\n",
	1002	+ logical, ptr);
997	1003	ix = curp->p_idx;
998	1004	}
999	1005
1000	1006	len = EXT_LAST_INDEX(curp->p_hdr) - ix + 1;
1001	1007	BUG_ON(len < 0);
1002	1008	if (len > 0) {
1003		- ext_debug("insert new index %d: "
	1009	+ ext_debug(inode, "insert new index %d: "
1004	1010	"move %d indices from 0x%p to 0x%p\n",
1005	1011	logical, len, ix, ix + 1);
1006	1012	memmove(ix + 1, ix, len * sizeof(struct ext4_extent_idx));
..	..	@@ -1049,8 +1055,12 @@
1049	1055	ext4_fsblk_t newblock, oldblock;
1050	1056	__le32 border;
1051	1057	ext4_fsblk_t ablocks = NULL; / array of allocated blocks */
	1058	+ gfp_t gfp_flags = GFP_NOFS;
1052	1059	int err = 0;
1053	1060	size_t ext_size = 0;
	1061	+
	1062	+ if (flags & EXT4_EX_NOFAIL)
	1063	+ gfp_flags \|= __GFP_NOFAIL;
1054	1064
1055	1065	/* make decision: where to split? */
1056	1066	/* FIXME: now decision is simplest: at current extent */
..	..	@@ -1063,12 +1073,12 @@
1063	1073	}
1064	1074	if (path[depth].p_ext != EXT_MAX_EXTENT(path[depth].p_hdr)) {
1065	1075	border = path[depth].p_ext[1].ee_block;
1066		- ext_debug("leaf will be split."
	1076	+ ext_debug(inode, "leaf will be split."
1067	1077	" next leaf starts at %d\n",
1068	1078	le32_to_cpu(border));
1069	1079	} else {
1070	1080	border = newext->ee_block;
1071		- ext_debug("leaf will be added."
	1081	+ ext_debug(inode, "leaf will be added."
1072	1082	" next leaf starts at %d\n",
1073	1083	le32_to_cpu(border));
1074	1084	}
..	..	@@ -1085,12 +1095,12 @@
1085	1095	* We need this to handle errors and free blocks
1086	1096	* upon them.
1087	1097	*/
1088		- ablocks = kcalloc(depth, sizeof(ext4_fsblk_t), GFP_NOFS);
	1098	+ ablocks = kcalloc(depth, sizeof(ext4_fsblk_t), gfp_flags);
1089	1099	if (!ablocks)
1090	1100	return -ENOMEM;
1091	1101
1092	1102	/* allocate all needed blocks */
1093		- ext_debug("allocate %d blocks for indexes/leaf\n", depth - at);
	1103	+ ext_debug(inode, "allocate %d blocks for indexes/leaf\n", depth - at);
1094	1104	for (a = 0; a < depth - at; a++) {
1095	1105	newblock = ext4_ext_new_meta_block(handle, inode, path,
1096	1106	newext, &err, flags);
..	..	@@ -1177,7 +1187,7 @@
1177	1187	goto cleanup;
1178	1188	}
1179	1189	if (k)
1180		- ext_debug("create %d intermediate indices\n", k);
	1190	+ ext_debug(inode, "create %d intermediate indices\n", k);
1181	1191	/* insert new index into current index block */
1182	1192	/* current depth stored in i var */
1183	1193	i = depth - 1;
..	..	@@ -1205,7 +1215,7 @@
1205	1215	fidx->ei_block = border;
1206	1216	ext4_idx_store_pblock(fidx, oldblock);
1207	1217
1208		- ext_debug("int.index at %d (block %llu): %u -> %llu\n",
	1218	+ ext_debug(inode, "int.index at %d (block %llu): %u -> %llu\n",
1209	1219	i, newblock, le32_to_cpu(border), oldblock);
1210	1220
1211	1221	/* move remainder of path[i] to the new index block */
..	..	@@ -1219,7 +1229,7 @@
1219	1229	}
1220	1230	/* start copy indexes */
1221	1231	m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++;
1222		- ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx,
	1232	+ ext_debug(inode, "cur 0x%p, last 0x%p\n", path[i].p_idx,
1223	1233	EXT_MAX_INDEX(path[i].p_hdr));
1224	1234	ext4_ext_show_move(inode, path, newblock, i);
1225	1235	if (m) {
..	..	@@ -1356,13 +1366,13 @@
1356	1366	EXT_FIRST_INDEX(neh)->ei_block =
1357	1367	EXT_FIRST_EXTENT(neh)->ee_block;
1358	1368	}
1359		- ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n",
	1369	+ ext_debug(inode, "new root: num %d(%d), lblock %d, ptr %llu\n",
1360	1370	le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max),
1361	1371	le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block),
1362	1372	ext4_idx_pblock(EXT_FIRST_INDEX(neh)));
1363	1373
1364	1374	le16_add_cpu(&neh->eh_depth, 1);
1365		- ext4_mark_inode_dirty(handle, inode);
	1375	+ err = ext4_mark_inode_dirty(handle, inode);
1366	1376	out:
1367	1377	brelse(bh);
1368	1378
..	..	@@ -1505,22 +1515,21 @@
1505	1515	}
1506	1516
1507	1517	/*
1508		- * search the closest allocated block to the right for *logical
1509		- * and returns it at @logical + it's physical address at @phys
1510		- * if *logical is the largest allocated block, the function
1511		- * returns 0 at @phys
1512		- * return value contains 0 (success) or error code
	1518	+ * Search the closest allocated block to the right for *logical
	1519	+ * and returns it at @logical + it's physical address at @phys.
	1520	+ * If not exists, return 0 and @phys is set to 0. We will return
	1521	+ * 1 which means we found an allocated block and ret_ex is valid.
	1522	+ * Or return a (< 0) error code.
1513	1523	*/
1514	1524	static int ext4_ext_search_right(struct inode *inode,
1515	1525	struct ext4_ext_path *path,
1516	1526	ext4_lblk_t logical, ext4_fsblk_t phys,
1517		- struct ext4_extent **ret_ex)
	1527	+ struct ext4_extent *ret_ex)
1518	1528	{
1519	1529	struct buffer_head *bh = NULL;
1520	1530	struct ext4_extent_header *eh;
1521	1531	struct ext4_extent_idx *ix;
1522	1532	struct ext4_extent *ex;
1523		- ext4_fsblk_t block;
1524	1533	int depth; /* Note, NOT eh_depth; depth from top of tree */
1525	1534	int ee_len;
1526	1535
..	..	@@ -1587,20 +1596,17 @@
1587	1596	* follow it and find the closest allocated
1588	1597	* block to the right */
1589	1598	ix++;
1590		- block = ext4_idx_pblock(ix);
1591	1599	while (++depth < path->p_depth) {
1592	1600	/* subtract from p_depth to get proper eh_depth */
1593		- bh = read_extent_tree_block(inode, block,
1594		- path->p_depth - depth, 0);
	1601	+ bh = read_extent_tree_block(inode, ix, path->p_depth - depth, 0);
1595	1602	if (IS_ERR(bh))
1596	1603	return PTR_ERR(bh);
1597	1604	eh = ext_block_hdr(bh);
1598	1605	ix = EXT_FIRST_INDEX(eh);
1599		- block = ext4_idx_pblock(ix);
1600	1606	put_bh(bh);
1601	1607	}
1602	1608
1603		- bh = read_extent_tree_block(inode, block, path->p_depth - depth, 0);
	1609	+ bh = read_extent_tree_block(inode, ix, path->p_depth - depth, 0);
1604	1610	if (IS_ERR(bh))
1605	1611	return PTR_ERR(bh);
1606	1612	eh = ext_block_hdr(bh);
..	..	@@ -1608,10 +1614,11 @@
1608	1614	found_extent:
1609	1615	*logical = le32_to_cpu(ex->ee_block);
1610	1616	*phys = ext4_ext_pblock(ex);
1611		- *ret_ex = ex;
	1617	+ if (ret_ex)
	1618	+ ret_ex = ex;
1612	1619	if (bh)
1613	1620	put_bh(bh);
1614		- return 0;
	1621	+ return 1;
1615	1622	}
1616	1623
1617	1624	/*
..	..	@@ -1633,17 +1640,16 @@
1633	1640	return EXT_MAX_BLOCKS;
1634	1641
1635	1642	while (depth >= 0) {
	1643	+ struct ext4_ext_path *p = &path[depth];
	1644	+
1636	1645	if (depth == path->p_depth) {
1637	1646	/* leaf */
1638		- if (path[depth].p_ext &&
1639		- path[depth].p_ext !=
1640		- EXT_LAST_EXTENT(path[depth].p_hdr))
1641		- return le32_to_cpu(path[depth].p_ext[1].ee_block);
	1647	+ if (p->p_ext && p->p_ext != EXT_LAST_EXTENT(p->p_hdr))
	1648	+ return le32_to_cpu(p->p_ext[1].ee_block);
1642	1649	} else {
1643	1650	/* index */
1644		- if (path[depth].p_idx !=
1645		- EXT_LAST_INDEX(path[depth].p_hdr))
1646		- return le32_to_cpu(path[depth].p_idx[1].ei_block);
	1651	+ if (p->p_idx != EXT_LAST_INDEX(p->p_hdr))
	1652	+ return le32_to_cpu(p->p_idx[1].ei_block);
1647	1653	}
1648	1654	depth--;
1649	1655	}
..	..	@@ -1743,9 +1749,9 @@
1743	1749	return err;
1744	1750	}
1745	1751
1746		-int
1747		-ext4_can_extents_be_merged(struct inode inode, struct ext4_extent ex1,
1748		- struct ext4_extent *ex2)
	1752	+static int ext4_can_extents_be_merged(struct inode *inode,
	1753	+ struct ext4_extent *ex1,
	1754	+ struct ext4_extent *ex2)
1749	1755	{
1750	1756	unsigned short ext1_ee_len, ext2_ee_len;
1751	1757
..	..	@@ -1759,23 +1765,11 @@
1759	1765	le32_to_cpu(ex2->ee_block))
1760	1766	return 0;
1761	1767
1762		- /*
1763		- * To allow future support for preallocated extents to be added
1764		- * as an RO_COMPAT feature, refuse to merge to extents if
1765		- * this can result in the top bit of ee_len being set.
1766		- */
1767	1768	if (ext1_ee_len + ext2_ee_len > EXT_INIT_MAX_LEN)
1768	1769	return 0;
1769		- /*
1770		- * The check for IO to unwritten extent is somewhat racy as we
1771		- * increment i_unwritten / set EXT4_STATE_DIO_UNWRITTEN only after
1772		- * dropping i_data_sem. But reserved blocks should save us in that
1773		- * case.
1774		- */
	1770	+
1775	1771	if (ext4_ext_is_unwritten(ex1) &&
1776		- (ext4_test_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN) \|\|
1777		- atomic_read(&EXT4_I(inode)->i_unwritten) \|\|
1778		- (ext1_ee_len + ext2_ee_len > EXT_UNWRITTEN_MAX_LEN)))
	1772	+ ext1_ee_len + ext2_ee_len > EXT_UNWRITTEN_MAX_LEN)
1779	1773	return 0;
1780	1774	#ifdef AGGRESSIVE_TEST
1781	1775	if (ext1_ee_len >= 4)
..	..	@@ -1853,7 +1847,8 @@
1853	1847	* group descriptor to release the extent tree block. If we
1854	1848	* can't get the journal credits, give up.
1855	1849	*/
1856		- if (ext4_journal_extend(handle, 2))
	1850	+ if (ext4_journal_extend(handle, 2,
	1851	+ ext4_free_metadata_revoke_credits(inode->i_sb, 1)))
1857	1852	return;
1858	1853
1859	1854	/*
..	..	@@ -1877,13 +1872,14 @@
1877	1872	}
1878	1873
1879	1874	/*
1880		- * This function tries to merge the @ex extent to neighbours in the tree.
1881		- * return 1 if merge left else 0.
	1875	+ * This function tries to merge the @ex extent to neighbours in the tree, then
	1876	+ * tries to collapse the extent tree into the inode.
1882	1877	*/
1883	1878	static void ext4_ext_try_to_merge(handle_t *handle,
1884	1879	struct inode *inode,
1885	1880	struct ext4_ext_path *path,
1886		- struct ext4_extent *ex) {
	1881	+ struct ext4_extent *ex)
	1882	+{
1887	1883	struct ext4_extent_header *eh;
1888	1884	unsigned int depth;
1889	1885	int merge_done = 0;
..	..	@@ -1954,7 +1950,7 @@
1954	1950
1955	1951	/*
1956	1952	* ext4_ext_insert_extent:
1957		- * tries to merge requsted extent into the existing extent or
	1953	+ * tries to merge requested extent into the existing extent or
1958	1954	* inserts requested extent as new one into the tree,
1959	1955	* creating new leaf in the no-space case.
1960	1956	*/
..	..	@@ -2009,7 +2005,7 @@
2009	2005
2010	2006	/* Try to append newex to the ex */
2011	2007	if (ext4_can_extents_be_merged(inode, ex, newext)) {
2012		- ext_debug("append [%d]%d block to %u:[%d]%d"
	2008	+ ext_debug(inode, "append [%d]%d block to %u:[%d]%d"
2013	2009	"(from %llu)\n",
2014	2010	ext4_ext_is_unwritten(newext),
2015	2011	ext4_ext_get_actual_len(newext),
..	..	@@ -2034,7 +2030,7 @@
2034	2030	prepend:
2035	2031	/* Try to prepend newex to the ex */
2036	2032	if (ext4_can_extents_be_merged(inode, newext, ex)) {
2037		- ext_debug("prepend %u[%d]%d block to %u:[%d]%d"
	2033	+ ext_debug(inode, "prepend %u[%d]%d block to %u:[%d]%d"
2038	2034	"(from %llu)\n",
2039	2035	le32_to_cpu(newext->ee_block),
2040	2036	ext4_ext_is_unwritten(newext),
..	..	@@ -2072,20 +2068,20 @@
2072	2068	if (le32_to_cpu(newext->ee_block) > le32_to_cpu(fex->ee_block))
2073	2069	next = ext4_ext_next_leaf_block(path);
2074	2070	if (next != EXT_MAX_BLOCKS) {
2075		- ext_debug("next leaf block - %u\n", next);
	2071	+ ext_debug(inode, "next leaf block - %u\n", next);
2076	2072	BUG_ON(npath != NULL);
2077		- npath = ext4_find_extent(inode, next, NULL, 0);
	2073	+ npath = ext4_find_extent(inode, next, NULL, gb_flags);
2078	2074	if (IS_ERR(npath))
2079	2075	return PTR_ERR(npath);
2080	2076	BUG_ON(npath->p_depth != path->p_depth);
2081	2077	eh = npath[depth].p_hdr;
2082	2078	if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) {
2083		- ext_debug("next leaf isn't full(%d)\n",
	2079	+ ext_debug(inode, "next leaf isn't full(%d)\n",
2084	2080	le16_to_cpu(eh->eh_entries));
2085	2081	path = npath;
2086	2082	goto has_space;
2087	2083	}
2088		- ext_debug("next leaf has no free space(%d,%d)\n",
	2084	+ ext_debug(inode, "next leaf has no free space(%d,%d)\n",
2089	2085	le16_to_cpu(eh->eh_entries), le16_to_cpu(eh->eh_max));
2090	2086	}
2091	2087
..	..	@@ -2111,7 +2107,7 @@
2111	2107
2112	2108	if (!nearex) {
2113	2109	/* there is no extent in this leaf, create first one */
2114		- ext_debug("first extent in the leaf: %u:%llu:[%d]%d\n",
	2110	+ ext_debug(inode, "first extent in the leaf: %u:%llu:[%d]%d\n",
2115	2111	le32_to_cpu(newext->ee_block),
2116	2112	ext4_ext_pblock(newext),
2117	2113	ext4_ext_is_unwritten(newext),
..	..	@@ -2121,7 +2117,7 @@
2121	2117	if (le32_to_cpu(newext->ee_block)
2122	2118	> le32_to_cpu(nearex->ee_block)) {
2123	2119	/* Insert after */
2124		- ext_debug("insert %u:%llu:[%d]%d before: "
	2120	+ ext_debug(inode, "insert %u:%llu:[%d]%d before: "
2125	2121	"nearest %p\n",
2126	2122	le32_to_cpu(newext->ee_block),
2127	2123	ext4_ext_pblock(newext),
..	..	@@ -2132,7 +2128,7 @@
2132	2128	} else {
2133	2129	/* Insert before */
2134	2130	BUG_ON(newext->ee_block == nearex->ee_block);
2135		- ext_debug("insert %u:%llu:[%d]%d after: "
	2131	+ ext_debug(inode, "insert %u:%llu:[%d]%d after: "
2136	2132	"nearest %p\n",
2137	2133	le32_to_cpu(newext->ee_block),
2138	2134	ext4_ext_pblock(newext),
..	..	@@ -2142,7 +2138,7 @@
2142	2138	}
2143	2139	len = EXT_LAST_EXTENT(eh) - nearex + 1;
2144	2140	if (len > 0) {
2145		- ext_debug("insert %u:%llu:[%d]%d: "
	2141	+ ext_debug(inode, "insert %u:%llu:[%d]%d: "
2146	2142	"move %d extents from 0x%p to 0x%p\n",
2147	2143	le32_to_cpu(newext->ee_block),
2148	2144	ext4_ext_pblock(newext),
..	..	@@ -2179,154 +2175,51 @@
2179	2175	return err;
2180	2176	}
2181	2177
2182		-static int ext4_fill_fiemap_extents(struct inode *inode,
2183		- ext4_lblk_t block, ext4_lblk_t num,
2184		- struct fiemap_extent_info *fieinfo)
	2178	+static int ext4_fill_es_cache_info(struct inode *inode,
	2179	+ ext4_lblk_t block, ext4_lblk_t num,
	2180	+ struct fiemap_extent_info *fieinfo)
2185	2181	{
2186		- struct ext4_ext_path *path = NULL;
2187		- struct ext4_extent *ex;
	2182	+ ext4_lblk_t next, end = block + num - 1;
2188	2183	struct extent_status es;
2189		- ext4_lblk_t next, next_del, start = 0, end = 0;
2190		- ext4_lblk_t last = block + num;
2191		- int exists, depth = 0, err = 0;
2192		- unsigned int flags = 0;
2193	2184	unsigned char blksize_bits = inode->i_sb->s_blocksize_bits;
	2185	+ unsigned int flags;
	2186	+ int err;
2194	2187
2195		- while (block < last && block != EXT_MAX_BLOCKS) {
2196		- num = last - block;
2197		- /* find extent for this block */
2198		- down_read(&EXT4_I(inode)->i_data_sem);
2199		-
2200		- path = ext4_find_extent(inode, block, &path, 0);
2201		- if (IS_ERR(path)) {
2202		- up_read(&EXT4_I(inode)->i_data_sem);
2203		- err = PTR_ERR(path);
2204		- path = NULL;
2205		- break;
2206		- }
2207		-
2208		- depth = ext_depth(inode);
2209		- if (unlikely(path[depth].p_hdr == NULL)) {
2210		- up_read(&EXT4_I(inode)->i_data_sem);
2211		- EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth);
2212		- err = -EFSCORRUPTED;
2213		- break;
2214		- }
2215		- ex = path[depth].p_ext;
2216		- next = ext4_ext_next_allocated_block(path);
2217		-
	2188	+ while (block <= end) {
	2189	+ next = 0;
2218	2190	flags = 0;
2219		- exists = 0;
2220		- if (!ex) {
2221		- /* there is no extent yet, so try to allocate
2222		- * all requested space */
2223		- start = block;
2224		- end = block + num;
2225		- } else if (le32_to_cpu(ex->ee_block) > block) {
2226		- /* need to allocate space before found extent */
2227		- start = block;
2228		- end = le32_to_cpu(ex->ee_block);
2229		- if (block + num < end)
2230		- end = block + num;
2231		- } else if (block >= le32_to_cpu(ex->ee_block)
2232		- + ext4_ext_get_actual_len(ex)) {
2233		- /* need to allocate space after found extent */
2234		- start = block;
2235		- end = block + num;
2236		- if (end >= next)
2237		- end = next;
2238		- } else if (block >= le32_to_cpu(ex->ee_block)) {
2239		- /*
2240		- * some part of requested space is covered
2241		- * by found extent
2242		- */
2243		- start = block;
2244		- end = le32_to_cpu(ex->ee_block)
2245		- + ext4_ext_get_actual_len(ex);
2246		- if (block + num < end)
2247		- end = block + num;
2248		- exists = 1;
2249		- } else {
2250		- BUG();
2251		- }
2252		- BUG_ON(end <= start);
2253		-
2254		- if (!exists) {
2255		- es.es_lblk = start;
2256		- es.es_len = end - start;
2257		- es.es_pblk = 0;
2258		- } else {
2259		- es.es_lblk = le32_to_cpu(ex->ee_block);
2260		- es.es_len = ext4_ext_get_actual_len(ex);
2261		- es.es_pblk = ext4_ext_pblock(ex);
2262		- if (ext4_ext_is_unwritten(ex))
2263		- flags \|= FIEMAP_EXTENT_UNWRITTEN;
2264		- }
2265		-
2266		- /*
2267		- * Find delayed extent and update es accordingly. We call
2268		- * it even in !exists case to find out whether es is the
2269		- * last existing extent or not.
