/*
* Add a reference to an extent in the rmap btree.
*/
int
xfs_rmap_alloc(
struct xfs_trans *tp,
struct xfs_buf *agbp,
xfs_agnumber_t agno,
xfs_agblock_t bno,
xfs_extlen_t len,
struct xfs_owner_info *oinfo)
{
struct xfs_mount *mp = tp->t_mountp;
struct xfs_btree_cur *cur;
int error;
if (!xfs_sb_version_hasrmapbt(&mp->m_sb))
return 0;
cur = xfs_rmapbt_init_cursor(mp, tp, agbp, agno);
error = xfs_rmap_map(cur, bno, len, false, oinfo);
if (error)
goto out_error;
xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
return 0;
out_error:
xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
return error;
}
/* Dispose of every block of every extent in the bitmap. */
int
xrep_reap_extents(
struct xfs_scrub *sc,
struct xfs_bitmap *bitmap,
const struct xfs_owner_info *oinfo,
enum xfs_ag_resv_type type)
{
struct xfs_bitmap_range *bmr;
struct xfs_bitmap_range *n;
xfs_fsblock_t fsbno;
int error = 0;
ASSERT(xfs_sb_version_hasrmapbt(&sc->mp->m_sb));
for_each_xfs_bitmap_block(fsbno, bmr, n, bitmap) {
ASSERT(sc->ip != NULL ||
XFS_FSB_TO_AGNO(sc->mp, fsbno) == sc->sa.agno);
trace_xrep_dispose_btree_extent(sc->mp,
XFS_FSB_TO_AGNO(sc->mp, fsbno),
XFS_FSB_TO_AGBNO(sc->mp, fsbno), 1);
error = xrep_reap_block(sc, fsbno, oinfo, type);
if (error)
goto out;
}
/*
* Per-extent log reservation for the btree changes involved in freeing or
* allocating an extent. In classic XFS there were two trees that will be
* modified (bnobt + cntbt). With rmap enabled, there are three trees
* (rmapbt). With reflink, there are four trees (refcountbt). The number of
* blocks reserved is based on the formula:
*
* num trees * ((2 blocks/level * max depth) - 1)
*
* Keep in mind that max depth is calculated separately for each type of tree.
*/
uint
xfs_allocfree_log_count(
struct xfs_mount *mp,
uint num_ops)
{
uint blocks;
blocks = num_ops * 2 * (2 * mp->m_ag_maxlevels - 1);
if (xfs_sb_version_hasrmapbt(&mp->m_sb))
blocks += num_ops * (2 * mp->m_rmap_maxlevels - 1);
if (xfs_sb_version_hasreflink(&mp->m_sb))
blocks += num_ops * (2 * mp->m_refc_maxlevels - 1);
return blocks;
}
/*
* Do we have enough reserve in this AG to handle a reflink? The refcount
* btree already reserved all the space it needs, but the rmap btree can grow
* infinitely, so we won't allow more reflinks when the AG is down to the
* btree reserves.
