#include <linux/delay.h>
#include <linux/sort.h>
#include <linux/jhash.h>
-#include <linux/kref.h>
#include <linux/kallsyms.h>
#include <linux/gfs2_ondisk.h>
+#include <linux/list.h>
+#include <linux/lm_interface.h>
+#include <linux/wait.h>
+#include <linux/rwsem.h>
#include <asm/uaccess.h>
#include "gfs2.h"
-#include "lm_interface.h"
#include "incore.h"
#include "glock.h"
#include "glops.h"
#include "super.h"
#include "util.h"
-/* Must be kept in sync with the beginning of struct gfs2_glock */
-struct glock_plug {
- struct list_head gl_list;
- unsigned long gl_flags;
-};
-
-struct greedy {
- struct gfs2_holder gr_gh;
- struct work_struct gr_work;
+struct gfs2_gl_hash_bucket {
+ struct hlist_head hb_list;
};
typedef void (*glock_examiner) (struct gfs2_glock * gl);
static int gfs2_dump_lockstate(struct gfs2_sbd *sdp);
static int dump_glock(struct gfs2_glock *gl);
+static int dump_inode(struct gfs2_inode *ip);
+static void gfs2_glock_xmote_th(struct gfs2_holder *gh);
+static void gfs2_glock_drop_th(struct gfs2_glock *gl);
+static DECLARE_RWSEM(gfs2_umount_flush_sem);
+
+#define GFS2_GL_HASH_SHIFT 15
+#define GFS2_GL_HASH_SIZE (1 << GFS2_GL_HASH_SHIFT)
+#define GFS2_GL_HASH_MASK (GFS2_GL_HASH_SIZE - 1)
+
+static struct gfs2_gl_hash_bucket gl_hash_table[GFS2_GL_HASH_SIZE];
+
+/*
+ * Despite what you might think, the numbers below are not arbitrary :-)
+ * They are taken from the ipv4 routing hash code, which is well tested
+ * and thus should be nearly optimal. Later on we might tweek the numbers
+ * but for now this should be fine.
+ *
+ * The reason for putting the locks in a separate array from the list heads
+ * is that we can have fewer locks than list heads and save memory. We use
+ * the same hash function for both, but with a different hash mask.
+ */
+#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) || \
+ defined(CONFIG_PROVE_LOCKING)
+
+#ifdef CONFIG_LOCKDEP
+# define GL_HASH_LOCK_SZ 256
+#else
+# if NR_CPUS >= 32
+# define GL_HASH_LOCK_SZ 4096
+# elif NR_CPUS >= 16
+# define GL_HASH_LOCK_SZ 2048
+# elif NR_CPUS >= 8
+# define GL_HASH_LOCK_SZ 1024
+# elif NR_CPUS >= 4
+# define GL_HASH_LOCK_SZ 512
+# else
+# define GL_HASH_LOCK_SZ 256
+# endif
+#endif
+
+/* We never want more locks than chains */
+#if GFS2_GL_HASH_SIZE < GL_HASH_LOCK_SZ
+# undef GL_HASH_LOCK_SZ
+# define GL_HASH_LOCK_SZ GFS2_GL_HASH_SIZE
+#endif
+
+static rwlock_t gl_hash_locks[GL_HASH_LOCK_SZ];
+
+static inline rwlock_t *gl_lock_addr(unsigned int x)
+{
+ return &gl_hash_locks[x & (GL_HASH_LOCK_SZ-1)];
+}
+#else /* not SMP, so no spinlocks required */
+static inline rwlock_t *gl_lock_addr(unsigned int x)
+{
+ return NULL;
+}
+#endif
/**
* relaxed_state_ok - is a requested lock compatible with the current lock mode?
* Returns: The number of the corresponding hash bucket
*/
-static unsigned int gl_hash(const struct lm_lockname *name)
+static unsigned int gl_hash(const struct gfs2_sbd *sdp,
+ const struct lm_lockname *name)
{
unsigned int h;
h = jhash(&name->ln_number, sizeof(u64), 0);
h = jhash(&name->ln_type, sizeof(unsigned int), h);
+ h = jhash(&sdp, sizeof(struct gfs2_sbd *), h);
h &= GFS2_GL_HASH_MASK;
return h;
void gfs2_glock_hold(struct gfs2_glock *gl)
{
- kref_get(&gl->gl_ref);
-}
-
-/* All work is done after the return from kref_put() so we
- can release the write_lock before the free. */
-
-static void kill_glock(struct kref *kref)
-{
- struct gfs2_glock *gl = container_of(kref, struct gfs2_glock, gl_ref);
- struct gfs2_sbd *sdp = gl->gl_sbd;
-
- gfs2_assert(sdp, gl->gl_state == LM_ST_UNLOCKED);
- gfs2_assert(sdp, list_empty(&gl->gl_reclaim));
- gfs2_assert(sdp, list_empty(&gl->gl_holders));
- gfs2_assert(sdp, list_empty(&gl->gl_waiters1));
- gfs2_assert(sdp, list_empty(&gl->gl_waiters2));
- gfs2_assert(sdp, list_empty(&gl->gl_waiters3));
+ atomic_inc(&gl->gl_ref);
}
/**
int gfs2_glock_put(struct gfs2_glock *gl)
{
- struct gfs2_gl_hash_bucket *bucket = gl->gl_bucket;
int rv = 0;
+ struct gfs2_sbd *sdp = gl->gl_sbd;
- write_lock(&bucket->hb_lock);
- if (kref_put(&gl->gl_ref, kill_glock)) {
- list_del_init(&gl->gl_list);
- write_unlock(&bucket->hb_lock);
+ write_lock(gl_lock_addr(gl->gl_hash));
+ if (atomic_dec_and_test(&gl->gl_ref)) {
+ hlist_del(&gl->gl_list);
+ write_unlock(gl_lock_addr(gl->gl_hash));
BUG_ON(spin_is_locked(&gl->gl_spin));
+ gfs2_assert(sdp, gl->gl_state == LM_ST_UNLOCKED);
+ gfs2_assert(sdp, list_empty(&gl->gl_reclaim));
+ gfs2_assert(sdp, list_empty(&gl->gl_holders));
+ gfs2_assert(sdp, list_empty(&gl->gl_waiters1));
+ gfs2_assert(sdp, list_empty(&gl->gl_waiters2));
+ gfs2_assert(sdp, list_empty(&gl->gl_waiters3));
glock_free(gl);
rv = 1;
goto out;
}
- write_unlock(&bucket->hb_lock);
+ write_unlock(gl_lock_addr(gl->gl_hash));
out:
return rv;
}
-/**
- * queue_empty - check to see if a glock's queue is empty
- * @gl: the glock
- * @head: the head of the queue to check
- *
- * This function protects the list in the event that a process already
- * has a holder on the list and is adding a second holder for itself.