2270		- */
2271		- next_del = ext4_find_delayed_extent(inode, &es);
2272		- if (!exists && next_del) {
2273		- exists = 1;
	2191	+ if (!ext4_es_lookup_extent(inode, block, &next, &es))
	2192	+ break;
	2193	+ if (ext4_es_is_unwritten(&es))
	2194	+ flags \|= FIEMAP_EXTENT_UNWRITTEN;
	2195	+ if (ext4_es_is_delayed(&es))
2274	2196	flags \|= (FIEMAP_EXTENT_DELALLOC \|
2275	2197	FIEMAP_EXTENT_UNKNOWN);
2276		- }
2277		- up_read(&EXT4_I(inode)->i_data_sem);
2278		-
2279		- if (unlikely(es.es_len == 0)) {
2280		- EXT4_ERROR_INODE(inode, "es.es_len == 0");
2281		- err = -EFSCORRUPTED;
2282		- break;
2283		- }
2284		-
2285		- /*
2286		- * This is possible iff next == next_del == EXT_MAX_BLOCKS.
2287		- * we need to check next == EXT_MAX_BLOCKS because it is
2288		- * possible that an extent is with unwritten and delayed
2289		- * status due to when an extent is delayed allocated and
2290		- * is allocated by fallocate status tree will track both of
2291		- * them in a extent.
2292		- *
2293		- * So we could return a unwritten and delayed extent, and
2294		- * its block is equal to 'next'.
2295		- */
2296		- if (next == next_del && next == EXT_MAX_BLOCKS) {
	2198	+ if (ext4_es_is_hole(&es))
	2199	+ flags \|= EXT4_FIEMAP_EXTENT_HOLE;
	2200	+ if (next == 0)
2297	2201	flags \|= FIEMAP_EXTENT_LAST;
2298		- if (unlikely(next_del != EXT_MAX_BLOCKS \|\|
2299		- next != EXT_MAX_BLOCKS)) {
2300		- EXT4_ERROR_INODE(inode,
2301		- "next extent == %u, next "
2302		- "delalloc extent = %u",
2303		- next, next_del);
2304		- err = -EFSCORRUPTED;
2305		- break;
2306		- }
2307		- }
2308		-
2309		- if (exists) {
2310		- err = fiemap_fill_next_extent(fieinfo,
	2202	+ if (flags & (FIEMAP_EXTENT_DELALLOC\|
	2203	+ EXT4_FIEMAP_EXTENT_HOLE))
	2204	+ es.es_pblk = 0;
	2205	+ else
	2206	+ es.es_pblk = ext4_es_pblock(&es);
	2207	+ err = fiemap_fill_next_extent(fieinfo,
2311	2208	(__u64)es.es_lblk << blksize_bits,
2312	2209	(__u64)es.es_pblk << blksize_bits,
2313	2210	(__u64)es.es_len << blksize_bits,
2314	2211	flags);
2315		- if (err < 0)
2316		- break;
2317		- if (err == 1) {
2318		- err = 0;
2319		- break;
2320		- }
2321		- }
2322		-
2323		- block = es.es_lblk + es.es_len;
	2212	+ if (next == 0)
	2213	+ break;
	2214	+ block = next;
	2215	+ if (err < 0)
	2216	+ return err;
	2217	+ if (err == 1)
	2218	+ return 0;
2324	2219	}
2325		-
2326		- ext4_ext_drop_refs(path);
2327		- kfree(path);
2328		- return err;
	2220	+ return 0;
2329	2221	}
	2222	+
2330	2223
2331	2224	/*
2332	2225	* ext4_ext_determine_hole - determine hole around given block
..	..	@@ -2381,15 +2274,15 @@
2381	2274	{
2382	2275	struct extent_status es;
2383	2276
2384		- ext4_es_find_delayed_extent_range(inode, hole_start,
2385		- hole_start + hole_len - 1, &es);
	2277	+ ext4_es_find_extent_range(inode, &ext4_es_is_delayed, hole_start,
	2278	+ hole_start + hole_len - 1, &es);
2386	2279	if (es.es_len) {
2387	2280	/* There's delayed extent containing lblock? */
2388	2281	if (es.es_lblk <= hole_start)
2389	2282	return;
2390	2283	hole_len = min(es.es_lblk - hole_start, hole_len);
2391	2284	}
2392		- ext_debug(" -> %u:%u\n", hole_start, hole_len);
	2285	+ ext_debug(inode, " -> %u:%u\n", hole_start, hole_len);
2393	2286	ext4_es_insert_extent(inode, hole_start, hole_len, ~0,
2394	2287	EXTENT_STATUS_HOLE);
2395	2288	}
..	..	@@ -2426,7 +2319,7 @@
2426	2319	err = ext4_ext_dirty(handle, inode, path);
2427	2320	if (err)
2428	2321	return err;
2429		- ext_debug("index is empty, remove it, free block %llu\n", leaf);
	2322	+ ext_debug(inode, "index is empty, remove it, free block %llu\n", leaf);
2430	2323	trace_ext4_ext_rm_idx(inode, leaf);
2431	2324
2432	2325	ext4_free_blocks(handle, inode, NULL, leaf, 1,
..	..	@@ -2520,105 +2413,156 @@
2520	2413	return 0;
2521	2414	}
2522	2415
	2416	+/*
	2417	+ * ext4_rereserve_cluster - increment the reserved cluster count when
	2418	+ * freeing a cluster with a pending reservation
	2419	+ *
	2420	+ * @inode - file containing the cluster
	2421	+ * @lblk - logical block in cluster to be reserved
	2422	+ *
	2423	+ * Increments the reserved cluster count and adjusts quota in a bigalloc
	2424	+ * file system when freeing a partial cluster containing at least one
	2425	+ * delayed and unwritten block. A partial cluster meeting that
	2426	+ * requirement will have a pending reservation. If so, the
	2427	+ * RERESERVE_CLUSTER flag is used when calling ext4_free_blocks() to
	2428	+ * defer reserved and allocated space accounting to a subsequent call
	2429	+ * to this function.
	2430	+ */
	2431	+static void ext4_rereserve_cluster(struct inode *inode, ext4_lblk_t lblk)
	2432	+{
	2433	+ struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
	2434	+ struct ext4_inode_info *ei = EXT4_I(inode);
	2435	+
	2436	+ dquot_reclaim_block(inode, EXT4_C2B(sbi, 1));
	2437	+
	2438	+ spin_lock(&ei->i_block_reservation_lock);
	2439	+ ei->i_reserved_data_blocks++;
	2440	+ percpu_counter_add(&sbi->s_dirtyclusters_counter, 1);
	2441	+ spin_unlock(&ei->i_block_reservation_lock);
	2442	+
	2443	+ percpu_counter_add(&sbi->s_freeclusters_counter, 1);
	2444	+ ext4_remove_pending(inode, lblk);
	2445	+}
	2446	+
2523	2447	static int ext4_remove_blocks(handle_t handle, struct inode inode,
2524	2448	struct ext4_extent *ex,
2525		- long long *partial_cluster,
	2449	+ struct partial_cluster *partial,
2526	2450	ext4_lblk_t from, ext4_lblk_t to)
2527	2451	{
2528	2452	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2529	2453	unsigned short ee_len = ext4_ext_get_actual_len(ex);
2530		- ext4_fsblk_t pblk;
2531		- int flags = get_default_free_blocks_flags(inode);
	2454	+ ext4_fsblk_t last_pblk, pblk;
	2455	+ ext4_lblk_t num;
	2456	+ int flags;
	2457	+
	2458	+ /* only extent tail removal is allowed */
	2459	+ if (from < le32_to_cpu(ex->ee_block) \|\|
	2460	+ to != le32_to_cpu(ex->ee_block) + ee_len - 1) {
	2461	+ ext4_error(sbi->s_sb,
	2462	+ "strange request: removal(2) %u-%u from %u:%u",
	2463	+ from, to, le32_to_cpu(ex->ee_block), ee_len);
	2464	+ return 0;
	2465	+ }
	2466	+
	2467	+#ifdef EXTENTS_STATS
	2468	+ spin_lock(&sbi->s_ext_stats_lock);
	2469	+ sbi->s_ext_blocks += ee_len;
	2470	+ sbi->s_ext_extents++;
	2471	+ if (ee_len < sbi->s_ext_min)
	2472	+ sbi->s_ext_min = ee_len;
	2473	+ if (ee_len > sbi->s_ext_max)
	2474	+ sbi->s_ext_max = ee_len;
	2475	+ if (ext_depth(inode) > sbi->s_depth_max)
	2476	+ sbi->s_depth_max = ext_depth(inode);
	2477	+ spin_unlock(&sbi->s_ext_stats_lock);
	2478	+#endif
	2479	+
	2480	+ trace_ext4_remove_blocks(inode, ex, from, to, partial);
	2481	+
	2482	+ /*
	2483	+ * if we have a partial cluster, and it's different from the
	2484	+ * cluster of the last block in the extent, we free it
	2485	+ */
	2486	+ last_pblk = ext4_ext_pblock(ex) + ee_len - 1;
	2487	+
	2488	+ if (partial->state != initial &&
	2489	+ partial->pclu != EXT4_B2C(sbi, last_pblk)) {
	2490	+ if (partial->state == tofree) {
	2491	+ flags = get_default_free_blocks_flags(inode);
	2492	+ if (ext4_is_pending(inode, partial->lblk))
	2493	+ flags \|= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER;
	2494	+ ext4_free_blocks(handle, inode, NULL,
	2495	+ EXT4_C2B(sbi, partial->pclu),
	2496	+ sbi->s_cluster_ratio, flags);
	2497	+ if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)
	2498	+ ext4_rereserve_cluster(inode, partial->lblk);
	2499	+ }
	2500	+ partial->state = initial;
	2501	+ }
	2502	+
	2503	+ num = le32_to_cpu(ex->ee_block) + ee_len - from;
	2504	+ pblk = ext4_ext_pblock(ex) + ee_len - num;
	2505	+
	2506	+ /*
	2507	+ * We free the partial cluster at the end of the extent (if any),
	2508	+ * unless the cluster is used by another extent (partial_cluster
	2509	+ * state is nofree). If a partial cluster exists here, it must be
	2510	+ * shared with the last block in the extent.
	2511	+ */
	2512	+ flags = get_default_free_blocks_flags(inode);
	2513	+
	2514	+ /* partial, left end cluster aligned, right end unaligned */
	2515	+ if ((EXT4_LBLK_COFF(sbi, to) != sbi->s_cluster_ratio - 1) &&
	2516	+ (EXT4_LBLK_CMASK(sbi, to) >= from) &&
	2517	+ (partial->state != nofree)) {
	2518	+ if (ext4_is_pending(inode, to))
	2519	+ flags \|= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER;
	2520	+ ext4_free_blocks(handle, inode, NULL,
	2521	+ EXT4_PBLK_CMASK(sbi, last_pblk),
	2522	+ sbi->s_cluster_ratio, flags);
	2523	+ if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)
	2524	+ ext4_rereserve_cluster(inode, to);
	2525	+ partial->state = initial;
	2526	+ flags = get_default_free_blocks_flags(inode);
	2527	+ }
	2528	+
	2529	+ flags \|= EXT4_FREE_BLOCKS_NOFREE_LAST_CLUSTER;
2532	2530
2533	2531	/*
2534	2532	* For bigalloc file systems, we never free a partial cluster
2535		- * at the beginning of the extent. Instead, we make a note
2536		- * that we tried freeing the cluster, and check to see if we
	2533	+ * at the beginning of the extent. Instead, we check to see if we
2537	2534	* need to free it on a subsequent call to ext4_remove_blocks,
2538	2535	* or at the end of ext4_ext_rm_leaf or ext4_ext_remove_space.
2539	2536	*/
2540	2537	flags \|= EXT4_FREE_BLOCKS_NOFREE_FIRST_CLUSTER;
	2538	+ ext4_free_blocks(handle, inode, NULL, pblk, num, flags);
2541	2539
2542		- trace_ext4_remove_blocks(inode, ex, from, to, *partial_cluster);
	2540	+ /* reset the partial cluster if we've freed past it */
	2541	+ if (partial->state != initial && partial->pclu != EXT4_B2C(sbi, pblk))
	2542	+ partial->state = initial;
	2543	+
2543	2544	/*
2544		- * If we have a partial cluster, and it's different from the
2545		- * cluster of the last block, we need to explicitly free the
2546		- * partial cluster here.
	2545	+ * If we've freed the entire extent but the beginning is not left
	2546	+ * cluster aligned and is not marked as ineligible for freeing we
	2547	+ * record the partial cluster at the beginning of the extent. It
	2548	+ * wasn't freed by the preceding ext4_free_blocks() call, and we
	2549	+ * need to look farther to the left to determine if it's to be freed
	2550	+ * (not shared with another extent). Else, reset the partial
	2551	+ * cluster - we're either done freeing or the beginning of the
	2552	+ * extent is left cluster aligned.
2547	2553	*/
2548		- pblk = ext4_ext_pblock(ex) + ee_len - 1;
2549		- if (*partial_cluster > 0 &&
2550		- *partial_cluster != (long long) EXT4_B2C(sbi, pblk)) {
2551		- ext4_free_blocks(handle, inode, NULL,
2552		- EXT4_C2B(sbi, *partial_cluster),
2553		- sbi->s_cluster_ratio, flags);
2554		- *partial_cluster = 0;
2555		- }
2556		-
2557		-#ifdef EXTENTS_STATS
2558		- {
2559		- struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2560		- spin_lock(&sbi->s_ext_stats_lock);
2561		- sbi->s_ext_blocks += ee_len;
2562		- sbi->s_ext_extents++;
2563		- if (ee_len < sbi->s_ext_min)
2564		- sbi->s_ext_min = ee_len;
2565		- if (ee_len > sbi->s_ext_max)
2566		- sbi->s_ext_max = ee_len;
2567		- if (ext_depth(inode) > sbi->s_depth_max)
2568		- sbi->s_depth_max = ext_depth(inode);
2569		- spin_unlock(&sbi->s_ext_stats_lock);
2570		- }
2571		-#endif
2572		- if (from >= le32_to_cpu(ex->ee_block)
2573		- && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
2574		- /* tail removal */
2575		- ext4_lblk_t num;
2576		- long long first_cluster;
2577		-
2578		- num = le32_to_cpu(ex->ee_block) + ee_len - from;
2579		- pblk = ext4_ext_pblock(ex) + ee_len - num;
2580		- /*
2581		- * Usually we want to free partial cluster at the end of the
2582		- * extent, except for the situation when the cluster is still
2583		- * used by any other extent (partial_cluster is negative).
2584		- */
2585		- if (*partial_cluster < 0 &&
2586		- *partial_cluster == -(long long) EXT4_B2C(sbi, pblk+num-1))
2587		- flags \|= EXT4_FREE_BLOCKS_NOFREE_LAST_CLUSTER;
2588		-
2589		- ext_debug("free last %u blocks starting %llu partial %lld\n",
2590		- num, pblk, *partial_cluster);
2591		- ext4_free_blocks(handle, inode, NULL, pblk, num, flags);
2592		- /*
2593		- * If the block range to be freed didn't start at the
2594		- * beginning of a cluster, and we removed the entire
2595		- * extent and the cluster is not used by any other extent,
2596		- * save the partial cluster here, since we might need to
2597		- * delete if we determine that the truncate or punch hole
2598		- * operation has removed all of the blocks in the cluster.
