void *
kmem_zone_zalloc(kmem_zone_t *zone, int flags)
{
void *ptr = kmem_zone_alloc(zone, flags);
memset(ptr, 0, zone->zone_unitsize);
return ptr;
}
/*
* ktrace_alloc()
*
* Allocate a ktrace header and enough buffering for the given
* number of entries.
*/
ktrace_t *
ktrace_alloc(int nentries, int sleep)
{
ktrace_t *ktp;
ktrace_entry_t *ktep;
ktp = (ktrace_t*)kmem_zone_alloc(ktrace_hdr_zone, sleep);
if (ktp == (ktrace_t*)NULL) {
/*
* KM_SLEEP callers don't expect failure.
*/
if (sleep & KM_SLEEP)
panic("ktrace_alloc: NULL memory on KM_SLEEP request!");
return NULL;
}
/*
* Special treatment for buffers with the ktrace_zentries entries
*/
if (nentries == ktrace_zentries) {
ktep = (ktrace_entry_t*)kmem_zone_zalloc(ktrace_ent_zone,
sleep);
} else {
ktep = (ktrace_entry_t*)kmem_zalloc((nentries * sizeof(*ktep)),
sleep);
}
if (ktep == NULL) {
/*
* KM_SLEEP callers don't expect failure.
*/
if (sleep & KM_SLEEP)
panic("ktrace_alloc: NULL memory on KM_SLEEP request!");
kmem_free(ktp, sizeof(*ktp));
return NULL;
}
spinlock_init(&(ktp->kt_lock), "kt_lock");
ktp->kt_entries = ktep;
ktp->kt_nentries = nentries;
ktp->kt_index = 0;
ktp->kt_rollover = 0;
return ktp;
}
/*
* Allocate and initialise an xfs_inode.
*/
struct xfs_inode *
xfs_inode_alloc(
struct xfs_mount *mp,
xfs_ino_t ino)
{
struct xfs_inode *ip;
/*
* if this didn't occur in transactions, we could use
* KM_MAYFAIL and return NULL here on ENOMEM. Set the
* code up to do this anyway.
*/
ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP);
if (!ip)
return NULL;
if (inode_init_always(mp->m_super, VFS_I(ip))) {
kmem_zone_free(xfs_inode_zone, ip);
return NULL;
}
/* VFS doesn't initialise i_mode! */
VFS_I(ip)->i_mode = 0;
XFS_STATS_INC(mp, vn_active);
ASSERT(atomic_read(&ip->i_pincount) == 0);
ASSERT(!spin_is_locked(&ip->i_flags_lock));
ASSERT(!xfs_isiflocked(ip));
ASSERT(ip->i_ino == 0);
mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
/* initialise the xfs inode */
ip->i_ino = ino;
ip->i_mount = mp;
memset(&ip->i_imap, 0, sizeof(struct xfs_imap));
ip->i_afp = NULL;
ip->i_cowfp = NULL;
ip->i_cnextents = 0;
ip->i_cformat = XFS_DINODE_FMT_EXTENTS;
memset(&ip->i_df, 0, sizeof(xfs_ifork_t));
ip->i_flags = 0;
ip->i_delayed_blks = 0;
memset(&ip->i_d, 0, sizeof(ip->i_d));
return ip;
}
/*
* Allocate and initialise an xfs_inode.
*/
STATIC struct xfs_inode *
xfs_inode_alloc(
struct xfs_mount *mp,
xfs_ino_t ino)
{
struct xfs_inode *ip;
/*
* if this didn't occur in transactions, we could use
* KM_MAYFAIL and return NULL here on ENOMEM. Set the
* code up to do this anyway.
