[linux-block.git] / fs / xfs / xfs_notify_failure.c

// SPDX-License-Identifier: GPL-2.0
/*
 * Copyright (c) 2022 Fujitsu.  All Rights Reserved.
 */

#include "xfs.h"
#include "xfs_shared.h"
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
#include "xfs_mount.h"
#include "xfs_alloc.h"
#include "xfs_bit.h"
#include "xfs_btree.h"
#include "xfs_inode.h"
#include "xfs_icache.h"
#include "xfs_rmap.h"
#include "xfs_rmap_btree.h"
#include "xfs_rtalloc.h"
#include "xfs_trans.h"
#include "xfs_ag.h"

#include <linux/mm.h>
#include <linux/dax.h>

struct xfs_failure_info {
	xfs_agblock_t		startblock;
	xfs_extlen_t		blockcount;
	int			mf_flags;
	bool			want_shutdown;
};

static pgoff_t
xfs_failure_pgoff(
	struct xfs_mount		*mp,
	const struct xfs_rmap_irec	*rec,
	const struct xfs_failure_info	*notify)
{
	loff_t				pos = XFS_FSB_TO_B(mp, rec->rm_offset);

	if (notify->startblock > rec->rm_startblock)
		pos += XFS_FSB_TO_B(mp,
				notify->startblock - rec->rm_startblock);
	return pos >> PAGE_SHIFT;
}

static unsigned long
xfs_failure_pgcnt(
	struct xfs_mount		*mp,
	const struct xfs_rmap_irec	*rec,
	const struct xfs_failure_info	*notify)
{
	xfs_agblock_t			end_rec;
	xfs_agblock_t			end_notify;
	xfs_agblock_t			start_cross;
	xfs_agblock_t			end_cross;

	start_cross = max(rec->rm_startblock, notify->startblock);

	end_rec = rec->rm_startblock + rec->rm_blockcount;
	end_notify = notify->startblock + notify->blockcount;
	end_cross = min(end_rec, end_notify);

	return XFS_FSB_TO_B(mp, end_cross - start_cross) >> PAGE_SHIFT;
}

static int
xfs_dax_failure_fn(
	struct xfs_btree_cur		*cur,
	const struct xfs_rmap_irec	*rec,
	void				*data)
{
	struct xfs_mount		*mp = cur->bc_mp;
	struct xfs_inode		*ip;
	struct xfs_failure_info		*notify = data;
	int				error = 0;

	if (XFS_RMAP_NON_INODE_OWNER(rec->rm_owner) ||
	    (rec->rm_flags & (XFS_RMAP_ATTR_FORK | XFS_RMAP_BMBT_BLOCK))) {
		notify->want_shutdown = true;
		return 0;
	}

	/* Get files that incore, filter out others that are not in use. */
	error = xfs_iget(mp, cur->bc_tp, rec->rm_owner, XFS_IGET_INCORE,
			 0, &ip);
	/* Continue the rmap query if the inode isn't incore */
	if (error == -ENODATA)
		return 0;
	if (error) {
		notify->want_shutdown = true;
		return 0;
	}

	error = mf_dax_kill_procs(VFS_I(ip)->i_mapping,
				  xfs_failure_pgoff(mp, rec, notify),
				  xfs_failure_pgcnt(mp, rec, notify),
				  notify->mf_flags);
	xfs_irele(ip);
	return error;
}

static int
xfs_dax_notify_ddev_failure(
	struct xfs_mount	*mp,
	xfs_daddr_t		daddr,
	xfs_daddr_t		bblen,
	int			mf_flags)
{
	struct xfs_failure_info	notify = { .mf_flags = mf_flags };
	struct xfs_trans	*tp = NULL;
	struct xfs_btree_cur	*cur = NULL;
	struct xfs_buf		*agf_bp = NULL;
	int			error = 0;
	xfs_fsblock_t		fsbno = XFS_DADDR_TO_FSB(mp, daddr);
	xfs_agnumber_t		agno = XFS_FSB_TO_AGNO(mp, fsbno);
	xfs_fsblock_t		end_fsbno = XFS_DADDR_TO_FSB(mp, daddr + bblen);
	xfs_agnumber_t		end_agno = XFS_FSB_TO_AGNO(mp, end_fsbno);

	error = xfs_trans_alloc_empty(mp, &tp);
	if (error)
		return error;

	for (; agno <= end_agno; agno++) {
		struct xfs_rmap_irec	ri_low = { };
		struct xfs_rmap_irec	ri_high;
		struct xfs_agf		*agf;
		xfs_agblock_t		agend;
		struct xfs_perag	*pag;

		pag = xfs_perag_get(mp, agno);
		error = xfs_alloc_read_agf(pag, tp, 0, &agf_bp);
		if (error) {
			xfs_perag_put(pag);
			break;
		}

		cur = xfs_rmapbt_init_cursor(mp, tp, agf_bp, pag);