- * The glmutex lock is what generally prevents processes from working
- * on the same glock at once, but the special case of adding a second
- * holder for yourself ("recursive" locking) doesn't involve locking
- * glmutex, making the spin lock necessary.
- *
- * Returns: 1 if the queue is empty
- */
-
-static inline int queue_empty(struct gfs2_glock *gl, struct list_head *head)
-{
- int empty;
- spin_lock(&gl->gl_spin);
- empty = list_empty(head);
- spin_unlock(&gl->gl_spin);
- return empty;
-}
-
/**
* search_bucket() - Find struct gfs2_glock by lock number
* @bucket: the bucket to search
* Returns: NULL, or the struct gfs2_glock with the requested number
*/
-static struct gfs2_glock *search_bucket(struct gfs2_gl_hash_bucket *bucket,
+static struct gfs2_glock *search_bucket(unsigned int hash,
const struct gfs2_sbd *sdp,
const struct lm_lockname *name)
{
struct gfs2_glock *gl;
+ struct hlist_node *h;
- list_for_each_entry(gl, &bucket->hb_list, gl_list) {
- if (test_bit(GLF_PLUG, &gl->gl_flags))
- continue;
+ hlist_for_each_entry(gl, h, &gl_hash_table[hash].hb_list, gl_list) {
if (!lm_name_equal(&gl->gl_name, name))
continue;
if (gl->gl_sbd != sdp)
continue;
- kref_get(&gl->gl_ref);
+ atomic_inc(&gl->gl_ref);
return gl;
}
* Returns: NULL, or the struct gfs2_glock with the requested number
*/
-static struct gfs2_glock *gfs2_glock_find(struct gfs2_sbd *sdp,
+static struct gfs2_glock *gfs2_glock_find(const struct gfs2_sbd *sdp,
const struct lm_lockname *name)
{
- struct gfs2_gl_hash_bucket *bucket = &sdp->sd_gl_hash[gl_hash(name)];
+ unsigned int hash = gl_hash(sdp, name);
struct gfs2_glock *gl;
- read_lock(&bucket->hb_lock);
- gl = search_bucket(bucket, sdp, name);
- read_unlock(&bucket->hb_lock);
+ read_lock(gl_lock_addr(hash));
+ gl = search_bucket(hash, sdp, name);
+ read_unlock(gl_lock_addr(hash));
return gl;
}
const struct gfs2_glock_operations *glops, int create,
struct gfs2_glock **glp)
{
- struct lm_lockname name;
+ struct lm_lockname name = { .ln_number = number, .ln_type = glops->go_type };
struct gfs2_glock *gl, *tmp;
- struct gfs2_gl_hash_bucket *bucket;
+ unsigned int hash = gl_hash(sdp, &name);
int error;
- name.ln_number = number;
- name.ln_type = glops->go_type;
- bucket = &sdp->sd_gl_hash[gl_hash(&name)];
-
- read_lock(&bucket->hb_lock);
- gl = search_bucket(bucket, sdp, &name);
- read_unlock(&bucket->hb_lock);
+ read_lock(gl_lock_addr(hash));
+ gl = search_bucket(hash, sdp, &name);
+ read_unlock(gl_lock_addr(hash));
if (gl || !create) {
*glp = gl;
gl->gl_flags = 0;
gl->gl_name = name;
- kref_init(&gl->gl_ref);
+ atomic_set(&gl->gl_ref, 1);
gl->gl_state = LM_ST_UNLOCKED;
+ gl->gl_hash = hash;
gl->gl_owner = NULL;
gl->gl_ip = 0;
gl->gl_ops = glops;
gl->gl_vn = 0;
gl->gl_stamp = jiffies;
gl->gl_object = NULL;
- gl->gl_bucket = bucket;
gl->gl_sbd = sdp;
gl->gl_aspace = NULL;
lops_init_le(&gl->gl_le, &gfs2_glock_lops);
if (error)
goto fail_aspace;
- write_lock(&bucket->hb_lock);
- tmp = search_bucket(bucket, sdp, &name);
+ write_lock(gl_lock_addr(hash));
+ tmp = search_bucket(hash, sdp, &name);
if (tmp) {
- write_unlock(&bucket->hb_lock);
