STATIC int
xfs_qm_flush_one(
struct xfs_dquot *dqp,
void *data)
{
struct list_head *buffer_list = data;
struct xfs_buf *bp = NULL;
int error = 0;
xfs_dqlock(dqp);
if (dqp->dq_flags & XFS_DQ_FREEING)
goto out_unlock;
if (!XFS_DQ_IS_DIRTY(dqp))
goto out_unlock;
xfs_dqflock(dqp);
error = xfs_qm_dqflush(dqp, &bp);
if (error)
goto out_unlock;
xfs_buf_delwri_queue(bp, buffer_list);
xfs_buf_relse(bp);
out_unlock:
xfs_dqunlock(dqp);
return error;
}
STATIC uint
xfs_buf_item_push(
struct xfs_log_item *lip,
struct list_head *buffer_list)
{
struct xfs_buf_log_item *bip = BUF_ITEM(lip);
struct xfs_buf *bp = bip->bli_buf;
uint rval = XFS_ITEM_SUCCESS;
if (xfs_buf_ispinned(bp))
return XFS_ITEM_PINNED;
if (!xfs_buf_trylock(bp)) {
/*
* If we have just raced with a buffer being pinned and it has
* been marked stale, we could end up stalling until someone else
* issues a log force to unpin the stale buffer. Check for the
* race condition here so xfsaild recognizes the buffer is pinned
* and queues a log force to move it along.
*/
if (xfs_buf_ispinned(bp))
return XFS_ITEM_PINNED;
return XFS_ITEM_LOCKED;
}
ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
trace_xfs_buf_item_push(bip);
if (!xfs_buf_delwri_queue(bp, buffer_list))
rval = XFS_ITEM_FLUSHING;
xfs_buf_unlock(bp);
return rval;
}
STATIC int
xfs_qm_dqiter_bufs(
struct xfs_mount *mp,
xfs_dqid_t firstid,
xfs_fsblock_t bno,
xfs_filblks_t blkcnt,
uint flags,
struct list_head *buffer_list)
{
struct xfs_buf *bp;
int error;
int type;
ASSERT(blkcnt > 0);
type = flags & XFS_QMOPT_UQUOTA ? XFS_DQ_USER :
(flags & XFS_QMOPT_PQUOTA ? XFS_DQ_PROJ : XFS_DQ_GROUP);
error = 0;
/*
* Blkcnt arg can be a very big number, and might even be
* larger than the log itself. So, we have to break it up into
* manageable-sized transactions.
* Note that we don't start a permanent transaction here; we might
* not be able to get a log reservation for the whole thing up front,
* and we don't really care to either, because we just discard
* everything if we were to crash in the middle of this loop.
*/
while (blkcnt--) {
error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp,
XFS_FSB_TO_DADDR(mp, bno),
mp->m_quotainfo->qi_dqchunklen, 0, &bp,
&xfs_dquot_buf_ops);
if (error)
break;
xfs_qm_reset_dqcounts(mp, bp, firstid, type);
xfs_buf_delwri_queue(bp, buffer_list);
xfs_buf_relse(bp);
/*
* goto the next block.
*/
bno++;
firstid += mp->m_quotainfo->qi_dqperchunk;
}
return error;
}
static int
xfs_ag_init_hdr(
struct xfs_mount *mp,
struct aghdr_init_data *id,
aghdr_init_work_f work,
const struct xfs_buf_ops *ops)
{
struct xfs_buf *bp;
bp = xfs_get_aghdr_buf(mp, id->daddr, id->numblks, 0, ops);
if (!bp)
return -ENOMEM;
(*work)(mp, bp, id);
xfs_buf_delwri_queue(bp, &id->buffer_list);
xfs_buf_relse(bp);
return 0;
}
STATIC uint
xfs_buf_item_push(
struct xfs_log_item *lip,
struct list_head *buffer_list)
{
struct xfs_buf_log_item *bip = BUF_ITEM(lip);
struct xfs_buf *bp = bip->bli_buf;
uint rval = XFS_ITEM_SUCCESS;
if (xfs_buf_ispinned(bp))
return XFS_ITEM_PINNED;
if (!xfs_buf_trylock(bp)) {
/*
* If we have just raced with a buffer being pinned and it has
* been marked stale, we could end up stalling until someone else
* issues a log force to unpin the stale buffer. Check for the
* race condition here so xfsaild recognizes the buffer is pinned
* and queues a log force to move it along.
