shrink struct dentry

struct dentry is one of the most critical structures in the kernel. So it's
sad to see it going neglected.

With CONFIG_PROFILING turned on (which is probably the common case at least
for distros and kernel developers), sizeof(struct dcache) == 208 here
(64-bit). This gives 19 objects per slab.

I packed d_mounted into a hole, and took another 4 bytes off the inline
name length to take the padding out from the end of the structure. This
shinks it to 200 bytes. I could have gone the other way and increased the
length to 40, but I'm aiming for a magic number, read on...

I then got rid of the d_cookie pointer. This shrinks it to 192 bytes. Rant:
why was this ever a good idea? The cookie system should increase its hash
size or use a tree or something if lookups are a problem. Also the "fast
dcookie lookups" in oprofile should be moved into the dcookie code -- how
can oprofile possibly care about the dcookie_mutex? It gets dropped after
get_dcookie() returns so it can't be providing any sort of protection.

At 192 bytes, 21 objects fit into a 4K page, saving about 3MB on my system
with ~140 000 entries allocated. 192 is also a multiple of 64, so we get
nice cacheline alignment on 64 and 32 byte line systems -- any given dentry
will now require 3 cachelines to touch all fields wheras previously it
would require 4.

I know the inline name size was chosen quite carefully, however with the
reduction in cacheline footprint, it should actually be just about as fast
to do a name lookup for a 36 character name as it was before the patch (and
faster for other sizes). The memory footprint savings for names which are
<= 32 or > 36 bytes long should more than make up for the memory cost for
33-36 byte names.

Performance is a feature...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

diff --git a/arch/powerpc/oprofile/cell/spu_task_sync.c b/arch/powerpc/oprofile/cell/spu_task_sync.c
index 2949126d28d1f8c9f45bef32d45b96df86f959fa..6b793aeda72ea041ac2a68f1ca2e2ece7f46badc 100644 (file)
--- a/arch/powerpc/oprofile/cell/spu_task_sync.c
+++ b/arch/powerpc/oprofile/cell/spu_task_sync.c
@@ -297,7 +297,7 @@ static inline unsigned long fast_get_dcookie(struct path *path)
 {
        unsigned long cookie;
 
-       if (path->dentry->d_cookie)
+       if (path->dentry->d_flags & DCACHE_COOKIE)
                return (unsigned long)path->dentry;
        get_dcookie(path, &cookie);
        return cookie;

diff --git a/drivers/oprofile/buffer_sync.c b/drivers/oprofile/buffer_sync.c
index 737bd9484822498d9db077640ed7acc0859babb9..65e8294a9e29db35cfcff8a493682804e9dcffe3 100644 (file)
--- a/drivers/oprofile/buffer_sync.c
+++ b/drivers/oprofile/buffer_sync.c
@@ -200,7 +200,7 @@ static inline unsigned long fast_get_dcookie(struct path *path)
 {
        unsigned long cookie;
 
-       if (path->dentry->d_cookie)
+       if (path->dentry->d_flags & DCACHE_COOKIE)
                return (unsigned long)path->dentry;
        get_dcookie(path, &cookie);
        return cookie;

diff --git a/fs/dcache.c b/fs/dcache.c
index a1d86c7f3e6644c8914edeae67a0ba02b3df6ada..fd244c7a7cc0e331101458caebcae512654461a8 100644 (file)
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -34,7 +34,6 @@
 #include <linux/bootmem.h>
 #include "internal.h"
 
-
 int sysctl_vfs_cache_pressure __read_mostly = 100;
 EXPORT_SYMBOL_GPL(sysctl_vfs_cache_pressure);
 
