6 files changed, 122 insertions, 86 deletions

diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c
index 3b11cb1e38fc..ee04ff5ee603 100644
--- a/fs/ocfs2/cluster/tcp.c
+++ b/fs/ocfs2/cluster/tcp.c
@@ -210,10 +210,6 @@ static inline void o2net_set_func_stop_time(struct o2net_sock_container *sc)
         sc->sc_tv_func_stop = ktime_get();
 }
 
-static ktime_t o2net_get_func_run_time(struct o2net_sock_container *sc)
-{
-        return ktime_sub(sc->sc_tv_func_stop, sc->sc_tv_func_start);
-}
 #else  /* CONFIG_DEBUG_FS */
 # define o2net_init_nst(a, b, c, d, e)
 # define o2net_set_nst_sock_time(a)
@@ -227,10 +223,14 @@ static ktime_t o2net_get_func_run_time(struct o2net_sock_container *sc)
 # define o2net_set_advance_stop_time(a)
 # define o2net_set_func_start_time(a)
 # define o2net_set_func_stop_time(a)
-# define o2net_get_func_run_time(a)             (ktime_t)0
 #endif /* CONFIG_DEBUG_FS */
 
 #ifdef CONFIG_OCFS2_FS_STATS
+static ktime_t o2net_get_func_run_time(struct o2net_sock_container *sc)
+{
+        return ktime_sub(sc->sc_tv_func_stop, sc->sc_tv_func_start);
+}
+
 static void o2net_update_send_stats(struct o2net_send_tracking *nst,
                                     struct o2net_sock_container *sc)
 {
diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c
index 0740752ca31e..b0c17490bbbc 100644
--- a/fs/ocfs2/dir.c
+++ b/fs/ocfs2/dir.c
@@ -322,21 +322,23 @@ static int ocfs2_check_dir_entry(struct inode * dir,
         const char *error_msg = NULL;
         const int rlen = le16_to_cpu(de->rec_len);
 
-        if (rlen < OCFS2_DIR_REC_LEN(1))
+        if (unlikely(rlen < OCFS2_DIR_REC_LEN(1)))
                 error_msg = "rec_len is smaller than minimal";
-        else if (rlen % 4 != 0)
+        else if (unlikely(rlen % 4 != 0))
                 error_msg = "rec_len % 4 != 0";
-        else if (rlen < OCFS2_DIR_REC_LEN(de->name_len))
+        else if (unlikely(rlen < OCFS2_DIR_REC_LEN(de->name_len)))
                 error_msg = "rec_len is too small for name_len";
-        else if (((char *) de - bh->b_data) + rlen > dir->i_sb->s_blocksize)
+        else if (unlikely(
+                 ((char *) de - bh->b_data) + rlen > dir->i_sb->s_blocksize))
                 error_msg = "directory entry across blocks";
 
-        if (error_msg != NULL)
+        if (unlikely(error_msg != NULL))
                 mlog(ML_ERROR, "bad entry in directory #%llu: %s - "
                      "offset=%lu, inode=%llu, rec_len=%d, name_len=%d\n",
                      (unsigned long long)OCFS2_I(dir)->ip_blkno, error_msg,
                      offset, (unsigned long long)le64_to_cpu(de->inode), rlen,
                      de->name_len);
+
         return error_msg == NULL ? 1 : 0;
 }
 
diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c
index 5391f28d940b..7540a492eaba 100644
--- a/fs/ocfs2/dlm/dlmdomain.c
+++ b/fs/ocfs2/dlm/dlmdomain.c
@@ -926,9 +926,10 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data,
 }
 
 static int dlm_match_regions(struct dlm_ctxt *dlm,
-                             struct dlm_query_region *qr)
+                             struct dlm_query_region *qr,
+                             char *local, int locallen)
 {
-        char *local = NULL, *remote = qr->qr_regions;
+        char *remote = qr->qr_regions;
         char *l, *r;
         int localnr, i, j, foundit;
         int status = 0;
@@ -957,13 +958,8 @@ static int dlm_match_regions(struct dlm_ctxt *dlm,
                 r += O2HB_MAX_REGION_NAME_LEN;
         }
 