2599		- * If that cluster is used by another extent, preserve its
2600		- * negative value so it isn't freed later on.
2601		- *
2602		- * If the whole extent wasn't freed, we've reached the
2603		- * start of the truncated/punched region and have finished
2604		- * removing blocks. If there's a partial cluster here it's
2605		- * shared with the remainder of the extent and is no longer
2606		- * a candidate for removal.
2607		- */
2608		- if (EXT4_PBLK_COFF(sbi, pblk) && ee_len == num) {
2609		- first_cluster = (long long) EXT4_B2C(sbi, pblk);
2610		- if (first_cluster != -*partial_cluster)
2611		- *partial_cluster = first_cluster;
2612		- } else {
2613		- *partial_cluster = 0;
	2554	+ if (EXT4_LBLK_COFF(sbi, from) && num == ee_len) {
	2555	+ if (partial->state == initial) {
	2556	+ partial->pclu = EXT4_B2C(sbi, pblk);
	2557	+ partial->lblk = from;
	2558	+ partial->state = tofree;
2614	2559	}
2615		- } else
2616		- ext4_error(sbi->s_sb, "strange request: removal(2) "
2617		- "%u-%u from %u:%u",
2618		- from, to, le32_to_cpu(ex->ee_block), ee_len);
	2560	+ } else {
	2561	+ partial->state = initial;
	2562	+ }
	2563	+
2619	2564	return 0;
2620	2565	}
2621		-
2622	2566
2623	2567	/*
2624	2568	* ext4_ext_rm_leaf() Removes the extents associated with the
..	..	@@ -2638,12 +2582,12 @@
2638	2582	static int
2639	2583	ext4_ext_rm_leaf(handle_t handle, struct inode inode,
2640	2584	struct ext4_ext_path *path,
2641		- long long *partial_cluster,
	2585	+ struct partial_cluster *partial,
2642	2586	ext4_lblk_t start, ext4_lblk_t end)
2643	2587	{
2644	2588	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2645	2589	int err = 0, correct_index = 0;
2646		- int depth = ext_depth(inode), credits;
	2590	+ int depth = ext_depth(inode), credits, revoke_credits;
2647	2591	struct ext4_extent_header *eh;
2648	2592	ext4_lblk_t a, b;
2649	2593	unsigned num;
..	..	@@ -2654,7 +2598,7 @@
2654	2598	ext4_fsblk_t pblk;
2655	2599
2656	2600	/* the header must be checked already in ext4_ext_remove_space() */
2657		- ext_debug("truncate since %u in leaf to %u\n", start, end);
	2601	+ ext_debug(inode, "truncate since %u in leaf to %u\n", start, end);
2658	2602	if (!path[depth].p_hdr)
2659	2603	path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
2660	2604	eh = path[depth].p_hdr;
..	..	@@ -2670,7 +2614,7 @@
2670	2614	ex_ee_block = le32_to_cpu(ex->ee_block);
2671	2615	ex_ee_len = ext4_ext_get_actual_len(ex);
2672	2616
2673		- trace_ext4_ext_rm_leaf(inode, start, ex, *partial_cluster);
	2617	+ trace_ext4_ext_rm_leaf(inode, start, ex, partial);
2674	2618
2675	2619	while (ex >= EXT_FIRST_EXTENT(eh) &&
2676	2620	ex_ee_block + ex_ee_len > start) {
..	..	@@ -2680,7 +2624,7 @@
2680	2624	else
2681	2625	unwritten = 0;
2682	2626
2683		- ext_debug("remove ext %u:[%d]%d\n", ex_ee_block,
	2627	+ ext_debug(inode, "remove ext %u:[%d]%d\n", ex_ee_block,
2684	2628	unwritten, ex_ee_len);
2685	2629	path[depth].p_ext = ex;
2686	2630
..	..	@@ -2688,7 +2632,7 @@
2688	2632	b = ex_ee_block+ex_ee_len - 1 < end ?
2689	2633	ex_ee_block+ex_ee_len - 1 : end;
2690	2634
2691		- ext_debug(" border %u:%u\n", a, b);
	2635	+ ext_debug(inode, " border %u:%u\n", a, b);
2692	2636
2693	2637	/* If this extent is beyond the end of the hole, skip it */
2694	2638	if (end < ex_ee_block) {
..	..	@@ -2701,8 +2645,8 @@
2701	2645	*/
2702	2646	if (sbi->s_cluster_ratio > 1) {
2703	2647	pblk = ext4_ext_pblock(ex);
2704		- *partial_cluster =
2705		- -(long long) EXT4_B2C(sbi, pblk);
	2648	+ partial->pclu = EXT4_B2C(sbi, pblk);
	2649	+ partial->state = nofree;
2706	2650	}
2707	2651	ex--;
2708	2652	ex_ee_block = le32_to_cpu(ex->ee_block);
..	..	@@ -2735,17 +2679,29 @@
2735	2679	credits += (ext_depth(inode)) + 1;
2736	2680	}
2737	2681	credits += EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb);
	2682	+ /*
	2683	+ * We may end up freeing some index blocks and data from the
	2684	+ * punched range. Note that partial clusters are accounted for
	2685	+ * by ext4_free_data_revoke_credits().
	2686	+ */
	2687	+ revoke_credits =
	2688	+ ext4_free_metadata_revoke_credits(inode->i_sb,
	2689	+ ext_depth(inode)) +
	2690	+ ext4_free_data_revoke_credits(inode, b - a + 1);
2738	2691
2739		- err = ext4_ext_truncate_extend_restart(handle, inode, credits);
2740		- if (err)
	2692	+ err = ext4_datasem_ensure_credits(handle, inode, credits,
	2693	+ credits, revoke_credits);
	2694	+ if (err) {
	2695	+ if (err > 0)
	2696	+ err = -EAGAIN;
2741	2697	goto out;
	2698	+ }
2742	2699
2743	2700	err = ext4_ext_get_access(handle, inode, path + depth);
2744	2701	if (err)
2745	2702	goto out;
2746	2703
2747		- err = ext4_remove_blocks(handle, inode, ex, partial_cluster,
2748		- a, b);
	2704	+ err = ext4_remove_blocks(handle, inode, ex, partial, a, b);
2749	2705	if (err)
2750	2706	goto out;
2751	2707
..	..	@@ -2785,7 +2741,7 @@
2785	2741	if (err)
2786	2742	goto out;
2787	2743
2788		- ext_debug("new extent: %u:%u:%llu\n", ex_ee_block, num,
	2744	+ ext_debug(inode, "new extent: %u:%u:%llu\n", ex_ee_block, num,
2789	2745	ext4_ext_pblock(ex));
2790	2746	ex--;
2791	2747	ex_ee_block = le32_to_cpu(ex->ee_block);
..	..	@@ -2799,18 +2755,23 @@
2799	2755	* If there's a partial cluster and at least one extent remains in
2800	2756	* the leaf, free the partial cluster if it isn't shared with the
2801	2757	* current extent. If it is shared with the current extent
2802		- * we zero partial_cluster because we've reached the start of the
	2758	+ * we reset the partial cluster because we've reached the start of the
2803	2759	* truncated/punched region and we're done removing blocks.
2804	2760	*/
2805		- if (*partial_cluster > 0 && ex >= EXT_FIRST_EXTENT(eh)) {
	2761	+ if (partial->state == tofree && ex >= EXT_FIRST_EXTENT(eh)) {
2806	2762	pblk = ext4_ext_pblock(ex) + ex_ee_len - 1;
2807		- if (*partial_cluster != (long long) EXT4_B2C(sbi, pblk)) {
	2763	+ if (partial->pclu != EXT4_B2C(sbi, pblk)) {
	2764	+ int flags = get_default_free_blocks_flags(inode);
	2765	+
	2766	+ if (ext4_is_pending(inode, partial->lblk))
	2767	+ flags \|= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER;
2808	2768	ext4_free_blocks(handle, inode, NULL,
2809		- EXT4_C2B(sbi, *partial_cluster),
2810		- sbi->s_cluster_ratio,
2811		- get_default_free_blocks_flags(inode));
	2769	+ EXT4_C2B(sbi, partial->pclu),
	2770	+ sbi->s_cluster_ratio, flags);
	2771	+ if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)
	2772	+ ext4_rereserve_cluster(inode, partial->lblk);
2812	2773	}
2813		- *partial_cluster = 0;
	2774	+ partial->state = initial;
2814	2775	}
2815	2776
2816	2777	/* if this leaf is free, then we should
..	..	@@ -2849,14 +2810,20 @@
2849	2810	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
2850	2811	int depth = ext_depth(inode);
2851	2812	struct ext4_ext_path *path = NULL;
2852		- long long partial_cluster = 0;
	2813	+ struct partial_cluster partial;
2853	2814	handle_t *handle;
2854	2815	int i = 0, err = 0;
2855	2816
2856		- ext_debug("truncate since %u to %u\n", start, end);
	2817	+ partial.pclu = 0;
	2818	+ partial.lblk = 0;
	2819	+ partial.state = initial;
	2820	+
	2821	+ ext_debug(inode, "truncate since %u to %u\n", start, end);
2857	2822
2858	2823	/* probably first extent we're gonna free will be last in block */
2859		- handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, depth + 1);
	2824	+ handle = ext4_journal_start_with_revoke(inode, EXT4_HT_TRUNCATE,
	2825	+ depth + 1,
	2826	+ ext4_free_metadata_revoke_credits(inode->i_sb, depth));
2860	2827	if (IS_ERR(handle))
2861	2828	return PTR_ERR(handle);
2862	2829
..	..	@@ -2876,7 +2843,8 @@
2876	2843	ext4_fsblk_t pblk;
2877	2844
2878	2845	/* find extent for or closest extent to this block */
2879		- path = ext4_find_extent(inode, end, NULL, EXT4_EX_NOCACHE);
	2846	+ path = ext4_find_extent(inode, end, NULL,
	2847	+ EXT4_EX_NOCACHE \| EXT4_EX_NOFAIL);
2880	2848	if (IS_ERR(path)) {
2881	2849	ext4_journal_stop(handle);
2882	2850	return PTR_ERR(path);
..	..	@@ -2912,8 +2880,8 @@
2912	2880	*/
2913	2881	if (sbi->s_cluster_ratio > 1) {
2914	2882	pblk = ext4_ext_pblock(ex) + end - ee_block + 1;
2915		- partial_cluster =
2916		- -(long long) EXT4_B2C(sbi, pblk);
	2883	+ partial.pclu = EXT4_B2C(sbi, pblk);
	2884	+ partial.state = nofree;
2917	2885	}
2918	2886
2919	2887	/*
..	..	@@ -2927,23 +2895,27 @@
2927	2895	if (err < 0)
2928	2896	goto out;
2929	2897
2930		- } else if (sbi->s_cluster_ratio > 1 && end >= ex_end) {
	2898	+ } else if (sbi->s_cluster_ratio > 1 && end >= ex_end &&
	2899	+ partial.state == initial) {
2931	2900	/*
2932		- * If there's an extent to the right its first cluster
2933		- * contains the immediate right boundary of the
2934		- * truncated/punched region. Set partial_cluster to
2935		- * its negative value so it won't be freed if shared
2936		- * with the current extent. The end < ee_block case
2937		- * is handled in ext4_ext_rm_leaf().
	2901	+ * If we're punching, there's an extent to the right.
	2902	+ * If the partial cluster hasn't been set, set it to
	2903	+ * that extent's first cluster and its state to nofree
	2904	+ * so it won't be freed should it contain blocks to be
	2905	+ * removed. If it's already set (tofree/nofree), we're
	2906	+ * retrying and keep the original partial cluster info
	2907	+ * so a cluster marked tofree as a result of earlier
	2908	+ * extent removal is not lost.
2938	2909	*/
2939	2910	lblk = ex_end + 1;
2940	2911	err = ext4_ext_search_right(inode, path, &lblk, &pblk,
2941		- &ex);
2942		- if (err)
	2912	+ NULL);
	2913	+ if (err < 0)
2943	2914	goto out;
2944		- if (pblk)
2945		- partial_cluster =
2946		- -(long long) EXT4_B2C(sbi, pblk);
	2915	+ if (pblk) {
	2916	+ partial.pclu = EXT4_B2C(sbi, pblk);
	2917	+ partial.state = nofree;
	2918	+ }
2947	2919	}
2948	2920	}
2949	2921	/*
..	..	@@ -2958,7 +2930,7 @@
2958	2930	le16_to_cpu(path[k].p_hdr->eh_entries)+1;
2959	2931	} else {
2960	2932	path = kcalloc(depth + 1, sizeof(struct ext4_ext_path),
2961		- GFP_NOFS);
	2933	+ GFP_NOFS \| __GFP_NOFAIL);
2962	2934	if (path == NULL) {
2963	2935	ext4_journal_stop(handle);
2964	2936	return -ENOMEM;
..	..	@@ -2978,8 +2950,7 @@
2978	2950	if (i == depth) {
2979	2951	/* this is leaf block */
2980	2952	err = ext4_ext_rm_leaf(handle, inode, path,
2981		- &partial_cluster, start,
2982		- end);
	2953	+ &partial, start, end);
2983	2954	/* root level has p_bh == NULL, brelse() eats this */
2984	2955	brelse(path[i].p_bh);
2985	2956	path[i].p_bh = NULL;
..	..	@@ -2989,7 +2960,7 @@
2989	2960
2990	2961	/* this is index block */
2991	2962	if (!path[i].p_hdr) {
2992		- ext_debug("initialize header\n");
	2963	+ ext_debug(inode, "initialize header\n");
2993	2964	path[i].p_hdr = ext_block_hdr(path[i].p_bh);
2994	2965	}
2995	2966
..	..	@@ -2997,7 +2968,7 @@
2997	2968	/* this level hasn't been touched yet */
2998	2969	path[i].p_idx = EXT_LAST_INDEX(path[i].p_hdr);
2999	2970	path[i].p_block = le16_to_cpu(path[i].p_hdr->eh_entries)+1;
3000		- ext_debug("init index ptr: hdr 0x%p, num %d\n",
	2971	+ ext_debug(inode, "init index ptr: hdr 0x%p, num %d\n",
3001	2972	path[i].p_hdr,
3002	2973	le16_to_cpu(path[i].p_hdr->eh_entries));
3003	2974	} else {
..	..	@@ -3005,18 +2976,18 @@
3005	2976	path[i].p_idx--;
3006	2977	}
3007	2978
3008		- ext_debug("level %d - index, first 0x%p, cur 0x%p\n",
	2979	+ ext_debug(inode, "level %d - index, first 0x%p, cur 0x%p\n",
3009	2980	i, EXT_FIRST_INDEX(path[i].p_hdr),
3010	2981	path[i].p_idx);
3011	2982	if (ext4_ext_more_to_rm(path + i)) {
3012	2983	struct buffer_head *bh;
3013	2984	/* go to the next level */
3014		- ext_debug("move to level %d (block %llu)\n",
	2985	+ ext_debug(inode, "move to level %d (block %llu)\n",
3015	2986	i + 1, ext4_idx_pblock(path[i].p_idx));
3016	2987	memset(path + i + 1, 0, sizeof(*path));
3017		- bh = read_extent_tree_block(inode,
3018		- ext4_idx_pblock(path[i].p_idx), depth - i - 1,
3019		- EXT4_EX_NOCACHE);
	2988	+ bh = read_extent_tree_block(inode, path[i].p_idx,
	2989	+ depth - i - 1,
	2990	+ EXT4_EX_NOCACHE);
3020	2991	if (IS_ERR(bh)) {
3021	2992	/* should we reset i_size? */
3022	2993	err = PTR_ERR(bh);
..	..	@@ -3047,25 +3018,28 @@
3047	3018	brelse(path[i].p_bh);
3048	3019	path[i].p_bh = NULL;
3049	3020	i--;
3050		- ext_debug("return to level %d\n", i);
	3021	+ ext_debug(inode, "return to level %d\n", i);
3051	3022	}
3052	3023	}
3053	3024
3054		- trace_ext4_ext_remove_space_done(inode, start, end, depth,
3055		- partial_cluster, path->p_hdr->eh_entries);
	3025	+ trace_ext4_ext_remove_space_done(inode, start, end, depth, &partial,
	3026	+ path->p_hdr->eh_entries);
3056	3027
3057	3028	/*
3058		- * If we still have something in the partial cluster and we have removed
3059		- * even the first extent, then we should free the blocks in the partial
3060		- * cluster as well. (This code will only run when there are no leaves
3061		- * to the immediate left of the truncated/punched region.)
	3029	+ * if there's a partial cluster and we have removed the first extent
	3030	+ * in the file, then we also free the partial cluster, if any
3062	3031	*/
3063		- if (partial_cluster > 0 && err == 0) {
3064		- /* don't zero partial_cluster since it's not used afterwards */
	3032	+ if (partial.state == tofree && err == 0) {
	3033	+ int flags = get_default_free_blocks_flags(inode);
	3034	+
	3035	+ if (ext4_is_pending(inode, partial.lblk))
	3036	+ flags \|= EXT4_FREE_BLOCKS_RERESERVE_CLUSTER;
3065	3037	ext4_free_blocks(handle, inode, NULL,
3066		- EXT4_C2B(sbi, partial_cluster),
3067		- sbi->s_cluster_ratio,
3068		- get_default_free_blocks_flags(inode));
	3038	+ EXT4_C2B(sbi, partial.pclu),
	3039	+ sbi->s_cluster_ratio, flags);
	3040	+ if (flags & EXT4_FREE_BLOCKS_RERESERVE_CLUSTER)
	3041	+ ext4_rereserve_cluster(inode, partial.lblk);
	3042	+ partial.state = initial;
3069	3043	}
3070	3044
3071	3045	/* TODO: flexible tree reduction should be here */
..	..	@@ -3186,7 +3160,7 @@
3186	3160	*
3187	3161	*
3188	3162	* Splits extent [a, b] into two extents [a, @split) and [@split, b], states
3189		- * of which are deterimined by split_flag.
	3163	+ * of which are determined by split_flag.
3190	3164	*
3191	3165	* There are two cases:
3192	3166	* a> the extent are splitted into two extent.