+ write_unlock(gl_lock_addr(hash));
glock_free(gl);
gl = tmp;
} else {
- list_add_tail(&gl->gl_list, &bucket->hb_list);
- write_unlock(&bucket->hb_lock);
+ hlist_add_head(&gl->gl_list, &gl_hash_table[hash].hb_list);
+ write_unlock(gl_lock_addr(hash));
}
*glp = gl;
if (gl->gl_aspace)
gfs2_aspace_put(gl->gl_aspace);
fail:
- kmem_cache_free(gfs2_glock_cachep, gl);
+ kmem_cache_free(gfs2_glock_cachep, gl);
return error;
}
gh->gh_flags = flags;
gh->gh_error = 0;
gh->gh_iflags = 0;
- init_completion(&gh->gh_wait);
-
- if (gh->gh_state == LM_ST_EXCLUSIVE)
- gh->gh_flags |= GL_LOCAL_EXCL;
-
gfs2_glock_hold(gl);
}
{
gh->gh_state = state;
gh->gh_flags = flags;
- if (gh->gh_state == LM_ST_EXCLUSIVE)
- gh->gh_flags |= GL_LOCAL_EXCL;
-
gh->gh_iflags &= 1 << HIF_ALLOCED;
gh->gh_ip = (unsigned long)__builtin_return_address(0);
}
kfree(gh);
}
+static void gfs2_holder_dispose_or_wake(struct gfs2_holder *gh)
+{
+ if (test_bit(HIF_DEALLOC, &gh->gh_iflags)) {
+ gfs2_holder_put(gh);
+ return;
+ }
+ clear_bit(HIF_WAIT, &gh->gh_iflags);
+ smp_mb();
+ wake_up_bit(&gh->gh_iflags, HIF_WAIT);
+}
+
+static int holder_wait(void *word)
+{
+ schedule();
+ return 0;
+}
+
+static void wait_on_holder(struct gfs2_holder *gh)
+{
+ might_sleep();
+ wait_on_bit(&gh->gh_iflags, HIF_WAIT, holder_wait, TASK_UNINTERRUPTIBLE);
+}
+
/**
* rq_mutex - process a mutex request in the queue
* @gh: the glock holder
list_del_init(&gh->gh_list);
/* gh->gh_error never examined. */
set_bit(GLF_LOCK, &gl->gl_flags);
- complete(&gh->gh_wait);
+ clear_bit(HIF_WAIT, &gh->gh_iflags);
+ smp_mb();
+ wake_up_bit(&gh->gh_iflags, HIF_WAIT);
return 1;
}
{
struct gfs2_glock *gl = gh->gh_gl;
struct gfs2_sbd *sdp = gl->gl_sbd;
- const struct gfs2_glock_operations *glops = gl->gl_ops;
if (!relaxed_state_ok(gl->gl_state, gh->gh_state, gh->gh_flags)) {
if (list_empty(&gl->gl_holders)) {
gfs2_reclaim_glock(sdp);
}
- glops->go_xmote_th(gl, gh->gh_state, gh->gh_flags);
+ gfs2_glock_xmote_th(gh);
spin_lock(&gl->gl_spin);
}
return 1;
set_bit(GLF_LOCK, &gl->gl_flags);
} else {
struct gfs2_holder *next_gh;
- if (gh->gh_flags & GL_LOCAL_EXCL)
+ if (gh->gh_state == LM_ST_EXCLUSIVE)
return 1;
next_gh = list_entry(gl->gl_holders.next, struct gfs2_holder,
gh_list);
- if (next_gh->gh_flags & GL_LOCAL_EXCL)
+ if (next_gh->gh_state == LM_ST_EXCLUSIVE)
return 1;
}
gh->gh_error = 0;
set_bit(HIF_HOLDER, &gh->gh_iflags);
- complete(&gh->gh_wait);
+ gfs2_holder_dispose_or_wake(gh);
return 0;
}
static int rq_demote(struct gfs2_holder *gh)
{
struct gfs2_glock *gl = gh->gh_gl;
- const struct gfs2_glock_operations *glops = gl->gl_ops;
if (!list_empty(&gl->gl_holders))
return 1;
list_del_init(&gh->gh_list);
gh->gh_error = 0;
spin_unlock(&gl->gl_spin);
- if (test_bit(HIF_DEALLOC, &gh->gh_iflags))
- gfs2_holder_put(gh);
- else
- complete(&gh->gh_wait);
+ gfs2_holder_dispose_or_wake(gh);
spin_lock(&gl->gl_spin);
} else {
gl->gl_req_gh = gh;
if (gh->gh_state == LM_ST_UNLOCKED ||
gl->gl_state != LM_ST_EXCLUSIVE)
- glops->go_drop_th(gl);
+ gfs2_glock_drop_th(gl);
else
- glops->go_xmote_th(gl, gh->gh_state, gh->gh_flags);
+ gfs2_glock_xmote_th(gh);