-        local = kmalloc(sizeof(qr->qr_regions), GFP_ATOMIC);
-        if (!local) {
-                status = -ENOMEM;
-                goto bail;
-        }
-
-        localnr = o2hb_get_all_regions(local, O2NM_MAX_REGIONS);
+        localnr = min(O2NM_MAX_REGIONS, locallen/O2HB_MAX_REGION_NAME_LEN);
+        localnr = o2hb_get_all_regions(local, (u8)localnr);
 
         /* compare local regions with remote */
         l = local;
@@ -1012,8 +1008,6 @@ static int dlm_match_regions(struct dlm_ctxt *dlm,
         }
 
 bail:
-        kfree(local);
-
         return status;
 }
 
@@ -1075,6 +1069,7 @@ static int dlm_query_region_handler(struct o2net_msg *msg, u32 len,
 {
         struct dlm_query_region *qr;
         struct dlm_ctxt *dlm = NULL;
+        char *local = NULL;
         int status = 0;
         int locked = 0;
 
@@ -1083,6 +1078,13 @@ static int dlm_query_region_handler(struct o2net_msg *msg, u32 len,
         mlog(0, "Node %u queries hb regions on domain %s\n", qr->qr_node,
              qr->qr_domain);
 
+        /* buffer used in dlm_mast_regions() */
+        local = kmalloc(sizeof(qr->qr_regions), GFP_KERNEL);
+        if (!local) {
+                status = -ENOMEM;
+                goto bail;
+        }
+
         status = -EINVAL;
 
         spin_lock(&dlm_domain_lock);
@@ -1112,13 +1114,15 @@ static int dlm_query_region_handler(struct o2net_msg *msg, u32 len,
                 goto bail;
         }
 
-        status = dlm_match_regions(dlm, qr);
+        status = dlm_match_regions(dlm, qr, local, sizeof(qr->qr_regions));
 
 bail:
         if (locked)
                 spin_unlock(&dlm->spinlock);
         spin_unlock(&dlm_domain_lock);
 
+        kfree(local);
+
         return status;
 }
 
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index 7eb4b8da5cbc..7642d7ca73e5 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -64,7 +64,7 @@ struct ocfs2_mask_waiter {
         unsigned long           mw_mask;
         unsigned long           mw_goal;
 #ifdef CONFIG_OCFS2_FS_STATS
-        unsigned long long      mw_lock_start;
+        ktime_t                 mw_lock_start;
 #endif
 };
 
@@ -431,44 +431,41 @@ static void ocfs2_remove_lockres_tracking(struct ocfs2_lock_res *res)
 #ifdef CONFIG_OCFS2_FS_STATS
 static void ocfs2_init_lock_stats(struct ocfs2_lock_res *res)
 {
-        res->l_lock_num_prmode = 0;
-        res->l_lock_num_prmode_failed = 0;
-        res->l_lock_total_prmode = 0;
-        res->l_lock_max_prmode = 0;
-        res->l_lock_num_exmode = 0;
-        res->l_lock_num_exmode_failed = 0;
-        res->l_lock_total_exmode = 0;
-        res->l_lock_max_exmode = 0;
         res->l_lock_refresh = 0;
+        memset(&res->l_lock_prmode, 0, sizeof(struct ocfs2_lock_stats));
+        memset(&res->l_lock_exmode, 0, sizeof(struct ocfs2_lock_stats));
 }
 