*/
if (xfs_buf_ispinned(bp))
return XFS_ITEM_PINNED;
return XFS_ITEM_LOCKED;
}
ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
trace_xfs_buf_item_push(bip);
/* has a previous flush failed due to IO errors? */
if ((bp->b_flags & XBF_WRITE_FAIL) &&
___ratelimit(&xfs_buf_write_fail_rl_state, "XFS:")) {
xfs_warn(bp->b_target->bt_mount,
"Detected failing async write on buffer block 0x%llx. Retrying async write.\n",
(long long)bp->b_bn);
}
if (!xfs_buf_delwri_queue(bp, buffer_list))
rval = XFS_ITEM_FLUSHING;
xfs_buf_unlock(bp);
return rval;
}
STATIC uint
xfs_buf_item_push(
struct xfs_log_item *lip,
struct list_head *buffer_list)
{
struct xfs_buf_log_item *bip = BUF_ITEM(lip);
struct xfs_buf *bp = bip->bli_buf;
uint rval = XFS_ITEM_SUCCESS;
if (xfs_buf_ispinned(bp))
return XFS_ITEM_PINNED;
if (!xfs_buf_trylock(bp))
return XFS_ITEM_LOCKED;
ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
trace_xfs_buf_item_push(bip);
if (!xfs_buf_delwri_queue(bp, buffer_list))
rval = XFS_ITEM_FLUSHING;
xfs_buf_unlock(bp);
return rval;
}
STATIC void
xfs_qm_dqreclaim_one(
struct xfs_dquot *dqp,
struct list_head *buffer_list,
struct list_head *dispose_list)
{
struct xfs_mount *mp = dqp->q_mount;
struct xfs_quotainfo *qi = mp->m_quotainfo;
int error;
if (!xfs_dqlock_nowait(dqp))
goto out_move_tail;
/*
* This dquot has acquired a reference in the meantime remove it from
* the freelist and try again.
*/
if (dqp->q_nrefs) {
xfs_dqunlock(dqp);
trace_xfs_dqreclaim_want(dqp);
XFS_STATS_INC(xs_qm_dqwants);
list_del_init(&dqp->q_lru);
qi->qi_lru_count--;
XFS_STATS_DEC(xs_qm_dquot_unused);
return;
}
/*
* Try to grab the flush lock. If this dquot is in the process of
* getting flushed to disk, we don't want to reclaim it.
*/
if (!xfs_dqflock_nowait(dqp))
goto out_unlock_move_tail;
if (XFS_DQ_IS_DIRTY(dqp)) {
struct xfs_buf *bp = NULL;
trace_xfs_dqreclaim_dirty(dqp);
error = xfs_qm_dqflush(dqp, &bp);
if (error) {
xfs_warn(mp, "%s: dquot %p flush failed",
__func__, dqp);
goto out_unlock_move_tail;
}
xfs_buf_delwri_queue(bp, buffer_list);
xfs_buf_relse(bp);
/*
* Give the dquot another try on the freelist, as the
* flushing will take some time.
*/
goto out_unlock_move_tail;
}
xfs_dqfunlock(dqp);
/*
* Prevent lookups now that we are past the point of no return.
*/
dqp->dq_flags |= XFS_DQ_FREEING;
xfs_dqunlock(dqp);
ASSERT(dqp->q_nrefs == 0);
list_move_tail(&dqp->q_lru, dispose_list);
qi->qi_lru_count--;
XFS_STATS_DEC(xs_qm_dquot_unused);
trace_xfs_dqreclaim_done(dqp);
XFS_STATS_INC(xs_qm_dqreclaims);
return;
/*
* Move the dquot to the tail of the list so that we don't spin on it.
