[linux.git] / fs / xfs / xfs_discard.c

/*
 * Copyright (C) 2010 Red Hat, Inc.
 * All Rights Reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it would be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#include "xfs.h"
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
#include "xfs_sb.h"
#include "xfs_mount.h"
#include "xfs_quota.h"
#include "xfs_inode.h"
#include "xfs_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_alloc.h"
#include "xfs_error.h"
#include "xfs_extent_busy.h"
#include "xfs_discard.h"
#include "xfs_trace.h"
#include "xfs_log.h"

STATIC int
xfs_trim_extents(
	struct xfs_mount	*mp,
	xfs_agnumber_t		agno,
	xfs_daddr_t		start,
	xfs_daddr_t		end,
	xfs_daddr_t		minlen,
	__uint64_t		*blocks_trimmed)
{
	struct block_device	*bdev = mp->m_ddev_targp->bt_bdev;
	struct xfs_btree_cur	*cur;
	struct xfs_buf		*agbp;
	struct xfs_perag	*pag;
	int			error;
	int			i;

	pag = xfs_perag_get(mp, agno);

	error = xfs_alloc_read_agf(mp, NULL, agno, 0, &agbp);
	if (error || !agbp)
		goto out_put_perag;

	cur = xfs_allocbt_init_cursor(mp, NULL, agbp, agno, XFS_BTNUM_CNT);

	/*
	 * Force out the log.  This means any transactions that might have freed
	 * space before we took the AGF buffer lock are now on disk, and the
	 * volatile disk cache is flushed.
	 */
	xfs_log_force(mp, XFS_LOG_SYNC);

	/*
	 * Look up the longest btree in the AGF and start with it.
	 */
	error = xfs_alloc_lookup_ge(cur, 0,
			    be32_to_cpu(XFS_BUF_TO_AGF(agbp)->agf_longest), &i);
	if (error)
		goto out_del_cursor;

	/*
	 * Loop until we are done with all extents that are large
	 * enough to be worth discarding.
	 */
	while (i) {
		xfs_agblock_t	fbno;
		xfs_extlen_t	flen;
		xfs_daddr_t	dbno;
		xfs_extlen_t	dlen;

		error = xfs_alloc_get_rec(cur, &fbno, &flen, &i);
		if (error)
			goto out_del_cursor;
		XFS_WANT_CORRUPTED_GOTO(mp, i == 1, out_del_cursor);
		ASSERT(flen <= be32_to_cpu(XFS_BUF_TO_AGF(agbp)->agf_longest));

		/*
		 * use daddr format for all range/len calculations as that is
		 * the format the range/len variables are supplied in by
		 * userspace.
		 */
		dbno = XFS_AGB_TO_DADDR(mp, agno, fbno);
		dlen = XFS_FSB_TO_BB(mp, flen);

		/*
		 * Too small?  Give up.
		 */
		if (dlen < minlen) {
			trace_xfs_discard_toosmall(mp, agno, fbno, flen);
			goto out_del_cursor;
		}

		/*
		 * If the extent is entirely outside of the range we are
		 * supposed to discard skip it.  Do not bother to trim
		 * down partially overlapping ranges for now.
		 */
		if (dbno + dlen < start || dbno > end) {
			trace_xfs_discard_exclude(mp, agno, fbno, flen);
			goto next_extent;
		}

		/*
		 * If any blocks in the range are still busy, skip the
		 * discard and try again the next time.
		 */
		if (xfs_extent_busy_search(mp, agno, fbno, flen)) {
			trace_xfs_discard_busy(mp, agno, fbno, flen);
			goto next_extent;
		}

		trace_xfs_discard_extent(mp, agno, fbno, flen);
		error = blkdev_issue_discard(bdev, dbno, dlen, GFP_NOFS, 0);
		if (error)
			goto out_del_cursor;
		*blocks_trimmed += flen;

next_extent:
		error = xfs_btree_decrement(cur, 0, &i);
		if (error)
			goto out_del_cursor;
	}

out_del_cursor:
	xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
	xfs_buf_relse(agbp);
out_put_perag:
	xfs_perag_put(pag);
	return error;
}