 static void ocfs2_update_lock_stats(struct ocfs2_lock_res *res, int level,
                                     struct ocfs2_mask_waiter *mw, int ret)
 {
-        unsigned long long *num, *sum;
-        unsigned int *max, *failed;
-        struct timespec ts = current_kernel_time();
-        unsigned long long time = timespec_to_ns(&ts) - mw->mw_lock_start;
-
-        if (level == LKM_PRMODE) {
-                num = &res->l_lock_num_prmode;
-                sum = &res->l_lock_total_prmode;
-                max = &res->l_lock_max_prmode;
-                failed = &res->l_lock_num_prmode_failed;
-        } else if (level == LKM_EXMODE) {
-                num = &res->l_lock_num_exmode;
-                sum = &res->l_lock_total_exmode;
-                max = &res->l_lock_max_exmode;
-                failed = &res->l_lock_num_exmode_failed;
-        } else
+        u32 usec;
+        ktime_t kt;
+        struct ocfs2_lock_stats *stats;
+
+        if (level == LKM_PRMODE)
+                stats = &res->l_lock_prmode;
+        else if (level == LKM_EXMODE)
+                stats = &res->l_lock_exmode;
+        else
                 return;
 
-        (*num)++;
-        (*sum) += time;
-        if (time > *max)
-                *max = time;
+        kt = ktime_sub(ktime_get(), mw->mw_lock_start);
+        usec = ktime_to_us(kt);
+
+        stats->ls_gets++;
+        stats->ls_total += ktime_to_ns(kt);
+        /* overflow */
+        if (unlikely(stats->ls_gets) == 0) {
+                stats->ls_gets++;
+                stats->ls_total = ktime_to_ns(kt);
+        }
+
+        if (stats->ls_max < usec)
+                stats->ls_max = usec;
+
         if (ret)
-                (*failed)++;
+                stats->ls_fail++;
 }
 
 static inline void ocfs2_track_lock_refresh(struct ocfs2_lock_res *lockres)
@@ -478,8 +475,7 @@ static inline void ocfs2_track_lock_refresh(struct ocfs2_lock_res *lockres)
 
 static inline void ocfs2_init_start_time(struct ocfs2_mask_waiter *mw)
 {
-        struct timespec ts = current_kernel_time();
-        mw->mw_lock_start = timespec_to_ns(&ts);
+        mw->mw_lock_start = ktime_get();
 }
 #else
 static inline void ocfs2_init_lock_stats(struct ocfs2_lock_res *res)
@@ -2774,8 +2770,15 @@ static void *ocfs2_dlm_seq_next(struct seq_file *m, void *v, loff_t *pos)
         return iter;
 }
 
-/* So that debugfs.ocfs2 can determine which format is being used */
-#define OCFS2_DLM_DEBUG_STR_VERSION 2
+/*
+ * Version is used by debugfs.ocfs2 to determine the format being used
+ *
+ * New in version 2
+ *      - Lock stats printed
+ * New in version 3
+ *      - Max time in lock stats is in usecs (instead of nsecs)
+ */
+#define OCFS2_DLM_DEBUG_STR_VERSION 3
 static int ocfs2_dlm_seq_show(struct seq_file *m, void *v)
 {
         int i;
@@ -2817,18 +2820,18 @@ static int ocfs2_dlm_seq_show(struct seq_file *m, void *v)
                 seq_printf(m, "0x%x\t", lvb[i]);
 