*/
static int
xfs_reflink_ag_has_free_space(
struct xfs_mount *mp,
xfs_agnumber_t agno)
{
struct xfs_perag *pag;
int error = 0;
if (!xfs_sb_version_hasrmapbt(&mp->m_sb))
return 0;
pag = xfs_perag_get(mp, agno);
if (xfs_ag_resv_critical(pag, XFS_AG_RESV_AGFL) ||
xfs_ag_resv_critical(pag, XFS_AG_RESV_METADATA))
error = -ENOSPC;
xfs_perag_put(pag);
return error;
}
static void
xfs_agfblock_init(
struct xfs_mount *mp,
struct xfs_buf *bp,
struct aghdr_init_data *id)
{
struct xfs_agf *agf = XFS_BUF_TO_AGF(bp);
xfs_extlen_t tmpsize;
agf->agf_magicnum = cpu_to_be32(XFS_AGF_MAGIC);
agf->agf_versionnum = cpu_to_be32(XFS_AGF_VERSION);
agf->agf_seqno = cpu_to_be32(id->agno);
agf->agf_length = cpu_to_be32(id->agsize);
agf->agf_roots[XFS_BTNUM_BNOi] = cpu_to_be32(XFS_BNO_BLOCK(mp));
agf->agf_roots[XFS_BTNUM_CNTi] = cpu_to_be32(XFS_CNT_BLOCK(mp));
agf->agf_levels[XFS_BTNUM_BNOi] = cpu_to_be32(1);
agf->agf_levels[XFS_BTNUM_CNTi] = cpu_to_be32(1);
if (xfs_sb_version_hasrmapbt(&mp->m_sb)) {
agf->agf_roots[XFS_BTNUM_RMAPi] =
cpu_to_be32(XFS_RMAP_BLOCK(mp));
agf->agf_levels[XFS_BTNUM_RMAPi] = cpu_to_be32(1);
agf->agf_rmap_blocks = cpu_to_be32(1);
}
agf->agf_flfirst = cpu_to_be32(1);
agf->agf_fllast = 0;
agf->agf_flcount = 0;
tmpsize = id->agsize - mp->m_ag_prealloc_blocks;
agf->agf_freeblks = cpu_to_be32(tmpsize);
agf->agf_longest = cpu_to_be32(tmpsize);
if (xfs_sb_version_hascrc(&mp->m_sb))
uuid_copy(&agf->agf_uuid, &mp->m_sb.sb_meta_uuid);
if (xfs_sb_version_hasreflink(&mp->m_sb)) {
agf->agf_refcount_root = cpu_to_be32(
xfs_refc_block(mp));
agf->agf_refcount_level = cpu_to_be32(1);
agf->agf_refcount_blocks = cpu_to_be32(1);
}
}
/*
* Get filesystem's extents as described in head, and format for
* output. Calls formatter to fill the user's buffer until all
* extents are mapped, until the passed-in head->fmh_count slots have
* been filled, or until the formatter short-circuits the loop, if it
* is tracking filled-in extents on its own.
*
* Key to Confusion
* ----------------
* There are multiple levels of keys and counters at work here:
* xfs_fsmap_head.fmh_keys -- low and high fsmap keys passed in;
* these reflect fs-wide sector addrs.
* dkeys -- fmh_keys used to query each device;
* these are fmh_keys but w/ the low key
* bumped up by fmr_length.
* xfs_getfsmap_info.next_daddr -- next disk addr we expect to see; this
* is how we detect gaps in the fsmap
records and report them.
* xfs_getfsmap_info.low/high -- per-AG low/high keys computed from
* dkeys; used to query the metadata.
*/
int
xfs_getfsmap(
struct xfs_mount *mp,
struct xfs_fsmap_head *head,
xfs_fsmap_format_t formatter,
void *arg)
{
struct xfs_trans *tp = NULL;
struct xfs_fsmap dkeys[2]; /* per-dev keys */
struct xfs_getfsmap_dev handlers[XFS_GETFSMAP_DEVS];
struct xfs_getfsmap_info info = { NULL };
bool use_rmap;
int i;
int error = 0;
if (head->fmh_iflags & ~FMH_IF_VALID)
return -EINVAL;
if (!xfs_getfsmap_is_valid_device(mp, &head->fmh_keys[0]) ||
!xfs_getfsmap_is_valid_device(mp, &head->fmh_keys[1]))
return -EINVAL;
use_rmap = capable(CAP_SYS_ADMIN) &&
xfs_sb_version_hasrmapbt(&mp->m_sb);
head->fmh_entries = 0;
/* Set up our device handlers. */
memset(handlers, 0, sizeof(handlers));
handlers[0].dev = new_encode_dev(mp->m_ddev_targp->bt_dev);
if (use_rmap)
handlers[0].fn = xfs_getfsmap_datadev_rmapbt;
else
handlers[0].fn = xfs_getfsmap_datadev_bnobt;
if (mp->m_logdev_targp != mp->m_ddev_targp) {
handlers[1].dev = new_encode_dev(mp->m_logdev_targp->bt_dev);
handlers[1].fn = xfs_getfsmap_logdev;
}
#ifdef CONFIG_XFS_RT
if (mp->m_rtdev_targp) {
handlers[2].dev = new_encode_dev(mp->m_rtdev_targp->bt_dev);
handlers[2].fn = xfs_getfsmap_rtdev_rtbitmap;
}
#endif /* CONFIG_XFS_RT */
xfs_sort(handlers, XFS_GETFSMAP_DEVS, sizeof(struct xfs_getfsmap_dev),
xfs_getfsmap_dev_compare);
/*
* To continue where we left off, we allow userspace to use the
* last mapping from a previous call as the low key of the next.