spin_lock(&gl->gl_spin);
}
return 0;
}
-/**
- * rq_greedy - process a queued request to drop greedy status
- * @gh: the glock holder
- *
- * Returns: 1 if the queue is blocked
- */
-
-static int rq_greedy(struct gfs2_holder *gh)
-{
- struct gfs2_glock *gl = gh->gh_gl;
-
- list_del_init(&gh->gh_list);
- /* gh->gh_error never examined. */
- clear_bit(GLF_GREEDY, &gl->gl_flags);
- spin_unlock(&gl->gl_spin);
-
- gfs2_holder_uninit(gh);
- kfree(container_of(gh, struct greedy, gr_gh));
-
- spin_lock(&gl->gl_spin);
-
- return 0;
-}
-
/**
* run_queue - process holder structures on a glock
* @gl: the glock
if (test_bit(HIF_DEMOTE, &gh->gh_iflags))
blocked = rq_demote(gh);
- else if (test_bit(HIF_GREEDY, &gh->gh_iflags))
- blocked = rq_greedy(gh);
else
gfs2_assert_warn(gl->gl_sbd, 0);
gfs2_holder_init(gl, 0, 0, &gh);
set_bit(HIF_MUTEX, &gh.gh_iflags);
+ if (test_and_set_bit(HIF_WAIT, &gh.gh_iflags))
+ BUG();
spin_lock(&gl->gl_spin);
- if (test_and_set_bit(GLF_LOCK, &gl->gl_flags))
+ if (test_and_set_bit(GLF_LOCK, &gl->gl_flags)) {
list_add_tail(&gh.gh_list, &gl->gl_waiters1);
- else {
+ } else {
gl->gl_owner = current;
gl->gl_ip = (unsigned long)__builtin_return_address(0);
- complete(&gh.gh_wait);
+ clear_bit(HIF_WAIT, &gh.gh_iflags);
+ smp_mb();
+ wake_up_bit(&gh.gh_iflags, HIF_WAIT);
}
spin_unlock(&gl->gl_spin);
- wait_for_completion(&gh.gh_wait);
+ wait_on_holder(&gh);
gfs2_holder_uninit(&gh);
}
int acquired = 1;
spin_lock(&gl->gl_spin);
- if (test_and_set_bit(GLF_LOCK, &gl->gl_flags))
+ if (test_and_set_bit(GLF_LOCK, &gl->gl_flags)) {
acquired = 0;
- else {
+ } else {
gl->gl_owner = current;
gl->gl_ip = (unsigned long)__builtin_return_address(0);
}
} else {
spin_unlock(&gl->gl_spin);
- new_gh = gfs2_holder_get(gl, state, LM_FLAG_TRY, GFP_KERNEL);
+ new_gh = gfs2_holder_get(gl, state, LM_FLAG_TRY, GFP_NOFS);
if (!new_gh)
return;
set_bit(HIF_DEMOTE, &new_gh->gh_iflags);
set_bit(HIF_DEALLOC, &new_gh->gh_iflags);
+ set_bit(HIF_WAIT, &new_gh->gh_iflags);
goto restart;
}
gfs2_holder_put(new_gh);
}
-void gfs2_glock_inode_squish(struct inode *inode)
-{
- struct gfs2_holder gh;
- struct gfs2_glock *gl = GFS2_I(inode)->i_gl;
- gfs2_holder_init(gl, LM_ST_UNLOCKED, 0, &gh);
- set_bit(HIF_DEMOTE, &gh.gh_iflags);
- spin_lock(&gl->gl_spin);
- gfs2_assert(inode->i_sb->s_fs_info, list_empty(&gl->gl_holders));
- list_add_tail(&gh.gh_list, &gl->gl_waiters2);
- run_queue(gl);
- spin_unlock(&gl->gl_spin);
- wait_for_completion(&gh.gh_wait);
- gfs2_holder_uninit(&gh);
-}
-
/**
* state_change - record that the glock is now in a different state
* @gl: the glock
int op_done = 1;
gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags));
- gfs2_assert_warn(sdp, queue_empty(gl, &gl->gl_holders));
+ gfs2_assert_warn(sdp, list_empty(&gl->gl_holders));
gfs2_assert_warn(sdp, !(ret & LM_OUT_ASYNC));
state_change(gl, ret & LM_OUT_ST_MASK);
if (prev_state != LM_ST_UNLOCKED && !(ret & LM_OUT_CACHEABLE)) {
if (glops->go_inval)
- glops->go_inval(gl, DIO_METADATA | DIO_DATA);
+ glops->go_inval(gl, DIO_METADATA);
} else if (gl->gl_state == LM_ST_DEFERRED) {
/* We might not want to do this here.
