/*
* build both the agf and the agfl for an agno given both
* btree cursors.
*
* XXX: yet more common code that can be shared with mkfs/growfs.
*/
static void
build_agf_agfl(xfs_mount_t *mp,
xfs_agnumber_t agno,
bt_status_t *bno_bt,
bt_status_t *bcnt_bt,
xfs_extlen_t freeblks, /* # free blocks in tree */
int lostblocks) /* # blocks that will be lost */
{
extent_tree_node_t *ext_ptr;
xfs_buf_t *agf_buf, *agfl_buf;
int i;
int j;
xfs_agfl_t *agfl;
xfs_agf_t *agf;
__be32 *freelist;
agf_buf = libxfs_getbuf(mp->m_dev,
XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)),
mp->m_sb.sb_sectsize/BBSIZE);
agf_buf->b_ops = &xfs_agf_buf_ops;
agf = XFS_BUF_TO_AGF(agf_buf);
memset(agf, 0, mp->m_sb.sb_sectsize);
#ifdef XR_BLD_FREE_TRACE
fprintf(stderr, "agf = 0x%p, agf_buf->b_addr = 0x%p\n",
agf, agf_buf->b_addr);
#endif
/*
* set up fixed part of agf
*/
agf->agf_magicnum = cpu_to_be32(XFS_AGF_MAGIC);
agf->agf_versionnum = cpu_to_be32(XFS_AGF_VERSION);
agf->agf_seqno = cpu_to_be32(agno);
if (agno < mp->m_sb.sb_agcount - 1)
agf->agf_length = cpu_to_be32(mp->m_sb.sb_agblocks);
else
agf->agf_length = cpu_to_be32(mp->m_sb.sb_dblocks -
(xfs_rfsblock_t) mp->m_sb.sb_agblocks * agno);
agf->agf_roots[XFS_BTNUM_BNO] = cpu_to_be32(bno_bt->root);
agf->agf_levels[XFS_BTNUM_BNO] = cpu_to_be32(bno_bt->num_levels);
agf->agf_roots[XFS_BTNUM_CNT] = cpu_to_be32(bcnt_bt->root);
agf->agf_levels[XFS_BTNUM_CNT] = cpu_to_be32(bcnt_bt->num_levels);
agf->agf_freeblks = cpu_to_be32(freeblks);
/*
* Count and record the number of btree blocks consumed if required.
*/
if (xfs_sb_version_haslazysbcount(&mp->m_sb)) {
/*
* Don't count the root blocks as they are already
* accounted for.
*/
agf->agf_btreeblks = cpu_to_be32(
(bno_bt->num_tot_blocks - bno_bt->num_free_blocks) +
(bcnt_bt->num_tot_blocks - bcnt_bt->num_free_blocks) -
2);
#ifdef XR_BLD_FREE_TRACE
fprintf(stderr, "agf->agf_btreeblks = %u\n",
be32_to_cpu(agf->agf_btreeblks));
#endif
}
#ifdef XR_BLD_FREE_TRACE
fprintf(stderr, "bno root = %u, bcnt root = %u, indices = %u %u\n",
be32_to_cpu(agf->agf_roots[XFS_BTNUM_BNO]),
be32_to_cpu(agf->agf_roots[XFS_BTNUM_CNT]),
XFS_BTNUM_BNO,
XFS_BTNUM_CNT);
#endif
if (xfs_sb_version_hascrc(&mp->m_sb))
platform_uuid_copy(&agf->agf_uuid, &mp->m_sb.sb_meta_uuid);
/* initialise the AGFL, then fill it if there are blocks left over. */
agfl_buf = libxfs_getbuf(mp->m_dev,
XFS_AG_DADDR(mp, agno, XFS_AGFL_DADDR(mp)),
mp->m_sb.sb_sectsize/BBSIZE);
agfl_buf->b_ops = &xfs_agfl_buf_ops;
agfl = XFS_BUF_TO_AGFL(agfl_buf);
/* setting to 0xff results in initialisation to NULLAGBLOCK */
memset(agfl, 0xff, mp->m_sb.sb_sectsize);
if (xfs_sb_version_hascrc(&mp->m_sb)) {
agfl->agfl_magicnum = cpu_to_be32(XFS_AGFL_MAGIC);
agfl->agfl_seqno = cpu_to_be32(agno);
platform_uuid_copy(&agfl->agfl_uuid, &mp->m_sb.sb_meta_uuid);
for (i = 0; i < XFS_AGFL_SIZE(mp); i++)
agfl->agfl_bno[i] = cpu_to_be32(NULLAGBLOCK);
}
freelist = XFS_BUF_TO_AGFL_BNO(mp, agfl_buf);
/*
* do we have left-over blocks in the btree cursors that should
* be used to fill the AGFL?