*/
ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP);
if (!ip)
return NULL;
if (inode_init_always(mp->m_super, VFS_I(ip))) {
kmem_zone_free(xfs_inode_zone, ip);
return NULL;
}
ASSERT(atomic_read(&ip->i_iocount) == 0);
ASSERT(atomic_read(&ip->i_pincount) == 0);
ASSERT(!spin_is_locked(&ip->i_flags_lock));
ASSERT(completion_done(&ip->i_flush));
ASSERT(ip->i_ino == 0);
mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
&xfs_iolock_active, "xfs_iolock_active");
/* initialise the xfs inode */
ip->i_ino = ino;
ip->i_mount = mp;
memset(&ip->i_imap, 0, sizeof(struct xfs_imap));
ip->i_afp = NULL;
memset(&ip->i_df, 0, sizeof(xfs_ifork_t));
ip->i_flags = 0;
ip->i_update_core = 0;
ip->i_delayed_blks = 0;
memset(&ip->i_d, 0, sizeof(xfs_icdinode_t));
ip->i_size = 0;
ip->i_new_size = 0;
return ip;
}
STATIC struct xfs_inode *
xfs_inode_alloc(
struct xfs_mount *mp,
xfs_ino_t ino)
{
struct xfs_inode *ip;
ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP);
if (!ip)
return NULL;
if (inode_init_always(mp->m_super, VFS_I(ip))) {
kmem_zone_free(xfs_inode_zone, ip);
return NULL;
}
ASSERT(atomic_read(&ip->i_pincount) == 0);
ASSERT(!spin_is_locked(&ip->i_flags_lock));
ASSERT(!xfs_isiflocked(ip));
ASSERT(ip->i_ino == 0);
mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
&xfs_iolock_active, "xfs_iolock_active");
ip->i_ino = ino;
ip->i_mount = mp;
memset(&ip->i_imap, 0, sizeof(struct xfs_imap));
ip->i_afp = NULL;
memset(&ip->i_df, 0, sizeof(xfs_ifork_t));
ip->i_flags = 0;
ip->i_delayed_blks = 0;
memset(&ip->i_d, 0, sizeof(xfs_icdinode_t));
return ip;
}
/*
* Allocate and initialise an xfs_inode.
*/
STATIC struct xfs_inode *
xfs_inode_alloc(
struct xfs_mount *mp,
xfs_ino_t ino)
{
struct xfs_inode *ip;
/*
* if this didn't occur in transactions, we could use
* KM_MAYFAIL and return NULL here on ENOMEM. Set the
* code up to do this anyway.
*/
ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP);
if (!ip)
return NULL;
ASSERT(atomic_read(&ip->i_iocount) == 0);
ASSERT(atomic_read(&ip->i_pincount) == 0);
ASSERT(!spin_is_locked(&ip->i_flags_lock));
ASSERT(completion_done(&ip->i_flush));
/* initialise the xfs inode */
ip->i_ino = ino;
ip->i_mount = mp;
memset(&ip->i_imap, 0, sizeof(struct xfs_imap));
ip->i_afp = NULL;
memset(&ip->i_df, 0, sizeof(xfs_ifork_t));
ip->i_flags = 0;
ip->i_update_core = 0;
ip->i_update_size = 0;
ip->i_delayed_blks = 0;
memset(&ip->i_d, 0, sizeof(xfs_icdinode_t));
ip->i_size = 0;
ip->i_new_size = 0;
/*
* Initialize inode's trace buffers.
*/
#ifdef XFS_INODE_TRACE
ip->i_trace = ktrace_alloc(INODE_TRACE_SIZE, KM_NOFS);
#endif
#ifdef XFS_BMAP_TRACE
ip->i_xtrace = ktrace_alloc(XFS_BMAP_KTRACE_SIZE, KM_NOFS);
#endif
#ifdef XFS_BTREE_TRACE
ip->i_btrace = ktrace_alloc(XFS_BMBT_KTRACE_SIZE, KM_NOFS);
#endif
#ifdef XFS_RW_TRACE
ip->i_rwtrace = ktrace_alloc(XFS_RW_KTRACE_SIZE, KM_NOFS);
#endif
#ifdef XFS_ILOCK_TRACE
ip->i_lock_trace = ktrace_alloc(XFS_ILOCK_KTRACE_SIZE, KM_NOFS);
#endif
#ifdef XFS_DIR2_TRACE
ip->i_dir_trace = ktrace_alloc(XFS_DIR2_KTRACE_SIZE, KM_NOFS);
#endif
/*
* Now initialise the VFS inode. We do this after the xfs_inode
* initialisation as internal failures will result in ->destroy_inode
* being called and that will pass down through the reclaim path and
* free the XFS inode. This path requires the XFS inode to already be
* initialised. Hence if this call fails, the xfs_inode has already
* been freed and we should not reference it at all in the error
* handling.