		/*
		 * Set the rmap range from ri_low to ri_high, which represents
		 * a [start, end] where we looking for the files or metadata.
		 */
		memset(&ri_high, 0xFF, sizeof(ri_high));
		ri_low.rm_startblock = XFS_FSB_TO_AGBNO(mp, fsbno);
		if (agno == end_agno)
			ri_high.rm_startblock = XFS_FSB_TO_AGBNO(mp, end_fsbno);

		agf = agf_bp->b_addr;
		agend = min(be32_to_cpu(agf->agf_length),
				ri_high.rm_startblock);
		notify.startblock = ri_low.rm_startblock;
		notify.blockcount = agend - ri_low.rm_startblock;

		error = xfs_rmap_query_range(cur, &ri_low, &ri_high,
				xfs_dax_failure_fn, &notify);
		xfs_btree_del_cursor(cur, error);
		xfs_trans_brelse(tp, agf_bp);
		xfs_perag_put(pag);
		if (error)
			break;

		fsbno = XFS_AGB_TO_FSB(mp, agno + 1, 0);
	}

	xfs_trans_cancel(tp);
	if (error || notify.want_shutdown) {
		xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_ONDISK);
		if (!error)
			error = -EFSCORRUPTED;
	}
	return error;
}

static int
xfs_dax_notify_failure(
	struct dax_device	*dax_dev,
	u64			offset,
	u64			len,
	int			mf_flags)
{
	struct xfs_mount	*mp = dax_holder(dax_dev);
	u64			ddev_start;
	u64			ddev_end;

	if (!(mp->m_super->s_flags & SB_BORN)) {
		xfs_warn(mp, "filesystem is not ready for notify_failure()!");
		return -EIO;
	}

	if (mp->m_rtdev_targp && mp->m_rtdev_targp->bt_daxdev == dax_dev) {
		xfs_debug(mp,
			 "notify_failure() not supported on realtime device!");
		return -EOPNOTSUPP;
	}

	if (mp->m_logdev_targp && mp->m_logdev_targp->bt_daxdev == dax_dev &&
	    mp->m_logdev_targp != mp->m_ddev_targp) {
		xfs_err(mp, "ondisk log corrupt, shutting down fs!");
		xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_ONDISK);
		return -EFSCORRUPTED;
	}

	if (!xfs_has_rmapbt(mp)) {
		xfs_debug(mp, "notify_failure() needs rmapbt enabled!");
		return -EOPNOTSUPP;
	}

	ddev_start = mp->m_ddev_targp->bt_dax_part_off;
	ddev_end = ddev_start + bdev_nr_bytes(mp->m_ddev_targp->bt_bdev) - 1;

	/* Ignore the range out of filesystem area */
	if (offset + len < ddev_start)
		return -ENXIO;
	if (offset > ddev_end)
		return -ENXIO;

	/* Calculate the real range when it touches the boundary */
	if (offset > ddev_start)
		offset -= ddev_start;
	else {
		len -= ddev_start - offset;
		offset = 0;
	}
	if (offset + len > ddev_end)
		len -= ddev_end - offset;

	return xfs_dax_notify_ddev_failure(mp, BTOBB(offset), BTOBB(len),
			mf_flags);
}