..	..	@@ -3212,8 +3186,7 @@
3212	3186	BUG_ON((split_flag & (EXT4_EXT_DATA_VALID1 \| EXT4_EXT_DATA_VALID2)) ==
3213	3187	(EXT4_EXT_DATA_VALID1 \| EXT4_EXT_DATA_VALID2));
3214	3188
3215		- ext_debug("ext4_split_extents_at: inode %lu, logical"
3216		- "block %llu\n", inode->i_ino, (unsigned long long)split);
	3189	+ ext_debug(inode, "logical block %llu\n", (unsigned long long)split);
3217	3190
3218	3191	ext4_ext_show_leaf(inode, path);
3219	3192
..	..	@@ -3321,6 +3294,10 @@
3321	3294
3322	3295	fix_extent_len:
3323	3296	ex->ee_len = orig_ex.ee_len;
	3297	+ /*
	3298	+ * Ignore ext4_ext_dirty return value since we are already in error path
	3299	+ * and err is a non-zero error code.
	3300	+ */
3324	3301	ext4_ext_dirty(handle, inode, path + path->p_depth);
3325	3302	return err;
3326	3303	out:
..	..	@@ -3380,7 +3357,7 @@
3380	3357	* Update path is required because previous ext4_split_extent_at() may
3381	3358	* result in split of original leaf or extent zeroout.
3382	3359	*/
3383		- path = ext4_find_extent(inode, map->m_lblk, ppath, 0);
	3360	+ path = ext4_find_extent(inode, map->m_lblk, ppath, flags);
3384	3361	if (IS_ERR(path))
3385	3362	return PTR_ERR(path);
3386	3363	depth = ext_depth(inode);
..	..	@@ -3449,9 +3426,8 @@
3449	3426	int err = 0;
3450	3427	int split_flag = EXT4_EXT_DATA_VALID2;
3451	3428
3452		- ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical"
3453		- "block %llu, max_blocks %u\n", inode->i_ino,
3454		- (unsigned long long)map->m_lblk, map_len);
	3429	+ ext_debug(inode, "logical block %llu, max_blocks %u\n",
	3430	+ (unsigned long long)map->m_lblk, map_len);
3455	3431
3456	3432	sbi = EXT4_SB(inode->i_sb);
3457	3433	eof_block = (EXT4_I(inode)->i_disksize + inode->i_sb->s_blocksize - 1)
..	..	@@ -3583,7 +3559,7 @@
3583	3559	}
3584	3560	if (allocated) {
3585	3561	/* Mark the block containing both extents as dirty */
3586		- ext4_ext_dirty(handle, inode, path + depth);
	3562	+ err = ext4_ext_dirty(handle, inode, path + depth);
3587	3563
3588	3564	/* Update path to point to the right extent */
3589	3565	path[depth].p_ext = abut_ex;
..	..	@@ -3601,9 +3577,6 @@
3601	3577	if (EXT4_EXT_MAY_ZEROOUT & split_flag)
3602	3578	max_zeroout = sbi->s_extent_max_zeroout_kb >>
3603	3579	(inode->i_sb->s_blocksize_bits - 10);
3604		-
3605		- if (IS_ENCRYPTED(inode))
3606		- max_zeroout = 0;
3607	3580
3608	3581	/*
3609	3582	* five cases:
..	..	@@ -3706,8 +3679,7 @@
3706	3679	unsigned int ee_len;
3707	3680	int split_flag = 0, depth;
3708	3681
3709		- ext_debug("%s: inode %lu, logical block %llu, max_blocks %u\n",
3710		- __func__, inode->i_ino,
	3682	+ ext_debug(inode, "logical block %llu, max_blocks %u\n",
3711	3683	(unsigned long long)map->m_lblk, map->m_len);
3712	3684
3713	3685	eof_block = (EXT4_I(inode)->i_disksize + inode->i_sb->s_blocksize - 1)
..	..	@@ -3716,7 +3688,7 @@
3716	3688	eof_block = map->m_lblk + map->m_len;
3717	3689	/*
3718	3690	* It is safe to convert extent to initialized via explicit
3719		- * zeroout only if extent is fully insde i_size or new_size.
	3691	+ * zeroout only if extent is fully inside i_size or new_size.
3720	3692	*/
3721	3693	depth = ext_depth(inode);
3722	3694	ex = path[depth].p_ext;
..	..	@@ -3753,8 +3725,7 @@
3753	3725	ee_block = le32_to_cpu(ex->ee_block);
3754	3726	ee_len = ext4_ext_get_actual_len(ex);
3755	3727
3756		- ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical"
3757		- "block %llu, max_blocks %u\n", inode->i_ino,
	3728	+ ext_debug(inode, "logical block %llu, max_blocks %u\n",
3758	3729	(unsigned long long)ee_block, ee_len);
3759	3730
3760	3731	/* If extent is larger than requested it is a clear sign that we still
..	..	@@ -3799,172 +3770,11 @@
3799	3770	return err;
3800	3771	}
3801	3772
3802		-/*
3803		- * Handle EOFBLOCKS_FL flag, clearing it if necessary
3804		- */
3805		-static int check_eofblocks_fl(handle_t handle, struct inode inode,
3806		- ext4_lblk_t lblk,
3807		- struct ext4_ext_path *path,
3808		- unsigned int len)
3809		-{
3810		- int i, depth;
3811		- struct ext4_extent_header *eh;
3812		- struct ext4_extent *last_ex;
3813		-
3814		- if (!ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))
3815		- return 0;
3816		-
3817		- depth = ext_depth(inode);
3818		- eh = path[depth].p_hdr;
3819		-
3820		- /*
3821		- * We're going to remove EOFBLOCKS_FL entirely in future so we
3822		- * do not care for this case anymore. Simply remove the flag
3823		- * if there are no extents.
3824		- */
3825		- if (unlikely(!eh->eh_entries))
3826		- goto out;
3827		- last_ex = EXT_LAST_EXTENT(eh);
3828		- /*
3829		- * We should clear the EOFBLOCKS_FL flag if we are writing the
3830		- * last block in the last extent in the file. We test this by
3831		- * first checking to see if the caller to
3832		- * ext4_ext_get_blocks() was interested in the last block (or
3833		- * a block beyond the last block) in the current extent. If
3834		- * this turns out to be false, we can bail out from this
3835		- * function immediately.
3836		- */
3837		- if (lblk + len < le32_to_cpu(last_ex->ee_block) +
3838		- ext4_ext_get_actual_len(last_ex))
3839		- return 0;
3840		- /*
3841		- * If the caller does appear to be planning to write at or
3842		- * beyond the end of the current extent, we then test to see
3843		- * if the current extent is the last extent in the file, by
3844		- * checking to make sure it was reached via the rightmost node
3845		- * at each level of the tree.
3846		- */
3847		- for (i = depth-1; i >= 0; i--)
3848		- if (path[i].p_idx != EXT_LAST_INDEX(path[i].p_hdr))
3849		- return 0;
3850		-out:
3851		- ext4_clear_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
3852		- return ext4_mark_inode_dirty(handle, inode);
3853		-}
3854		-
3855		-/**
3856		- * ext4_find_delalloc_range: find delayed allocated block in the given range.
3857		- *
3858		- * Return 1 if there is a delalloc block in the range, otherwise 0.
3859		- */
3860		-int ext4_find_delalloc_range(struct inode *inode,
3861		- ext4_lblk_t lblk_start,
3862		- ext4_lblk_t lblk_end)
3863		-{
3864		- struct extent_status es;
3865		-
3866		- ext4_es_find_delayed_extent_range(inode, lblk_start, lblk_end, &es);
3867		- if (es.es_len == 0)
3868		- return 0; /* there is no delay extent in this tree */
3869		- else if (es.es_lblk <= lblk_start &&
3870		- lblk_start < es.es_lblk + es.es_len)
3871		- return 1;
3872		- else if (lblk_start <= es.es_lblk && es.es_lblk <= lblk_end)
3873		- return 1;
3874		- else
3875		- return 0;
3876		-}
3877		-
3878		-int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk)
3879		-{
3880		- struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3881		- ext4_lblk_t lblk_start, lblk_end;
3882		- lblk_start = EXT4_LBLK_CMASK(sbi, lblk);
3883		- lblk_end = lblk_start + sbi->s_cluster_ratio - 1;
3884		-
3885		- return ext4_find_delalloc_range(inode, lblk_start, lblk_end);
3886		-}
3887		-
3888		-/**
3889		- * Determines how many complete clusters (out of those specified by the 'map')
3890		- * are under delalloc and were reserved quota for.
3891		- * This function is called when we are writing out the blocks that were
3892		- * originally written with their allocation delayed, but then the space was
3893		- * allocated using fallocate() before the delayed allocation could be resolved.
3894		- * The cases to look for are:
3895		- * ('=' indicated delayed allocated blocks
3896		- * '-' indicates non-delayed allocated blocks)
3897		- * (a) partial clusters towards beginning and/or end outside of allocated range
3898		- * are not delalloc'ed.
3899		- * Ex:
3900		- * \|----c---=\|====c====\|====c====\|===-c----\|
3901		- * \|++++++ allocated ++++++\|
3902		- * ==> 4 complete clusters in above example
3903		- *
3904		- * (b) partial cluster (outside of allocated range) towards either end is
3905		- * marked for delayed allocation. In this case, we will exclude that
3906		- * cluster.
3907		- * Ex:
3908		- * \|----====c========\|========c========\|
3909		- * \|++++++ allocated ++++++\|
3910		- * ==> 1 complete clusters in above example
3911		- *
3912		- * Ex:
3913		- * \|================c================\|
3914		- * \|++++++ allocated ++++++\|
3915		- * ==> 0 complete clusters in above example
3916		- *
3917		- * The ext4_da_update_reserve_space will be called only if we
3918		- * determine here that there were some "entire" clusters that span
3919		- * this 'allocated' range.
3920		- * In the non-bigalloc case, this function will just end up returning num_blks
3921		- * without ever calling ext4_find_delalloc_range.
3922		- */
3923		-static unsigned int
3924		-get_reserved_cluster_alloc(struct inode *inode, ext4_lblk_t lblk_start,
3925		- unsigned int num_blks)
3926		-{
3927		- struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
3928		- ext4_lblk_t alloc_cluster_start, alloc_cluster_end;
3929		- ext4_lblk_t lblk_from, lblk_to, c_offset;
3930		- unsigned int allocated_clusters = 0;
3931		-
3932		- alloc_cluster_start = EXT4_B2C(sbi, lblk_start);
3933		- alloc_cluster_end = EXT4_B2C(sbi, lblk_start + num_blks - 1);
3934		-
3935		- /* max possible clusters for this allocation */
3936		- allocated_clusters = alloc_cluster_end - alloc_cluster_start + 1;
3937		-
3938		- trace_ext4_get_reserved_cluster_alloc(inode, lblk_start, num_blks);
3939		-
3940		- /* Check towards left side */
3941		- c_offset = EXT4_LBLK_COFF(sbi, lblk_start);
3942		- if (c_offset) {
3943		- lblk_from = EXT4_LBLK_CMASK(sbi, lblk_start);
3944		- lblk_to = lblk_from + c_offset - 1;
3945		-
3946		- if (ext4_find_delalloc_range(inode, lblk_from, lblk_to))
3947		- allocated_clusters--;
3948		- }
3949		-
3950		- /* Now check towards right. */
3951		- c_offset = EXT4_LBLK_COFF(sbi, lblk_start + num_blks);
3952		- if (allocated_clusters && c_offset) {
3953		- lblk_from = lblk_start + num_blks;
3954		- lblk_to = lblk_from + (sbi->s_cluster_ratio - c_offset) - 1;
3955		-
3956		- if (ext4_find_delalloc_range(inode, lblk_from, lblk_to))
3957		- allocated_clusters--;
3958		- }
3959		-
3960		- return allocated_clusters;
3961		-}
3962		-
3963	3773	static int
3964	3774	convert_initialized_extent(handle_t handle, struct inode inode,
3965	3775	struct ext4_map_blocks *map,
3966	3776	struct ext4_ext_path **ppath,
3967		- unsigned int allocated)
	3777	+ unsigned int *allocated)
3968	3778	{
3969	3779	struct ext4_ext_path path = ppath;
3970	3780	struct ext4_extent *ex;
..	..	@@ -3985,8 +3795,7 @@
3985	3795	ee_block = le32_to_cpu(ex->ee_block);
3986	3796	ee_len = ext4_ext_get_actual_len(ex);
3987	3797
3988		- ext_debug("%s: inode %lu, logical"
3989		- "block %llu, max_blocks %u\n", __func__, inode->i_ino,
	3798	+ ext_debug(inode, "logical block %llu, max_blocks %u\n",
3990	3799	(unsigned long long)ee_block, ee_len);
3991	3800
3992	3801	if (ee_block != map->m_lblk \|\| ee_len > map->m_len) {
..	..	@@ -4024,14 +3833,12 @@
4024	3833	ext4_ext_show_leaf(inode, path);
4025	3834
4026	3835	ext4_update_inode_fsync_trans(handle, inode, 1);
4027		- err = check_eofblocks_fl(handle, inode, map->m_lblk, path, map->m_len);
4028		- if (err)
4029		- return err;
	3836	+
4030	3837	map->m_flags \|= EXT4_MAP_UNWRITTEN;
4031		- if (allocated > map->m_len)
4032		- allocated = map->m_len;
4033		- map->m_len = allocated;
4034		- return allocated;
	3838	+ if (*allocated > map->m_len)
	3839	+ *allocated = map->m_len;
	3840	+ map->m_len = *allocated;
	3841	+ return 0;
4035	3842	}
4036	3843
4037	3844	static int
..	..	@@ -4040,14 +3847,13 @@
4040	3847	struct ext4_ext_path **ppath, int flags,
4041	3848	unsigned int allocated, ext4_fsblk_t newblock)
4042	3849	{
4043		- struct ext4_ext_path path = ppath;
	3850	+ struct ext4_ext_path __maybe_unused path = ppath;
4044	3851	int ret = 0;
4045	3852	int err = 0;
4046	3853
4047		- ext_debug("ext4_ext_handle_unwritten_extents: inode %lu, logical "
4048		- "block %llu, max_blocks %u, flags %x, allocated %u\n",
4049		- inode->i_ino, (unsigned long long)map->m_lblk, map->m_len,
4050		- flags, allocated);
	3854	+ ext_debug(inode, "logical block %llu, max_blocks %u, flags 0x%x, allocated %u\n",
	3855	+ (unsigned long long)map->m_lblk, map->m_len, flags,
	3856	+ allocated);
4051	3857	ext4_ext_show_leaf(inode, path);
4052	3858
4053	3859	/*
..	..	@@ -4059,41 +3865,38 @@
4059	3865	trace_ext4_ext_handle_unwritten_extents(inode, map, flags,
4060	3866	allocated, newblock);
4061	3867
4062		- /* get_block() before submit the IO, split the extent */
	3868	+ /* get_block() before submitting IO, split the extent */
4063	3869	if (flags & EXT4_GET_BLOCKS_PRE_IO) {
4064	3870	ret = ext4_split_convert_extents(handle, inode, map, ppath,
4065	3871	flags \| EXT4_GET_BLOCKS_CONVERT);
4066		- if (ret <= 0)
4067		- goto out;
	3872	+ if (ret < 0) {
	3873	+ err = ret;
	3874	+ goto out2;
	3875	+ }
	3876	+ /*
	3877	+ * shouldn't get a 0 return when splitting an extent unless
	3878	+ * m_len is 0 (bug) or extent has been corrupted
	3879	+ */
	3880	+ if (unlikely(ret == 0)) {
	3881	+ EXT4_ERROR_INODE(inode,
	3882	+ "unexpected ret == 0, m_len = %u",
	3883	+ map->m_len);
	3884	+ err = -EFSCORRUPTED;
	3885	+ goto out2;
	3886	+ }
4068	3887	map->m_flags \|= EXT4_MAP_UNWRITTEN;
4069	3888	goto out;
4070	3889	}
4071	3890	/* IO end_io complete, convert the filled extent to written */
4072	3891	if (flags & EXT4_GET_BLOCKS_CONVERT) {
4073		- if (flags & EXT4_GET_BLOCKS_ZERO) {
4074		- if (allocated > map->m_len)
4075		- allocated = map->m_len;
4076		- err = ext4_issue_zeroout(inode, map->m_lblk, newblock,
4077		- allocated);
4078		- if (err < 0)
4079		- goto out2;
4080		- }
4081		- ret = ext4_convert_unwritten_extents_endio(handle, inode, map,
	3892	+ err = ext4_convert_unwritten_extents_endio(handle, inode, map,
4082	3893	ppath);
4083		- if (ret >= 0) {
4084		- ext4_update_inode_fsync_trans(handle, inode, 1);
4085		- err = check_eofblocks_fl(handle, inode, map->m_lblk,
4086		- path, map->m_len);
4087		- } else
4088		- err = ret;
4089		- map->m_flags \|= EXT4_MAP_MAPPED;
4090		- map->m_pblk = newblock;
4091		- if (allocated > map->m_len)
4092		- allocated = map->m_len;
4093		- map->m_len = allocated;
4094		- goto out2;
	3894	+ if (err < 0)
	3895	+ goto out2;
	3896	+ ext4_update_inode_fsync_trans(handle, inode, 1);
	3897	+ goto map_out;
4095	3898	}
4096		- /* buffered IO case */
	3899	+ /* buffered IO cases */
4097	3900	/*
4098	3901	* repeat fallocate creation request
4099	3902	* we already have an unwritten extent
..	..	@@ -4116,62 +3919,39 @@
4116	3919	goto out1;
4117	3920	}
4118	3921
4119		- /* buffered write, writepage time, convert*/
	3922	+ /*
	3923	+ * Default case when (flags & EXT4_GET_BLOCKS_CREATE) == 1.
	3924	+ * For buffered writes, at writepage time, etc. Convert a
	3925	+ * discovered unwritten extent to written.
	3926	+ */
4120	3927	ret = ext4_ext_convert_to_initialized(handle, inode, map, ppath, flags);
4121		- if (ret >= 0)
4122		- ext4_update_inode_fsync_trans(handle, inode, 1);
4123		-out:
4124		- if (ret <= 0) {
	3928	+ if (ret < 0) {
4125	3929	err = ret;
4126	3930	goto out2;
4127		- } else
4128		- allocated = ret;
	3931	+ }
	3932	+ ext4_update_inode_fsync_trans(handle, inode, 1);
	3933	+ /*
	3934	+ * shouldn't get a 0 return when converting an unwritten extent
	3935	+ * unless m_len is 0 (bug) or extent has been corrupted
	3936	+ */
	3937	+ if (unlikely(ret == 0)) {
	3938	+ EXT4_ERROR_INODE(inode, "unexpected ret == 0, m_len = %u",
	3939	+ map->m_len);
	3940	+ err = -EFSCORRUPTED;
	3941	+ goto out2;
	3942	+ }
	3943	+
	3944	+out:
	3945	+ allocated = ret;
4129	3946	map->m_flags \|= EXT4_MAP_NEW;
4130		- /*
4131		- * if we allocated more blocks than requested
4132		- * we need to make sure we unmap the extra block
4133		- * allocated. The actual needed block will get
4134		- * unmapped later when we find the buffer_head marked
4135		- * new.