*/
if (!inode_init_always(mp->m_super, VFS_I(ip)))
return NULL;
/* prevent anyone from using this yet */
VFS_I(ip)->i_state = I_NEW|I_LOCK;
return ip;
}
/*
* Look up an inode by number in the given file system.
* The inode is looked up in the hash table for the file system
* represented by the mount point parameter mp. Each bucket of
* the hash table is guarded by an individual semaphore.
*
* If the inode is found in the hash table, its corresponding vnode
* is obtained with a call to vn_get(). This call takes care of
* coordination with the reclamation of the inode and vnode. Note
* that the vmap structure is filled in while holding the hash lock.
* This gives us the state of the inode/vnode when we found it and
* is used for coordination in vn_get().
*
* If it is not in core, read it in from the file system's device and
* add the inode into the hash table.
*
* The inode is locked according to the value of the lock_flags parameter.
* This flag parameter indicates how and if the inode's IO lock and inode lock
* should be taken.
*
* mp -- the mount point structure for the current file system. It points
* to the inode hash table.
* tp -- a pointer to the current transaction if there is one. This is
* simply passed through to the xfs_iread() call.
* ino -- the number of the inode desired. This is the unique identifier
* within the file system for the inode being requested.
* lock_flags -- flags indicating how to lock the inode. See the comment
* for xfs_ilock() for a list of valid values.
* bno -- the block number starting the buffer containing the inode,
* if known (as by bulkstat), else 0.
*/
int
xfs_iget(
xfs_mount_t *mp,
xfs_trans_t *tp,
xfs_ino_t ino,
uint flags,
uint lock_flags,
xfs_inode_t **ipp,
xfs_daddr_t bno)
{
xfs_ihash_t *ih;
xfs_inode_t *ip;
xfs_inode_t *iq;
xfs_vnode_t *vp;
ulong version;
int error;
/* REFERENCED */
int newnode;
xfs_chash_t *ch;
xfs_chashlist_t *chl, *chlnew;
vmap_t vmap;
SPLDECL(s);
XFS_STATS_INC(xs_ig_attempts);
ih = XFS_IHASH(mp, ino);
again:
read_lock(&ih->ih_lock);
for (ip = ih->ih_next; ip != NULL; ip = ip->i_next) {
if (ip->i_ino == ino) {
vp = XFS_ITOV(ip);
VMAP(vp, vmap);
/*
* Inode cache hit: if ip is not at the front of
* its hash chain, move it there now.
* Do this with the lock held for update, but
* do statistics after releasing the lock.
*/
if (ip->i_prevp != &ih->ih_next
&& rwlock_trypromote(&ih->ih_lock)) {
if ((iq = ip->i_next)) {
iq->i_prevp = ip->i_prevp;
}
*ip->i_prevp = iq;
iq = ih->ih_next;
iq->i_prevp = &ip->i_next;
ip->i_next = iq;
ip->i_prevp = &ih->ih_next;
ih->ih_next = ip;
write_unlock(&ih->ih_lock);
} else {
read_unlock(&ih->ih_lock);
}
XFS_STATS_INC(xs_ig_found);
/*
* Get a reference to the vnode/inode.
* vn_get() takes care of coordination with
* the file system inode release and reclaim
* functions. If it returns NULL, the inode
* has been reclaimed so just start the search
* over again. We probably won't find it,
* but we could be racing with another cpu
* looking for the same inode so we have to at
* least look.
*/
if (!(vp = vn_get(vp, &vmap))) {
XFS_STATS_INC(xs_ig_frecycle);
goto again;
}
if (lock_flags != 0) {
ip->i_flags &= ~XFS_IRECLAIM;
xfs_ilock(ip, lock_flags);
}
newnode = (ip->i_d.di_mode == 0);
if (newnode) {
xfs_iocore_inode_reinit(ip);
}
ip->i_flags &= ~XFS_ISTALE;
vn_trace_exit(vp, "xfs_iget.found",
(inst_t *)__return_address);
goto return_ip;
}
}
/*
* Inode cache miss: save the hash chain version stamp and unlock
* the chain, so we don't deadlock in vn_alloc.
*/
XFS_STATS_INC(xs_ig_missed);
version = ih->ih_version;
read_unlock(&ih->ih_lock);
/*
* Read the disk inode attributes into a new inode structure and get
* a new vnode for it. This should also initialize i_ino and i_mount.