Look at moving to the inode glops. */
if (glops->go_inval)
- glops->go_inval(gl, DIO_DATA);
+ glops->go_inval(gl, 0);
}
/* Deal with each possible exit condition */
spin_lock(&gl->gl_spin);
list_del_init(&gh->gh_list);
if (gl->gl_state == gh->gh_state ||
- gl->gl_state == LM_ST_UNLOCKED)
+ gl->gl_state == LM_ST_UNLOCKED) {
gh->gh_error = 0;
- else {
+ } else {
if (gfs2_assert_warn(sdp, gh->gh_flags &
(LM_FLAG_TRY | LM_FLAG_TRY_1CB)) == -1)
fs_warn(sdp, "ret = 0x%.8X\n", ret);
gfs2_glock_put(gl);
- if (gh) {
- if (test_bit(HIF_DEALLOC, &gh->gh_iflags))
- gfs2_holder_put(gh);
- else
- complete(&gh->gh_wait);
- }
+ if (gh)
+ gfs2_holder_dispose_or_wake(gh);
}
/**
*
*/
-void gfs2_glock_xmote_th(struct gfs2_glock *gl, unsigned int state, int flags)
+void gfs2_glock_xmote_th(struct gfs2_holder *gh)
{
+ struct gfs2_glock *gl = gh->gh_gl;
struct gfs2_sbd *sdp = gl->gl_sbd;
+ int flags = gh->gh_flags;
+ unsigned state = gh->gh_state;
const struct gfs2_glock_operations *glops = gl->gl_ops;
int lck_flags = flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB |
LM_FLAG_NOEXP | LM_FLAG_ANY |
LM_FLAG_PRIORITY);
unsigned int lck_ret;
+ if (glops->go_xmote_th)
+ glops->go_xmote_th(gl);
+
gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags));
- gfs2_assert_warn(sdp, queue_empty(gl, &gl->gl_holders));
+ gfs2_assert_warn(sdp, list_empty(&gl->gl_holders));
gfs2_assert_warn(sdp, state != LM_ST_UNLOCKED);
gfs2_assert_warn(sdp, state != gl->gl_state);
- if (gl->gl_state == LM_ST_EXCLUSIVE && glops->go_sync)
- glops->go_sync(gl, DIO_METADATA | DIO_DATA | DIO_RELEASE);
-
gfs2_glock_hold(gl);
gl->gl_req_bh = xmote_bh;
const struct gfs2_glock_operations *glops = gl->gl_ops;
struct gfs2_holder *gh = gl->gl_req_gh;
- clear_bit(GLF_PREFETCH, &gl->gl_flags);
-
gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags));
- gfs2_assert_warn(sdp, queue_empty(gl, &gl->gl_holders));
+ gfs2_assert_warn(sdp, list_empty(&gl->gl_holders));
gfs2_assert_warn(sdp, !ret);
state_change(gl, LM_ST_UNLOCKED);
if (glops->go_inval)
- glops->go_inval(gl, DIO_METADATA | DIO_DATA);
+ glops->go_inval(gl, DIO_METADATA);
if (gh) {
spin_lock(&gl->gl_spin);
gfs2_glock_put(gl);
- if (gh) {
- if (test_bit(HIF_DEALLOC, &gh->gh_iflags))
- gfs2_holder_put(gh);
- else
- complete(&gh->gh_wait);
- }
+ if (gh)
+ gfs2_holder_dispose_or_wake(gh);
}
/**
*
*/
-void gfs2_glock_drop_th(struct gfs2_glock *gl)
+static void gfs2_glock_drop_th(struct gfs2_glock *gl)
{
struct gfs2_sbd *sdp = gl->gl_sbd;
const struct gfs2_glock_operations *glops = gl->gl_ops;
unsigned int ret;
+ if (glops->go_drop_th)
+ glops->go_drop_th(gl);
+
gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags));
- gfs2_assert_warn(sdp, queue_empty(gl, &gl->gl_holders));
+ gfs2_assert_warn(sdp, list_empty(&gl->gl_holders));
gfs2_assert_warn(sdp, gl->gl_state != LM_ST_UNLOCKED);
- if (gl->gl_state == LM_ST_EXCLUSIVE && glops->go_sync)
- glops->go_sync(gl, DIO_METADATA | DIO_DATA | DIO_RELEASE);
-
gfs2_glock_hold(gl);
gl->gl_req_bh = drop_bh;
if (gh->gh_flags & LM_FLAG_PRIORITY)
do_cancels(gh);
- wait_for_completion(&gh->gh_wait);
-
+ wait_on_holder(gh);
if (gh->gh_error)
return gh->gh_error;
gfs2_assert_withdraw(sdp, test_bit(HIF_HOLDER, &gh->gh_iflags));
- gfs2_assert_withdraw(sdp, relaxed_state_ok(gl->gl_state,
- gh->gh_state,
+ gfs2_assert_withdraw(sdp, relaxed_state_ok(gl->gl_state, gh->gh_state,
gh->gh_flags));
if (test_bit(HIF_FIRST, &gh->gh_iflags)) {
struct gfs2_holder *existing;
BUG_ON(!gh->gh_owner);
+ if (test_and_set_bit(HIF_WAIT, &gh->gh_iflags))
+ BUG();
existing = find_holder_by_owner(&gl->gl_holders, gh->gh_owner);
if (existing) {
print_symbol(KERN_WARNING "original: %s\n", existing->gh_ip);
printk(KERN_INFO "pid : %d\n", existing->gh_owner->pid);
- printk(KERN_INFO "lock type : %d lock state : %d\n",
+ printk(KERN_INFO "lock type : %d lock state : %d\n",
existing->gh_gl->gl_name.ln_type, existing->gh_gl->gl_state);
print_symbol(KERN_WARNING "new: %s\n", gh->gh_ip);
printk(KERN_INFO "pid : %d\n", gh->gh_owner->pid);
- printk(KERN_INFO "lock type : %d lock state : %d\n",
+ printk(KERN_INFO "lock type : %d lock state : %d\n",
gl->gl_name.ln_type, gl->gl_state);
BUG();
}
if (gh->gh_flags & LM_FLAG_PRIORITY)
list_add(&gh->gh_list, &gl->gl_waiters3);
else
- list_add_tail(&gh->gh_list, &gl->gl_waiters3);
+ list_add_tail(&gh->gh_list, &gl->gl_waiters3);
}
/**
}
}
- clear_bit(GLF_PREFETCH, &gl->gl_flags);
-
- if (error == GLR_TRYFAILED && (gh->gh_flags & GL_DUMP))
- dump_glock(gl);
-
return error;
}