const struct dax_holder_operations xfs_dax_holder_operations = {
	.notify_failure		= xfs_dax_notify_failure,
};
Commit	Line	Data
6f643c57 SR	1	// SPDX-License-Identifier: GPL-2.0
	2	/*
	3	* Copyright (c) 2022 Fujitsu. All Rights Reserved.
	4	*/
	5
	6	#include "xfs.h"
	7	#include "xfs_shared.h"
	8	#include "xfs_format.h"
	9	#include "xfs_log_format.h"
	10	#include "xfs_trans_resv.h"
	11	#include "xfs_mount.h"
	12	#include "xfs_alloc.h"
	13	#include "xfs_bit.h"
	14	#include "xfs_btree.h"
	15	#include "xfs_inode.h"
	16	#include "xfs_icache.h"
	17	#include "xfs_rmap.h"
	18	#include "xfs_rmap_btree.h"
	19	#include "xfs_rtalloc.h"
	20	#include "xfs_trans.h"
6614a3c3	21	#include "xfs_ag.h"
6f643c57 SR	22
	23	#include <linux/mm.h>
	24	#include <linux/dax.h>
	25
e033f40b	26	struct xfs_failure_info {
6f643c57 SR	27	xfs_agblock_t startblock;
	28	xfs_extlen_t blockcount;
	29	int mf_flags;
e033f40b	30	bool want_shutdown;
6f643c57 SR	31	};
	32
	33	static pgoff_t
	34	xfs_failure_pgoff(
	35	struct xfs_mount *mp,
	36	const struct xfs_rmap_irec *rec,
e033f40b	37	const struct xfs_failure_info *notify)
6f643c57 SR	38	{
	39	loff_t pos = XFS_FSB_TO_B(mp, rec->rm_offset);
	40
	41	if (notify->startblock > rec->rm_startblock)
	42	pos += XFS_FSB_TO_B(mp,
	43	notify->startblock - rec->rm_startblock);
	44	return pos >> PAGE_SHIFT;
	45	}
	46
	47	static unsigned long
	48	xfs_failure_pgcnt(
	49	struct xfs_mount *mp,
	50	const struct xfs_rmap_irec *rec,
e033f40b	51	const struct xfs_failure_info *notify)
6f643c57 SR	52	{
	53	xfs_agblock_t end_rec;
	54	xfs_agblock_t end_notify;
	55	xfs_agblock_t start_cross;
	56	xfs_agblock_t end_cross;
	57
	58	start_cross = max(rec->rm_startblock, notify->startblock);
	59
	60	end_rec = rec->rm_startblock + rec->rm_blockcount;
	61	end_notify = notify->startblock + notify->blockcount;
	62	end_cross = min(end_rec, end_notify);
	63
	64	return XFS_FSB_TO_B(mp, end_cross - start_cross) >> PAGE_SHIFT;
	65	}
	66
	67	static int
	68	xfs_dax_failure_fn(
	69	struct xfs_btree_cur *cur,
	70	const struct xfs_rmap_irec *rec,
	71	void *data)
	72	{
	73	struct xfs_mount *mp = cur->bc_mp;
	74	struct xfs_inode *ip;
e033f40b	75	struct xfs_failure_info *notify = data;
6f643c57 SR	76	int error = 0;
	77
	78	if (XFS_RMAP_NON_INODE_OWNER(rec->rm_owner) \|\|
	79	(rec->rm_flags & (XFS_RMAP_ATTR_FORK \| XFS_RMAP_BMBT_BLOCK))) {
e033f40b DW	80	notify->want_shutdown = true;
e033f40b DW	81	return 0;
6f643c57 SR	82	}
	83
	84	/* Get files that incore, filter out others that are not in use. */
	85	error = xfs_iget(mp, cur->bc_tp, rec->rm_owner, XFS_IGET_INCORE,
	86	0, &ip);
	87	/* Continue the rmap query if the inode isn't incore */
	88	if (error == -ENODATA)
	89	return 0;
e033f40b DW	90	if (error) {
	91	notify->want_shutdown = true;
	92	return 0;
	93	}
6f643c57 SR	94
	95	error = mf_dax_kill_procs(VFS_I(ip)->i_mapping,
	96	xfs_failure_pgoff(mp, rec, notify),
	97	xfs_failure_pgcnt(mp, rec, notify),
	98	notify->mf_flags);
	99	xfs_irele(ip);
	100	return error;
	101	}
	102
	103	static int
	104	xfs_dax_notify_ddev_failure(
	105	struct xfs_mount *mp,
	106	xfs_daddr_t daddr,
	107	xfs_daddr_t bblen,
	108	int mf_flags)
	109	{
e033f40b	110	struct xfs_failure_info notify = { .mf_flags = mf_flags };
6f643c57 SR	111	struct xfs_trans *tp = NULL;
	112	struct xfs_btree_cur *cur = NULL;
	113	struct xfs_buf *agf_bp = NULL;
	114	int error = 0;
	115	xfs_fsblock_t fsbno = XFS_DADDR_TO_FSB(mp, daddr);
	116	xfs_agnumber_t agno = XFS_FSB_TO_AGNO(mp, fsbno);
	117	xfs_fsblock_t end_fsbno = XFS_DADDR_TO_FSB(mp, daddr + bblen);