*/
if (bno_bt->num_free_blocks > 0 || bcnt_bt->num_free_blocks > 0) {
/*
* yes, now grab as many blocks as we can
*/
i = j = 0;
while (bno_bt->num_free_blocks > 0 && i < XFS_AGFL_SIZE(mp)) {
freelist[i] = cpu_to_be32(
get_next_blockaddr(agno, 0, bno_bt));
i++;
}
while (bcnt_bt->num_free_blocks > 0 && i < XFS_AGFL_SIZE(mp)) {
freelist[i] = cpu_to_be32(
get_next_blockaddr(agno, 0, bcnt_bt));
i++;
}
/*
* now throw the rest of the blocks away and complain
*/
while (bno_bt->num_free_blocks > 0) {
(void) get_next_blockaddr(agno, 0, bno_bt);
j++;
}
while (bcnt_bt->num_free_blocks > 0) {
(void) get_next_blockaddr(agno, 0, bcnt_bt);
j++;
}
if (j > 0) {
if (j == lostblocks)
do_warn(_("lost %d blocks in ag %u\n"),
j, agno);
else
do_warn(_("thought we were going to lose %d "
"blocks in ag %u, actually lost "
"%d\n"),
lostblocks, j, agno);
}
agf->agf_flfirst = 0;
agf->agf_fllast = cpu_to_be32(i - 1);
agf->agf_flcount = cpu_to_be32(i);
#ifdef XR_BLD_FREE_TRACE
fprintf(stderr, "writing agfl for ag %u\n", agno);
#endif
} else {
agf->agf_flfirst = 0;
agf->agf_fllast = cpu_to_be32(XFS_AGFL_SIZE(mp) - 1);
agf->agf_flcount = 0;
}
libxfs_writebuf(agfl_buf, 0);
ext_ptr = findbiggest_bcnt_extent(agno);
agf->agf_longest = cpu_to_be32((ext_ptr != NULL) ?
ext_ptr->ex_blockcount : 0);
ASSERT(be32_to_cpu(agf->agf_roots[XFS_BTNUM_BNOi]) !=
be32_to_cpu(agf->agf_roots[XFS_BTNUM_CNTi]));
libxfs_writebuf(agf_buf, 0);
/*
* now fix up the free list appropriately
* XXX: code lifted from mkfs, should be shared.
*/
{
xfs_alloc_arg_t args;
xfs_trans_t *tp;
struct xfs_trans_res tres = {0};
int error;
memset(&args, 0, sizeof(args));
args.tp = tp = libxfs_trans_alloc(mp, 0);
args.mp = mp;
args.agno = agno;
args.alignment = 1;
args.pag = xfs_perag_get(mp,agno);
libxfs_trans_reserve(tp, &tres,
xfs_alloc_min_freelist(mp, args.pag), 0);
error = libxfs_alloc_fix_freelist(&args, 0);
xfs_perag_put(args.pag);
if (error) {
do_error(_("failed to fix AGFL on AG %d, error %d\n"),
agno, error);
}
libxfs_trans_commit(tp);
}
#ifdef XR_BLD_FREE_TRACE
fprintf(stderr, "wrote agf for ag %u\n", agno);
#endif
}
/*
* Scan the AGs starting at startag looking for an AG that isn't in use and has
* at least minlen blocks free.