 #ifdef CONFIG_OCFS2_FS_STATS
-# define lock_num_prmode(_l)            (_l)->l_lock_num_prmode
-# define lock_num_exmode(_l)            (_l)->l_lock_num_exmode
-# define lock_num_prmode_failed(_l)     (_l)->l_lock_num_prmode_failed
-# define lock_num_exmode_failed(_l)     (_l)->l_lock_num_exmode_failed
-# define lock_total_prmode(_l)          (_l)->l_lock_total_prmode
-# define lock_total_exmode(_l)          (_l)->l_lock_total_exmode
-# define lock_max_prmode(_l)            (_l)->l_lock_max_prmode
-# define lock_max_exmode(_l)            (_l)->l_lock_max_exmode
-# define lock_refresh(_l)               (_l)->l_lock_refresh
+# define lock_num_prmode(_l)            ((_l)->l_lock_prmode.ls_gets)
+# define lock_num_exmode(_l)            ((_l)->l_lock_exmode.ls_gets)
+# define lock_num_prmode_failed(_l)     ((_l)->l_lock_prmode.ls_fail)
+# define lock_num_exmode_failed(_l)     ((_l)->l_lock_exmode.ls_fail)
+# define lock_total_prmode(_l)          ((_l)->l_lock_prmode.ls_total)
+# define lock_total_exmode(_l)          ((_l)->l_lock_exmode.ls_total)
+# define lock_max_prmode(_l)            ((_l)->l_lock_prmode.ls_max)
+# define lock_max_exmode(_l)            ((_l)->l_lock_exmode.ls_max)
+# define lock_refresh(_l)               ((_l)->l_lock_refresh)
 #else
-# define lock_num_prmode(_l)            (0ULL)
-# define lock_num_exmode(_l)            (0ULL)
+# define lock_num_prmode(_l)            (0)
+# define lock_num_exmode(_l)            (0)
 # define lock_num_prmode_failed(_l)     (0)
 # define lock_num_exmode_failed(_l)     (0)
 # define lock_total_prmode(_l)          (0ULL)
@@ -2838,8 +2841,8 @@ static int ocfs2_dlm_seq_show(struct seq_file *m, void *v)
 # define lock_refresh(_l)               (0)
 #endif
         /* The following seq_print was added in version 2 of this output */
-        seq_printf(m, "%llu\t"
-                   "%llu\t"
+        seq_printf(m, "%u\t"
+                   "%u\t"
                    "%u\t"
                    "%u\t"
                    "%llu\t"
diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
index d9bfa904d006..f97a213223e8 100644
--- a/fs/ocfs2/ioctl.c
+++ b/fs/ocfs2/ioctl.c
@@ -45,6 +45,22 @@ static inline void __o2info_set_request_error(struct ocfs2_info_request *kreq,
 #define o2info_set_request_error(a, b) \
                 __o2info_set_request_error((struct ocfs2_info_request *)&(a), b)
 
+static inline void __o2info_set_request_filled(struct ocfs2_info_request *req)
+{
+        req->ir_flags |= OCFS2_INFO_FL_FILLED;
+}
+
+#define o2info_set_request_filled(a) \
+                __o2info_set_request_filled((struct ocfs2_info_request *)&(a))
+
+static inline void __o2info_clear_request_filled(struct ocfs2_info_request *req)
+{
+        req->ir_flags &= ~OCFS2_INFO_FL_FILLED;
+}
+
+#define o2info_clear_request_filled(a) \
+                __o2info_clear_request_filled((struct ocfs2_info_request *)&(a))
+
 static int ocfs2_get_inode_attr(struct inode *inode, unsigned *flags)
 {
         int status;
@@ -136,7 +152,8 @@ int ocfs2_info_handle_blocksize(struct inode *inode,
                 goto bail;
 
         oib.ib_blocksize = inode->i_sb->s_blocksize;
-        oib.ib_req.ir_flags |= OCFS2_INFO_FL_FILLED;
+
+        o2info_set_request_filled(oib);
 
         if (o2info_to_user(oib, req))
                 goto bail;
@@ -160,7 +177,8 @@ int ocfs2_info_handle_clustersize(struct inode *inode,
                 goto bail;
 
         oic.ic_clustersize = osb->s_clustersize;
-        oic.ic_req.ir_flags |= OCFS2_INFO_FL_FILLED;
+
+        o2info_set_request_filled(oic);
 
         if (o2info_to_user(oic, req))
                 goto bail;
@@ -184,7 +202,8 @@ int ocfs2_info_handle_maxslots(struct inode *inode,
                 goto bail;
 
         oim.im_max_slots = osb->max_slots;
-        oim.im_req.ir_flags |= OCFS2_INFO_FL_FILLED;
+
+        o2info_set_request_filled(oim);
 
         if (o2info_to_user(oim, req))
                 goto bail;
@@ -208,7 +227,8 @@ int ocfs2_info_handle_label(struct inode *inode,
                 goto bail;
 