* This is identified by a non-zero length in the low key. We
* have to increment the low key in this scenario to ensure we
* don't return the same mapping again, and instead return the
* very next mapping.
*
* If the low key mapping refers to file data, the same physical
* blocks could be mapped to several other files/offsets.
* According to rmapbt record ordering, the minimal next
* possible record for the block range is the next starting
* offset in the same inode. Therefore, bump the file offset to
* continue the search appropriately. For all other low key
* mapping types (attr blocks, metadata), bump the physical
* offset as there can be no other mapping for the same physical
* block range.
*/
dkeys[0] = head->fmh_keys[0];
if (dkeys[0].fmr_flags & (FMR_OF_SPECIAL_OWNER | FMR_OF_EXTENT_MAP)) {
dkeys[0].fmr_physical += dkeys[0].fmr_length;
dkeys[0].fmr_owner = 0;
if (dkeys[0].fmr_offset)
return -EINVAL;
} else
dkeys[0].fmr_offset += dkeys[0].fmr_length;
dkeys[0].fmr_length = 0;
memset(&dkeys[1], 0xFF, sizeof(struct xfs_fsmap));
if (!xfs_getfsmap_check_keys(dkeys, &head->fmh_keys[1]))
return -EINVAL;
info.next_daddr = head->fmh_keys[0].fmr_physical +
head->fmh_keys[0].fmr_length;
info.formatter = formatter;
info.format_arg = arg;
info.head = head;
/* For each device we support... */
for (i = 0; i < XFS_GETFSMAP_DEVS; i++) {
/* Is this device within the range the user asked for? */
if (!handlers[i].fn)
continue;
if (head->fmh_keys[0].fmr_device > handlers[i].dev)
continue;
if (head->fmh_keys[1].fmr_device < handlers[i].dev)
break;
/*
* If this device number matches the high key, we have
* to pass the high key to the handler to limit the
* query results. If the device number exceeds the
* low key, zero out the low key so that we get
* everything from the beginning.
*/
if (handlers[i].dev == head->fmh_keys[1].fmr_device)
dkeys[1] = head->fmh_keys[1];
if (handlers[i].dev > head->fmh_keys[0].fmr_device)
memset(&dkeys[0], 0, sizeof(struct xfs_fsmap));
error = xfs_trans_alloc_empty(mp, &tp);
if (error)
break;
info.dev = handlers[i].dev;
info.last = false;
info.agno = NULLAGNUMBER;
error = handlers[i].fn(tp, dkeys, &info);
if (error)
break;
xfs_trans_cancel(tp);
tp = NULL;
info.next_daddr = 0;
}
if (tp)
xfs_trans_cancel(tp);
head->fmh_oflags = FMH_OF_DEV_T;
return error;
}
/*
* Don't defer an rmap if we aren't an rmap filesystem.
*/
static bool
xfs_rmap_update_is_needed(
struct xfs_mount *mp)
{
return xfs_sb_version_hasrmapbt(&mp->m_sb);
}
/*
* Figure out how many blocks to reserve for an AG repair. We calculate the
* worst case estimate for the number of blocks we'd need to rebuild one of
* any type of per-AG btree.