4136		- */
4137		- if (allocated > map->m_len) {
4138		- clean_bdev_aliases(inode->i_sb->s_bdev, newblock + map->m_len,
4139		- allocated - map->m_len);
4140		- allocated = map->m_len;
4141		- }
4142		- map->m_len = allocated;
4143		-
4144		- /*
4145		- * If we have done fallocate with the offset that is already
4146		- * delayed allocated, we would have block reservation
4147		- * and quota reservation done in the delayed write path.
4148		- * But fallocate would have already updated quota and block
4149		- * count for this offset. So cancel these reservation
4150		- */
4151		- if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
4152		- unsigned int reserved_clusters;
4153		- reserved_clusters = get_reserved_cluster_alloc(inode,
4154		- map->m_lblk, map->m_len);
4155		- if (reserved_clusters)
4156		- ext4_da_update_reserve_space(inode,
4157		- reserved_clusters,
4158		- 0);
4159		- }
4160		-
4161	3947	map_out:
4162	3948	map->m_flags \|= EXT4_MAP_MAPPED;
4163		- if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0) {
4164		- err = check_eofblocks_fl(handle, inode, map->m_lblk, path,
4165		- map->m_len);
4166		- if (err < 0)
4167		- goto out2;
4168		- }
4169	3949	out1:
	3950	+ map->m_pblk = newblock;
4170	3951	if (allocated > map->m_len)
4171	3952	allocated = map->m_len;
4172		- ext4_ext_show_leaf(inode, path);
4173		- map->m_pblk = newblock;
4174	3953	map->m_len = allocated;
	3954	+ ext4_ext_show_leaf(inode, path);
4175	3955	out2:
4176	3956	return err ? err : allocated;
4177	3957	}
..	..	@@ -4287,7 +4067,7 @@
4287	4067	* down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
4288	4068	* (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
4289	4069	*
4290		- * return > 0, number of of blocks already mapped/allocated
	4070	+ * return > 0, number of blocks already mapped/allocated
4291	4071	* if create == 0 and these are pre-allocated blocks
4292	4072	* buffer head is unmapped
4293	4073	* otherwise blocks are mapped
..	..	@@ -4301,18 +4081,16 @@
4301	4081	struct ext4_map_blocks *map, int flags)
4302	4082	{
4303	4083	struct ext4_ext_path *path = NULL;
4304		- struct ext4_extent newex, ex, ex2;
	4084	+ struct ext4_extent newex, *ex, ex2;
4305	4085	struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
4306		- ext4_fsblk_t newblock = 0;
4307		- int free_on_err = 0, err = 0, depth, ret;
	4086	+ ext4_fsblk_t newblock = 0, pblk;
	4087	+ int err = 0, depth, ret;
4308	4088	unsigned int allocated = 0, offset = 0;
4309	4089	unsigned int allocated_clusters = 0;
4310	4090	struct ext4_allocation_request ar;
4311	4091	ext4_lblk_t cluster_offset;
4312		- bool map_from_cluster = false;
4313	4092
4314		- ext_debug("blocks %u/%u requested for inode %lu\n",
4315		- map->m_lblk, map->m_len, inode->i_ino);
	4093	+ ext_debug(inode, "blocks %u/%u requested\n", map->m_lblk, map->m_len);
4316	4094	trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags);
4317	4095
4318	4096	/* find extent for this block */
..	..	@@ -4320,7 +4098,7 @@
4320	4098	if (IS_ERR(path)) {
4321	4099	err = PTR_ERR(path);
4322	4100	path = NULL;
4323		- goto out2;
	4101	+ goto out;
4324	4102	}
4325	4103
4326	4104	depth = ext_depth(inode);
..	..	@@ -4336,7 +4114,7 @@
4336	4114	(unsigned long) map->m_lblk, depth,
4337	4115	path[depth].p_block);
4338	4116	err = -EFSCORRUPTED;
4339		- goto out2;
	4117	+ goto out;
4340	4118	}
4341	4119
4342	4120	ex = path[depth].p_ext;
..	..	@@ -4359,8 +4137,8 @@
4359	4137	newblock = map->m_lblk - ee_block + ee_start;
4360	4138	/* number of remaining blocks in the extent */
4361	4139	allocated = ee_len - (map->m_lblk - ee_block);
4362		- ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk,
4363		- ee_block, ee_len, newblock);
	4140	+ ext_debug(inode, "%u fit into %u:%d -> %llu\n",
	4141	+ map->m_lblk, ee_block, ee_len, newblock);
4364	4142
4365	4143	/*
4366	4144	* If the extent is initialized check whether the
..	..	@@ -4368,12 +4146,18 @@
4368	4146	*/
4369	4147	if ((!ext4_ext_is_unwritten(ex)) &&
4370	4148	(flags & EXT4_GET_BLOCKS_CONVERT_UNWRITTEN)) {
4371		- allocated = convert_initialized_extent(
4372		- handle, inode, map, &path,
4373		- allocated);
4374		- goto out2;
4375		- } else if (!ext4_ext_is_unwritten(ex))
	4149	+ err = convert_initialized_extent(handle,
	4150	+ inode, map, &path, &allocated);
4376	4151	goto out;
	4152	+ } else if (!ext4_ext_is_unwritten(ex)) {
	4153	+ map->m_flags \|= EXT4_MAP_MAPPED;
	4154	+ map->m_pblk = newblock;
	4155	+ if (allocated > map->m_len)
	4156	+ allocated = map->m_len;
	4157	+ map->m_len = allocated;
	4158	+ ext4_ext_show_leaf(inode, path);
	4159	+ goto out;
	4160	+ }
4377	4161
4378	4162	ret = ext4_ext_handle_unwritten_extents(
4379	4163	handle, inode, map, &path, flags,
..	..	@@ -4382,7 +4166,7 @@
4382	4166	err = ret;
4383	4167	else
4384	4168	allocated = ret;
4385		- goto out2;
	4169	+ goto out;
4386	4170	}
4387	4171	}
4388	4172
..	..	@@ -4407,7 +4191,7 @@
4407	4191	map->m_pblk = 0;
4408	4192	map->m_len = min_t(unsigned int, map->m_len, hole_len);
4409	4193
4410		- goto out2;
	4194	+ goto out;
4411	4195	}
4412	4196
4413	4197	/*
..	..	@@ -4424,7 +4208,6 @@
4424	4208	get_implied_cluster_alloc(inode->i_sb, map, ex, path)) {
4425	4209	ar.len = allocated = map->m_len;
4426	4210	newblock = map->m_pblk;
4427		- map_from_cluster = true;
4428	4211	goto got_allocated_blocks;
4429	4212	}
4430	4213
..	..	@@ -4432,20 +4215,18 @@
4432	4215	ar.lleft = map->m_lblk;
4433	4216	err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
4434	4217	if (err)
4435		- goto out2;
	4218	+ goto out;
4436	4219	ar.lright = map->m_lblk;
4437		- ex2 = NULL;
4438	4220	err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright, &ex2);
4439		- if (err)
4440		- goto out2;
	4221	+ if (err < 0)
	4222	+ goto out;
4441	4223
4442	4224	/* Check if the extent after searching to the right implies a
4443	4225	* cluster we can use. */
4444		- if ((sbi->s_cluster_ratio > 1) && ex2 &&
4445		- get_implied_cluster_alloc(inode->i_sb, map, ex2, path)) {
	4226	+ if ((sbi->s_cluster_ratio > 1) && err &&
	4227	+ get_implied_cluster_alloc(inode->i_sb, map, &ex2, path)) {
4446	4228	ar.len = allocated = map->m_len;
4447	4229	newblock = map->m_pblk;
4448		- map_from_cluster = true;
4449	4230	goto got_allocated_blocks;
4450	4231	}
4451	4232
..	..	@@ -4499,124 +4280,79 @@
4499	4280	ar.flags \|= EXT4_MB_USE_RESERVED;
4500	4281	newblock = ext4_mb_new_blocks(handle, &ar, &err);
4501	4282	if (!newblock)
4502		- goto out2;
4503		- ext_debug("allocate new block: goal %llu, found %llu/%u\n",
4504		- ar.goal, newblock, allocated);
4505		- free_on_err = 1;
	4283	+ goto out;
4506	4284	allocated_clusters = ar.len;
4507	4285	ar.len = EXT4_C2B(sbi, ar.len) - offset;
	4286	+ ext_debug(inode, "allocate new block: goal %llu, found %llu/%u, requested %u\n",
	4287	+ ar.goal, newblock, ar.len, allocated);
4508	4288	if (ar.len > allocated)
4509	4289	ar.len = allocated;
4510	4290
4511	4291	got_allocated_blocks:
4512	4292	/* try to insert new extent into found leaf and return */
4513		- ext4_ext_store_pblock(&newex, newblock + offset);
	4293	+ pblk = newblock + offset;
	4294	+ ext4_ext_store_pblock(&newex, pblk);
4514	4295	newex.ee_len = cpu_to_le16(ar.len);
4515	4296	/* Mark unwritten */
4516		- if (flags & EXT4_GET_BLOCKS_UNWRIT_EXT){
	4297	+ if (flags & EXT4_GET_BLOCKS_UNWRIT_EXT) {
4517	4298	ext4_ext_mark_unwritten(&newex);
4518	4299	map->m_flags \|= EXT4_MAP_UNWRITTEN;
4519	4300	}
4520	4301
4521		- err = 0;
4522		- if ((flags & EXT4_GET_BLOCKS_KEEP_SIZE) == 0)
4523		- err = check_eofblocks_fl(handle, inode, map->m_lblk,
4524		- path, ar.len);
4525		- if (!err)
4526		- err = ext4_ext_insert_extent(handle, inode, &path,
4527		- &newex, flags);
	4302	+ err = ext4_ext_insert_extent(handle, inode, &path, &newex, flags);
	4303	+ if (err) {
	4304	+ if (allocated_clusters) {
	4305	+ int fb_flags = 0;
4528	4306
4529		- if (err && free_on_err) {
4530		- int fb_flags = flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE ?
4531		- EXT4_FREE_BLOCKS_NO_QUOT_UPDATE : 0;
4532		- /* free data blocks we just allocated */
4533		- /* not a good idea to call discard here directly,
4534		- * but otherwise we'd need to call it every free() */
4535		- ext4_discard_preallocations(inode);
4536		- ext4_free_blocks(handle, inode, NULL, newblock,
4537		- EXT4_C2B(sbi, allocated_clusters), fb_flags);
4538		- goto out2;
	4307	+ /*
	4308	+ * free data blocks we just allocated.
	4309	+ * not a good idea to call discard here directly,
	4310	+ * but otherwise we'd need to call it every free().
	4311	+ */
	4312	+ ext4_discard_preallocations(inode, 0);
	4313	+ if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)
	4314	+ fb_flags = EXT4_FREE_BLOCKS_NO_QUOT_UPDATE;
	4315	+ ext4_free_blocks(handle, inode, NULL, newblock,
	4316	+ EXT4_C2B(sbi, allocated_clusters),
	4317	+ fb_flags);
	4318	+ }
	4319	+ goto out;
4539	4320	}
4540	4321
4541		- /* previous routine could use block we allocated */
4542		- newblock = ext4_ext_pblock(&newex);
4543		- allocated = ext4_ext_get_actual_len(&newex);
4544		- if (allocated > map->m_len)
4545		- allocated = map->m_len;
4546		- map->m_flags \|= EXT4_MAP_NEW;
4547		-
4548	4322	/*
4549		- * Update reserved blocks/metadata blocks after successful
4550		- * block allocation which had been deferred till now.
	4323	+ * Reduce the reserved cluster count to reflect successful deferred
	4324	+ * allocation of delayed allocated clusters or direct allocation of
	4325	+ * clusters discovered to be delayed allocated. Once allocated, a
	4326	+ * cluster is not included in the reserved count.
4551	4327	*/
4552		- if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
4553		- unsigned int reserved_clusters;
4554		- /*
4555		- * Check how many clusters we had reserved this allocated range
4556		- */
4557		- reserved_clusters = get_reserved_cluster_alloc(inode,
4558		- map->m_lblk, allocated);
4559		- if (!map_from_cluster) {
4560		- BUG_ON(allocated_clusters < reserved_clusters);
4561		- if (reserved_clusters < allocated_clusters) {
4562		- struct ext4_inode_info *ei = EXT4_I(inode);
4563		- int reservation = allocated_clusters -
4564		- reserved_clusters;
4565		- /*
4566		- * It seems we claimed few clusters outside of
4567		- * the range of this allocation. We should give
4568		- * it back to the reservation pool. This can
4569		- * happen in the following case:
4570		- *
4571		- * * Suppose s_cluster_ratio is 4 (i.e., each
4572		- * cluster has 4 blocks. Thus, the clusters
4573		- * are [0-3],[4-7],[8-11]...
4574		- * * First comes delayed allocation write for
4575		- * logical blocks 10 & 11. Since there were no
4576		- * previous delayed allocated blocks in the
4577		- * range [8-11], we would reserve 1 cluster
4578		- * for this write.
4579		- * * Next comes write for logical blocks 3 to 8.
4580		- * In this case, we will reserve 2 clusters
4581		- * (for [0-3] and [4-7]; and not for [8-11] as
4582		- * that range has a delayed allocated blocks.
4583		- * Thus total reserved clusters now becomes 3.
4584		- * * Now, during the delayed allocation writeout
4585		- * time, we will first write blocks [3-8] and
4586		- * allocate 3 clusters for writing these
4587		- * blocks. Also, we would claim all these
4588		- * three clusters above.
4589		- * * Now when we come here to writeout the
4590		- * blocks [10-11], we would expect to claim
4591		- * the reservation of 1 cluster we had made
4592		- * (and we would claim it since there are no
4593		- * more delayed allocated blocks in the range
4594		- * [8-11]. But our reserved cluster count had
4595		- * already gone to 0.
4596		- *
4597		- * Thus, at the step 4 above when we determine
4598		- * that there are still some unwritten delayed
4599		- * allocated blocks outside of our current
4600		- * block range, we should increment the
4601		- * reserved clusters count so that when the
4602		- * remaining blocks finally gets written, we
4603		- * could claim them.
4604		- */
4605		- dquot_reserve_block(inode,
4606		- EXT4_C2B(sbi, reservation));
4607		- spin_lock(&ei->i_block_reservation_lock);
4608		- ei->i_reserved_data_blocks += reservation;
4609		- spin_unlock(&ei->i_block_reservation_lock);
4610		- }
	4328	+ if (test_opt(inode->i_sb, DELALLOC) && allocated_clusters) {
	4329	+ if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) {
4611	4330	/*
4612		- * We will claim quota for all newly allocated blocks.
4613		- * We're updating the reserved space after the
4614		- * correction above so we do not accidentally free
4615		- * all the metadata reservation because we might
4616		- * actually need it later on.
	4331	+ * When allocating delayed allocated clusters, simply
	4332	+ * reduce the reserved cluster count and claim quota
4617	4333	*/
4618	4334	ext4_da_update_reserve_space(inode, allocated_clusters,
4619	4335	1);
	4336	+ } else {
	4337	+ ext4_lblk_t lblk, len;
	4338	+ unsigned int n;
	4339	+
	4340	+ /*
	4341	+ * When allocating non-delayed allocated clusters
	4342	+ * (from fallocate, filemap, DIO, or clusters
	4343	+ * allocated when delalloc has been disabled by
	4344	+ * ext4_nonda_switch), reduce the reserved cluster
	4345	+ * count by the number of allocated clusters that
	4346	+ * have previously been delayed allocated. Quota
	4347	+ * has been claimed by ext4_mb_new_blocks() above,
	4348	+ * so release the quota reservations made for any
	4349	+ * previously delayed allocated clusters.