spin_unlock(&gl->gl_spin);
}
-/**
- * gfs2_glock_prefetch - Try to prefetch a glock
- * @gl: the glock
- * @state: the state to prefetch in
- * @flags: flags passed to go_xmote_th()
- *
- */
-
-static void gfs2_glock_prefetch(struct gfs2_glock *gl, unsigned int state,
- int flags)
-{
- const struct gfs2_glock_operations *glops = gl->gl_ops;
-
- spin_lock(&gl->gl_spin);
-
- if (test_bit(GLF_LOCK, &gl->gl_flags) || !list_empty(&gl->gl_holders) ||
- !list_empty(&gl->gl_waiters1) || !list_empty(&gl->gl_waiters2) ||
- !list_empty(&gl->gl_waiters3) ||
- relaxed_state_ok(gl->gl_state, state, flags)) {
- spin_unlock(&gl->gl_spin);
- return;
- }
-
- set_bit(GLF_PREFETCH, &gl->gl_flags);
- set_bit(GLF_LOCK, &gl->gl_flags);
- spin_unlock(&gl->gl_spin);
-
- glops->go_xmote_th(gl, state, flags);
-}
-
-static void greedy_work(void *data)
-{
- struct greedy *gr = data;
- struct gfs2_holder *gh = &gr->gr_gh;
- struct gfs2_glock *gl = gh->gh_gl;
- const struct gfs2_glock_operations *glops = gl->gl_ops;
-
- clear_bit(GLF_SKIP_WAITERS2, &gl->gl_flags);
-
- if (glops->go_greedy)
- glops->go_greedy(gl);
-
- spin_lock(&gl->gl_spin);
-
- if (list_empty(&gl->gl_waiters2)) {
- clear_bit(GLF_GREEDY, &gl->gl_flags);
- spin_unlock(&gl->gl_spin);
- gfs2_holder_uninit(gh);
- kfree(gr);
- } else {
- gfs2_glock_hold(gl);
- list_add_tail(&gh->gh_list, &gl->gl_waiters2);
- run_queue(gl);
- spin_unlock(&gl->gl_spin);
- gfs2_glock_put(gl);
- }
-}
-
-/**
- * gfs2_glock_be_greedy -
- * @gl:
- * @time:
- *
- * Returns: 0 if go_greedy will be called, 1 otherwise
- */
-
-int gfs2_glock_be_greedy(struct gfs2_glock *gl, unsigned int time)
-{
- struct greedy *gr;
- struct gfs2_holder *gh;
-
- if (!time || gl->gl_sbd->sd_args.ar_localcaching ||
- test_and_set_bit(GLF_GREEDY, &gl->gl_flags))
- return 1;
-
- gr = kmalloc(sizeof(struct greedy), GFP_KERNEL);
- if (!gr) {
- clear_bit(GLF_GREEDY, &gl->gl_flags);
- return 1;
- }
- gh = &gr->gr_gh;
-
- gfs2_holder_init(gl, 0, 0, gh);
- set_bit(HIF_GREEDY, &gh->gh_iflags);
- INIT_WORK(&gr->gr_work, greedy_work, gr);
-
- set_bit(GLF_SKIP_WAITERS2, &gl->gl_flags);
- schedule_delayed_work(&gr->gr_work, time);
-
- return 0;
-}
-
/**
* gfs2_glock_dq_uninit - dequeue a holder from a glock and initialize it
* @gh: the holder structure
static int glock_compare(const void *arg_a, const void *arg_b)
{
- struct gfs2_holder *gh_a = *(struct gfs2_holder **)arg_a;
- struct gfs2_holder *gh_b = *(struct gfs2_holder **)arg_b;
- struct lm_lockname *a = &gh_a->gh_gl->gl_name;
- struct lm_lockname *b = &gh_b->gh_gl->gl_name;
- int ret = 0;
+ const struct gfs2_holder *gh_a = *(const struct gfs2_holder **)arg_a;
+ const struct gfs2_holder *gh_b = *(const struct gfs2_holder **)arg_b;
+ const struct lm_lockname *a = &gh_a->gh_gl->gl_name;
+ const struct lm_lockname *b = &gh_b->gh_gl->gl_name;
if (a->ln_number > b->ln_number)
- ret = 1;
- else if (a->ln_number < b->ln_number)
- ret = -1;
- else {
- if (gh_a->gh_state == LM_ST_SHARED &&
- gh_b->gh_state == LM_ST_EXCLUSIVE)
- ret = 1;
- else if (!(gh_a->gh_flags & GL_LOCAL_EXCL) &&
- (gh_b->gh_flags & GL_LOCAL_EXCL))
- ret = 1;
- }
-
- return ret;
+ return 1;
+ if (a->ln_number < b->ln_number)
+ return -1;
+ BUG_ON(gh_a->gh_gl->gl_ops->go_type == gh_b->gh_gl->gl_ops->go_type);
+ return 0;
}
/**
gfs2_glock_dq_uninit(&ghs[x]);
}
-/**
- * gfs2_glock_prefetch_num - prefetch a glock based on lock number
- * @sdp: the filesystem
- * @number: the lock number
- * @glops: the glock operations for the type of glock
- * @state: the state to acquire the glock in
- * @flags: modifier flags for the aquisition
- *
- * Returns: errno
- */
-
-void gfs2_glock_prefetch_num(struct gfs2_sbd *sdp, u64 number,
- const struct gfs2_glock_operations *glops,
- unsigned int state, int flags)
-{
- struct gfs2_glock *gl;
- int error;
-
- if (atomic_read(&sdp->sd_reclaim_count) <
- gfs2_tune_get(sdp, gt_reclaim_limit)) {
- error = gfs2_glock_get(sdp, number, glops, CREATE, &gl);
- if (!error) {
- gfs2_glock_prefetch(gl, state, flags);
- gfs2_glock_put(gl);
- }
- }
-}
-
/**
* gfs2_lvb_hold - attach a LVB from a glock
* @gl: The glock in question
if (!gl)
return;
- if (gl->gl_ops->go_callback)
- gl->gl_ops->go_callback(gl, state);
handle_callback(gl, state);
spin_lock(&gl->gl_spin);
/**
* gfs2_glock_cb - Callback used by locking module
- * @fsdata: Pointer to the superblock
+ * @sdp: Pointer to the superblock
* @type: Type of callback
* @data: Type dependent data pointer
*
* a journal from another client needs to be recovered.
