Merge Linux 2.6.23

13 files changed, 77 insertions, 86 deletions

diff --git a/fs/xfs/linux-2.6/kmem.h b/fs/xfs/linux-2.6/kmem.h
index b4acc7f3c37..e6ea293f303 100644
--- a/fs/xfs/linux-2.6/kmem.h
+++ b/fs/xfs/linux-2.6/kmem.h
@@ -103,7 +103,7 @@ extern void *kmem_zone_zalloc(kmem_zone_t *, unsigned int __nocast);
 static inline int
 kmem_shake_allow(gfp_t gfp_mask)
 {
-	return (gfp_mask & __GFP_WAIT);
+	return (gfp_mask & __GFP_WAIT) != 0;
 }
 
 #endif /* __XFS_SUPPORT_KMEM_H__ */
diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index fd4105d662e..5f152f60d74 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -181,6 +181,7 @@ xfs_setfilesize(
 		ip->i_d.di_size = isize;
 		ip->i_update_core = 1;
 		ip->i_update_size = 1;
+		mark_inode_dirty_sync(vn_to_inode(ioend->io_vnode));
 	}
 
 	xfs_iunlock(ip, XFS_ILOCK_EXCL);
@@ -652,7 +653,7 @@ xfs_probe_cluster(
 
 		for (i = 0; i < pagevec_count(&pvec); i++) {
 			struct page *page = pvec.pages[i];
-			size_t pg_offset, len = 0;
+			size_t pg_offset, pg_len = 0;
 
 			if (tindex == tlast) {
 				pg_offset =
@@ -665,16 +666,16 @@ xfs_probe_cluster(
 				pg_offset = PAGE_CACHE_SIZE;
 
 			if (page->index == tindex && !TestSetPageLocked(page)) {
-				len = xfs_probe_page(page, pg_offset, mapped);
+				pg_len = xfs_probe_page(page, pg_offset, mapped);
 				unlock_page(page);
 			}
 
-			if (!len) {
+			if (!pg_len) {
 				done = 1;
 				break;
 			}
 
-			total += len;
+			total += pg_len;
 			tindex++;
 		}
 
diff --git a/fs/xfs/linux-2.6/xfs_globals.c b/fs/xfs/linux-2.6/xfs_globals.c
index bb72c3d4141..81565dea9af 100644
--- a/fs/xfs/linux-2.6/xfs_globals.c
+++ b/fs/xfs/linux-2.6/xfs_globals.c
@@ -46,7 +46,7 @@ xfs_param_t xfs_params = {
 	.inherit_nosym	= {	0,		0,		1	},
 	.rotorstep	= {	1,		1,		255	},
 	.inherit_nodfrg	= {	0,		1,		1	},
-	.fstrm_timer	= {	1,		50,		3600*100},
+	.fstrm_timer	= {	1,		30*100,		3600*100},
 };
 
 /*
diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index 4528f9a3f30..491d1f4f202 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -415,8 +415,10 @@ xfs_fs_write_inode(
 
 	if (vp) {
 		vn_trace_entry(vp, __FUNCTION__, (inst_t *)__return_address);
-		if (sync)
+		if (sync) {
+			filemap_fdatawait(inode->i_mapping);
 			flags |= FLUSH_SYNC;
+		}
 		error = bhv_vop_iflush(vp, flags);
 		if (error == EAGAIN)
 			error = sync? bhv_vop_iflush(vp, flags | FLUSH_LOG) : 0;
diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c
index 2d274b23ade..6ff0f4de163 100644
--- a/fs/xfs/quota/xfs_qm.c
+++ b/fs/xfs/quota/xfs_qm.c
@@ -120,7 +120,8 @@ xfs_Gqm_init(void)
 	 * Initialize the dquot hash tables.
 	 */
 	udqhash = kmem_zalloc_greedy(&hsize,
-				     XFS_QM_HASHSIZE_LOW, XFS_QM_HASHSIZE_HIGH,
+				     XFS_QM_HASHSIZE_LOW * sizeof(xfs_dqhash_t),
+				     XFS_QM_HASHSIZE_HIGH * sizeof(xfs_dqhash_t),
 				     KM_SLEEP | KM_MAYFAIL | KM_LARGE);
 	gdqhash = kmem_zalloc(hsize, KM_SLEEP | KM_LARGE);
 	hsize /= sizeof(xfs_dqhash_t);
diff --git a/fs/xfs/support/debug.h b/fs/xfs/support/debug.h
index a27a7c8c052..855da040864 100644
--- a/fs/xfs/support/debug.h
+++ b/fs/xfs/support/debug.h
@@ -34,10 +34,10 @@ extern void cmn_err(int, char *, ...)
 extern void assfail(char *expr, char *f, int l);
 