	4350	+ */
	4351	+ lblk = EXT4_LBLK_CMASK(sbi, map->m_lblk);
	4352	+ len = allocated_clusters << sbi->s_cluster_bits;
	4353	+ n = ext4_es_delayed_clu(inode, lblk, len);
	4354	+ if (n > 0)
	4355	+ ext4_da_update_reserve_space(inode, (int) n, 0);
4620	4356	}
4621	4357	}
4622	4358
..	..	@@ -4628,14 +4364,13 @@
4628	4364	ext4_update_inode_fsync_trans(handle, inode, 1);
4629	4365	else
4630	4366	ext4_update_inode_fsync_trans(handle, inode, 0);
4631		-out:
4632		- if (allocated > map->m_len)
4633		- allocated = map->m_len;
	4367	+
	4368	+ map->m_flags \|= (EXT4_MAP_NEW \| EXT4_MAP_MAPPED);
	4369	+ map->m_pblk = pblk;
	4370	+ map->m_len = ar.len;
	4371	+ allocated = map->m_len;
4634	4372	ext4_ext_show_leaf(inode, path);
4635		- map->m_flags \|= EXT4_MAP_MAPPED;
4636		- map->m_pblk = newblock;
4637		- map->m_len = allocated;
4638		-out2:
	4373	+out:
4639	4374	ext4_ext_drop_refs(path);
4640	4375	kfree(path);
4641	4376
..	..	@@ -4674,7 +4409,14 @@
4674	4409	}
4675	4410	if (err)
4676	4411	return err;
4677		- return ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1);
	4412	+retry_remove_space:
	4413	+ err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1);
	4414	+ if (err == -ENOMEM) {
	4415	+ cond_resched();
	4416	+ congestion_wait(BLK_RW_ASYNC, HZ/50);
	4417	+ goto retry_remove_space;
	4418	+ }
	4419	+ return err;
4678	4420	}
4679	4421
4680	4422	static int ext4_alloc_file_blocks(struct file *file, ext4_lblk_t offset,
..	..	@@ -4684,7 +4426,7 @@
4684	4426	struct inode *inode = file_inode(file);
4685	4427	handle_t *handle;
4686	4428	int ret = 0;
4687		- int ret2 = 0;
	4429	+ int ret2 = 0, ret3 = 0;
4688	4430	int retries = 0;
4689	4431	int depth = 0;
4690	4432	struct ext4_map_blocks map;
..	..	@@ -4743,15 +4485,12 @@
4743	4485	epos = new_size;
4744	4486	if (ext4_update_inode_size(inode, epos) & 0x1)
4745	4487	inode->i_mtime = inode->i_ctime;
4746		- } else {
4747		- if (epos > inode->i_size)
4748		- ext4_set_inode_flag(inode,
4749		- EXT4_INODE_EOFBLOCKS);
4750	4488	}
4751		- ext4_mark_inode_dirty(handle, inode);
	4489	+ ret2 = ext4_mark_inode_dirty(handle, inode);
4752	4490	ext4_update_inode_fsync_trans(handle, inode, 1);
4753		- ret2 = ext4_journal_stop(handle);
4754		- if (ret2)
	4491	+ ret3 = ext4_journal_stop(handle);
	4492	+ ret2 = ret3 ? ret3 : ret2;
	4493	+ if (unlikely(ret2))
4755	4494	break;
4756	4495	}
4757	4496	if (ret == -ENOSPC &&
..	..	@@ -4762,6 +4501,10 @@
4762	4501
4763	4502	return ret > 0 ? ret2 : ret;
4764	4503	}
	4504	+
	4505	+static int ext4_collapse_range(struct file *file, loff_t offset, loff_t len);
	4506	+
	4507	+static int ext4_insert_range(struct file *file, loff_t offset, loff_t len);
4765	4508
4766	4509	static long ext4_zero_range(struct file *file, loff_t offset,
4767	4510	loff_t len, int mode)
..	..	@@ -4780,9 +4523,6 @@
4780	4523
4781	4524	trace_ext4_zero_range(inode, offset, len, mode);
4782	4525
4783		- if (!S_ISREG(inode->i_mode))
4784		- return -EINVAL;
4785		-
4786	4526	/* Call ext4_force_commit to flush all data in case of data=journal. */
4787	4527	if (ext4_should_journal_data(inode)) {
4788	4528	ret = ext4_force_commit(inode->i_sb);
..	..	@@ -4791,7 +4531,7 @@
4791	4531	}
4792	4532
4793	4533	/*
4794		- * Round up offset. This is not fallocate, we neet to zero out
	4534	+ * Round up offset. This is not fallocate, we need to zero out
4795	4535	* blocks, so convert interior block aligned part of the range to
4796	4536	* unwritten and possibly manually zero out unaligned parts of the
4797	4537	* range.
..	..	@@ -4814,7 +4554,7 @@
4814	4554	inode_lock(inode);
4815	4555
4816	4556	/*
4817		- * Indirect files do not support unwritten extnets
	4557	+ * Indirect files do not support unwritten extents
4818	4558	*/
4819	4559	if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) {
4820	4560	ret = -EOPNOTSUPP;
..	..	@@ -4822,7 +4562,7 @@
4822	4562	}
4823	4563
4824	4564	if (!(mode & FALLOC_FL_KEEP_SIZE) &&
4825		- (offset + len > i_size_read(inode) \|\|
	4565	+ (offset + len > inode->i_size \|\|
4826	4566	offset + len > EXT4_I(inode)->i_disksize)) {
4827	4567	new_size = offset + len;
4828	4568	ret = inode_newsize_ok(inode, new_size);
..	..	@@ -4831,11 +4571,13 @@
4831	4571	}
4832	4572
4833	4573	flags = EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT;
4834		- if (mode & FALLOC_FL_KEEP_SIZE)
4835		- flags \|= EXT4_GET_BLOCKS_KEEP_SIZE;
4836	4574
4837	4575	/* Wait all existing dio workers, newcomers will block on i_mutex */
4838	4576	inode_dio_wait(inode);
	4577	+
	4578	+ ret = file_modified(file);
	4579	+ if (ret)
	4580	+ goto out_mutex;
4839	4581
4840	4582	/* Preallocate the range including the unaligned edges */
4841	4583	if (partial_begin \|\| partial_end) {
..	..	@@ -4899,18 +4641,11 @@
4899	4641	}
4900	4642
4901	4643	inode->i_mtime = inode->i_ctime = current_time(inode);
4902		- if (new_size) {
	4644	+ if (new_size)
4903	4645	ext4_update_inode_size(inode, new_size);
4904		- } else {
4905		- /*
4906		- * Mark that we allocate beyond EOF so the subsequent truncate
4907		- * can proceed even if the new size is the same as i_size.
4908		- */
4909		- if ((offset + len) > i_size_read(inode))
4910		- ext4_set_inode_flag(inode, EXT4_INODE_EOFBLOCKS);
4911		- }
4912		- ext4_mark_inode_dirty(handle, inode);
4913		-
	4646	+ ret = ext4_mark_inode_dirty(handle, inode);
	4647	+ if (unlikely(ret))
	4648	+ goto out_handle;
4914	4649	/* Zero out partial block at the edges of the range */
4915	4650	ret = ext4_zero_partial_blocks(handle, inode, offset, len);
4916	4651	if (ret >= 0)
..	..	@@ -4919,6 +4654,7 @@
4919	4654	if (file->f_flags & O_SYNC)
4920	4655	ext4_handle_sync(handle);
4921	4656
	4657	+out_handle:
4922	4658	ext4_journal_stop(handle);
4923	4659	out_mutex:
4924	4660	inode_unlock(inode);
..	..	@@ -4947,14 +4683,9 @@
4947	4683	* range since we would need to re-encrypt blocks with a
4948	4684	* different IV or XTS tweak (which are based on the logical
4949	4685	* block number).
4950		- *
4951		- * XXX It's not clear why zero range isn't working, but we'll
4952		- * leave it disabled for encrypted inodes for now. This is a
4953		- * bug we should fix....
4954	4686	*/
4955	4687	if (IS_ENCRYPTED(inode) &&
4956		- (mode & (FALLOC_FL_COLLAPSE_RANGE \| FALLOC_FL_INSERT_RANGE \|
4957		- FALLOC_FL_ZERO_RANGE)))
	4688	+ (mode & (FALLOC_FL_COLLAPSE_RANGE \| FALLOC_FL_INSERT_RANGE)))
4958	4689	return -EOPNOTSUPP;
4959	4690
4960	4691	/* Return error if mode is not supported */
..	..	@@ -4963,29 +4694,37 @@
4963	4694	FALLOC_FL_INSERT_RANGE))
4964	4695	return -EOPNOTSUPP;
4965	4696
4966		- if (mode & FALLOC_FL_PUNCH_HOLE)
4967		- return ext4_punch_hole(inode, offset, len);
4968		-
	4697	+ ext4_fc_start_update(inode);
	4698	+ inode_lock(inode);
4969	4699	ret = ext4_convert_inline_data(inode);
	4700	+ inode_unlock(inode);
4970	4701	if (ret)
4971		- return ret;
	4702	+ goto exit;
4972	4703
4973		- if (mode & FALLOC_FL_COLLAPSE_RANGE)
4974		- return ext4_collapse_range(inode, offset, len);
	4704	+ if (mode & FALLOC_FL_PUNCH_HOLE) {
	4705	+ ret = ext4_punch_hole(file, offset, len);
	4706	+ goto exit;
	4707	+ }
4975	4708
4976		- if (mode & FALLOC_FL_INSERT_RANGE)
4977		- return ext4_insert_range(inode, offset, len);
	4709	+ if (mode & FALLOC_FL_COLLAPSE_RANGE) {
	4710	+ ret = ext4_collapse_range(file, offset, len);
	4711	+ goto exit;
	4712	+ }
4978	4713
4979		- if (mode & FALLOC_FL_ZERO_RANGE)
4980		- return ext4_zero_range(file, offset, len, mode);
	4714	+ if (mode & FALLOC_FL_INSERT_RANGE) {
	4715	+ ret = ext4_insert_range(file, offset, len);
	4716	+ goto exit;
	4717	+ }
4981	4718
	4719	+ if (mode & FALLOC_FL_ZERO_RANGE) {
	4720	+ ret = ext4_zero_range(file, offset, len, mode);
	4721	+ goto exit;
	4722	+ }
4982	4723	trace_ext4_fallocate_enter(inode, offset, len, mode);
4983	4724	lblk = offset >> blkbits;
4984	4725
4985	4726	max_blocks = EXT4_MAX_BLOCKS(len, offset, blkbits);
4986	4727	flags = EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT;
4987		- if (mode & FALLOC_FL_KEEP_SIZE)
4988		- flags \|= EXT4_GET_BLOCKS_KEEP_SIZE;
4989	4728
4990	4729	inode_lock(inode);
4991	4730
..	..	@@ -4998,7 +4737,7 @@
4998	4737	}
4999	4738
5000	4739	if (!(mode & FALLOC_FL_KEEP_SIZE) &&
5001		- (offset + len > i_size_read(inode) \|\|
	4740	+ (offset + len > inode->i_size \|\|
5002	4741	offset + len > EXT4_I(inode)->i_disksize)) {
5003	4742	new_size = offset + len;
5004	4743	ret = inode_newsize_ok(inode, new_size);
..	..	@@ -5009,17 +4748,23 @@
5009	4748	/* Wait all existing dio workers, newcomers will block on i_mutex */
5010	4749	inode_dio_wait(inode);
5011	4750
	4751	+ ret = file_modified(file);
	4752	+ if (ret)
	4753	+ goto out;
	4754	+
5012	4755	ret = ext4_alloc_file_blocks(file, lblk, max_blocks, new_size, flags);
5013	4756	if (ret)
5014	4757	goto out;
5015	4758
5016	4759	if (file->f_flags & O_SYNC && EXT4_SB(inode->i_sb)->s_journal) {
5017		- ret = jbd2_complete_transaction(EXT4_SB(inode->i_sb)->s_journal,
5018		- EXT4_I(inode)->i_sync_tid);
	4760	+ ret = ext4_fc_commit(EXT4_SB(inode->i_sb)->s_journal,
	4761	+ EXT4_I(inode)->i_sync_tid);
5019	4762	}
5020	4763	out:
5021	4764	inode_unlock(inode);
5022	4765	trace_ext4_fallocate_exit(inode, offset, max_blocks, ret);
	4766	+exit:
	4767	+ ext4_fc_stop_update(inode);
5023	4768	return ret;
5024	4769	}
5025	4770
..	..	@@ -5037,26 +4782,15 @@
5037	4782	loff_t offset, ssize_t len)
5038	4783	{
5039	4784	unsigned int max_blocks;
5040		- int ret = 0;
5041		- int ret2 = 0;
	4785	+ int ret = 0, ret2 = 0, ret3 = 0;
5042	4786	struct ext4_map_blocks map;
5043		- unsigned int credits, blkbits = inode->i_blkbits;
	4787	+ unsigned int blkbits = inode->i_blkbits;
	4788	+ unsigned int credits = 0;
5044	4789
5045	4790	map.m_lblk = offset >> blkbits;
5046	4791	max_blocks = EXT4_MAX_BLOCKS(len, offset, blkbits);
5047	4792
5048		- /*
5049		- * This is somewhat ugly but the idea is clear: When transaction is
5050		- * reserved, everything goes into it. Otherwise we rather start several
5051		- * smaller transactions for conversion of each extent separately.
5052		- */
5053		- if (handle) {
5054		- handle = ext4_journal_start_reserved(handle,
5055		- EXT4_HT_EXT_CONVERT);
5056		- if (IS_ERR(handle))
5057		- return PTR_ERR(handle);
5058		- credits = 0;
5059		- } else {
	4793	+ if (!handle) {
5060	4794	/*
5061	4795	* credits to insert 1 extent into extent tree
5062	4796	*/
..	..	@@ -5081,74 +4815,57 @@
5081	4815	"ext4_ext_map_blocks returned %d",
5082	4816	inode->i_ino, map.m_lblk,
5083	4817	map.m_len, ret);
5084		- ext4_mark_inode_dirty(handle, inode);
5085		- if (credits)
5086		- ret2 = ext4_journal_stop(handle);
	4818	+ ret2 = ext4_mark_inode_dirty(handle, inode);
	4819	+ if (credits) {
	4820	+ ret3 = ext4_journal_stop(handle);
	4821	+ if (unlikely(ret3))
	4822	+ ret2 = ret3;
	4823	+ }
	4824	+
5087	4825	if (ret <= 0 \|\| ret2)
5088	4826	break;
5089	4827	}
5090		- if (!credits)
5091		- ret2 = ext4_journal_stop(handle);
5092	4828	return ret > 0 ? ret2 : ret;
5093	4829	}
5094	4830
5095		-/*
5096		- * If newes is not existing extent (newes->ec_pblk equals zero) find
5097		- * delayed extent at start of newes and update newes accordingly and
5098		- * return start of the next delayed extent.
5099		- *
5100		- * If newes is existing extent (newes->ec_pblk is not equal zero)
5101		- * return start of next delayed extent or EXT_MAX_BLOCKS if no delayed
5102		- * extent found. Leave newes unmodified.
5103		- */
5104		-static int ext4_find_delayed_extent(struct inode *inode,
5105		- struct extent_status *newes)
	4831	+int ext4_convert_unwritten_io_end_vec(handle_t handle, ext4_io_end_t io_end)
5106	4832	{
5107		- struct extent_status es;
5108		- ext4_lblk_t block, next_del;
	4833	+ int ret = 0, err = 0;
	4834	+ struct ext4_io_end_vec *io_end_vec;
5109	4835
5110		- if (newes->es_pblk == 0) {
5111		- ext4_es_find_delayed_extent_range(inode, newes->es_lblk,
5112		- newes->es_lblk + newes->es_len - 1, &es);
5113		-
5114		- /*
5115		- * No extent in extent-tree contains block @newes->es_pblk,
5116		- * then the block may stay in 1)a hole or 2)delayed-extent.
5117		- */
5118		- if (es.es_len == 0)
5119		- /* A hole found. */
5120		- return 0;
5121		-
5122		- if (es.es_lblk > newes->es_lblk) {
5123		- /* A hole found. */
5124		- newes->es_len = min(es.es_lblk - newes->es_lblk,
5125		- newes->es_len);
5126		- return 0;
5127		- }
5128		-
5129		- newes->es_len = es.es_lblk + es.es_len - newes->es_lblk;
	4836	+ /*
	4837	+ * This is somewhat ugly but the idea is clear: When transaction is
	4838	+ * reserved, everything goes into it. Otherwise we rather start several
	4839	+ * smaller transactions for conversion of each extent separately.
	4840	+ */
	4841	+ if (handle) {
	4842	+ handle = ext4_journal_start_reserved(handle,
	4843	+ EXT4_HT_EXT_CONVERT);
	4844	+ if (IS_ERR(handle))
	4845	+ return PTR_ERR(handle);
5130	4846	}
5131	4847
5132		- block = newes->es_lblk + newes->es_len;
5133		- ext4_es_find_delayed_extent_range(inode, block, EXT_MAX_BLOCKS, &es);
5134		- if (es.es_len == 0)
5135		- next_del = EXT_MAX_BLOCKS;
5136		- else
5137		- next_del = es.es_lblk;
	4848	+ list_for_each_entry(io_end_vec, &io_end->list_vec, list) {
	4849	+ ret = ext4_convert_unwritten_extents(handle, io_end->inode,
	4850	+ io_end_vec->offset,
	4851	+ io_end_vec->size);
	4852	+ if (ret)
	4853	+ break;
	4854	+ }
5138	4855
5139		- return next_del;
	4856	+ if (handle)
	4857	+ err = ext4_journal_stop(handle);
	4858	+
	4859	+ return ret < 0 ? ret : err;
5140	4860	}
5141		-/* fiemap flags we can handle specified here */
5142		-#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC\|FIEMAP_FLAG_XATTR)
5143	4861
5144		-static int ext4_xattr_fiemap(struct inode *inode,
5145		- struct fiemap_extent_info *fieinfo)
	4862	+static int ext4_iomap_xattr_fiemap(struct inode inode, struct iomap iomap)
5146	4863	{
5147	4864	__u64 physical = 0;
5148		- __u64 length;
5149		- __u32 flags = FIEMAP_EXTENT_LAST;
	4865	+ __u64 length = 0;
5150	4866	int blockbits = inode->i_sb->s_blocksize_bits;
5151	4867	int error = 0;
	4868	+ u16 iomap_type;
5152	4869
5153	4870	/* in-inode? */
5154	4871	if (ext4_test_inode_state(inode, EXT4_STATE_XATTR)) {
..	..	@@ -5163,102 +4880,138 @@
5163	4880	EXT4_I(inode)->i_extra_isize;
5164	4881	physical += offset;
5165	4882	length = EXT4_SB(inode->i_sb)->s_inode_size - offset;
5166		- flags \|= FIEMAP_EXTENT_DATA_INLINE;
5167	4883	brelse(iloc.bh);
5168		- } else { /* external block */
	4884	+ iomap_type = IOMAP_INLINE;
	4885	+ } else if (EXT4_I(inode)->i_file_acl) { /* external block */
5169	4886	physical = (__u64)EXT4_I(inode)->i_file_acl << blockbits;
5170	4887	length = inode->i_sb->s_blocksize;
	4888	+ iomap_type = IOMAP_MAPPED;
	4889	+ } else {
	4890	+ /* no in-inode or external block for xattr, so return -ENOENT */
	4891	+ error = -ENOENT;
	4892	+ goto out;
5171	4893	}
5172	4894
5173		- if (physical)
5174		- error = fiemap_fill_next_extent(fieinfo, 0, physical,
5175		- length, flags);
5176		- return (error < 0 ? error : 0);
	4895	+ iomap->addr = physical;
	4896	+ iomap->offset = 0;
	4897	+ iomap->length = length;
	4898	+ iomap->type = iomap_type;
	4899	+ iomap->flags = 0;
	4900	+out:
	4901	+ return error;
	4902	+}
	4903	+
	4904	+static int ext4_iomap_xattr_begin(struct inode *inode, loff_t offset,
	4905	+ loff_t length, unsigned flags,
	4906	+ struct iomap iomap, struct iomap srcmap)
	4907	+{
	4908	+ int error;
	4909	+
	4910	+ error = ext4_iomap_xattr_fiemap(inode, iomap);
	4911	+ if (error == 0 && (offset >= iomap->length))
	4912	+ error = -ENOENT;
	4913	+ return error;
	4914	+}
	4915	+
	4916	+static const struct iomap_ops ext4_iomap_xattr_ops = {
	4917	+ .iomap_begin = ext4_iomap_xattr_begin,
	4918	+};
	4919	+
	4920	+static int ext4_fiemap_check_ranges(struct inode inode, u64 start, u64 len)
	4921	+{
	4922	+ u64 maxbytes;
	4923	+
	4924	+ if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
	4925	+ maxbytes = inode->i_sb->s_maxbytes;
	4926	+ else
	4927	+ maxbytes = EXT4_SB(inode->i_sb)->s_bitmap_maxbytes;
	4928	+
	4929	+ if (*len == 0)
	4930	+ return -EINVAL;
	4931	+ if (start > maxbytes)
	4932	+ return -EFBIG;
	4933	+
	4934	+ /*
	4935	+ * Shrink request scope to what the fs can actually handle.
