Track negative entries v3

Track negative dentries by recording the generation number of the parent
directory in d_fsdata. The generation number for the parent directory is
recorded in the inode_info, which increments every time the lock on the
directory is dropped.

If the generation number of the parent directory and the negative dentry
matches, there is no need to perform the revalidate, else a revalidate
is forced. This improves performance in situations where nodes look for
the same non-existent file multiple times.

Thanks Mark for explaining the DLM sequence.

Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.de>
Signed-off-by: Joel Becker <joel.becker@oracle.com>

6 files changed, 42 insertions, 5 deletions

diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c
index b4957c7d9fe2..edaded48e7e9 100644
--- a/fs/ocfs2/dcache.c
+++ b/fs/ocfs2/dcache.c
@@ -40,6 +40,14 @@
 #include "inode.h"
 #include "super.h"
 
+void ocfs2_dentry_attach_gen(struct dentry *dentry)
+{
+        unsigned long gen =
+                OCFS2_I(dentry->d_parent->d_inode)->ip_dir_lock_gen;
+        BUG_ON(dentry->d_inode);
+        dentry->d_fsdata = (void *)gen;
+}
+
 
 static int ocfs2_dentry_revalidate(struct dentry *dentry,
                                    struct nameidata *nd)
@@ -51,11 +59,20 @@ static int ocfs2_dentry_revalidate(struct dentry *dentry,
         mlog_entry("(0x%p, '%.*s')\n", dentry,
                    dentry->d_name.len, dentry->d_name.name);
 
-        /* Never trust a negative dentry - force a new lookup. */
+        /* For a negative dentry -
+         * check the generation number of the parent and compare with the
+         * one stored in the inode.
+         */
         if (inode == NULL) {
-                mlog(0, "negative dentry: %.*s\n", dentry->d_name.len,
-                     dentry->d_name.name);
-                goto bail;
+                unsigned long gen = (unsigned long) dentry->d_fsdata;
+                unsigned long pgen =
+                        OCFS2_I(dentry->d_parent->d_inode)->ip_dir_lock_gen;
+                mlog(0, "negative dentry: %.*s parent gen: %lu "
+                        "dentry gen: %lu\n",
+                        dentry->d_name.len, dentry->d_name.name, pgen, gen);
+                if (gen != pgen)
+                        goto bail;
+                goto valid;
         }
 
         BUG_ON(!osb);
@@ -96,6 +113,7 @@ static int ocfs2_dentry_revalidate(struct dentry *dentry,
                 goto bail;
         }
 
+valid:
         ret = 1;
 
 bail:
@@ -227,6 +245,12 @@ int ocfs2_dentry_attach_lock(struct dentry *dentry,
         if (!inode)
                 return 0;
 
+        if (!dentry->d_inode && dentry->d_fsdata) {
+                /* Converting a negative dentry to positive
+                   Clear dentry->d_fsdata */
+                dentry->d_fsdata = dl = NULL;
+        }
+
         if (dl) {
                 mlog_bug_on_msg(dl->dl_parent_blkno != parent_blkno,
                                 " \"%.*s\": old parent: %llu, new: %llu\n",
@@ -452,6 +476,7 @@ static void ocfs2_dentry_iput(struct dentry *dentry, struct inode *inode)
 
 out:
         iput(inode);
+        ocfs2_dentry_attach_gen(dentry);
 }
 
 /*
diff --git a/fs/ocfs2/dcache.h b/fs/ocfs2/dcache.h
index f5dd1789acf1..b79eff709958 100644
--- a/fs/ocfs2/dcache.h
+++ b/fs/ocfs2/dcache.h
@@ -64,5 +64,6 @@ void ocfs2_dentry_move(struct dentry *dentry, struct dentry *target,
                        struct inode *old_dir, struct inode *new_dir);
 
 extern spinlock_t dentry_attach_lock;
+void ocfs2_dentry_attach_gen(struct dentry *dentry);
 
 #endif /* OCFS2_DCACHE_H */
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index 5e02a893f46e..e8d94d722ecb 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -3635,10 +3635,18 @@ static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres,
 {
         struct inode *inode;
         struct address_space *mapping;
+        struct ocfs2_inode_info *oi;
 
         inode = ocfs2_lock_res_inode(lockres);
         mapping = inode->i_mapping;
 
+        if (S_ISDIR(inode->i_mode)) {
+                oi = OCFS2_I(inode);
+                oi->ip_dir_lock_gen++;
+                mlog(0, "generation: %u\n", oi->ip_dir_lock_gen);
+                goto out;
+        }
+
         if (!S_ISREG(inode->i_mode))
                 goto out;
 
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index eece3e05d9d0..f935fd6600dd 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -335,6 +335,7 @@ void ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe,
                     else
                             inode->i_fop = &ocfs2_dops_no_plocks;
                     i_size_write(inode, le64_to_cpu(fe->i_size));
+                    OCFS2_I(inode)->ip_dir_lock_gen = 1;
                     break;
             case S_IFLNK:
                     if (ocfs2_inode_is_fast_symlink(inode))
diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h
index 0bc477a3aeb8..1c508b149b3a 100644
--- a/fs/ocfs2/inode.h
+++ b/fs/ocfs2/inode.h
@@ -67,6 +67,7 @@ struct ocfs2_inode_info
         /* Only valid if the inode is the dir. */
         u32                             ip_last_used_slot;
         u64                             ip_last_used_group;
+        u32                             ip_dir_lock_gen;
 
         struct ocfs2_alloc_reservation  ip_la_data_resv;
 };
diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c
index a00dda2e4f16..e7bde21149ae 100644
--- a/fs/ocfs2/namei.c
+++ b/fs/ocfs2/namei.c
@@ -171,7 +171,8 @@ bail_add:
                         ret = ERR_PTR(status);
                         goto bail_unlock;
                 }
-        }
+        } else
+                ocfs2_dentry_attach_gen(dentry);
 
 bail_unlock:
         /* Don't drop the cluster lock until *after* the d_add --