*/
static int
xfs_filestream_pick_ag(
struct xfs_inode *ip,
xfs_agnumber_t startag,
xfs_agnumber_t *agp,
int flags,
xfs_extlen_t minlen)
{
struct xfs_mount *mp = ip->i_mount;
struct xfs_fstrm_item *item;
struct xfs_perag *pag;
xfs_extlen_t longest, free = 0, minfree, maxfree = 0;
xfs_agnumber_t ag, max_ag = NULLAGNUMBER;
int err, trylock, nscan;
ASSERT(S_ISDIR(VFS_I(ip)->i_mode));
/* 2% of an AG's blocks must be free for it to be chosen. */
minfree = mp->m_sb.sb_agblocks / 50;
ag = startag;
*agp = NULLAGNUMBER;
/* For the first pass, don't sleep trying to init the per-AG. */
trylock = XFS_ALLOC_FLAG_TRYLOCK;
for (nscan = 0; 1; nscan++) {
trace_xfs_filestream_scan(ip, ag);
pag = xfs_perag_get(mp, ag);
if (!pag->pagf_init) {
err = xfs_alloc_pagf_init(mp, NULL, ag, trylock);
if (err && !trylock) {
xfs_perag_put(pag);
return err;
}
}
/* Might fail sometimes during the 1st pass with trylock set. */
if (!pag->pagf_init)
goto next_ag;
/* Keep track of the AG with the most free blocks. */
if (pag->pagf_freeblks > maxfree) {
maxfree = pag->pagf_freeblks;
max_ag = ag;
}
/*
* The AG reference count does two things: it enforces mutual
* exclusion when examining the suitability of an AG in this
* loop, and it guards against two filestreams being established
* in the same AG as each other.
*/
if (xfs_filestream_get_ag(mp, ag) > 1) {
xfs_filestream_put_ag(mp, ag);
goto next_ag;
}
longest = xfs_alloc_longest_free_extent(mp, pag,
xfs_alloc_min_freelist(mp, pag));
if (((minlen && longest >= minlen) ||
(!minlen && pag->pagf_freeblks >= minfree)) &&
(!pag->pagf_metadata || !(flags & XFS_PICK_USERDATA) ||
(flags & XFS_PICK_LOWSPACE))) {
/* Break out, retaining the reference on the AG. */
free = pag->pagf_freeblks;
xfs_perag_put(pag);
*agp = ag;
break;
}
/* Drop the reference on this AG, it's not usable. */
xfs_filestream_put_ag(mp, ag);
next_ag:
xfs_perag_put(pag);
/* Move to the next AG, wrapping to AG 0 if necessary. */
if (++ag >= mp->m_sb.sb_agcount)
ag = 0;
/* If a full pass of the AGs hasn't been done yet, continue. */
if (ag != startag)
continue;
/* Allow sleeping in xfs_alloc_pagf_init() on the 2nd pass. */
if (trylock != 0) {
trylock = 0;
continue;
}
/* Finally, if lowspace wasn't set, set it for the 3rd pass. */
if (!(flags & XFS_PICK_LOWSPACE)) {
flags |= XFS_PICK_LOWSPACE;
continue;
}
/*
* Take the AG with the most free space, regardless of whether
* it's already in use by another filestream.
*/
if (max_ag != NULLAGNUMBER) {
xfs_filestream_get_ag(mp, max_ag);
free = maxfree;
*agp = max_ag;
break;
}
/* take AG 0 if none matched */
trace_xfs_filestream_pick(ip, *agp, free, nscan);
*agp = 0;
return 0;
}
trace_xfs_filestream_pick(ip, *agp, free, nscan);
if (*agp == NULLAGNUMBER)
return 0;
err = -ENOMEM;
item = kmem_alloc(sizeof(*item), KM_MAYFAIL);
if (!item)
goto out_put_ag;
item->ag = *agp;
item->ip = ip;
err = xfs_mru_cache_insert(mp->m_filestream, ip->i_ino, &item->mru);
if (err) {
if (err == -EEXIST)
err = 0;
goto out_free_item;
}
return 0;
out_free_item:
kmem_free(item);
out_put_ag:
xfs_filestream_put_ag(mp, *agp);
return err;
}