*/
xfs_extlen_t
xrep_calc_ag_resblks(
struct xfs_scrub *sc)
{
struct xfs_mount *mp = sc->mp;
struct xfs_scrub_metadata *sm = sc->sm;
struct xfs_perag *pag;
struct xfs_buf *bp;
xfs_agino_t icount = NULLAGINO;
xfs_extlen_t aglen = NULLAGBLOCK;
xfs_extlen_t usedlen;
xfs_extlen_t freelen;
xfs_extlen_t bnobt_sz;
xfs_extlen_t inobt_sz;
xfs_extlen_t rmapbt_sz;
xfs_extlen_t refcbt_sz;
int error;
if (!(sm->sm_flags & XFS_SCRUB_IFLAG_REPAIR))
return 0;
pag = xfs_perag_get(mp, sm->sm_agno);
if (pag->pagi_init) {
/* Use in-core icount if possible. */
icount = pag->pagi_count;
} else {
/* Try to get the actual counters from disk. */
error = xfs_ialloc_read_agi(mp, NULL, sm->sm_agno, &bp);
if (!error) {
icount = pag->pagi_count;
xfs_buf_relse(bp);
}
}
/* Now grab the block counters from the AGF. */
error = xfs_alloc_read_agf(mp, NULL, sm->sm_agno, 0, &bp);
if (!error) {
aglen = be32_to_cpu(XFS_BUF_TO_AGF(bp)->agf_length);
freelen = be32_to_cpu(XFS_BUF_TO_AGF(bp)->agf_freeblks);
usedlen = aglen - freelen;
xfs_buf_relse(bp);
}
xfs_perag_put(pag);
/* If the icount is impossible, make some worst-case assumptions. */
if (icount == NULLAGINO ||
!xfs_verify_agino(mp, sm->sm_agno, icount)) {
xfs_agino_t first, last;
xfs_agino_range(mp, sm->sm_agno, &first, &last);
icount = last - first + 1;
}
/* If the block counts are impossible, make worst-case assumptions. */
if (aglen == NULLAGBLOCK ||
aglen != xfs_ag_block_count(mp, sm->sm_agno) ||
freelen >= aglen) {
aglen = xfs_ag_block_count(mp, sm->sm_agno);
freelen = aglen;
usedlen = aglen;
}
trace_xrep_calc_ag_resblks(mp, sm->sm_agno, icount, aglen,
freelen, usedlen);
/*
* Figure out how many blocks we'd need worst case to rebuild
* each type of btree. Note that we can only rebuild the
* bnobt/cntbt or inobt/finobt as pairs.
*/
bnobt_sz = 2 * xfs_allocbt_calc_size(mp, freelen);
if (xfs_sb_version_hassparseinodes(&mp->m_sb))
inobt_sz = xfs_iallocbt_calc_size(mp, icount /
XFS_INODES_PER_HOLEMASK_BIT);
else
inobt_sz = xfs_iallocbt_calc_size(mp, icount /
XFS_INODES_PER_CHUNK);
if (xfs_sb_version_hasfinobt(&mp->m_sb))
inobt_sz *= 2;
if (xfs_sb_version_hasreflink(&mp->m_sb))
refcbt_sz = xfs_refcountbt_calc_size(mp, usedlen);
else
refcbt_sz = 0;
if (xfs_sb_version_hasrmapbt(&mp->m_sb)) {
/*
* Guess how many blocks we need to rebuild the rmapbt.
* For non-reflink filesystems we can't have more records than
* used blocks. However, with reflink it's possible to have
* more than one rmap record per AG block. We don't know how
* many rmaps there could be in the AG, so we start off with
* what we hope is an generous over-estimation.
*/
if (xfs_sb_version_hasreflink(&mp->m_sb))
rmapbt_sz = xfs_rmapbt_calc_size(mp,
(unsigned long long)aglen * 2);
else
rmapbt_sz = xfs_rmapbt_calc_size(mp, usedlen);
} else {
rmapbt_sz = 0;
}
trace_xrep_calc_ag_resblks_btsize(mp, sm->sm_agno, bnobt_sz,
inobt_sz, rmapbt_sz, refcbt_sz);
return max(max(bnobt_sz, inobt_sz), max(rmapbt_sz, refcbt_sz));
}