*/
out_unlock_move_tail:
xfs_dqunlock(dqp);
out_move_tail:
list_move_tail(&dqp->q_lru, &qi->qi_lru_list);
trace_xfs_dqreclaim_busy(dqp);
XFS_STATS_INC(xs_qm_dqreclaim_misses);
}
/*
* Initialise a new set of inodes. When called without a transaction context
* (e.g. from recovery) we initiate a delayed write of the inode buffers rather
* than logging them (which in a transaction context puts them into the AIL
* for writeback rather than the xfsbufd queue).
*/
int
xfs_ialloc_inode_init(
struct xfs_mount *mp,
struct xfs_trans *tp,
struct list_head *buffer_list,
xfs_agnumber_t agno,
xfs_agblock_t agbno,
xfs_agblock_t length,
unsigned int gen)
{
struct xfs_buf *fbuf;
struct xfs_dinode *free;
int blks_per_cluster, nbufs, ninodes;
int version;
int i, j;
xfs_daddr_t d;
xfs_ino_t ino = 0;
/*
* Loop over the new block(s), filling in the inodes.
* For small block sizes, manipulate the inodes in buffers
* which are multiples of the blocks size.
*/
if (mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(mp)) {
blks_per_cluster = 1;
nbufs = length;
ninodes = mp->m_sb.sb_inopblock;
} else {
blks_per_cluster = XFS_INODE_CLUSTER_SIZE(mp) /
mp->m_sb.sb_blocksize;
nbufs = length / blks_per_cluster;
ninodes = blks_per_cluster * mp->m_sb.sb_inopblock;
}
/*
* Figure out what version number to use in the inodes we create. If
* the superblock version has caught up to the one that supports the new
* inode format, then use the new inode version. Otherwise use the old
* version so that old kernels will continue to be able to use the file
* system.
*
* For v3 inodes, we also need to write the inode number into the inode,
* so calculate the first inode number of the chunk here as
* XFS_OFFBNO_TO_AGINO() only works within a filesystem block, not
* across multiple filesystem blocks (such as a cluster) and so cannot
* be used in the cluster buffer loop below.
*
* Further, because we are writing the inode directly into the buffer
* and calculating a CRC on the entire inode, we have ot log the entire
* inode so that the entire range the CRC covers is present in the log.
* That means for v3 inode we log the entire buffer rather than just the
* inode cores.
*/
if (xfs_sb_version_hascrc(&mp->m_sb)) {
version = 3;
ino = XFS_AGINO_TO_INO(mp, agno,
XFS_OFFBNO_TO_AGINO(mp, agbno, 0));
/*
* log the initialisation that is about to take place as an
* logical operation. This means the transaction does not
* need to log the physical changes to the inode buffers as log
* recovery will know what initialisation is actually needed.
* Hence we only need to log the buffers as "ordered" buffers so
* they track in the AIL as if they were physically logged.
*/
if (tp)
xfs_icreate_log(tp, agno, agbno, XFS_IALLOC_INODES(mp),
mp->m_sb.sb_inodesize, length, gen);
} else if (xfs_sb_version_hasnlink(&mp->m_sb))
version = 2;
else
version = 1;
for (j = 0; j < nbufs; j++) {
/*
* Get the block.
*/
d = XFS_AGB_TO_DADDR(mp, agno, agbno + (j * blks_per_cluster));
fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
mp->m_bsize * blks_per_cluster,
XBF_UNMAPPED);
if (!fbuf)
return ENOMEM;
/* Initialize the inode buffers and log them appropriately. */
fbuf->b_ops = &xfs_inode_buf_ops;
xfs_buf_zero(fbuf, 0, BBTOB(fbuf->b_length));
for (i = 0; i < ninodes; i++) {
int ioffset = i << mp->m_sb.sb_inodelog;
uint isize = xfs_dinode_size(version);
free = xfs_make_iptr(mp, fbuf, i);
free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC);
free->di_version = version;
free->di_gen = cpu_to_be32(gen);
free->di_next_unlinked = cpu_to_be32(NULLAGINO);
if (version == 3) {
free->di_ino = cpu_to_be64(ino);
ino++;
uuid_copy(&free->di_uuid, &mp->m_sb.sb_uuid);
xfs_dinode_calc_crc(mp, free);
} else if (tp) {
/* just log the inode core */
xfs_trans_log_buf(tp, fbuf, ioffset,
ioffset + isize - 1);
}
}
if (tp) {
/*
* Mark the buffer as an inode allocation buffer so it
* sticks in AIL at the point of this allocation
* transaction. This ensures the they are on disk before
* the tail of the log can be moved past this
* transaction (i.e. by preventing relogging from moving
* it forward in the log).