 #define ASSERT_ALWAYS(expr)	\
-	(unlikely((expr) != 0) ? (void)0 : assfail(#expr, __FILE__, __LINE__))
+	(unlikely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__))
 
 #ifndef DEBUG
-# define ASSERT(expr)	((void)0)
+#define ASSERT(expr)	((void)0)
 
 #ifndef STATIC
 # define STATIC static noinline
@@ -49,8 +49,10 @@ extern void assfail(char *expr, char *f, int l);
 
 #else /* DEBUG */
 
-# define ASSERT(expr)	ASSERT_ALWAYS(expr)
-# include <linux/random.h>
+#include <linux/random.h>
+
+#define ASSERT(expr)	\
+	(unlikely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__))
 
 #ifndef STATIC
 # define STATIC noinline
diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c
index aea37df4aa6..26d09e2e1a7 100644
--- a/fs/xfs/xfs_da_btree.c
+++ b/fs/xfs/xfs_da_btree.c
@@ -1975,7 +1975,6 @@ xfs_da_do_buf(
 		error = mappedbno == -2 ? 0 : XFS_ERROR(EFSCORRUPTED);
 		if (unlikely(error == EFSCORRUPTED)) {
 			if (xfs_error_level >= XFS_ERRLEVEL_LOW) {
-				int	i;
 				cmn_err(CE_ALERT, "xfs_da_do_buf: bno %lld\n",
 					(long long)bno);
 				cmn_err(CE_ALERT, "dir: inode %lld\n",
diff --git a/fs/xfs/xfs_filestream.c b/fs/xfs/xfs_filestream.c
index ce2278611bb..36d8f6aa11a 100644
--- a/fs/xfs/xfs_filestream.c
+++ b/fs/xfs/xfs_filestream.c
@@ -350,9 +350,10 @@ _xfs_filestream_update_ag(
 /* xfs_fstrm_free_func(): callback for freeing cached stream items. */
 void
 xfs_fstrm_free_func(
-	xfs_ino_t	ino,
-	fstrm_item_t	*item)
+	unsigned long	ino,
+	void		*data)
 {
+	fstrm_item_t	*item  = (fstrm_item_t *)data;
 	xfs_inode_t	*ip = item->ip;
 	int ref;
 
@@ -438,7 +439,7 @@ xfs_filestream_mount(
 	grp_count = 10;
 
 	err = xfs_mru_cache_create(&mp->m_filestream, lifetime, grp_count,
-	                     (xfs_mru_cache_free_func_t)xfs_fstrm_free_func);
+	                     xfs_fstrm_free_func);
 
 	return err;
 }
@@ -467,8 +468,7 @@ void
 xfs_filestream_flush(
 	xfs_mount_t	*mp)
 {
-	/* point in time flush, so keep the reaper running */
-	xfs_mru_cache_flush(mp->m_filestream, 1);
+	xfs_mru_cache_flush(mp->m_filestream);
 }
 
 /*
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 9d4c4fbeb3e..9bfb69e1e88 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -2185,13 +2185,13 @@ xlog_state_do_callback(
 			}
 			cb = iclog->ic_callback;
 
-			while (cb != 0) {
+			while (cb) {
 				iclog->ic_callback_tail = &(iclog->ic_callback);
 				iclog->ic_callback = NULL;
 				LOG_UNLOCK(log, s);
 
 				/* perform callbacks in the order given */
-				for (; cb != 0; cb = cb_next) {
+				for (; cb; cb = cb_next) {
 					cb_next = cb->cb_next;
 					cb->cb_func(cb->cb_arg, aborted);
 				}
@@ -2202,7 +2202,7 @@ xlog_state_do_callback(
 			loopdidcallbacks++;
 			funcdidcallbacks++;
 