	4936	+ */
	4937	+ if (len > maxbytes \|\| (maxbytes - len) < start)
	4938	+ *len = maxbytes - start;
	4939	+ return 0;
5177	4940	}
5178	4941
5179	4942	int ext4_fiemap(struct inode inode, struct fiemap_extent_info fieinfo,
5180		- __u64 start, __u64 len)
	4943	+ u64 start, u64 len)
5181	4944	{
5182		- ext4_lblk_t start_blk;
	4945	+ int error = 0;
	4946	+
	4947	+ if (fieinfo->fi_flags & FIEMAP_FLAG_CACHE) {
	4948	+ error = ext4_ext_precache(inode);
	4949	+ if (error)
	4950	+ return error;
	4951	+ fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
	4952	+ }
	4953	+
	4954	+ /*
	4955	+ * For bitmap files the maximum size limit could be smaller than
	4956	+ * s_maxbytes, so check len here manually instead of just relying on the
	4957	+ * generic check.
	4958	+ */
	4959	+ error = ext4_fiemap_check_ranges(inode, start, &len);
	4960	+ if (error)
	4961	+ return error;
	4962	+
	4963	+ if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
	4964	+ fieinfo->fi_flags &= ~FIEMAP_FLAG_XATTR;
	4965	+ return iomap_fiemap(inode, fieinfo, start, len,
	4966	+ &ext4_iomap_xattr_ops);
	4967	+ }
	4968	+
	4969	+ return iomap_fiemap(inode, fieinfo, start, len, &ext4_iomap_report_ops);
	4970	+}
	4971	+
	4972	+int ext4_get_es_cache(struct inode inode, struct fiemap_extent_info fieinfo,
	4973	+ __u64 start, __u64 len)
	4974	+{
	4975	+ ext4_lblk_t start_blk, len_blks;
	4976	+ __u64 last_blk;
5183	4977	int error = 0;
5184	4978
5185	4979	if (ext4_has_inline_data(inode)) {
5186		- int has_inline = 1;
	4980	+ int has_inline;
5187	4981
5188		- error = ext4_inline_data_fiemap(inode, fieinfo, &has_inline,
5189		- start, len);
5190		-
	4982	+ down_read(&EXT4_I(inode)->xattr_sem);
	4983	+ has_inline = ext4_has_inline_data(inode);
	4984	+ up_read(&EXT4_I(inode)->xattr_sem);
5191	4985	if (has_inline)
5192		- return error;
	4986	+ return 0;
5193	4987	}
5194	4988
5195	4989	if (fieinfo->fi_flags & FIEMAP_FLAG_CACHE) {
5196	4990	error = ext4_ext_precache(inode);
5197	4991	if (error)
5198	4992	return error;
	4993	+ fieinfo->fi_flags &= ~FIEMAP_FLAG_CACHE;
5199	4994	}
5200	4995
5201		- /* fallback to generic here if not in extents fmt */
5202		- if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)))
5203		- return generic_block_fiemap(inode, fieinfo, start, len,
5204		- ext4_get_block);
	4996	+ error = fiemap_prep(inode, fieinfo, start, &len, 0);
	4997	+ if (error)
	4998	+ return error;
5205	4999
5206		- if (fiemap_check_flags(fieinfo, EXT4_FIEMAP_FLAGS))
5207		- return -EBADR;
	5000	+ error = ext4_fiemap_check_ranges(inode, start, &len);
	5001	+ if (error)
	5002	+ return error;
5208	5003
5209		- if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) {
5210		- error = ext4_xattr_fiemap(inode, fieinfo);
5211		- } else {
5212		- ext4_lblk_t len_blks;
5213		- __u64 last_blk;
5214		-
5215		- start_blk = start >> inode->i_sb->s_blocksize_bits;
5216		- last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
5217		- if (last_blk >= EXT_MAX_BLOCKS)
5218		- last_blk = EXT_MAX_BLOCKS-1;
5219		- len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
5220		-
5221		- /*
5222		- * Walk the extent tree gathering extent information
5223		- * and pushing extents back to the user.
5224		- */
5225		- error = ext4_fill_fiemap_extents(inode, start_blk,
5226		- len_blks, fieinfo);
5227		- }
5228		- return error;
5229		-}
5230		-
5231		-/*
5232		- * ext4_access_path:
5233		- * Function to access the path buffer for marking it dirty.
5234		- * It also checks if there are sufficient credits left in the journal handle
5235		- * to update path.
5236		- */
5237		-static int
5238		-ext4_access_path(handle_t handle, struct inode inode,
5239		- struct ext4_ext_path *path)
5240		-{
5241		- int credits, err;
5242		-
5243		- if (!ext4_handle_valid(handle))
5244		- return 0;
	5004	+ start_blk = start >> inode->i_sb->s_blocksize_bits;
	5005	+ last_blk = (start + len - 1) >> inode->i_sb->s_blocksize_bits;
	5006	+ if (last_blk >= EXT_MAX_BLOCKS)
	5007	+ last_blk = EXT_MAX_BLOCKS-1;
	5008	+ len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1;
5245	5009
5246	5010	/*
5247		- * Check if need to extend journal credits
5248		- * 3 for leaf, sb, and inode plus 2 (bmap and group
5249		- * descriptor) for each block group; assume two block
5250		- * groups
	5011	+ * Walk the extent tree gathering extent information
	5012	+ * and pushing extents back to the user.
5251	5013	*/
5252		- if (handle->h_buffer_credits < 7) {
5253		- credits = ext4_writepage_trans_blocks(inode);
5254		- err = ext4_ext_truncate_extend_restart(handle, inode, credits);
5255		- /* EAGAIN is success */
5256		- if (err && err != -EAGAIN)
5257		- return err;
5258		- }
5259		-
5260		- err = ext4_ext_get_access(handle, inode, path);
5261		- return err;
	5014	+ return ext4_fill_es_cache_info(inode, start_blk, len_blks, fieinfo);
5262	5015	}
5263	5016
5264	5017	/*
..	..	@@ -5274,7 +5027,8 @@
5274	5027	{
5275	5028	int depth, err = 0;
5276	5029	struct ext4_extent ex_start, ex_last;
5277		- bool update = 0;
	5030	+ bool update = false;
	5031	+ int credits, restart_credits;
5278	5032	depth = path->p_depth;
5279	5033
5280	5034	while (depth >= 0) {
..	..	@@ -5284,13 +5038,26 @@
5284	5038	return -EFSCORRUPTED;
5285	5039
5286	5040	ex_last = EXT_LAST_EXTENT(path[depth].p_hdr);
	5041	+ /* leaf + sb + inode */
	5042	+ credits = 3;
	5043	+ if (ex_start == EXT_FIRST_EXTENT(path[depth].p_hdr)) {
	5044	+ update = true;
	5045	+ /* extent tree + sb + inode */
	5046	+ credits = depth + 2;
	5047	+ }
5287	5048
5288		- err = ext4_access_path(handle, inode, path + depth);
	5049	+ restart_credits = ext4_writepage_trans_blocks(inode);
	5050	+ err = ext4_datasem_ensure_credits(handle, inode, credits,
	5051	+ restart_credits, 0);
	5052	+ if (err) {
	5053	+ if (err > 0)
	5054	+ err = -EAGAIN;
	5055	+ goto out;
	5056	+ }
	5057	+
	5058	+ err = ext4_ext_get_access(handle, inode, path + depth);
5289	5059	if (err)
5290	5060	goto out;
5291		-
5292		- if (ex_start == EXT_FIRST_EXTENT(path[depth].p_hdr))
5293		- update = 1;
5294	5061
5295	5062	while (ex_start <= ex_last) {
5296	5063	if (SHIFT == SHIFT_LEFT) {
..	..	@@ -5321,7 +5088,7 @@
5321	5088	}
5322	5089
5323	5090	/* Update index too */
5324		- err = ext4_access_path(handle, inode, path + depth);
	5091	+ err = ext4_ext_get_access(handle, inode, path + depth);
5325	5092	if (err)
5326	5093	goto out;
5327	5094
..	..	@@ -5360,6 +5127,7 @@
5360	5127	int ret = 0, depth;
5361	5128	struct ext4_extent *extent;
5362	5129	ext4_lblk_t stop, *iterator, ex_start, ex_end;
	5130	+ ext4_lblk_t tmp = EXT_MAX_BLOCKS;
5363	5131
5364	5132	/* Let path point to the last extent */
5365	5133	path = ext4_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL,
..	..	@@ -5413,10 +5181,15 @@
5413	5181	* till we reach stop. In case of right shift, iterator points to stop
5414	5182	* and it is decreased till we reach start.
5415	5183	*/
	5184	+again:
	5185	+ ret = 0;
5416	5186	if (SHIFT == SHIFT_LEFT)
5417	5187	iterator = &start;
5418	5188	else
5419	5189	iterator = &stop;
	5190	+
	5191	+ if (tmp != EXT_MAX_BLOCKS)
	5192	+ *iterator = tmp;
5420	5193
5421	5194	/*
5422	5195	* Its safe to start updating extents. Start and stop are unsigned, so
..	..	@@ -5446,24 +5219,35 @@
5446	5219	}
5447	5220	}
5448	5221
	5222	+ tmp = *iterator;
5449	5223	if (SHIFT == SHIFT_LEFT) {
5450	5224	extent = EXT_LAST_EXTENT(path[depth].p_hdr);
5451	5225	*iterator = le32_to_cpu(extent->ee_block) +
5452	5226	ext4_ext_get_actual_len(extent);
5453	5227	} else {
5454	5228	extent = EXT_FIRST_EXTENT(path[depth].p_hdr);
5455		- if (le32_to_cpu(extent->ee_block) > 0)
	5229	+ if (le32_to_cpu(extent->ee_block) > start)
5456	5230	*iterator = le32_to_cpu(extent->ee_block) - 1;
5457		- else
5458		- /* Beginning is reached, end of the loop */
	5231	+ else if (le32_to_cpu(extent->ee_block) == start)
5459	5232	iterator = NULL;
5460		- /* Update path extent in case we need to stop */
5461		- while (le32_to_cpu(extent->ee_block) < start)
	5233	+ else {
	5234	+ extent = EXT_LAST_EXTENT(path[depth].p_hdr);
	5235	+ while (le32_to_cpu(extent->ee_block) >= start)
	5236	+ extent--;
	5237	+
	5238	+ if (extent == EXT_LAST_EXTENT(path[depth].p_hdr))
	5239	+ break;
	5240	+
5462	5241	extent++;
	5242	+ iterator = NULL;
	5243	+ }
5463	5244	path[depth].p_ext = extent;
5464	5245	}
5465	5246	ret = ext4_ext_shift_path_extents(path, shift, inode,
5466	5247	handle, SHIFT);
	5248	+ /* iterator can be NULL which means we should break */
	5249	+ if (ret == -EAGAIN)
	5250	+ goto again;
5467	5251	if (ret)
5468	5252	break;
5469	5253	}
..	..	@@ -5478,8 +5262,9 @@
5478	5262	* This implements the fallocate's collapse range functionality for ext4
5479	5263	* Returns: 0 and non-zero on error.
5480	5264	*/
5481		-int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len)
	5265	+static int ext4_collapse_range(struct file *file, loff_t offset, loff_t len)
5482	5266	{
	5267	+ struct inode *inode = file_inode(file);
5483	5268	struct super_block *sb = inode->i_sb;
5484	5269	ext4_lblk_t punch_start, punch_stop;
5485	5270	handle_t *handle;
..	..	@@ -5495,12 +5280,8 @@
5495	5280	if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
5496	5281	return -EOPNOTSUPP;
5497	5282
5498		- /* Collapse range works only on fs block size aligned offsets. */
5499		- if (offset & (EXT4_CLUSTER_SIZE(sb) - 1) \|\|
5500		- len & (EXT4_CLUSTER_SIZE(sb) - 1))
5501		- return -EINVAL;
5502		-
5503		- if (!S_ISREG(inode->i_mode))
	5283	+ /* Collapse range works only on fs cluster size aligned regions. */
	5284	+ if (!IS_ALIGNED(offset \| len, EXT4_CLUSTER_SIZE(sb)))
5504	5285	return -EINVAL;
5505	5286
5506	5287	trace_ext4_collapse_range(inode, offset, len);
..	..	@@ -5520,7 +5301,7 @@
5520	5301	* There is no need to overlap collapse range with EOF, in which case
5521	5302	* it is effectively a truncate operation
5522	5303	*/
5523		- if (offset + len >= i_size_read(inode)) {
	5304	+ if (offset + len >= inode->i_size) {
5524	5305	ret = -EINVAL;
5525	5306	goto out_mutex;
5526	5307	}
..	..	@@ -5533,6 +5314,10 @@
5533	5314
5534	5315	/* Wait for existing dio to complete */
5535	5316	inode_dio_wait(inode);
	5317	+
	5318	+ ret = file_modified(file);
	5319	+ if (ret)
	5320	+ goto out_mutex;
5536	5321
5537	5322	/*
5538	5323	* Prevent page faults from reinstantiating pages we have released from
..	..	@@ -5573,9 +5358,10 @@
5573	5358	ret = PTR_ERR(handle);
5574	5359	goto out_mmap;
5575	5360	}
	5361	+ ext4_fc_start_ineligible(sb, EXT4_FC_REASON_FALLOC_RANGE);
5576	5362
5577	5363	down_write(&EXT4_I(inode)->i_data_sem);
5578		- ext4_discard_preallocations(inode);
	5364	+ ext4_discard_preallocations(inode, 0);
5579	5365
5580	5366	ret = ext4_es_remove_extent(inode, punch_start,
5581	5367	EXT_MAX_BLOCKS - punch_start);
..	..	@@ -5589,7 +5375,7 @@
5589	5375	up_write(&EXT4_I(inode)->i_data_sem);
5590	5376	goto out_stop;
5591	5377	}
5592		- ext4_discard_preallocations(inode);
	5378	+ ext4_discard_preallocations(inode, 0);
5593	5379
5594	5380	ret = ext4_ext_shift_extents(inode, handle, punch_stop,
5595	5381	punch_stop - punch_start, SHIFT_LEFT);
..	..	@@ -5598,7 +5384,7 @@
5598	5384	goto out_stop;
5599	5385	}
5600	5386
5601		- new_size = i_size_read(inode) - len;
	5387	+ new_size = inode->i_size - len;
5602	5388	i_size_write(inode, new_size);
5603	5389	EXT4_I(inode)->i_disksize = new_size;
5604	5390
..	..	@@ -5606,11 +5392,12 @@
5606	5392	if (IS_SYNC(inode))
5607	5393	ext4_handle_sync(handle);
5608	5394	inode->i_mtime = inode->i_ctime = current_time(inode);
5609		- ext4_mark_inode_dirty(handle, inode);
	5395	+ ret = ext4_mark_inode_dirty(handle, inode);
5610	5396	ext4_update_inode_fsync_trans(handle, inode, 1);
5611	5397
5612	5398	out_stop:
5613	5399	ext4_journal_stop(handle);
	5400	+ ext4_fc_stop_ineligible(sb);
5614	5401	out_mmap:
5615	5402	up_write(&EXT4_I(inode)->i_mmap_sem);
5616	5403	out_mutex:
..	..	@@ -5626,8 +5413,9 @@
5626	5413	* by len bytes.
5627	5414	* Returns 0 on success, error otherwise.