/*
 * trim a range of the filesystem.
 *
 * Note: the parameters passed from userspace are byte ranges into the
 * filesystem which does not match to the format we use for filesystem block
 * addressing. FSB addressing is sparse (AGNO|AGBNO), while the incoming format
 * is a linear address range. Hence we need to use DADDR based conversions and
 * comparisons for determining the correct offset and regions to trim.
 */
int
xfs_ioc_trim(
	struct xfs_mount		*mp,
	struct fstrim_range __user	*urange)
{
	struct request_queue	*q = bdev_get_queue(mp->m_ddev_targp->bt_bdev);
	unsigned int		granularity = q->limits.discard_granularity;
	struct fstrim_range	range;
	xfs_daddr_t		start, end, minlen;
	xfs_agnumber_t		start_agno, end_agno, agno;
	__uint64_t		blocks_trimmed = 0;
	int			error, last_error = 0;

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;
	if (!blk_queue_discard(q))
		return -EOPNOTSUPP;
	if (copy_from_user(&range, urange, sizeof(range)))
		return -EFAULT;

	/*
	 * Truncating down the len isn't actually quite correct, but using
	 * BBTOB would mean we trivially get overflows for values
	 * of ULLONG_MAX or slightly lower.  And ULLONG_MAX is the default
	 * used by the fstrim application.  In the end it really doesn't
	 * matter as trimming blocks is an advisory interface.
	 */
	if (range.start >= XFS_FSB_TO_B(mp, mp->m_sb.sb_dblocks) ||
	    range.minlen > XFS_FSB_TO_B(mp, XFS_ALLOC_AG_MAX_USABLE(mp)) ||
	    range.len < mp->m_sb.sb_blocksize)
		return -EINVAL;

	start = BTOBB(range.start);
	end = start + BTOBBT(range.len) - 1;
	minlen = BTOBB(max_t(u64, granularity, range.minlen));

	if (end > XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks) - 1)
		end = XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)- 1;

	start_agno = xfs_daddr_to_agno(mp, start);
	end_agno = xfs_daddr_to_agno(mp, end);

	for (agno = start_agno; agno <= end_agno; agno++) {
		error = xfs_trim_extents(mp, agno, start, end, minlen,
					  &blocks_trimmed);
		if (error)
			last_error = error;
	}

	if (last_error)
		return last_error;

	range.len = XFS_FSB_TO_B(mp, blocks_trimmed);
	if (copy_to_user(urange, &range, sizeof(range)))
		return -EFAULT;
	return 0;
}

int
xfs_discard_extents(
	struct xfs_mount	*mp,
	struct list_head	*list)
{
	struct xfs_extent_busy	*busyp;
	int			error = 0;

	list_for_each_entry(busyp, list, list) {
		trace_xfs_discard_extent(mp, busyp->agno, busyp->bno,
					 busyp->length);

		error = blkdev_issue_discard(mp->m_ddev_targp->bt_bdev,
				XFS_AGB_TO_DADDR(mp, busyp->agno, busyp->bno),
				XFS_FSB_TO_BB(mp, busyp->length),
				GFP_NOFS, 0);
		if (error && error != -EOPNOTSUPP) {
			xfs_info(mp,
	 "discard failed for extent [0x%llu,%u], error %d",
				 (unsigned long long)busyp->bno,
				 busyp->length,
				 error);
			return error;
		}
	}