*/
xfs_trans_inode_alloc_buf(tp, fbuf);
if (version == 3) {
/*
* Mark the buffer as ordered so that they are
* not physically logged in the transaction but
* still tracked in the AIL as part of the
* transaction and pin the log appropriately.
*/
xfs_trans_ordered_buf(tp, fbuf);
xfs_trans_log_buf(tp, fbuf, 0,
BBTOB(fbuf->b_length) - 1);
}
} else {
fbuf->b_flags |= XBF_DONE;
xfs_buf_delwri_queue(fbuf, buffer_list);
xfs_buf_relse(fbuf);
}
}
return 0;
}
STATIC int
xfs_qm_dqiter_bufs(
struct xfs_mount *mp,
xfs_dqid_t firstid,
xfs_fsblock_t bno,
xfs_filblks_t blkcnt,
uint flags,
struct list_head *buffer_list)
{
struct xfs_buf *bp;
int error;
int type;
ASSERT(blkcnt > 0);
type = flags & XFS_QMOPT_UQUOTA ? XFS_DQ_USER :
(flags & XFS_QMOPT_PQUOTA ? XFS_DQ_PROJ : XFS_DQ_GROUP);
error = 0;
/*
* Blkcnt arg can be a very big number, and might even be
* larger than the log itself. So, we have to break it up into
* manageable-sized transactions.
* Note that we don't start a permanent transaction here; we might
* not be able to get a log reservation for the whole thing up front,
* and we don't really care to either, because we just discard
* everything if we were to crash in the middle of this loop.
*/
while (blkcnt--) {
error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp,
XFS_FSB_TO_DADDR(mp, bno),
mp->m_quotainfo->qi_dqchunklen, 0, &bp,
&xfs_dquot_buf_ops);
/*
* CRC and validation errors will return a EFSCORRUPTED here. If
* this occurs, re-read without CRC validation so that we can
* repair the damage via xfs_qm_reset_dqcounts(). This process
* will leave a trace in the log indicating corruption has
* been detected.
*/
if (error == -EFSCORRUPTED) {
error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp,
XFS_FSB_TO_DADDR(mp, bno),
mp->m_quotainfo->qi_dqchunklen, 0, &bp,
NULL);
}
if (error)
break;
/*
* A corrupt buffer might not have a verifier attached, so
* make sure we have the correct one attached before writeback
* occurs.
*/
bp->b_ops = &xfs_dquot_buf_ops;
xfs_qm_reset_dqcounts(mp, bp, firstid, type);
xfs_buf_delwri_queue(bp, buffer_list);
xfs_buf_relse(bp);
/* goto the next block. */
bno++;
firstid += mp->m_quotainfo->qi_dqperchunk;
}
return error;
}
/*
* Write a modified dquot to disk.
* The dquot must be locked and the flush lock too taken by caller.
* The flush lock will not be unlocked until the dquot reaches the disk,
* but the dquot is free to be unlocked and modified by the caller
* in the interim. Dquot is still locked on return. This behavior is
* identical to that of inodes.
*/
int
xfs_qm_dqflush(
xfs_dquot_t *dqp,
uint flags)
{
struct xfs_mount *mp = dqp->q_mount;
struct xfs_buf *bp;
struct xfs_disk_dquot *ddqp;
int error;
ASSERT(XFS_DQ_IS_LOCKED(dqp));
ASSERT(!completion_done(&dqp->q_flush));
trace_xfs_dqflush(dqp);
/*
* If not dirty, or it's pinned and we are not supposed to block, nada.
*/
if (!XFS_DQ_IS_DIRTY(dqp) ||
((flags & SYNC_TRYLOCK) && atomic_read(&dqp->q_pincount) > 0)) {
xfs_dqfunlock(dqp);
return 0;
}
xfs_qm_dqunpin_wait(dqp);
/*
* This may have been unpinned because the filesystem is shutting
* down forcibly. If that's the case we must not write this dquot
* to disk, because the log record didn't make it to disk!