-			ASSERT(iclog->ic_callback == 0);
+			ASSERT(iclog->ic_callback == NULL);
 			if (!(iclog->ic_state & XLOG_STATE_IOERROR))
 				iclog->ic_state = XLOG_STATE_DIRTY;
 
@@ -3242,10 +3242,10 @@ xlog_ticket_put(xlog_t		*log,
 #else
 	/* When we debug, it is easier if tickets are cycled */
 	ticket->t_next     = NULL;
-	if (log->l_tail != 0) {
+	if (log->l_tail) {
 		log->l_tail->t_next = ticket;
 	} else {
-		ASSERT(log->l_freelist == 0);
+		ASSERT(log->l_freelist == NULL);
 		log->l_freelist = ticket;
 	}
 	log->l_tail	    = ticket;
@@ -3463,7 +3463,7 @@ xlog_verify_iclog(xlog_t	 *log,
 	s = LOG_LOCK(log);
 	icptr = log->l_iclog;
 	for (i=0; i < log->l_iclog_bufs; i++) {
-		if (icptr == 0)
+		if (icptr == NULL)
 			xlog_panic("xlog_verify_iclog: invalid ptr");
 		icptr = icptr->ic_next;
 	}
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index fddbb091a86..8ae6e8e5f3d 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1366,7 +1366,7 @@ xlog_recover_add_to_cont_trans(
 	int			old_len;
 
 	item = trans->r_itemq;
-	if (item == 0) {
+	if (item == NULL) {
 		/* finish copying rest of trans header */
 		xlog_recover_add_item(&trans->r_itemq);
 		ptr = (xfs_caddr_t) &trans->r_theader +
@@ -1412,7 +1412,7 @@ xlog_recover_add_to_trans(
 	if (!len)
 		return 0;
 	item = trans->r_itemq;
-	if (item == 0) {
+	if (item == NULL) {
 		ASSERT(*(uint *)dp == XFS_TRANS_HEADER_MAGIC);
 		if (len == sizeof(xfs_trans_header_t))
 			xlog_recover_add_item(&trans->r_itemq);
@@ -1467,12 +1467,12 @@ xlog_recover_unlink_tid(
 	xlog_recover_t		*tp;
 	int			found = 0;
 
-	ASSERT(trans != 0);
+	ASSERT(trans != NULL);
 	if (trans == *q) {
 		*q = (*q)->r_next;
 	} else {
 		tp = *q;
-		while (tp != 0) {
+		while (tp) {
 			if (tp->r_next == trans) {
 				found = 1;
 				break;
@@ -1495,7 +1495,7 @@ xlog_recover_insert_item_backq(
 	xlog_recover_item_t	**q,
 	xlog_recover_item_t	*item)
 {
-	if (*q == 0) {
+	if (*q == NULL) {
 		item->ri_prev = item->ri_next = item;
 		*q = item;
 	} else {
@@ -1899,7 +1899,7 @@ xlog_recover_do_reg_buffer(
 			break;
 		nbits = xfs_contig_bits(data_map, map_size, bit);
 		ASSERT(nbits > 0);
-		ASSERT(item->ri_buf[i].i_addr != 0);
+		ASSERT(item->ri_buf[i].i_addr != NULL);
 		ASSERT(item->ri_buf[i].i_len % XFS_BLI_CHUNK == 0);
 		ASSERT(XFS_BUF_COUNT(bp) >=
 		       ((uint)bit << XFS_BLI_SHIFT)+(nbits<<XFS_BLI_SHIFT));
diff --git a/fs/xfs/xfs_mru_cache.c b/fs/xfs/xfs_mru_cache.c
index 7deb9e3cbbd..e0b358c1c53 100644
--- a/fs/xfs/xfs_mru_cache.c
+++ b/fs/xfs/xfs_mru_cache.c
@@ -206,8 +206,11 @@ _xfs_mru_cache_list_insert(
 	 */
 	if (!_xfs_mru_cache_migrate(mru, now)) {
 		mru->time_zero = now;
-		if (!mru->next_reap)
-			mru->next_reap = mru->grp_count * mru->grp_time;
+		if (!mru->queued) {
+			mru->queued = 1;
+			queue_delayed_work(xfs_mru_reap_wq, &mru->work,
+			                   mru->grp_count * mru->grp_time);
+		}
 	} else {
 		grp = (now - mru->time_zero) / mru->grp_time;
 		grp = (mru->lru_grp + grp) % mru->grp_count;
@@ -271,29 +274,26 @@ _xfs_mru_cache_reap(
 	struct work_struct	*work)
 {
 	xfs_mru_cache_t		*mru = container_of(work, xfs_mru_cache_t, work.work);
-	unsigned long		now;
+	unsigned long		now, next;
 