	return 0;
}
Commit	Line	Data
a46db608 CH	1	/*
	2	* Copyright (C) 2010 Red Hat, Inc.
	3	* All Rights Reserved.
	4	*
	5	* This program is free software; you can redistribute it and/or
	6	* modify it under the terms of the GNU General Public License as
	7	* published by the Free Software Foundation.
	8	*
	9	* This program is distributed in the hope that it would be useful,
	10	* but WITHOUT ANY WARRANTY; without even the implied warranty of
	11	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
	12	* GNU General Public License for more details.
	13	*
	14	* You should have received a copy of the GNU General Public License
	15	* along with this program; if not, write the Free Software Foundation,
	16	* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
	17	*/
	18	#include "xfs.h"
6ca1c906	19	#include "xfs_format.h"
239880ef DC	20	#include "xfs_log_format.h"
239880ef DC	21	#include "xfs_trans_resv.h"
7fd36c44	22	#include "xfs_sb.h"
a46db608 CH	23	#include "xfs_mount.h"
a46db608 CH	24	#include "xfs_quota.h"
a46db608	25	#include "xfs_inode.h"
a4fbe6ab DC	26	#include "xfs_btree.h"
a4fbe6ab DC	27	#include "xfs_alloc_btree.h"
a46db608 CH	28	#include "xfs_alloc.h"
a46db608 CH	29	#include "xfs_error.h"
efc27b52	30	#include "xfs_extent_busy.h"
a46db608 CH	31	#include "xfs_discard.h"
a46db608 CH	32	#include "xfs_trace.h"
239880ef	33	#include "xfs_log.h"
a46db608 CH	34
	35	STATIC int
	36	xfs_trim_extents(
	37	struct xfs_mount *mp,
	38	xfs_agnumber_t agno,
a66d6363 DC	39	xfs_daddr_t start,
	40	xfs_daddr_t end,
	41	xfs_daddr_t minlen,
a46db608 CH	42	__uint64_t *blocks_trimmed)
	43	{
	44	struct block_device *bdev = mp->m_ddev_targp->bt_bdev;
	45	struct xfs_btree_cur *cur;
	46	struct xfs_buf *agbp;
	47	struct xfs_perag *pag;
	48	int error;
	49	int i;
	50
	51	pag = xfs_perag_get(mp, agno);
	52
	53	error = xfs_alloc_read_agf(mp, NULL, agno, 0, &agbp);
	54	if (error \|\| !agbp)
	55	goto out_put_perag;
	56
	57	cur = xfs_allocbt_init_cursor(mp, NULL, agbp, agno, XFS_BTNUM_CNT);
	58
	59	/*
	60	* Force out the log. This means any transactions that might have freed
	61	* space before we took the AGF buffer lock are now on disk, and the
	62	* volatile disk cache is flushed.
	63	*/
	64	xfs_log_force(mp, XFS_LOG_SYNC);
	65
	66	/*
	67	* Look up the longest btree in the AGF and start with it.
	68	*/
a66d6363	69	error = xfs_alloc_lookup_ge(cur, 0,
b1c770c2	70	be32_to_cpu(XFS_BUF_TO_AGF(agbp)->agf_longest), &i);
a46db608 CH	71	if (error)
	72	goto out_del_cursor;
	73
	74	/*
	75	* Loop until we are done with all extents that are large
	76	* enough to be worth discarding.
	77	*/
	78	while (i) {
a66d6363 DC	79	xfs_agblock_t fbno;
	80	xfs_extlen_t flen;
	81	xfs_daddr_t dbno;
	82	xfs_extlen_t dlen;
a46db608 CH	83
	84	error = xfs_alloc_get_rec(cur, &fbno, &flen, &i);
	85	if (error)
	86	goto out_del_cursor;
c29aad41	87	XFS_WANT_CORRUPTED_GOTO(mp, i == 1, out_del_cursor);
b1c770c2	88	ASSERT(flen <= be32_to_cpu(XFS_BUF_TO_AGF(agbp)->agf_longest));
a46db608	89
a66d6363 DC	90	/*
	91	* use daddr format for all range/len calculations as that is
	92	* the format the range/len variables are supplied in by
	93	* userspace.
	94	*/
	95	dbno = XFS_AGB_TO_DADDR(mp, agno, fbno);
	96	dlen = XFS_FSB_TO_BB(mp, flen);
	97
a46db608 CH	98	/*
	99	* Too small? Give up.
	100	*/
a66d6363	101	if (dlen < minlen) {
a46db608 CH	102	trace_xfs_discard_toosmall(mp, agno, fbno, flen);
	103	goto out_del_cursor;
	104	}
	105
	106	/*
	107	* If the extent is entirely outside of the range we are
	108	* supposed to discard skip it. Do not bother to trim
	109	* down partially overlapping ranges for now.
	110	*/
a66d6363	111	if (dbno + dlen < start \|\| dbno > end) {
a46db608 CH	112	trace_xfs_discard_exclude(mp, agno, fbno, flen);
	113	goto next_extent;
	114	}
	115
	116	/*
	117	* If any blocks in the range are still busy, skip the
	118	* discard and try again the next time.
	119	*/
4ecbfe63	120	if (xfs_extent_busy_search(mp, agno, fbno, flen)) {
a46db608 CH	121	trace_xfs_discard_busy(mp, agno, fbno, flen);
	122	goto next_extent;
	123	}
	124