*/
if (XFS_FORCED_SHUTDOWN(mp)) {
dqp->dq_flags &= ~XFS_DQ_DIRTY;
xfs_dqfunlock(dqp);
return XFS_ERROR(EIO);
}
/*
* Get the buffer containing the on-disk dquot
*/
error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, dqp->q_blkno,
mp->m_quotainfo->qi_dqchunklen, 0, &bp);
if (error) {
ASSERT(error != ENOENT);
xfs_dqfunlock(dqp);
return error;
}
/*
* Calculate the location of the dquot inside the buffer.
*/
ddqp = bp->b_addr + dqp->q_bufoffset;
/*
* A simple sanity check in case we got a corrupted dquot..
*/
error = xfs_qm_dqcheck(mp, &dqp->q_core, be32_to_cpu(ddqp->d_id), 0,
XFS_QMOPT_DOWARN, "dqflush (incore copy)");
if (error) {
xfs_buf_relse(bp);
xfs_dqfunlock(dqp);
xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE);
return XFS_ERROR(EIO);
}
/* This is the only portion of data that needs to persist */
memcpy(ddqp, &dqp->q_core, sizeof(xfs_disk_dquot_t));
/*
* Clear the dirty field and remember the flush lsn for later use.
*/
dqp->dq_flags &= ~XFS_DQ_DIRTY;
xfs_trans_ail_copy_lsn(mp->m_ail, &dqp->q_logitem.qli_flush_lsn,
&dqp->q_logitem.qli_item.li_lsn);
/*
* Attach an iodone routine so that we can remove this dquot from the
* AIL and release the flush lock once the dquot is synced to disk.
*/
xfs_buf_attach_iodone(bp, xfs_qm_dqflush_done,
&dqp->q_logitem.qli_item);
/*
* If the buffer is pinned then push on the log so we won't
* get stuck waiting in the write for too long.
*/
if (xfs_buf_ispinned(bp)) {
trace_xfs_dqflush_force(dqp);
xfs_log_force(mp, 0);
}
if (flags & SYNC_WAIT)
error = xfs_bwrite(bp);
else
xfs_buf_delwri_queue(bp);
xfs_buf_relse(bp);
trace_xfs_dqflush_done(dqp);
/*
* dqp is still locked, but caller is free to unlock it now.
*/
return error;
}
STATIC uint
xfs_inode_item_push(
struct xfs_log_item *lip,
struct list_head *buffer_list)
{
struct xfs_inode_log_item *iip = INODE_ITEM(lip);
struct xfs_inode *ip = iip->ili_inode;
struct xfs_buf *bp = NULL;
uint rval = XFS_ITEM_SUCCESS;
int error;
if (xfs_ipincount(ip) > 0)
return XFS_ITEM_PINNED;
if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED))
return XFS_ITEM_LOCKED;
/*
* Re-check the pincount now that we stabilized the value by
* taking the ilock.
*/
if (xfs_ipincount(ip) > 0) {
rval = XFS_ITEM_PINNED;
goto out_unlock;
}
/*
* Stale inode items should force out the iclog.
*/
if (ip->i_flags & XFS_ISTALE) {
rval = XFS_ITEM_PINNED;
goto out_unlock;
}
/*
* Someone else is already flushing the inode. Nothing we can do
* here but wait for the flush to finish and remove the item from
* the AIL.
*/
if (!xfs_iflock_nowait(ip)) {
rval = XFS_ITEM_FLUSHING;
goto out_unlock;
}
ASSERT(iip->ili_fields != 0 || XFS_FORCED_SHUTDOWN(ip->i_mount));
ASSERT(iip->ili_logged == 0 || XFS_FORCED_SHUTDOWN(ip->i_mount));
spin_unlock(&lip->li_ailp->xa_lock);
error = xfs_iflush(ip, &bp);
if (!error) {
if (!xfs_buf_delwri_queue(bp, buffer_list))
rval = XFS_ITEM_FLUSHING;
xfs_buf_relse(bp);
}
spin_lock(&lip->li_ailp->xa_lock);
out_unlock:
xfs_iunlock(ip, XFS_ILOCK_SHARED);
return rval;
}
