/*
* Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
- * Copyright (C) 2004-2005 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
*
* This copyrighted material is made available to anyone wishing to use,
* modify, copy, or redistribute it subject to the terms and conditions
- * of the GNU General Public License v.2.
+ * of the GNU General Public License version 2.
*/
#include <linux/sched.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
-#include <asm/semaphore.h>
+#include <linux/gfs2_ondisk.h>
+#include <linux/crc32.h>
#include "gfs2.h"
+#include "lm_interface.h"
+#include "incore.h"
#include "bmap.h"
#include "glock.h"
#include "log.h"
#include "lops.h"
#include "meta_io.h"
+#include "util.h"
+#include "dir.h"
#define PULL 1
-static inline int is_done(struct gfs2_sbd *sdp, atomic_t *a)
-{
- int done;
- gfs2_log_lock(sdp);
- done = atomic_read(a) ? 0 : 1;
- gfs2_log_unlock(sdp);
- return done;
-}
-
-static void do_lock_wait(struct gfs2_sbd *sdp, wait_queue_head_t *wq,
- atomic_t *a)
-{
- gfs2_log_unlock(sdp);
- wait_event(*wq, is_done(sdp, a));
- gfs2_log_lock(sdp);
-}
-
-static void lock_for_trans(struct gfs2_sbd *sdp)
-{
- gfs2_log_lock(sdp);
- do_lock_wait(sdp, &sdp->sd_log_trans_wq, &sdp->sd_log_flush_count);
- atomic_inc(&sdp->sd_log_trans_count);
- gfs2_log_unlock(sdp);
-}
-
-static void unlock_from_trans(struct gfs2_sbd *sdp)
-{
- gfs2_assert_warn(sdp, atomic_read(&sdp->sd_log_trans_count));
- if (atomic_dec_and_test(&sdp->sd_log_trans_count))
- wake_up(&sdp->sd_log_flush_wq);
-}
-
-void gfs2_lock_for_flush(struct gfs2_sbd *sdp)
-{
- gfs2_log_lock(sdp);
- atomic_inc(&sdp->sd_log_flush_count);
- do_lock_wait(sdp, &sdp->sd_log_flush_wq, &sdp->sd_log_trans_count);
- gfs2_log_unlock(sdp);
-}
-
-void gfs2_unlock_from_flush(struct gfs2_sbd *sdp)
-{
- gfs2_assert_warn(sdp, atomic_read(&sdp->sd_log_flush_count));
- if (atomic_dec_and_test(&sdp->sd_log_flush_count))
- wake_up(&sdp->sd_log_trans_wq);
-}
-
/**
* gfs2_struct2blk - compute stuff
* @sdp: the filesystem
unsigned int first, second;
blks = 1;
- first = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_log_descriptor)) / ssize;
+ first = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_log_descriptor)) /
+ ssize;
if (nstruct > first) {
- second = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_meta_header)) / ssize;
- blks += DIV_RU(nstruct - first, second);
+ second = (sdp->sd_sb.sb_bsize -
+ sizeof(struct gfs2_meta_header)) / ssize;
+ blks += DIV_ROUND_UP(nstruct - first, second);
}
return blks;
first = NULL;
for (;;) {
- if (first &&
- (head->prev != first ||
- gfs2_ail1_empty_one(sdp, first_ai, 0)))
+ if (first && (head->prev != first ||
+ gfs2_ail1_empty_one(sdp, first_ai, 0)))
break;
for (tmp = head->prev; tmp != head; tmp = tmp->prev) {
int gfs2_log_reserve(struct gfs2_sbd *sdp, unsigned int blks)
{
- LIST_HEAD(list);
unsigned int try = 0;
if (gfs2_assert_warn(sdp, blks) ||