	125	trace_xfs_discard_extent(mp, agno, fbno, flen);
2451337d	126	error = blkdev_issue_discard(bdev, dbno, dlen, GFP_NOFS, 0);
a46db608 CH	127	if (error)
	128	goto out_del_cursor;
	129	*blocks_trimmed += flen;
	130
	131	next_extent:
	132	error = xfs_btree_decrement(cur, 0, &i);
	133	if (error)
	134	goto out_del_cursor;
	135	}
	136
	137	out_del_cursor:
	138	xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
	139	xfs_buf_relse(agbp);
	140	out_put_perag:
	141	xfs_perag_put(pag);
	142	return error;
	143	}
	144
a66d6363 DC	145	/*
	146	* trim a range of the filesystem.
	147	*
	148	* Note: the parameters passed from userspace are byte ranges into the
	149	* filesystem which does not match to the format we use for filesystem block
	150	* addressing. FSB addressing is sparse (AGNO\|AGBNO), while the incoming format
	151	* is a linear address range. Hence we need to use DADDR based conversions and
	152	* comparisons for determining the correct offset and regions to trim.
	153	*/
a46db608 CH	154	int
	155	xfs_ioc_trim(
	156	struct xfs_mount *mp,
	157	struct fstrim_range __user *urange)
	158	{
2f42d612	159	struct request_queue *q = bdev_get_queue(mp->m_ddev_targp->bt_bdev);
a46db608 CH	160	unsigned int granularity = q->limits.discard_granularity;
a46db608 CH	161	struct fstrim_range range;
a66d6363	162	xfs_daddr_t start, end, minlen;
a46db608 CH	163	xfs_agnumber_t start_agno, end_agno, agno;
	164	__uint64_t blocks_trimmed = 0;
	165	int error, last_error = 0;
	166
	167	if (!capable(CAP_SYS_ADMIN))
b474c7ae	168	return -EPERM;
be715140	169	if (!blk_queue_discard(q))
b474c7ae	170	return -EOPNOTSUPP;
a46db608	171	if (copy_from_user(&range, urange, sizeof(range)))
b474c7ae	172	return -EFAULT;
a46db608 CH	173
	174	/*
	175	* Truncating down the len isn't actually quite correct, but using
a66d6363	176	* BBTOB would mean we trivially get overflows for values
a46db608 CH	177	* of ULLONG_MAX or slightly lower. And ULLONG_MAX is the default
	178	* used by the fstrim application. In the end it really doesn't
	179	* matter as trimming blocks is an advisory interface.
	180	*/
a672e1be	181	if (range.start >= XFS_FSB_TO_B(mp, mp->m_sb.sb_dblocks) \|\|
2f42d612 JL	182	range.minlen > XFS_FSB_TO_B(mp, XFS_ALLOC_AG_MAX_USABLE(mp)) \|\|
2f42d612 JL	183	range.len < mp->m_sb.sb_blocksize)
b474c7ae	184	return -EINVAL;
a672e1be	185
a66d6363 DC	186	start = BTOBB(range.start);
	187	end = start + BTOBBT(range.len) - 1;
	188	minlen = BTOBB(max_t(u64, granularity, range.minlen));
a46db608	189
a66d6363 DC	190	if (end > XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks) - 1)
a66d6363 DC	191	end = XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)- 1;
a46db608	192
a66d6363 DC	193	start_agno = xfs_daddr_to_agno(mp, start);
a66d6363 DC	194	end_agno = xfs_daddr_to_agno(mp, end);
a46db608 CH	195
a46db608 CH	196	for (agno = start_agno; agno <= end_agno; agno++) {
2451337d	197	error = xfs_trim_extents(mp, agno, start, end, minlen,
a46db608 CH	198	&blocks_trimmed);
	199	if (error)
	200	last_error = error;
	201	}
	202
	203	if (last_error)
	204	return last_error;
	205
	206	range.len = XFS_FSB_TO_B(mp, blocks_trimmed);
	207	if (copy_to_user(urange, &range, sizeof(range)))
b474c7ae	208	return -EFAULT;
a46db608 CH	209	return 0;
a46db608 CH	210	}
e84661aa CH	211
	212	int
	213	xfs_discard_extents(
	214	struct xfs_mount *mp,
	215	struct list_head *list)
	216	{
4ecbfe63	217	struct xfs_extent_busy *busyp;
e84661aa CH	218	int error = 0;
	219
	220	list_for_each_entry(busyp, list, list) {
	221	trace_xfs_discard_extent(mp, busyp->agno, busyp->bno,
	222	busyp->length);
	223
2451337d	224	error = blkdev_issue_discard(mp->m_ddev_targp->bt_bdev,
e84661aa CH	225	XFS_AGB_TO_DADDR(mp, busyp->agno, busyp->bno),
	226	XFS_FSB_TO_BB(mp, busyp->length),
	227	GFP_NOFS, 0);
2451337d	228	if (error && error != -EOPNOTSUPP) {
e84661aa CH	229	xfs_info(mp,
	230	"discard failed for extent [0x%llu,%u], error %d",
	231	(unsigned long long)busyp->bno,
	232	busyp->length,
	233	error);
	234	return error;
	235	}
	236	}
	237
	238	return 0;
	239	}