 	ASSERT(mru && mru->lists);
 	if (!mru || !mru->lists)
 		return;
 
 	mutex_spinlock(&mru->lock);
-	now = jiffies;
-	if (mru->reap_all ||
-	    (mru->next_reap && time_after(now, mru->next_reap))) {
-		if (mru->reap_all)
-			now += mru->grp_count * mru->grp_time * 2;
-		mru->next_reap = _xfs_mru_cache_migrate(mru, now);
-		_xfs_mru_cache_clear_reap_list(mru);
+	next = _xfs_mru_cache_migrate(mru, jiffies);
+	_xfs_mru_cache_clear_reap_list(mru);
+
+	mru->queued = next;
+	if ((mru->queued > 0)) {
+		now = jiffies;
+		if (next <= now)
+			next = 0;
+		else
+			next -= now;
+		queue_delayed_work(xfs_mru_reap_wq, &mru->work, next);
 	}
 
-	/*
-	 * the process that triggered the reap_all is responsible
-	 * for restating the periodic reap if it is required.
-	 */
-	if (!mru->reap_all)
-		queue_delayed_work(xfs_mru_reap_wq, &mru->work, mru->grp_time);
-	mru->reap_all = 0;
 	mutex_spinunlock(&mru->lock, 0);
 }
 
@@ -352,7 +352,7 @@ xfs_mru_cache_create(
 
 	/* An extra list is needed to avoid reaping up to a grp_time early. */
 	mru->grp_count = grp_count + 1;
-	mru->lists = kmem_alloc(mru->grp_count * sizeof(*mru->lists), KM_SLEEP);
+	mru->lists = kmem_zalloc(mru->grp_count * sizeof(*mru->lists), KM_SLEEP);
 
 	if (!mru->lists) {
 		err = ENOMEM;
@@ -374,11 +374,6 @@ xfs_mru_cache_create(
 	mru->grp_time  = grp_time;
 	mru->free_func = free_func;
 
-	/* start up the reaper event */
-	mru->next_reap = 0;
-	mru->reap_all = 0;
-	queue_delayed_work(xfs_mru_reap_wq, &mru->work, mru->grp_time);
-
 	*mrup = mru;
 
 exit:
@@ -394,35 +389,25 @@ exit:
  * Call xfs_mru_cache_flush() to flush out all cached entries, calling their
  * free functions as they're deleted.  When this function returns, the caller is
  * guaranteed that all the free functions for all the elements have finished
- * executing.
- *
- * While we are flushing, we stop the periodic reaper event from triggering.
- * Normally, we want to restart this periodic event, but if we are shutting
- * down the cache we do not want it restarted. hence the restart parameter
- * where 0 = do not restart reaper and 1 = restart reaper.
+ * executing and the reaper is not running.
  */
 void
 xfs_mru_cache_flush(
-	xfs_mru_cache_t		*mru,
-	int			restart)
+	xfs_mru_cache_t		*mru)
 {
 	if (!mru || !mru->lists)
 		return;
 
-	cancel_rearming_delayed_workqueue(xfs_mru_reap_wq, &mru->work);
-
 	mutex_spinlock(&mru->lock);
-	mru->reap_all = 1;
-	mutex_spinunlock(&mru->lock, 0);
+	if (mru->queued) {
+		mutex_spinunlock(&mru->lock, 0);
+		cancel_rearming_delayed_workqueue(xfs_mru_reap_wq, &mru->work);
+		mutex_spinlock(&mru->lock);
+	}
 
-	queue_work(xfs_mru_reap_wq, &mru->work.work);
-	flush_workqueue(xfs_mru_reap_wq);
+	_xfs_mru_cache_migrate(mru, jiffies + mru->grp_count * mru->grp_time);
+	_xfs_mru_cache_clear_reap_list(mru);
 