5628	5415	*/
5629		-int ext4_insert_range(struct inode *inode, loff_t offset, loff_t len)
	5416	+static int ext4_insert_range(struct file *file, loff_t offset, loff_t len)
5630	5417	{
	5418	+ struct inode *inode = file_inode(file);
5631	5419	struct super_block *sb = inode->i_sb;
5632	5420	handle_t *handle;
5633	5421	struct ext4_ext_path *path;
..	..	@@ -5645,13 +5433,9 @@
5645	5433	if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))
5646	5434	return -EOPNOTSUPP;
5647	5435
5648		- /* Insert range works only on fs block size aligned offsets. */
5649		- if (offset & (EXT4_CLUSTER_SIZE(sb) - 1) \|\|
5650		- len & (EXT4_CLUSTER_SIZE(sb) - 1))
	5436	+ /* Insert range works only on fs cluster size aligned regions. */
	5437	+ if (!IS_ALIGNED(offset \| len, EXT4_CLUSTER_SIZE(sb)))
5651	5438	return -EINVAL;
5652		-
5653		- if (!S_ISREG(inode->i_mode))
5654		- return -EOPNOTSUPP;
5655	5439
5656	5440	trace_ext4_insert_range(inode, offset, len);
5657	5441
..	..	@@ -5672,20 +5456,24 @@
5672	5456	goto out_mutex;
5673	5457	}
5674	5458
5675		- /* Check for wrap through zero */
5676		- if (inode->i_size + len > inode->i_sb->s_maxbytes) {
	5459	+ /* Check whether the maximum file size would be exceeded */
	5460	+ if (len > inode->i_sb->s_maxbytes - inode->i_size) {
5677	5461	ret = -EFBIG;
5678	5462	goto out_mutex;
5679	5463	}
5680	5464
5681		- /* Offset should be less than i_size */
5682		- if (offset >= i_size_read(inode)) {
	5465	+ /* Offset must be less than i_size */
	5466	+ if (offset >= inode->i_size) {
5683	5467	ret = -EINVAL;
5684	5468	goto out_mutex;
5685	5469	}
5686	5470
5687	5471	/* Wait for existing dio to complete */
5688	5472	inode_dio_wait(inode);
	5473	+
	5474	+ ret = file_modified(file);
	5475	+ if (ret)
	5476	+ goto out_mutex;
5689	5477
5690	5478	/*
5691	5479	* Prevent page faults from reinstantiating pages we have released from
..	..	@@ -5715,6 +5503,7 @@
5715	5503	ret = PTR_ERR(handle);
5716	5504	goto out_mmap;
5717	5505	}
	5506	+ ext4_fc_start_ineligible(sb, EXT4_FC_REASON_FALLOC_RANGE);
5718	5507
5719	5508	/* Expand file to avoid data loss if there is error while shifting */
5720	5509	inode->i_size += len;
..	..	@@ -5725,7 +5514,7 @@
5725	5514	goto out_stop;
5726	5515
5727	5516	down_write(&EXT4_I(inode)->i_data_sem);
5728		- ext4_discard_preallocations(inode);
	5517	+ ext4_discard_preallocations(inode, 0);
5729	5518
5730	5519	path = ext4_find_extent(inode, offset_lblk, NULL, 0);
5731	5520	if (IS_ERR(path)) {
..	..	@@ -5789,6 +5578,7 @@
5789	5578
5790	5579	out_stop:
5791	5580	ext4_journal_stop(handle);
	5581	+ ext4_fc_stop_ineligible(sb);
5792	5582	out_mmap:
5793	5583	up_write(&EXT4_I(inode)->i_mmap_sem);
5794	5584	out_mutex:
..	..	@@ -5797,8 +5587,8 @@
5797	5587	}
5798	5588
5799	5589	/**
5800		- * ext4_swap_extents - Swap extents between two inodes
5801		- *
	5590	+ * ext4_swap_extents() - Swap extents between two inodes
	5591	+ * @handle: handle for this transaction
5802	5592	* @inode1: First inode
5803	5593	* @inode2: Second inode
5804	5594	* @lblk1: Start block for first inode
..	..	@@ -5859,7 +5649,7 @@
5859	5649	}
5860	5650	ex1 = path1[path1->p_depth].p_ext;
5861	5651	ex2 = path2[path2->p_depth].p_ext;
5862		- /* Do we have somthing to swap ? */
	5652	+ /* Do we have something to swap ? */
5863	5653	if (unlikely(!ex2 \|\| !ex1))
5864	5654	goto finish;
5865	5655
..	..	@@ -5991,3 +5781,356 @@
5991	5781	}
5992	5782	return replaced_count;
5993	5783	}
	5784	+
	5785	+/*
	5786	+ * ext4_clu_mapped - determine whether any block in a logical cluster has
	5787	+ * been mapped to a physical cluster
	5788	+ *
	5789	+ * @inode - file containing the logical cluster
	5790	+ * @lclu - logical cluster of interest
	5791	+ *
	5792	+ * Returns 1 if any block in the logical cluster is mapped, signifying
	5793	+ * that a physical cluster has been allocated for it. Otherwise,
	5794	+ * returns 0. Can also return negative error codes. Derived from
	5795	+ * ext4_ext_map_blocks().
	5796	+ */
	5797	+int ext4_clu_mapped(struct inode *inode, ext4_lblk_t lclu)
	5798	+{
	5799	+ struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
	5800	+ struct ext4_ext_path *path;
	5801	+ int depth, mapped = 0, err = 0;
	5802	+ struct ext4_extent *extent;
	5803	+ ext4_lblk_t first_lblk, first_lclu, last_lclu;
	5804	+
	5805	+ /* search for the extent closest to the first block in the cluster */
	5806	+ path = ext4_find_extent(inode, EXT4_C2B(sbi, lclu), NULL, 0);
	5807	+ if (IS_ERR(path)) {
	5808	+ err = PTR_ERR(path);
	5809	+ path = NULL;
	5810	+ goto out;
	5811	+ }
	5812	+
	5813	+ depth = ext_depth(inode);
	5814	+
	5815	+ /*
	5816	+ * A consistent leaf must not be empty. This situation is possible,
	5817	+ * though, _during_ tree modification, and it's why an assert can't
	5818	+ * be put in ext4_find_extent().
	5819	+ */
	5820	+ if (unlikely(path[depth].p_ext == NULL && depth != 0)) {
	5821	+ EXT4_ERROR_INODE(inode,
	5822	+ "bad extent address - lblock: %lu, depth: %d, pblock: %lld",
	5823	+ (unsigned long) EXT4_C2B(sbi, lclu),
	5824	+ depth, path[depth].p_block);
	5825	+ err = -EFSCORRUPTED;
	5826	+ goto out;
	5827	+ }
	5828	+
	5829	+ extent = path[depth].p_ext;
	5830	+
	5831	+ /* can't be mapped if the extent tree is empty */
	5832	+ if (extent == NULL)
	5833	+ goto out;
	5834	+
	5835	+ first_lblk = le32_to_cpu(extent->ee_block);
	5836	+ first_lclu = EXT4_B2C(sbi, first_lblk);
	5837	+
	5838	+ /*
	5839	+ * Three possible outcomes at this point - found extent spanning
	5840	+ * the target cluster, to the left of the target cluster, or to the
	5841	+ * right of the target cluster. The first two cases are handled here.
	5842	+ * The last case indicates the target cluster is not mapped.
	5843	+ */
	5844	+ if (lclu >= first_lclu) {
	5845	+ last_lclu = EXT4_B2C(sbi, first_lblk +
	5846	+ ext4_ext_get_actual_len(extent) - 1);
	5847	+ if (lclu <= last_lclu) {
	5848	+ mapped = 1;
	5849	+ } else {
	5850	+ first_lblk = ext4_ext_next_allocated_block(path);
	5851	+ first_lclu = EXT4_B2C(sbi, first_lblk);
	5852	+ if (lclu == first_lclu)
	5853	+ mapped = 1;
	5854	+ }
	5855	+ }
	5856	+
	5857	+out:
	5858	+ ext4_ext_drop_refs(path);
	5859	+ kfree(path);
	5860	+
	5861	+ return err ? err : mapped;
	5862	+}
	5863	+
	5864	+/*
	5865	+ * Updates physical block address and unwritten status of extent
	5866	+ * starting at lblk start and of len. If such an extent doesn't exist,
	5867	+ * this function splits the extent tree appropriately to create an
	5868	+ * extent like this. This function is called in the fast commit
	5869	+ * replay path. Returns 0 on success and error on failure.
	5870	+ */
	5871	+int ext4_ext_replay_update_ex(struct inode *inode, ext4_lblk_t start,
	5872	+ int len, int unwritten, ext4_fsblk_t pblk)
	5873	+{
	5874	+ struct ext4_ext_path path = NULL, ppath;
	5875	+ struct ext4_extent *ex;
	5876	+ int ret;
	5877	+
	5878	+ path = ext4_find_extent(inode, start, NULL, 0);
	5879	+ if (IS_ERR(path))
	5880	+ return PTR_ERR(path);
	5881	+ ex = path[path->p_depth].p_ext;
	5882	+ if (!ex) {
	5883	+ ret = -EFSCORRUPTED;
	5884	+ goto out;
	5885	+ }
	5886	+
	5887	+ if (le32_to_cpu(ex->ee_block) != start \|\|
	5888	+ ext4_ext_get_actual_len(ex) != len) {
	5889	+ /* We need to split this extent to match our extent first */
	5890	+ ppath = path;
	5891	+ down_write(&EXT4_I(inode)->i_data_sem);
	5892	+ ret = ext4_force_split_extent_at(NULL, inode, &ppath, start, 1);
	5893	+ up_write(&EXT4_I(inode)->i_data_sem);
	5894	+ if (ret)
	5895	+ goto out;
	5896	+ kfree(path);
	5897	+ path = ext4_find_extent(inode, start, NULL, 0);
	5898	+ if (IS_ERR(path))
	5899	+ return -1;
	5900	+ ppath = path;
	5901	+ ex = path[path->p_depth].p_ext;
	5902	+ WARN_ON(le32_to_cpu(ex->ee_block) != start);
	5903	+ if (ext4_ext_get_actual_len(ex) != len) {
	5904	+ down_write(&EXT4_I(inode)->i_data_sem);
	5905	+ ret = ext4_force_split_extent_at(NULL, inode, &ppath,
	5906	+ start + len, 1);
	5907	+ up_write(&EXT4_I(inode)->i_data_sem);
	5908	+ if (ret)
	5909	+ goto out;
	5910	+ kfree(path);
	5911	+ path = ext4_find_extent(inode, start, NULL, 0);
	5912	+ if (IS_ERR(path))
	5913	+ return -EINVAL;
	5914	+ ex = path[path->p_depth].p_ext;
	5915	+ }
	5916	+ }
	5917	+ if (unwritten)
	5918	+ ext4_ext_mark_unwritten(ex);
	5919	+ else
	5920	+ ext4_ext_mark_initialized(ex);
	5921	+ ext4_ext_store_pblock(ex, pblk);
	5922	+ down_write(&EXT4_I(inode)->i_data_sem);
	5923	+ ret = ext4_ext_dirty(NULL, inode, &path[path->p_depth]);
	5924	+ up_write(&EXT4_I(inode)->i_data_sem);
	5925	+out:
	5926	+ ext4_ext_drop_refs(path);
	5927	+ kfree(path);
	5928	+ ext4_mark_inode_dirty(NULL, inode);
	5929	+ return ret;
	5930	+}
	5931	+
	5932	+/* Try to shrink the extent tree */
	5933	+void ext4_ext_replay_shrink_inode(struct inode *inode, ext4_lblk_t end)
	5934	+{
	5935	+ struct ext4_ext_path *path = NULL;
	5936	+ struct ext4_extent *ex;
	5937	+ ext4_lblk_t old_cur, cur = 0;
	5938	+
	5939	+ while (cur < end) {
	5940	+ path = ext4_find_extent(inode, cur, NULL, 0);
	5941	+ if (IS_ERR(path))
	5942	+ return;
	5943	+ ex = path[path->p_depth].p_ext;
	5944	+ if (!ex) {
	5945	+ ext4_ext_drop_refs(path);
	5946	+ kfree(path);
	5947	+ ext4_mark_inode_dirty(NULL, inode);
	5948	+ return;
	5949	+ }
	5950	+ old_cur = cur;
	5951	+ cur = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex);
	5952	+ if (cur <= old_cur)
	5953	+ cur = old_cur + 1;
	5954	+ ext4_ext_try_to_merge(NULL, inode, path, ex);
	5955	+ down_write(&EXT4_I(inode)->i_data_sem);
	5956	+ ext4_ext_dirty(NULL, inode, &path[path->p_depth]);
	5957	+ up_write(&EXT4_I(inode)->i_data_sem);
	5958	+ ext4_mark_inode_dirty(NULL, inode);
	5959	+ ext4_ext_drop_refs(path);
	5960	+ kfree(path);
	5961	+ }
	5962	+}
	5963	+
	5964	+/* Check if cur is a hole and if it is, skip it /
	5965	+static int skip_hole(struct inode inode, ext4_lblk_t cur)
	5966	+{
	5967	+ int ret;
	5968	+ struct ext4_map_blocks map;
	5969	+
	5970	+ map.m_lblk = *cur;
	5971	+ map.m_len = ((inode->i_size) >> inode->i_sb->s_blocksize_bits) - *cur;
	5972	+
	5973	+ ret = ext4_map_blocks(NULL, inode, &map, 0);
	5974	+ if (ret < 0)
	5975	+ return ret;
	5976	+ if (ret != 0)
	5977	+ return 0;
	5978	+ cur = cur + map.m_len;
	5979	+ return 0;
	5980	+}
	5981	+
	5982	+/* Count number of blocks used by this inode and update i_blocks */
	5983	+int ext4_ext_replay_set_iblocks(struct inode *inode)
	5984	+{
	5985	+ struct ext4_ext_path path = NULL, path2 = NULL;
	5986	+ struct ext4_extent *ex;
	5987	+ ext4_lblk_t cur = 0, end;
	5988	+ int numblks = 0, i, ret = 0;
	5989	+ ext4_fsblk_t cmp1, cmp2;
	5990	+ struct ext4_map_blocks map;
	5991	+
	5992	+ /* Determin the size of the file first */
	5993	+ path = ext4_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL,
	5994	+ EXT4_EX_NOCACHE);
	5995	+ if (IS_ERR(path))
	5996	+ return PTR_ERR(path);
	5997	+ ex = path[path->p_depth].p_ext;
	5998	+ if (!ex) {
	5999	+ ext4_ext_drop_refs(path);
	6000	+ kfree(path);
	6001	+ goto out;
	6002	+ }
	6003	+ end = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex);
	6004	+ ext4_ext_drop_refs(path);
	6005	+ kfree(path);
	6006	+
	6007	+ /* Count the number of data blocks */
	6008	+ cur = 0;
	6009	+ while (cur < end) {
	6010	+ map.m_lblk = cur;
	6011	+ map.m_len = end - cur;
	6012	+ ret = ext4_map_blocks(NULL, inode, &map, 0);
	6013	+ if (ret < 0)
	6014	+ break;
	6015	+ if (ret > 0)
	6016	+ numblks += ret;
	6017	+ cur = cur + map.m_len;
	6018	+ }
	6019	+
	6020	+ /*
	6021	+ * Count the number of extent tree blocks. We do it by looking up
	6022	+ * two successive extents and determining the difference between
	6023	+ * their paths. When path is different for 2 successive extents
	6024	+ * we compare the blocks in the path at each level and increment
	6025	+ * iblocks by total number of differences found.
	6026	+ */
	6027	+ cur = 0;
	6028	+ ret = skip_hole(inode, &cur);
	6029	+ if (ret < 0)
	6030	+ goto out;
	6031	+ path = ext4_find_extent(inode, cur, NULL, 0);
	6032	+ if (IS_ERR(path))
	6033	+ goto out;
	6034	+ numblks += path->p_depth;
	6035	+ ext4_ext_drop_refs(path);
	6036	+ kfree(path);
	6037	+ while (cur < end) {
	6038	+ path = ext4_find_extent(inode, cur, NULL, 0);
	6039	+ if (IS_ERR(path))
	6040	+ break;
	6041	+ ex = path[path->p_depth].p_ext;
	6042	+ if (!ex) {
	6043	+ ext4_ext_drop_refs(path);
	6044	+ kfree(path);
	6045	+ return 0;
	6046	+ }
	6047	+ cur = max(cur + 1, le32_to_cpu(ex->ee_block) +
	6048	+ ext4_ext_get_actual_len(ex));
	6049	+ ret = skip_hole(inode, &cur);
	6050	+ if (ret < 0) {
	6051	+ ext4_ext_drop_refs(path);
	6052	+ kfree(path);
	6053	+ break;
	6054	+ }
	6055	+ path2 = ext4_find_extent(inode, cur, NULL, 0);
	6056	+ if (IS_ERR(path2)) {
	6057	+ ext4_ext_drop_refs(path);
	6058	+ kfree(path);
	6059	+ break;
	6060	+ }
	6061	+ ex = path2[path2->p_depth].p_ext;
	6062	+ for (i = 0; i <= max(path->p_depth, path2->p_depth); i++) {
	6063	+ cmp1 = cmp2 = 0;
	6064	+ if (i <= path->p_depth)
	6065	+ cmp1 = path[i].p_bh ?
	6066	+ path[i].p_bh->b_blocknr : 0;
	6067	+ if (i <= path2->p_depth)
	6068	+ cmp2 = path2[i].p_bh ?
	6069	+ path2[i].p_bh->b_blocknr : 0;
	6070	+ if (cmp1 != cmp2 && cmp2 != 0)
	6071	+ numblks++;
	6072	+ }
	6073	+ ext4_ext_drop_refs(path);
	6074	+ ext4_ext_drop_refs(path2);
	6075	+ kfree(path);
	6076	+ kfree(path2);
	6077	+ }
	6078	+
	6079	+out:
	6080	+ inode->i_blocks = numblks << (inode->i_sb->s_blocksize_bits - 9);
	6081	+ ext4_mark_inode_dirty(NULL, inode);
	6082	+ return 0;
	6083	+}
	6084	+
	6085	+int ext4_ext_clear_bb(struct inode *inode)
	6086	+{
	6087	+ struct ext4_ext_path *path = NULL;
	6088	+ struct ext4_extent *ex;
	6089	+ ext4_lblk_t cur = 0, end;
	6090	+ int j, ret = 0;
	6091	+ struct ext4_map_blocks map;
	6092	+
	6093	+ /* Determin the size of the file first */
	6094	+ path = ext4_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL,
	6095	+ EXT4_EX_NOCACHE);
	6096	+ if (IS_ERR(path))
	6097	+ return PTR_ERR(path);
	6098	+ ex = path[path->p_depth].p_ext;
	6099	+ if (!ex) {
	6100	+ ext4_ext_drop_refs(path);
	6101	+ kfree(path);
	6102	+ return 0;
	6103	+ }
	6104	+ end = le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex);
	6105	+ ext4_ext_drop_refs(path);
	6106	+ kfree(path);
	6107	+
	6108	+ cur = 0;
	6109	+ while (cur < end) {
	6110	+ map.m_lblk = cur;
	6111	+ map.m_len = end - cur;
	6112	+ ret = ext4_map_blocks(NULL, inode, &map, 0);
	6113	+ if (ret < 0)
	6114	+ break;
	6115	+ if (ret > 0) {
	6116	+ path = ext4_find_extent(inode, map.m_lblk, NULL, 0);
	6117	+ if (!IS_ERR_OR_NULL(path)) {
	6118	+ for (j = 0; j < path->p_depth; j++) {
	6119	+
	6120	+ ext4_mb_mark_bb(inode->i_sb,
	6121	+ path[j].p_block, 1, 0);
	6122	+ ext4_fc_record_regions(inode->i_sb, inode->i_ino,
	6123	+ 0, path[j].p_block, 1, 1);
	6124	+ }
	6125	+ ext4_ext_drop_refs(path);
	6126	+ kfree(path);
	6127	+ }
	6128	+ ext4_mb_mark_bb(inode->i_sb, map.m_pblk, map.m_len, 0);
	6129	+ ext4_fc_record_regions(inode->i_sb, inode->i_ino,
	6130	+ map.m_lblk, map.m_pblk, map.m_len, 1);
	6131	+ }
	6132	+ cur = cur + map.m_len;
	6133	+ }
	6134	+
	6135	+ return 0;
	6136	+}