gfs2_assert_warn(sdp, blks <= sdp->sd_jdesc->jd_blocks))
return -EINVAL;
- for (;;) {
- gfs2_log_lock(sdp);
-
- if (list_empty(&list)) {
- list_add_tail(&list, &sdp->sd_log_blks_list);
- while (sdp->sd_log_blks_list.next != &list) {
- DECLARE_WAITQUEUE(__wait_chan, current);
- set_current_state(TASK_UNINTERRUPTIBLE);
- add_wait_queue(&sdp->sd_log_blks_wait,
- &__wait_chan);
- gfs2_log_unlock(sdp);
- schedule();
- gfs2_log_lock(sdp);
- remove_wait_queue(&sdp->sd_log_blks_wait,
- &__wait_chan);
- set_current_state(TASK_RUNNING);
- }
- }
-
- /* Never give away the last block so we can
- always pull the tail if we need to. */
- if (sdp->sd_log_blks_free > blks) {
- sdp->sd_log_blks_free -= blks;
- list_del(&list);
- gfs2_log_unlock(sdp);
- wake_up(&sdp->sd_log_blks_wait);
- break;
- }
-
+ mutex_lock(&sdp->sd_log_reserve_mutex);
+ gfs2_log_lock(sdp);
+ while(sdp->sd_log_blks_free <= blks) {
gfs2_log_unlock(sdp);
-
gfs2_ail1_empty(sdp, 0);
- gfs2_log_flush(sdp);
+ gfs2_log_flush(sdp, NULL);
if (try++)
gfs2_ail1_start(sdp, 0);
+ gfs2_log_lock(sdp);
}
+ sdp->sd_log_blks_free -= blks;
+ /* printk(KERN_INFO "reserved %u blocks (%u left)\n", blks, sdp->sd_log_blks_free); */
+ gfs2_log_unlock(sdp);
+ mutex_unlock(&sdp->sd_log_reserve_mutex);
- lock_for_trans(sdp);
+ down_read(&sdp->sd_log_flush_lock);
return 0;
}
void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks)
{
- unlock_from_trans(sdp);
gfs2_log_lock(sdp);
sdp->sd_log_blks_free += blks;
+ /* printk(KERN_INFO "released %u blocks (%u left)\n", blks, sdp->sd_log_blks_free); */
gfs2_assert_withdraw(sdp,
sdp->sd_log_blks_free <= sdp->sd_jdesc->jd_blocks);
gfs2_log_unlock(sdp);
+ up_read(&sdp->sd_log_flush_lock);
}
static uint64_t log_bmap(struct gfs2_sbd *sdp, unsigned int lbn)
int new = 0;
uint64_t dbn;
int error;
+ int bdy;
- error = gfs2_block_map(get_v2ip(sdp->sd_jdesc->jd_inode), lbn, &new, &dbn, NULL);
+ error = gfs2_block_map(sdp->sd_jdesc->jd_inode, lbn, &new, &dbn, &bdy);
+ if (!(!error && dbn)) {
+ printk(KERN_INFO "error=%d, dbn=%llu lbn=%u", error, (unsigned long long)dbn, lbn);
+ }
gfs2_assert_withdraw(sdp, !error && dbn);
return dbn;
if (list_empty(&sdp->sd_ail1_list))
tail = sdp->sd_log_head;
else {
- ai = list_entry(sdp->sd_ail1_list.prev,
- struct gfs2_ail, ai_list);
+ ai = list_entry(sdp->sd_ail1_list.prev, struct gfs2_ail,
+ ai_list);
tail = ai->ai_first;
}
gfs2_log_lock(sdp);
sdp->sd_log_blks_free += dist - ((pull) ? 1 : 0);
+ /* printk(KERN_INFO "pull tail refunding %u blocks (%u left) pull=%d\n", dist - ((pull) ? 1 : 0), sdp->sd_log_blks_free, pull); */
gfs2_assert_withdraw(sdp,
sdp->sd_log_blks_free <= sdp->sd_jdesc->jd_blocks);
gfs2_log_unlock(sdp);
unsigned int tail;
uint32_t hash;
- atomic_inc(&sdp->sd_log_flush_ondisk);
+ /* printk(KERN_INFO "log write header start (flags=%08x, pull=%d)\n", flags, pull); */
bh = sb_getblk(sdp->sd_vfs, blkno);
lock_buffer(bh);
lh = (struct gfs2_log_header *)bh->b_data;
memset(lh, 0, sizeof(struct gfs2_log_header));