*/
error = xfs_iread(mp, tp, ino, &ip, bno);
if (error) {
return error;
}
error = xfs_vn_allocate(mp, ip, &vp);
if (error) {
return error;
}
vn_trace_exit(vp, "xfs_iget.alloc", (inst_t *)__return_address);
xfs_inode_lock_init(ip, vp);
xfs_iocore_inode_init(ip);
if (lock_flags != 0) {
xfs_ilock(ip, lock_flags);
}
/*
* Put ip on its hash chain, unless someone else hashed a duplicate
* after we released the hash lock.
*/
write_lock(&ih->ih_lock);
if (ih->ih_version != version) {
for (iq = ih->ih_next; iq != NULL; iq = iq->i_next) {
if (iq->i_ino == ino) {
write_unlock(&ih->ih_lock);
xfs_idestroy(ip);
XFS_STATS_INC(xs_ig_dup);
goto again;
}
}
}
/*
* These values _must_ be set before releasing ihlock!
*/
ip->i_hash = ih;
if ((iq = ih->ih_next)) {
iq->i_prevp = &ip->i_next;
}
ip->i_next = iq;
ip->i_prevp = &ih->ih_next;
ih->ih_next = ip;
ip->i_udquot = ip->i_gdquot = NULL;
ih->ih_version++;
write_unlock(&ih->ih_lock);
/*
* put ip on its cluster's hash chain
*/
ASSERT(ip->i_chash == NULL && ip->i_cprev == NULL &&
ip->i_cnext == NULL);
chlnew = NULL;
ch = XFS_CHASH(mp, ip->i_blkno);
chlredo:
s = mutex_spinlock(&ch->ch_lock);
for (chl = ch->ch_list; chl != NULL; chl = chl->chl_next) {
if (chl->chl_blkno == ip->i_blkno) {
/* insert this inode into the doubly-linked list
* where chl points */
if ((iq = chl->chl_ip)) {
ip->i_cprev = iq->i_cprev;
iq->i_cprev->i_cnext = ip;
iq->i_cprev = ip;
ip->i_cnext = iq;
} else {
ip->i_cnext = ip;
ip->i_cprev = ip;
}
chl->chl_ip = ip;
ip->i_chash = chl;
break;
}
}
/* no hash list found for this block; add a new hash list */
if (chl == NULL) {
if (chlnew == NULL) {
mutex_spinunlock(&ch->ch_lock, s);
ASSERT(xfs_chashlist_zone != NULL);
chlnew = (xfs_chashlist_t *)
kmem_zone_alloc(xfs_chashlist_zone,
KM_SLEEP);
ASSERT(chlnew != NULL);
goto chlredo;
} else {
ip->i_cnext = ip;
ip->i_cprev = ip;
ip->i_chash = chlnew;
chlnew->chl_ip = ip;
chlnew->chl_blkno = ip->i_blkno;
chlnew->chl_next = ch->ch_list;
ch->ch_list = chlnew;
chlnew = NULL;
}
} else {
if (chlnew != NULL) {
kmem_zone_free(xfs_chashlist_zone, chlnew);
}
}
mutex_spinunlock(&ch->ch_lock, s);
/*
* Link ip to its mount and thread it on the mount's inode list.
*/
XFS_MOUNT_ILOCK(mp);
if ((iq = mp->m_inodes)) {
ASSERT(iq->i_mprev->i_mnext == iq);
ip->i_mprev = iq->i_mprev;
iq->i_mprev->i_mnext = ip;
iq->i_mprev = ip;
ip->i_mnext = iq;
} else {
ip->i_mnext = ip;
ip->i_mprev = ip;
}
mp->m_inodes = ip;
XFS_MOUNT_IUNLOCK(mp);
newnode = 1;
return_ip:
ASSERT(ip->i_df.if_ext_max ==
XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t));
ASSERT(((ip->i_d.di_flags & XFS_DIFLAG_REALTIME) != 0) ==
((ip->i_iocore.io_flags & XFS_IOCORE_RT) != 0));
*ipp = ip;
/*
* If we have a real type for an on-disk inode, we can set ops(&unlock)
* now. If it's a new inode being created, xfs_ialloc will handle it.
*/
XVFS_INIT_VNODE(XFS_MTOVFS(mp), vp, XFS_ITOBHV(ip), 1);
return 0;
}