	118	xfs_agnumber_t end_agno = XFS_FSB_TO_AGNO(mp, end_fsbno);
	119
	120	error = xfs_trans_alloc_empty(mp, &tp);
	121	if (error)
	122	return error;
	123
	124	for (; agno <= end_agno; agno++) {
	125	struct xfs_rmap_irec ri_low = { };
	126	struct xfs_rmap_irec ri_high;
6f643c57 SR	127	struct xfs_agf *agf;
6f643c57 SR	128	xfs_agblock_t agend;
6614a3c3	129	struct xfs_perag *pag;
6f643c57	130
6614a3c3 LT	131	pag = xfs_perag_get(mp, agno);
	132	error = xfs_alloc_read_agf(pag, tp, 0, &agf_bp);
	133	if (error) {
	134	xfs_perag_put(pag);
6f643c57	135	break;
6614a3c3	136	}
6f643c57	137
6614a3c3	138	cur = xfs_rmapbt_init_cursor(mp, tp, agf_bp, pag);
6f643c57 SR	139
	140	/*
	141	* Set the rmap range from ri_low to ri_high, which represents
	142	* a [start, end] where we looking for the files or metadata.
	143	*/
	144	memset(&ri_high, 0xFF, sizeof(ri_high));
	145	ri_low.rm_startblock = XFS_FSB_TO_AGBNO(mp, fsbno);
	146	if (agno == end_agno)
	147	ri_high.rm_startblock = XFS_FSB_TO_AGBNO(mp, end_fsbno);
	148
	149	agf = agf_bp->b_addr;
	150	agend = min(be32_to_cpu(agf->agf_length),
	151	ri_high.rm_startblock);
	152	notify.startblock = ri_low.rm_startblock;
	153	notify.blockcount = agend - ri_low.rm_startblock;
	154
	155	error = xfs_rmap_query_range(cur, &ri_low, &ri_high,
	156	xfs_dax_failure_fn, &notify);
	157	xfs_btree_del_cursor(cur, error);
	158	xfs_trans_brelse(tp, agf_bp);
6614a3c3	159	xfs_perag_put(pag);
6f643c57 SR	160	if (error)
	161	break;
	162
	163	fsbno = XFS_AGB_TO_FSB(mp, agno + 1, 0);
	164	}
	165
	166	xfs_trans_cancel(tp);
e033f40b DW	167	if (error \|\| notify.want_shutdown) {
	168	xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_ONDISK);
	169	if (!error)
	170	error = -EFSCORRUPTED;
	171	}
6f643c57 SR	172	return error;
	173	}
	174
	175	static int
	176	xfs_dax_notify_failure(
	177	struct dax_device *dax_dev,
	178	u64 offset,
	179	u64 len,
	180	int mf_flags)
	181	{
	182	struct xfs_mount *mp = dax_holder(dax_dev);
	183	u64 ddev_start;
	184	u64 ddev_end;
	185
fd63612a	186	if (!(mp->m_super->s_flags & SB_BORN)) {
6f643c57 SR	187	xfs_warn(mp, "filesystem is not ready for notify_failure()!");
	188	return -EIO;
	189	}
	190
	191	if (mp->m_rtdev_targp && mp->m_rtdev_targp->bt_daxdev == dax_dev) {
b14d067e	192	xfs_debug(mp,
6f643c57 SR	193	"notify_failure() not supported on realtime device!");
	194	return -EOPNOTSUPP;
	195	}
	196
	197	if (mp->m_logdev_targp && mp->m_logdev_targp->bt_daxdev == dax_dev &&
	198	mp->m_logdev_targp != mp->m_ddev_targp) {
	199	xfs_err(mp, "ondisk log corrupt, shutting down fs!");
	200	xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_ONDISK);
	201	return -EFSCORRUPTED;
	202	}
	203
	204	if (!xfs_has_rmapbt(mp)) {
b14d067e	205	xfs_debug(mp, "notify_failure() needs rmapbt enabled!");
6f643c57 SR	206	return -EOPNOTSUPP;
	207	}
	208
	209	ddev_start = mp->m_ddev_targp->bt_dax_part_off;
	210	ddev_end = ddev_start + bdev_nr_bytes(mp->m_ddev_targp->bt_bdev) - 1;
	211
	212	/* Ignore the range out of filesystem area */
	213	if (offset + len < ddev_start)
	214	return -ENXIO;
	215	if (offset > ddev_end)
	216	return -ENXIO;
	217
	218	/* Calculate the real range when it touches the boundary */
	219	if (offset > ddev_start)
	220	offset -= ddev_start;
	221	else {
	222	len -= ddev_start - offset;
	223	offset = 0;
	224	}
	225	if (offset + len > ddev_end)
	226	len -= ddev_end - offset;
	227
	228	return xfs_dax_notify_ddev_failure(mp, BTOBB(offset), BTOBB(len),
	229	mf_flags);
	230	}
	231
	232	const struct dax_holder_operations xfs_dax_holder_operations = {
	233	.notify_failure = xfs_dax_notify_failure,
	234	};