lh->lh_header.mh_magic = cpu_to_be32(GFS2_MAGIC);
- lh->lh_header.mh_type = cpu_to_be16(GFS2_METATYPE_LH);
- lh->lh_header.mh_format = cpu_to_be16(GFS2_FORMAT_LH);
- lh->lh_sequence = be64_to_cpu(sdp->sd_log_sequence++);
- lh->lh_flags = be32_to_cpu(flags);
- lh->lh_tail = be32_to_cpu(tail);
- lh->lh_blkno = be32_to_cpu(sdp->sd_log_flush_head);
+ lh->lh_header.mh_type = cpu_to_be32(GFS2_METATYPE_LH);
+ lh->lh_header.mh_format = cpu_to_be32(GFS2_FORMAT_LH);
+ lh->lh_sequence = cpu_to_be64(sdp->sd_log_sequence++);
+ lh->lh_flags = cpu_to_be32(flags);
+ lh->lh_tail = cpu_to_be32(tail);
+ lh->lh_blkno = cpu_to_be32(sdp->sd_log_flush_head);
hash = gfs2_disk_hash(bh->b_data, sizeof(struct gfs2_log_header));
lh->lh_hash = cpu_to_be32(hash);
sdp->sd_log_idle = (tail == sdp->sd_log_flush_head);
log_incr_head(sdp);
+
+ /* printk(KERN_INFO "log write header out\n"); */
}
static void log_flush_commit(struct gfs2_sbd *sdp)
struct list_head *head = &sdp->sd_log_flush_list;
struct gfs2_log_buf *lb;
struct buffer_head *bh;
+#if 0
unsigned int d;
d = log_distance(sdp, sdp->sd_log_flush_head, sdp->sd_log_head);
gfs2_assert_withdraw(sdp, d + 1 == sdp->sd_log_blks_reserved);
+#endif
while (!list_empty(head)) {
lb = list_entry(head->next, struct gfs2_log_buf, lb_list);
}
/**
- * gfs2_log_flush_i - flush incore transaction(s)
+ * gfs2_log_flush - flush incore transaction(s)
* @sdp: the filesystem
* @gl: The glock structure to flush. If NULL, flush the whole incore log
*
*/
-void gfs2_log_flush_i(struct gfs2_sbd *sdp, struct gfs2_glock *gl)
+void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl)
{
struct gfs2_ail *ai;
- atomic_inc(&sdp->sd_log_flush_incore);
+ down_write(&sdp->sd_log_flush_lock);
+
+ if (gl) {
+ gfs2_log_lock(sdp);
+ if (list_empty(&gl->gl_le.le_list)) {
+ gfs2_log_unlock(sdp);
+ up_write(&sdp->sd_log_flush_lock);
+ return;
+ }
+ gfs2_log_unlock(sdp);
+ }
ai = kzalloc(sizeof(struct gfs2_ail), GFP_NOFS | __GFP_NOFAIL);
INIT_LIST_HEAD(&ai->ai_ail1_list);
INIT_LIST_HEAD(&ai->ai_ail2_list);
- gfs2_lock_for_flush(sdp);
- down(&sdp->sd_log_flush_lock);
-
gfs2_assert_withdraw(sdp,
sdp->sd_log_num_buf == sdp->sd_log_commited_buf);
gfs2_assert_withdraw(sdp,
sdp->sd_log_num_revoke == sdp->sd_log_commited_revoke);
- if (gl && list_empty(&gl->gl_le.le_list)) {
- up(&sdp->sd_log_flush_lock);
- gfs2_unlock_from_flush(sdp);
- kfree(ai);
- return;
- }
-
sdp->sd_log_flush_head = sdp->sd_log_head;
sdp->sd_log_flush_wrapped = 0;
ai->ai_first = sdp->sd_log_flush_head;
else if (sdp->sd_log_tail != current_tail(sdp) && !sdp->sd_log_idle)
log_write_header(sdp, 0, PULL);
lops_after_commit(sdp, ai);
-
sdp->sd_log_head = sdp->sd_log_flush_head;
- if (sdp->sd_log_flush_wrapped)
- sdp->sd_log_wraps++;
+
+ /* printk(KERN_INFO "sd_log_num_hdrs %u\n", sdp->sd_log_num_hdrs); */
+ sdp->sd_log_blks_free -= sdp->sd_log_num_hdrs;
sdp->sd_log_blks_reserved =
sdp->sd_log_commited_buf =
+ sdp->sd_log_num_hdrs =