*/
-void gfs2_glock_cb(lm_fsdata_t *fsdata, unsigned int type, void *data)
+void gfs2_glock_cb(void *cb_data, unsigned int type, void *data)
{
- struct gfs2_sbd *sdp = (struct gfs2_sbd *)fsdata;
+ struct gfs2_sbd *sdp = cb_data;
switch (type) {
case LM_CB_NEED_E:
struct lm_async_cb *async = data;
struct gfs2_glock *gl;
+ down_read(&gfs2_umount_flush_sem);
gl = gfs2_glock_find(sdp, &async->lc_name);
if (gfs2_assert_warn(sdp, gl))
return;
if (!gfs2_assert_warn(sdp, gl->gl_req_bh))
gl->gl_req_bh(gl, async->lc_ret);
gfs2_glock_put(gl);
+ up_read(&gfs2_umount_flush_sem);
return;
}
}
}
-/**
- * gfs2_iopen_go_callback - Try to kick the inode/vnode associated with an
- * iopen glock from memory
- * @io_gl: the iopen glock
- * @state: the state into which the glock should be put
- *
- */
-
-void gfs2_iopen_go_callback(struct gfs2_glock *io_gl, unsigned int state)
-{
-
- if (state != LM_ST_UNLOCKED)
- return;
- /* FIXME: remove this? */
-}
-
/**
* demote_ok - Check to see if it's ok to unlock a glock
* @gl: the glock
static int demote_ok(struct gfs2_glock *gl)
{
- struct gfs2_sbd *sdp = gl->gl_sbd;
const struct gfs2_glock_operations *glops = gl->gl_ops;
int demote = 1;
if (test_bit(GLF_STICKY, &gl->gl_flags))
demote = 0;
- else if (test_bit(GLF_PREFETCH, &gl->gl_flags))
- demote = time_after_eq(jiffies, gl->gl_stamp +
- gfs2_tune_get(sdp, gt_prefetch_secs) * HZ);
else if (glops->go_demote_ok)
demote = glops->go_demote_ok(gl);
atomic_inc(&sdp->sd_reclaimed);
if (gfs2_glmutex_trylock(gl)) {
- if (queue_empty(gl, &gl->gl_holders) &&
+ if (list_empty(&gl->gl_holders) &&
gl->gl_state != LM_ST_UNLOCKED && demote_ok(gl))
handle_callback(gl, LM_ST_UNLOCKED);
gfs2_glmutex_unlock(gl);
*/
static int examine_bucket(glock_examiner examiner, struct gfs2_sbd *sdp,
- struct gfs2_gl_hash_bucket *bucket)
+ unsigned int hash)
{
- struct glock_plug plug;
- struct list_head *tmp;
- struct gfs2_glock *gl;
- int entries;
-
- /* Add "plug" to end of bucket list, work back up list from there */
- memset(&plug.gl_flags, 0, sizeof(unsigned long));
- set_bit(GLF_PLUG, &plug.gl_flags);
-
- write_lock(&bucket->hb_lock);
- list_add(&plug.gl_list, &bucket->hb_list);
- write_unlock(&bucket->hb_lock);
-
- for (;;) {
- write_lock(&bucket->hb_lock);
+ struct gfs2_glock *gl, *prev = NULL;
+ int has_entries = 0;
+ struct hlist_head *head = &gl_hash_table[hash].hb_list;
- for (;;) {
- tmp = plug.gl_list.next;
-
- if (tmp == &bucket->hb_list) {
- list_del(&plug.gl_list);
- entries = !list_empty(&bucket->hb_list);
- write_unlock(&bucket->hb_lock);
- return entries;
- }
- gl = list_entry(tmp, struct gfs2_glock, gl_list);
-
- /* Move plug up list */
- list_move(&plug.gl_list, &gl->gl_list);
-
- if (test_bit(GLF_PLUG, &gl->gl_flags))
- continue;
-
- /* examiner() must glock_put() */
+ read_lock(gl_lock_addr(hash));
+ /* Can't use hlist_for_each_entry - don't want prefetch here */
+ if (hlist_empty(head))
+ goto out;
+ gl = list_entry(head->first, struct gfs2_glock, gl_list);
+ while(1) {
+ if (gl->gl_sbd == sdp) {
gfs2_glock_hold(gl);
-
- break;
+ read_unlock(gl_lock_addr(hash));
+ if (prev)
+ gfs2_glock_put(prev);
+ prev = gl;
+ examiner(gl);
+ has_entries = 1;
+ read_lock(gl_lock_addr(hash));
}
-
- write_unlock(&bucket->hb_lock);
-
- examiner(gl);
+ if (gl->gl_list.next == NULL)
+ break;
+ gl = list_entry(gl->gl_list.next, struct gfs2_glock, gl_list);
}
+out:
+ read_unlock(gl_lock_addr(hash));
+ if (prev)
+ gfs2_glock_put(prev);
+ return has_entries;
}
/**
static void scan_glock(struct gfs2_glock *gl)
{
- if (gl->gl_ops == &gfs2_inode_glops)
- goto out;