         memcpy(oil.il_label, osb->vol_label, OCFS2_MAX_VOL_LABEL_LEN);
-        oil.il_req.ir_flags |= OCFS2_INFO_FL_FILLED;
+
+        o2info_set_request_filled(oil);
 
         if (o2info_to_user(oil, req))
                 goto bail;
@@ -232,7 +252,8 @@ int ocfs2_info_handle_uuid(struct inode *inode,
                 goto bail;
 
         memcpy(oiu.iu_uuid_str, osb->uuid_str, OCFS2_TEXT_UUID_LEN + 1);
-        oiu.iu_req.ir_flags |= OCFS2_INFO_FL_FILLED;
+
+        o2info_set_request_filled(oiu);
 
         if (o2info_to_user(oiu, req))
                 goto bail;
@@ -258,7 +279,8 @@ int ocfs2_info_handle_fs_features(struct inode *inode,
         oif.if_compat_features = osb->s_feature_compat;
         oif.if_incompat_features = osb->s_feature_incompat;
         oif.if_ro_compat_features = osb->s_feature_ro_compat;
-        oif.if_req.ir_flags |= OCFS2_INFO_FL_FILLED;
+
+        o2info_set_request_filled(oif);
 
         if (o2info_to_user(oif, req))
                 goto bail;
@@ -283,7 +305,7 @@ int ocfs2_info_handle_journal_size(struct inode *inode,
 
         oij.ij_journal_size = osb->journal->j_inode->i_size;
 
-        oij.ij_req.ir_flags |= OCFS2_INFO_FL_FILLED;
+        o2info_set_request_filled(oij);
 
         if (o2info_to_user(oij, req))
                 goto bail;
@@ -305,7 +327,7 @@ int ocfs2_info_handle_unknown(struct inode *inode,
         if (o2info_from_user(oir, req))
                 goto bail;
 
-        oir.ir_flags &= ~OCFS2_INFO_FL_FILLED;
+        o2info_clear_request_filled(oir);
 
         if (o2info_to_user(oir, req))
                 goto bail;
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 51cd6898e7f1..4e3d3c1363f3 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -147,6 +147,17 @@ struct ocfs2_lock_res_ops;
 
 typedef void (*ocfs2_lock_callback)(int status, unsigned long data);
 
+#ifdef CONFIG_OCFS2_FS_STATS
+struct ocfs2_lock_stats {
+        u64             ls_total;       /* Total wait in NSEC */
+        u32             ls_gets;        /* Num acquires */
+        u32             ls_fail;        /* Num failed acquires */
+
+        /* Storing max wait in usecs saves 24 bytes per inode */
+        u32             ls_max;         /* Max wait in USEC */
+};
+#endif
+
 struct ocfs2_lock_res {
         void                    *l_priv;
         struct ocfs2_lock_res_ops *l_ops;
@@ -182,15 +193,9 @@ struct ocfs2_lock_res {
         struct list_head         l_debug_list;
 
 #ifdef CONFIG_OCFS2_FS_STATS
-        unsigned long long       l_lock_num_prmode;        /* PR acquires */
-        unsigned long long       l_lock_num_exmode;        /* EX acquires */
-        unsigned int             l_lock_num_prmode_failed; /* Failed PR gets */
-        unsigned int             l_lock_num_exmode_failed; /* Failed EX gets */
-        unsigned long long       l_lock_total_prmode;      /* Tot wait for PR */
-        unsigned long long       l_lock_total_exmode;      /* Tot wait for EX */
-        unsigned int             l_lock_max_prmode;        /* Max wait for PR */
-        unsigned int             l_lock_max_exmode;        /* Max wait for EX */
-        unsigned int             l_lock_refresh;           /* Disk refreshes */
+        struct ocfs2_lock_stats  l_lock_prmode;         /* PR mode stats */
+        u32                      l_lock_refresh;        /* Disk refreshes */
+        struct ocfs2_lock_stats  l_lock_exmode;         /* EX mode stats */
 #endif
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
         struct lockdep_map       l_lockdep_map;