sdp->sd_log_commited_revoke = 0;
gfs2_log_lock(sdp);
}
gfs2_log_unlock(sdp);
- up(&sdp->sd_log_flush_lock);
sdp->sd_vfs->s_dirt = 0;
- gfs2_unlock_from_flush(sdp);
+ up_write(&sdp->sd_log_flush_lock);
kfree(ai);
}
static void log_refund(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
{
- unsigned int reserved = 1;
+ unsigned int reserved = 0;
unsigned int old;
gfs2_log_lock(sdp);
gfs2_assert_withdraw(sdp, ((int)sdp->sd_log_commited_revoke) >= 0);
if (sdp->sd_log_commited_buf)
- reserved += 1 + sdp->sd_log_commited_buf + sdp->sd_log_commited_buf/503;
+ reserved += sdp->sd_log_commited_buf;
if (sdp->sd_log_commited_revoke)
reserved += gfs2_struct2blk(sdp, sdp->sd_log_commited_revoke,
sizeof(uint64_t));
+ if (reserved)
+ reserved++;
old = sdp->sd_log_blks_free;
sdp->sd_log_blks_free += tr->tr_reserved -
(reserved - sdp->sd_log_blks_reserved);
+ gfs2_assert_withdraw(sdp, sdp->sd_log_blks_free >= old);
gfs2_assert_withdraw(sdp,
- sdp->sd_log_blks_free >= old);
- gfs2_assert_withdraw(sdp,
- sdp->sd_log_blks_free <= sdp->sd_jdesc->jd_blocks);
+ sdp->sd_log_blks_free <= sdp->sd_jdesc->jd_blocks +
+ sdp->sd_log_num_hdrs);
sdp->sd_log_blks_reserved = reserved;
lops_incore_commit(sdp, tr);
sdp->sd_vfs->s_dirt = 1;
- unlock_from_trans(sdp);
-
- kfree(tr);
+ up_read(&sdp->sd_log_flush_lock);
gfs2_log_lock(sdp);
if (sdp->sd_log_num_buf > gfs2_tune_get(sdp, gt_incore_log_blocks)) {
gfs2_log_unlock(sdp);
- gfs2_log_flush(sdp);
+ gfs2_log_flush(sdp, NULL);
} else
gfs2_log_unlock(sdp);
}
void gfs2_log_shutdown(struct gfs2_sbd *sdp)
{
- down(&sdp->sd_log_flush_lock);
+ down_write(&sdp->sd_log_flush_lock);
- gfs2_assert_withdraw(sdp, !atomic_read(&sdp->sd_log_trans_count));
gfs2_assert_withdraw(sdp, !sdp->sd_log_blks_reserved);
gfs2_assert_withdraw(sdp, !sdp->sd_log_num_gl);
gfs2_assert_withdraw(sdp, !sdp->sd_log_num_buf);
gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke);
gfs2_assert_withdraw(sdp, !sdp->sd_log_num_rg);
gfs2_assert_withdraw(sdp, !sdp->sd_log_num_databuf);
+ gfs2_assert_withdraw(sdp, !sdp->sd_log_num_hdrs);
gfs2_assert_withdraw(sdp, list_empty(&sdp->sd_ail1_list));
sdp->sd_log_flush_head = sdp->sd_log_head;
log_write_header(sdp, GFS2_LOG_HEAD_UNMOUNT, 0);
- gfs2_assert_withdraw(sdp, sdp->sd_log_blks_free ==
- sdp->sd_jdesc->jd_blocks);
- gfs2_assert_withdraw(sdp, sdp->sd_log_head == sdp->sd_log_tail);
- gfs2_assert_withdraw(sdp, list_empty(&sdp->sd_ail2_list));
+ /* printk(KERN_INFO "sd_log_blks_free %u, sd_jdesc->jd_blocks %u\n", sdp->sd_log_blks_free, sdp->sd_jdesc->jd_blocks); */
+ gfs2_assert_warn(sdp, sdp->sd_log_blks_free == sdp->sd_jdesc->jd_blocks);
+ gfs2_assert_warn(sdp, sdp->sd_log_head == sdp->sd_log_tail);
+ gfs2_assert_warn(sdp, list_empty(&sdp->sd_ail2_list));
sdp->sd_log_head = sdp->sd_log_flush_head;
- if (sdp->sd_log_flush_wrapped)
- sdp->sd_log_wraps++;
sdp->sd_log_tail = sdp->sd_log_head;
- up(&sdp->sd_log_flush_lock);
+ up_write(&sdp->sd_log_flush_lock);
}