+ if (gl->gl_ops == &gfs2_inode_glops && gl->gl_object)
+ return;
if (gfs2_glmutex_trylock(gl)) {
- if (queue_empty(gl, &gl->gl_holders) &&
- gl->gl_state != LM_ST_UNLOCKED &&
- demote_ok(gl))
+ if (list_empty(&gl->gl_holders) &&
+ gl->gl_state != LM_ST_UNLOCKED && demote_ok(gl))
goto out_schedule;
gfs2_glmutex_unlock(gl);
}
-out:
- gfs2_glock_put(gl);
return;
out_schedule:
gfs2_glmutex_unlock(gl);
gfs2_glock_schedule_for_reclaim(gl);
- gfs2_glock_put(gl);
}
/**
{
unsigned int x;
- for (x = 0; x < GFS2_GL_HASH_SIZE; x++) {
- examine_bucket(scan_glock, sdp, &sdp->sd_gl_hash[x]);
- cond_resched();
- }
+ for (x = 0; x < GFS2_GL_HASH_SIZE; x++)
+ examine_bucket(scan_glock, sdp, x);
}
/**
}
if (gfs2_glmutex_trylock(gl)) {
- if (queue_empty(gl, &gl->gl_holders) &&
+ if (list_empty(&gl->gl_holders) &&
gl->gl_state != LM_ST_UNLOCKED)
handle_callback(gl, LM_ST_UNLOCKED);
-
gfs2_glmutex_unlock(gl);
}
-
- gfs2_glock_put(gl);
}
/**
for (;;) {
cont = 0;
-
- for (x = 0; x < GFS2_GL_HASH_SIZE; x++)
- if (examine_bucket(clear_glock, sdp, &sdp->sd_gl_hash[x]))
+ for (x = 0; x < GFS2_GL_HASH_SIZE; x++) {
+ if (examine_bucket(clear_glock, sdp, x))
cont = 1;
+ }
if (!wait || !cont)
break;
t = jiffies;
}
+ down_write(&gfs2_umount_flush_sem);
invalidate_inodes(sdp->sd_vfs);
+ up_write(&gfs2_umount_flush_sem);
msleep(10);
}
}
printk(KERN_INFO " num = %llu %llu\n",
(unsigned long long)ip->i_num.no_formal_ino,
(unsigned long long)ip->i_num.no_addr);
- printk(KERN_INFO " type = %u\n", IF2DT(ip->i_di.di_mode));
+ printk(KERN_INFO " type = %u\n", IF2DT(ip->i_inode.i_mode));
printk(KERN_INFO " i_flags =");
for (x = 0; x < 32; x++)
if (test_bit(x, &ip->i_flags))
spin_lock(&gl->gl_spin);
- printk(KERN_INFO "Glock 0x%p (%u, %llu)\n",
- gl,
- gl->gl_name.ln_type,
+ printk(KERN_INFO "Glock 0x%p (%u, %llu)\n", gl, gl->gl_name.ln_type,
(unsigned long long)gl->gl_name.ln_number);
printk(KERN_INFO " gl_flags =");
- for (x = 0; x < 32; x++)
+ for (x = 0; x < 32; x++) {
if (test_bit(x, &gl->gl_flags))
printk(" %u", x);
+ }
printk(" \n");
- printk(KERN_INFO " gl_ref = %d\n", atomic_read(&gl->gl_ref.refcount));
+ printk(KERN_INFO " gl_ref = %d\n", atomic_read(&gl->gl_ref));
printk(KERN_INFO " gl_state = %u\n", gl->gl_state);
printk(KERN_INFO " gl_owner = %s\n", gl->gl_owner->comm);
print_symbol(KERN_INFO " gl_ip = %s\n", gl->gl_ip);
printk(KERN_INFO " reclaim = %s\n",
(list_empty(&gl->gl_reclaim)) ? "no" : "yes");
if (gl->gl_aspace)
- printk(KERN_INFO " aspace = 0x%p nrpages = %lu\n",
- gl->gl_aspace,
+ printk(KERN_INFO " aspace = 0x%p nrpages = %lu\n", gl->gl_aspace,
gl->gl_aspace->i_mapping->nrpages);
else
printk(KERN_INFO " aspace = no\n");
static int gfs2_dump_lockstate(struct gfs2_sbd *sdp)
{
- struct gfs2_gl_hash_bucket *bucket;
struct gfs2_glock *gl;
+ struct hlist_node *h;
unsigned int x;
int error = 0;
for (x = 0; x < GFS2_GL_HASH_SIZE; x++) {
- bucket = &sdp->sd_gl_hash[x];
- read_lock(&bucket->hb_lock);
+ read_lock(gl_lock_addr(x));
- list_for_each_entry(gl, &bucket->hb_list, gl_list) {
- if (test_bit(GLF_PLUG, &gl->gl_flags))
+ hlist_for_each_entry(gl, h, &gl_hash_table[x].hb_list, gl_list) {
+ if (gl->gl_sbd != sdp)
continue;
error = dump_glock(gl);
break;
}
- read_unlock(&bucket->hb_lock);
+ read_unlock(gl_lock_addr(x));
if (error)
break;
return error;
}
+int __init gfs2_glock_init(void)
+{
+ unsigned i;
+ for(i = 0; i < GFS2_GL_HASH_SIZE; i++) {
+ INIT_HLIST_HEAD(&gl_hash_table[i].hb_list);
+ }
+#ifdef GL_HASH_LOCK_SZ
+ for(i = 0; i < GL_HASH_LOCK_SZ; i++) {
+ rwlock_init(&gl_hash_locks[i]);
+ }
+#endif
+ return 0;
+}
+