-	mutex_spinlock(&mru->lock);
-	WARN_ON_ONCE(mru->reap_all != 0);
-	mru->reap_all = 0;
-	if (restart)
-		queue_delayed_work(xfs_mru_reap_wq, &mru->work, mru->grp_time);
 	mutex_spinunlock(&mru->lock, 0);
 }
 
@@ -433,8 +418,7 @@ xfs_mru_cache_destroy(
 	if (!mru || !mru->lists)
 		return;
 
-	/* we don't want the reaper to restart here */
-	xfs_mru_cache_flush(mru, 0);
+	xfs_mru_cache_flush(mru);
 
 	kmem_free(mru->lists, mru->grp_count * sizeof(*mru->lists));
 	kmem_free(mru, sizeof(*mru));
diff --git a/fs/xfs/xfs_mru_cache.h b/fs/xfs/xfs_mru_cache.h
index 624fd10ee8e..dd58ea1bbeb 100644
--- a/fs/xfs/xfs_mru_cache.h
+++ b/fs/xfs/xfs_mru_cache.h
@@ -32,11 +32,9 @@ typedef struct xfs_mru_cache
 	unsigned int		grp_time;  /* Time period spanned by grps.  */
 	unsigned int		lru_grp;   /* Group containing time zero.   */
 	unsigned long		time_zero; /* Time first element was added. */
-	unsigned long		next_reap; /* Time that the reaper should
-					      next do something. */
-	unsigned int		reap_all;  /* if set, reap all lists */
 	xfs_mru_cache_free_func_t free_func; /* Function pointer for freeing. */
 	struct delayed_work	work;      /* Workqueue data for reaping.   */
+	unsigned int		queued;	   /* work has been queued */
 } xfs_mru_cache_t;
 
 int xfs_mru_cache_init(void);
@@ -44,7 +42,7 @@ void xfs_mru_cache_uninit(void);
 int xfs_mru_cache_create(struct xfs_mru_cache **mrup, unsigned int lifetime_ms,
 			     unsigned int grp_count,
 			     xfs_mru_cache_free_func_t free_func);
-void xfs_mru_cache_flush(xfs_mru_cache_t *mru, int restart);
+void xfs_mru_cache_flush(xfs_mru_cache_t *mru);
 void xfs_mru_cache_destroy(struct xfs_mru_cache *mru);
 int xfs_mru_cache_insert(struct xfs_mru_cache *mru, unsigned long key,
 				void *value);
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index 1a5ad8cd97b..60345922990 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -1082,6 +1082,9 @@ xfs_fsync(
 	if (XFS_FORCED_SHUTDOWN(ip->i_mount))
 		return XFS_ERROR(EIO);
 
+	if (flag & FSYNC_DATA)
+		filemap_fdatawait(vn_to_inode(XFS_ITOV(ip))->i_mapping);
+
 	/*
 	 * We always need to make sure that the required inode state
 	 * is safe on disk.  The vnode might be clean but because
@@ -3769,12 +3772,16 @@ xfs_inode_flush(
 			sync_lsn = log->l_last_sync_lsn;
 			GRANT_UNLOCK(log, s);
 
-			if ((XFS_LSN_CMP(iip->ili_last_lsn, sync_lsn) <= 0))
-				return 0;
+			if ((XFS_LSN_CMP(iip->ili_last_lsn, sync_lsn) > 0)) {
+				if (flags & FLUSH_SYNC)
+					log_flags |= XFS_LOG_SYNC;
+				error = xfs_log_force(mp, iip->ili_last_lsn, log_flags);
+				if (error)
+					return error;
+			}
 
-			if (flags & FLUSH_SYNC)
-				log_flags |= XFS_LOG_SYNC;
-			return xfs_log_force(mp, iip->ili_last_lsn, log_flags);
+			if (ip->i_update_core == 0)
+				return 0;
 		}
 	}
 
@@ -3788,9 +3795,6 @@ xfs_inode_flush(
 	if (flags & FLUSH_INODE) {
 		int	flush_flags;
 
-		if (xfs_ipincount(ip))
-			return EAGAIN;
-
 		if (flags & FLUSH_SYNC) {
 			xfs_ilock(ip, XFS_ILOCK_SHARED);
 			xfs_iflock(ip);