@@ -948,9 +947,6 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name)
        dentry->d_op = NULL;
        dentry->d_fsdata = NULL;
        dentry->d_mounted = 0;
-#ifdef CONFIG_PROFILING
-       dentry->d_cookie = NULL;
-#endif
        INIT_HLIST_NODE(&dentry->d_hash);
        INIT_LIST_HEAD(&dentry->d_lru);
        INIT_LIST_HEAD(&dentry->d_subdirs);

diff --git a/fs/dcookies.c b/fs/dcookies.c
index 855d4b1d619aec6244b182c3669b6efe2ec1f726..180e9fec4ad832763bbf6344ecf74265bc1d144b 100644 (file)
--- a/fs/dcookies.c
+++ b/fs/dcookies.c
@@ -93,10 +93,15 @@ static struct dcookie_struct *alloc_dcookie(struct path *path)
 {
        struct dcookie_struct *dcs = kmem_cache_alloc(dcookie_cache,
                                                        GFP_KERNEL);
+       struct dentry *d;
        if (!dcs)
                return NULL;
 
-       path->dentry->d_cookie = dcs;
+       d = path->dentry;
+       spin_lock(&d->d_lock);
+       d->d_flags |= DCACHE_COOKIE;
+       spin_unlock(&d->d_lock);
+
        dcs->path = *path;
        path_get(path);
        hash_dcookie(dcs);
@@ -119,14 +124,14 @@ int get_dcookie(struct path *path, unsigned long *cookie)
                goto out;
        }
 
-       dcs = path->dentry->d_cookie;
-
-       if (!dcs)
+       if (path->dentry->d_flags & DCACHE_COOKIE) {
+               dcs = find_dcookie((unsigned long)path->dentry);
+       } else {
                dcs = alloc_dcookie(path);
-
-       if (!dcs) {
-               err = -ENOMEM;
-               goto out;
+               if (!dcs) {
+                       err = -ENOMEM;
+                       goto out;
+               }
        }
 
        *cookie = dcookie_value(dcs);
@@ -251,7 +256,12 @@ out_kmem:
 
 static void free_dcookie(struct dcookie_struct * dcs)
 {
-       dcs->path.dentry->d_cookie = NULL;
+       struct dentry *d = dcs->path.dentry;
+
+       spin_lock(&d->d_lock);
+       d->d_flags &= ~DCACHE_COOKIE;
+       spin_unlock(&d->d_lock);
+
        path_put(&dcs->path);
        kmem_cache_free(dcookie_cache, dcs);
 }

diff --git a/include/linux/dcache.h b/include/linux/dcache.h
index a37359d0bad11613d7baa9ee0401fa79106c5576..c66d22487bf8c53e7dfb0ee6c34b8cbe726cae56 100644 (file)
--- a/include/linux/dcache.h
+++ b/include/linux/dcache.h
@@ -75,14 +75,22 @@ full_name_hash(const unsigned char *name, unsigned int len)
        return end_name_hash(hash);
 }
 
-struct dcookie_struct;
-
-#define DNAME_INLINE_LEN_MIN 36
+/*
+ * Try to keep struct dentry aligned on 64 byte cachelines (this will
+ * give reasonable cacheline footprint with larger lines without the
+ * large memory footprint increase).
+ */
+#ifdef CONFIG_64BIT
+#define DNAME_INLINE_LEN_MIN 32 /* 192 bytes */
+#else
+#define DNAME_INLINE_LEN_MIN 40 /* 128 bytes */
+#endif
 
 struct dentry {
        atomic_t d_count;
        unsigned int d_flags;           /* protected by d_lock */
        spinlock_t d_lock;              /* per dentry lock */
+       int d_mounted;
        struct inode *d_inode;          /* Where the name belongs to - NULL is
                                         * negative */
        /*
@@ -107,10 +115,7 @@ struct dentry {
        struct dentry_operations *d_op;
        struct super_block *d_sb;       /* The root of the dentry tree */
        void *d_fsdata;                 /* fs-specific data */
-#ifdef CONFIG_PROFILING
-       struct dcookie_struct *d_cookie; /* cookie, if any */
-#endif
-       int d_mounted;
+
        unsigned char d_iname[DNAME_INLINE_LEN_MIN];    /* small names */
 };
 
@@ -177,6 +182,8 @@ d_iput:             no              no              no       yes
 
 #define DCACHE_INOTIFY_PARENT_WATCHED  0x0020 /* Parent inode is watched */
 
+#define DCACHE_COOKIE          0x0040  /* For use by dcookie subsystem */
+
 extern spinlock_t dcache_lock;
 extern seqlock_t rename_lock;